Evidence That Aspartic Acid 301 Is a Critical Substrate-Contact Residue in the Active Site of Cytochrome P450 2D6*

Model building studies have intimated a role for aspartic acid 301 in the substrate binding of cytochrome P450 2D6 (CYP2D6). We have tested this hypothesis by generating a range of CYP2D6 mutants substituting a variety of amino acids at this site. The mutant proteins, which included substitution with a negatively charged glutamic acid residue or neutral asparagine, alanine, or glycine residues, were expressed in Saccharomyces cerevisiae . In addition, a mutant where aspartic acid 301 was deleted was also tested. All the mutants expressed approximately equivalent amounts of recombinant apoprotein and, apart from the alanine 301 and the aspartic acid 301 deletion mutants, gave carbon monoxide difference spectra of similar magnitude to the wild type. In the cases of the alanine and deletion mutants, the amount of holoprotein was significantly reduced or absent relative to the amount of apoprotein, indicating restricted heme incorporation. The glutamic acid mutant was shown to have similar catalytic properties to the wild type enzyme toward the

Model building studies have intimated a role for aspartic acid 301 in the substrate binding of cytochrome P450 2D6 (CYP2D6). We have tested this hypothesis by generating a range of CYP2D6 mutants substituting a variety of amino acids at this site. The mutant proteins, which included substitution with a negatively charged glutamic acid residue or neutral asparagine, alanine, or glycine residues, were expressed in Saccharomyces cerevisiae. In addition, a mutant where aspartic acid 301 was deleted was also tested. All the mutants expressed approximately equivalent amounts of recombinant apoprotein and, apart from the alanine 301 and the aspartic acid 301 deletion mutants, gave carbon monoxide difference spectra of similar magnitude to the wild type. In the cases of the alanine and deletion mutants, the amount of holoprotein was significantly reduced or absent relative to the amount of apoprotein, indicating restricted heme incorporation. The glutamic acid mutant was shown to have similar catalytic properties to the wild type enzyme toward the substrates debrisoquine and metoprolol; however, some differences in regioselectivity and ligand binding were observed. The mutants containing neutral amino acids at position 301 exhibited marked reductions in catalytic activity. At low substrate concentrations little, if any, activity toward debrisoquine and metoprolol was measured. However, at a higher substrate concentration (2 mM) some activity was observed (about 10 -20% of wild type levels). Consistent with the above findings, the debrisoquine-induced spin changes in the mutant proteins were mark-edly reduced. These data collectively demonstrate that aspartic acid 301 plays an important role in determining the substrate specificity and activity of CYP2D6 and provide experimental evidence supporting the role of this amino acid in forming an electrostatic interaction between the basic nitrogen atom in CYP2D6 substrates and the carboxylate group of aspartic acid 301.
Cytochrome P450 2D6 (CYP2D6) 1 mediates the metabolism of over 30 drugs of wide therapeutic use including many antiarrhythmics, antidepressants, ␤-adrenergic antagonists, neuroleptics, and analgesics (1). Although structurally diverse, all known ligands (substrates and inhibitors) of CYP2D6 possess a basic nitrogen, usually either an amine or a guanidino group, which is presumed to be protonated when the ligand is bound in the active site of the enzyme (2). Furthermore, substratetemplate models (3)(4)(5) have revealed that this basic nitrogen is normally located 5-7 Å from the site of oxidation in the substrate molecule. A pharmacophore based on competitive inhibitors of CYP2D6 (6) also complies with this model; the positively charged nitrogen of such inhibitors is distanced up to 7.5 Å from a flat hydrophobic region of the inhibitor. Based on this conformity, it has been proposed that the basic nitrogen of CYP2D6 substrates and inhibitors interacts with a negatively charged residue, such as aspartate or glutamate, in the active site of enzyme and that this electrostatic interaction facilitates binding and orientation of the ligand in the active site (3)(4)(5)(6)(7).
Recent computer-derived homology models of the active site of CYP2D6 (4,8,9), 2 based on alignment with the crystal structures of the bacterial P450s CYP101 (P450 cam ) or CYP102 (P450 BM-3 ) have identified aspartic acid 301 (Asp 301 ) as a candidate residue for the proposed electrostatic interaction with the ligand. Such homology models and a recent structure-based alignment (10) locate Asp 301 in the central region of the I-helix of CYP2D6. This region maps to one of the substrate-recognition sites (SRS4) identified by Gotoh (11) as being important in substrate binding in the CYP2 family of P450s; these substrate-recognition site regions may also be predictive across other P450 families (10).
In three bacterial P450s for which crystal structures are known (12)(13)(14), the central region of the I-helix is one of the most spatially conserved areas of the P450 core (10). In these P450s, it is located close to the heme moiety and runs across the distal face of the heme, completely or partially covering pyrrole ring B (10). Several residues in the central region of the I-helix have also been shown by mutation to play a role in substrate specificity and/or reaction kinetics (15)(16)(17). Recent work also indicates that residues in this region may play a role in the supply of catalytic protons via helix-associated solvent molecules (10).
In this report we provide experimental evidence, through the use of site-directed mutagenesis, that Asp 301 is of critical importance in the efficient oxidation of substrates by CYP2D6. The data are consistent with the proposal that Asp 301 forms an * This work was supported by Wellcome Trust Grant 038735. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  ion pair with the basic nitrogen of CYP2D6 ligands, facilitating binding and orientation in the active site.
Recombinant DNA Manipulation and Generation of Mutant CYP2D6 cDNAs-Manipulations leading to the generation and isolation of recombinant constructs were performed according to standard protocols as described by Sambrook et al. (21). Mutant CYP2D6 cDNAs were generated by site-directed mutagenesis using the dut Ϫ ung Ϫ E. coli strain CJ236 according to the method of Kunkel et al. (19). Point mutations at bp 903 C3 A, 901 G3 A, 902 A3 C, and 902 A3 G and a deletion of bp 901-903 were performed; when expressed these cDNAs give rise to the CYP2D6 mutants D301E, D301N, D301A, D301G, and D301␦, respectively. The mutated cDNAs were ligated into the BglII site of the yeast expression vector pMA91 and transformed into S. cerevisiae AH22 cells as described previously (20). All constructs were sequenced (Sequenase 2.0 kit, Amersham International, UK) before and after subcloning in pMA91 to confirm the presence of the desired mutation and the lack of any additional changes in the CYP2D6 sequence. Sequencing also confirmed that the ATG start codon of the CYP2D6 gene of each construct was distanced minimally (12 bp) from the end of the phosphoglycerate kinase promoter of pMA91.
Yeast Culture and Microsome Preparation-Transformation of S. cerevisiae was by electroporation (22). Yeast transformants were grown in batch culture in 1 liter of selective synthetic medium (0.67% w/v yeast minimal medium without amino acids, 0.04% w/v histidine, 3% w/v glucose) for 42 h (stationary phase) at 200 rpm and 30°C. After harvesting by centrifugation, the yeast cells were disrupted mechanically with glass beads (0.45-0.50 mm) at 4000 rpm for 40 s with liquid carbon dioxide in a cell homogenizer (B. Braun Medical Limited, Aylesbury, UK). The microsomal fraction was prepared by centrifugation of the homogenate at 15,000 ϫ g for 30 min at 4°C, followed by a further centrifugation at 100,000 ϫ g for 2 h at 4°C. The microsomal pellet was washed twice with microsomal buffer (0.1 M potassium phosphate buffer (pH 7.4) containing 0.65 M sorbitol and 0.1 mM EDTA), and aliquots were immediately flash-frozen in liquid nitrogen and stored at 80°C prior to use.
Immunoblotting and Spectrophotometric Measurements-The relative apoprotein content of microsomes was estimated by separation on a 10% SDS-polyacrylamide gel (23) and electrophoretic transfer to Hybond TM ECL nitrocellulose membrane (Amersham International, UK). Blots were probed with rabbit anti-human CYP2D6 polyclonal antibodies and visualized by ECL using a streptavidin-horseradish peroxide conjugate and luminol (Amersham International, UK). The holoprotein content of microsomes was quantified by carbon monoxidedifference spectroscopy according to the method of Omura and Sato (24). Values of apparent K d and ␦A max of debrisoquine and quinidine were determined from ligand-induced difference spectra as described by Jefcoate (25).

RESULTS
Yeast cells transformed with wild type and Asp 301 mutant forms of CYP2D6, with the exception of D301␦, expressed similar amounts of immunodetectable microsomal bound CYP2D6 apoprotein (Fig. 1). The CYP2D6 holoprotein (heme-containing) content of microsomes, as determined by carbon monoxidedifference spectroscopy, varied depending on the mutant (representative spectra are shown in Fig. 2). No holoprotein was detectable in microsomes from the D301␦ mutant, and a marked decrease (90%) was found in microsomes prepared from D301A (8 Ϯ 4 pmol/mg protein) relative to wild type (49 Ϯ 15 pmol/mg protein). The holoprotein content of microsomes prepared from the other mutants (D301E, 45 Ϯ 22 pmol/mg protein; D301N, 29 Ϯ 9 pmol/mg protein; D301G, 40 Ϯ 12 pmol/mg protein) was comparable with that of wild type (Fig.  2). (The preceding mean Ϯ S.D. P450 values were determined from at least 10 microsomal preparations in each case.) A Soret absorption maximum of 448 nm was observed with the carbon monoxide complex of wild type and D301E microsomes. This was shifted slightly to 450 nm with microsomes prepared from the D301N and D301G mutants and to 454 nm with the D301A mutant. Variable amounts of P420 were detectable in microsomes prepared from the D301A mutant only (Fig. 2).
Catalytic activity, as assessed by the oxidation of 250 M debrisoquine and 40 M (Ϯ)-metoprolol, was virtually absent (1-2% of wild type) in microsomes prepared from yeast cells expressing the mutant forms of CYP2D6, with the exception of D301E, which retained rates of activity comparable with that of the wild type (see Table II). However, the regioselective oxidation of (Ϯ)-metoprolol, as assessed by the ratio of formation of ODM and ␣-OH metabolites, was significantly different with microsomes prepared from the D301E mutant compared with the wild type (8.5:1 and 3.8:1, respectively; p Ͻ 0.005) ( Table I). An alteration in the regioselective oxidation of metoprolol was also apparent with the R-(ϩ)-and S-(Ϫ)-enantiomers (Table  II). In contrast, enantioselective oxidation was not altered by the substitution of Asp 301 with Glu (Table III). Thus, although O-demethylation was significantly R-enantioselective and ␣-hydroxylation showed a preference for S-(Ϫ)-metoprolol, the pattern and extent of this enantioselectivity were similar with microsomes prepared from wild type and D301E mutant (Table II).
When the substrate concentration of (Ϯ)-metoprolol was increased 50-fold from 40 M to 2 mM, significant formation of ODM and ␣-OH metabolites was observed in microsomes derived from all of the Asp 301 mutants, with the exception of the D301␦ (Table III). In contrast to the insignificant level of oxidation of metoprolol at 40 M, rates of formation of ODM and ␣-OH equivalent to 20 -22% and 10 -20% of wild type rates, respectively, were observed with microsomes derived from the D301N, D301A, and D301G mutants. Although highly variable, the regioselective oxidation of the substrate was also altered in the D301N and D301A mutants, while D301G retained the same regioselectivity as wild type (Table III). The regioselectivity of metoprolol oxidation by D301E at a concen-tration of 2 mM was the same as that determined at 40 M (Tables I and III). DISCUSSION The results demonstrate that Asp 301 is a critical residue in the catalytic function of CYP2D6. Substitution of this carboxylate residue with a similar functional moiety (Glu) did not influence the catalytic competence of the enzyme significantly, although a subtle change in the regioselective oxidation of metoprolol and a 10-fold reduction in quinidine binding was observed. In contrast, substitution of Asp 301 with neutral amino acids (Asn, Ala, Gly), differing in size and polarity, resulted in marked reductions in catalytic activity. While it appears that a negative charge at amino acid residue 301 is important for efficient CYP2D6 catalysis, the precise role of the negative charge cannot be ascertained from these results alone. Nevertheless, the data do support the proposal that the carboxylate anion of Asp 301 forms a charge pair with the positively charged substrate nitrogen and, by doing so, facilitates the binding and orientation of the ligand in the active site.
The ligand binding role of Asp 301 is substantiated by the observation that replacement of the negatively charged residue with a neutral side chain results in substantial decreases in the binding capacity of debrisoquine (loss of type I spectrum) and quinidine (1000-fold greater K d value). The requirement of a negative charge at position 301 for substrate binding can also be inferred from the catalytic data, which show that substantially higher concentrations of metoprolol were required to achieve significant catalytic activity in the mutants in which Asp 301 was replaced with a neutral residue. Although full kinetic analyses have not been conducted, these results are indicative of a decrease in the affinity of the enzyme for the substrate. In addition, a reduction in the V max value of the mutated enzyme cannot be precluded.
The altered regioselective oxidation of metoprolol in the D301E mutant compared with wild type suggests a slightly different orientation of the substrate in the active site of the enzyme. As no gross change in the integrity of the active site was apparent in this mutant (as evidenced by a normal Soret absorption maximum of 448 nm, good heme incorporation, and retention of catalytic activity), the altered regioselectivity could be due to a subtle difference in the location of the substrate oxidation sites relative to the (Fe-O) 3ϩ entity, as a consequence of the extension of the carboxylate residue by a methylene group. Implicit in such a rationale is an interaction between a carboxylate residue in the active site of the enzyme and a positive charge of the substrate molecule. Thus the observed alteration in the regioselective oxidation of metoprolol by the D301E mutant adds weight to the proposal that Asp 301 serves as a negatively charged substrate-contact residue in the active site of CYP2D6. The lack of effect of the D301E substitution on the enantioselective oxidation of metoprolol indicates that a residue(s) other than Asp 301 is a determinant of CYP2D6 chiral selectivity. A candidate amino acid residue for such a role is Ser-304. 2 When Asp 301 was replaced with a neutral residue (Asn, Ala, or Gly), the structural integrity of the active site was also perturbed to varying degrees (as seen by the slight shift in the Soret absorption maximum of the carbon monoxide complex and in the different extent of heme incorporation). Thus an alternative explanation for the requirement of an anionic residue at position 301 for substrate binding is that this amino acid helps to maintain the integrity of the active site and that in its absence the topography of the site is altered. The different effects of the Ala and Gly substitutions on heme incorporation are difficult to explain. However, while both residues can be classed as neutral, Gly with only a hydrogen atom as a side  Role of Asp 301 in CYP2D6 Catalysis 29057 chain can adopt a much wider range of main chain conformations than other residues and hence may accommodate and minimize potentially deleterious structural changes in the active site. The greater conformational flexibility of glycine is due to the lack of steric hindrance between its side chain atoms and the polypeptide main chain, thus allowing greater rotation around the C␣-CЈ and the N-C␣ bonds of a glycine residue (27).
In this context, it may be significant that the central region of the I-helix of CYP2D6, in which Asp 301 is located, comprises a sequence of non-polar residues (Leu 295 -Val 308 ) interspersed at regular intervals with polar residues (Arg 296 , Asp 301 , Ser 304 ). Modeling indicates that these polar residues all point in the direction of the proposed active site and potentially form one of its boundaries. 3 Thus, the introduction of a non-polar residue (Ala) into this polar zone may drastically perturb the local environment adjacent to the active site and directly or indirectly influence heme incorporation. This may not arise when Gly is the substituent due to the adaptable nature of this residue resulting in minimal conformational change of the active site. Koymans et al. (8) have proposed Asp 100 as an alternative carboxylate residue for the interaction with the basic nitrogen of CYP2D6 ligands. However, a recent homology model of the active site 2 was unable to rationalize the involvement of this residue because of its peripheral location in the site. Furthermore, site-directed mutagenesis studies of Asp 100 have confirmed that the substitution of this residue with neutral amino acids (Asn or Ala) does not adversely influence the catalytic competence of the enzyme. 4 In the absence of a substrate-bound crystal structure of CYP2D6, the precise role(s) of Asp 301 in CYP2D6 catalysis will remain unproven. Nevertheless, the present data strongly support the proposal that Asp 301 is a ligand-binding residue in the active site of CYP2D6, interacting via a charge pair with the positively charged nitrogen of CYP2D6 ligands.