The complex translocation (9;14;14) involving IGH and CEBPE genes suggests a new subgroup in B-lineage acute lymphoblastic leukemia

Abstract Many subtypes of acute lymphoblastic leukemia (ALL) are associated with specific chromosomal rearrangements. The complex translocation t(9;14;14), a variant of the translocation (14;14)(q11;q32), is a rare but recurrent chromosomal abnormality involving the immunoglobulin heavy-chain (IGH) and CCAAT enhancer-binding protein (CEBPE) genes in B-lineage ALL (B-ALL) and may represent a new B-ALL subgroup. We report here the case of a 5-year-old girl with B-ALL, positive for CD19, CD38 and HLA-DR. A direct technique and G-banding were used for chromosomal analysis and fluorescentin situ hybridization (FISH) with BAC probes was used to investigate a possible rearrangement of the IGH andCEBPE genes. The karyotype exhibit the chromosomal aberration 46,XX,del(9)(p21),t(14;14)(q11;q32). FISH with dual-color break-apartIGH-specific and CEPBE-specific bacterial artificial chromosome (BAC) probes showed a complex t(9;14;14) associated with a deletion of cyclin-dependent kinase inhibitor 2A (CDKN2A) and paired box gene 5 (PAX5) at 9p21-13 and duplication of the fusion gene IGH-CEBPE.


Introduction
Acute lymphoblastic leukemia (ALL) is a malignant clonal proliferation of lymphoid progenitor cells, most commonly of the B-cell lineage (B-ALL). In pediatric populations, ALL accounts for 81% of childhood leukemias, with leukemia in general accounting for one third of cancers diagnosed in children up to 14 years of age (Howlader et al., 2014;Woo et al., 2014).
Deletion of PAX5 and CDKN2A was reported by Kim et al. (2011); their comprehensive studies using FISH, G-banding and immunohistochemistry (IHC) showed that PAX5 deletion was common in childhood and adult B-ALL. To our knowledge, the t(14;14) has been reported in only six cases of B-ALL (Berger et al., 2001;Han et al., 2008). Here, we report for the first time, the simultaneous involvement of an IGH (14q32)/CEBPE (14q11) fusion gene and a PAX5/CDK2NA concurrent deletion (9p13p21) in a complex translocation t(9;14;14) in a case of childhood B-ALL.
The first chemotherapy protocol (FRALLE 93) was started. After induction and consolidation, the complete first remission (2% blast cells) was achieved 2.5 years after admission. Five months later, she relapsed with 92% blast cells. A second chemotherapy protocol was started (COPRALL 2001), but two months later the patient presented a nosocomial infection. Following a third protocol (VANDA), 1.5 months later, a second complete remission was obtained with no blast cells detected. As no compatible family member was found, bone marrow transplant was not considered as an option for treatment. One year later, the blood analysis showed an infection with Staphylococcus and Clostridium difficile with 91% of blast cells. A fourth chemotherapy protocol was started, but unfortunately six months later the patient passed away.

Chromosomal analysis
Chromosomal analysis of a bone marrow sample was done using a direct technique (Shiloh and Cohen, 1978). This method was based on short (25 min) incubation, immediately following aspiration, in a solution containing hypotonic KCl and colcemid that omitted the use of tissue culture medium. A conventional G-banding method was used for karyotyping. Clonal karyotype anomalies were described according to ISCN (Shaffer et al., 2013).

Fluorescence in situ hybridization
FISH was used to investigate whether t(14;14)(q11; q32) involved rearrangement of the genes IGH and CEBPE and was done as previously described by Akasaka et al. (2007). DNA was extracted from a BAC clone using a QIAGEN plasmid midi kit (Qiagen, Hilden, Germany) following the manufacturer's protocol. BAC DNA was labeled by nick translation (Roche Diagnostics, Mannheim, Germany) using a nick translation test kit (Abbott/Vysis, USA). Pretreatment of the probe and hybridization were done as previously described .
We used a break-apart LSI CDKN2A BAC clone (RP11-149I2/70L8) (Welcome Trust Sanger Institute, http://www.sanger.ac.uk) to detect the deletion of CDKN2A gene (P16) on 9p21. We also used a CEP9 probe (Abbott/Vysis, USA) to detect the deletion of chromosome 9p and a LSI MYB probe for chromosome 6 as an internal control.
The FISH signal was amplified and detected by using a conventional system that included a first layer of FITC-Avidin, a second layer of biotinylated-anti-Avidin and a third layer of FITC-Avidin (Cambio, Cambridge, UK). The BAC probe was initially hybridized to normal metaphases to confirm its location (data not shown). The FISH signal was captured using a Leica DMRXA fluorescence micro-scope (Leica, Wetzlar, Germany) and Q-FISH imaging software (Metasystems, Altlussheim, Germany) was used to scan and capture the images. At least 20 metaphases and/or 100 interphase nuclei were analyzed for each test. Each metaphase was counterstained with 4'-6-diamidino-2-phenylindole (DAPI) (Roche Diagnostics, Laval, QC, Canada).

Fluorescence in situ hybridization
In each analyzed cell, we observed two abnormal derivative chromosomes 14 ( Figures 1B and 2). Two FISH signals were observed on the large derivative chromosome 14: an orange signal at the translocation breakpoint 14q32 (3' part of the IGH break-apart probe, 442F20) and an orange/green fusion signal at the normal IGH locus (442F20 and DJ998D24). One green signal corresponded to the non-rearranged 14q11 locus and the second green signal (5' part of the CEBPE gene) to the rearranged IGH locus (14q32). The former locus was translocated from the small derivative 14 (3' part of the CEBPE gene). The small derivative chromosome 14 showed a single green signal at the translocation breakpoint 14q11 (5' part of the IGH breakpoint probe DJ998D24). No normal cells were seen in this analysis.
To detect rearrangement of the CEBPE gene on the 14q11 locus, we used a FITC-labeled BAC green probe (RP11-147E17 and RP11-68M15) (http://www.sanger.ac.uk). Figures 1C and 2 show that two green FISH signals were detected on the large derivative chromosome 14; a single green signal (3' part of the CEBPE BAC probe) was also detect at the translocation breakpoint (14q11) on the small derivative chromosome 14. Another green signal was observed on the derivative chromosome 9 (9p21) on the 3' part of the CEBPE BAC probe. To investigate the breakpoint on chromosome 9, we used a dual-color break-apart PAX5 BAC probe (RP11-243F8, RP11-297B17 and RP11-344B23) (http://www.sanger.ac.uk). Only one orange/green signal was seen on a normal chromosome 9, indicating that the PAX5 gene on the other chromosome 9 was deleted. Figures 1D and 2 show the metaphase FISH analysis using the CEP9 probe for the two chromosomes 9, with two green signals on the centromeres: one on normal chromosome 9 and the other on derivative chromosome 9 (9p-). An MYB SpectrumAqua probe was used on both normal chromosomes 6 as an internal control and showed two aquablue signals. Figures 1E and 2 show that metaphase FISH analysis using the CEP9 probe (9p11-q11) confirmed deletion of the PAX5 locus. Interphase nuclei FISH analysis using break-apart CDKN2A probe for the two chromosomes 9 yielded two green signals for CEP9 and only one red signal for CDKN2A at 9p21 (data not shown).

Discussion
Based on GTG banding alone the interpretation of the karyotype was 46,XX,del(9)(p21), t(14;14)(q11;q32). However, in our case, the FISH results using the BAC locus probe specific for CEBPE and the break-apart probe specific for IGH (442F20 and DJ998D24) showed signals corresponding to the 3' of CEBPE and 5' of IGH on deleted chromosome 9, suggesting the presence of an IGH-CEBPE fusion gene. Based on the FISH results, the most probable interpretation of this karyotype was a complex translocation t(9;14;14) associated with a large deletion within 9p and a duplication involving at least the fusion gene IGH-CEBPE.
Chromosome in situ hybridization with BAC specific for the CEBPE and IGH genes revealed a hybridization profile compatible with rearrangement of the CEBPE (14q11.2) and IGH (14q32) loci. This finding suggested the presence of IGH-CEBPE fusion on the small derivative chromosome 14 and CEBPE-IGH fusion on the derivative large chromosome 14, a conclusion in agreement with Han et al. (2008), who demonstrated the involvement of IGH and CEBPE genes in t(14;14)(q11;q32) in B-ALL.
Intra-chromosomal translocations involving IGH and CEBPE have been described in childhood ALL and result in the upregulation of CEBPE expression, suggesting that CEBPE plays a possible role in the development of B-ALL (Akasaka et al., 2007). The presence of an IGH-CEBPE fusion on 9p suggests that a duplication and large deletion occurred simultaneously with a translocation involving 9p12, 14q11 and 14q32. This complex single rearrangement event led to the formation of an IGH-CEBPE fusion gene and concurrent deletion of PAX5 and CDKN2A on 9p. Simultaneous deletion of PAX5 and CDKN2A is a common event in leukemogenesis and most ALL patients with a deletion of PAX5 have a concurrent deletion of CDKN2A (Kim et al., 2009).
Although we cannot exclude that the translocation t(14;14) and deletion 9p are two independent events, we believe that the presence of a second set of IGH-CEBPE fusion genes at the breakpoint of 9p reflects the activity of a DNA repair mechanism such as non-homologous end joining (NHEJ). This pathway repairs double-strand breaks with no homologous sequence and usually underlies deletions and duplications at the breakpoints of the two broken DNA ends to be tied (Zhang et al., 2009a,b). If NHEJ is involved, then only one event was needed to produce a double set of IGH-CEBPE and the concurrent deletion of PAX5 and CDKN2A. The occurrence of all these aberrations probably potentiated the aggressive refractory leukemia in our patient. Our case increases the number of B-ALL patients with t(14;14) in the literature to seven. Table 1 summarizes the clinical, hematological, immunophenotypic and genetic findings of these patients.

Conclusion
Our B-ALL finding revealed a complex translocation t(9;14;14)(p12;q11;q32) accompanied by the formation of an IGH-CEBPE fusion gene and its duplication, and the concurrent deletion of PAX5 and CDKN2A on 9p. To our knowledge, this is the first report to identify four important steps of leukemogenesis simultaneously in one event.