Although the Hippo pathway and associated genes are vital for the development of LIHC [6, 22], the prognostic role and molecular mechanisms of Hippo pathway-related lncRNAs in LIHC progression remain unclear. The results of previous studies suggest that the mining of multi-data molecular subtypes is the basis for personalized clinical interventions in cancer [11, 23]. Cancer subtypes differ in clinical factors, such as overall survival time, which is often an important indicator of personalized therapy [24]. In this study, according to the expression of 88 Hippo pathway-related prognostic lncRNAs, two molecular subtypes (clusters A and B) were identified. Patients in cluster A had longer survival than those in cluster B; therefore, the clinical characteristics of these two molecular subtypes were compared to investigate possible reasons for this difference. We found that the cluster A subtype had relatively more patients with pathologic T1 (55.79% vs. 25.93%), stage I (52.63% vs. 24.69%), and non-vascular invasive tumors (60.35% vs. 40.74%). This might explain the favorable prognosis of patients with the cluster A subtype.
Additionally, the immune response is closely related to the molecular mechanisms of cancer subtypes [25, 26]. We found that the infiltrating abundance of the seven immune cells was significantly different between subtypes A and B. For example, the abundance of tumor-infiltrating activated memory CD4+ T cells was elevated in cluster A, while the abundance of Tregs and M2 macrophages was significantly higher in cluster B. Various antitumor biological effects of CD4+ T cells have been revealed in a previous study [27]. Activation of CD4+ T cells caused by tumor invasion can inhibit inflammation, which is used in the prognosis of head and neck squamous cell carcinoma [28]. Tumor cells have been demonstrated to play a direct role in the expansion of CD4+ T cells, thus further inhibiting antitumor immunity due to T cells [29]. Accumulating evidence suggests that Treg cells play a vital role in inhibiting natural killer cell immune responses in human cancer [30]. Yu et al. showed that tumors with an increased number of Treg cells were associated with poor prognosis in LIHC [31]. A previous study indicated that Treg cells foster tumor progression and predict adverse outcomes in LIHC [32]. Moreover, M2-derived exosomes can accelerate the migration and invasion of colon cancer cells [33]. CHI3L1 proteins produced by tumor M2 macrophages play crucial roles in the progression of tumors in humans [34]. Furthermore, 14 immune checkpoint genes, including CTLA-4 and PD1/PD-L1, were significantly upregulated in cluster B. As an inhibitory co-receptor, CTLA-4 not only interferes with the activation of T cells but is also overexpressed in patients with LIHC [35]. Checkpoint antibody inhibitors, including PD1 and PD-L1 inhibitors, are also common inhibitors with certain tumor-suppressive effects [36]. These findings further explain the differences in the overall survival of patients between the two subtypes.
Among the 88 Hippo pathway-related prognostic lncRNAs, 12 independent prognostic lncRNAs were identified, and LASSO regression identified six valuable independent prognostic lncRNAs, including JMJD1C-AS1, LINC01410, LINC01503, RBM5-AS1, RHPN1-AS1, and TMEM220-AS1. A recent study showed that RHPN1-AS1 promotes malignant progression and predicts poor clinical outcomes in liver cancer [37]. It has also been found to promote tumor cell progression in human cancer by participating in the miR-299–3p/FGF2 axis [38]. A previous study showed that LINC01410 promotes angiogenesis and metastasis in human cancers [39]. A recent study indicated that upregulated LINC01410 presents a poor prognosis in cholangiocarcinoma patients and can be used as a prognostic gene for cancer [40]. Wang et al. showed that upregulated LINC01503 contributes to cancer cell progression through the MAPK/ERK pathway, which is considered a therapeutic target for LIHC [41]. RBM5-AS1 is involved in the realization of related molecular functions in colon CSCs [42]. Du et al. indicated that the downregulation of TMEM220-AS1 with copy deletion was associated with a poor prognosis of LIHC, which could be used as a promising prognostic biomarker [43]. These studies suggest the important roles of these lncRNAs. Based on these six lncRNAs, a prognosis risk model was developed to assign patients into two risk -groups, in which patients in the high-risk group had shorter survival times than low-risk patients. Additionally, the risk score was found to have an independent prognostic value in patients with LIHC. This further confirms the prognostic value of these lncRNAs.
Genes that were dysregulated between the two risk groups were also analyzed. Most gene expression levels were elevated in the high-risk group. GSEA revealed that pathways in cancer and the cell cycle were significantly associated with the high-risk group. Under normal circumstances, the damage to endogenous and exogenous DNA generated during cancer development can be repaired by the cell cycle pathway [44,45]. A previous study showed that oncogenic H2AZ1 plays an established role in accelerating the cell cycle transition during hepatocarcinogenesis [46]. In addition, the pathway in cancer is meaningful because it helps reverse, delay, or prevent the occurrence of tumors from a therapeutic point of view [47]. Our study showed that cancer and cell cycle pathways were significantly related to the high-risk group, indicating that the genes in this group contributed to the progression of LIHC.
In conclusion, two molecular subtypes of LIHC based on the Hippo pathway-related prognostic lncRNAs were identified in this study. These two subtypes differ in terms of overall survival, clinical pathology, infiltration abundance of 7 immune cells, and expression of checkpoint genes such as CTLA-4 and PD-1/L1. Moreover, a prognostic risk model was developed using six independent prognostic lncRNAs (JMJD1C-AS1, LINC01410, LINC01503, RBM5-AS1, RHPN1-AS1, and TMEM220-AS1). This risk model can independently predict the prognosis of LIHC.