1. ¹Leukaemia Research Fund Cytogenetics Group, Cancer Sciences Division, University of Southampton, Southampton, UK
2. ²Leukaemia Research Fund Centre, Institute of Cancer Research, Chester Beatty Laboratories, London, UK
3. ³Department of Haematology, Royal Manchester Children's Hospital, Manchester, UK

Correspondence: Dr CJ Harrison, Leukaemia Research Fund Cytogenetics Group, Cancer Sciences Division, MP 822 Duthie Building, Southampton General Hospital, Tremona Road, Southampton SO16 6YD, UK. Fax: + 44 023 8079 6432

Received 19 November 2002; Accepted 5 December 2002.

TO THE EDITOR

The translocation, t(12;21)(p13;q22), is the most common chromosomal rearrangement in children with B-cell precursor acute lymphoblastic leukemia (ALL), occurring in approximately 25% of patients.^1,2 It is invisible by G-banded chromosomal analysis and was discovered fortuitously by fluorescence in situ hybridization (FISH) using whole-chromosome painting probes in 1994³. The translocation fuses the 5' part of the ETV6 gene on the short arm of chromosome 12 (12p) with most of the AML1 gene on the long arm of chromosome 21 (21q), to produce the chimeric fusion transcript ETV6/AML1 on the derived chromosome 21⁴. Thus, FISH with probes specific for these regions or reverse transcriptase polymerase chain reaction (RT-PCR) with appropriate primers can be used in the detection of t(12;21). From detailed cytogenetic and FISH studies, a small number of variant and complex chromosomal rearrangements have been reported that involve chromosomes 12 and 21 in the production of the ETV6/AML1 fusion gene.^5,6 This is the first report of the ETV6/AML1 fusion produced by a visible insertion of most of 12p into chromosome 21.

A 5-year-old boy presented with hepatomegaly, a white blood cell count of 28 10⁹/l and 63% blasts on the blood film. Bone marrow examination showed increased cellularity with 96% lymphoblasts of L1 morphology. The blast cells were PAS positive, Sudan black and acid phosphate negative. The immunophenotype showed the leukemic cell population to be TdT, CD10 and CD19 positive, consistent with a diagnosis of common ALL. The patient was entered into the UK Medical Research Council ALL treatment trial, UKALLXI, for children aged 1–14 years. He has remained in continuous complete remission for 8 years.

Metaphases from the diagnostic bone marrow were examined by G-banded chromosomal analysis. Parallel multiplex-FISH (M-FISH) was performed with the 24-color chromosome painting kit (SpectraVysion, Vysis, UK) on the same cell suspension. Interphase nuclei were examined by FISH using the dual-color probe kit LSI TEL/AML1 ES (Vysis, UK). Other probes used for metaphase FISH analysis included directly labeled subtelomeric probes (Appligene Oncor, Qbiogene, UK) to chromosomes 1, 10, 12, 13 and 21, and cosmid exon-specific probes. There were nine cosmids covering the ETV6 gene, 179A6 (exon 1), 15A4 (exon/intron 1), 67C6 (intron 1), 50F4 (exon 2), 132B11 (intron 2), 242E1 (exon 3), 163E7 (exons 3-5), 54D5 (exons 5-8), 148B6 (exon 8), and two covering the AML1 gene, ICRF c103 CO664 (exons 1-5) and H11086 (exons 6 and 7). The probes for M-FISH, subtelomeres and specific gene exons were, when possible, sequentially applied to the same metaphases in order to obtain a complete cytogenetic picture.

For the molecular tests, mRNA was extracted from leukemic blood cells taken at diagnosis, reverse transcribed and screened for TEL/AML1 fusion sequences by RT-PCR as previously described.⁷

Interphase FISH revealed the ETV6/AML1 yellow fusion in 62% of cells together with one green signal indicating the retention of the ETV6 gene on the second copy of chromosome 12 and two red signals of similar size for AML1.

The complementary application of cytogenetics and sequential FISH revealed a highly complex karyotype: 48,XY,der(1)t(1;10)(q1; p1?1),+4,+10,der(10)t(10;13)(p1?1;q3?2)t(1;10)(q1;q1)t(1;12)(qter; pter)x2,der(12)del(12)(p11)t(1;12)(qter;p11),der(13)del(13)(q1 q32)t(10;13)(q1;q1),der(21)inv ins(21;12)(21pter>21q21::12p13> 12p13::21q22.1>21q21: :12p11>12p13::21q22.1>21qter)/49, idem,+21. The italics indicate those abnormalities identified only by the subtelomeric probes. G-banding and M-FISH provided clear evidence of an insertion of a substantial amount of 12p into chromosome 21 and complex rearrangements involving the pericentromeric regions of chromosomes 1, 10 and 13. Five derived (der) chromosomes were defined in the karyotype (Figure 1). On three of these chromosomes, namely der(1)t(1;10), der(13)t(10;13) and der(21)ins(21;12), the signals of the 10p, 10q and 21q subtelomeric probes were observed in the expected locations (Figure 1). However, the der(10) and the der(12) had exchanged telomeres. The arm of the der(10) composed of 1q terminated with a 12p subtelomeric probe signal and the deleted 12p with the 1q subtelomeric probe signal (Figure 1). The subtelomeric probe 21q signal was seen on the der(21) confirming that 12p material had been inserted into 21q. Its presence on one or two normal copies of chromosome 21 verified that the second AML1 signal at interphase, represented an extra copy of chromosome 21.

Figure 1.

Partial karyograms illustrating the chromosomal abnormalities, involving chromosomes 1, 10, 12, 13 and 21. In each box the normal chromosome is on the left and the derived chromosome on the right. (a–d) Chromosome 1 and der(1): (a) G-banded, (b) M-FISH, (c) subtelomeric probes for chromosome 1 (1p in green and 1q in red), (d) 10p subtelomere. (e–i) Chromosome 10 and der(10): (e) G-banded, (f) M-FISH, (g) subtelomeric probes for chromosome 10 (10p in green and 10q in red), (h) 12p subtelomere, (i)13q subtelomere. (j–m) Chromosome 12 and der(12): (j) G-banded, (k) M-FISH, (l) subtelomeric probes for chromosome 12 (12p in green and 12q in red), (m) 1q subtelomere. (n–q) Chromosome 13 and der(13): (n) G-banded, (o) M-FISH, (p) subtelomeric probe for 13q, (q) 10q subtelomere. (r–t) Chromosome 21 and ins(21;12): (r) G-banded, (s) M-FISH, (t) 21q subtelomere.

Full figure and legend (138K)

Metaphase FISH with the exon-specific probes to AML1 showed the gene to be intact on the der(21) and on the one/two normal copies of chromosome 21. There was no reciprocal AML1 signal on the der(12), consistent with the findings from the 21q subtelomeric probe. The signals of the probes specific to the exons of ETV6 were all observed within the insertion of 12p into 21q. The origin of the gene from the der(12) was assumed, since there was a visible deletion and all the ETV6 exons were missing from it. Dual-color hybridization of the probes 179A6 and 148B6, specific for exons 1 and 8 of ETV6 showed a distinct separation of these two signals on the der(21). On a normal ETV6 gene, these probes when applied together usually appear closely apposed as a fusion signal. This dual-color hybridization also revealed that the orientation of the inserted material containing ETV6 appeared to be inverted in comparison with the position usually observed on an intact 12p. In this patient, the 5' end (exon 1) was in a centromeric rather than a telomeric position, while the 3' end (exon 8) was telomeric. At the same time, it was demonstrated that exon 8 was deleted from the second ETV6 allele on the homologous chromosome 12. Exons 1–7 were present on this chromosome, which appeared normal by cytogenetics and M-FISH.

Leukemic cells were positive for the ETV6/AML1 fusion transcript by RT-PCR. The amplicon was of the predicted size for the major ETV6/AML1 mRNA product (300 bp) and sequencing of the band confirmed that it was a (bona fide) ETV6 (exon 5)-AML1 exon 2 fusion mRNA. This result provided important verification that the usual ETV6/AML1 fusion event had occurred, since side by side positioning of AML1 and ETV6 on the der(21) would also have produced a colocalization of the AML1 and ETV6 signals with the commercial probe, indistinguishable from a true fusion signal.

As a result of the standard t(12;21), the breakpoints producing the ETV6/AML1 fusion occur between exons 5 and 6 of ETV6 and 1 and 2 of AML1, thus exons 1–5 of ETV6 are usually translocated to the derived chromosome 21 and exon 1 of AML1 moves to the derived chromosome 12. In this patient, we demonstrated that exons 1–8 of ETV6 were inserted into the der(21). However, to achieve the correct orientation of ETV6 and AML1 for the molecular positivity observed in this case, the first event must have been the insertion of ETV6 into the q arm of chromosome 21 centromeric to AML1. This would require the second event to be breakage within both AML1 and ETV6 in the usual introns, followed by an inversion of the material between the two breaks (Figure 2). After the inversion, exons 1 and 8 would become separated by the chromosome 21 material between the 12p insertion and AML1. The two exons would thus be further apart than in the intact gene, consistent with the discrete nature of the signals observed for the probes specific to exons 1 and 8.

Figure 2.

Diagram of the proposed sequence of events resulting in ETV6/AML1 fusion. (1) Direct insertion of 12p into 21q centromeric to AML1. (2) Breakage within both AML1 and ETV6 in the usual introns and inversion of the material between the two breaks. (3) This results in separation of exons 1 and 8 of ETV6 by chromosome 21 material between the 12p insertion and AML1.

Full figure and legend (50K)

In summary, the complementary use of interphase FISH, G-banding and M-FISH, together with subtelomeric and exon-specific probes, has been successful in the accurate characterization of the complex karyotype of a patient with ETV6/AML1 positive ALL. Although the involvement of chromosomes 1, 10, 12, 13 and 21 were evident from the G-banded and M-FISH karyograms, subtelomeric probes were required to uncover a cryptic exchange of telomeres within this complex rearrangement. Specific probes were necessary to discover the insertion and inversion events giving rise to the ETV6/AML1 fusion and to reveal the deletion of exon 8 from the apparently normal chromosome 12. Although ETV6/AML1 fusion by insertion of 21q into 12p has been previously described,⁶ we are the first to report the opposite phenomenon, namely insertion of 12p into 21q, producing the same result.