Triosephosphate isomerase 1 may be a risk predictor in laryngeal squamous cell carcinoma: a multi-centered study integrating bulk RNA, single-cell RNA, and protein immunohistochemistry

Laryngeal cancer is a common malignant tumor of the respiratory system of origin. According to the latest cancer statistics report, there would be 184,615 new cases of laryngeal tumors and 99,840 deaths worldwide in 2020 [1]. In China, it is predicted that there will be 30,832 new cases of laryngeal tumors and 14,404 deaths in 2022 [2]. Moreover, laryngeal cancer severely affects the quality of life of patients [3], accompanied by impairment of speech, breathing, and swallowing ability. Therefore, both globally and in China, laryngeal cancer is featured by a high incidence, high mortality rate, and significant impact. The most common type of laryngeal cancer is laryngeal squamous cell carcinoma (LSCC). Some studies have shown that some common factors, such as dietary factors, environmental factors, and alcohol or tobacco abuse [4], are the major causal factors for the occurrence of LSCC [5]. Although great progress has been made in recent years in chemotherapy and radiotherapy, the prognosis of LSCC patients is still unsatisfactory, and there is an urgent need for clear molecular mechanisms related to the occurrence of LSCC to help improve the treatment and prognosis [6, 7]. However, numerous molecular mechanisms associated with LSCC remain unclear, specifically regarding the reprogramming of metabolism in LSCC. Further investigations are necessary to unveil the intricate mechanisms governing the development of LSCC [8‐10].

In the context of cancer, including LSCC, a significant phenomenon known as the Warburg effect involves a metabolic reprogramming characterized by an increased reliance on glycolysis [11]. Of particular interest in this context is triosephosphate isomerase 1 (TPI1), an enzyme that plays a crucial role in glycolysis. TPI1 functions by catalyzing the conversion of dihydroxyacetone phosphate (DHAP) to D-type glyceraldehyde-3-phosphate (G3P) and vice versa. Remarkably, TPI1 is not only involved in metabolic processes [12, 13] but also contributes to tumorigenesis through non-metabolic mechanisms [14]. In terms of its metabolic influence, TPI1 has been implicated in promoting oral cancer [15], liver cancer occurrence [12, 13], and metastasis by modulating glycolytic levels within cancer cells. On the non-metabolic front, TPI1 has been found to enhance overall histone acetylation during the cell cycle [14], a critical factor in cancer-related cellular processes [16]. Despite these intriguing findings, the role of TPI1 in LSCC remains unexplored, which has captured our attention.

Herein, we studied the expression of TPI1 in LSCC using LSCC transcriptome data, in-house tissue microarrays (TMA), and related gene pathway analysis, and explored the possible mechanisms of TPI1 occurrence in LSCC single cells. All these results will contribute to a deeper understanding of the mechanisms of LSCC occurrence and have significant implications for the prevention, treatment, and improved prognosis of LSCC.

It is well known that the development of many cancers is associated with metabolism reprograming at the cellular level [29]. As a common initial pathway for both aerobic and anaerobic oxidation, glycolysis is indeed associated with the development of many cancers [30]. In the process of glycolytic reprograming, TPI1, one of the key enzymes, have been shown to be closely associated with the deterioration of many cancers, such as intrahepatic cholangiocarcinoma [31, 32], pancreatic cancer [33], breast cancer [34], and lung cancer [35, 36]. However, abnormal TPI1 levels have never been reported to be associated with LSCC. In the present study, the comprehensive patterns of TPI1 overexpression were confirmed using worldwide bulk RNA data and intra-group protein immunohistochemistry, as well as scRNA-seq result, which reflects the strong evidence-based philosophy throughout our study. Furthermore, we probed the intricate co-expression network of TPI1 in promoting glycolysis and cell cycle in both LSCC tissue and single cells. Our study provides important clues for depicting the molecular landscape of TPI1 underlying LSCC.

TPI1 activity was significantly enhanced in LSCC tissue and single cells at the mRNA and protein levels. Based on the evidence-based concept, worldwide gene microarrays and RNA-seq datasets were collected to complete the quantitatively expression analysis of TPI1 mRNA in LSCC tissue. Although the results were statistically heterogeneous, the results showed the overexpression of TPI1 in 251 LSCC tissue samples, as is opposed to 136 non-LSCC tissue samples, and the plausibility was confirmed by SROC curves. More importantly, compared with cytoplasm localization in normal laryngeal cells, TPI1 protein was observed to be accumulated in the nucleus of LSCC cells. Surprisingly, it was reported that the translocation of TPI1 to cell nucleus could induce the chemoresistance of lung adenocarcinoma cells [37]. Herein, a translocation from cytoplasm to nucleus was also observed for TPI1 protein in LSCC cells. Moreover, higher mRNA expression levels of TPI1 were predicted to produce poorer survival outcomes in LSCC patients. Taken together, it is conceivable that the elevated activity and nuclear translocation of TPI1 may promote the occurrence and progression of LSCC.

Furthermore, the complicated molecular mechanisms of TPI1 underlying LSCC were addressed using both bulk RNA and scRNA-seq analysis. It is noted that glycolysis and cell cycle pathways were significantly enriched in both LSCC tissue and single cells. In the present study, we not only identified a LSCC-specific gene co-expression module from the hdWGCNA network, but also determined the putative transcriptional regulators and hub genes of them.

As is well known, glycolysis is regarded as a hallmark of cancers. Head and neck squamous cell carcinoma exhibited strong mitochondrial and glycolytic reserving ability [38, 39]. To support the huge energy consumption, head and neck cancer cells become highly glycolytic [40]. This makes glycolysis the predominant source coupled with oxidative phosphorylation, which promotes the metastasis of head and neck cancer cells. In the cascade reaction of glycolysis, TPI1 enzyme catalyzes the isomerization of DHAP and G3P. Additionally, DHAP can also be converted into glycerol 3-phosphate, which is a pivotal substance linking glucose metabolism and fat metabolism [41]. Large amounts of studies have reported that enhanced glycolysis or glycolytic reprograming could result in the invasion and chemoresistance of LSCC cells [10, 42, 43]. For instance, upregulated glycolysis was correlated to the progression and immune escape in head and neck squamous cell carcinoma [44]. In-depth investigation demonstrated that glycolysis and viability were promoted in LSCC cells by mediating the miR-377-3p/lactate dehydrogenase A axis [45]. Based on such findings, glycolysis blockage provides a possibility for treating squamous cell carcinoma of the head and neck [46]. Surprisingly, the inhibition of TPI1 and glucose-6-phosphate isomerase was demonstrated to attenuate the proliferation and invasion ability of MDA-MB-231 cells [47]. Taken together, it was reasonable to propose that upregulated TPI1 may facilitates the development of LSCC cells by enhancing glycolysis.

Uncontrollable cell cycle progression acts as a trigger for the development of cancers [48]. Herein, our study put novel insights into the association between TPI1 and cell cycle. The co-expression gene set of TPI1 was identified in LSCC tissue and single cells, which were consistently mapped to the cell cycle pathway. More excitingly, for the intercellular communication network of LSCC single cells, the malignant laryngeal epithelial cells and fibroblasts indeed sent out an outgoing communication signal by coordinating with the MK signaling pathway, which was considered to be required for cell cycle progression [49‐51]. Although there is no report showing the precise functional mechanisms of TPI1 in the cell cycle of LSCC cells, few studies have demonstrated the association between them in other solid tumors. For example, cyclin-dependent kinase 2 could phosphorylate TPI1 Ser 117 and promote the nuclear translocation of TPI1 [14]. Given the enhanced nuclear accumulation of TPI1 protein in the LSCC cells, we could speculate that the cell-cycle-related genes co-expressed with TPI1 may promote the translocation of TPI1 protein and facilitate its activity in glycolytic reprograming. Additionally, TPI1 promotes the progression of breast cancer cells by regulating cell division cycle proteins and activating phosphoinositide 3-kinase/AKT serine/threonine kinase 1/mammalian target of rapamycin pathway [52]. Highly expressed TPI1 and the other glycolytic regulators were also suggested to be involved in cell cycle of prostate cancer [53]. When the glycolytic module (contains TPI1) is inhibited, there is a decrease in the S phase and an increase in the G2/M phase of the cell cycle and a decrease in the aggressiveness of hepatocellular carcinoma [54]. Moreover, TPI1 downregulation inhibits DNA replication in human fibroblasts and others and delays the entry of cells into the S phase ⁽⁵⁵⁾. Therefore, it is suggested that there may be a bidirectional promotion mechanism between TPI1 and the cell cycle of malignant laryngeal squamous epithelial cells, which ultimately triggers LSCC progression. Further studies are needed to determine how TPI1 regulates the cell cycle, cell division, and other approaches to promote the mechanism of LSCC.

However, our limitations are also obvious. For example, although overexpression of TPI1 predicts poor survival outcome in the TCGA-LSCC cohort, its prognostic value in the other three cohorts was either insignificant or, in some cases, contradictory. Two factors may contribute to these discrepancies. Firstly, there were several confounding factors that could have influenced the prognosis of LSCC patients. For example, in the GSE65858 cohort, variations in treatment strategies, human papillomavirus status, and clinical stages were observed among different patients. Secondly, the limited sample size in these cohorts might have compromised the representativeness and reliability of the prognostic value of TPI1. Further validation of the prognostic value of TPI1 in LSCC is warranted in larger cohorts. Additionally, the molecular biological functions of TPI1 in LSCC have not been experimentally confirmed. The possible role that TPI1 plays in the communication between LSCC cells and immune-stromal cells is less known. Functional experiments must be carried out to certify the finding of the present study. Despite that, the results of our work reveal the upregulation of TPI1 at both the mRNA and protein levels. More importantly, we emphasized the important role TPI1 plays in promoting glycolysis and cell cycle of LSCC single cells, which will enrich our understanding of the occurrence and progression of LSCC. We expect further studies in the future to fill in our shortcomings.

TPI1 is dramatically highly expressed in LSCC than in normal tissue, and the high expression of TPI1 may promote LSCC deterioration through its metabolic and non-metabolic functions, which will facilitate our future understanding of the detailed mechanisms of TPI1 in the development of LSCC.