This thesis describes the expression and purification of recombinant wildtype human hypoxanthine-guanine phosphoribosyltransferase (HPRT) together with several mutant forms generated by site-directed mutagenesis. The General Introduction provides a review of the literature on HPRT in the context of its enzyme family, the PRTases, its role in purine metabolism, human deficiency disease states and mutations detected. It also describes the substrate specificities of the 6-oxopurine enzymes from various sources including mammalian, bacterial and parasite and the possible use of HPRT as an anti-parasitic drug target.
Wildtype human HPRT was expressed in an E. coli expression system. The freshly prepared recombinant enzyme had a similar specific activity (25 U/mg) and substrate Km values to freshly purified erythrocyte enzyme isolated from human blood. The erythrocyte enzyme shows an increase in Km for PRib-PP as the specific activity decreases. For the recombinant protein, the Km value for the substrate PRib-PP was found to vary less predictably with specific activity on storage.
There are four cysteine residues in human HPRT at positions 23, 66, 106 and 206. Earlier chemical modification studies had suggested that cysteine residues may be important in human HPRT. The Km for PRib-PP was found to increase as cysteine residue C23 was modified by iodoacetate. Site-directed mutants of human HPRT with each cysteine residue replaced by alanine were therefore produced, purified and characterised. The four single cysteine mutants were all catalytically active, demonstrating that cysteine residues are not catalytically important despite being linked with substrate binding in chemical modification studies. The single mutant C66A was the most unstable of the four single mutants. This corroborates with a speculated requirement for C66 in mammalian HPRT, since mutations at C66 are the most frequent cysteine mutants in HPRT deficient cell lines. C206A was more stable than C66A, and C23A was more stable again, though both were still more unstable than wildtype HPRT. CI06A was the most stable of the four single mutants. All of the single mutant proteins demonstrated variation in the Km for PRib-PP on storage. A triple mutant with C23, C106 and C206 all changed to alanine was found to be stable in the presence of substrate and to show an invariant Km for PRib-PP which was similar to that of erythrocyte HPRT.
There was evidence for disulfide bond formation in recombinant HPRT. This is despite the cysteine residues being remote from each other in the crystal structure. The triple mutant showed no evidence of disulfide bonding. Analysis of the single mutants revealed that the intra-subunit disulfide bonding involved C206 in most cases and at least two of C23, C106 and C66. Inter-subunit disulfide bonding involved C206 and C66 predominantly. Inter-subunit disulfide bonding resulted in greater loss of activity than intra-subunit disulfide bonding. Whether these disulfides have any relevance in vivo is speculative.
The C23A mutant was resistant to iodoacetate inhibition, confirming that C23 modification is the source of inactivation. How C23 modification increases the Km for PRib-PP and how PRib-PP prevents modification with iodoacetate is not obvious as C23 is remote from the PRib-PP binding site in the crystal structure. A change in conformation of the protein must occur on PRib-PP binding, which prevents access to C23. This change must also be inhibited by carboxymethylation of C23. The conformation of the free enzyme must then be different to the PRib-PP or GMP bound forms seen in the crystal structures.
A mutation found in a Lesch-Nyhan patient, HPRTreading (C206Y), was reproduced in the recombinant system. Lesch-Nyhan syndrome is associated with very low levels of HPRT activity. The mutant protein produced was found to be catalytically active, with specific activity and substrate Km values similar to wildtype protein. However the protein was unstable on storage. Supematants of E. coli cell lysates after the expression of the mutant recombinant protein had only about 10% of the activity found in similar expressions of the wildtype protein. It is therefore probable that the Lesch-Nyhan syndrome in this case is due to the protein being unstable in the mammalian system.