Abstract
22 min readThe interleukin-1 receptor-associated kinases (IRAKs) are important downstream signaling components of Toll-like receptors (TLRs). To date, four mammalian IRAKs have been found, namely IRAK-1, IRAK-2, IRAK-4, and IRAK-M. Herein, we show a detailed analysis of the genomic region encompassing the murine Irak2 gene and the molecular cloning of four isoforms of Irak2 (designated Irak2a, Irak2b, Irak2c, and Irak2d) generated by alternative splicing at the 5′-end of the gene. This alternative splicing has direct effects on the expression of the N-terminal death domain and/or inter-domain. No evidence of similar alternative splicing was found for the human IRAK2 gene. When overexpressed, Irak2a and Irak2b potentiated NF-κB activation by lipopolysaccharide. Importantly, Irak2c and Irak2d were inhibitory. The promoter for Irak2c differed from that of the other Irak2 isoforms in that it contained putative NF-κB binding sites. Lipopolysaccharide induced the expression of Irak2c, indicating a possible negative feedback effect on the signaling pathway. Alternative splicing of the Irak2 gene in mice will therefore generate agonistic or antagonistic Irak2 isoforms, which is likely to have consequences for the regulation of TLR signaling. These observations identify another distinguishing feature between mice and humans in the TLR system that is likely to be due to differences in the selective pressure imposed by pathogens on each species during evolution. The interleukin-1 receptor-associated kinases (IRAKs) are important downstream signaling components of Toll-like receptors (TLRs). To date, four mammalian IRAKs have been found, namely IRAK-1, IRAK-2, IRAK-4, and IRAK-M. Herein, we show a detailed analysis of the genomic region encompassing the murine Irak2 gene and the molecular cloning of four isoforms of Irak2 (designated Irak2a, Irak2b, Irak2c, and Irak2d) generated by alternative splicing at the 5′-end of the gene. This alternative splicing has direct effects on the expression of the N-terminal death domain and/or inter-domain. No evidence of similar alternative splicing was found for the human IRAK2 gene. When overexpressed, Irak2a and Irak2b potentiated NF-κB activation by lipopolysaccharide. Importantly, Irak2c and Irak2d were inhibitory. The promoter for Irak2c differed from that of the other Irak2 isoforms in that it contained putative NF-κB binding sites. Lipopolysaccharide induced the expression of Irak2c, indicating a possible negative feedback effect on the signaling pathway. Alternative splicing of the Irak2 gene in mice will therefore generate agonistic or antagonistic Irak2 isoforms, which is likely to have consequences for the regulation of TLR signaling. These observations identify another distinguishing feature between mice and humans in the TLR system that is likely to be due to differences in the selective pressure imposed by pathogens on each species during evolution. The Toll-like receptors (TLRs) 1The abbreviations used are: TLR, Toll-like receptor; aa, amino acid(s); DD, death domain; IRAK, interleukin-1 receptor-associated kinase; LPS, lipopolysaccharide; MEF, murine embryonic fibroblast; MyD88, myeloid differentiation factor 88; nt, nucleotide(s); RACE, rapid amplification of cDNA ends; RT, reverse transcriptase; UTR, untranslated region. are a family of molecules tailored to respond to microbial pathogens, with particular TLRs able to recognize and bind to specific pathogen-associated molecular patterns. Once activated, TLRs recruit cytoplasmic adapter molecules such as myeloid differentiation factor 88 (MyD88), MyD88 adaptor-like protein (Mal; also known as TIRAP), TIR domain-containing adaptor protein inducing interferon-β (TRIF; also termed TICAM-1), and TRIF-related adaptor molecule (TRAM; also termed TIRP or TICAM-2), which, in turn, initiate signaling cascades that result in biological responses geared toward the elimination of pathogens during infection (reviewed in Refs. 1Dunne A. O'Neill L. Science's STKE. 2003; (http://stke.sciencemag.org/cgi/content/full/sigtrans;2003/171/re3)PubMed Google Scholar and 2O'Neill L. Fitzgerald K. Bowie A. Trends Immunol. 2003; 24: 286-290Abstract Full Text Full Text PDF PubMed Scopus (416) Google Scholar). Critical to the TLR signaling cascade are the interleukin-1 receptor-associated kinases (IRAKs). The first human IRAK to be cloned was IRAK1 (3Cao Z. Henzel W.J. Gao X. Science. 1996; 271: 1128-1131Crossref PubMed Scopus (773) Google Scholar), followed by IRAK2 (4Muzio M. Ni J. Feng P. Dixit V.M. Science. 1997; 278: 1612-1615Crossref PubMed Scopus (984) Google Scholar), IRAK-M (5Wesche H. Gao X. Li X. Kirschning C.J. Stark G.R. Cao Z. J. Biol. Chem. 1999; 274: 19403-19410Abstract Full Text Full Text PDF PubMed Scopus (340) Google Scholar), and IRAK4 (6Li S. Strelow A. Fontana E.J. Wesche H. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 5567-5572Crossref PubMed Scopus (543) Google Scholar). The IRAKs share sequence homology to the Drosophila melanogaster protein kinase Pelle, and all contain a death domain (DD), which is used for protein-protein interactions with the DDs of other molecules. For example, IRAK2 uses its DD to mediate its interaction with MyD88 (4Muzio M. Ni J. Feng P. Dixit V.M. Science. 1997; 278: 1612-1615Crossref PubMed Scopus (984) Google Scholar). The IRAKs also have putative kinase domains, although IRAK1 has dispensable kinase activity because interleukin-1-induced NF-κB activation could still be driven by a kinase-inactive mutant (7Li X. Commane M. Burns C. Vithalani K. Cao Z. Stark G.R. Mol. Cell. Biol. 1999; 19: 4643-4652Crossref PubMed Scopus (187) Google Scholar). In addition, both IRAK2 and IRAK-M are catalytically inactive due to the absence of certain key residues within their putative kinase domains (4Muzio M. Ni J. Feng P. Dixit V.M. Science. 1997; 278: 1612-1615Crossref PubMed Scopus (984) Google Scholar, 5Wesche H. Gao X. Li X. Kirschning C.J. Stark G.R. Cao Z. J. Biol. Chem. 1999; 274: 19403-19410Abstract Full Text Full Text PDF PubMed Scopus (340) Google Scholar). Adding further complexity to the signaling cascades initiated by the TLRs are the recent findings that some of the genes encoding components of TLR signaling are alternatively spliced, thus generating multiple isoforms. One example is the murine MyD88 gene, which encodes a full-length MyD88 (MyD88L) and a shorter form (MyD88S) generated by the splicing out of exon 3; the removal of this exon causes the deletion in the mature polypeptide of the intermediate domain (8Janssens S. Burns K. Tschopp J. Beyaert R. Curr. Biol. 2002; 12: 467-471Abstract Full Text Full Text PDF PubMed Scopus (232) Google Scholar). Both forms of MyD88 are differentially expressed and exhibit differences in their ability to induce NF-κB activation and IRAK phosphorylation, with MyD88S being inhibitory (8Janssens S. Burns K. Tschopp J. Beyaert R. Curr. Biol. 2002; 12: 467-471Abstract Full Text Full Text PDF PubMed Scopus (232) Google Scholar). MyD88S can, however, mediate the activation of c-Jun N-terminal kinase (9Janssens S. Burns K. Vercammen E. Tschopp J. Beya FEBS Lett. 2003; 548: 103-107Crossref PubMed Scopus (151) Google Scholar). Another example is the human IRAK1 gene, which encodes two isoforms generated by the differential usage of a splice acceptor site within exon 12 (10Jensen L.E. Whitehead A.S. J. Biol. Chem. 2001; 276: 29037-29044Abstract Full Text Full Text PDF PubMed Scopus (70) Google Scholar). In contrast to the full-length isoform (designated IRAK1a), the slightly shorter isoform (IRAK1b) is kinase-inactive and displays no change in its protein levels following interleukin-1 stimulation (10Jensen L.E. Whitehead A.S. J. Biol. Chem. 2001; 276: 29037-29044Abstract Full Text Full Text PDF PubMed Scopus (70) Google Scholar). Here, we report the identification and annotation of the murine Irak2 gene, which generates four alternatively spliced isoforms of Irak2 that contain various deletions of the N-terminal third of the mature protein. Of these, Irak2a and Irak2b enhance the activity of an NF-κB reporter, whereas Irak2c and Irak2d are inhibitory. Our results therefore reveal a level of control of TLR signaling that involves differential splicing of murine Irak2. Sequence Analysis—The complete nucleotide sequence of the murine Irak2 gene and flanking regions was obtained from the National Center for Biotechnology Information (NCBI) mouse genomic data base (contig NT_039353 found at www.ncbi.nlm.nih.gov), following BLAST analysis with the human IRAK2 cDNA (GenBank™ accession number AF026273). Other sequences used in this manuscript that were also obtained from the GenBank™ were the human IRAK2 genomic locus (contig NT_005927) and Irak2 cDNA (accession number AJ440756). Transcription factor binding site searches were performed using TFSEARCH (molsun1.cbrc.aist.go.jp/research/db/TFSEARCH.html) (11Heinemeyer T. Wingender E. Reuter I. Hermjakob H. Kel A. Kel. O. Ignatieva E. Ananko E. Podkolodnaya O. Kolpakov F.A. Podkolodny N.L. Kolchanov N.A Nucleic Acids Res. 1998; 26: 362-367Crossref PubMed Scopus (1325) Google Scholar) and MatInspector Release Professional (genomatix.gsf.de/mat_fam) (12Quandt K. Frech K. Karas H. Wingender E. Werner T. Nucleic Acids Res. 1995; 23: 4878-4884Crossref PubMed Scopus (2424) Google Scholar). Global alignment of the human and mouse genomic sequences of conserved synteny was performed with the program AVID 2N. Bray, A. Fabrikant, J. Lord, J. Schwartz, I. Dubchak, and L. Pachter, unpublished work. using a window size of 100 bp and a conservation level of 70%; the results were viewed with the program VISTA (www-gsd.lbl.gov/vista/) (13Bray N. Dubchak I. Pachter L. Genome Res. 2003; 13: 97Crossref PubMed Scopus (365) Google Scholar, 14Dubchak I. Brudno M. Loots G.G. Pachter L. Mayor C. Rubin E.M. Frazer K.A. Genome Res. 2000; 10: 1304-1306Crossref PubMed Scopus (272) Google Scholar, 15Mayor C. Brudno M. Schwartz J.R. Poliakov A. Rubin E.M. Frazer K.A. Pachter L.S. Dubchak I. Bioinformatics. 2000; 16: 1046-1047Crossref PubMed Scopus (794) Google Scholar). The identification of transcribed nucleotide sequences and repeat sequences in the genomic sequence was performed using the NIX application (hgmp.mrc.ac.uk). The translation of putative open reading frames was carried out using MacVector version 7.1 Oxford Molecular Ltd, and amino acid alignments were performed using ClustalW 1.8 (searchlauncher.bcm.tmc.edu). Domain predictions were determined using PROSITE (www.expasy.ch/prosite/) (16Sigrist C.J. Cerutti L. Hulo N. Gattiker A. Falquet L. Pagni M. Bairoch A. Bucher P. Brief. Bioinformatics. 2002; 3: 265-274Crossref PubMed Scopus (676) Google Scholar). 5′- and 3′-SMART™ Rapid Amplification of cDNA Ends (RACE)—5′- and 3′-SMART™ RACE-ready cDNA libraries were generated with or without reverse transcriptase according to the manufacturer's recommendations (Clontech) using 1 μg of polyadenylated (poly(A)+) mRNA from normal C57Bl/6 murine embryonic fibroblasts (MEFs), thymus, placenta, or liver (Clontech) or from 1 μg of total human placental or spleen mRNA (Clontech). 5′- and 3′-RACE reactions were carried out according to the manufacturer's recommendations (Clontech) using 100 units of Advantage 2 polymerase, 0.2 pmol of universal primer mix, and 0.2 pmol of gene-specific primer (Table I) in a final volume of 50 μl. PCRs were amplified as described previously (17Hardy M.P. Hertzog P.J. Owczarek C.M. Biochem. J. 2002; 365: 355-367Crossref PubMed Google Scholar). To obtain specific RACE products, primary RACE PCR products were diluted 1:50 with 10 mm Tris-EDTA buffer and further amplified with Advantage 2 polymerase using 0.2 pmol of Nested universal primer (Clontech) and 0.2 pmol of Nested gene-specific primer (Table I) under conditions described previously (17Hardy M.P. Hertzog P.J. Owczarek C.M. Biochem. J. 2002; 365: 355-367Crossref PubMed Google Scholar). PCR products were analyzed by agarose gel electrophoresis and transferred to GeneScreen Plus nylon membranes, and their specificity was determined by hybridization with an internal oligonucleotide. Positive RACE PCR products were gel-purified, cloned into pGEM-T (Promega, Madison, WI), and sequenced.Table IOligonucleotides usedNameSequence (5′ → 3′)UsageIrak2-15′-TGGACCTCCTGTGTCACCTGGAACTCT-3′RT-PCRIrak2-25′-CAACGTCATCTGTTTCCTCCGGGGTAT-3′RT-PCRIrak2-35′-GCTGCATCTCTGCCTGTAGGAATCTGTCCA-3′5′-RACEIrak2-45′-GCTTGGACGACATCTGCTTCACTCCAGAAA-3′5′-RACEIrak2-55′-CTTCGGGATGCGTCCCCAGGCAGA-3′RT-PCRIrak2-65′-GCTGCATCTCTGCCTGTAGGAATCTGTCCA-3′RT-PCRIrak2-75′-ACTGGATGCAGTTCGGGAAGCCGGTTC-3′RT-PCRIrak2-85′-CCATGGCTTGCTACATCTACCAGCTGCCGT-3′RT-PCRIrak2-95′-CATCAGGGTCCAAAGAGCTCGCTGCTGTCC-3′RT-PCRIrak2-105′-CTTCAGTCTGCTCCAGGAAGACCAGCA-3′3′-RACEIrak2-115′-GTTTCTGAGGCCACAGGCTCATCTTCCAAT-3′3′ -RACEIrak2-125′-GGCCCATCATGGCTGGGGCCCAGCGGCAGC-3′Northern blotIrak2-135′-CTCCTCCACGCTGGCATTGCGCCTCCGCAG-3′Northern blotIrak2-145′-GCGGATCCATGGCTTGCTACATCTACCAGC-3′ConstructsIrak2-155′-CTGGATCCATGGCTGGGGCCCAGCGGCAGC-3′ConstructsIrak2-165′-GTCGGGAACGTCGTAGGGGTAGGGTCCAAAGAGCTCGCT-3′ConstructsIrak2-175′-GCGGAATTCCTAGGCGTAGTCGGGAACGTC-3′ConstructsIrak2-185′-AGATTGTCTTGAGCTGGAAGCCGGTTCCTG-3′RT-PCRHuI2-15′-TTCCTGCTTAGGAATGGAGGTGCTGAAGTC-3′5′-RACEHuI2-25′-AAGTGGGGAGGTCGCTTCTCAAGGAATGAG-3′5′-RACEHuI2-35′-CTGCTACATCTACCAGCTGCCCTCCTG-3′RT-PCRhuI2-45′-CAACACAGGCCTCAGAGGAGGAGTCAG-3′RT-PCRGAPDH-FaGlyceraldehyde-3-phosphate dehydrogenase (forward).5′-GAACGGGAAGCTTGTCATCAA-3′RT-PCRGAPDH-RbGlyceraldehyde-3-phosphate dehydrogenase (reverse).5′-CTAAGCAGTTGGTGGTGCAG-3′RT-PCRa Glyceraldehyde-3-phosphate dehydrogenase (forward).b Glyceraldehyde-3-phosphate dehydrogenase (reverse). Open table in a new tab Cell Culture and Reagents—Murine 3T3 fibroblasts and RAW264.7 cells were grown in RPMI 1640 medium (Sigma) supplemented with 10% fetal calf serum, 1% penicillin/streptomycin, and 1% l-glutamine (Sigma). The human embryonic kidney HEK293 cell line stably transfected with TLR4 (HEK293-TLR4) was a gift from Dr. Katherine Fitzgerald (University of Massachusetts Medical School) and was grown in Dulbecco's modified Eagle's medium (Sigma) supplemented as above with the addition of 600 μg/ml G418 (Sigma). Lipopolysaccharide (LPS) from Escherichia coli serotype O26:B6 was obtained from Sigma and used at 1 μg/ml. The pRL-TK vector was obtained from Promega, and the pCDNA3.0 vector was from Invitrogen (Carlsbad, CA) The pGL3-NF-κB construct bearing five repeats of the κB-consensus was a gift from Dr. R. Hofmeister (Universitaet Regensburg, Regensburg Germany). The human pCDNA-IRAK2 construct was obtained from Dr. Marta Muzio (Mario Negri Institute, Milan, Italy). Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)—Total RNA was extracted from 5 × 106 RAW264.7 cells treated with 1 μg/ml LPS for 0, 1, 3, 6, 9, and 24 h using TRI REAGENT™ according to the manufacturer's recommendations (Sigma). To each 10-cm2 dish of LPS-treated cells was added 1 ml of TRI-REAGENT™, and the homogenous lysates were then centrifuged at 12,000 × g for 10 min. Each supernatant was added to 0.2 ml of chloroform and incubated at room temperature for 15 min following vigorous mixing. The RNA phase was then separated by centrifugation at 12,000 × g for 15 min, added to 0.5 ml of isopropanol, mixed, allowed to stand for 10 min then further centrifuged at 12,000 × g for 10 min. The RNA pellets were then washed with 1 ml of 70% ethanol and resuspended in 50 μl of H2O. cDNA was generated from total RNA using Superscript™ III reverse transcriptase according to the manufacturer's recommendations (Invitrogen). To 9 μl of H2O was added 1 μl 50 μm oligo(dT)20 (Promega), 1 μg of total RNA, 1 μl of 10 mm dNTP, and the mixture was incubated at 65 °C for 5 min and then on ice for a further 5 min. To this mixture was then added 4 μl of 5× first-strand buffer (Invitrogen), 1 μl of 0.1 m dithiothreitol (Invitrogen), 1 μl of RNase inhibitor (1 unit/μl; Promega), and 1 μl of 200 units/μl Superscript™ III reverse transcriptase (Invitrogen) for plus RT libraries or 1 μl of H2O for minus RT libraries; the final mixture was incubated at 50 °C for 60 min and then inactivated at 70 °C for 15 min. RT-PCR from RACE-ready cDNA libraries was performed on a PerkinElmer Life Sciences 2400 thermocycler in a reaction involving 2.5 μl 10× Vent (Roche Applied Science), 0.5 μl of Vent polymerase (2 units/μl; Roche Applied Science), 1 μl 10 mm dNTP, 10 μm each forward and reverse primer (MWG-Biotech; Table I), 2 μl of cDNA (plus or minus RT), and 17 μl of H2O. PCR cycling conditions were 95 °C for 1 min and then 30 cycles at 95 °C for 1 min, 64-68 °C for 1 min, and 72 °C for 3 min. Amplified fragments were then cloned into pGEM-T and sequenced. For reverse transcribed cDNAs generated from LPS-treated RAW264.7 cells, the following reaction was used involving 2.5 μl of 10× Taq buffer (Promega), 3 μl of 25 mm MgCl2 (Promega), 1 μl of 2.5 mm dNTP, 1 μl of 10 μm each forward and reverse primers (Table I), cDNA (plus or minus RT), 0.175 μl of Taq (5 units/μl; Promega) and 14.325 μl of H2O. PCR cycling conditions for these reactions were 94 °C for 30 s and then 35 cycles at 94 °C for 10 s, 67 °C for 10 s, and 72 °C for 30 s followed by a final extension at 72 °C for 2 min. For glyceraldehyde-3-phosphate dehydrogenase gene RT-PCR, identical conditions were used except for an annealing temperature of 55 °C and 25 cycles instead of 35. Northern Blotting—Northern blots were carried out using a murine eight-tissue blot (Clontech) according to the manufacturer's recommendations. Blots were hybridized at 68 °C for 90 min with a 1038-bp 32P-labeled Irak2 cDNA fragment (nucleotides (nt) 482-1527) corresponding to exons 4-11 inclusive of the Irak2 gene. Blots were then washed at 50 °C with 0.2× SSC 0.1% SDS and exposed to Kodak X-OMAT film for 2 days at -70 °C. Generation of Irak2 Expression Constructs—The entire open reading fames of Irak2a, Irak2b, Irak2c, and Irak2d were amplified by PCR from a murine liver cDNA library (Clontech) using the primers Irak2-14 (for Irak2a, 2b, and 2d), Irak2-15 (for Irak2c), and Irak2-16 as a common reverse primer (Table I). PCR was carried out on a PerkinElmer Life Sciences 2400 thermocycler using the following conditions: 2.5 μl 10× Vent; 1 μl of 10 mm dNTP; 0.5 μl of Vent polymerase (2 units/μl); 1 μl of 20 ng/μl cDNA; 1 μl each of 10 μm 5′ and 3′ primers (Table I), and 18 μl of H2O. The following cycling conditions were then used: 95 °C for 1 min; 35 cycles at 95 °C for 1 min; 68 °C for 1 min; and 72 °C for 3 min. All Irak2 constructs were generated to encode in-frame hemagglutinin A tags (YPYDVPDYA) at their C termini using the primer Irak2-17 and contained BamHI and EcoRI linkers at their 5′- and 3′-ends, respectively; these linkers were subsequently cleaved with their respective restriction endonucleases (New England Biolabs, Hertfordshire, UK), subcloned into pCDNA3.0, and sequenced. Transient Transfection—Cells were seeded at ∼1 × 105 cells/ml into 24-well plates in complete growth media and cultured at 37 °C (in the presence of 5% CO2) until 50-80% confluent. To 17 μl of serum-free media was added 3 μl of GeneJuice™ (Novagen, La Jolla, CA); following mixing and a 5 min incubation at RT, the solution was added to 450 ng of plasmid DNA, gently mixed, incubated for a further 15 min at RT, then added dropwise to each well of the 24-well plate. This was performed in triplicate for each assay point. The composition of the transfected DNA typically consisted of 100 ng of Renilla luciferase (pRL-TK), 200 ng of NF-κB-luciferase reporter, and 25-150 ng of Irak2 expression plasmids, and the remainder was empty vector (pCDNA3.0) up to 150 ng. Following transfection, cells were either left untreated at 37 °C (5% CO2) for 24 h or treated with LPS at 1 μg/ml for 16 h. Optimal treatment times had been determined previously by time course, and total incubation time for all transfected cells was 24 h. Cells were then lysed in 300 μl/well of 1× passive lysis buffer (Promega) for 15 min at room temperature, and 30 μl each of the lysate was analyzed in duplicate for firefly luciferase and Renilla luciferase activity using a Mediators PhL luminometer. Reporter activity was then determined as a function of firefly luciferase activity divided by the Renilla luciferase activity. Differences between means were calculated using Student's t test analysis. Identification of the Murine Irak2 Gene—To identify the murine orthologue of the human IRAK2 gene, a BLAST search of the murine genomic DNA data base using the human IRAK2 cDNA sequence of Muzio et al. (4Muzio M. Ni J. Feng P. Dixit V.M. Science. 1997; 278: 1612-1615Crossref PubMed Scopus (984) Google Scholar) was performed. A region of significant homology was found on Mus musculus (MMU) chromosome 6 at position E3. This putative Irak gene was not murine Irak1, Irak4, or Irak-M, because these genes lay on other chromosomes; this gene was therefore likely to be Irak2. To obtain a complete annotation of the putative Irak2 gene, ∼120-kb pairs of murine genomic DNA encompassing this gene was analyzed in detail using the NIX suite of programs, which are able to identify putative exons, transcriptional units, polyadenylation signal sequences, CpG islands, and repetitive elements within the genomic sequence to be analyzed. As shown in Fig. 1, the NIX programs were able to readily identify the murine Irak2 gene and the positions of the majority of the individual exons of the Irak2 gene with the exception of exons 3 and 11. This prediction was confirmed by BLASTN and BLASTX analysis and alignment of the huIRAK2 cDNA sequence against the murine genomic sequence. The putative Irak2 gene encompasses ∼55 kb of murine genomic sequence and is transcribed toward the center of MMU chromosome 6 (Fig. 1). ∼7 kb 5′ of the putative exon 1 of Irak2 lies the well predicted murine Von Hippel-Lindau disease tumor suppressor gene (VHL; GenBank™ accession number P40338), a multiexon gene of ∼6-kb transcribed in the same orientation as Irak2 (Fig. 1). Immediately 3′ of Irak2 lay another predicted gene with the designation Kiaa0218 (Fig. 1) with strong homology to human KIAA0218, a gene encoding a putative deoxyribonuclease (GenBank™ accession number Q93075). This murine transcriptional unit also appears to be transcribed in the same orientation as VHL and Irak2. The human homologues of all three murine genes are found in exactly the same positions on Homo sapiens chromosome 3p25.3, transcribed toward the centromere (data not shown). Located further 3′ of murine Kiaa0218, transcribed in the opposite orientation, was a putative transcriptional unit (the Ghr gene) encoding the murine Ghrelin precursor (also known as the growth hormone secretagogue, the growth hormone-releasing peptide, the motilin-related peptide, and the M46 protein; GenBank™ accession no. Q9EQX0) (Fig. 1). Identification of Multiple Irak2 cDNAs—To obtain Irak2 cDNA sequence information, primers were designed (Irak2-1 and Irak2-2; Table I) within two regions of murine genomic sequence with high homology to the human IRAK2 cDNA; these primers were also located at either end of the putative gene (exons 2 and 12) to amplify as much cDNA as possible. Surprisingly, RT-PCR amplification from a MEF cDNA using these primers library yielded two fragments of ∼1200 and 1350 bp (Fig. 2A), which strongly suggested differential splicing of the murine Irak2 gene. No products were amplified from minus RT MEF cDNA, demonstrating that these products were not the result of genomic DNA contamination (data not shown). Upon sequencing of these PCR products, it was determined that the larger cDNA fragment (designated Irak2a) was encoded by exons 2-12, inclusive, of the putative Irak2 gene. The amplified cDNA sequence also matched the predicted exons, thus providing confirmation of the efficacy of the prediction programs used. The smaller cDNA fragment lacked a sequence encoded by exon 3 but was otherwise identical to the larger amplified cDNA; this cDNA therefore appeared to be a variant of Irak2 (we have designated this cDNA Irak2b). To further annotate the genomic structure of Irak2, identify its 5′ and 3′ ends, and identify additional Irak2 variants, RACE PCR was performed using murine thymus, placenta, liver, and MEF RACE cDNA libraries, with similar results obtained. 5′-RACE and nested 5′-RACE PCR was first performed using primers located in exons 6 and 5, respectively (Irak2-3 and Irak2-4; Table I). Two products of ∼650 and 500 bp were amplified (Fig. 2B); the larger fragment, when sequenced, corresponded to Irak2a, and the smaller fragment corresponded to the exon 3-deficient Irak2b. A smaller PCR product of ∼400 bp was also observed (Fig 2B); upon sequencing, this product turned out to be nonspecific. 5′-RACE also identified the putative exon 1 and two distinct 5′ terminations at positions 1 and 33 of the Irak2a cDNA. Also identified by 5′-RACE were several 650- and 500-bp clones (Fig. 2B) that generated further variant Irak2 cDNA sequences. Several of the 650-bp cDNA fragments generated utilized the 3′ 420 bp of intron 3 before reading directly into exon 4; we designated this isoform Irak2c. Several 500-bp cDNA fragments were also found to contain an Irak2 cDNA sequence similar to that of Irak2a but lacking a sequence encoded by exon 2; this isoform has been designated Irak2d. To confirm the existence of the Irak2c and Irak2d cDNAs, RT-PCR was performed using a MEF cDNA library (Fig. 2C) and isoform-specific primers (Irak2-5 and Irak-6 for Irak2c; Irak2-7 and Irak-6 for Irak2d). The Irak2c and Irak2d RT-PCR reactions generated 400- and 510-bp PCR products, respectively; these were sequenced and the results confirmed the presence of these isoforms. As final confirmation, the open reading frames of all putative Irak2 isoforms were amplified by RT-PCR using Irak2-8 and Irak2-9 primers for Irak2a, 2b and 2d cDNAs, and Irak2-5 (exon 4′) and Irak2-9 for Irak2c (Data not shown). The Irak2a is the same cDNA as that previously described (18Rosati O. Martin M.U. Biochem. Biophys. Res. Commun. 2002; 297: 52-58Crossref PubMed Scopus (16) Google Scholar). Amplification of the Irak2d cDNA also revealed a 30-nt deletion in exon 12 in addition to the deleted exon 2 (data not shown). For all RT-PCR reactions, no PCR products were generated using a minus RT MEF cDNA library (data not shown). To locate the 3′-end of the Irak2 gene, 3′-RACE was performed using the Irak2-10 and Irak2-11 primers located in exons 11 and 12, respectively (Table I). A single PCR fragment of ∼350-bp was generated (Fig. 2D), which, upon sequencing, was found to contain an Irak2 cDNA sequence encoding exons 12 and 13. All four Irak2 cDNA sequences were confirmed against the murine genomic sequence data base by BLASTN and BLASTX and have been deposited into GenBank™ under the accession numbers AY162378, AY162379, AY162380, and AY162381. Gene Structure and Alternative Splicing of Murine Irak2—The murine Irak2 gene is composed of 13 exons and 12 introns and encodes a predicted full-length protein of 622 amino acids (aa) (Fig. 3A and Table II). Full-length Irak2 contains a well predicted kinase domain (aa 206-471) and a weakly predicted DD (aa 14-93). Exon 1 contains the common 5′-untranslated region (UTR) of Irak2a, 2b, and 2d, as well as their common initiating codon (ATG), and also encodes the N terminus (aa 1-13). Exon 2 encodes the entire DD; exons 3 and 4 encode what we now designate as the α and β subdomains of the interdomain (aa 94-205), respectively; exons 5-11 encode the kinase domain; exons 12 and 13 encode the C terminus (aa 472-622); and the termination codon for all Irak2 variants lies in exon 13 (Fig. 3A). Irak2c has its own 5′-UTR (designated exon 4′) that is encoded by a 5′ continuation of exon 4. Intron sizes range from 83 bp (intron 5) to almost 18 kb (intron 2), and the intron-exon boundaries of Irak2 conform to the GT-AG rule (Fig. 3A) (19Shapiro M. Senapathy P. Nucleic Acids Res. 1987; 15: 7155-7174Crossref PubMed Scopus (1971) Google Scholar) (Table II). The three possible codon disruption phases are present in the Irak2 gene splice junctions; introns 4, 5, 7, 9, 10, and 11 disrupt exons between amino acids (phase 0), introns 1, 2, and 3 interrupt a codon between the first and second nucleotide (phase 1), and introns 6 and 8 disrupt a codon between the second and third nucleotide (phase 2) (Table II).Table IIIntron-exon boundaries of the murine Irak2 gene3′ Splice acceptor siteExonBase pairs5′ Splice acceptor siteIntronBase pairsSplice phase1154GGATGCAGTTCGgtgagtgaca19,0361W M Q FtgccttccagCTTCCTACGTGA2183TTGTCTTGAGCTgtgagtaact217,8801A S Y VI V L ScttgttccagGGAAGCCGGTTC3144TTCTAGACACAGgtgctctcct34,3521W K P VL L D TtcctttctagGGCCCATCATGG495CCCCAGTCTAAGgtaaatccac42,4450G P I MP Q S KcaatctccagTATTGCAGTACT5201AAGCTCAGGGAGgtgaggtgag5830Y C S TK L R EtttcccacagGTGGCCGGCTCC665GCTCTGCCTCAGgtaagcttcc62,5432V A G SL C L RctgggttcagATGCTGCCACGC7115CTCTGGGCTCAGgtaaaccagt78440C C H AL W A QtttctcttagGGCAACTCAGAC8110CAATGTTAAGAGgtgagagggc81,9502G N S DN V K StttcccgcagTGCCAACGTCTT9196AGCTGTGGAATTgtaagcattt93,8880A N V LS C G ItcctacccagGTATTGGCCGAG1063CCAGTTTACCTGgtaagagatc103,8300V L A EP V Y LcttcccatagAAGGATTTGCTT11201AGCGTGGAGGAGgtgagctctc112,3590K D L LS V E EctcctccaagGCACGAGTCTCC12290CACCCCAGAACGgtaagcttgg122,609A R V StgaacatcagCTACAGAGACTT13131 Open table in a new tab The alternative splicing of Irak2 isoforms is shown in detail in Fig. 3B. Irak2a utilizes every exon consecutively. However, Irak2b and Irak2d are generated by the deletion of exons 3 and 2, respectively, in a process called exon skipping. Irak2d also utilizes an alternative splice acceptor site 30 bp into exon 12, which causes the deletion of 10 amino acids (designated the C-box) in the C terminus (Fig. 3D). The function of
.jpg)
Discussion(0)
No comments yet. Be the first to comment.