Neuroblastoma (Nb) is a malignancy of the sympathetic nervous system which affects children in their first decade. It is the most common extra-cranial solid tumor in children with an incidence of approximately 1 in 8–10 000 live births annually and accounts for approximately 10% of all children's cancers. Ganglioneuroblastoma is a relatively benign form of Nb and consists of a mixture of fibrils, mature and maturing ganglion cells, as well as undifferentiated neuroblasts. During routine cytogenetic analysis of patients with different manifestations of neuroblastoma we have identified one patient with ganglioneuroblastoma that carries an apparently balanced t(1:13)(q21:q12) reciprocal translocation. Positional cloning of the translocation breakpoint on chromosome 13 resulted in the mapping of the breakpoint between coding exon 2 and exon 3 of WAVE3, a member of WASP gene family. Although the breakpoint region on chromosome 1 was localized to within 2 kb of genomic sequence, no gene was found to be interrupted on this chromosome. The WAVE3 transcript is mainly expressed in the nervous system and, like all the members of the WASP gene family, WAVE3 is a key element in actin polymerization and cytoskeleton organization. WAVE3, therefore, is important for cell differentiation and motility and its expression is lost in a number of low grade and stage 4S tumors. From analysis of its expression pattern and function, WAVE3 is a candidate tumor suppressor gene, at least in some forms of neuroblastoma.
Neuroblastoma (Nb) is the most common extracranial solid tumor in childhood, accounting for 8% of all pediatric malignancy. Family studies (Kushner et al., 1986) and epidemiological analyses (Knudson and Strong, 1972) suggests that the Nb tumor phenotype segregates as an autosomal dominant trait with reduced (63%) penetrance. This led to the proposal that tumorigenesis was a consequence of a loss of function of a critical gene(s) in embryonic tumor cells in a way similar to that proposed for retinoblastoma (Knudson, 1971). However, because Nb families are rare (Kushner et al., 1986), it has not yet been possible to assign any Nb predisposition gene to a particular chromosome region using conventional linkage studies. This analysis may be further complicated by the possibility that Nb may, in fact, represent several different clinical entities. Thus, ganglioneuroblastoma, which is considered a relatively benign tumor, consists of both immature neuroblasts as well as fully differentiated neuronal cells. Ganglioneuroma, on the other hand, consists entirely of differentiated cells. What is not known is whether these two variants represent distinct diseases or whether ganglioneuroma arises as a result of differentiation of cells from a ganglioneuroblastoma. The most aggressive forms of Nb within the disease spectrum consist of undifferentiated neuroblasts. It has been consistently demonstrated that undifferentiated Nb cells can differentiate in vitro when exposed to a variety of inducing agents (Ponzoni et al., 1991; Lanciotti et al., 1992), indicating that differentiation signals in these cells can overcome the malignant phenotype. Thus, whether neuroblastoma represents the end point or a continuous progression from the more benign forms, or whether they are different diseases, is not clear. The indication that Nb in fact probably represents several different diseases has been suggested from survival studies (Castleberry et al., 1999). The final conundrum comes from the special form of the disease called ‘stage 4S’. This tumor presents as a low-grade primary tumor in children under the age of 12 months, but is associated with distant metastases primarily to the liver and bone marrow. The feature that makes this form of Nb stand out is the fact that these tumors can spontaneously resolve themselves (D'Angio et al., 1971). Clearly, in this sub-type, there has been a massive expansion of primitive neuroblasts, although these cells can still respond to differentiation or apoptotic signals. Again whether this represents a distinct form of the disease or arises through a completely different pathway has not been determined.
Analysis of constitutional chromosome abnormalities in individuals who develop particular tumor types has frequently indicated the site of predisposition gene. This has been the case for retinoblastoma (Yunis and Ramsay, 1978), Wilms' tumor (Riccardi et al., 1978), neurofibromatosis (Fountain et al., 1989; O'Connell et al., 1989), and familial adenomatous polyposis (Varesco et al., 1989). These structural abnormalities are predominantly deletions, but where translocations are identified, the breakpoints usually interrupt the predisposition gene (reviewed in Mitchell, 1991). A few cases of Nb with constitutional chromosome abnormalities have been described (Anderson et al., 2001; Fryns, 1996; Laureys et al., 1990, Nagano et al., 1980; Sanger et al., 1984). In particular, two patients with stage 3 and 4 tumors were shown to carry constitutional chromosome changes involving 1p35-36. Laureys et al. (1990) reported a patient with a constitutional reciprocal translocation which was later defined (van Roy et al., 1997) although no genes have been reported to be involved in this rearrangement to date. We (Mead and Cowell, 1995) described a patient with stage 4S disease, and a constitutional t(1;10)(p22;q21) translocation. The chromosome 1 translocation breakpoint does not lie within the consensus LOH region seen in advanced stage tumors but does lie in the LOH region reported for stage 4S tumors (Mora et al., 2000). Cloning the breakpoint from this translocation identified an in-frame fusion of two genes resulting in the truncation and inactivation of the NB4S/EVI5 gene (Roberts et al., 1998a,b). The association of different chromosome rearrangements with different subtypes of the disease was further demonstrated in the report by Michalski et al. (1992) who described a patient with ganglioneuroblastoma and a constitutional t(1;13)(q23;q12) reciprocal chromosome translocation. The breakpoint on chromosome 1 was in a different position compared with the patients with stage 3, 4 and 4S. Recent reports suggest that amplification of this region is frequently seen in low-grade tumors and so may represent a subtype-specific gene locus for Nb (Hirai et al., 1999). We have now cloned the breakpoints associated with this 1;13 translocation and shown that WAVE3 on 13q12, which is though to be involved in actin-polymerization (Suetsugu et al., 1999) is physically interrupted and inactivated as a result of this translocation event. We have shown that the WAVE3 transcript is down regulated in three sporadic neuroblastoma tumors. No known or predicted gene is associated with the 1q23 breakpoint.
Mapping the breakpoint on chromosome 1
To study the de novo constitutional t(1;13)(q22;q12) translocation in patient DG a somatic cell hybrid, DGF27 (Michalski and Cowell, 1993), was isolated which retained the derivative chromosome 1, (1pter-q21;13q12-qter).
Analysis of the 1q21-23 breakpoint using the DGF27 somatic cell hybrid established that the breakpoint lay between markers WI7842 and WI4536 (Figure 1). Both of these markers were present in YAC 955E11. FISH analysis showed that this YAC crossed the breakpoint (data not shown). Using the WI7842 and WI4536 markers we then screened a BAC library to create a contig across the critical region. This strategy involved sequential end sequencing of BACs as they were isolated and use of PCR primer pairs generated from these end sequences to re-screen hybrid DGF27. BACs from the telomeric end of the contig (B535J20) were isolated using WI4536 and, using end clones, overlapping BAC clones B27F19 (containing IB3262) and B180E12 were isolated (Figure 1). None of these crossed the breakpoint by PCR analysis. Using marker WI7842 we isolated BAC B326C19 and PCR primers from the end sequences of this BAC were then used to isolate B565A21 and B498K12. Since B498K12 also contained WI7842, we were able to orientate this contig of three BACs and used the end clone of B498K12 to isolate B338G16. Again all of the end clone sequences were shown by PCR to be present in hybrid DGF27 showing that they lay above the breakpoint. Finally PCR primers from the distal end of B338G16A were used to isolate B236C22 and analysis of DGF27 using the end clone sequences from this BAC were shown to lie on either side of the translocation breakpoint. A summary of this mapping effort is shown in Figure 1. The fact that B236C22 crosses the breakpoint was confirmed by FISH using chromosomes from the lymphoblastoid cell line prepared from patient DG (data not shown). B236C22 was shown by pulse field gel analysis to be approximately 110 kb long (data not shown) and the entire BAC 236C22 (GDB accession # AC011666) was sequenced. BLAST analysis identified four genes in the region (Figure 2); an anonymous EST, a cDNA clone (IMAGE1676497) of unknown function and two members of the S100 family of genes, S100A9 and S100A12. PCR primers were designed from the 3′UTR and 5′UTR of S100A12 and S100A9 but neither crossed the translocation breakpoint. All seven known exons of cDNA clone 1676497 which map centromeric to S100A9 gene were also absent in hybrid DGF27 demonstrating that none of these three genes are involved the translocation event. Similarly, the EST sequence lay above the breakpoint. No other genes were identified in BAC 236C22 using in silico analysis.
To critically define the breakpoint location, therefore, primer pairs were designed along the sequence of BAC 236C22 beyond cDNA clone 1676497 at 2 kb intervals and were tested for presence or absence in hybrid DGF27 (Figure 2). In this way we were able to map the translocation breakpoint on chromosome 1 to within a 2 kb genomic sequence as shown in Figure 2. This region contains interspersed highly repeated elements of the Alu families. No known gene or EST were found in the sequence around the translocation breakpoint.
Mapping the breakpoint on chromosome 13
To determine whether a gene on chromosome 13 was disrupted by the breakpoint we used FISH and PCR to study the 13q12 breakpoint. The original analysis of the 13q12 region (Michalski and Cowell, 1993) was limited by the availability of markers in the region of the breakpoint. Using STS markers from the Whitehead YAC contigs W13.0 and W13.1 (Figure 3) which map to the proximal and distal ends of band 13q12, respectively, we showed that WI4676 from the most distal end of contig W13.0 was absent in DGF27 whereas marker WI6155 from the most proximal end of contig W13.1 was present (Figure 3). These results were confirmed using FISH analysis with BAC clones B172I24 and B144H23 which contain markers WI4676 and WI6155, respectively (data not shown). Contigs W13.0 and W13.1 were not physically linked and so it was difficult to estimate the size of the gap between the two contigs. Analysis of the BAC contigs in the sequence databases from the 13q12 region showed that BAC clones 172I24 and 144H23 were present in BAC contigs NT_024513 and NT_05884, respectively (Figure 3). These two contigs are 0.61 Mbp and 1.37 Mbp long respectively, with only one 1.2 Mbp contig (NT_00998) between them. Since several functional genes are present within these contigs, we designed PCR primer pairs from the 5′ and 3′ ends of each of these genes and analysed DGF27. This strategy demonstrated that WAVE3 was interrupted by the 13q12 breakpoint. The availability of the genomic DNA sequence allowed us to define the exon/intron structure (see below) of this gene which allowed us to further demonstrate that the breakpoint was in the intron between exon 2 and exon 3 (Figure 4a). The fact that WAVE3 gene crosses the breakpoint was confirmed by Southern analysis of DNA from DGF27 using a DNA probe which spanned exons 1–4 of the WAVE3 transcript (Figure 4b).
Genomic organization of WAVE3
The product of WAVE3 (Acc. # AB026543) was first identified as a protein which has similarity to the Wiskott-Aldrich syndrome protein (WASP) and the Neural-WASP (N-WASP) protein, especially in the C-terminal region (Suetsugu et al., 1999). A putative leucine zipper motif is found at the N-terminal end of the gene and a proline-rich domain is found at the C-terminal end adjacent to verprolin, cofilin and acidic (VCA) domains. This gene family was shown to be involved in cytoskeleton actin polymerization and reorganization, and cell motility processes (Takenawa and Miki, 2001). The open reading frame of WAVE3 is 1509-base pair long and encodes for a predicted 502-amino acid polypeptide sequence with a calculated molecular mass of 55 kD which has a high degree of identity with the other WAVE proteins in the C terminal region (Suetsugu et al., 1999). We used sequence alignment against genomic sequences from the public databases to determine the genomic structure of WAVE3. Alignment against the EST database allowed the identification of a new EST clone (Acc. # BF115452), the sequence of which extended the WAVE3 cDNA 179 nucleotides into the 5′ untranslated region and which identified 2 new exons (New acc. # AF454702). Thus, the updated genomic structure of WAVE3 spans more than 150 kb of genomic sequence and contains 10 exons, 8 which encode for the WAVE3 protein (Table 1).
Sequence analysis at the breakpoint junction
We were able to map the translocation breakpoint on chromosome 1 to within 2 kb of genomic sequence, which was shown to be composed entirely of repetitive sequences from the Alu family. The breakpoint in 13q12 lies within intron 2 which is ∼2350 bp long and contains mostly repetitive elements of the same Alu family which are present in the junction region on chromosome 1. It appears, therefore, that this chromosome translocation event has arisen as a result of recombination between homologous sequences.
Mutation analysis of WAVE3
PCR products from each of the 10 exons of WAVE3 (Table 2) were analysed from 28 sporadic Nb tumors from different stages using IPS and DHPLC (WAVE). We identified only one SSCP change in exon 7 which was present in nine different tumors. The same change was detected by WAVE analysis. Subsequent sequencing of the corresponding PCR products identified a single nucleotide change corresponding to a T to C transition at nucleotide 678 or the ORF which did not affect the amino-acid at that position (H226H). The nucleotide change was confirmed using the StyI and NcoI restriction enzymes whose recognition sites were destroyed as a result of the T→C change. This nucleotide change, therefore, represents a polymorphism, rather than a disease causing mutation.
Expression of WAVE3 in neuroblastoma tumors and cell lines
The expression pattern of WAVE3 was determined using a specific probe which spans exons 1–4 of the transcript. The 443-bp probe which shows less than 60% homology to WAVE1 and WAVE2 was generated from a human fetal brain cDNA library using primers 5′-TGCCTTTAGTGAAGAGGAACA-3′ and 5′-CAGCCCATCCTTCTTGTCAT-3′ and was hybridized to multiple tissue Northern blots (Clontech) containing poly(A) selected RNA from normal human adult tissues. A single major transcript of ∼4.4 kb was expressed in all the tissues tested (Figure 5). A high level of expression was detected in the brain and testis while a very low level of expression was detected in the thymus, spleen, colon and small intestine. The 4.4 Kb transcript was either not expressed, or present at low levels in the placenta and peripheral blood. A much smaller 1.4 kb isoform was exclusively expressed in the testis.
To investigate the expression levels of WAVE3 in neuroblastoma we analysed primary tumors and cell lines using RT–PCR and primer pairs (5′-TGCCTTTAGTGAAGAGGAACA-3′ and 5′-ATTCGAATAGCAGCGAGGAG-3′) designed from the WAVE3 transcript to amplify a 1348-bp product spanning seven of the eight exons of the open reading frame. No RT–PCR product was detected in normal lymphoblasts, which may account for why we were unable to detect a PCR product for WAVE3 in the lymphoblastoid cell line (ET72A) carrying the 1;13 translocation (Figure 6). Whether the translocation event completely abrogated the expression of WAVE3 in the tumor from patient DG is not known since no tumor sample, which could be used for RNA extraction, was available. Despite a fairly ubiquitous expression in solid tissues, RT–PCR analysis of 17 neuroblastoma tumors showed that expression of WAVE3 was barely detectable in two tumors from stage 4S neuroblastoma and showed a very low level of expression in a third tumor from stage A neuroblastoma (Figure 6). Since these tumors likely contain normal cells it is possible that the very low levels of expression are derived from these contaminating cells.
Cloning of the mouse Wave3 gene
We used oligonucleotide primers specific to the human WAVE3 transcript in an RT–PCR reaction to amplify a 413 bp product using RNA extracted from NIH3T3 mouse cells. The sequence of the 413-bp fragment was aligned with the public mouse EST Database (NCBI) which allowed the identification of three mouse EST clones (Acc. # BG694902, BG293850 and BG295757), the partial sequence of which showed almost 100% identity to the 413 bp PCR product. We then generated the complete sequence of these three clones which could be assembled into a single, 2.4-kb long contig encompassing the open reading frame of the mouse Wave3 transcript (Acc. # AF454703). The mouse Wave3 gene showed 85% identity at the nucleotide level to its human counterpart within the open reading frame, and 94% identity at the protein level (Figure 7).
Constitutional chromosome translocations have repeatedly identified genes interrupted by the rearrangements, although in many cases it is not immediately clear how these genes are involved in the phenotype. In this report we demonstrate that WAVE3 is truncated as a result of a balanced reciprocal translocation in a patient with ganglioneuroblastoma. Since only 1–2% of the genome codes for genes it is unlikely that this is a random event that involves a coding region by chance. In fact, all of the constitutional chromosome translocation breakpoints we have described to date (Mitchell and Cowell, 1989; Roberts et al., 1998b) as well as tumor-specific translocations (Still and Cowell, 1998; Chernova et al., 1998, 2001), have involved breakpoints within the coding regions of genes, often on both sides of the breakpoints. From the molecular evidence it seems likely that WAVE3 is inactivated as a result of this translocation but, since it is not normally expressed in hematopoietic cells, we cannot verify this using the only tissue we have available from patient DG, which is a lymphoblastoid cell line. This patient was treated for ganglioneuroblastoma over 20 years ago and no tissue is available for analysis. Interestingly, the three tumors which showed absence of WAVE3 expression were also derived from low grade tumors. It is unlikely that the rearrangement generates a chimeric gene, as shown in another Nb patient described by Roberts et al. (1998b), since no gene could be found on chromosome 1. Consequently, the translocation appears to be an inactivating event. The fact that WAVE3 can be inactivated in Nb was demonstrated by the absence or reduced of expression in three out of 17 tumors analysed. Surprisingly, none of these tumors showed mutations in the coding region although there are many other ways in which expression can be silenced which have not yet been explored. In this analysis there was no association with MYCN amplification status. Another possibility is that the truncated WAVE3 gene product acts as a dominant negative mutation in neural crest cells, since the exons which are retained under the control of the WAVE3 promoter still encode the leucine zipper motif which is associated with protein–protein binding capabilities. From the structure of the rearrangement, however, it is hard to imagine how this would be processed in the absence of a clear in-frame fusion with chromosome 1 sequences.
WAVE3 is a member of the WAVE family, which has clearly been implicated in the actin polymerization process (Takenawa and Miki, 2001). It should be noted, however, that although the functions of the other members of the WAVE family have been more fully investigated, the exact role of WAVE3 is still not known. WAVE3, like all WAVE proteins, is usually present in an inactive form in which the VCA region is masked (Miki et al., 1998). In the presence of an appropriate signal the VCA region is exposed which allows the actin polymerization process to take place via recruitment of the actin-related proteins (Arp2/3) complex and monomeric actin (Miki et al., 1998). It has been shown that WAVE2 is activated via a specific interaction between the SH3 domain of IRSp53 and the prolin rich domain (PRD) of WAVE2, downstream of Rac signaling pathway (Miki et al., 2000). The SH3-containing proteins which are involved in the activation of WAVE1 and WAVE3 are yet to be identified, although it has been suggested that all three WAVE proteins co-localize, where they are expressed in the same tissue, and work in a cooperative manner during the acting polymerization and reorganization events (Westphal et al., 2000). Clearly, loss of WAVE3 function may well affect normal differentiation and migration, which are important events in the life of neural crest cells. Loss of these functions could give rise to localized accumulation of cells interrupted in the migration process. Thus, cells with a proliferative potential but unable to migrate correctly, might form enlarged ganglia of undifferentiated cells which can later differentiate when the appropriate signals are received, which is the manifestation of ganglioneuroblastoma.
During our analysis we noted that WAVE2 is located in 1p35–36, which is a region that frequently undergoes LOH and deletion in advanced stage Nb. We were unable, however, to demonstrate any loss of expression in the cohort of Nb tumors that were available to us. Mutation analysis in these same tumors using D-HPLC also failed to find any inactivating mutations in the coding region (data not shown), although several intron polymorphisms were identified. WAVE2, however, is expressed in every tissue unlike WAVE3 which might suggest that it is essential for cell viability and therefore cannot represent a target for tumorigenesis. Clearly a better understanding of the role of WAVE3 in actin polymerization is needed to be able to establish its role in the development of Nb.
Materials and methods
Isolation of BACs
BAC clones were obtained by screening the Research Genetics Human BAC Library Pools (Release IV) or by searching the Genome Survey Sequence (GSS) databases of the NCBI's BLAST server (Altschul et al., 1990). Individual oligonucleotide primer pairs were used to screen the pools of DNA provided using standard PCR conditions optimized for each oligonucleotide pair. Individual BACs were purchased from Research Genetics and tested against the original primers for authenticity. BAC DNA was prepared for restriction enzyme analysis and end-sequencing as described by Roberts et al. (1998a) and end-sequencing was achieved directly from the linearized BAC using T7 and SP6 vector primers.
STSs, PCR assays, and physical mapping
YAC contigs of the 1q21 and 13q12 regions, and marker information is available through the Whitehead Genome Center. PCR assays used Taq DNA polymerize (Qiagen, CA, USA) and initial denaturation at 94°C for 4 min followed by 35 cycles of 95°C for 20 s, 55–65°C for 30 s, and 72°C for 30 s. Template genomic DNA varied from 25 to 100 ng per 20 μl reaction.
Genomic DNA was isolated from tumor samples as described by Sambrook et al. (1989). Incorporation PCR SSCP (IPS) was performed as described in Sossey-Alaoui et al. (1999), using primers specific to each exon of WAVE3 (Table 2). We also used the WAVE instrument (Transgenomics, NE, USA) to perform Denaturing High Performance Liquid Chromatography (DHPLC) as described by Sossey-Alaoui et al. (2001), and search for mutations in WAVE3.
Northern blots (Clontech, CA, USA) containing poly(A+) RNAs from different human adult and fetal tissues were hybridized with a 413-bp cDNA probe (nt 2 to nt 414) from the coding region of the WAVE3 transcript. Filters were washed at high stringency according to the manufacturer's (Clontech, CA, USA) instructions.
Computational sequence analysis
High throughput genomic sequencing of BACs was performed as described by Kitamura et al. (2000). Genomic sequences were analysed in blocs of 20 kb of sequence using the BLASTN program against the dbEST database (Altschul et al., 1990). Alignments of predicted genes against various databases were made using the Genome Analysis Pipeline (Oak Ridge National Laboratory: http://www.ornl/GP), which predicts both genes and exons using GRAIL (Xu et al., 1994) and GENESCAN (Burge and Karlin, 1997), as well as providing integrated GRAIL annotated features and BLAST and BEAUTY analyses.
Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ . 1990 J. Mol. Biol. 215: 403–410
Anderson J, Kempski H, Hill L, Rampling D, Gordon T, Michalski A . 2001 Br. J. Cancer 85: 493–496
Burge C, Karlin S . 1997 J. Mol. Biol. 268: 78–94
Castleberry RP . 1999 J. Pediatr. Hematol. Oncol. 21: 178–180
Chernova OB, Hunyadi A, Malaj E, Pan H, Crooks C, Roe B, Cowell JK . 2001 Oncogene 20: 5378–5392
Chernova OB, Somerville RP, Cowell JK . 1998 Oncogene 17: 2873–2881
D'Angio GJ, Evans AE, Koop CE . 1971 Lancet 1: 1046–1049
Fountain JW, Wallace MR, Bruce MA, Seizinger BR, Menon AG, Gusella JF, Michels VV, Schmidt MA, Dewald GW, Collins FS . 1989 Science 244: 1085–1087
Fryns JP . 1996 Genet. Couns. 7: 73
Hirai M, Yoshida S, Kashiwagi H, Kawamura T, Ishikawa T, Kaneko M, Ohkawa H, Nakagawara A, Miwa M, Uchida K . 1999 Genes Chromosomes Cancer 25: 261–269
Kitamura E, Su G, Alaoui-Soussy K, Malaj E, Lewis J, Pan H, Hawthorn LA, Roe B, Cowell JK . 2000 Oncogene 19: 5772–5780
Knudson AG . 1971 Proc. Natl. Acad. Sci. USA 68: 820–823
Knudson AG, Strong LC . 1972 Am. J. Hum. Genet. 24: 514–532
Kushner BH, Gilbert F, Helson L . 1986 Cancer 57: 1887–1893
Lanciotti M, Montaldo PG, Folghera S, Lucarelli E, Cornaglia-Ferraris P, Ponzoni M . 1992 Cell. Mol. Neurobiol. 12: 225–240
Laureys G, Speleman F, Opdenakker G, Benoit Y, Leroy J . 1990 Genes Chromosomes Cancer 2: 252–254
Mead RS, Cowell JK . 1995 Cancer Genet. Cytogenet. 81: 151–157
Michalski AJ, Cotter FE, Cowell JK . 1992 Oncogene 7: 1595–1602
Michalski AJ, Cowell JK . 1993 Genomics 18: 141–143
Miki H, Suetsugu S, Takenawa T . 1998 EMBO J. 17: 6932–6941
Miki H, Yamaguchi H, Suetsugu S, Takenawa T . 2000 Nature 408: 732–735
Mitchell CD . 1991 Br. Med. Bull. 47: 136–156
Mitchell CD, Cowell JK . 1989 Oncogene 4: 253–257
Mora J, Cheung NK, Kushner BH, LaQuaglia MP, Kramer K, Fazzari M, Heller G, Chen L, Gerald WL . 2000 J. Mol. Diagn. 2: 37–46
Nagano H, Kano Y, Kobuchi S, Kajitani T . 1980 Jinrui Idengaku Zasshi 25: 39–45
O'Connell P, Leach R, Cawthon RM, Culver M, Stevens J, Viskochil D, Fournier RE, Rich DC, Ledbetter DH, White R . 1989 Science 244: 1087–1088
Ponzoni M, Lanciotti M, Montaldo PG, Cornaglia-Ferraris P . 1991 Cell. Mol. Neurobiol. 11: 397–413
Riccardi VM, Sujasnsky E, Smith AC, Francke U . 1978 Pediatrics 61: 604–610
Roberts T, Chernova O, Cowell JK . 1998a Cancer Genet. Cytogenet. 100: 10–20
Roberts T, Chernova O, Cowell JK . 1998b Hum. Mol. Genet. 7: 1169–1178
Sambrook J, Fritsch EF, Maniatis T . 1989 Molecular cloning: a laboratory manual 2nd edn. Nolan C (ed) Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press
Sanger WG, Howe J, Fordyce R, Purtilo DT . 1984 Cancer Genet. Cytogenet. 11: 153–159
Sossey-Alaoui K, Kitamura E, Cowell JK . 2001 Int. J. Mol. Med. 7: 543–546
Sossey-Alaoui K, Lyon JA, Jones L, Abidi FE, Hartung AJ, Hane B, Schwartz CE, Stevenson RE, Srivastava AK . 1999 Genomics 60: 330–340
Still IH, Cowell JK . 1998 Blood 92: 1456–1458
Suetsugu S, Miki H, Takenawa T . 1999 Biochem. Biophys. Res. Commun. 260: 296–302
Takenawa T, Miki H . 2001 J. Cell. Sci. 114: 1801–1809
Van Roy N, Laureys G, Van Gele M, Opdenakker G, Miura R, van der Drift P, Chan A, Versteeg R, Speleman F . 1997 Eur. J. Cancer 33: 1974–1978
Varesco L, Thomas HJ, Cottrell S, Murday V, Fennell SJ, Williams S, Searle S, Sheer D, Bodmer WF, Frischauf AM, Soloman E . 1989 Proc. Natl. Acad. Sci. USA 86: 10118–10122
Westphal RS, Soderling SH, Alto NM, Langeberg LK, Scott JD . 2000 EMBO J 19: 4589–4600
Xu Y, Einstein JR, Mural RJ, Shah M, Uberbacher EC . 1994 Proc. Int. Conf. Intell. Syst. Mol. Biol. 2: 376–384
Yunis JJ, Ramsay N . 1978 Am. J. Dis. Child 132: 161–163
The authors wish to acknowledge NIH grant NS35791 which supported this work as well as the Roswell Park Cancer Center Support Grant (P30 CA 16056–26). We are grateful to the Children's Oncology Group for providing tumor samples for this study. We would also like to thank Eiko Kitamura, Ph.D. and Karen Head, MS and Lisa Wylie for their technical assistance.
About this article
Cite this article
Sossey-Alaoui, K., Su, G., Malaj, E. et al. WAVE3, an actin-polymerization gene, is truncated and inactivated as a result of a constitutional t(1;13)(q21;q12) chromosome translocation in a patient with ganglioneuroblastoma. Oncogene 21, 5967–5974 (2002). https://doi.org/10.1038/sj.onc.1205734
- actin polymerization
Scientific Reports (2021)
WAVE3 phosphorylation regulates the interplay between PI3K, TGF-β, and EGF signaling pathways in breast cancer
Future Oncology (2019)
The American Journal of Pathology (2019)
The STAT3 inhibitor pyrimethamine displays anti-cancer and immune stimulatory effects in murine models of breast cancer
Cancer Immunology, Immunotherapy (2018)