In this study, we demonstrate for the first time that NLRP3 inflammasome plays an important role in 5-FU-based chemoresistance of OSCC. Our data support the hypothesis that activation of NLRP3 inflammasome could serve as a driver of OSCC development in patients receiving preoperative 5-FU-based adjuvant chemotherapy. Furthermore, our results indicate that lower NLRP3 expression is related to a less aggressive OSCC phenotype, and decreased NLRP3 inflammasome activation strengthens the antitumor effect of 5-FU in OSCC cells. We also find that increased intracellular ROS induced by 5-FU is a key regulator of NLRP3 inflammasome expression and activation in OSCC cells. Collectively, these results indicate that high expression and activation of NLRP3 inflammasome might confer a more malignant phenotype of OSCC, and regulate the chemoresistance of OSCC cells to 5-FU-based therapy.
It has been acknowledged that chronic inflammation is an important factor in carcinogenesis and tumor progression [
21,
22], and cancer-related inflammation has been recognized as the seventh hallmark of cancer [
23]. Tumor incidence and development are determined by oncogene and tumor suppressor gene aberrations as well as tumor microenvironment [
24]. Cytokines and chemokines in the tumor microenvironment were once considered to be released mainly by immune cells, however, many studies have demonstrated that tumor cells themselves also produce these inflammatory mediators [
25,
26]. Abnormal expression and activation of PRRs in tumor cells play critical roles in this process. Mastorci et al. have reported that mantle cell lymphoma (MCL) cell lines and primary MCL cells express high levels of Toll-like receptor (TLR) 2 and TLR5, which increases IL-4 and IL-6 production and promotes the development of MCL [
27]. Schwartz et al. have found that TLR3 is constitutively expressed in human pancreatic cancer and melanoma cells, and inhibition of TLR signaling reduces interferon-β, IL-6 and chemokine CXC ligand 10 production, which results in decreased proliferation and migration of tumor cells [
28]. Mehmeti et al. have shown a strong correlation between expression of TLR4 and proinflammatory mediators like IL-6 and IL-8 in primary breast cancer, and TLR4 protein expression is correlated with decreased survival [
29]. Similarly, our previous study revealed an important role of TLR4 in OSCC [
30]. We found that TLR4 and myeloid differentiation primary response gene 88 (MyD88) were highly expressed in OSCC cells and activation of TLR4 induced production of IL-6, IL-8 and vascular endothelial growth factor. OSCC cells were also found to be resistant to cisplatin-mediated apoptosis after lipopolysaccharide treatment. In gastric cancer, it has been shown that the TLR/MyD88 pathway is needed for the inflammatory response in tumor tissues, and plays a role in maintenance of stemness in gastric tumor cells [
31]. Besides TLRs, other PRRs/NLRs show aberrant function and expression in tumors [
32,
33]. To date, 22 human and 34 mouse NLRs have been identified, including NLRP1, NLRP3, NLRP6 and NLRC4 etc. [
34]. NLRP3, as part of the inflammasome protein complex, is better characterized in cancer compared with the others. Tarassishin et al. found that increased IL-1β processing by upregulated NLRP3 expression in human astrocytes and astrogliomas conferred them a mesenchymal phenotype, including increased migratory capacity, unique gene signature and proinflammatory signaling [
35]. Chow et al. showed that NLRP3-inflammasome-deficient mice had reduced incidence of methylcholanthrene-induced fibrosarcoma, and deficiency of NLRP3 contributed to decreased tumor metastasis via increased activation of natural killer cells [
36]. Li et al. have demonstrated that the NLRP3 gene signature may also serve as a predictive biomarker for glioma patients [
37]. However, its clinical significance and potential role in OSCC remains unknown. In the present study, we confirmed that NLRP3 inflammasome involved in OSCC development after 5-FU-based chemotherapy, and patients with higher NLRP3 expression had poorer survival than those with lower NLRP3 expression.
5-FU is an antimetabolite chemotherapeutic agent that is widely used for the treatment of cancer [
38]. However, 5-FU resistance with different causes threatens the clinical outcome of OSCC patients, and the exact mechanism is unclear [
39]. Recently, the role of NLRP3 inflammasome in this process has aroused the interests of researchers. Bruchard et al. found that increased activation of NLRP3 inflammasome in myeloid-derived suppressor cells (MDSCs) by 5-FU limited its antitumor efficacy [
40]. Ghiringhelli et al. showed that the 5-FU-derived activation of NLRP3 inflammasome in MDSCs promoted tumor angiogenesis by eliciting a T helper 17 response [
41]. In this study, we found that expression of NLRP3 and IL-1β was significantly upregulated in WSU-HN6 and CAL27 cells following 5-FU treatment. We also found that downregulating NLRP3 expression significantly enhanced the decrease of cell viability caused by 5-FU. These findings were confirmed in
Nlrp3
−/−
and
Caspase1
−/−
mice. With the 4-NQO-induced tongue squamous carcinoma model (a widely used OSCC model), we demonstrated that deficiency of NLRP3 inflammasome enhanced the anti-tumor effect of 5-FU.
Some recent investigations have recognized a growing number of diverse stimuli involved in NLRP3 inflammasome activation, and ROS is the most important one [
42,
43]. Increased ROS stress has been observed in a wide spectrum of human cancers and is associated with oncogenic signals such as c-myc and Ras [
44,
45]. Suzuki et al. found that treatment of human pancreatic cancer stem cells with 5-FU increased the intracellular ROS [
46]. Similarly, Darsigny et al. found that 5-FU treatment led to efficient production of intracellular ROS in colorectal cancer cells [
47]. Given that 5-FU induces intracellular ROS production in many cancer cells, we speculated whether 5-FU would increase ROS in OSCC cells and activate NLRP3 inflammasome. Our results demonstrated that intracellular ROS levels were indeed upregulated in 5-FU-treated OSCC cells, and later activated NLRP3 inflammasome, which promoted IL-1β production and 5-FU chemoresistance. When we treated OSCC cells with NAC, an antioxidant that suppressed intracellular ROS, the effect of 5-FU on NLRP3 inflammasome was reversed. However, the mechanisms of chemoresitance to 5-FU treatment lie on multiple aspects. For example, Zhang et al have demonstrated that tumor associated macrophages (TAMs) became activated during treatment with 5-FU and secreted putrescine that protected the colorectal cancer cells against chemotherapy with 5-FU by attenuating JNK-caspase-3 pathway-mediated cell apoptosis [
48]. Therefore, whether a similar work mechanism existed in TAMs of OSCC required needs further study.