The PBX Homeobox 1 (PBX1) protein is a novel molecular target for multiple myeloma therapy

Abstract

Gain of chr1q (gain1q) is a secondary genetic event present in approximately 30% of patients with multiple myeloma (MM), an incurable blood cancer of plasma cells (PC), and one of the top 3 high-risk markers of adverse prognosis. However, the molecular mechanisms underlying this event remain unclear. The transcription factor (TF) PBX homeobox 1 (PBX1), encoded on chr1q, is a master regulator of early haematopoiesis and a prominent oncogene in leukaemia and other malignancies. Herein, I hypothesized that PBX1 has a central role in orchestrating proliferative regulatory networks that underpin the high-risk prognostic features associated with gain1q in MM.

First, combined genomic (WES/WGS) analysis of the MMRF Compass dataset revealed that poor prognosis of gain1q MM is exhibited only in patients with genetic amplification that extends beyond the traditionally considered 1q21 region, to include the PBX1 locus. RNA-seq and immunohistochemical (IHC) analysis confirmed that PBX1 expression is restricted to haemopoietic stem cells, progenitors and megakaryocyte/erythrocyte lineages in healthy individuals, while expression of PBX1 is “turned-off” across B-cell development, including PC. Ectopic expression of PBX1 was identified in myeloma PC of ~15% of MM patients (RNA-seq data from 3 independent cohorts). Validating IHC analysis showed expression of PBX1 in bone marrow myeloma PC at clonal (7/32) or sub-clonal level (14/32 MM patients). Upregulation of PBX1 was significantly associated with high-risk clinical variables, including presence of clinical symptoms at diagnosis, low albumin levels, advanced ISS stage, elevated proliferation myeloma PC index, progressed and relapsed disease. Moreover, significant correlation between PBX1 amplification and overexpression was confirmed in two MM patient cohorts.

Next, genetic depletion of PBX1 using two validated shRNAs demonstrated prominent addiction of gain1q human myeloma cell lines (1q HMCLs) to PBX1, both in vitro (MM1S, U266 cell lines) and in vivo (MM1S cells). Transcriptomic profiling of PBX1-depleted cells from two 1q HMCLs defined prominent involvement of PBX1 in transcriptional regulation of key oncogenic pathways, including cell cycle regulation, apoptosis, DNA replication and DNA damage responses. Flow-cytometric analysis validated G1-phase cell cycle arrest of PBX1-depleted cells. ChIP-seq analysis against PBX1 in the same 1q HMCLs, with reference to in-house built ChromHMM epigenomic maps, identified thousands of binding sites across the MM genome, predominantly detected in active regulatory regions (~60-80% occupancy on active promoter and enhancer loci). Additional analysis using H3K27ac ChIP-seq data across 8 primary MM samples and 9 cell lines showed PBX1 recruitment in approximately 70% of all identified super-enhancer (SE) regions. Clinical stratification based on the gain1q status revealed higher occupancy of PBX1 on gain1q-specific SEs, while H3K27ac signal on PBX1 sites was significantly higher in gain1q-positive versus -negative cells, supporting extensive epigenetic reprogramming that favours PBX1 recruitment in gain1q MM cells.

Integrated cistrome-transcriptome analysis defined more than 1,000 genes to be directly regulated by PBX1 in two 1q HMCLs, with primary involvement in hallmark oncogenic processes (cell cycle progression, DNA damage responses, metabolism and ER stress). Importantly, PBX1 was found to directly bind onto and control expression of E2F1, E2F2, FOXM1 and its downstream target NEK2, while enrichment analysis on public ChIP-seq datasets suggested significant overlap between PBX1 and FOXM1 transcriptional programmes in 1q HMCLs. Genetic depletion of FOXM1 confirmed prominent addiction of MM1S cells to FOXM1 expression; transcriptomic analysis revealed approximately 800 genes, mainly involved in cell cycle control, to be regulated by FOXM1. Flow cytometry-based analysis on FOXM1-depleted cells confirmed involvement of FOXM1 in regulation of the G2-M cell cycle phase. Epigenetic analysis demonstrated direct regulation of NEK2 by FOXM1 and the presence of a FOXM1 autoregulatory loop in MM1S cells. However, PBX1 expression levels remained unaffected upon FOXM1 depletion, suggestive of a unidirectional epigenetic connection between these two TFs in gain1q MM cells. In addition, transcriptional profiling of two 1q HMCLs upon PBX1 overexpression validated FOXM1, NEK2 and E2F1 as its direct transcriptional targets. Notably, overexpression of PBX1 in two 1q HMCLs demonstrated significantly increased resistance to the specific anti-FOXM1 inhibitor thiostrepton.

Finally, differential TF repertoire and wiring profiling of gain1q-positive versus -negative primary MM samples via combined RNA-seq/ATAC-seq analysis uncovered a molecular circuitry of higher complexity and TF density in gain1q MM cells. Strikingly, both PBX1 and FOXM1 were identified as top over-expressed and hyper-wired TFs in gain1q-positive versus -negative cells, suggesting a central role of the PBX1-FOXM1 axis in transcriptional regulation of gain1q myeloma cells. Additional analysis of transcriptome data from two independent MM patient cohorts (>1,200 patients combined), validated significant co-expression of PBX1 and FOXM1 with their transcriptional targets, and further integration with genomics data revealed presence of the PBX1 and PBX1-FOXM1 programmes in MM patients with PBX1 and 1q genetic amplification. Of note, all 10 genes previously identified to comprise a high-risk signature in MM cells were found as direct regulatory targets of PBX1 in this study, suggesting a direct involvement of PBX1 to the high-risk phenotype of gain1q MM cells. Survival analysis displayed significant correlation between activation of the PBX1 molecular programme with poor prognosis in nine solid tumour patient cohorts, proposing a pan-cancer role of the PBX1 circuitry to promote high-risk disease.

Overall, these data highlight PBX1 as a novel TF regulator in gain1q MM cells and support a critical PBX1-FOXM1 regulatory axis that promotes cell cycle progression and high-risk disease in gain1q MM cells and solid tumours, providing new molecular perspectives for drug development.