(1998) 12, 1099–1106  1998 Stockton Press All rights reserved 0887-6924/98 $12.00 http://www.stockton-press.co.uk/leu Abnormalities of the ETV6 occur in the majority of patients with aberrations of the short arm of 12: a combined PCR and Southern blotting analysis HE O’Connor1, TA Butler1, R Clark1, S Swanton1, CJ Harrison1, LM Secker-Walker1 and L Foroni2

1Department of and Cytogenetics and 2Molecular Genetics Laboratories, The Royal Free Hospital and School of Medicine, Pond Street, London NW3 2QG, UK

Involvement of the ETV6 gene, located at 12p13, has been myelomonocytic leukemia (CMML).1 It now appears that investigated in 20 patients with an abnormality of the short arm ETV6 is a ‘promiscuous’ gene disrupted by fusion with one of of (abn 12p) detected cytogenetically. Patients in the study had c/pre-B acute lymphoblastic leukemia (ALL) many partner . Fusion of the ETV6 gene to the MN1 gene (nine children and three adults), T-ALL (three adults), acute located on chromosome 22q11 has been shown in myelo- myeloid leukemia (AML) (two adults), biphenotypic acute leuke- pproliferative disorders (MPD).4 The ETV6 gene has also been mia (Bip-L) (one adult), myelodysplasia (MDS) (one adult) and shown to fuse with the ABL gene (on 9q34) in a patient with chronic myelomonocytic leukemia (CMML) (one child). Abnor- t(9;12;14) and acute undifferentiated leukemia.3 Recently, malities of 12p comprised deleted (del)(12p) alone (seven translocations involving ETV6 have been reported in patients cases), add(12p) alone (seven cases), del(12p) and add(12p) 15 (one case) and balanced translocations of 12p to 1p13, 1q31, with and a t(3;12)(q26;p13). Pre- 10q11, 14q11 and 15q15 (one case of each). A novel, exon- liminary reports of fusion products cloned by rapid amplifi- specific RT-PCR assay identified breakpoints in ETV6 in nine cation of cDNA ends (RACE-PCR) have described fusions in of 19 cases, and showed breakpoints in intron 5 (seven cases which exons 1, or exons 1 and 2 of ETV6, were fused to the of children with c-ALL), in intron 4 (in one adult with Bip-L) and EVI1 gene on chromosome 3. Fusion of ETV6 and AML1 is in intron 2 (in one adult with AML). RT-PCR for the ETV6/AMLI brought about by t(12;21),2,11 a translocation that cannot be fusion (tested in 19 cases) was positive using standard primers in five cases (four of which had shown rearrangements in detected by classic cytogenetic analysis. This genetic intron 5) and occurred as a variant fusion in a sixth case (also rearrangement has been described in B-lineage acute lym- positive for a rearrangement in intron 5) using 3′ RACE PCR. phoblastic leukemia (ALL) with an incidence of 16–30% in Southern blotting confirmed rearrangements in intron 5 in the childhood2,10,16,17 and of 3% in adults with ALL.18,19 Patients five cases available for analysis and revealed a rearrangement with this translocation have been shown to have a good in intron 5 in one of 10 cases with no evidence of intron 5 prognosis.16,20–22 involvement by RT-PCR. Rearrangements in intron 5 of ETV6 were found in eight of nine cases of children with c-ALL of Twenty patients with cytogenetically detectable 12p abnor- which six carried the ETV6/AMLI fusion. Heterozygosity within malities were analyzed for rearrangements/deletions of the intron 5 (revealed by the genomic probe B1) was found in seven ETV6 gene. We have developed a novel one-step exon- of 11 cases tested. Deletion of one allele was indicated in three specific RT-PCR assay to investigate ETV6 rearrangements and cases with del(12p) and one case with add(12p). This study, to identify the exons adjoining the breakpoint. using a combination of ETV6 exon-specific RT-PCR, RT-PCR for ETV6/AMLI and Southern blotting has shown that rearrange- ment and/or deletion of ETV6 may occur in up to 70% of Materials and methods patients with abn 12p. Furthermore, 90% of children in this study with an abn 12p and c-ALL, carried a rearrangement of ETV6 in intron 5. Patient data Keywords: ETV6 gene; 12p abnormality; leukemia; ETV6/AMLI Patients in the study had c/pre-B ALL (nine children and three adults), T-ALL (three adults), (AML) Introduction (two adults), biphenotypic acute leukemia (Bip-L) (one adult), myelodysplasia (MDS) (one adult) and chronic myelomono- As many as 10% of all have abnormalities of the cytic leukemia (CMML) (one child). Children were aged short arm of chromosome 12 (abn 12p). The 12p aberrations between 3 and 11 years and adults were aged between 17 include translocations, deletions, [del(12p)], inversions, and 53 years. The age, sex, diagnosis, karyotype and fluor- dicentric and derived, [der(12p)]. . Reciprocal escence in situ hybridization (FISH) results for each patient translocations between 12p and a number of other chromo- are shown in Table 1. somal bands have been identified by karyotypic analysis. Part- ner chromosome bands include: 1p13, 2p11, 2q11, 3q26, Cytogenetic and FISH analysis 5q33, 7q11, 9p11, 11q23, 13q14, 14q11, 16q13, 17p11, 1–13 17q21, 20q11 and 22q12. Cytogenetic analysis had been carried out following 24-h cul- To date, the only gene shown to be involved in 12p 1,2 ture, using standard techniques. Karyotypes were described rearrangements is the ETV6 gene located at band 12p13. according to ISCN.23 In 10 cases, FISH using chromosome- This gene is a member of the ETS gene family of DNA binding 14 specific, whole chromosome paints (wcp) was applied to proteins that recognise the core motifs C/AGGAA/T. ETV6 clarify the 12p abnormality. was first identified as the fusion partner of the PDGFR␤ gene in the translocation t(5;12)(q33;p12) in a patient with chronic Nucleic acid extraction

Correspondence: L Foroni; Fax: 171 794 0645 DNA was extracted from (unseparated or mono- Received 19 January 1998; accepted 26 March 1998 nuclear cell fractions) using a method previously described.24 Analysis of the abnormalities of the ETV6 gene HE O’Connor et al 1100 Table 1 RT-PCR analysis of patients with 12p abnormalities

Case Sex/Age Diagnosis Cytogenetic analysis FISH wcp Exon ETV6/ SB ETV6 SB for ETV6 ETV6c no. (Years) abn12p [no. cells]/ specific AMLI intron 5 B1 [no. normal cells;a RT-PCR RT-PCR poly- morphism

1 F/9 c-ALL Rel del(12)(p11)[18]/[2] t(12;21) Exon 5-6 + NA NA ?G/R 2 F/8 c-ALL Rel del(12)(p13)[17]/[1] NA Exon 5-6 + G/R 6.5kb+6.5kb ?G/R 3 F/4 c-ALL add(12)(p13)[20]/[10] ins(12;1)(p13;?) Exon 5-6 + G↓/R 6.5kb+6.5kb D/R t(12;21) 4 M/4 c-ALL add(12)(p13)[6]/[5] der(12)t(11;12) Exon 5-6 + NA NA ?G/R der(12)t(12;21) 5 F/3 c-ALL add(12)(p13)[7]/[11] NA normal + NA NA ?G/R 6 F/3 c-ALL add(12)(p13) t(3;12) Exon 5-6 +b G↓/R NA D/R del(12)(p13)[11] der(?21)t(12;21) 7 M/11 c-ALL Rel t(1;12)(p13;p13)[18]/[2] NA Exon 5-6 − G/R 6.5kb+6.5kb ?G/R 8 M/5 c-ALL t(12;14)(p11;q11)[4]/[5] t(12;14) Exon 5-6 − G/R NA ?G/R der(21)t(12;21) 9 M/3 c-ALL del(12)(p12)[2]/[4] not t(12;21) normal − G/G 5.5 kb+6.5kb G/G 10 F/38 c-ALL Rel t(10;12)(q11;p12)[14]/[3] t(10;12) normal − G/G NA G/?G 11 M/27 c-ALL del(12)(p11)[15]/[1] NA normal − G/R↓ 5.5kb↓+6.5kb D/R 12 M/25 pre-B ALL add(12)(p13)[14]/[6] dup(12) normal − G/G 5.5kb+6.5kb G/G 13 F/22 T-ALL Rel del(12)(p11)[3]/[30] NA normal − NA NA ?G/?G 14 M/38 T-ALL add(12)(p13)[8]/[3] NA normal − G/G 5.5kb+6.5kb G/G 15 M/17 T-ALL del(12)(p12)[3]/[17] NA normal − G/G 5.5kb↓+6.5kb G/D 16 M/53 MDS add(12)(p13) NA normal − G/G 5.5kb↓+6.5kb G/D 17 M/9 CMML del(12)(p11)[10] NA NA NA G/G 5.5kb↓+6.5kb G/D 18 M/16 AML add(12)(p13)[7]/[1] NA normal − G/G NA G/?G 19 M/51 AML t(1;12)(q31;p13)[11] NA Exon 2-3 − G/G NA G/R 20 M/31 Bip-L t(12;15)(p12;q15)[8]/[1] NA Exon 4-5 − G/G 6.5kb+6.5kb ?G/R

aIncludes all cells with normal 12p; bETV6/AML1 with alternative splicing; cRT-PCR and Southern blotting; ↓, reduced intensity; D, deleted; G, germ-line; ?G, unable to test for hemizygosity (deletion); R, rearranged; , relapse.

RNA was isolated after guanidine thiocyanate disruption of size. As a result, given the small size of exons 1, 2 and 3, the cells, followed by caesium chloride gradient separation25 rearrangements in introns 1, 2 and 3 were identified by ampli- or acid phenol extraction.26 fication between exons 1 and 2, 1 and 3, and 1 and 4, respect- ively. Rearrangements in introns 6 and 7 were identified by amplification between exons 6 and 8. Amplification between RT-PCR analysis for ETV6 rearrangement exons 6 and 8 would not distinguish between rearrangements in introns 6 and 7. Consequently, a further primer to exon 7 could be used to identify the breakpoint in intron 6 or 7. Accurate assessment of a reduction in band intensity Development of exon-specific RT-PCR analysis brought about by a translocation was achieved following a series of tests. Each primer pair was first tested with 1, 2.5 and Bone marrow cDNA was synthesized from 2 ␮g of total RNA 5 ␮l cDNA from normal bone marrow using between 25 and as previously described.25 One twentieth of the reaction (5 ␮l 35 cycles. For each primer pair semi-quantitative results were cDNA) was submitted to PCR amplification in 50 ␮l with obtained using 30 cycles (data not shown). Thirty cycles were 250 ng primers, 0.2 mM dNTPs and 2.5 U of Taq (Life Techno- sufficient for all patient samples except cases 2 and 19, which logies, Inchinnan, UK) in the buffer recommended by the sup- required 28 and 32 cycles, respectively. All reactions were plier. To prevent false negative results all cDNA preparations carried out using a denaturation step at 94°C for 5 min fol- were verified by amplification of G6PD as previously lowed by 30 cycles of denaturation at 94°C for 1 min, described.25 annealing at 64°C for 1 min and extension at 72°C for 1 min Forward and reverse primers for exons 1 to 8 of the ETV6 with a final extension at 72°C for 10 min. PCR products were gene were used in six simultaneous RT-PCR reactions to amp- electrophoresed through a 1.5% agarose gel, containing lify across the exons of the ETV6 cDNA as illustrated in 1 ␮l/ml ethidium bromide and photographed under UV light. Figure 1a. Figure 1b shows the nucleotide sequence of the All tests were performed at least three times for each patient primers used. to ensure consistent and reproducible results. A break in any intron of ETV6 as a result of a translocation will interrupt the continuity of the exon organisation. This will RT-PCR analysis for ETV6/AML1 fusion result in a reduction/failure in amplification of products when using primers for exons flanking the breakpoint region. How- RNA prepared from bone marrow was investigated for the ever, exons upstream of the breakpoint itself will still amplify ETV6/AML1 fusion by RT-PCR. The primers used were: ETF5; as the ETV6 remains functional.1,11 Exons down- CTC CGA GGA CGG GCT GCA TAG GGA AGG (bp 900– stream of the breakpoint will amplify if the reciprocal product 927) specific to ETV6 exon 5 and AML3; AAG AGC TCG CTC is expressed. Primers were designed to have comparable ATC TTG CCT GG (bp 833–855) for exon 2 of the AMLI annealing temperatures and to amplify products of similar gene.17 Analysis of the abnormalities of the ETV6 gene HE O’Connor et al 1101

Figure 1 ETV6 genomic organisation and RT-PCR strategy. (a) Exon-specific RT-PCR analysis to identify rearrangements in the ETV6 gene. The forward (F) and backward (B) primers are indicated as arrows. The breakpoints of identified fusions are shown in bold. (b) Nucleotide sequence of the primers used. Restriction sites artificially introduced into the primers sequence for cloning purposes (underlined and italics) or as part of the ETV6 gene sequence (bold) are indicated.

3′ RACE PCR and two genomic probes for intron 5 were used, as described below. This procedure was carried out as described by the 3′ RACE PCR protocol supplied with the kit (Life Technologies) with the following modifications: the primer used to generate the Probes for Southern blotting cDNA contained random hexamers at the 3′ end as opposed to dT (5′ GGC CAC GCG TCG ACT AGT ACT nnn nnn 3′). RT-PCR amplification from normal BM cDNA was used to In the first round PCR, only the gene-specific primer (GSP) generate the 5′ and 3′ ETV6 cDNA probes. Amplification was 5Fb (5′ GGG AGA CTA GGA CTT GGG GGC CGT GAG G carried out using primers 1F and 8B (Figure 1b) with denatur- 3′) were used with the following PCR conditions: 30 cycles of ation at 94°C for 1 min, 66°C annealing for 1 min and exten- 1 min denaturation at 94°C, 1 min annealing at 64°C and 2.5 sion at 72°C for 2 min. A 1.3 kb band was cloned and min extension at 72°C. This was completed by a final 10 min sequenced. The 5′ subclone containing exons 2–5 (bp 136 to extension at 72°C. In the second round PCR both the internal bp 675 according to the published sequence1 was generated GSP 5Fa: (5′ TAG GAA TCC GAG GCT CCT GCC CGA CGT by digestion with XhoI and SacI. The 3′ subclone containing ATC CCT TCC CTT C 3′) and the RACE PCR primer (5′ GGC exons 5–8 (bp 672 to bp 1383) was generated by digestion CAC GCG TCG ACT AGT AC 3′) were used with the following with the same enzymes (XhoI artificially introduced at the 3′ PCR conditions: an initial denaturation of 5 min at 94°C fol- end). The cDNA probes would only identify rearrangements lowed by 30 cycles of, 1 min denaturation at 94°C, 1 min which occurred close to exon boundaries, ie on the same annealing at 64°C and 2 min extension at 72°C. This was restricted fragment as the exon. Two genomic probes (B1 and completed by a final 10 min extension at 72°C. The PCR pro- D) which map to intron 5 of ETV6 were used. These were a ducts were cleaned and digested with the restriction enzymes kind gift from Dr R Berger, Paris, and have been previously EcoRI and SalI. The digested products were electrophoresed described.17 The genomic probe, B1 was used with EcoRI on a 1.5% agarose gel. PCR products between 400–600 bp digests to detect heterozygosity for an EcoRI polymorphism were isolated, cloned into the plasmid pBSK+ and sequenced. and hence to detect ETV6 deletions.17

Southern blot analysis of ETV6 Results

DNA samples (10 ␮g) were digested to completion with the Cytogenetic analysis enzymes EcoRI, HindIII, BamHI or SacI. The restricted DNA was size fractionated on a 0.8% agarose gel and visualized Abnormalities of 12p detected by G-banding and by FISH by ethidium bromide staining. Following transfer to Hybond- analysis, when available are shown in Table 1. From cyto- N membranes (Amersham, Buckinghamshire, UK) using stan- genetic analysis, del(12p) was seen as the sole abnormality in dard Southern blotting techniques,27 filters were hybridized at seven cases and accompanied by an abnormality of the 65°C using a dextran sulphate-based hybridization buffer.24 second 12p in one case. Five cases had a balanced translo- Probes were labelled to high specific activity by random hex- cation of 12p with partner sites at 1p13, 14p11, 10q11, 1q31 anucleotide priming (Amersham) with ␣-32P-dCTP. Filters and 15q15. The remaining seven cases had an add(12p) were washed at 65°C for 20 min in 0.3 × SSC, 0.1% SDS with indicative of an unbalanced translocation of 12p which could one change of buffer.28 Southern analysis was used, where not be fully described. FISH, using whole chromosome paints possible, to confirm the data from the exon-specific RT-PCR (wcp) to 12 and 21, was applied in five cases and showed assay and to detect heterozygosity to a polymorphic site in fusion between chromosomes 12 and 21 in four cases; a bal- ETV6 and hence allelic deletion. For this reason two cDNA anced translocation t(12;21) was shown in cases 2 and 3 and Analysis of the abnormalities of the ETV6 gene HE O’Connor et al 1102 an unbalanced translocation, der(12)t(12;21), in case 4. In (Table 1). A normal pattern was detected in the remaining 10 case 6, due to limited material, FISH with wcp 12 only was patients (cases 5, 9–16, 18). applied and the result suggested der(21)t(12;21). The use of whole chromosome paints identified the der(12p) partner in three cases as chromosome 1 (case 3), chromosome 3 (case 6), and chromosome 11 (case 4) and confirmed the partner Investigation of the t(12;21) ETV6/AML1 chromosome as 14 and 10 in cases 8 and 10, respectively.

ETV6/AML1 rearrangements were detected, using previously RT-PCR analysis for ETV6 rearrangement published primers17 in five out of 19 patients investigated (cases 1–5). A sixth case (case 6), had an alternative spliced Control bone marrow from hematologically healthy individ- form of ETV6/AML1 fusion detected using 3′ RACE PCR uals, using the combination of primers illustrated in Figure 1a (figure 3). In this patient, the ETB6 exon 5 was fused in frame and listed in Figure 1b, showed six bands of equal intensity to AML1 alternative exon 2 which was then spliced to exon (Figure 2a). RT-PCR analysis was carried out in 19 of 20 5 (Figure 3). No product was obtained upon repeated RT-PCR patients (case 17 had insufficient material for this study and amplification using the conventional primers in this patient. was not investigated). In nine patients a reduction in one of The remaining 18 patients were tested for this newly identified the RT-PCR-amplified products was seen indicating a break- fusion form but were found to be negative. Five of the six point within ETV6 (Figure 2b–d). Breakpoints were localized patients with ETV6/AML1 fusion had been found to have within the ETV6 gene to intron 5 in seven patients, (cases 1– rearrangements in intron 5, the sixth case, however had been 4 and 6–8), to intron 2 (case 19), and to intron 4 (case 20) normal by exon specific RT-PCR.

Figure 2 RT-PCR analysis of the ETV6 gene using cDNA from bone marrow of patients with 12p rearrangements. RT-PCR analysis of the ETV6 gene using cDNA from bone marrow of patients with 12p rearrangements. A, control; B, patient 20; C, patient 7; D, patient 19; and E, patient 8. Each bands results from PCR amplification across two exons as indicated. In Figure 2b, amplification has failed between exons 4–5 suggesting that a rearrangement has occurred in intron 4 in patient 20. In Figure 2c the band resulting from the amplification between exons 5 and 6 is reduced suggesting that the organisation of the ETV6 gene was disrupted within intron 5 in patient 7. In patient No. 19, a reduction in the intensity of amplification was observed between exons 1–3 and exons 1–4 (Figure 2d) while amplification between exons 1 and 2 was normal, thus suggesting a breakpoint in intron 2. A reduction in amplification between exons 5–6 and 6–7 was observed in patient 8 (Figure 2e) suggesting the presence of a breakpoint in intron 5 and a lack of expression of the reciprocal product containing ETV6 exons 6 and 7. Analysis of the abnormalities of the ETV6 gene HE O’Connor et al 1103 Deletion of ETV6 detected by EcoRI/SacI for B1 RFLP

Our analysis indicated that the RFLP identified by the probe B1 occurred in 50% of hematologically healthy individuals and patients tested (data not shown). Seven of 11 patients tested were heterozygous for the B1 site (5.5 kb and 6.5 kb on the SacIorEcoRI digest). In four patients reduction of the 5.5 kb band (cases 11 and 15–17) suggested deletion of one allele (Figure 6). Three of these patients had cytogenetically detectable deletions of 12p (cases 11, 15 and 17), while case 16 had add(12).

Abnormalities of 12p and rearrangements of ET6 shown by exon-specific RT-PCR, by RT-PCR for ETV6/AML1 fusion or by Southern blotting to exon 5 or by B1 heterozygosity

Figure 3 ETV6/AML1 fusion product in patient 6. 3′ RACE PCR In summary, rearrangements of the ETV6 gene were shown in amplification using primers described in the Materials and methods 11/19 cases with abn 12p. These abnormalities were ident- indicated the presence of a fusion between exon 5 of ETV6 and exon ified by G-banding as partial arm deletions only (three cases), 2ofAML1. A splice between exon 2 and exon 5 of AML1 is suggested by sequence analysis of the clones obtained. Primers derived from as balanced translocations (four cases) or as translocations of this sequence were used for PCR amplification on patient’s material 12p which could not be fully described, ie add(12p) (four and the expected size fragment was obtained. No other patient tested cases). Nine patients had rearrangements in intron 5 of which in this series shows this type of alternative ETV6/AML1 product. six carried the ETV6/AML1 fusion product. Discrepancies between the molecular methods were seen in two cases. Case 5 was positive for the ETV6/AML1 fusion but had not shown rearrangement with exon-specific RT-PCR and case 11 showed rearrangement by Southern blotting only. Cases 7, 8 Southern blot analysis for rearrangement of ETV6 and 11 showed rearrangements in intron 5 which were not due to ETV6/AML1 fusion. The possibility that one allele of ETV6 might be deleted was investigated in the seven cases Southern blots with the enzyme BamHI were available for 16 which were polymorphic using the B1 probe. Four of these patients, of these 11 patients also had either EcoRI or SacI showed a deletion of the 5.5 kb band. In total, six patients had Southern blots. Southern blotting confirmed rearrangement of a deletion of ETV6, three of which also had a rearrangement of intron 5 in five available cases which had shown intron 5 the gene. involvement by exon-specific PCR (cases 1, 3, 6, 7 and 8) (Figure 4). In addition, Southern blotting showed a rearrange- ment in intron 5 in case 11 which had been normal on exon- Discussion specific PCR. In this case, the intensity of the rearranged frag- ment was less than that of the germ-line fragment, suggesting The detection of chromosomal abnormalities and subsequent that rearrangement had occurred in a minor clone and conse- identification of aberrant fusion genes in leukemia is quently it would have been undetectable by the RT-PCR assay important for diagnosis and patient management. The detailed (Figure 5). Rearrangements in these six cases were identified molecular analysis of patients with hematological malignancy using BamHI digests. In five cases the rearrangement was seen and a clonal abnormality of the short arm of chromosome 12 with the genomic probe B1, but not with the genomic probe presented here has revealed a marked association of this kary- D which suggested that a small deletion of intronic region had otypic abnormality with abnormalities of the ETV6 gene. This occurred during the genomic rearrangements. In one remain- study indicates that rearrangement of the ETV6 gene in the ing case (case 8) the rearrangement was identified only by the bone marrow may be as high as 60%. This incidence com- cDNA probes indicating that this deletion also included the pares with that reported by Sato et al29 who described ETV6 region covered by probe B1. Deletion of the germ-line allele rearrangements in 12/23 patients with 12p abnormalities. In was suggested by the presence of a faint germ-line band - our patient group, rearrangements of ETV6 were most frequent tive to the rearranged band and was detected in two cases in children with c-ALL where eight of nine children had a with ETV6 rearrangement (cases 3 and 6). However in the rearrangement in intron 5 of ETV6. In particular the ETV6- remaining three cases (cases 1, 7 and 8), the rearranged band AML1 fusion was present in 67% of children with c-ALL. The was of equal or slightly less intensity than the germ-line band incidence of the ETV6-AML1 fusion in childhood B-lineage suggesting ETV6 was present on the untranslocated allele, or ALL in an unselected series of childhood cases has been the deletion had occurred in a very small percentage of cells. between 15 and 30%.10,16,17 Our study suggests that this cryp- No rearrangements were identified elsewhere in the gene but tic translocation occurs predominantly in association with a this did not exclude rearrangements in other introns of ETV6 cytogenetically detectable abnormality of 12p. None of the as the probes used only covered a small portion of the ETV6 adults with ALL had the ETV6-AML1 fusion in this study. gene. No rearrangements were seen in cases 19 and 20 due The exon-specific RT-PCR analysis enables rearrangements to lack of available probes to introns 2 and 4. In four patients to be detected anywhere within the gene and in addition indi- (Nos 2, 4, 5 and 13) insufficient material was available for cates where the breakpoint has occurred. The assay enabled Southern blot analysis. identification of rearrangements in patients where no small Analysis of the abnormalities of the ETV6 gene HE O’Connor et al 1104

Figure 4 Southern blot analysis of the ETV6 gene in patients with 12p rearrangements. (a) Southern analysis of genomic DNA from controls and patient 7 probed with cDNA from the 3′ end of ETV6. The 20 kb and 8.5 kb germ-line fragment is present in the patient and the controls. The 14 kb rearranged BamHI fragment is arrowed. (b) Southern analysis of genomic DNA from controls and patient 6 probed with B1 which contains sequence of intron 5. Note the 14 kb rearranged BamHI fragment which is stronger than the 20 kb germ-line fragment suggesting a partial deletion of germline ETV6. (c) Patient 6 probed with cDNA from the 5′ end of ETV6 which shows the reciprocal 7 kb rearranged fragment. ␭HindIII molecular weight markers are indicated.

Figure 5 Southern blot analysis of the ETV6 gene in patient with Figure 6 Southern blot analysis to identify ETV6 deletions in 12p deletions. Southern analysis of genomic DNA digested with patients who are heterozygous for an EcoRI/SacI polymorphism. BamHI from controls and patients 2 and 11 probed with B1 which Southern blot analysis of the ETV6 gene in patients with 12p contains sequence of intron 5. Rearrangements are detected in two rearrangements. Total human DNA digested with SacI and probed patients with deletions of 12p cytogenetically. In one patient (case with the polymorphic probe B1. Note the two bands which occur in 11) the rearranged band was markedly fainter than the germ-line band 50% of cases. The patient (case 17) with a deletion of one ETV6 gene suggesting that the arrangement occurs in a minor clone. is arrowed.

deletion of the other ETV6 allele was detected by Southern cases) as previously reported.7,30 Patients with breakpoints in analysis. However in two cases, a rearrangement which was intron 2 and intron 4 of ETV6 were also identified in this study. identified by ETV6AML1 RT-PCR or Southern blotting was not Fusions between ETV6 introns 2 and 4 and the MN1 gene detected by exon-specific RT-PCR. This is probably due to the and the PDGFR␤ gene in patients with t(12;22) and t(5;12), size of the clone, detection of which would be beyond the respectively, have also been described.1,4 A breakpoint in sensitivity limits of the assay. In such cases FISH analysis intron 1 had also been reported.15 Investigation by RT-PCR would be more effective in detecting rearrangements. Break- using conventional primers for the presence of ETV6-AML1 points were concentrated in intron 5 (seven of nine positive fusion, revealed the fusion in four of the seven patients with Analysis of the abnormalities of the ETV6 gene HE O’Connor et al 1105 a breakpoint in intron 5 of ETV6 and in a fifth patient in whom allele in 5–20% of ALL patients. The region of loss, however, no breakpoint had been detected by the exon-specific included not only ETV6 but also the gene KIP1 and possibly approach. A possible explanation for the discrepancy would other as yet unknown genes.31,32 The target gene(s) for these be the presence of the in a minor clone, as dis- deletions is as yet unidentified. Two further deletions were cussed above. A sixth patient had an alternative spliced detected in patients who had rearrangements. fusion, detected by 3′ RACE PCR, which was unique in our It has been suggested that ETV6 is a ‘promiscuous’ gene and series of patients. In this transcript, exon 5 of ETV6 is fused FISH analysis using single copy probes has revealed a number to AMLI exon 5 as opposed to exon 3. FISH analysis using of other translocations of ETV6 with a variety of chromosome whole chromosome paints, when available, identified a trans- bands including 10q24, 9p23, Xq28 and 1q21 and 6q23.33–35 location t(12;21) in patients with ETV6-AML1 as expected. The translocation t(9;12) has led to the identification of a Evidence from FISH analysis indicated that both homologues novel fusion in which the HLH domain of ETV6 is fused to of chromosome 12 were structurally changed in two of four the JAK2 kinase in ALL and AML.36,37 Our findings confirm cases tested and that the cryptic t(12;21) translocation had that ETV6 may be involved in translocations with a number occurred on the cytogenetically normal 12p. These cases (Nos of other chromosomes; 1p13, 1q31, 14q11 and 15q15. The 1 and 6) also had a translocation of 12p to chromosome 1, five patients with novel ETV6 rearrangements will be analyzed and to chromosome 3, respectively. Two further cases had using RACE-PCR in an attempt to clone novel partner genes. translocations between chromosomes 12 and 21, and 12 and 11 or 14 (cases 4 and 8) but there was insufficient evidence to discover whether both homologues of chromosome 12 were Acknowledgements involved or whether a three-way translocation resulting in ETV6/AML1 fusion had occurred. Cases with both t(12;21) We would like to thank all clinicians and cytogenetics labora- and ins(12;1) or (3;12) showed a marked reduction in the tories who have contributed to this study. We are grateful to germline ETV6 allele on Southern blot analysis indicating that the Leukemia Research Fund and to the Kay Kendall Leukemia the translocation of 12p to chromosomes 1 and 3 was each Fund for supporting this study. accompanied by a deletion of ETV6 (cases 3 and 6, respectively). In a further patient carrying the ETV6/AML1 fusion (case 2), the rearranged band detected by Southern References analysis was weaker than the germline band. This does not preclude the possibility that a deletion of ETV6 had occurred 1 Golub TR, Barker GF, Lovett M, Gilliland DG. Fusion of PDGF in a minor clone which would be undetectable by Southern beta to a novel ets-like gene, tel, in chronic myelomono- analysis. In these cases, FISH analysis would be more pro- cytic leukemia with t(5;12) chromosomal translocation. Cell 1994; ductive in detecting any deletions. Deletion of the second 77: 307–316. 10 2 Golub TR, Barker GF, Bohlander SK, Hiebert SW, Ward DC, Bray- allele in cases with t(12;21) has been reported. Ward P, Morgan E, Raimondi SC, Rowley JD, Gilliland DG. Fusion Two children with c-ALL were shown to have a breakpoint of TEL gene on 12p13 to the AML1 gene on 21q22 in acute lym- in intron 5 but not to have the ETV6-AMLI fusion. In these phoblastic leukemia. Proc Natl Acad Sci USA 1995; 95: 4917– patients, a translocation of 12p with 1p13 and 14q11, respect- 4921. ively, had been demonstrated. Neither has previously been 3 Golub TR, Goga A, Barker GF, Agar DE, Mclaughlin J, Bohlander described in association with ETV6 but they form likely candi- SK, Rowley JD, Witte ON, Gilliland DG. Oligomerization of the ABL by the Ets protein TEL in human leukemia. date sites for novel fusion genes. Both these cases are being Mol Cell Biol 1996; 16: 4107–4116. further investigated for possible fusion partners to ETV6.In 4 Buijs A, Sherr S, van Baal S, van der Plas D, Geurts van Kessel A, one of these cases, FISH analysis had shown a derived 21 Riegman P, Lekanne Deprex R, Zwarthoff E, Hagemeijer A, Gros- containing DNA from chromosome 12 but this did not appear veld G. Translocation (12;22)(p13;q11) in myeloproliferative dis- to be reciprocal. This raises the possibility that genes other orders results in fusion of ETS-like TEL gene on 12p13 to the MNI than ETV6 and AML1 may be important in some cases. gene on 22q11. 1995; 10: 1511–1519. 5 Butler TA, Secker-Walker LM. Whole chromosome FISH identifies Two adults, one with acute biphenotypic leukemia and one chromosome 12p translocation partners in leukaemia. Br J Haema- with acute myeloid leukemia, with a translocation of 12p13 tol 1997; 97 (Suppl. 1): 198a (Abstr.). to 1q31 and 15q15 showed breakpoints in intron 2 and intron 6 Johansson B, Mertens F, Mitelman F. Cytogenetic deletion maps 4, respectively. These two locations are therefore candidate of hematologic : circumstantial evidence for tumor sup- sites for new fusion partners to ETV6. The identification of pressor loci. Genes Chromosom Cancer 1993; 8: 205–218. ETV6 rearrangements, using exon-specific RT-PCR, conven- 7 Papadopoulos P, Ridge SA, Boucher CA, Stocking C, Wiedemann LM. The novel activation of ABL by fusion to an ets-related gene, tional RT-PCR analysis and Southern blotting to exon 5, were TEL1. Cancer Res 1995; 55: 34–38. concordant in all but two cases. Failure to detect a rearrange- 8 Mitelman F, Kaneko Y, Trent J. Report of the Committee on Chro- ment by the exon-specific method in each case may have mosome Changes in Neoplasia, Human Gene Mapping 11 (1991). been explained by the presence of a relatively small clone Cytogenet Cell Genet 1991; 58: 1053–1079. below the level of detection by RT-PCR. 9 Martineau M, Clark R, Hawkins JM, Moorman AV, Butler TA, An attempt was made to look for deletions of ETV6 alleles Secker-Walker LM. Abnormal 12p acute lymphoblastic leukaemia in the LRF UK Cancer Cytogenetics Group karyotype database. in patients heterozygous for the B1 site. Among seven of the Cytogenet Cell Genet 1997; 77: 134. informative cases, three had a reduced amount of the 5.5 kb 10 Romana SP, Le Coniat M, Berger R. t(12;21): A new recurrent homologue indicating deletion. These were found in one translocation in acute lymphoblastic leukemia.Genes Chromosom patient with add(12p) and in three patients with cytogenet- Cancer 1994; 9: 186–191. ically visible del(12p). In these three cases the cytogenetic evi- 11 Romana SP, Mauchauffe M, Le Coniat M, Chumakov I, Le Paslier dence indicated that the region deleted was considerably D, Berger R, Bernard OA. The t(12;21) of acute lymphoblastic leu- kemia results in a tel-AML1 gene fusion. Blood 1995; 85: greater than that covered by a single gene and the relevance 3662–3670. of this gene loss in these cases is therefore uncertain. Studies 12 UKCCG. Translocations involving 12p in acute myeloid leukemia: on loss of heterozygosity have described deletion of one ETV6 association with prior myelodysplasia and exposure to mutagenic Analysis of the abnormalities of the ETV6 gene HE O’Connor et al 1106 agents. United Kingdom Cancer Cytogenetics Group (UKCCG). 26 Chomczynski P, Sacchi N. Single-step method of RNA isolation Genes Chromosom Cancer 1992; 5: 252–254. by acid guanidinium thiocyanate-phenol-chloroform extraction. 13 UKCCG. United Kingdom Cancer Cytogenetics Group. Translo- Analyt Biochem 1987; 162: 156–159. cations involving 9p and/or 12p in acute lymphoblastic leukemia. 27 Sambrook J, Fritsch EF, Maniatis T. Analysis and cloning of eukary- Genes Chromosom Cancer 1992; 5: 255–259. otic genomic DNA. In: Ford N, Nolan C, Ferguson M (eds). Mol- 14 Wasylyk B, Hahn SL, Giovane A. The Ets family of transcription ecular Cloning: A Laboratory Manual - 2nd edition. Cold Spring factors (published erratum appears in Eur J Biochem 1993; 215: Harbor Laboratory Press: Cold Spring Harbor, NY, 1989; pp 907). Eur J Biochem 1993; 211: 7–18. 9.38–9.46. 15 Peeters P, Wlodarska I, Baens M, Criel A, Selleslag D, Hagemeijer 28 Jabbar SA, Ganeshaguru K, Wickremasinghe RG, Hoffbrand AV, A, Van den Berghe H, Marynen P. Fusion of ETV6 to MDS1/EVI1 Foroni L. Deletion of chromosome 13 (band q14) but not trisomy as a result of t(3;12)(q26;p13) in myeloproliferative disorders. Can- 12 is a clonal event in B-chronic lymphocytic leukaemia (CLL). cer Res 1997; 57: 564–569. Br J Haematol 1995; 90: 476–478. 16 Shurtleff SA, Buijs A, Behm FG, Rubnitz JE, Raimondi SC, Hancock 29 Janssen JWG, Ridge SA, Papadopoulos P, Cotter F, Ludwig W, ML, Chan GC-F, Pui C, Grosveld G. TEL/AML1 fusion resulting Fonatsch C, Rieder H, Ostertag W, Bartram CR, Wiedemann LM. from a cryptic t(12;21) is the most common genetic lesion in pedi- The fusion of TEL and ABL in human acute lymphoblastic leu- atric ALL and defines a subgroup of patients with an excellent kaemia is a rare event. Br J Haematol 1995; 90: 222–224. prognosis. Leukemia 1995; 9: 1985–1989. 30 Sato Y, Bohlander SK, Kobayashi H, Shalini R, Suto T, Davis EM, 17 Romana SP, Poirel H, Le Coniat M, Flexor MA, Mauchauffe M, Espinosa R III, Hoopes R, Monygomery KT, Kucherlapati RS, Jonveaux P, Berger R, Bernard OA. High frequency of t(12;21) in LeBeau MM, Rowley ID. Heterogeneity in the breakpoints in bal- childhood B-lineage acute lymphoblastic leukemia. Blood 1995; anced rearrangements involving band 12p13 in hematologic 86: 4263–4269. malignancies identified by fluorescence in situ hybridisation: Tel 18 Raynaud S, Mauvieux L, Cayuela JM, Bastard C, Bilhou-Nabera (ETV6) is involved in half. Blood 1997; 90: 4886–4893. C, Debuire B, Bories D, Boucheix C, Charrin C, Fiere D, Gabert J. 31 Stegmaier K, Pendse S, Barker GF, Bray-Ward P, Ward DC, TEL/AML1 fusion gene is a rare event in adult acute lymphoblastic Montgomery KT, Krauter KS, Reynolds C, Sklar J, Donnelly M, leukemia. Leukemia 1996; 10: 1529–1530. Bohlander SK, Rowley JD, Sallan SE, Gilliland DG, Golub TR. Fre- 19 Aguiar RTC, Sohal J, Van Rhee F, Carapetti M, Franklin IM, Gold- quent loss of heterozygosity at the TEL gene in acute lym- stone AH, Goldman JM, Cross NCP. TEL/AML1 fusion in acute phoblastic leukemia of childhood. Blood 1995; 86: 38–44. lymphoblastic leukemia of adults. Br J Haematol 1996; 95: 32 Sato Y, Suto Y, Pietenpol J, Golub TR, Gilliland DG, Davis EM, 673–677. LeBeau MM, Roberts JM, Vogelstein B, Rowley JD et al. TEL and 20 Borkhardt A, Ludwig W, Mangioni S, Burci L, Schrappe M, Valsec- KIP1 define the smallest region of deletions on 12p13 in hemato- chi MG, Riehm H, Lampert F, Basso G, Masera G, Harbott J, poietic malignancies. Blood 1995; 86: 1525–1533. Biondi A. Does the TEL/AML1 expression define a subgroup of 33 Doheny KF, Rasmussen SA, Rutberg J, Semenza GL, Stamberg J, pediatric ALL with a good prognosis? The results of the German Schwartz M, Batista DA, Stetten G, Thomas GH. Segregation of a (BFM) and Italian (AEOP) ALL Study Groups. Blood 1996; 88: 451a (Abstr.). familial balanced (12;10) insertion resulting in 21 McLean WM, Ringold S, Neuberg D, Stegmaier K, Tantravahi R, dup(10)(q21.2q22.1) and del(10)(q21.2q22.1) in first cousins. Am Ritz J, Koeffler HP, Takeuchi S, Janssen JWG, Seriu T, Bartram CR, J Med Gen 1997; 69: 188–193. Sallan SE, Gilliland DG, Golub TR. TEL/AMLI dimerizes and is 34 Berger R, Le Coniat M, Daniel M, Lessard M, Berthou C, Maryen associated with favorable outcome in childhood acute lympho- P, Bernard OA. Chromosome abnormalities of the short arm of blastic leukemia. Blood 1996; 11: 4252–4258. chromosome 12 in hematopoietic malignancies: a report includ- 22 Nakao M, Yokota S, Horiike S, Taniwaki M, Kashima K, Sonoda ing three novel translocations involving the ETV6 gene. Leukemia Y, Koizumi S, Takaue Y, Matsushita T, Fujimoto T, Misawa S. 1997; 11: 1400–1403. Detection and quantification of TEL/AML1 fusion transcripts by 35 Suto Y, Sato Y, Smith SD, Rowley JD, Bohlander SK. A polymerase chain reaction in childhood acute lymphoblastic leu- t(6;12)(q23;p13) results in the fusion of ETV6 to a novel gene, STL, kemia. Leukemia 1996; 10: 1463–1470. in a B cell ALL cell line. Genes Chromosom Cancer 1997; 18: 23 ISCN (1995). An International System for Human Cytogenetic 254–268. Nomenclature (1995). S Karger: Basel, pp 1–114. 36 Lacronique V, Boureax A, Della Vale V, Poiret H, Quang CT, 24 Foroni L, Mason P, Luzzatto L. Ig and T cell receptor gene Mauchauffe M, Berthon C, Lessard M, Berger R, Ghysdaed J, rearrangements. In: Catovsky D (ed). Churchill Livingstone: UK, Bernard OA. A TEL-JAK2 fusion product with constitutive kinase 1991, pp 339–391. activity in human leukaemia. Science 1997; 278: 1309–1321. 25 Devaraj PE, Foroni L, Janossy G, Hoffbrand AV, Secker-Walker 36 Peeters P, Raynaud SD, Cools J, Wlodarska I, Grosgeorgr J, Phillip LM. Expression of the E2A-PBX1 fusion transcripts in P, Monpoux F, Rompaey T, Baens M, Van der Berghe H, Martnen t(1;19)(q23;p13) and der(19(1;19) at diagnosis and in remission of P. Fusion of TEL, the-ETS variant gene 6 (ETV6) to the receptor- acute lymphoblastic leukemia with different B lineage immuno- associated kinase JAK2 as a result of t(9;12) in a lymphoid and . Leukemia 1995; 9: 821–825. t(9;15;12) in a myeloid leukaemia. Blood 1997; 90: 2535–2545.