The Role of Small RNAs in C4 Photosynthesis E.L.C. Gage Magdalene College University of Cambridge A thesis submitted for the degree of Doctor of Philosophy June 2012 Contents Declaration ........................................................................................................................... i Acknowledgements ............................................................................................................ iii Abstract ............................................................................................................................... v List of Figures .................................................................................................................... vii List of Tables .................................................................................................................... viii Abbreviations ..................................................................................................................... ix 1. Introduction ...................................................................................................................... 1 1.1: A Requirement for Improved Crop Productivity ....................................................... 1 1.2: The Effects of Photorespiration ................................................................................. 1 1.3: The C4 Cycle .............................................................................................................. 3 1.4: miRNA Regulation of the C4 Cycle ......................................................................... 14 1.5: miRNA Regulation .................................................................................................... 15 1.6: Experimental Approach ........................................................................................... 17 1.7: Project Aims .............................................................................................................. 21 2. Materials and Methods ................................................................................................. 23 2.1: General Plant Cultivation........................................................................................ 23 2.2: Molecular Techniques ............................................................................................ 24 2.3: Quantification of Protein ....................................................................................... 30 2.4: Plasmid Design and Construction ......................................................................... 32 2.5: Microbial Techniques ..............................................................................................35 2.6: Transformation of Cleome gynandra ..................................................................... 37 2.7: Transformation of Arabidopsis thaliana ................................................................ 39 2.8: Analysis of Transgenics .......................................................................................... 40 2.9: Anatomical Analysis ................................................................................................ 41 2.10: Photosynthetic Analysis ........................................................................................ 42 2.11: Hybridisations of C. hassleriana and C. gynandra ................................................ 43 2.11.2: In vitro Ovule Culture ......................................................................................... 44 3. Do miRNAs regulate the C4 Cycle? Analysis of sRNAs and their Inhibition ............. 45 3.1: Introduction ............................................................................................................. 45 3.1.4: Experimental Design ............................................................................................ 48 3.3: Discussion ............................................................................................................... 76 4. Changes in Gene Regulation during Cotyledon Development ................................... 87 4.1: Introduction ............................................................................................................ 87 4.2: Results ..................................................................................................................... 90 4.3: Discussion ............................................................................................................... 96 5. The Generation of Hybrids within Cleome ................................................................. 103 5.1: Introduction ............................................................................................................ 103 5.2: Results .................................................................................................................... 106 5.3: Discussion ............................................................................................................... 110 6. Final Discussion ............................................................................................................ 115 6.1: The potential for miRNAs to regulate the C4 Cycle .............................................. 115 6.2: miRNAs actively regulate Development in C. gynandra ...................................... 115 6.3: sRNAs may act as Regulators of the C4 Cycle in C. gynandra .............................. 116 6.4: Light Influences Development in C. gynandra .................................................... 120 6.5: Final Conclusions ................................................................................................... 121 7. References ..................................................................................................................... 123 Appendix 1: Primer List ..................................................................................................... 141 Appendix 2: sRNA Target List ......................................................................................... 144 Appendix 3: Plasmid Maps .............................................................................................. 147 i Declaration This dissertation is the result of my own work and includes nothing which is the outcome of work done in collaboration except where specifically indicated in the text. It does not exceed the word limit set out by the Degree Committee for Biology and is not substantially the same as any work that has been, or is being submitted to any other university for any degree, diploma or any other qualification. E.L.C. Gage ii iii Acknowledgements I wish to thank my supervisor Julian Hibberd for his invaluable instruction and support throughout this project. Secondly, John Grey, John Carr and Howard Griffiths have all provided excellent advice and motivation throughout the project. Thanks are due to Kaiser Kajala, Naomi Brown and Holly Astley for advice on molecular techniques, Christine Newell for instruction on in vitro techniques, Padubidri Shiaprasad for advice and assistance on small RNA sequencing, Risha Narayan and Kim Rutherford for assistance with the bioinformatic analysis, Maria Herse and Jessica Royles for advice regarding photosynthetic analysis and to Susan Stanley and Britta Kuempers for advice on histological techniques. Thanks are due to all the members of the Molecular Physiology group for making this project a uniquely enjoyable experience and for putting up with my eccentricities as a scientist. Helen Woodfield and Jane Knerova have made working in the lab a constant pleasure, while Ben Williams and Marek Szecowka never fail to brighten my day. Considerable thanks are due to my parents, Harriet Dickinson, Alice Chang, Peter Brierley, Miranda de Graaf and Cheryl Snowdon for their unfailing support, advice and all-round-greatness. Finally, I would like to thank the Biotechnology and Biological Sciences Research Council, the Department of Plant Sciences, Magdalene College and the Cambridge Philosophical Society for the funding and support that made this work possible. Si scientia modo per ruinam confecta est, completa scientia modo per ruinam hominis ipsius confecta est. Immolatio operi factus sum, sed scientia mentis, corporis et plantae perfecta est. iv v Abstract The C4 cycle represents a series of biochemical and anatomical modifications that are targeted to overcome the effects of photorespiration caused by the oxygenase capability of Ribulose Bisphosphate Carboxylase/Oxygenase (RuBisCO). The cycle has evolved independently in over 60 lineages, which suggests that recruitment of genes into the C4 cycle is a relatively easy process. However, the mechanisms by which the anatomy and cell-specificity of the components of the C4 cycle is achieved is poorly understood. Preliminary work in maize indicated several components of the C4 cycle may be targeted by microRNAs (miRNAs). To explore this, a library of sRNA sequences from mature leaf tissue of the model C4 species Cleome gynandra L. was generated and then searched against a list of expressed sequence tag sequences for candidate genes of the C4 cycle. To complement this, transgenic C. gynandra containing the viral p19 protein, which is capable of suppressing miRNA activity, were produced. A limited subset of the candidate C4 genes showed a high level of sRNA read alignment. In C. gynandra plants expressing p19 photosynthesis was compromised and transcripts of several genes (most notably RbcS and RCA) were upregulated. These data were complemented by examining the effect of illumination on developing C. gynandra cotyledons, and attempts to generate a hybrid between C. gynandra and the C3 C. hassleriana Chodat. RbcS also showed elevated abundance in etiolated cotyledons, although this rapidly declined after illumination. The remainder of the C4 genes profiled accumulated in etiolated tissue, but were upregulated within 6 hours of illumination. Therefore, this study has illustrated that miRNA activity may play a role in maintaining the C4 photosynthetic cycle at optimum efficiency, although it has not been possible to identify at which point(s) this regulation is applied. Secondly, RbcS appears to be subject to multiple regulatory mechanisms in C. gynandra, and it is possible that miRNAs have a role in negatively regulating expression of RbcS. vi vii List of Figures Figure Contents Page 1.1 Outline of the NAD-ME subtype of the C4 Cycle 4 1.2 The distribution of C4 lineages 7 1.3 miRNA Formation and activity 15 2.1 Sequencing and Processing of the sRNA Library 28 2.2 Design of p19 expression constructs 47 3.1 Function of the pOp promoter 50 3.2 Size distribution of reads in the C. gynandra sRNA library 50 3.3 Distribution of sRNA hits against target sequences 54 3.4 Alignment patterns of sRNA hits against target sequences 58 3.5 Alignments of potential miRNA sequences to NAD-ME2 transcripts 58 3.6 Products of a 5'RACE of PPDK transcripts 59 3.7 p19 accumulation in transgenic Arabidopsis lines 60 3.8 p19 expression alters leaf morphology in Arabidopsis 61 3.9 p19 expression alters reproductive anatomy in Arabidopsis 62 3.10 A correlation between p19 abundance and Arabidopsis rosette diameter 62 3.11 p19 expression in transgenic C. gynandra lines 63 3.12 p19 expression effects leaf anatomy in C. gynandra 64 3.13 p19 expression effects leaf anatomy in C. gynandra 64 3.14 Creation of the pOp::p19 construct 65 3.15 GUS expression in Arabidopsis lines containing pOp::p19 67 3.16 p19 transcript and protein abundance in pOp::p19 Arabidopsis lines 67 3.17 p19 expression in pOp::p19 Arabidopsis lines alters leaf morphology 69 3.18 Grafting transgenic C. gynandra scions 710 3.19 GUS expression in C. gynandra transgenics containing pOp::p19 71 3.20 p19 expression in pOp::p19 C. gynandra trangenics 71 3.21 p19 expression does not alter Kranz development in C. gynandra 73 3.22 Photosynthetic response of C. gynandra transgenics expressing p19 73 3.23 Changes in transcript abundance in C. gynandra plants expressing p19 74 3.24 Protein abundance in C. gynandra transgenics expressing p19 75 4.1 Etiolated cotyledons possess Kranz anatomy 90 4.2 Etiolated cotyledons possess Kranz anatomy 90 4.3 Chlorophyll abundance in C. gynandra cotyledons 91 4.4 Light-induced changes in transcript abundance in C. gynandra cotyledons 93 4.5 Light-induced changes in transcript abundance in C. gynandra cotyledons 94 4.6 Light-induced changes in protein abundance in C. gynandra cotyledons 96 5.1 Reproductive anatomy of C. gynandra and C. hassleriana 107 5.2 Stigma receptivity in C. gynandra and C. hassleriana 108 5.3 Pollen tube growth in heterospecific pollen crosses 108 5.4 Hybrid embryo growth in plant 119 5.5 Response of hybrid embryos to in vitro culture 110 5.6 Stages of Embryo development 111 Chapter1: Introduction viii List of Tables Table Contents Page 1.1 sRNA reads aligning against C4 genes in Maize 13 3.1 Statistics of sRNA library composition 51 3.2 Numbers of sRNA reads aligning against candidate EST sequences 53 3.3 Distribution of mismatched sRNA hits against candidate EST sequences 55/56 3.4 Conserved sequences present in predicted miRNA sequences 58 3.5 p19 expression effects photosynthesis in C. gynandra 63 Chapter1: Introduction ix Abbreviations Gene and transcript names are given in italics while protein names are given in Roman text. In the following list only the protein name is given in the gene and protein names correspond. This is supplementary to abbreviations stated in the text. ABA Abscisic Acid AlaAT Alanine Aminotransferase AspAT Asparate Aminotransferase At Arabidopsis thaliana bp Base Pair BS Bundle Sheath CA Carbonic Anhydrase CaMV Cauliflower Mosaic Virus cDNA Complementary DNA Cg Cleome gynandra CO2 Carbonic Dioxide Chp chloroplastic Cs Cleome hassleriana cyt cytosolic DNA Deoxyribonucleic Acid dNTP Deoxyribonucleotide triphosphate EDTA Ethylene Diaminetetra-acetic Acid gDNA Genomic DNA GUS β-Glucuronidase HCO3- Bicarbonate LB Lysogeny Broth LHC Light Harvesting Complex LHCA1 Light Harvesting Complex of PSI M Mesophyll MDH Malate Dehydrogenase miRNA micro RNA mRNA Messenger RNA MS Murashige-Skoog medium mt mitochondrial NAD Nicotinamide Adenine Dinucleotide NAD-ME NAD-Malic Enzyme NADP Nicotinamide Adenine Dinucleotide Phosphate NADP-ME NADP-Malic Enzyme NHD1 Sodium : Hydrogen Antiporter nosT Nopaline Synthase Terminator OAA Oxaloacetate PCR Polymerase Chain Reaction PEP Phosphoenolpyruvate PEPC Phosphoenolpyruvate Carboxylase PEPCK Phosphoenolpyruvate Carboxykinase PFD Photon Flux Density Pi Inorganic Phosphate PIP Plasma membrane Intrinsic Protein Chapter1: Introduction x PM Plasma Membrane PPCK PEPC kinase PPDK Pyruvate,orthophosphate dikinase PPi Pyrophosphate PPT Plastidial PEP/Pi Translocator PSBQ2 Subunit Q of PSII PSI Photosystem I PSII Photosystem II PyT1 Chloroplastic Pyruvate Transporter qRT-PCR Quantitative Reverse Transcription Polymerase Chain Reaction RACE Rapid Amplification of cDNA Ends RbcS Gene encoding RuBisCO small subunit RCA RuBisCO Activase RNA Ribonucleic Acid RT-PCR Reverse Transcription Polymerase Chain Reaction RuBisCO Ribulose 1,5-Bis-Phosphate Carboxylase Oxygenase sRNA Small RNA SOC Super Optimal Broth, Catabolite Repression TBE Tris-Borate-EDTA buffer Tris Tris(hydroxymethyl)aminomethane uidA Gene encoding β-Glucuronidase UTR Untranslated Region X-Gluc 5-Bromo-4-Chloro-3-Indolyl-Beta-D Glucuronic Acid, Cyclohexylammonium Salt 1 1. Introduction 1.1: A Requirement for Improved Crop Productivity Over 200 years ago it was proposed that the demands of the human population would outstrip agricultural output (Malthus, 1798) although this stark prediction has not yet been fully realised. The human population has increased by around 125% since 1950, with conservative estimates of a further 65% increase over the next 50 years (Fischer & Heilig, 1997). The increase in population has so far been sustained through the “Green Revolution” that combined high-yielding varieties of wheat and rice with increased mechanisation and agrichemical use to increase crop yields (Evenson & Gollin, 2003). However, the rate of yield increase has slowed since the 1990’s to the extent that the increase in food production is now below that of the rate of predicted population increase (McCalla, 1994; Khush, 1999). Besides a reduction in the rate of yield increase, a number of additional factors may negatively impact food production. For example, while climate change may lead to a slight increase in yields at high latitudes, significant reductions are likely to be seen nearer the equator affecting the developing world where increases in population are likely to be greatest (Parry et al., 2004). A reduction in yield may be brought about by reduced growing seasons, decreased water availability or increased losses to pests and disease (Parry et al., 1999; Wallace, 2004). As a result of these factors, sustained increases in crop yields are required. This is especially the case for crops grown in southern Africa and Asia as these regions are likely to be subject to the greatest climate-change related effects (Lobell et al., 2008). The yield potential of crops under ideal conditions is determined by the efficiency of light capture, the conversion of light into biomass and the proportion of biomass partitioned into the grain (Evans & Fischer, 1999). The latter two components are considered to have been developed to an optimum by breeding, and it has been suggested that increasing photosynthetic efficiency is likely to be the key area for improving crop yields (Long et al., 2006). In addition, the predicted effects of climate change on water availability may mean that increases in crop water use efficiency will be required if production levels are to be sustained (Wallace, 2000). 1.2: The Effects of Photorespiration One potential method to increase photosynthetic efficiency may be to reduce the effects of photorespiration. A key step in photosynthesis is the carboxylation of ribulose bisphosphate through the activity of Ribulose Bisphosphate Carboxylase/Oxygenase (RuBisCO). Due to the 2 active site conformation of RuBisCO, Ribulose BisPhosphate (RuBP) can be either carboxylated or oxygenated (Andrews & Lorimer, 1987). The phosphoglycolate that is produced by the oxygenation reaction is wasteful, and its generation and subsequent metabolism is collectively termed photorespiration. At low temperatures, the comparative rate of oxygenation and carboxylation of RuBP makes the generation of phosphoglycolate negligible, but once temperatures exceed 30°C photorespiration can reduce photosynthesis by 30-40% at ambient CO2 concentrations (Ehleringer et al., 1991). One solution to this problem is to concentrate CO2 around RuBisCO to the point that the oxygenation reaction is inhibited (Chollet & Ogren, 1975), a mechanism which is employed in the C4 cycle (Hatch & Slack, 1966). The cycle operates to initially fix HCO3- using phosphoenolpyruvate carboxylase (PEPC) to form a 4C acid which is subsequently decarboxylated in a separate compartment close to RuBisCO for refixing and entry into the Calvin cycle. PEPC is specific to bicarbonate, for which it shows a high affinity (Maruyama et al., 1966) so that the operation of this cycle can effectively supress photorespiration. The ability to overcome the oxygenation problem of RuBisCO generates a competitive advantage in high light, high temperature environments with the effect that C4 species can become dominant in warmer climates (Sage et al., 1999). Such increased photosynthetic efficiency means C4 photosynthesis increases the radiation use efficiency by c.50% compared to C3 taxa in the same environment (Kiniry et al., 1989). The C4 pathway is utilized by maize, sugarcane and sorghum, representing some of our most productive crops (Brown, 1999). Besides allowing increased rates of carbon fixation, the C4 cycle allows increased nitrogen use efficiency (Gallaher & Brown, 1977). For example, maize shows higher nitrogen use efficiency compared with wheat and rice (both C3) irrespective of temperature and therefore the extent of photorespiration (Schmitt & Edwards, 1980). The increased efficiency of CO2 fixation seen in the C4 system allows for decreased stomatal conductance in C4 species relative to C3 species. Therefore, the C4 cycle can also increase water use efficiency. This is thought to impact on the dominance of C4 photosynthesis in hot, dry environments (Knapp et al., 1994; Sage et al., 1999). Growth of C4 species tends to be less competitive at cooler temperatures, either due to evolutionary weakness of plants which evolved under hotter conditions (Long, 2004) or the increased metabolic cost of the C4 pathway: a need to regenerate the metabolic precursors increases the net ATP cost of operating the C4 cycle above that of the C3 system, lowering quantum yield (Ehleringer & Björkman, 1977). This means that the C4 system ceases to have competitive advantage under low temperatures when C3 plants show reduced 3 photorespiration, but C4 species can still become dominant in low temperature ecosystems by virtue of the other physiological benefits of the C4 cycle. For example, C4 species have become dominant in north European saline habitats as a result of the increased water use efficiency associated with the C4 cycle (Collins & Jones, 1986). Therefore, operation of the C4 cycle can confer significant benefits in overcoming environmental limitations in the form of elevated temperature, nutrient scarcity or limited water availability. Given the concerns regarding food production (see above), the C4 cycle is considered a trait that could improve yields (Long et al., 2006). The ability of the C4 cycle to confer increased water use efficiency in particular has led to the suggestion that the C4 cycle may be helpful in promoting yield increases in rice (Hibberd et al., 2008). Rice represents the principal foodstuff in Asia, a region which is already subject to food scarcity, the problems caused by the high irrigation requirements of the crop are likely to be exacerbated by the effects of climate change (Shehy et al., 2007). Therefore, the introduction of the C4 cycle may help to maintain a yield increase in line with population requirements. The C4 system represents a number of significant anatomical and biochemical modifications from the C3 state. However, the genetic origin of the C4 system is not yet fully understood, and it will be necessary to develop this understanding if the C4 system is to be replicated in a previously C3 species. 1.3: The C4 Cycle 1.3.1: Operation of the C4 Cycle While C4 cycles based on a single-cell system have been described (e.g. Bienertia cycloptera, Voznesenskaya et al., 2005), in most cases the cycle is shared between two cell types. Initial carbon fixation occurs in the mesophyll (M) to form a C4 acid, typically aspartate or malate. The 4C acid then diffuses into specially modified bundle sheath (BS) cells where it is decarboxylated and the CO2 produced is refixed by RuBisCO. The C3 by-products of the decarboxylation reaction then diffuse back into the M to be used in the regeneration of the cycle precursors (Hatch, 1971). The BS cells surrounding the vasculature are enlarged and contain increased numbers of chloroplasts that can show altered ultrastructure. Such specialised so-called “Kranz” anatomy was first described by Haberlandt (1882) and the anatomical modifications are closely linked to the underpinning biochemistry of the cycle (Dengler & Nelson, 1999). BS cell walls are frequently thickened and contain a suberized lamella to limit CO2 leakage, although the wall is permeated by an increased number of plasmodesmatal linkages to the M to promote diffusion of metabolites through the symplast. 4 Starch formation is predominately limited to the BS, where the chloroplasts may show reduced granal stacking and asymmetric arrangement (discussed below). The M resemble normal C3 cells, although these may be radialised to increase contact with the BS. Finally, leaf venation is increased to reduce the diffusive pathway distance between the M and BS. Some variation is seen in the orchestration of the cycle, and three principal biochemical subtypes have been described which vary primarily in the method of decarboxylation in the BS (reviewed by Kanai & Edwards, 1999; Furbank, 2011). The presence of Kranz anatomy allows for each cell types to be adapted to a separate role in the C4 cycle, as outlined in Figure 1.2. CO2 is initially fixed in the M through the carboxylation of phosphoenolpyruvate (PEP) using phosphoenolpyruvate carboxylase (PEPC). The oxaloacetate (OAA) that is formed is converted into either aspartate or malate. In the NADP- malic enzyme (NADP-ME) subtype, malate is produced and diffuses into the BS where it is decarboxylated by NADP-ME in the chloroplast. The pyruvate that is produced as a by- product is returned back to the M where it is used to regenerate PEP using pyruvate, orthophosphate dikinase (PPDK) in the chloroplast. In the NAD-ME subtype, the OAA is converted into aspartate via aspartate aminotransferase which diffuses into the BS, is transaminated back into OAA before reduction to malate via malate dehydrogenase which is then decarboxylated by NAD-ME, reactions which all occur in the mitochondria. Again, pyruvate is produced as a by-product, but this is transferred back into the M in the form of alanine via alanine aminotransferase. In the PEP carboxykinase (PEPCK) subtype, two alternate reactions can occur. Aspartate can be formed from OAA, and this aspartate diffuses into the BS where it is transaminated back into OAA before decarboxylation into PEP and CO2 via PEPCK. Alternately, the OAA can be converted into malate via NADP-malate dehydrogenase in the M chloroplasts, and the malate is then transferred into the BS mitochondria before being metabolised into pyruvate using NAD-malic enzyme. The pyruvate is then returned to the M as alanine where it is used to regenerate PEP using PPDK. While diffusive transport is considered sufficient for the bulk of metabolite exchange between the two cell types in the C4 cycle, active transport across organellar membranes is also necessary. For example, the sodium-dependant pyruvate transporter BASS2 and a PEP/phosphate translocator are required for pyruvate/PEP transport across the chloroplast envelope (Weber et al., 2010; Furumoto et al., 2011). 5 Figure 1.1: Outline of the NAD-ME subtype of the C4 Cycle. The fixation of CO2 in the mesophyll (M) initially forms oxaloacetate (OAA) which then diffuses into the bundle sheath (BS) in the form of aspartate (ASP) before reduction to malate (MAL) and decarboxylation with NAD-malic enzyme (NAD- ME), creating a CO2-enriched environment which reduces photorespiratory reactions from occurring. The pyruvate (PYR) that is also produced diffuses back to the M in the form of alanine (ALN) where it is used to regenerate the phosphoenolpyruvate (PEP) consumed in the initial carboxylation step. Sites of active transport are marked with a black dot, as discussed in the text. CA: carbonic anhydrase, PEPC: PEP carboxylase, AspAT: aspartate aminotransferase, MDH: malate dehydrogenase, AlnAT: alanine aminotransferase, PPDK: pyruvate, orthophosphate dikinase 6 1.3.2: Evolution of the C4 Cycle Despite the significant changes that are likely to have occurred during the evolution of the C4 cycle, it is found in 66 independent angiosperm lineages (Sage, 2012). While the first fossil evidence of the C4 cycle has been dated to the Miocene at 12.5 Mya (Nambudiri et al., 1978), molecular clock analysis has placed the first occurrences of the C4 cycle to 32 Mya, a point which coincides with the rapid decline in atmospheric CO2 that occurred during the Oligocene around 35 Mya (Kellogg, 1999; Christin et al., 2008). This would indicate that evolution of C4 photosynthesis was in response to increased oxygen concentrations rather than temperature increases, although the delayed response of c.3 million years was necessary for evolutionary preconditioning events to occur, such as gene duplication events prior to recruitment into novel photosynthetic roles or the anatomical modifications required for efficient operation of the cycle (Monson & Rawsthorne, 2004). Once sufficient environmental conditions existed to promote evolution of the C4 cycle (i.e. depressed atmospheric CO2 concentrations and/or elevated temperatures), then a number of step-wise evolutionary stages are considered likely to have occurred (Sage, 2004). A measure of general preconditioning would be have been necessary. It has been proposed that the enlargement of gene families via duplication events allowed for divergent evolution needed for the C4 cycle (Monson, 1999). However, recent evidence suggests that evolution of trans-acting factors (as discussed below) may be important in the evolution of the C4 cycle, bypassing the requirement for gene duplication events, which may explain why gene families in C4 lineages are not significantly enlarged (Williams et al., In Press). This would have been followed by anatomical alterations such as reduced intervein distance or enlargement of the BS although such modifications may have served other ecological benefits such as enhanced leaf water status (Dengler et al., 1993; Sage, 2004). Increased capacity of BS organelles would then enable them to assume a greater photosynthetic role, along with increased plasmodesmatal links with the M to facilitate increased transport of metabolites. The development of a glycine shuttle is then proposed (Monson, 1999) so that photorespiratory products would accumulate in the bundle sheath for metabolic breakdown. Once metabolite transport is established, increased activity of PEPC and PPDK in the mesophyll and a decarboxylase in the bundle sheath, twined with a reduction of RuBisCO activity in the mesophyll would facilitate a transition from a C3- like to a C4-like system. Finally, optimisation of individual C4 components and whole-plant optimisation would allow for the obtainment of full C4 status. 7 Genes required for the C4 cycle existed previously, but typically these were involved with alternative non-photosynthetic processes such as the proposed role of PPDK (utilised in the regeneration of PEP used in the initial fixation step) in nitrogen remobilisation from leaves during senescence (Taylor et al., 2010; Aoyagi et al., 1984; Liu et al., 2008). Expression of these genes would have to be modified to alter abundance, acquire cell-type specific expression, and in some instances altered reaction kinetics (Monson et al., 2004; Sheen, 1999), a process that would have to be concurrent with anatomical modification. However, while no novel elements are required for the utilisation of the C4 cycle, considerable changes in gene expression are required leading to cell-type specific accumulation of enzymes, changes in enzyme kinetics and regulation and the development of the M/BS anatomy (Sage, 1999). Despite the complexity of the evolutionary requirements of the C4 cycle, it is predicted to have evolved at least 66 separate times (Sage et al., 2011, Figure 1.2) which would indicate that the changes in gene expression and anatomy required for the effective operation of the C4 cycle do not require any substantial evolutionary steps. The evolutionary steps towards the C4 cycle must either infer neutral or positive competitive advantage to prevent negative selection pressures from eliminating them, a hypothesis that is supported by the abundance of C3/C4 intermediates throughout various C4 lineages, as discussed below. These may represent discrete evolutionary stages or prematurely terminated evolutionary pathways, but their persistence would indicate evolutionary fitness (Monson & Moore, 1989; Sage et al., 2011). The multiple points of evolution have given rise to considerable variation in the operation of the cycle. Anatomically, the origin and number of BS layers, extent of suberin deposition, and the position and extent of granal development of BS chloroplasts varies greatly between lineages (reviewed by Dengler & Nelson, 1999). While this may be due to evolutionary divergence, such variation may be attributable to ecological factors. For example, Cleome angustifolia (Cleomaceae) shows modified Kranz anatomy to incorporate water storage tissues while other C4 members of the genus show the more standard atriplicoid Kranz anatomy (Koteyeva et al., 2011). 8 Figure 1.2: The distribution of C4 lineages within the angiosperms. 47 of the predicted 62 independent origins of the C4 cycle are overlaid in red upon the phylogeny of 9400 angiosperms of Smith et al. (1999). The values quoted with each clade indicate the number of independent origins of the C4 system within each clade. Figure adapted from Sage et al., 2011. Biochemical variation is seen in the operation of the cycle, with separate lineages being classified on the basis of the principal enzyme utilised at the decarboxylase step which may either be NAD-Malic Enzyme (NAD-ME), NADP-Malic Enzyme (NADP-ME) or PEP carboxykinase (PEPCK), the biochemistry of which is discussed below. The presence of a particular decarboxylation subtype demonstrates some patterning with evolutionary lineage, although there is also some correlation between C4 subtype and precipitation levels in which each subtype is predominant indicating that decarboxylase variation may serve additional ecological roles (Cabido et al., 2008). Furthermore, the decarboxylation subtype may correlate with anatomical features, such as the reduction in granal stacking and PSII components seen in the BS chloroplasts of NADP-ME species to reduce oxygen evolution (Meierhoff & Westhoff, 1993) or the centripetal accumulation of chloroplasts in the BS of NAD-ME species in an attempt to limit CO2 leakage in the absence of a suberized lamella (Sage, 2004). 9 The stepwise nature of the evolution of the C4 cycle, and the ability of each stage to be competitive in its own right, is supported by a number of stable C3/C4 intermediates (e.g. Edwards et al., 1982). The ability of the C4 cycle to supress photorespiration leads to a significant depression in the CO2 compensation point (Γ) from 4-5.5 kPa CO2 in C3 species to 0-0.5 kPa CO2, with C3/C4 species showing values intermediate to these values, although this can be in a light-dependant manner (Hunt et al., 1987). Such intermediates can contain Kranz- like anatomy with the formation of distinct BS cells with increased organelle abundance, although this does not correlate with an increased activity of C4 enzymes, indicating that the biochemical partitioning has not occurred (Edwards & Ku, 1987). A number of “C4-like” members of Flaveria show a more intermediate metabolism with increased carbon fixation into C4 acids and more extensive Kranz formation (Moore, 1989). The presence of C3, C4, and intermediate species within the same genus has led to a number of hybridisation attempts between photosynthetic types within Panicum, Flaveria and Atriplex (as reviewed by Brown & Bouton, 1993). Panicum C3/C4 x C3 hybrids showed intermediate characteristics in the F1 generation, and while segregation in the F2 hybrids was sufficient to recover the parental C3/C4 anatomy, Γ did not approach that of the C3/C4 parent, indicating that strict compartmentalisation is required for true repression of photorespiration (Bouton et al., 1986). Hybrids between C3 and C4 members of Atriplex showed similar intermediary traits in anatomy, physiology and expression patterns of C4 enzymes. For example, C3 dominance was shown on the expression of PEPC in Atriplex C3 x C4 hybrids (Pearcy & Björkman, 1970), PPDK levels resembled F. floridana (C3/C4) when crossed with F. brownii (C4-like) (Cheng et al., 1987) and NADP-ME expression was significantly depressed in hybrids between F. trinervia and a number of C3/C4 intermediates (summarised by Brown & Bouton, 1993). RuBisCO activity varied, with activity in separate hybrids resembling both the C3 (Pearcy & Björkman, 1970) and C4 (Björkman et al., 1969) parent, although expression of RuBisCO is complicated by the biparental inheritance of the SSU/LSU. A general trend in these data is that C3 is predominant, although this could be attributable to several causes. Repression of genes may be seen in the C3 parent which has been lost during the evolution of the C4 lineage, or the trans-acting factors (see below) required for altered levels of expression are absent from the C3 parent and are inherited independently. The variation in both biochemical and anatomical traits seen in hybrids between C3 and C4 species indicates that traits of the C4 cycle are independently inherited due to their apparent lack of segregation. 10 1.3.3: Regulation of the C4 Cycle While recruitment of genes into the C4 cycle was heavily dependent on altering regulatory methods to bring about modified expression patterns, our understanding of the regulatory mechanisms functioning in the current C4 cycle is far from complete. The regulation that underpins the cell-type specific protein accumulation necessary for the C4 cycle is likely to operate at multiple levels, varying between gene, developmental stage and lineage. At the transcriptional level, exclusion of PEPC transcripts from Flaveria bidentis BS cells is dependent on promoter activity (Stockhaus et al., 1997) which is attributable to the presence of the mesophyll enhancing module 1 (MEM1), a 41bp sequence that is sufficient to inhibit expression in the BS. The MEM1 of the C3 F. pringlei did not direct M-specific expression in F. bidentis when fused to the native C4 Ppc promoter, but inhibition in the BS was recovered by insertion of the F. bidentis MEM1 sequence (Gowik et al., 2004; Akyildiz et al., 2007). Separate regions within the promoter may be responsible for elevated transcription and cell-type specificity. The -211 to +434 region of the RbcSm3 promoter in maize is required for light- induced expression, while a region between -907 and -445 in combination with a sequence within the 3’ region of the gene is required for suppression of RbcS transcription in the M (Viret et al., 1994). Transcriptional control may not relate to the promoter sequence itself: 240bp sequences within the Cleome gynandra NAD-ME1 and 2 transcripts are sufficient to grant BS-specific accumulation (Brown et al., 2011). This element retains this function when placed in the antisense direction which would infer their mode of action is unlikely to be based on interaction with the mRNA molecules (Williams et al., unpublished). In maize, illumination may promote the activity of Dof1/2 transcription factors which have been implicated in coordinating the patterning of Ppc transcripts in maize (Yanagisawa, 2000; Kalamajka et al., 2003). In addition, Langdale et al. (1991) proposed a demethylation event 3.3kb upstream of the Ppc promoter is linked to M-specific expression in response to light. However, Tolley et al. (2011) did not identify any methylated sites within the 0.6kb region that is required for M-specific expression (Kaush et al., 2001), which would infer the absence of PEPC in maize BS cells is not attributable to modifications to DNA methylation. While promoter components such as these have been identified as inferring C4-like expression patterns, the trans-acting factors and the sequence specific targets with which they interact remain largely undefined. The Golden2 and Golden-like1 transcription factors have been suggested as potential trans factors in maize due to their ability to target high numbers of photosynthesis-related genes in Arabidopsis (Hall et al., 1998; Waters et al., 2009), although 11 whether they are capable of directing the C4 expression patterns of the previously non- photosynthetic genes has yet to be explicitly determined. Transcriptional control is likely to be combined with additional levels of control. For example, while higher rates of chloroplastic rbcL and psbB transcription are seen in maize BS cells, the rate difference does not correlate with observed differences in transcript abundance between the M and BS (Kubicki et al., 1994). Post-transcriptional processes may involve increased transcript stability, transport, elevated decay rates or alternative processing methods. Transcript-specific cis-acting elements have been described in a number of instances, and the placement of the cis-acting element within the coding regions could infer post-transcriptional methods are in action. For example, PPDK and CA4 transcripts have also been shown to contain sequences in the UTR regions that confer M-specific accumulation (Kajala et al., 2011). UTR regions in RbcS transcripts of A. hypochondriacus generate translational enhancement in BS cells while triggering transcript degradation in the M (Patel et al., 2004). Cis-acting elements in F. bidentis RbcS UTRs may also direct BS-specific accumulation either by enhanced BS stability or decay rates of transcripts in the M (Patel et al., 2006). Beyond the transcriptional level, control may be exerted at the translational level. In the non- photosynthetic foliar regions of Amaranthus tricolor, a loss of transcript specificity does not correspond to a loss of BS-specificity of both subunits of RuBisCO (McCormac et al., 1997), while in dark-grown A. hypochondriacus both subunits remain BS specific despite a collapse of transcript specificity (Berry et al., 1985) so that while transcriptional-level regulation has ceased, a formerly redundant method of control is being exerted at the translational level. When grown in light, chloroplastic rbcL in still transcribed but fails to show protein accumulation in the M of A. hypochondriacus (Boinski et al., 1993). Trans-acting factors may be involved in translational processes, such as the BSD2 zinc-binding protein which is believed to prevent aggregation and misfolding of large subunit polypeptides prior to incorporation into the RuBisCO holoenzyme (Brutnell et al., 1999). Transcript regulation may be combined with translational control. Variation is seen in rbcL transcript length in A. hypochondriacus which impacts the binding of a 47kDa protein required for ribosome attachment (McCormac et al., 2001) while 5’UTR regions of RbcS function as a translational enhancer in the same species (Patel et al., 2004). Post-translational regulation may also be possible, such as the inactivation of PPDK by phosphorylation by a regulatory protein correlating with developmental age or illumination 12 (Chastain et al., 2006; 2008). In addition to the transcriptional control of maize PEPC discussed above, the mature protein is subject to reversible phosphorylation by PEPC kinase, the abundance of which is regulated at the transcriptional level in response to light (Jiao & Chollet, 1988; Hartwell et al., 1999) and PPDK demonstrates reversible phosphorylation in light (Astley et al., 2011) 1.3.4: Development of the C4 Cycle During leaf development, the onset of cellular patterning associated with the C4 cycle shows considerable variation between lineages in terms of timing and response to developmental or environmental signals. RuBisCO, PPDK, PEPC, and NAD- malic enzyme accumulation in maize is tightly linked to light exposure combined with acropetal vein development as demonstrated by a lack of accumulation mutants that lack chlorophyll (Langdale et al., 1987; Nelson et al., 1984). Similarly, CA shows an increase in development but requires light exposure for high levels of expression (Burnell et al., 1990). While light is necessary for C4 development in maize, the patterning of RuBisCO, PEPC and PPDK occurs as a response to developmental signals in etiolated Amaranth cotyledons (Wang et al., 1993a), indicating that light is not required for this process. In A. hypochondraicus, PEPC and PPDK are localised to the M from an early stage in development, while both subunits of RuBisCO only become localised to the BS once a developmental barrier has been crossed and transcriptional suppression has become active (Wang et al., 1992). In contrast, partitioning of the large subunit of RuBisCO occurs before the maturation of the BS in Atriplex rosea (Hudson et al., 1992). Therefore, if light is not required for activation of the C4 cycle, then alternative signals must be responded to. The acquisition of BS-specific RuBisCO accumulation Amaranthus is tightly correlated to the transition from photosynthate sink to source (Wang et al., 1993b), although whether this correlation is due to active photosynthetic metabolism or coincident developmental events remains to be explored. PEPC expression in maize has been shown to be depressed by glucose and acetate application (Kausch et al., 2001), and various metabolites have been shown to negatively regulate photosynthesis genes (Sheen, 1990) so the former hypothesis is not without merit. Finally, there is some suggestion that the acquisition of C4 traits is spatially regulated in maize by a transmissible signal from the vasculature (Langdale et al., 1988; Langdale & Nelson, 1991), although this hypothesis remains to be validated. 13 Besides the biochemical adaptation, the formation of Kranz anatomy is considered a general prerequisite for the evolution of an effective C4 cycle (Sage, 2004), with the exception of the unicellular systems discussed above. In maize, the close vascularisation is present from leaf initiation with M/BS initialisation closely following the same pattern of vein maturation in both an acropetal and basipetal direction, depending on vein position (Langdale & Nelson, 1991; Langdale et al., 1987). Dicot C4 species (e.g. Amaranthus) do not show such a developmental gradient, but instead the development of Kranz anatomy occurs after the formation of the secondary veins in a basipetal direction, an event which tends to coincide with the onset of C4 biochemical patterning and the transition from source to sink tissues (Ramsperger et al., 1996; Wang et al., 1993b). The genetic origin of Kranz anatomy has yet to be determined. Hybridisation experiments in Atriplex shows intermediate and non- segregating forms of Kranz anatomy are expressed in F1 and F2 hybrids between C3 and C4 taxa, which may infer this is a polygenic trait as the failure of the Kranz-like anatomy to show segregation between the F1 and F2 generations would infer independent polygenic control (Björkman et al., 1970; 1971). As such, the regulation methods employed in establishing the expression patterns associated with the C4 cycle are varied at multiple levels, and while there is some understanding of the cis-elements required the trans factors that interact with them are poorly understood. The abundance of post-transcriptional regulatory methods may provide some insight into the evolution of the C4 cycle, however. Modified gene expression may be brought about by altering either the cis-regulating sequence or the activity of the trans-acting component with which in interacts (Love et al., 2007). The latter path would enable recruitment of genes into the C4 cycle without modification of the coding sequence itself, a situation which has been demonstrated in the C4 cycle previously. For example, the UTR sequences of NAD-ME, PPDK and CA transcripts required to grant M-specific expression are found in C3 orthologs (Brown et al, 2011; Kajala et al., 2011) which would suggest that the cis elements required for M specific expression are ancestral and that the trans factor itself has undergone recruitment into the cycle. As noted previously, understanding of the trans-acting factors in the C4 cycle is poor, limiting the extent to which their evolution and regulation can be studied. Therefore, it is considered that the identification of trans-acting factors, especially of a suppressive or silencing role, would be of considerable benefit in extending our understanding of the C4 cycle. 14 1.4: miRNA Regulation of the C4 Cycle A class of trans-acting factors which have been increasingly shown to have regulatory roles in plant systems are a number of small non-coding RNAs (sRNA, Hamilton & Baulcombe 1999). These are short sequences of RNA around 24nt in length which interact with a target gene through possession of a complementary sequence to either the genomic or transcript sequence, bringing about negative repression through a number of routes (see below). As a measure of complementarity between an sRNA and its target it required, prediction of sRNA- target genes can be carried out by comparison of sRNA sequences aligning against a gene of interest (Ambros, 2004). When sRNA libraries were aligned against photosynthesis genes in Arabidopsis, Rice and Maize, a disproportionate accumulation of sRNA sequences against these genes was demonstrated in maize (Table 1.1). A large number of sRNA sequences aligning against a gene may indicate that an underlying sRNA is present which is capable of negative regulation. The absence of high sequence aggregations in both C3 species may suggest that sRNAs have evolved to negatively regulate components of the C4 cycle in maize, promoting cell-type specific accumulation of protein. Protein Encoded Arabidopsis (C3) Rice (C3) Maize (C4) CA 0 0 2 PEPC 0 0 11 PPCK 0 0 10 MDH - 0 1 OMT1 - 0 2 NADP-ME 0 2 4 RbcS1 0 1 6 RbcS2 1 1 3 PRK 0 0 7 PPDK 0 1 2 This hypothesis is attractive on a number of levels. MicroRNAs (miRNAs), a subset of sRNAs, have been heavily implicated in a number of developmental and regulatory roles (discussed below). Given this background, miRNAs may play a role in the development of Kranz anatomy during leaf maturation. If a regulatory miRNA were to accumulate in a cell-type specific manner after leaf maturation exclusion of a gene could be achieved via post-transcriptional regulation. The non-mobile nature of miRNAs (Parizotto et al., 2004) would mean that Table 1.1: In silico predictions of sRNA sequences aligning against key photosynthesis genes in both C3 and C4 backgrounds. (Tolley et al., unpublished data taken from a study restricted to known photosynthesis genes) 15 separate miRNA populations in the M and BS could be in place, facilitating cell-type specific suppression of gene expression. Despite high levels of miRNA sequence conservation between diverse lineages, a rapid rate of proto-miRNA sequence generation and evolution has been implicated via inverted duplication events occurring during gene family expansion (Allan et al., 2004; Fahlgren et al., 2007). If such rapid evolution did occur then it may provide a mechanism by which rapid C4 evolution could occur along divergent lineages. 1.5: miRNA Regulation Various classes of small non-coding RNAs between 20–28nt have been shown to direct target- specific gene regulation (Hamilton & Baulcombe, 1999; Ambros, 2004; Bartel et al., 2004). Classification within general sRNA regulation is derived from the origin of the regulatory fragment, which may involve different precursors and processing methods (reviewed by Brodersen & Voinnet, 2006). In short, a double-stranded RNA (dsRNA) molecule is cleaved to produce single-stranded RNA (ssRNA) fragments of 20-28nt which can then be incorporated into an RNA-Induced Silencing Complex (RISC), capable of directing a regulatory effect. Three principal sRNA forms are active in plants: short interfering RNA (siRNA) which target exogenous transcripts, and trans-acting short interfering RNA (tasi-RNA) and micro RNA (miRNA), both of which target endogenous transcripts. These differ both on methods of biosynthesis and the origin of the dsRNA utilised to produce the sRNA. The dsRNA precursor may be produced from a single-stranded target via an RNA-dependant RNA polymerase (RDRP), as occurs in the suppression of ssRNA viral transcripts or the post-transcriptional silencing of exogenous transgenes, or it may be formed by a transcript which contains an inverted repeat sequence which allows the transcript to fold to produce a secondary structure which contains a double stranded region (a hairpin bend). siRNAs may be generated from natural dsRNA (e.g. viral genomes) , or through the action of an RDRP prior to cleavage. The production of tasiRNA occurs from endogenous genes, the transcripts of which are targeted by a miRNA. One of the cleavage products is used to produce dsRNA using RDRP which is subsequently processed into a functional miRNA. miRNAs represent the greatest proportion of endogenous regulatory sRNAs in plants (Schwab et al., 2009), and have received the greatest extent of scientific exploration to date. The predominant method of regulation is through targeted transcript cleavage although translational repression and miRNA-directed DNA methylation also have been described (Brodersen et al., 2008; Chellappan et al., 2010). Such miRNAs demonstrate a regulatory effect on numerous developmental processes including leaf formation (Chuck et al., 2007) and organ 16 polarity (Kidner & Martienssen, 2004). The roles of miRNA in leaf development are well documented. The miRNA-mediated regulation of PHABULOSA\PHAVOLUTA during leaf formation has provided significant insights into the mechanisms of miRNA regulatory pathways (Mallory et al., 2004; McHale et al., 2004), while tasiRNA have been shown to be capable of creating tissue patterns during development based on differential spatial distribution (Chitwood et al., 2009). The role of miRNAs in regulating organ formation is continued with determined roles for miRNAs in petal and root formation (e.g. Chen, 2004; Llave et al., 2002). Besides developmental roles, miRNAs have been implicated in a range of abiotic stress responses including phosphate starvation (Fujii et al., 2005), salt stress (Ding et al., 2009), arsenic toxicity (Srivastava et al., 2005) and mechanical stress (Lu et al., 2005). In these instances upregulation of miRNAs has been detected as part of a given stress response. For example, Arabidopsis plants that were subject to copper starvation showed an upregulation of miR398 which targets copper/zinc superoxide dismutases, subsequently downregulating these genes in response to copper depletion (Beauclair et al., 2010). Figure 1.3: miRNA formation and activity. Pri-miRNA transcripts contain an inverted complementary sequence capable of forming a hairpin fold-back structure containing a stretch of dsRNA (A). This acts as a substrate for Dicer-like 1 (DCL1), which cleaves the hairpin into a short dsRNA fragment which is stabilised by methylation by Hua-Enhancer 1 (HEN1) (B). The dsRNA fragment is then exported out of the nucleus via HASTY (C) where it is processed into the single-stranded miRNA and incorporated into the RNA-induced silencing complex (RISC), the chief component of which is the Argonaute 1 (AGO1) endonuclease. The complementarity of the miRNA is utilised to guide the RISC to the target transcript (E) which is then cleaved before subsequent degradation (F). (After Brodersen & Voinnet, 2006) miRNAs are derived from longer length transcripts (including those taken from intergene, UTR and intronic regions) (Figure 1.3) which are processed into short double-stranded RNA (dsRNA) duplexes by the DICER-LIKE1 protein (DCL1) (Kurihara & Watanabe, 2004) prior to stabilisation by methylation via HUA-ENHANCER1 (HEN1; Yang et al., 2006). The resultant 17 duplex is then exported to the cytosol via HASTY (HST, Bollman et al., 2003) whereupon the duplex is separated and the functional miRNA is incorporated into an RNA-induced silencing complex (RISC), of which the RNA slicer ARGONAUTE1 (AGO1) is an integral component. Transcript targeting is achieved via near-perfect complementarity between miRNA and its target mRNA (Rhoades et al., 2002) which will promote cleavage and subsequent degradation of the target transcript (Llave et al., 2002a). Translational repression tends to be linked to less- perfect base pairing, although there are some exceptions such as the translational repression of APETALA2 (Chen, 2004). A significant proportion of plant miRNA targets have been classified as transcription factors (Rhoades et al., 2002) and regulation on a spatial or temporal scale has been demonstrated through regulation of the promoter sequences of the pre-miRNA or parent transcript (Obernosterer et al., 2006; Chitwood et al., 2007) or through the expression of RNA-binding proteins which specifically target certain pre-miRNAs (Viswanathan et al., 2008). In addition to posttranscriptional processes, miRNA has been illustrated as having an impact on the epigenetic level, such as the chromatin methylation of PHABULOSA by miR165/6 (Bao et al., 2004). While some sRNA forms have been shown to be mobile (moving between cells) these tend to be assiociated with the suppression of exogenous viral or transgene transcripts by siRNAs (Voinnet, 2009; Molnar et al., 2010). miRNAs linked with developmental roles show zones of activity which appear to coincide entirely with the zones of transcription of miRNA-encoding genes, so a lack of mobility is demonstrated (Parizotto et al., 2004). This would be a requirement for regulation of the C4 cycle: if miRNAs were mobile their application to cell-type specificity would be limited. 1.6: Experimental Approach 1.6.1: miRNA Prediction miRNA sequences may be identified by computational analysis of the target genome to predict pre-miRNA sequences based on ability to form dsRNA duplexes (e.g. Jones-Rhoades & Bartel, 2004), although this requires a well-annotated genome if it is to be effective. Alternatively, by sequencing the sRNA population of the target organism it is possible to identify miRNAs by comparing sRNA reads against the sequences of target genes. Such combined computational/cloning based strategies have been applied in the identification of miRNA sequences in a range of circumstances such as the identification of mechanical stress- responsive miRNAs in Populus (Lu et al., 2005) and varied stresses in Arabidopsis (Sunkar & Zhu, 2004). While direct the computational analysis of miRNA precursors can yield predictions, the latter authors note the detection of miRNAs by cloning methods that would 18 not have been predicted in silico. Therefore, the construction of a library of miRNA sequences via the cloning methods of Havecker (2011) was considered to be the primary step in profiling regulatory miRNAs active in a C4 leaf. This strategy has two key limitations however. Firstly, a well-annotated genome is required for full miRNA prediction. A key part of validating a potential miRNA is the presence of a precursor transcript within the target genome, the flanking regions of which must be capable of generating a minimal free-energy folding hairpin structure capable of generating a pre- miRNA duplex (Ambros et al., 2004). Secondly, the lack of a gene-specific understanding of the origin of many C4-traits would hamper the identification of regulatory miRNA on the basis of sequence complementarity. As a result of this, a second strategy was implemented utilising viral miRNA inhibitors. 1.6.2: Viral Inhibition of sRNA Activity The biosynthesis of short-interfering RNAs (siRNA) by a pathway closely related to that of miRNA production is considered to be an immune response in plants to virus infection (Voinnet 2001; Vance & Vaucheret, 2001). In response to this, a number of viral proteins have been demonstrated to supress the activity of si/miRNAs as a pathogenicity factor. p21 of the Beet Yellow Virus (Reed et al., 2003), AC4 of the African Cassava Mosaic Virus (Chellappan et al., 2005) and p19 of the Tomato Bushy Stunt Virus (Voinnet et al., 2003, Lakatos et al., 2004) have been shown to inhibit miRNA function by acting as competitors with DCL/HEN1 for miRNA*/miRNA dsRNA duplexes before they can be processed into a functional RISC complex. A forth inhibitor, P1/HC-Pro of the Turnip Mosaic Virus has been well studied (Kasschau et al., 2003) although its exact method of function is still unclear (Lakatos et al., 2006). Due to the similarity between biosynthesis pathways of siRNAs and miRNAs, expression of such viral inhibitors can interfere with the function of endogenous miRNAs (Chapman et al., 2004), promoting a phenotype that matches that of dcl-1 mutants in Arabidopsis (Kasschau et al., 2003; Dunoyer et al., 2004). Due to the competitive nature of the inhibitory mechanism a dose-dependent response was generally seen, although p19 was shown to be the most potent inhibitor (Lakatos et al., 2006, Dunoyer et al., 2004). Cytosolic p19 selectively binds to 21nt dsRNA (Vargason et al., 2003) prior to RISC formation, but has no effect on existing functional RISC complexes (Lakatos et al., 2006). In addition, Uhrig et al. (2004) reported a relocalisation effect of p19 upon nuclear ALY proteins, although the effects of this interaction are unclear. 19 Given the efficiency with which p19 has been shown to inhibit sRNA biosynthetic pathways, its reduced size (19 kDa) and well-documented phenotype in Arabidopsis, it was chosen as an inhibitory protein for introduction into C. gynandra. The benefits of establishing a stable transgenic line would be three-fold. Firstly, leaf development could be examined to identify any effects of miRNA inhibition on the formation of Kranz anatomy. Secondly, gross phenotypic quantification may highlight any effects at the physiological level attributable to interference by p19 expression. Finally, components of the C4 cycle can be quantified at the transcript and protein level to screen for any p19-induced changes in abundance that could be associated with a relief from miRNA inhibition. 1.6.3: Inducible Expression Systems Due to the effects of p19 expression on development as discussed above, it was considered likely that high levels of constitutive expression may impact leaf development and reduce pollen fertility. Inducible promoters have been used previously to allow for temporal control of genes (e.g. CLAVATA1/2, Schoof et al., 2000) and so the placement of p19 under the control of a chemically inducible promoter was investigated. The dexamethasone-inducible pOp/LhGR system has been shown to generate specific induction of transgenes in tobacco (Samalova et al., 2005) and Arabidopsis (Craft et al., 2005) and so this system was selected for inducible expression of p19. The ability to induce expression of p19 would allow transgenic lines to be cultivated until maturity and then p19 to be induced over a short timescale. This would limit the impact of inhibition of the small RNA pathway to an established C4 system, allowing for effects of such inhibition to be assessed. Long-term exposure to dexamethasone during earlier plant development would also enable the effect of removing small RNA regulation on leaf anatomy to be examined. 1.6.4: A Combined Experimental Approach Following the considerations discussed previously, a twin-tailed approach was taken to examine the potential for sRNAs to regulate the C4 cycle. The production and analysis of a library of small RNA sequences present in mature leaf tissue would allow the identification of potential regulatory sequences on the basis of complementary target/miRNA pairing through computational methods. Any such sequence could then be verified through a number of experimental methods to identify the sites and nature of miRNA activity. This work was carried out in conjunction with the analysis of transgenic lines expressing the viral p19 inhibitor of miRNA biosynthesis. The use of this inhibitor both supports the computational 20 prediction of miRNAs and extends to the scope of the study beyond that which would be possible with a bioinformatic approach alone. The unknown genetic origin of many aspects of the C4 cycle would prevent a full exploration of the effects of miRNAs upon the cycle by computational analysis alone and so it was hoped that the utilisation of the p19 inhibitor would allow an exploration of these areas. In addition the utilisation of such an inhibitor would provide an additional method of validating predictions made by the computational analysis: should a miRNA be targeting a component of the system, then a disruption of miRNA synthesis may cause detectable changes in the expression of the target gene. 1.6.5: Choice of Model Species A considerable amount of knowledge has been accumulated relating to the control of C4 gene expression in maize with examples of regulation occurring at multiple levels as discussed above. The C4 cycle was initially described from maize and sugarcane (reviewed by Hatch, 1999) and the utilisation of maize as a tool to study C4 has been promoted by its importance as a crop plant and technical benefits such as the ease of preparation of purified mesophyll or bundle sheath cellular extracts (Kanai & Edwards, 1973). Genetic resources for maize include an annotated genome sequence (Schnable et al., 2009), transcriptome (Li et al., 2010) and plastid proteome data (Majeran et al., 2008). Small RNA sequencing attempts have enabled miRNA identification in maize (Zhang et al., 2009) with roles in leaf development (Juarez et al., 2004) and stress metabolism (Ding et al., 2009). Maize, however, is deficient in a number of areas as a model species. These include the lack of phylogenetically proximal C3 taxa, limiting the extent to which the value of potential regulators of the C4 cycle can be examined. While a method for maize transformation has been described (Frame et al., 2002) it is laborious and its long life cycle would limit the extent to which stable expressing lines could be established within the timeframe of a PhD program. Various other taxa have been used as model C4 species, including Amaranthus, Flaveria and Atriplex. The presence of C3, C4 and C3/C4 intermediates within Flaveria have enabled various evolutionary trajectories that may lead to C4 to be explored (for example, Holaday et al., 1985; Moore et al., 1989) and a method for transforming the C4 F. bidentis has been described (Chitty et al., 1994). Similar examples of gene-specific regulatory mechanisms have been described in Amaranthus and Atriplex (as discussed above), and transformation systems exist for C4 member of both genera (Jofre-Garfias et al., 1997; Uchida et al., 2003). While the transcriptomes for C3 and C4 members of Flaveria have been published (Gowick et al., 2011) the genetic resources for Amaranthus, Flaveria and Atriplex are still sparse and the application 21 of the resources accumulated for Arabidopsis would be difficult given the phylogenetic distances between the respective taxa. These taxa also have relatively long lifecycles, and only Flaveria has been used as a model C4 species in recent literature. In order to exploit the genetic resources associated with Arabidopsis, Brown et al. (2005) proposed the utilisation members of Cleome (Cleomeaceae) in order to examine the control of C4 photosynthesis. Cleomeaceae has recently been resolved as being sister to Brassicaceae (Hall et al., 2002), and possesses a number of C4 species (Marshall et al., 2007; Voznesenskaya et al., 2007). While a predominately C3 genus, Cleome includes the C3/C4 intermediate C. paradoxa and two NAD-ME subtype C4 species (C. gynandra and C. oxalidea). A third C4 member, C. angustifolia shows novel leaf anatomy which represents an adaptation of traditional Kranz-like anatomy to incorporate additional water storage tissues (Koteyeva et al., 2011). While a genomic sequence is not available for C. gynandra, transcriptome data is available for C. gynandra and the C3 C. hassleriana (Braütigam et al., 2011) and a phylogeny for the genus has been reconstructed (Inda et al., 2008). C. gynandra is transformable (Newell et al., 2010) and has relatively short generation times. The phylogenetic proximity of Cleome to Arabidopsis has enabled the transfer of genetic systems between the two to screen for conserved regulatory elements (Brown et al., 2011; Kajala et al., 2011). Due to the flexibility of Cleome as a potential model system, it was decided that this system would be utilised to examine the potential roles of regulatory small RNAs in the C4 cycle. 1.7: Project Aims 1.7.1: miRNA Identification Two approaches were undertaken to identify potential regulatory roles for sRNAs in the development and maintenance of the C4 cycle. First, a sRNA library was compiled from total RNA extracts prepared from mature C. gynandra leaf tissue. This data was then explored using target sequences of known photosynthesis genes to identify potential candidate targets. This was combined with a study of C. gynandra transgenics expressing the p19 viral inhibitor of miRNA synthesis in order to identify effects of miRNA inhibition upon the C4 cycle. It was considered that the effects of such a disruption of miRNA activity could be examined at multiple levels. Firstly, if a gene was being actively excluded from a given cell type due to suppression by a miRNA, an inhibition of the miRNA may lead to overaccumulation of the gene at the transcript or protein level. Secondly, it was considered that p19 activity may destabilise the C4 cycle to the extent that net photosynthesis would be affected and that examination of the physiological effects of p19 expression would expand the scope of this 22 study beyond those components of the C4 cycle with a known genetic origin. Finally, expression of p19 during leaf development would enable the roles that miRNAs play during leaf development in C. gynandra to be explored, specifically with regards to the formation of Kranz anatomy 1.7.2: Hybridisation Studies As discussed above, the generation of hybrids between C3, C4 and C3/C4 species has been utilised to explore the effects of a mixed genetic background upon the performance of the C4 cycle and the formation of Kranz anatomy. If fertile hybrid lines can be generated hitherto- undescribed genes required for the C4 cycle may be identified through segregation studies. Successful hybridisations with Cleome were reported by Rajndrudu & Rama’Das (1982), and so attempts were made to hybridise C. gynandra with the C3 C. hassleriana. 1.7.3: The Effects of Light during the Development of C. gynandra Cotyledons The onset of the C4 system in developing leaves may respond to a number of environmental or developmental triggers (as discussed above), and it was considered that an analysis of the effects of light exposure in C. gynandra may yield insights into possible environmentally- linked controls of the C4 cycle during early leaf development. Therefore, a study was undertaken to quantify changes in gene expression after light exposure of etiolated cotyledons. Chapter 2: Materials & Methods 23 2. Materials and Methods 2.1: General Plant Cultivation Individual lines of Cleome gynandra, C. hassleriana and Arabidopsis thaliana were obtained from laboratory stocks. Cultivation was carried out on Levington M3 potting compost (Scotts Miracle-Grow, Godalming, UK) supplemented with medium-coarse vermiculite (William Sinclair Horticulture Ltd., Lincoln, UK) (3:1). All soil was treated before use with Intercept® (Scotts Miracle-Gro) at a concentration of 200 mg L-1 per 5 L of soil to prevent sciarid fly infestation. Young plants were kept under propagator lids (Stewart Ltd., Croydon, UK) until established and the first pair of true leaves had formed. Arabidopsis thaliana: Seeds from the Columbia-0 ecotype were sown directly onto soil prior to stratification at 4 °C for 3 days in darkness before transfer to a growth chamber at 20 °C, a photon flux density of 200 µmol m-2 s-1 with either a short day (12 hours) or long day (16 hours) photoperiod. Plants grown for transformation by floral dipping were sown directly onto muslin in pots 9.5 x 8.5 cm and were thinned to c.25 plants per pot. Plants grown for vegetative analysis were grown in 24 well trays under a short day photoperiod. Cleome gynandra: Seeds were sown onto moist filter paper discs (Whatman grade I, Whatman Ltd., Maidstone, UK) in dishes sealed with Parafilm® (Alpha Labs Ltd., Eastleigh, UK) were incubated at 30 °C for 24 hours in darkness. Germinated seedlings were transferred to soil in 9.5 x 8.5 cm pots and were cultivated at 22 °C, of 350 µmol m-2 s-1 with a 16 or 12 hour photoperiod. For light shift experiments, seeds were germinated in the normal way before transfer to Cg media (see below) for in vitro culture. The work was carried out under “green- safe” light filtered through a HT124 filter (Stage Electric Ltd., Bristol, UK), and the pots wrapped in aluminium foil before incubation at room temperature of 7 days. C. hassleriana: Cultivation matched that of C. gynandra, except that germination was achieved by incubation for seven days at 16 hours light at 28 °C followed by 8 hours in darkness at 18 °C. Once fully established, the plants were transferred to 12 cm square pots to allow for maximal growth before flowering. Chapter 2: Materials & Methods 24 2.2: Molecular Techniques 2.2.1: DNA Amplification Amplification of target DNA sequences was achieved through the use of the polymerase chain reaction (PCR, Kleppe et al., 1971, Saiki et al., 1985). A TC-312 thermocycler (Techne, Staffordshire, UK) was used to perform the reactions using either BioTAQ® DNA polymerase (Bioline) or Phusion High-fidelity DNA polymerase (Finnzymes, Espoo, Finland). Under general conditions, 0.2mM dNTPs (Bioline) and 2 ng µl-1 of primers (Invitrogen) were used against 1-100 ng of template DNA. When Phusion was employed, the reaction mix was heated to 98 °C for 30 s prior to the addition of 0.02 U µl-1 of the polymerase and the initiation of the reaction cycle. A typical reaction sequence is as follows: initial denaturatisation at 94 °C for 3 minutes, the 25-30 cycles of 94 °C for 30 s, primer annealing at 55 °C for 30 s, extension at 72 °C for 1-2 minutes, then final extension for 7-10 minutes at 72 °C. Polymerase choice, PCR component concentration and cycling conditions were optimised in each case according to amplicon length, primer annealing temperature and the manufacturer’s guidelines and are stated in the text accordingly. Agarose Gel-phase Electrophoresis DNA fragments and RNA extracts were visualised after size-separation on gels of 0.8–2% agarose (Melford Laboratories Ltd., Suffolk, UK) made with ½ strength tris-borate-EDTA buffer (TBE): 44.5 mM tris(hydroxymethyl)aminomethane (Tris, Melford), 44.5 mM boric acid (Fisher Scientific) and 1 mM diaminoehtetetra-acetic acid disodium salt (EDTA, Fisher Scientific) at pH 8.0. After heating to dissolution, the agarose gel was cooled before pouring into casting moulds (Bio-Rad, Hercules, USA) followed by the addition of ethidium bromide (Sigma) to a concentration of 0.5 µg ml-1 (Sambrook et al., 1989). DNA samples were prepared for loading by the addition of DNA loading buffer (40% sucrose (Fisher Scientific), 0.4% bromophenol blue (Sigma), which was used at a 1:5 concentration. RNA samples were prepared with RNA loading buffer: 50% glycerol (Fisher Scientific), 0.4% bromophenol blue (Sigma), 1 mM EDTA (Fisher Scientific), 48% formamide (Sigma) and 0.05% TBE) used at a 1:5 concentration and the samples were heated to 65 °C for 5 minutes before loading. Hyperladder® II (Bioline) was used as a size marker, and the gels were run for c.40 minutes at 80 V in a ½x TBE running buffer in Wide Mini-Sub cell tanks (Bio-Rad). Gels were then visualised on an Alpha Imager UV transilluminator (Alpha Innotech, San Leandro, USA) before photographing with a MultiImage light cabinet and Alpha-Ease FC Imaging System software (Alpha Innotech). Chapter 2: Materials & Methods 25 Purification of DNA Bands Gel extraction of DNA fragments separated by electrophoresis was performed using the QIAquick Gel Extraction Kit (Qiagen) following the manufacturer’s guidelines. PCR products were run out on a 1% agarose gel as detailed above, and the required bands excised using a fresh razor blade. The resulting DNA extracts were eluted into 25µl of water before being quantified via spectrophotometry. 2.2.2: RNA Preparation and cDNA Synthesis Young, fully developed leaf material was harvested at 10am and flash-frozen in liquid nitrogen before storage at -80 °C. Total RNA extraction was achieved through phenol/chloroform extraction (Chomczynski & Sacchi, 1987) using TriPure® Isolation Reagent (Roche Diagnostics Ltd., Burgess Hill). The RNA was precipitated from the aqueous phase after extraction with a 1:1 mix with 0.8 M sodium citrate, 1.2 M sodium chloride (all Fisher Scientific) and incubation at -20 °C for 16 hours. The RNA was pelleted by centrifugation at 16000g for 30 minutes at 4 °C before cleaning in 70% and 100% ethanol (Fisher Scientific) and resuspension in water. All water used in RNA-orientated reactions was treated with diethylpyrocarbonate in a 1 in 1000 dilution and autoclaved prior to use to inhibit RNase activity. After resuspension, the RNA was treated with RQ1 DNase (Promega, Madison, WI, USA) before column cleaning and purification (RNeasy® Kit, (Qiagen, Hilden, Germany). The RNA was the quantified directly using the Nanodrop® spectrophotometer (Thermo Scientific, Waltham, MA, USA). The quality of the RNA was confirmed either by microcapillary electrophoresis (2100 Bioanalyser, Agilent Technologies Ltd., Santa Clara, USA) or by electrophoresis on a 1% agarose gel. In the latter case, quality was assessed via the appearance of the rRNA bands, while in the former a RNA Integrity Number (RIN, Schroeder et al., 2006) greater than 7 was considered good quality. The RNA was then stored at -80 °C prior to use. 1 µg of total RNA was used for cDNA synthesis using SuperScript® II reverse transcriptase (Invitrogen, Carlsbad, CA, USA) following the manufacture’s guidelines using oligo (dT)15 (Roche) to prime the reaction. The cDNA was diluted 1:5 before storage at -18 °C. 2.2.3: Quantitative RT-PCR Relative transcript abundance of a given gene was assessed with quantitative real-time PCR (qRT-PCR, Bustin et al., 2000). cDNA was obtained following the method given above, before qRT-PCR reactions were carried out using a Rotor-Gene Q centrifugal real-time thermocycler (Qiagen) and associated analytical software. cDNA synthesis was performed using 1µg of total RNA, the products of which were diluted 15-fold before use as a template in a 10µl reaction mix Chapter 2: Materials & Methods 26 based on SYBR Green Jump Ready Mix (Sigma). Initial denaturisation was carried out at 95°C for 5 minutes before 40 cycles of 95 °C for 20s, 58 °C for 30s, 72 °C for 30s and 75 °C for 5 s, at which point the sample fluorescence was recorded. After the completion of cycling, a melting curve was produced between 55-95 °C recording every 1°. Primers for qRT-PCR were designed to have a melting temperature of 60 °C (±½ °C) and to produce an amplicon of 100-300bp. Actin was chosen as a basic reference gene due to consistent levels of expression in C. gynandra and C. hassleriana (Bräutigam et al., 2010), so ACTIN7 was used for Cleome qRT-PCR reactions, although ACTIN8 was used in analysis of Arabidopsis transcripts. Primer pairs against C4 genes in Cleome were taken from Bräutigam et al. (2010), except for those developed for this study which were designed using the IDT PrimerQuest software (http://eu.idtdna.com/Scitools/Applications/Primerquest) prior to synthesis by Invitrogen. All primer sequences are given Appendix 1. In each instance, Ct values were produced for three technical replicates of at least three biological replicates and were calculated using a threshold of 0.02 for all transcripts so that fluorescence was recorded in the exponential phase of amplification. Reference actin qRT-PCR reactions were run alongside each gene of interest, and the comparative 2-ΔΔCt value was calculated to allow for quantification of transcript abundance (Livak & Schmittgen, 2001), and standard errors were calculated from each combination of biological replicates. If a given reaction produced Ct values smaller than 12, or with variation greater than 0.25 between technical replicates the reaction was repeated with a 1 in 10 dilution of the original template. 2.2.4: Small RNA Library Preparation The sRNA population of C. gynandra was sequenced following the methods of Havecker (2011) using materials supplied with the Illumina® sRNA TruSeq™ library preparation kit (see Figure 2.1). RNA was extracted from mature leaf material following the methods described above until the stage of ethanol precipitation. In short, the 18-25nt fraction of the total leaf RNA was purified prior to ligation of 5’ and 3’ adaptors before reverse transcription and PCR amplification of the library. The sRNA fraction was purified using a 15% TBU denaturing polyacrylamide gel comprised of 42% urea (w/v, Fisher Scientific), 1x TBE and 15% acrylamide (19:1 acrylamide:bisacrylamide, Severn Biotech Ltd., Worcestershire, UK) set with 0.05% ammonium peroxodisulphate (APS, w/v, BDH) and 0.05% N,N,N’,N’tetramethylethylenediamine (TEMED, v/v, Severn Biotech Ltd.). The gel was prerun at 200 V in 1x TBE buffer and the lanes washed out prior to loading of 10 µg of total RNA that had been prepared for loading as described above. In addition to the Chapter 2: Materials & Methods 27 standard running ladder, a 20-100bp low ladder (Sigma) and Hyperladder II® were run alongside the sample. The gel was run in an AE-6450 Dual Mini Vertical PAGE system (Atto, Tokyo, Japan) at 150 V until the bromophenol blue front had reached the bottom of the gel. The ladder portion of the gel was then placed in a 1xTBE solution containing 0.225 µg ml-1 ethidium bromide for 5 minutes before being briefly rinsed in fresh TBE. The gel was visualised on a M26 UV transilluminator (Alpha Innotech) at 360nm and the RNA section of the gel corresponding to 15-30nt was excised with a fresh razor blade. The gel fragment was shredded by brief centrifugation through a microfuge tube that had been punctured with a 21- gauge needle before the addition of 500 µl 0.3 M NaCl. The gel was left at a slow rotation at 4 °C for 16 hours to elute the RNA. The gel/NaCL mix was then transferred to a Spin-X® cellulose/acetate filter (Sigma) and spun at 12000g for 2 minutes, followed by a second washing using 100µl of 0.3 M NaCL. The elutate was combined with 600 µl 100% isopropanol and 3 µl of a co-precipitant stain GlycoBlue™ (Ambion, Austin, USA) before incubation at -80 °C for 1 hour. The RNA precipitate was then pelleted and washed as described above before resuspension in 5.7 µl of RNase-free water. The 5.7 µl RNA eluate was heated to 90 °C for 30 s before snap cooling on ice prior to the ligation of the 5’ adaptor following the manufacturer’s guidelines. The ligation reaction consisted of the 5’ adaptor at 1.3 µM, 10 units of T4 RNA ligase, 40 units of RNAaseOUT and 1x RNA ligation buffer. The reaction mix was incubated at 37 °C for one hour before termination of the reaction by the addition of 10 µl of RNA loading buffer. The reaction mix was then run on a 15% TBU gel at 200 V, before excision of a gel slice corresponding to the 40-60nt region. The gel slice was macerated, the RNA eluted in 0.3 M NaCl before precipitated as described above. The resultant RNA pellet was then resuspended into 6.4 µl of RNase-free water. The full RNA product was then used for the 3’ adaptor ligation reaction, following the same procedure as given above, although the 3’ adaptor concentration was at 0.5 µM. The product of the 3’ ligation reaction was separated out using a 10% TBU gel, and the fraction corresponding to 60- 80nt was excised and used for RNA elution into 4.5 µl of RNase free water, as outlined above. The full 4.5 µl RNA elutate was used for the reverse transcription reaction. The reaction was carried out using SuperScript™ II reverse transcriptase following the manufacturer’s guidelines using a siRNA RT primer (Illumina) to prime the reaction. 4µl of the RT-PCR product was used in an 80 µl PCR reaction mix using Phusion® polymerase as described above, using Illumina®-supplied primers at a concentration of 0.625 µM. The reaction mix was subdivided into four separate reactions. Initial denaturisation at 98 °C for 30 s was followed by 15 cycles of Chapter 2: Materials & Methods 28 98 °C for 2 s, 58 °C for 30 s and 72 °C for 20 s before a final extension at 72 °C for 5 minutes. The separate reactions were then recombined prior to the addition of 8 µl 3 M sodium acetate, 200 µl ethanol and 3 µl GlycoBlue™ before incubation at -80 °C for one hour to precipitate the PCR products. After centrifugation at 18ooog, the pellet was washed in 100% ethanol before air drying and resuspension in 20 µl of sterile water and 3µl of DNA loading buffer as described above prior to loading on a 6% TBE gel. The gel was run at 100V until the xylene cyanol front reached the bottom of the gel before staining with ethidium bromide and excision of a gel slice corresponding to 92bp from the sample lane. The gel was then macerated, and the DNA extracted using the QIAquick gel extraction kit (Qiagen) following the manufacturer’s guidelines. The concentration and quality of the amplified library was verified as described above before submission for Illumina® sequencing-by-synthesis method version 1. Figure 2.1: Sequencing and Processing of the sRNA Library Total leaf RNA was extracted (A) before purification of the 15-30nt fraction using TBU-PAGE (B) was carried out. Separate reactions to ligate the 5’ (C) and 3’ (D) adaptor were then performed, each followed by purification of the 40-60nt and 60-80nt fraction respectively. The ligated RNA was then used for cDNA synthesis (E) to provide a ssDNA template for PCR amplification (F). The amplified library was subject to Illumina® sequencing v.1. After sequencing, the library was processed to remove noise reads (determined by length, composition and adaptor content) to provide a unique read library that was used as a basis for miRNA prediction based on direct and mismatched complementarity to candidate EST sequences. The reads were then subject to the given criteria for the in silico prediction of miRNA sequences as outlined. A: Total Leaf RNA Extract B: sRNA Fraction (15-30nt) C: 5’ Adaptor Ligation (40-60nt) D: 3’ Adaptor Ligation (60-80nt) E: cDNA Synthesis F: Library Amplification Illumina ® Sequencing Candidate EST sequences searched against the sRNA library for directly complementary Reads High scoring candidate sequences searched against the sRNA library for Reads with up to 4 Mismatches Alignment of Reads against candidate sequences Mismatch Positions and Read Frequency Identified Candidate Genes and Proposed miRNA Sequences subject to Experimental Validation Chapter 2: Materials & Methods 29 2.2.5: Library Processing On receipt of the library, the sequences were processed to produce a library of adaptor-free, non-redundant reads using the SIROCCO ADDAPTS processing pipeline of Nag et al. (In Submission). The sequences were first filtered for reads of lengths between 15-50nt, as the sequences obtained are comprised of the sRNA fused to the 3’ adapter. The starting 8nt of the 3’ adaptor were then identified, and the sequences upstream from this site cleaved to leave the sRNA sequence. This was then subject to further filtering to produce a library of unique reads between 15-35nt in length. Due to the lack of a genome for C. gynandra, the library was initially examined using selected target EST sequences of the transcriptome data of Braütigam et al. (2011) as given in Appendix 2. Initial searches were made for direct complementarity using the algorithms of Moxon et al. (2008). Those sequences which showed an above average accumulation of hits against the library were then reanalysed using a modified form of the Bowtie alignment algorithm (Langmead et al., 2009) to allow for alignment of sRNA sequences possessing up to 4 mismatches against an EST sequence of interest. Such sequences were then examined for high levels of abundance in the unfiltered library and for fulfilment of the miRNA targeting requirements as discussed above to identify any potential regulatory components (see Figure 2.1). Alignment visualisation was carried out using the Integrative Genomics Viewer of Robinson et al. (2011). 2.2.6: Cleavage Product Identification via 5’ RACE Validation of a predicted cleavage-directing miRNA was carried out using 5’ Rapid Amplification of cDNA Ends (RACE, Frohman et al., 1988) following methods of Llave et al. (2002b) and Kasschau et al. (2003) using the FirstChoice® RLM RACE kit (Ambion). Nested primer pairs were developed against the midsection of the target transcript (see Appendix 1), and were initially used following the full manufacturer’s protocol to test for effectiveness of the primer pairs in amplifying the target transcript. The FirstChoice® kit includes treatment with Calf Intestinal Phosphatase to remove 5’ terminal phosphates of non-degraded transcripts prior to treatment with Tobacco Acid Pyrophosphatase, 5’ adaptor ligation and RT- PCR. 10 µg of total RNA taken from mature C. gynandra leaf material was used for each reaction. The product of the RT reaction was then used for two rounds of PCR using nested primers to obtain a high level of product specificity using hot start reactions mediated by the Phusion® high fidelity polymerase as described above. The nested reverse primers were designed to match the forward primers supplied with the FirstChoice® kit, although melting temperature optimisation was carried out using Chapter 2: Materials & Methods 30 temperature-graded PCRs in a Techne TC-512 gradient thermocycler. The PCR products were run out on a 1% (w/v) agarose gel, and bands of the predicted size were excised from the gel prior to DNA extraction using the QIAquick gel extraction kit (Qiagen). The products were then cloned using the ZeroBlunt® TOPO PCR cloning kit (see below), and sequenced to obtain target verification. Once the products were fully identified, the process was repeated following the modified procedure of Llave et al. (2002b). 2.3: Quantification of Protein 2.3.1: Protein Extract Preparation Material for protein extraction was flash frozen in liquid nitrogen before storage at -80°C. The material was then finely ground using a micropestle under liquid nitrogen before the addition of an extraction buffer comprised of 0.1 M sodium chloride, 1 mM magnesium chloride and 0.05 M tris at pH 7.8 to which had been added EDTA-free protease inhibitor cocktail (Roche). The material was vortexed briefly before pelleting at 4000g for 15 minutes at 4 °C before the removal of the supernatant. Total protein content was quantified via the methods of Bradford (1976) and Jones et al. (1989) using the Bio-Rad Protein Assay Reagent (Bio-Rad), following the manufacturers protocol. The supernatant was then mixed 5:1 with protein loading buffer comprised of 0.1 M dithiothreitol (DTT, Sigma), 0.3 M sodium dodecyl sulphate (SDS, Melford), 0.2 M tris, 1 M sucrose and 0.4% bromophenol blue before storage at -20 °C. 2.3.2: Separation of Proteins via Electrophoresis Total leaf protein was separated via SDS/polyacrylamide gel electrophoresis after the methods of Maizel (1966) and Shapiro et al. (1966). A stacking gel of 5% acrylamide (37.5:1 acrylamide:bisacrylamide, (Severn Biotech Ltd.), 125 mM tris, 3.5 mM SDS, 0.05% ammonium peroxodisulphate and 0.05% TEMED at pH 6.8 was poured over a set resolving gel of varying percentage. 14% gels were used for resolving c.20 kDa proteins, while a 10% gel was used for proteins in the range of 50-100 kDa. Besides acrylamide, the resolving gel was composed of 375 mM tris, 3.5 mM SDS, 0.06% APS and 0.1% TEMED at pH 8.8. Electrophoretic separation was carried out using an AE-6450 Dual Mini Vertical PAGE system (Atto, Tokyo, Japan) using a running buffer comprised of 25mM tris, 0.2M glycine (Fisher Scientific) and 1.75 mM SDS at pH 8.5. Twin gels were run in tandem at 50 mA until the bands reached the bottom of the stacking gel, after which the current was increased to 70 mA until the bands reached the bottom of the gel. One gel was stained with Coomassie stained using InstantBlue™ (Expedeon, Cambridge, UK) for one hour to allow for protein visualisation, after which the gel was dried between two layers of gel drying film (Promega) for long-term preservation. Before loading, Chapter 2: Materials & Methods 31 aliquots of protein samples were heated to 95 °C for 5 minutes to ensure sufficient denaturisation. 2.3.3: Immunoblot Analysis The second gel produced by electrophoresis was used for immunoblotting following the methods of Towbin et al. (1979) and Renart et al. (1979). Transfer to a membrane was achieved through the use of a Mini Trans-Blot Cell (Bio-Rad) which was run using a running buffer of 50 mM tris, 40 mM glycine, 1.3 mM SDS and 20% methanol (Fisher Scientific) at 100 V for 1 hour, whilst cooling with an ice pack. The separated proteins were transferred onto Protran™ nitrocellulose transfer and immobilisation membrane (Perkin Elmer, Massachusetts, USA) with a pore size of 0.2 µm. The membrane was then incubated in blocking solution of 5% milk powder (Marvel™ dried skimmed milk powder, Premier International Foods Ltd., Lincolnshire, UK) in tween/tris-buffered saline (TTBS, 50 mM tris, 0. 2 M sodium chloride with 0.1% Tween-20 at pH 7.5) for 1 hour at room temperature while shaking at 50 rpm. After rinsing briefly in TTBS, the primary antibody was applied at the appropriate dilution in TTBS and left to hybridise over night at 4°C while shaking at 50 rpm. The membrane was then washed in blocking solution five times (for five minutes each) before application of the secondary antibody, also in blocking solution, for 1 hour while shaking at room temperature. In all cases, the secondary antibody used was anti-Rabbit IgG biotinylated antibody raised in Donkey (GE Healthcare, Wisconsin, USA) which was used in a 1 in 1000 dilution. The membrane was then washed in blocking solution for a further five times before rinsing in TTBS before application of streptavidin-biotinylated horseradish peroxidase complex (GE Healthcare) in a 1 in 1000 dilution in TTBS for 30 minutes. The membrane was then rinsed three times in TTBS before two washes in sterile water, each for five minutes while shaking. Enhanced chemiluminescent detection was carried out using the Western Lightning™ Plus- ECL substrate (Perkin Elmer), medical X-ray film (Konica Minolta, Tokyo, Japan) and a compact X4 X-ray film processor (Xograph Imaging Systems, Gloucestershire, UK) following the manufacturer’s guidelines in each instance. Numerical data was derived from western blot images were calculated using the ImageJ (National Institutes of Health, USA) using the ‘Rectangular Selections’, ‘Histogram Plot’ and ‘Measure’ functions. Chapter 2: Materials & Methods 32 2.4: Plasmid Design and Construction 2.4.1: Design of Plasmids The 35S::p19::nos construct was obtained as a 13.5 kp binary vector (Voinnet et al., 2003) while the pOpON2.1 promoter (Moore et al., 2006) was obtained as a 20.1 kb binary vector with an empty destination site for the Gateway™ cloning system (Invitrogen). The p19 coding sequence was cloned from the 35S::p19::nos construct and inserted into the Gateway™ destination site of the pOp promoter following the methods given below. The final plasmid blueprint is given in Figure 2.1, while full restriction site maps are given in Appendix 3. Figure 2.2: p19 Expression Constructs A: 35S::p19::nos: A 13.5kb binary vector carrying streptomycin resistance regulated by the octopine synthase (OCS) promoter, kanamycin resistance (neomycin transferase, nptII) regulated by the nopaline synthase (nos) promoter (P) and terminator (T) placed upstream of p19 promoted by the CaMV 35S promoter and terminated by nos (T). PvuI restriction sites are marked, which yield fragments of 2781 and 10677bp upon digestion. B: pOp::p19: A 19.3kb binary vector carrying the constitutively expressed spectinomycin (SpecR, regulated by T7), nptII (regulated by Nos (P) and (T)), and LhGR transcription factor, promoted by 35S. β-glucuronidase (uidA) and p19 are promoted by the pOp promoter but are transcriptionally isolated by the presence of a 35S and OCR polyadenylation signal respectively. Nhe1 restriction sites are marked, which yield fragments of 7006, 10931 and 1373 upon digestion. Chapter 2: Materials & Methods 33 2.4.2: Vector Production The pOpON2.1 binary vector was transformed into chemically competent Escherichia coli DH5α cells (as described below) which were used to inoculate solid LB plates containing 100 µg ml-1 specintomycin (Melford) before incubation at 37 °C for 16 hours. A single colony was then used to inoculate 5ml LB liquid cultures with the same selective antibiotic which was incubated at 37 °C for 16 hours while shaking before being used to prepare the raw pOpON2.1 plasmid via alkaline lysis and elution into TE buffer (10 mM Tris, 1 mM EDTA, pH 7.5). The plasmid was then verified by restriction enzyme digestion and sequence analysis as described below and above respectively. The plasmid was then stored at -20 °C until use. The p19 coding sequence was amplified as a blunt-ended PCR product using Phusion® high- fidelity polymerase as described above. Primers were designed to amplify the entirety of the p19 coding sequence from the 35s::p19::nos construct with the addition of a CACC sequence to the 5’ terminus after the ATG site (see Appendix 1). After verification by sequencing, the p19 amplicon was inserted into the pENTR/D-TOPO® vector using the Zero Blunt® TOPO® PCR cloning kit (Invitrogen) following the manufacturers guidelines to produce the pENT/p19 entry vector. The product of the cloning reaction was transformed into chemically competent E. coli DH5α cells which were cultured on solid LB media containing kanamycin (Melford) at 50 µg ml-1 as a selective antibiotic. Single colonies were then used to inoculate 5ml LB liquid cultures with the same selective antibiotic which we incubated at 37 °C for 16 hours before being used to prepare the raw pENTR/p19 plasmid via alkaline lysis and elution into TE buffer. The prepared plasmid was then verified by restriction enzyme digestion and sequencing before storage at -20 °C until use. 2.4.3: Gateway™ Recombination of Vectors Recombination between the pOpON2.1 destination vector and pENTR/p19 entry vector was achieved through the use of Gateway™ LR clonase II Enzyme Mix (Invitrogen), following the manufacturers guidelines. The LR clonase mediates recombination between a region boarded by att sites in the entry and destination vectors. The insertion of the CACC sequence before the p19 ATG site ensures this occurs in a directional manner so that the p19 insert would be placed downsteam of the pOp promoter in a 5’ → 3’ orientation. Upon completion of the LR reaction, a lethality gene is removed from the destination vector and is replaced with the required insert so that only complete recombination products will proliferate in subsequent cultures. The products of the LR reaction were transformed into E. coli DH5α cells, before subculturing and plasmid preparation as described below. The resultant pOp::p19 construct Chapter 2: Materials & Methods 34 was then verified by restriction digests and sequencing prior to transformation into corresponding Agrobacterium strains as described below. 2.4.4: Plasmid Isolation and Verification Plasmids were obtained via alkaline lysis (Bimboim & Doly, 1979) from cultures of Escherichia coli DH5α cells produced by the methods discussed below. 5 ml liquid cultures of LB broth containing the appropriate antibiotics were established from streak plates and incubated at 37 °C for 20 hours. 2ml aliquots of cultures were pelleted twice in a bench-top centrifuge (260D, Denville Scientific Inc., Metuchen, USA) before plasmid extraction using the QIAprep Spin Miniprep kit (Qiagen) following the manufacturer’s protocol. Final elution of the plasmid yield was made in either water or TE buffer (10 mM Tris-Cl (Melford), 1mM EDTA (Fisher Scientific) at pH 7.5 with NaOH) depending on the eventual downstream application. The extraction yield was then spectrophotometrically determined (50 μg ml-1 dsDNA A260 = 1.0). Plasmids were verified through a process of restriction digestion (see below) to test for fragment size, followed by sequencing via the chain-termination method (Sanger & Coulson, 1975, Sanger et al., 1977). This was carried out by the DNA Sequencing Facility (Department of Biochemistry, University of Cambridge) using a 3730xl DNA Analyser (Applied Biosystems, Foster City, USA) using 1½ μg of raw plasmid and primers designed against the predicted sequence. Chromatogram and sequence analysis was carried out using BioEdit (Hall, 1999) and BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi). Purified plasmids were subject to restriction digestion to release DNA fragments of sizes based on the analysis of their predicted sequence. 10 µl reactions were carried out using enzyme preparations and corresponding buffer solutions (NEB, Massachusetts, USA) and 0.5 µg of DNA. The reaction was performed following the manufacturer’s guidelines before inactivation by heating to 65 °C for 15 minutes prior to visualisation on an agarose gel of a corresponding strength. The enzymes used with each plasmid are discussed above. Chapter 2: Materials & Methods 35 2.5: Microbial Techniques 2.5.1: Production of Competent Escherichia coli DH5α Cells Competent DH5α cells were prepared following a method derived from Hanahan (1983). Cells were plated out on Luria-Bertani (LB) medium comprised of 1% Bacto™tryptone (BD, Franklin Lakes, USA), 1% (w/v) sodium chloride (Fisher Scientific), 0.5% (w/v) yeast extract (Oxoid, Cambridge, UK) solidified with 1.5% (w/v) agar (Melford), pH 7.5 with NaOH. After incubation at 37 °C for 20 hours a single colony was used to inoculate 5ml of Ψ broth comprised of 2% Bacto™ tryptone (w/v), 0.5% yeast extract (w/v), 0.4% magnesium sulphate (w/v, Sigma), 10mM potassium chloride (w/v, Sigma) at pH 7.6 via NaOH. After incubation at 37 °C for 16 hours at a rotation of 200 rpm, 1ml of the culture was used to inoculate a further 100ml of Ψ broth which was then cultured until such time that the growth rate had entered the mid-log phase. The culture was cooled on ice for 15 minutes before centrifugation at 1000g at 4 °C and resuspension in chilled TFBI (100mM rubidium chloride, 50mM magnesium chloride, 30mM potassium acetate, 10mM calcium chloride, 15% glycerol (v/v) taken to pH 5.8 with 0.2 M acetic acid (all Sigma)). After 15 minutes incubation on ice, the suspensions were centrifuged again before resuspension in 8ml of TBII (10 mM MOPS, 10 mM rubidium chloride, 75 mM calcium chloride, 15% glycerol (v/v) at pH 7.0 with NaOH (all Sigma)). 200μl aliquots of the resuspended cells were flash-frozen on dry ice before storage at -80 °C. The efficacy of the preparation was tested by a trial transformation with selective antibiotic genes. 2.5.2: Transformation of Competent Escherichia coli DH5α Cells Vectors were inserted into DH5α cells using heat shock transformation (Hananhan, 1983). 5 μl of chemically competent DH5α cells were incubated with 2 μl of the product of a cloning reaction on ice for 20 minutes before heating to 42 °C for 1 minute and then rapid cooling on ice for 3 minutes. 500 μl of SOC medium was added (Super Optimal broth with Catabolite repression, 2% Bacto™tryptone (w/v), 0.5% yeast extract (w/v), 0.05% sodium chloride (w/v), 25mM potassium chloride (w/v), 0.01M magnesium chloride (w/v) (all Sigma) with 20mM glucose (Fisher Scientific) at pH 7.0 with NaOH) before incubation at 37 °C for 1 hour while shaking at 200 rpm. 200 and 300 μl of the resultant culture was spread on LB plates containing either 50 μg ml-1 kanamycin sulphate, 100 μg ml-1 streptomycin sulphate or 50 μg ml-1 spectinomycin dihydrochloride (All Duchefa Biochemie, The Netherlands). Colonies were screened for successful transgenic colonies after 20 hours incubation at 37 °C. Chapter 2: Materials & Methods 36 Confirmation of plasmid transfer was initially carried out via PCR screen before subsequent verification through sequencing of the purified plasmid (see 2.2.4). Primers complementary to the insert were used in an amplification reaction using 0.5 units BioTaq® DNA polymerase (Bioline) in a 10μl reaction mix containing 1 mM magnesium chloride (Bioline), 0.25 mM dNTPs (Bioline) with a primer concentration of 2 ng ul-1 (Invitrogen). Colonies were selected using a sterile pipette tip were added to the reaction mix after streaking on a fresh LB plates containing the appropriate antibiotics. The thermocycler conditions were as follows: initial denaturation at 94 °C, then 27 cycles of 94 °C for 30s, 55 °C for 30s, 72 °C for 1 minute followed by a final extension period of 7 minutes at 72 °C. Streaks were identified as positive by visualisation of an amplification product on a 1% agarose gel as described previously. To establish bacterial stocks, positive streak colonies were used to inoculate 5ml LB liquid cultures containing the corresponding antibiotic and incubated at 37 °C for 20 hours while shaking. A 500μl aliquot of culture was combined with an equal volume of sterile 80% glycerol (v/v, Sigma) before storage at -80 °C. 2.5.3: Production of Electrocompetent Agrobacterium tumefaciens Cells The production of electrocompetent A. tumefaciens cells was achieved via a protocol derived from Fisher & Guiltinan (1995) for both the LBA4404 and GV3101 strains. Cells taken from a central stock of each strain were streaked onto solid LB media before incubation at 30 °C for 24 hours. A single colony was then used to incubate a 5ml culture of LB medium containing either 25 µg ml-1 rifampicin (GV3101, Melford) or 100 µg ml-1 streptomycin sulphate (LBA4404) which was incubated at 30°C for 20 hours while shaking at 200rpm. 1ml of the resultant culture was then used to inoculate a 100ml liquid LB culture with selective antibiotics which was incubated for a further 3 hours until the bacteria reached the mid-log phase of growth with an OD560 of 0.45-0.6. The culture was then pelleted by centrifugation at 4000g for 15 minutes at 4°C before resuspension in 100ml chilled 10% glycerol. This process was repeated with 50 ml, 20 ml and 4 ml of 10% glycerol. The cells were then separated into 45 µl aliquots before flash freezing on dry ice and storage at -80 °C. 2.5.4: Plasmid Entry into Agrobacterium tumefaciens Cells Plasmids were inserted into either strain of A. tumefaciens by electroporation (Schen & Forde, 1989). 250 ng of purified plasmid DNA was added to a thawing 45 µl aliquot of cells and transferred to an electroporation cuvette (Bio-Rad) after 5 minutes incubation on ice. The cells were then subject to an electroporation pulse using a Bio-Rad Gene Pulser set to 400 Ω, 2.5 mV with a capacitance of 25 mF. Directly after electroporation 1ml of SOC media was applied Chapter 2: Materials & Methods 37 and the cells incubated at 30 °C while shaking at 180rpm for three hours. At this point, the cells were spread over solidified LB media containing the corresponding selective antibiotics (streptomycin for LBA4404, rifampicin for GV3101, kanamycin for 35s::p19::Nos, spectinomycin for pOp::p19) and incubated at 30 °C for 36–48 hours to produce colonies. These were then subject to assessment by PCR screen to identify the positive colonies before the establishment of glycerol stocks as described for E. coli above. 2.6: Transformation of Cleome gynandra The transformation of C. gynandra was performed in vitro from cotyledon and hypocotyl explant material following the method derived by Newell et al. (2010). All reagents were supplied by Duchefa Biochemie, unless otherwise stated. Agrobacterium tumefaciens strain LBA4404 was used for C. gynandra transformations. 2.6.1: Provision of Plant Material and Bacterial Stocks Fresh seeds taken from unopened pods were sown directly onto moist sterile filter paper in sealed plates and incubated in darkness at 30 °C for 30 hours. Germinating seeds with a hypocotyl greater than 7mm were transferred to 9cm-diameter sterile pots containing MS medium (Murashige & Skoog, 1962) supplemented with the vitamins of B5 medium (Gamborg et al., 1968) and 1% sucrose (Fisher Scientific) at pH 6.0 and set with 0.8% agar (Melford) at a density of 30 seeds per pot. The pots were then incubated at 24 °C at a light intensity of 50 µmol m-2 s-1 with a 16 hour photoperiod for a period of 7 days. 200µl of stock cultures were used to inoculate 50ml of YEP medium (An et al., 1988), comprising of 1% Bacto™Peptone (BD), 1% yeast extract (Oxoid) and 0.5% sodium chloride (Fisher Scientific) at pH 7.0 in a sterile 250ml conical flask. In all instances kanamycin sulphate (50 µg ml-1) was used as a selective antibiotic, with the addition of streptomycin sulphate (100 µg ml-1) for the 35S::p19::nos construct and spectinomycin dihydrochloride (50 μg ml-1) for the pOp::p19 construct. The culture was incubated at 30 °C for 24 hours while shaking at 200rpm. The culture was cooled on ice for 15 minutes before centrifugation at 1000g at 4°C and resuspension in chilled SIM (Simplified Induction Medium, Alt-Mörbe et al., 1989) comprised of 2% sucrose (Fisher Scientific), 0.59% trisodium citrate and 0.2mM acetosyringone (Acros Organics) at pH 5.5. Chapter 2: Materials & Methods 38 2.6.2: Transformation and Regeneration Transformation of C. gynandra was achieved following the methods of Newell et al., (2010). Explant material was prepared by excising each cotyledon at the base, followed by a second incision at the cotyledon apex to provide a second wound site. The root tip and apical bud was removed from the hypocotyl. Both hypocotyl and cotyledon (abaxial surface downwards) material were placed into the Agrobacterium inoculum at a density of 25 seedlings per 10ml of inoculum and left for 30 minutes with gentle shaking. The explant material was then transferred to co-culture plates comprised of 1/10 MS medium supplemented with 3% sucrose, 1mgl-1 benzylaminopurine (BAP), 0.1mgl-1 napthalene acetic acid (NAA) at pH 5.5, set with 0.8% agar. The solidified medium was covered with a disc of sterile filter paper moistened with 1ml of liquid MSO (MS medium with 3% sucrose at pH 5.5) on top of which the explant material was placed, with the cotyledons placed abaxial surface placed downwards. Co-culture plates were incubated for 2 days under the conditions given above before the explant material was transferred to regenerative media comprised of MS medium supplemented with the vitamin cocktail of Nitsch & Nitsch (1969), 0.5 g l-1 2-[N-morpholino]ethanesulfonic acid (MES, Sigma), 0.5 g l-1 poly(vinyl-polyrrolidone) (PVP, Sigma), 40 mg l-1 adenine hemisulphate, 10 mg l-1 silver nitrate (added after autoclaving), 0.1 mg l-1 NAA, 1 mg l-1 BAP, 2% sucrose at pH 5.7 and solidified with 0.8% agar. Kanamycin sulphate (60 mg l-1) was added to the cooled media as a selective antibiotic in addition to carbenicillin (500 mg l-1) to prevent Agrobacterium overgrowth. The cultures were grown under the conditions given above for three weeks, at which point they were subcultured onto fresh selective medium. After 5 weeks, explant material showing regenerative callus was transferred to selective A6 medium (De Block et al., 1989). After a further three weeks callus balls showing significant leaf formation were transferred to A7 medium (De Block et al., 1989) to promote shoot formation, and were maintained on this media indefinitely with subculturing occurring every three weeks. 2.6.3: Recovery of Seed Shoots greater than 1.5 cm in height and showing little to no hyperhydricity were selected for grafting attempts. Wild-type C. gynandra plants were grown at 24 °C with a light intensity of 350 µmol m-2 s-1 with a photoperiod of 12 hours for 6–8 weeks until a stem greater than 7.5cm was produced while still showing vegetative growth. The stem apex was removed after the third node, and the remaining leaves and auxiliary buds removed. Two diagonal slits were made in the exposed rootstock stem creating a narrow channel into which the transgenic Chapter 2: Materials & Methods 39 scion could be inserted. A plastic collar was wrapped around the join to strengthen the graft, and was secured with dental thread. A clear plastic bag was then placed entirely around the plant to maintain a humid atmosphere, which was in turn covered with a non-woven agricultural textile to prevent excessive light damage affecting the scion. Once the scion established and demonstrated positive growth, the bag and covering were incrementally removed to allow progressive acclimatisation to the ambient growth chamber conditions. The scion was then allowed to mature naturally before being both self-fertilised and providing pollen to artificially adjacent WT individuals. 2.7: Transformation of Arabidopsis thaliana The production of transgenic Arabidopsis lines was achieved through the use of the floral dip method (Clough & Bent, 1998). Col-O A. thaliana seeds were cultivated as described above, and were used for transformation shortly after bolting, which typically occurred after 4 weeks of cultivation. Glycerol stocks were used to establish 500 ml liquid LB cultures in 2 L conical flasks containing the corresponding antibiotics, which were inoculated with 200µl of glycerol- stored cultures before incubation at 30 °C for 24 hours while shaking at 200 rpm. The cultures were then pelleted by centrifugation at 3000g at 4 °C for 15 minutes before resuspension in 250ml of a 5% sucrose solution containing 0.05% Silwett L-77 (GE Silicones, West Virginia, USA). The plants were immersed inflorescence-side down in the bacterial suspension for 30s while gently shaking. After inoculation, the plants were wrapped in cling film before replacement in the growth room out of direct light for 24 hours, after which the plants were returned to normal growth conditions (see 2.1) until seed pod formation was advanced and the plants were allowed to dry out prior to harvesting. Seed recovered from dipped plants was cleaned by filtration before being subject to antibiotic selection following the methods of Harrison et al. (2006). Surface sterilisation was achieved by washing the seeds in 70% ethanol (Fisher Scientific) for three minutes followed by 15% sodium hypochlorite (Fisher Scientific) with 0.1% Tween®-20 (Sigma) for 20 minutes, after which the seeds were washed three times in sterile water. The seeds were then mixed in a 0.5% agarose gel (Melford) and poured over a 25 cm plate of MS medium supplemented with 1% sucrose and kanamycin sulphate at 60mgl-1. The plates were sealed with parafilm and stratified at 4°C in darkness for three days before light pulsing for 4 hours at 200 µmol m-2 s-1. After four days of culturing at 24°C in darkness, the plates were transferred back into light with a 16 day photoperiod for a further week. Healthy, developing plants were then potted up and cultivated in the fashion discussed previously (2.1). Chapter 2: Materials & Methods 40 2.8: Analysis of Transgenics 2.8.1: Verification of Transgenic Recovery Individuals from recovered seed and in vitro transgenics were tested for the presence of the target transgene through PCR screening using primer pairs complementary to the transgene. Total genomic DNA was tested through the use of the Extract-N-Amp™ Plant PCR Kit (Sigma) following the manufacturer’s guidelines. 2.8.2: Dexamethasone Induction Induction of transgenic lines containing the pOpON2.1 promoter was achieved using the methods of Samalova et al. (2005). 20 mM ethanolic dexamethasone (dex, Sigma) stocks were kept at -20 °C until dilution to a working concentration of 100mM with sterile water. Transient induction was achieved by inserting cut leaf petioles into the induction solution. For in vitro induction, dex was added directly to the cooled media before pouring. Soil-grown plants were either induced by misting of the leaf surfaces with a dex solution containing 0.1% Tween-20 or by application to the soil surface every three days to the point soil saturation. Successful induction was validated by GUS screening. 2.8.3: β-glucuronidase (GUS) Assay Stable lines of both Arabidopsis and C. gynandra containing the pOpON promoter system were assessed for GUS expression after induction. The uidA gene, expressed concurrently with the gene of interest after induction of the pOpON promoter, encodes for β-glucuronidase which cleaves the cyclohexylammonium salt of 5-bromo-4-chloro-3-indolyl-β-D-glucuronic acid (X-Gluc) to produce glucuronic acid and chloro-bromoindigo, the precipitate of which is coloured deep blue and is preserved through leaf clearing processes in ethanol and sodium hydroxide (Jefferson et al., 1987). The GUS stain was comprised of 100mM disodium hydrogen orthophosphate (Fisher Scientific), 0.5mM potassium ferricyanide (Fisher Scientific), 0.5 mM potassium ferrocyanide (Sigma), 10mM EDTA, 0.06% Triton X-100 (Sigma) and 2 mM X-GlcA (Sigma) (Jefferson, 1987). Penetration of the GUS stain into foliar tissues was achieved by three rounds of vacuum infiltration, each of one minute long. The plant material was then left to incubate at 37 °C for 16 hours before fixing and clearing. Chapter 2: Materials & Methods 41 2.9: Anatomical Analysis 2.9.1: Clearing of Leaf Tissue Leaf tissue, either freshly harvested or treated with GUS stain, was cleared for light microscopy with ethanol treatment. The material was first submerged in ethanol/acetic acid (3:1, both Fisher Scientific) for one hour before replacement with fresh 96% ethanol for 24 hours, both at 37°C. The material was then transferred to fresh 70% ethanol before storage at 4°C until required. Prior to examination the tissue was rehydrated by submersion in 50% ethanol followed by distilled water, each for 10 minutes. 2.9.2: Preparation of Material for Histology Material was fixed and embedded for histological examination using a method adapted from Jackson (1992) and Langdale (1993). Leaf tissue was harvested directly from the plant and cut into sections 2x10mm before fixing in 4% paraformaldehyde with 0.1% Tween®20 and 0.1% Triton X-100 (all Sigma) before three rounds of 1 minute long vacuum infiltration on ice. The paraformaldehyde was replaced after infiltration and the material incubated for 24 hours at 4 °C. The samples were then taken up an ethanol dehydration series of 30%, 40%, 50%, 60%, 70%, 85% ethanol on ice for 1 hour at each step before overnight incubation in 95% ethanol with 0.1% eosin (Sigma) to stain the tissue deep red. The material was then washed twice in 100% ethanol at 4°C for 30 minutes before transfer to a Leica TP1020 automated embedding machine (Leica Microsystems, Wetzlar, Germany). An overnight run was performed which consisted of three 1 hour washes in 100% ethanol, 1 hour wash in 50% ethanol/50% histoclear (National Diagnostics, Georgia, USA), three 1 hour washes in 100% histoclear and two 1½ hour treatments in molten Paraplast® embedding media (Sigma). The ethanol and histoclear washes were carried out at room temperature under a vacuum, while the paraplast treatment was carried out at 60°C, also under a vacuum. The samples were then aligned in the correct orientation in molten paraplast and set by rapid cooling in Peel-A-Way® embedding moulds (Polysciences, Inc., Pennsylvania, USA). The resultant wax blocks were then cut down for sectioning before mounting on plastic cassettes (Histosette®, Quebec, Canada) prior to sectioning on a Leica RM2135 rotary microtome. Sections cut to a thickness of 7µm using a fresh razorblade cleaned with 100% ethanol were floated on a layer of water on top of Polysine+™ adhesive slides (VWR International, Pennsylvania, USA) and left to dry overnight at 42 °C. Dry sections were then stripped of wax by the following process: two washes of 100% histoclear, 80% histoclear (made with 20% ethanol), then 50% histoclear, 30% histoclear, two washes of 100% ethanol followed Chapter 2: Materials & Methods 42 by 95%, 80%, 50%, 30% ethanol before a final water wash. Each treatment was for 2 minutes, and the process was performed at room temperature. The tissue was then placed in filtered 0.1% toluidine blue (Sigma) for 3-4 minutes before rinsing twice in water and examination using an Olympus BX41 microscope (Olympus, Tokyo, Japan) and were photographed using a MicroPublisher 3.3RTV camera (Q Imaging, Surrey, Canada). 2.10: Photosynthetic Analysis 2.10.1: Photosynthetic Pigment Quantification The abundance of protochlorophyll (the precursor of chlorophyll a, Virgin, 1955) and chlorophyll a/b in C. gynandra cotyledons were determined by spectrophotometry following the extinction coefficients of Moran & Porath (1980). Three pairs of light-grown or ten pairs of etiolated cotyledons were incubated in 1.5 ml N,N-dimethylforamide (DMF, Sigma) for 24 hours in darkness at 4 °C before quantification in a Helios Gamma spectrophotometer (Unicam Ltd.). Readings were taken at 603, 625, 647 and 664 nm to enable accurate determination of the photosynthetic components utilising the following equations of Moran (1982). Absorption at 664 nm was kept below 0.8 by diluting extracts accordingly. Chlorophyll a = 12.81 A664 - 2.16 A647 + 1.44 A625 - 4.91 A603 Chlorophyll b = -4.93 A664 + 26.01 A647 + 3.74 A625 - 15.55 A603 Protochlorophyll = -2.52 A664 - 0.79 A647 + 36.55 A625 - 27.08 A603 2.10.2: Photosynthetic Capacity Analysis Transgenic lines were cultivated in soil under the conditions given above (2.1) for a period of 4 weeks before induction with dexamethasone application for 1 week prior to photosynthetic quantification. Photosynthetic capacity was assessed using a 6400-40 leaf chamber fluorometer attached to a LI6400-XT portable open gas exchange system under illumination from a variable 6400-02B LED light source (all Li-COR Biosciences, Nebraska, USA). Measurements were taken from the central leaflet of the 3rd youngest leaf, which was allowed to acclimatise to a saturating light level of 1500 µmol quanta m-2 s-1 at an external CO2 concentration of 400 µmol mol-1. The chamber area was 2cm2, and was maintained at a leaf temperate of 22 °C while the vapour pressure deficit was kept below 1.3kPa. A steady state was determined when the conductance remained in a constant for 5 minutes, after which the response to changing ACi and PAR levels were recorded. The photosynthetic rate was recorded relative to internal [CO2] produced from external concentrations of 400, 400, 300, 200, 100, 50, 400, 400, 400, 500, 600, 800, 1000, 1200, 1500, 1800 and 400 µmol mol-1. Light Chapter 2: Materials & Methods 43 responses were measured at levels of 3000, 2500, 2000, 1800, 1600, 1400, 1200, 100, 800, 600, 400, 200, 100, 50, 25 and 2000 µmol quanta m-2 s-1. 2.11: Hybridisations of C. hassleriana and C. gynandra 2.11.1: Generation of Crosses Cross pollination of C. gynandra and C. hassleriana plants was carried out following methods adapted from Rajendrudu & Rama’Das (1982). The reproductive biology of both taxa were examined as a preliminary study prior to establishling crosses. Given the high levels of self- fertility that have been reported in Cleome (Cane, 2008) it was necessary to emasculate the flowers on the day before anthesis, a point which was determined by the extension of the petals to the point that the anther filaments were visible at the base of the petal in C. hassleriana, and the anther filaments were drawn out of the flower bud, and the anthers removed before dehiscence. Anthesis of C. gynandra was determined by a loss of green pigmentation from the flower. The emasculated flowers were covered with thin paper bags (“Glassine” bags, Eastern Polytene & Paper Ltd., St. Ives, UK), and all intact flowers surrounding the emasculated buds were removed. 20 plants of each species, each with 3-6 mature flowers were used for crossing. The following morning, the stigmatic surfaces of the emasculated flowers were pollinated with pollen grains taken from a freshly dehisced anther of the opposite species before rebagging and subsequent cultivation until maturity. The receptivity of the stigmatic surface was assessed by submersion of the stigma into 6% (v/v) hydrogen peroxide (Fischer), and evidence of oxygen evolution on the stigmatic surface was taken as an indicator of receptivity (Dafni & Maués, 1998). Pollen germination and tube growth was examined through the use of fluorescent visualisation following the method of Martin (1959). The gynoecium was removed from sample flowers and fixed in 3:1 ethanol:acetic acid for 16 hours before clearing in 4 M sodium hydroxide for one hour. The gynoecia were then washed three times in a 50 mM potassium phosphate solution buffered to pH 8 with potassium hydroxide, before submersion in a 0.1% (w/v) aniline blue stain for two hours at 60 °C. The stain was prepared by dissolution of aniline blue diammonium salt (Sigma) in 50 mM potassium phosphate buffer (pH 8) before being left to decolourise in darkness at 4 °C for one week. After rinsing in fresh phosphate buffer the gynoecia were placed into 10% glycerol (v/v) before visualisation under UV light (λ = c.350nm) using a Leica M165FC dissecting microscope and photographing using the accompanying camera and software. Under such conditions, the callose-rich pollen grain and tube fluoresce while the surrounding stylar tissue acquires a dark blue hue. Chapter 2: Materials & Methods 44 2.11.2: In vitro Ovule Culture In order to attempt in vitro culture of developing hybrid ovules, a method was adapted from Quazi (1988) and Zhang et al. (2003). 7–10 days after pollination (see Chapter 5), the gynoecia were removed and were sterilised by washing in 70% ethanol for 5 minutes before soaking in 12.5% bleach for 20 minutes before washing three times in sterile water. The ovaries were then dissected along the repulm, and the developing embryos excised under sterile conditions and were transferred to Embryo Rescue (ER) media. At this stage the undeveloped embryos could easily be identified and only maturing embryos were extracted. The ER medium was comprised of MS salts with Gamborg B5 vitamins (Murashige & Skoog, 1962; Gamborg et al., 1968) supplemented with 3% sucrose (w/v), 1 g l-1 casein hydrolysate (Sigma), 0.5 g l-1 MES, 0.5 g l-1 PVP, 40 mg l-1 adenine hemisulphate at pH 5.7. The medium was either solidified with agar at 0.75% (w/v), or used in a liquid state. In the latter case, the liquid medium was placed in the bottom of a sterile 30 ml tube, into which was inserted a filter paper bridge (Whatman) onto which was placed the recovered embryos. In addition, cultures were established in slowly rotating liquid medium, which was cultured while shaking at 30rpm. The medium was supplemented with 0, 0.25 or 0.5 mg l-1 NAA and 0, 0.5, 1.0, 1.5, 2.0 and 3.0 mg l-1 BAP. The cultures were then incubated at the conditions given above (2.1) with subculturing occurring every three weeks. After 12 weeks, the weight of each callus ball was measured using an AB104/S balance (Mettler-Toledo, Inc., Columbus, USA). Chapter 3: miRNA Regulation of the C4 Cycle 45 3. Do miRNAs regulate the C4 Cycle? Analysis of sRNAs and their Inhibition 3.1: Introduction The primary goal of this study was to test the hypothesis that miRNAs regulate the development or maintenance of the C4 pathway. If a miRNA targeting the C4 pathway was identified, its function in a C3 system could be tested in order to understand how the miRNA was recruited into regulating the C4 pathway. This hypothesis is based on the report that a disproportionate accumulation of sRNA sequences align against certain components of the C4 system in maize (Tolley et al., unpublished), and so attempts to identify any regulatory miRNAs that target components of the C4 cycle in C. gynandra were undertaken. 3.1.1: miRNA Identification A novel miRNA can be defined under the basis of a number of criteria (Ambros et al., 2003; Bonnet et al., 2004). Firstly, most plant miRNAs show between 1–4 mismatches with their target transcript, although these mismatches must not occur within the proposed cleavage site. Secondly, the miRNA sequence must match precisely with a locus on the genome of the target organism which is capable of generating the mature miRNA: the flanking regions of the miRNA coding sequence must be capable of generating a minimal free-energy folding hairpin structure capable of generating a pre-miRNA duplex. miRNA sequences can be identified by computational analysis of the target genome by identification of regions coding for pre-miRNA hairpins (e.g. Wang et al., 2005) or by cloning and sequencing of the small RNA (sRNA) fraction of the transcriptome between 15–35nt. In both instances, the targets of the miRNAs must be identified computationally by exploring the genome for target transcripts. Combined computational/cloning based strategies have been utilised to identify miRNA sequences during mechanical stress-responsive miRNAs in Populus (Lu et al., 2005), oxidative stress in rice (Li et al., 2011) and varied stresses in Arabidopsis (Sunkar & Zhu, 2004). While the direct computational analysis of genomic sequences for miRNA precursors can yield predictions, the latter authors report the detection of miRNAs by cloning methods that would not have been predicted in silico. However, as the genome of C. gynandra has not been sequenced, global prediction of functional sRNA using these methods is not possible. Therefore, the construction of a library of sRNA sequences present in mature C. gynandra was undertaken as an initial step in the identification of miRNA sequences which may interact with components of the C4 cycle. Chapter 3: miRNA Regulation of the C4 Cycle 46 Construction of sRNA libraries from total leaf RNA is well-established (Havecker, 2011). The production of libraries from M or BS cells would enable sRNAs with differential abundance in these cell types to be identified. Methods exist for cell-type specific RNA extraction in maize (Kanai & Edwards, 1973), but when this method was attempted in C. gynandra a high level of contamination from the opposing cell type was observed (Gage, unpublished). The use of Laser Capture Microdissection (Emmert-Buck et al., 1996) has been applied to C. gynandra (Aubury et al., in prep.), but the low yields of RNA from this method would make this approach unsuitable for isolation of sRNA. As a result, the construction of a library of sRNA reads from total mature leaf RNA extracts of C. gynandra was undertaken. Previous studies have predicted miRNA sequences utilising Expressed Sequence Tag (EST) data (Prabu & Mandal, 2010; Wu et al., 2010). An EST library is available for C. gynandra (Bräutigam et al., 2011), and so ESTs for candidate genes were used as targets for the computation prediction of miRNA sequences present in the sRNA population of C. gynandra. Hits against target sequences were examined for abundance and targeting criteria that have been described for plant miRNAs (Rhoades et al., 2002; Bonnet et al., 2004), and were compared against previously described miRNA sequences to identify any conserved sequences from Arabidopsis. Predicted miRNAs can experimentally validated by the detection of cleavage products of their predicted target transcripts that would indicate miRNA activity. Those transcripts which show a high level of sRNA accumulation to specific loci can be screened for such cleavage products using 5’ Rapid Amplification of cDNA Ends (RACE, Frohman et al., 1988). The cleavage reaction will produce a transcript missing the 5’ cap so that additional treatment with Tobacco Acid Pyrophosphatase (TAP) prior to adaptor ligation is not required (unlike full length transcripts). The point of cleavage can then be determined by subsequent sequencing. Using this approach, it should be possible to detect sRNA sequences which directly target the components of the C4 cycle. In addition to this approach, the effects of expressing an inhibitor of miRNA activity in C. gynandra were examined in order to test whether miRNAs may play a role in regulating aspects of the C4 cycle for which a genetic origin has yet to be described, such as the formation of Kranz anatomy. 3.1.2: Viral Inhibition of miRNA Activity As discussed previously (Section 1.4.2), a number of plant viruses have evolved mechanisms to suppress siRNA. Due to the similarity between the biosynthesis pathways of siRNAs and Chapter 3: miRNA Regulation of the C4 Cycle 47 miRNAs, expression of such viral inhibitors can interfere with the function of endogenous miRNAs (Chapman et al., 2004), promoting a phenotype that matches that of dcl-1 mutants in Arabidopsis (Kasschau et al., 2003; Dunoyer et al., 2004). Variation is seen in the activity between different viral inhibitor proteins, and the p19 protein of the TOMBUS viral family is one of the most potent inhibitors (Lakatos et al., 2006; Dunoyer et al., 2004). Cytosolic p19 selectively binds to 21nt dsRNA (Vargason et al., 2003) prior to RISC formation, but has no effect on existing functional RISC complexes (Lakatos et al., 2006). In addition, Uhrig et al. (2004) reported a relocalisation effect of p19 upon nuclear ALY proteins, although the effects of this interaction are unclear. As p19 has been shown to inhibit sRNA biosynthetic pathways it was chosen as an inhibitory protein for introduction into C. gynandra. The benefits of establishing a stable transgenic line would be three-fold. Firstly, leaf development could be examined to identify any effects of miRNA inhibition on the formation of Kranz anatomy. Secondly, analysis of photosynthetic performance may highlight any effects of disruption of miRNA activity. Finally, components of the C4 cycle can be quantified at the transcript and protein level to screen for any p19-induced changes in abundance that could be associated with a relief from miRNA inhibition. This would test the hypothesis that if a component of the C4 cycle was subject to miRNA-mediated repression, then an over-accumulation of the component could be identifiable when the miRNA activity is suppressed by the presence of p19. 3.1.3: Temporal Control of p19 Expression Due to the effects of p19 expression on leaf development it was considered likely that high levels of constitutive p19 expression may limit the recovery of a viable transgenic line. Inducible promoters have been used previously to allow temporal control of genes (e.g. CLAVATA1/2, Schoof et al., 2000) and so the placement of p19 under control of a chemically inducible promoter was investigated. The dexamethasone-inducible pOpON inducible promotion system has been shown to generate specific induction of transgenes in tobacco (Samalova et al., 2005) and Arabidopsis (Craft et al., 2005) and so this system was selected for temporal control of p19. The utilisation of such a promoter would ensure that transgenic lines could develop normally prior to the induction of p19 activity. The pOpON system is comprised of two components (Figure 3.1). Firstly, the transgene (in this case p19) is placed downstream of the chimeric pOp promoter comprised of lac operators upstream of a 35S promoter with the addition of the Tobacco Mosaic Virus Ω translational enhancer in the 5’ UTR region. This construct is inserted into the target organism along with Chapter 3: miRNA Regulation of the C4 Cycle 48 the constitutively expressed LhGR synthetic promoter which is comprised of a lac DNA- binding domain placed between a GAL4 transcription activation domain taken from Saccharomyces cerevisiae and GR, the glucocorticoid binding domain taken from rat. The presence of the GR domain ensures that LhGR is bound to the heat-shock protein HSP90 until ligation of dexamethasone to the GR domain allows release of LhGR and subsequent activation of the pOp promoter (Moore et al., 2006). Due to the absence of dexamethasone from a standard cultivation system, transgene expression is effectively silent in the absence of induction. In contrast, high levels of transgene expression can be seen within 4 hours of dexamethasone application, either locally or systemically (Samalova et al., 2005). Other glucocorticoid-inducible promoter systems have been reported to cause developmental defects unrelated to transgene expression (Kang et al., 1999), Samalova et al. (2005), although this is not the case in the pOp/LhGR system. Therefore, this system should allow silencing of miRNAs via p19 induction in mature leaves to investigate the importance of the sRNA pathway in maintaining C4 photosynthesis. Furthermore, long term exposure to dexamethasone during early plant development should enable the effect of removing small RNA activity on C4 leaf anatomy to be examined. Figure 3.1: Function of the pOp promoter. The LhGR transcription factor is constitutively expressed, but when present in the cytosol it is bound to the heat-shock protein hsp90 (A). For induction, externally applied dexamethasone ligates to the GR domain of LhGR, releasing it from hsp90 and allowing it to pass into the nucleus (B) where it binds to the lac regions of the pOp promoter (C). The promoter drives transcription in both directions, leading to the expression of p19 and GUS in tandem (D). 3.1.4: Experimental Design The sRNA population of mature leaf tissue of C. gynandra was sequenced in order to generate a library of reads as described previously. The EST sequences of candidate components of the C4 cycle were then used as targets for exploring the library in order to identify miRNAs which Chapter 3: miRNA Regulation of the C4 Cycle 49 may target components of the C4 cycle on the basis of sequence complementarity. Potential miRNA sequences were then verified by testing for the products of cleavage of their target transcript. This work was underpinned by exploring the effects of inhibition of miRNA biosynthesis upon the C4 cycle. The 35S::p19::nos construct of Voinnet et al. (2003) was transformed into C. gynandra in vitro using the methods of Newell et al. (2010) in order to assess the effects of constitutive expression of p19. In addition, the same construct was transformed into Arabidopsis via the floral dip method (Clough & Bent, 1998) in order to validate the effects of p19 expression upon leaf development described from previous studies. Agrobacterium- mediated transgene insertion occurs at random sites (Townsend et al., 1994), and as such generates variation in the extent of transgene expression due to positional effects of neighbouring loci (van Leeuwen et al., 2001). In both C. gynandra and Arabidopsis the T0 generation were assessed in order to examine dose-dependent effects of p19 expression by exploiting this variation. In parallel to this, p19 was cloned into the pOpON2.1 promoter system (Moore et al., 2006). This construct was then transformed into C. gynandra in vitro. Transgenic seed was obtained by grafting T0 transgenic shoots (shoots recovered directly from in vitro culture) onto established WT rootstocks, and this T1 generation was assessed for any impacts of p19 activity upon the C4 cycle, either through changes in anatomy, photosynthetic capacity or transcript/protein abundance. The pOp::p19 construct was also transformed into Arabidopsis in order to examine the effects of inducible p19 expression on the C3 leaf. The benefit of the pOpON system is that the effects of p19 expression can be examined at a number of levels. The effects of p19 expression on plant development at the cotyledon stage and beyond will be possible with the lines containing p19 under the control of the CaMV 35S promoter, but induction of mature plants will enable assessment of the effects of p19 expression on a developed, functioning C4 leaf. Initially, induced plants were subject to gas exchange analysis in order to determine the effect of p19 expression on net photosynthesis by evaluating changes in CO2 uptake under varying light and CO2 regimes (Farquhar et al., 1980; Caemmerer & Farquhar, 1981). Plants operating the C4 cycle show reduced carbon compensation points and increased carboxylation efficiencies (as discussed previously). Therefore, if p19 activity impacted on the C4 cycle, net photosynthesis should become more C3-like. If a miRNA is directly regulating the accumulation of a component of the C4 cycle, its method of acting is most likely to be via the negative regulation of the target gene in the Chapter 3: miRNA Regulation of the C4 Cycle 50 opposing cell type. In order to assess this, induced plants were examined for changes in transcript abundance using quantitative real-time PCR (qRT-PCR, Bustin et al., 2000) and in protein abundance using the western blot method of Maizel (1966) and Shapiro et al. (1966). Given the wide number of roles that miRNAs have been linked with, it is likely that the expression of p19 is going to have a range of effects on multiple levels. By working with developed leaves it is hoped that developmental effects are going to be limited. However, it will be impossible to separate out any additional metabolic effects on a biochemical level. By working at the level of gas exchange analysis, and at transcript/protein abundance it is hoped that as greater focus on photosynthetic effects as possible will be achieved. It was hoped that through the combination of sRNA cloning, computational prediction of targets and examination of p19 transgenic lines, any miRNAs which are targeting the C4 cycle in C. gynandra could be identified. The use of the p19 inhibition is beneficial on two levels. Firstly, it may provide support for computationally-predicted miRNA sequences. Secondly, it will enable a broader examination of miRNA-mediated regulation of the C4 cycle in that it will be possible to evaluate any effects of p19 expression upon traits for which no genetic basis has been determined, such as the formation of Kranz anatomy. Chapter 3: miRNA Regulation of the C4 Cycle 51 3.2: Results 3.2.1: sRNA Library Compilation and Analysis A library of small RNA sequences present in mature C. gynandra leaf tissue was cloned and sequenced, yielding 2.4 million reads of which 660 thousand were unique (Table 3.1). Sizes ranged between 15-25nt in length and these demonstrated a distribution typical of the sRNA fraction, with the greatest abundance of reads being 24nt in length (Figure 3.2). Number of Reads Total Number of Nucleotides GC Content (%) Percentage (%) Raw Reads 2,437,396 87,746,256 46.43 100.00 Reads rejected due to Length 1,394,192 50,190,912 45.17 57.20 Reads rejected due to Noise 17,914 644,904 46.77 0.73 Reads rejected during Trimming 8,588 81,250 44.94 0.35 Duplicate Reads 354,548 7,981,925 44.95 14.53 Unique Reads 662,548 15,135,415 44.23 27.18 Table 3.1: Statistics relating to the sRNA library compiled for C. gynandra. 27.18% of the total reads were non-redundant and within the 15-25nt size threshold defined during the filtration process. 0 10 20 30 40 50 60 0 1 2 3 4 5 10 15 20 30 40 50 60 N u m b er o f G en es Number of Complementary Hits Figure 3.2: Size distribution of reads in the C. gynandra sRNA library. Read length was between 20-24nt, with a strong bias towards 24nt after adaptor removal. The sRNA library demonstrates a distribution and composition typical of this sequencing method. These data were obtained in collaboration with P. Shivaprasad. P P D K P E P C R b cS 1A Figure 3.3: Reads from the C. gynandra sRNA library aligning against target EST sequences. The majority of EST sequences had no sRNA hits, with less than 15% of the targets obtaining more than 10 hits from the sRNA library. The decline in sequence number within each size threshold drops off rapidly. Chapter 3: miRNA Regulation of the C4 Cycle 52 Analysis of the library was preceded with a series of processing steps to produce a library of reads. The sequences were first filtered for reads of lengths between 15-50nt, as the sequences obtained are comprised of the sRNA fused to the 3’ adapter. The starting 8nt of the 3’ adaptor were then identified, and the sequences upstream from this site cleaved to leave the sRNA sequence. This was then subject to further filtering to produce a library of unique reads between 15-35nt in length. These steps were achieved using the SIROCCO ADDAPTS processing pipeline of Nag et al. (in submission). The quality of the library was verified by the size distribution of its content (Figure 3.2) and by the identification of two known miRNA sequences. miR398 and miR165/166 which regulate copper superoxide dismutase (Jagadeeswaran et al., 2009) and PHABULOSA (Mallory et al., 2004) respectively. Sequences corresponding to these miRNAs were present in the sRNA library, which was taken as confirming its quality. As a result of the lack of an annotated genomic sequence for C. gynandra, the library was initially explored used EST sequences of the transcriptome data of Braütigam et al. (2011). Initially, candidate target sequences were searched against the library for directly complementary reads using the algorithms of Moxon et al. (2008) (Figure 3.3, Table 3.2) The candidate EST sequences (given in Appendix 2) were then used as targets in a search for direct complementarity to sequences in the non-redundant read library, the results of which are summarised in Figure 3.3. Only 15% of the ESTs had more than 10 hits, with the top scorers being represented by transcripts of PPDK (60), an isoform of PEPC (40) and RbcS1A (18). Given the rapid decline in hit rate seen within these data, these sequences were analysed with an additional set of the genes important for the C4 pathway components for mismatches within sRNA/target pairings (see Appendix 2). Sequencing of the sRNAs present in C. gynandra leaf material indicated that a disproportionate accumulation of sRNA that map to some components of the C4 cycle. Functional plant miRNAs tend to show a measure of imperfection in pairing with their targets, as discussed previously. Therefore, a smaller subset of candidate genes were searched with relaxed targeting criteria to allow mismatched pairings. While increased numbers of reads aligning against the target sequences were seen, this increase was not equally distributed between genes (Table 3.3). While an average of 39 unique reads aligned against each EST, ALAAT1, ASP1, CA4, DIT2.1, MDH, PEPC, PPCK and PyT1 sequences showed below average Chapter 3: miRNA Regulation of the C4 Cycle 53 sRNA alignment while RbcS1A, RCA and both isoforms of PPDK and NAD-ME had considerably above average alignment scores. Table 3.3: sRNA reads aligning against photosynthesis genes in C. gynandra. The number of unique reads aligning against each target is unevenly distributed, with PPDK, NAD-ME2, CA4, RbcS1A and RCA showing alignment to an above average number of sRNA reads. ALAAT1: Alanine aminotransferase 1, ASP1: Asparate aminotransferase 1, DiT2.1: Dicarboxylate Transporter 1, MDH: Malate Dehydrogenase, NAD-ME: NAD-malic enzyme, PEPC2: PEP carboxylase, PPCK1: PEP carboxylase kinase 1, PPDK: pyruvate, orthophosphate dikinase (including chloroplastic isoform and genomic DNA), PyT1: pyruvate transporter, RbcS1A: SSU of RuBisCO, RCA: RuBisCO Activase. When the sRNA reads were aligned against the target sequences, a number of patterns were evident (Figure 3.4 and Table 3.4). sRNA reads mapped against some target sequences in a very diffuse pattern (e.g. NADME2, ALAAT1), while others showed a highly focused aggregation of sRNA reads to a narrow range of sites (e.g. CA4, MDH, RCA). The direction, nature and mismatch content of sRNA alignments were grouped into a number of predominant classes (Table 3.4). The diffusive pattern of some alignments contained a number of loci at which multiple sRNA reads aligned although this was most pronounced in NADME1&2 and RCA. Frequently, antisense alignment clusters contained both perfect and mismatched alignment loci and several sRNA reads aligned against a series of concurrent loci so that a contiguous sequence greater than 30nt was achieved, frequently with greater than 4 mismatches (e.g. alignment J in Figure 3.4). Due to the nature of the sequencing method involved, the reads are unidirectional and so primary analysis was concerned with antisense reads, as discussed previously. However, alignment parameters allowed sense alignments to be recorded in order to detect any aberrant accumulation patterns. In addition, the genomic sequence of a gene was also searched where available, and sense alignments in these conditions were considered. In several targets, perfect alignment (or alignments in which mismatches were not supported by all aligned reads) against a number of loci were recovered, and this was most pronounced in ASP1, PyT1 and RCA. Perfectly aligned reads containing Total Reads Unique Reads ALAAT1 25 25 ASP1 19 17 CA4 29 29 DiT2.1 12 12 MDH 30 11 NAD-ME1 39 38 NAD-ME2 64 60 PEPC2 6 6 PPCK 27 27 PPDK 47 41 PPDK (Chp) 33 29 PPDK (gDNA) 323 224 PyT1 24 24 RbcS1A 31 31 RCA 51 39 Chapter 3: miRNA Regulation of the C4 Cycle 54 mismatches and no mismatches are given in Table 3.4, although contiguous sequences are also stated. Significant regions of antisense alignment (Figure 3.4) were found in MDH (I: 20nt, 22 reads), CA4 (J: 35nt, 17 reads), NAD-ME1 (A: 35nt, 11 reads), RCA (G: 15nt, 7 reads). A region of 26nt covered by 9 reads was seen in NAD-ME2 (B), although this was part of a greater sequence aggregation. Two significant sense alignments were also detected: 11 reads aligned to a 20nt region in RCA (H), and 11 reads aligned against a 42nt region of the chloroplastic isoform of PPDK (C). Interestingly, alignments against the genomic sequence of PPDK were found which were not detected in the chloroplastic isoform: 20 antisense reads aligned against a 50nt region (D), and 125 sense reads aligned against a region 29nt long (E), both of which rest in intronic regions. A perfect sense alignment was also recovered (F). Chapter 3: miRNA Regulation of the C4 Cycle 55 Figure 3.4: Alignments of sRNA reads against target sequences in C. gynandra. A: NAD-ME1, B: NAD- ME2, C: PPDK (chloroplastic isoform), D, E & F: PPDK (genomic DNA), G&H: RCA, I: MDH, J: CA4. The scale bar at the top of each plot refers to the alignment position from the transcription start site. Right handed arrows are in the sense direction to the target, with left-handed arrows representing antisense. Grey alignments indicate perfect pairing, while a coloured band highlights single-base substitutions. The arrow in E represents the additional 105 reads which align at this position. Chapter 3: miRNA Regulation of the C4 Cycle 56 Antisense Sense Target Perfect Mismatch >30 nt Perfect Mismatch >30 nt № L № L № L № L № L № L ALT1 1 24 4 44 3 20 4 29 ASP1 7 25 1 18 1 18 2 22 3 18 DIT2.1 2 26 8 32 PyT1 1 19 2 44 6 27 5 37 NAD- ME1 1 18 11 37 1 24 1 24 2 31 1 28 2 25 2 29 6 28 3 20 NAD- ME2 1 20 4 37 2 18 2 25 11 48 2 24 5 23 2 26 3 27 3 34 7 40 PEPC2 PPCK1 5 47 3 39 5 34 3 13 4 28 5 19 Table 3.4: sRNA read alignments against target sequences in C. gynandra. The number of reads aligning at a single site is given (№) with the total length of the alignment (L). Matches are separated on the basis of sense or antisense complementarity, and on the basis of perfect or mismatched alignment. The sequences which aligned to form a contiguous sequence greater than 30nt in length are also denoted. Those alignment clusters discussed in the text are given in bold. (Continued Overleaf) Chapter 3: miRNA Regulation of the C4 Cycle 57 Antisense Sense Target Perfect Mismatch >30 nt Perfect Mismatch >30 nt № L № L № L № L № L № L CA4 1 19 17 36 1 18 2 18 PPDK 2 17 2 25 2 24 2 32 4 17 2 26 3 40 14 33 17 33 PPDK 1 18 4 28 1 21 11 41 (Chp) 1 22 1 26 2 28 2 28 PPDK 1 18 2 21 3 34 1 20 2 21 17 37 (gDNA) 1 19 4 26 4 32 1 24 2 21 4 33 1 20 5 26 7 36 125 29 3 24 1 21 7 23 20 50 10 28 MDH 2 22 2 20 Rbcs1A 3 28 6 43 RCA 1 18 6 42 1 10 3 37 3 18 10 51 7 15 2 33 Table 3.4 (Continued): sRNA read alignments against target sequences. The number of reads aligning at a single site is given (№) with the total length of the alignment (L). Matches are separated on the basis of sense or antisense complementarity, and on the basis of perfect or mismatched alignment. The sequences which aligned to form a contiguous sequence greater than 30nt in length are also denoted. Those alignment clusters discussed in the text are given in bold. Chapter 3: miRNA Regulation of the C4 Cycle 58 Mismatches were highly conserved within a given alignment, and it was only within extended regions of alignment coverage that multiple mismatches were recorded (e.g. 3.4-J). Sequences that were abundant were searched against a database of mature miRNAs (miRBase, Kozomara & Griffiths-Jones, 2011). Five sequences were identified which showed similarity to the alignments recovered in this analysis (Table 3.5). Target Position Sequence Reference miRNA Reference Origin PPDK Chp 2263 + GACGAAUAUGAACGAAGAAGA miR5293 Medicago truncatula Rbcs1A 419 + UGUUAGCGGAGCUUGAAGAGG miR4398 Glycine max NAD-ME2 1014 + CAAAGGAGAUGGGCU miR399k Populus trichocarpa CAAAGGAGAUGGGCU miR399d Vitis vinifera CAAAGGAGAUGGGCU miR399f Ricinus communis 1303 + ACUGAUCAGAAAGCAGAGGAC miR2111 Vitis vinifera ALT1 124 - CUUCUUCAUUUGUUUCUACUU miR5655 Arabidopsis thaliana Table 3.5: Known miRNAs which show similarity to reads aligning to potential C4 photosynthesis target genes in C. gynandra. The first site of alignment against the target sequence and orientation is given under Position, with the read itself quoted under Sequence. Bases given in red show a mismatch against the established miRNA sequence. The reference and species of origin according to miRBASE (see text) are also given. While single sequences have been published for miR5293, miR4398 and miR5655, sequences for miR399 and miR2111 (both of which potentially target NAD-ME2) have been described in a sufficient number of species to produce alignments (Figure 3.5). A high level of sequence conservation is seen in miR2111 between lineages, and the only region present in the proposed C. gynandra NAD-ME sequence is a GAGG motif in the 5’ region. A broader spectrum of variation is seen in miR399, both between and within lineages. Greater similarity was seen between the C. gynandra sRNA read aligning to position 1014 in NAD-ME2 and miR399, with the first 10bp matching the conserved sequence, and the structure present in the 5’ region bearing some resemblance to the miR399 sequence. However, the reduced length of the sRNA read does not allow for a full comparison to be made. The sequence of the proposed target site in NAD-ME2 does not match miR399 at either the 5’ or 3’ end, so even if the sRNA read was part of a full length sequence it would not be possible to align against NAD-ME2 and show a conserved sequence with miR399. Given the abundance of sRNA sequences aligning against PPDK, it was chosen as a candidate to identify potential cleavage products via 5’ RACE. Initially, attempts were made to produce Chapter 3: miRNA Regulation of the C4 Cycle 59 sequences for the first 250nt of the chloroplastic isoform of PPDK. These were successful, yielding a PCR product of the predicted length (Figure 3.6) which corresponded to the C. gynandra PPDK sequence of Bräutigam et al. (2011). However, subsequent attempts to increase the sequence obtainable via the 5’RACE protocol failed to produce amplicons that were distinguishable from non-specific products despite reaction optimisation. Figure 3.6: Amplification products of a 5’RACE reaction using nested primers against PPDK run out on a 1% agarose gel. The bands indicated correspond to the region of cytosolic isoform and chloroplastic isoform of PPDK after the addition of the 20nt RACE adaptor. The smearing on the gel is due to non- specific primer binding. Figure 3.5: Alignments of potential miRNA sequences against Cg NAD-ME2 show varying levels of similarity with described miR399 sequences. The proposed miRNA sequence (Cg BXC10X) underneath the target site (NAD-ME2, top), followed by published sequences for miR2111 (A) and miR399 (B). Sequence notation corresponds to the miRNA reference number on miRBase (see text). Neither sequence shows sufficient similarity to the conserved structure of the published miRNA, nor is a corresponding sequence between either miR399/miR2111 present in NAD-ME2 transcripts. Chapter 3: miRNA Regulation of the C4 Cycle 60 3.2.2: The Effects of p19 Expression 3.2.2.1: Constitutively Expressed p19 in Arabidopsis The 35S::p19 construct (Voinnet et al., 2001) was transformed into Arabidopsis in order to quantify the effects of constitutive expression of p19 upon leaf morphology. T0 individuals were recovered at a rate of 1.8% although a loss of between 10-20% of individuals was seen within two weeks of transfer to soil. p19 protein was detectable by immunoblotting (Figure 3.7) and a strong phenotype was recordable in the T0 generation (Figure 3.8). Individuals showed reduced vigour to the point of dwarfism (3.8-A) and leaves showed increased lobe formation and serration of the leaf margin. In the most severe lines shortened leaves with very limited blade expansion were seen (3.8-B). Internal leaf anatomy of these lines was disorganised so that the standard arrangement of palisade and spongy mesophyll was not identifiable, although the formation of the bundle sheath was still evident (3.8-C). While premature flowering was recorded in weakly expressing individuals, a number of variations in inflorescence anatomy were seen (Figure 3.9). The strongest leaf phenotype corresponded with delayed or inhibited flowering, in which extension of the raceme halted prematurely or failed completely. Increased cauline leaf formation was seen, although these showed similar developmental modifications to the rosette leaves (3.9-B). The primary raceme was subject to premature termination (3.9-C,D), although this did not always prevent development of the axillary racemes. The flowers produced on the ascendant racemes formed reduced sepals and petals, which were narrowed and curled (3.9-E,F). The stamens showed reduced anther and filament devolvement, and pollen dehiscence was not detected (3.9-G). The presence of p19 at both the transcript and protein level was detectable, and a correlation was found between the extent of p19 accumulation and the severity of the recorded phenotype within the T0 individuals (Figure 3.10). As a result of the altered floral anatomy caused by p19 expression, fertility was significantly impacted so that no T1 seed was recovered. Figure 3.7: p19 accumulation in stable lines of transgenic Arabidopsis. 20µg of total soluble leaf protein was loaded onto a 14% denaturing acrylamide gel. A signal is detectable at the 19 kDa point in lines 1.8–1.5 after conjugation with antibodies raised against p19 which is absent from the WT ColO control (i). Equal protein loading is shown by Coomassie staining (ii). Chapter 3: miRNA Regulation of the C4 Cycle 61 Figure 3.8: p19 expression alters leaf morphology in Arabidopsis. Varied expression is seen in T0 lines (A), with extreme instances of the phenotype showing reduced vigour and stunting (lines 1.6 - 1.8). In addition to reduced leaf growth, increased serration of the leaf margin is seen in the more developed individuals (B, lines 1.1 – 1.4, 1.9) relative to the ColO WT leaf. While the palisade (PM) and spongy (SM) mesophyll is identifiable in sections of ColO leaf tissue (C), an increased level of disorganisation in p19- expressing lines (C – 1.9) renders these tissues indistinguishable, although the bundle sheath (BS) is still evident. Scale = 1cm (A,B), 100µm (C). SM PM BS Chapter 3: miRNA Regulation of the C4 Cycle 62 Figure 3.9: p19 expression effects reproductive anatomy in Arabidopsis. Increased formation of cauline leaves was observed, with the leaves showing the same increase in serration shown in the rosette leaves (B) relative to that of the WT Col-O (A) individuals. Premature termination of the primary raceme was seen although this did not always prevent the formation of axillary racemes (C,D). The flowers of p19- expressing individuals (E,F-ii) showed reduced, misformed petals and sepals when compared with ColO individuals (E,F-i). The anthers of p19-expressing lines (G-ii) failed to mature and pollen dehiscence was not observed by the stage that dehiscence was evident in ColO individuals (G-i). Scale = 1cm (A- D), 2mm (E,F,G). Chapter 3: miRNA Regulation of the C4 Cycle 63 Figure 3.10: Arabidopsis rosette growth shows a negative correlation with p19 transcript abundance. Increased abundance of p19 leads to an increased severity of the dwarfing phenotype. 3.2.2.2: Constitutive Expression of p19 in C. gynandra Transformation of the 35S::p19 construct into C. gynandra was successful. Regenerative callus was produced from cotyledon and hypocotyl explant material within three weeks of culture, and early leaf formation was detectable within six weeks. While c.75% of explants produced 1– 3 callus masses, leaf regeneration was only seen in 30–50% of the callus produced. Significant variation was seen between subsequent transformation attempts and explant age was found to be most significant with an optimum age of 6 days from initial sowing. Transgenic lines were identified on the basis of p19 presence at the genomic and transcript level (Figure 3.11). The antisera raised against p19 showed cross-reactivity to a c.20kDa residue in WT C. gynandra material, so it was not possible to test for the presence of p19 protein. Figure 3.11: p19 in transgenic C. gynandra lines. PCR screens of genomic DNA (A) showed the presence of p19 in lines expressing the radialised phenotype (1.1, 1.3) which were absent in regenerative lines resembling WT (1.7) and in WT material. PPDK was used as a positive control. The presence of p19 at the transcript level was detected through the use of RT-PCR (B). Total leaf RNA was used for cDNA synthesis to provide a template for PCR reactions to screen for the presence of p19 transcripts. These were detectable in those lines expressing the radialised phenotype (1.1–1.3), but were absent from the WT. Primers for ACT7 were used as a positive control for effective cDNA synthesis. Chapter 3: miRNA Regulation of the C4 Cycle 64 The in vitro regeneration of leaf material from non-transgenic control lines showed the formation of small but intact leaves of (Figure 3.12–WT). However, p19-expressing lines showed the formation of long, thin, radialised leaves which showed little to no mesophyll development (3.12–1.1, 1.3). These showed indeterminate growth, with a tortuous habit which showed little to no polarity. No shoots were recovered from these lines. The internal anatomy of p19-expressing lines showed a significant reduction in the extent of vein formation (Figure 3.13). While normal Kranz anatomy was seen in the leaves regenerated by non-transgenic material, the extent of primary and secondary vein formation was greatly reduced in p19 expressing lines. The reduction in mesophyll development was also notable. Figure 3.12: p19 expression in C. gynandra causes significant changes to leaf anatomy. The regeneration of WT C. gynandra material in vitro shows the formation of small leaves and shoots from callus (1.7). Transgenic lines 1.1 and 1.3 showed the production of thin, radialised leaves with little to no mesophyll development, and the formation and shoots was not seen. Scale = 1cm. Figure 3.13: p19 expression causes a reduction in leaf venation and leaf expansion in C. gynandra. The radialised leaves of p19-expressing lines are narrow and thinned and demonstrate little mesophyll formation, while normal leaf development is seen in WT material. The extent of primary and secondary branching of the vasculature is decreased, and as a result the mesophyll area between such venation is equally decreased. Scale = 1mm. Chapter 3: miRNA Regulation of the C4 Cycle 65 3.2.3.3: Creation of the pOp::p19 System As a result of the significant morphological changes caused by constitutive expression of p19 in C. gynandra, p19 was placed under the control of the pOpON2.1 dexamethasone inducible promoter. The stages involved in the production of this construct are shown in Figure 3.14. The coding sequence of p19 was amplified from the original 35S::p19::nos construct (3.14-A) using primers which would extend the 5’ end of the coding region to include a CACC sequence to ensure that insertion of the amplified fragment could occur in the correct orientation when placed into the entry vector (3.14-B). Plasmid composition was verified through restriction enzyme digestion. A region of 1.2kb sits between the recombination sites of the pOpON destination vector, and this is replaced with the 0.5kb p19 coding sequence. As a result, a 2.2kb fragment of the pOpON destination vector released by digestion with Nhe1 was reduced to 1.5kb after recombination between the entry and destination vectors. Integration of the p19 fragment into pOp::p19 was confirmed by restriction digestion (3.14-D), followed by sequencing. Figure 3.14: Creation of the pOp::p19 Construct Initially, the coding sequences of p19 was amplified from the 35S::p19 construct. A 5’ CACC sequence was added to allow for directional cloning, giving a total fragment size of 530bp, as visualised on a 1% agarose gel (A). This was inserted into a TOPO® entry vector, which was propagated in E. coli before extraction and digestion with Nhe1. A fragment of 5340bp containing the p19 insert can be detected as a product of digestion, shown here on a 0.9% agarose gel (B). The pOpON destination vector produces fragments of 2208, 7006 and 10931bp after digestion with Asc1 and Not1, shown here on a 0.9% agarose gel (C). The p19 entry and pOpON destination vectors were subject to recombination, producing the pOpON::p19 construct. Upon digestion with Nhe1, the previous 2.2kp fragment of the pOpON destination vector is reduced to 1.5kb due to the recombination reaction inserting the shortened p19 sequence, as visualised here on a 0.9% agarose gel (D). Lanes A to D show the 1.5kb fragment containing p19, while lane E contains the larger 2.2kb fragment produced by digestion of the empty pOpON destination vector shown in C. A B C D E Chapter 3: miRNA Regulation of the C4 Cycle 66 3.2.3.4: pOp::p19 in Arabidopsis The pOp::p19 construct was transformed into Arabidopsis using the floral dip method, and a transgenic recovery rate of around 0.25% was seen. T0 individuals were grown on soil in the absence of dexamethasone for four weeks before induction was attempted. Initially, a single leaf was excised and placed in a 50 µM aqueous dexamethasone solution for four hours before GUS staining. The dual-orientation promoter system of pOpON would ensure that both uidA and p19 are expressed together, so the accumulation of GUS was taken as indicative of p19 accumulation. GUS expression and accumulation was detectable within the four hour induction period, as indicated by the accumulation of GUS in the excised leaf tissue, and subsequent RNA extraction followed by qRT-PCR showed the accumulation of p19 transcripts, although p19 was not detectable in the absence of dexamethasone (Figure 3.15). T0 individuals that had been cultivated on soil for four weeks were then subjected to a week-long period of induction through soil irrigation with 30 µM dexamethasone on alternate days. After seven days, the youngest fully developed leaf was removed and used for RNA and protein extraction prior to profiling for p19 transcript and protein accumulation. Transcripts and protein of p19 were detectable only after induction with dex (Figure 3.15, 3.16). Figure 3.15: GUS expression is inducible by exposure to dexamethasone in transgenic Arabidopsis lines containing the pOp::p19 construct. Scale = 1cm. A: A pair of leaves that have been removed from the same T0 individual and have been incubated in either a water control (i) or 50 μM dexamethasone solution (ii) for four hours before GUS staining. GUS accumulation is specific to the induced leaf, indicating that GUS is not expressed in the absence of dexamethasone, but is capable of obtaining high levels of accumulation within 4 hours of dexamethasone exposure. B: The accumulation of p19 transcripts follows a similar pattern. Leaves taken from a single T0 individual show induction of p19 after exposure to dexamethasone, generating a level of transcript accumulation not seen in the water-treated control or in a WT ColO individual. A B Chapter 3: miRNA Regulation of the C4 Cycle 67 Figure 3.16: Induction of p19 and GUS in Arabidopsis lines containing pOp::p19. Multiple T0 lines were irrigated with 30 µM dexamethasone solution for a week prior to sampling. The abundance of p19 transcripts relative to that of ACT7 is given (dotted column) alongside the amount of protein given relative to the chemiluminescence of a ColO control sample as determined by immunoblotting (hollow circles). Prolonged cultivation in the presence of dexamethasone led to serration of the leaf margin and the leaves showed reduced elongation (Figure 3.17). The severity of this effect correlated with the extent of p19 protein accumulation (Figure 3.16) and matched that seen in the lines containing p19 under the control of the 35S promoter described above. Induced plants reverted to a phenotype resembling that of ColO after the removal of dexamethasone, although a time lag of 1–2 weeks was seen before a return to normal leaf morphology, and in some instances altered floral morphology was seen a month after irrigation with dexamethasone was halted. Figure 3.17: Chemically-induced expression of p19 effects leaf architecture in Arabidopsis. Transgenic lines grown on soil irrigated with dexamethasone for six weeks showed the formation of leafs with increased abundance of lobes and serration of the leaf margin. Control lines had a normal leaf phenotype, and the serrated leaf margin was only seen after dexamethasone application. The line number given correlates that that stated in Figure 3.16. Scale = 1cm. Chapter 3: miRNA Regulation of the C4 Cycle 68 3.2.3.5: pOp::p19 in C. gynandra The pOp::p19 construct was transformed into C. gynandra via Agrobacterium-mediated in vitro transformation (Newell et al., 2010). Between one and three regenerative callus balls were recovered from each explant within 4 weeks of culture, although organogenesis was seen within six weeks in about 60% of those recovered. The regeneration of shoots was seen within 12 weeks of culture at an average rate of two to five shoots per 20 callus balls. The induction of root formation was routinely unsuccessful, and only one intact plant was recovered. This was derived from callus tissue rather than a shoot/hypocotyl structure, and so was unsuitable for prolonged culture. Due to the lack of root induction, it was necessary to establish grafts between transgenic shoot scions and wild type root stocks that had already been established in soil. The establishment of such grafts faced a number of significant technical challenges. The in vitro grown material tended to be somewhat hyperhydratic as a result of prolonged culture, which compounded the difficulties of establishing vascularisation between root stock and scion, leading to increased risk of desiccation. Secondly, while the sucrose supplementation of the culture media was reduced in the later stages of culture to promote self-sufficiency of the scions, the ambient light levels were constantly low-level which led to a risk of photodamage when exposed to the elevated light levels under which C. gynandra is grown in soil. Finally, C. gynandra plants only form substantial stems when approaching flowering. The axillary buds showed rapid induction after removal of the apical bud, often reaching functional maturity before the establishment of vasculature between root stock and scion was fully developed. To avoid problems relating to the maturity of the root stock, all lower leaves and axillary bud material was removed both at the point of grafting and at regular intervals until the graft had established. The plants were transferred to low light, high humidity environments to prevent desiccation and photodamage. After establishment, it was found that a carefully graduated process of hardening off was necessary before exposure of the graft to normal cultivation conditions. Such a transfer, however, was necessary as it was found the prolonged growth under low light/high humidity promoted the formation of small, short lived leaves and inhibited flowering, as illustrated in Figure 3.18. This was necessary during establishment of the graft, but transfer to normal growing conditions led to normal leaf development and flowering. The production of vegetative stem tissue was promoted by growth under short day conditions, and Chapter 3: miRNA Regulation of the C4 Cycle 69 chemical fertilisation. Increased stem formation is seen in C. hasseleriana, but attempts to establish interspecific grafts between C. gynandra and C. hasseleriana were unsuccessful. Figure 3.18: An established graft between a transgenic C. gynandra scion (S) and a WT root stock (R). The site the graft is indicated with an arrow. A vertical slit was made in the root stock, into which a notch was cut. The base of the scion was cut into a wedge, inserted into the notch and the graft secured with a plastic collar. All axillary buds of the rootstock were removed at the point of grafting, and the plant was left in a high humidity, low light environment to protect the scion from desiccative or photoxidative damage. This environment caused the leaves of the established graft to be small, curled and short lived (illustrated here) but a return to normal leaf morphology was seen upon return to standard cultivation conditions. Scale = 1cm. Despite these modifications to the grafting procedure, only one transgenic pOp::p19 line was recovered. This was back crossed with both itself and WT C. gynandra individuals in order to provide T1 seed. As an initial test of the presence of the pOp::p19 construct, a subset of leaves on the T0 individual were painted with a 30 μM dexamethasone solution before GUS staining after 24 hours. Those leaves which had been treated with dexamethasone showed GUS activity in a pattern similar to that seen in the Arabidopsis transgenics, while the leaves which had not been induced did not show GUS activity (Figure 3.19-A). Once the efficiency of transient surface-applied dexamethasone induction was demonstrated, the effects of induction through soil irrigation were explored. Dexamethasone requires ethanol as a solvent, but the extent of dilution resulted in concentrations of 0.01% in the induction solution. The application of blank ethanol-containing mock induction (MI) solutions to the leaf surface was found to promote a measure of pitting and necrosis which may have been attributable to the effect of ethanol or optical effects of water droplets on the leaf surface promoting photodamage. Secondly, the induction of p19 expression in leaf primordia was desirable in order to assess the effects of p19 expression on the development of Kranz anatomy, but the nature of the apical meristem prevented direct surface application of dexamethasone. To avoid these effects, induction through the transpiration stream was investigated by irrigating the roots with a dexamethasone solution. This promoted systemic expression of GUS in both the cauline leaves and within developing leaf primordia (Figure 3.19-B,C). Chapter 3: miRNA Regulation of the C4 Cycle 70 Figure 3.19: GUS expression is inducible in transgenic C. gynandra plants. Localised expression within an individual is possible by direct application to the leaf surface (A) as shown by induced (A-ii) and mock induced (A-i) leaves taken from the same individual. Systemic induction is possible through soil irrigation with dexamethasone, leading to GUS expression in all leaves (C, shown in age order). Induction through the transpiration stream is sufficient to generate GUS expression in developing leaf primordia, shown here accumulating in the veins prior to full vascular development (B). The developmental gradient of Kranz anatomy can be seen from the tip (X) to the leaf base (Y). Scale = 1cm (A.C), 1mm (B). X Y Chapter 3: miRNA Regulation of the C4 Cycle 71 Once GUS induction was confirmed, attempts were made to assess the extent of p19 accumulation after induction. Individuals from the T1 generation were induced and sampled for the presence of p19 transcripts after four and 24 hours, and a high level of p19 expression was seen in induced (I) individuals between four and 24 hours which was absent in the mock induced (MI) individuals (Figure 3.20). Given the correlation between GUS expression and p19 accumulation, and that GUS expression was detectable in the developing primordia of induced C. gynandra plants, it was considered likely that p19 could be induced before leaf development. The development of Kranz anatomy occurs in a basipetal direction when the leaf is 4-7mm in length, a period of development which can be seen occurring in Figure 3.19-B. Transgenic lines were cultivated in the presence of dexamethasone for a month before the developing leaves were examined. Samples were taken beyond the stage at which Kranz anatomy should have developed, and Kranz anatomy was evident in both the induced and MI lines (Figure 3.21). In addition, the developed leaves appeared normal, and no significant morphological phenotype was recovered. Figure 3.20: p19 expression is inducible in transgenic C. gynandra individuals when irrigated with dexamethasone solution. No expression was seen at time point 0, and expression remained negligible in the mock induced individuals after 24 hours (MI – 24). Induced individuals (I) showed a marked increase in p19 accumulation between 4 and 24 hours. Figure 3.21: p19 expression does not alter the development of Kranz anatomy in C. gynandra. The leaves of plants cultivated on dexamethasone-containing soil showed fully developed Kranz anatomy in both mock- induced (A) and induced (B) plants. In addition, general leaf development appeared unaffected. Scale = 1mm. Chapter 3: miRNA Regulation of the C4 Cycle 72 Once successful systemic induction of p19 expression was established through soil-based application of dexamethasone, a course of week-long induction of p19 was carried out before photosynthesis was assessed. Successful induction was confirmed via GUS staining. p19 expression had considerable impact on photosynthesis in C. gynandra (Figure 3.22, Table 3.6). While quantum efficiency was constant between induced and mock-induced (MI) individuals, the level of dark respiration was increased from -1.2 to -2.0 µmol CO2 m-2 s-1 in p19-expressing individuals (p = 0.01). The increased dark respiration combined with the constant quantum efficiency increased the light compensation point in p19 expressing individuals by 17 µmol m-2 s-1 (p = 0.03). Analysis of the response of photosynthesis to internal concentrations of CO2 allowed both carboxylation efficiency and the CO2 compensation point to be derived. Induction of p19 reduced carboxylation efficiency (an increase of 0.05 µmol mol-1 from the MI individuals, p = 0.08) and increased the CO2 compensation point (an increase of 13 µmol Q m-2 s-1 from the MI individuals, p = 0.03). Finally, the PMax of induced individuals was reduced by 3.0 µmol CO2 m -2 s-1, but this was not statistically significant. Chapter 3: miRNA Regulation of the C4 Cycle 73 Figure 3.22: p19 expression effects photosynthesis in C. gynandra. Under changing light intensity (A) quantum efficiency is not altered, but an increase in dark respiration is evident. Under changing CO2 concentrations (B) the carbon compensation point is increased, while carboxylation efficiency is depressed. A B Chapter 3: miRNA Regulation of the C4 Cycle 74 Quantum Efficiency Carboxylation Efficiency Dark Respiration Light Compensation Point Γ PMax MI 0.05 ± 0.003 0.13 ± 0.018 -1.22 ± 0.22 26.40 ± 4.61 8.97 ± 0.75 20.85 ± 1.77 p19 0.05 ± 0.002 0.08* ± 0.011 -2.01* ± 0.08 43.57* ± 2.10 21.98* ± 2.56 18.80 ± 1.32 0 10 20 30 40 50 60 70 80 RbcS1A LHCA1 PPDK PyTR1 R el at iv e T ra n sc ri p t A b u n d an ce p19 WT 0 1 2 3 4 5 6 7 8 PIP RCA NAD-ME2 T ra n sc t T ra n sc ri p t A b u n d an ce * * * * Figure 3.23: Changes in transcript abundance in C. gynandra are detectable that may be attributable to p19 expression. Some genes do not show significant changes in abundance between p19-expressing and mock-induced individuals (e.g. PYR, CA4, PSBQ2), while PEPC, RbcS1A, RCA and ASP2 show levels of abundance which are significantly elevated relative to the MI individuals. Those transcripts which showed a difference with a p value less than 0.1 are marked with an asterisk (*). Transcript abundance is relative to ACT7. Table 3.5: p19 expression effects photosynthesis in C. gynandra. While the quantum efficiency (µmol CO2 µmol m -2 s -1 ) remains constant between p19-expressing and mock induced (MI) individuals, an increase in dark respiration (µmol CO2 m -2 s -1 ) is seen which leads to a corresponding increase in the light compensation point (µmol Q m -2 s -1 ) at 400 µmol mol -1 CO2. A significant difference is not seen in the maximum rate of photosynthesis (PMax, µmol CO2 µmol Q m -2 s -1 ). The carboxylation efficiency (Γ, µmol mol -1 ) is increased with p19 expression, while the carboxylation efficiency (mol CO2 m -2 s -1 ) is decreased. Those values that showed a p value below 0.05 between MI and p19 values are marked (*). Chapter 3: miRNA Regulation of the C4 Cycle 75 Concurrent with the photosynthetic analysis, RNA and protein samples were prepared for quantification. Analysis of relative transcript abundance (Figure 3.23) showed that most components of the C4 cycle maintained comparable levels of expression between p19- expressing and MI individuals. Although some average expression values differed, these were not statistically significant (e.g. RbcS1A, PEPC). However, two genes showed statistically significantly elevated levels of expression when p19 was expressed compared with the MI individuals: RCA (+ 53%) and RbcS (+ 23%). While such differences were detectable at the transcript level, no significant variation was seen in the abundance of protein (Figure 3.24). While antisera against all components of the pathway were not available, no changes were detected in the rate of accumulation of either subunit of RuBisCO. No antisera were available for RCA or ASP2, so the differences in transcript abundance that had been detected for these two genes could not be assessed. Figure 3.24: p19 expression does not affect protein accumulation in C. gynandra. Protein abundance for PPDK, NAD-ME, PEPC and the SSU/LSU of RuBisCO are constant between individuals expressing p19 (+) and the MI controls (-). 10µg of total protein extracts prepared from mature leaf tissue were run out on 10% denaturing polyacrylamide gels prior to treatment with an appropriate antisera. The molecular weight of the expected band is given next to each blot, and a Coomassie-stained parallel gel to demonstrate equal protein loading rates. Chapter 3: miRNA Regulation of the C4 Cycle 76 3.3: Discussion 3.3.1: miRNA Prediction A library of sRNA sequences in C. gynandra was compiled with good coverage between 18- 25nts. The presence of previously described miRNA sequences within the library showed not only that sRNAs in C. gynandra are conserved when compared with other species, but also that the method used to compile the library was sufficient to obtain reads for functional miRNAs. An initial search indicated that hits against potential target genes were over- represented in some C4 genes (RbcS, RCA, both isoforms of PPDK and of NAD-ME) which corresponded with the over-representation of sRNA hits against C4 genes in maize (Tolley et al., unpublished). These reads may represent mRNA decay products or endogenous sRNAs (Kim & Nam, 2006). The absence of genomic DNA sequence for C. gynandra meant that it was not possible to identify miRNA sequences on the basis of a separately encoded hairpin precursor. The presence of such a precursor can be identified by RNA blots (Berezikov et al., 2005), but this approach would be challenging due to the large number of candidates. Therefore, it was necessary to examine the alignment patterns of mismatch reads against ESTs. 3.3.2: Over-Representation of Hits against Certain C4 Genes A number of distribution patterns along target transcripts were evident. Some sequences (e.g. PEPC2) showed a diffuse pattern of reads with little overlap which was taken to indicate that the presence of an underlying miRNA sequence was unlikely. In some cases, an aggregation of reads aligned against a single site created a concatenated sequence greater than 24nt (e.g. NAD-ME2), and these may contain an underlying miRNA sequence. However, numerous alignments in the uni- or bidirectional orientation were recovered and in some instances these contained sequences in both orientations. The nature of the sequencing method used should ensure that read orientation is retained, and so reads in the sense direction were classified as probable decay products. While sense alignments may indicate DNA-based activity such as methylation (Chellappan et al., 2010, Bao et al., 2004), a broad screen of methylation sites would be challenging. In addition, methylation of base DNA would need to occur during the acquisition of C4 competency, and so these miRNAs could not be regulated to maintain process in a mature C4 leaf. The genomic sequence of PPDK showed high read alignment in both sense (17 & 125 reads) and antisense (20 reads) directions. These sites were within introns and so are unlikely to represent a regulatory miRNA targeting the transcript. Interestingly, this pattern of over- Chapter 3: miRNA Regulation of the C4 Cycle 77 accumulation is present in the C. gynandra sRNA library compiled by Aubury (unpublished) but is absent from comparable data in Arabidopsis. These data may indicate a potential role for methylation, or transcript decay products. While attempts were made to screen for cleavage products of PPDK, the apparent inability to recover products greater than 500bp for a 5’RACE reaction limits the extent to which this method can be utilised to screen for cleavage products. The major sRNA alignment clusters against PPDK occurred far into the 3’ region of the transcript (Figure 3.4: C-F), and it is likely that considerable optimisation of the RACE procedure would be required before a satisfactory level of amplification could be achieved to test for cleavage products at these sites. 22 antisense reads aligned to a single site in the transcripts of MDH, giving a sequence of 20nt with a pair of highly supported mismatches and so represent a good candidate miRNA sequence. A large amount of sequence aggregation exceeding the 24nt threshold was seen in other targets, such as the 36nt region covered by 17 reads in CA4. Similar aggregations were seen against RbcS, NAD-ME and ALT1. The sequences covering these loci were examined for conserved sequences present in previously described miRNAs, as the high level of conservation within miRNA families has facilitated miRNA prediction previously (Zhang et al., 2005). Those sequences targeting NAD-ME2 bear the closest resemblance to miR399 and miR2111 which have been described as playing roles in regulating phosphatase transporters (Jones- Rhoades & Bartel, 2004) and phosphate metabolism (Hsieh et al., 2009) respectively. miR5293 showed close relatedness to a sRNA read aligning against PPDK, although the activity of miR5293 has been related to the formation of mycorrhizal relationships in Medicago (Devers et al., 2011). sRNA reads aligning against RbcS and ALAAT1 showed similarity to miR4398 and miR5655 respectively, although the function of the targeted sequences for these miRNAs have yet to be established. While variation is seen in miR2111 and miR399 between divergent lineages, the extent of this variation is not sufficient to indicate that the proposed miRNA sequences are members of either family. Equally, the proposed target sites do not show complementarity to either miRNA family, making it unlikely that NAD-ME is targeted by members of the miR211 or miR399 families. While sequence conservation can be effective in identifying miRNAs (e.g. Zeng et al., 2010) this approach is limited. For example, Lu et al. (2005; 2008) described over 50 miRNA sequences from Populus trichocarpa via sRNA cloning that were not found in Arabidopsis. Therefore, it is conceivable that novel miRNAs are active in C. gynandra which target Chapter 3: miRNA Regulation of the C4 Cycle 78 components of the C4 cycle, but show no similarity to reads present in Arabidopsis. However, given the close phylogenetic proximity of Cleome to Arabidopsis (unlike Populus) some measure of sequence conservation is to be expected. Regulatory miRNAs could be recruited by two routes into the C4 cycle: mutation of the miRNA to target a C4 gene or by modification of the C4 gene to possess a miRNA targeting site. In either scenario, some measure of ancestral conservation would be expected. 3.3.3: The Identification of Potential Regulatory miRNA Sequences in C. gynandra While the approach outlined above did not specifically identify miRNA sequences that regulate the C4 cycle, an uneven distribution of sRNA reads aligning against the components of the C4 cycle was found, and this was similar to that reported in maize (Tolley et al;., unpublished). This pattern of sRNA alignments may indicate underlying miRNA activity in a number of instances. While it is possible that novel miRNA sequences are active in C. gynandra, the lack of a sequenced genome means that miRNAs can only be compared to EST sequences. A key area of C4 regulation which is poorly understood is the nature of trans-acting factors which may bring about the cell-type specific patterning associated with the C4 cycle. Given that the vast proportion of miRNAs target transcription factor families (Carrington & Ambros, 2003), if miRNAs are involved with regulating the C4 cycle then they are most likely to target the trans acting factors rather than the specific components. Therefore, a lack of understanding of the transcription factors involved with the C4 cycle means that it is not possible to explore the most likely area of miRNA activity using bioinformatics with data taken from a sRNA library. 3.3.4: Establishing Transgenic Lines In vitro transformation of the 35S::p19 construct into C. gynandra was successful, but the severity of the phenotype prevented establishment of transgenic lines. The successful utilisation of the inducible pOp::p19 construct in Arabidopsis therefore provided a method of by-passing this. Recovery of C. gynandra transgenics containing the pOp::p19 construct was limited to a single line. Two transgenic calli generated roots in vitro (this represents a recovery rate of less than 0.1%) and time restrictions meant that obtaining more rooted plantlets was not possible. While root formation can be impacted by a wide range of factors including osmotic conditions, nutrient composition, growth regulator choice and explant material (Marks & Simpson, 2000; Conner & Falloon, 1993; Parra & Amo-Marco, 1996), the formation of roots seen in this study indicates that the current system is capable of root induction. It is likely that root formation is triggered either by variation in media composition (e.g. localised Chapter 3: miRNA Regulation of the C4 Cycle 79 depletion of components, or accumulation of breakdown products) or callus stress, although further exploration of these aspects would be worthwhile if this study were to be extended. As an alternative, transgenic scions were grafted onto wild type root stocks. While optimisation of the process was undertaken, it was still difficult to promote graft establishment, meaning that only one transgenic line was recovered. Therefore, the conclusions drawn from this study should be considered in this light. While the transformation of the pOp::p19 into Arabidopsis was primarily carried out to test its functionality, it remains an effective system to explore the effects of the inhibition of miRNA activity in a spatial and temporal fashion. 3.3.5: Constitutive expression of p19 alters leaf and flower anatomy in Arabidopsis Expression of p19 led to the formation of individuals with heavily lobed, serrated leaves which showed stunting, premature flowering and the formation of flowers with malformed sepals, petals and anthers which were effectively sterile. The depressed rate of transgenic recovery at the T0 stage could be attributable to the effect of strongly expressing p19 lines in that highly expressing individuals were incapable of surviving after transfer to soil. In vitro cultivation of transgenics was attempted to bypass this, but this was prevented by the secondary phenotypic effects caused by prolonged in vitro culture of Arabidopsis. Numerous pathways in leaf regulation are subject to miRNA regulation. For example, transgenic lines overexpressing the TCP3 transcription factor show the formation of serrated leaf margins, matching that shown in the p19-expressing lines generated by this study (Koyama et al., 2007). The TCP family is negatively regulated by miR164, the precursor to which is encoded by the JAW locus, and loss-of-function jaw mutants produce serrated leaf margins and altered leaf morphology (Palatnik et al., 2003). Therefore, the effects of p19 expression can be attributed to its effects on miRNA activity rather than a direct effect on the p19 protein itself. The effects of p19 expression on internal leaf anatomy have not been described previously, and this study indicates that p19 expression has a significant effect on internal leaf anatomy (Figure 3.8-C). While WT Col-O leaf material shows the standard arrangement of palisade and spongy mesophyll, this arrangement was greatly disrupted in p19-expressing individuals. In addition, p19 expressing lines demonstrated irregular and enlarged epidermal cells. Again, the loss of miRNA-mediated regulation of PHAVOLUTA promotes disorganised adaxial mesophyll and palisade development in Nicotiana (McHale & Koning, 2004), and over-expression of Chapter 3: miRNA Regulation of the C4 Cycle 80 TCP3 causes undifferentiated, rounded epidermal cells (Koyama et al., 2007). Therefore, the effects of p19 expression on internal leaf anatomy are considered to be via interference with the miRNA regulatory pathway. The expression of p19 causes a phenotype similar that which has been reported for miRNA inhibition by previous authors (e.g. Dunoyer et al., 2004). The latter authors note, however, that the p19 phenotype remained constant despite variation in expression levels (unlike that of P1-HCPro which showed a dose-dependent response). This contrasts with the results of this study which illustrate a correlation between the phenotype severity and the extent of p19 accumulation. Due to the activity of p19 as a competitive inhibitor (Vargason et al., 2003; Lakatos et al., 2004), it is conceivable that elevated p19 accumulation will increase its effectiveness as a competitor for dsRNA, hence increasing of the resultant phenotype. Flowering was seen only in weakly expressing Arabidopsis lines, and altered floral morphology was evident in these instances. p19-expressing lines showed a reduction in stem formation with maximum cauline leaf development, leading to a dense, conical growth habit (Figure 3.9- C,D) which eventually produced a prematurely terminated raceme, and pollen dehiscence was not recorded. Similar phenotypes have been reported in other instances of p19 and P1-HCPro expression (Dunoyer et al., 2004; Lakatos et al., 2006), further supporting the hypothesis that the p19-induced phenotype is attributable to its effects on miRNA activity. 3.3.6: p19 Expression can alter Leaf Anatomy of C. gynandra C. gynandra containing 35S::p19 showed a strong phenotype with stick-like, radialised leaves and little to no mesophyll development. The lack of expansion of the leaf blade in p19- expressing transgenics is considered to be attributable to the lack of vein development. Mesophyll development does occur, but the reduction in venation has prevented the expansion of the leaf blade away from leaf midrib. This phenotype correlated with the accumulation of p19 at the transcript level, and so was presumably related to the activity of p19 at the protein level. The radialisation of leaf tissue correlates with loss-of-function mutants of PHABULOSA and PHAVOLUTA (Mallory et al., 2004) which are subject to miRNA regulation, and to mutants deficient in AGO1 (Kidner & Martienssen, 2004). The severity of the phenotype recovered in transgenic individuals exceeds that seen in the Arabidopsis 35S::p19 expressing lines. While it is difficult to draw comparisons between soil- and in vitro-grown material, it is conceivable that carbohydrate supplementation of the culture media used in C. gynandra cultures was sufficient to promote growth of organs that would be unsustainable in a soil- grown individual due to the lack of photosynthetic tissue production. The initial in vitro Chapter 3: miRNA Regulation of the C4 Cycle 81 cultures demonstrated that p19 was effective in C. gynandra, but the severity of the phenotype meant that it was not possible to recover stable expressing lines from the established cultures. While RNA and protein quantification would have been possible, any changes may have been caused by the in vitro cultivation process rather than the effect of p19 on the C4 cycle. 3.3.7: p19 Expression can be induced in C. gynandra and Arabidopsis Placing p19 under the control of the inducible pOpON promoter (Samalova et al., 2005) facilitated transient expression of p19 in both Arabidopsis and C. gynandra. Testing the system with Arabidopsis showed that GUS accumulation was detectable within 4 hours of the application of 20–50μM dexamethasone solution to the leaf surfaces, and p19 was detectable at the transcript and protein level only after induction. Prolonged induction led to a phenotype matching that seen in plants containing p19 under the control of the CaMV 35S promoter. Surface application of dexamethasone led to the development of necrotic patches and curvature of the leaf in C. gynandra, but these effects were seen in mock-induced individuals and so the effects were attributed to the ethanol used as a solvent rather than dexamethasone itself, supporting the findings of Samalova et al. (2005). As root irrigation was used for induction throughout the study, these problems were bypassed. It was not possible to assess the impact of a functional pOp promoter in the absence of p19 expression. Craft et al., (2005) and Sheldon et al. (2008) reported no detrimental effects of pOp expression in control lines, but testing of an empty pOpON vector would be advisable if this study were extended. In C. gynandra, systemic expression of GUS was detectable within 24 hours of soil irrigation with dexamethasone, to the extent that it could be detected in the apical bud and developing leaf primordia (Figure 3.9-B). The distribution of GUS was unequal, however, with highest levels of expression seen in the oldest and youngest leaves (Figure 3.9-C). Samalova et al. (2005) reported an unequal distribution of GUS expression, although the greatest GUS expression was seen in mid-aged leaves, with reduced expression in the youngest and eldest leaves. It is conceivable that differences in transpiration rates between leaves may be responsible for this variation. While the origin of the unequal GUS expression remains obscure, the leaf tissue sampled for this experiment was taken from the youngest mature leaf, a site which should have shown maximal p19 expression. 3.3.8: Induced expression of p19 does not affect the development of Kranz Anatomy The accumulation of GUS in primordia of plants irrigated with dexamethasone was detectable at a stage before full development of Kranz anatomy had occurred (3.19-B). Prolonged Chapter 3: miRNA Regulation of the C4 Cycle 82 induction, however, did not yield any modification in leaf anatomy either in terms of gross leaf structure or formation of Kranz anatomy. Due to the low RNA yields from primordial tissue it was not possible to validate p19 expression in the developing leaves, but given the correlation between the accumulation of GUS and p19 seen in other tissues, it is likely that p19 was being expressed in the developing leaves. As p19 is capable of producing a vegetative phenotype in C. gynandra when controlled by the 35S promoter, but follows a dose-dependent response in Arabidopsis, it is conceivable that the lack of a vegetative phenotype in the pOp promoter line is due to relatively low levels of p19 expression. 3.3.9: p19 expression negatively impacts photosynthesis in C. gynandra Analysis of photosynthetic gas exchange of the p19-expressing and mock induced (MI) C. gynandra individuals demonstrated a number of significant differences. The MI individuals had a carbon compensation point (Γ) of 8.9 μmol m-2 s-1, which is comparable to values published by Voznesenskaya et al. (2007) and Marshall et al. (2007). The average Γ values reported for other C3 members of Cleome given by these authors is between 40–65 μmol m-2 s1. However, in p19-expressing individuals, Γ is increased to an average of 22.0 μmol m-2 s-1, and while this does not approach the Γ of C3 taxa (Marshall et al., 2007), this is significantly higher than in functional C4 leaves. The cause of this increase appears to be two-fold: firstly, the level of dark respiration in C4 p19-expressing individuals is increased by 0.8 μmol CO2 m-2 s-1, and the carboxylation efficiency (CE) is decreased by 0.05 mol m-2 s-1. The quantum efficiency was not significantly different between MI and p19-expressing individuals, so initial photochemistry is not being disrupted. While the effects of p19 expression on photosynthesis have been quantified, the cause of these effects is yet to be identified, and a number of components may be responsible for the changes in Γ and CE. For example, Γ may increase due to changes in light-dependent non- photorespiratory CO2 production (e.g. nitrate metabolism, Guo et al., 2005), or to changes in photorespiration. Furthermore, dark respiration values are relatively equal between C3 and C4 species (Byrd et al., 1992), so while a return to C3-like values is seen in the CE, the increase in dark respiration may represent an effect independent of C3 and C4 photosynthesis. Therefore, it would be advisable to quantify the extent of photorespiration (Brooks & Farquhar, 1985) in order to separate these effects. The observed changes in CE may be due to changes in the activity of RuBisCO or other components of the C4 cycle. Changes in RuBisCO activity can lead to a depression in CE (Jacob & Lawlor, 1992), while CE and dark respiration can be altered by metabolic instability. Chapter 3: miRNA Regulation of the C4 Cycle 83 For example, exposure to ozone or heavy metal toxins can depress CE due to direct effects on photosynthetic pathways, while dark respiration increased in line with elevated protein turnover rates (McLaughlin et al., 1991; Reich, 1983; Parys et al., 1998). Therefore, the photosynthetic effects observed in this study may be due to single or multiple causes, which may or may not be directly attributable to photosynthesis. If p19 is affecting photosynthesis directly, it may destabilise the patterns of gene expression associated with the C4 cycle. For example, a collapse of cell-type specific expression caused by negative regulation via a miRNA may be detectable at the transcript or protein level. The establishment of strict C4 patterning is considered integral to the optimisation of the C4 pathway (Hibberd & Covshoff, 2010), so it conceivable that a partial reversal of this may lead to increased inefficiency manifesting as the changes in Γ and CE recorded here. 3.3.10: p19 Expression alters Transcript, but not Protein Accumulation If the changes in Γ and CE are attributable to reduced negative regulation of a component of the C4 cycle, then the altered patterns of accumulation may be detectable at the transcript or protein level. At the transcript level, multiple genes show equal expression between the induced and MI individuals, indicating that p19 expression does not universally effect endogenous gene expression. While a number of genes showed a slight increase in relative abundance after p19 expression, transcripts of RCA, RbcS and ASP2 show a statistically significant difference between p19 expressing a MI lines, with RCA and RbcS showing an increase in response to p19 expression. Both transcripts showed an abnormally high level of read accumulation during the analysis of the sRNA library. While no definitive miRNA sequences were identified from these alignments, they still represent potential sites of underlying miRNA activity, and the increase in transcript abundance seen with p19 expression may signify a reduction in negative repression. Interestingly, the upregulation of RuBisCO and/or RCA may explain the observed changes in gas exchange properties seen after p19 induction. The effects of overexpressing a single component of the C4 cycle has not be fully explored to date, and so it is not possible to state that the changes in photosynthesis observed in this study are derived from changes in overexpression of a C4 component as a result of a removal of negative regulation by a miRNA. Sage (2004) notes that as part of optimisation of the pathway during its early evolution, significant changes in expression of both existing and newly recruited genes in order to maximise the efficiency of the cycle. Therefore, any disruption of the cycle caused by an over accumulation of a single component which upsets the core metabolite flow could lower net Chapter 3: miRNA Regulation of the C4 Cycle 84 photosynthetic efficiency. Quantification of enzyme activity was not included in this study, but screening the core components of the C4 cycle for changes in activity after p19 expression may enable the identification of any induced changes, attributable to the specific enzyme or its regulators that may occur in a background of p19 expression. While correlations may be seen between changes in transcript abundance and the small RNAs identified from sequencing, these effects are not seen at the protein level. Of the proteins assessed by this study, relatively equal levels of accumulation were seen between induced and MI plants, although abundance of RCA was not examined. Some discrepancy is seen between accumulation of transcripts and protein level, which may be indicative of two potential scenarios. Firstly, these genes may be subject to multiple levels of control: a failure of regulation at the transcript level is superseded by a second, redundant, regulatory pathway at the protein level, a system which has been described in Amaranthus (McCormac et al., 1997). Alternatively, the changes in photosynthetic response may be due to changes in abundance or activity of a variety of other genes which have not been explored in this study. A key extension for this work, in addition to profiling enzyme activities, would be the utilisation of in situ immunolocalisation or RNA hybridisation to identify any chances in the cell specificity of the components of the C4 cycle. This would not only bypass the problems associated with whole-tissue qRT-PCR analysis, but would provide definitive evidence of any loss of expression patterning as a result of p19. In addition, utilisation of a single transgenic line in this study means that it has not been possible to explore any dose-dependent response of photosynthesis. Therefore, it would be advisable to expand this study to include additional lines. Additional lines of exploration would be possible if two key areas were addressed. Firstly, the provision of an annotated genome for C. gynandra would allow for the identification of both target and miRNA-precursor site identification (Wang et al., 2005; Lu et al., 2005). The ability to screen for both sequences would not only enable more effective prediction of miRNAs, but also of their targets. Secondly, an improved transformation protocol for C. gynandra would enable a number of transgenic routes to be explored. For example, fusion of proposed target sites to uidA sequences could be a rapid way of testing for M/BS specificity of proposed miRNAs (e.g. Millar & Gubler, 2005). Mutants deficient in miRNA silencing pathways have been used to identify regulatory miRNAs previously (e.g. Vaucheret et al., 2004), although these examinations have mostly focused on developmental pathways. Generation of such mutants in C. gynandra may be prohibitively complex, and it was hoped that similar effects Chapter 3: miRNA Regulation of the C4 Cycle 85 could be achieved through the use of the p19 protein in this study. Finally Srivastava et al. (2012) identified stress-induced miRNAs in Brassica juncea through microarray analysis using probes against known miRNA sequences, and it is possible that this method could be applied to C. gynandra upon further exploration of the sRNA library. These mechanisms could be adapted to C. gynandra, and may prove effective in obtaining more finite miRNA predictions. 3.3.11: Concluding Remarks Overall, these data may provide support for the hypothesis that miRNA-mediated regulation may play a role in regulating C4 photosynthesis. This has been demonstrated in a transgenic C. gynandra line that has shown a reduction in carboxylation efficiency and an increase in the carbon compensation point in response to inhibition of miRNA biosynthesis by the p19 protein. Two genes (RbcS and RCA) show high levels of sRNA alignment and statistically significant increases in relative transcript abundance after p19 expression, although this does not appear to persist to the protein level, indicating that these changes may not be the origin of the observed changes in photosynthetic capacity. Through the utilisation of sRNA profiling, it has been demonstrated that a disproportionate accumulation is seen in sRNA reads aligning against certain genes of the C4 cycle in C. gynandra. While it has not been possible to specifically identify a conserved miRNA sequence targeting known components of the C4 cycle, a number of focused accumulations of sRNA reads are sufficiently disproportionate to warrant further examination to test whether an underlying miRNA sequence may be present. While a specific regulatory miRNA has been identified by this study, there are indications that miRNAs may have a role to play in establishing or maintaining the C4 photosynthetic cycle. Several areas of this study could be extended to further in order to evaluate the areas highlighted by the data here reported. Chapter 3: miRNA Regulation of the C4 Cycle 86 Chapter 4: Changes in Gene Regulation during Cotyledon Development 87 4. Changes in Gene Regulation during Cotyledon Development 4.1: Introduction 4.1.1: Both Environmental and Developmental Factors may Influence the C4 Cycle One approach to examine gene regulation associated with the C4 cycle is to investigate changes which occur during development. The onset of C4 photosynthesis may occur at a predetermined developmental stage (Wang et al., 1993b) or in response to environmental stimulation (e.g. Langdale et al., 1988). While variation may be seen within and between lineages, the changes in gene expression that occur during maturation of a C4 leaf or cotyledon may yield insights into methods of regulation of the C4 cycle. Illumination can trigger significant changes in gene expression, and is both sensed and transduced in a variety of ways (Desprez et al., 1998, Ma et al., 2001). For example, specific wavelengths can be detected via photosensitive pigments such as phytochrome and cryptochrome (e.g. Neff et al., 1994; Lin, 1998). Alternatively, light responses may be mediated by electrochemical gradients across the thylakoid membrane such as in the translational repression of psaA&B (Klein & Mullet, 1987; Kim & Mullet, 2003). Light has been shown to have a significant regulatory effect on the C4 cycle in various species. In maize, neither PEPC nor PPDK accumulate to detectable levels in darkness and light is necessary for cell-type specific upregulation of both enzymes (Langdale et al., 1984, Matsuoka et al., 1993). RuBisCO accumulates at the same developmental stage in light- and dark-grown leaves of maize, albeit at lower levels in darkness (Nelson et al., 1984), with proteins of both subunits accumulating in both the mesophyll and bundle sheath until exposure to light triggers cell specificity (Langdale et al., 1988). Similarly, light is required for the correct cell-type patterning of PPDK, PEPC and the small subunit of RuBisCO in Flaveria, indicating that light-mediated responses are required for the acquisition of full C4 competency (Shu et al., 1999). Finally, dark-grown cotyledons of Amaranthus fail to show NAD-ME accumulation, and high levels of transcript accumulation are seen only after illumination (Long & Berry, 1996). Several instances have been described where promoter regions give both a light-mediated response and cell specificity, for example the PPDK and PEPC1 promoters of maize (Matsuoka et al., 2003; Kausch et al., 2001). However, enhanced expression and cell type-specific repression may be attributable to separate components. Shu et al. (1999) note that a lack of coordinated light-induced expression between PPDK and RuBisCO suggests that light- mediated effects on the C4 cycle are likely to be effected via indirect light-induced developmental signals rather than a direct effect of light upon the C4 cycle itself. This Chapter 4: Changes in Gene Regulation during Cotyledon Development 88 correlates with the observation that red light itself is insufficient to exclude RbcS transcripts from mesophyll cells in maize (Bilang & Bogorad, 1996) and the finding that RuBisCO specificity occurs as a response to metabolic changes associated with illumination rather than in response to light itself (Langdale & Kidner, 1994). In contrast to Flaveria and maize, PPDK and PEPC show mesophyll-specific expression in Amaranthus in darkness (Wang et al., 1993a), indicating that light is not required for the correct cell-type expression patterning in these species. In light-grown leaves, the restriction of both RuBisCO transcripts and protein to the bundle sheath of Amaranthus occurs concurrently with the sink-to-source transition (Wang et al., 1993b). However, the development of C4-like expression patterns of both subunits occurs in etiolated cotyledons indicating that light is not required for this transition (Wang et al., 1993a). Light does influence the accumulation of RuBisCO, however, as transcripts of both subunits are subject to translational repression in darkness via modified ribosome interaction (Berry et al., 1986; 1990). While similar processes are present in C3 plants (Kim & Mullet, 2003), the use of multiple levels of regulation is likely a common theme in the regulation of the C4 cycle (Hibberd & Covshoff, 2010). A lack of congruence between transcript and protein abundance is also seen in Amaranth meristems and developing leaf primordial (Ramsperger et al., 1996). For example, NAD-ME transcripts in Amaranthus only develop cell specificity once immature leaves have expanded beyond a given threshold, although the NAD-ME α subunit polypeptide shows specificity throughout this period (Long & Berry, 1996). It therefore appears that a common theme in these processes is the development of the C4 expression pattern over an existing C3-like pattern. In conditions where a transcript is undetectable in darkness it is not possible to state whether its upregulation in light automatically generates cell specificity, or whether additional regulatory process are in operation to provide high levels of cell-type specific expression. 4.1.2: The Development of the C4 Cycle in C. gynandra Given that previous studies of the effects of de-etiolation upon gene expression have enabled the identification of the mechanisms of regulation associated with the C4 cycle (e.g. Wang et al., 1993a), a similar study was undertaken in C. gynandra to provide an additional C4 dicot species in which the extent of light-induction of gene regulation is described. Photosynthetic tissue grown in darkness will lack the necessary stimuli to trigger gene responses which are mediated either by direct light exposure or the metabolic changes associated with the acquisition of photosynthetic ability. After exposure to light, any changes in gene expression Chapter 4: Changes in Gene Regulation during Cotyledon Development 89 in response to illumination can then be detected. In maize, the production of leaf tissue is seen in darkness (Langdale et al., 1988), but etiolation tends to limit cotyledon development and prevents the formation of leaf tissue in some dicots. Attempts to produced etiolated leaf material in C. gynandra were unsuccessful, which indicates that leaf formation is inhibited in dark-grown C. gynandra plants, a similar situation to that described in Amaranthus (Wang, et al., 1993a). However, functional Kranz anatomy is present in the cotyledons of Amaranthus and Flaveria, and this has been utilised in previous studies of the effects of etiolation on the C4 cycle (e.g. Shu et al., 1999; Wang et al., 1993a). Kranz anatomy is present in the cotyledons of the C4 members of Cleome (Koteyeva et al., 2011), and so cotyledon material of C. gynandra was used in this study in the absence of etiolated true leaf material. 4.1.3: Experimental Design In order to examine the effects of illumination on cotyledon development in C. gynandra, concurrent experiments were established. Seedlings germinated in darkness were transferred to MS medium supplemented with 1% sucrose and set with 0.8% agar (Brown et al., 2011; Kajala et al., 2011). Seeds kept in darkness were manipulated under a green safe lamp, and while previous reports have suggested such wavelengths are sufficient to produce a light- inductive response (Baskin & Baskin, 1979), the limited wavelengths and exposure was deemed sufficient to prevent undue influence. Seedlings were then placed in light, or in darkness prior to exposure to light after seven days, with sampling of tissue occurring during the first six hours after light exposure. This period was deemed sufficient to identify changes in gene expression linked directly to illumination. Beyond this point, developmental and metabolite flux changes are likely to be considerable and so it will no longer be possible to separate any causative mechanisms in observed changes in gene expression based on the data obtainable. Light grown seedlings were harvested concurrently so that any influence of circadian rhythms or developmentally-linked changes could be taken into account. Changes in gene expression were then assessed through the use of qRT-PCR and immunoblotting to quantify abundance of transcripts and proteins respectively. In addition, the abundance of chlorophyll was examined as an indicator of the extent of cotyledon development. Finally, the anatomy of cotyledons grown in light and in darkness were examined to verify the presence of Kranz anatomy in both treatments. This experiment was performed three times, with three replicates in both light treatments in each instance. Chapter 4: Changes in Gene Regulation during Cotyledon Development 90 4.2: Results 4.3.1: Effects of Etiolation on Cotyledons Cultivation in darkness triggered the formation of stereotypical etiolated cotyledons: hypocotyl growth was advanced while the cotyledons failed to expand and retained yellow pigmentation. The seed coat did drop from the cotyledons, however, and etiolated cotyledons demonstrated full Kranz anatomy comparable with that of light-grown seedlings (Figure 4.1; 4.2). The intervein distance was reduced in etiolated cotyledons, and cell density appeared to be increased. Therefore, the development of Kranz anatomy occurs early in cotyledon formation independent of illumination, and cotyledon expansion is achieved through expansion of existing mesophyll cells, increasing the intervein distance as the cotyledons develop. The presence of stomata on both the adaxial and abaxial epidermis was also notable. The presence of the cotyledons after seven days means that the assessment of light exposure at this point is justified. Figure 4.2: Internal anatomy of light- grown and etiolated (dark-grown) cotyledons of C. gynandra. The development of enlarged bundle sheath (BS) cells distinct from the mesophyll (M) is evident in both cotyledons, indicative that the cotyledons develop a functional Kranz anatomy. Scale = 100µm. Images produced in collaboration with B. Kumpers. BS BS Figure 4.1 Etiolated cotyledons possess Kranz anatomy. Cotyledons grown in darkness (D) form Kranz anatomy, which is characterised by high levels of venation. This is comparable with light-grown cotyledons (L) showing normal development of Kranz anatomy. Scale = 25µm. SG SG SG SG M M Chapter 4: Changes in Gene Regulation during Cotyledon Development 91 Total chlorophyll concentrations showed an increase within six hours of light exposure in etiolated cotyledons (Figure 4.3). While concentrations of chlorophyll remained relatively constant in light grown cotyledons at around 1.53 µg per mg of fresh tissue, chlorophyll was undetectable in etiolated cotyledons until two hours after exposure to light. The increase in total chlorophyll appeared linear at a rate of 0.06 µg mg-1 hr-1. Mature leaf tissue of C. gynandra showed an average chlorophyll content of 2.25 (± 0.31) µg per mg of fresh tissue. Figure 4.3: An increase in total chlorophyll concentration can be seen in etiolated cotyledons within six hours of illumination. Chlorophyll abundance remains relatively constant in light-grown seedlings, while it was not detectable in etiolated cotyledons at the point of transfer into illumination (t=0). Chapter 4: Changes in Gene Regulation during Cotyledon Development 92 6.3.2: Transcriptional Effects Exposure to light had a significant effect on transcript accumulation in etiolated C. gynandra cotyledons (Figure 4.4 & 4.5). Only LHCA1 showed a complete absence in etiolated tissue, but it was detectable after 2 hours of light exposure. PPDK, CA4, NAD-ME and the transporters DiT2 and Bass2 were present in etiolated tissue, but showed a significant upregulation after light exposure. In the case of PPDK, CA4 and Bass2 this upregulation leads to levels which exceeded that seen in light-grown seedlings. PEPC and RCA accumulation fluctuated in etiolated tissue, although PEPC showed a level of accumulation greater than that seen in the light-grown tissue at t=6. Variation in transcript abundance was also seen in light-grown cotyledons, although this was most pronounced in DiT2 and Bass2. Overlaid on this was a general decline in transcript abundance seen throughout the course of the study. In light- grown material, the abundance at t=6 was around half that seen at t=0 in LHCA1, NAD-ME, RCA and CA4. The increase in expression seen in the etiolated material appears transient in NAD-ME and CA4, and levels began to decline within a period of time after the initial peak in expression. The extent of this decline, however, was within the boundaries set by variation seen within the result and was not considered to be significant. The only gene which showed increased levels of accumulation in etiolated seedlings was RbcS1A, which showed a level of accumulation some 1.5x that seen in light-grown cotyledons. Within the first 2 hours of light exposure this valued had decreased to below that of the light-grown seedlings, although a gradual increase in expression was seen between t=2 and t=6, by which point levels of accumulation approximated that seen in the light grown tissue. Chapter 4: Changes in Gene Regulation during Cotyledon Development 93 NAD-ME PPDK RbcS1A RCA DiT2 Bass2 DiT1 CA4 PSBQ PEPC LHCA1 Figure 4.4: Light induction causes changes in transcript accumulation. Of the genes profiled, only LHCA is not present in etiolated cotyledons. Upregulation of gene expression in response to illumination is seen, but the extent and point of onset of this increase varies between gene. Only RbcS shows an initial decline in accumulation levels after light exposure. Etiolated cotyledons are shown with a broken line, while light-grown seedlings are given a solid line. Chapter 4: Changes in Gene Regulation during Cotyledon Development 94 Figure 4.5: Light induction causes changes in transcript accumulation. Transcript abundance (relative to ACT7, and between light and dark grown plants) are given. A general increase is seen in transcript abundance after illumination (PEPC, PPDK, NAD-ME2) and while these increases may be transient (e.g PEPC) some transcripts remain elevated (e.g. CA4). RbcS shows an initial decline in accumulation after illumination. 0 0.5 1 1.5 2 2.5 0 2 4 6 PPDK 0 1 2 3 0 2 4 6 CA4 0 0.5 1 1.5 0 2 4 6 NAD-ME2 0 0.5 1 1.5 2 0 2 4 6 Bass2 0 0.5 1 1.5 0 2 4 6 DiT20 1 2 3 4 0 2 4 6 DiT1 0 0.5 1 1.5 2 0 2 4 6 RCA 0 0.5 1 1.5 2 2.5 0 2 4 6 PEPC 0 0.2 0.4 0.6 0.8 1 0 2 4 6 PSBQ2 0 0.5 1 1.5 2 0 2 4 6 RbcS Chapter 4: Changes in Gene Regulation during Cotyledon Development 95 4.3.3: Changes in Protein Accumulation The increase in transcript accumulation associated with illumination corresponds to the abundance of protein (Figure 4.6), although the levels of protein accumulation appear more constant throughout the course of sampling. PEPC and PPDK were detectable in dark-grown cotyledons, although at lower levels, indicating that light exposure was required for increased levels of accumulation. This is seen to a greater extent in NAD-ME, which was largely undetectable in etiolated material. Both subunits of RuBisCO were present in etiolated material, although the SSU appeared to be at a lower level of abundance compared with light- grown material indicating that increased expression is seen in light-grown cotyledons. The LSU appears more constant, although the abundance of this protein made it difficult to fully quantify. Figure 4.6: Changes in protein accumulation occur after illumination. PEPC, PPDK and NAD-ME are detectable in etiolated cotyledons, but at lower levels than light grown individuals. NAD-ME shows significant depression in dark grown cotyledons, but is still detectable. Similarly, an increase in the abundance of the SSU of RuBisCO is seen between are relatively constant between dark and light grown samples. Equal loading of total protein is shown by in the Coomassie-stained gel (Com.). Chapter 4: Changes in Gene Regulation during Cotyledon Development 96 4.3: Discussion 4.3.1: Post-germination Cotyledon Development in C. gynandra While some C4 species produce cotyledons that are C3 (e.g. Salsola gemmascens, Pyankov et al., 2000), others produce cotyledons with a functional C4 cycle (e.g. Flaveria (Shu et al., 1999), S. laricina (Pyankov et al., 2000). These data demonstrate that C. gynandra is in the latter group, forming developed Kranz anatomy in its cotyledons, an observation that matches the observations of Koteyeva et al. (2011). While the contrasting origins and functions of cotyledons when compared with mature leaf material may result in differing photosynthetic types, a limited number of mutations may yield leaf-like development of cotyledons (Meinke et al., 1992) which would suggest that genetic control is malleable. In addition, if the C4 cycle can be considered an ecological adaptation (Sage, 2004), then it is to be expected that cotyledons would possess the same photosynthetic type as that present in mature plants to enable maximal exploitation of a given ecological niche. However, from this data it is not possible to define whether Kranz formation occurs during cotyledon formation or maturation after germination. The latter instance has been demonstrated in S. richeteri (Voznesenskaya et al., 2003), and it may be worth extending the time course over which samples are prepared in order to properly quantify this. While Kranz anatomy is considered likely to be present in cotyledons prior to germination, the reduced development of the mesophyll in etiolated cotyledons suggests that maturation of the cotyledon is achieved predominately by the expansion of these cells. The vein density of the light-grown cotyledons is similar to that of mature C. gynandra leaves but this is lower than that of etiolated cotyledons (Figure 4.1, 4.2). Again, this matches expected developmental pathways: if Kranz anatomy is present prior to maturation then it would be necessary to expand existing tissues rather than forming de novo structures. Such a process has been observed in Arabidopsis, whereby cell numbers within the cotyledon remain relatively constant during development, while expansion is achieved through increases in cell volume with a corresponding reduction in cell density (Mansfield & Briarty, 1996). A reduction in the ratio of mesophyll to bundle sheath cells is seen during the expansion of Amaranthus tricolor leaves (Wang & Nii, 2001) so similar developmental processes have been recorded in other C4 dicots. The response of C. gynandra cotyledons to darkness is typical of dicotyledons: the elongation of the hypocotyl is considerable, while the cotyledons do not expand and do not accumulate chlorophyll. The accumulation of chlorophyll in response to illumination is recordable within Chapter 4: Changes in Gene Regulation during Cotyledon Development 97 six hours of illumination (Figure 4.3). Little to no change is evident in the abundance of chlorophyll of light-grown seedlings over the duration of this experiment, which would infer that a steady state of chlorophyll abundance may have been reached at around 1.53 μg mg-1 of fresh weight. This is lower than that seen in mature leaf material, which showed total chlorophyll content of 2.25 μg mg-1 of fresh weight, although this value is comparable to that reported for C. gynandra previously (Kulya et al., 2011). A diverse range of species reach maxima of chlorophyll production around the 7–8 day point of cotyledon development (Moore et al., 1972) so this situation in C. gynandra is not without precedent, although the reduced abundance of chlorophyll may depress the photosynthetic capacity of the cotyledons relative to the mature plant. 4.3.2: Light regulates Components of the C4 cycle in C. gynandra The exposure of etiolated cotyledons of C. gynandra to light triggers significant changes in gene expression. CA4, NAD-ME2, LHCA1 show a rapid increase in transcript abundance within 2 hours of illumination, reaching or exceeding the level of accumulation seen in light-grown cotyledons (Figure 4.4). LHCA1 shows a more significant increase than the other genes examined in this study, but given its relative absence in etiolated cotyledons this may be a proportionate response. The absence of LHCA1 from etiolated cotyledons corresponds with the absence of chlorophyll from these tissues, and so it is likely that the upregulation the light- harvesting complexes (LHC) is concurrent with the increased abundance of chlorophyll although the changes in LHC abundance may be triggered by changes in redox potential rather than light sensitivity per se (Maxwell et al., 1995). An up-regulation of LHCA1 occurs within the first two hours after light exposure, while an increase in chlorophyll abundance was not detectable until four hours after exposure. This would suggest that the regulation of LHCA1 is not linked to changes in redox potential due to the lack of functioning chlorophyll during the initial period of its upregulation. A general increase in transcript accumulation is seen, although the majority of transcripts profiled in this study appear to have reached a plateau of expression within 6 hours of light exposure. However, only RbcS, LHCA1 and PSBQ2 are at levels equal to or lower than the light-grown samples which may indicate that the elevated level of expression is transient, and will be subject to a decline once the cotyledons become photosynthetically active. All the genes examined by this study show a measure of accumulation at the transcript level before illumination, with the exception of LHCA1. A general increase in abundance is seen after illumination, which indicates that light is required to induce expression to levels seen in Chapter 4: Changes in Gene Regulation during Cotyledon Development 98 photosynthetically active tissue. The exception to this is PSBQ which appears to be relatively constant, albeit it at a lower level than light grown material. In addition, RbcS shows a sharp decrease in abundance within the first 2 hours of illumination, which may suggest that light has a suppressive effect on the accumulation of its transcripts, potentially similar to that which has been described in maize (Langdale et al., 1988) 4.3.3: Circadian Regulation may be active in C. gynandra Interestingly, a general decline was seen in the levels of gene expression seen in light-grown cotyledons over the course the experiment (Figure 4.4). As t=0 occurred between 9 and 10am in the morning, this may correspond to a decline in expression linked to a pre-dawn or dawn- promoted upregulation of expression. Circadian regulation has been shown to have a significant impact on the expression of genes associated with photosynthesis such as the production of chlorophyll and LHC-binding proteins (Gehring et al., 1977; Paulsen & Bogorad, 1988). The decline in transcription levels observed here may be linked to circadian rhythms promoting peak expression at dawn with a subsequent decline over time. While light-grown seedlings were subject to a long-day photoperiod, etiolated seedlings were kept in constant temperature and darkness from the point of germination so that a circadian rhythm may not have been entrained, although circadian rhythms have been described in Arabidopsis seedlings deprived of both light and temperature stimuli (Salomé et al., 2008). A significant proportion of genes may be placed under circadian control (Michael & McClung, 2003), and a number of key C4 genes in maize follow a circadian rhythm which peaks around dawn (Khan et al., 2010). Therefore, it is conceivable that the general decline in transcription seen in this experiment may be linked to intrinsic rhythms. It would be interesting to address whether such non-photosynthesis genes are under circadian control in a C3 environment prior to their recruitment into the C4 cycle as this could represent an additional evolutionary requirement. It is beyond the scope of this study to explore the effects of circadian rhythms upon the C4 cycle. However, should this study be extended, it would be beneficial to examine gene expression levels over a longer time period to properly quantify this decline, and to assess what, if any, impact circadian regulation has upon these genes.. In the context of this exploration, however, the scale of differences in expression between the etiolated and light- grown cotyledons is likely to have exceeded any changes in expression linked to the circadian clock. It would be beneficial to adjust the culturing process so that light grown seedlings were grown under continuous illumination to remove the impact of any periodicity upon expression patterns should this study be extended. Chapter 4: Changes in Gene Regulation during Cotyledon Development 99 Of those genes that show a reduction in accumulation, either initially (e.g. RbcS) or after an initial increase after illumination (e.g. DiT2), a feature which may alter the interpretation of this data is that of RNA stability. Not only do these reductions indicate that transcription has been depressed, but that the stability of these transcripts show half-life values within 2-4 hours based on the reduction in relative abundance implied by these data. The control of mRNA stability has been implicated as a key feature of gene regulation, and it is likely that significant variation will occur between transcripts (Abler & Green, 1996). Transcript stability in C. gynandra has yet to be fully explored, and significant variation exists in published values for Arabidopsis (Gutiérrez et al., 2002; Narsai et al., 2007) so it is not possible to explore this area. Should such data become available, however, it may be possible to test whether the changes in accumulation seen in this study are attributable to alteration in transcript stability either between genes or in response to illumination. 4.3.4: Metabolic Flux may also direct Development in C. gynandra. Of the transporters examined in this study, DiT1/2 and Bass2 show similar up-regulation after illumination. The onset of this increase occurs between 2–4 hours after illumination, representing a delay from that seen in the majority of genes profiled by this study. The data indicates that the regulation of these genes is not linked directly to light exposure (given that their non-C4 roles are unlikely to be associated with photosynthetic activities), and so it is conceivable that their regulation is targeted more towards the concentrations of metabolites with which they interact (Bräutigam et al., 2008). A role for metabolite-led development has been described in Amaranthus as the C3 to C4 switch corresponds with a gain of net productivity (Wang et al., 1993b). Therefore, these may indicate that metabolite flux may have an impact on regulation of the onset of the C4 cycle in C. gynandra, but testing this is far beyond the scope of this study. 4.3.5: Protein Accumulation in Etiolated Cotyledons While this study has highlighted that an upregulation of many of the components of the C4 cycle at the transcript level occurs in response to light, the actual cause of this response remains to be examined. Those genes such as LHCA1 which show the most rapid response to illumination may be attributable to the direct effect of illumination, but those that show a more delayed response (e.g. DiT2, PPDK) may be subject to additional regulation based on metabolic flux, or the activity of intermediate regulators. At the protein level, expression patterns largely mirror those seen at the transcript level. PPDK, PEPC and NAD-ME are detectable in etiolated cotyledons, but at levels significantly Chapter 4: Changes in Gene Regulation during Cotyledon Development 100 depressed from that of light-grown seedlings. The difference in abundance of PEPC between the treatments is less than that seen in NAD-ME and PPDK, which reflects the lower levels of transcript accumulation seen at t=0 in the latter two instances. The presence of NAD-ME, PPDK and PEPC in etiolated cotyledons would suggest that light is not required for their expression, but that light is required to achieve high levels of accumulation. This mirrors the situation in Amaranthus, where both PPDK and PEPC show localisation to the mesophyll in etiolated cotyledons (Wang et al., 1993a). However, a lack of correlation between transcript and protein abundance appears to be present in the expression of RbcS, which is also the only gene examined here which shows a reduction within the first two hours of light exposure. Levels drop from about 1.5x that of light-grown cotyledons to around 0.5x, before returning to c.1x after six hours. This may indicate that the steady-state level of RbcS expression is elevated in darkness, and a light- mediated mechanism reduces expression of RbcS upon illumination. At the protein level, however, the SSU appears to increase in abundance after illumination, which would indicate an incongruity between transcript and protein accumulation in darkness. This would mean that if increased levels of RbcS transcripts are present during etiolation, then a secondary level of regulation exists which prevents increased abundance at the protein level. In Amaranthus, both subunits show the correct patterning in darkness (albeit at greatly reduced levels), although multiple levels of regulation via transcription and translation have been reported (Wang et al., 1993a; McCormac et al., 2000). For the data presented here is not possible to define whether the changes in abundance of either protein or RNA are concurrent with a change in cell-specificity, but this would be worthwhile exploring further. A key extension of this study would be to develop in situ localisation of RNA and protein accumulation. Primarily, this would provide better quantification of any changes in gene expression patterns. The data utilised to date has been derived from whole-cotyledon extracts so that it has not been possible to explore changes in the BS or M cells in isolation. In situ localisations would enable visualisation of any changes that occur in either protein or transcript abundance in a cell type-specific fashion, improving the clarity of the examination. Additionally, it would enable assessment to be carried out on a cotyledon-specific basis. While lower levels of PPDK and PEPC were recorded in etiolated plants, the up-regulation of both genes may be independent of the acquisition of cell specificity. In Amaranthus, both show the correct cell patterning from the onset of cotyledon development in both light and dark (Wang et al., 1993a). It could be argued that as both genes have been recruited from non- Chapter 4: Changes in Gene Regulation during Cotyledon Development 101 photosynthetic roles, the acquisition of direct light-mediated regulation would require an additional step in the recruitment process. However, PPDK expression is induced by light in maize, an effect attributable to cis-acting elements in its promoter (Matsuoka et al., 1993). In addition, both enzymes are subject to light-mediated post-translation regulation (Vidal & Chollet, 1997; Astley et al., 2011), so it is likely that additional transcriptional or post- transcriptional controls are in place which are directly regulated by illumination. Finally, the development of C. gynandra lines expressing an inhibitor of sRNA synthesis may provide an opportunity for the study of the effects of sRNA/miRNA activity during the de-etiolation process. miRNAs have been implicated in light sensitivity in roots (Sorin et al., 2005), but whether this extends to light-mediated physiological changes in foliar tissue remains to be explored. Replicating this experiment using cotyledons expressing the p19 inhibitor that was transformed into C. gynandra as part of this would enable testing of a potential miRNA- mediated role in the decline in transcript accumulation that has been demonstrated in this study. 4.3.6: Concluding Remarks This study has confirmed the presence of Kranz anatomy in the cotyledons of C. gynandra, and that light is not required for its development. Genes associated with light harvesting, and the formation of chlorophyll, are only demonstrated after illumination of etiolated cotyledons. The other components of the C4 cycle are present at the transcript level prior to illumination, although a measure of upregulation occurs after light exposure indicating that environmental stimuli are required if accumulation at the transcript level is to reach levels comparable with that seen in light grown cotyledons. Accumulation of PEPC, PPDK, NAD-ME polypeptides match the increase in accumulation after illumination seen at the transcript level. A decrease in the accumulation of RbcS transcripts is seen within the first two hours of light exposure, however, although levels of the SSU remain lower than that seen in light-grown cotyledons. This may indicate that a secondary level of regulation is in effect which uncouples the levels of transcript and peptide accumulation of SSU which is relieved upon illumination. While the data presented here are insufficient to fully test the effects of illumination upon cell- specificity, a number of areas have been highlighted that may benefit from further exploration. 102 Chapter 5: The Generation of Hybrids within Cleome 103 5. The Generation of Hybrids within Cleome 5.1: Introduction 5.1.1: The Assessment of Hybrids between Photosynthetic Types The occurrence of C3, C4 and C3/4 intermediates within close phylogenetic groups offers an opportunity to examine the C4 cycle through hybridisation studies. The generation of hybrids could allow the assessment of C4 traits in a C3 environment and vice versa in order to be able to identify the regulatory processes that have been modified during the evolution of a given trait into the C4 cycle. While this experimental approach has been used on a gene-specific basis (e.g. Brown et al., 2011) this can be a slow and laborious process, and is not applicable to traits for which no genetic basis has been identified. The generation of an artificial C3 x C4 hybrid lineage would enable the identification of loci associated with the C4 cycle via mapping of an introgression line, a method which has been previously used to identify features of complex polygenic traits (e.g. Eshed & Zamir, 1995). Analysis of hybrids would also enable the identification of linked traits within the C4 cycle which may indicate a shared regulatory or evolutionary process, as well as assessing whether components of the C4 cycle would be dominant or recessive in a C3 leaf. The hybridisation studies published to date tend to show hybrids with reduced vigour and levels of photosynthesis which are depressed towards if not below that of the C3 parent (e.g. Byrd et al., 1992; Pearcy & Björkman, 1970), although C4-like dominance is seen in some single-gene instances (e.g. NADP-malate dehydrogenase, Holaday et al., 1988). The dominance of the C3 cycle may be attributable to C3 representing the “default” state over which the C4 cycle is placed (which would correlate with the delayed onset of C4 photosynthesis during leaf development discussed in Section 1.1.6). The dominance of C3 or C4 traits in the opposing environment may test this, in addition to enabling the identification of trans-acting factors, and so it is considered important to explore this aspect in a wider context. In addition, the presence of heterologous SSU/LSU combinations may enable screening for altered RuBisCO kinetics. C3 and C4 species show a measure of altered catalytic efficiency (Sage, 2001), and so the production of hybrid holoenzymes may enable this aspect to be explored further. It will be necessary to carry out reciprocal crosses to screen for differences in parental dominance. Brown et al. (1992) reported paternal dominance patterns of δ13C values in reciprocal crosses between Flaveria brownii and F. linearis, so production of bidirectional crosses would be required to fully test any hypothesis of dominance. Chapter 5: The Generation of Hybrids within Cleome 104 A number of successful hybridisation attempts have been made between members of Panicum (C3 x C3/C4, Brown et al., 1985; Sternberg et al., 1986), Flaveria (C3 x C3/C4, C3 x C4, Brown et al., 1986, Holaday et al., 1988; Bryd et al., 1992) and Atriplex (Björkman et al., 1970), although these studies were generally restricted to analysis of gas exchange parameters, enzyme activity and anatomical structures with no global assessment of changes in transcript or protein abundance of components of the C4 cycle. To test whether interspecific crosses between C3 and C4 members of Cleome could be generated to address these questions, hybridisations between C. gynandra and C. hassleriana were undertaken. Rajndrudu & Rama’Das (1982) reported successful hybridisation between C. gynandra and C. viscosa L. (C3) using conventional breeding techniques, and while their study was limited to seed coat morphology it does indicate the potential for interspecific crosses within Cleome. C. hassleriana was chosen as the C3 parent in this instance due to the availability of molecular resources for this species (Bräutigam et al., 2011). 5.1.2: Production of Hybrid Lines A number of barriers may be present to prevent the production of hybrids at both the pre- and postzygotic level. At the prezygotic level, heterospecific pollen grains may fail to germinate or may show abnormal growth through the stylar tissue (Chen & Gibson, 1972), preventing fertilisation. After fertilisation the hybrid embryos may abort before full development can be achieved, most likely as the result of a halting of development of the endosperm leading to a cessation of the supply of resources to the developing seed (Briggs et al., 1987; Abbo & Ladizinsky, 1991). Once hybrids are obtained, they may demonstrate sterility or unviability, potentially through an uncoupling of coadapted subspecific gene complexes or negative interactions between divergent loci (Wu & Palopoli, 1994; Li et al., 1997). In addition, successful hybridisation events may be asymmetrical, so that the cross is unidirectional in that viable plants are obtained with a specific maternal/paternal mix (Smith, 1970). This is especially relevant in this situation due to the benefit of carrying out reciprocal crosses. Rajndrudu & Rama’Das (1982) were unsuccessful in obtaining viable seed when C. gynandra was the maternal partner when crossed with C. viscosa, although only a 5.5% success rate was reported in the reciprocal cross. 5.1.3: Experimental Approach In order to minimise the impact of fertility barriers upon the generation of hybrids within Cleome, it was decided to supplement traditional crossing methods with a regime of in vitro cultivation of excised hybrid embryos in order to minimise any negative effects of continued Chapter 5: The Generation of Hybrids within Cleome 105 development in planta. Such methods have improved the recovery of interspecific hybrids in Brassica (Bajaj et al., 1986; Quazi, 1988), and given that C. gynandra (Newell et al., 2010) and C. spinosa (Albarello et al., 2006) have been successfully cultured in vitro it was hoped that such techniques could be adapted to interspecific hybrids within Cleome. In addition to in plant development of embryos, a three tailed approach was undertaken to attempt in vitro culturing of hybrid embryos. Liquid suspension cultures of Quazi (1988), rotating liquid cultures of Harberd (1969) and gel-phase cultures of Zhang et al. (2003) were adapted, with the primary focus of ensuring adequate contact between the medium and embryo. Additionally, a hormone regime matrix was established to identify optimum concentrations to ensure either continued embryo development or the formation of regenerative callus. The crossing methods employed by Rajndrudu & Rama’Das (1982) first involved the emasculation and isolation of flowers prior to anthesis, an approach which is considered valid given the reported occurrence of self-compatibility within Cleome (Cane, 2008). It was also necessary to identify the reproductive morphology of both species as hermaphroditic, andromonoecious (flowers lacking gynoecia) and staminate flowers (flowers which originate as hermaphroditic, but show halted pistal growth) have been reported within Cleome (Cane, 2008; Machado et al., 2006). Successful ovule penetration by the pollen tube was assessed using the methods of Martin (1959) to enable visualisation of tube growth, although this will be used as a tentative indicator of successful fertilization as pollen-tube penetration may not directly correspond to fertilisation success rates (Rieseberg et al., 1995). Reciprocal heterospecific crosses between C. gynandra and C. hassleriana were therefore attempted. In addition to maturation in planta, attempts to recover mature embryos via in vitro embryo cultured were undertaken. The unknown hormonal requirements of C. hassleriana and of the potential hybrids required a hormone matrix to be established to highlight optimum concentrations for embryo maturation. In addition, the potentially conflicting requirements of nutrient and gas exchange requirements were screened through the use of three alternative culture systems. Solid-state cultures were achieved by solidifying the media with 0.9% agar, while liquid-phase culturing was achieved through the direct placement of the excised embryos into agitated liquid medium. Finally, liquid-suspension cultures were achieved through the placement of the excised embryos onto filter paper bridges placed over a vessel of liquid medium, which could then be drawn up to the suspended embryo. The latter two methods were considered to represent the two extremes of nutrient/gas supply, although these were chosen in an attempt to overcome any problems Chapter 5: The Generation of Hybrids within Cleome 106 associated with the unequal distribution of nutrients or by-products within the solid-state medium. 5.2: Results 5.2.1: Reproductive Biology C. gynandra and C. hassleriana both demonstrated the formation of heteromorphic flower types (Figure 5.1). The development of hermaphrodite flowers interspaced with staminate flowers which show aborted gynoecium development at c.3mm of growth was seen in layers progressing up a simple raceme. In both instances the dehiscence of the anthers occurred 2-4 days after anthesis. The petals of C. gynandra were greatly reduced and shortened, although the gynoecium and anthers were borne on an extend androgynophore. Such extension was absent from C. hassleriana, but the extension of both petals, anther filaments and style ensure that the projection of the reproductive organs away from the petals was roughly equal between the two species (5.1-A,B). The anthers and stigma of C. hassleriana remain enclosed within the flower bud until anthesis, although the anther filaments and style curved outwards forming a hook shape prior to anthesis (5.1-B). In C. gynandra, the stigma and anther tips are visible above the edge of the developing petal whorl, resulting in the exposure of the stigma considerably prior to anthesis (5.1-C). When treated with hydrogen peroxide, the stigmatic tissue of C. hassleriana demonstrated receptivity before emergence from the flower bud, and C. gynandra showed receptivity before the extension of the androgynophore away from the petals (Figure 5.2). 5.2.2: Pollen Grain Germination and Seed Set Pollen tube tissue was readily discernible from that of the style under UV light after aniline staining (Figure 5.3). The gynoecia of isolated plants of both species that had received a heterospecific pollen load showed high levels of pollen grain germination. Tube growth showed no abnormal growth through the style, and growth was sufficient to reach the ovules. A subset of plants that had been cross pollinated were left to set seed naturally. While no seed production was seen in maternal C. gynandra plants, maternal C. hassleriana pods developed at a rate comparable with plants which had received a homospecific pollen load, although the number of seeds set was depressed to less than 25% of that seen in WT crosses. When the seeds were harvested, they showed no difference in size from WT seeds to C. hassleriana, although the surface was discoloured, the seed coat was thinned and brittle, and the seeds were incompletely filled. No successful germination occurred from seeds obtained from heterospecific crosses. Chapter 5: The Generation of Hybrids within Cleome 107 Figure 5.1: Sexually dimorphic flowers are present in Cleome. Staminate (i) and hermaphrodite (ii) flowers are produced by both C. gynandra (A) and C. hassleriana (B). The gynoecium of each type is marked with an arrow. The gynoecium of the staminate flowers shows halted development resulting in a reduced ovary and stigma at the base of the anther filaments in both species. The vestigial ovary is identifiable by its retention of green pigmentation while the other floral organs acquire pink/purple pigmentation during maturation. The C. hassleriana flowers are shown prior to anthesis, with the hook of the elongating style evident in the hermaphrodite flower. In the developing flowers of C. gynandra (C), the anthers (Y) and stigma (X) project above the petal whorl prior to anthesis. Scale: 1cm (A,B), 1mm (C). Chapter 5: The Generation of Hybrids within Cleome 108 Figure 5.2: Oxygen evolution at the stigmatic surface of C. gynandra (A) and C. hassleriana (B). Both stigmas were removed shortly before anthesis and submerged in 6% hydrogen peroxide solution. The production of oxygen bubbles was taken as an indicator of stigma receptivity. Scale = 250µm. Figure 5.3: Heterospecific pollen tube growth in C. gynandra (A) and C. hassleriana (B). The growth of the pollen tube (t) from the germinating grains on the stigmatic surface (p) down the style into the ovary (o) is distinguishable from the stylar tissue by staining with aniline blue. No pollen tubes were evident in unfertilised flowers. The gynoecia were flattened between glass slides before examination. Scale: 500µm. Chapter 5: The Generation of Hybrids within Cleome 109 5.2.3: Embryo Extraction and Culture The method of embryo extraction discussed above was sufficient for the establishment of aseptic cultures. Maternal C. gynandra crosses did not produce any viable embryos, although between 5 and 10 embryos per ovary were recoverable from maternal C. hassleriana crosses. The 7-day window after fertilisation was sufficient to produce enlarged embryos that were approaching the globular/heart boundary and were discernible from the unfertilized ovules (Figure 5.4). Figure 5.4: Embryo growth in C. hasseleriana. The ovary was cut down the replum, revealing the extent of ovule growth 7 days after fertilisation. A pair of unfertilised ovules (i) can be identified relative to the developing fertilised embryo (ii). The stigmatic tissue and remaining embryos/ovules were removed prior to photography. Scale = 1mm. The induction of callus was successful in the solid-phase culture, although considerable variation was seen in the extent of callus formation (Figure 5.5). The extent of callus formation at levels of 0.5 mg l-1 NAA was considerably higher over 0 and 0.25 mg l-1 NAA, although a slight increase in callus weight was seen at 0.25 mg l-1 NAA when placed in tandem with increased concentrations of BAP. Increases of callus weight correlated well with BAP concentration, which showed a peak at 1 mg l-1, with NAA concentrations of 0.5 mg l-1 NAA. For the duration of this experiment, no organogenesis was demonstrated, and senescence of the callus occurred sporadically after 16 weeks. Callus formation was only recordable in the solid- state cultures as no growth was recorded and senescence occurred within two weeks of the establishment of the liquid-suspension and liquid phase cultures. Chapter 5: The Generation of Hybrids within Cleome 110 Figure 5.5: Response of embryos cultured in vitro to varying BAP concentrations in tandem with 0, o.25 and 0.5 mg l -1 of NAA. Limited growth was seen in cultures with 0 and 0.25 mg l -1 NAA at all concentrations of BAP. Maximal callus regeneration was seen at concentrations of 1.0 mg l -1 BAP/ 0.5 mg l -1 NAA. 5.3: Discussion 5.3.1: Reproductive Biology of C. gynandra and C. hassleriana The production of heteromorphic flower types by C. gynandra and C. hassleriana match similar reports of other members of the genus (Stout, 1923; Cane, 2008) although the male flowers were found to be staminate (flowers in which the gynoecium halted development) rather than andromonoecious (the gynoecium is absent entirely) which agrees with the observation of Cane (2008). Cane also notes the receptivity of the stigma from the point of anthesis as was found in both C. gynandra and C. spinosa, with the delayed dehiscence of the anthers potentially acting as a method of limiting the extent of selfing. While the stigma of C. hassleriana was receptive prior to anthesis, its encapsulation in the developing bud means that its isolation and cross pollination can be performed with minimal risk of contamination with homospecific pollen. The lack of seed recovered from crosses in which C. gynandra was the maternal partner matches the report of Rajendrudu & Rama’Das (1982) that C. gynandra was incapable of acting as a maternal parent in such crosses. The data presented here suggests that this is not due to a failure of pollen tube penetration. This would infer that additional post-zygotic barriers are in place which halts the development of the embryo after fertilisation, as discussed above. The Chapter 5: The Generation of Hybrids within Cleome 111 failure of C. hassleriana to produce viable seed through normal crossing methods also indicates that post-zygotic barriers are active within Cleome. The embryos produced in the maternal C. hassleriana crosses show viability at the 7 day harvesting threshold, while no such development is present in crosses in which C. gynandra is the maternal partner. This would indicate that abortion of the hybrid embryos occurs early in development before the globule stage has been passed, as has been reported in other interspecific crosses (e.g. Barone et al., 1992). Figure 5.6: Developmental patterning of the fertilised zygote. Once the octant stage has passed into the global stage, enlargement of the cotyledon precursor cells shows passage into the heart stage, before further development leads to torpedo formation. (After Steeves & Sussex, 1989) 5.3.2: Receptivity of Hybrid Material to In Vitro Culture The rapid senescence of excised embryos placed in either form of liquid culture limited the extent to which this experimental system could be evaluated. The success of the solid-state culture medium in obtaining callus may indicate the negative effects of liquid culture are due to physical parameters rather than biochemical ones. Due to the nature of liquid culture, it was not possible to ensure the embryos remained on the surface of the liquid medium, and such submersion is likely to have significant impact on the gas exchange properties of the embryos. The relative quantity of atmosphere within the culture vessel, and especially oxygen abundance, has been shown to be significant in promoting embryo growth in other systems (Dunwell, 1979) and so it is feasible that a lack of aeration prevented development in vitro. If gas exchange limits were a feature in the failure of the liquid-based cultures, then the opposing situation may be seen in the liquid-suspension cultures. The lack of growth in the liquid-suspension cultures is considered most likely to be attributable to a lack of nutrients resulting from poor contact with the culture media. Cultivation in the solid-state media enabled the embryos to be partiality imbedded in the media, which may have allowed for sufficient exposure to the culture vessel atmosphere while ensuring good contact with the nutrient source. Chapter 5: The Generation of Hybrids within Cleome 112 An aspect that was difficult to examine using embryo material as explant matter was the effect of the osmotic potential of the media. Variation in the osmotic potential of the media due to changes in the nature or abundance of the carbohydrate source (Biahoua & Bonneau, 1998) or alternative osmoticum (e.g. sorbitol, Finkelstein & Crouch, 1986) may help to promote embryo development in both the liquid and solid-phase culture. Finally, the choice of synthetic hormone analogue may have an impact on the extent of callus development due to a variety interactions between the carbohydrate source, hormone analogue and sterilisation method which can occur (Mendoza & Kaeppler, 2002). As such, a considerable number of aspects may have an impact on the level of embryo development in vitro. The diversity of these factors, and the large number of possible interactions between them limited the extent to which they could be examined in this experiment. While it would be advisable to explore these aspects further should this study be extended, it may be advisable to optimise the culture process using fertilised embryos obtained from homospecific crosses in order to ensure that the failure of the culturing system is not due to base genetic incompatibility present in the hybrid material. The key variable that has been assessed by this experiment is the effect of varying hormone regimes upon the recovery of viable material. Optimal callus recovery was seen at 0.25 mg l-1 NAA and 1.5 mg l-1 of BAP. This range is significantly higher than the concentrations required for the propagation of C. gynandra described by Newell et al. (2010), and concentrations that have shown to be effective in Brassica embryo culture. The ranges used within this experiment are comparable to those used by Bajaj et al. (1986) and the range found optimum by Zhang et al. (2003) reported peak rates of callus production 0.3 mg l-1 NAA and 2 mg l-1 of BAP. This concentration of BAP exceeds that which was found to be optimum in this experiment for callus growth in Cleome hybrids, although the NAA concentration is comparable. Quazi (1988) reported successful maturation and germination of hybrid Brassica embryos in vitro on a relatively simple media lacking hormone supplements, although the embryos were harvested 22 days after pollination. The results of this experiment indicate that embryo development does not continue in the absence of supplemental hormones. It may be the case that the embryos were harvested at a stage where there is still some dependence on maternally-derived signals such as auxin (Weijers & Jürgens, 2004) although it would be hoped that any such signal could be replaced by supplemental hormone addition to the culture media. Liu et al. (1993) reported successful culture of proembryos that were extracted at the 8–36 cell stage and in vitro fertilisation of excised ovules has been effective in Chapter 5: The Generation of Hybrids within Cleome 113 a number of taxa, including Brassica (Zenkteler, 1990). Stewart & Hsu (1978) report increased cotton embryo recovery after 22 days onto media supplemented with auxin, although it is noted the external application of hormones may predispose the embryos to form callus rather than develop normally. If this were to be the case in Cleome hybrids then the level of hormonal supplementation required to promote embryo development would coincide with that required for callus induction, so the organogenesis step may be inevitable. The production of potentially hybrid callus as demonstrated by this experiment may still prove informative if organogenesis can be induced, despite the complications of achieving whole plant regeneration from in vitro cultures of C. gynandra (see section 3.3.3.5). If this work is continued further, the hormone regime should be more closely explored to identify optimal concentrations to promote embryo development, callus production and promote organogenesis to increase the probability of hybrid recovery. 5.3.3: Genetic Incompatibility within Cleome Difficulties in obtaining viable hybrids from C. gynandra x C. hasseleriana crosses may be due to genetic incompatibility rather than experimental methods. Rajndrudu & Rama’Das (1982) reported successful hybridisation between C. gynandra and C. viscosa although the genetic distance between these two taxa is relatively limited according to the phylogeny of Inda et al. (2008). In contrast, C. hasseleriana is placed at considerable genetic distance from both C. gynandra and C. viscosa. The increased divergence between C. gynandra and C. hasseleriana relative to C. viscosa may have facilitated a higher level of evolution of heterologous loci which negatively interact in the hybrid environment leading to detrimental heterosis effects (Orr & Turelli, 2001). The application of such theory (as described by the Dobzhansky-Muller model) has been described in other plant systems (e.g. Tomato, Moyle & Nakazato, 2010) so it is conceivable that such barriers may exist within Cleome. In addition to genetic incompatibility, effects may also be seen due to differences in karyology. C. gynandra shows a base chromosome number (x = 16 or 17) that differs from both C. viscosa (x = 10) and C. hasseleriana (x = 10), although the latter taxon is tetraploid (Ruiz-Zapata et al., 1996). It is possible that chromosomal variation prevents hybridisation, as diploid x tetraploid crosses have been reported to show lower than expected progeny yields in other plant systems (Griffiths et al., 1971). A possible method of overcoming both genetic incompatibility and problems associated with the in vitro culture of excised embryos may be the generation of somatic hybrids. The production of somatic hybrids via the fusion of isolated protoplasts has been demonstrated Chapter 5: The Generation of Hybrids within Cleome 114 between members of Brassicaceae (e.g. Forsberg et al., 1994; Terada et al., 1987). The utilisation of this method would have the additional benefit of producing “cybrids” containing a mixed parental inheritance of organelles that would not have been possible in sexual crosses. Utilisation of such methods would require an effective protocol for the isolation and regeneration of protoplasts from Cleome, and this has not been explored experimentally to date. Additionally, the inability to recover root tissue would place a bottleneck on the recovery of hybrid material as discussed previously. Therefore, this method would require considerable experimental exploration before it could be practically utilised. This experiment has illustrated that it is possible to obtain potentially hybrid callus material from crosses between C. gynandra and C. hassleriana in which the latter taxon is used at the maternal parent. The inability to obtain viable embryos from the reciprocal cross matches that which has been reported previously (e.g Smith, 1970) and may represent asymmetrical hybridisation potential between these two taxa. Genetic and chromosomal aspects of the cross may act as incompatibility barriers between the cross, although such barriers may be overcome through in vitro culture. While normal embryonic development in vitro has been not been achieved, the production of callus has been demonstrated with a potential for the recovery of hybrid individuals should organogenesis be realised through further exploration of the effects of supplementary hormone control. Chapter 6: Final Discussion 115 6. Final Discussion 6.1: The potential for miRNAs to regulate the C4 Cycle This study has explored the roles that sRNA may play in regulating photosynthesis in C. gynandra. Since their original description as a mechanism of viral defence (Hamilton & Baulcombe, 1999), miRNAs have been implicated in an ever-increasing number of regulatory roles including leaf development (e.g. Kidner & Martienssen, 2004; Chuck et al., 2007), phase change (e.g. Aukerman & Sakai 2003) and responses to plant growth regulators (e.g. Reyes & Chua 2007). Besides developmental roles, miRNAs have been described in regulating aspects of plant metabolism, notably in response to abiotic stress (e.g. Lu et al., 2005; Ding et al., 2009) and nutrient availability (Beauclair et al., 2010). A role for miRNA activity in the C4 cycle was initially indicated by the disproportionate alignment of sRNA sequences to components of the C4 cycle in maize (Tolley et al., unpublished). While no previous role for sRNA activity in the regulation of photosynthesis has been published, computational prediction of miRNAs in Glycine max has identified miRNAs which may target proteins associated with photosynthesis (Katara et al., 2010) and a broader role for miRNA activity in plants has been suggested given the extent to which such regulatory processes are active in animals (Rhoades et al., 2002). miRNAs are attractive as potential regulatory elements of the C4 cycle for a number of reasons. Firstly, miRNAs have been shown to function in a cell-specific fashion (Parizotto et al., 2004) so that BS/M differentiation could enable a unique miRNA population to be present in each cell type, allowing for selective inhibition of a target gene in a cell-specific fashion. Secondly, as trans-acting factors a variety of separate evolutionary pathways would enable cell-type specific gene expression to be achieved without direct modification of the transcriptional regulation of the target gene itself. Finally, miRNAs appear to be subject to rapid rates of evolutionary turnover, and their limited size may mean that significant changes in their target range can be caused by a limited number of base substitutions (Fahlgren et al., 2007). Given these considerations, an exploration of the roles that sRNAs play in regulation in C. gynandra was considered justified, and the results of this study support the hypothesis that sRNAs may regulate the C4 cycle in C. gynandra. 6.2: miRNAs actively regulate Development in C. gynandra The expression of the viral p19 inhibitor of miRNA biosynthesis in transgenic Arabidopsis and C. gynandra lines has highlighted the extent to which miRNAs are involved in regulating leaf development in both species. In Arabidopsis, expression of p19 lead to reduced vigour, the formation of leaves with increased serration of the leaf margin and modified inflorescence Chapter 6: Final Discussion 116 structure (Figure 3.8 & 3.9), the extent of which correlated with the abundance of p19 protein which varied between lines (Figure 3.10). The severity of this phenotype is considered to be the maximum obtainable in unsupported plants grown on soil. In contrast, C. gynandra transgenic lines demonstrating a more advanced foliar phenotype were obtained by growth in vitro when supported by exogenous supplies of nutrients and growth regulators. These cultures demonstrated the formation of radialised leaves with little to no mesophyll development (Figure 3.12). These effects are consistent with previously reported phenotypes where miRNA activity has been disrupted either by the presence of viral inhibitors (Dunoyer et al., 2004; Chapman et al., 2004) or a loss-of-function mutation of miRNA-encoding genes (Palatnik et al., 2003). The same phenotype was found in Arabidopsis when p19 was placed under inducible control, although a phenotype was not recovered in the C. gynandra transgenic line. The accumulation of p19 transcripts in these plants was considerably less than the relative abundance of p19 in the Arabidopsis lines and so the extent of p19 expression may have been insufficient to trigger the foliar phenotype even though p19 expression in the developing primorida was inferred from GUS accumulation patterns (Figure 3.19). From these data it can be inferred that miRNA-led regulatory pathways are active in C. gynandra that are comparable to those present in Arabidopsis, and the utilisation of the p19 protein as a method of studying miRNA-mediated effects is effective. 6.3: sRNAs may act as Regulators of the C4 Cycle in C. gynandra 6.3.1: Computational Prediction of miRNAs in C. gynandra A key stage in the assessment of potential sRNA regulators of the C4 cycle was the development of a library of sRNA sequences present in mature photosynthetic tissue, and the compilation of a sequence library has enabled the core components of the C4 cycle to be subject to miRNA prediction. While the vast proportion of candidate sequences showed little to no alignment to sRNA reads, a limited subset of candidates showed above-average alignment counts (NAD-ME1&2, PPDK, RCA, MDH and CA4). When the alignment of mismatched sequences against these transcripts was explored, it was found that the distribution of hits along each sequence was not equal, and that sRNA reads aligned against a limited number of loci. These positions tended to exceed the 21-24nt size window for miRNA functionality, and the presence of a conserved sequence from previously described miRNAs was not identified. Therefore, while a specific miRNA sequence has not been predicted, the unequal distribution of sRNA reads aligning against C4 genes reported in maize (Tolley et al., unpublished) has also been demonstrated in C. gynandra. Chapter 6: Final Discussion 117 The full utilisation of this resource is not yet possible, however, and this limits the scope of this study. Traditional miRNA prediction methods are dependent upon an annotated target genome (Rhoades et al., 2002) or candidate EST sequence lists (Prabu et al., 2010), the former of which is unavailable for C. gynandra. Therefore, direct prediction of targets or precursor miRNA coding sequences is not possible, and the incomplete genetic blueprint for the C4 system means that a full exploration of potential candidates is not possible with extant resources. The analysis that is possible, however, supports the suggestion that sRNAs exert a regulatory influence on the C4 cycle. A primary extension would be to obtain similar sequence data for the C3 C. hassleriana. The availability of an annoted EST database for this species (Bräutigam et al., 2011) would enable a replication of the analysis performed on the C. gynandra sRNA library as well as direct comparison of sRNA sequences. Not only would this enable identification of sRNA sequences which are over-represented in C. gynandra, but this may facilitate the detection of novel miRNA sequences which do not any show conserved sequence identity with Arabidopsis. Secondly, a more comparative approach may enable the identification of miRNA targets which play a role in the C4 cycle but which have hitherto unidentified roles. The foliar tissue used to construct the library was taken from fully expanded and photosynthetically active leaves (net CO2 fixation was demonstrable) so any regulatory miRNAs targeting components of the C4 cycle could be expected to be present in the subsequent RNA extracts. An additional hypothesis that was to be tested by this experiment was that miRNAs play a role in regulating the development of Kranz anatomy early in leaf development. This, however, cannot be tested using the library compiled for this experiment as the genes underlying Kranz anatomy have not been identified in C. gynandra. In addition to anatomical development, C. gynandra shows some measure of age-related variation in the extent and nature of C4 cycle, especially with regards to the decarboxylation steps (Sommer et al., 2012). Therefore, a key extension would be a compilation of similar sRNA libraries along a developmental gradient from prior to the onset of Kranz formation to the point of leaf maturity. This would enable the identification of sRNA sequences which showed a high level of abundance before or during the development of Kranz anatomy that may serve as potential regulators of the maturation process. Target prediction using such libraries would be limited by the lack of known target sequences, and so a second extension would be necessary to include data from a C3 relative of C. gynandra, most likely C. hassleriana. The availability of data relating to sRNA reads present in both a C3 and a C4 species along a developmental gradient would enable the identification of sequences which Chapter 6: Final Discussion 118 show altered abundance during the development of Kranz anatomy relative to the maturation of a C3 leaf. 6.3.2: The Efficacy of p19 as a System to test miRNA Activity Besides direct target prediction, this study has validated the feasibility of utilising the viral p19 inhibitor of miRNA biosynthesis to examine the roles of miRNA action in C. gynandra. Constitutive expression of p19 in situ lead to significant morphological changes, reducing the leaves to a radialised structure with little to no mesophyll development, a finding which is concurrent with previously reported effects of miRNA-related disruption during leaf formation (Palatnik et al., 2003). This confirms that the developmental control of leaf formation in C. gynandra follows parallel routes to that seen in Arabidopsis. The utilisation of the pOp promoter enabled p19 expression to be examined in developed tissue, and while p19 expression was implied in developing leaf primordia (as inferred by GUS activity) no change in leaf development was observed. This may suggest that inhibition of sRNA activity does not impact the formation of Kranz anatomy. However, a dose-dependent response was evident in the effects of p19 expression in Arabidopsis, and the relative levels of p19 accumulation in the C. gynandra transgenics are below the threshold where developmental effects were seen in transgenic Arabidopsis lines developed for this study. This issue should be addressed by the establishment of additional transgenic lines of C. gynandra containing the pOp::p19 construct which would show elevated levels of p19 expression to better quantify the effects of p19 expression upon leaf development. Interestingly, the dose-dependent response of transgenic Arabidopsis lines to p19 expression demonstrated by this study contrasts with that reported by previous studies, which reported a consistent phenotype despite varied p19 expression (Dunoyer et al., 2004). 6.3.3: The Effects of p19 Expression in C. gynandra The key physiological effect of p19 expression demonstrated by this experiment was the depression in carboxylation efficiency (CE) and increase in the level of dark respiration in p19- expressing individuals relative to WT plants. The exact cause of the decrease in CE and increase of Γ is unclear from the data presented here. If the changes in these parameters are attributable to a relaxation of negative regulation, then the effects could be attributable to the increased activity of a component of the pathway: such an increase may indicate a loss of cell- type specificity leading to an overall reduction in efficiency. The effects of overexpression of components of the C4 system have yet to be fully explored, although hybrid systems have been useful in exploring the effects of destabilised C4-expression profiles. Chapter 6: Final Discussion 119 Hybrids that were generated between various species of Flaveria showed an increase in the Γ values, and while this did not fully return to that of the C3 parent, it represented a substantial increase from that of the C4 or C4-like parent (Byrd et al., 1992; Holaday et al., 1988). A similar increase in Γ, and a marked decrease in CE was seen in C3xC4 hybrids within Atriplex, to the extent that some hybrid lines showed photosynthetic characteristics which were reduced below that seen in either parent (Björkman et al., 1970; Pearcy et al., 1970). While the latter authors were unable to attribute the depressed CE to the activity of a single component of the C4 cycle, it was felt the mixed genetic environment of the C3xC4 hybrid may lead to a combination of gene expression profiles that have a negative net effect on photosynthesis. Given the frequency of such incidences among C3xC4 hybrids (see Chapter 5), and the high level of orchestration that is believed to be required for efficient operation of the C4 cycle (Sage, 2004) this is considered to be the most likely hypothesis. Given these effects, it is unfortunate that the complete regeneration of hybrids from the C. gynandra x C. hasseleriana crosses was not achieved by this study, although the validity of this approach has been confirmed and it would be advisable to optimise this process should the study be continued. No changes in gene accumulation capable of causing the changes in Γ or CE have been recovered by this study, however. An increase in the relative abundance of RbcS and RCA transcripts was recordable, although levels of SSU polypeptides appeared to remain constant. Therefore, it is considered that the changes in photosynthesis are either attributable to changes in the activity of the components of the pathway (as manifested by changes in abundance of a trans-regulatory factor not profiled in this study) or to changes in the abundance of a component of the pathway hitherto unquantified. Therefore, an essential extension of this study would be to identify the reason for the increase in dark respiration in order to address the origin of these changes as discussed previously. If the effects of non- photorespiratory CO2 production can be discounted, then the direct effects of photosynthesis can be explored. Given that no change in protein abundance sufficient to cause these effects is present in the data presented here, it is considered that the changes in photosynthetic capacity are due to either changes in a component not profiled by this study, or to changes in the activity of the existing components of the cycle. Numerous components of the C4 cycle have been described as being subject to significant post-translational modification as a regulatory method (e.g. PPDK, Astley et al., 2011). While the known regulatory proteins were included in this analysis, it is possible that additional mechanisms of transient regulation of enzyme activity have been affected by p19 expression. Hypothetically, a relaxation of miRNA- mediated negative regulation of a secondary regulatory protein may lead to increased activity Chapter 6: Final Discussion 120 of its target which in turn may lead to the changes in photosynthesis reported in this study. This would require significant experimentation to validate this hypothesis, however, and this study has highlighted the areas which would most benefit for further exploration in order to fully explore this possibility. 6.4: Light Influences Development in C. gynandra Besides the effects of p19 expression, an increased accumulation of RbcS transcripts was also demonstrated in etiolated cotyledons of C. gynandra (Figure 4.5). In Amaranthus, RbcS transcripts do not associate ribosomes in darkness, and light-mediated increases in the SSU is attributable to a relaxation of translational inhibition rather than an increase in transcription rates (Berry et al., 1986; 1990), a light exerts a suppressive effect on RuBisCO expression in maize (Langdale et al., 1988). Therefore, the reduced abundance of the SSU in etiolated cotyledons of C. gynandra relative to transcript abundance correlates with this observation, although the light-mediated nature of this response means that is unlikely to be the cause of the discrepancy between transcript and peptide accumulation in the p19-expressing individuals. Secondly, it does not sufficiently explain the decline in RbcS transcripts that occurred within the first two hours of illumination. It is possible that the decline in transcript abundance indicates an acquisition of cell-specificity of RbcS (due to the exclusion of RbcS transcripts from the mesophyll). However, the developmentally-linked control of RbcS accumulation seen in Amaranthus (Wang et al., 1992; 1993a) suggests that such regulation is most probable in C. gynandra, and that cell-specificity is likely to have already been in place prior to illumination. Additionally, a collapse of cell-specificity of RbcS transcripts does not lead to a corresponding collapse of BS-specific SSU accumulation in A. tricolor (McCormac et al., 1997) so it is likely that RbcS is subject to multiple levels of regulation. The elevated levels of RbcS accumulation in p19-expressing plants appears to be similar to the increased abundance of RbcS in etiolated cotyledons, although again this does not appear to persist to the protein level. The rate of decline seen in RbcS transcripts after illumination would infer a transcript half-life for RbcS that would be in the order of two hours. While transcript stability data is not available for C. gynandra, this would be considerably faster than values reported for other transcripts in Arabidopsis (Gutiérrrez et al., 2002; Narsai et al., 2007), inferring that negative regulation of transcript abundance would be required if this decline were attributable to illumination alone. It would be attractive to speculate that such negative regulation could take the form of miRNA-mediated decay, but the data presented here is not sufficient to address Chapter 6: Final Discussion 121 this hypothesis. While a number of sRNA reads aligned against RbcS, no underlying miRNA sequence was identifiable on the basis of conserved sequence or adherence to accepted targeting rules, and the frequency of aligned reads was considerably below that seen in other candidate genes profiled. However, this hypothesis might be supported by the increased abundance of RbcS transcripts after induction of p19 which may indicate a relaxation of active negative suppression beyond the transcriptional level. If these two conditions are linked, then it could be hypothesised that a light-induced miRNA may exert some regulatory effect upon RbcS. While a direct miRNA response to light has not been fully established, UV-B radiation has shown to modulate miRNA expression (Zhou et al., 2007), and a role of miRNA activity in regulating adventitious rooting via influence on light-dependant auxin homeostasis (Sorin et al., 2005). This is largely speculation, however, and it would be necessary to undergo significant experimentation to justify this. As noted previously, an ability to perform in situ RNA hybridisations and immunolocalisations would enable a specific gain or collapse of cell- specificity in response to illumination or expression of p19 to be identified. In addition, the repetition of the light-induction experiments with cotyledons of transgenic lines expressing p19 would enable a further examination of the decline in RbcS accumulation after illumination. 6.5: Final Conclusions Overall, this has study has demonstrated the depth of possible regulatory mechanisms that may be active in C. gynandra. The disproportionate alignment of sRNA sequences against C4 genes in maize has also been identified in C. gynandra. Not only were high numbers of sRNA reads aligning against a limit subset of candidate genes but the distribution of these reads was not equal. No specific miRNA sequence was isolated, but alignments produced by this study may contain underlying miRNA sequences which could be identified with further study. A role of miRNA activity in leaf formation of C. gynandra is also supported by the significant phenotype caused by the expression of the p19 miRNA biosynthesis inhibitor. The severity of this phenotype required the use of a transient expression system so that the effects of p19 on a pre-existing functional C4 system could be examined. The effectiveness of chemically inducing p19 expression was demonstrated, although the low levels of p19 expression seen under these conditions prevented a full exploration of any potential effects of an inhibition of miRNA activity on the formation of Kranz anatomy. When p19 is transiently expressed in C. gynandra, carboxylation efficiency was reduced while the extent dark respiration increased which would indicate the photosynthesis in C. gynandra Chapter 6: Final Discussion 122 is comprised when miRNA inactivity is reduced. While some changes were detected in transcript abundance these did not persist to the protein level, indicating that the effects of p19 upon photosynthesis are due either to a secondary cause not profiled in this study, or due to altered activity of the existing components of the C4 system. RbcS appears to be negatively regulated at a number of levels due to a discrepancy between transcript and protein accumulation seen in both the p19-expressing systems and in etiolated cotyledons, although it is not possible to identify the nature of this regulation based upon the data obtained in this study. 123 7. References Abbo, S., Ladizinsky, G. (1991) Anatomical aspects of hybrid embryo abortion in the genus Lens L. Botanical Gazette 152(3): 316–320. Abler, M.L., Green, P.J. (1996) Control of mRNA stability in higher plants. Plant Molecular Biology 32(1– 2): 63–78. Akyildiz, M., Gowik, U., Engelmann, S., Koczor, M., Streubel, M., Westhoff, P. (2007) Evolution and function of a cis-regulatory module for mesophyll-specific gene expression in the C4 dicot Flaveria trinervia. The Plant Cell 19(11): 3391–402. Albarello, N., Simões, C., Rosas, P.F.G., Castro, T.C., Gianfaldoni, M.G., Callado, C.H., Mansur, E. (2006) In vitro propagation of Cleome spinosa (Capparaceae) using explants from nursery-grown seedlings and axenic plants. In Vitro Cellular & Developmental Biology - Plant 42(6): 601–606. Alt-Mörbe, J., Kühlmann H., Schröder, J. (1989) Differences in induction of Ti plasmid virulence genes virG and virD, and continued control of virD expression by four external factors. Molecular Plant- Microbe Interactions 2(6):301–308. Ambros, V. (2004) The functions of animal microRNAs. Nature 431: 350–355. An, G., Ebert, P.R., Mitra, A., Ha, S.B. (1988) Binary vectors. In: S.B. Gelvin and R.A. Schilperoort (Eds.), Plant Molecular Biology Manual, Kluwer Academic Publishers, Dordrecht, Netherlands, pp. A3: 1–19. Andrews, T.J., Lorimer, G.H. (1987) RuBisCO: structure, mechanisms and prospects for improvement. The Biochemistry of Plants 10: 131–218. Aoyagi, K., Bassham, J. A., Greene, F.C. (1984) Pyruvate orthophosphate dikinase gene expression in developing wheat seeds. Plant Physiology 75(2): 393–6. Astley, H.M., Parlsey, K., Aubry, S., Chastain, C.J., Burnell, J.N., Webb, M.E., Hibberd, J.M. (2011) The pyruvate, orthophosphate dikinase regulatory proteins of Arabidopsis are both bifunctional and interact with the catalytic and nucleotide-binding domains of pyruvate, orthophosphate dikinase. The Plant Journal 68(6): 1070–1080. Aukerman, M.J, Sakai, H. (2003) Regulation of flowering time and floral organ identity by a MicroRNA and its APETALA2-like target genes. The Plant cell 15(11): 2730–41. Bajaj, Y.P.S., Mahajan, S.K., Labana, K.S. (1986) Interspecific hybridization of Brassica napus and B. juncea through ovary, ovule and embryo culture. Euphytica 35(1): 103–109. Bao, N., Lye, K., Barton, M.K. (2004) MicroRNA binding sites in Arabidopsis class III HD-ZIP mRNAs are required for methylation of the template chromosome. Developmental Cell 7(5): 653–662. Barone, A., Giudice, A., Ng, N.Q. (1992) Barriers to interspecific hybridization between Vigna unguiculata and Vigna vexillata. Sexual Plant Reproduction 5(3): 195–200. Bartel, D.P., Lee, R., Feinbaum, R. (2004) MicroRNAs: Genomics, Biogenesis, Mechanism, and Function Genomics. Genes 116: 281–297. Baskin, J.M., Baskin, C.C. (1979) Promotion of germination of Stellaria media seeds by light from a greensafe lamp. New Phytologist 82(2): 381–383. 124 Beauclair, L., Yu, A., Bouché, N. (2010) microRNA-directed cleavage and translational repression of the copper chaperone for superoxide dismutase mRNA in Arabidopsis. The Plant Journal  62(3): 454–62. Berezikov, E., Gurye, V., van de Belt, J., Wienholds, E., Plasterk, R.H.A. (2005) Phylogenetic shadowing and computational identification of human microRNA genes. Cell 120(1): 21–24. Berry, J.O., Breiding, D.E., Klessig, D.F. (1990) Light-mediated control of translational initiation of ribulose-1, 5-bisphosphate carboxylase in amaranth cotyledons. The Plant Cell 2(8): 795–803. Berry, J.O., Nikolau, B.J., Carr, J.P., Klessig, D.F. (1986) Translational regulation of light-induced ribulose 1,5-bisphosphate carboxylase gene expression in amaranth. Molecular Cellular Biology 6(7): 2347–2353. Biahoua, A., Bonneau, L. (1999) Control of in vitro somatic embryogenesis of the spindle tree (Euonymus europaeus L.) by the sugar type and the osmotic potential of the culture medium. Plant Cell Reports 19(2): 185–190. Bilang, R., Bogorad, L. (1996) Light-dependent developmental control of RbcS gene expression in epidermal cells of maize leaves. Plant Molecular Biology 31(4): 831–841. Bimboim, H.C., Doly, J. (1979) A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Research 7(6): 1513–1523. Björkman, O., Gauhl, E., Nobs, M.A. (1969) Comparative studies of Atriplex species with and without B- carboxylation photosynthesis and their first-generation hybrid. Carnegie Institute of Washington Yearbook 68: 620–633. Björkman, O., Nobs, M. A., Berry, J.A. (1971) Further studies on hybrids between C3 and C4 species of Atriplex. Carnegie Institute of Washington Yearbook 70: 507–511. Björkman, O., Pearcy, R W., Nobs, M.A. (1970) Hybrids between Atriplex species with and without β- carboxylation photosynthesis. Photosynthetic characteristics. Carnegie Institute of Washington Yearbook 69:640–48 Boinski, J.J., Wang, J.L., Xu, P., Hotchkiss, T., Berry, J. O. (1993) Post-transcriptional control of cell type- specific gene expression in bundle sheath and mesophyll chloroplasts of Amaranthus hypochondriacus. Plant Molecular Biology 22(3): 397–410. Bollman, K.M. (2003) HASTY, the Arabidopsis ortholog of exportin 5/MSN5, regulates phase change and morphogenesis. Development 130(8): 1493–1504. Bonnet, E., Wuyts, J., Rouzé, P. Van de Peer, Y. (2004) Detection of 91 potential conserved plant microRNAs in Arabidopsis thaliana and Oryza sativa identifies important target genes. Proceedings of the National Academy of Sciences of the United States of America 101(31): 11511–11516. Bouton, J.H., Brown, R.H., Evans, P.T., Jernstedt, J.A. (1986) Photosynthesis, leaf anatomy, and morphology of progeny from hybrids between C3 and C3/C4 Panicum species. Plant Physiology 80(2): 487–492. Bradford, M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72: 248–254. Bräutigam, A., Hoffmann-Benning, S. (2008) Comparative proteomics of chloroplast envelopes from C3 and C4 plants reveals specific adaptations of the plastid envelope to C4 photosynthesis and candidate proteins required for maintaining C4 metabolite fluxes. Plant Physiology 148(1): 568–579. 125 Bräutigam, A., Kajala, K., Wullenweber, J., Sommer, M., Gagneul, D., Weber, K.L., Carr, K.M. (2011) An mRNA blueprint for C4 photosynthesis derived from comparative transcriptomics of closely related C3 and C4 species. Plant Physiology 155(1): 142–156. Briggs, C.L., Westoby, M., Selkirk, P.M., Oldfield, R.J. (1987) Embryology of early abortion due to limited maternal resources in Pisum sativum L. Annuals of Botany 59(6): 611–619. Brodersen, P., Sakvarelidze-Achard, L., Bruun-Rasmussen, M., Dunoyer, P., Yamamoto, Y. Y., Sieburth, L., Voinnet, O. (2008) Widespread translational inhibition by plant miRNAs and siRNAs. Science 320: 1185-1190 Brooks, A., Farquhar, G.D. (1985) Effect of temperature on the CO2/O2 specificity of ribulose-1,5- bisphosphate carboxylase/oxygenase and the rate of respiration in the light. Planta 165(3): 397–406. Brown, N.J., Newell, C.A., Stanley, S., Chen, J.E., Perrin, A.J., Kajala, K., Hibberd, J.M. (2011) Independent and parallel recruitment of preexisting mechanisms underlying C₄ photosynthesis. Science 331: 1436–9. Brown, R.H., Bassett, C.L., Cameron, R.G., Evans, P.T., Bouton, J.H., Black, C.C., Sternberg, L.O. (1986) Photosynthesis of F(1) hybrids between C4 and C3-C4 species of Flaveria. Plant Physiology 82(1): 211–217. Brown, R.H., Bouton, J.H. (1993) Physiology and genetics of interspecific hybrids between photosynthetic types. Annual Review of Plant Physiology 44: 435–456. Brown, R.H., Bouton, J.H., Evans, P.T., Malter, H.E., Rigsby, L.L. (1985) Photosynthesis, morphology, leaf anatomy, and cytogenetics of hybrids between C3 and C3/C4 Panicum species. Plant Physiology 77(3): 653–658. Brown, R.H., Byrd, G.T., Black, C.C. (1992) Degree of C4 Photosynthesis in C4 and C3-C4 Flaveria species and their hybrids: II Inhibition of apparent photosynthesis by a phosphoenolpyruvate carboxylase inhibitor. Plant Physiology 100(2): 947–950. Brutnell, T.P., Sawers, R.J., Mant, A., Langdale, J.A. (1999) BUNDLE SHEATH DEFECTIVE2, a novel protein required for post-translational regulation of the rbcL gene of maize. The Plant Cell 11(5): 849–64. Burnell, J.N., Suzuki, I., Sugiyama, T. (1990) Light induction and the effect of nitrogen status upon the activity of carbonic anhydrase in maize leaves. Plant Physiology 94(1): 384–387. Bustin, S. (2000) Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays. Journal of Molecular Endocrinology 25(2): 169–193. Byrd, G.T., Brown, R.H., Bouton, J.H., Bassett, C.L., Black, C.C. (1992) Degree of C4 photosynthesis in C4 and C3-C4 Flaveria species and Their Hybrids : I. CO2 Assimilation and metabolism and activities of phosphoenolpyruvate carboxylase and NADP-malic enzyme. Plant Physiology 100(2): 939–46. Byrd, G.T., Sage, R.F., Brown, R.H. (1992) A comparison of dark respiration between C3 and C4 plants. Plant Physiology 100(1): 191–198. Cabido, M., Pons, E., Cantero, J.J., Lewis, J.P., Anton, A. (2007) Photosynthetic pathway variation among C4 grasses along a precipitation gradient in Argentina. Journal of Biogeography 35(1): 131–140. Caemmerer, S., Farquhar, G.D. (1981) Some relationships between the biochemistry of photosynthesis and the gas exchange of leaves. Planta 153(4): 376–387. 126 Cane, J.H. (2008) Breeding biologies, seed production and species-rich bee guilds of Cleome lutea and Cleome serrulata (Cleomaceae). Plant Species Biology 23(3): 152–158. Carrington, J.C., Ambros, V. (2003) Role of microRNAs in plant and animal development. Science 301: 336–338. Chapman, E.J., Prokhnevsky, A.I., Gopinath, K., Dolja, V.V., Carrington, J.C. (2004) Viral RNA silencing suppressors inhibit the microRNA pathway at an intermediate step. Genes & development 18(10): 1179– 86. Chastain, C.J., Heck, J.W., Colquhoun, T.A., Voge, D.G., Gu, X.Y. (2006) Posttranslational regulation of pyruvate, orthophosphate dikinase in developing rice (Oryza sativa) seeds. Planta 224(4): 924–34. Chastain, C.J., Xu, W., Parsley, K., Sarath, G., Hibberd, J.M., Chollet, R. (2008) The pyruvate, orthophosphate dikinase regulatory proteins of Arabidopsis possess a novel, unprecedented Ser/Thr protein kinase primary structure. The Plant Journal 53(5): 854-863. Chellappan, P., Xia, J., Zhou, X., Gao, S., Zhang, X., Coutino, G., azquez, F., Zhang, W., Jin, H. (2010) siRNAs from miRNA sites mediate DNA methylation of target genes. Nucleic Acids Research 38(20): 6883–6894. Chen, C.C., Gibson, P.B. (1972) Barriers to hybridization of Trifolium repens with related species. Canadian Journal of Genetics and Cytology 14(2): 381–389. Chen, X. (2004) A microRNA as a translational repressor of APETALA2 in Arabidopsis flower development. Science 303: 2022–2025. Cheng, S.H., Keefe, D., Mets, L., Ku, M.S.B. (1987) Photosynthetic characteristics of reciprocal backcrosses between F1 hybrid (C3-C4 x C4) and C4 Flaveria species. Plant Physiology 83:4. Chitwood, D.H., Timmermans, M.C.P. (2007) Target mimics modulate miRNAs. Nature Genetics 39(8): 935–936. Chitwood, D.H.Nogueira, F.T., Howell, M.D. (2009) Pattern formation via small RNA mobility. Genes and Development 23: 549–554. Chollet, R., Ogren, W.L. (1975) Regulation of photorespiration in C3 and C4 species. The Botanical Review 41(2): 137–179. Chomczynski, P., Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Analytical Biochemistry 162: 156–159. Christin, P.A., Salamin, N., Muasya, A.M., Roalson, E.H., Russier, F., & Besnard, G. (2008) Evolutionary switch and genetic convergence on rbcL following the evolution of C4 photosynthesis. Molecular biology and Evolution 25(11): 2361–2368. Chuck, G., Cigan, M., Saeteurn, K., Hake, S. (2007) The heterochronic maize mutant Corngrass1 results from overexpression of a tandem microRNA. Nature Genetics 39(4): 544–549. Clough, S.J., Bent, A.F. (1998) Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. The Plant Journal 16(6): 735–743. Collins, R.P., Jones, M.B. (1986) The influence of climatic factors on the distribution of C4 species in Europe. 64: 121–129. 127 Conner, A.J., Falloon, P.G. (1993) Osmotic versus nutritional effects when rooting in vitro asparagus minicrowns on high sucrose media. Plant Science 89(1): 101–106. Craft, J., Samalova, M., Baroux, C., Townley, H., Martinez, A., Tsiantis, M., Moore, I. (2005) New pOp/LhG4 vectors for stringent glucocorticoid-dependent transgene expression in Arabidopsis. The Plant Journal 41(6): 899–918. Dafni, A., Maués, M.M. (1998) A rapid and simple procedure to determine stigma receptivity. Sexual Plant Reproduction 11(3), 177–180. De Block, M., De Brouwer, D., & Tenning, P. (1989) Transformation of Brassica napus and Brassica oleracea using Agrobacterium tumefaciens and the expression of the bar and neo genes in the transgenic plants. Plant Physiology 91(2): 694–701. Dengler N.G., Dengler R.E., Donnelly P.M., Hattersley P.W. (1993) Quantitative leaf anatomy of C3 and C4 grasses (Poaceae): bundle sheath and mesophyll surface area relationships. Annals of Botany 73: 241– 255. Dengler N.G., Nelson T. (1999) Leaf structure and development in C4 plants. In: Sage RF, Monson RK, eds. C4 Plant Biology. San Diego, CA, USA: Academic Press, pp.133–172. Desprez, T., Amselem, J., Caboche, M., Hofte, H. (1998) Differential gene expression in Arabidopsis monitored using cDNA arrays. The Plant Journal 14(5): 643–652. Devers, E.A., Branscheid, A., May, P., Krajinski, F. (2011) Stars and symbiosis: microRNA- and microRNA*-mediated transcript cleavage involved in arbuscular mycorrhizal symbiosis. Plant physiology 156(4): 1990–2010. Ding, D., Zhang, L., Wang, H., Liu, Z., Zhang, Z., Zheng, Y. (2009) Differential expression of miRNAs in response to salt stress in maize roots. Annals of Botany 103(1): 29–38. Dunoyer, P., Lecellier, C., Parizotto, E.A., Himber, C., Voinnet, O. (2004) Probing the microRNA and small interfering RNA pathways with virus-encoded suppressors of RNA silencing. The Plant Cell 16(5): 1235–50. Dunwell, J.M. (1979) Anther culture in Nicotiana tabacum: The role of the culture vessel atmosphere in pollen embryo induction and growth. Journal of Experimental Botany 30(3): 419–428. Edwards, G.E., Ku, M.S.B. (1987) Biochemistry of C3-C4 Intermediates. The Biochemistry of Plants 10: 275–325. Edwards, G.E., Ku, M.S.B., Hatch, M.D. (1982) Photosynthesis in Panicum milioides, a species with reduced photorespiration. Plant Cell Physiology 23(7): 1185–1195. Ehleringer, J., Björkman, O. (1977) Quantum yields for CO2 uptake in C3 and C4 plants. Plant Physiology 59(1): 86–90. Ehleringer, J.R., Sage, R.F., Flanagan, L.B., Pearcy, R.W. (1991) Climate change and the evolution of C4 photosynthesis. Trends in Ecology & Evolution 6(3): 95–99. Emmert-Buck, M.R, Bonner, R.F., Smith, P.D., Chaqui, R.F., Zhuang, Z., Goldstein, S.R., Weiss, R.A., Liotta, L.A. (1996) Laser Capture Microdissection. Science 274: 998-1001. 128 Eshed, Y., Zamir, D. (1995) An introgression line population of Lycopersicon pennellii in the cultivated tomato enables the identification and fine mapping of yield-associated QTL. Genetics 141: 1147–1162. Evans, L.T., Fischer, R.A. (1999) Yield potential: its definition and significance. Crop Science 39(6): 1544 – 1551. Evenson, R.E, Gollin, D. (2003) Assessing the impact of the Green Revolution, 1960 – 2000. Science 300: 75 –762. Farquhar, G.D., Caemmerer, S., Berry, J.A. (1980) A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta 149(1): 78–90. Finkelstein, R.R., Crouch, M.L. (1986) Rapeseed embryo development in culture on high osmoticum is similar to that in seeds. Plant Physiology 81(3): 907–912. Fisher, D.K., Guiltinan, M.J. (1995) Rapid, efficient production of homozygous transgenic tobacco plants with Agrobacterium tumefaciens: A seed-to-seed protocol. Plant Molecular Biology Reporter 13(3): 278– 289. Fischer, G., Heilig G.K. (1997) Population momentum and the demand on land and water resources. Philosophical Transactions of the Royal Society B 352(1356): 869–889. Forsberg, J., Landgren, M., Glimelius, K. (1994) Fertile somatic hybrids between Brassica napus and Arabidopsis thaliana. Plant Science 95(2): 213–223. Frohman, M.A. (1988) Rapid production of full-length cDNAs from rare transcripts: Amplification using a single gene-specific oligonucleotide primer. Proceedings of the National Academy of Sciences 85(23): 8998–9002. Fujii, H., Chiou, T.-J., Lin, S.-I., Aung, K., Zhu, J.-K. (2005) A miRNA involved in phosphate-starvation response in Arabidopsis. Current Biology 15(22): 2038–2043. Furbank, R.T. (2011) Evolution of the C4 photosynthetic mechanism: are there really three C4 acid decarboxylation types? Journal of Experimental Botany 62(9): 3103–3108. Furumoto, T., Yamaguchi, T., Ohshima-Ichie, Y., Nakamura, M., Tsuchida-Iwata, Y., Shimamura, M., Ohnishi, J. (2011) A plastidial sodium-dependent pyruvate transporter. Nature 476: 472–475. Gallaher, R.N., Brown, R.H. (1977) Starch storage in C4 vs. C3 grass leaf cells as related to nitrogen deficiency. Crop Science 17(1): 85. Gamborg, O.L., Miller, R.A., Ojima, K. (1968) Nutrient requirements of suspension cultures of soybean root cells. Experimental Cell Research 50(1): 151–158. Gehring, H., Kasemir, H., Mohr. H. (1977) The capacity of chlorophyll-a biosynthesis in the mustard seedling cotyledons as modulated by phytochrome and circadian rhythmicity. Planta 133(3): 295–302. Gowik, U., Burscheidt, J., Akyildiz, M., Schlue, U., Koczor, M., Streubel, M., Westhoff, P. (2004) cis- Regulatory elements for mesophyll-specific gene expression in the C4 plant Flaveria trinervia, the promoter of the C4 phosphoenolpyruvate carboxylase. The Plant Cell 16: 1077–1090. Griffiths, D.J., Pegler, R.A.D., Tonguthaisri, T. (1971) Cross compatibility between diploid and tetraploid perennial ryegrass (Lolium perenne L.). Euphytica 20(1): 102–112. 129 Guo, S., Schinner, K., Sattelmacher, B., Hansen, U.P. (2005) Different apparent CO2 compensation points in nitrate- and ammonium-grown Phaseolus vulgaris and the relationship to non- photorespiratory CO2 evolution. Physiologia Plantarum 123(3): 288–301. Gutierrez, R.A., Ewing, R.M., Cherry, J.M., Green, P.J. (2002) Identification of unstable transcripts in Arabidopsis by cDNA microarray analysis: rapid decay is associated with a group of touch- and specific clock-controlled genes. Proceedings of the National Academy of Sciences of the United States of America 99(17): 11513–11518. Haberlandt, D.G. (1882) Vergleichende anatomie des assimilatorischen gewebesystems der pflanzen. Pringsheim Jahrbuch fur Wiss. Botanischer Boded 13. Hall, L.N., Rossini, L., Cribb, L., Langdale, J.A. (1998) GOLDEN 2: a novel transcriptional regulator of cellular differentiation in the maize leaf. The Plant Cell, 10(6): 925–936. Hall, T.A. 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41: 95-98. Hamilton, A.J., Baulcombe, D.C. (1999) A species of small antisense RNA in posttranscriptional gene silencing in plants. Science 286: 950–952. Hanahan, D. (1983) Studies on transformation of Escherichia coli with plasmids. Journal of Molecular Biology 166(4): 557–580. Harberd, D.J. (1969) A simple effective embryo culture technique for Brassica. Euphytica 18(3): 425–429. Harrison, S.J., Mott, E.K., Parsley, K., Aspinall, S., Gray, J.C., Cottage, A. (2006) A rapid and robust method of identifying transformed Arabidopsis thaliana seedlings following floral dip transformation. Plant Methods 2(1): 19. Hartwell, J., Gill, A., Nimmo, G.A., Wilkins, M.B., Jenkins, G.I., Nimmo, H.G. (1999) Phosphoenolpyruvate carboxylase kinase is a novel protein kinase regulated at the level of expression. The Plant Journal, 20(3): 333–342. Hatch, M.D. (1971) The C4 pathway of photosynthesis. Evidence for an intermediate pool of carbon dioxide and the identity of the donor C 4 -dicarboxylic acid. The Biochemical journal, 125(2): 425–32. Hatch, M.D., Slack, C.R. (1966) Photosynthesis by sugar-cane leaves. A new carboxylation reaction and the pathway of sugar formation. The Biochemical Journal, 101(1), 103–11. Havecker, E. (2011) Detection of small RNAs and microRNAs using deep sequencing technology. Methods in Molecular Biology 732: 55–68. Hibberd, J.M., Sheehy, J.E., Langdale, J.A. (2008) Using C4 photosynthesis to increase the yield of rice- rationale and feasibility. Current Opinion in Plant Biology, 11(2): 228–31. Holaday, A.S., Brown, R.H., Bartlett, J.M., Sandlin, E.A., Jackson, R.C. (1988) Enzymic and photosynthetic characteristics of reciprocal F1 hybrids of Flaveria pringlei C3 and Flaveria brownii (C4- Like Species). Plant Physiology 87(2): 484–490. Hsieh, L.C., Lin, S.I., Shih, A.C.C., Chen, J.W., Lin, W.Y., Tseng, C.Y., Li, W.H., Chiou, T.J. (2009) Uncovering small RNA-mediated responses to phosphate deficiency in Arabidopsis by deep sequencing. Plant Physiology 151(4): 2120–2132. 130 Hudson, G., Dengler, R., Hattersley, P., Dengler, G. (1992) Cell-specific expression of Rubisco small subunit and Rubisco activase genes in C3 and C4 species of Atriplex. Australian Journal of Plant Physiology: 19(1): 89. Hunt, S., Smith, A.M., Woolhouse, H.W. (1987) Evidence for a light-dependent system for reassimilation of photorespiratory CO2, which does not include a C4 cycle, in the C3/C4 intermediate species Moricandia arvensis. Planta 171(2): 227–234. Inda, L.A., Torrecilla, P., Catalán, P., Ruiz-Zapata, T. (2008) Phylogeny of Cleome L. and its close relatives Podandrogyne Ducke and Polanisia Raf. (Cleomoideae, Cleomaceae) based on analysis of nuclear ITS sequences and morphology. Plant Systematics and Evolution 274(1-2): 111–126. Jackson, D.P. (1992) In situ hybridization in plants. In Molecular Plant Pathology: A Practical Approach (D.J. Bowles, S.J. Gurr and M.McPhereson, Eds) Oxford University Press, Oxford, pp. 163–174. Jacob, J., Lawlor, D.W. (1992) Dependence of photosynthesis of sunflower and maize leaves on phosphate supply, ribulose-1,5-bisphosphate carboxylase/oxygenase activity, and ribulose-1,5- bisphosphate pool size. Plant Physiology 98(3): 801–807. Jefferson, R.A. (1987) Assaying chimeric genes in plants: The GUS gene fusion system. Plant Molecular Biology Reporter 5(4): 387–405. Jefferson, R.A., Kavanagh, T.A., Bevan, M.W. (1987) GUS fusions: beta-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. The EMBO journal 6(13): 3901–3907. Jiao, J., Chollet, R. (1988) Light/dark regulation of maize leaf phosphoenolpyruvate carboxylase by in vivo phosphorylation. Archives of Biochemistry and Biophysics 261(2): 409–417. Jones, C.G., Daniel Hare, J., Compton, S.J. (1989) Measuring plant protein with the Bradford assay. Journal of Chemical Ecology 15(3): 979–992. Jones-Rhoades, M.W., Bartel, D.B. (2004) Computational identification of plant microRNAs and their targets, including a stress-induced miRNA. Molecular Cell 14(6): 787–799. Kajala, K., Brown, N.J., Williams, B.P., Borrill, P., Taylor, L.E., & Hibberd, J.M. (2011) Multiple Arabidopsis genes primed for recruitment into C4 photosynthesis. The Plant Journal 69(1): 47–56.  Kalamajka, R., Hahnen, S., Cavalar, M., Töpsch, S., Weier, D., Peterhänsel C. (2003) Restriction accessibility in isolated nuclei reveals light-induced chromatin reorganization at the PEPC promoter in maize. Plant Molecular Biology 52(3): 669–678. Kanai, R., Edwards, G.E. (1973) Separation of mesophyll protoplasts and bundle sheath cells from maize leaves for photosynthetic studies. Plant Physiology 51(6): 1133–1137. Kanai, R., Edwards, G.E., (1999) Structure-Function of the C4 Syndromes In: Sage RF, Monson RK, eds. C4 Plant Biology. San Diego, CA, USA: Academic Press, pp.49–80. Kang, H.G., Fang, Y., Singh, K.B. (1999) A glucocorticoid-inducible transcription system causes severe growth defects in Arabidopsis and induces defense-related genes. The Plant Journal 20(1): 127–133. Kasschau, K.D., Xie, Z., Allen, E., Llae, C., Chapman, E.J., Krizan, K.A., Carrington, J.C. (2003) P1/HC- Pro, a viral suppressor of RNA silencing, interferes with Arabidopsis development and miRNA function. Developmental Cell 4(2): 205–217. 131 Katara, P., Gautam, B., Kuntal, H., Sharma, V. (2010) Prediction of miRNA targets, affected proteins and their homologs in Glycine max. Bioinformation 5(4): 162–5. Khush, G.S. (1999) Green revolution: preparing for the 21 st centurary. Genome 42(4): 646 – 655. Kausch, A.P., Page Owen, T., Zachwieja, S.J., Flynn, A.R., & Sheen, J. (2001) Mesophyll-specific, light and metabolic regulation of the C4 PPCZm1 promoter in transgenic maize. Plant Molecular Biology 45(1): 1– 15. Kellogg, E.A. (1999) Phylogenetic aspects of the evolution of C4 photosynthesis In: Sage RF, Monson RK, eds. C4 Plant Biology. San Diego, CA, USA: Academic Press, pp.411–439. Khan, S., Rowe, S.C, Harmon, F.G. (2010) Coordination of the maize transcriptome by a conserved circadian clock. BMC Plant Biology 10(1):126. Kiniry, J.R., Jones, C.A., O’Toole, J.C., Blanchet, R., Cabelguenne, M., Spanel, D.A. (1989) Radiation-use efficiency in biomass accumulation prior to grain filling for five grain-crop species. Field Crops Research 20(1): 51–64. Kidner, C.A., Martienssen, R.A. (2004) Spatially restricted microRNA directs leaf polarity through ARGONAUTE1. Nature 428: 81–84. Kim, J. (2003) A Mechanism for Light-Induced Translation of the rbcL mRNA Encoding the Large Subunit of Ribulose-1,5-bisphosphate Carboxylase in Barley Chloroplasts. Plant and Cell Physiology 44(5): 491–499. Kim, V.N., Nam, J.W. (2006) Genomics of microRNA. Trends in Genetics 22(3): 165–73. Kiniry, J.R., Jones, C.A., O’Toole, J.C., Blanchet, R., Cabelguenne, M., Spanel, D.A. (1989) Radiation-use efficiency in biomass accumulation prior to grain-filling for five grain-crop species. Field Crops Research 20(1), 51-64. Klein, R.R., Mullet, J.E. (1987) Control of gene expression during higher plant chloroplast biogenesis. Protein synthesis and transcript levels of psbA, psaA-psaB, and rbcL in dark-grown and illuminated barley seedlings. Journal of Biological Chemistry 262(9): 4341–4348. Kleppe, K., Ohtsuka, E., Kleppe, R., Molineux, I., Khorana, H.G. (1971) Studies on polynucleotides. Journal of Molecular Biology 56(2): 341–361. Knapp, A.K., Cocke, M., Hamerlynch, E.P., Owensby, C.E. (1994) Effect of elevated CO2 on stomatal density and distribution in a C4 grass and a C3 forb under field conditions. Annuals of Botany 74(6): 595–599. Koteyeva, N.K., Voznesenskaya, E.V., Roalson, E.H., & Edwards, G.E. (2011) Diversity in forms of C4 in the genus Cleome (Cleomaceae). Annals of Botany 107(2): 269–283. Koyama, T., Furutani, M., Tasaka, M., Ohme-Takagi, M. (2007) TCP transcription factors control the morphology of shoot lateral organs via negative regulation of the expression of boundary-specific genes in Arabidopsis. The Plant Cell 19(2): 473–484. Kubicki, A., Steinmaller, K., Westhoff, P., Disseldorf, H. (1994) Differential transcription of the plastome-encoded genes in the mesophyll and bundle-sheath chloroplasts of the monocotyledonous NADP-malic enzyme-type C4 plants maize and Sorghum. Plant Molecular Biology 25(4): 669–679. 132 Kulya, J., Lontom,W., Bunnag, S. Theerakulpisut, P. (2011) Cleome gynandra L. (a C4 plant) shows higher tolerance of salt stress than its C3 close relative, C. viscosa L. Advances in Agriculture & Botanics 3(1): 59– 66. Kurihara, Y., Watanabe, Y. (2004) Arabidopsis micro-RNA biogenesis through Dicer-like 1 protein functions. Proceedings of the National Academy of Sciences of the United States of America 101(34): 12753–12758. Lakatos, L., Csorba, T., Pantaleo, V., Chapman, E., Carrington, J.C., Liu, Y.P., Calino, L.F., Lopez-Moya, J., Burgyán, J. (2006) Small RNA binding is a common strategy to suppress RNA silencing by several viral suppressors. The EMBO Journal 25(12): 2768–2780. Lakatos, L., Szittya, G., Silhavy, D., Burgyán, J. (2004) Molecular mechanism of RNA silencing suppression mediated by p19 protein of tombusviruses. The EMBO Journal 23(4):876–884. Langdale, J.A. (1993) In situ hybridization. In: The Maize Handbook. (Eds: V. Walbot and M. Freeling) pp. 165–180. Langdale, J.A., Kidner, C.A. (1994) bundle sheath defective, a mutation that disrupts cellular differentiation in maize leaves. Cell 681: 673–681. Langdale, J.A., Metzler, M.C., Nelson, T. (1987) The argentia mutation delays normal development of photosynthetic cell-types in Zea mays. Developmental Biology 122(1): 243–255. Langdale, J.A., Nelson, T. (1991) Spatial regulation of photosynthetic development in C4 plants. Trends in Genetics 7(6): 191–196. Langdale, J.A., Taylor, W.C., Nelson, T. (1991) Cell-specific accumulation of maize phosphoenolpyruvate carboxylase is correlated with demethylation at a specific site greater than 3 kb upstream of the gene. Molecular & General Genetics 225(1): 49–55. Langdale, J.A., Zelitch, I., Miller, E., Nelson, T. (1988) Cell position and light influence C4 versus C3 patterns of photosynthetic gene expression in maize. The EMBO Journal 7(12): 3643–3651. Langmead, B., Trapnell, C., Pop, M., Salzberg, S.L. (2009) Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biology 10:R25 Li, T., Li, H., Zhang, Y.X., Liu, J.Y. (2011) Identification and analysis of seven H₂O₂-responsive miRNAs and 32 new miRNAs in the seedlings of rice (Oryza sativa L. ssp. indica). Nucleic acids Research 39(7): 2821–2833. Li., Z., Pinson, S.R.M., Paterson, A.H., Park, W.D., Stansel, J.W. (1997) Genetics of hybrid sterility and hybrid breakdown in an intersubspecific rice (Oryza sativa L.) population. Genetics 145: 1139–1148. Lin, C. (1998) Enhancement of blue-light sensitivity of Arabidopsis seedlings by a blue light receptor cryptochrome. Proceedings of the National Academy of Sciences 95(5): 2686–2690. Liu, C.M., Xu, Z.H., Chua, N.H. (1993) Proembryo culture: in vitro development of early globular-stage zygotic embryos from Brassica juncea. The Plant Journal 3(2): 291–300. Liu, J., Wu, Y.H., Yang, J.J., Liu, Y.D., & Shen, F.F. (2008) Protein degradation and nitrogen remobilization during leaf senescence. Journal of Plant Biology 51(1): 11–19. 133 Livak, K.J., Schmittgen, T.D. (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25(4): 402–408. Llave, C., Kasschau, K. D., Rector, M. A., Carrington, J. C. (2002a) Endogenous and Silencing-Associated Small RNAs in Plants. Society 14: 1605–1619. Llave, C., Xie, Z., Kasschau, K.D., Carrington, J.C. (2002) Cleavage of Scarecrow-like mRNA targets directed by a class of Arabidopsis miRNA. Science 297: 2053–2056. Long, S.P., Ainsworth, E.A., Leakey, A.D.B, Nosberger, J., Ort, D.R. (2006) Food for thought: lower-than- expected crop yield simulation with rising CO2 concentrations. Science 312: 1918–1921. Lobell, D.B., Burke, M.B, Tebaldi, C., Mastrandrea, M.D., Falcon, W.P., Naylor, R.L. (2008) Prioritising climate change adaptation needs for food security in 2030. Science 319: 607–610. Long, J.J., Berry, J.O. (1996) Tissue-specific and light-mediated expression of the C4 photosynthetic NAD-dependent malic enzyme of Amaranth mitochondria. Plant Physiology 112(2): 473–482. Love, A.C., Andrews, M.E., Raff, R.A. (2007) Gene expression patterns in a novel animal appendage: the sea urchin pluteus arm. Evolution & Development 9(1): 51–68. Lu, S., Sun, Y.H., Chiang, V.L. (2008) Stress-responsive microRNAs in Populus. The Plant Journal 55(1): 131–151. Lu, S., Sun, Y.H., Shi, R., Clark, C., Li, L., Chiang, V.L. (2005) Novel and mechanical stress-responsive MicroRNAs in Populus trichocarpa that are absent from Arabidopsis. The Plant Cell 17(8): 2186–203. Ma, L., Li, J., Qu, L., Hager, J., Chen, Z., Zhao, H., Deng, X.W. (2001) Light control of Arabidopsis development entails coordinated regulation of genome expression and cellular pathways. Plant Cell 13(12): 2589–2607. Machado, I.C., Lopes, A.V., Leite, A.V., de Brito Neves, C. (2006) Cleome spinosa (Capparaceae): polygamodioecy and pollination by bats in urban and Caatinga areas, northeastern Brazil. Botanische Jahrbücher 127(1): 14. Maizel, J.V. (1966) Acrylamide-gel electrophorograms by mechanical fractionation: Radioactive adenovirus proteins. Science 151: 988–990. Mallory, A.C., Reinhart, B.J., Jones-Rhoades, M.W., Tang, G., Zamore, P.D., Barton, M.K., Bartel, D.P. (2004) MicroRNA control of PHABULOSA in leaf development: importance of pairing to the microRNA 5’ region. The EMBO Journal 23(16): 3356–64. Mansfield, S.G., Briarty, I.G. (1996) The dynamics of seedling and cotyledon cell development in Arabidopsis thaliana during reserve mobilisation. International Journal of Plant Science 157(3): 280–295. Marks, T., Simpson, S. (2000) Interaction of explant type and indole-3-butyric acid during rooting in vitro in a range of difficult and easy-to-root woody plants. Plant Cell, Tissue and Organ Culture 62(1): 67–74. Marshall, D.M., Muhaidat, R., Brown, N.J., Liu, Z., Stanley,S., Griffiths, H., Sage, R.F. (2007) Cleome, a genus closely related to Arabidopsis, contains species spanning a developmental progression from C3 to C4 photosynthesis. The Plant Journal 51(5): 886–896. 134 Martin, F.W. (1959) Staining and observing pollen tubes in the style by means of fluorescence. Biotechnic & Histochemistry 34(3): 125–128. Matsuoka, M. (1993) Tissue-specific light-regulated expression directed by the pomoter of a C4 gene, maize pyruvate,orthophosphate dikinase, in a C3 Plant, Rice. Proceedings of the National Academy of Sciences 90(20): 9586–9590. Malthus, T.R. (1798) An essay on the principle of population in Oxford World’s Classic Reprints, Chapter 1, pp.13. Maxwell, D.P., Laudenbach, D.E., Huner, N.P.A. (1995) Redox regulation of light-harvesting complex II and cab mRNA abundance in Dunaliella salina. Plant Physiology 109(3): 787–795. McCalla, A.F. (1994) Agriculture and food needs to 2025: Why we should be concerned. McCormac, D., Boinski, J.J., Ramsperger, V.C., Berry, J.O. (1997) C4 Gene expression in photosynthetic and nonphotosynthetic leaf regions of Amaranthus tricolor. Plant Physiology 114(3): 801–815. McCormac, D.J., Litz, H., Wang, J., Gollnick, P.D., & Berry, J.O. (2001) Light-associated and processing- dependent protein binding to 5’ regions of rbcL mRNA in the chloroplasts of a C4 plant. The Journal of Biological Chemistry 276(5): 3476–83. McHale, N.A., Koning, R.E. (2004) MicroRNA-directed cleavage of Nicotiana sylvestris PHAVOLUTA mRNA regulates the vascular cambium and structure of apical meristems. The Plant Cell 16(7): 1730– 1740. McLaughlin, S.B., Andersen, C.P., Hanson, P.J., Tjoelker, M.G., Roy, W.K. (1991) Increased dark respiration and calcium deficiency of red spruce in relation to acidic deposition at high-elevation southern Appalachian Mountain sites. Canadian Journal of Forest Research 21(8): 1234–1244. Meierhoff, K., Westhoff, P. (1993) Differential biogenesis of photosystem II in mesophyll and bundle- sheath cells of monocotyledonous NADP-malic enzyme-type C4 plants: the non-stoichiometric abundance of the subunits of photosystem II in the bundle-sheath chloroplasts and the translational. Planta 191(1): 23–33. Meinke, D.W. (1992) A Homoeotic Mutant of Arabidopsis thaliana with Leafy Cotyledons. Science 258: 1647–1650. Mendoza, M.G., Kaeppler, H.F. (2002) Auxin and sugar effects on callus induction and plant regeneration frequencies from mature embryos of wheat (Triticum aestivum L.). In Vitro Cellular & Developmental Biology - Plant 38(1): 39–45. Michael, T.P., McClung, C.R. (2003) Enhancer trapping reveals widespread circadian clock transcriptional control in Arabidopsis. Plant Physiology 132(2): 629–639. Millar, A.A., Gubler, F. (2005) The Arabidopsis GAMYB-Like genes, MYB33 and MY65, are microRNA- regulated genes that redundantly facilitate anther development. The Plant Cell 17(3): 705–721. Monson, R., Rawsthorne, S. (2004) CO2 assimilation in C3-C4 intermediate plants. Photosynthesis 9: 533–550. Monson, R.K., Moore, B.D. (1989) On the significance of C3-C4 intermediate photosynthesis to the evolution of C4 photosynthesis. Plant, Cell & Enironment 12(7): 689–699. 135 Moore, B.D., Ku, M.S.B., Edwards, G.E. (1989) Expression of C4-like photosynthesis in several species of Flaveria. Plant, Cell and Environment 12(5): 541–549. Moore, I., Samalova, M., Kurup, S. (2006) Transactivated and chemically inducible gene expression in plants. The Plant Journal 45(4): 651–683. Moore, K.G., Bentley, K., Lovell, P.H. (1972) A comparative study of chlorophyll production in cotyledons. Journal of Experimental Botany 23(2): 432–439. Moran, R., Porath, D. (1980) Chlorophyll determination in intact tissues using n,n-dimethylformamide. Plant Physiology 65(3): 478–479. Moxon, S., Schwach, F., Dalmay, T., Maclean, D., Studholme, D.J., Moulton, V. (2008) A toolkit for analysing large-scale plant small RNA datasets. Bioinformatics 24(19): 2252–2253. Moyle, L.C., Nakazato, T. (2010) Hybrid incompatibility snowballs between Solanum species. Science 329: 1521-1523. Murashige, T., Skoog, F. (1962) A revised medium for rapid growth and bioassays with Tobacco tissue cultures. Physiologia Plantarum 15(3): 473–497. Nambudiri, E.M.V., Tidwell, W.D., Smith, B.N., Hebbert, N.P. (1978) A C4 plant from the Pliocene. Nature 276: 816–817. Narsai, R., Howell, K.A., Millar, A.H., O’Toole, N., Smmall, I., Whelan, J. (2007) Genome-wide analysis of mRNA decay rates and their determinants in Arabidopsis thaliana. The Plant cell 19(11): 3418–3436. Neff, M.M., Van Volkenburgh, E. (1994) Light-stimulated cotyledon expansion in Arabidopsis seedlings (the role of Phytochrome B). Plant Physiology 104(3): 1027–1032. Nelson, T. (1984) Light-regulated gene expression during maize leaf development. The Journal of Cell Biology 98(2):558–564. Newell, C.A., Brown, N.J., Liu, Z., Pflug, A., Gowik, U., Westhoff, P., & Hibberd, J.M. (2010) Agrobacterium tumefaciens-mediated transformation of Cleome gynandra L., a C4 dicotyledon that is closely related to Arabidopsis thaliana. Journal of experimental Botany 61(5): 1311-1319. Nitsch, J.P., Nitsch, C. (1969) Haploid plants from pollen grains. Science 163: 85–87. Obernosterer, G., Leuschner, P.J.F., Alenius, M., Martinez, J. (2006) Post-transcriptional regulation of microRNA expression. RNA 12(7): 1161-1167. Orr, H. A., & Turelli, M. (2001) The evolution of postzygotic isolation: Accumulating Dobzhansky- Muller incompatibilities. Evolution 55(6): 1085–1094. Palatnik, J.F., Allen, E., Wu, X., Schommer, C., Schwab, R., Carrington, J.C.C, Weigel, D. (2003) Control of leaf morphogenesis by microRNAs. Nature 425: 257–263. Parizotto, E.A., Dunoyer, P., Rahm, N., Himber, C., Voinnet, O. (2004) In vivo investigation of the transcription, processing, endonucleolytic activity, and functional relevance of the spatial distribution of a plant miRNA. Genes & Development 18(18): 2237–2242. Parra, R., Amo-Marco, J.B. (1996) Effect of plant growth regulators and basal media on in vitro shoot proliferation and rooting of Myrtus communis L. Biologia Plantarum 38(2): 161–168. 136 Parry, M., Rosenzweig, C., Iglesias, A., Fischer, G., Livermore, M. (1999) Climate change and world food security: a new assessment. Global Environmental Change 9 (S1): 51–67. Parys, E., Jastrzebski, H. (2006) Light-enhanced dark respiration in leaves, isolated cells and protoplasts of various types of C4 plants. Journal of Plant Physiology 163(6): 638–647. Parys, E., Romanowska, E., Siedlecka, M., Poskuta, J.W. (1998) The effect of lead on photosynthesis and respiration in detached leaves and in mesophyll protoplasts of Pisum sativum. Acta Physiologiae Plantarum 20(3): 313–322. Patel, M., Corey, A.C., Yin, L.P., Ali, S., Taylor, W.C., Berry, J.O. (2004) Untranslated regions from C4 Amaranth AhRbcS1 mRNAs confer translational enhancement and preferential bundle sheath cell expression in transgenic C4 Flaveria bidentis. Gene Expression 136: 3550–3561. Patel, M., Siegel, A.J., Berry, J.O. (2006) Untranslated regions of FbRbcS1 mRNA mediate bundle sheath cell-specific gene expression in leaves of a C4 plant. The Journal of Biological Chemistry 281(35): 25485– 25491. Paulsen, H., Bogorad, L. (1988) Diurnal and circadian rhythms in the accumulation and synthesis of mRNA for the light-harvesting chlorophyll a/b-Binding protein in Tobacco. Plant Physiology 88(4): 1104–1109. Pearcy, R.W., Björkman, O. (1970) Hybrids between Atriplex species with and without ß-carboxylation photosynthesis: Biochemical characteristics. Carnegie Institute of Washington Yearbook 69: 632–640. Prabu, G.R., Mandal, A.K.A. (2010) Computational identification of miRNAs and their target genes from expressed sequence tags of tea (Camellia sinensis). Genomics, Proteomics & Bioinformatics 8(2): 113–121. Pyankov, V.I., Voznesenskaya, E.V., Kuz’min, A.N., Ku, M.S.B, Ganko, E., Franceschi, V.R., Black, C.C. (2000) Occurrence of C3 and C4 photosynthesis in cotyledons and leaves of Salsola species (Chenopodiaceae). Photosynthesis Research 63(1): 69–84. Quazi, M.H. (1988) Interspecific hybrids between Brassica napus L. and B. oleracea L. developed by embryo culture. Theoretical and Applied Genetics 75(2): 309–318. Rajendrudu, G., Rama’Das, V.S. (1982) Seed surface study of possible hybrids between C3 and C4 species of Cleome (Capparidaceae) Current Science 51(16): 795–797. Ramsperger, V.C., Summers, R.G., Berry, J.O. (1996) Photosynthetic gene expression in meristems and during initial leaf development in a C4 dicotyledonous plant. Plant Physiology 111(4): 999–1010. Reich, P.B. (1983) Effects of low concentrations of O3 on net photosynthesis, dark respiration, and chlorophyll contents in aging hybrid poplar leaves. Plant Physiology 73(2): 291–296. Renart, J. (1979) Transfer of proteins from dels to diazobenzyloxymethyl-paper and detection with antisera: A method for studying antibody specificity and antigen structure. Proceedings of the National Academy of Sciences 76(7): 3116–3120. Reyes, J.L., Chua, N.H. (2007) ABA induction of miR159 controls transcript levels of two MYB factors during Arabidopsis seed germination. The Plant Journal 49(4): 592–606. Rhoades, M.W., Reinhart, B.J., Lim, L.P., Burge, C.B., Bartel, B., Bartel, D.P. (2002) Prediction of plant microRNA targets. Cell 110(4), 513–520. 137 Rieseberg, L.H., Desrochers, A.M., Youn, S.J. (1995) Interspecific pollen competition as a reproductive barrier between sympatric species of Helianthus (Asteraceae). American Journal of Botany 82(4): 515– 519. Robinson, J.T., Thorvaldsdóttir, H., Winckler, W., Guttman, M., Lander, E.S., Getz, G., Mesirov, J.P. (2011) Integrative genomics viewer. Nature Biotechnology 29(1): 24–26. Ruiz-Zapata, T., Huérfano, A.A., Xena de Enrech, N. (1996) Contribución al studio citotaxonómico del género Cleome L. (Capparidaceae). Fyton 59: 85–94. Sage, R.F. (2004) The evolution of C4 photosynthesis. New Phytologist 161(2): 341–370. Sage, R.F., Christin, P.A., Edwards, E.J. (2011) The C4 plant lineages of planet Earth. Journal of Experimental Botany 62(9): 3155–3169. Sage, R.F., Monson, R.K. (1999) C4 Plant Biology. San Diego, CA, USA: Academic Press. Saiki, R., Scharf, S., Faloona, F., Mullis, K., Horn, G., Erlich, H., & Arnheim, N. (1985) Enzymatic amplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230: 1350–1354. Salomé, P.A., Xie, Q., McClung, C.R. (2008) Circadian timekeeping during early Arabidopsis development. Plant Physiology 147(3): 1110–1125. Samalova, M., Brzobohaty, B., Moore, I. (2005) pOp6/LhGR: a stringently regulated and highly responsive dexamethasone-inducible gene expression system for tobacco. The Plant Journal 41(6): 919– 935. Sambrook, J., Fritsch, E., Maniatis, T. (1989) Quantitation of DNA and RNA In: Ford N, Nolan C, Ferguson M, eds. Molecular Cloning. A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, 1989:E.5 (Appendix) Sanger, F., Coulson, R.A. (1975) A rapid method for determining sequences in DNA by primed synthesis with DNA polymerase. Journal of Molecular Biology 94(3): 441–448. Sanger, F., Nicklen, S., Coulson, A.R. (1977) DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences 74(12): 5643–5467. Schoof, H., Lenhard, M., Haecker, A., Mayer, K.F.X., Jurgens, G., Laux, T. (2000) The stem cell population of Arabidopsis shoot meristems is maintained by a regulatory loop between the CLAVATA and WUSCHEL Genes. Cell 100(6): 635–644. Schroeder, A., Mueller, O., Stocker, S., Salowsky, R., Leiber, M., Gassmann, M., Lightfoot, S. (2006) The RIN: an RNA integrity number for assigning integrity values to RNA measurements. BMC Molecular Biology 7(1): 3. Shapiro, A.L., Scharff, M.D., Maizel, J.V., Uhr, J.W. (1966) Polyribosomal synthesis and assembly of the H and L chains of gamma globulin. Proceedings of the National Academy of Sciences of the United States of America 56(1): 216–21. Sheen, J. (1990) Metabolic repression of transcription in higher plants. The Plant Cell 2(10): 1027–38. Shu, G. 1999. Light Induction of cell type differentiation and cell-type-specific gene expression in cotyledons of a C4 Plant, Flaveria trinervia. Plant Physiology 121(3): 731–741. 138 Smith, E.D. (1970) Pollen competition and relatedness in Haplopappus section Isopappus (Compositae) II. American Journal of Botany 57(7): 874–880. Sommer, M.A., Bräutigam, A., Weber, A.P.M. (2012) The dicotyledonous NAD malic enzyme C4 plant Cleome gynandra displays age-dependent plasticity of C4 decarboxylation biochemistry. Plant Biology Published Online (June 2012) Sorin, C., Bussel, J.D., Camus, I., Ljung, K., Kowalczyk, M., Geiss, G., McKhann, H., Garcion, C., Vaucheret, H., Sandberg, G., Bellini, C. (2005) Auxin and light control of adventitious rooting in Arabidopsis require ARGONAUTE1. The Plant Cell 17(5): 1343–59. Srivastava, S., Srivastava, A.K., Suprasanna, P., D’Souza, S.F. (2012) Identification and profiling of arsenic stress-induced microRNAs in Brassica juncea. Journal of Experimental Botany Advanced Online Publication, 16 th November 2012. Steeves, T.A., Sussex, I.M. (1989) Plant Developmental Patterning pp 6–26. Cambridge University Press, UK. Sternberg, L.S., Deniro, M.J., Sloan, M.E., Black, C.C. (1986) Compensation point and isotopic characteristics of C3/C4 intermediates and hybrids in Panicum. Plant Physiology 80(1): 242–245. Stewart, J.M., Hsu, C.L. (1978) Hybridization of diploid and tetraploid cottons through in-ovulo embryo culture. Journal of Hereditary 69(6): 404–408. Stockhaus, J., Schlue, U., Koczor, M., Chitty, J.A., Taylor, W.C.,Westhoff, P. (1997) The promoter of the gene encoding the C4 Form of phosphoenolpyruvate carboxylase directs mesophyll-specific expression in transgenic C4 Flaveria spp. The Plant Cell 9(4): 479–489. Stout, A.A. (1923) Alternation of sexes and intermittent production of fruit in the spider flower (Cleome spinosa). American Journal of Botany 10(2): 57–66. Sunkar, R., Zhu, J.K. (2004) Novel and stress-regulated microRNAs and other small RNAs from Arabidopsis. The Plant Cell 16(8): 2001–2019. Taylor, L., Nunes-Nesi, A., Parsley, K., Leiss, A., Leach, G., Coates, S., Wingler, A. (2010) Cytosolic pyruvate,orthophosphate dikinase functions in nitrogen remobilization during leaf senescence and limits individual seed growth and nitrogen content. The Plant Journal 62(4): 641–652. Terada, R., Yamashita, Y., Nishibayashi, S., Shimamoto, K. (1987) Somatic hybrids between Brassica oleracea and B. Campestris: selection by the use of iodoacetamide inactivation and regeneration abilty. Theoretical and Applied Genetics 73(3): 379–384. Tolley, B.J., Woodfield, H., Wanchana, S., Bruskiewich, R., & Hibberd, J.M. (2011) Light-regulated and cell-specific methylation of the maize PEPC promoter. Journal of Experimental Botany 63(3): 1381–1390. Towbin, H., Staehelin, T., Gordon J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proceedings of the National Academy of Science 76(9): 4350–4354. Townsend, J.A., Thomas, L.A. (1994) Factors which influence the Agrobacterium-mediated transformation of soybean. Keystone Symposium on Molecular and Cellular Biology. Journal of Cell Biochemistry Supplement 18A. 139 Uhrig, J.F., Canto, T., Marshall, D., MacFarlane, S.A. (2004) Relocalization of nuclear ALY proteins to the cytoplasm by the tomato bushy stunt virus P19 pathogenicity protein. Plant Physiology 135(4): 2411– 2423. van Leeuwen, W. (2001) Characterization of position-induced spatial and temporal regulation of transgene promoter activity in plants. Journal of Experimental Botany 52(358): 949–959. Vargason, J.M., Szittya, G., Burgyán, J., Hall, T.M.T. (2003) Size selective recognition of siRNA by an RNA silencing suppressor. Cell 115(7): 799–811. Vidal, J., Chollet, R. (1997) Regulatory phosphorylation of C4 PEP carboxylase. Trends in Plant Science 2(6): 230–237. Viret, J. (1994) Transcriptional photoregulation of cell-type-preferred expression of maize rbcS-m3: 3’ and 5' Sequences are involved. Proceedings of the National Academy of Sciences, 91(18): 8577–8581. Virgin, H.I. (1955) Protochlorophyll formation and greening in etiolated Barley leaves. Physiologica Plantarum 8(3): 630 – 643.dx Viswanathan, S.R., Daley, G.Q., Gregory, R.I. (2008) Selective blockade of microRNA processing by Lin28. Science 320: 97–100. Voinnet, O., Rivas, S., Mestre, P., Baulcombe, D. (2003) An enhanced transient expression system in plants based on suppression of gene silencing by the p19 protein of tomato bushy stunt virus. The Plant Journal 33(5): 949–956. Voinnet, O. (2005) Non-cell autonomous RNA silencing. FEBS Letters 579: 5858–5871. Voznesenskaya, E.V., Franceschi, R., Artyushea, E.G., Black, C.C., Pyankov, I., Edwards, G.E. (2003) Development of the C4 photosynthetic apparatus in cotyledons and leaves of Salsola richteri (Chenopodiaceae). International Journal of Plant Sciences 164(4):471–487. Voznesenskaya, E.V., Koteyeva, N.K., Chuong, S.D.X., Akhani, H., Edwards, G.E., Franceschi, V.R. (2005) Differentiation of cellular and biochemical features of the single-cell C4 syndrome during leaf development in Bienertia cycloptera (Chenopodiaceae). American journal of Botany 92(11): 1784–1795. Voznesenskaya, E.V., Koteyeva, N.K., Chuong, S.D.X., Ivanova, A N., Barroca, J., Craven, L.A., Edwards, G.E. (2007) Physiological, anatomical and biochemical characterisation of photosynthetic types in genus Cleome (Cleomaceae). Functional Plant Biology 34(4): 247. Wallace, J.S. (2000) Increasing agricultural water use efficiency to meet future food production. Agriculture, Ecosystems & Environment 82: 105–119 Wang, J.L., Long, J.J., Hotchkiss, T., Berry, J.O. (1993a) C4 photosynthetic gene expression in light- and dark-grown Amaranth cotyledons. Plant Physiology 102(4), 1085–1093. Wang, J.L., Turgeon, R., Carr, J.P., Berry, J.O. (1993b) Carbon Sink-to-Source Transition Is Coordinated with Establishment of Cell-Specific Gene Expression in a C4 Plant. The Plant Cell 5(3): 289–296. Wang, J.L., Klessig, D.F., Berry J.O. (1992) Regulation of C4 gene expression in developing Amaranth leaves. The Plant Cell 4: 173–184. Wang, X., Zhang, J., Li, F., Gu, J., He, T., Zhang, X., Li, Y. (2005) MicroRNA identification based on sequence and structure alignment. Bioinformatics 21(18): 3610–3614. 140 Wang, Y., Nii, N. (2002) Anatomical development and biochemical alteration in ribulose bisphosphate carboxylase-oxygenase accumulation during leaf enlargement in Amaranthus tricolor. Journal of the Japanese Society for Horticultural Science 70(6): 675–681. Waters, M.T., Wang, P., Korkaric, M., Capper, R.G., Saunders, N.J., Langdale, J.A. (2009) GLK transcription factors coordinate expression of the photosynthetic apparatus in Arabidopsis. The Plant Cell 21(4): 1109–1128. Weber, A.P.M., von Caemmerer, S. (2010) Plastid transport and metabolism of C3 and C4 plants- comparative analysis and possible biotechnological exploitation. Current opinion in Plant Biology 13(3): 257–265. Weijers, D., Jürgens G. (2005) Auxin and embryo axis formation: the ends in sight? Current Opinion in Plant Biology 8(1): 32–37. Wu, C., Palopoli, M.F. (1994) Genetics of postmating reproductive isolation in animals. Annual Review of Genetics 27: 283–308. Wu, X.M., Liu, M.Y., Xu, Q., Guo, W.W. (2010) Identification and characterization of microRNAs from citrus expressed sequence tags. Tree Genetics & Genomes 7(1): 117–133. Xie, Z., Kasschau, K.D., & Carrington, J.C. (2003) Negative feedback regulation of Dicer-Like1 in Arabidopsis by microRNA-Guided mRNA degradation. Current Biology 13(9): 784–789. Yanagisawa, S. (2000) Dof1 and Dof2 transcription factors are associated with expression of multiple genes involved in carbon metabolism in maize. The Plant Journal  21(3): 281–288. Yang, Z., Ebright, Y.W., Yu, B., Chen, X. (2006) HEN1 recognizes 21-24 nt small RNA duplexes and deposits a methyl group onto the 2’ OH of the 3' terminal nucleotide. Nucleic acids research 34(2): 667– 675. Zenkteler, M. (1990) In vitro fertilization and wide hybridization in higher plants. Critical Reviews in Plant Sciences 9(3): 267–279. Zhang, G.Q., Zhou, W.J., Gu, H.H., Song, W.J., Momoh, E.J.J. (2003) Plant regeneration from the hybridization of Brassica juncea and B. napus through Eembryo culture. Journal of Agronomy and Crop Science 189(5): 347–350. Zhang, Y. (2005) miRU: an automated plant miRNA target prediction server. Nucleic acids Research 33: W701–704. Zhou, X., Wang, G., Zhang, W. (2007) UV-B responsive microRNA genes in Arabidopsis thaliana. Molecular Systems Biology 3: 103. Appendix 1: Primer List 141 Appendix 1: Primer List The sequences of primers used in this study are tabulated below, along with a brief explanation of their purpose. The origin of each primer is given, adhering to the following code: KK: Kaisa Kajala (of Bräutigam et al., 2011), NB: Naomi Brown, HA: Holly Astley, EG: E. Gage, AM: Ambion, Inc., IN: Invitrogen, Inc. Primer Sequence Purpose Origin ACT7 F TCCGACCCGATGTGATGTTATGGT qRT-PCR of ACTIN7 KK ACT7 R CAATCACTTTCCGGCTGCAACCAA qRT-PCR of ACTIN7 KK CA4 F AGACCACTCCTTCCGTTGCTTTCT qRT-PCR of CA4 KK CA4 R CACGAACACATCTCTGTGGTGTTAGGAC qRT-PCR of CA4 KK LHCA1 F GACATCATCATCCCTATTTCCCTTTCTCCC qRT-PCR of LHCA1 KK LHCA1 R AGTGATCAAATTACAGAGAGAGGAGGGCT qRT-PCR of LHCA1 KK NADME2 F AGGATCGTGAAGGATGTTGAGGCT qRT-PCR of NADME2 KK NADME2 R TTCCTGAATTCCGGTATGGCGTCT qRT-PCR of NADME2 KK PEPC F TCGACATTGCCCTCATCGGATTCT qRT-PCR of PEPC KK PEPC R AATCCAAGAACCTTCTCCGGGTCA qRT-PCR of PEPC KK PPCK F CTTGTCAATGGTCTTGCAGGCGAA qRT-PCR of PPCK KK PPCK R CCCAAACACTTCACAACAGCAGCA qRT-PCR of PPCK KK PPDK F GCAGGACTTGATTATGTCTCTTGCTC qRT-PCR of PPDK KK PPDK R GGGACAGCCATCTAACATTCTTCC qRT-PCR of PPDK KK PYR F TGCGGCATTCGGTATTGGCTATTG qRT-PCR of PyT1 KK PYR R TAACGGCAGAAGGAACGGCTACAA qRT-PCR of PyT1 KK RbcS1A F TGGATTCGACAACTCCCGTCAAGT qRT-PCR of RbcS1A KK RbcS1A R TTACAGCCAGAAGGCCGTGTGATA qRT-PCR of RbcS1A KK RCA F AGAGAGTCCAACTGGCCGATAAGT qRT-PCR of RCA KK RCA R AACTGGCAGATTCACTTGCTGTGC qRT-PCR of RCA KK 35S F AAGCTTGCATGCCTGCAGGTCA Sequencing of TOPO Insert NB 35S R TCCTCTCCAAATGAAATGAAC Sequencing of TOPO Insert NB 5’ RACE Adapter GCUGAUGAAUGAACACUGCGUUUGC UGGCUUUGAUGAAA Adapter for 5’RACE AM 5’RACE Inner CGCGGATCCGAACACTGCGTTTGCTG GCTTTGATG Adapter primer for 5’RACE Inner PCR AM 5’RACE Outer GCTGATGGCGATGAATGAACACTG Adapter primer for 5’RACE Outer PCR AM Appendix 1: Primer List 142 Primer Sequence Purpose Origin ACT8 F CAGATGTGGATCGCAAAGGCA qRT-PCR of ACTIN8 - At HA ACT8 R AAACCCGGCTTGAGAAATGGTC qRT-PCR of ACTIN8 - At HA ASP2 F TGTACAGCGATTGGACCATTGAGC qRT-PCR of ASP2 EG ASP2 F TCACCAGGTGTGCCTTTAGCT qRT-PCR of ASP3 EG CA F GTCAGAGCCCAAAGTATCTGGTATTTGC qRT-PCR of CA KK CA R TCACAGCTTCCTTCTCACATTGGGTGC qRT-PCR of CA KK NADME F ATTGCAACAACATCGTCTCGACGGTG qRT-PCR of NAD-ME NB NADME R AATAGAAAACGCCTCTTGTGGAGTGC qRT-PCR of NAD-ME NB p19 F AACAGTGAACGTTGGGATGGAGGA qRT-PCR of p19 EG p19 R ACATCCGATCTGGTCGAAACCGAA qRT-PCR of p19 EG p19W F CACCATGGAACGAGCTATACAAGG Cloning of p19 Coding Sequence EG p19W R TTACTCGCTTTCGAAGGTCT Cloning of p19 Coding Sequence EG pOpp19 F AGAATGAACCGAAACCGGCGGTA Directional Sequencing of pOp::p19 EG pOpp19 R GGCAAGCTGTAGCAGTTCTTGCTTT Directional Sequencing of pOp::p19 EG PPCK F CTGGTCCTCTTTCTGAGGCTCAAACG qRT-PCR of PPCK KK PPCK R CCATGGATGCCTTAGTGATTGCTCTGC qRT-PCR of PPCK KK PPDK F2 CAAGCACGTGCGATTTTCCAAGCAGC qRT-PCR of PPDK KK PPDK R CACATATTCCAACCTTGAGGTTAGGTCG qRT-PCR of PPDK KK PPDK InA R CACCGAATTTTACCACTTTACCA Internal Primer for PPDK 5' RACE EG PPDK InB R AAAGTGGTGAGGATGGGTTG Internal Primer for PPDK 5' RACE EG PPDK InC R TGTAAAGTGGGTAAAATTCGGTG Internal Primer for PPDK 5' RACE EG PPDK InD R CCTTCCAAGCCTCTTCTCCT Internal Primer for PPDK 5' RACE EG PPDK InE R AAGCAAAACGTTCTCCACTTTT Internal Primer for PPDK 5' RACE EG PPDK InF R TCCTAGGTTAAGAACTGTATCCATC Internal Primer for PPDK 5' RACE EG PPDK InG R ATGCCGGGTATGATGGATACAGTTC Internal Primer for PPDK 5' RACE EG PPDK InH R GATATGTTCGGCAATGTTGTGATGATGG Internal Primer for PPDK 5' RACE EG PPDK InI R GAACTGTATCCATCATACCCGGCAT Internal Primer for PPDK 5' RACE EG PPDK InJ R CCATCACAACATTGCCGAACATATC Internal Primer for PPDK 5' RACE EG PPDK InK R CACTGCTATCTTCACCGCACCTTT Internal Primer for PPDK 5' RACE EG PPDK InL R CTGCCTTGATCGCCAATTCCAA Internal Primer for PPDK 5' RACE EG Appendix 1: Primer List 143 Primer Sequence Purpose Origin PPDK IRnA F GAGATGGCCAGCATAGGTCT Internal Primer for PPDK 5' RACE EG PPDK IRnB F CAACCCATCCTCACCACTTTTCTT Internal Primer for PPDK 5' RACE EG PYR F TCATGACAACATGTTCAACCATTGGAGC qRT-PCR of PyT1 KK PYR R CCACCGAGCGCCATACAGACAACGC qRT-PCR of PyT1 KK Spec F ATTCATTCAAGCCGACACCGCTTC Spectinomycin resistance (sequencing) EG Spec R CGTCGTGCACAACAATGGTGACTT Spectinomycin resistance (sequencing) EG T3 ATTAACCCTCACTAAAGGGA PCR screen of TOPO vectors IN T7 TAATACGACTCACTATAGGG PCR screen of TOPO vectors IN uidA TGATCGCGTCAGCGCCGTC Verification of uidA NB Appendix 2: sRNA Target List 144 Appendix 2: sRNA Target List The data of sequences used in the initial search of the sRNA library are given below. “Contig” refers to the sequence reference number in the EST database of Bräutigam et al. (2011). The EST identification is given, along with the number of sRNA hits against each sequence. Contig Blast Results Hits 5 RbcS1A 4 11 Rbcs1A 13 15 RbcS3B 4 16 PPDK 60 30 PEPC2 40 32 CA1 2 33 NAD(P)H:Plastoquinone Dehydrogenase subunit O 7 55 NDH-M (subunit NDH-M OF NAD(P)H:Plastoquinone Dehydrogenase complex) 0 63 Rbcs1A 7 90 RbcS1a 7 98 CA1 0 114 Rbcs1A 12 115 Rbcs1A 9 120 Rbcs1A 10 144 RbcS3B 4 148 MDH 4 164 Rbcs1A 18 166 CA 0 212 RCA 1 213 RCA 0 233 Starch Synthase 3 242 NADPH dehydrogenase/electron transporter, 3 528 PEPCK1 2 1078 NDH-N of NAD(P)H:plastoquinone dehydrogenase complex (Ndh complex) 0 1223 PEPCK1 3 1708 RCA 0 1963 NAD(P)H:Plastoquinone Dehydrogenase subunit O 6 2049 PDK (Pyruvate Dehydrogenase Kinase) 0 2052 MDH 1 2081 Bundle-sheath defective protein 2 family / bsd2 family 0 2085 Malate dehydrogenase (NAD), mitochondrial 0 2236 DIT1 (Dicarboxylate transporter; oxoglutarate:malate antiporter) 1 2459 (NADP-malic enzyme 2); malate dehydrogenase (oxaloacetate- decarboxylating) 0 2466 Pyruvate dexcarboxylase family protein 0 2488 RuBisCO Large subunit, putative 0 Appendix 2: sRNA Target List 145 Contig Blast Results Hits 2637 ASP3 (Asparatate aminotransferase 5) 1 2891 PPDK 2 2916 PEPC2 1 2924 Rbcs1A 8 2955 Rbcs1A 18 2957 PEPC1 6 2958 CA family protein 0 2964 Rbcs1A 12 2984 PEPC2 1 3005 RbcS1A 4 3058 PEPC1 6 3070 PPDK 5 3098 PEPC2 2 3099 PEPC2 2 3136 PEPC2 1 3138 Rbcs1A 12 3167 CA family protein 0 3212 ASP3 (Asparatate aminotransferase 1) 0 3213 ASP1 (Asparatate aminotransferase 1) 5 3214 ASP1 (Asparatate aminotransferase 1) 5 3221 RCA 4 3486 PEPCK2 1 4520 MDH 0 5745 RuBisCO Large subunit, putative 0 5822 MDH 0 7150 Pyruvate dexcarboxylase family protein 0 7482 (NADP-malic enzyme 2); malate dehydrogenase (oxaloacetate- decarboxylating) 0 7993 MDH 0 8199 PEPC1 0 8389 DIT2.1 (Dicarboxylate transporter; oxoglutarate:malate antiporter) 0 8584 (NADP-malic enzyme 4); malate dehydrogenase (oxaloacetate- decarboxylating) 1 8736 pyruvate decarboxylase 0 9100 ASP3 (Asparatate aminotransferase 5) 0 9469 DIT2.1 (Dicarboxylate transporter; oxoglutarate:malate antiporter) 0 9719 RBCS1A; ribulose-bisphosphate carboxylase 0 10016 DIT2.1 (Dicarboxylate transporter; oxoglutarate:malate antiporter) 1 10211 PEPKR1 (PEPC-related kinase1) 0 10422 ALAAT2 (Alanine transferase 2) 0 10609 ALAAT2 (Alanine transferase 2) 0 11028 RbcS3B 0 12043 MDH 0 Appendix 2: sRNA Target List 146 Contig Blast Results Hits 12274 MDH 0 13440 PEPCK1 0 13495 PEPC3 0 13682 ALAAT2 (Alanine aminotransferase 2) 0 13684 ALAAT2 (Alanine aminotransferase 2) 1 13685 ALAAT2 (Alanine aminotransferase 2) 0 13686 ALAAT1 (Alanine aminotransferase 1) 0 13689 ALAAT1 (Alanine aminotransferase 1) 0 13691 ALAAT2 (Alanine aminotransferase 2) 0 13692 ALAAT2 (Alanine aminotransferase 2) 0 13693 ALAAT2 (Alanine aminotransferase 2) 0 13812 Malate dehydrogenase, cytosolic, putative 0 14367 Pyruvate decarboxylase 0 14704 DIT2.1 (Dicarboxylate transporter; oxoglutarate:malate antiporter) 0 14714 PEPKR1 (PEPC-related kinase1) 0 14821 RCA 0 14941 Pyruvate dexcarboxylase family protein 0 14951 ASP3 (Asparatate aminotransferase 3) 0 15046 (NADP-malic enzyme 3); malate dehydrogenase (oxaloacetate- decarboxylating) 0 15107 ALAAT2 (Alanine aminotransferase 2) 0 15566 PEPC2 1 15934 RCA 0 16083 PEPC3 0 16651 PEPC2 0 16712 ASP3 (Asparatate aminotransferase 2) 0 17220 pyruvate decarboxylase family protein 0 Appendix 3: Plasmid Maps 147 Appendix 3: Plasmid Maps 35S::p19::nos (Voinnet et al., 2003) Appendix 3: Plasmid Maps 148 Empty pOpON2.1 Vector (Moore et al., 2006) pOp::p19 Expression Vector 149