(Q42;P15) Breakpoint Associated with Wilms' Tumour

The parathyroid hormone-responsive B1 gene is interrupted by a t(1;7)(q42;p15) breakpoint associated with Wilms’ tumour

Ellen G Vernon1, Karim Malik1, Paul Reynolds2, Rachel Powlesland1, Anthony R Dallosso1, Sally Jackson1, Karla Henthorn3, Eric D Green3 and Keith W Brown*,1

1CLIC Research Unit, Department of Pathology and Microbiology, University of Bristol, School of Medical Sciences, University Walk, Bristol BS8 1TD, UK; 2MGH Cancer Centre, Boston, MA, USA; 3Genome Technology Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD 20892, USA

Wilms’ tumour (WT) has a diverse and complex molecular suggests the involvement of other genetic loci in the aetiology, with several different loci identified by cytoge- development of WT (Brown and Malik, 2001). Investi- netic and molecular analyses. One such locus is on gations of inherited predisposition syndromes and loss chromosome 7p, where cytogenetic abnormalities and loss of heterozygosity (LOH) studies in sporadic tumours of heterozygosity (LOH) indicate the presence of a have identified several additional loci, including 11p15 Wilms’ tumour suppressor gene. In order to isolate a (designated WT2) (Henry et al., 1989; Koufos et al., candidate gene for this locus, we have characterized the 1989; Ping et al., 1989; Reeve et al., 1989), 1p (Grundy breakpoint regions at a novel constitutional chromosome et al., 1994), 7p (Wilmore et al., 1994; Miozzo et al., translocation (t(1;7)(q42;p15)), found in a child with WT 1996; Grundy et al., 1998; Powlesland et al., 2000; and skeletal abnormalities. We identified two genes that Perotti et al., 2001), 11q (Radice et al., 1995) and 16q were interrupted by the translocation: the parathyroid (Grundy et al., 1994). hormone-responsive B1 gene (PTH-B1) at 7p and Most of the above changes, apart from those at 11p, obscurin at 1q. With no evidence for LOH at 1q42, we are thought to be involved in the progression of the focused on the characterization of PTH-B1. We detected disease rather than the initiating event in WT develop- novel alternately spliced isoforms of PTH-B1, which were ment. However, chromosome 7p deletions, 7q duplica- expressed in a wide range of adult and foetal tissues. tions, or a combination of these two events have been Importantly, expression of two isoforms were disrupted in observed in a large subset of WTs (15–25%) occurring the WT of the t(1;7) patient. We also identified an as a result of acquired rearrangements, that is, deletions, additional splice isoform expressed only in 7p LOH translocations and long-arm isochromosomes (reviewed tumours. The disruption of PTH-B1 by the t(1;7), in Rivera, 1995). Since some of these occur as the sole together with aberrant splicing in sporadic WTs, suggests cytogenetic abnormality, this implies that chromosome that PTH-B1 is a candidate for the 7p Wilms’ tumour 7 aberrations can occur as a primary event in WT suppressor gene. development. Oncogene (2003) 22, 1371–1380. doi:10.1038/sj.onc. We therefore became interested in one patient (WT21) 1206332 who developed WT, with a nephrogenic rest in the contralateral kidney. This bilateral disease suggested a Keywords: Wilms’ tumour; chromosome 7; PTH-B1 genetic predisposition to WT and importantly, the gene patient did not have a WT1 mutation (Brown et al., 1993; Powlesland et al., 2000), but did have a novel constitutional balanced chromosome translocation; t(1;7)(q42;p15) (Hewitt et al., 1991). In addition to Introduction WT, this patient had congenital skeletal abnormalities similar to the thrombocytopenia and absent radii (TAR) Wilms’ tumour (WT) is the most common paediatric syndrome, a disease not previously associated with WT. malignancy of the kidney. It has a diverse and complex Our past data showed that the WT in this patient molecular aetiology, with several different loci identified (WT21) was monosomic for 7p and trisomic for 7q, by cytogenetic and molecular analyses, although only because of the presence of an isochromosome 7(q) and one gene (WT1) has so far been cloned (Gessler et al., the retention of both translocated chromosomes (Wil- 1990; Brown and Malik, 2001). However, WT1 is only more et al., 1994). We also previously localized the mutated in approximately 20% of all WTs and this breakpoint on 7p between STS markers sWSS355 and sWSS1449 (Reynolds et al., 1996) and showed that 9% *Correspondence: KW Brown; E-mail: [email protected] of sporadic WTs had LOHin a region encompassing Received 15 July 2002; revised 31 October 2002; accepted 12 December this breakpoint (Powlesland et al., 2000). Other groups 2002 have shown similar findings for 7p LOH, with the 1;7 translocation and Wilms tumour EG Vernon et al 1372 minimal regions of LOHoverlapping 7p15 (Miozzo et al., 1996; Grundy et al., 1998). Taken together, these data suggest the presence of a tumour suppressor gene on 7p15. The work presented in this paper describes the cloning and characterization of the constitutional t(1;7)(q42;p15) breakpoints found in patient WT21, and the identification of a gene at 7p not previously associated with WT, which is disrupted in the WT of the translocation patient.

Results

Fine mapping of the t(1;7) breakpoint We previously used FISHmapping to localize the 7p breakpoint of the constitutional t(1;7)(q42;p15) in patient WT21 to within a 440 kb YAC (yWSS1081) on 7p15, in-between STS markers sWSS355 and sWSS1449 (Reynolds et al., 1996). We used this YAC to screen a chromosome 7 cosmid library and isolated 27 positive cosmid clones. These clones were grouped (A–C) relative to an overlapping YAC, yWSS3805, and markers sWSS355 and sWSS1449 (Figure 1a). FISH mapping of representative cosmids from each group allowed their positioning relative to the breakpoint (data not shown). Three cosmids from group A were localized as proximal (1D10), distal (2A05) or spanning (2G01) the breakpoint, thus identifying cosmid 2G01 as containing the 7p breakpoint region. A standard restriction mapping approach was used to identify smaller fragments of cosmid 2G01 that con- tained the breakpoint. These fragments were subcloned, 32P-dCTP labelled and used to probe a Southern blot of WT21 constitutional DNA and normal control DNA in Figure 1 Mapping the translocation breakpoint of WT21. order to (a) verify single copy probes, (b) identify the (a) Horizontal black lines/dashes represent chromosome 7p YACs fragment spanning the breakpoint and (c) exclude yWSS1081 and yWSS3805; vertical lines show positions of the STS rearrangements of 2G01 (Figure 1b). A 0.58 kb SacI/ 7p markers sWSS1449 and sWSS355. The boxes denote three EcoRI (E500) fragment of 2G01 hybridized to abnormal groups of cosmids where numbers in brackets equate to numbers of clones per group (representative cosmids in group A: 1D10, 2G01, junction fragments in WT21 constitutional DNA. 2A05; group B: 2F07 and group C: 2C08). FISHmapping results BLAST analysis with the E500 sequence localized the are shown in italics under each box. proximal: prox, distal: dist, breakpoint within the chromosome 7p BAC contig djs94 spanning: span. The large arrow indicates the 7p15 breakpoint. (B180 kb, Genbank Accession number: AC008080). CEN: centromeric, TEL: telomeric. (b) Southern blot showing normal DNA (N) versus WT21 (21) LCL genomic DNA, where the Database scanning showed no known exons within the arrows denote the abnormal junction fragments. The E500 E500 fragment and exon trapping of 2G01 did not fragment from cosmid 2G01 was 32P-dCTP labelled and used to reveal any robust candidate genes. Rapid amplification probe the blot. Size markers are shown on the right of cDNA ends (RACE) and cDNA library screens using trapped exons were also unsuccessful in identifying candidate genes. breakpoint. Sequencing of both 1 : 7 and 7 : 1 fragments To identify the precise location of the breakpoint on revealed a minor 2 bp ‘GA’ deletion at their junctions both 1q and 7p, a subgenomic library (average insert (Figure 2b: boxed). The breakpoint was precisely size 15–23 kb) was constructed using WT21 constitu- mapped by alignment with the draft human genome tional DNA and probed with E500. Two independent sequence to 7p14.3 (at position 126 575 bp of AC008080) E500-hybridizing clones were selected (Figure 2a: T4-1, and 1q42.13 (at 90 455 bp of AL359510). PCR analysis T1-2) and Southern blotted (again with the E500 probe). of WT21 normal kidney DNA (Figure 2c: first lane) and Fragments of 3 and 4 kb were identified (arrowed in LCL DNA (second lane) confirmed the presence of both Figure 2a), subcloned, sequenced and found to contain the translocated derivative chromosomes (panels i and the breakpoint in the same orientation (Figure 2b: ii), normal copies of chromosome 7 (panel iii) and partial sequence shown). The reciprocal abnormal chromosome 1 (panel iv). The loss of the remaining junction fragment was isolated by PCR using primers normal copy of 7p within the tumour of WT21 was also specific for chromosomes 1 and 7 that flanked the confirmed by PCR analysis (Figure 2c, panel iii, third

Oncogene 1;7 translocation and Wilms tumour EG Vernon et al 1373 further, because our previous studies did not detect any LOHat 1q in the WT of patient WT21 or in other sporadic WTs (Wilmore et al., 1994). Chromosome 7p database searches revealed three ESTs (B22 kb from the breakpoint) derived from independent cDNA libraries, which were found to be homologous to the parathyroid hormone-responsive B1 (PTH-B1) gene (NM_014451). PTH-B1 was originally identified in osteoblastic SaoS-2/B10 cells, where expression was rapidly downregulated (within 1 h) in response to parathyroid hormone (PTH) (Adams et al., 1999). The gene product(s) of PTH-B1 bear no strong homology to any other known proteins or protein domains. These ESTs represent two previously unchar- acterized alternative splice isoforms of PTH-B1 and we have also subsequently cloned three additional isoforms by RT–PCR from foetal kidney RNA, making six possible PTH-BI isoforms in total (Figure 3b and c). In total, there appear to be 24 exons of the PTH-B1 gene, extending over approximately 700 kbp of genomic DNA (Figure 3a). Our mapping of the t(1;7) breakpoint on 7p places it in intron 22 of the gene (Figure 3a), thus interrupting PTH-B1 by translocating exons 1–22 and 23–24 to different derivative chromosomes. Isoforms 2–6 differ relative to the original published PTH-B1 sequence (NM_014451; referred to here as isoform 1) at two positions (1367 and 2456 bp; Figure 4). At 1367 bp (Figure 4a), isoforms 2, 3 and 4 contain two additional exons (13 and 14) that code for 40 amino acids, 13 of which (within exon 13) are capable of being phosphorylated (Figure 4a: bold). In addition, isoform 4 has an unspliced intron at this location (intron 12) that Figure 2 Cloning and confirmation of the breakpoint in WT21. precedes exon 13. This intron codes for four additional (a) Southern blot of two representative phage clones (T4-1/T1-2) amino acids, which are followed by an in-frame isolated from a sub-genomic WT21 LCL library. The E500 2G01 fragment (used as a probe) identified 3 and 4 kb inserts (arrowed), termination codon, presumably leading to the produc- which were found to contain the breakpoint in the same tion of a truncated PTH-B1 protein. Isoform 5 has only orientation; (b) partial sequence of the t(7;1) translocation break- exon 14 inserted at 1367 bp, which gives an in-frame point isolated from the subgenomic library. Boxed ‘GA’ indicates insertion of extra five amino acids. precise position of breakpoint. (c) PCR confirmation of the At 2456 bp (Figure 4b) isoforms 2, 5 and 6 have breakpoint in WT21. All panels contain the same genomic DNA templates, and show PCR products for (i) t(7;1) and (ii) t(1;7) completely different C-terminal ends from isoform 1. derivative chromosomes, (iii) normal copies of chromosome 7 Isoform 2 contains an unspliced intron (intron 21) that (shaded boxes) and (iv) chromosome 1 (white boxes). WT21 is the codes for 10 alternative C-terminal amino acids followed t(1;7) translocation patient, WT77 is a WT patient without 7p by a termination codon (Figure 4b). In isoform 5, exon LOH. NK ¼ normal kidney, LCL ¼ lymphoblastoid cell line, 22 has been replaced with exons 23 and 24, which code WT ¼ Wilms’ tumour for different 38 amino acids at the C-terminus. Sequence analysis shows that exon 23 is part of a mammalian apparent long terminal repeat retrotransposon (MaLR) lane). These results concur with previous data from this that is approximately 342 bp in length (Smit, 1993). group showing 7p LOHin the WT of WT21 (Powles- Isoform 6 has exon 22 replaced by exon 24 alone, et al land ., 2000). leading to the deletion of 10 amino acids from the C- terminus, because the first codon in exon 24 is a stop Two genes are interrupted by the t(1;7) breakpoint codon (Figure 4b). The two unspliced introns (12 and 21) were found in The sequences flanking the breakpoints were analysed large cDNA clones or PCR products in which they were using BLAST and NIX search programmes (http:// flanked by normally spliced exons, demonstrating that www.hgmp.mrc.ac.uk/). Chromosome 1q analysis re- these unspliced introns were not artefacts caused by vealed that the breakpoint bisected the first intron of the genomic DNA contamination. obscurin gene (AJ002535). Obscurin encodes a sarco- Both exons 23 and 24 map to the other side of meric Rho guanine nucleotide exchange factor (GEF) the WT21 breakpoint from exons 1–22. Thus the involved in sarcomere assembly (Young et al., 2001). t(1;7) breakpoint on 7p is predicted to interrupt iso- However, alterations in this this gene was not pursued forms 5 and 6 of PTH-B1. The derivative translocation

Oncogene 1;7 translocation and Wilms tumour EG Vernon et al 1374

Figure 3 Structure of the PTH-B1 gene and alternate splices. (a) Genomic map of chromosome 7p showing the approximate positions of the exons (vertical lines) of the PTH-B1 gene. Black triangles denote 7p STS markers and the arrow represents the breakpoint (TEL: telomeric and CEN: Centromeric). (b) Diagram of the six different PTH-B1 isoforms. Thick vertical lines denote exons present in only some PTH-B1 isoforms, while the ﬁlled blocks between exons (in isoforms 2 and 4) depict unspliced introns. Exons 21 and 22 have been separated to show which isoforms lack these exons. (c) Table showing the presence/absence of additional exons/introns in the PTH-B1 isoforms. ND ¼ not determined (RT–PCR products did not extend to these exons)

chromosomes are not predicted to result in any aberrant Primer pair pFOR5I and 2230109F (in exon 24) were gene fusions, because in the translocation chromosomes used to identify isoforms 5 (with exon 23) and 6 (without the obscurin and PTH-B1 genes are transcribed in exon 23). All these isoforms were expressed in a wide opposite directions. variety of adult and foetal tissues, apart from placenta where isoforms 1–4 were not expressed and adult liver where isoform 5 was not expressed (Figure 5). PTH-B1 isoforms are ubiquitously expressed in many RT–PCR analysis was performed on 13 WTs (seven adult and foetal tissues, apart from WT21 Wilms’ tumour of which had 7p LOH) using primer pairs F3INT (in Some of the PTH-B1 splice isoforms can be distin- exon 20) and NR3 (in exon 22) for isoforms 1–4, and guished by RT–PCR analysis using primers speciﬁc for F3INT and 2230109F (in exon 24) for isoforms 5 and 6 their 30 ends (Figure 5). Primer pair pFOR5I (in exon (Figure 6). Apart from three WTs, all isoforms were 19) and NR3 (in exon 22) were used to differentiate expressed in every case. In WT21 there was no between the isoform containing unspliced intron 21 expression of isoforms 5 and 6 (Figure 6a, lane 2), as (isoform 2) and those that do not (isoforms 1, 3 and 4). predicted from the interruption of the PTH-B1 gene by

Oncogene 1;7 translocation and Wilms tumour EG Vernon et al 1375

Figure 4 Amino-acid comparisons of the alternate PTH-B1 isoforms. (a) Differences between the PTH-B1 isoforms in the exon 12–15 region (sequence shown starts at base 1359 in NM_014451) and (b) differences between the PTH-B1 isoforms in the exon 21–24 region (sequence shown starts at base 2439 in NM_014451). Dashes indicate sequence absent in that isoform. Asterisks indicate sequence present but noncoding (because preceded by stop codon). Residues shown in bold have the potential to be phosphorylated the translocation. In another two 7p LOHWTs, isoform Tumour histology data were available for six of the 2 was not expressed (Figure 6b, bottom panel: WT42 samples shown in Figure 6b. Of these, WT42, 59 and 73 and WT82). However, these 7p LOH WTs did possess were blastemal predominant, whereas WT30, 57 and 82 novel tumour-specific splices of exon 21 (Figure 7a, i). were triphasic. Thus, the two tumours with aberrant These intraexonic splices appear to result in a 48 base splicing (WT42 and 82) had different histology, suggest- (16 amino acids) in-frame deletion within exon 21 ing that this splicing was not merely a reflection of the (Figure 7b). There was no evidence of any genomic pattern of differentiation in the tumours. deletion in this region in WT42 or 82 (Figure 7a, ii). In In total, therefore, three out of nine 7p LOHtumours addition, multiple novel bands were also observed showed changes in PTH-B1 expression. Thus, in in one 7p LOHWT, which hybridized to isoform 5- addition to the PTH-B1 gene being interrupted by the specific probes (indicated by asterisks in Figure 6b, WT21 translocation, two other WTs (42 and 82) also WT42). displayed aberrant splicing of PTH-B1.

Oncogene 1;7 translocation and Wilms tumour EG Vernon et al 1376

Figure 5 Expression patterns of alternative splice isoforms of PTH-B1 in foetal and adult tissues. Expression in various tissues detected by RT–PCR using primers speciﬁc for 30 end of PTH-B1 that distinguish between the alternate splice isoforms (depicted diagrammatically on the left). Top panel: isoforms 5 and 6; middle panel: isoforms 1 to 4; bottom panel: HPRT control. In each PTH- B1 panel, ethidium bromide stained gels are shown above Southern blots of these gels probed with 32P-dCTP-labelled PTH-B1 (isoform 1) or 2230109EST cDNA. Sizes of PCR products are shown in brackets

Discussion and foetal tissues. We have also demonstrated that both isoforms 5 and 6 are disrupted by the t(1;7) constitu- The work in this paper set out to characterize the areas tional breakpoint of WT21, and therefore not expressed surrounding the translocation breakpoints found in WT in the WT of WT21. Although these isoforms are patient WT21, in an attempt to identify novel genes on expressed in other 7p LOHWTs, additional tumour- 7p involved in WT. This region had been implicated in specific isoforms appear to be present in two other WTs WT development by our previous studies (Wilmore (WT42 and 82). In this subset of WT patients, we et al., 1994; Reynolds et al., 1996; Powlesland et al., showed mutually exclusive expression of isoform 2 and a 2000) as well as the work of other groups (Miozzo et al., novel isoform with an intraexonic splice in exon 21. The 1996; Grundy et al., 1998). splicing alterations were tumour-specific (i.e. not germ Cloning and sequencing of the WT21 breakpoint line) and neither patient had any congenital abnormal- allowed exact positioning of the t(1;7) breakpoints of ities, unlike patient WT21 in which PTH-B1 was WT21 to 1q42.13 between markers SHGC-76460 and interrupted by a germline chromosomal defect. This is SHGC-146605, and to 7p14.3 between STS-T78597 and the first report of the expression patterns of new SHGC-172305. Two genes are interrupted by the isoforms of the PTH-B1 gene and their aberrant breakpoints: the obscurin gene at 1q42 and the expression in cancer. parathyroid hormone-responsive B1 gene (PTH-B1)at While the function of PTH-B1 gene has yet to be 7p14. elucidated, it is thought to encode for a globular protein Obscurin was originally identified in a yeast 2-hybrid that may be involved in intracellular signalling (Adams screen as an interacting partner of the scaffolding et al., 1999). Our amino-acid comparisons of the PTH- protein titin (Young et al., 2001). However, obscurin is B1 alternative transcripts detailed in this paper suggest unlikely to be a candidate for a WT suppressor gene that each protein isoform could have distinct cellular because (a) we did not detect any LOHof 1q in our WTs roles, possibly within different sub-cellular compart- and (b) the WT of WT21 retains both copies of 1q ments. The splicing defects that we have demonstrated (Wilmore et al., 1994). It is possible that the disruption in two 7p LOHWTs could therefore affect certain of one copy of obscurin may contribute to the TAR-like specific cellular functions of PTH-B1. Changes in symptoms observed in this patient. Further work is relative levels of alternative spliced isoforms that affect required to elucidate the role of obscurin in this patient tumorigenesis are not uncommon (Nissim-Rafinia and and in others with similar skeletal abnormalities. Kerem, 2002) and include changes in the relative levels Our efforts to identify other potential candidates of WT1 isoforms in WT (Baudry et al., 2000). involved in WT genesis were therefore focused on a Interestingly, WT patients with associated hypercal- cluster of chromosome 7 ESTs that locate near the caemia have elevated PTHlevels, and removal of the WT21 7p breakpoint. These were found to possess a tumour resets normal cellular levels (de Kraker and 99% sequence identity to the PTH-B1 gene (Adams Voute, 1979). We have shown that PTH-B1 is ubiqui- et al., 1999). Our analyses show that the PTH-B1 gene is tously expressed and it is known that PTHcan rapidly multiply spliced in kidney as well as in many other adult decrease the expression of PTH-B1 (Adams et al., 1999).

Oncogene 1;7 translocation and Wilms tumour EG Vernon et al 1377

Figure 6 Expression of alternative splice isoforms of PTH-B1 in WTs. Expression in WTs detected by RT–PCR using primers speciﬁc for 30 end of PTH-B1, as described in Figure 5. (a) Comparison of WT21 WT (cultured cells) and other cultured WTs. (b) RT–PCR on WT samples (from frozen tissue). Asterisks denote novel transcripts that hybridize to 2230109EST cDNA probes. WTs showing 7p LOHare in bold. Sizes of PCR products are shown in brackets

Thus high levels of PTHassociated with hypercalcaemia the developing kidney of patient WT21 could then have in WT patients could affect the expression of PTH-B1 resulted in WT. A precedent for this is seen with WAGR and might contribute to the tumour phenotype in some patients, where the loss of one functional copy of WT1 cases. results in the development of a particular phenotype (i.e. One could speculate that the constitutional disruption genitourinary abnormalities), while loss of the remain- of isoforms 5 and 6 in patient WT21 may be involved in ing copy of WT1 results in WT. However, it is important their skeletal abnormalities, because PTH-B1 is thought to note that most, but not all, WTs of WAGR patients to be important in PTHeffects on osteoblasts (Adams show somatic mutation in the remaining WT1 copy, et al., 1999). The loss of the second copy of PTH-B1 in suggesting the involvement of other genes yet to be

Oncogene 1;7 translocation and Wilms tumour EG Vernon et al 1378

Figure 7 Exon 21 tumour-specific alternative splice in 7p LOHWTs. ( a) (i) RT–PCR showing expression of PTH-B1 isoform 6, or the smaller tumour-specific product, in foetal kidney (lanes 1 and 2), WT42 WT (lanes 2 and 3) and WT82 (lanes 4–5). Primers used in RT- PCRs were specific for exon 21 (F3INT/pREV6I). Gels show representative cloned RT-PCR products. (ii) Genomic PCR using F3INT/I22REV primer pair to detect deletions of exon 21 or flanking sequences. (b) Sequence of the novel 48 base deletion in isoform 6. Bases between, and including, one of the boxed regions (5 bp repeat) are eliminated in this RNA

identified in WT development (Little and Wells, 1997). agarose gel and transferred onto Hybond-N+. This filter was At least two other loci on 7p (7p13–14 and 7p22) prehybridized (7% SDS, 5 mg/ml Bovine Serum Albumin m m identified by different groups appear to be the target for (Sigma), 0.5 NaOPO4 pH7.2, 1 m EDTA) for 3 h at 651C, chromosomal rearrangements or deletions in WTs then hybridized with the relevant probe for 18 h at 651C (Sawyer et al., 1993; Rivera, 1995; Grundy et al., 1998; (Figure 1). Probes were generated by digesting 2G01 cosmid Lobbert et al., 1998; Perotti et al., 2001). It is possible, DNA with various restriction enzymes (SacI, HindIII, XbaI, EcoRI). These fragments were separated on a 0.6% gel, some therefore, that there are multiple genes on 7p, any one of of which were gel-extracted, 32P-dCTP labelled and purified which can predispose to WT development. Alterna- (Amersham Pharmacia Biotech’s Multiprime labelling DNA tively, haploinsufficiency of several genes on 7p might kit and Nickt columns, respectively). These fragments were collectively contribute to WT development. Finding the also subcloned into puc18. critical gene(s) may therefore necessitate the use of complementary methods, such as the use of animal Chromosome 1:7 breakpoint genomic cloning models (Balmain, 2002). WT21 LCL genomic DNA (100 mg) was partially digested with 2.4 U Sau3A (New England Biolabs) and subsequently size fractionated on a 12 ml 10–40% sucrose density gradient Materials and methods (Wilson and Goulding, 1986) for 21 h at 78 900 g. Fractions containing WT21 LCL DNA fragments within a 15–23 kb Tissue samples range were pooled, dialysed, precipitated (Sambrook et al., Fresh normal kidney and WT tissues were snap frozen in 1989) and resuspended in 10 ml distilled H2O. liquid nitrogen and then stored at À701C. DNA and RNA Size-fractionated inserts were ligated and packaged using were extracted from frozen tissues using standard methods. Stratagene’s Lambda DASHII/BamHI vector system, accord- One patient (WT21) had a constitutional t(1;7) with skeletal ing to the manufacturer’s instructions. The resulting library abnormalities (Hewitt et al., 1991). All other cases were was titered and plated to a density of 50 000 PFU (plaque 3 5 sporadic with no obvious genetic predisposition or congenital forming units)/15 cm agar plate (4 Â 10 PFU total). Plaque defects. The 7p LOHstatus of these samples has been lifts were performed on these plates using Schleicher & Schuell determined previously (Powlesland et al., 2000). Optitran (0.45 mm) membranes as previously described (Ben- ton and Davis, 1977). Standard conditions were used (see above) to pre-hybridize/hybridize the filters and a 32P-dCTP- Cell culture labelled E500 DNA fragment (SacI/EcoRI fragment from WT21 lymphoblastoid cell line (LCL) and cultured WT cells 2G01) used as a probe. From these filters eight individual were maintained as previously described (Brown et al., 1989; clones were isolated. In each case, the phage was eluted from Wilmore et al., 1994; Reynolds et al., 1996). Characterization the agar plugs (Sambrook et al., 1989) and taken through of cultured WT cells by LOHand cytogenetic analysis have secondary and tertiary screens. shown that these cells are tumour-derived (Brown et al., 1989; Phage DNA minipreps were performed on individual clones Wilmore et al., 1994). (Sambrook et al., 1989) and the resulting plasmid DNA was digested using a number of enzymes. Fragments were separated on a 0.8% agarose gel and Southern blotted using Southern blotting and identification of a 0.5 kb 2G01 breakpoint E500 as a probe. Two positive SacI/XbaI fragments (approx. 4 fragment and 3 kb, respectively. Figure 2a) were identified, cloned into Genomic DNA was extracted from normal kidney tissue and puc18 and positive clones sequenced (Figure 2b. Durham WT21 LCL cells (as previously described in Sambrook et al., University sequencing service). Reciprocal inserts were 1989) and 5 mg was digested with KpnI, PvuII or SacI (24 h, obtained by PCR using 100 ng WT21 genomic DNA, 1 mm 371C). The resulting fragments were separated on a 0.6% of primers (unless otherwise stated) BP1FA (chromosome 1

Oncogene 1;7 translocation and Wilms tumour EG Vernon et al 1379 forward primer: 50-CTGCAGTGAGCTGTGAGTATG- parameters following cDNA synthesis were as follows: 941C 30, positioned at 90 770 bp within AL359510) and E500- for 5 min; 35 cycles of 941C for 30 s, 591C for 30 s, 721C for S1 (chromosome 7 forward primer: 50-GAGGTCTGGT- 1 min and 1 cycle of 721C for 5 min (reaction contents CACTTGGAAAAT-30, positioned at 126 747 bp within described above). Approximately 10 ml of each RT–PCR was AC008080), 0.2 mm dNTPs (Amersham Pharmacia Biotech), run on a 2% agarose gel, transferred onto Hybond N+ 1 Â Supertaq buffer and 0.1 U Supertaq enzyme (HT Bio- membrane (Amersham Pharmacia Biotech) and hybridized technologies). PCR cycling parameters were as follows: 941C according to standard Southern blotting protocols (Sambrook for 5 min; 35 cycles of 941C for 30 s, 581C for 30 s, 721C for et al., 1989). Selected PCR products were also cloned into 1 min and 721C for 5 min. In addition, PCRs were also carried pGEM-T easy vector system according to the manufacturer’s out to check WT21 LCL for normal copies of chromosome 1 instructions (Promega). Primer pair F3INT and I22REV (in (BP1FA and BP1RA: 50-GCTCGGCCTGTTTCTTCATC-30), intron 21) (50-CAAATTCAGCTGCACCAGTGC-30) were chromosome 7 (E500-S1 and S13-1A: 50-AAGGCATG- also used to amplify exon 21/flanking genomic DNA for CAAAATGGAACTGC-30), t(7;1) (S13-1A and BP1RA) and sequencing (Figure 7b). the t(1;7) fragment (BP1FA and E500-S1). Generation of probes for RT–PCR blots RNA extraction and RT–PCR The PTH-B1 coding sequence (76–2530 bp of NM_014451) Total RNA was obtained using the TRI REAGENTt was amplified by PCR from foetal kidney cDNA using protocol (Sigma), and reverse transcription was performed Expandt Long template PCR system (Roche) according to using the Moloney murine leukaemia virus reverse transcrip- the manufacturer’s instructions. Primer pair NF1 (in exon 2) tase, RNase Hminus point mutant (M-MLV RT [H À]: (50-GCTTTATGGGATACCTAAGG-30) and NR3 primers Promega). Each RT reaction was set up in a total of 25 ml were used. PCR cycling parameters were as follows: six cycles according to the manufacturer’s instructions using 0.5 mgof of 941C for 15 s, 611C (decreasing at 11C per cycle) for 30 s, total RNA, 0.5 ml of ribonuclease inhibitor (30 000–50 000 U/ 721C for 2 min; 10 cycles of 941C for 15 s, 551C for 30 s, 721C ml: Sigma) and 500 ng of Oligo(dT)15 primers (Promega). The for 2 min; 20 cycles of 941C for 15 s, 551C for 30 s, 721C for RT was performed at 421C for 1 h and 1 ml of the cDNA 2 min (increasing 20 s per cycle); one cycle of 721C for 5 min. produced (1/25th) was sufficient for one PCR reaction. PCR products were gel-purified and cloned into PGEM-T easy The primers used for the subsequent PCRs were as follows: vector system according to the manufacturer’s instructions (Promega). Fragments were also 32P-dCTP labelled and used to 0 F3INT (in exon 20): 5 -GATTATCCAAAGGTGGCC- hybridize to PTH-B1 and isoforms 1 and 2 filters as previously 0 GTC-3 described above. IMAGE clone 2230109 (AI624582) was 0 2230109F (in exon 24): 5 -GCATCTTGATAAATGT- digested with NotI/EcoRI and the resulting 1.8 kb band was 0 CAGCTTG-3 resolved on a 0.8% agarose gel, extracted and 32P-dCTP 0 0 NR3 (in exon 22): 5 -GAGATGGGGATGTTCCTCTC-3 labelled as described above. This probe was used to hybridize 0 pFOR5I (in exon 19): 5 -GCTTAGTGCTGACCAG- to isoforms 5 and 6 RT–PCR filters (as described above). GTTG-30 pREV6I (in exon 21): 50-GATGTCTGTGGTTGGTAAA- CAG-30 50-HPRT: 50-CTTGCTGGTGAAAAGGACCCCACGA-30 Acknowledgements 30-HPRT: 50-GTCAAGGGCATATCCTACAACAAAC-30 This work was supported by funds from the Association for International Cancer Research (AICR), Cancer and Leukae- All adult tissue cDNAs used in Figure 5 were obtained from mia in Childhood (CLIC) and the National Kidney Research Origene technologies. Unless otherwise stated PCR cycling Fund (NKRF).

References

Adams AE, Rosenblatt M and Suva LJ. (1999) Bone, 24, 305– Grundy RG, Pritchard J, Scambler P and Cowell JK. (1998). 313. Oncogene, 17, 395–400. Balmain A. (2002). Nature, 417, 235–237. Henry I, Grandjouan S, Couillin P, Barichard F, Huerre Baudry D, Hamelin M, Cabanis MO, Fournet JC, Tournade Jeanpierre C, Glaser T, Philip T, Lenoir G, Chaussain JL MF, Sarnacki S, Junien C and Jeanpierre C. (2000). Clin. and Junien C. (1989). Proc. Natl. Acad. Sci. USA, 86, 3247– Cancer Res., 6, 3957–3965. 3251. Benton WD and Davis RW. (1977). Science, 196, 180–182. Hewitt M, Lunt PW and Oakhill A. (1991). J. Med. Genet., 28, Brown KW and Malik KTA. (2001). Exp. Rev. Mol. Med., 14, 411–412. http://www-ermm.cbcu.cam.ac.uk/01003027h.htm. Koufos A, Grundy P, Morgan K, Aleck KA, Hadro T, Brown KW, Shaw AP, Poirier V, Tyler SJ, Berry PJ, Mott MG Lampkin BC, Kalbakji A and Cavenee WK. (1989). Am. J. and Maitland NJ. (1989). Br. J. Cancer, 60, 25–29. Hum. Genet., 44, 711–719. Brown KW, Wilmore HP, Watson JE, Mott MG, Berry PJ Little M and Wells C. (1997). Hum. Mutat., 9, 209–225. and Maitland NJ. (1993). Gene Chromosome Cancer, 8, Lobbert RW, Klemm G, Gruttner H-P, Harms D, Winter- 74–79. pacht A and Zabel BU. (1998). Gene Chromosome Cancer, de Kraker J and Voute PA. (1979). Helv. Paediatr. Acta, 34, 21, 347–350. 459–460. Miozzo M, Perotti D, Minoletti F, Mondini P, Pilotti S, Gessler M, Poustka A, Cavenee W, Neve RL, Orkin SHand Luksch R, Fossatibellani F, Pierotti MA, Sozzi G and Bruns GA. (1990). Nature, 343, 774–778. Radice P. (1996). Genomics, 37, 310–315. Grundy PE, Telzerow PE, Breslow N, Moksness J, Huff V and Nissim-Raﬁnia M and Kerem B. (2002). Trends Genet., 18, Paterson MC. (1994). Cancer Res., 54, 2331–2333. 123–127.

Oncogene 1;7 translocation and Wilms tumour EG Vernon et al 1380 Perotti D, Testi MA, Mondini P, Pilotti S, Green ED, Pession Rivera H. (1995). Cancer Genet. Cytogenet., 81, 97–98. A, Sozzi G, Pierotti MA, Fossati-Bellani F and Radice P. Sambrook J, Fritsch EF and Maniatis T. (1989). Molecular (2001). Gene Chromosome Cancer, 31, 42–47. Cloning. A Laboratory Manual. Cold Spring Harbor Ping AJ, Reeve AE, Law DJ, Young MR, Boehnke M and Laboratory Press: New York. Feinberg AP. (1989). Am. J. Hum. Genet., 44, 720–723. Sawyer JR, Winkel EW, Redman JF and Roloson GJ. (1993). Powlesland RM, Charles AK, Malik KTA, Reynolds PA, Cancer Genet. Cytogenet., 69, 57–59. Pires S, Boavida M and Brown KW. (2000). Br. J. Cancer, Smit AF. (1993). Nucleic Acids Res., 21, 1863–1872. 82, 323–329. Wilmore HP, White GFJ, Howell RT and Brown KW. (1994). Radice P, Perotti D, Debenedetti V, Mondini P, Radice MT, Cancer Genet. Cytogenet., 77, 93–98. Pilotti S, Luksch R, Bellani FF and Pierotti MA. (1995). Wilson K and Goulding KH. (1986). A Biologist’s Guide to Genomics, 27, 497–501. Principles and Techniques of Practical Biochemistry. Edward Reeve AE, Sih SA, Raizis AM and Feinberg AP. (1989). Mol. Arnold: London. Cell. Biol., 9, 1799–1803. Young P, Ehler E and Gautel M. (2001). J. Cell Biol., 154, Reynolds PA, Powlesland RM, Keen TJ, Inglehearn CF, 123–136. Cunningham AF, Green ED and Brown KW. (1996). Gene Chromosome Cancer, 17, 151–155.

Oncogene