NDUFS1 links mitochondrial respiration to aggressive breast cancer behavior

Elevated expression of the mitochondrial protein NDUFS1 was associated with advanced breast cancer, poorer patient outcomes and aggressive cellular behavior in a study combining patient data, breast cancer cell models and orthotopic mouse tumors. The research identified a signaling route through which NDUFS1 supports proliferation, migration, invasion and stemness-associated phenotypes: increased mitochondrial respiration promotes the PROX1–c-Myc pathway and reduces the cell-cycle inhibitor p21. The authors describe the findings as preclinical evidence and caution that the results do not establish a clinical treatment strategy for patients with breast cancer.

NDUFS1 is the largest subunit of complex I in the mitochondrial electron transport chain, where it is essential for normal complex I function. The chain supports oxidative phosphorylation, the process by which mitochondria generate much of the cellular energy used for growth and division. The study notes that cancer cells often favor glycolysis even when oxygen is available, a pattern commonly called the Warburg effect, but mitochondrial oxidative phosphorylation can remain intact or become elevated in particular tumor contexts. Breast cancer, the authors write, is heterogeneous in its reliance on mitochondrial metabolism, making it important to determine whether particular respiratory components contribute to disease progression.

The investigators first examined NDUFS1 expression in publicly available breast cancer datasets and in tissue samples collected at Kaohsiung Medical University Hospital in Taiwan. NDUFS1 messenger RNA levels were higher in breast tumors than in normal breast tissue in the Oncomine analysis. An analysis of the harmonized TCGA-TARGET-GTEx dataset also showed higher expression across breast cancer subtypes, compared with 178 normal breast samples. Immunohistochemical analysis of hospital specimens showed higher NDUFS1 protein expression in cancer tissue than in normal tissue and an additional increase in later-stage tumors. In the hospital cohort, high NDUFS1 expression was significantly associated with stage, T stage and lymph-node metastasis; adjusted odds ratios were 3.36 for T2–T4 versus T1 disease and 6.34 for N1–3 versus N0 disease.

Survival analyses produced a parallel clinical pattern. In publicly available breast cancer microarray data, patients with high NDUFS1 messenger RNA expression had significantly poorer recurrence-free and overall survival. The same association appeared in overall-survival data from TCGA-BRCA. Multivariable Cox analysis indicated that high NDUFS1 expression remained independently associated with worse overall survival after adjustment for age, stage and breast cancer subtype, or for age, TNM stage and receptor status. These observations are associations rather than evidence that NDUFS1 directly causes poor outcomes, and the study combined several different datasets and a single-center tissue cohort.

Functional experiments were performed in two human breast cancer lines: ER-positive MCF-7 cells, used as a luminal A model, and triple-negative MDA-MB-231 cells. Reducing NDUFS1 with short hairpin RNA significantly reduced XTT-measured proliferation, migration and invasion in both lines. The opposite effects occurred when NDUFS1 was overexpressed. In MCF-7 cells, NDUFS1 knockdown also reduced tumorsphere formation and lowered the expression of KLF4, c-Myc and SOX2, proteins associated with stemness-related features. NDUFS1 overexpression increased c-Myc in both cell lines. The association also appeared in patient material: NDUFS1 expression correlated positively with MYC messenger RNA in TCGA-BRCA tumors and with c-Myc protein in immunohistochemical measurements.

Respiration measurements indicated that these effects were connected to mitochondrial function. NDUFS1 knockdown reduced basal oxygen consumption, oxygen consumption rate, ATP production, maximal respiration and spare respiratory capacity in MDA-MB-231 and MCF-7 cells. Overexpression increased basal oxygen consumption and ATP production. Extracellular acidification also declined after knockdown, rather than rising as a compensatory sign of increased glycolysis. To test whether respiration contributed to the malignant phenotypes, the researchers used metformin, a mitochondrial complex I inhibitor. At a concentration reported not to affect cancer-cell viability, metformin reversed the increases in proliferation and migration caused by NDUFS1 overexpression. This supports a functional role for complex I activity, although metformin can also affect other signaling pathways and the study did not establish that the drug acted exclusively through NDUFS1.

Bulk RNA sequencing of MDA-MB-231 cells after NDUFS1 knockdown was used to search for a mechanism connecting respiration to the observed behaviors. Three genes associated with migration, stemness or proliferation—PROX1, MEF2C and WNT5A—were reduced after knockdown, with PROX1 selected for further analysis because its protein level changed consistently with NDUFS1 in both breast cancer lines. NDUFS1 and PROX1 expression were positively correlated in TCGA-BRCA data and in tumor immunohistochemistry. In cells, reducing PROX1 lowered c-Myc, whereas reducing c-Myc did not lower PROX1, placing PROX1 upstream of c-Myc in the proposed pathway. Knockdown of either protein increased p21, while metformin reversed the NDUFS1-associated increases in PROX1 and c-Myc. The authors therefore proposed an NDUFS1–PROX1–c-Myc–p21 axis, but acknowledged that the precise metabolic link remains unresolved.

Several possible mechanisms were considered. Mitochondrial oxidative phosphorylation can affect metabolite availability and histone acetylation, and reduced citrate production could theoretically contribute to decreased PROX1 expression. Changes in reactive oxygen species and redox-sensitive pathways were another possibility, but the direction of known effects in those pathways did not fully match the observations. Ferroptosis, an iron-dependent form of programmed cell death associated with lipid peroxidation, was also raised as a potential contributor because NDUFS1 participates in iron–sulfur cluster biology. The authors explicitly identified this as a question for future experiments rather than a demonstrated mechanism. They likewise noted that metformin may influence AMPK–mTOR, TGF-β and prostaglandin-related signaling, so the cell experiments do not prove that every effect runs through the PROX1–c-Myc pathway.

The findings were extended to a syngeneic orthotopic model using luciferase-expressing 4T1 mouse breast cancer cells implanted in the mammary fat pads of BALB/c mice. Ndufs1 knockdown reduced proliferation, migration, invasion and tumorsphere formation in the cells before implantation. Mice receiving the knockdown cells developed tumors with significantly lower bioluminescence, final volume and weight than controls. NDUFS1, PROX1 and c-Myc protein levels were lower in the resulting tumors, and tumor-level NDUFS1 expression correlated with both PROX1 and c-Myc. No obvious metastases were detected by bioluminescence imaging, which the authors said could reflect an immune response to the xenogeneic luciferase reporter and therefore limited metastasis assessment.

Giving the complex I inhibitors metformin or rotenone intraperitoneally five times per week for four weeks reduced bioluminescence, tumor volume and tumor weight in control mice to levels comparable to those in the Ndufs1-knockdown group. Immunohistochemistry also showed lower NDUFS1, PROX1 and c-Myc in treated tumors. These mouse experiments support a tumor-suppressing effect of complex I inhibition in this model, but they do not demonstrate clinical efficacy in people. The study itself cites a randomized trial involving 3,649 patients that did not find a benefit from metformin for invasive disease-free survival in breast cancer. NDUFS1 expression may eventually help identify metabolically high tumors for appropriately designed trials, but biomarker-guided treatment remains a hypothesis requiring prospective testing.

Overall, the study provides mechanistic and preclinical evidence that NDUFS1 supports breast cancer malignancy by sustaining mitochondrial oxidative phosphorylation and enhancing PROX1–c-Myc signaling. Its strongest links are the consistent changes in tumor and cell phenotypes after NDUFS1 manipulation, the respiratory measurements and the suppression of those effects by complex I inhibition. Important gaps remain: the study did not directly demonstrate that PROX1 and c-Myc are required for NDUFS1-driven proliferation, migration or invasion; it did not establish how reduced respiration alters PROX1; and it did not test the drugs’ effects on breast cancer metastasis in vivo. The work therefore identifies NDUFS1 as a candidate biomarker and therapeutic vulnerability, rather than a ready-made target or a replacement for established breast cancer treatment.

Subject of Research: Preclinical investigation of NDUFS1-driven mitochondrial respiration and breast cancer progression

Article Title: NDUFS1 promotes malignant breast cancer behaviors through enhancement of mitochondrial respiration and PROX1/c-Myc signaling

Article References: Wang, Y.-Y., Chen, P.-Y., Vinzons, L. U., Tseng, C.-H., Chen, Y.-K., Chu-Sung Hu, S., Lo, S., Hou, M.-F., Wu, C.-H., Chen, Y.-C., & Yuan, S.-S. F. (2026). NDUFS1 promotes malignant breast cancer behaviors through enhancement of mitochondrial respiration and PROX1/c-Myc signaling. iScience, 29(11), Article 117942. https://doi.org/10.1016/j.isci.2026.117942

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117942

Keywords: breast cancer, NDUFS1, mitochondrial respiration, oxidative phosphorylation, PROX1, c-Myc, complex I inhibition, metformin, tumor growth, promotes, malignant, breast