Construction and Validation of a Combined Anoikis and Autophagy Prognostic Signature for Hepatocellular Carcinoma

doi:10.21203/rs.3.rs-4798023/v1

Download PDF

Article

Construction and Validation of a Combined Anoikis and Autophagy Prognostic Signature for Hepatocellular Carcinoma

https://doi.org/10.21203/rs.3.rs-4798023/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Given the poor prognosis of patients with hepatocellular carcinoma (HCC), it is crucial to investigate possible new biomarkers to aid in prognostication and customised treatment. Accordingly, we analysed differentially expressed anoikis- and autophagy-related genes (DE-AARGs) associated with poor outcomes in actual cases of HCC. Analysis of differentially expressed genes (DEGs) was performed based on mRNA expression patterns and clinicopathological information found in the Cancer Genome Atlas Liver Hepatocellular Carcinoma (TCGA-LIHC) database. Further validation of TCGA results was performed using the International Cancer Genome Consortium database. AARGs signatures were constructed by applying Univariate COX regression and the Least Absolute Shrinkage and Selection Operator method. We identified 13 AARGs, of which 9 showed significant associations with overall survival. Three AARGs (BIRC5, MAPK3, and BAK1) were selected to establish an AARGs signature. We assessed the prognostic capacity of the AARGs signature through various statistical methods. The molecular mechanisms underpinning this phenomenon were further studied using Gene Set Enrichment Analyses (GSEA). The prognostic ability of the signature was also examined in terms of clinical characteristics, immune landscape, immune checkpoint-blocking response, stemness, and chemotherapy response. Immunohistochemical staining was used to compare the protein expression levels of AARGs between normal liver tissue and HCC tissues. The high-risk group had higher tumour staging, shorter survival time, and worse prognosis than the low-risk group. In addition, high-risk patients showed inhibition of anoikis, a high autophagy index, and a suppressed immune system. The nomogram showed a strong prognostic capability for predicting overall survival in patients with HCC. With this study, a new AARGs-based signature has been developed to reliably predict patient prognosis for HCC.

Biological sciences/Cancer/Gastrointestinal cancer/Liver cancer

Biological sciences/Cancer/Cancer genomics

Biological sciences/Cancer/Cancer microenvironment

Biological sciences/Cancer/Cancer therapy

Biological sciences/Cancer/Tumour immunology

anoiki

autophagy

prognosis

differentially expressed genes

Ninety percent of all liver cancers are primary hepatocellular carcinoma (HCC), which is the leading cause of disease-related deaths globally [1]. HCC is among the most lethal malignancies globally, with less than 50% one-year survival rate and only 10% five-year survival rate [2]. According to the current guidelines, radical resection is the most effective treatment of HCC, and partial hepatectomy is the most common form of resection in cases of liver cancer, including HCC [3]. Because symptoms of HCC are not immediately apparent, most cases are diagnosed at an advanced stage, and approximately 80% of patients with HCC experience disease recurrence after treatment [4], the above reasons confirm that liver cancer is a malignant tumor with low cure rate. However, with the development of medicine and the arrival of the era of tumor immunity, immunological combination therapies based on immune checkpoint inhibitors can be used effectively to treat advanced liver cancer [5, 6]. Hois treatment is not suitable for all patients with HCC. Consequently, identifying new prognostic biomarkers is important for improving the quality of life of patients with HCC.

Anoikis occurs when cell–cell or cell–ECM attachments rupture, maintaining tissue homeostasis by removing misplaced or dislodged cells [7]. Under physiological conditions, the process of anoikis prevents displaced cells from adhering to other substrates for abnormal proliferation [8]. Therefore, after separation from the extracellular matrix (ECM), tumour cells survive by inhibiting anoikis, a particular type of programmed cell death [9, 10]. In order for metastsing, chemotherapy resistance, and tumour proliferation to occurr, cancer cells must acquire resistance to anoikis-like apoptotic cell death [11, 12]. Tumour cells with the capacity to metastasise have developed mechanisms to block anoikis [13]. To date, relatively few studies have been published on the relevance of anoikis in the clinical treatment of cancers. The stage of non-small cell lung cancer tumours and poor outcome are positively associated with the expression of Fas apoptotic inhibitory molecule 2 (FAIM2), an anoikis-related gene, and FAIM2 knockdown possibly inhibits anoikis resistance [14]. McCormick et al. discovered that epigenetic alterations in head and neck squamous cell carcinoma (HNSCC) enhance the downregulation of receptor-interacting serine/threonine-protein kinase 1 (RIPK1) expression during tumour progression, allowing tumour cells to avoid anoikis, thus promoting carcinogenesis [15]. Zhang et al. suggested that miR-26a inhibits ITGA5 expression in aggressive HCC cells and, at least in part, governs anoikis sensitivity in tumour cells. Consequently, miR-26a overexpression in HCC cells promotes anoikis [16]. The three biomarkers established by Chen et al. as distinctive anoikis-related hallmarks of HCC may be useful in predicting survival outcomes, reflecting immunological conditions, and assessing the efficacy of immunotherapy [17]. In summary, anoikis can affects the occurrence and development of tumors and the prognosis of patients through related mechanisms.

Autophagy is a complex multistep cellular process that conserves energy and is essential for cell development, survival, and homeostasis [18]. In its most prevalent form, autophagy is a process whereby cytosolic components are sequestered in double-membraned autophagosomes and degraded upon fusion with lysosomes [19]. Autophagy is a double-edged process that accelerates and hinders cell death. Different cancer types and stages of development can cause autophagy to function in both tumour-promoting and tumour-suppressing pathways [20]. Through the suppression of MHC-I, the activation of autophagy may reduce cancer-adaptive immune responses while enhancing dendritic cell immunity and antigen presentation [21, 22]. Abnormal expression of autophagy-related genes (ARGs) is believed to be the key factor influencing the prognosis of HCC. Fang et al. analysed autophagy-related genes in HCC progression and found six biomarkers that enable the construction of a novel nomogram to predict overall survival (OS) in patients with HCC [23]. These results highlight the fact that autophagy plays an important role in the onset and progression of cancer.

The results of several studies have linked the processes of anoikis and autophagy. Long et al. indicated that valosin-containing protein (VCP) induces autophagy to enhance cell survival, possibly through an anti-anoikis mechanism [24]. Palorini et al. suggested that levels of ATP and reactive oxygen species (ROS) may be higher in cells when protein kinase A (PKA) activity stimulates mitochondrial activity. Activation of Src in conjunction with autophagy contributes to anoikis resistance [25]. Based on these results, it appears that autophagy and anoikis are involved in cancer development. Further research is required to better understand how these cellular processes affect disease outcomes in cancer.

To elucidate the connection between anoikis, autophagy and cancer prognosis, we utilized the The International Cancer Genome Consortium (ICGC) and TCGA databases to develop a prognostic signature combining DRGs and anoikis- and autophagy-related genes (AARGs). This signature is significantly associated with HCC prognosis, anoikis status, autophagy status, and immune status. To construct a recurrence-forecasting model to determine the clinical net benefit that effectively directs clinical decisions, we created two additional signatures: a nomogram merging gene signature and a signature incorporating all independent prognostic factors. To investigate the probable mechanism of these AARGs, we also carried out functional analyses using Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and gene set enrichment analysis (GSEA). Using the CIBERSORT algorithm and TIMER database, we evaluated the influence of this gene profile on immunological infiltration. Immunohistochemistry and cell experiments confirmed the accuracy of this signature. In conclusion, these findings may contribute to the further exploration of the functional role of AARGs in HCC pathogenesis, improving clinical practice in the treatment of HCC.

2.1. Data Acquisition, Differential Expression Analysis, and Intersection Identification

Analysis of the clinical features and gene expression profiles of HCC samples was conducted using data gathered from TCGA (https://portal.gdc.cancer.gov/) and ICGC databases (https://dcc.icgc.org/). Patients with incomplete clinicopathological data were excluded from further analyses. To ensure that the comparison of expression levels is fair, unaffected by variations in gene length. We compared the fragments per kilobase million (FPKM) results with the transcripts per kilobase million (TPM) values, we downloaded RNAS sequencing data from the GDC site [26]. GeneCards (https://www.genecard.org/) was used to select anoikis-related genes (AnRGs) [27]. The Human Autophagy Database (HADb) (http://www.autophagy.lu/) was searched for autophagy-related genes (AuRGs). Human genes and proteins linked to autophagy are listed in depth and are up-to-date in HADb [28]. To determine the differentially expressed genes (DEGs) in HCC, using transcription assessing mRNA data from the TCGA collection, we carried out the differential expression of gene analysis using the "limma" package of the R programming language. [29]. For the next step of the analysis, DEGs with absolute |log2 fold change (FC)| > 1 and P-value < 0.05 were included for further analysis.

In the next step of the analysis, the DEGs that were found to intersect the TCGA set, the AnRGs acquired from GeneCards, and the AuRGs obtained from HADb were screened out as differentially expressed anoikis-autophagy-related genes (DE-AARGs) and used for subsequent analysis. Furthermore, we used cBioPortal (http://www.cbioportal.org/) to compare the mutations in DE-AARGs in HCC [30, 31]. Using the "maftools" package in R software, TCGA-mutation waterfall plot analysis and visualisation were carried out [32].

2.2. Development and Validation of a Prognostic Signature based on AARGs

We first used univariate COX regression to identify potential prognostic anoikis-autophagy-related genes (AARGs) linked to overall survival (OS) in TCGA samples. The range of prognostic genes was then limited using Least Absolute Shrinkage and Selection Operator (LASSO) analysis. Based on the "glmnet" R package's Lasso regression coefficients, we removed overfitting genes and calculated risk scores [33]. The risk score was calculated based on the following formula: Risk score = (βmRNA1 × expression level of mRNA1) + (βmRNA2 × expression level of mRNA2) + (βmRNA3 × expression level of mRNA3) + … + (βmRNAn × expression level of mRNAn). The risk scores of all patients with HCC in TCGA dataset were determined using the aforementioned formula. Patients were then divided into high- and low-risk groups using the median score. In this way, we gained the AARGs signature. Survival analysis was performed using the Log-rank method. Differences in survival rates were compared using the Kaplan–Meier survival curve. In addition, as an internal validation of stability, the receiver operating characteristic (ROC) curve area under the curve (AUC) was used to evaluate this signature-discriminating capability. Lastly, we validated the degree of generalisability of the AARGs signature using ICGC data as an independent external validation cohort.

2.3. Stratified Analysis by Clinical Characteristics

To determine whether the AARGs signature was a statistically significant prognostic factor regardless of the clinical data, we performed a stratified analysis based on T stage, stage, HBV, and HCV. We then used Kaplan–Meier survival analysis to compare the OS between the low- and high-risk prognostic groups in various subgroups. Further analysis of the TCGA database was conducted to compare the AARGs values between the two groups stratified by T, M, N, Stage, and Grade. We then performed Kaplan–Meier analyses to confirm that patients expressing different levels of AARGs were likely to have poor outcomes. In addition, we evaluated the robustness of the AARGs by performing a stratified life-course study for patients divided into various subgroups. By examining the Kaplan–Meier curves for each group, we were able to establish a correlation between the expression of AARGs and the overall prognosis of the patients. This analysis provides further evidence to support the hypothesis that aberrant expression of these specific genes is a significant factor contributing to poor patient outcomes.

2.4. Constructing and Validating a Predictive Nomogram

Additionally, the risk score and a few clinical traits, including age, sex, TNM stage, tumour grade, and the presence or absence of HBV and HCV, were combined to conduct univariate and multivariate Cox regression analyses to identify independent risk variables among patients with HCC. Using the “RMS” package of R software [34], we created a nomogram by integrating the independent risk factors for OS in TCGA cohort, such as the risk score, T stage, Stage, HBV, and HCV.

To distinguish between the actual outcomes and those predicted by the nomogram, calibration curves of the OS statistical likelihood at 1, 3, and 5 years were plotted. We employed the "timeROC" R package to plot the time ROC to assess the nomogram's ability to accurately predict patient outcomes [35]. The decision curve analysis (DCA) curve was plotted using the "RMDA" R program to assist in clinical decision-making for the acquisition of the best net benefit [36].

2.5. Pathway and Functional Enrichment Analyses

These two databases were used to perform operational analyses of Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) annotations and enrichment analyses of DEGs. This process was performed using the “clusterProfiler” package, and the “GOplot” package of “R” software was used to visualise the results [37, 38]. Through gene set enrichment analysis (GSEA), we examined the immunological state to determine the roles associated with the high- and low-risk groups. [39]. The selection criteria included enriched pathway items with P-values < 0.05, and false discovery rate (FDR) P-values < 0.25.

2.6. Immune Profiles of Gene Signature

Immune infiltration was estimated using TIMER [40], CIBERSORT [41], CIBERSORT-ABS, quanTIseq [42], xCell [43], MCP-counter [44], and EPIC [45] algorithms. The proportions of 22 different types of invading immune cells are shown as a bar graph for each HCC sample. Therefore, human leukocyte antigen (HLA) alleles in patients with tumours directly influence the effectiveness of immunotherapy. We analysed the association between HLA score and risk score groups and plotted a bar plot to visualise the data. Malta et al. calculated stemness indices by extracting transcriptomic and epigenetic information from tumour samples [46]. We then obtained tumour stemness data from TCGA database and assessed the stem-cell-like levels of tumour cells. According to a study by Charoentong et al., the immunophenoscore (IPS) is a biomarker for predicting how the body responds to checkpoint blockade and is used to predict the likelihood of patients responding to treatment with antibodies targeting CTLA-4 and PD-1 [47]. A robust response to checkpoint blockade has been detected in tumours that display microsatellite instability (MSI) [48]. By using the “Spearmans” method, we investigated the relationship between HCC patient's risk module gene expression and MSI score. We used tumour immune dysfunction and exclusion (TIDE) scores to examine tumour immune escape mechanisms, including tumour-infiltrating cytotoxic T lymphocytes (CTLs) and immune suppressor-mediated rejection of CTLs [49]. A poorer response to immune checkpoint blockade (ICB) therapy was predicted by higher TIDE scores [50].

2.7. Drug Sensitivity Analysis

Axitinib [51], A-443654 [52], BI-2536 [53], CGP-6047, cisplatin [54], epothilone B [55], etoposide [56], foretinib [57], fefitinib [58], gemcitabine [59], GSK212645, and JW-7-52-1 are considered crucial drugs to treat cancer. Moreover, we studied the expression of target genes of various drugs obtained from the DrugBank [60] (www.drugbank.ca) database. We then used the R package "pRRophetic" to identify the sensitivity and resistance of chemotherapeutic drugs and small molecule drugs. This software analyses the Genomics of Drug Sensitivity in Cancer (GDSC) database to determine the half-maximal inhibitory concentration (IC50) by ridge regression (www.cancerrxgene.org) [61]. It is defined as the 50% concentration at which apoptotic tumor cells are induced to die; the lower the IC50 value, the higher the drug-inducing ability and the more effective the drug is at treating tumors.

2.8. Verified the AARGs Expression in Other Database

To verified our signature, we choosed other database for analysis. Gene expression profile interactive analysis (GEPIA) (http://gepia.cancer-pku.cn/) [62], Cancer Cell Line Encyclopedia (CCLE) database (https://portals.broadinstitute.org/ccle/home) [63] and The Human Protein Atlas (THPA) database (https://www.proteinatlas.org/search) [64] which were validated the expression of three genes in tissue, cell and protein levels. Then we selected GSE 112790, GSE 78737 and GSE 14520 from GEO database to verify the expression in mRNA level.

2.9. Immunohistochemistry Analysis of Prognostic AARGs Protein Expression in Normal Liver and HCC Tissues

The Research Ethics Committee of Shandong Cancer Hospital approved the study protocol (approval number: SDTHEC2024003174). And our study conforms with The Code of Ethics of the World Medical Association (Declaration of Helsin-ki), printed in the British Medical Journal (18 July 1964). Written informed consent was obtained from individual or guardian participants. From the Shandong Province Tumour Hospital, we obtained nine pairs of HCC samples and adjacent normal liver tissues that had been fixed in 10% formalin and embedded in paraffin under pressure. Degreasing and immunohistochemical staining of the optimal tissue sections were performed. Antibodies against BIRC5 (Abcam, ab76424), MAPK3 (Cell Signaling, 4695) and BAK1 (Cell Signaling, 12105) were used in this study. Additionally, the immune profiles of the gene signatures allowed us to obtain immunohistochemical images of the three prognostic AARGs.

2.10. Statistical Analysis

All analytical methods and R packages were implemented using R (v4.4.0) software. Statistical significance was set at P < 0.05. significant. Log-rank tests and Kaplan–Meier analyses were used to compare the OS of the two risk groups. Using Mann–Whitney U tests, clinical characteristics were compared between the high- and low-risk groups. To determine whether the AARGs signature was independent of other clinical characteristics in patients with HCC, multivariate and univariate COX regression analyses were performed.

3.1. Identification of DE-AnAuRGs

A detailed flowchart of the experiments is shown in Fig. 1. The TCGA database included 50 normal and 371 HCC samples for RNA-seq analysis. A total of 2,897 differentially expressed genes (DEGs) were identified after the expressed gene datasets of HCC samples and normal subjects in TCGA were evaluated (Fig. 2a). The set of DEGs was overlapped with 434 AnRGs extracted from GeneCards and 232 AuRGs from the Human Autophagy Database, resulting in 13 differentially expressed (DE) anoikis–autophagy-related genes (AARGs): MAPK3, BAX, BIRC5, ITGA6, BAK1, CAPNS1, CDKN2A, PARP1, DAPK2, TSC2, IKBKE, ITGB4, and SPHK1 (Fig. 2b).

We performed an enrichment analysis of the dataset in the KEGG database and GO functional analysis to assess genetic differences between the two subgroups, in particular, whether bioactivity and signalling pathways are associated with risk scores. Based on the KEGG analysis, the DE-AARGs were found to be mainly associated with human papillomavirus infection, apoptosis, platinum drug resistance, and the intrinsic apoptotic signalling pathway (Fig. 2c). The most enriched GO terms were the intrinsic apoptotic signalling pathway, appendage development, limb development, and nuclear envelope (Fig. 2d).

On examining the changes in these genes, we discovered that 12 had a mutation rate of less than 1%, with CDKN2A having the highest rate (26%). The two most frequent mutations were missense and truncated mutations (Fig. 2e).

3.2. Development of the AARGs Signature for TCGA Cohort

We used the univariate Cox method to identify nine representative prognostic AARGs from TCGA cohort. Moreover, a significantly prognostic genes was constructed through multivariate Cox regression analysis of these nine prognostic-related genes, comprising four AARGs: BIRC5, MAPK3, BAK1, and IKBKE. (Figs. 3a). We then used LASSO-penalised regression to eliminate the overfitting of genes from significantly prognostic genes (Fig. 3b-c). Based on this analysis, three AARGs (BIRC5, MAPK3, and BAK1) were identified. To investigate the predictive significance of the candidate genes and their relationship with clinical and pathological findings, additional analyses of TCGA database were carried out. The mRNA expression levels of BIRC5, MAPK3, and BAK1 in patients with HCC were positively correlated with T stage, tumour stage, and tumour grade (Fig. 3d-l).

The following formula was used to construct the AARGs signature: Risk Score = [BIRC5 × (0.183637) + MAPK3 × (0.038556) + BAK1 × (0.065712)]. The patients were then divided into high- and low-risk groups according to the median risk score. Our proposed AARGs signature distinguishes the risk score and clinical state between HCC samples from TCGA cohort (Fig. 4a); a higher AARGs signature risk score signifies shorter overall survival and higher mortality among patients with HCC. Log-rank tests and Kaplan–Meier curves showed that patients in the high-AARGs signature risk group had poor outcomes (P < 0.001), including a greater mortality rate and shorter life (Fig. 4b). As shown in Fig. 4c, there is a strong predictive ability of the AARGs signature for the overall survival (OS) of patients with HCC (area under the curve of 1, 3, and 4 years was 0.734, 0.675, and 0.647, respectively).

3.3. Validation of the AARGs signature in the ICGC cohort

The AARGs signature scores for the 231 patients in the ICGC cohort were calculated using the same formula as above, and patients were categorised into low- or high-risk groups according to their median score values. Patient survival status and risk scores showed that patients in the high-risk group had shorter survival periods and lower recovery rates (Fig. 4d). According to data from the ICGC cohort, survival rates were lower, and survival times were shorter than those in TCGA, and patients with high-risk scores showed lower survival rates (Fig. 4e). As shown by the ROC analysis, the AARGs signature had high predictive power in the ICGC cohort. AUC values were 0.808 (1-year), 0.787 (3-year), and 0.777 (4-year) (Fig. 4f). Next, we tested whether patients in the low-risk groups had better outcomes than those in the high-risk groups; Kaplan–Meier analysis was performed on AARGs (P < 0.05) (Fig. 4g-i).

3.4. Differential Expression of AARGs in Clinical Subgroup

Using different clinical subgroups, we compared the expression of the three AARGs in patients with HCC to determine whether clinicopathological traits were connected to them, as in TCGA and ICGC cohorts. Patients with high-risk scores for high expression of the three AARGs in TCGA cohort were validated in the ICGC cohort (Fig. 5). In TCGA cohort, female patients showed high BIRC5 expression (P < 0.05). Patients in the HBV subgroup and BIRC5 and BAK1 in the C1 group had high expression (P < 0.05, P < 0.01, respectively). In the HCV subgroup, patients in the C1 group had high BAK1 expression (P < 0.05). All three of the genes were highly expressed in the high-risk group. (P < 0.001; Fig. 5a). In the ICGC cohort, female patients showed high expression of BAK1(P < 0.05). Patients in the TNM stage group and WHO grade group had high expression of three AARGs in T Stages 3–4 (BIRC5, MAPK3, and BAK1; P < 0.05, P < 0.01, and P < 0.001, respectively) and Grades 3–4 (BIRC5, MAPK3, and BAK1; P < 0.001, P < 0.01, and ns, respectively). In the ICGC cohort, HCC patients in the high-risk group showed high expression levels of all three genes (Fig. 5b). Additionally, the heatmap showed three AARGs that were expressed to a much greater degree in the high-risk group than in the low-risk group. With increasing risk scores, the three AARGs were more highly expressed. In addition, the expression levels of these genes were associated with sex, HBV infection, and HCV infection (Fig. 5c).

3.5. GSEA Analysis of Immune Status

To evaluate the immune status of TCGA cohort of patients with HCC, the GSEA gene sets were analysed in terms of CD8-T cell, IL4, Culture, CD4, and STEM. As can be seen from the results of the GSEA, the high-risk group exhibited a significantly enriched immune response (Fig. 5d, e and Supplementary Figure S1).

3.6. Stratification Analysis of the AARGs Signature

In order to determine whether the AARGs signature score was related to any clinical traits. We conducted comparatively analyse across the various subgroups which subsequent to categorization based on clinical characteristics in HCC patients. HCC patients with a worse T, Stage, or Grade tended to have higher AARGs risk scores. In the HBV group, patients infected with HBV (referred to as C1) had higher AARGs risk scores than the other patients (Fig. 6a-c). Furthermore, we performed survival analysis across several clinical groups to determine whether the AARGs signature has prognostic value in various subgroups of patients with HCC. Our Kaplan-Meier analysis suggested that the AARGs signature score was highly predictive in several clinical subgroups. The OS rate was better in low-risk HCC patients (those categorized as T I, Stage I, and III, and Grades 2 and 3) than in high-risk HCC patients (Fig. 6e-i). Similar results were obtained for HBV C1 (Fig. 6j). The aforementioned findings showed that the AARGs signature score established in this study is a reliable indicator of prognosis for patients with HCC across a wide range of clinical subgroups.

3.7. Constructing and Validating the Nomogram based on AARGs Signature

The multivariate and univariate Cox regression models, incorporating AARGs signature scores and other clinicopathological characteristics were assessed to determine whether the AARGs signature has independent prognostic value. Based on the results of the univariate Cox regression analysis, the Risk score (HR = 3.760 [2.210–6.399, 95%CI], p < 0.001), T stage (HR = 2.540 [1.785–3.613, 95%CI], p < 0.001), M stage (HR = 4.032 [1.267–12.831, 95%CI], p = 0.018), Total stage (HR = 2.449 [1.689–3.549, 95%CI], p < 0.001), HBV (HR = 1.803 [1.274–2.550, 95%CI], p < 0.001), and HCV (HR = 2.439 [1.692–3.517, 95%CI], p < 0.001) were found to be risk factors for OS in patients with HCC (Fig. 7a). Then we selected the item which was p < 0.05 to next multivariate regression analysis. Multivariate regression analysis showed that risk score (HR = 3.073 [1.763–65.356, 95%CI], p < 0.001) was the significant independent prognostic factor in HCC patients. No more than this, M stage (HR = 4.191 [1.227–14.308, 95%CI], p = 0.022), HBV (HR = 1.888 [1.108–3.217, 95%CI], p = 0.019), HCV (HR = 2.401 [1.300-4.436, 95%CI], p = 0.005) also was the independent prognostic factor in our research (Fig. 7b). Based on the above results, the AARGs signature was independently associated with the outcomes of patients with HCC.

A nomogram was created in TCGA cohort based on the AARGs signature and other independent prognostic factor criteria to improve the use of the AARGs signature in the clinic. (Fig. 7c).In this nomogram plot, we founded that high-Riskscore, The calibration curves demonstrate that the anticipated survival rates for 1, 3, and 5 years were very similar to the actual proportions (Fig. 7d). In addition, the nomogram had a significantly better prognostic predictive ability than did other factors (including AARGs signature score, M stage, HBV, and HCV) for patients with HCC, according to the ROC curves. In the TCGA cohort, the 1-, 3-, and 5-years AUCs were 0.648, 0.695, and 0.666, respectively. (Fig. 8e-g). DCA demonstrated a higher net benefit from using the nomogram for predicting adverse pathology compared to the M stage, HBV, and HCV scores (Fig. 7h-j). These findings demonstrate that our nomogram, based on the AARGs signature and clinical and pathological factors, is a reliable predictor of OS in patients with HCC.

3.8. Predictive Role of the AARGs Prognostic Signature in Immunotherapeutic Response

An additional investigation was conducted to determine whether the prognostic signature of the AARGs was associated with the efficacy of immunotherapy. The correlations between high- or low-risk groups and immune cells were examined. A CIBERSORT analysis of the TCGA cohort has revealed that whereas naive B cells, T cell CD4 + memory resting, and monocytes were more invasive in the low-risk group, T cell follicular helper, T cell regulatory (Tregs), and macrophage M0 were more invasive in the high-risk group (P < 0.0001, P < 0.01, P < 0.05, P < 0.0001, P < 0.0001, P < 0.0001, P < 0.05, and P < 0.0001) (Fig. 8a). We then compared the expression of HLA between the low- and high-risk groups. The expression of immunity-related terms was significantly higher in the high-risk group (Fig. 8b). Using the Wilcoxon test, we assessed the expression of several immunological checkpoints (ICs) in patients with low- and high-risk scores. Considerable increases in the expression of CD274, CTLA4, HAVCR2, LAG3, PDCD1, and TIGHT were observed in the group with a high risk score. (P < 0.05, P < 0.0001, P < 0.0001, P < 0.0001, P < 0.0001, and P < 0.0001) (Fig. 8c). However, regarding the expression of the IPS score, an increase in the low-risk score group occurred only when NEG-PD1-NEG was expressed. (P < 0.0001) (Fig. 8d). Our additional analysis revealed a substantial positive correlation between risk score and the degree of MSI, stemness, stroma, and ICBs. (P = 0.03, P < 0.01, P < 0.01, P < 0.01) (Fig. 8e, g, i, k). In addition, the degree of stromal invasion negatively correlated with the risk score (Fig. 8i). We compared the score of immune terms between patients with low- and high-risk groups. The results revealed that the high-risk group had considerably higher levels of MSI, Stemness, and TIDE than did the other groups. (P < 0.05, P < 0.0001, P < 0.0001, and P < 0.0001) (Fig. 8f, h, j, l). Additionally, the low-risk group showed a higher association with stromal cells than did the high-risk group. (Fig. 8j). A higher TIDE score indicates a poorer response to ICB. This suggests that the high-risk group may have an inferior responsiveness to immunotherapy. Therefore, early therapy is preferable in high-risk patients.

3.9. Chemotherapy Response with AARGs Signature

Additionally, we investigated whether chemotherapy was more effective in the high-risk group than in the low-risk group. We obtained information on axitinib, A-443654, BI-2536, CGP-6047, cisplatin, epothilone B, etoposide, foretinib, gefitinib, gemcitabine, GSK212645, and JW-7-52-1 from databases. Compared to the low-risk group, the high-risk group had lower IC50 values for axitinib (P < 0.01), A-443654 (P < 0.01), BI-2536 (P < 0.001), CGP-6047 (P < 0.001), cisplatin (P < 0.001), epothilone B (P < 0.001), etoposide (P < 0.001), foretinib (P < 0.001), gefitinib (P < 0.001), gemcitabine (P < 0.001), GSK212645 (P < 0.001), and JW-7-52-1 (P < 0.01) (Fig. 9a–l). These results indicate that chemotherapy has a greater effect on high-risk patients than on low-risk patients.

3.10. Validation of the AARGs Signature

To validate the accuracy of our signature, we evaluated the expression levels of the three genes in the CCLE, GEO, GEPIA, and HPA databases. We used GEPIA to identify the different expression of three genes between HCC and normal tissue in our study. HCC patients were found to have upregulated BIRC5, MAPK3, and BAK1 (Fig. 10a). The CCLE database contains data from 947 human cancer cell lines that have been deep-sequenced. At the HCC cell level, MAPK3 and BAK1 were expressed at low levels, as can be seen from the CCLE data (Fig. 10b). In GSE 112790, GSE 78737 and GSE 14520 from GEO database, we found that three genes was high expression in HCC patients (Fig. 10c-e). To confirm this result we searched for immunohistochemical data for BIRC5, MAPK3 and BAK1 in the HPA database. The immunohistochemical data show that HCC tissues exhibited significantly higher levels of BIRC5 and MAPK3 than adjacent tissues (Fig. 10f-i). But unfortunately, in the databases only have one type of HCC tissues with the expression of BAK1 (Fig. 10j).

2.11. Protein expression differences between normal liver tissue and HCC tissue

As shown in Fig. 11, we used imageJ to process our immunohistochemical staining pictures, in this way we calculated the positive sign in pixel count. Based on our study data, the expression of BIRC5 (Fig. 11a,b), MAPK3 (Fig. 11c,d) and BAK1 (Fig. 11e,f) was higher in tumor tissue than normal tissue. This results could verify the initial analysis which was patients with high-risk scores for high expression of the three AARGs.

Liver cancer commonly develops in hepatocytes and is called hepatocellular carcinoma (also known as primary liver cancer). As advances have been made in precision medicine, including oncology, improvements in clinical outcomes are expected [65]. Currently, a growing number of HCC patients are receiving cellular immunotherapy, indicating that HCC treatment has entered the era of immunotherapy. However, the complexity of patient conditions leads to an insufficient therapeutic response, drug resistance, and relapse during treatment. Currently, gene signatures, a type of biological function pattern created from the expression data of numerous genes, can be used to predict the prognosis and progression of several types of cancers [66, 67]. Here we constructed the AARGs signature associated with anoikis and autophagy and used the signature to predict the prognosis, disease progression, and therapy effectiveness in patients with HCC. The survival of HCC patients was accurately predicted using our signature.

Thirteen AARGs with differential expression were identified in this study, three of which were employed to create the AARGs signature using LASSO and Cox regression analyses. HCC patients could be classified into low- and high-risk groups with this signature. The ICGC, an independent validation cohort, confirmed the performance of this signature and demonstrated robust HCC survival prediction accuracy. Interestingly, in both the TCGA and ICGC cohorts, patients in the high-risk category exhibited lower tumour differentiation, progressed TNM stages, shorter survival durations, and higher fatality rates. HCC involves intricate mechanisms that are affected by a variety of variables, not simply AARGs, that aid in tumourigenesis. Consequently, we thought that additional important genes should be included in future signatures. Then we conducted the subgroup analyses and determined that the model could predict prognosis with precision, independent of individual clinical factor variations. In addition, model performance was validated in the TCGA and ICGC cohorts. The AARGs signature was found to be a significant independent prognostic factor. ROC analysis was performed to assess patient outcomes associated with time. In TCGA cohort, compared to long-term survival (4-year OS, AUC = 0.647), this model was fairly accurate at forecasting short-term survival (1-year OS, AUC = 0.734). Overall, the AARGs signature we developed in this study was demonstrated to be superior to the other prognostic signatures. To provide these patients with appropriate therapeutic strategies, such as using immunotherapy and changing the follow-up rate of individuals with high scores to track their performance, we can also use this sign to distinguish patients who have poor prognoses due to late-stage carcinoma stage at the time of clinical presentation.

It has previously been demonstrated that three genes [baculoviral IAP repeat containing 5 (BIRC5), mitogen-activated protein kinase 3 (MAPK3), and BRI1-associated kinase 1 (BAK1] ) in this signature are linked to the development of cancer. Jin et al. revealed a positive correlation exists between BIRC5 (survivalin) expression and NF-B expression in HCC tissues. It has been reported that both of these biomarkers are associated with poor prognoses in patients with HCC. Therefore, BIRC5 may be used as a biomarker to predict the prognosis of patients with HCC based on their biological characteristics [68]. Deng et al. revealed that MAPK1/3 regulates the degradation of ULK1 via the E3 ligase BTRC-mediated proteasomal degradation of its Lys48-linked ubiquitination, which contributes to the deregulation of ULK1 in breast cancer [69]. Xian et al. demonstrated that MIEF1 loss also results in mitochondrial depolarisation mediated by BAK1. As a consequence of pro-apoptotic effectors BAX and BAK1, MIEF1 plays a critical role in mitochondrial-dependent apoptosis [70]. In the present study, we combined these three genes to construct a prognostic signature. Further investigations are required to determine the prognostic value of these three AARGs in HCC and their underlying mechanisms.

Current research suggests that the tumour microenvironment (TME) is a setting that promotes the survival and growth of tumour cells [71]. The TME is composed of infiltrating stromal cells, immune cells, tumour cells, cytokines, and stromal microenvironments [72] and plays an essential role in tumour proliferation, angiogenesis, invasion, and metastasis. Consequently, more detailed information on the TME can lead to potential therapies to benefit patients with advanced HCC [73]. In the present study, we divided patients into high- and low-risk groups based on their signatures. According to our findings, patients with high-risk scores had superior immune responses than those in the low-risk group because they had more infiltrating Macrophages M0, T cell follicular helper cells, and T cell regulatory cells. Immunotherapy, especially ICB, has resulted in a paradigm shift in HCC treatment [74]. There was a substantial difference between the two patient groups receiving effective care when the immune checkpoint genes (CD274, CTLA4, HAVCR2, LAG3, PDCD1, and TIGHT) were stratified by risk score. However, in this signature, high-risk patients have poor treatment options for ICB therapy. We conducted GSEA for these two groups of patients to evaluate their immune status. This result is similar to the above results. We performed drug sensitivity analysis of our signature using the IC50 which can identify the sensitivity of chemotherapeutic and small-molecule drugs. The responses to various chemotherapeutic treatments (etoposide, foretinib, gefitinib, gemcitabine) varied significantly between the high- and low-risk groups. In our analysis, low-risk patients had an increased monocyte count, T cell CD4 + memory, and naive B cell count, which confirms previous reports of immunological profiles in HCC cases [75–77], such profiles are indicative of of negative prognosis. in these patients. The high-risk groups show aggressive tumour growth caused by stemness [78] and CTLA4 [79]. Overall, these findings are consistent with our findings.

There are currently a few published prognostic models based on AnRGs [17] or AuRGs [23]. Anoikis or autophagy alone does not provide good accuracy or sensitivity for prognosis because of the diversity of HCC and the complexity of its microenvironment. In our study, we first combined AnRGs and AuRGs to construct an AARGs signature that is robust in predicting the prognosis and progression of HCC. By establishing a nomogram, the AARGs signature could be easily used in clinical settings. We conducted an immune analysis of our signature to provide individualised treatment options for patients with HCC. However, this study was limited by the following factors: 1) There is always some bias in retrospective studies, which is why additional verification of this signature through multicentre datasets is warranted; 2) To better understand these three AARGs, comprehensive functional studies are required to shed light on the detailed mechanisms. The next stage of our research will involve analysing the interaction mechanisms and clinical and bioinformatics data.

Ethics approval and consent to participate:

Consent for publication:

All authors had been obtained from Consent for publication.

Availability of data and materials:

If you would like to review the materials, data, or protocols that support the findings of this study, you can contact the corresponding authors. Corresponding author: Zengjun Li, Department of endoscopy, Shandong Cancer Hospital and Institute, Shandong First Medical University and Shandong Academy of Medical Sciences, No. 440 Jiyan Road, Jinan City 250117, Shandong Province, China; E-mail: [email protected].

Competing interests:

Neither the authors nor their collaborators have any commercial or financial relationships that might be construed as conflict of interest.

Funding:

This work was supported by the Natural Science Fund of Shandong Province (ZR2020LZL003).

Author Contributions:

Zengjun Li, Hao Wang conceived and designed the study. Hao Wang, Kai Cui, Mingang He and Yang Gao analysed the data and generated the figures. Bingbing Ren and Changsheng Yan performed the immunohistochemical staining. Hao Wang and Mingang HE wrote the initial draft of the manuscript. Zengjun Li conceptualized and revised the manuscript. A version of the article was submitted with the approval of all authors.

Acknowledgements:

All authors are to be commended for their collaboration.

Informed Consent Statement:

Informed consent was provided by all participants before the survey started, by indicating that they agreed with the statement of consent.

Yang X, Zhang D, Liu S, Li X, Hu W, Han C. KLF4 suppresses the migration of hepatocellular carcinoma by transcriptionally upregulating monoglyceride lipase. American journal of cancer research. 2018;8(6):1019-29.
McGlynn KA, London WT. The global epidemiology of hepatocellular carcinoma: present and future. Clinics in liver disease. 2011;15(2):223-43, vii-x.
de Boer MT, Molenaar IQ, Porte RJ. Impact of blood loss on outcome after liver resection. Digestive surgery. 2007;24(4):259-64.
Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer Statistics, 2021. CA: a cancer journal for clinicians. 2021;71(1):7-33.
Mayakonda A, Lin DC, Assenov Y, Plass C, Koeffler HP. Maftools: efficient and comprehensive analysis of somatic variants in cancer. Genome research. 2018;28(11):1747-56.
Tajbakhsh A, Rivandi M, Abedini S, Pasdar A, Sahebkar A. Regulators and mechanisms of anoikis in triple-negative breast cancer (TNBC): A review. Critical reviews in oncology/hematology. 2019;140:17-27.
Han HJ, Sung JY, Kim SH, Yun UJ, Kim H, Jang EJ, et al. Fibronectin regulates anoikis resistance via cell aggregate formation. Cancer letters. 2021;508:59-72.
Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic acids research. 2015;43(7):e47.
Zhou X, Li L, Guo X, Zhang C, Du Y, Li T, et al. HBXIP induces anoikis resistance by forming a reciprocal feedback loop with Nrf2 to maintain redox homeostasis and stabilize Prdx1 in breast cancer. NPJ breast cancer. 2022;8(1):7.
Xiao T, Xu Z, Zhou Y, Zhang H, Geng J, Liang Y, et al. Loss of TP53I11 Enhances the Extracellular Matrix-independent Survival by Promoting Activation of AMPK. IUBMB life. 2019;71(2):183-91.
Venetsanakos E, Mirza A, Fanton C, Romanov SR, Tlsty T, McMahon M. Induction of tubulogenesis in telomerase-immortalized human microvascular endothelial cells by glioblastoma cells. Experimental cell research. 2002;273(1):21-33.
Middlemas DS, Lindberg RA, Hunter T. trkB, a neural receptor protein-tyrosine kinase: evidence for a full-length and two truncated receptors. Molecular and cellular biology. 1991;11(1):143-53.
Zhong X, Rescorla FJ. Cell surface adhesion molecules and adhesion-initiated signaling: understanding of anoikis resistance mechanisms and therapeutic opportunities. Cellular signalling. 2012;24(2):393-401.
She K, Yang W, Li M, Xiong W, Zhou M. FAIM2 Promotes Non-Small Cell Lung Cancer Cell Growth and Bone Metastasis by Activating the Wnt/β-Catenin Pathway. Frontiers in oncology. 2021;11:690142.
McCormick KD, Ghosh A, Trivedi S, Wang L, Coyne CB, Ferris RL, et al. Innate immune signaling through differential RIPK1 expression promote tumor progression in head and neck squamous cell carcinoma. Carcinogenesis. 2016;37(5):522-9.
Zhang X, Cheng SL, Bian K, Wang L, Zhang X, Yan B, et al. MicroRNA-26a promotes anoikis in human hepatocellular carcinoma cells by targeting alpha5 integrin. Oncotarget. 2015;6(4):2277-89.
Chen Y, Lin QX, Xu YT, Qian FJ, Lin CJ, Zhao WY, et al. An anoikis-related gene signature predicts prognosis and reveals immune infiltration in hepatocellular carcinoma. Frontiers in oncology. 2023;13:1158605.
Mizushima N, Levine B, Cuervo AM, Klionsky DJ. Autophagy fights disease through cellular self-digestion. Nature. 2008;451(7182):1069-75.
Scott RC, Juhász G, Neufeld TP. Direct induction of autophagy by Atg1 inhibits cell growth and induces apoptotic cell death. Current biology : CB. 2007;17(1):1-11.
Li X, He S, Ma B. Autophagy and autophagy-related proteins in cancer. Molecular cancer. 2020;19(1):12.
Li Y, Hahn T, Garrison K, Cui ZH, Thorburn A, Thorburn J, et al. The vitamin E analogue α-TEA stimulates tumor autophagy and enhances antigen cross-presentation. Cancer research. 2012;72(14):3535-45.
Yamamoto K, Venida A, Yano J, Biancur DE, Kakiuchi M, Gupta S, et al. Autophagy promotes immune evasion of pancreatic cancer by degrading MHC-I. Nature. 2020;581(7806):100-5.
Fang Q, Chen H. Development of a Novel Autophagy-Related Prognostic Signature and Nomogram for Hepatocellular Carcinoma. Frontiers in oncology. 2020;10:591356.
Long XH, Zhou YF, Lan M, Huang SH, Li Liu Z, Shu Y. Valosin-containing protein promotes metastasis of osteosarcoma through autophagy induction and anoikis inhibition via the ERK/NF-κβ/beclin-1 signaling pathway. Oncology letters. 2019;18(4):3823-9.
Palorini R, Votta G, Pirola Y, De Vitto H, De Palma S, Airoldi C, et al. Protein Kinase A Activation Promotes Cancer Cell Resistance to Glucose Starvation and Anoikis. PLoS genetics. 2016;12(3):e1005931.
Wu XN, Su D, Mei YD, Xu MQ, Zhang H, Wang ZY, et al. Identified lung adenocarcinoma metabolic phenotypes and their association with tumor immune microenvironment. Cancer immunology, immunotherapy : CII. 2021;70(10):2835-50.
Rebhan M, Chalifa-Caspi V, Prilusky J, Lancet D. GeneCards: integrating information about genes, proteins and diseases. Trends in genetics : TIG. 1997;13(4):163.
Deng Y, Zhu L, Cai H, Wang G, Liu B. Autophagic compound database: A resource connecting autophagy-modulating compounds, their potential targets and relevant diseases. Cell proliferation. 2018;51(3):e12403.
Walter W, Sánchez-Cabo F, Ricote M. GOplot: an R package for visually combining expression data with functional analysis. Bioinformatics (Oxford, England). 2015;31(17):2912-4.
Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Science signaling. 2013;6(269):pl1.
Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer discovery. 2012;2(5):401-4.
Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics : a journal of integrative biology. 2012;16(5):284-7.
Friedman J, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models via Coordinate Descent. Journal of statistical software. 2010;33(1):1-22.
Kramer AA, Zimmerman JE. Assessing the calibration of mortality benchmarks in critical care: The Hosmer-Lemeshow test revisited. Critical care medicine. 2007;35(9):2052-6.
Malta TM, Sokolov A, Gentles AJ, Burzykowski T, Poisson L, Weinstein JN, et al. Machine Learning Identifies Stemness Features Associated with Oncogenic Dedifferentiation. Cell. 2018;173(2):338-54.e15.
Kerr KF, Brown MD, Zhu K, Janes H. Assessing the Clinical Impact of Risk Prediction Models With Decision Curves: Guidance for Correct Interpretation and Appropriate Use. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2016;34(21):2534-40.
Fabbri L, Dufies M, Lacas-Gervais S, Gardie B, Gad-Lapiteau S, Parola J, et al. Identification of a new aggressive axis driven by ciliogenesis and absence of VDAC1-ΔC in clear cell Renal Cell Carcinoma patients. Theranostics. 2020;10(6):2696-713.
Blanche P, Dartigues JF, Jacqmin-Gadda H. Estimating and comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Statistics in medicine. 2013;32(30):5381-97.
Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proceedings of the National Academy of Sciences of the United States of America. 2005;102(43):15545-50.
Li T, Fan J, Wang B, Traugh N, Chen Q, Liu JS, et al. TIMER: A Web Server for Comprehensive Analysis of Tumor-Infiltrating Immune Cells. Cancer research. 2017;77(21):e108-e10.
Newman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, et al. Robust enumeration of cell subsets from tissue expression profiles. Nature methods. 2015;12(5):453-7.
Finotello F, Mayer C, Plattner C, Laschober G, Rieder D, Hackl H, et al. Molecular and pharmacological modulators of the tumor immune contexture revealed by deconvolution of RNA-seq data. Genome medicine. 2019;11(1):34.
Aran D, Hu Z, Butte AJ. xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome biology. 2017;18(1):220.
Becht E, Giraldo NA, Lacroix L, Buttard B, Elarouci N, Petitprez F, et al. Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome biology. 2016;17(1):218.
Racle J, Gfeller D. EPIC: A Tool to Estimate the Proportions of Different Cell Types from Bulk Gene Expression Data. Methods in molecular biology (Clifton, NJ). 2020;2120:233-48.
Li X, Li Y, Yu X, Jin F. Identification and validation of stemness-related lncRNA prognostic signature for breast cancer. Journal of translational medicine. 2020;18(1):331.
Charoentong P, Finotello F, Angelova M, Mayer C, Efremova M, Rieder D, et al. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell reports. 2017;18(1):248-62.
Cristescu R, Mogg R, Ayers M, Albright A, Murphy E, Yearley J, et al. Pan-tumor genomic biomarkers for PD-1 checkpoint blockade-based immunotherapy. Science (New York, NY). 2018;362(6411).
Jiang P, Gu S, Pan D, Fu J, Sahu A, Hu X, et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nature medicine. 2018;24(10):1550-8.
Cohen R, Hain E, Buhard O, Guilloux A, Bardier A, Kaci R, et al. Association of Primary Resistance to Immune Checkpoint Inhibitors in Metastatic Colorectal Cancer With Misdiagnosis of Microsatellite Instability or Mismatch Repair Deficiency Status. JAMA oncology. 2019;5(4):551-5.
Kudo M, Kang YK, Park JW, Qin S, Inaba Y, Assenat E, et al. Regional Differences in Efficacy, Safety, and Biomarkers for Second-Line Axitinib in Patients with Advanced Hepatocellular Carcinoma: From a Randomized Phase II Study. Liver cancer. 2018;7(2):148-64.
Han EK, Leverson JD, McGonigal T, Shah OJ, Woods KW, Hunter T, et al. Akt inhibitor A-443654 induces rapid Akt Ser-473 phosphorylation independent of mTORC1 inhibition. Oncogene. 2007;26(38):5655-61.
Sebastian M, Reck M, Waller CF, Kortsik C, Frickhofen N, Schuler M, et al. The efficacy and safety of BI 2536, a novel Plk-1 inhibitor, in patients with stage IIIB/IV non-small cell lung cancer who had relapsed after, or failed, chemotherapy: results from an open-label, randomized phase II clinical trial. Journal of thoracic oncology : official publication of the International Association for the Study of Lung Cancer. 2010;5(7):1060-7.
Li P, Song R, Yin F, Liu M, Liu H, Ma S, et al. circMRPS35 promotes malignant progression and cisplatin resistance in hepatocellular carcinoma. Molecular therapy : the journal of the American Society of Gene Therapy. 2022;30(1):431-47.
Li F, Huang T, Tang Y, Li Q, Wang J, Cheng X, et al. Utidelone inhibits growth of colorectal cancer cells through ROS/JNK signaling pathway. Cell death & disease. 2021;12(4):338.
Hande KR. Etoposide: four decades of development of a topoisomerase II inhibitor. European journal of cancer (Oxford, England : 1990). 1998;34(10):1514-21.
Fujino T, Suda K, Koga T, Hamada A, Ohara S, Chiba M, et al. Foretinib can overcome common on-target resistance mutations after capmatinib/tepotinib treatment in NSCLCs with MET exon 14 skipping mutation. Journal of hematology & oncology. 2022;15(1):79.
Paez JG, Jänne PA, Lee JC, Tracy S, Greulich H, Gabriel S, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science (New York, NY). 2004;304(5676):1497-500.
Muggia F, Diaz I, Peters GJ. Nucleoside and nucleobase analogs in cancer treatment: not only sapacitabine, but also gemcitabine. Expert opinion on investigational drugs. 2012;21(4):403-8.
Knox C, Law V, Jewison T, Liu P, Ly S, Frolkis A, et al. DrugBank 3.0: a comprehensive resource for 'omics' research on drugs. Nucleic acids research. 2011;39(Database issue):D1035-41.
Yang W, Soares J, Greninger P, Edelman EJ, Lightfoot H, Forbes S, et al. Genomics of Drug Sensitivity in Cancer (GDSC): a resource for therapeutic biomarker discovery in cancer cells. Nucleic acids research. 2013;41(Database issue):D955-61.
Tang Z, Li C, Kang B, Gao G, Li C, Zhang Z. GEPIA: a web server for cancer and normal gene expression profiling and interactive analyses. Nucleic acids research. 2017;45(W1):W98-w102.
Nusinow DP, Szpyt J, Ghandi M, Rose CM, McDonald ER, 3rd, Kalocsay M, et al. Quantitative Proteomics of the Cancer Cell Line Encyclopedia. Cell. 2020;180(2):387-402.e16.
Uhlen M, Oksvold P, Fagerberg L, Lundberg E, Jonasson K, Forsberg M, et al. Towards a knowledge-based Human Protein Atlas. Nature biotechnology. 2010;28(12):1248-50.
Mathai RA, Vidya RVS, Reddy BS, Thomas L, Udupa K, Kolesar J, et al. Potential Utility of Liquid Biopsy as a Diagnostic and Prognostic Tool for the Assessment of Solid Tumors: Implications in the Precision Oncology. Journal of clinical medicine. 2019;8(3).
Cheng Q, Huang C, Cao H, Lin J, Gong X, Li J, et al. A Novel Prognostic Signature of Transcription Factors for the Prediction in Patients With GBM. Frontiers in genetics. 2019;10:906.
Peng Z, Chen Y, Cao H, Zou H, Wan X, Zeng W, et al. Protein disulfide isomerases are promising targets for predicting the survival and tumor progression in glioma patients. Aging. 2020;12(3):2347-72.
Jin Y, Chen J, Feng Z, Fan W, Wang Y, Li J, et al. The expression of Survivin and NF-κB associated with prognostically worse clinicopathologic variables in hepatocellular carcinoma. Tumour biology : the journal of the International Society for Oncodevelopmental Biology and Medicine. 2014;35(10):9905-10.
Deng R, Zhang HL, Huang JH, Cai RZ, Wang Y, Chen YH, et al. MAPK1/3 kinase-dependent ULK1 degradation attenuates mitophagy and promotes breast cancer bone metastasis. Autophagy. 2021;17(10):3011-29.
Xian H, Liou YC. Loss of MIEF1/MiD51 confers susceptibility to BAX-mediated cell death and PINK1-PRKN-dependent mitophagy. Autophagy. 2019;15(12):2107-25.
Ovais M, Mukherjee S, Pramanik A, Das D, Mukherjee A, Raza A, et al. Designing Stimuli-Responsive Upconversion Nanoparticles that Exploit the Tumor Microenvironment. Advanced materials (Deerfield Beach, Fla). 2020;32(22):e2000055.
Kessenbrock K, Plaks V, Werb Z. Matrix metalloproteinases: regulators of the tumor microenvironment. Cell. 2010;141(1):52-67.
Pitt JM, Marabelle A, Eggermont A, Soria JC, Kroemer G, Zitvogel L. Targeting the tumor microenvironment: removing obstruction to anticancer immune responses and immunotherapy. Annals of oncology : official journal of the European Society for Medical Oncology. 2016;27(8):1482-92.
Llovet JM, Castet F, Heikenwalder M, Maini MK, Mazzaferro V, Pinato DJ, et al. Immunotherapies for hepatocellular carcinoma. Nature reviews Clinical oncology. 2022;19(3):151-72.
Li X, Yao W, Yuan Y, Chen P, Li B, Li J, et al. Targeting of tumour-infiltrating macrophages via CCL2/CCR2 signalling as a therapeutic strategy against hepatocellular carcinoma. Gut. 2017;66(1):157-67.
Chen D, Li X, Li H, Wang K, Tian X. Identification of Immune-Related Prognostic mRNA and lncRNA in Patients with Hepatocellular Carcinoma. Journal of oncology. 2022;2022:5313149.
He Y, Luo J, Zhang G, Jin Y, Wang N, Lu J, et al. Single-cell profiling of human CD127(+) innate lymphoid cells reveals diverse immune phenotypes in hepatocellular carcinoma. Hepatology (Baltimore, Md). 2022;76(4):1013-29.
Yamashita T, Wang XW. Cancer stem cells in the development of liver cancer. The Journal of clinical investigation. 2013;123(5):1911-8.
Wang L, Li N, Fan X, Wang X, Zhang X, Zhang K, et al. Circulating CTLA-4 levels and CTLA4 polymorphisms associate with disease condition and progression and hepatocellular carcinoma patients' survival in chronic hepatitis B virus infection. International immunopharmacology. 2020;82:106377.

No competing interests reported.

supplement.jpg

Download PDF

Reviewers invited by journal
24 Oct, 2024
Editor assigned by journal
22 Oct, 2024
Editor invited by journal
06 Aug, 2024
Submission checks completed at journal
02 Aug, 2024
First submitted to journal
24 Jul, 2024

You are reading this latest preprint version

Construction and Validation of a Combined Anoikis and Autophagy Prognostic Signature for Hepatocellular Carcinoma

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials and Methods

2.1. Data Acquisition, Differential Expression Analysis, and Intersection Identification

2.2. Development and Validation of a Prognostic Signature based on AARGs

2.3. Stratified Analysis by Clinical Characteristics

2.4. Constructing and Validating a Predictive Nomogram

2.5. Pathway and Functional Enrichment Analyses

2.6. Immune Profiles of Gene Signature

2.7. Drug Sensitivity Analysis

2.8. Verified the AARGs Expression in Other Database

2.9. Immunohistochemistry Analysis of Prognostic AARGs Protein Expression in Normal Liver and HCC Tissues

2.10. Statistical Analysis

3. Results

3.1. Identification of DE-AnAuRGs

3.2. Development of the AARGs Signature for TCGA Cohort

3.3. Validation of the AARGs signature in the ICGC cohort

3.4. Differential Expression of AARGs in Clinical Subgroup

3.5. GSEA Analysis of Immune Status

3.6. Stratification Analysis of the AARGs Signature

3.7. Constructing and Validating the Nomogram based on AARGs Signature

3.8. Predictive Role of the AARGs Prognostic Signature in Immunotherapeutic Response

3.9. Chemotherapy Response with AARGs Signature

3.10. Validation of the AARGs Signature

2.11. Protein expression differences between normal liver tissue and HCC tissue

4. Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1