Structure and Molecular Biology of the Pyruvate Dehydrogenase Complex from Bacillus stearothermophilus

Abstract

The pyruvate dehydrogenase multienzyme complex (PDHC) from the thennophilic, Gram-positive bacterium Bacillus stearothermophilus is assembled around a core of 60 copies of the dihydrolipoamide acetyltransferase (E2p) component, organized with icosahedral symmetry. The peripheral subunits, attached to this inner-core, are dihydrolipoamide dehydrogenase (E3) and pyruvate decarboxylase (Elp), the latter comprising two polypeptide chains, Elpa and Elpf3. The molemcular weights of the individual components as well as the morphology of the E2 particle resemble those from the eukaryotic PDHC. These characteristics make the B. stearothermophilus PDHC a good model for those 2-oxo acid dehydrogenase complexes bearing a different inner-core symmetry from the well-studied PDHC and 2-oxoglutarate dehydrogenase complex (20GDHC) from Escherichia coli, which have octahedral E2 cores. This thesis presents evidence that a similar multidomain structure to that found in the E2 chains of the PDHC and 20GDHC from E. coli occurs in the B. stearothermophilus E2p subunit, although significant differences are apparent. After limited proteolysis, a folded fragment containing 52 amino acid residues was identified as being responsible for the binding of the E3 subunits, although, unlike the corresponding region of the E2 chains of the E.coli complexes, the fragment was also responsible for binding Elp molecules. The amino acid sequence of this peptide, together with the sequences of some other proteolytic E2p-derivatives, could be combined to produce the first 211 residues of the amino acid sequence from the N-terminus of the E2p subunit. This provided the complete primary structure of the single lipoyl domain, the peripheral subunits (Elp/E3)-binding domain, and the regions of polypeptide chain, probably highly flexible in nature, that link the domains to each other and to the inner-core domain. It also permitted the identification of several proteolytically sensitive sites and more complete interpretation of the data currently available from 1H-n.m.r. spectroscopy of this complex. To characterize further the multidomain structure of the E2p chain and to determine the primary structure of the three remaining subunits of the B. stearothermophilus PDHC, cloning and nucleotide sequence analysis of the structural genes encoding the various components were attempted. Two different plasmids, pBst42 (in pUC13) and pBst27 (in pUC18), were isolated from libraries prepared with EcoRI-cut genomic B. stearothermophilus DNA, carrying the genes encoding the four PDHC subunits. Plasmid pBst42 was found to contain the structural genes for E 1 pa and E 1 pJ3 and the 5' end of the gene for the E2p subunit. Plasmid pBst27 contained the 3' end of the E2p gene and the entire gene for the E3 subunit. The genes coding for the PDHC components lie in close proximity to each other, with Elpa, Elpf3, E2fand E3 polarity, immediately downstream of a region showing possible multiple promoter sequences. All four genes are preceded by correctly placed ribosome binding sites which share homology with the 3' end of the 16S ribosomal RNA from B. stearothermophilus. The E2p gene is linked closely to the gene for the E3 component, the latter having no separate promoter and ending with a putative p-independent transcription terminator. The protein sequences deduced from the nucleotide sequences of the four structural genes were in agreement with N-terminal amino acid sequence information obtained from peptides isolated from the intact subunits and with published amino acid composition data. The deduced amino acid sequence of the E3 subunit was highly similar to its E. coli counterpart. The same was true of the E2p subunit, but the best overall homology was obtained when the Bacillus sequence was compared with that of the E2 component of the branched-chain dehydrogenase complex (BCDHC) from Pseudomonas putida. The Bacillus Elpa and ElpB subunits were poorly related to the E.coli Elp and Elo polypeptide chains (the El components of the PDHC and 20GDHC, respectively) but extensive regions of homology could be found with their counterparts of the human and P. putida BCDHC and the human PDHC. One of these regions encompassed phosphorylation sites 1 and 2 of the Ela subunits of the mammalian PDHC and BCDHC, although B. stearothermophilus PDHC does not appear to be regulated by phosphorylation. A common sequence motif, approximately 30 residues in length, was also found in all the Ela subunits sequenced so far, including the Bacillus Elpa chain, and detected in the amino acid sequence of a wide range of thiamin pyrophosphate (TPP)-utilising enzymes. This motif may constitute the cofactor binding site. Its occurrence in the Ela subunits is consistent with this possibility since the binding site for TPP is thought to reside in the Elpa and not the ElpB subunits of the mammalian PDHC. The Elpa and Elpf3 subunits of the B. stearothermophilus PDHC were found to be expressed in E.coli from the original pUC13 clone, pBst42. Expression of the thermostable E3 component was also achieved in E. coli, under the control of the Ipp promoter in the vector pBSTNAV. The E2p gene, split during the cloning procedures, was reconstructed from pBst42 and pBst27. Expression of the encoded polypeptide was obtained under the control of the tac promoter in pKK223-3, although the in vivo produced E2p seemed to be degraded. Availability of the structural genes for the B. stearothermophilus PDHC components opens the way for the study of the catalytic and structural properties of this icosahedral 2-oxo acid dehydrogenase complex using the techniques of protein engineering.