By continuing to browse this site you agree to us using cookies as described in About Cookies
Wiley Online Library is migrating to a new platform powered by Atypon, the leading provider of scholarly publishing platforms. The new Wiley Online Library will be migrated over the weekend of March 17 & 18. You should not experience any issues or loss of access during this time. For more information, please visit our migration page: http://www.wileyactual.com/WOLMigration/
Author for correspondence: Mette T. Skaanild, Department of Veterinary Pathobiology, Ridebanevej 9, DK-1870 Frederiksberg C, Denmark (fax +45 35 35 35 14, e-mail email@example.com).
Abstract: CYP2A6 in man catalyzes the oxidation of nicotine-forming cotinine and 7-hydroxylation of coumarin, which is used as test substrate for CYP2A6 in man. Large interindividual differences are found in man and some are due to genetic polymorphism. The 7-hydroxylation of coumarin is present in pigs, and an inter-individual variation has been found that might be due to polymorphisms. To enable the finding of polymorphism in pigs, the minipig cDNA was sequenced. Two cDNAs were found and translated to a 494 and a 487 amino acid long protein, both cDNAs were found in all but one pig. The 494 a.a. protein showed high homology to the human and 100% homology to the conventional pig CYP2A19 protein. In the wild type protein, all 6 substrate recognition sites were found, whereas the short protein only contained the first 5 substrate recognition sites. SSCP analysis revealed 3 polymorphisms. In order to study the effect of these polymorphisms on enzyme activity, microsomes were incubated with nicotine and coumarin. The polymorphisms appeared to have no effect on either enzyme activity as the specific enzyme activity towards nicotine and coumarin were approximately the same for all pigs. The specificity of pig CYP2A was investigated and it was found that the formation of cotinine correlated with the immunochemical level of CYP2A as did the coumarin hydroxylation. Anti-human CYP2A inhibitory antibody inhibited coumarin 7-hydroxylation by about 90% and formation of cotinine by 44–60% and 85–100% at substrate concentrations of 500 μM and 50 μM respectively, showing that coumarin and nicotine (50 μM) are very specific substrates for CYP2A in pigs, whereas the CYP2A only is responsible for about 50% of the cotinine formation at a 500 μM nicotine incubation concentration. These results show that the large interindividual differences in porcine CYP2A activity are not caused by polymorphisms but transcriptional regulation and the coumarin 7-hydroxylation is as specific a reaction for porcine CYP2A as for human CYP2A6.
Cytochrome P450 2A (CYP2A) represent 4% of human hepatic P450 (Shimada et al. 1994; Guengereich et al. 1995). The CYP2A family includes CYP2A6, CYP2A7 and CYP2A13 (Fernandez-Salguero et al. 1995; Hoffman et al. 1995) where coumarin has been found to be a high affinity substrate for human CYP2A6. Human CYP2A6 catalysis the 7-hydroxylation of coumarin, whereas neither CYP2A7 nor CYP2A13 can catalysis this reaction in any tissue (Pelkonen et al. 1995). In humans, large interindividual differences in both coumarin hydroxylase activity and the formation of cotinine from nicotine have been found, due to genetic polymorphisms in the CYP2A6 gene (Inoue et al. 2000; Tricker 2003). CYP2A6 catalysed 70–80% of the conversion of nicotine to cotinine via a C-oxidation (Benowits & Jacobs 1994). Substrate selectivity to CYP2A in different species has been evaluated by molecular modelling and crystallography (Lewis & Lake 2002; Lewis et al. 2003). These studies revealed 6 substrate recognition sites (SRS) in the CYP2A protein. Quantitative structure-activity relationships (QSAR) of wild type and mutants showed importance of the amino acids (a.a.) phenylalanine in SRS 2, arginine and threonine in SRS 4 for coumarin binding. According to Xu et al. 2002 the phenylalanine F480 in SRS 6 is the major substrate interacting site for nicotine and isoleucine I471 and arginine R485 are responsible for the folding of the protein around SRS 6. The coumarin 7-hydroxylase activity has also been detected in mouse, catalysed by Cyp2a5 (van Iersel et al. 1994) and in pigs, conventional as well as minipigs. (Skaanild & Friis 1999). It has also been shown that the porcine CYP2A is involved in the metabolism of 3-methylindole or skatole (Diaz & Squires 2000) one of the major contributors to the boar taint, that can be observed in 5–10% of intact male pig (Bæk et al. 1995). The activity of porcine CYP2A show large interindividual differences both for coumarin as well as 3-methylindole metabolism, but it is at present not known if these differences are due to polymorphisms in the porcine CYP2A gene.
The objective of this work is therefore to further analyse the CYP2A in pigs with regard to cDNA sequence, polymorphisms and enzyme activity.
Materials and Methods
Animals. Eight Göttingen minipig (4 female, 4 male, age 4 month) and 12 conventional pigs (4 female, 4 male, 4 castrates, age 3.5 month) have previously been described. (Skaanild & Friis 1999).
Chemicals. All antibodies were obtained from Gentest (MA, USA), sequencing and synthesis of all primers were done by TAG Copenhagen A/S (Denmark). RNA isolation kit was supplied by Quiagen (Hilden, Germany). All other chemicals were of analytical grade obtained either from Amersham Biosciences or Sigma (St. Louis, MO, USA).
Isolation and sequencing of cDNA. Total RNA was isolated from female Göttingen minipig hepatocytes using Qiagen RNAeasy mini kit according to protocol. Amplification of 5′ and 3′ cDNA ends was done using the GeneRacer kit supplied by Invitrogen life technologies. The amplification was done according to kit protocol using 3 μg total RNA. The gene specific primers (GSP) used for the amplification were designed from the CYP2A PCR product described earlier (Skaanild & Friis 1999). This fragment was isolated and sequenced. The following GSP were used: 3′ enden 5′ TCCACGAGATCCAGAGATTCGGAGACA3′ and 5`enden 5′ CTCATCCAGGAAGTGCTGGGGGTTGT 3′. Pfu turbo DNA polymerase from Stratagene was used for high fidelity PCR and the PCR reactions were set up according to kit protocol. Two PCR reactions were set up for each fragment 1) a reaction with RACE primers and GSP and 2) a reaction using the first PCR reaction as template with nested RACE primers and GSP. A touch down cycle reaction with annealing temperatures from 69 ° to 64/° was used for high specificity annealing. PCR fragments of about 700 and 1300 base pair (bp) long for the 3′ end and 5′ end respectively were expected. These products were isolated from the PCR reactions and both strands were sequenced by TAGCopenhagen A/S. The isolation of RNA and RACE amplification and sequencing were done 2 times.
Polymorphism assay using SSCP. Total RNA was isolated as already described and reverse transcription was performed with first strand synthesis kit from Amersham Biotech according to protocol using 2 μg of total RNA in each reaction. The cDNA was divided into 4 fragment 5′ end, 2 middle parts and 3′ end (1, 2, 3, 4) 383 bp, 557 bp, 470 bp and 537 bp respectively (fig. 1). After PCR again using touch down cycling the PCR products were analysed using PAGE gel electrophoresis looking for single strand conformation polymorphisms. The samples were denatured by addition of an equal volume 50 mM NaOH, 1 mM EDTA and then incubated at 95 ° for 5 min. The gels were run using an Amersham Biotech multiphor unit and precast CleanGels and the bands were visualised using silver staining according to kit protocol (Amersham Biosciences).
The frequency of the 3′ deletion was analysed setting up PCR reactions using a primer on each side of the deletion (primer G and GSP3′). This will give rise to two PCR fragments of different sizes, one about 587 bp and the other about 418 bp.
Isolation of liver microsomes. Isolation of microsomes was performed according to Olsen et al. (1997). Briefly the liver was homogenized in 50 mM Tris-HCl buffer containing 0.25 M sucrose and 1 mM EDTA. The homogenate was centrifuged and the supernatant was transferred to new tubes and centrifuged once more at 105,000 ×g at 4 ° for 60 min. The pellet containing the microsomes were homogenized in storage buffer and frozen in liquid nitrogen.
Microsomal protein concentration was determined using a modified Lowry method (Petterson 1977).
Enzyme assay. The microsomal mixture for all assays consisted of a buffer containing 32 mM K-phosphate pH 7.46, 2.5 mM MgCl2, 15 mM glucose-6-phosphate, 10 U glucose-6-P-dehydrogenase/ml, 1.1 mM NADP and 0.91 mg microsomal protein in a total volume of 1.075 ml. The mixture was pre incubated for 5 min. at 37 ° before the test substrate was added.
Nicotine C-oxidation assay. The formation of cotinine was measured according to Yamazaki et al. (1999) with some modifications. Briefly the microsomal incubation mixture (final total volume 525 μl) was added cytosol (3.3 mg prot./ml) and the reaction was started by addition of nicotine (500 μM). The samples were incubated at 37 ° for 60 min. and then the reaction was added 550 μl internal standard (200 ng caffeine/ml MeOH) and stopped by adding 50 μl of 1 M NaCl2, 1 M Na2CO3 pH 10.5 buffer and 3 ml CH2Cl2. The samples were mixed, centrifuged and 2.5 ml of the organic phase was dried. The residue was dissolved in 200 μl 0.01 N HCl and 50 μl of this suspension was analysed by HPLC using the following conditions: Column: Symetri 054215 (250×4.6 mm, Waters); Column temperature: 30 °; eluent: MeOH and 0.05 M CH3COOH (1:4); flow rate 1 ml/min.; Detection: UV at 260 nm. To determine the amount of cytosol necessary for maximal cotinine formation, incubations with different concentrations of cytosol were set up. The results show that the reaction reached the maximum when 200 μl cytosol (3.3 mg protein/ml) were added.
Coumarin 7-hydroxylase assay. To initiate the reaction 50 μl of a 2.2 mM coumarin solution in water was added and the reaction was incubated for further 10 min. then stopped by addition of 1.1 ml methanol. The samples were analyzed by HPLC as described earlier (Skaanild & Friis 1999).
Immunoblotting. Microsomes isolated from the pig livers were used for blotting according to Skaanild & Friis (1999). Briefly, microsomal protein was separated by PAGE SDS gel electrophoresis and blotted to Hybond-ECL nitro-cellulose membranes. Membranes were hybridised (1 hr) with a diluted primary anti-human CYP antibody. Antibody binding was detected by chemiluminescence using a biotinylated secondary antibody followed by a streptavidin-horseradish peroxide conjugate. After development, the blots were exposed to HyperfilmECL.
Inhibition assay. Polyclonal anti-human CYP2A6 was added to the microsome solution (20 μl/mg protein) and pre incubated for 5 min. at 37 °, where after the reactions were started by adding either of the substrates (final concentrations 100 μM coumarin and 500 μM or 50 μM nicotine). The reactions were then analysed as descried for coumarin hydroxylase and cotinine formation.
Isolation and sequencing of cDNA.
In order to analyse the porcine CYP2A gene for polymorphisms it was necessary to sequence the minipig CYP2A. For this purpose total RNA was isolated from liver cells from a female minipig as the females have a much higher expression of this gene than male minipigs (Skaanild & Friis 1999). The RNA (mRNA) was reverse transcribed and a 3′ end and 5′ end RACE (rapid amplification of cDNA end) was performed. The 3′ end PCR reaction gave rise to two different fragment of about 450 bp and 850 bp respectively, whereas only one 5′ end PCR fragment of about 1300 bp long was synthesised. After sequencing of the PCR fragments, two cDNAs of 1620 bp and 1785 bp could be obtained (GenBank accession number AY380866). Blast analysis revealed that the long or wild type cDNA was very homologues to the pig CYP2A19 sequence (GeneBank accession number AB052255) and the cDNA encoded a protein that was 99% (493 of 494 amino acids) homologous to the CYP2A19. The short cDNA sequences were also very similar to the conventional pig CYP2A19 sequence and from base pair 1 to 1426 there is a 100% homology with the conventional base pairs 20 to 1446, where after there is a deletion of 169 bp in the minipig cDNA compared to the conventional. Base pairs 1427 to 1578 are again 100% homologous to the conventional base pairs 1615 to 1766. When compared to the human CYP2A6 cDNA (GeneBank accession number NP 0007359) 88% homology was found from base pair 20 to 1406 and again the minipig cDNA had a deletion compared to the human cDNA. This short cDNA sequence encoded a protein of 487 amino acids, 7 amino acids shorter than the human and conventional counterpart (fig. 2). The beginning of this protein from amino acids 1–466 was nearly identical to both the conventional pig CYP2A19 (464 of 466) and the human CYP2A6 (409 of 466), whereas the rest of the protein sequence was different. The human, conventional pig and minipig wild-type proteins all contain 6 substrate recognition sites whereas only SRS 1 to 5 was found in the short minipig proteins (fig. 2). The 3 amino acid in SRS 2 and 4 important for coumarin binding were conserved in both conventional pig and minipigs and the major nicotine interacting amino acid phenylalanine F480 is present in both the conventional and wild-type minipig protein, but not in the short protein. The frequency of this deletion was analyzed and all but 1 pig contained both the wild type and the deleted PCR fragment.
Polymorphism assay and SSCP.
After sequencing, primers for the polymorphism study were designed as shown in fig. 1 to divide the cDNA in fragment sizes ideal for single strand conformation polymorphism (SSCP). The SSCP analysis of the porcine CYP2A revealed 1 polymorphism (pig 9759) in fragment 1 and 1(pig 9743) polymorphism in fragment 3. Fragment 4 gave rise to one polymorphism that was found in pig 54919.
Nicotine C-oxidation assay.
In order to investigate the influence of the 3′ end deletion and the 3 polymorphisms, the formation of cotinine from nicotine was measured, as the major nicotine recognition site is situated in SRS 6. The formation of cotinine from nicotine involve a nicotine C-oxidation catalysed by CYP2A and a conversion of the metabolite to cotinine catalysed by cytosolic aldehyde oxidase. The Km and Vm were 380 μM and 678 pmol cot./mg prot/min. respectively. The cotinine formation rate for the pigs varied from 212–921 pmol/mg prot./min. when incubated with 500 μM nicotine (table 1).
Table 1. This table shows the coumarin 7-hydroxylase activity, the cotinine formation activity, their specific activities, and CYP2A protein level.
Coumarin-7 hydrox. pmol/mg protein/min.
Cotinine pmol/mg protein/min.
Imm. level U/mg prot.
Cou.hydr.act/ U apoprotein pmol/U/min.
Cotinine/ U apoprotein pmol/U/min.
ND no detectable band. * SSCP Polymophism. ** No deleted PCR fragment.
The microsomes were incubated with a final concentration of 100 μM coumarin and the formation of 7-hydroxy coumarin was found to vary between 0.4–500 pmol/mg protein/min, the male minipigs giving the lowest formation rates (table 1).
The immunoblotting, using anti-human CYP2A6, gave responses from 40–172 U/mg protein for the conventional pigs and 118–160 U/mg protein for female minipigs, whereas for the male minipigs no bands could be observed (table 1). The specific activity (enzyme activity/apoprotein level) could then be calculated and it showed insignificant pig to pig variation (table 1). No specific activity could be calculated for male minipigs as no immunoblotting results were obtained. The CYP2A immunochemical level correlated well with both cotinine formation (r2=0.7) and coumarin 7-hydroxylation (r2=0.9), whereas these activities did not correlate with CYP2B apoprotein level (fig. 3).
To determine how specific CYP2A is for the coumarin 7-hydroxylation and formation of cotinine, the enzyme reactions were incubated with inhibitory antibody against human CYP2A6. The inhibition analysis revealed that the anti-human CYP2A6 inhibited the coumarin 7-hydroxylation reaction from 87.5–91.9% (table 2). One of the minipigs, however, gave a much lower inhibition, but that is due to the nearly 0 enzyme activity. The inhibition of cotinine formation on the other hand was only 44% to 60% when using 500 μM nicotine whereas the inhibition varied from 85–100% when using 50 μM nicotine substrate concentration (table 2).
Table 2. This table shows the inhibition of coumarin 7-hydroxylase activity and cotinine formation after adding anti-human CYP2A6.
In human CYP2A6 at least 17 different polymorphisms have been established (www.imm.ki.se/CYPallales/cyp2a6). They give rise to either no, decreased or unchanged enzyme activity and are responsible for some of the large interindividual differences found in humans. The porcine cytochrome CYP2A that catalyses the coumarin 7-hydroxylation also show large interindividual differences (Skaanild & Friis 1999) as does the metabolism of 3-methylindole. To further study the reasons for these differences in enzyme activities, the minipig CYP2A was sequenced as only the CYP2A sequence of the conventional pig is known. Two cDNA sequences were obtained, a wild type and a deletion of 169 bp compared to the conventional pig sequence. The sequences upstream the deletion was very similar for the 3 porcine sequences and the human sequence. When the 3 porcine protein sequences were compared, it could be seen that they all contained SRS 1–5, whereas SRS 6 was different in the short minipig protein due to the deletion at the end of the cDNA sequence. Both the wild type and deleted 3′ end was found in all but 1 pig. The SSCP analysis revealed further 3 polymorphisms in different segments of the sequence. The influence of these polymorphisms and the deletion was studied further, measuring the formation of cotinine from nicotine. The absence of SRS 6 in the deletion indicated that this protein can not metabolize nicotine as SRS 6 contains the nicotine recognition site. It was shown that the human CYP2A6 test substrate nicotine can also be metabolized by pig microsomes. The Km was found to be 380 μM, which is higher than the human Km of 64.9±32.7 μM (Messina et al. 1997). The porcine cotinine formation rate range between 212–922 pmol/mg protein/min. compared to the human Vmax values of 500±500 pmol/mg protein/min. (Messina et al. 1997). This shows that the formation rate of cotinine in humans and pigs is identical, whereas the Km in pigs is higher than in humans. The difference in Km values may be caused either by differences in the binding strength between enzyme and substrate or because another enzyme also binds and metabolises nicotine. Correlation analysis revealed that the formation of cotinine correlates with the apoprotein concentration of CYP2A, indicating that this enzyme may be responsible for this reaction. However, it can be seen from the graph that although there is no apoprotein level, there is still a high cotinine formation rate of 207 pmol cotinine/mg protein/min., indicating that another isoenzyme is also involved in the cotinine formation. This is in accordance with results published by Yamazaki et al. (1999) using both human microsomes and human recombinant CYPs. They found that at low substrate concentrations (50 μM) the CYP2A6 was the predominant isoenzyme, but at higher substrate concentration (500 μM) also CYP2B6 and CYP2D6 played an important role. The correlation analysis showed that human CYP2A6 correlated well with nicotine C-oxidation as did the porcine CYP2A. Neither human CYP2B6 nor porcine CYP2B protein levels correlated with the cotinine formation. The specific activity (activity/U apoprotein) of the cotinine formation was about the same in all pigs showing that the polymorphisms do not have any effect on specific activity. This could either be because the antibody used in immunoblotting could not recognize the deleted protein or the short PCR fragment may be a 3′ end of either a CYP2A7 or CYP2A13 like cDNA. Both these cDNAs are very homologous to CYP2A6 (96% and 94% respectively) and neither of these genes shows any CYP2A6 activity. However, the significance of the deletion is not known. No differences were found in the specific activity of the nicotine oxidation and coumarin hydroxylation for the polymorphism found in this study. This indicates that they do not have any effect on the interindividual variation in enzyme activity. The variation must therefore be due to transcriptional regulation, if the substrates are as specific to porcine CYP2A as the correlation analysis indicates. This was further investigated using inhibitory anti-human CYP2A6. The antibody inhibited the coumarin 7-hydroxylation with about 91%, which is the same as seen in human coumarin assay where the same antibody can inhibit the assay up to 95% (Le Gal et al. 2003). The cotinine formation on the other hand could only be inhibited by 40–60% using a substrate concentration of 500 μM indicating again that another isoenzyme may be involved. At a 50 μM substrate concentration the antibody inhibited the reaction 86–100%. These results are in accordance with results obtained using human microsomes (Yamazaki et al. 1999), where a 33–60% inhibition with anti-human CYP2A6 was found using a substrate concentration of 500 μM, and cotinine formation can be inhibited up to 80% with the same antibody (Le Gal et al. 2003).
The conclusion of this study is that the observed differences in porcine coumarin hydroxylation and cotinine formation are not due to polymorphisms, but to transcriptional regulation and that the coumarin 7-hydroxylation activity is as specific in pigs as in humans for CYP2A.