Abstract
Common marmosets (Callithrix jacchus), small New World primates, are increasingly attracting attention as potentially useful animal models for drug development. However, characterization of cytochrome P450 (P450) 3A enzymes involved in the metabolism of a wide variety of drugs has not investigated in marmosets. In this study, sequence homology, tissue distribution, and enzymatic properties of marmoset P450 3A4 ortholog, 3A5 ortholog, and 3A90 were investigated. Marmoset P450 3A forms exhibited high amino acid sequence identities (88–90%) to the human and cynomolgus monkey P450 3A orthologs and evolutionary closeness to human and cynomolgus monkey P450 3A orthologs compared with other P450 3A enzymes. Among the five marmoset tissues examined, P450 3A4 ortholog mRNA was abundant in livers and small intestines where P450 3A4 ortholog proteins were immunologically detected. Three marmoset P450 3A proteins heterologously expressed in Escherichia coli membranes catalyzed midazolam 1′- and 4-hydroxylation, alprazolam 4-hydroxylation, nifedipine oxidation, and testosterone 6β-hydroxylation, similar to cynomolgus monkey and human P450 3A enzymes. Among the marmoset P450 3A enzymes, P450 3A4 ortholog effectively catalyzed midazolam 1′-hydroxylation, comparable to microsomes from marmoset livers and small intestines. Correlation analyses with 23 individual marmoset liver microsomes suggested contributions of P450 3A enzymes to 1′-hydroxylation of both midazolam (human P450 3A probe) and bufuralol (human P450 2D6 probe), similar to cynomolgus monkey P450 3A enzymes. These results indicated that marmoset P450 3A forms had functional characteristics roughly similar to cynomolgus monkeys and humans in terms of tissue expression patterns and catalytic activities, suggesting marmosets as suitable animal models for P450 3A–dependent drug metabolism.
Introduction
Common marmosets (Callithrix jacchus) have attracted increasing attention as a potentially useful nonhuman primate model in fields such as pharmacokinetics, toxicology, neuroscience, stem-cell research, immunology, and infectious disease because of their genetic closeness to humans, small body size (weighing 350–400 g on average), high reproductive efficiency (typically producing twins), early sexual maturity (reached within 18 months of age), and applicability of transgenic technologies (Orsi et al., 2011; Sasaki, 2015). Cynomolgus monkeys (Macaca fascicularis) are the widely used nonhuman primate species for pharmacokinetics and drug-safety testing in pharmaceutical companies.
Cytochrome P450s (P450s), the major drug-metabolizing enzymes comprising multiple subfamilies, catalyze the oxidative biotransformation of potentially toxic compounds, including drugs and new chemical compounds (Wrighton and Stevens, 1992). In humans, it has been reported that approximately 75% of the drugs on the market are cleared by P450s, and P450 3A enzymes are significantly involved in the metabolism of more than 50% of the drugs (Williams et al., 2004). In humans, the P450 3A subfamily consists of four members: P450 3A4, 3A5, 3A7, and 3A43 forms. Human P450 3A enzymes reportedly metabolize midazolam, alprazolam, nifedipine, and testosterone (Yamazaki et al., 2002; Ohtsuka et al., 2010). P450 3A4 and 3A5 mRNAs are highly expressed in livers, followed by small intestines (among 10 human tissues) (Nishimura et al., 2003), and their protein expression was also detected in livers (Yamazaki et al., 1995) and small intestines (Paine et al., 2006). P450 3A4 protein expression was approximately 10-fold higher than that of P450 3A5 in human livers (Yamaori et al., 2004, 2005; Wang et al., 2008).
Marmoset P450 3A forms identified to date are P450 3A4 ortholog (formerly 3A21), 3A5 ortholog, and 3A90 (http://drnelson.uthsc.edu/CytochromeP450.html). Our previous study showed that marmoset P450 3A4 ortholog and 3A90 enzymes effectively catalyzed midazolam 1′-hydroxylation, similar to human and cynomolgus monkey P450 3As (Uehara et al., 2016a). Marmoset and cynomolgus monkey P450 3A4 ortholog also catalyzed omeprazole 5-hydroxylation and sulfoxidation reactions with high capacity (Uehara et al., 2016b). Quantitative analysis of gene expression for common marmoset transcriptomes indicated that P450 3A4 ortholog and 3A5 ortholog/3A90 mRNAs were expressed in livers and small intestines, similar to human P450 3A mRNA (Shimizu et al., 2014). P450 3A4 and 3A5 ortholog-like proteins were detected in marmoset livers (Schulz et al., 2001). Despite the potential importance as a nonhuman primate model in drug-metabolism and toxicological research, the molecular characteristics of marmoset P450 3A forms has not been analyzed in detail.
More than 20 marmoset P450 cDNAs have been identified so far; these P450s have high sequence similarities (>85%) to their orthologous human P450s (Uno et al., 2016). Overall, substrate specificity and tissue expression of orthologous P450 enzymes are similar between marmosets and humans, except for some enzymes belonging to the P450 2 family. In this study, gene cluster organization, sequence similarity, tissue distribution, and enzymatic properties of marmoset P450 3A4 ortholog, 3A5 ortholog, and 3A90 were investigated. This work is of importance for understanding the metabolic characteristics of marmosets as animal models in drug development.
Materials and Methods
Chemicals and Enzymes.
Alprazolam, midazolam, and testosterone were purchased from Wako Pure Chemicals (Osaka, Japan). 4-Hydroxyalprazolam was purchased from Enzo Life Sciences (Farmingdale, NY). Nifedipine, oxidized nifedipine, 1′-hydroxymidazolam, 4-hydroxymidazolam, and 6β-hydroxytestosterone were purchased from Sigma-Aldrich (Tokyo, Japan). Bufuralol and 1′-hydroxybufuralol were purchased from Toronto Research Chemicals (Toronto, Canada). Oligonucleotides were synthesized at Sigma Genosys (Ishikari, Japan). Pooled liver microsomes from marmosets (five males, sexually mature) and humans (74 males and 76 females, 18–82 years old) were purchased from Corning Life Sciences (Woburn, MA). Pooled liver microsomes from cynomolgus monkeys (eight males, 3–8 years old) and pooled intestine microsomes from cynomolgus monkeys (15 males, 2–5 years old) and humans (four males and six females, 14–65 years old) were purchased from Xenotech (Lenexa, KS). Pooled microsomes of brains, lungs, livers, kidneys, and small intestines were prepared from tissue samples of 20 marmosets (10 males and 10 females, >2 years old) raised at the Central Institution for Experimental Animals (Kawasaki, Japan) as described previously (Uehara et al., 2016c). Individual liver microsomes were prepared from 23 marmosets (14 males and nine females, >2 years old). This study was reviewed and approved by the Institutional Animal Care and Use Committee (Central Institution for Experimental Animals). Anti-human P450 3A4 antibodies (WB-3A4) and anti-human P450 3A5 antibodies (WB-3A5) were purchased from Corning Life Sciences. Anti-human P450 2D6 antibodies were purchased from Nosan Corporation (Yokohama, Japan). Anti-human protein disulfide isomerase antibodies (H-160) and goat anti-rabbit IgG-horseradish peroxidase were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). All other regents used were the highest quality commercially available.
Bioinformatics.
The structure of the marmoset P450 3A gene cluster was analyzed by BLAT (UCSC Genome Bioinformatics, University of California, Santa Cruz, CA) and by GGGenome (DNA Data Bank of Japan, National Institute of Genetics, Mishima, Japan). The amino acid sequence similarity of marmoset P450 3A forms compared with other species P450 3A members was determined by BLAST (National Center for Biotechnology Information, Bethesda, MD). The amino acid sequence alignment was performed using the Genetyx system (Software Development, Tokyo, Japan). The phylogenetic tree was constructed with the aligned sequences by the neighbor-joining method using DNASIS Pro (Hitachi Software, Tokyo, Japan). P450 amino acid sequences used were from GenBank: human P450 3A4 (NP_059488), 3A5 (NP_000768), 3A7 (NP_000756), 3A43 (NP_073731), and 2D6 (NP_000097); chimpanzee P450 3A4 (NP_001116247), 3A5 (NP_001087246), and 3A7 (NP_001087243); orangutan P450 3A43 (ABU85093) and 3A67 (ABU85096); cynomolgus monkey P450 3A4 (NP_001271463), 3A5 (NP_001306440), 3A7 (NP_001306436), and 3A43 (NP_001306434); rhesus monkey P450 3A4 (NP_001035504), 3A5 (NP_001035309), and 3A7 (NP_001182687); marmoset P450 3A4 ortholog (NP_001191369), 3A5 ortholog (NP_001191371), and 3A90 (NP_001191372); dog P450 3A12 (NP_001003340) and 3A26 (NP_001003338); pig P450 3A22 (NP_001182438), 3A29 (NP_999588), 3A39 (NP_999587), and 3A46 (NP_001128296); rabbit P450 3A6 (NP_001164739); guinea pig P450 3A14 (NP_001166587), 3A15 (NP_001166588), and 3A17 (NP_001166540); rat P450 3A2 (NP_695224), 3A9 (NP_671739), 3A18 (NP_665725), 3A23 (NP_037237), and 3A62 (NP_001019403); and mouse P450 3A11 (NP_031844), 3A13 (NP_031845), 3A16 (NP_031846), 3A25 (NP_062766), 3A41 (NP_001098629), 3A44 (NP_796354), 3A57 (NP_001093650), and 3A59 (NP_001098630).
Quantitative Reverse-Transcription Polymerase Chain Reaction.
The P450 3A mRNA distribution in marmoset tissues was analyzed by real-time reverse-transcription (RT) polymerase chain reaction (PCR) as described previously (Uehara et al., 2016d). In brief, total RNAs were extracted from brains, lungs, livers, kidneys, and small intestines, each pooled from 12 adult marmosets (six males and six females, >2 years old), using an RNeasy Mini Kit (Qiagen, Valencia, CA) and were used to synthesize cDNA using SuperScript III RT reverse transcriptase (Invitrogen, Carlsbad, CA) with random hexamers in a 20-µl reaction. Quantitative RT-PCR was performed with a SYBR Green–based detection system using gene-specific primers: 5′-GCTTTTGGAAGTTTGACATGGA-3′ and 5′-CAGGCTGTCGACCATCATAAATC-3′ for marmoset P450 3A4 ortholog, 5′- GTGAAGAAGTTCCTAAAATTTGATTTCC-3′ and 5′- GGGGTAAGGAACGGGAAGAA-3′ for marmoset P450 3A5 ortholog, and 5′- CCTAAAATTTGATGTATTAGCTCCACTG-3′ and 5′- GGATAAGGAACGGAAAGAGTACTACTGA-3′ for marmoset P450 3A90. The reaction mixture contained 400 nM each primer, 12.5 µl of Power SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA), and 2 µl of template DNA in a total volume of 25 µl. The PCR conditions were as follows: an initial denaturation for 10 minutes at 95°C, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. Real-time PCR was performed with an ABI PRISM 7300 sequence detection system (Applied Biosystems). This study performed absolute quantification, displaying high PCR efficiency (>93%) comparable for three P450 genes with a high correlation coefficient (r > 0.99). Standard curves were created by absolute amounts (102–106 copies) with a 10-fold dilution series of purified PCR products of marmoset P450 3A cDNAs. The expression level of each P450 3A mRNA was normalized to the level of 18S ribosomal RNA measured using Eukaryotic 18S ribosomal RNA Endogenous Control (Applied Biosystems).
Heterologous Protein Expression in Escherichia coli.
Marmoset P450 3A4 ortholog, 3A5 ortholog, and 3A90 cDNAs were cloned from liver by RT-PCR as described previously (Uehara et al., 2015c). To enhance the efficiency of expression for recombinant marmoset P450 3A proteins, the N-terminal modification was performed for the expression plasmids constructed with the P450 3A cDNAs by PCR using the following forward and reverse primers: cjCYP3A4 (5exp1) 5′-CATATGGCTCTGTTATTAGCAGTTTTCCTGGTGCTTCTGTATCTA-3′ and cjCYP3A4 (3exp1) 5′-TCTAGATTAGGCTCCACTTACAGTCC-3′ for marmoset P450 3A4 ortholog, cjCYP3A5 (5exp1) 5′-CATATGGCTAAGAAAACGAGCTCTAAAGGTAAGCTTATTCCAGGACCCGCACCT-3′ and cjCYP3A5 (3exp1) 5′-TCTAGATTATTCTCCACTTAGGGTTC-3′ for marmoset P450 3A5 ortholog, and cjCYP3A4 (5exp1) 5′-CATATGGCTCTGTTATTAGCAGTTTTCCTGGTGCTTCTGTATCTA-3′ and cjCYP3A5 (3exp1) 5′-TCTAGATTATTCTCCACTTAGGGTTC-3′ for marmoset P450 3A90. PCR amplification was carried out for 30 cycles (denaturation at 98°C for 15 seconds, annealing at 60°C for 30 seconds, and extension at 68°C for 2 minutes) using KOD-Plus-Neo DNA polymerase (Toyobo, Osaka, Japan) with an ABI GeneAmp PCR System 2720 thermocycler (Applied Biosystems). PCR products were cloned into pGEM-T easy vectors (Promega, Madison, WI) and subsequently subcloned into pCW vectors using the restriction sites of the NdeI and XbaI sites (underlined). Recombinant P450 3A proteins were prepared in an Escherichia coli DH5α expression system, and the concentration of P450 and NADPH-P450 reductase in each membrane preparation was measured as described previously (Yamazaki et al., 2002). Recombinant proteins were produced on the yield of approximately 2 μM culture medium. Recombinant marmoset P450 2D6 and 2D8; recombinant cynomolgus monkey 3A4 and 3A5; and recombinant human P450 3A4, 3A5, and 2D6 were prepared as described previously (Yamazaki et al., 2002; Uno et al., 2010; Uehara et al., 2015a).
Western Blotting.
Recombinant P450 proteins (1.0 pmol) or liver microsomes (10 μg) were subjected to 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis, and then electrophoretically transferred to a polyvinylidene difluoride membrane (Merck Millipore, Billerica, MA). After blocking with 0.5% nonfat milk in Tris-buffered saline (50 mM Tris, 138 mM NaCl, 2.7 mM KCl) containing 0.05% Tween 20 (v/v) at room temperature for 30 minutes, the membrane was probed with anti-human P450 3A4 antibodies (1:2000), anti-human P450 3A5 antibodies (1:5000), anti-human P450 2D6 antibodies (1:10,000), or anti-human protein disulfide isomerase antibodies (1:200) at room temperature for 1 hour, and then with goat-anti-rabbit IgG antibodies (1:5000) at room temperature for 20 minutes. The antigen-antibody complex was visualized by an ECL Prime Western Blotting Detection System (GE Healthcare, Buckinghamshire, UK).
Enzyme Assay.
Midazolam 1′- and 4-hydroxylation, nifedipine oxidation, testosterone 6β-hydroxylation, and bufuralol 1′-hydroxylation activities by recombinant P450 proteins and tissue microsomes were measured as described previously (Yamazaki et al., 2002; Uehara et al., 2015b). For alprazolam hydroxylation, the incubation mixture consisted of 40 pmol/ml recombinant protein or 0.5 mg/ml microsomes (liver or small intestine), 200 μM alprazolam, an NADPH-generating system (0.25 mM NADP+, 2.5 mM glucose 6-phosphate, and 0.25 units/ml glucose 6-phosphate dehydrogenase), and 100 mM potassium phosphate buffer (pH 7.4) in a total volume of 0.25 ml. Reactions were started by adding the NADPH-generating system, and were performed with 100 rpm/min of shaking at 37°C for 10 minutes. Reactions were terminated by the addition of methanol (0.40 ml), and then centrifuged at 20,000g for 10 minutes. The resulting supernatant was analyzed by reversed-phase high-performance liquid chromatography (HPLC) using a Prominence-I LC-2030C HPLC system with a fluorescence detector (Shimadzu, Kyoto, Japan). HPLC analysis was performed on a C18 column (L-column2 ODS, 5 μm, 150 × 4.6 mm; Chemicals Evaluation and Research Institute, Tokyo, Japan) using isocratic elution by methanol/acetonitrile/10 mM potassium phosphate buffer (pH 7.4) (24:33:43, v/v/v) at a flow rate of 1.0 ml/min with monitoring of the absorbance at 220 nm. Metabolite concentrations were quantified based on standard curves prepared with reference standards. Kinetic parameters for midazolam 1′- and 4-hydroxylation and nifedipine oxidation were estimated from the fitted curves using Michaelis-Menten equations, substrate inhibition equations, or Hill equations (Emoto et al., 2001; Shimizu, et al., 2007; Okada, et al., 2009) using the KaleidaGraph program (Synergy Software, Reading, PA). Linear regression analysis was performed with Prism (Graphpad Software, La Jolla, CA).
Results
Determination of P450 3A Gene Cluster Structure and Amino Acid Sequence Identity in Marmosets.
The structure of the marmoset P450 3A gene cluster was determined using marmoset genomic sequence by BLAT. The marmoset P450 3A cluster was localized in marmoset chromosome 2 (12,136,811–12,255,660) (Fig. 1). Three marmoset P450 3A genes (P450 3A4 ortholog, 3A5 ortholog, and 3A90) form a cluster with a total length of 118,850 bp (containing two gaps) between ZSCAN25 and TRIM4 genes on the short arm of chromosome 2. No marmoset P450 3A genes had a one-to-one orthologous relationship to human or cynomolgus monkey P450 3A genes. A marmoset P450 3A43 ortholog was not found in the genome as analyzed by BLAT or GGGenome. The amino acid sequences of marmoset P450 3A4 ortholog, 3A5 ortholog, and 3A90 had six substrate recognition sites and a heme-binding site, similar to human and cynomolgus monkey P450 3A forms (Fig. 2). Amino acid sequences of marmoset P450 3A4 ortholog, 3A5 ortholog, and 3A90 showed high degrees of sequence identity (>88%) to those of human and cynomolgus monkey P450 3A forms (Table 1). Phylogenetic analysis using amino acid sequences of the P450 3A forms from 12 species showed that marmoset P450 3A forms were evolutionarily closer to the P450 3A forms of primates, including humans, chimpanzees, orangutans, rhesus monkeys, and cynomolgus monkeys, than those of dogs, pigs, guinea pigs, rats, and mice (Fig. 3).
P450 3A cluster in marmosets, cynomolgus monkeys, and humans. The structure of the P450 3A gene cluster in marmoset, cynomolgus monkey, and human genome was analyzed by BLAT. Three marmoset P450 3A genes were located adjacent to ZSCAN25 and TRIM4 in the genome region corresponding to human and cynomolgus P450 3A genes. This schematic diagram is not proportionate to actual size and distance on the chromosome.
Multiple alignment of P450 3A amino acid sequences from marmosets, cynomolgus monkeys, and humans. P450 3A amino acid sequences from marmosets, cynomolgus monkeys, and humans were aligned using Genetyx. Asterisks and dots under amino acid alignment indicate regions conserved and roughly conserved among the three species, respectively. Solid and broken lines indicate substrate recognition sites and heme-binding domain, respectively.
Similarity of amino acid sequences of three marmoset P450 3A forms compared with P450 3As from other species
A phylogenetic tree of P450 3A amino acid sequences in various species. Phylogenetic analysis was performed using P450 3A amino acid sequences of marmoset (cj), human (h), chimpanzee (chim), orangutan (ora), cynomolgus monkey (mf), rhesus monkey (mm), dog (d), pig (p), rabbit (rab), guinea pig (cp), rat (r), and mouse (m) by the neighbor-joining method. Human P450 2D6 was used as an outgroup. Three marmoset P450 3A forms are shown in bold. The scale bar indicates the evolutionary distance of 0.1 amino acid substitutions per site.
Tissue Distribution of Three P450 3A mRNAs and Proteins in Marmosets.
Expression levels of marmoset P450 3A mRNAs in brains, lungs, livers, kidneys, and small intestines were analyzed by real-time RT-PCR. All three marmoset P450 3A mRNAs were abundantly expressed in livers and small intestines among the five tissues examined (Fig. 4). The expression level of P450 3A4 ortholog mRNA was >5-fold (livers) and >3-fold (small intestines) higher than those of P450 3A5 ortholog and 3A90 mRNAs, respectively, indicating that P450 3A4 ortholog was the major P450 3A form in livers and small intestines, the organs responsible for drug metabolism. P450 3A5 ortholog mRNA was >7-fold higher in small intestines than livers, different from P450 3A4 ortholog and 3A90 mRNAs expressed in these tissues at comparable levels. Tissue distribution of P450 3A proteins in marmosets was investigated by Western blotting using anti-human P450 3A4 and 3A5 antibodies capable of detecting selectively recombinant marmoset P450 3A proteins (Fig. 5A). P450 3A4 ortholog and 3A5 ortholog/3A90 proteins were detected in marmoset livers and small intestines (Fig. 5B); abundant P450 3A4 ortholog protein was detected in marmoset livers. P450 3A4 ortholog and 3A5 ortholog/3A90 proteins were constitutively expressed in livers from five individual marmosets (Fig. 5C).
Tissue distribution of P450 3A mRNAs in five marmoset tissues. Expression levels of marmoset P450 3A4 ortholog, 3A5 ortholog, and 3A90 mRNAs were normalized to 18S ribosomal RNA levels in each tissue (a pool of 12 marmosets, six males and six females). For graphical presentation, P450 3A4 ortholog mRNA level was adjusted to 1, and the relative expression levels of other P450 3A mRNAs to P450 3A4 ortholog mRNA are shown. Data are the mean ± standard deviation values of triplicate determinations.
Immunoblots of marmoset P450 3A proteins in marmoset tissues. (A) Recombinant marmoset P450 3A proteins (1.0 pmol) were selectively detected by immunoblotting using anti-human P450 3A4 and 3A5 antibodies. Expression levels of P450 3A proteins in marmoset tissue microsomes (10 μg) (B) and individual liver microsomes (lanes 1, 2, and 5, males; lanes 3 and 4, females) (C) were investigated. Protein disulfide isomerase (PDI) was used as a loading control.
Enzymatic Activities of Marmoset P450 3A Proteins.
To assess the enzymatic function of marmoset P450 3A enzymes, drug oxidation activities by recombinant P450 3A proteins and tissue microsomes were measured using typical human P450 3A probe substrates (midazolam, alprazolam, nifedipine, and testosterone). Liver and small intestine microsomes from marmosets catalyzed midazolam 1′- and 4-hydroxylation, alprazolam 4-hydroxylation, nifedipine oxidation, and testosterone 6β-hydroxylation in a manner similar to those of humans and cynomolgus monkeys (Table 2); liver microsomes from marmoset and cynomolgus monkeys catalyzed these reactions more strongly compared with those from humans. All marmoset P450 3A4 ortholog, 3A5 ortholog, and 3A90 enzymes also metabolized above typical human P450 3A probe substrates. Catalytic activities of the P450 3A4 ortholog enzyme were the highest among the marmoset P450 3A enzymes analyzed. Bufuralol 1′-hydroxylation activities, a typical human P450 2D probe activity, were higher for marmoset P450 3A4 ortholog, 3A5 ortholog, and 3A90 than human P450 3A4 and 3A5, similar to cynomolgus monkey P450 3A4 and 3A5 reported previously (Emoto et al., 2011). Kinetic analyses indicated that marmoset liver microsomes more effectively catalyzed midazolam 1′-hydroxylation (Vmax/Km, 0.74 ml/min/mg protein) with substrate inhibition kinetic, compared with those of humans and cynomolgus monkeys (Vmax/Km, 0.38 and 0.33 ml/min/mg protein, respectively) (Fig. 6; Table 3). Marmoset P450 3A4 ortholog enzyme was the most effective catalytic enzyme (Vmax/Km, 14 ml/min/nmol) for midazolam 1′-hydroxylation with substrate inhibition kinetic among the marmoset, cynomolgus monkey, and human P450 3A enzymes, and showed a low Km value (2.5 μM) comparable to liver microsomes (8.1 μM) and small intestine microsomes (6.6 μM) from marmosets. Similarly, kinetic analyses indicated that marmoset liver microsomes effectively catalyzed nifedipine oxidation (Vmax/Km, 0.24 ml/min/mg protein) comparable to those from humans and cynomolgus monkeys (Vmax/Km, 0.25 and 0.35 ml/min/mg protein, respectively) (Fig. 6; Table 4). Marmoset P450 3A4 ortholog enzyme also had a high Vmax/Km (2.3 ml/min/nmol) for nifedipine oxidation and showed a low Km value (33 μM), roughly corresponding to those of liver microsomes (42 μM) and small intestine microsomes (84 μM) from marmosets. Vmax/Km values of marmoset P450 3A5 ortholog and 3A90 enzymes for midazolam 1′- and 4-hydroxylation and nifedipine oxidation were of different orders of magnitude compared with that of marmoset P450 3A4 ortholog.
Drug-metabolizing activities by recombinant P450 enzymes and tissue microsomes from marmosets, cynomolgus monkeys, and humans
Units of drug oxidation rates by tissue microsomes and recombinant P450 proteins are nmol/min/mg protein and nmol/min/nmol of P450, respectively. Each activity was measured at a substrate concentration of 10 μM (midazolam), 200 μM (alprazolam), 20 μM (nifedipine), 50 μM (testosterone), and 100 μM (bufuralol). Values represent the mean ± standard deviation from triplicate measurements.
Enzyme kinetics for midazolam hydroxylation and nifedipine oxidation by recombinant marmoset, cynomolgus monkey, and human P450 3A enzymes and liver and small intestine microsomes from marmosets, cynomolgus monkeys, and humans. Kinetic analyses were performed for midazolam 1′- (closed) and 4-hydroxylation (open) and nifedipine oxidation by liver (A and B) and small intestine (C and D) microsomes from marmosets (triangles), cynomolgus monkeys (squares), and humans (circles). Kinetic analyses were performed for 1′- (closed) and 4-midazolam hydroxylation (open) and nifedipine oxidation by recombinant P450 3A of marmosets (E and F) (P450 3A4 ortholog, circles; P450 3A5 ortholog, squares; P450 3A90, triangles), cynomolgus monkeys (G and H) (P450 3A4, circles; P450 3A5, squares), and humans (I and J) (P450 3A4, circles; P450 3A5, squares).
Kinetic parameters of midazolam 1′- and 4-hydroxylation by recombinant P450 3A enzymes and tissue microsomes from marmosets, cynomolgus monkeys, and humans
Kinetic parameters were determined by nonlinear regression analysis (mean ± standard error, n = 17 points of substrate concentrations of 0.5–625 μM) using the equation v = Vmax × [S]/(Km + [S]) for Michaelis-Menten or v = Vmax × [S]/(Km + [S] + [S]2/Ks) for substrate inhibition (Shimizu et al., 2007; Okada et al., 2009). Units of enzyme activities for tissue microsomes and recombinant P450 proteins are nmol/min/mg protein and nmol/min/nmol of P450, respectively. Units of Vmax/Km for tissue microsomes and recombinant P450 proteins are ml/min/nmol and ml/min/mg protein, respectively. The units of Km and Ks values are μM.
Kinetic parameters of nifedipine oxidation by recombinant P450 3A enzymes and tissue microsomes from marmosets, cynomolgus monkeys, and humans
Kinetic parameters were determined by nonlinear regression analysis (mean ± standard error, n = 12 points of substrate concentrations of 1.1–200 μM) using the equation v = Vmax × [S]/(Km + [S]) for Michaelis-Menten or v = Vmax × [S]n/(S50 n + [S]n) for Hill equation (Emoto et al., 2001; Okada et al., 2009). Units of enzyme activities for tissue microsomes and recombinant P450 proteins are nmol/min/mg protein and nmol/min/nmol of P450, respectively. Units of Vmax/Km for tissue microsomes and recombinant P450 proteins are ml/min/nmol and ml/min/mg protein, respectively. The units of Km and S50 values are μM.
Midazolam 1′-hydroxylation activities were significantly correlated with P450 3A4 ortholog contents (r = 0.76, p < 0.05; Fig. 7B), but not with P450 3A5 ortholog/3A90 contents (r = 0.23) in 23 individual marmoset liver microsomes (Fig. 7C). P450 3A4 ortholog, 3A5 ortholog/3A90, and sum of P450 3A contents in individual marmoset liver microsomes were 77–218 (average ± S.D., 150 ± 45), 42–112 (average ± S.D., 83 ± 18), and 137–321 (average ± S.D., 233 ± 57) pmol/mg protein, respectively. The total contribution of P450 3A4 to P450 3A protein was assumed to be 64 ± 7% (range 51–74%). These results indicated that P450 3A4 ortholog was the major hepatic P450 3A enzyme in marmosets, similar to humans (Westlind-Johnsson et al., 2003). Hydroxylation activities of midazolam and bufuralol were also observed (Fig. 7D). Moreover, bufuralol 1′-hydroxylation activities in marmoset liver microsomes were correlated with midazolam 1′-hydroxylation activities (r = 0.81, p < 0.01; Fig. 7D), marmoset P450 2D contents (r = 0.50, p < 0.05; Fig. 7E), and sum of P450 3A contents (r = 0.83, p < 0.01; Fig. 7F), although bufuralol 1′-hydroxylation by cDNA-expressed marmoset P450 3As was lower than cDNA-expressed marmoset P450 2D6 (a major hepatic P450 2D enzyme) (Table 2). When considered together with high levels of P450 3A proteins in livers, P450 3A enzymes might play another important role for bufuralol 1′-hydroxylation by marmoset livers.
Correlations between activities of midazolam (A-C) and bufuralol (D-F) 1′-hydroxylation and P450 3A and 2D contents in liver microsomes from 23 marmosets. P450 3A4 ortholog (B), P450 3A5 ortholog/90 (C), and P450 2D (E) contents were estimated based on the immunochemically determined data. The sum of P450 3A (P450 3A4 ortholog + P450 3A5 ortholog/3A90) (A and F) in 23 individual marmoset liver microsomes was also calculated. Midazolam and bufuralol 1′-hydroxylation activities in individual marmoset liver microsomes were measured in duplicate at substrate concentrations of 5.0 and 1.0 μM, respectively.
Discussion
In marmosets, three P450 3A genes have been identified (Qiu et al., 2008); however, the molecular characteristics have not been clarified. In this study, we investigated the sequence identity, tissue distribution, and enzymatic function of marmoset P450 3A forms. Marmoset P450 3A forms were highly identical (>88%) to the human and cynomolgus monkey P450 3A orthologs (Table 1). A phylogenetic tree created using amino acid sequences showed that marmoset P450 3A forms were clustered with P450 3A forms from other primates, different from P450 3A forms found in preclinical animal species, including dogs, pigs, and rodents (Fig. 3). The high similarity of P450 3A amino acid sequences between marmosets, cynomolgus monkeys, and humans suggested a possibility that the enzymatic function of P450 3A was conserved among these primate species.
Among the marmoset P450 3A enzymes, P450 3A4 ortholog most abundantly expressed in livers and small intestines effectively catalyzed the oxidation of midazolam and nifedipine (Tables 3 and 4), and showed a significant correlation between P450 3A4 ortholog contents and midazolam 1′-hydroxylation activities in 23 individual marmoset liver microsomes (r = 0.76, p < 0.05; Fig. 7B), suggesting that P450 3A4 ortholog was the major catalyst for P450 3A–dependent drug metabolism in livers and small intestines. Moreover, Km, Vmax, and Vmax/Km values for the oxidation of midazolam and nifedipine were similar between marmoset, cynomolgus monkey, and human P450 3A4 enzymes. By analysis of site-directed mutagenesis based on a three-dimensional homology model of human P450 3A4, Phe-108, Ser-119, Ile-120, Leu-211, Asp-214, Ile-301, Phe-304, Ala-305, Thr-309, Ala-370, and Leu-373 were identified as key residues for substrate binding and regioselectivity (He et al., 1997; Fowler et al., 2000, 2002; Khan et al., 2002). These amino acid residues on human P450 3A4 were completely shared with marmoset and cynomolgus monkey P450 3A4 ortholog, suggesting that the enzymatic function of P450 3A4 ortholog enzyme was highly conserved between marmosets, cynomolgus monkeys, and humans. Therefore, marmosets might be a suitable model for evaluating the P450 3A–dependent drug metabolism in preclinical studies.
Oxidation of bufuralol, a typical human P450 2D probe, was faster in marmoset livers than human livers (Table 2). Bufuralol 1′-hydroxylation activities were also higher for marmoset P450 3A enzymes than human P450 3A enzymes, although those activities by marmoset P450 2D6 enzymes (major P450 2D enzyme responsible for bufuralol 1′-hydroxylation in marmoset liver in terms of expression level and enzyme kinetics) were comparable to human P450 2D6 (Uehara et al., 2015a). Correlation analyses with 23 individual marmoset liver microsomes (Fig. 7) suggested contributions of P450 3A enzymes to 1′-hydroxylation of midazolam (human P450 3A probe) and bufuralol (human P450 2D6 probe). Similarly, cynomolgus monkey P450 3A4 and/or 3A5 enzymes were faster than human P450 3A enzymes in bufuralol 1′-hydroxylation and dextromethorphan O-dealkylation (Emoto et al., 2011; Selvakumar et al., 2014). Hence, P450 3A enzymes might account for the higher velocity of bufuralol 1′-hydroxylation in marmoset and cynomolgus monkey livers.
In addition to drugs, P450 enzymes have various important physiologic functions, including the metabolism of steroids, bile acids, vitamins, and prostaglandins (Nebert and Dalton, 2006). In humans, P450 3A4 plays major roles in testosterone 16β-, 6β-, and 2β-hydroxylation and progesterone 16α-, 6β-, and 2β-hydroxylation (Yamazaki and Shimada, 1997; Niwa et al., 2015). Progesterone metabolism in humans is most similar to that in cynomolgus monkeys and least similar to that in rats (Swinney, 1990). In this study, P450 3A4 ortholog, 3A5 ortholog, and 3A90 enzymes catalyzed testosterone 6β-hydroxylation, similar to human P450 3A enzymes (Table 2). To understand the physiologic similarity between marmosets and humans, further study is needed to investigate the role of P450 enzymes in the metabolism of various endogenous compounds.
In humans, P450 3A4 and 3A5 mRNAs are abundant in livers and small intestines among the 10 human tissues (Nishimura et al., 2003), and these proteins are expressed in human livers and small intestines (Gibbs et al., 1999; Paine et al., 2006). Marmoset P450 3A4 ortholog, 3A5 ortholog, and 3A90 mRNAs and proteins were also expressed in livers and small intestines among the five tissues (Figs. 4 and 5), suggesting a possibility that the basal transcriptional regulation of P450 3A genes was shared between marmosets and humans. Indeed, the sequences of the proximal promoter and the xenobiotic-responsive enhancer module of the marmoset P450 3A4 ortholog gene are highly identical (88%) to those of the human P450 3A4 gene (Koehler et al., 2006). Several putative transcriptional factor-binding sites conserved between marmoset P450 3A4 ortholog and human P450 3A4 promoter, including the CCAAT/enhancer-binding protein binding site in the proximal promoter and hepatic nuclear factor-4 binding site in the xenobiotic-responsive enhancer module, might play an important role in basal transcriptional regulation. The similarity of tissue expression patterns possibly accounted for by common transcriptional regulation in P450 3A genes suggests that the marmoset, again, would potentially be a suitable model for preclinical safety testing in relation to P450 3A enzymes.
The marmoset P450 3A cluster contained two genes highly identical to human P450 3A5 genes, P450 3A5 ortholog and 3A90 (Fig. 1). By phylogenomics analysis of primate P450 3A locus structure, marmoset P450 3A4 ortholog shared a common ancestry with catarrhine P450 3A4, whereas marmoset P450 3A5 ortholog and 3A90 shared a common ancestry with catarrhine P450 3A5, suggesting that P450 3A5 expanded in only New World monkeys, in contrast to the repeated duplication of P450 3A4-like genes (P450 3A4, 3A7, and 3A67) in catarrhines (Qiu et al., 2008). In comparison with the recombinant marmoset P450 3A4 ortholog, recombinant marmoset 3A5 ortholog and 3A90 moderately metabolized multiple typical human P450 3A probe substrates, such as midazolam, alprazolam, nifedipine, and testosterone (Table 2). It would be of great interest to compare the physiologic significance of P450 3A5 ortholog and 3A90 with that of P450 3A4 ortholog in marmosets by exhaustive analysis of substrate specificity using xenobiotics and endogenous compounds.
In conclusion, amino acid sequences of three marmoset P450 3A forms showed high sequence identities (>87%) with P450 3A forms of cynomolgus monkeys, great apes, and humans, and phylogenetically had a close relationship with the human counterparts. P450 3A4 ortholog mRNA was abundant in marmoset livers and small intestines among three P450 3A mRNAs; P450 3A4 ortholog and 3A5 ortholog/3A90 proteins were also detected in these organs contributing to drug metabolism. Recombinant P450 3A4 ortholog, 3A5 ortholog, and 3A90 enzymes prepared in E. coli membranes catalyzed typical human P450 3A probe substrates, suggesting that P450 3A function was highly conserved between marmosets and humans. The significant correlation relationship between P450 3A4 ortholog contents and midazolam 1′-hydroxylation activities in marmoset livers showed that the P450 3A4 ortholog enzyme greatly contributed to midazolam 1′-hydroxylation in marmoset liver microsomes. These results indicated that marmoset P450 3A forms have functional similarities with those of humans in terms of tissue expression and enzymatic properties.
Acknowledgments
The authors thank Drs. Norie Murayama and Makiko Shimizu for their technical help, and Lance Bell for advice on English writing.
Authorship Contributions
Participated in research design: Uehara, Uno, Yamazaki.
Conducted experiments: Uehara, Uno, Nakanishi, Ishii.
Contributed new reagents or analytic tools: Inoue Sasaki.
Performed data analysis: Uehara, Uno, Yamazaki.
Wrote or contributed to the writing of the manuscript: Uehara, Uno, Yamazaki.
Footnotes
- Received December 29, 2016.
- Accepted February 10, 2017.
↵1 S.U. and Y.U. contributed equally to this work.
This work resulted from “Construction of System for Spread of Primate Model Animals” under the Strategic Research Program for Brain Sciences of Japan Agency for Medical Research and Development. S.U. was also supported in part by the Japan Society for the Promotion of Science Grant-in-Aid for Young Scientists B [Grant 15K18934].
Abbreviations
- HPLC
- high-performance liquid chromatography
- P450
- cytochrome P450
- PCR
- polymerase chain reaction
- RT
- reverse transcription
- Copyright © 2017 by The American Society for Pharmacology and Experimental Therapeutics
