In T-cell acute lymphoblastic leukemia (T-ALL), LMO2 (Lim domain only 2) is frequently overexpressed due to chromosomal translocations or epigenetic changes. LMO2 is a molecular scaffold connecting LDB1 (Lim domain binding 1) to DNA-binding transcription factors. LDB1 is a major chromatin architectural factor involved in long-range enhancer-promoter interactions. Given the prevalence of aberrant LMO2 expression in T-ALL, we hypothesized that LDB1 promotes T-ALL proliferation/survival by connecting oncogenic enhancers to their target genes.

We performed LDB1 ChIP-seq in 2 T-ALL cell lines with high LMO2 expression: LOUCY (early thymic precursor [ETP]; SET::NUP214) and KOPT K1 (non-ETP; TCRG::LMO2). LDB1 is predominantly recruited to cis-regulatory elements, with 40-60% of peaks localizing to enhancers. However, LDB1 enhancer occupancy differed significantly between cell lines, suggesting subtype-specific regulation of transcriptional programs. Indeed, motif enrichment analysis at LDB1-occupied enhancers demonstrated distinct recruiting factors in LOUCY versus KOPT K1. To enable rapid LDB1 degradation, we tagged endogenous LDB1 with dTAG in both cell lines. Treatment with dTAG-V1 ligand for 4hrs depleted LDB1 by >90%. To assess LDB1-dependent enhancer-promoter connectivity, we treated cells with dTAG-V1 for 4hrs and performed Micro-C sequenced to a depth of ~1 billion valid contacts. Globally, compartmentalization and topologically associated domains remained largely intact in the absence of LDB1. However, we observed widespread changes to loops involving LDB1-occupied enhancers and/or promoters. Importantly, histone modifications and chromatin accessibility at enhancers remained intact upon acute LDB1 depletion, suggesting that architectural changes were directly due to loss of LDB1's looping function rather than indirect consequences of enhancer loss.

To explore how architectural changes impact gene expression, we measured nascent transcription using TT-seq upon acute LDB1 depletion and observed 328 and 134 differentially expressed genes in LOUCY and KOPT K1, respectively. LDB1-dependent transcripts in LOUCY included critical hematopoietic oncogenes (HHEX, MYB, MYCN, ERG, MEF2C) while those in KOPT K1 were enriched in immune activation pathways (DUSP6, BACH2, CD69). Reduced transcriptional output at these loci was associated with weakened enhancer-promoter connectivity. We analyzed 26 H3K27ac HiChIP datasets from patients with T-ALL and found that LDB1-dependent loops at these oncogenic loci are recapitulated and often correlate with T-ALL differentiation state. These findings suggest that LDB1 connects enhancers with oncogenic driver genes in both cultured and primary T-ALL cells.

To determine if LDB1 is required for T-ALL proliferation/survival, we passaged LOUCY and KOPT K1 cells in the presence of dTAG-V1 for 15 days and observed loss of cell expansion. Loss of LDB1 in LOUCY induced apoptosis, while in KOPT K1 it elicited a senescent phenotype characterized by beta-galactosidase activation, cell cycle exit, and senescence-associated gene expression. To determine which gene expression changes accounted for these phenotypes, we restored expression of several LDB1-dependent genes in LDB1-depleted cells using CRISPRa. Reactivation of oncogenes like MYB partially rescued cell expansion, suggesting that LDB1 dependency in T-ALL may in part be conveyed through MYB.

Finally, we compared LDB1-dependent transcriptional signatures from LOUCY and KOPT K1 to transcriptomic data from >1300 patients with T-ALL. The LOUCY signature was enriched in several high-risk T-ALL subtypes, including the recently defined ETP-like and LMO2 γδ-like entities. The KOPT K1 signature was conversely enriched in more differentiated subtypes such as TAL1/DP-like, together suggesting that LDB1 may be required to maintain transcriptional identity of T-ALL subtypes in vivo. Notably, LOUCY signature enrichment conferred worse overall survival (p=0.026) and event-free survival (p=0.00032), driven in part due to LDB1 regulation of high-risk transcriptional features in the ETP-like subtype.

Together, our study highlights LDB1 as a critical architectural factor that connects T-ALL subtype-specific enhancers and promoters to regulate transcriptional programs required for T-ALL cell expansion. Analyses of patient-derived data corroborate our findings and implicate the prognostic and therapeutic value of LDB1 regulatory networks.

This content is only available as a PDF.
Sign in via your Institution