Deep inside every solid tumor lies a paradox that has fascinated cancer biologists for decades: the very lack of oxygen that should starve a growing mass of cells instead often makes it more aggressive. A new study published in the British Journal of Cancer adds a striking piece to this puzzle, showing that in oral squamous cell carcinoma, the most common cancer of the mouth, oxygen deprivation switches on a specific inflammatory signaling chain that pushes tumor cells to multiply faster. The work, led by Kaixiang Wu, Shupeng Zhang and Xiaodong Feng and their colleagues at institutions including Shandong First Medical University, Qingdao University and Shanghai Jiao Tong University, maps a three-part molecular relay, running from the protein HMGB1 through the NLRP3 inflammasome to the enzyme caspase-1, and demonstrates that each link in that chain matters for tumor growth in living animals.

The starting point of the investigation was a molecule with a double life. High mobility group box 1, or HMGB1, is a nuclear protein that normally helps organize DNA and regulate gene transcription inside healthy cells. But when cells are stressed, injured or dying, HMGB1 can be released into the surrounding tissue, where it acts as a danger signal, an alarm molecule that summons and activates the immune system. In cancer, this alarm function can backfire spectacularly. Chronic inflammation is a recognized hallmark of malignancy, and HMGB1 has previously been implicated in promoting tumor growth, invasion and metastasis in cancers ranging from hepatocellular carcinoma to colorectal and pancreatic tumors. What remained unclear was precisely how the oxygen-starved environment inside tumors regulates HMGB1, and what that regulation means for cancers of the oral cavity.

To answer that question, the researchers began with human tissue. They profiled HMGB1 expression in clinical specimens of oral squamous cell carcinoma and compared it with clinicopathological information from the same patients. The results were unambiguous: HMGB1 was constitutively upregulated in the tumor tissues, and the degree of overexpression tracked with clinically ominous features. Tumors with higher HMGB1 levels tended to be larger, and elevated expression was positively associated with lymph node metastasis, the process by which oral cancer cells spread to the neck and dramatically worsen prognosis. When the team constructed Kaplan-Meier survival curves, patients whose tumors expressed the most HMGB1 fared significantly worse than those with lower levels, marking the protein as a candidate prognostic biomarker for the disease.

With the clinical correlation established, the investigators turned to controlled experiments in cell culture to determine whether hypoxia was driving the HMGB1 surge. When oral squamous cell carcinoma cells were exposed to hypoxic conditions, their HMGB1 expression rose markedly. More importantly, the team traced what happened downstream. Hypoxia also increased the levels of NLRP3, a sensor protein that assembles into a multi-subunit complex known as the inflammasome, and of cleaved caspase-1, the proteolytic enzyme that the NLRP3 inflammasome activates. Alongside these inflammatory components, the cells accumulated HIF-1α, the master transcriptional regulator of the hypoxic response. The picture that emerged was of a coordinated pathway: low oxygen elevates HMGB1, and HMGB1 in turn sustains the activation of the NLRP3 inflammasome and its enzymatic output through caspase-1.

The functional consequences of this signaling cascade were tested by silencing HMGB1 in the cancer cells. The results showed that the protein is not merely a passive byproduct of hypoxia but an active driver of malignant behavior. When HMGB1 was knocked down, the hypoxia-induced expression of NLRP3, cleaved caspase-1 and HIF-1α all fell, indicating that HMGB1 sits upstream of the entire axis. Hypoxia alone had markedly enhanced the proliferation of the oral cancer cells, and it also induced pyroptosis, an inflammatory form of programmed cell death executed by gasdermin pores that release interleukin-1 family cytokines. Intriguingly, hypoxia alleviated a G2/M phase arrest in the cell cycle, freeing the cells to progress through division more readily. All of these pro-tumor effects were reversed when HMGB1 was depleted, demonstrating that the molecule is required for hypoxia to exert its full malignant influence on oral cancer cells.

The interplay between hypoxia, inflammation and cell death revealed here is more nuanced than a simple growth signal. Pyroptosis is classically viewed as an anti-cancer process, since it inflames the tumor bed and can alert immune cells to the presence of abnormal tissue. Yet the new findings suggest that in the hypoxic context of oral squamous cell carcinoma, the HMGB1-driven inflammasome activity coexists with, and may even support, enhanced proliferation. The authors propose that the net effect of the pathway is pro-tumor: the same caspase-1 activation that triggers inflammatory death in some cells appears to accompany conditions that release surviving cells from cell cycle constraints and accelerate their expansion. This kind of dual role echoes a broader theme in tumor immunology, in which inflammatory pathways that evolved to fight infection can be co-opted by cancers to remodel their microenvironment in self-serving ways.

Critical evidence came from animal models, where the team could test whether the pathway mattered in the complexity of a living organism rather than a culture dish. In murine xenograft models, in which human oral cancer cells are implanted into mice, knocking down HMGB1 inhibited tumor growth. Separately, the researchers examined the downstream links of the axis using models deficient in NLRP3 or caspase-1, and found that loss of either component delayed the progression of oral cancer. To extend the relevance beyond transplanted tumors, the team also employed a 4-NQO-induced carcinogenesis protocol, a chemically induced model that recapitulates the stepwise development of oral squamous cell carcinoma in mice, from premalignant lesions to invasive cancer. Together, these in vivo experiments established that the HMGB1/NLRP3/caspase-1 axis is not an artifact of cell culture but a functional driver of tumor progression in intact animals.

The study builds on a substantial body of prior work implicating both HMGB1 and the NLRP3 inflammasome in head and neck malignancies. Earlier research from some of the same collaborators had shown that NLRP3 promotes tumor growth and metastasis in oral squamous cell carcinoma, that a microRNA called miR-22 suppresses oral cancer cell proliferation by targeting NLRP3, and that NLRP3 expression in tumor-associated macrophages predicts poor prognosis in head and neck squamous cell carcinoma. Other groups had connected hypoxia-induced HMGB1 to tumor growth in liver cancer through Toll-like receptor 9, and to the self-renewal of glioma stem cells through the receptor RAGE. The new study unifies these threads by placing HMGB1 upstream of the inflammasome in a hypoxia-dependent pathway specific to oral cancer, and by validating the entire axis from clinical specimens through cell biology to animal models.

Therapeutically, the findings point to a strategy that differs from conventional approaches. Because HMGB1 is upregulated in tumor tissue and correlates with poor outcomes, it offers both a biomarker for identifying high-risk patients and a target for intervention. Blocking the pathway at any of its three nodes, whether by neutralizing HMGB1, inhibiting NLRP3 assembly or suppressing caspase-1 activity, delayed cancer progression in the experimental systems. Inflammasome inhibitors are already an active area of drug development for inflammatory and metabolic diseases, and HMGB1-targeting agents have been explored in contexts from acute liver failure to sepsis. The new data suggest that repurposing or refining such inhibitors for oral squamous cell carcinoma, particularly for tumors with documented hypoxia and high HMGB1 expression, could be a viable avenue. The authors also note that hypoxia itself is a known barrier to radiotherapy and immunotherapy, so a pathway that links hypoxia to inflammatory tumor promotion may help explain why some oral cancers resist current treatments.

Oral squamous cell carcinoma remains a disease with a heavy global burden, particularly in regions where tobacco use, alcohol consumption and betel quid chewing are prevalent, and survival rates for advanced disease have improved only modestly in recent decades. A molecular axis that connects the tumor’s oxygen-starved core to inflammatory proliferation, and that is measurable in clinical specimens and targetable in animal models, offers a fresh angle of attack. As with all preclinical research, the distance between xenograft mice and patients is considerable, and questions remain about how best to inhibit the pathway without suppressing the beneficial inflammatory responses that guard against infection. But the study provides a clear mechanistic blueprint: hypoxia raises HMGB1, HMGB1 activates the NLRP3 inflammasome, caspase-1 carries the signal forward, and oral cancer cells proliferate. Interrupting that relay, the researchers conclude, may offer a viable therapeutic strategy for a cancer that urgently needs new options.

Subject of Research: Hypoxia-induced HMGB1/NLRP3/Caspase-1 inflammatory signaling in oral squamous cell carcinoma progression

Article Title: Hypoxia-induced HMGB1/NLRP3/Caspase-1 signaling axis enhances proliferation of oral squamous cell carcinoma

Article References: Wu, K., Zhang, S., Cao, L., Wang, J., Wang, M., Zheng, Z., & Feng, X. (2026). Hypoxia-induced HMGB1/NLRP3/Caspase-1 signaling axis enhances proliferation of oral squamous cell carcinoma. British Journal of Cancer. https://doi.org/10.1038/s41416-026-03632-6

Image Credits: AI Generated

DOI: 10.1038/s41416-026-03632-6

Keywords: oral squamous cell carcinoma, hypoxia, HMGB1, NLRP3 inflammasome, caspase-1, pyroptosis, HIF-1α, tumor microenvironment, prognostic biomarker, inflammation, xenograft model, cancer therapy