Experimental AMPK activator traps esophageal cancer cells in S phase, priming them for chemotherapy

Esophageal cancer carries one of the bleakest prognoses in oncology: it ranks as the sixth leading cause of cancer-related mortality worldwide, five-year survival falls below twenty percent, and chemotherapy for advanced disease is frequently undermined by drug resistance. A new cell-culture study published in Frontiers in Oncology reports that AICAR, a well-characterized pharmacological activator of the cellular energy sensor AMP-activated protein kinase (AMPK), can do something unusual in this context. In two esophageal squamous cell carcinoma (ESCC) cell lines, TE1 and TE13, AICAR treatment drove a time-dependent accumulation of cells in S phase, the stage of the cell cycle in which DNA is replicated. Normal esophageal epithelial cells, by contrast, showed no significant cell-cycle shift under the same conditions. The researchers then showed that this drug-induced synchronization could be exploited: pretreating the cancer cells with AICAR made them markedly more vulnerable to SN38, the active metabolite of the chemotherapy drug irinotecan, which preferentially kills cells in S phase.

The study, conducted by a team led by Xin Zhang of Shantou University Medical College with collaborators including Thomas Ming Swi Chang of McGill University, treated three cell lines with 500 micromolar AICAR over periods of 24, 48, and 72 hours. Proliferation assays using CCK8 reagent showed significant, time-dependent suppression of growth in all three lines, and immunostaining for Ki67, a marker of active cell division, confirmed reduced proliferation across the board. That antiproliferative effect alone is not surprising; AMPK activation is known to shut down energy-consuming anabolic processes when cells sense metabolic stress. What caught the researchers’ attention was the divergence downstream of that shared effect. The cancer cells and the normal cells responded to the same signaling events with fundamentally different cell-cycle fates, and untangling why became the central question of the paper.

Flow-cytometric analysis of DNA content revealed the scale of that divergence. In TE1 cells, the S-phase fraction rose from 33.85 percent in untreated controls to 52.87 percent after 72 hours of AICAR exposure, while TE13 cells showed an even more dramatic accumulation, climbing from 47.42 percent to 73.83 percent, both statistically significant changes. HEEC cells, the normal esophageal epithelial line, remained essentially stable, with the S-phase fraction moving from 37.85 percent to 42.79 percent, a difference the authors reported as not statistically significant. The normal cells exhibited only a marginal, transient increase in the G0/G1 population at intermediate time points, which returned to baseline by 72 hours. In other words, AICAR did not merely slow the cancer cells down; it trapped them at a specific, therapeutically vulnerable point in their division cycle, and left normal cells largely untouched in this respect.

To explain the mechanism, the researchers mapped the signaling cascades activated by AICAR using Western blotting. As expected, AICAR increased phosphorylation of AMPK at threonine 172 in all three cell lines without changing total AMPK abundance. Activated AMPK phosphorylated TSC2, which in turn suppressed the mTORC1 complex, seen as reduced phosphorylation of the downstream effectors p70S6K and S6. This canonical energy-sensing pathway, which curtails protein synthesis and ribosome biogenesis, was uniformly suppressed in every cell type examined and plausibly accounts for the broad antiproliferative effect. The authors note that mTORC1 inhibition also impairs de novo nucleotide synthesis, depleting the pool of deoxyribonucleotides needed for DNA replication, which would stall DNA elongation and help explain why cells already in S phase could not progress further.

The more surprising half of the mechanism involved the second mTOR complex. Activated AMPK phosphorylated the transcription factor FOXO3a at serine 413, enhancing its transcriptional activity, and FOXO3a in turn upregulated Rictor, a core structural component of mTORC2. Rising Rictor levels drove increased phosphorylation of AKT at serine 473 in all three cell lines, indicating sustained mTORC2 activation running in parallel with mTORC1 suppression. The authors interpret this apparent paradox, an energy sensor simultaneously throttling anabolic growth while boosting pro-survival signaling, as an adaptive compensatory strategy: by blocking protein and lipid synthesis while preserving essential AKT survival signaling, the cell can weather acute energy stress without immediately triggering apoptosis. While bidirectional regulation of the two mTOR complexes by AMPK has been described before, the authors state this is the first study to link the phenomenon to drug-induced S-phase arrest in ESCC.

Full AKT activation, however, requires phosphorylation at two residues, threonine 308 and serine 473, and here the cell-type differences became decisive. In TE1 and TE13 cells, AICAR induced sustained elevation of phospho-AKT at both sites, along with increased phosphorylation of FOXO3a at serine 253, a modification that promotes FOXO3a’s export from the nucleus and degradation. HEEC cells showed no significant change at either position. Hyperactivated AKT in the cancer cells suppresses the FOXO3a-driven transcription of the cyclin-dependent kinase inhibitors p21Cip1 and p27Kip1, weakening the G1/S checkpoint and allowing cells to stream into S phase, where the concurrent mTORC1 inhibition then held them captive. In the cancer lines, the levels of p21Cip1, p27Kip1, and p53 progressively declined with prolonged treatment, and phosphorylated Rb rose, consistent with a compromised G1 checkpoint.

The upstream determinant of this divergence, the researchers argue, is PTEN, a phosphatase that opposes PI3K/AKT signaling and restrains AKT phosphorylation at threonine 308. In HEEC cells, AICAR treatment progressively increased PTEN expression, keeping AKT in only a partially activated state. In TE1 and TE13 cells, PTEN levels remained largely unchanged or slightly decreased, permitting full dual-site AKT activation. The authors connect this to p53 status: PTEN is a direct transcriptional target of p53, and the p53-PTEN axis operates as a positive feedback loop in which p53 drives PTEN transcription while PTEN stabilizes p53. HEEC cells carry functional p53, so AICAR treatment raised both p53 and PTEN, constructing a protective brake on AKT. The TE1 and TE13 lines carry dysfunctional p53, breaking the loop and leaving AKT unrestrained. Notably, this S-phase outcome also departs from the G0/G1 arrest typically reported for AMPK activation in breast, liver, glioma, and renal cancers, suggesting ESCC follows a distinct regulatory pattern.

The therapeutic payoff of the finding lies in chemosensitization. Irinotecan, a topoisomerase I inhibitor used as a second-line option for ESCC, exerts its cytotoxicity primarily on S-phase cells, but its clinical utility is limited by adverse effects such as myelosuppression and hepatotoxicity. The researchers pretreated TE1 and TE13 cells with AICAR for 48 hours, withdrew the drug, and then administered SN38. Ki67 staining confirmed that proliferation recovered within 24 hours of AICAR withdrawal, and Annexin V/propidium iodide staining showed that the combination group underwent a markedly higher rate of apoptosis than either agent alone, with significantly elevated cleaved caspase-3 confirming the apoptotic execution. AICAR pretreatment also sensitized the cells to 5-fluorouracil, another S-phase-specific agent: apoptosis in the combination group at 1 microgram per milliliter of 5-Fu was comparable, though not significantly different, to that achieved with 5-Fu alone at a fivefold higher concentration, suggesting a substantial reduction in the effective dose might be possible.

Several important caveats frame these results. The entire study was performed in cell culture, on two cancer lines and one normal line, and no animal or human data yet exist; the authors explicitly note that in vivo experiments are needed to optimize dosing and evaluate the efficacy and safety of the combination strategy. AICAR is also known to produce AMPK-independent off-target effects, so genetic knockout or knockdown of AMPK will be required to confirm the pathway is fully responsible for the observed phenotypes. The proposed AMPK/FOXO3/Rictor/mTORC2/AKT axis rests primarily on measurements of protein phosphorylation, and gain- or loss-of-function experiments manipulating FOXO3, Rictor, and p53 will be needed to establish causality. With those limitations acknowledged, the study offers a conceptually appealing template: use a metabolic drug to synchronize tumor cells at a vulnerable cell-cycle stage while sparing normal tissue, then strike with an agent tailored to that stage. Whether that sequence survives the passage from dish to organism remains the critical next question for this line of research.

Subject of Research: How the AMPK activator AICAR induces S-phase arrest in esophageal squamous cell carcinoma cells via AMPK/mTORC1 and mTORC2/AKT signaling and enhances chemosensitivity in cell culture

Article Title: AICAR induces S phase cell cycle arrest in ESCC by targeting the AMPK/mTORC1 and mTORC2/AKT pathway

Article References: Zou, M., Zhang, T., He, L., Li, J., Lin, Z., Jia, C., Ma, K., Wang, L., Xu, S., Cai, B., Wu, L., Ren, C., Chang, T. M. S., Sun, P., & Zhang, X. (2026). AICAR induces S phase cell cycle arrest in ESCC by targeting the AMPK/mTORC1 and mTORC2/AKT pathway. Frontiers in Oncology, 16, Article 1975805. https://doi.org/10.3389/fonc.2026.1975805

Image Credits: AI Generated

DOI: 10.3389/fonc.2026.1975805

Keywords: AICAR, AMPK, esophageal squamous cell carcinoma, cell cycle arrest, S phase, mTORC1, mTORC2, AKT, PTEN, p53, SN38, chemosensitization