Structure/function studies of 5a-reductase.

Abstract

This thesis reports structure-function assessments made using site-directed mutagenesis of the human enzyme 5alpha-reductase (5AR), an enzyme crucial for normal masculine development. These assessments utilised the differences between the two forms of the enzyme in affinity for testosterone and sensitivity to the competitive inhibitor Finasteride.Although first described in the 1950s, the enzyme has never been isolated and biochemical studies generated conflicting data. In some reports the enzyme was said to operate at a pH of 5.5 whereas others considered the pH optimum was more alkaline than this. Similarly, the affinity for testosterone was reported as being around 3muM and around 0.5muM and different sensitivities to the competitive inhibitor Finasteride also were reported. These differences were resolved in the period 1989 1991 by isolation of cDNAs coding for two forms of 5AR. These were termed 5AR1 and 5AR2. It is now known that 5AR1 operates at the more alkaline pH, has the lower affinity for testosterone and is the more resistant to Finasteride. 5AR1 is present to a greater extent in the periphery with 5AR2 predominating in the prostate.Isolation of the cDNAs coding for 5AR1 and 5AR2 and the differing characteristics of the two enzymes, have provided the means to investigate structure-function relationships using site-directed mutagenesis. This approach enables identification of residues important for the binding of substrate and inhibitor by the two forms of human 5AR.At the commencement of these studies the residues -AVFA- had been identified as comprising part of the substrate/inhibitor binding site of human 5AR1. As the analogous residues, -VSIV- of rat 5AR1 also had been shown to contribute to this binding, it had been suggested that the analogous residues -GALA- in human 5AR2 may form its substrate/inhibitor binding domain. There was however no experimental evidence to support this suggestion. Similarly the roles of particular residues identified as being involved in substrate/inhibitor binding by 5AR had not been subjected to significant study. Accordingly the experiments conducted were designed to address several unanswered questions. The most significant of these questions was are the residues -GALA- involved in substrate/inhibitor binding by human 5AR2 and if not what residues are in fact involved in this binding.Rat 5AR1 exhibits a different mechanism of Finasteride inhibition than does rat 5AR2and both of the human enzymes. This difference had been attributed to expression of a cysteine at position 146 in rat 5AR1 rather than the arginine expressed at the analogous position in rat 5AR2 and the human enzymes. The studies reported here for the mutation of the relevant arginine to cysteine do not substantiate the suggestion that cysteine 146 is responsible for the unique mechanism of inhibition by Finasteride shown by rat 5AR1.Although residues G34 and H231 in human 5AR2 had been shown important for substrate binding by this enzyme, the roles of these residues in inhibitor binding had not been examined nor had the roles of the analogous residues in human 5ARI (G39 and H236). Experiments designed to determine these factors demonstrated a requirement of both G34 and H231 for substrate and inhibitor binding by human 5AR2. The analogous residues in 5AR1 were however only required for the binding of inhibitor.The tetrapeptide -AVFA- had been shown to be involved in substrate/inhibitor binding by human 5AR1 but no assessment had been made of the relative roles of the individual amino acids comprising this tetrapeptide. Replacing residues in human 5AR1 with the corresponding residues from rat 5AR1 demonstrated a requirement in human 5AR1 of A26 and V27 for inhibitor but not substrate binding, possibly as a result of their conformational and electrostatic effects respectively. The branched chain residue V27 was found to be required for both substrate and inhibitor binding and the terminal alanine (A29) did not appear to play a significant role in the binding of either substrate or inhibitor. These studies also permitted an indirect assessment of the contribution of all four residues of the tetrapeptide -VSIV- to substrate binding by rat 5AR1 and showed that while the tripeptide -SIV- is important for inhibitor binding, the initial valine is not.As mentioned above the main impetus of these studies was to determine if residues -GALA- were involved in substrate/inhibitor binding by human 5AR2. Characterisation of mutants in which -GALA- in human 5AR2 was substituted with -AVFA- from 5AR1 and vice versa, showed these residues do not comprise the substrate/inhibitor binding site of human 5AR2.From studies of a chimeric constructs of human 5AR1 and 5AR2 residues involved in substrate/inhibitor binding by human 5AR2 were mapped to the first 40 amino acid residues. Mutations of 6 residue clusters showed the hexapeptide -ATLVAL(r15-20) in 5AR2 to contribute significantly to substrate/inhibitor binding. Further studies of this hexapeptide showed that residues -ATL- but not -VAL- form part of the substrate/inhibitor binding domain of human 5AR2.