Introduction:
Acute lymphoblastic leukemia (T-ALL) is an aggressive hematopoietic malignancy in children and adolescents that is associated with high rates of treatment failure and early relapse. T-ALL patients frequently harbor NOTCH1 activating mutations as the driving oncogene in this disease. A multitude of strategies preventing NOTCH1 cleavage and activation, such as Gamma-secretase inhibitors (GSIs) have been developed. Despite promising pre-clinical data, the rapid development of Notch1 inhibitor resistance in early clinical trials, prevented the translation of these inhibitors into the clinical setting. In previous work, our group demonstrated that T-ALL resistant to NOTCH1 inhibition carry altered epigenetic states conferring unique dependency on epigenetic modifiers, such as BRD4. The goal of this study was to study enhancer rewiring in Notch1 inhibitor resistant T-ALL in vivo and its relationship to apoptotic priming.
Methods:
After reaching 5% of circulating leukemic blasts, five established T-ALL PDX models with aberrant NOTCH1 expression were divided into to two treatment groups each (8 mice per group). 1 group received the Notch inhibitor DBZ (Dibenzazepine; 10 μM/kg intraperitoneal every other day) and the other group was treated with vehicle. Short-term effect of DBZ in vivo was assessed after 1 week of treatment, when 3 mice per group were sacrificed and leukemic blasts were isolated from spleen and bone marrow. The remaining 5 mice were monitored for disease burden (by flow cytometry staining for human CD45+) and followed for survival. After reaching moribund state, animals were sacrificed, spleens and bone marrows were collected and prepared for further analyses.
To assess DBZ efficacy in vivo, the presence of active NOTCH1 (ICN1) in spleen and bone marrow was analyzed by Immunohistochemistry analysis (IHC).
Enhancer landscapes were identified by chromatin-immunoprecipitation followed by sequencing (ChIP-Seq) for Histone 3 Lysine 27 acetylation (H3K27ac). A custom computational pipeline that incorporates algorithms for demultiplexing, alignment, normalization, peak calling, and computation of signal intensities within peaks was used to call differential peaks and intersect with RNA-sequencing results.
BH3 profiling was performed on leukemic blasts isolated from spleen to measure overall mitochondrial priming and to identify anti-apoptotic dependencies.
Results:
In four out of five T-ALL PDX models, IHC analysis of spleen and bone marrow demonstrated a drastic downregulation of active NOTCH1 upon DBZ treatment, validating the efficacy of the used inhibitor. Weak ICN1 staining that remained unchanged upon DBZ treatment, was observed in 1 of the models, resulting in the exclusion of this strain from further functional analysis.
Survival analysis of the four T-ALL PDX models expressing ICN1, revealed the presence of two Notch inhibitor sensitive and two refractory strains. The latter strains developed DBZ resistance rapidly after starting treatment (less than 10 days). One sensitive strain eventually developed resistance, while the second showed long-term disease control. Transcriptional profiling (bulk RNA-seq) of Notch inhibitor refractory strains versus sensitive identified the intrinsic apoptotic pathway as one of the most deferentially deregulated GSEA signatures.
H3K27ac ChIPseq analysis at pretreatment (baseline), showed increased signal intensity of H3K27ac peaks at BCL2 and MCL1 enhancers in the refractory strains compared to sensitive. Upon DBZ treatment, while the enhancer state in refractory T-ALL remained unchanged, in the sensitive strains the signal intensity of H3K27ac peaks within the BCL2 and MCL1 loci decreased. Mitochondrial BH3 profiling at baseline demonstrated BCL-2 dependency (measured via BAD peptide) in sensitive strains and MCL-1 dependency (measured via MS1 peptide) in refractory strains. Upon DBZ treatment, sensitive strains showed a decrease in BCL-2 dependency and compensatory switch to MCL1-dependency, while dependency profile remained unchanged in refractory T-ALL.
Conclusions:
Our results suggest that enhancer rewiring near anti-apoptotic genes is critical for Notch inhibitor resistance. Combining BH3 profiling with enhancer profiling may allow to predict drug responses in vivo and may contribute to the identification of novel therapeutic targets for combination therapy in resistant disease.
Letai:Zeno Pharmaceuticals, Vivid Bioscience, Flash Therapeutics, Dialectic Therapeutics: Membership on an entity's Board of Directors or advisory committees, Other: Cofounder or Advisory Board member; AbbVie, AstraZeneca, Novartis: Consultancy, Research Funding. Weinstock:Celgene: Research Funding. Lohr:T2 Biosystems: Honoraria; Celgene: Research Funding.
Author notes
Asterisk with author names denotes non-ASH members.