Tumor and adjacent normal tissues of COAD, rectum adenocarcinoma (READ), and liver hepatocellular carcinoma (LIHC) were obtained from the Department of General Surgery, Xiangya Hospital. The research protocols were approved by the Research Ethics Committee of Xiangya Hospital. All participants gave written consent of their tissue samples and medical information to be analyzed for scientific research. The total RNA of tissues was isolated and purified by AG RNAex Pro Reagent, AG21101 (Accurate Biology, China), following the manufacturer’s instructions. Quantitative PCR (qPCR) analysis of samples was performed using EvoM-MLV RT Kit, AG11711 (Accurate Biology, China) following the manufacturer’s instructions. ACTB1 was used as the internal control. The primers were designed as follows:

ACTB1:

Forward 5′-GGC​ATT​CAC​GAG​ACC​ACC​TAC-3′.

Reverse 5′-CGA​CAT​GAC​GTT​GTT​GGC​ATA​C-3′.

ISYNA1:

Forward 5′-CAG​CAC​CGG​GTT​TTT​GTG​G-3′.

Reverse 5′-TCC​TTA​GAG​CGG​AAC​TGC​AAT-3′.

CD274 (PDL1):

Forward 5′-TGG​CAT​TTG​CTG​AAC​GCA​TTT-3′.

Reverse 5′- TGC​AGC​CAG​GTC​TAA​TTG​TTT​T-3′.

The expression profiles and clinical information of the Cancer Genome Atlas (TCGA) and Cancer Cell Line Encyclopedia (CCLE) were downloaded from the UCSC Xena (https://xenabrowser.net/datapages/) database. The infiltration level of different immune cells in TCGA was downloaded from the ImmuCellAI database (http://bioinfo.life.hust.edu.cn/ImmuCellAI#!/).

The pan-cancer expression of ISYNA1 in TCGA cohort was performed using the TIMER2 database (http://timer.comp-genomics.org/). The cBioPortal database was used to analyze the CNA and mutation status of ISYNA1. The correlation between ISYNA1 expression and infiltration level CAFs/TAMs was performed using the TIMER2 database.

The association between the ISYNA1 expression level and all mRNAs was analyzed using TCGA RNAseq data, and Pearson’s correlation coefficient was calculated. The mRNAs correlated with ISYNA1 (p < 0.05) were ranked and subjected to the GSEA. The analysis was conducted using the R package “clusterProfiler.”

The TME gene sets were downloaded, and the scores of each gene set in TCGA pan-cancer were calculated using the method from this published study (Zeng et al., 2019).

First, the correlation between ISYNA1 expression and infiltration level CAFs/TAMs was performed using the TIMER2 database. Second, we downloaded the immune cell infiltration data of TCGA pan-cancer from the ImmuCellAI database and performed a correlation analysis between ISYNA1 expression and infiltration levels of 24 immune cells.

We downloaded the RNA expression profiles of cancer cell lines and IC50 information of 192 anticancer drugs from the GDSC database (https://www.cancerrxgene.org/). Then, we calculated the association between ISYNA1 expression and IC50 of anticancer drugs.

First, we investigated ISYNA1 expression using the TIMER2 database. We found that ISYNA1 was highly expressed in 12 among 33 cancers types, including BLCA, BRCA, CHOL, COAD, ESCA, GBM, HNSC, LIHC, LUAD, LUSC, STAD, and THCA. In addition, ISYNA1 was lowly expressed in only four cancer types, such as KICH, KIRC, KIRP, and PRAD (Figure 1A). To evaluate the expression of ISYNA1 only in tumor tissues from TCGA cohort, we found that ISYNA1 expression was highest in tumor tissues of UCS and lowest in tumor tissues of COAD compared with other tumor tissues (Figure 1B). In normal tissues from GTEx database, we found that ISYNA1 expression was highest in testis tissue and lowest in blood (Figure 1C).

Regarding the paired tumor and adjacent normal tissues in TCGA cohort, we observed that ISYNA1 was over-expressed in 10 cancers, such as BRCA, CHOL, ESCA, HNSC, LIHC, LUAD, LUSC, STAD, THCA, and BLCA (Figures 2A–J). In addition, ISYNA1 was lowly expressed in KICH, KIRC, KIRP, and PRAD (Figures 2K–N), which was consistent with the above results. In addition, we verified the expression of ISYNA1 by qRT-PCR in COAD, READ, and LIHC. Results revealed that ISYNA1 was highly expressed in COAD and READ (Figures 2O–Q).

We further investigated ISYNA1 expression at different tumor stages. The results revealed that ISYNA1 expression was different in the various tumor stages in COAD, ESCA, HNSC, LIHC, READ, STAD, THCA, and ACC (Figures 3A–H).

Mutation, DNA methylation, and CNA are main factors that influence gene expression. Thus, we assessed the mutations, DNA methylation, and CNA of ISYNA1. We revealed that the highest alteration frequency of ISYNA1 (>6%) was observed in OV patients, in which the “amplification” was the primary type (Figure 4A). For the correlation between ISYNA1 and CNA, we found that ISYNA1 expression was positively correlated with CNA in 17 among 33 cancers types (Figure 4B), indicating that the high CNA was one of the main causes of high ISYNA1 expression in pan-cancer, except in COAD. In addition, we also found that the promoter methylation level of ISYNA1 was negatively correlated with ISYNA1 expression in 31 among 33 cancers types (Figure 4C), including COAD (r = −0.33), indicating that the promoter methylation level of ISYNA1 mainly affects ISYNA1 mRNA expression in OCAD.

To assess the prognostic value of ISYNA1 in pan-cancer, the univariate Cox regression analysis (UniCox) and Kaplan–Meier survival analysis were conducted. Results of the UniCox indicated that ISYNA1 acts as a risk factor for overall survival (OS) of patients with ACC, KIRC, LGG, and SKCM, and acts as a protective factor for OS in patients with OV, PAAD, and PCPG (Figure 5A). The Kaplan–Meier OS analysis proved that the elevated expression of ISYNA1 predicted worse OS of patients with ACC, COAD, LAML, LGG, READ, SKCM, and STAD. In contrast, a high ISYNA1 expression predicted longer OS time in patients with MESO, OV, and PCPG (Figure 5B).

We also assessed the prognostic value of ISYNA1 in predicting disease-free interval (DFI), progression-free interval (PFI), and disease-specific survival (DSS) of tumor patients using univariate Cox regression analysis. For DFI, a high ISYNA1 expression predicted shorter DFI times in patients with CESC and PRAD (Figure 6A). For PFI, a high ISYNA1 expression predicted shorter PFI times in patients with ACC, COAD, LGG, and PRAD, and longer PFI time in patients with PAAD and THYM (Figure 6B). For DSS, a high ISYNA1 expression predicted a worse DSS status in patients with ACC, COAD, LGG, and SKCM, and a better DSS status in patients with OV and PCPG (Figure 6C).

To explore the potential pathways for the potential involvement of ISYNA1 in tumor progression, we further conducted the GSEA of ISYNA1 using TCGA pan-cancer data. Interestingly, we found common GSEA results for various tumor types. The immune regulation relevant pathways, such as “adaptive immune system,” “innate immune system,” “signaling by interleukins,” and “cytokine signaling in immune system,” were simultaneously enriched in LGG, MESO, PRAD, and SKCM (Figure 7).

We further analyzed the role of ISYNA1 in TME. Through difference analysis and correlation analysis, we found that ISYNA1 was positively correlated with TME-related pathways mainly in OV, LGG, LIHC, and COAD (Figures 8A–E). For example, the scores of immune system-related pathways and tumor stroma-related pathways were higher in the high-ISYNA1 expression group in COAD (Figure 8E). By analyzing the association between ISYNA1 expression and immune cell infiltration using the TIMER2 database, we found that ISYNA1 expression was positively associated with the infiltration of CAFs and TAMs in COAD (Figures 9A,B). Using immune cell infiltration data from the ImmuCellAI database, we found that ISYNA1 expression was positively correlated with immunosuppressive cells, such as TAMs and iTreg (Figure 9C). These results indicated that high ISYNA1 expression may contribute to the tumor immunosuppressive microenvironment. To further explore the relationship between ISYNA1 and TIME, we conducted a correlation analysis among immunosuppressive genes, immune checkpoints, and ISYNA1. The results showed that ISYNA1 expression was positively correlated with immune checkpoints, such as PDCD1, TIGIT, LAG3, CTLA4, and KLRB1, in COAD, LGG, LIHC, and OV (Figures 10A–D). We also observed that ISYNA1 was positively correlated with MHC genes (Figure 10E), chemokines and chemokine receptors (Figures 10F,G), and immunosuppressive genes (Figure 10H) in COAD. Interestingly, there was no correlation between ISYNA1 and CD274 (PD-L1) expression, which was validated by qRT-PCR in COAD, READ, and LIHC (Supplementary Figures S1A–C). These results revealed that ISYNA1 expression predicted the immunosuppressive microenvironment.

We downloaded half-inhibitory concentration (IC50) values of 192 drugs and gene expression profiles of 809 cell lines from the Genomics of Drug Sensitivity in Cancer database (GDSC: https://www.cancerrxgene.org/) and analyzed Spearman’s correlation between ISYNA1 expression and IC50 values. Among 192 anticancer drugs, the IC50 values of 26 drugs were positively correlated with ISYNA1 expression, while only the IC50 values of three drugs were negatively correlated with ISYNA1 expression (Supplementary Table S1, Figures 11A–D). These results revealed that patients with high ISYNA1 expression were predicted to be resistant to most anticancer drugs.