APM-1, a novel human gene, identified by aberrant co-transcription with papillomavirus oncogenes in a cervical carcinoma cell line, encodes a BTB/POZ-zinc finger protein with growth inhibitory activity
1998; Springer Nature; Volume: 17; Issue: 1 Linguagem: Inglês
10.1093/emboj/17.1.215
ISSN1460-2075
Autores Tópico(s)TGF-β signaling in diseases
ResumoArticle1 January 1998free access APM-1, a novel human gene, identified by aberrant co-transcription with papillomavirus oncogenes in a cervical carcinoma cell line, encodes a BTB/POZ-zinc finger protein with growth inhibitory activity Stella Reuter Stella Reuter Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Matthias Bartelmann Matthias Bartelmann Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Markus Vogt Markus Vogt Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Caroline Geisen Caroline Geisen Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Inge Napierski Inge Napierski Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Tomas Kahn Tomas Kahn Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Hajo Delius Hajo Delius Deutsches Krebsforschungszentrum, Oligo-Sequenzierungsgruppe, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Peter Lichter Peter Lichter Deutsches Krebsforschungszentrum, Organisation komplexer Genome, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Sandra Weitz Sandra Weitz Deutsches Krebsforschungszentrum, Organisation komplexer Genome, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Bernd Korn Bernd Korn Molekulare Genomanalyse, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Elisabeth Schwarz Corresponding Author Elisabeth Schwarz Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Stella Reuter Stella Reuter Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Matthias Bartelmann Matthias Bartelmann Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Markus Vogt Markus Vogt Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Caroline Geisen Caroline Geisen Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Inge Napierski Inge Napierski Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Tomas Kahn Tomas Kahn Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Hajo Delius Hajo Delius Deutsches Krebsforschungszentrum, Oligo-Sequenzierungsgruppe, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Peter Lichter Peter Lichter Deutsches Krebsforschungszentrum, Organisation komplexer Genome, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Sandra Weitz Sandra Weitz Deutsches Krebsforschungszentrum, Organisation komplexer Genome, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Bernd Korn Bernd Korn Molekulare Genomanalyse, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Elisabeth Schwarz Corresponding Author Elisabeth Schwarz Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany Search for more papers by this author Author Information Stella Reuter1, Matthias Bartelmann1, Markus Vogt1, Caroline Geisen1, Inge Napierski1, Tomas Kahn1, Hajo Delius2, Peter Lichter3, Sandra Weitz3, Bernd Korn4 and Elisabeth Schwarz 1 1Deutsches Krebsforschungszentrum, Angewandte Tumorvirologie, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany 2Deutsches Krebsforschungszentrum, Oligo-Sequenzierungsgruppe, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany 3Deutsches Krebsforschungszentrum, Organisation komplexer Genome, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany 4Molekulare Genomanalyse, Im Neuenheimer Feld 242, D-69120 Heidelberg, Germany ‡S.Reuter and M.Bartelmann contributed equally to this work *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:215-222https://doi.org/10.1093/emboj/17.1.215 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Integration of human papillomavirus (HPV) DNA into the host cell genome is an important step in cervical carcinogenesis. In tumour cells with integrated HPV DNA, transcription of viral oncogenes E6 and E7 continues into the flanking cellular sequences thereby producing viral–cellular fusion transcripts. Analysis of cellular sequences flanking the integrated HPV68 DNA in the cervical carcinoma cell line ME180 revealed homozygosity of the mutant allele in ME180 cells. We speculated that this could indicate the existence of a cellular tumour suppressor gene in the integration region. We report here the identification of a novel human gene, named APM-1, which is co-transcribed with the HPV68 E6 and E7 genes and is present in the 3′-cellular part of the ME180 viral–cellular fusion transcripts. The APM-1 gene encodes a protein with a BTB/POZ domain and four zinc fingers, and is located at chromosome 18q21. APM-1 transcripts are detected in normal cervical keratinocytes, but not in the majority of cervical carcinoma cell lines analysed. The APM-1 gene caused a reduction of clonal cell growth in vitro of HeLa and CaSki tumour cells. These characteristics make APM-1, the first novel human gene identified in a HPV integration region, a likely candidate for the postulated tumour suppressor gene. Introduction Cancer development is a multifactorial and multistep process in which sequential genetic changes accumulate causing the activation of cellular proto-oncogenes and the loss or inactivation of tumour suppressor genes (Vogelstein and Kinzler, 1993; Lanfrancone et al., 1994). In addition, viruses play an essential role as initiators and drivers of the tumorigenic process in some types of human tumours (zur Hausen, 1996). Human papillomaviruses (HPV) are causally involved in the genesis of cervical carcinomas (zur Hausen, 1994). Types HPV16 and HPV18 are the prototypes of the cancer-associated papillomaviruses, but other related HPV types have also been detected in malignant tumours and cell lines. The oncogenic potential of HPV16 and HPV18 has been attributed primarily to the two early genes, E6 and E7. Their products are multifunctional immortalizing and growth-promoting proteins. Major pathways by which they interfere with normal cell regulation are the inactivation of two key tumour suppressors, the p53 protein (by E6) and the retinoblastoma gene product pRB (by E7) (Hoppe-Seyler and Scheffner, 1997; Tommasino and Jansen-Duerr, 1997). In addition to HPV infection and viral oncogene expression, mutations of host cell genes have to accumulate for a fully malignant phenotype of cervical epithelial cells. The detection of chromosome losses and aberrations and of losses of allele heterozygosity reflect the requirement for multiple genetic changes in cervical carcinogenesis (Mullokandov et al., 1996). However, the human genes involved have still to be identified. The HPV DNA genome persists as an extrachromosomal episome in the nucleus of the infected non-malignant cells. In cancer cells and cell lines, however, the viral DNA is found integrated into the host cell genome. Integration can occur into different chromosomal locations with a preference for fragile sites and near oncogenes such as the myc genes (Cannizzaro et al., 1988; Couturier et al., 1991). Integration of HPV DNA is considered to be an important step in cervical carcinogenesis. Recent studies supporting this view have shown that integration leads to increased stability of the mRNAs encoding the viral E6 and E7 oncogenes (Jeon and Lambert, 1995), to increased levels of viral oncogene expression and selective growth advantage of the cells (Jeon et al., 1995). Conversely, expression of the E6 and E7 oncogenes has been shown to result in an increased frequency of DNA integration (Kessis et al., 1996). In the integrated HPV genomes, the early region which harbours the E6 and E7 oncogenes is usually truncated and lacks the 3′-terminal part with the poly(A) signal (Schwarz et al., 1985). Therefore, transcription initiated at the viral early promoter continues into the flanking cellular sequences, and RNA processing results in multiply spliced fusion transcripts composed of 5′-terminal viral sequences with the E6 and E7 oncogenes and 3′-terminal co-transcribed cellular sequences (Schneider-Gädicke and Schwarz, 1986). Besides affecting the expression of the viral oncogenes, the integration event may also cause alterations of cellular genes within the target region. For example, integration of HPV DNA near myc genes has been reported to enhance the proto-oncogene expression (Couturier et al., 1991). Since the integrated HPV DNA is a tag for the chromosomal target region, it may also provide a useful tool for the identification of novel cancer-relevant cellular genes (Kahn et al., 1996). In the present study, we have analysed in the human cervical carcinoma cell line ME180 the cellular target region of the integrated HPV DNA (Reuter et al., 1991) which has meanwhile been classified as an HPV68 subtype (Longuet et al., 1996). We report here the identification of a novel human gene, APM-1, which is present in the viral–cellular fusion transcripts due to co-transcription with the HPV68 E6 and E7 genes. The APM-1 gene encodes a protein with zinc fingers and a BTB/POZ domain that exerts the activities of a cell growth inhibitor. Results Homozygosity of the mutant allele carrying the integrated HPV68 DNA in the cell line ME180 From a genomic DNA library of the cell line ME180, the integrated HPV68 DNA had previously been isolated together with flanking cellular sequences at both sides (Reuter et al., 1991). To examine the cellular target region of HPV68 DNA integration, the cloned flanking cellular sequences were used as radiolabelled probes in genomic Southern blot hybridizations (Figure 1). Remarkably, the hybridization patterns of ME180 DNA revealed homozygosity of the mutant allele carrying the integrated HPV68 DNA, a result suggesting that loss of heterozygosity has occurred after the HPV DNA integration event. Since allele loss is characteristic for loci harbouring tumour suppressor genes (Knudson, 1993), we speculated that the cellular target region of HPV68 DNA integration might harbour a tumour suppressor gene. Figure 1.Southern blot analysis of different cervical carcinoma cell lines with hybridization probes representing the left and right flanking cellular sequences of the integrated HPV68 DNA in ME180 cells. (A) The genomic DNA was cleaved with SacI, and the filters were hybridized either with probe 2.5, a 2.5 kb SacI–BamHI fragment containing the left flanking cellular sequences, or with probe 3.5, a 3.5 kb BamHI–SacI fragment containing the right flanking cellular sequences. (B) The genomic DNA was cleaved with HindIII, and the filter was hybridized with probe 2.5. Lane M in (A) and the bars in (B) indicate the positions of the λ HindIII size marker fragments. The names of the cell lines are given above the lanes. DNA isolated from a biopsy of normal human placenta was also included. Download figure Download PowerPoint Transcribed human sequences in the target region of HPV68 integration To search for the suspected gene, we first examined by Northern blot analysis whether the cloned cellular sequences flanking the integrated HPV68 DNA hybridize with human mRNAs. Since this was not the case (data not shown), we isolated the genomic region encompassing the ME180-specific site of HPV68 DNA integration from a cosmid library of normal human leukocyte DNA. One cosmid clone, designated INT68 (because it contains the chromosomal target site of HPV68 DNA integration), was used for the further analyses. To examine whether the cosmid insert harbours transcribed sequences, Northern hybridizations with the radiolabelled INT68 DNA probe were performed (Figure 2). In ME180 poly(A)+-RNA, the cosmid probe produced very strong hybridization signals which were indistinguishable from the signals produced with an HPV68 E6/E7 probe (not shown). In the poly(A)+-RNAs of the other cell lines tested, two transcripts of ∼4 kb and ∼3 kb length were detected. These results indicated first, that the cosmid INT68 includes sequences that are co-transcribed with the HPV68 genes E6 and E7 to produce the viral–cellular fusion transcripts in ME180 cells, and second, that the cosmid INT68 harbours sequences that are transcribed in other cell lines in the absence of integrated HPV68 DNA and thus should be integral parts of authentic cellular mRNAs. Figure 2.Identification of transcripts originating from the cosmid INT68 region which harbours the integration site of HPV68 DNA in the cell line ME180. A filter with poly(A)+-RNA from different cervical carcinoma cell lines was hybridized with the complete cosmid DNA as radiolabelled probe. Hybridization of repetitive sequences was blocked with Cot-1 DNA. The filter was rehybridized with a human β-actin probe. The horizontal bars indicate the positions of the 28S and 18S rRNAs. Download figure Download PowerPoint To localize the transcribed sequences within the cosmid insert, the INT68 DNA was dissected into SacI fragments which were used as individual probes in Northern hybridizations. The viral–cellular fusion transcripts of ME180 cells hybridized with a SacI fragment of 5.5 kb (not shown) which was subsequently cloned and partially sequenced (see Figure 4B for the SacI map of cosmid INT68). Comparison of the partial sequences with the database-stored sequences revealed that the fragment contains sequence motifs with the potential to encode Cys2-His2 zinc fingers of the Krueppel type. The INT68 zinc finger sequences were not identical with any of the already known human zinc finger-encoding genes. Furthermore, a 0.5 kb subfragment containing the zinc finger motifs also hybridized with the fusion transcripts of ME180 cells (not shown). These results indicated that a novel human gene encoding a zinc finger protein is located in the INT68 region and is co-transcribed with the HPV68 E6 and E7 genes in ME180 cells. We named the human gene APM-1 to indicate that its transcription is affected by papillomavirus DNA integration in ME180 cells. Fusion transcript structure and exon–intron composition of the human APM-1 gene To characterize the ME180-specific fusion transcripts, more than 20 positive clones were isolated from a ME180 cDNA library and were structurally characterized by PCR and hybridization analysis. The results indicated that the co-transcribed cellular sequences of all cDNA clones are identical, and that the transition from the 5′-viral into the 3′-co-transcribed cellular sequences occurs in all cDNA clones at a splice donor located at the beginning of HPV68 ORF E1. One representative cDNA clone, Mefust-4, was selected for complete sequence analysis. According to the nucleotide sequence, the fusion transcript contains a 5′-terminal HPV68 segment of 839 bp with the early genes E6 and E7 followed by a 3′-terminal cellular segment of 3779 bp. The cellular part includes an open reading frame with the coding potential for a protein of 619 amino acid residues (Figure 3A). The deduced protein contains four Krueppel-type Cys2-His2 zinc fingers, of which two-and-a-half are encoded in the 0.5 kb subfragment of the 5.5 kb SacI fragment of cosmid INT68. These results show that the ORF in the co-transcribed cellular part of the fusion transcripts harbours the APM-1 gene. Figure 3.Amino acid sequence of the APM-1 protein. (A) Complete amino acid sequence of APM-1 as deduced from the nucleotide sequence of the viral–cellular fusion transcript cDNA of cell line ME180. The complete fusion transcript cDNA sequence is available under DDBJ/EMBL/GenBank accession number Y14591. The N-terminal BTB/POZ domain is indicated by underlining, the acidic region by a broken line, and the four C2H2 zinc fingers by heavy underlining. The vertical arrow marks the cellular exon–exon junction. (B) Comparison of the N-terminal 120 amino acid residues of APM-1 with the N-terminal BTB/POZ domains of other zinc finger proteins of human origin (indicated by h after the gene name). White letters in black columns denote the positions of amino acid residues which are identical in all seven proteins. Grey shading marks positions with identical amino acids in at least four of the seven proteins compared. For the other human proteins, the sequence accession numbers and positions shown are: Q05516, pos. 1–115 (PLZF), X 82018, pos. 1–110 (ZID), P41182, pos. 1–118 (BCL-6), L16896, pos. 1–108 (ZFPJS), P24278, pos. 1–108 (KUP), and L41919, pos. 1–123 (HIC-1). Download figure Download PowerPoint Figure 4.Structural organization of the viral–cellular fusion transcripts of ME180 cells and of the human genomic region harbouring the APM-1 gene and the HPV68 integration site. (A) The ORFs present in the 4618 bp sequence of the fusion transcript cDNA (DDBJ/EMBL/GenBank accession number Y14591) are shown by boxes (5′-end: first ATG codon, 3′-end: termination codon). Viral sequences are indicated by diagonal hatching. The cDNA is composed of three exons. Exon 1 (pos. 1–839) is of viral origin, exon 2 (pos. 840–2063) and exon 3 (pos. 2064–4618) are of human origin. The HPV68 E6 gene (pos. 1–477) encodes a protein of 158 amino acids, and the HPV68 E7 gene (pos. 485–817) a protein of 110 amino acids. The two viral ORFs are followed by a small fusion ORF E1-C (pos. 824–1060) potentially encoding a protein composed of six amino acid residues derived from the HPV68 E1 gene, and 72 amino acid residues of cellular origin. The APM-1 coding region (pos. 856–2715) extends over two exons. The stop codon which defines the APM-1 ORF at the 5′-side is located in the viral exon 5 bp upstream of the viral–cellular exon junction. The first ATG codon providing the putative translational start site (see Figure 3) is located in the cellular part 17 bp downstream of the exon junction. The positions of the encoded BTB and zinc finger domains are indicated. The ATG codon of the HPV68 E6 gene is set as starting point for numbering. The cDNA clone Mefust-4 used for sequence analysis does not include the nucleotides from positions 1–12 which have been taken from the HPV68 DNA sequence (Reuter et al., 1991; accession number M73258). The EST sequences with DDBJ/EMBL/GenBank accession numbers H44758 and H44759 of the Homo sapiens cDNA clone 187988 are nearly identical to positions 1366–1688 and 1887–2215 of the fusion transcript cDNA sequence indicating that cDNA clone 187988 is derived from an APM-1 transcript. (B) SacI restriction map of the cosmid INT68. The numbers indicate the fragment sizes in kb. The integration site of HPV68 DNA in ME180 cells is located in the SacI fragment 6.3. The positions of the fusion transcript exons within the INT68 region (exons 2 and 3 containing the APM-1 coding region) and within the integrated HPV68 DNA (exon 1) are given on top. The introns are shown by the zig-zag lines. Download figure Download PowerPoint Computer-based alignment analysis revealed that the N-terminus of the deduced APM-1 protein shows strong similarity to a conserved motif called the BTB or POZ domain (Figure 3B). The BTB/POZ domain is a newly characterized motif of ∼120 amino acid residues which is evolutionary conserved, primarily found at the N-terminus of zinc finger proteins, and involved in protein–protein interactions (Bardwell and Treisman, 1994; Albagli et al., 1995; Chen et al., 1995). The ATG triplet located 17 bp downstream of the viral–cellular exon junction can be assigned as the potential translational start codon of the APM-1 gene based on the sequence alignment and because it is in a good sequence context for translational initiation according to the Kozak rules (Kozak, 1996). The BTB/POZ domain is followed by a highly acidic region (Figure 3A). Comparison of the cDNA sequence with the partial and complete nucleotide sequences of the genomic SacI fragments 5.5 kb and 4.5 kb (4.5a) and of the adjacent terminus of the 3.3 kb fragment allowed identification of the exon junctions in the fusion transcript cDNA and the exon locations within cosmid INT68. The results are summarized in Figure 4. The cDNA clone Mefust-4 includes a 3′-terminal A20 stretch. Fifteen of these A–residues are found in the respective genomic sequence, followed by a potential polyadenylation signal AATAAA and a potential splice donor sequence. These data indicate that the fusion transcripts may either be terminated at this polyadenylation signal or, alternatively, may contain an additional 3′-untranslated exon. APM-1 transcript patterns To analyse expression of the APM-1 gene, Northern analyses with poly(A)+-RNA from cervical carcinoma cell lines and cultured normal cervical epithelial cells were performed. Figure 5 gives a synopsis of the results. Using various subfragments of the INT68 genomic segment and of the Mefust-4 cDNA as radiolabelled probes (depicted in Figure 5, panels k and l), altogether three cellular mRNA species with sizes of ∼5 kb, 4 kb and 3 kb, respectively, were detected in the cell lines that do not harbour the integrated HPV68 DNA. The different hybridization profiles indicate that the 5 kb transcript is the most relevant of these three mRNA species with regard to the APM-1 gene. Poly(A)+-RNA of the lung carcinoma cell line LX-1 was included in the Northern experiments as a positive control, since the 5 kb mRNA was discernible in this cell line with a strong hybridization signal. One important feature of the 5 kb mRNA is that it hybridized with all the different cDNA probes derived from the APM-1 coding region and the 3′-untranslated part (Figure 5, panels d–i). These hybridization properties indicate that the exons 2 and 3 of the fusion transcript are present in identical form in the cellular 5 kb mRNA. We concluded that the 5 kb transcript is the authentic cellular mRNA for the APM-1 gene. Another feature of the 5 kb transcript is the restricted presence in the cervical carcinoma cell lines. The 5 kb mRNA was detectable in the cultured normal cervical keratinocytes (Figure 5, panel i), but in only two out of the 11 cell lines tested (C4-II and MRI-H196; Figure 5, panels d, h and i). The levels of the 5 kb mRNA in the normal cervical cells and the two carcinoma cell lines are much lower than in the LX-1 cell line. Figure 5.Synopsis of Northern blot hybridizations. Poly(A)+-RNA was isolated from different cervical carcinoma cell lines and from cultured normal human ectocervical keratinocytes (ect.k.). The RNA filters were hybridized with probes derived either from cosmid INT68 (panels a–c) or from the fusion transcript cDNA (panels d–i). The results of rehybridization of the filters with a GAPDH probe are shown at the bottom part of each panel. The horizontal bars indicate the positions of the 28S and 18S rRNA (top part) and of the 18S rRNA (bottom part). The small letters a to i also indicate the locations of the fragments used as hybridization probes within the SacI map of cosmid INT68 (panel k) and within the cDNA (panel l). For a description of the genes and exons legend to Figure 4. Download figure Download PowerPoint The two other mRNA species, with sizes of ∼3 kb and 4 kb, differ in two important aspects from the 5 kb APM-1 mRNA. First, they seem to be ubiquitously present, in particular the 4 kb species. Second, they did not hybridize with the second APM-1 coding exon (probes c and g) or the BTB/POZ domain segment (probe e), but only with a probe containing the 3′-part of the first APM-1 coding exon (probe f). These hybridization properties suggest that the 3 kb and 4 kb transcripts do not harbour the complete coding region of the APM-1 gene and the 3′-untranslated part. It should be mentioned here that the cellular mRNAs detected with the INT68 cosmid probe in the initial Northern analysis (see Figure 2) are the 4 kb and 3 kb transcripts. For additional analysis of the cellular APM-1 mRNAs, we used cosmid INT68 to employ a strategy for cosmid-specific selection of cDNA clones (Korn et al., 1992, 1997). Sixteen cDNA clones with insert sizes ranging between 150 bp and 924 bp were isolated and sequenced (data not shown). Six cDNAs were derived from human fetal liver, 10 from fetal brain, but none from adult skeletal muscle. In most cases, the nucleotide sequences of the cDNA inserts were identical to the respective parts of the APM-1 sequences present in the ME180 fusion transcripts. Interestingly, in one liver cDNA clone the first APM-1 coding exon (i.e. exon 2 of the fusion transcript) is preceded by a 5′-terminal sequence of 127 bp which probably constitutes part of a 5′-non-coding exon of the cellular APM-1 mRNA (not shown). The 127 bp segment did not hybridize with the INT68 cosmid DNA (not shown). Chromosomal localization of the HPV68 integration site and of the APM-1 gene To determine the chromosomal localization of the HPV68 integration site and of the APM-1 gene, fluorescence in situ hybridizations to metaphase chromosomes of normal human cells and ME180 cells were performed. Hybridization of the biotin-labelled INT68 cosmid probe to normal human metaphase chromosomes and banding of the chromosomes by DAPI revealed the specific localization of the target sequences on the long arm of chromosome 18 within band 18q21 (Figure 6). Figure 6.Chromosomal localization of the HPV68 integration site and the human APM-1 gene by fluorescence in situ hybridization. The left panel shows a complete metaphase cell of the cell line ME180 after in situ hybridization with the INT68 cosmid probe (green) and the HPV68 probe (red). The right panels show chromosome 18 homologues from normal human metaphase cells after hybridization with INT68 (green) and with the probe L1.84 specific for chromosome 18 centromeres (red). All panels show digitized images resulting from electronic overlay of two or three fluorescence images obtained from the hybridization signal and the DAPI stain, respectively. Download figure Download PowerPoint The copy number of the viral integration locus in ME180 cells was assessed by dual colour fluorescence in situ hybridization with the biotin-labelled INT68 cosmid and a digoxigenin-labelled HPV68 restriction fragment. In the analysed metaphase cells, two copies of chromosome 18 were found which contain the human genomic DNA of the integration site and the viral sequences. The highly specific co-localization of both sequences is illustrated in Figure 6. No INT68 sequences without co-localization with HPV68 DNA were found. These data demonstrate that ME180 cells contain two copies of the mutant chromosome 18 harbouring the integrated HPV68 DNA, and lack copies of the normal (unoccupied) chromosome 18. Effect of APM-1 on cell growth The present study was initiated by the speculation that a cellular tumour suppressor gene might be located in the genomic target region of HPV68 DNA integration in the cell line ME180. We tested this assumption for the identified APM-1 gene using a colony formation assay. After inserting the APM-1 gene and the complete fusion transcript cDNA, Mefust, into the eukaryotic expression vector pSG5, each of the expression constructs was co-transfected with the G418 selection plasmid pSV2-neo into cells of the HeLa and CaSki cervical carcinoma cell lines, both of which do not contain detectable amounts of the endogenous APM-1 transcript (see Figure 5). The pSG5 vector alone and the cDNA antisense construct were used as controls. The results of the colony formation assays are shown in Figure 7. The clonal growth of both HeLa and CaSki cells was strongly reduced by the transfected APM-1 expression plasmid. Compared with the pSG5 control, the Mefust fusion transcript cDNA also exerted a growth inhibitory effect which, however, was weaker than that of the APM-1 construct and was comparable with the effect of the antisense construct. The results indicate that the APM-1 gene encodes a protein with growth inhibitory activity for cervical carcinoma cell lines. Figure 7.Colony formation assay to examine the influence of the APM-1 gene on tumour cell growth. HeLa or CaSki cells were transfected with either pSG5 vector alone (1), the expression plasmid pSG5–APM-1 using two different DNA preparations (2, 3), the antisense construct pSG5–APM-1-as (4), or the expression plasmid pSG5–Mefust (5). Two additional independent experiments were performed with HeLa cells and one additional experiment with CaSki cells. All experiments yielded comp
Referência(s)