Murine cervical cancer cell study links KNG1–S1P axis to reprogrammed myeloid-derived suppressor cells

A multi-omics study combining immune infiltration analysis, mouse cell models, proteomics, metabolomics and functional experiments found that cervical cancer cell-conditioned medium altered the metabolism of myeloid-derived suppressor cells, with changes in sphingolipid metabolism converging on a kininogen-1–sphingosine 1-phosphate network. These preclinical findings support the KNG1–sphingosine 1-phosphate axis as a potential target for further investigation, rather than establishing a clinical treatment.

Researchers first used TIMER 2.0 to estimate immune-cell infiltration in TCGA cervical cancer data. They combined the infiltration estimates with survival and gene set enrichment analyses, which identified MDSCs as an important component of the cervical cancer immune microenvironment. The analysis also associated high MDSC infiltration with poor prognosis. Gene set enrichment analysis showed lower MDSC infiltration in the high-immunological-score group than in the low-immunological-score group, with a normalized enrichment score of −2.214, an enrichment score of −0.5199 and P<0.001.

Clinical tissue analysis supported the association between MDSCs and cervical cancer. Investigators examined 50 paraffin-embedded cervical cancer tissues and 10 normal cervical tissues collected at the First Affiliated Hospital of Sun Yat-sen University between January 2015 and December 2017. Flow cytometry detected CD33-positive, CD11b-positive MDSCs in 4.4% of cervical cancer samples and 0.13% of normal cervical samples, with three normal tissues and 10 cancer tissues included in that comparison. Immunofluorescence analysis of all 60 tissue samples found an approximately twofold increase in MDSCs in cervical cancer relative to normal tissue.

The experimental work used mouse bone marrow–derived MDSCs exposed to supernatants from the murine U14 and HPV-16 E6/E7-expressing TC1 cervical cancer cell lines. These conditioned groups were designated U14_Co and TC1_Co. CCK-8 assays showed greater proliferation in both conditioned groups than in controls, while transwell assays showed increased migration. Quantitative reverse-transcription PCR also showed higher inducible nitric oxide synthase, or iNOS, expression in the tumor-conditioned MDSCs. These assays model interactions between cancer-cell factors and MDSCs in vitro and do not establish the same effects in patients.

Proteomic profiling of control, TC1_Co and U14_Co MDSCs, with three samples per group, identified shared changes in Serpinc1, Kng1 and Slpi. Gene Ontology enrichment in both conditioned comparisons included terms related to lipoprotein-lipase activity and protein–lipid complexes, pointing to altered lipid metabolism. Metabolomic analyses found enrichment across several lipid-related pathways, including sphingolipid metabolism, linoleic acid metabolism, glycerophospholipid metabolism, fatty acid biosynthesis and arachidonic acid metabolism. Sphingolipid metabolism, fatty acid biosynthesis and glycerophospholipid metabolism were shared findings in the metabolic analyses of both conditioned systems.

Integrated protein–metabolite analysis yielded 73 common proteins and 132 metabolites across the two comparisons. Glycerophospholipid metabolism was enriched in both systems, but no shared differentially expressed proteins could be linked to that pathway across the models. By contrast, sphingolipid metabolism contained shared protein changes that converged on sphingosine 1-phosphate, or S1P. The network connected S1P with proteins including AKT2, Cers6, Degs1, Pten and Kng1 in the TC1_Co model and with Sphk2, Degs1, Psap and Gba1 in the U14_Co model. Kng1 was the only candidate in these lists that was upregulated in both conditioned groups in the proteomic data, and western blotting confirmed increased KNG1 protein expression.

The researchers then performed gain- and loss-of-function experiments in the tumor-conditioned mouse MDSCs. Small interfering RNA was used to knock down Kng1, while a plasmid was used to overexpress it, with functional assays conducted 48 hours after transfection. KNG1 knockdown reduced proliferation, migration, iNOS expression and S1P concentration in both TC1_Co and U14_Co MDSCs. Conversely, KNG1 overexpression increased proliferation and rescued the reductions in migration and iNOS expression caused by depletion. ELISA measurements of culture supernatants showed that knockdown also reduced S1P levels. These experiments support a functional role for KNG1 in the measured MDSC phenotypes and S1P production in the two cell models.

The study has important limitations. Its conditioned-medium system used supernatants from murine cervical cancer cell lines and did not reproduce the full cellular interactions or three-dimensional architecture of an intact tumor microenvironment. The direct mechanism by which KNG1 influences S1P production was not resolved; the authors note that it could involve sphingosine kinases, substrate availability or indirect signaling, and those possibilities remain to be tested. The study therefore identifies a mechanistically supported KNG1–S1P network in preclinical MDSC models, but its clinical relevance will require validation in orthotopic tumor models, patient-derived samples and human contexts.

Overall, the research links cervical cancer cell conditioning to enhanced MDSC proliferation, migration and iNOS expression in vitro. The multi-omics analysis points to sphingolipid metabolic reprogramming as a feature of that response, while the functional experiments identify KNG1 and S1P as a connected regulatory axis. The authors nominate this axis as a potential therapeutic target for cervical cancer, but the current evidence comes from database analyses, clinical tissue comparisons and murine cell experiments rather than a demonstrated treatment effect in patients.

Subject of Research: Preclinical multi-omics study of sphingolipid metabolic reprogramming in cervical cancer-conditioned myeloid-derived suppressor cells

Article Title: Multi-omics profiling reveals sphingolipid metabolism reprogramming of tumor-conditioned MDSCs in cervical cancer

Article References: Mai, Q., Chi, C., Wang, X., Chen, Y., Zou, Q., Chen, Q., Zeng, F., Wei, M., Chen, Y., Wang, A., Liao, Y., Xiao, Y., Li, X., Yan, Q., Zhan, L., Wei, H., Jing, X., Du, Q., & Liu, J. (2026). Multi-omics profiling reveals sphingolipid metabolism reprogramming of tumor-conditioned MDSCs in cervical cancer. BMC Cancer, 26(1), Article 1122. https://doi.org/10.1186/s12885-026-16914-2

Image Credits: AI Generated

DOI: 10.1186/s12885-026-16914-2

Keywords: cervical cancer, myeloid-derived suppressor cells, sphingolipid metabolism, KNG1, sphingosine 1-phosphate, multi-omics, tumor microenvironment, immune suppression, mouse cell models, profiling, reveals, sphingolipid