Alternative Splicing Determines the Function of CYP4F3 by Switching Substrate Specificity
2001; Elsevier BV; Volume: 276; Issue: 41 Linguagem: Inglês
10.1074/jbc.m104818200
ISSN1083-351X
AutoresPeter Christmas, Jeffrey P. Jones, Christopher Patten, Dan A. Rock, Yimin Zheng, Shing‐Ming Cheng, Brittany M. Weber, Nadia Carlesso, David T. Scadden, Allan E. Rettie, Roy J. Soberman,
Tópico(s)Drug Transport and Resistance Mechanisms
ResumoDiversity of cytochrome P450 function is determined by the expression of multiple genes, many of which have a high degree of identity. We report that the use of alternate exons, each coding for 48 amino acids, generates isoforms of human CYP4F3 that differ in substrate specificity, tissue distribution, and biological function. Both isoforms contain a total of 520 amino acids. CYP4F3A, which incorporates exon 4, inactivates LTB4 by ω-hydroxylation (Km = 0.68 μm) but has low activity for arachidonic acid (Km = 185 μm); it is the only CYP4F isoform expressed in myeloid cells in peripheral blood and bone marrow. CYP4F3B incorporates exon 3 and is selectively expressed in liver and kidney; it is also the predominant CYP4F isoform in trachea and tissues of the gastrointestinal tract. CYP4F3B has a 30-fold higherKm for LTB4 compared with CYP4F3A, but it utilizes arachidonic acid as a substrate for ω-hydroxylation (Km = 22 μm) and generates 20-HETE, an activator of protein kinase C and Ca2+/calmodulin-dependent kinase II. Homology modeling demonstrates that the alternative exon has a position in the molecule which could enable it to contribute to substrate interactions. The results establish that tissue-specific alternative splicing of pre-mRNA can be used as a mechanism for changing substrate specificity and increasing the functional diversity of cytochrome P450 genes. Diversity of cytochrome P450 function is determined by the expression of multiple genes, many of which have a high degree of identity. We report that the use of alternate exons, each coding for 48 amino acids, generates isoforms of human CYP4F3 that differ in substrate specificity, tissue distribution, and biological function. Both isoforms contain a total of 520 amino acids. CYP4F3A, which incorporates exon 4, inactivates LTB4 by ω-hydroxylation (Km = 0.68 μm) but has low activity for arachidonic acid (Km = 185 μm); it is the only CYP4F isoform expressed in myeloid cells in peripheral blood and bone marrow. CYP4F3B incorporates exon 3 and is selectively expressed in liver and kidney; it is also the predominant CYP4F isoform in trachea and tissues of the gastrointestinal tract. CYP4F3B has a 30-fold higherKm for LTB4 compared with CYP4F3A, but it utilizes arachidonic acid as a substrate for ω-hydroxylation (Km = 22 μm) and generates 20-HETE, an activator of protein kinase C and Ca2+/calmodulin-dependent kinase II. Homology modeling demonstrates that the alternative exon has a position in the molecule which could enable it to contribute to substrate interactions. The results establish that tissue-specific alternative splicing of pre-mRNA can be used as a mechanism for changing substrate specificity and increasing the functional diversity of cytochrome P450 genes. cytochrome P450 leukotriene B4 prostaglandin H2 hydroxyeicosatetraenoic acid glyceraldehyde-3-phosphate dehydrogenase reverse transcription polymerase chain reaction arachidonic acid high performance liquid chromatography base pair(s) Cytochrome P450 (CYP)1 monooxygenases catalyze the oxidation of a broad spectrum of lipophilic substrates that include endogenous products such as cholesterol, steroids, and fatty acids, or xenobiotics such as drugs. Fifty-five human CYP genes have been classified into 17 families and 40 subfamilies. 2Web address: drnelson.utmem.edu/CytochromeP450.html. Phylogenetic studies indicate that all CYPs derive from duplication and divergence of an ancestral gene, and this radiation of CYP genes accounts for the diverse substrate utilization by the superfamily. To date, alternative splicing of CYP pre-mRNAs is not considered a mechanism that contributes to this functional diversity. CYP-dependent oxidation of arachidonic acid generates biologically active eicosanoids that function as intracellular mediators in normal physiology and disease. Members of the CYP2C and CYP2J subfamilies act as arachidonic acid epoxygenases and generate eicosatrienoic acids (EETs), which regulate ion channels and are candidates for endothelium-derived hyperpolarizing factor (1Fisslthaler B. Popp R. Kiss L. Potente M. Harder D.R. Fleming I. Busse R. Nature. 1999; 401: 493-497Crossref PubMed Scopus (812) Google Scholar, 2Node K. Huo Y. Ruan X. Yang B. Spiecker M. Ley K. Zeldin D.C. Liao J.K. Science. 1999; 285: 1276-1279Crossref PubMed Scopus (1035) Google Scholar). CYP4 enzymes catalyze ω-hydroxylation of fatty acids including arachidonic acid and are a potential source of 20-HETE. 20-HETE is a potent activator of protein kinase C and Ca2+/calmodulin-dependent kinase II and has roles in regulating vascular tone, natriuresis, and cell proliferation (3Nowicki S. Chen S.L. Aizman O. Cheng X.J. Li D. Nowicki C. Nairn A. Greengard P. Aperia A. J. Clin. Invest. 1997; 99: 1224-1230Crossref PubMed Scopus (152) Google Scholar, 4Ma Y.-H. Gebremedhin D. Schwartzman M.L. Falck J.R. Clark J.E. Masters B.S. Harder D.R. Roman R.J. Circ. Res. 1993; 72: 126-136Crossref PubMed Google Scholar, 5Amlal H. LeGoff C. Vernimmen C. Soleimani M. Paillard M. Bichara M. Am. J. Physiol. 1998; 274: C1047-C1056Crossref PubMed Google Scholar, 6Muthalif M.M. Benter I.F. Karzoun N. Fatima S. Harper J. Uddin M.R. Malik K.U. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 12701-12706Crossref PubMed Scopus (186) Google Scholar, 7McGiff J.C. Quilley J. Am. J. Physiol. 1999; 277: R607-R623PubMed Google Scholar). Enzymes in the CYP4A subfamily have high activity for arachidonic acid in animals (8Loughran P.A. Roman L.J. Aitken A.E. Miller R.T. Masters B.S. Biochemistry. 2000; 39: 15110-15120Crossref PubMed Scopus (21) Google Scholar). The identity and distribution of enzymes generating 20-HETE in humans is poorly understood. This determination is complicated by wide-ranging Kmvalues of different enzymes for arachidonic acid, distinct patterns of tissue expression of the relevant enzymes, close homology between subfamily members, and variation between species. The human enzyme CYP4A11 has been shown to generate 20-HETE (9Palmer C.N. Richardson T.H. Griffin K.J. Hsu M.H. Muerhoff A.S. Clark J.E. Johnson E.F. Biochim. Biophys. Acta. 1993; 1172: 161-166Crossref PubMed Scopus (84) Google Scholar), but it exhibits very low activity for arachidonic acid (10Hoch U. Zhang Z. Kroetz D.L. Ortiz de Montellano P.R. Arch. Biochem. Biophys. 2000; 373: 63-71Crossref PubMed Scopus (83) Google Scholar). Recently it was suggested that ω-hydroxylation of arachidonic acid in human liver and kidney is mediated primarily by CYP4F2 (11Powell P.K. Wolf I. Jin R. Lasker J.M. J. Pharmacol. Exp. Therap. 1998; 285: 1327-1336PubMed Google Scholar, 12Lasker J.M. Chen W.B. Wolf I. Bloswick B.P. Wilson P.D. Powell P.K. J. Biol. Chem. 2000; 275: 4118-4126Abstract Full Text Full Text PDF PubMed Scopus (281) Google Scholar). In addition to activating arachidonic acid, CYPs can inactivate LTB4 and prostaglandins by ω-hydroxylation. CYP4F3 was originally identified as the enzyme that catalyzes ω-oxidation of the 5-lipoxygenase pathway product LTB4 with a lowKm (0.5–1.0 μm) in human neutrophils (13Shak S. Goldstein I.M. J. Biol. Chem. 1984; 259: 10181-10187Abstract Full Text PDF PubMed Google Scholar, 14Soberman R.J. Harper T.W. Murphy R.C. Austen K.F. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 2292-2295Crossref PubMed Scopus (54) Google Scholar, 15Kikuta Y. Kusunose E. Endo K. Yamamoto S. Sogawa K. Fujii-Kuriyama Y. Kusunose M. J. Biol. Chem. 1993; 268: 9376-9380Abstract Full Text PDF PubMed Google Scholar). LTB4 functions as a chemoattractant of neutrophils and monocytes (16Ford-Hutchinson A.W. Bray M.A. Doig M.V. Shipley N.E. Smith M.J.H. Nature. 1980; 286: 26465Crossref Scopus (1633) Google Scholar, 17Migliorisi G. Folkes E. Pawlowski N. Cramer E.B. Am. J. Pathol. 1987; 127: 157-167PubMed Google Scholar), and CYP4F3 has been projected to play a role in the termination of LTB4-mediated inflammation. We identified an alternative splice form of CYP4F3 in liver (18Christmas P. Ursino S.R. Fox J.W. Soberman R.J. J. Biol. Chem. 1999; 274: 21191-21199Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar) and designated the two isoforms CYP4F3A (the original isoform detected in neutrophils) and CYP4F3B (the isoform detected in liver). Both isoforms have 520 amino acids but are distinguished by the alternate use of exons that code for amino acids 67–114. CYP4F3B contains exon 3, whereas CYP4F3A contains exon 4. These exons are identical in size but code for sequences that share only 27% amino acid identity. CYP4F3B has lower activity for LTB4 and is not expressed in myeloid cells, whereas CYP4F3A is not expressed in liver. We analyzed the tissue distribution and kinetic properties of CYP4F3B to determine its functional significance. CYP4F3B has an expression pattern that is distinct from CYP4F3A and is the predominant CYP4F enzyme in liver and other non-hematopoietic tissues. It has aKm for arachidonic acid of 22 μm and generates 20-HETE as the major product of ω-hydroxylation. In contrast, arachidonic acid is a very poor substrate for CYP4F3A. We used molecular modeling to predict the position of amino acids 67–114 within the molecule. These studies suggest that exons 3 and 4 code for a region that contributes to the active site and substrate access channel. Selection between these exons determines the ability of the enzyme to either inactivate LTB4 (CYP4F3A) or activate arachidonic acid (CYP4F3B). The results demonstrate that tissue-specific alternative splicing of pre-mRNA must now be considered a mechanism for generating functional diversity of cytochrome P450s. Microsomes containing human CYPs (specific content of 640 pmol of 4F3B/mg of protein or 36 pmol of 4F3A/mg of protein), P450 reductase, and cytochrome b5(SupersomesTM) were prepared from baculovirus-infected BTI-TN-5B1-4 cells by Gentest Corporation. To assay the conversion of LTB4 to 20-hydroxy-LTB4, reaction mixtures containing Supersomes (10 pmol of CYP enzyme), LTB4, 1 mm NADPH, and 100 mm KPO4buffer, pH 7.4 were incubated in a final volume of 0.1 ml for 20 min at 37 °C. The reaction was terminated with 0.1 ml of 94% acetonitrile, 6% glacial acetic acid and centrifuged at 10,000 × gfor 3 min. The supernatant (60–85 μl) was injected into a Zorbax C18-SB reverse phase HPLC column (4.6 × 250 mm, 5 μm) coupled to a 5-μm Waters Sentry guard column and separated with a linear gradient of 30% acetonitrile, 2 mm perchloric acid to 70% methanol over 24 min at a flow rate of 1.0 ml/min. The product was detected by absorbance at 270 nm using a Waters 2487 UV detector and its identity confirmed by comparing the retention time of a 20-hydroxy-LTB4 standard (Cayman). Kinetic results were analyzed by non-linear regression (Sigma Plot Software, SPSS inc., Chicago, IL) using eight substrate concentrations in a range of 0.2–20 μm (4F3A) or 2–80 μm (4F3B). To assay the conversion of arachidonic acid to 20-HETE, reaction mixtures containing Supersomes (10 pmol of CYP enzyme), arachidonic acid, 0.15 μCi of [14C]arachidonic acid (810 Ci/mmol), 1 mmNADPH, and 100 mm KPO4 buffer, pH 7.4 were incubated in a final volume of 0.1 ml for 10 min at room temperature. The reactions were terminated and injected into a C18reverse phase HPLC column as before and separated with a linear gradient of 0.1% trifluoroacetic acid in H20 to 0.1% trifluoroacetic acid in acetonitrile over 24 min at a flow rate of 1.0 ml/min. The product was detected by on-line monitoring with a RadioFlow detector (Flo-One Beta 150TR, Packard BioScience). The retention time of a 20-HETE standard (Cayman) was determined by absorbance at 205 nm. Kinetic results were analyzed by linear regression using 8 substrate concentrations in a range of 1.0–150 μm (4F3A) or 0.5–60 μm (4F3B). The kinetic experiments were performed under conditions where less than 10% of the substrate was converted to product. CYP4F3 cDNA probes were labeled with [α-32P]dCTP by random priming. The full-length coding region of CYP4F3A and a partial cDNA betweenNsiI and EagI sites (nucleotides 561–1268) were used to generate probes of 1560 bp and 708 bp, respectively. Multiple Tissue Northern (MTN) Blots from CLONTECH were hybridized with the 1560-bp probe for the full-length coding region (5 ng/ml, ∼106 cpm/ng) in 10 ml of ExpressHyb (CLONTECH) for 1 h at 68 °C. The blots were washed in 2× SSC, 0.05% SDS at room temperature for 4 × 10 min and then 0.1× SSC, 0.1% SDS at 50 °C for 2× 20 min. A Human Multiple Tissue Expression (MTE) Array fromCLONTECH was hybridized with the 708-bp probe (2 ng/ml, ∼2 × 106 cpm/ng) for 6 h at 65 °C in 10 ml of ExpressHyb supplemented with 300 μg of sheared salmon testes DNA (Sigma), 60 μg C0t-1 DNA (Roche Molecular Biochemicals), and 100 μl of 20× SSC. The partial cDNA probe was used because it gave lower background than the full-length probe, and nucleotides 561–1268 were selected for this probe to ensure that both isoforms of CYP4F3 would be recognized with equal efficiency. The array was washed in 2× SSC, 1% SDS at 65 °C for 5× 20 min and then 0.1× SSC, 0.5% SDS at 55 °C for 2× 20 min. The MTN Blots and MTE Array were exposed to Kodak XAR film overnight at −70 °C with an intensifying screen. Hybridization signals on the MTE Array were quantified with a phosphorimager (Cyclone, Packard BioScience) and expressed relative to peripheral blood leukocytes. RNA was extracted from freshly isolated human bone marrow cells (1–2 × 106 cells) using the RNeasy Kit (QIAGEN) with QIAshredder columns for cell homogenization. Total RNA was purified from peripheral blood neutrophils as previously described (18Christmas P. Ursino S.R. Fox J.W. Soberman R.J. J. Biol. Chem. 1999; 274: 21191-21199Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar). Total and poly(A)+ RNA from liver, fetal liver, kidney, prostate, ileum, and trachea were from CLONTECH. First-strand cDNA synthesis for PCR was performed using the cDNA Cycle Kit (Invitrogen) with AMV reverse transcriptase and random primers. The cDNA was purified by phenol/chloroform extraction and ethanol precipitation. Primers B, C, G, and H (TableI) were used to detect CYP4F3A (primer pair B-C), CYP4F3B (B-G), and CYP4F2 (H-G) by isoform-specific PCR as previously described (18Christmas P. Ursino S.R. Fox J.W. Soberman R.J. J. Biol. Chem. 1999; 274: 21191-21199Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar). The PCR conditions were 94 °C for 1 min, 52 °C for 1 min, 72 °C for 1 min; 30 cycles were followed by 1 cycle with a 10 min extension time. Primers specific for CYP4F8 (K-L) and CYP4F11 (M-N) were used under the same PCR conditions except for annealing temperatures of 65 °C and 62 °C, respectively.Table ISummary of primersNameDirectioncDNAPositionSequence (5′ to 3′)Asense4F3A, 4F3B, 4F2, 4F1176–97GGGGCCTCCTGGCTCCTGGCCCBsense4F3A, 4F3B165–184GAAACGGAATTGGTTCTTGGCantisense4F3A340–322GAGCAAAGAGCACAGGCTTDantisense4F3A, 4F3B, 4F2, 4F11524–507TGCATGATGTTCACACTCEantisense4F3A, 4F3B1212–1186TTGGGTGCAGCAGCGAGAGACGGCAGGFsense4F3B277–307GGCCCCATCTTCCCCGTCATCCGTTTTTGCCGantisense4F3B, 4F2340–322AGGCGTTGATGACAGACCGHsense4F2165–184AAGACGGAACTGGTTTTGGGIsense4F2277–306GGACCCATCTCCCCCCTCCTCAGTTTGTGCJantisense4F21212–1186CTGGGTGACATGGCGGGAGATGACCGGKsense4F8281–313CCATCACTCCCATCATCAACTTGTGCCACCCTGLantisense4F81212–1187CTGGGTGCAGCCGCGGGCGAATGTAGMsense4F11203–229CTCCCACGGAAGAGGGCATGAAGACATNantisense4F11349–325CAGCAGCTGAGGCACTGGTGATAGGGDHFsenseGAPDH31–61TTTGGTCGTATTGGGCGCCTGGTCACCAGGGGDHRantisenseGAPDH488–458ACCTTGGCCAGGGGTGCTAAGCAGTTGGTGGSequences of primers referred to in text are shown. cDNA positions are relative to the ATG initiation codon (A = +1). Open table in a new tab Sequences of primers referred to in text are shown. cDNA positions are relative to the ATG initiation codon (A = +1). cDNA samples generated as described above were used to estimate the relative levels of expression of different CYP4F gene products. A variety of primers specific for CYP4F gene products (CYP4F3A, CYP4F3B, CYP4F2) and internal standards (GAPDH, cyclophilin A, 18 S rRNA) were tested in semiquantitative PCR experiments. Primer pair E-F (CYP4F3B-specific, 962-bp product), or I-J (CYP4F2-specific, 962-bp product), worked well in combination with the internal standard primers GDHF and GDHR (GAPDH-specific, 458-bp product). The PCR conditions were 94 °C for 1 min, 68 °C for 1 min, 72 °C for 1 min, 20 cycles. This cycle number was selected after comparing the linear range of 32P incorporation for the different cDNAs and primer pairs used, and it represents a point within the linear range for all samples. Each reaction contained 5 μCi of [α-32P]dCTP, unlabeled dNTPs at a final concentration of 0.2 mm each, and 20 pmol of each primer. PCR products were separated by electrophoresis on a 6% sequencing gel (SequaGel System, National Diagnostics). The gel was dried onto filter paper, and the products were visualized by exposure to Kodak XAR film and quantified with a phosphorimager. The relative CYP4F isoform expression level in each sample was determined by comparing the intensity of the CYP4F and GAPDH signals. The results of four experiments using different RNA batches were used to determine the mean ± S.D. A limited number of primers can be designed that are specific for CYP4F3A, and these failed to give PCR products when combined with primers for the internal standards required in quantitative PCR experiments. We attempted to use a variety of primer sequences, primer lengths, and annealing temperatures for CYP4F3A and internal standards that included GAPDH, cyclophilin A, and 18 S rRNA + competimers (Ambion) to eliminate this problem. None of these was successful, though the reasons for the inhibition were not clear. An alternative method for quantifying the relative levels of CYP4F3A in tissues was developed using a generic CYP4F primer pair (A-D), which generates a 449-bp PCR product from CYP4F3A, CYP4F3B, CYP4F2, and CYP4F11. PCR was performed: 94 °C for 1 min, 52 °C for 1 min, 72 °C for 1 min; 20 cycles were followed by 1 cycle with a 10 min extension time. The PCR products obtained from each tissue were then ligated into pCR2.1-TOPO vector (Invitrogen) and transformed into TOP10 cells. Individual colonies were picked for isolation of plasmid DNA and sequence analysis. A minimum of 50 colonies from each tissue were analyzed. Two methods of analysis of plasmid DNA were performed, which gave identical results: plasmid DNA from 50% of colonies were sequenced; alternatively, restriction digests were used to predict the outcome of sequence analysis. The 449-bp PCR products generated by primer pair A-D contain sites for both MscI andHincII (CYP4F2), HincII only (CYP4F11),MscI only (CYP4F3B), or neither enzyme (CYP4F3A). A comparison of the restriction pattern generated with MscIversus HincII predicts the identity of the PCR product. CYP4F8 and CYP4F12 are excluded from this analysis. Initial alignments of P450BM3 and CYP4A1 from rat were taken from the work of Hasemannet al. (19Hasemann C.A. Kurumbail R.G. Boddupalli S.S. Peterson J.A. Deisenhofer J. Structure. 1995; 3: 41-62Abstract Full Text Full Text PDF PubMed Scopus (653) Google Scholar). CYP4A1 was aligned with the CYP4F enzymes with the multiple sequence alignment program ClustalW and the PAM250 similarity matrix. A second alignment was performed on amino acids 50–150 using ClustalW and the sequences for 4A7, 4A11, 4B1, 4F2, 4F3A, 4F3B, 4F12, and BM3. A three-dimensional homology model was constructed from these alignments using the program Modeler (20Sali A. Blundell T.L. J. Mol. Biol. 1993; 234: 779-815Crossref PubMed Scopus (10788) Google Scholar). This program uses a combination of molecular dynamics and restraints based on the known structures of homologous enzymes. Thus, this program will give a three-dimensional representation of an alignment that has a predicted structure similar to the known structure. The degree of similarity between target and model is dependent on the degree of implied homology in the alignments. Thus, structure in areas with inserts will be determined solely by the molecular dynamics force field, and areas without inserts will give a weighted average of the restraints and the force field. Five members of the human CYP4F subfamily have been described: 4F3 (15Kikuta Y. Kusunose E. Endo K. Yamamoto S. Sogawa K. Fujii-Kuriyama Y. Kusunose M. J. Biol. Chem. 1993; 268: 9376-9380Abstract Full Text PDF PubMed Google Scholar), 4F2 (21Kikuta Y. Kusunose E. Kondo T. Yamamoto S. Kinoshita H. Kusunose M. FEBS Lett. 1994; 348: 70-74Crossref PubMed Scopus (79) Google Scholar), and 4F8 (22Bylund J. Finnstrom N. Oliw E.H. Biochem. Biophys. Res. Commun. 1999; 261: 169-174Crossref PubMed Scopus (51) Google Scholar) have 520 amino acids, and 4F11 (23Cui X. Nelson D.R. Strobel H.W. Genomics. 2000; 68: 161-166Crossref PubMed Scopus (57) Google Scholar) and 4F12 (24Bylund J. Bylund M. Oliw E.H. Biochem. Biophys. Res. Commun. 2001; 280: 892-897Crossref PubMed Scopus (79) Google Scholar, 25Hashizume T. Imaoka S. Hiroi T. Terauchi Y. Fujii T. Miyazaki H. Kamataki T. Funae Y. Biochem. Biophys. Res. Commun. 2001; 280: 1135-1141Crossref PubMed Scopus (71) Google Scholar) have 524. The proteins share ∼ 80% amino acid identity. The genes exhibit a highly similar organization and were originally assigned 13 exons, with the ATG initiation codon at the beginning of exon 2 (22Bylund J. Finnstrom N. Oliw E.H. Biochem. Biophys. Res. Commun. 1999; 261: 169-174Crossref PubMed Scopus (51) Google Scholar, 23Cui X. Nelson D.R. Strobel H.W. Genomics. 2000; 68: 161-166Crossref PubMed Scopus (57) Google Scholar, 26Kikuta Y. Kato M. Yamashita Y. Miyauchi Y. Tanaka K. Kamada N. Kusunose M. DNA Cell Biol. 1998; 17: 221-230Crossref PubMed Scopus (38) Google Scholar, 27Kikuta Y. Miyauchi Y. Kusunose E. Kusunose M. DNA Cell Biol. 1999; 18: 723-730Crossref PubMed Scopus (41) Google Scholar). Our analysis of the 4F3 gene identified an additional exon which can participate in alternative splicing reactions (18Christmas P. Ursino S.R. Fox J.W. Soberman R.J. J. Biol. Chem. 1999; 274: 21191-21199Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar). The 4F3 and 4F2 genes are now considered to have 14 exons as shown in Fig.1. Exons 3 and 4 in the 4F3gene are identical in size (145 bp) and are mutually exclusive, each coding for amino acids 67–114. The cDNA of 4F3 originally isolated from human neutrophils (now designated 4F3A) contains exon 4, which lacks a homologous counterpart in other 4F cDNAs. 4F3B contains exon 3, which shows high identity with the exon coding for amino acids 67–114 in all other 4F members. Exon 3 in 4F3B has 83, 75, 66, and 60% amino acid identity with the corresponding exon in 4F2, 4F8, 4F11, and 4F12, respectively. This identity is highest for 4F2, and in all cases is much greater than the 27% seen between exons 3 and 4 within the 4F3 gene. Analysis of genomic sequences determined that a homolog of 4F3 exon 4 is present in the 4F2gene (18Christmas P. Ursino S.R. Fox J.W. Soberman R.J. J. Biol. Chem. 1999; 274: 21191-21199Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar), but not the 4F8, 4F11, or4F12 genes. A reassignment of 14 exons to the 4F2gene, analogous to 4F3, is shown in Fig. 1. To determine the significance of alternative splicing we investigated whether 4F3B has a distinct substrate preference and tissue distribution compared with 4F3A, and whether a novel isoform of 4F2 containing exon 4 is expressed. The 4F3B enzyme converts LTB4 to 20-hydroxy-LTB4 (Fig.2A) with aKm of 20.6 μm using a substrate concentration range of 1–80 μm. Substrate inhibition was observed at higher concentrations of LTB4 (onset > 80 μm) for 4F3B, but not 4F3A. The Km of 4F3A for LTB4 was 0.68 μm, consistent with previous reports (14Soberman R.J. Harper T.W. Murphy R.C. Austen K.F. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 2292-2295Crossref PubMed Scopus (54) Google Scholar, 15Kikuta Y. Kusunose E. Endo K. Yamamoto S. Sogawa K. Fujii-Kuriyama Y. Kusunose M. J. Biol. Chem. 1993; 268: 9376-9380Abstract Full Text PDF PubMed Google Scholar). 4F3B utilizes arachidonic acid as a substrate and generates 20-HETE as its major product (Fig.2B). Minor products with retention times of 16.8 min and 13.7 min likely correspond with the oxidation products 20-aldehyde-AA and 20-carboxy-AA, respectively, analogous to the conversion of 20-hydroxy-LTB4 to 20-aldehyde-LTB4 and 20-carboxy-LTB4 (13Shak S. Goldstein I.M. J. Biol. Chem. 1984; 259: 10181-10187Abstract Full Text PDF PubMed Google Scholar, 28Soberman R.J. Sutyak J.P. Okita R.T. Wendelborn D.F. Roberts II, L.J. Austen K.F. J. Biol. Chem. 1988; 263: 7996-8002Abstract Full Text PDF PubMed Google Scholar). These products were not observed with higher concentrations of arachidonic acid or at shorter times of incubation and appear to result from the sequential oxidation of 20-HETE. 4F3B has a Km for arachidonic acid of 22 μm and a Vmax of 13.3 pmol of product/min/pmol of P450. As a comparison, the Km of 4F3A for arachidonic acid was determined to be 185.6 μmwith a Vmax of 11.5 pmol of product/min/pmol of P450. A summary of the kinetic data obtained for 4F3 isoforms using the Supersome system is shown in Table II. 4F3B has a V/K value that is 44-fold lower than 4F3A for LTB4, and 10-fold higher than 4F3A for arachidonic acid.Table IILTB4 and arachidonic acid ω-hydroxylation by CYP4F3 isoformsLTB4Arachidonic acidKmVmax2-bVmax values measured as pmol of product/min/pmol of P450.V/KKmVmaxV/Kμmmin−1μmmin−1CYP4F3A0.68 ± 0.252-aValues ± S.D.,n = 3.32.8 ± 1548.2185.6 ± 1.511.5 ± 0.30.06CYP4F3B20.6 ± 10.723.3 ± 1.11.122.0 ± 813.3 ± 5.60.62-a Values ± S.D.,n = 3.2-b Vmax values measured as pmol of product/min/pmol of P450. Open table in a new tab To analyze the tissue distribution of mRNA products of the CYP4F subfamily, a human MTE array was hybridized with a32P-labeled cDNA probe (nucleotides 561–1268) for 4F3A (Fig. 3A). There is 100% identity with 4F3B within the 708-bp region of the probe, 95% identity with 4F2, and ∼90% identity with 4F8, 4F11, and 4F12. Strong hybridization signals were detected in peripheral blood leukocytes, bone marrow, liver, kidney, fetal liver, and prostate. The signals were quantified and expressed relative to the measurement for peripheral blood leukocytes (Table III). Weaker hybridization signals were detected in trachea, ileum, and other gastrointestinal tract tissues including duodenum, jejunum, and colon. No hybridization was observed in the negative control grids in column 12 containing yeast total RNA, yeast tRNA, Escherichia colirRNA, and E. coli DNA. A positive signal was obtained from human genomic DNA (100 ng of DNA, relative signal strength = 0.2; 500 ng of DNA, relative signal strength = 0.8) detecting the CYP4F gene family on chromosome 19.Table IIITissue distribution and quantification of CYP42 and CYP4F3 isoformsTissue Source of RNAExpression array CYP4F hybridization signal relative to blood leukocytesQuantitative RT-PCR Isoform-specific PCR signal relative to GAPDH internal control (±S.D., n = 4)Clonal analysis % colonies analyzed (n = 50) following RT-PCR with isoform-generic primers and TA cloningCYP4F3BCYP4F2CYP4F3ACYP4F3BCYP4F2Blood leukocyte1.010000Bone marrow1.810000Liver16.70.48 ± 0.10.12 ± 0.0708812Kidney3.00.15 ± 0.050.2 ± 0.0504060Fetal liver2.10.1 ± 0.030.04 ± 0.0267420Prostate1.70.08 ± 0.030.02 ± 0.0170246Ileum0.20.02 ± 0.010.01 ± 0.007205426Trachea0.20.015 ± 0.010.036640 Open table in a new tab Northern blotting confirmed the restricted tissue distribution of CYP4F mRNA, and identified transcripts of ∼5 and ∼2.4 kb in both liver and kidney, but not heart, brain, placenta, lung, skeletal muscle, or pancreas (Fig. 3B). 4F3 mRNA transcripts of 5.034 and 2.339 kb correspond to alternative polyadenylation signals in the 4F3 gene (18Christmas P. Ursino S.R. Fox J.W. Soberman R.J. J. Biol. Chem. 1999; 274: 21191-21199Abstract Full Text Full Text PDF PubMed Scopus (68) Google Scholar, 26Kikuta Y. Kato M. Yamashita Y. Miyauchi Y. Tanaka K. Kamada N. Kusunose M. DNA Cell Biol. 1998; 17: 221-230Crossref PubMed Scopus (38) Google Scholar) and are consistent with the pattern observed. The identities of the CYP4F transcripts expressed in each tissue were determined by isoform-specific RT-PCR (Fig.4A). Distinct patterns of expression were observed in different tissues. 4F3A was the only transcript that could be detected in RNA from peripheral blood neutrophils and bone marrow after 30 cycles of PCR. Fetal liver, kidney, prostate, and ileum contain 4F3B and 4F2 in addition to 4F3A. Trachea contains 4F3A and 4F3B, but not 4F2. Liver contains 4F3B and 4F2 but not 4F3A. The expression of exon 4 from the 4F2 gene was not detected in any of the tissues examined. Primers specific for 4F8 (K-L) or 4F11 (M-N) were included in RT-PCR experiments (data not shown). 4F8 could only be detected in prostate and required more than 30 cycles of PCR to give a visible band; 4F11 was detected in liver, kidney, prostate, ileum, and trachea. The relative levels of 4F3B and 4F2 in tissues were determined by quantitative RT-PCR using GAPDH primers as an internal standard (Fig.4B). The results are summarized in Table III. In addition, a sample of PCR products generated by the isoform-generic primer pair A-D (recognizes 4F3A, 4F3B, 4F2, and 4F11) was cloned and sequenced. Sequence analysis of cloned PCR products provides confirmation of isoform identity, and analysis of 50 PCR products (50 colonies from each tissue) predicted a similar ratio of 4F3B to 4F2 to that obtained by quantitative PCR (Table III). When determined by quantitative PCR, the levels of 4F3B and 4F2 are similar in kidney, but 4F3B exceeds 4F2 in liver, prostate, fetal liver, and ileum by 4-, 4-, 2.5-, and 2-fold, res
Referência(s)