POLN, a Nuclear PolA Family DNA Polymerase Homologous to the DNA Cross-link Sensitivity Protein Mus308
2003; Elsevier BV; Volume: 278; Issue: 34 Linguagem: Inglês
10.1074/jbc.m305646200
ISSN1083-351X
AutoresF. Marini, Nayun Kim, Anthony Schuffert, Richard D. Wood,
Tópico(s)Carcinogens and Genotoxicity Assessment
ResumoThe Drosophila Mus308 gene is unusual in encoding both a family A DNA polymerase domain and a DNA/RNA helicase domain. A mus308 mutation was shown to result in increased sensitivity to DNA cross-linking agents, leading to the hypothesis that Mus308 functions in the repair of DNA interstrand cross-links. Recently a mammalian ortholog of Mus308, POLQ, has been identified. We report here the identification, cloning, and characterization of POLN and its gene product, a new mammalian DNA polymerase also related to Mus308. The human cDNA encodes a protein of 900 amino acid residues. The region starting from residue 419 shares 33% identity (48% similarity) with the equivalent region of Escherichia coli DNA polymerase I. POLN is expressed in human cell lines with numerous alternatively spliced transcripts, and a full-length human coding region that comprises 24 exons within 160 kilobases of genomic DNA. Expression analysis by northern blotting and in situ hybridization showed highest expression of full-length POLN in human and mouse testis. POLN localized to the nucleus when expressed as a enhanced green fluorescent protein (GFP)-tagged protein in human fibroblasts. GFP-tagged recombinant POLN had DNA polymerase activity on activated calf thymus DNA and on a singly primed template. The Drosophila Mus308 gene is unusual in encoding both a family A DNA polymerase domain and a DNA/RNA helicase domain. A mus308 mutation was shown to result in increased sensitivity to DNA cross-linking agents, leading to the hypothesis that Mus308 functions in the repair of DNA interstrand cross-links. Recently a mammalian ortholog of Mus308, POLQ, has been identified. We report here the identification, cloning, and characterization of POLN and its gene product, a new mammalian DNA polymerase also related to Mus308. The human cDNA encodes a protein of 900 amino acid residues. The region starting from residue 419 shares 33% identity (48% similarity) with the equivalent region of Escherichia coli DNA polymerase I. POLN is expressed in human cell lines with numerous alternatively spliced transcripts, and a full-length human coding region that comprises 24 exons within 160 kilobases of genomic DNA. Expression analysis by northern blotting and in situ hybridization showed highest expression of full-length POLN in human and mouse testis. POLN localized to the nucleus when expressed as a enhanced green fluorescent protein (GFP)-tagged protein in human fibroblasts. GFP-tagged recombinant POLN had DNA polymerase activity on activated calf thymus DNA and on a singly primed template. A growing number of mammalian DNA polymerases have been identified within recent years, and initial characterization suggests that they have specialized roles in DNA replication or repair. Based on sequence relationships, the enzymes are currently classified into four families designated A, B, X, and Y (1Burgers P.M. Koonin E.V. Bruford E. Blanco L. Burtis K.C. Christman M.F. Copeland W.C. Friedberg E.C. Hanaoka F. Hinkle D.C. Lawrence C.W. Nakanishi M. Ohmori H. Prakash L. Prakash S. Reynaud C.A. Sugino A. Todo T. Wang Z. Weill J.C. Woodgate R. J. Biol. Chem. 2001; 276: 43487-43490Abstract Full Text Full Text PDF PubMed Scopus (284) Google Scholar). Founder members of each family are Escherichia coli DNA polymerase (pol) 1The abbreviations used are: pol, DNA polymerase; ICL, interstrand crosslink; RACE, rapid amplification of cDNA ends; GFP, enhanced green fluorescent protein; EST, expressed sequence tag. I (family A), mammalian pol α (family B), mammalian pol β (family X), and E. coli pol V (family Y). Nuclear enzymes in families B, X, and Y have received much recent attention, with family Y including enzymes that can bypass DNA template lesions. In Drosophila melanogaster, a family A enzyme is believed to be encoded by the Mus308 gene (2Harris P.V. Mazina O.M. Leonhardt E.A. Case R.B. Boyd J.B. Burtis K.C. Mol. Cell. Biol. 1996; 16: 5764-5771Crossref PubMed Scopus (99) Google Scholar). The COOH-terminal portion of Mus308 is predicted to encode a DNA polymerase, whereas the NH2-terminal portion predicts seven characteristic motifs found in DNA and RNA helicases. An apparent human ortholog of Mus308, designated POLQ or DNA pol θ, is encoded in the human genome (NCBI accession number NM_006596), and a cDNA representing the COOH-terminal part of the gene has been isolated (3Sharief F.S. Vojta P.J. Ropp P.A. Copeland W.C. Genomics. 1999; 59: 90-96Crossref PubMed Scopus (97) Google Scholar). We recently isolated a mammalian DNA helicase gene designated HEL308, which is homologous to the NH2-terminal portion of Drosophila Mus308 (4Marini F. Wood R.D. J. Biol. Chem. 2002; 277: 8716-8723Abstract Full Text Full Text PDF PubMed Scopus (74) Google Scholar). Mutations in the Mus308 gene lead to marked sensitivity to DNA interstrand cross-linking agents (5Boyd J.B. Sakaguchi K. Harris P.V. Genetics. 1990; 125: 813-819Crossref PubMed Google Scholar). Such interstrand DNA cross-links (ICLs) can be caused by some environmental and chemotherapeutic agents and are potent inhibitors of DNA replication and transcription. In mammalian systems, ICL repair takes place and can be observed in cells, but the mechanisms are not well understood. In E. coli, ICLs can be repaired by the coordinated action of nucleotide excision repair and homologous recombination (6Cole R.S. Proc. Natl. Acad. Sci. U. S. A. 1973; 70: 1064-1068Crossref PubMed Scopus (254) Google Scholar, 7Van Houten B. Microbiol. Rev. 1990; 54: 18-51Crossref PubMed Google Scholar). In this pathway, the prototype A family enzyme pol I plays a crucial role. With the aim of discovering enzymes related to Mus308 and pol I that may be involved in DNA crosslink repair, we sought mammalian DNA polymerases belonging to the A family of DNA polymerases and identified a new mammalian DNA polymerase, POLN (DNA pol ν). Isolation of Human and Mouse POLN Genes—Searches of public data bases revealed a mouse expressed sequence tag, AA475002 (NCBI accession number) homologous to the COOH-terminal part of the Drosophila mus308 polymerase domain. From this 1193-bp sequence, primers were designed to clone the murine coding sequence by 3′- and 5′-rapid amplification of cDNA ends (BD Biosciences SMART RACE cDNA amplification kit). Total RNA was prepared from mouse leukemia L1210 cells using an RNeasy kit (Qiagen). Two different amplification reactions were performed to exclude mistakes obtained by PCR amplification, and verified fragments were assembled. For the human cDNA, primers were designed from NCBI sequence AF044578 to isolate a 1756-bp COOH-terminal fragment by PCR from a testis cDNA library (BD Biosciences). To obtain and analyze the NH2-terminal region of human POLN, total RNA was extracted from K562 cells and 833K cells. Reverse transcription and nested PCR was performed and the products cloned using a Zero Blunt TOPO PCR cloning kit (Invitrogen). Inserts were screened by PCR using vector primers M13F and M13R. Insert sizes varied because of alternative splicing and several clones from each set were sequenced. Cell Growth and Expression of GFP-POLN—SV40-transformed human MRC5VA fibroblasts were grown in minimal Eagle's medium supplemented with 10% fetal calf serum. The POLN open reading frame was amplified using Pfu DNA polymerase and primers containing terminal EcoRI and BamHI sites. The PCR product was digested with EcoRI and BamHI and ligated into the EcoRI, BamHI sites of pEGFP-C1 (BD Biosciences). This produced an in-frame fusion of POLN to the COOH terminus of enhanced green fluorescent protein (GFP). A QuikChange mutagenesis kit (Stratagene) was used for introducing substitution mutations. Constructs and mutations were confirmed by sequencing. Cells were transfected using FuGENE 6 (Roche Molecular Biochemicals). After 48 h, cells were incubated in selection medium containing 800 μg/ml G418. Stable transfectants were isolated after 2 weeks selection. To visualize GFP-tagged POLN, cells grown on coverslips were rinsed in PBS and fixed with 2% paraformaldehyde for 10 min. After washing with PBS, cells were mounted with Glycergel (Dako) and photographed with an Olympus Fluoview BX61 confocal microscope and a CCD camera. DNA Polymerase Assays—Exponentially growing MRC5VA cells were transfected with either GFP or GFP-POLN and 24 h later were washed in phosphate-buffered saline. Cells (2 × 107 per assay point) were lysed in 500 μl of buffer A (50 mm Tris-HCl, pH 7.5, 50 mm β-glycerophosphate, 100 mm NaCl, 0.1% Tween 20, 5% glycerol, 1 mm phenylmethylsulfonyl fluoride, 1 mm dithiothreitol, and EDTA-free protease inhibitor mixture from Roche Diagnostics) for 30 min on ice. Each cleared lysate (22,000 × g for 20 min) was immunoprecipitated at 4 °C for 2 h with 5 μg of mouse monoclonal 3E6 anti-GFP antibody (Molecular Probes) that had been cross-linked to 35 μl of protein A-Sepharose CL-4B beads (Amersham Biosciences). Immunoprecipitates on a Micro Bio-Spin column (Bio-Rad) were washed three times with 1 ml lysis buffer and one time with 1 ml of buffer B (10 mm Tris-HCl, pH 7.9, 50 mm NaCl, 10 mm MgCl2, 1 mm dithiothreitol, 0.2 mg/ml bovine serum albumin). Rabbit polyclonal antibody PA434, raised against an NH2-terminal 500-amino acid fragment of human POLN produced in E. coli, was used for immunoblotting to detect the presence of various constructs of GFP-POLN. For reactions with activated calf thymus DNA (8Rush J. Konigsberg W.H. Prep. Biochem. 1989; 19: 329-340PubMed Google Scholar), 10 μl of protein-bound beads was incubated with 20 μl of buffer B, supplemented with 0.1 mm dATP, 0.1 mm dTTP, 0.1 mm dGTP, and 0.02 mm dCTP, 2 μCi of [α-32P]dCTP, and 25 μg of activated calf thymus DNA (Sigma), at 37 °C for 30 min. Reactions were stopped with 10 μl of 40 mm EDTA, 2% sodium pyrophosphate and placed on ice. As a positive control 0.5 unit of exo-free DNA pol I was used. 5 μl of each reaction mixture was spotted onto DE81 paper (Whatman), which was washed three times with 0.5 m Na2HPO4 for 5 min and once with ethanol. The paper was dried at room temperature and radioactivity quantified with the use of a Fuji phosphorimager. For primer extension, a 16-mer (5′-CTT GAA AAC ATA GCG A-3′) was end-labeled with T4 polynucleotide kinase and [γ-32P]ATP and annealed to a 34-mer template (5′-AGC GTC TTA ATC TAA GCT ATC GCT ATG TTT TCA AG-3′) (9Tissier A. McDonald J.P. Frank E.G. Woodgate R. Genes Dev. 2000; 14: 1642-1650PubMed Google Scholar). Reaction mixtures (20 μl) in buffer B with 0.1 mm of each dNTP were incubated for 30 min at 37 °C with 10 μl of immunoprecipitated beads or 0.1 unit of pol I (exo-) as indicated. Reactions were terminated with 95% formamide and 2 mm EDTA and resolved in a 20% polyacrylamide gel. Northern Hybridization—Human and mouse multiple tissue Northern blots and human cancer cell line Northern blots of poly(A)+ RNA (Clontech) were hybridized according to the manufacturer's procedures. 100 ng of DNA probe, labeled with the Ready-To-Go DNA labeling beads kit (Amersham Biosciences), was used in each case. Labeled probes were purified with Sephadex G-25 quick spin columns (Roche Molecular Biochemicals). For mouse PolN, an 1193-bp cDNA fragment derived by PCR from EST AA475002 containing motifs 3–6 of the DNA polymerase domain was used as probe. For mouse Hel308, a 507-bp cDNA fragment containing motif VI of the helicase domain was used. The fragment was obtained by 5′-rapid amplification of cDNA ends (SMART RACE cDNA amplification kit, Clontech), subcloned into pT-AdV (Clontech), and sequenced. EcoRI sites in the multiple cloning sites of the vector were used to obtain the fragment used for hybridization. Primers used for the RACE reaction were designed from EST AA517170. For human POLN the 1756-bp COOH-terminal cDNA fragment described above was subcloned into pT-AdV and sequenced. EcoRI sites in the multiple cloning site of the vector were used to release the fragment used for hybridization. For human HEL308 a 580-bp fragment derived by PCR amplification from EST AA625285 and containing motif VI of the helicase domain was used. The 2.0-kb human β-actin control probe (BD Biosciences) strongly hybridizes with mouse and rat actin mRNA and was used for all blots tested. In Situ Hybridization—A 930-bp fragment, including polymerase motifs 3–6 of mouse PolN from EST AA475002, was subcloned into the EcoRI and XbaI sites of pGEM-3Z. A 631-bp fragment, including helicase motif VI of mouse Hel308 from EST AA517170, was subcloned into the XmaI and XbaI sites of pGEM-3Z. 35S-Labeled antisense riboprobes were synthesized with T7 RNA polymerase for both PolN and Hel308 (35S at ∼800 Ci/mmol; Amersham Biosciences). Methods for probe synthesis, mouse tissue preparation, pretreatment, hybridization, washing, and coating of slides for autoradiography were essentially as described in Ref. 10Senior P.V. Critchley D.R. Beck F. Walker R.A. Varley J.M. Development (Camb.). 1988; 104: 431-446PubMed Google Scholar. Post-hybridization washes included an RNase A step to degrade imperfectly hybridized regions, followed by washes to reach a stringency of 0.5 × SSC at 65 °C. Autoradiography was for 18 days, after which sections were developed and Giemsa-stained. Slides were examined using an Olympus BH2 microscope under conventional (bright field) and epi-illumination (dark field) conditions that reveal autoradiographic silver grains as bright particles. As a control to show the presence of hybridizable mRNA, sections of all blocks were hybridized separately with a probe for β-actin mRNA, generated with SP6 RNA polymerase on DraI-linearized plasmid (11Zandvliet D.W. Hanby A.M. Austin C.A. Marsh K.L. Clark I.B. Wright N.A. Poulsom R. Biochim. Biophys. Acta. 1996; 1307: 239-247Crossref PubMed Scopus (46) Google Scholar). Isolation of a Novel A-family DNA Polymerase Gene from Mammalian Genomes—Searching for mammalian homologs of E. coli DNA pol I, we found a mouse expressed sequence tag clone (AA475002) and a human sequence (AF044578) related to pol I but unlinked to POLQ. Reverse transcriptase-PCR was used to obtain longer mouse and human cDNA fragments and complete sequence information (Fig. 1). Because of the strong DNA polymerase similarity and activity described below, the gene was designated as POLN (symbol, pol ν), with agreement from the HUGO nomenclature committee. Both the mouse and human genes span unusually long regions of the genomes and have consequently escaped accurate computational annotation. The human genomic region comprises 24 exons that inhabit 160 kb of genomic DNA, with the longest two introns measuring 26 and 32 kb. The COOH-terminal part of the human gene encoded by exons 14–24 was confirmed by repeated isolation of the same sequence, whereas the NH2-terminal part of the gene had numerous alternatively spliced forms. Of 20 human cDNA clones derived from RNA of cell lines K562 and 833K and sequenced, 14 represented various alternative splices that would produce an inactive protein by exon skipping or inclusion of alternative exons (see Supplemental Material for details). Six of 20 clones (the most frequent class) were for a human cDNA with the sequence in Fig. 1, encoding a protein of 900 amino acids. There was also evidence of alternative splicing in the mouse sequence. Full sequencing of two murine reverse transcriptase-PCR clones lacked one exon encoding residues 539–556. The predicted mouse cDNA encodes a protein of 866 residues (the nucleotide sequences reported in this paper have been submitted to the GenBank™/EBI Data Bank with accession numbers AY136549 and AY135562) (Fig. 1). The proteins predicted by the human and mouse cDNAs are 70% identical and 75% similar over their entire lengths. In the human cDNA, the ATG encoding the first Met residue lies within a sequence compatible with the Kozak consensus for start of translation, with a G at position +4 and an A at position –3. In the mouse cDNA, Met residue 3 (Fig. 1A) is in a sequence context that matches the Kozak consensus better in the two critical positions +4 and –3 than does Met residue 1. The site of transcript polyadenylation in the human sequence was found by cDNA sequencing and occurred 151 nucleotides 3′ of the termination codon, preceded by a likely polyadenylation signal sequence (AGUAAA). The COOH-terminal half of mammalian POLN, beginning at residue 419 of the human protein, is highly related to PolA family enzymes from bacteria and bacteriophage, as well as to the predicted DNA polymerase domains of D. melanogaster Mus308 and human POLQ. Within this pol similarity region, human POLN and E. coli pol I have 33% identity (48% similarity), while human POLN and POLQ have 29% identity (43% similarity) within the equivalent region (see Supplemental Material for details). Motifs found to be conserved throughout the procaryotic PolA family (12Patel P.H. Suzuki M. Adman E. Shinkai A. Loeb L.A. J. Mol. Biol. 2001; 308: 823-837Crossref PubMed Scopus (160) Google Scholar) are also found in POLN as indicated in Fig. 1. With reference to the fingers-thumb-palm structural nomenclature for family A polymerases, the POLN proteins contain well conserved features. These include Motif 1 (at the top of the thumb subdomain), Motif 2a (on one of the thumb helices), Motif 2b (contributing to palm and thumb subdomain contacts with DNA), Motif 3 (also called Motif A), which forms a pocket for the incoming dNTP, Motif 4 (also called Motif B), Motif 5 (also called Motif C), which coordinates a divalent cation, and Motif 6, which interacts with the first template base. From the cDNA cloning and sequencing results we concluded that the mammalian POLN cDNAs are expressed and have the potential to encode active DNA polymerases. There is no sequence indication of 3′ to 5′ or 5′ to 3′ nuclease domains and no significant similarity of residues preceding the pol domain (1–418) to other proteins in the public data base. Chromosomal Localization—Fluorescence in situ hybridization was performed to identify the genomic location of mouse PolN and Hel308 genes. A genomic DNA cosmid library from mouse 129/Ola spleen cells (RessourcenZentrum Primaer-Datenbank) was probed with a 400-bp cDNA fragment derived from EST AA475002 by PCR, containing part of the PolN DNA polymerase domain. Sequencing of 4.3 kbp of a positive genomic clone revealed that it contained motif 3 of PolN. The cosmid was used to hybridize mouse metaphases and showed that PolN mapped to chromosome 5, region C1 (not shown). This region is syntenic to the human chromosome 4p16.2 region where POLN is located in the human genome. The related human POLQ gene is on chromosome 3q13. The same mouse cosmid library was screened with a 500-bp cDNA fragment containing helicase motif VI of Hel308, obtained from NCBI EST number AA517170. Three cosmids were confirmed to contained mouse Hel308 gene fragments and all mapped to chromosome 5, region E by fluorescence in situ hybridization analysis (not shown). This region of the mouse genome is syntenic to human chromosome 4q21 where human HEL308 is encoded. Human POLN Localizes to Nuclei—Analysis of POLN by the PSORT II program (psort.nibb.ac.jp/) predicts nuclear localization, with no mitochondrial signal sequence. Putative nuclear localization signals are present starting at residues 134 and 741 of human POLN (Fig. 1). To experimentally analyze the cellular localization of human POLN, the cDNA encoding the GFP protein was fused in-frame to the amino terminus of POLN (GFP-POLN), and transfected into SV40-transformed MRC5VA fibroblasts. After G418 selection, single clones were isolated. In all clones tested, the tagged protein was found only in cell nuclei, with exclusion from nucleoli. Control clones harboring a construct expressing only GFP showed diffuse distribution of fluorescence in both cytoplasm and nucleus (Fig. 2). POLN Encodes a DNA Polymerase—To determine whether POLN had DNA polymerase activity as predicted from the sequence, human GFP-POLN was expressed in mammalian cells and immunopurified. As a negative control, a single point mutation was created to change an Asp codon (GAC) in the highly conserved polymerase motif 3 (Fig. 1) to an Ala codon (GCC). Substitution of the corresponding Asp abrogates polymerase activity in bacterial pol I (13Patel P.H. Loeb L.A. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 5095-5100Crossref PubMed Scopus (86) Google Scholar). This mutant (D624A) form of POLN was strongly predicted to be catalytically inactive. Recently, Shimazaki et al. (14Shimazaki N. Yoshida K. Kobayashi T. Toji S. Tamai K. Koiwai O. Genes Cells. 2002; 7: 639-651Crossref PubMed Scopus (112) Google Scholar) found that recombinant DNA polymerase λ (pol λ) contains a proline-rich region. Deletion of this region increased pol λ activity by 3–18-fold. The COOH-terminal 35 amino acids of human POLN include 11 proline residues, not present in the mouse PolN sequence (Fig. 1). To test for an effect of the proline-rich domain on activity, a point mutation was introduced to eliminate the last 39 residues of human POLN by changing Trp codon 861 (TGG) to a stop codon (TGA). POLN and the mutant forms were expressed as NH2-terminal GFP tagged proteins by transfection of the human fibroblast cell line MRC5VA. Expression and nuclear localization of the normal and mutant GFP fusion proteins was confirmed by fluorescence microscopy (not shown). From total cell extracts of the transiently transfected MRC5VA cells, tagged POLN was immunoprecipitated using anti-GFP antibody conjugated to protein A-Sepharose beads. POLN and mutant forms were recovered in similar amounts in the immunoprecipitates (Fig. 3A, IPP lanes). Human POLN had DNA polymerase activity as assayed with activated calf thymus DNA as substrate (Fig. 3B). The D624A mutant protein showed no detectable activity above background. POLN was also able to extend DNA on a singly primed template (Fig. 3C). A ladder of products was observed, with the predominant product being extension of the primer by 18 nucleotides to the end of the template. The D624A mutant protein (Fig. 3C, D>A) showed no DNA polymerase activity on the singly primed template in comparison with immunoprecipitations from non-transfected cells or cells containing GFP alone. Deletion of the proline-rich COOH-terminal tail from POLN did not measurably change DNA polymerase activity (Fig. 3, B and C, ΔP constructs). Expression of POLN and HEL308 in Mammalian Cells— Northern hybridization was used to begin an investigation of the expression pattern of POLN in human and mouse cells. Whether expression of POLN and HEL308 overlaps is a point of interest, because each is homologous to portions of POLQ. For this reason, hybridization was done in parallel to investigate HEL308 expression. Northern hybridization of a human tissue Poly(A)+ RNA blot was performed with a probe complementary to the COOH-terminal 1700 nucleotides of human POLN. A full-length POLN transcript is predicted to be a minimum of 3.3 kb long, based on the coding sequence plus 5′- and 3′-untranslated regions. A POLN transcript of about 4 kb was detected in testis tissue (Fig. 4A). A shorter transcript of about 2 kb was also present in most tissues, and was the predominant form detected in heart, liver, skeletal muscle, kidney, and pancreas. This presumably reflects a frequent alternative splice event. Longer transcripts were also apparent in skeletal muscle and heart (Fig. 4A). Overall, expression of POLN was relatively weak and required long exposure times to detect hybridization in human tissues. A full-length HEL308 transcript is predicted to be about 3 kb long. Northern hybridization detected a single transcript of this size in many tissues (Fig. 4A). Expression was strongest in testis and next highest in heart, skeletal muscle, and ovary. A panel of human cancer cell lines was also surveyed by northern hybridization using the same probes (Fig. 4B). All lines examined contained the predominant POLN 2-kb transcript, with the longer, full-length transcript apparent only in the K562 cell line, originating from a chronic myelogenous leukemia. HEL308 transcripts were detected in all cell lines tested, although at varying levels. Northern hybridization of RNA from mouse tissues detected expression of PolN only in testis tissue, with a transcript size of about 4 kb (Fig. 4C). A Hel308 transcript of 3–4 kb was apparent in mouse liver and kidney tissue as well as testis. Because of the predominance of expression of POLN and HEL308 in testis tissue, in situ hybridization was used to more closely examine the production of PolN and Hel308 transcripts in the mouse. Expression of PolN and Hel308 in Testis—Thirty sections from different mouse organs were analyzed by in situ hybridization. Convincing evidence of expression of both Hel308 and PolN was seen by this technique only in seminiferous tubules of the adult testis (Fig. 5). In mammals, the seminiferous epithelium presents germ cells at different stages of development with specific cell types integrated in concentric layers (15Russell L.D. Ettlin R.A. SinhaHikin A.P. Clegg E.D. Histological and Histopathological Evaluation of the Testis. Cache River Press, Clearwater, FL1990Google Scholar, 16Oakberg E.F. Am. J. Anat. 1956; 99: 391-413Crossref PubMed Scopus (738) Google Scholar). Dark field conditions (Fig. 5, B, D, and F) reveal autoradiographic silver grains as bright particles. These show that, while β-actin is expressed in all cell types (Fig. 5F), PolN and Hel308 are expressed only in a subset of cells within the adult testis. PolN expression appeared maximal in meiotic spermatocytes and in post-meiotic round spermatids (Fig. 5, A and B, and data not shown). There appeared to be a graded reduction in labeling across the spermatid layers. Expression was seen neither in elongated spermatids nor in the mature spermatozoa. Labeling was low or absent in the B type spermatogonia and Sertoli cells. There was no expression in interstitial cells including Leydig cells. In situ analysis for Hel308 showed weaker hybridization (Fig. 5, C and D), with maximal signal from primary spermatocytes, but not round spermatids. Sertoli cells, Leydig cells, and other cell types within the testis were not labeled. A Distinct Mus308/POLN Subfamily in Higher Eukaryotes— POLN and the related enzyme POLQ are found in eukaryotic cells, but are not detected in yeast or other fungi. A multiple sequence alignment was performed with 42 PolA family sequences representing a wide variety of bacterial pol I enzymes, bacteriophage DNA polymerases, DNA pol γ enzymes from fungi and higher eukaryotes, and enzymes similar to POLN and POLQ. A tree view of such a sequence alignment reveals relationships between the DNA polymerase region of POLN and that of other family A DNA polymerases (Fig. 6). Bacterial Family A DNA polymerases similar to E. coli pol I comprise one branch. A separate, more distantly related branch is composed of the mitochondrial DNA pol γ enzymes. DNA polymerases from many bacteriophage diverge more widely. The POLN and POLQ enzymes cluster in a distinct branch to form a separate subfamily (Fig. 6). Among nuclear DNA polymerases, POLN is the 14th currently proposed to have activity, the others being pols α, β, δ, ϵ, ζ, η, θ, ι, κ, λ, μ, Rev1, and terminal deoxynucleotidyl transferase. Maga et al. (17Maga G. Shevelev I. Ramadan K. Spadari S. Hubscher U. J. Mol. Biol. 2002; 319: 359-369Crossref PubMed Scopus (39) Google Scholar) detected DNA polymerase activity for a purified 100-kDa protein postulated to be a fragment of human POLQ, but definitive identification as such and proof that full-length POLQ is an active nuclear DNA polymerase have yet to be demonstrated. Similarly, the Drosophila Mus308 gene has been recognized for some time, but clear evidence of the activity of its gene product has not been presented. Lesions that link together the two strands of DNA are among the most toxic forms of DNA damage, and their repair is a challenging problem. Such interstrand cross-links are potent inhibitors of DNA replication and transcription and can induce mutations and DNA rearrangements. Cells are exposed to environmental agents that cause ICLs, such as furocoumarins, and agents that form ICLs are widely used in cancer chemotherapy and phototherapy. In E. coli and in Saccharomyces cerevisiae, both nucleotide excision repair and homologous recombination are involved in interstrand cross-link repair (18Jachymczyk W.J. von Borstel R.C. Mowat M.R. Hastings P.J. Mol. Gen. Genet. 1981; 182: 196-205Crossref PubMed Scopus (158) Google Scholar). To complete repair, E. coli uses DNA polymerase I and the UvrD DNA helicase (19Sladek F.M. Munn M.M. Rupp W.D. Howard-Flanders P. J. Biol. Chem. 1989; 264: 6755-6765Abstract Full Text PDF PubMed Google Scholar, 20Van Houten B. Gamper H. Hearst J.E. Sancar A. J. Biol. Chem. 1988; 263: 6553-6560Abstract Full Text PDF Google Scholar). In mammalian cells comparable processes may also operate (21Thompson L.F. Mutat. Res. 1996; 363: 77-88Crossref PubMed Scopus (97) Google Scholar, 22Hoy C. Thompson L. Mooney C. Salazar E. Cancer Res. 1985; 45: 1737-1743PubMed Google Scholar, 23Howlett N.G. Taniguchi T. Olson S. Cox B. Waisfisz Q. De Die-Smulders C. Persky N. Grompe M. Joenje H. Pals G. Ikeda H. Fox E.A. D'Andrea A.D. Science. 2002; 297: 606-609Crossref PubMed Scopus (964) Google Scholar). Eukaryotes also appear to have additional pathways of cross-link repair which use DNA polymerases that can bypass lesions (24De Silva I.U. McHugh P.J. Clingen P.H. Hartley J.A. Mol. Cell. Biol. 2000; 20: 7980-7990Crossref PubMed Scopus (386) Google Scholar, 25Grossmann K.F. Ward A.M. Matkovic M.E. Folias A.E. Moses R.E. Mutat. Res. 2001; 487: 73-83Crossref PubMed Scopus (87) Google Scholar, 26Zheng H. Wang X. Warren A.J. Legerski R.J. Nairn R.S. Hamilton J.W. Li L. Mol. Cell. Biol. 2003; 23: 754-761Crossref PubMed Scopus (130) Google Scholar). Mammalian POLQ is predicted to encode both a DNA helicase and a DNA polymerase. Because HEL308 and POLN are, respectively, most similar to the helicase and polymerase domains of POLQ, it is possible that POLN and HEL308 are related to POLQ by an ancient gene duplication event. The close similarity between mouse and human POLN, and between mouse and human POLQ (Fig. 6), suggest that continued selection for distinct functions of the two enzymes has operated during mammalian evolution. It is possible that POLN and HEL308 may work together in the cell. Our investigation of the expression pattern of POLN and HEL308 showed a partial overlap, consistent with this possibility. We looked for co-immunoprecipitation of additional human proteins in cells expressing GFP-POLN by mass spectrometric analysis of proteins after two-dimensional electrophoresis. Although several heat-shock proteins were co-immunoprecipitated, HEL308 was not found. 2F. Marini, J. Minden, and R. D. Wood, unpublished data. Proliferating cell nuclear antigen, often associated with DNA polymerases as a cofactor, was also absent from the GFP-POLN immunoprecipitate. The precise function of POLN in the cell is yet unknown, a situation that is also presently true for most nuclear DNA polymerases. A possible function of the POLN/POLQ polymerase subfamily in DNA repair function was suggested recently by Shima et al. (27Shima N. Hartford S.A. Duffy T. Wilson L.A. Schimenti K.J. Schimenti J.C. Genetics. 2003; 163: 1031-1040PubMed Google Scholar). Mouse PolQ was proposed as a candidate for a gene that leads to spontaneous chromosome instability as measured by formation of micronuclei in reticulocytes. It is also intriguing that chromosome 4p16.2, where human POLN is located, is found to be deleted in about 50% of breast carcinomas (28Shivapurkar N. Sood S. Wistuba II Virmani A.K. Maitra A. Milchgrub S. Minna J.D. Gazdar A.F. Cancer Res. 1999; 59: 3576-3580PubMed Google Scholar). Considering its high expression in human and mouse testis tissue, another potential role for POLN is in meiosis. Previous studies have found that the mouse Polk, Poll, and Poli genes, respectively, encoding the specialized DNA polymerases dinB (pol κ), pol λ, and pol ι, are also highly expressed in the seminiferous tubules of the mouse testis (29Velasco-Miguel S. Richardson J.A. Gerlach V.L. Lai W.C. Gao T. Russell L.D. Hladik C.L. White C.L. Friedberg E.C. DNA Repair (Amst.). 2003; 2: 91-106Crossref PubMed Scopus (72) Google Scholar, 30McDonald J.P. Rapic-Otrin V. Epstein J.A. Broughton B.C. Wang X. Lehmann A.R. Wolgemuth D.J. Woodgate R. Genomics. 1999; 60: 20-30Crossref PubMed Scopus (181) Google Scholar, 31Aoufouchi S. Flatter E. Dahan A. Faili A. Bertocci B. Storck S. Delbos F. Cocea L. Gupta N. Weill J.C. Reynaud C.A. Nucleic Acids Res. 2000; 28: 3684-3693Crossref PubMed Google Scholar). However, the expression patterns of mRNA of the four polymerases in testis is somewhat different. While PolN and DinB are expressed in meiotic spermatocytes and in post-meiotic round spermatids, Poli is expressed in earlier stages of spermatogenesis. The expression pattern of Poll in seminiferous tubules is uniform. A high level of expression in testis does not necessarily restrict PolN or other DNA polymerases to a specialized role in this tissue. For example, POLI is thought to participate in somatic hypermutation of immunoglobulin genes during B lymphocyte development (32Faili A. Aoufouchi S. Flatter E. Gueranger Q. Reynaud C.A. Weill J.C. Nature. 2002; 419: 944-947Crossref PubMed Scopus (176) Google Scholar). With respect to POLN, we found cDNAs expressed in human (K562) and mouse (L1210) cell lines, of hematopoeitic origin, implying that POLN may have roles in tissues other than testis. Gathering of more information on the biochemical characteristics of POLN may be expected to provide clues regarding its cellular functions. We thank Richard Poulsom for in situ hybridization analysis and Jill Williamson and Denise Sheer for fluorescence in situ hybridization mapping (all at the Imperial Cancer Research Fund, now Cancer Research UK). We are grateful to Jonathan Minden and Lei Gong (Carnegie Mellon University) for guidance and facilities in mass spectrometric analysis and Cathy Joyce (Yale University) and our laboratory colleagues for discussions. Download .pdf (.13 MB) Help with pdf files
Referência(s)