Constructing a recombinant model of the human pyruvate dehydrogenase complex

The human pyruvate dehydrogenase complex (PDC) is a large macromolecular assembly involved in the oxidative decarboxylation of pyruvate yielding acetyl CoA as the end product of this reaction, which subsequently enters the tricarboxylic acid (TCA) cycle. PDC is composed of multiple copies of various enzyme subunits, termed E1, E2 and E3. Human PDC also contains an additional component, E3BP, which has evolved in order to bind E3 to the core of the complex. The individual components of human PDC have now been cloned and overexpressed in E. coli. A His-tag has been engineered into the N-terminus of each protein to facilitate the rapid purification of these subunits using affinity chromatography. With the exception of E1, an alpha2beta2 heterotetramer, all recombinant proteins are soluble and produced in high yield. By using antibodies specific only for lipoylated E2 and E3BP from PDC, and by assaying the catalytically active subunits for activity, it has been found that these proteins are correctly folded and have been produced in active form. The use of a detergent, N-lauroylsarcosine, was required to produce a soluble E1. However, this component is active, as determined by enzymatic assay, under these conditions. Gel filtration studies have shown that E2 and E3BP must be coexpressed in E. coli in order for them to assemble into the stable E2/E3BP core complex that is central to the structure and organisation of human PDC. When these two proteins are expressed individually and then mixed they cannot form a stable core assembly. This suggests that the association between E2 and E3BP occurs in a co-translational manner and presumably requires their initial association as folding intermediates. Circular dichroism and fluorescence studies have been employed to examine the stability of the independently expressed E3BP. These studies have shown that each domain of E3BP, the N-terminal lipoyl domain, subunit-binding domain and C-terminal inner domain, unfolds at discrete concentrations of GdmCl. This indicates that while each domain is capable of independent folding, they also unfold independently of one another. In the native complex it is unclear how many E3 dimers are associated with human PDC. Isothermal titration calorimetry was utilised in order to assess the stoichiometry of binding between the recombinant E3BP and E3. These studies were performed using both full-length E3BP and a truncated construct, which consists of the subunit-binding domain expressed as a GST-fusion protein. These results suggest that one E3 dimer binds to two E3BP subunits; thus there would be six E3 dimers present per complex. The association constant for these two proteins was in the nanomolar range, indicative of very tight binding as expected. The binding affinity of E3 to E2 was also assessed using this technique. Truncated constructs of E2, specifically the subunit-binding domain expressed as a GST-fusion protein and the E2 didomain, a His-tagged protein containing the lipoyl domain and the subunit-binding domain of E2, were utilised in these studies. It was found that while E2 preferentially binds E1, it has also retained a residual affinity for the E3 subunit. Binding between E2 and E3 is approx. 100-1000 fold weaker than that between E3 and E3BP. The results described here support previous findings from our laboratory, obtained using an alternative technique, surface plasmon resonance and provide a molecular basis as to why E3BP- deficient patients retain residual PDC activity. While pursuing the main aim of this research, to reconstitute a recombinant human pyruvate dehydrogenase complex in vitro, a number of findings have produced interesting results. The unexpected 2:1 stoichiometry determined for the interaction between E3 and E3BP suggests that E3 dimers may form a network of crossbridges linking pairs of E3BP monomers across the 12 faces of the core. Production of sufficient quantities of active E1 is required in order to investigate whether a similar 2:1 stoichiometry exists between the alpha2beta2 E1 heterotetramer and the E2 didomain. If this is indeed the case, this introduces a new level of structure into the human pyruvate dehydrogenase complex, which has not been recognised previously