Plasmodesmata are nanoscale, membrane lined pores that span the cell wall of neighbouring plant cells. The presence of these structures enables cell-cell movement of substances important for plant growth and development. This thesis describes the nature of plasmodesmata and explores their importance at various biological scales within the model plant Arabidopsis thaliana. The first two experimental chapters focus on selected molecular components of individual plasmodesmata. I show that sphingolipids, a lipid class enriched in the membranes of these structures, have a functional role in shaping the pores. Sphingolipid alterations resulting from mutation in the PHLOEM UNLOADING MEDIATOR gene or in the LONG CHAIN BASE KINASE 1 gene result in plasmodesmata morphological changes with consequences on cell-cell permeability. Moving to the protein component of plasmodesmata, I highlight that the family of MULTIPLE C2 AND TRANSMEMBRANE REGION PROTEINS influences long distance movement in the phloem, a vascular conduit in plants. I also show that loss of selected members of the family causes severe developmental phenotypes. Additionally, my results suggest potential interactions between sphingolipids and these proteins. The following experimental chapters look at increasingly larger biological scales. Employing serial block electron microscopy, I develop computational tools to study the distribution and density of plasmodesmata over entire cellular interfaces, and the thickness of the wall they span. The proof of concept examples interestingly show differences in spatial patterns, depending on the interface, and suggest an influence of sphingolipids on cell wall thickness. I move to the tissue scale, focusing on movement in the phloem, and identify EMS-MUTAGENIZED BRI1 SUPPRESSOR 3 as a suppressor of restricted phloem unloading in lines that over-accumulate callose, a polysaccharide occluding plasmodesmata. The protein might specifically control correct folding of callose synthase enzymes within the endoplasmic reticulum. Lastly, looking at the whole plant level, I show that induced closure of plasmodesmata within a specific phloem vascular domain, that of SISTER OF APL, can have reproducible impacts on the rate of growth of axillary buds into branches. The effect is observed in a range of genetic backgrounds and might involve PIN-FORMED proteins.