Harnessing the MYB-dependent TAL1 5’super-enhancer for targeted therapy in T-ALL

The acquisition of genetic abnormalities engendering oncogene dysregulation underpins cancer development. Certain proto-oncogenes possess several dysregulation mechanisms, yet how each mechanism impacts clinical outcome is unclear. Using T-cell acute lymphoblastic leukemia (T-ALL) as an example, we show that patients harboring 5’super-enhancer (5’SE) mutations of the TAL1 oncogene identifies a specific patient subgroup with poor prognosis irrespective of the level of oncogene dysregulation. Remarkably, the MYB dependent oncogenic 5’SE can be targeted using Mebendazole to induce MYB protein degradation and T-ALL cell death. Of note Mebendazole treatment demonstrated efficacy in vivo in T-ALL preclinical models. Our work provides proof of concept that within a specific oncogene driven cancer, the mechanism of oncogene dysregulation rather than the oncogene itself can identify clinically distinct patient subgroups and pave the way for future super-enhancer targeting therapy. Supplementary Information The online version contains supplementary material available at 10.1186/s12943-022-01701-x.


Background
For many years a major focus of cancer research has been the identification of genetic alterations leading to oncogene dysregulation. Crucially, this has led to the discovery of several targeted therapies such as the tyrosine kinase inhibitor Imatinib in Chronic Myeloid Leukemia (CML), therapies targeting HER2 receptor in breast cancers or BRAF V600E inhibitors in melanoma. Recent studies have evidenced that certain oncogenes have several distinct dysregulation mechanisms, including mutations in non-coding intergenic regions causing ectopic super-enhancer activation [1][2][3]. Whether different molecular mechanisms affecting oncogene dysregulation might have clinical implications remains unclear. An illustration of this would be the TAL1 (T-cell Acute Lymphocytic Leukemia Protein 1) gene. TAL1 is a major transcription factor dysregulated in more than 50% of T-ALL [4]. Although some effort to determine the clinical impact of TAL1 dysregulation in T-ALL has been made, conclusions remain contradictory [5,6]. Like several oncogenes, TAL1 can be overexpressed by chromosomal rearrangement placing its expression under the control of strong cis-regulatory elements. Most commonly reported is the SIL-TAL1 fusion transcript resulting from 90 kb interstitial microdeletions fusing the 5'portion of the gene to the 3'region of its neighboring gene STIL (SCL-interrupting locus) [7]. A much rarer chromosomal rearrangement involves the translocation of TAL1 into T-cell Receptor Delta (TCRD) and Beta (TCRB) loci accounting for only ~ 1-2% of T-ALL [1]. Besides chromosomal rearrangement, we and others discovered the second most recurrent dysregulation mechanism involving novel intergenic mutations upstream of the TAL1 promoter that lead to oncogenic super-enhancer formation [1,2]. These mutations nucleate the formation of the superenhancer by creating de novo MYB transcription factor binding sites. MYB in turn recruits transcription co-activators and the transcription factor complex in close proximity to the mutation, thus driving aberrant TAL1 expression. Here, we report a comprehensive study of the clinical importance and prognostic impact of 5'-TAL1 super-enhancer mutations in T-ALL. We demonstrate proof of concept that the mechanism of oncogene dysregulation rather than oncogene dysregulation itself can have significant clinical implications and that uncovering the molecular basis for oncogene dysregulation can pave the way to new therapeutic targets beyond direct pharmacological inhibition of oncogenes.

5% of T-ALL have 5'TAL1 super-enhancer (5'SE) mutations
The occurrence of 5'SE mutations was assessed in a large cohort of 443 unselected T-ALLs treated in the GRAALL-2003/2005 (Adult) and FRALLE-2000 (Pediatric) clinical trials. Sanger Sequencing identified 20 5'SE mutated patients accounting for 5% of T-ALL. The microinsertions were of variable size but all mapped to the same genomic position as previously reported [1,2] (1:47,239,295 hg38) and were predicted to create a neomorphic binding site for MYB transcription factor (Fig. 1A). Of note, we observed a comparable level of TAL1 expression between 5'SE, TAL1-TCRD and SIL-TAL1 rearranged T-ALLs (Fig. 1B, C), however due to the limited number of TAL1-TCRD cases (n = 4) preventing robust statistical analysis, we focused subsequent analyses on 5'SE and SIL-TAL cases. As expected, 5'SE mutations, and SIL-TAL1 were both mutually exclusive with other major T-ALL driver oncogenes such as TLX1, TLX3, CALM-AF10 (Supplementary Tables 1 and 2) [8,9]. Furthermore, analysis of 5'SE patients' oncogenetic landscape revealed a similar mutational profile to SIL-TAL1 patients with a low rate of co-mutations ( Fig. S1A-B + S2A + B).

Mebendazole demonstrates anti-leukemic activity in 5'SE T-ALL due to MYB-mediated TAL1 degradation
Mebendazole is a readily available and well tolerated anti-helminth drug that has anti-tumoral activity in a wide range of cancers, including in Acute Myeloid Leukemia via a MYB protein degradation mechanism [10].  (Fig. 1F + S4A + B). As expected, Mebendazole induced MYB degradation and reduced TAL1 mRNA and protein expression in 5'SE T-ALLs but had minimal effect on TAL1 expression in SIL-TAL1 T-ALLs despite MYB degradation (Fig. 1G). These results confirm 5'SE T-ALLs dependency on the MYB-TAL1 axis for their survival and identified a potential specific targeted therapy for 5'SE patients.

Mebendazole delays tumor progression in vivo
To test Mebendazole's ability to hinder leukemic progression in vivo, we injected NSG-mice with Jurkat cells carrying a native 12 bp 5'SE microinsertion [1] that we transduced to express the luciferase gene. Mice were administered Mebendazole in preventive and curative settings ( Fig. 2A). Treatment with Mebendazole delayed tumor progression in treated mice compared to control mice. Significantly less bioluminescence, reflecting bulk leukemic engraftment was detected in preventive (p = 0.03) and curative mice (p = 0.01) compared to vehicle control (Fig. 2B). Likewise, hCD45 staining of bone marrow cells revealed significantly reduced leukemic burden in preventive mice (p = 0.01) and curative mice (p = 0.02) compared to control mice (Fig. 2C). Furthermore, Mebendazole improved the overall survival of both preventive (p = 0.005) and curative (p = 0.001) treated mice compared to control mice (Fig. 2E). We also tested Mebendazole's efficacy in a more clinically relevant model using a 5'SE mutated PDX. Following the same treatment settings ( Fig. 2A), hCD45 peripheral blood staining similarly showed reduced leukemic burden in treated mice with significantly fewer leukemic blasts detected in preventive (p = 0.03) and curative mice (p = 0.02) compared to control (Fig. 2D). Importantly, Mebendazole administration resulted in a significantly improved survival in both treatment settings (preventive and curative p = 0.01) (Fig. 2F). Collectively these results demonstrate the efficacy of Mebendazole in reducing leukemic burden and delaying leukemic engraftment by specifically targeting MYB-dependent TAL1 + T-ALLs.

Discussion
The work presented here challenges the paradigm of cancer treatment which has hitherto focused on the identification of genetic alterations underlying expression and maintenance of malignant phenotypes. We have shown that the molecular mechanisms engendering oncogene dysregulation, rather than the level of oncogene overexpression itself, can identify subgroups of poor prognosis. Clinical Management of T-ALL remains a challenge, especially for relapsed/refractory T-ALL which are associated with extremely poor prognosis. Despite a high response rate after first-line therapy, about 20% of pediatric and 40% of adult T-ALL patients will suffer from relapse [11]. Even though clinical testing of targeted therapy has dramatically increased recently, such treatment options are limited for T-ALL due to the uniqueness of initiating events and oncogenic drivers implicated in T-ALL leukemogenesis. New approaches are needed for the design of personalized medicines in high-risk T-ALL [12]. Several oncogenes are considered 'undruggable' primarily because of their critical functions in developmental and physiological contexts, or because of technical constraints designing specific and efficient molecules [11,13]. Hence, a void exists in the development of effective personalized medicines for such high-risk T-ALL and oncogene-driven malignancies.
Our study has shown proof-of-concept that a mechanism of oncogene dysregulation such as the previously reported 5'SE [1,2], is associated with poor clinical outcome and can be efficiently targeted to suppress oncogenic signaling. While super-enhancer (SE) dysregulation sustains oncogenesis, it also creates an exploitable vulnerability. Among candidate targets, bromodomain and extra-terminal domain (BET) protein BRD4 has been implicated as a core component of SE activation in cancer. Its targeting has shown promising results in several hematological malignancies [14]. Other critical SE regulators such as CDK7 are currently being evaluated to disrupt SE-driven oncogene dysregulation [15]. In line with this, our study provides a strong rationale for the development of novel therapies targeting the dysregulation mechanism such as SEs to efficiently suppress oncogene driven cancers.

Conclusions
In this study we have shown that within a specific oncogene driven cancer the underlining molecular mechanism responsible for oncogene dysregulation can have significant clinical implications rather than the level of oncogene overexpression. Importantly, we demonstrate that Mebendazole can be repurposed to induce MYB mediated TAL1 degradation and induce cell death in MYB-dependent 5'SE T-ALLs, highlighting the importance of understanding the molecular basis of oncogene dysregulation which can uncover suitable targets such as the 5'SE, exploitable for the development of targeted therapy.