Inhibition of TLX3 and NKX2-5 in t(5;14)(q35;q32) T-ALL after Blocking Remote 3′-BCL11B Enhancer Sequences with Matching DNA Oligos Reveals Coregulation by PU.1 and HMGA1.

In T-cell acute lymphoblastic leukaemia (T-ALL) alternative t(5;14)(q35;q32.2) forms effect leukemic dysregulation of either TLX3 or NKX2-5 homeobox genes at 5q35 by juxtaposition with 3′-BCL11B at 14q32.2. Putative regulatory sequences underlying ectopic homeobox gene activation in t(5;14), and their mode of action have remained poorly understood mainly because breakpoints at 14q32.2 are widely scattered over the ~1 Mbp genomic desert region. We pooled cytogenetic data from t(5;14) cell lines together with published clinical data to refine the BCL11B downstream breakpoint cluster region (bcr). Ectopic homeobox gene dysregulation was investigated by DNA-i(nhibitory-treatments) with 26-mer double-stranded DNA oligo(nucleotide)s directed against putative enhancers using NKX2-5 expression as endpoints. Enhancer targets were provisionally identified from orphan T-cell DNase-I hypersensive sites (DNaseI-HS) located between 3′-BCL11B and VRK1. NKX2-5 downregulation in t(5;14) PEER cells was almost entirely restricted to DNAi targeting enhancers within the distal bcr and was dose- and sequence-dependent. Interestingly, enhancers near 3′-BCL11B regulated that gene only. These data imply that enhancer-promoter distances and/or locations are important for long-range gene regulation. Chromatin immunoprecipation assays showed that the four most effectual NKX2-5 ectopic enhancers were hyperacetylated. These enhancers clustered ~1 Mbp downstream of BCL11B, within a region displaying multiple regulatory stigmata, including a TCRA-enhancer motif, and abyssal sequence-conservation (“5-Way Regulatory Potential”). Paradoxically, although TLX3/NKX2-5 promoter/exon-1 regions were hypo-acetylated, their expression decreased after TSA treatment, implying extrinsic regulation by factor(s) subject to acetylation-control. PU.1 is known to get transcriptionally repressed by TSA and potentially binds TLX3/NKX2-5 upstream promoter regions. Knockdown of PU.1 effected downregulation of both homeobox genes. Moreover, genomic analysis showed preferential enrichment near validated ectopic enhancers of binding sites for the PU.1-cofactor HMGA1, knockdown of which also inhibited NKX2-5 in PEER cells. Analysis of nuclear matrix attachment (NMA) in PEER cells showed enhanced attachment near to the most effectual enhancer cluster which was alleviated by TSA-treatment. Interestingly, the juxtapositional genomic regions of “active” ins(14;5) rearrangements driving NKX2-5 expression exhibited tight NMA, forming structures reminiscent of “active chromatin hubs”. These findings lead us to propose that HMGA1 and PU.1 co-regulate ectopic homeobox gene expression in t(5;14) T-ALL by interactions mediated at the nuclear matrix, possibly mediated by SATB1 binding. Our data document homeobox gene dysregulation by a novel regulatory region at 3′-BCL11B responsive to HDAC-inhibition and highlight a novel class of potential therapeutic target amid “junk” DNA.

Duke Scholars

Author Gregory E. Crawford Pediatrics, Medical Genetics