- Email: [email protected]
DPC4 splice variants in neuroblastoma
Cancer Letters 122 (1998) 187–193
DPC4 splice variants in neuroblastoma Hajime Kageyama a, Naohiko Seki b, Shigeru Yamada a, Shigeru Sakiyama a, Akira Nakagawara a,* a
Division of Biochemistry, Chiba Cancer Center Research Institute, 666-2 Chuoh-ku, Chiba 260, Japan b Laboratory of Gene Structure I, Kazusa DNA Research Institute, Kisarazu, Chiba 292, Japan Received 22 July 1997; received in revised form 20 August 1997; accepted 21 August 1997
Abstract One of the loci for neuroblastoma suppressor genes is chromosome 18q21 where the DPC4 tumor suppressor gene, as well as the DCC and MADR2 genes, is located. DPC4 is a molecule of the TGF-b signal which regulates differentiation of the neural crest precursor cells from which neuroblastoma originates. During the search for the significance of DPC4 as a candidate neuroblastoma suppressor gene, we found that there are at least two variant forms of the DPC4 transcripts by using the reverse-transcriptase-PCR procedure. The subsequent sequencing analysis has revealed that one is missing exons 5 and 6 and the other is missing exons 4–6. Both splice variants were frequently observed in neuroblastomas and at low levels in normal tissues. Though the functional role of the DPC4 splice variants is unknown, they might be important in regulating the TGF-b signaling not only in neuroblastomas but also in other tumors and normal tissues. 1998 Elsevier Science Ireland Ltd. Keywords: DPC4; Alternative splicing; Neuroblastoma
1. Introduction Recent advances in cancer genetics have revealed that inactivation of tumor suppressor genes plays a major role in tumor development or progression. DPC4, originally identified as a candidate tumor suppressor gene of pancreatic cancers, frequently exhibits homozygous deletion, as well as nonsense or missense mutations in some other cancers [1,2]. The DPC4 gene is mapped at chromosome 18q21.1 and encodes a protein homologous to a product of a Drosophila
* Corresponding author. Tel.: +81 43 2645431; fax: +81 43 2654459; e-mail: [email protected]
melanogaster gene (Mad) which is implicated in a transforming growth factor-b (TGF-b)-like signaling pathway. Chromosome 18q21 is also a locus for a possible tumor suppressor gene of neuroblastoma which is one of the most common childhood tumors and is derived from the sympathoadrenal lineage of the neural crest. Takita et al. [3] reported that the DCC locus on 18q21.1 is deleted in neuroblastomas at a relatively high incidence (31%), but the mutation of the DCC gene is not so common in human neuroblastomas (Y. Hayashi, pers. commun.). Since the DPC4 gene is located close to the DCC gene, we decided to examine the role of DPC4 in neuroblastoma. During the study we found that there are at least two alternatively
0304-3835/98/$19.00 1998 Elsevier Science Ireland Ltd. All rights reserved PII S0304-3835 (97 )0 0389-3
188
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193
spliced forms of DPC4 mRNAs in both neuroblastomas and non-malignant tissues.
2. Materials and methods 2.1. Tumor tissues and cell lines Fresh tumor samples were obtained from the patients with neuroblastoma at surgery and were stored at −80°C until use. The tumors were staged according to the system of Evans et al. [4]. The derivations, descriptions and culture conditions of cell lines were described previously [5].
primer for the 3′-untranslated region of DPC4 mRNA (R0: 5′-GCCTGACAAGTTCTGAAGAGG3′). The PCR amplification for generating the DPC4 cDNA fragments was performed by using combinations of the forward primers F1 (5′-TCTGAGTCTAATGCTACCAGCACTGCC-3′), F2 and F3 (5′CCTTGCAACGTTAGCTGTTG-3′) and the reverse primers R1, R2 (5′-GTAGTCCACCATCCTGATAAGG-3′) and R3 (5′-GCAAGCTCATTGTGAACAGG3′). The amplification products were run on 1.5% agarose gel and visualized by ethidium bromide staining. The RT-PCR products were subsequently cloned into the pCRII vector by using a TA cloning kit (Invitrogen, San Diego, CA) and processed for DNA sequence analysis.
2.2. Northern blot analysis 2.4. DNA sequencing Total RNA was extracted from 0.2 to 0.5 g of primary neuroblastoma tissues or cultured cell lines according to the method of Chomczynski and Sacchi [6]. We resolved 15 mg of each RNA on 1% agaroseformaldehyde gels and transferred the RNA by blotting to a nylon membrane (Hybond-N+, Amersham, Arlington Heights, IL). Blots were hybridized, washed and exposed to X-ray film as described previously. The human adult and fetal multiple tissue blots for Northern analysis were obtained from Clontech Laboratories (Palo Alto, CA). The 1092 bp DPC4 probe was made by a PCR amplification procedure using a forward primer F2 (5′-CTGGTCAGCCAGCTACTTACC-3′) and a reverse primer R1 (5′-TCTGTCTGCTAGGAGCAAGG-3′) with cDNAs, which were reversely transcribed from the total RNA obtained from a case 63 neuroblastoma tissue, as templates. 2.3. Reverse-transcriptase (RT)-PCR and Southern blot analysis Total RNA (1 mg) from primary tumors or cell lines, or 0.2 mg of poly A+ RNA obtained from normal tissues (Clontech Laboratories, Palo Alto, CA) was used to generate complementary DNA fragments for DPC4. The reverse-transcriptase reaction was performed in the reaction mixture containing 50 units of superscript II (Life Technologies, Tokyo), 1× RT buffer, 10 mM DTT, 250 mM dNTPs, 10 units of RNAsin (Toyobo, Tokyo) and a specific antisense
DNA sequencing was performed using the dideoxy chain termination method (SequiTherm Long-Read Cycle Sequencing Kit-LC, Epicentre Technologies, Madison, WI) on an automated DNA sequencer Model 4000L (Li-COR, Lincoln, NE). The sequence information was analyzed using Genetyx (Software Development, Tokyo).
3. Results 3.1. Expression of DPC4 mRNA DPC4 mRNA expression was examined by Northern blot analysis in both neuroblastomas and normal tissues, since the expression has not been previously reported in detail. In fetal tissues, both 3 and 7 kb transcripts were strongly expressed in lung and kidney and weakly expressed in liver and brain. In adult normal tissues, the 3 kb transcript was predominantly expressed at high levels in skeletal muscle, at intermediate levels in heart, kidney, thymus, prostate, testis and ovary, at low levels in brain, placenta, lung, liver, pancreas, spleen, small intestine and colon and at undetectable levels in leukocytes. The 7 kb transcript was detectable in the adult skeletal muscle (data not shown). Fig. 1 shows the representatives of Northern blot for DPC4 expression in neuroblastomas. All 29 primary neuroblastomas examined (11 in stage 1, five in stage 2, five in stage 3 and eight in stage 4)
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193
189
Fig. 1. Northern blot analysis for mRNA expression of DPC4 in primary neuroblastomas and neuroblastoma cell lines. Total RNA (15 mg) was loaded on each lane. The RNA blots were hybridized with a 32P-labeled DPC4 probe, followed by hybridization with glyceraldehyde-3phosphate dehydrogenase (GAPDH) as the control for the amount of loaded RNA. N-myc is amplified in SK-N-BE, SK-N-DZ, RTBM1, NLF and LAN5 neuroblastoma cell lines.
expressed the 3 kb transcript at similar levels, while neuroblastoma cell lines expressed it at various levels. 3.2. Alternative splicing of DPC4 Since in some tumor suppressor genes, such as WT1 and BRCA1, RNA was abnormally processed and it resulted in a production of a functionally changed protein [7,8], we next tested by means of an RTPCR procedure whether or not there are different sizes of the DPC4 transcripts in neuroblastomas. Total RNA obtained from two representative tumors with stage 1 (case 63) and stage 4 (case 52) was allowed for the reverse-transcriptase reaction to make complementary DNAs of DPC4 by using the specific primer R0 at the 3′-UTR. The PCR amplification using the F1 primer (exon 4) and either R1 or R2 (both 3′-UTR) primer gave two bands (approximately 1200 and 970 bp, respectively) on an agarose gel (Fig. 2A), whereas the forward primer F2 paired with either R1 or R2 gave a single band which corresponded to the product size expected from the wild-type DPC4. We then performed a PCR amplification using a primer pair, F3 and R3, which allows amplification of the region spanning exons 1–7 (Fig. 2B). It gave a few smaller bands than expected and the sizes of the common bands were about 800 and 580 bp. The Southern
blot analysis showed that at least two smaller sizes of the bands hybridized to the DPC4 probe. The sequencing of both products obtained from the C-52 primary neuroblastoma demonstrated that the clone 52-1 deleted exons 4–6 (codons 152–301) and the clone 52-2 deleted exons 5 and 6 (codons 223–301) (Fig. 3A,B), according to the information about the DPC4 gene from the GenBank database (accession number U44378). Both possible splice variants were also faintly observed in normal tissues of the lung and small intestine. Because the deletion in both transcripts occurred in frame, two possible splice variants may be translated to proteins that lack amino acids corresponding to codons 152–301 and codons 223–301. The putative truncated proteins reserve both MH1 and MH2 domains which are conserved in SMAD family proteins [9], but lack a domain that is unique to DPC4 (Fig. 3B).
4. Discussion The different proteins that derive from the same gene as a result of alternative splicing are functionally related, but in some cases they exhibit distinctly counteracting biological activities. These include tumor
190
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193
Fig. 2. Identification of alternative splicing of DPC4 in neuroblastomas and normal tissues. (A) Ethidium bromide-stained 1% agarose gel of the RT-PCR products generated with DPC4-specific primer pairs; expected product sizes are 1203 bp from the pair F1 (codons 738–764 at exon 4) and R1 (codons 1922–1941 at exon 11), 1096 bp from the pair F1 and R2 (codons 1813–1834 at exon 11), 1092 bp from the pair F2 (codons 849–909 at exon 5) and R1 and 985 bp from the pair F2 and R2, according to the GenBank database (accession number U44378). Total RNA obtained from two primary neuroblastomas (stages 1 and 4) was used for the RT-PCR reactions. Only when the F1 primer was used were two different bands detectable. The size marker used was HindIII-digested l DNA. (B) DPC4 splice variants in primary neuroblastomas, neuroblastoma cell lines and human normal tissues. (Upper) Ethidium bromide-stained 2% agarose gel of the RT-PCR products generated with DPC4-specific primer pair F3 (codons 31–50 at exon 1) and R3 (codons 1035–1054 at exon 7). Arrows indicate the sizes of the products of wild-type (upper), about 800 bp (middle) and about 580 bp (lower). The size marker used was pUC118 digested with both HinfI and NciI. (Lower) The gel in the upper panel was blotted on the Nylon membrane, which was subsequently probed with a 32Plabeled DNA fragment of DPC4.
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193
suppressor genes such as WT1 and BRCA1 as well as genes for bcl-x and hepatocyte growth factor [7,8, 10,11]. We here demonstrate that at least two splice variants of DPC4 mRNA are frequently expressed in human neuroblastomas; one deleted exons 5 and 6 and the other deleted exons 4–6. DPC4 has recently been found to be a molecule in the signaling pathway of the TGF-b superfamily [12,13]. Since both variants are also expressed in some non-malignant tissues, they could function as general modifiers of cell function in both malignant and non-malignant cells responsive to growth factors of the TGF-b superfamily.
191
Neuroblastoma originates from the sympathoadrenal progenitor cells which are derived from neural crest cells [14]. Recently, Shah et al. [15] suggested that the fate of neural crest stem cells during early embryogenesis is strongly influenced by TGF-b and bone morphogenic protein 2 (BMP2), which are growth factors of the TGF-b superfamily. However, the functions of BMP2 and TGF-b are different in that the former induces neurogenesis in neural crest cells, whereas the latter promotes smooth muscle differentiation. The responsiveness to TGF-b is reserved in some neuroblastoma cell lines, but the response pat-
Fig. 3. (A) Sequence of aberrant RT-PCR products of DPC4 from case 52 primary neuroblastoma. The sequences of subcloned RT-PCR products (clones 52-1 and 52-2) showed loss of exons 4–6 and exons 5 and 6, respectively. (B) Diagram of aberrant mRNA transcripts of DPC4. Splice sites are indicated based on the GenBank database (accession number U44378). Arrows indicate the location and direction of primers used for RT-PCR analyses.
192
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193
terns are diverse [16]. In addition, many lines do not respond to TGF-b phenotypically. These findings suggest that neuroblastoma may have some kind of disregulation in the TGF-b signal transduction pathway. The chromosome locus at 18q21 is where DCC, a candidate tumor suppressor gene for colorectal cancer, is localized [17]. The DCC locus is reported to be deleted in about 30% of neuroblastomas by a single study of the loss of heterozygosity [3]. Furthermore, loss of DCC expression is significantly associated with the dissemination of neuroblastoma [18]. DCC could therefore also be a candidate tumor suppressor gene for neuroblastoma, although that needs to be confirmed. The DPC4 gene, which was originally reported as a tumor suppressor gene for pancreatic carcinomas, is localized at chromosome 18q21, quite close to the DCC gene [1,2]. However, recent extensive studies suggest that mutations or deletions in other cancers are uncommon (,10%) [19–22]. Our preliminary study of neuroblastomas showed a missense mutation in only one of 15 neuroblastoma cell lines (unpublished data). This suggests that if DPC4 could function as a tumor suppressor for neuroblastoma, the mechanism might be other than mutations or deletions of the gene. Accumulating evidence demonstrates that intracellular signals from TGF-b or BMP are mediated by the formation of hetero-oligomeric complexes of DPC4 and Smad1 (Madr1) or Smad2 (Madr2) [9,12,13,23]. Loss of the oligomerization may be critical in the biological responses induced by growth factors of the TGF-b superfamily. From that point of view, the presence of alternatively spliced variants of DPC4 in neuroblastomas might be important. Two splice variants we report here lose exons 5 and 6 or exons 4–6, both of which are in frame but outside of the conserved domains of MH1 and MH2. The functional roles of MH1 and MH2 are still unknown, although truncation of C-terminal 39 amino acids is reported to act dominant-negatively on the TGF-b-induced response [13]. Thus, it is possible that the presence of either or both forms of DPC4 splice variants may have an influence on oligomerization between intact DPC4 and Smad proteins, resulting in inhibition of TGF-b signaling. Intriguingly, our present results suggest that the advanced stage neuroblastomas express more DPC4 splice variants than the low stages tumors. However, this needs to be confirmed by a study based
on the quantitative analysis of a large number of samples. In summary, we found at least two alternatively spliced variants of DPC4 mRNAs in human neuroblastomas as well as some normal tissues at low levels. Both sites spliced out are between MH1 and MH2 domains and may affect the oligomerization of DPC4 and Smads. The TGF-b signaling is important in regulating the growth and differentiation of cancer cells as well as normal cells. The frequent expression of DPC4 splice variants may play a role in an aberrant regulation of neuroblastoma growth.
5. Addendum During the preparation of this manuscript, Johan P. de Winter et al. published a DPC4 splice variant which was cloned from a breast tumor cell line, MDA-MB-231 (Oncogene 14 (1997) 1891–1899). The splice variant was the same as one of two variants found in this manuscript. References [1] S.A. Hahn, M. Schutte, A.T. Hoque, C.A. Moskaluk, L.T. da Costa, E. Rozenblum, C.L. Weinstein, A. Fischer, C.J. Yeo, R.H. Hruban, S.E. Kern, DPC4, A candidate tumor suppressor gene at human chromosome 18q21.1, Science 271 (1996) 350–353. [2] S.A. Hahn, A.T. Hoque, C.A. Moskaluk, L.T. da Costa, M. Schutte, E. Rozenblum, A.B. Seymour, C.L. Weinstein, C.J. Yeo, R.H. Hruban, S.E. Kern, Homozygous deletion maps at 18q21.1 in pancreatic cancer, Cancer Res. 56 (1996) 490– 494. [3] J. Takita, Y. Hayashi, T. Kohno, M. Shiseki, N. Yamaguchi, R. Hanada, K. Yamamoto, J. Yokota, Allelotype of neuroblastoma, Oncogene 11 (1995) 1829–1834. [4] A.E. Evans, G.J. D’Angio, J.A. Randolph, A proposed staging for children with neuroblastoma: children’s cancer study group, Cancer 27 (1971) 374–378. [5] A. Nakagawara, C.G. Azar, N.J. Scavarda, G.M. Brodeur, Expression and function of TRK-B and BDNF in human neuroblastomas, Mol. Cell. Biol. 14 (1994) 759–767. [6] P. Chomczynski, N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction, Anal. Biochem. 162 (1987) 156–159. [7] A.L. Menke, N. Riteco, R.C.A. van Ham, C. de Bruyne, F.J. Rauscher III, A.J. van der Eb, A.G. Jochemsen, Wilms’ tumor 1 splice variants have opposite effects on the tumorigenicity of adenovirus-transformed baby-rat kidney cells, Oncogene 12 (1996) 537–546.
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193 [8] M. Lu, S.D. Conzen, C.N. Cole, B.A. Arrick, Characterization of functional messenger RNA splice variants of BRCA1 expressed in nonmalignant and tumor-derived breast cells, Cancer Res. 56 (1996) 4578–4581. [9] K. Eppert, S.W. Scherer, H. Ozcelik, R. Pirone, P. Hoodless, H. Kim, L.-C. Tsui, B. Bapat, S. Gallinger, I.L. Andrulis, G.H. Thomsen, J.L. Wrana, L. Attisano, MADR2 maps to 18q21 and encodes a TRGb-regulated MAD-related protein that is functionally mutated in colorectal carcinoma, Cell 86 (1996) 543–552. [10] L.H. Boise, M. Gonzalez-Garcia, C.E. Postema, L. Ding, T. Lindsten, L.A. Turka, X. Mao, G. Nunez, C.B. Thompson, bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death, Cell 74 (1993) 597–608. [11] V. Cioce, K.G. Csaky, A.M. Chan, D.P. Bottaro, W.G. Taylor, R. Jensen, S.A. Aaronson, J.S. Rubin, Hepatocyte growth factor (HGF)/NK1 is a naturally occurring HGF/scatter factor variant with partial agonist/antagonist activity, J. Biol. Chem. 271 (1996) 13110–13115. [12] G. Lagna, A. Hata, A. Hemmati-Brivanlou, J. Massague, Partnership between DPC4 and SMAD proteins in TGFbeta signalling pathways, Nature 383 (1996) 832–836. [13] Y. Zhang, X. Feng, R. We, R. Derynck, Receptor-associated Mad homologues synergize as effectors of the TGF-beta response, Nature 383 (1996) 168–172. [14] G.M. Brodeur, A. Nakagawara, Molecular basis of clinical heterogeneity in neuroblastoma, Am. J. Pediatr. Hematol. Oncol. 14 (1992) 111–116. [15] N.M. Shah, A.K. Groves, D.J. Anderson, Alternative neural crest cell fates are instructively promoted by TGFbeta superfamily members, Cell 85 (1996) 331–343. [16] P.S. Cohen, J.J. Letterio, C. Gaetano, J. Chan, K. Matsumoto, M.B. Sporn, C.J. Thiele, Induction of transforming growth factor beta 1 and its receptors during all-trans-retinoic acid
[17]
[18]
[19]
[20]
[21]
[22]
[23]
193
(RA) treatment of RA-responsive human neuroblastoma cell lines, Cancer Res. 55 (1995) 2380–2386. S. Thiagalingam, C. Lengauer, F.S. Leach, M. Schutte, S.A. Hahn, J. Overhauser, J.K. Willson, S. Markowitz, S.R. Hamilton, S.E. Kern, K.W. Kinzler, B. Vogelstein, Evaluation of candidate tumour suppressor genes on chromosome 18 in colorectal cancers, Nat. Genet. 13 (1996) 343–346. M.A. Reale, M. Reyes-Mugica, W.E. Pierceall, M.C. Rubinstein, L. Hedrick, S.L. Cohn, A. Nakagawara, G.M. Brodeur, E.R. Fearon, Loss of DCC expression in neuroblastoma is associated with disease dissemination, Clin. Cancer Res. 2 (1996) 1097–1102. M. Schutte, R.H. Hruban, L. Hedrick, K.R. Cho, G.M. Nadasdy, C.L. Weinstein, G.S. Bova, W.B. Isaacs, P. Cairns, H. Nawroz, D. Sidransky, R.A.Casero Jr., , P.S. Meltzer, S.A. Hahn, S.E. Kern, DPC4 gene in various tumor types, Cancer Res. 56 (1996) 2527–2530. M. Nagatake, Y. Takagi, H. Osada, K. Uchida, T. Mitsudomi, S. Saji, K. Shimokata, T. Takahashi, T. Takahashi, Somatic in vivo alterations of the DPC4 gene at 18q21 in human lung cancers, Cancer Res. 56 (1996) 2718–2720. S.K. Kim, Y. Fan, V. Papadimitrakopoulou, G. Clayman, W.N. Hittelman, W.K. Hong, R. Lotan, L. Mao, DPC4, a candidate tumor suppressor gene, is altered infrequently in head and neck squamous cell carcinoma, Cancer Res. 56 (1996) 2519–2521. M.T. Barrett, M. Schutte, S.E. Kern, B.J. Reid, Allelic loss and mutational analysis of the DPC4 gene in esophageal adenocarcinoma, Cancer Res. 56 (1996) 4351–4353. P.A. Hoodless, T. Haerry, S. Abdollah, M. Stapleton, M.B. O’Connor, L. Attisana, J.L. Wrana, MADR1, a MAD-related protein that functions in BMP2 signaling pathways, Cell 85 (1996) 489–500.
DPC4 splice variants in neuroblastoma Hajime Kageyama a, Naohiko Seki b, Shigeru Yamada a, Shigeru Sakiyama a, Akira Nakagawara a,* a
Division of Biochemistry, Chiba Cancer Center Research Institute, 666-2 Chuoh-ku, Chiba 260, Japan b Laboratory of Gene Structure I, Kazusa DNA Research Institute, Kisarazu, Chiba 292, Japan Received 22 July 1997; received in revised form 20 August 1997; accepted 21 August 1997
Abstract One of the loci for neuroblastoma suppressor genes is chromosome 18q21 where the DPC4 tumor suppressor gene, as well as the DCC and MADR2 genes, is located. DPC4 is a molecule of the TGF-b signal which regulates differentiation of the neural crest precursor cells from which neuroblastoma originates. During the search for the significance of DPC4 as a candidate neuroblastoma suppressor gene, we found that there are at least two variant forms of the DPC4 transcripts by using the reverse-transcriptase-PCR procedure. The subsequent sequencing analysis has revealed that one is missing exons 5 and 6 and the other is missing exons 4–6. Both splice variants were frequently observed in neuroblastomas and at low levels in normal tissues. Though the functional role of the DPC4 splice variants is unknown, they might be important in regulating the TGF-b signaling not only in neuroblastomas but also in other tumors and normal tissues. 1998 Elsevier Science Ireland Ltd. Keywords: DPC4; Alternative splicing; Neuroblastoma
1. Introduction Recent advances in cancer genetics have revealed that inactivation of tumor suppressor genes plays a major role in tumor development or progression. DPC4, originally identified as a candidate tumor suppressor gene of pancreatic cancers, frequently exhibits homozygous deletion, as well as nonsense or missense mutations in some other cancers [1,2]. The DPC4 gene is mapped at chromosome 18q21.1 and encodes a protein homologous to a product of a Drosophila
* Corresponding author. Tel.: +81 43 2645431; fax: +81 43 2654459; e-mail: [email protected]
melanogaster gene (Mad) which is implicated in a transforming growth factor-b (TGF-b)-like signaling pathway. Chromosome 18q21 is also a locus for a possible tumor suppressor gene of neuroblastoma which is one of the most common childhood tumors and is derived from the sympathoadrenal lineage of the neural crest. Takita et al. [3] reported that the DCC locus on 18q21.1 is deleted in neuroblastomas at a relatively high incidence (31%), but the mutation of the DCC gene is not so common in human neuroblastomas (Y. Hayashi, pers. commun.). Since the DPC4 gene is located close to the DCC gene, we decided to examine the role of DPC4 in neuroblastoma. During the study we found that there are at least two alternatively
0304-3835/98/$19.00 1998 Elsevier Science Ireland Ltd. All rights reserved PII S0304-3835 (97 )0 0389-3
188
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193
spliced forms of DPC4 mRNAs in both neuroblastomas and non-malignant tissues.
2. Materials and methods 2.1. Tumor tissues and cell lines Fresh tumor samples were obtained from the patients with neuroblastoma at surgery and were stored at −80°C until use. The tumors were staged according to the system of Evans et al. [4]. The derivations, descriptions and culture conditions of cell lines were described previously [5].
primer for the 3′-untranslated region of DPC4 mRNA (R0: 5′-GCCTGACAAGTTCTGAAGAGG3′). The PCR amplification for generating the DPC4 cDNA fragments was performed by using combinations of the forward primers F1 (5′-TCTGAGTCTAATGCTACCAGCACTGCC-3′), F2 and F3 (5′CCTTGCAACGTTAGCTGTTG-3′) and the reverse primers R1, R2 (5′-GTAGTCCACCATCCTGATAAGG-3′) and R3 (5′-GCAAGCTCATTGTGAACAGG3′). The amplification products were run on 1.5% agarose gel and visualized by ethidium bromide staining. The RT-PCR products were subsequently cloned into the pCRII vector by using a TA cloning kit (Invitrogen, San Diego, CA) and processed for DNA sequence analysis.
2.2. Northern blot analysis 2.4. DNA sequencing Total RNA was extracted from 0.2 to 0.5 g of primary neuroblastoma tissues or cultured cell lines according to the method of Chomczynski and Sacchi [6]. We resolved 15 mg of each RNA on 1% agaroseformaldehyde gels and transferred the RNA by blotting to a nylon membrane (Hybond-N+, Amersham, Arlington Heights, IL). Blots were hybridized, washed and exposed to X-ray film as described previously. The human adult and fetal multiple tissue blots for Northern analysis were obtained from Clontech Laboratories (Palo Alto, CA). The 1092 bp DPC4 probe was made by a PCR amplification procedure using a forward primer F2 (5′-CTGGTCAGCCAGCTACTTACC-3′) and a reverse primer R1 (5′-TCTGTCTGCTAGGAGCAAGG-3′) with cDNAs, which were reversely transcribed from the total RNA obtained from a case 63 neuroblastoma tissue, as templates. 2.3. Reverse-transcriptase (RT)-PCR and Southern blot analysis Total RNA (1 mg) from primary tumors or cell lines, or 0.2 mg of poly A+ RNA obtained from normal tissues (Clontech Laboratories, Palo Alto, CA) was used to generate complementary DNA fragments for DPC4. The reverse-transcriptase reaction was performed in the reaction mixture containing 50 units of superscript II (Life Technologies, Tokyo), 1× RT buffer, 10 mM DTT, 250 mM dNTPs, 10 units of RNAsin (Toyobo, Tokyo) and a specific antisense
DNA sequencing was performed using the dideoxy chain termination method (SequiTherm Long-Read Cycle Sequencing Kit-LC, Epicentre Technologies, Madison, WI) on an automated DNA sequencer Model 4000L (Li-COR, Lincoln, NE). The sequence information was analyzed using Genetyx (Software Development, Tokyo).
3. Results 3.1. Expression of DPC4 mRNA DPC4 mRNA expression was examined by Northern blot analysis in both neuroblastomas and normal tissues, since the expression has not been previously reported in detail. In fetal tissues, both 3 and 7 kb transcripts were strongly expressed in lung and kidney and weakly expressed in liver and brain. In adult normal tissues, the 3 kb transcript was predominantly expressed at high levels in skeletal muscle, at intermediate levels in heart, kidney, thymus, prostate, testis and ovary, at low levels in brain, placenta, lung, liver, pancreas, spleen, small intestine and colon and at undetectable levels in leukocytes. The 7 kb transcript was detectable in the adult skeletal muscle (data not shown). Fig. 1 shows the representatives of Northern blot for DPC4 expression in neuroblastomas. All 29 primary neuroblastomas examined (11 in stage 1, five in stage 2, five in stage 3 and eight in stage 4)
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193
189
Fig. 1. Northern blot analysis for mRNA expression of DPC4 in primary neuroblastomas and neuroblastoma cell lines. Total RNA (15 mg) was loaded on each lane. The RNA blots were hybridized with a 32P-labeled DPC4 probe, followed by hybridization with glyceraldehyde-3phosphate dehydrogenase (GAPDH) as the control for the amount of loaded RNA. N-myc is amplified in SK-N-BE, SK-N-DZ, RTBM1, NLF and LAN5 neuroblastoma cell lines.
expressed the 3 kb transcript at similar levels, while neuroblastoma cell lines expressed it at various levels. 3.2. Alternative splicing of DPC4 Since in some tumor suppressor genes, such as WT1 and BRCA1, RNA was abnormally processed and it resulted in a production of a functionally changed protein [7,8], we next tested by means of an RTPCR procedure whether or not there are different sizes of the DPC4 transcripts in neuroblastomas. Total RNA obtained from two representative tumors with stage 1 (case 63) and stage 4 (case 52) was allowed for the reverse-transcriptase reaction to make complementary DNAs of DPC4 by using the specific primer R0 at the 3′-UTR. The PCR amplification using the F1 primer (exon 4) and either R1 or R2 (both 3′-UTR) primer gave two bands (approximately 1200 and 970 bp, respectively) on an agarose gel (Fig. 2A), whereas the forward primer F2 paired with either R1 or R2 gave a single band which corresponded to the product size expected from the wild-type DPC4. We then performed a PCR amplification using a primer pair, F3 and R3, which allows amplification of the region spanning exons 1–7 (Fig. 2B). It gave a few smaller bands than expected and the sizes of the common bands were about 800 and 580 bp. The Southern
blot analysis showed that at least two smaller sizes of the bands hybridized to the DPC4 probe. The sequencing of both products obtained from the C-52 primary neuroblastoma demonstrated that the clone 52-1 deleted exons 4–6 (codons 152–301) and the clone 52-2 deleted exons 5 and 6 (codons 223–301) (Fig. 3A,B), according to the information about the DPC4 gene from the GenBank database (accession number U44378). Both possible splice variants were also faintly observed in normal tissues of the lung and small intestine. Because the deletion in both transcripts occurred in frame, two possible splice variants may be translated to proteins that lack amino acids corresponding to codons 152–301 and codons 223–301. The putative truncated proteins reserve both MH1 and MH2 domains which are conserved in SMAD family proteins [9], but lack a domain that is unique to DPC4 (Fig. 3B).
4. Discussion The different proteins that derive from the same gene as a result of alternative splicing are functionally related, but in some cases they exhibit distinctly counteracting biological activities. These include tumor
190
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193
Fig. 2. Identification of alternative splicing of DPC4 in neuroblastomas and normal tissues. (A) Ethidium bromide-stained 1% agarose gel of the RT-PCR products generated with DPC4-specific primer pairs; expected product sizes are 1203 bp from the pair F1 (codons 738–764 at exon 4) and R1 (codons 1922–1941 at exon 11), 1096 bp from the pair F1 and R2 (codons 1813–1834 at exon 11), 1092 bp from the pair F2 (codons 849–909 at exon 5) and R1 and 985 bp from the pair F2 and R2, according to the GenBank database (accession number U44378). Total RNA obtained from two primary neuroblastomas (stages 1 and 4) was used for the RT-PCR reactions. Only when the F1 primer was used were two different bands detectable. The size marker used was HindIII-digested l DNA. (B) DPC4 splice variants in primary neuroblastomas, neuroblastoma cell lines and human normal tissues. (Upper) Ethidium bromide-stained 2% agarose gel of the RT-PCR products generated with DPC4-specific primer pair F3 (codons 31–50 at exon 1) and R3 (codons 1035–1054 at exon 7). Arrows indicate the sizes of the products of wild-type (upper), about 800 bp (middle) and about 580 bp (lower). The size marker used was pUC118 digested with both HinfI and NciI. (Lower) The gel in the upper panel was blotted on the Nylon membrane, which was subsequently probed with a 32Plabeled DNA fragment of DPC4.
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193
suppressor genes such as WT1 and BRCA1 as well as genes for bcl-x and hepatocyte growth factor [7,8, 10,11]. We here demonstrate that at least two splice variants of DPC4 mRNA are frequently expressed in human neuroblastomas; one deleted exons 5 and 6 and the other deleted exons 4–6. DPC4 has recently been found to be a molecule in the signaling pathway of the TGF-b superfamily [12,13]. Since both variants are also expressed in some non-malignant tissues, they could function as general modifiers of cell function in both malignant and non-malignant cells responsive to growth factors of the TGF-b superfamily.
191
Neuroblastoma originates from the sympathoadrenal progenitor cells which are derived from neural crest cells [14]. Recently, Shah et al. [15] suggested that the fate of neural crest stem cells during early embryogenesis is strongly influenced by TGF-b and bone morphogenic protein 2 (BMP2), which are growth factors of the TGF-b superfamily. However, the functions of BMP2 and TGF-b are different in that the former induces neurogenesis in neural crest cells, whereas the latter promotes smooth muscle differentiation. The responsiveness to TGF-b is reserved in some neuroblastoma cell lines, but the response pat-
Fig. 3. (A) Sequence of aberrant RT-PCR products of DPC4 from case 52 primary neuroblastoma. The sequences of subcloned RT-PCR products (clones 52-1 and 52-2) showed loss of exons 4–6 and exons 5 and 6, respectively. (B) Diagram of aberrant mRNA transcripts of DPC4. Splice sites are indicated based on the GenBank database (accession number U44378). Arrows indicate the location and direction of primers used for RT-PCR analyses.
192
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193
terns are diverse [16]. In addition, many lines do not respond to TGF-b phenotypically. These findings suggest that neuroblastoma may have some kind of disregulation in the TGF-b signal transduction pathway. The chromosome locus at 18q21 is where DCC, a candidate tumor suppressor gene for colorectal cancer, is localized [17]. The DCC locus is reported to be deleted in about 30% of neuroblastomas by a single study of the loss of heterozygosity [3]. Furthermore, loss of DCC expression is significantly associated with the dissemination of neuroblastoma [18]. DCC could therefore also be a candidate tumor suppressor gene for neuroblastoma, although that needs to be confirmed. The DPC4 gene, which was originally reported as a tumor suppressor gene for pancreatic carcinomas, is localized at chromosome 18q21, quite close to the DCC gene [1,2]. However, recent extensive studies suggest that mutations or deletions in other cancers are uncommon (,10%) [19–22]. Our preliminary study of neuroblastomas showed a missense mutation in only one of 15 neuroblastoma cell lines (unpublished data). This suggests that if DPC4 could function as a tumor suppressor for neuroblastoma, the mechanism might be other than mutations or deletions of the gene. Accumulating evidence demonstrates that intracellular signals from TGF-b or BMP are mediated by the formation of hetero-oligomeric complexes of DPC4 and Smad1 (Madr1) or Smad2 (Madr2) [9,12,13,23]. Loss of the oligomerization may be critical in the biological responses induced by growth factors of the TGF-b superfamily. From that point of view, the presence of alternatively spliced variants of DPC4 in neuroblastomas might be important. Two splice variants we report here lose exons 5 and 6 or exons 4–6, both of which are in frame but outside of the conserved domains of MH1 and MH2. The functional roles of MH1 and MH2 are still unknown, although truncation of C-terminal 39 amino acids is reported to act dominant-negatively on the TGF-b-induced response [13]. Thus, it is possible that the presence of either or both forms of DPC4 splice variants may have an influence on oligomerization between intact DPC4 and Smad proteins, resulting in inhibition of TGF-b signaling. Intriguingly, our present results suggest that the advanced stage neuroblastomas express more DPC4 splice variants than the low stages tumors. However, this needs to be confirmed by a study based
on the quantitative analysis of a large number of samples. In summary, we found at least two alternatively spliced variants of DPC4 mRNAs in human neuroblastomas as well as some normal tissues at low levels. Both sites spliced out are between MH1 and MH2 domains and may affect the oligomerization of DPC4 and Smads. The TGF-b signaling is important in regulating the growth and differentiation of cancer cells as well as normal cells. The frequent expression of DPC4 splice variants may play a role in an aberrant regulation of neuroblastoma growth.
5. Addendum During the preparation of this manuscript, Johan P. de Winter et al. published a DPC4 splice variant which was cloned from a breast tumor cell line, MDA-MB-231 (Oncogene 14 (1997) 1891–1899). The splice variant was the same as one of two variants found in this manuscript. References [1] S.A. Hahn, M. Schutte, A.T. Hoque, C.A. Moskaluk, L.T. da Costa, E. Rozenblum, C.L. Weinstein, A. Fischer, C.J. Yeo, R.H. Hruban, S.E. Kern, DPC4, A candidate tumor suppressor gene at human chromosome 18q21.1, Science 271 (1996) 350–353. [2] S.A. Hahn, A.T. Hoque, C.A. Moskaluk, L.T. da Costa, M. Schutte, E. Rozenblum, A.B. Seymour, C.L. Weinstein, C.J. Yeo, R.H. Hruban, S.E. Kern, Homozygous deletion maps at 18q21.1 in pancreatic cancer, Cancer Res. 56 (1996) 490– 494. [3] J. Takita, Y. Hayashi, T. Kohno, M. Shiseki, N. Yamaguchi, R. Hanada, K. Yamamoto, J. Yokota, Allelotype of neuroblastoma, Oncogene 11 (1995) 1829–1834. [4] A.E. Evans, G.J. D’Angio, J.A. Randolph, A proposed staging for children with neuroblastoma: children’s cancer study group, Cancer 27 (1971) 374–378. [5] A. Nakagawara, C.G. Azar, N.J. Scavarda, G.M. Brodeur, Expression and function of TRK-B and BDNF in human neuroblastomas, Mol. Cell. Biol. 14 (1994) 759–767. [6] P. Chomczynski, N. Sacchi, Single-step method of RNA isolation by acid guanidinium thiocyanate–phenol–chloroform extraction, Anal. Biochem. 162 (1987) 156–159. [7] A.L. Menke, N. Riteco, R.C.A. van Ham, C. de Bruyne, F.J. Rauscher III, A.J. van der Eb, A.G. Jochemsen, Wilms’ tumor 1 splice variants have opposite effects on the tumorigenicity of adenovirus-transformed baby-rat kidney cells, Oncogene 12 (1996) 537–546.
H. Kageyama et al. / Cancer Letters 122 (1998) 187–193 [8] M. Lu, S.D. Conzen, C.N. Cole, B.A. Arrick, Characterization of functional messenger RNA splice variants of BRCA1 expressed in nonmalignant and tumor-derived breast cells, Cancer Res. 56 (1996) 4578–4581. [9] K. Eppert, S.W. Scherer, H. Ozcelik, R. Pirone, P. Hoodless, H. Kim, L.-C. Tsui, B. Bapat, S. Gallinger, I.L. Andrulis, G.H. Thomsen, J.L. Wrana, L. Attisano, MADR2 maps to 18q21 and encodes a TRGb-regulated MAD-related protein that is functionally mutated in colorectal carcinoma, Cell 86 (1996) 543–552. [10] L.H. Boise, M. Gonzalez-Garcia, C.E. Postema, L. Ding, T. Lindsten, L.A. Turka, X. Mao, G. Nunez, C.B. Thompson, bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death, Cell 74 (1993) 597–608. [11] V. Cioce, K.G. Csaky, A.M. Chan, D.P. Bottaro, W.G. Taylor, R. Jensen, S.A. Aaronson, J.S. Rubin, Hepatocyte growth factor (HGF)/NK1 is a naturally occurring HGF/scatter factor variant with partial agonist/antagonist activity, J. Biol. Chem. 271 (1996) 13110–13115. [12] G. Lagna, A. Hata, A. Hemmati-Brivanlou, J. Massague, Partnership between DPC4 and SMAD proteins in TGFbeta signalling pathways, Nature 383 (1996) 832–836. [13] Y. Zhang, X. Feng, R. We, R. Derynck, Receptor-associated Mad homologues synergize as effectors of the TGF-beta response, Nature 383 (1996) 168–172. [14] G.M. Brodeur, A. Nakagawara, Molecular basis of clinical heterogeneity in neuroblastoma, Am. J. Pediatr. Hematol. Oncol. 14 (1992) 111–116. [15] N.M. Shah, A.K. Groves, D.J. Anderson, Alternative neural crest cell fates are instructively promoted by TGFbeta superfamily members, Cell 85 (1996) 331–343. [16] P.S. Cohen, J.J. Letterio, C. Gaetano, J. Chan, K. Matsumoto, M.B. Sporn, C.J. Thiele, Induction of transforming growth factor beta 1 and its receptors during all-trans-retinoic acid
[17]
[18]
[19]
[20]
[21]
[22]
[23]
193
(RA) treatment of RA-responsive human neuroblastoma cell lines, Cancer Res. 55 (1995) 2380–2386. S. Thiagalingam, C. Lengauer, F.S. Leach, M. Schutte, S.A. Hahn, J. Overhauser, J.K. Willson, S. Markowitz, S.R. Hamilton, S.E. Kern, K.W. Kinzler, B. Vogelstein, Evaluation of candidate tumour suppressor genes on chromosome 18 in colorectal cancers, Nat. Genet. 13 (1996) 343–346. M.A. Reale, M. Reyes-Mugica, W.E. Pierceall, M.C. Rubinstein, L. Hedrick, S.L. Cohn, A. Nakagawara, G.M. Brodeur, E.R. Fearon, Loss of DCC expression in neuroblastoma is associated with disease dissemination, Clin. Cancer Res. 2 (1996) 1097–1102. M. Schutte, R.H. Hruban, L. Hedrick, K.R. Cho, G.M. Nadasdy, C.L. Weinstein, G.S. Bova, W.B. Isaacs, P. Cairns, H. Nawroz, D. Sidransky, R.A.Casero Jr., , P.S. Meltzer, S.A. Hahn, S.E. Kern, DPC4 gene in various tumor types, Cancer Res. 56 (1996) 2527–2530. M. Nagatake, Y. Takagi, H. Osada, K. Uchida, T. Mitsudomi, S. Saji, K. Shimokata, T. Takahashi, T. Takahashi, Somatic in vivo alterations of the DPC4 gene at 18q21 in human lung cancers, Cancer Res. 56 (1996) 2718–2720. S.K. Kim, Y. Fan, V. Papadimitrakopoulou, G. Clayman, W.N. Hittelman, W.K. Hong, R. Lotan, L. Mao, DPC4, a candidate tumor suppressor gene, is altered infrequently in head and neck squamous cell carcinoma, Cancer Res. 56 (1996) 2519–2521. M.T. Barrett, M. Schutte, S.E. Kern, B.J. Reid, Allelic loss and mutational analysis of the DPC4 gene in esophageal adenocarcinoma, Cancer Res. 56 (1996) 4351–4353. P.A. Hoodless, T. Haerry, S. Abdollah, M. Stapleton, M.B. O’Connor, L. Attisana, J.L. Wrana, MADR1, a MAD-related protein that functions in BMP2 signaling pathways, Cell 85 (1996) 489–500.