KNSTRN, also known as C15orf23, is a mitosis-related protein and directly binds to microtubules[17]. The role of KNSTRN in tumorigenesis is rarely investigated in the past. In recent years, Studies reported functional links between KNSTRN and endometrial cancer and lung adenocarcinoma[16, 19]. In the present study, we first sought to identify the expression and prognostic value of KNSTRN gene in HCC. We found that KNSTRN was up-regulated in tumor tissues and high KNSTRN expression was closely associated with poorer overall survival (OS) and progress free interval (PFI) in HCC. Furthermore, high KNSTRN expression was linked to the worse histological grade features in HCC cohorts. The area under the curve (AUC) values of 1-, 3-, and 5-year predicted survival rates were all greater than 0.6. Simultaneously, the AUC value of the diagnostic efficiency was 0.935. These findings strongly suggested that KNSTRN was an oncogenic molecule of tumor progression, a potential prognostic and diagnostic biomarker for HCC.
The tumor microenvironment (TME) and tumor immune evasion are correlated with cancer prognoses and therapeutic[23]. First, tumor infiltration of immune cells leads to T-cell anergy or dysfunctional T-cell phenotypes[24, 25], which promote tumor escape of the host immune system, tumor progression, invasion, metastasis, and therapeutic resistance[24]. In this study, given that KNSTRN is abnormally highly expressed in LIHC, so we speculate that KNSTRN may be involved in regulating the tumor immune response. We, therefore, explored the possible correlation between the infiltration level of various immune cells and the expression of the KNSTRN gene in LIHC(TCGA data). As shown in Fig. 5A-C, a significant positive correlation was observed between KNSTRN expression levels and the proportion of T follicular helper cells (TFH), T helper cells, and type 2 T helper cells (Th2) in the HCC tissues. These results are consistent with previous studies which have shown that TFH cells regulate tumor growth and progression via CXCR5, the chemokine receptor[26, 27]. Th2 induce M2 macrophage polarization to promote tumor progression by cytokines[28]. Moreover, Th2 plays a significant role in the resistance mechanisms against cancer immunotherapy[29].
One of the most important features of immunosuppression is T cell exhaustion. In our study, upregulated KNSTRN expression was strongly associated with T-cell exhaustion markers (Fig. 5E-J), including PD-1(PDCD1), PD-L1(CD274), CTLA4, TIM-3(HAVCR2), LAG3 and TOX. PD1/PDL1, as the important immune checkpoint component, has been verified to regulate the function of Tumor-infiltrating lymphocytes. Notably, although PD1/PDL1 checkpoint blockade therapy is widely performed to various malignancies including HCC, some studies reported that PD-1 has a critical role in tumor antigen tolerance, leading to poor therapeutic effect in some patients with PD1 therapy[30–32]. Concurrently, combinations of inhibitory receptors such as PD-1 and TIM-3 coregulate exhausted T cells and cooperate to induce deficiency of CD8 + T cells[33]. Further, activation of TIM-3 and CTLA-4 can suppress T-cell mediated immune responses[34]. Also, Wherry et al. have reported that TOX + cells can express multiple inhibitory receptors such as LAG3 and PDCD-1, and suggest that TOX is a major molecule which regulates the differentiation of TEX at the transcriptional and epigenetic levels[35]. Additionally, T-Cell exclusion is dependent on the infiltration of immunosuppressive cells, including CAFs, Tregs, M2-TAMs, and MDSCs[36]. Fascinatingly, as shown in Fig. 5D, we found that KNSTRN expression levels correlated well with expression of immunosuppressive cells in HCC. Thus, we speculated that T-cell exclusion is the major mechanism through which KNSTRN regulates tumor escape of immune cells, thereby contributing to HCC growth and progression.
Furthermore, the correlation between the KNSTRN and immune markers of various immune cells in HCC is shown as Table 2. In our study, we found that the expression level of KNSTRN was significantly correlated with the immune marker of M2 macrophage in HCC, such as CD163, immunoglobulin domain containing 4 (VSIG4), and Colony stimulating factor 1 receptor (CSF1R) (P < 0.001). This shows that KNSTRN has a potential function to regulate macrophage polarization in HCC. Moreover, the expression of KNSTRN was also markedly positively correlated with immune markers of Neutrophil and DCs in HCC, including ITGAX, CD1C, NRP1, CCR7, ITGAM, and CD59 (P < 0.05, Table 2). The results after adjustment of tumor purity revealed that the expression level of KNSTRN was significantly correlated with most of the immune markers of different T cells in HCC. In summary, these findings indicate that KNSTRN plays crucial but different ways on the regulation of the TME.
However, cancer development and therapeutic limitation cannot simply be attributed to a single gene. Therefore, we constructed the PPI network for the screened out the hub gene in the KNSTRN regulation. Intriguingly, we found significant correlations (r > 0.8) between the KNSTRN expression level and expression levels of SPAG5, CENPE, CCNB1, IQGAP1, CDC20, NUF2, BUB1B, and CCNB2 in HCC (Fig. 4F and Supplementary File.5). Studies have shown that SPAG5 serves a promising prognostic factor in HCC and acts as an oncogene[37–40]. Human centromere-associated protein (CENPE) can be used as potential biomarkers for clinical diagnosis[41]. CCNB1 and CCNB2 were independent risk factors of HCC[42]. IQGAP1 is known to be overexpressed in HCC[43]. Zhao et al. and Lai et al. reported that CDC20 as tumor promoters in HCC[44, 45]. Upregulation of BUB1B, CCNB1, and CDC20 in tumor tissues predicted worse overall survival and disease-free survival in HCC patients[46–48]. Nuf2 is highly expressed in several human cancers and promotes tumorigenesis[49–51]. Taken together, these related genes, studies above suggest that KNSTRN might play a key role in the deterioration of HCC.
DNA methylation during carcinogenesis has an impact on not only gene expression, but also the prognosis of cancer patients[52]. Notably, our data revealed that methylation level of KNSTRN-cg08036289 was significantly correlated with its expression. Moreover, hypermethylation CpG site KNSTRN‐cg08036289 (P = 0.042) was associated with an improved prognosis in HCC.
Nevertheless, we attempted to perform GO and KEGG functional enrichment analysis to further clarify the underlying biological function of KNSTRN gene. Among tumor cells, cell cycle disorder is one of the most remarkable features. In this study, KEGG pathway analysis showed that KNSTRN was mainly related to Cell cycle, Oocyte meiosis and Cellular senescence signaling pathway. Furthermore, we found that the KNSTRN co-expression is mainly involved in organelle fission, chromosomal region and tubulin binding using GO functional enrichment analysis. The GO and KEGG analysis revealed that KNSTRN may play a role in influencing Cell cycle and regulating Cellular senescence.
The majority of studies are based on miRNA binding to 3’UTR of target mRNA to degrade or repress the corresponding gene expression[53, 54]. Interestingly, just like how every sword has two sides of the blade, the growing evidence has shown that miRNAs can also establish positive roles and promote gene expression under certain circumstances[20, 55–57]. Moreover, Yu et al.[22] defined a subset of RNA-nuclear activating miRNA (NamiRNA) with an activation function in the nucleus. Encouragingly, several researches also proposed a novel regulatory network of NamiRNA-enhancer gene activation[58, 59], and this novel regulatory axis is crucial for cell identification throughout the transformation from normal to cancerous cells[21]. To ascertain whether KNSTRN was positively modulated by Nuclear activating miRNAs, we first predicted the NamiRNA that could potentially bind to KNSTRN and finally found hsa-miR-107. Based on the action mechanism of Nuclear activating miRNA in regulation of target gene expression, there should be positive correlation between hsa-miR-107 and KNSTRN. In our study, the correlation analysis ultimately revealed that KNSTRN was significantly positively correlated with hsa-miR-107 (r = 0.119, P < 0.05), as seen in Fig. 6B. Next, the expression and prognostic value of hsa-miR-107 in HCC were determined. As presented in Figs. 6C and 6D, hsa-miR-107 was markedly up-regulated in HCC and its upregulation was nagatively linked to patients, prognosis (HR = 0.62, P < 0.05). All these findings suggest that hsa-miR-107 might be the most potential regulatory NamiRNA of KNSTRN in HCC. Finally, a possible NamiRNA-enhancer network of hsa-miR-107 activates the KNSTRN expression was constructed in LIHC, as was presented in Fig. 7. Of course, although the NamiRNA-enhancer-KNSTRN regulatory axis was obtained through bioinformatics analysis, more experiments are needed to confirm our prediction. The NamiRNA − enhancer − target gene activation network should deserve more attention.