... J.E. Ricci (2026) Tumor metabolic adaptation induced by L-asparaginase reveals a vulnerability to PARP1/2 inhibitor in B-cell lymphomas

Tumor metabolic adaptation induced by L-asparaginase reveals a vulnerability to PARP one/two inhibitor in B-cell lymphomas

Secondary resistance to the amino-acid-depleting agent L-asparaginase remains poorly understood. Using ASNase-sensitive B-cell lymphoma models, we investigate tumor relapse during treatment. Through in vitro and in vivo metabolic profiling, here we show that ASNase triggers a metabolic reprogramming characterized by increased de novo serine biosynthesis driven by phosphoglycerate dehydrogenase. This response mitigates treatment-induced oxidative stress and associated DNA damage, enabling malignant cells to survive. We evidence that ASNase-treated malignant cells exhibit features of replication stress and increase activity of poly(ADP-ribose) polymerase, revealing a dependence on DNA repair. Combining ASNase with the clinically approved PARP inhibitor Olaparib enhances the antineoplastic effect of each monotherapy in vitro and in vivo. Moreover, this combination shows effectiveness in homologous recombination-proficient colorectal cancer cells, suggesting broader therapeutic potential. Overall, our study identifies tumor metabolic and genomic vulnerabilities induced by ASNase and supports a rational combination strategy using clinically approved drugs.

Cancer cells exhibit enhanced metabolism to meet expanding energetic and biosynthetic demands, a vulnerability that prompted several laboratories to develop metabolic inhibitors for clinical use in oncology. Despite successful evaluations on preclinical models of cancers, only a few of these inhibitors have achieved FDA approval, due in part to significant toxicities and development of resistance mechanisms. The intrinsic mechanisms by which tumor cells adapt to metabolic targeting in vivo remain poorly understood, thereby limiting the development of additional therapeutic strategies for patients who exhibit resistance to routinely used anti-metabolic agents in the clinic.

L-asparaginase is the only clinically approved drug targeting cancer cell addiction for a particular amino acid, successfully integrated into multi-agent chemotherapy regimens to treat both childhood and adult T-cell and B-cell acute lymphoblastic leukemias and NK/T-cell lymphomas. By catalyzing the hydrolysis of the non-essential amino acid asparagine in the bloodstream, ASNase triggers starvation and selective apoptosis of asparagine-addicted cancer cells. To date, clinical use of ASNase is restricted to ALL and NKTCL, as these hematological malignancies commonly harbor an epigenetic silencing of the gene encoding the asparagine synthetase that catalyzes the ATP-dependent conversion of L-aspartate and L-Glutamine into L-asparagine and L-Glutamate. Consequently, most ALL and NKTCL are asparagine-auxotrophic, which makes them highly vulnerable to ASNase treatment. Therefore, asparagine synthetase has long been considered a primary determinant of cancer cells' sensitivity to ASNase treatment. However, recent studies have shown that ASNase treatment also exhibits anti-tumor efficacy in a variety of other malignancies displaying basal heterogenous asparagine synthetase expression. Moreover, our recent work demonstrated that extracellular asparagine prevents its de novo biosynthesis in B-cell lymphomas, regardless of asparagine synthetase expression levels. These findings challenge the conventional view that asparagine synthetase expression alone is a reliable factor predicting tumor sensitivity to ASNase treatment in asparagine synthetase-expressing malignancies.

Several mechanisms involved in tumor resurgence during ASNase treatment have been described. Due to its bacterial nature, immunization against ASNase may occur during therapy, leading to neutralization of the enzyme activity. This issue is being addressed through the development of second generation of ASNase formulations with reduced immunogenicity. Additional mechanisms primarily involve the upregulation and/or activation of asparagine synthetase in malignant cells, driven by activation of multiple disdistinct signaling pathways that converge to an ATF4-dependent transcriptional program promoting cell growth. Nutrient release by stromal cells, which fuels leukemic blasts, along with specific tumor metabolic adaptations, also accounts for additional reported mechanisms of resistance to ASNase treatment. However, none have yet led to clinically viable alternative treatments for patients who fail ASNase therapy.

We recently demonstrated the antitumor efficacy of an ASNase-based anti-metabolic strategy in patients with refractory/relapsed diffuse large B-cell lymphoma. Despite complete responses achievement, all patients eventually relapsed during or after completed therapy, suggesting that a subset of malignant cells survived and adapted to circumvent treatment.

Our study aimed to uncover additional cancer cell intrinsic metabolic alterations that contribute to tumor relapse during ASNase treatment, with the intent to reveal unknown targetable vulnerabilities. Using a preclinical mouse model of B-cell lymphomas, we modeled secondary resistance following an initial antineoplastic response to ASNase treatment both in vitro and in vivo. Through comprehensive metabolomic profiling of B-cell lymphoma, complemented by in-depth analysis of altered metabolic pathways, using the stable isotope tracing in both in vitro and in vivo settings, we evidenced that ASNase treatment induces a dynamic metabolic reprogramming characterized by increased de novo serine biosynthesis in malignant cells. Using pharmacological inhibition, shRNA-mediated silencing, and overexpression of phosphoglycerate dehydrogenase-the rate-limiting enzyme of the de novo serine biosynthesis-we demonstrated that phosphoglycerate dehydrogenase plays a critical role in mediating the tumor's adaptive response to ASNase. Mechanistically, phosphoglycerate dehydrogenase activity mitigates ASNase-induced oxidative stress and related DNA damage, thereby enabling malignant cells to resume proliferation during treatment. Importantly, ASNase-induced DNA damage enhances reliance of malignant B cells on poly(ADP ribose) polymerase one/two activity, revealing a therapeutic vulnerability to the clinically approved PARP inhibitor, Olaparib, which impairs DNA repair mechanisms.

Results

ASNase-sensitive B-cell lymphomas undergo metabolic reprogramming during ASNase therapy both in vitro and in vivo

Previously, we demonstrated that B-cell lymphomas relying on oxidative phosphorylation metabolism for energy production are sensitive to E-coli L-asparaginase therapy, whereas glycolytic-dependent B-cell lymphoma exhibit resistance. In the present study, we engrafted wild-type C57BL/six mice with oxidative phosphorylation-dependent Eu-Myc cells (malignant B cells) isolated from two individual transgenic Eu-MycTg/plus mice. Seven days later, mice were treated either with Vehicle or ASNase every forty-eight hours until the lymphoma reached the ethical endpoint. ASNase treatment showed significant antitumor efficacy resulting in delayed B-cell lymphoma development. This represents the initial antineoplastic response to treatment. However, despite continued treatment, all mice eventually developed B-cell lymphoma, a result consistently observed across two independent Eu-Myc clones showing equivalent initial sensitivity to ASNase treatment in vivo. This secondary response reflects a phase of therapeutic failure, mimicking tumor relapse following an initial favorable response, as observed in patients with refractory/relapsed diffuse large B-cell lymphoma treated with innovative ASNase-based anti-metabolic therapy.

One mechanism of tumor resurgence during ASNase treatment in clinical settings is the production of anti-ASNase antibodies, which neutralize the enzyme and are often associated with allergic episodes in patients. To rule out decreased ASNase activity as the cause of tumor relapse, mice received a final bolus of Vehicle or ASNase at endpoint. Four hours post administration, BCL-bearing mice treated with ASNase exhibited undetectable plasma asparagine concentration and elevated plasma aspartate concentration compared to Vehicle-treated controls, indicating sustained ASNase activity despite disease relapse.

Regardless of the tumor entity or energetic dependency, therapy-resistant cancer cells frequently enhance mitochondrial energetic functions to promote survival. Consistently, we observed a significant increase in the contribution of OxPhos to ATP production in BCL that progressed during ASNase treatment in vivo, suggesting a metabolic reprogramming to meet elevated ATP demands.

We next conducted targeted metabolomic analysis in BCL undergoing Vehicle or ASNase treatment in vivo. Principal component analysis and Venn diagrams showed that the metabolic profile of BCL treated with ASNase in vivo was distinct from that of Vehicle-treated BCL, irrespective of the Eu-Myc clone used. Nineteen significantly deregulated metabolites were common to both datasets (Eu-Myc four hundred six cell- or Eu-Myc six hundred eighty-eight cell-derived BCL) and exhibited similar directional regulation in response to ASNase treatment, with fourteen metabolites increased and five metabolites decreased. These shared metabolites likely reflect core metabolic responses to ASNase therapy that are independent of the Eu-Myc clone origin, suggesting conserved metabolic vulnerabilities in our in vivo model.

Metabolite Set Enrichment Analysis using the Kyoto Encyclopedia of Genes and Genomes pathway computational tool, revealed deregulated nucleotide and amino acid metabolism in ASNase-treated BCL. Among proteinogenic amino acids, serine, glycine, tyrosine, and asparagine were the only significantly and consistently deregulated amino acids in ASNase-treated BCL, meeting the fold change threshold (ASNase/Vehicle) greater than one point two five and raw P-value less than zero point zero five, across both datasets. Indeed, BCL treated with ASNase in vivo exhibit a two- to three-fold increase in steady-state serine and glycine levels in ASNase-treated BCL compared to Vehicle-treated BCL.

To confirm that ASNase treatment triggered serine metabolism remodeling in malignant cells, Eu-Myc cells were cultured in DMEM medium lacking asparagine but containing supraphysiological concentrations of serine and glycine, with (+) or without (-)

supplementation of asparagine and ASNase. A concentration of zero point zero zero three International Units per milliliter ASNase was used to avoid hydrolysis of glutamine into L-Glutamate. ASNase treatment significantly increased the death of Eu-Myc cells in twenty-four hours. However, Eu-Myc cells surviving this nutritional stress resumed proliferation over time, a process accompanied by a significant increase in intracellular serine and glycine levels following twenty-four hours of ASNase treatment. To specifically evaluate metabolic adaptations to asparagine deprivation-since ASNase-mediated release of aspartate and ammonium can each influence the tumor metabolic response-we cultured Eu-Myc cells in asparagine-free medium. This experimental setup ensured that the observed metabolic changes were directly attributable to asparagine depletion, thereby reinforcing the robustness of our conclusions. Our findings indicate that ASNase treatment induces a metabolic response in BCL, both in vitro and in vivo, characterized by elevated serine and glycine levels in malignant B cells.