One Enzyme, Two Faces: ASRGL1 Emerges as a Context-Dependent Player in Cancer

A single metabolic enzyme can behave like an accomplice of the tumor in one organ and like a protective factor in another. That paradox sits at the center of a new review by researchers at Nanjing Medical University and Soochow University, published on 21 October 2026 in Frontiers in Cell and Developmental Biology. The authors, Huiping Zeng, Zekang Li, Renhao Geng and Lujun Chen, systematically assembled the evidence on asparaginase and isoaspartyl peptidase 1, or ASRGL1, a human enzyme that has long lived in the shadow of its famous bacterial cousins used to treat leukemia. Their synthesis shows that ASRGL1 is aberrantly expressed in a striking range of malignancies, from nasopharyngeal and cervical cancer to hepatocellular, colorectal and renal cell carcinoma, yet its biological and prognostic implications vary substantially between tumor types. The review positions ASRGL1 as a candidate biomarker whose meaning is inseparable from tissue context, molecular subtype and metabolic state.

To understand why the enzyme matters, it helps to start with its biochemistry. The ASRGL1 gene sits on chromosome 11q12.3 and spans seven exons and six introns, encoding a 308-amino-acid protein of 31.9 kilodaltons that belongs to the N-terminal nucleophilic hydrolase superfamily. The protein is produced as an inactive precursor and must activate itself: intramolecular processing between the residues Gly167 and Thr168 exposes a new N-terminal threonine, Thr168, which becomes the catalytic nucleophile and generates the mature enzyme with its conserved alpha-beta-beta-alpha fold. Once active, ASRGL1 carries two enzymatic capacities. Its L-asparaginase activity hydrolyzes L-asparagine into L-aspartate and ammonia, and its beta-aspartyl peptidase activity cleaves isoaspartyl dipeptides, damaged residues that arise from non-enzymatic protein wear and can drive misfolding and aggregation. L-asparaginase therapy has been a mainstay of acute lymphoblastic leukemia treatment for decades, but human ASRGL1 itself has mainly been studied as a potential alternative enzyme and has never been established as a clinical agent.

Beyond cancer, the enzyme’s day job appears to be protein homeostasis. By recycling amino acids and clearing isoaspartyl damage, ASRGL1 helps maintain normal function of the nervous and reproductive systems, and it has been suggested to cooperate with isoaspartyl methyltransferase in preventing the accumulation of toxic modified peptides in mammalian brain and other tissues. In mouse models, mutation or loss of the enzyme leads to retinal degeneration, and it has been associated with infertility and neurodegenerative changes. In normal human tissue, ASRGL1 is predominantly expressed in the testis, brain, prostate and proliferating endometrium. Against that baseline, the review documents aberrant expression in nasopharyngeal carcinoma, breast cancer, hepatocellular carcinoma, colorectal cancer, renal cell carcinoma, cervical cancer, prostate cancer and endometrial cancer, with expression trends, functional tendencies and prognostic associations that differ sharply from one tumor to the next.

Hepatocellular carcinoma provides the most consistent pro-tumor picture. ASRGL1 levels are significantly higher in liver tumor tissue than in adjacent normal tissue, and the upregulation positively correlates with pathological grade, T stage, clinical stage and lymph node metastasis. Kaplan-Meier analyses reported in the underlying studies linked high ASRGL1 expression to poorer recurrence-free survival, progression-free survival, disease-specific survival and overall survival. A ceRNA network analysis identified ASRGL1 as one of 29 genes tied to survival and prognosis in this cancer, closely correlated with the infiltration levels of multiple immune cell types. Co-expression and gene set enrichment analyses further connect ASRGL1-associated genes to DNA replication, microtubule cytoskeleton organization, spliceosome development and homologous recombination. Mechanistically, work in hepatocellular carcinoma cells showed that ASRGL1 knockdown promotes p53 protein expression, reduces Cyclin B and Cyclin-dependent kinase 1, and disrupts the interaction between those two cell-cycle partners, arresting cells in the G2/M phase.

Breast cancer tells a more hormone-entangled story. An estrogen response element has been identified in the promoter region of the ASRGL1 gene, through which sex hormones induce ASRGL1 expression in BT474 breast cancer cells. In an immunohistochemical study of 148 invasive ductal carcinoma samples and 105 non-neoplastic breast specimens, ASRGL1 levels were higher in estrogen receptor and progesterone receptor positive tumors than in normal breast tissue, while expression was downregulated in receptor-negative tumors. Yet ASRGL1 expression was not significantly associated with tumor size, grade, stage, lymph node status or distant metastasis, nor with disease-free or overall survival, limiting its value as a standalone prognostic marker. Nasopharyngeal carcinoma is even more contentious: one research group detected high ASRGL1 expression in tumor tissues by RT-qPCR and western blotting, while a single-cell transcriptomic and metabolomic analysis found the enzyme primarily expressed in fibroblasts, epithelial cells and myeloid cells at levels significantly lower than in adjacent normal tissue, with low expression tied to the malignant phenotype. In cervical cancer, by contrast, ASRGL1 is consistently higher than in adjacent normal tissue and was identified through RNA-seq screening and RT-qPCR validation as a potential biomarker of malignant progression in cervical lesions.

Colorectal cancer exposes the sharpest contradiction. In functional assays, ASRGL1 messenger RNA was significantly higher in the highly metastatic KM12L4A cell line than in moderately metastatic KM12SM cells, and knockdown markedly suppressed proliferation of the highly metastatic cells while sparing the others, suggesting a selective growth advantage. Yet bioinformatic analyses of patient datasets point the other way: ASRGL1 emerged among Th17 cell-related hub genes from the TCGA database, was highly expressed in a low-risk group enriched for lipid and vitamin metabolism pathways, and higher expression correlated with favorable prognosis. Renal cell carcinoma follows a similar protective pattern. In a six-gene metabolism-related prognostic model built from TCGA data, ASRGL1 was downregulated in tumor subsets with a low-metabolism phenotype and poor prognosis, and immunohistochemistry showed strong or moderate staining in normal kidney tissue but markedly reduced intensity in tumors, with higher expression correlating with longer survival. In prostate cancer, an eight-gene risk-scoring model predicted biochemical recurrence better than any single clinical parameter, and high ASRGL1 expression was significantly associated with poor prognosis, positive correlation with TNM stage, Gleason score and lymph node invasion, and possible links to amino acid metabolism or the TMPRSS2-ERG fusion subtype.

Endometrial cancer offers the most clinically developed dataset. ASRGL1 is highly expressed in normal endometrium but heterogeneously lost in endometrioid endometrial carcinoma, and low expression is associated with higher pathological stage, grade, lymph node metastasis, depth of myometrial invasion and worse disease-specific survival. A two-cohort study confirmed this pattern, and a prospective cohort study found ASRGL1 highly expressed in premalignant lesions but frequently low or absent in metastatic lesions, with an 85 percent concordance rate between preoperative curettage and postoperative surgical specimens, a basis for preoperative risk assessment. LASSO regression identified combined ASRGL1 and p53 immunohistochemistry as an independent survival predictor, and a five-gene model incorporating ASRGL1 was constructed for lymph node metastasis risk in early-stage disease. However, a retrospective analysis of 775 patients found that molecular classification was the only independent prognostic factor, and the independent prognostic value of low ASRGL1 expression was attenuated after multivariable adjustment, a caution that the review’s authors treat as central rather than incidental.

The mechanistic chapters of the review weave these threads together. In cervical cancer SiHa cells, ASRGL1 knockdown caused S-phase arrest, reduced CDK2 and cyclin A2, increased the pro-apoptotic protein Bax and decreased Bcl-2, implicating the mitochondria-mediated apoptotic pathway. In nasopharyngeal carcinoma cell lines, silencing ASRGL1 elevated early and late apoptosis, increased cleaved caspase-3 and cleaved caspase-9, and, for the first time, triggered reactive oxygen species accumulation and ferroptosis, with altered ACSL4 and FTH1 expression and decreased glutathione peroxidase 4. In hepatocellular carcinoma, overexpression of CDK1 reversed the proliferation, migration and invasion losses caused by ASRGL1 knockdown, supporting a CDK1-dependent axis. Metabolically, ASRGL1 depletion reduced ATP levels and blunted amino acid utilization in nasopharyngeal carcinoma cells; in endometrial cancer, low ASRGL1 cooperated with high PHGDH expression in the glutamine-to-asparagine pathway, and a PHGDH inhibitor suppressed proliferation and migration of low-ASRGL1 cells. Correlative analyses also tie ASRGL1 expression to B cell, CD4+ T cell, macrophage, neutrophil and dendritic cell infiltration in hepatocellular carcinoma, where high expression conferred a 1.56-fold increased risk of overall survival deterioration.

The authors are careful about what the evidence can and cannot support. Although ASRGL1’s catalytic properties are biochemically well established, current cancer studies have not conclusively shown that its enzymatic activity directly drives the observed malignant phenotypes; changes in ATP or metabolomic profiles after depletion establish association, not catalytic causality, and catalytic-mutant and metabolite-rescue experiments will be needed. The immune findings are primarily correlative, upstream regulation of ASRGL1 expression is largely unknown, and no small-molecule inhibitors or specific antibodies have yet been developed or validated in preclinical models. Most prognostic evidence derives from retrospective bioinformatic analyses of bulk-tissue data, so prospective validation and integration with established molecular classifiers are required before clinical use. Whether ASRGL1 ultimately proves to be a therapeutically actionable vulnerability or primarily a context-dependent biomarker remains an open question, but the review makes a compelling case that this self-activating hydrolase deserves a far more prominent place on the cancer metabolism map.

Subject of Research: Context-dependent roles and mechanisms of the enzyme ASRGL1 in tumorigenesis and cancer progression

Article Title: Emerging roles and mechanistic insights of ASRGL1 in tumorigenesis and cancer progression

Article References: Zeng, H., Li, Z., Geng, R., & Chen, L. (2026). Emerging roles and mechanistic insights of ASRGL1 in tumorigenesis and cancer progression. Frontiers in Cell and Developmental Biology, 14, Article 1955895. https://doi.org/10.3389/fcell.2026.1955895

Image Credits: AI Generated

DOI: 10.3389/fcell.2026.1955895

Keywords: ASRGL1, asparagine metabolism, metabolic reprogramming, tumor biomarker, hepatocellular carcinoma, endometrial cancer, cell cycle, ferroptosis, tumor microenvironment, prognosis, Ntn hydrolase, cancer biology