© 1992 Oxford University Press
Human Molecular Genetics, Vol. 1, No. 9
677-680
Human apurinic endonuclease gene (APE): structure and genomic mapping (chromosome 14q11.2 —12) Lynn Harrison, Gian Ascione, Joan C.Menninger1, David C.Ward1 and Bruce Demple* Department of Molecular and Cellular Toxicology, Harvard School of Public Health, 665 Huntingdon Avenue, Boston, MA 02115 and departments of Molecular Biophysics and Biochemistry and of Genetics, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA Received October 12, 1992; Accepted October 28, 1992
GenBank accession no. M99703
ABSTRACT multifunctional class II AP endonuclease is also present in most mammalian cells and has been purified extensively from HeLa extracts (10). The cellular function of the human enzymes in the repair of oxidative and other DNA lesions is starting to be addressed. The cDNA encoding the major human AP endonuclease has been independently cloned from a placental library (APE; ref. 13), a tumor cell library (HAP1; ref. 14), and a bone marrow library (APEX; ref. 15). The proteins predicted from these three cDNAs are essentially identical and are members of a family of enzymes that includes E.coli exonuclease HI, which functions in repair of oxidative and alkylation damages (12, 16). The possible involvement of AP endonucleases in human genetic disease has been addressed for specific syndromes (e.g., ataxia telangiectasia (10)) at the level of enzyme activity. Information on the location and structure of the APE gene in the human genome would help focus efforts to define such a connection, both by eliminating some candidate syndromes and by suggesting others not previously considered. We report here the genomic cloning of the APE from a human placental library and the mapping of the gene to chromosome 14ql 1.2 — 12.
INTRODUCTION Cellular DNA is under constant assault from both environmental and intracellular agents that cause mutagenic damage. Key endogenous mutagens in eukaryotes include oxygen radicals generated by aerobic metabolism (1,2) and some non-oxidative agents (3). Accurate repair of these DNA damages is necessary to maintain genetic integrity and avert carcinogenesis. Abasic (AP) sites are a key type of damage produced by many agents directly (4) and indirectly, through the action of DNA Nglycosylases on damaged bases (5-7). Unrepaired AP sites, which represent a loss of genetic information, block DNA synthesis and are mutagenic (4, 5). The main cellular enzymes that initiate repair of AP sites are the hydrolytic (class II) AP endonucleases, which incise DNA at the immediate 5' side of AP sites (5-8). The major human AP endonuclease is a multifunctional, 37-kDa protein that also removes nucleotide fragments from 3' termini at DNA strand breaks induced by oxidative damage (9, 10). Such termini, notably 3'-phosphates and 3'-phosphoglycolate esters, are present at the DNA strand breaks induced by ionizing radiation in vitro (11) and by hydrogen peroxide in vivo (12). A separate • To whom correspondence should be addressed
RESULTS Molecular cloning, sequence and structure analysis Screening a human placental genomic library with the APE cDNA (13) yielded two positive clones (XAPE1 and XAPE2) from 90,000 plaques. These clones contained human genomic inserts 18 kb (XAPE1) and 14 kb (XAPE2), respectively. The 14-kb insert was subcloned into pACYC177 (18) to produce the plasmid pAPE2. A physical map of the pAPE2 insert is shown in Fig. 1A. The position of APE within the 14-kb genomic fragment was established initially by a combination of restriction mapping and Southern blotting (data not shown), and confirmed unambiguously by DNA sequencing (see below). The genomic DNA was sequenced from the Pstl site just upstream of APE across a segment of 3 kb (GenBank accession number M99703). A comparison of this sequence with that of the APE cDNA (13)revealedthe positions and sizes of the introns (Fig. IB). The APE gene contains five exons, the first of which is not translated, and four introns. All four introns are bracketed by consensus splice donor/acceptor sequences (GT/AG). The
Downloaded from http://hmg.oxfordjournals.org/ at McMaster University Library on June 29, 2015
Abasic (AP) sites in DNA are produced spontaneously and by many genotoxic agents. Hie repair of such damages is initiated by AP endonucleases, which are evidently ubiquitous. We employed the recently cloned cDNA, APE, that encodes the major human AP endonuclease, to isolate large genomic fragments that contain the intact APE gene. Hie sequence of 3 kb encompassing APE was determined (GenBank Accession No. M99703). The APE gene contains four small introns (ranging 130 to 566 bp) and five exons, the first of which is untranslated. Hie 0.5 kb of DNA sequence upstream of APE did revealed only a possible CCAAT box, but no other regulatory sites or a TATA box, consistent with the constitutive expression of AP endonuclease activity observed in other studies. The location of APE in the human genome was mapped to chromosome 14, bands q l l . 2 - 1 2 , by fluorescence in situ hybridization of meta phase cells with DNA from the genomic clones and subclones. Although this locus has not been associated causally with genetic diseases of DNA repair, some translocations that affect 14qll.2-12 could compromise APE and lead to genetic instability.
678 Human Molecular Genetics, Vol. 1, No. 9 A
1 kb
PLHBS2
S H
P
H SI
I
I I
S
P
APE
ATG
B
Figure 1. A. Physical map of APE genomic clone. Plasmid pAPE2 was digested with Xhol (X), Smal (S), Hiruffll (H), Pstl (P), Sacl (SI) or BamHl (B) to obtain the physical map of the 14.0-kb genomic DNA insert from XAPE2. The location of APE within the insert was established by a combination of Southern blotting (17) and comparison to the cDNA map, and verified by the sequencing of the entire APE gene (see text). The fragment subcloned into pLHBS2 is indicated. B. Structure of the APE gene. The region shown corresponds to that determined by DNA sequencing, and the exon/intron boundaries identified by comparison to the cDNA sequence (13). Hatched boxes, untranslated exons; filled boxes, coding exons; open boxes, introns; thin horizontal line, sequence outside reported cDNAs.
junctions of introns 1 and 4 contain short direct repeats (Fig. 2A), which are also frequently observed at splice sites (19). The sequences of the exons match the placental APE cDNA sequence, except for changes at four positions. Two of these affect the predicted Ape protein, but not the previously noted homology to the other AP endonucleases (13): bp 1225-1227 and bp 2668-2670 are both GGC codons, rather than GCG, encoding glycines rather than alanines. Two other changes lie outside the coding sequence (the C at position 833 was not reported in the cDNA, and T was found at position 2817 instead of the G reported for the cDNA). All of these changes are consistent with the cDNA sequences reported recently by others (20, 21). The placental cDNA also contained a deletion of 66 bp relative to the cDNAs from three tumor cell lines (14, 20, 21). This difference is probably not due to an alternatively spliced mRNA: consensus splice junctions are not present at the deletion endpoints (bp 536—784 in the genomic sequence, which contains intron 1), but short direct repeats (GCAA; bp 520-533 and 779-782) could have mediated this deletion during the cloning and amplification of the cDNA in bacterial cells (17). The DNA sequence ~ 0.5 kb upstream of the APE start codon was analyzed for possible control sequences (Fig. 2B). Since the APE mRNA is - 1 . 6 kb (ref. 21, and our unpublished data), this segment must include the start site for APE transcription. No 'TATA' box was observed in this upstream region. Apart from a CCAAT sequence at bp 340-344 (underlined in Fig. 2B), no known transcriptional control sites were discovered by this analysis. However, the upstream region of APE is relatively GCrich, with 12 CpG sites contained in bp 303-407 alone. Chromosome mapping The location of APE in the human genome was determined by fluorescence in situ hybridization (22). Both the XAPE1 and the XAPE2 phage probes produced clear signals at high stringency. These genomic probes gave strong hybridization in all metaphase spreads to chromosome 14, in bands q l l . 2 - 1 2 . Occasionally, a fainter hybridization to chromosome 4 (the ql3 band) was also
INTRON 1 163bp INTRON 2
Junction 596 1
566bp INTRON 4 130bp
Junction 778 1
905 1
1114
1303 1
1868 1 nCJkOrmo
210bp INTRON 3
3'
2062 1 OCUiOOTO. . . .
1
2191 1 TlTWlmi
B *«tl
60 I I I I I cc CTGCAATA0CACTGGGA ACCCCCCMCTCCCCCGASaa 61 120 I I CTCCACOGACCTCTCTTC 121 180 I I CTATCO1111 TlCTAICIC! I ICCCCTgiTCaLCCCCaCCXTTCTCCACTttTITrTT'lCC 161 240 I I 241 I
300 I
301
360 I
I 361 I
420 I
421 I
480 I
491 I
lindltt
CGCGGTTCCTC111 i GCTCATAAGAgyagTTCCCTGG(aCTCTG*ACSGCAACCTT. . .
Figure 2. Intron/exon boundaries and upstream region of the APE gene. A. Splice site junctions. Intron sequences are shown in bold type, exons in small capital letters, and direct repeat sequences underlined. B. Upstream sequence. The DNA sequence is shown from the Pstl site (upstream of the APE) to the HindlU site (located in exon 1). The Pstl, Hindm and CCAAT motif (bp 340-344) sites are shown in italics.
Downloaded from http://hmg.oxfordjournals.org/ at McMaster University Library on June 29, 2015
1 kb
Human Molecular Genetics, Vol. 1, No. 9 679
DISCUSSION The structural and sequence analysis of APE presented here reveals an unusually compact human gene, due primarily to its small introns. The ~ 30-kDa C-terminal segment of the predicted Ape protein, which belongs to a family of AP endonucleases that includes E. coli exonuclease in (13), is encoded over the last three exons of APE, while the first exon is noncoding. The ~0.5 kb sequence upstream of APE did not reveal obvious regulatory sites, except for a CCAAT box. APE might not be regulated as part of a known control system, and the high frequency of CpG dinucleotides upstream of APE and the lack of a TATA box is a common feature of 'housekeeping' genes (23). The AP endonuclease is expressed at a relatively high level in HeLa cells and in T-lymphoblasts ( - 7 X 1 0 6 molecules per cell; ref. 10). Moreover, the level of the /tP£-encoded enzyme was not significantly increased in HeLa cells (or in CHO cells) by the oxidative stress exerted by high levels of oxygen (10). The APE AP endonuclease may perform functions that are important in a variety of circumstances. The likely role of the AP endonuclease in correcting mutagenic DNA damages from both environmental and endogenous sources, as demonstrated for the major enzyme in yeast (3), would correspond to such a function. The Ape protein has also recendy been shown to restore DNA binding activity to oxidized Jun/Fos heterodimers in vitro (21), although the in vivo significance of this proposed redox reaction has not been established. Such dual roles might help account for the abundance of Ape in human cells. The site of APE in the human genome, the ql 1.2 —12 band of chromosome 14*, is not the locus of any known human genetic disease thought to involve DNA repair (Human Genome Database, Johns Hopkins University, Baltimore, Md.; most recent search 7 August 1992). Purine nucleoside phosphorylase deficiency maps to 14ql2, but this defect seems unrelated to the AP endonuclease, which lacks phosphate-dependent activities and acts only on DNA, not nucleosides (10, 24). However, 14ql 1.2 — 12 is frequently involved in cytologically-detectable
B
Figure 3. Chromosome localization of APE genomic clones. A. Metaphase spread after hybridization with biotinylated pLHBS2 and DAPI staining. The FTTC hybridization signal has been pseudo-colored red (white arrows) and the DAPI staining pattern green. B. A G-banded idiognun of human chromosome 14 displaying the range of hybridization signal locations observed on 15 metaphase spreads probed with pLHBS2. C. A composite of chromosomes 14 displaying signals produced by pLHBS2 probes. Pseudo-coloring as in A.
chromosome rearrangements (translocations, inversions) in patients with the cancer-prone genetic disease ataxia telangiectasia (AT). The mutations responsible for the three most common complementation groups of AT map to the long arm of chromosome 11 (25). Some of the rearrangements found in AT lymphocytes involving 14ql 1.2 include T-cell receptor genes (26, 27). These rearrangements correlate with the predisposition toward T cell leukemia associated with AT (25). Other AT-associated cancers (25, 28), or AT-independent neoplasias might be connected to translocations in this region which inactivate or disregulate the APE gene, and thus elevate die spontaneous mutation rate (29). This possibility can now be addressed experimentally by examining tumor cells with other known rearrangements involving 14ql 1.2 —12 for effects on the structure or expression of APE.
* As this manuscript was being prepared, we were made aware of independent experiments that confirm the DNA sequence and the chromosome mapping presented here, except that APE was assigned solely to 14ql2; that work subsequently appeared in Zhao, B., Grandy, D.K., Hagerup, J.M., Magenis, R.E., Smith, L., Chauhan, B.C., Henner, W.D. (1992) Nud. Acids Res. 20, 4097-4098.
Downloaded from http://hmg.oxfordjournals.org/ at McMaster University Library on June 29, 2015
detected. Since only single-copy hybridization was observed in Southern blots probed with the APE cDNA (ref. 21, and our unpublished data), it seemed likely that the secondary hybridization to chromosome 4 was due to one of the APE introns or to a DNA segment adjacent to APE on the genomic probe fragments. Indeed, when the subclone pLHBS2 was used as an in situ probe, hybridization was observed only to chromosome 14 (Fig. 3A). Plasmid pLHBS2 contains most of the APE gene: the 8.1-kb Sacl-Xhol fragment that includes the 3' end of exon 2, and all of exons 3 - 5 and introns 2 - 4 (Fig. 1). Thus, the faint hybridization to chromosome 4ql3 was evidently due to DNA upstream of APE, to intron 1, or to exons 1 or 2. The possibility has not been eliminated that an APE pseudogene or a gene related to APE resides on chromosome 4. The fine mapping to bands ql 1.2 —12 (Fig. 3B) was obtained by comparing signals in a composite made from from several metaphases (Fig. 3C). The horizontal bar indicated in the ideogram in Fig. 3B indicates the range of signals obtained with fifteen different metaphase spreads.
680 Human Molecular Genetics, Vol. 1, No. 9
REFERENCES 1. Ames, B.N. (1983) Science 221,1256-1264. 2. Sies, H. (1991) American J. Med. 91, 31S-38S. 3. Ramotar, D., Popoff, S.C., Gralla, E.B. and Demple, B. (1991) Mol. Cell. Biol. 11, 4537-4544. 4. Loeb, L.A., and Preston, B.D. (1986) Annu. Rev. Gtnei. 20, 201-230. 5. Wallace, S.S. (1988) Envir. Molec. Mulagen. 12, 431-77. 6. Sancar, A., and Sancar, G.B. (1988) Annu. Rev. Biochem. 57, 29-67. 7. Demple, B., and Levin, J.D. (1991) in Sies, H. (ed.), Oxidative Stress: Oxidants and Antioxidants. Academic Press, London, pp. 119-154. 8. Levin, J.D., and Demple, B. (1990) Nudeic Acids Res. 18, 5069-5075. 9. Demple, B., Greenberg, J.T., Johnson, A.W., and Levin, J.D. (1988) In Friedberg, E.C., and Hanawalt, P.C. (ed.), Mechanisms and Consequences of DNA Damage Processing, UCLA symposia on Molecular and Cellular Biology, New Series. Alan R. Liss, New York, Vol. 83, pp. 159-166. 10. Chen, D.S., Herman, T., and Demple, B. (1991) Nucleic Acids Res. 19, 5907-5914. 11. Henner, W.D., Rodriguez, L.O., Hecht, S.M., Haseltine, W.A. (1983)7. Biol. Chem. 258, 711-713. 12. Demple, B., Johnson, A.W., and Fung, D. (1986). Proc. Nail. Acad. Sci. (USA) 83, 7731-7735. 13. Demple, B., Herman, T., Chen, D.S. (1991) Proc. Nail. Acad. Sci. (USA) 88, 11450-11454. 14. Robson, C.N., and Hickson, I.D. (1991) Nucleic Acids Res. 19,5519-5523. 15. Seki.S., Hatsushika, M., Watanabe, S., Akiyama, K., Nagao, K., Tsutsui, K. (1992) Biochim. Biophys. Ada, in press. 16. Cunningham, R.P., Saporito, S.M., Spitzer, S.G., and Weiss, B. (1986) J. Bacteriol. 168, 1120-1127. 17. Sambrook, J., Fritsch, E.F., Maniatis, T. (1989) Molecular cloning: A laboratory manual. (Second edition.) Cold Spring Harbor Laboratory Press. 18. Chang, A.C.Y., and Cohen, S.N. (1978) J. Bacteriol. 134, 1141 -1156. 19. Brunak, S., Engelbrecht, J., and Knudsen, S. (1991) /. Mol. Biol. 220, 49-65. 20. Cheng, X.B., Bunville, J., Patterson, T.A. (1992) Nucleic Acids Res. 20, 370. 21. Xanthoudakis, S., Miao, G., Wang, F., Pan, Y.-C. E., Curran, T (1992) EMBOJ. 11, 3323-3335. 22. LJchter, P. Tang,C.-J.C, CaU,K., Hermanson, G., Evans, G.A., Housman, D., Ward, D.C. (1990) Science 247,64-69. 23. Dynan, W.S. (1986) Trends in Genetics 2, 196-197. 24. Kane, C M . , and Linn, S. (1981) J. Biol. Chem. 256, 3405-3414. 25. Gatti, R.A., Boder, E., Vinters, H.V., Sparkes, R.S., Norman, A., and Lange, K. (1991) Medicine 70, 99-117. 26. Stem, M.-H., Zhang, F., Griscelli, C , Thomas, G., Aurias, A (1988) Hum. Genet. 78, 33-36. 27. Humphreys, M.W., Nevin, N.C., and WooMridge, M.A. (1989) Human Genetics 83, 79-82. 28. Swift, M., Morrell, D., Massey, R.B., and Chase, C.L. (1991) New Eng. J. Med. 325, 1831-1836. 29. Loeb, L.A. (1991) Cancer Res. 51, 3075-3079. 30. Brigatti, DJ., Meyerson, D., Leary, J J., Spalholtz, B., Travis, S.Z., Fong, C.K.Y., Hsiung, C D . , and Ward, D.C. (1983) Virology 126, 32-50. 31. Baldini, A., and Ward, D.C. (1991) Genomics 9, 770-774.
The human placental genomk library in the EMBL3 vector was obtained from Clontech (product no. HL 1067J). The vectors pACYC177 (18) and Bluescript SK (17) were obtained from New England BioLabs and Stratagene, respectively. Recombinant plasmids were propagated in E. coli strain DH5or bacteria cultured at 37 °C in Luria-Bertani (LB) broth or agar supplemented with the appropriate antibiotic (17). Restriction enzymes were purchased from New England BioLabs or BRL, Klertow fragment DNA polymerase from New England BioLabs, and Sequenase™ from US Biochemicals; all the enzymes were used according to the supplier's instructions. Molecular cloning and sequence analysis of a fragment mumming the human APE geae A placental geuomic library in the EMBL3 vector was screened using the 1.3-kb human APE cDNA (13) as a probe. Two positive clones, designated XAPE1 and XAPE2, were identified and purified through three successive screens. XAPE1 and XAPE2 had human genomic inserts of 18 kb and 14 kb, respectively. Digestion with Xhol released 14-kb and 4-kb fragments from XAPE1, and a single 14-kb fragment from XAPE2. The 14-kb fragments from XAPE1 and XAPE2 were subcloned into pACYC177, to produce the plasmids pAPEl and pAPE2, respectively. Restriction mapping indicated that the genomic insert in pAPEl overlaps the 5' - 9 kb of the insert in pAPE2, but includes — 5 kb of additional upstream sequence (data not shown). Fragments of 4.1 kb and 8.1 kb, each containing a portion of the APE gene, were isolated by digestion of pAPE2 with Hindm, or with Xhol and Sad, and ligated into the MncflH-digested or SacVXholdigested Bluescript SK to yield pLHBSl and pLHBS2, respectively. pAPE2, pLHBSl and pLHBS2 were used to sequence the APE gene using the dideoxy chain termination method (17) with Sequenase and the M13 reverse, - 2 0 and - 4 0 primers (US Biochemicals) and additional synthetic oligonudeotides (5'-ACTCCTGGCAATGTCAG-3' and oligonucleotides corresponding positions 138-122, 421-405, 419-436, 511-495, 676-692, 837-819, 967-983, 1090-1108, 1127-1111, 1454-1470, 1647-1631, 1980-1963, 1988-1971, 2303-2287, 2362-2378, 2510-2494, 2609-2625, 2718-2702, 2935-2919 in the sequence deposited in GenBank; Accession number M99703). Chromosome mapping The chromosomal location of APE was determined by fluorescence in situ hybridization (22), initially with probes from XAPE1 and XAPE2, and finally with pLHBS2 (Fig. 3). Briefly, purified DNA (17) was labeled with biotin-11-dUTP (Sigma) by nick translation (30). The size of the labeled probe ranged 200-500 nudeotides, and unincorporated nuckotides were separated from the probe DNA using Sephadex G-50 spin columns. Metaphase chromosome spreads were prepared by standard techniques. Denaturation and prcannealing of 2 - 4 >ig Cot I DNA (BRL), 10 »ig of DNase-treated salmon sperm DNA and 150 ng of biotinylated probe DNA was performed as described (22). Following 16—18 hr incubation at 37°C and post-hybridization washes, the slides were incubated with blocking solution and the biotin-labeled probe was detected using 5 |ig/mlfluoresceinisothiocyanate (FITO-conjugated avidin (Vector Laboratories). The banding pattern used for chromosome identification and visual localization of the probes was produced by diamidinopnenylindole (DAPI) staining (0.2 /ig/ml, 10 min), which produces a G-like banding pattern. Fluorescence signals were imaged using a Zeiss Axioscope 20 epifluorescence microscope equipped with a cooled CCD camera (Photometries, CH220), and image acquisition was performed using an Apple Macintosh II computer. The gray scale images were pseudo-colored and merged electronically (31). Image processing and merging was performed on a Macintosh FJci computer using Gene Join Max Pix (software by Tun Rand, laboratory of D.C.W., Yak). Photographs were taken directly from the computer monitor.
ACKNOWLEDGEMENTS We thank Dr. Davis S.Chen for advice, Dr. Tom Graf (Molecular Biology Computer Research Resource, Harvard) for his assistance in searching the Human Genome Database, and Drs. W.D.Hermer and T.Curran for communicating results before publication. This work was supported by grants from the NTH (GM40000 to B.D. and HG00272 to D.C.W.) and from the Markey Foundation (to the Department of Molecular and Cellular Toxicology, Harvard School of Public Health).
ABBREVIATIONS AP, abasic or arxirink/apyrimkunic; kDa, kilodalton; AT, ataxia telangicctasia; LB, Luria-Bertani broth; FITC, ftuorescein isothiocyanate, DAPI, diamidinopnenylindole.
Downloaded from http://hmg.oxfordjournals.org/ at McMaster University Library on June 29, 2015
MATERIALS AND METHODS Materials
Human Molecular Genetics, Vol. 1, No. 9
677-680
Human apurinic endonuclease gene (APE): structure and genomic mapping (chromosome 14q11.2 —12) Lynn Harrison, Gian Ascione, Joan C.Menninger1, David C.Ward1 and Bruce Demple* Department of Molecular and Cellular Toxicology, Harvard School of Public Health, 665 Huntingdon Avenue, Boston, MA 02115 and departments of Molecular Biophysics and Biochemistry and of Genetics, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA Received October 12, 1992; Accepted October 28, 1992
GenBank accession no. M99703
ABSTRACT multifunctional class II AP endonuclease is also present in most mammalian cells and has been purified extensively from HeLa extracts (10). The cellular function of the human enzymes in the repair of oxidative and other DNA lesions is starting to be addressed. The cDNA encoding the major human AP endonuclease has been independently cloned from a placental library (APE; ref. 13), a tumor cell library (HAP1; ref. 14), and a bone marrow library (APEX; ref. 15). The proteins predicted from these three cDNAs are essentially identical and are members of a family of enzymes that includes E.coli exonuclease HI, which functions in repair of oxidative and alkylation damages (12, 16). The possible involvement of AP endonucleases in human genetic disease has been addressed for specific syndromes (e.g., ataxia telangiectasia (10)) at the level of enzyme activity. Information on the location and structure of the APE gene in the human genome would help focus efforts to define such a connection, both by eliminating some candidate syndromes and by suggesting others not previously considered. We report here the genomic cloning of the APE from a human placental library and the mapping of the gene to chromosome 14ql 1.2 — 12.
INTRODUCTION Cellular DNA is under constant assault from both environmental and intracellular agents that cause mutagenic damage. Key endogenous mutagens in eukaryotes include oxygen radicals generated by aerobic metabolism (1,2) and some non-oxidative agents (3). Accurate repair of these DNA damages is necessary to maintain genetic integrity and avert carcinogenesis. Abasic (AP) sites are a key type of damage produced by many agents directly (4) and indirectly, through the action of DNA Nglycosylases on damaged bases (5-7). Unrepaired AP sites, which represent a loss of genetic information, block DNA synthesis and are mutagenic (4, 5). The main cellular enzymes that initiate repair of AP sites are the hydrolytic (class II) AP endonucleases, which incise DNA at the immediate 5' side of AP sites (5-8). The major human AP endonuclease is a multifunctional, 37-kDa protein that also removes nucleotide fragments from 3' termini at DNA strand breaks induced by oxidative damage (9, 10). Such termini, notably 3'-phosphates and 3'-phosphoglycolate esters, are present at the DNA strand breaks induced by ionizing radiation in vitro (11) and by hydrogen peroxide in vivo (12). A separate • To whom correspondence should be addressed
RESULTS Molecular cloning, sequence and structure analysis Screening a human placental genomic library with the APE cDNA (13) yielded two positive clones (XAPE1 and XAPE2) from 90,000 plaques. These clones contained human genomic inserts 18 kb (XAPE1) and 14 kb (XAPE2), respectively. The 14-kb insert was subcloned into pACYC177 (18) to produce the plasmid pAPE2. A physical map of the pAPE2 insert is shown in Fig. 1A. The position of APE within the 14-kb genomic fragment was established initially by a combination of restriction mapping and Southern blotting (data not shown), and confirmed unambiguously by DNA sequencing (see below). The genomic DNA was sequenced from the Pstl site just upstream of APE across a segment of 3 kb (GenBank accession number M99703). A comparison of this sequence with that of the APE cDNA (13)revealedthe positions and sizes of the introns (Fig. IB). The APE gene contains five exons, the first of which is not translated, and four introns. All four introns are bracketed by consensus splice donor/acceptor sequences (GT/AG). The
Downloaded from http://hmg.oxfordjournals.org/ at McMaster University Library on June 29, 2015
Abasic (AP) sites in DNA are produced spontaneously and by many genotoxic agents. Hie repair of such damages is initiated by AP endonucleases, which are evidently ubiquitous. We employed the recently cloned cDNA, APE, that encodes the major human AP endonuclease, to isolate large genomic fragments that contain the intact APE gene. Hie sequence of 3 kb encompassing APE was determined (GenBank Accession No. M99703). The APE gene contains four small introns (ranging 130 to 566 bp) and five exons, the first of which is untranslated. Hie 0.5 kb of DNA sequence upstream of APE did revealed only a possible CCAAT box, but no other regulatory sites or a TATA box, consistent with the constitutive expression of AP endonuclease activity observed in other studies. The location of APE in the human genome was mapped to chromosome 14, bands q l l . 2 - 1 2 , by fluorescence in situ hybridization of meta phase cells with DNA from the genomic clones and subclones. Although this locus has not been associated causally with genetic diseases of DNA repair, some translocations that affect 14qll.2-12 could compromise APE and lead to genetic instability.
678 Human Molecular Genetics, Vol. 1, No. 9 A
1 kb
PLHBS2
S H
P
H SI
I
I I
S
P
APE
ATG
B
Figure 1. A. Physical map of APE genomic clone. Plasmid pAPE2 was digested with Xhol (X), Smal (S), Hiruffll (H), Pstl (P), Sacl (SI) or BamHl (B) to obtain the physical map of the 14.0-kb genomic DNA insert from XAPE2. The location of APE within the insert was established by a combination of Southern blotting (17) and comparison to the cDNA map, and verified by the sequencing of the entire APE gene (see text). The fragment subcloned into pLHBS2 is indicated. B. Structure of the APE gene. The region shown corresponds to that determined by DNA sequencing, and the exon/intron boundaries identified by comparison to the cDNA sequence (13). Hatched boxes, untranslated exons; filled boxes, coding exons; open boxes, introns; thin horizontal line, sequence outside reported cDNAs.
junctions of introns 1 and 4 contain short direct repeats (Fig. 2A), which are also frequently observed at splice sites (19). The sequences of the exons match the placental APE cDNA sequence, except for changes at four positions. Two of these affect the predicted Ape protein, but not the previously noted homology to the other AP endonucleases (13): bp 1225-1227 and bp 2668-2670 are both GGC codons, rather than GCG, encoding glycines rather than alanines. Two other changes lie outside the coding sequence (the C at position 833 was not reported in the cDNA, and T was found at position 2817 instead of the G reported for the cDNA). All of these changes are consistent with the cDNA sequences reported recently by others (20, 21). The placental cDNA also contained a deletion of 66 bp relative to the cDNAs from three tumor cell lines (14, 20, 21). This difference is probably not due to an alternatively spliced mRNA: consensus splice junctions are not present at the deletion endpoints (bp 536—784 in the genomic sequence, which contains intron 1), but short direct repeats (GCAA; bp 520-533 and 779-782) could have mediated this deletion during the cloning and amplification of the cDNA in bacterial cells (17). The DNA sequence ~ 0.5 kb upstream of the APE start codon was analyzed for possible control sequences (Fig. 2B). Since the APE mRNA is - 1 . 6 kb (ref. 21, and our unpublished data), this segment must include the start site for APE transcription. No 'TATA' box was observed in this upstream region. Apart from a CCAAT sequence at bp 340-344 (underlined in Fig. 2B), no known transcriptional control sites were discovered by this analysis. However, the upstream region of APE is relatively GCrich, with 12 CpG sites contained in bp 303-407 alone. Chromosome mapping The location of APE in the human genome was determined by fluorescence in situ hybridization (22). Both the XAPE1 and the XAPE2 phage probes produced clear signals at high stringency. These genomic probes gave strong hybridization in all metaphase spreads to chromosome 14, in bands q l l . 2 - 1 2 . Occasionally, a fainter hybridization to chromosome 4 (the ql3 band) was also
INTRON 1 163bp INTRON 2
Junction 596 1
566bp INTRON 4 130bp
Junction 778 1
905 1
1114
1303 1
1868 1 nCJkOrmo
210bp INTRON 3
3'
2062 1 OCUiOOTO. . . .
1
2191 1 TlTWlmi
B *«tl
60 I I I I I cc CTGCAATA0CACTGGGA ACCCCCCMCTCCCCCGASaa 61 120 I I CTCCACOGACCTCTCTTC 121 180 I I CTATCO1111 TlCTAICIC! I ICCCCTgiTCaLCCCCaCCXTTCTCCACTttTITrTT'lCC 161 240 I I 241 I
300 I
301
360 I
I 361 I
420 I
421 I
480 I
491 I
lindltt
CGCGGTTCCTC111 i GCTCATAAGAgyagTTCCCTGG(aCTCTG*ACSGCAACCTT. . .
Figure 2. Intron/exon boundaries and upstream region of the APE gene. A. Splice site junctions. Intron sequences are shown in bold type, exons in small capital letters, and direct repeat sequences underlined. B. Upstream sequence. The DNA sequence is shown from the Pstl site (upstream of the APE) to the HindlU site (located in exon 1). The Pstl, Hindm and CCAAT motif (bp 340-344) sites are shown in italics.
Downloaded from http://hmg.oxfordjournals.org/ at McMaster University Library on June 29, 2015
1 kb
Human Molecular Genetics, Vol. 1, No. 9 679
DISCUSSION The structural and sequence analysis of APE presented here reveals an unusually compact human gene, due primarily to its small introns. The ~ 30-kDa C-terminal segment of the predicted Ape protein, which belongs to a family of AP endonucleases that includes E. coli exonuclease in (13), is encoded over the last three exons of APE, while the first exon is noncoding. The ~0.5 kb sequence upstream of APE did not reveal obvious regulatory sites, except for a CCAAT box. APE might not be regulated as part of a known control system, and the high frequency of CpG dinucleotides upstream of APE and the lack of a TATA box is a common feature of 'housekeeping' genes (23). The AP endonuclease is expressed at a relatively high level in HeLa cells and in T-lymphoblasts ( - 7 X 1 0 6 molecules per cell; ref. 10). Moreover, the level of the /tP£-encoded enzyme was not significantly increased in HeLa cells (or in CHO cells) by the oxidative stress exerted by high levels of oxygen (10). The APE AP endonuclease may perform functions that are important in a variety of circumstances. The likely role of the AP endonuclease in correcting mutagenic DNA damages from both environmental and endogenous sources, as demonstrated for the major enzyme in yeast (3), would correspond to such a function. The Ape protein has also recendy been shown to restore DNA binding activity to oxidized Jun/Fos heterodimers in vitro (21), although the in vivo significance of this proposed redox reaction has not been established. Such dual roles might help account for the abundance of Ape in human cells. The site of APE in the human genome, the ql 1.2 —12 band of chromosome 14*, is not the locus of any known human genetic disease thought to involve DNA repair (Human Genome Database, Johns Hopkins University, Baltimore, Md.; most recent search 7 August 1992). Purine nucleoside phosphorylase deficiency maps to 14ql2, but this defect seems unrelated to the AP endonuclease, which lacks phosphate-dependent activities and acts only on DNA, not nucleosides (10, 24). However, 14ql 1.2 — 12 is frequently involved in cytologically-detectable
B
Figure 3. Chromosome localization of APE genomic clones. A. Metaphase spread after hybridization with biotinylated pLHBS2 and DAPI staining. The FTTC hybridization signal has been pseudo-colored red (white arrows) and the DAPI staining pattern green. B. A G-banded idiognun of human chromosome 14 displaying the range of hybridization signal locations observed on 15 metaphase spreads probed with pLHBS2. C. A composite of chromosomes 14 displaying signals produced by pLHBS2 probes. Pseudo-coloring as in A.
chromosome rearrangements (translocations, inversions) in patients with the cancer-prone genetic disease ataxia telangiectasia (AT). The mutations responsible for the three most common complementation groups of AT map to the long arm of chromosome 11 (25). Some of the rearrangements found in AT lymphocytes involving 14ql 1.2 include T-cell receptor genes (26, 27). These rearrangements correlate with the predisposition toward T cell leukemia associated with AT (25). Other AT-associated cancers (25, 28), or AT-independent neoplasias might be connected to translocations in this region which inactivate or disregulate the APE gene, and thus elevate die spontaneous mutation rate (29). This possibility can now be addressed experimentally by examining tumor cells with other known rearrangements involving 14ql 1.2 —12 for effects on the structure or expression of APE.
* As this manuscript was being prepared, we were made aware of independent experiments that confirm the DNA sequence and the chromosome mapping presented here, except that APE was assigned solely to 14ql2; that work subsequently appeared in Zhao, B., Grandy, D.K., Hagerup, J.M., Magenis, R.E., Smith, L., Chauhan, B.C., Henner, W.D. (1992) Nud. Acids Res. 20, 4097-4098.
Downloaded from http://hmg.oxfordjournals.org/ at McMaster University Library on June 29, 2015
detected. Since only single-copy hybridization was observed in Southern blots probed with the APE cDNA (ref. 21, and our unpublished data), it seemed likely that the secondary hybridization to chromosome 4 was due to one of the APE introns or to a DNA segment adjacent to APE on the genomic probe fragments. Indeed, when the subclone pLHBS2 was used as an in situ probe, hybridization was observed only to chromosome 14 (Fig. 3A). Plasmid pLHBS2 contains most of the APE gene: the 8.1-kb Sacl-Xhol fragment that includes the 3' end of exon 2, and all of exons 3 - 5 and introns 2 - 4 (Fig. 1). Thus, the faint hybridization to chromosome 4ql3 was evidently due to DNA upstream of APE, to intron 1, or to exons 1 or 2. The possibility has not been eliminated that an APE pseudogene or a gene related to APE resides on chromosome 4. The fine mapping to bands ql 1.2 —12 (Fig. 3B) was obtained by comparing signals in a composite made from from several metaphases (Fig. 3C). The horizontal bar indicated in the ideogram in Fig. 3B indicates the range of signals obtained with fifteen different metaphase spreads.
680 Human Molecular Genetics, Vol. 1, No. 9
REFERENCES 1. Ames, B.N. (1983) Science 221,1256-1264. 2. Sies, H. (1991) American J. Med. 91, 31S-38S. 3. Ramotar, D., Popoff, S.C., Gralla, E.B. and Demple, B. (1991) Mol. Cell. Biol. 11, 4537-4544. 4. Loeb, L.A., and Preston, B.D. (1986) Annu. Rev. Gtnei. 20, 201-230. 5. Wallace, S.S. (1988) Envir. Molec. Mulagen. 12, 431-77. 6. Sancar, A., and Sancar, G.B. (1988) Annu. Rev. Biochem. 57, 29-67. 7. Demple, B., and Levin, J.D. (1991) in Sies, H. (ed.), Oxidative Stress: Oxidants and Antioxidants. Academic Press, London, pp. 119-154. 8. Levin, J.D., and Demple, B. (1990) Nudeic Acids Res. 18, 5069-5075. 9. Demple, B., Greenberg, J.T., Johnson, A.W., and Levin, J.D. (1988) In Friedberg, E.C., and Hanawalt, P.C. (ed.), Mechanisms and Consequences of DNA Damage Processing, UCLA symposia on Molecular and Cellular Biology, New Series. Alan R. Liss, New York, Vol. 83, pp. 159-166. 10. Chen, D.S., Herman, T., and Demple, B. (1991) Nucleic Acids Res. 19, 5907-5914. 11. Henner, W.D., Rodriguez, L.O., Hecht, S.M., Haseltine, W.A. (1983)7. Biol. Chem. 258, 711-713. 12. Demple, B., Johnson, A.W., and Fung, D. (1986). Proc. Nail. Acad. Sci. (USA) 83, 7731-7735. 13. Demple, B., Herman, T., Chen, D.S. (1991) Proc. Nail. Acad. Sci. (USA) 88, 11450-11454. 14. Robson, C.N., and Hickson, I.D. (1991) Nucleic Acids Res. 19,5519-5523. 15. Seki.S., Hatsushika, M., Watanabe, S., Akiyama, K., Nagao, K., Tsutsui, K. (1992) Biochim. Biophys. Ada, in press. 16. Cunningham, R.P., Saporito, S.M., Spitzer, S.G., and Weiss, B. (1986) J. Bacteriol. 168, 1120-1127. 17. Sambrook, J., Fritsch, E.F., Maniatis, T. (1989) Molecular cloning: A laboratory manual. (Second edition.) Cold Spring Harbor Laboratory Press. 18. Chang, A.C.Y., and Cohen, S.N. (1978) J. Bacteriol. 134, 1141 -1156. 19. Brunak, S., Engelbrecht, J., and Knudsen, S. (1991) /. Mol. Biol. 220, 49-65. 20. Cheng, X.B., Bunville, J., Patterson, T.A. (1992) Nucleic Acids Res. 20, 370. 21. Xanthoudakis, S., Miao, G., Wang, F., Pan, Y.-C. E., Curran, T (1992) EMBOJ. 11, 3323-3335. 22. LJchter, P. Tang,C.-J.C, CaU,K., Hermanson, G., Evans, G.A., Housman, D., Ward, D.C. (1990) Science 247,64-69. 23. Dynan, W.S. (1986) Trends in Genetics 2, 196-197. 24. Kane, C M . , and Linn, S. (1981) J. Biol. Chem. 256, 3405-3414. 25. Gatti, R.A., Boder, E., Vinters, H.V., Sparkes, R.S., Norman, A., and Lange, K. (1991) Medicine 70, 99-117. 26. Stem, M.-H., Zhang, F., Griscelli, C , Thomas, G., Aurias, A (1988) Hum. Genet. 78, 33-36. 27. Humphreys, M.W., Nevin, N.C., and WooMridge, M.A. (1989) Human Genetics 83, 79-82. 28. Swift, M., Morrell, D., Massey, R.B., and Chase, C.L. (1991) New Eng. J. Med. 325, 1831-1836. 29. Loeb, L.A. (1991) Cancer Res. 51, 3075-3079. 30. Brigatti, DJ., Meyerson, D., Leary, J J., Spalholtz, B., Travis, S.Z., Fong, C.K.Y., Hsiung, C D . , and Ward, D.C. (1983) Virology 126, 32-50. 31. Baldini, A., and Ward, D.C. (1991) Genomics 9, 770-774.
The human placental genomk library in the EMBL3 vector was obtained from Clontech (product no. HL 1067J). The vectors pACYC177 (18) and Bluescript SK (17) were obtained from New England BioLabs and Stratagene, respectively. Recombinant plasmids were propagated in E. coli strain DH5or bacteria cultured at 37 °C in Luria-Bertani (LB) broth or agar supplemented with the appropriate antibiotic (17). Restriction enzymes were purchased from New England BioLabs or BRL, Klertow fragment DNA polymerase from New England BioLabs, and Sequenase™ from US Biochemicals; all the enzymes were used according to the supplier's instructions. Molecular cloning and sequence analysis of a fragment mumming the human APE geae A placental geuomic library in the EMBL3 vector was screened using the 1.3-kb human APE cDNA (13) as a probe. Two positive clones, designated XAPE1 and XAPE2, were identified and purified through three successive screens. XAPE1 and XAPE2 had human genomic inserts of 18 kb and 14 kb, respectively. Digestion with Xhol released 14-kb and 4-kb fragments from XAPE1, and a single 14-kb fragment from XAPE2. The 14-kb fragments from XAPE1 and XAPE2 were subcloned into pACYC177, to produce the plasmids pAPEl and pAPE2, respectively. Restriction mapping indicated that the genomic insert in pAPEl overlaps the 5' - 9 kb of the insert in pAPE2, but includes — 5 kb of additional upstream sequence (data not shown). Fragments of 4.1 kb and 8.1 kb, each containing a portion of the APE gene, were isolated by digestion of pAPE2 with Hindm, or with Xhol and Sad, and ligated into the MncflH-digested or SacVXholdigested Bluescript SK to yield pLHBSl and pLHBS2, respectively. pAPE2, pLHBSl and pLHBS2 were used to sequence the APE gene using the dideoxy chain termination method (17) with Sequenase and the M13 reverse, - 2 0 and - 4 0 primers (US Biochemicals) and additional synthetic oligonudeotides (5'-ACTCCTGGCAATGTCAG-3' and oligonucleotides corresponding positions 138-122, 421-405, 419-436, 511-495, 676-692, 837-819, 967-983, 1090-1108, 1127-1111, 1454-1470, 1647-1631, 1980-1963, 1988-1971, 2303-2287, 2362-2378, 2510-2494, 2609-2625, 2718-2702, 2935-2919 in the sequence deposited in GenBank; Accession number M99703). Chromosome mapping The chromosomal location of APE was determined by fluorescence in situ hybridization (22), initially with probes from XAPE1 and XAPE2, and finally with pLHBS2 (Fig. 3). Briefly, purified DNA (17) was labeled with biotin-11-dUTP (Sigma) by nick translation (30). The size of the labeled probe ranged 200-500 nudeotides, and unincorporated nuckotides were separated from the probe DNA using Sephadex G-50 spin columns. Metaphase chromosome spreads were prepared by standard techniques. Denaturation and prcannealing of 2 - 4 >ig Cot I DNA (BRL), 10 »ig of DNase-treated salmon sperm DNA and 150 ng of biotinylated probe DNA was performed as described (22). Following 16—18 hr incubation at 37°C and post-hybridization washes, the slides were incubated with blocking solution and the biotin-labeled probe was detected using 5 |ig/mlfluoresceinisothiocyanate (FITO-conjugated avidin (Vector Laboratories). The banding pattern used for chromosome identification and visual localization of the probes was produced by diamidinopnenylindole (DAPI) staining (0.2 /ig/ml, 10 min), which produces a G-like banding pattern. Fluorescence signals were imaged using a Zeiss Axioscope 20 epifluorescence microscope equipped with a cooled CCD camera (Photometries, CH220), and image acquisition was performed using an Apple Macintosh II computer. The gray scale images were pseudo-colored and merged electronically (31). Image processing and merging was performed on a Macintosh FJci computer using Gene Join Max Pix (software by Tun Rand, laboratory of D.C.W., Yak). Photographs were taken directly from the computer monitor.
ACKNOWLEDGEMENTS We thank Dr. Davis S.Chen for advice, Dr. Tom Graf (Molecular Biology Computer Research Resource, Harvard) for his assistance in searching the Human Genome Database, and Drs. W.D.Hermer and T.Curran for communicating results before publication. This work was supported by grants from the NTH (GM40000 to B.D. and HG00272 to D.C.W.) and from the Markey Foundation (to the Department of Molecular and Cellular Toxicology, Harvard School of Public Health).
ABBREVIATIONS AP, abasic or arxirink/apyrimkunic; kDa, kilodalton; AT, ataxia telangicctasia; LB, Luria-Bertani broth; FITC, ftuorescein isothiocyanate, DAPI, diamidinopnenylindole.
Downloaded from http://hmg.oxfordjournals.org/ at McMaster University Library on June 29, 2015
MATERIALS AND METHODS Materials
