Introduction

Top Abstract Introduction Materials & Methods Results & Discussion References

The cytochrome P450 superfamily consists of multiple P450² gene products that play an important role in the detoxification and metabolic activation of a large number of endogenous and exogenous substances (Nelson, 1995). Members of the 2B subfamily have provided an excellent framework to study P450 structure/function because of their high sequence homology and distinct substrate specificities. Approaches that have been instrumental in the identification of substrate contact residues include the isolation and analysis of allelic variants with altered substrate specificity, the analysis of site-directed mutants constructed on the basis of sequence differences between closely related members of the same subfamily, or predictions based on computer homology modeling. Among the determinants of substrate specificity that have been identified by these means are residues 114, 206, 209, 302, 363, 367, 477, 478, and 480 (Kedzie et al., 1991; He et al., 1992; Aoyama et al., 1989; Halpert and He, 1993; He et al., 1994; He et al., 1995; Szklarz et al., 1995).

Many of the residues important in substrate specificity are also involved in the susceptibility of P450s to mechanism-based inactivation. Orientation of the inactivator within the enzyme active site is critical to determining whether the reactive intermediate will be formed and whether this intermediate will react with the heme or apoprotein to result in covalent modification and enzyme inactivation. Thus mechanism-based inactivators have proved to be valuable probes of P450 function by differentiating between closely related enzymes. In particular, chloramphenicol and related analogs are highly selective, as these compounds are able to distinguish between 2B1 enzymes with single amino acid substitutions at residue 478 (Kedzie et al., 1991; He et al., 1992; Halpert and He, 1993) or between 2B1 and 2B2 variants (Strobel and Halpert, 1997). Aryl acetylenic compounds have also been shown to be mechanism-based inactivators of P450 enzymes, resulting in modification of either the heme moiety or the apoprotein (Ortiz de Montellano and Kunze, 1980). Many of these compounds are also selective in their inactivation (Hopkins et al., 1992; Foroozesh et al., 1997; Roberts et al., 1996); for example, 9-ethynylphenanthrene inactivates 2B1, 2B4, 2B11, but not 1A1 (Hopkins et al., 1992).

The current study characterizes heterologously expressed 2B1 cDNA clones isolated from Wistar-Kyoto (WKY) rats by examination of their substrate specificity and susceptibility to mechanism-based inactivators. We have previously shown that microsomes from WKY rats exhibited approximately twofold-higher 16 androstenedione hydroxylase activity, compared with other phenobarbital-induced microsomal preparations (Kedzie et al., 1991). However, the molecular basis for this difference remained unclear. Since alterations in androstenedione hydroxylase activity may indicate the presence of an allelic variant of 2B1, we have isolated and heterologously expressed P450 2B1 from a WKY rat liver cDNA library. Characterization of the 2B1 WKY variant that contained a single Glu-322  Val alteration revealed that it did not significantly differ from wild-type 2B1 in its activities or susceptibility to inactivation. In addition, the affect of an alteration at residue 103 was examined. Although this residue is not predicted to be a substrate contact residue based on computer homology modeling, the Val-103  Ala substitution decreased the susceptibility of 2B1 to inactivation by 9-ethynylphenanthrene.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results & Discussion References

Materials. Primers were synthesized by the University of Arizona Macromolecular Structure Facility (Tucson, AZ) or by National Biosciences, Inc. (Plymouth, MN). Restriction endonucleases, Taq polymerase, media for bacterial growth, and the 3' RACE kit were purchased from Gibco-BRL (Grand Island, NY). The Sequenase 2.0 dideoxy sequencing kit was purchased from the United States Biochemical Corp. (Cleveland, OH). The E. coli strain Topp3 was purchased from Stratagene (La Jolla, CA). Androstenedione, NADPH, diauroyl-L-3-phosphatidyl choline, chloramphenicol, BioMax MR-1 film, and CHAPS were purchased from Sigma Chemical Co. (St. Louis, MO). Guanidinium thiocyanate was obtained from Fluka BioChemika (Ronkonkoma, NY). HEPES was purchased from Calbiochem Co. (La Jolla, CA). [4-¹⁴C]Androstenedione was purchased from DuPont-New England Nuclear (Boston, MA). TLC plates [silica gel, 250 mM, Si 250PA (19C)] were purchased from J.T. Baker, Inc. (Phillipsburg, NJ). Rat NADPH cytochrome P450 reductase was expressed in E. coli as described previously (Harlow and Halpert, 1997). N-(2-p-nitrophenethyl)chlorofluoroacetamide and 9-EPh were synthesized as described previously (Hopkins et al., 1992; Halpert et al., 1990; Hall et al., 1990). All other chemicals and supplies used were of the highest grade commercially available.

Isolation of P450 2B1 cDNA Clones. Two male WKY rats (180-220 g) were obtained from Harlan Sprague-Dawley (Indianapolis, IN). To induce hepatic P450s, rats were given 0.1% sodium phenobarbital (Mallinckrodt, Inc., St. Louis, MO) in the drinking water for 4 consecutive days. Rats were sacrificed and the livers were harvested. Hepatic mRNA isolation by the guanidinium thiocyanate method, cDNA library construction using the 3' RACE kit, and cloning of 2B1 cDNAs into pSE380 were performed as previously described (Kedzie et al., 1991; Strobel and Halpert, 1997). PCR amplification conditions were as previously described (Strobel and Halpert, 1997) with the following exceptions: Taq DNA polymerase was used instead of Deep Vent Polymerase, and annealing temperatures were 45°C for the first five cycles, followed by 55°C for the remaining 25 cycles. Two full-length 2B1 cDNA clones were sequenced by the University of Arizona Macromolecular Structure Facility or by double-stranded sequencing using Sequenase. For the construction of 2B1 E322V with the Val-103  Ala substitution, the NcoI/BamHI fragment encoding residues 1-165 from the cDNA clone 2B1 E322V/V103A/E424K (containing this substitution in addition to one at codon 424) was ligated into pSE2B1 E322V (encoding wild-type sequence at codons 103 and 424). Sequencing confirmed the presence of an alanine codon at position 103 and a glutamic acid codon at position 424.

Expression of P450 WKY 2B1 Clones in E. coli. The 2B1 cDNA clones were expressed in Topp3 cells as described previously (He et al., 1995; John et al., 1994). The total P450 content was measured by reduced CO spectrum (Omura and Sato, 1964), and expression levels of P450 were determined by measuring this spectra from sonicated whole-cell preparations. CHAPS-solubilized membrane preparations were performed as previously described (John et al., 1994).

Purification of Recombinant P450 2B1. Recombinant P450 2B1 enzymes were purified using the two-column protocol as previously described (Fang et al., 1997). In brief, 30 nmol of P450 from a CHAPS-solubilized membrane preparation was applied onto an equilibrated n-octylamino-Sepharose column. P450 fractions were eluted by buffer containing 0.1% Emulgen 913. After dialysis overnight, the pooled P450-containing fractions were applied to a hydroxyapatite column for removal of detergent. After elution, P450-containing fractions were dialyzed, aliquoted, and stored at 70°C until use. This procedure typically yields protein with a specific content of 13-14 nmol/mg (Fang et al., 1997). The purity of P450 2B1 was confirmed by sodium dodecyl sulfate/polyacrylamide gel electrophoresis (Laemmli, 1970).

Purification of WKY Rat Liver P450 2B1. The WKY rat liver P450 2B1 enzyme was purified from phenobarbital-induced liver microsomes as described previously (Halpert et al., 1982). In brief, microsomal P450 was solubilized with 0.6% sodium cholate, applied onto an n-octylamino-Sepharose column, and P450 enzymes were eluted with buffer containing 0.08% Lubrol PX. The eluted fractions were analyzed at 417 nm, pooled, dialyzed overnight, then applied onto a DEAE-Sephacel column. A linear gradient of 25 mM to 100 mM NaCl in buffer was used to elute and separate the P450 isozymes. The purity of eluted P450 2B1 was confirmed by sodium dodecyl sulfate/polyacrylamide gel electrophoresis.

Androstenedione Hydroxylase and 7-Ethoxycoumarin O-Deethylase Assays. Androstenedione hydroxylase activity was measured as described (He et al., 1995; John et al., 1994; Szklarz et al., 1996). The final 100 µl reaction mixture contained 10 pmol of purified recombinant P450, 20 pmol of reductase (Harlow and Halpert, 1997), 20 pmol of rat liver cytochrome b₅, 30 µg/ml of dilauroyl-L-3-phosphatidyl choline, 1 mM NADPH, 2.5 nmol of [¹⁴C]-androstenedione in 50 mM HEPES (pH 7.6), 15 mM MgCl₂, and 0.1 mM EDTA. After addition of 1 mM NADPH, reactions proceeded for 5 min. Hydroxylated metabolites of androstenedione were developed on TLC plates by two cycles of chromatography in ethyl acetate/chloroform (2:1, v/v). 7-Ethoxycoumarin O-deethylase activity was assayed as described previously (He et al., 1995; Grimm et al., 1994). After reconstitution, P450 (10 pmol) was incubated with substrate for 5 min at 37°C. The concentration of the product (7-hydroxycoumarin) was determined from its fluorescence with an excitation wavelength of 366 nm and emission wavelength of 454 nm (Miller and Halpert, 1986).

Mechanism-Based Inactivation. Inactivation studies were performed as described previously (Kedzie, 1991; He et al., 1992; He et al., 1995). 9-EPh was added from a 100 µM dimethylsulfoxide stock solution to a final concentration of 1 µM, and chloramphenicol and N-(2-p-nitrophenethyl)chlorofluoroacetamide were added from 5 mM and 25 mM methanol stock solutions to a final concentration of 50 and 250 µM, respectively. Reconstituted P450 samples were preincubated with or without inhibitor for 2 or 5 min at 37°C. Reactions were started by the addition of 1 mM NADPH and were allowed to proceed for various times, at which point 10 µl and 25 µl aliquots were removed and added to 190 µl or 75 µl of [¹⁴C]androstenedione, for experiments with 9-EPh and chloramphenicol, respectively. Aliquots were removed at either 20, 40, 60, 90, 120, and 180 sec, or 60, 90, 120, 180, 300 sec. One third of the reaction mixture was spotted onto a TLC plate, and the 16-OH androstenedione product was quantitated by scintillation counting. Rate constants for inactivation were calculated by linear regression analysis of the natural logarithm of the residual activity as a function of time. The extent of reversible (competitive) inhibition was estimated from the decrease in the extrapolated activity at zero preincubation time, compared with the methanol or dimethylsulfoxide control.

Mass Spectrometry Analysis. The mass spectrometry analysis of hepatic purified 2B1 from WKY rats was as previously described (Strobel and Halpert, 1997) except that 2 nmol of hepatic purified 2B1 protein was dialyzed against 0.2M ammonium bicarbonate (pH 8.0) prior to trypsin digestion. The tryptic fragments of interest were sequenced using positive ion electrospray tandem mass spectrometry with collision-induced dissociation, as previously described (Strobel and Halpert, 1997).

	Results and Discussion

Top Abstract Introduction Materials & Methods Results & Discussion References

Isolation and Characterization of the P450 2B1 E322V Variant. We previously observed higher androstenedione 16-hydroxylase activities in microsomes from WKY rats (Kedzie et al., 1991). As this activity is attributed to 2B1 (Wood et al., 1983), and 2B1 variants with altered catalytic activities have been previously isolated, we constructed a cDNA library and obtained P450 2B1 clones from WKY rats. When these 2B1 cDNA sequences were compared with the first reported 2B1 sequence (Fujii-Kuriyama et al., 1983), a silent mutation was found at position 354 (CAC to CAT), and a second mutation was found at position 322 (GAG to GTG), resulting in a change from a glutamic acid to a valine codon. This WKY 2B1 variant will be designated 2B1 E322V. The same E322V alteration has also been observed in 2B2 cDNA clones (Mizukami et al., 1983). To establish whether this alteration affected 2B1 activity, metabolism of ethoxycoumarin and androstenedione was examined. These two substrates have been previously used to assay 2B1 function (Kedzie et al., 1991; He et al., 1992; Halpert and He, 1993; He et al., 1994). For these experiments, 2B1 was obtained either from heterologous expression in E. coli, or from WKY rat liver. As seen in table 1, no significant differences were observed between the activities of either heterologously expressed 2B1 E322V or WKY hepatic purified 2B1, compared with wild-type 2B1. Consistant with this result, residue 322 is not within one of the predicted substrate recognition sites (Gotoh, 1992). Since the 16-androstenedione hydroxylase activity of either 2B1 E322V or WKY hepatic purified 2B1 was not significantly higher than that of wild-type 2B1, the basis for the increased activity observed with microsomes remains unknown.

Characterization of Additional 2B1 E322V Enzymes Containing a V103A Alteration. The predicted amino acid sequence of one of the cDNA clones obtained from the cDNA library contained two substitutions in addition to the E322V alteration: Val-103  Ala (GTG to GCG) and Glu-424  Lys (GAA to AAA). Partial sequence analysis of 4 clones from the same PCR reaction and 46 clones from an independent RT-PCR reaction failed to yield a second clone with an alanine codon at position 103. Thus we were unable to establish whether this clone represents a minor 2B1 variant or a PCR-generated mutant. However, as residue 103 is within a substrate recognition site (SRS-1) and differs between 2B1 and 2B6 (Yamano et al., 1989), it was of interest to determine whether this alteration would affect the catalytic activities of 2B1. In addition, alterations at this position of 2B1 have not been previously examined. The alteration at residue 424 would not be predicted to affect substrate specificity, as it does not lie within an SRS region. The cDNA clone containing the V103A, E322V, and E424K substitutions is designated 2B1 E322V/V103A/E424K. An additional clone was generated that contained only the V103A and E322V alterations, and this clone is designated 2B1 E322V/V103A.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Effect of preincubation with 1 µM 9-EPh on androstenedione 16-hydroxylase activity in P450 2B1 E322 and 2B1 E322V/V103A. Open circles, dimethylsulfoxide control (for 2B1 E322V/V103A); closed circles, 2B1 E322V/V103A; closed squares, 2B1 E322V. Reactions were initiated by addition of 1 mM NADPH. At the indicated times, 10-µl aliquots representing 10 pmol of P450 were removed and assayed. The lines shown were generated by linear regression analysis of the natural logarithm of the residual androstenedione hydroxylase activity as a function of time. Control values were all < 0.01 min1.

Mass Spectrometry. Mass spectrometry was used to determine whether 2B1 E322V/V103A/E424K could be detected in the WKY hepatic purified 2B1 preparation. After digestion with trypsin, peptides were separated by HPLC and analyzed by mass spectrometry. The tryptic peptide containing residue 103 has a predicted mass of 1187.4 for 2B1 E322V, and a predicted mass of 1159.4 for 2B1 E322V/V103A. Fragments with m/z ratios of 1188 and 594.5 were obtained, consistent with a tryptic fragment from 2B1 E322V containing residues 99-109. The presence of valine at position 103 as well as the rest of the sequence in this peptide was confirmed by tandem mass spectrometry (data not shown). However, no abundant ions with the predicted m/z ratio for this tryptic fragment from 2B1 E322V/V103A/E424K could be detected. Along with the inability to find an additional cDNA clone, these results suggest that the 2B1 E322V/V103A/E424K is not likely to be a predominant variant of WKY rats. A less likely alternative is that the 2B1 E322V/V103A/E424K protein may not copurify with 2B1, and thus would not be represented in these experiments.

Computer Modeling. Since the Val  Ala alteration at residue 103 influences inactivation by 9-EPh, location of this residue relative to the active site was examined in the 2B1 homology model (fig. 2). However, as 9-EPh is an estimated 10-11Å from residue 103, it is unlikely to contact the inactivator directly. Previous studies of 2B1 inactivation by 9-EPh have demonstrated that the peptide Phe297-Leu307 is the site of covalent modification (Roberts et al., 1995). This region is part of helix I, which is not in close approximation to residue 103. The V103A alteration may cause an increase in the size of the active site, thereby altering the orientation of 9-EPh and, thus, its ability to covalently modify the enzyme. This may be accomplished by altering the conformation of the sheet next to helix B', which contains the active site residues 363 and 367. In addition, helix B' is the most highly variable region among P450s and thus difficult to predict by modeling. Repositioning of this helix in the enzyme model may bring residue 103 closer to the active site and into a position more suggestive of a direct role on substrate orientation.

View larger version (73K): [in this window] [in a new window]   Fig. 2.   9-Ethynylphenanthrene docked into the active site of the P450 2B1 enzyme. Residue 103 of 2B1 is approximately 10-11 Å from 9-EPh.