Enrichment in Metabolic Pathway Activation Corresponds to Immunoglobulin Gene Diversity across B Cell Developmental Stages in B-Lymphoblastic Leukemia

Abstract

Background: Intraclonal genetic variation in the immunoglobulin heavy chain variable gene (IGHV) has been widely documented in B-lymphoblastic leukemia (B-ALL), but its biologic significance as a presumed reflection of the B cell stage of leukemic transformation has not been systematically evaluated. Using ultradeep, targeted sequencing of the rearranged IgH gene in 22 patients with B-ALL, we previously demonstrated that there are subsets of patients with either extensive or minimal diversity in recombination activating gene (RAG)-mediated variable (V), diversity (D), and joining (J) gene rearrangements and that gene expression profiling (GEP) reveals enrichment of oxidative phosphorylation (OXPHOS), glycolysis, and fatty acid metabolism Hallmark pathways among cases with V-DJ rearrangement diversity (Fries et al. ASH 2022). We observed enrichment in these same metabolic signatures among leukemia cells with a pre-B cell phenotypic signature found to be predictive of relapse (Good et al. Nat Med 2018; Liu et al. ASH 2021). To understand the functional implications of this metabolic activation observed among cases with IgH diversity, we applied cytometry by time of flight (CyTOF) phenotyping to determine if IgH clonal composition reflects the metabolic and proliferative activation inherent to B cell stage of origin which may have relevance to disease behavior and outcome. Methods: Using viably-frozen, pre-treatment bone marrow (BM) and/or peripheral blood (PB) biospecimens from a cohort of 18 patients with B-ALL with known IgH clonal composition (7 with extensive V-DJ diversity, 8 without diversity, and 3 lacking any dominant clonal sequence), we performed deep phenotyping by mass cytometry and compared signaling pathway activation patterns between cohorts and compared to healthy controls. Samples were analyzed by mass cytometry with a 40-marker panel including B cell developmental phenotypic and metabolic proteins. Developmental classification was performed as previously described (Good et al. 2018). To test statistical significance in frequency of each subpopulation between two groups, we applied a multiple unpaired t-test with correction by Holm-Sidak method. Results: We discovered that all leukemia samples - regardless of IgH clonal composition - were significantly enriched for phenotypic features reflecting a pro-BII cell state, whereas enrichment of other more or less mature cell stages varied according to IgH composition. Developmental classification demonstrated that cases lacking a clonal IgH rearrangement (N=3) - presumed to have derived from a B cell stage preceding an initial DJ joining event - were enriched at the pro-BII population (P<0.0001), as were cases with extensive V-DJ diversity (N=7; P=0.019), and those lacking in diversity (N=8; P=0.0007) compared to healthy BM. However, cases with V-DJ diversity were uniformly enriched for OXPHOS, glycolytic, and pentose phosphate pathway (PPP)-associated protein expression compared to those without V-DJ diversity. In cells with a pro-BII phenotype, differentially expressed proteins between diverse and non-diverse cases included proteins essential for glycolysis (GLUT1, ENO1, PKM1), PPP (PGD, TKT), and OXPHOS (SDHA, CS, ATPA5, CytC). Conclusions: Cellular metabolism is a known hallmark of cancer proliferation, leukemia stem cell maintenance, and therapeutic escape, just as genetic diversification is integral to tumor evolution. Using primary B-ALL specimens, we show that metabolic pathway activation coincides with extensive RAG-mediated IGHV gene diversification across B cell developmental phenotypes. Contrary to the canonical assumption in B-ALL that V-DJ rearrangement and diversity is tied to developmental state, our analysis suggests that all cases are enriched at the proB-preB transition, but with differences in cellular metabolism that distinguish the diverse from homogeneous rearrangement state. These data suggest that IgH composition is not merely a passive genetic signature exploitable for tracking purposes in B-ALL, but may also inherently reflect underlying biology relevant to clinical outcomes.