t(15;21) translocations leading to the concurrent downregulation of RUNX1 and its transcription factor partner genes SIN3A and TCF12 in myeloid disorders

Through a combined approach integrating RNA-Seq, SNP-array, FISH and PCR techniques, we identified two novel t(15;21) translocations leading to the inactivation of RUNX1 and its partners SIN3A and TCF12. One is a complex t(15;21)(q24;q22), with both breakpoints mapped at the nucleotide level, joining RUNX1 to SIN3A and UBL7-AS1 in a patient with myelodysplasia. The other is a recurrent t(15;21)(q21;q22), juxtaposing RUNX1 and TCF12, with an opposite transcriptional orientation, in three myeloid leukemia cases. Since our transcriptome analysis indicated a significant number of differentially expressed genes associated with both translocations, we speculate an important pathogenetic role for these alterations involving RUNX1. Electronic supplementary material The online version of this article (doi:10.1186/s12943-015-0484-0) contains supplementary material, which is available to authorized users.

Translocations involving RUNX1 are known to decrease the function of the encoded protein in myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML) [1]. For those involving chromosome 15, SV2B was the only RUNX1 partner gene identified in AML [2].
We report on two novel t(15;21) alterations leading to the concurrent disruption of RUNX1 and SIN3A or TCF12 (Additional file 1: Table S1). Another interrupted gene is the UBL7-AS1 long noncoding RNA gene.
The results obtained in case 1 clearly suggest a role as a tumor suppressor (TS) gene not only for RUNX1 (the shorter RUNX1 encoded by the chimeric RUNX1/UBL7-AS1 should behave as a dominant negative mutant of the wild-type RUNX1), but also for SIN3A. We speculate that the inactivation of both proteins should have led to an abnormal activation of RUNX1/SIN3A target genes, leading to myelodysplasia. Even if haploinsufficiency was never reported for SIN3A, its role as a TS has been described in other tumors [4]. Notably, the SIN3A corepressor was known to interact with RUNX1 [5], leading to the transcriptional inactivation of their target genes [6].
In cases 2 and 3, FISH indicated that RUNX1 was interrupted within intron 7. Additionally, in case 4, a 600kb deletion removed the 5′ portion of RUNX1 starting from intron 6 ( Fig. 2a and Additional file 1: Table S1). In all cases, RUNX1 was joined with an opposite transcriptional orientation to intron 3 of TCF12 (NM_003205), a basic helix-loop-helix transcription factor (Fig. 2b) fused with MLL in MDS [7] and recurrently mutated in myeloproliferative disorders [8]. Notably, in case 2, RT-qPCR indicated the downregulation of both RUNX1 and TCF12 (Fig. 2c) and IPA analysis disclosed significant deregulation of the AML pathway. The number of altered pathways was slightly higher than in case 1 (146 vs. 136) and mostly overlapped the previously described categories. Particularly, RUNX1 was shown to control many differentially expressed genes involved in cell cycle regulation, inflammatory response, and transcription regulation (Additional file 8: Figure S3 and Additional file 9: Table S6).
We thus suggest a TS role for TCF12 in myeloid disorders, as already described in colon carcinoma [9]. The concurrent inactivation of RUNX1 and TCF12 should mimic the same leukemogenic effect of the t(8;21) RUNX1/CBFA2T1 fusion protein. E proteins, like TCF12,  (FISH) results that allowed the mapping of t(15;21) translocation breakpoints on der(21) (a) and der(15) (b), using the consistently colored probes listed for cases 2 (first column) and 4 (second column). RP11-662O10 was used only in case 4. Cases 2 and 3 shared the same breakpoints (data not shown). On the right, the map of the BAC probes used in the FISH experiments, according to GRCh37/hg19, and identifying both translocation and deletion breakpoints, is shown. The black and orange dashed lines indicate the breakpoints in cases 2 and 4, respectively. The grey rectangle encompasses the deleted region flanking the translocation breakpoint in case 4. c Evaluation of RUNX1 and TCF12 expression levels: exon-specific reverse transcription quantitative polymerase chain reaction results of runt-related transcription factor 1 (RUNX1; left) and transcription factor 12 (TCF12; right) in case 2. Asterisks indicate statistically significant results (P < .05) are inactivated through their interaction with the domain TAFH of CBFA2T1, leading to the inhibition of the p300/ CBP histone acetyltransferase recruitment at their target genes' promoters and consequently to the lack of activation of genes with E-box promoters [10].
To summarize, we here identified three novel RUNX1 partner genes, including two transcription factors and a long noncoding RNA, in 2 t(15;21) translocations. Both of the t(15;21) translocations resulted in the concurrent inactivation of RUNX1 and one related transcription factor (SIN3A or TCF12), leading to the potential haploinsufficiency of both involved genes. Moreover, the IPA analyses clearly indicated that the AML pathway was significantly deregulated in our samples, and showed that RUNX1, SIN3A, or TCF12 have a crucial impact on differentially expressed genes. The analysis of additional cases harboring t(15;21) translocations will be helpful to better understand the pathogenetic impact of these alterations in myeloid neoplasms.