Pan-Cancer Analysis of the Prognostic and Immunological Role of SMG5: A Biomarker for Cancers Free

SMG5 gene, located on chromosome 1, encodes the SMG5 protein containing multiple structural domains, including a C-terminal sequence, an N-terminal sequence, a central fusion sequence, and a DUF1768 sequence [1]. SMG5 protein is a key component of the RNA degradation pathway [2], mainly involved in nonsense-mediated mRNA decay (NMD) regulation, and is a member of the NMD complex, acting in concert with SMG7, SMG6, UPF1, and other proteins to complete the recognition and degradation of mRNAs containing early termination codons [2‒4]. Moreover, it is also involved in the formation and regulation of the NMD complex, regulating the rate and efficiency of mRNA degradation through the interaction of specific structural domains with other proteins, thus ensuring the normal translation and stability of mRNA [5, 6].

In addition to its important role in the RNA degradation pathway, SMG5 protein also exerts important effects on other biological processes [7]. For example, SMG5 protein also play a role in cell division, mitotic progression, as well as chromosome stability and segregation, and abnormal SMG5 function may contribute to the development and progression of cancer [8, 9]. The researchers have revealed that SMG5 is involved in the proliferation and metastasis of tumor cells [10]. A study has found that SMG5 expression level is significantly elevated in a variety of cancers, including liver, pancreatic, and gastric cancer, and are closely associated with tumor cell proliferation and metastasis [11‒13]. Furthermore, it is also potential for SMG5 to promote tumor cell proliferation and metastasis through the activation of signaling pathways, such as the Wnt/β-catenin pathway [2]. High expression of SMG5 is strongly correlated with a poor prognosis of some tumors, making it a new target for tumor therapy [14]. Several studies have shown that targeting SMG5 can inhibit tumor cell proliferation and metastasis and promote tumor cell apoptosis [12]. Researchers have elucidated that the use of SMG5 siRNA works on the proliferation and migration ability of lung cancer cells and promotes apoptosis [15]. In the meantime, it has also been shown that targeting SMG5 can enhance the efficacy of chemotherapy by increasing the sensitivity of tumor cells to chemotherapeutic agents [13]. Immunotherapy has been a high-profile tumor treatment in recent years, and SMG5 is also believed to play a role in immunotherapy [13]. Researchers have found that SMG5 is associated with the recognition and degradation of tumor-associated antigens, and targeting SMG5 can promote the recognition and clearance of tumor cells by immune cells and improve the efficacy of immunotherapy [16]. In summary, numerous studies have illustrated that SMG5 expression affects cancer development, and targeting SMG5 therapy can effectively mitigate cancer progression.

To the authors’ knowledge, no studies have reported the role of SMG5 in pan-cancer to date. Therefore, in this study, a systematic analysis of SMG5 expression, prognosis, and immune function in pan-cancer was conducted by retrieving multiple databases to explore the role of SMG5 in cancer development and further provide new insights and theoretical basis for the search of new tumor biomarkers and potential therapeutic targets.

As shown in Figure 1, this study focused on the analysis of SMG5 expression in tumor and normal tissues. Specifically, we used transcriptional RNA sequencing (RNA-seq) data from The Cancer Genome Atlas (TCGA) to analyze 33 cancer types, and data from the Cancer Cell Line Encyclopedia (CCLE, https://portals.broadinstitute.org/ccle/) database were used to identify each tumor cell line and the expression levels in 21 tissues. In addition, gene expression data related to 31 different tissues were retrieved from the Genotype-Tissue Expression database (GTEx, https://commonfund.nih.gov/GTEx). Afterward, the data were converted to Transcripts Per Million (TPM) format, log2 transformed, and statistical tests were performed to estimate the differences in SMG5 expression levels between normal and tumor tissues [17].

Gene Expression Profiling Interactive Analysis (GEPIA, http://gepia.cancer-pku.cn/) is a free interactive tool for gene expression profiling using samples from the TCGA and GTEx databases. Kaplan-Meier (KM) analysis was used in the survival module of GEPIA to assess the prognostic value of the SMG5 gene in cancer, including overall survival (OS, time from diagnosis to death from any cause) and disease-free survival (DFS, time without signs of cancer after treatment). In addition, an online database was used to validate the relationship between SMG5 expression and overall cancer prognosis. Firstly, the survival analysis was carried out by Cox proportional hazard regression test. The results were visualized by drawing the forest map using Sangerbox 3.0 mapping tool (http://sangerbox.com/) [18]. Besides, the median SMG5 expression was selected as the survival value analyzed by cloud platform. Finally, the Oncolnc database (http://www.oncolnc.org/) was used to conduct a prognostic analysis of each cancer again.

Tumor Immune Estimation Resource (TIMER, https://cistrome.shinyapps.io/timer/) is a tool that enables systematic analysis of immune infiltration in different types of cancer using inverse fold product statistics [19] and can be used to infer tumor-infiltrating lymphocyte (TIL) counts based on gene expression data [20]. This study examined the correlation between SMG5 expression levels and the number of six types of immune infiltrating cells (i.e., differentiated CD4⁺ T cells, CD8⁺ T cells, B cells, neutrophils, dendritic cells, and macrophages) using the TIMER algorithm. To investigate the relationship characteristics between immune cell infiltration and SMG5 expression levels, TIMER, CIBERSORT, and xCELL algorithms were applied for correlation analysis.

Tumor-immune system interactions database (TISIDB, http://cis.hku.hk/TISIDB/index.php) [21] is an automated platform, based on a large number of high-throughput screens, exome and RNA sequencing data from immunotherapy patient cohorts, and data from TCGA, designed to analyze tumor-immune system interactions [21]. We investigated the correlation of SMG5 expression with immune and molecular subtypes, TME, immunomodulator-related genes, and major histocompatibility complex (MHC) molecules using the TISIDB system. To explore the relationship between TME and SMG5 expression levels, stromal and immune cell scores in malignant tumor tissues was calculated using the ESTIMATE algorithm, and the correlation between SMG5 and immune scores was assessed using Pearson’s correlation coefficients.

Tumor mutational burden (TMB), microsatellite instability (MSI), and neoantigen load were calculated using single nucleotide variant data from patients to estimate the relationship between SMG5 expression and these factors. In addition, the expression of SMG5 in TME and the association of SMG5 expression with the expression of immune checkpoint (ICP) genes and immunotherapy biomarkers were analyzed through the Sangerbox website. The ESTIMATE score was used to evaluate the relationship between SMG5 expression and stromal and immune cells in malignant tumor tissues [22].

SMG5 mRNA expression levels and cell sensitivity data z-scores (GI50) in 60 different cancer cell lines were retrieved by accessing the NCI-60 database using the CellMiner interface (https://discover.nci.nih.gov/cellminer/) [23], and Pearson’s correlation analysis was used to analyze the correlation between SMG5 expression levels and drug scores.

With the gene expression data from the TCGA cohort, a differential expression analysis was conducted on the SMG5 gene, in which there were more than 20 cancers, except for the absence of normal tissue data (Fig. 2a). Comparing the TCGA and GTEx groups, SMG5 was highly expressed in 23 cancers, including low-grade glioma (LGG), and lowly expressed in 4 cancers (Fig. 2b; Table 1). Besides, the CCLE and GTEx data revealed differentially expression of SMG5 in cancer cell lines (online suppl. Fig. S1; for all online suppl. material, see https://doi.org/10.1159/000533421).

In order to investigate the link between SMG5 expression and the prognosis of cancer, we conducted OS and DFS prognostic analyses on cancer patients with high SMG5 expression. Cox analysis suggested that SMG5 might be a risk factor for LGG (p < 0.001), liver hepatocellular carcinoma (LIHC, p < 0.001), acute myeloid leukemia (LAML, p < 0.001), pheochromocytoma and paraganglioma (PCPG, p < 0.001), adrenocortical carcinoma (ACC, p = 0.02), and kidney chromophobe (KICH, p = 0.04) (Fig. 3a). KM survival analysis showed that high SMG5 expression was significantly associated with poorer OS in patients with ACC (p = 0.0095), LGG (p < 0.001), and LIHC (p < 0.001), and with better OS in patients with pancreatic adenocarcinoma (PAAD, p = 0.039). However, KM survival analysis of DFS showed ACC (p = 0.001), LGG (p < 0.001), PRAD (p = 0.023), and uveal melanoma (UVM, p = 0.016) with poorer prognosis in the SMG5 high expression group (online suppl. Fig. S2). Based on the results of these three prognostic analyses, we found that only patients with ACC and LGG had a worse prognosis when SMG5 was highly expressed. In addition, the KM survival in high and low SMG5 expression groups was assessed by the OncoLnc database, and the results indicated that among patients with ACC and LGG, only LGG (p < 0.001) patients with high SMG5 expression had poor prognosis. Similarly, the prognostic analysis of the GEPIA, Sangerbox, and Ualcan databases was similar to the prognostic analysis of the SMG5 gene described above. Taken together, SMG5 may be a risk gene for most cancers and has a significant prognostic impact on LGG patients in particular (Fig. 3b–d).

Then link between SMG5 expression and the immune cell abundance was assessed. First, the relation in SMG5 expression with lymphocyte and immunomodulatory factor abundance was compared. Afterward, immune-related characteristics of 28 TILs from Charoentong’s study were evaluated [24], showing the relationship between the abundance of TILs and the expression, copy number, and methylation of SMG5. More precisely, in LGG, 14 of 28 TILs were significantly positively correlated with SMG5 expression, and 4 types of TILs were negatively correlated with SMG5 expression. Most of the TILs were negatively correlated with the copy number of SMG5. Ten TILs were negatively correlated with the methylation of SMG5, one TIL was positively correlated with the methylation of SMG5, and the correlation of other cell types with the methylation of SMG5 was not significant (online suppl. Fig. S3a). We then analyzed the correlations of three immunomodulators (including immunosuppressants, immunostimulants, and MHC molecules) with the expression, copy number, and methylation of SMG5. Scatter plots demonstrated the strongest positive and negative correlations of immunosuppressants and immunostimulants with SMG5 expression, copy number, and methylation. The results showed significant correlations between the expression, copy number, and methylation of SMG5 and some of the genes related to immunosuppressants, immunostimulants, and MHC molecules, respectively. Collectively, SMG5 expression affects the expression of immunomodulator genes in various cancers (online suppl. Fig. S3B, C, D).

Subsequently, the results of the analysis of the correlation between SMG5 expression and immune cell infiltration showed that in LGG, tumor purity (p < 0.001) was significantly negatively correlated with SMG5 expression, and the levels of B cell (p < 0.001), CD8⁺ T cell (p < 0.001), CD4⁺ T cell (p < 0.001), Macrophage (p < 0.001), Neutrophil (p < 0.001) and Dendritic Cell (p < 0.001) infiltration level were significantly positively correlated with SMG5 expression (Fig. 4a). We then analyzed the prognostic value of these immune cells, indicating that LGG patients had a poor prognosis for the presence of high abundance of B cells, CD8⁺ T cells, CD4⁺ T cells, macrophages, and neutrophils (Fig. 4b).

According to an analysis on the TISIDB website, SMG5 expression was significantly associated with the immune subtypes of 30 tumors and the molecular subtypes of 17 tumors. More importantly, SMG5 expression levels were closely associated with the four immune subtypes of LGG, with SMG5 highly expressed in C4 and lowly expressed in C5 (Fig. 5a; online suppl. Fig. S4). Molecular subtypes of LGG were also significantly correlated with SMG5 expression (Fig. 5b; online suppl. Fig. S5). Based on these findings, we conclude that SMG5 expression varies between immune subtypes and molecular subtypes of cancers, with significant differences, especially in LGG.

The correlation of SMG5 expression with TMB, MSI, and neoantigens was examined in order to study the role of SMG5 in the tumor microenvironment (TME). Studies have reported that antitumor immunity is correlated with TMB, MSI, and neoantigen immunotherapy efficacy in TME [25]. The results of our study showed that SMG5 expression correlated positively with TMB and MSI in LGG (Fig. 6a, b), and no significant correlation of SMG5 expression with neoantigens was detected (Fig. 6c). Afterward, the relationship between SMG5 expression and the three estimates was explored (Fig. 6d). Compared to other cancer immunotherapy markers, SMG5 expression was correlated positively with LGG immune microenvironment markers. Our results further supported the hypothesis that SMG5 may affect LGG’s antitumor immunity by regulating TME compositions and immune mechanisms.

Researchers have found that ICP genes play a crucial role in the regulation of immune cell infiltration and cancer therapy [23]. We further analyzed the relationship between SMG5 expression and ICP genes and found that all tumors had a part of ICP genes associated with SMG5 expression, and 29 out of 47 ICP genes in LGG were associated with SMG5 expression (Fig. 7; online suppl. Table S1, S2), highlighting the potential of SMG5 in immunotherapy. Therefore, SMG5 may act as a pan-cancer biomarker or a new LGG immunotherapeutic target.

We analyzed the correlation of SMG5 expression with drug sensitivity to more than 200 chemotherapeutic agents in the NCI-60 database. We found that while SMG5 expression levels were positively correlated with drug sensitivity of procarbazine and maltoglucoside (Fig. 8a, e), they were negatively correlated with oxaliplatin, everolimus, and paclitaxel drug sensitivity (Fig. 8b–d).

Cancer is one of the main causes of death in the world, leading to >10 million deaths each year [26]. In recent years, researchers have worked to find biomarkers for pan-cancer [27, 28], but the role of SMG5 in pan-cancer has not been reported. Studies based on biogenic analysis and pathological diagnosis have found that SMG5 is highly expressed in patients with hepatocellular carcinoma, and there is a high degree of consistency between predicted and actual diagnosis status [12, 14]. Another study has confirmed that lncRNA DDIT4-AS1 effectively inhibits the binding of SMG5 to PP2A, leading to increased p-UPF1 levels in pancreatic ductal adenocarcinoma cells, increased stemness of pancreatic ductal adenocarcinoma cell, and suppression of sensitivity to the potent drug gemcitabine [13]. In other tumors, although there are no clear studies on the association between SMG5 expression and tumor cells, UPF1 phosphorylation plays an important role in promoting its cellular activity [29], and SMG5 is an important NMD factor involved in the regulation of UPF1 phosphorylation [30]. Therefore, SMG5 gene expression in 27 cancers was analyzed using TCGA, GTEx data, and we found that SMG5 was highly expressed in 23 of these 27 cancers, and the results of LIHC and PAAD were consistent with previous studies. This suggested that SMG5 may be a potential oncogene for most cancers, and its high expression may promote tumor development by enhancing cell stemness.

Prognostic analysis expands new ideas for the clinical prognostic management of patients and the search for potential targets for the treatment of cancer [31]. Tang et al. [12] have pointed out that SMG5 high expression is associated with shorter survival in HCC patients. More interestingly, SMG5 high expression is related to the poorer prognosis in black male with gastric cancer [11]. In our study, after the analysis and validated by multiple databases and algorithms, SMG5 expression was associated with significant prognostic outcomes in many cancers, and in particular, LGG patients with high SMG5 expression were shown a poorer prognosis in all algorithms. Taken together, SMG5 may play a pro-cancer role in tumors and is closely associated with cancer prognosis, especially LGG. Differently, in DFS, high SMG5 expression was associated with the better prognosis in patients with PAAD, which may be related to the underlying diseases and the number of patients included in databases [32].

Tumor-infiltrating immune cells play an important role in tumor development and can counteract or promote the occurrence of tumors [33]. Previous studies identified a significant correlation between SMG5 expression and immune cells and their degree of infiltration in patients with LIHC in the TCGA-LIHC cohort [12, 34]. In inflammatory myofibroblastic tumors, SMG5 has been proved to regulate NMD activity as an NMD factor, and downregulation of NMD activity leads to increased levels of chemokines, which in turn promote immune cell infiltration [35, 36]. Similarly, in our study, SMG5 expression correlated significantly and positively with the level of LGG immune cell infiltration including B cells, CD8⁺ T cells, CD4⁺ T cells, macrophages, neutrophils, and dendritic cells, and negatively with tumor purity. The higher the abundance of immune infiltrating cells, the poorer the prognosis of LGG patients. These provided evidence for SMG5 as a possible novel therapeutic target for LGG immunotherapy.

Several studies have reported that different immune and molecular subtypes [33] can drive immunotherapy outcomes [37]. In colon cancer, six immune subtypes are associated with different clinicopathological features, molecular alterations, specific enrichment of supervised gene expression profiles, and dysregulation of signaling pathways [38]. In breast cancer brain metastases, HER2 and TNBC molecular subtypes provide evidence that brain metastatic disease supports dysregulation of metabolic reprogramming and immune response pathways, revealing pathways with biological and potential therapeutic implications [39]. In addition, we explored the expression of SMG5 in different immune and molecular subtypes of human cancers to determine its potential mechanism of action. The results showed that SMG5 expression was significantly different in different immune and molecular subtypes of most tumors. In LGG, SMG5 expression levels were closely correlated with its 4 immune subtypes, significantly higher in C4 and lower in C5, and differential expression was also found in its 6 molecular subtypes. These suggested that SMG5 may be a promising pan-cancer diagnostic biomarker and involved in immune regulation. More importantly, this may provide insights into the future development of immunotherapy for LGG patients.

The total number of mutations within the coding region of the genome is referred to as the TMB that has been identified as a potential biomarker for various types of cancers [40, 41]. Previous studies have shown that TMB can be used as a biomarker to improve the effectiveness of immunotherapy in non-small-cell lung and colorectal cancers [42]. MSI is characterized by abnormal microsatellite sequences caused by impaired DNA mismatch repair, which is associated with an increased risk of developing cancer [43]. High-frequency MSI in colorectal cancer is an independent predictor of clinical features and prognosis [44]. Both TMB and MSI are also intrinsically linked to the sensitivity of ICP inhibitors [27]. Our study showed that SMG5 expression was positively correlated with TMB and MSI of LGG, indicating that SMG5 expression levels affect TMB and MSI of LGG and patients’ response to ICP inhibition therapy. This may provide a new theoretical reference for LGG immunotherapy. Furthermore, numerous studies have highlighted the importance of the immune microenvironment in cancer development. Some studies have reported that stromal scores and immune cell scores in tumor tissues may be associated with clinical features and that the ESTIMATE Immunity Score is expected to serve as a predictor of response to immunotherapy [45]. In our study, we assessed the relationship between SMG5 expression and immune score and stromal score of tumors using the ESTIMATE algorithm and found that SMG5 expression was significantly associated with both immune score and stromal score in most cancers. Similar results were found in LGG as well. These findings underscore the value of studying the interaction between SMG5 expression and the immune microenvironment and provide new perspectives to advance more effective therapeutic interventions. In addition, we further investigated the effect of SMG5 expression on ICPs in cancer patients. SMG5 expression was found to be significantly correlated with the expression of most ICPs in tumor tissues, and 29 out of 47 ICP genes in LGG were positively correlated with SMG5 expression. And many studies have also shown that cancer treatment can be greatly improved through immunotherapy [46‒48]. Obstructing the ICP pathway shows great promise for anti-LGG immunity [49].

Finally, since the treatment of LGG in recent years mainly relies on radiotherapy techniques and chemotherapeutic agents [50, 51], we predicted LGG-related drug sensitivity based on the expression of SMG5 in cell lines and retrieval of data about NCI-60 cell lines and showed that SMG5 expression levels may correlate with the chemosensitivity of several FDA-approved drugs (e.g., procarbazine, oxaliplatin, everolimus, paclitaxel, and vermoxanin), and all these findings tentatively inferred that SMG5 may play a potentially critical role in the chemosensitivity or drug resistance of cancer cells, and therefore, it may serve as a promising target against drug resistance in various cancers.

In conclusion, we found that SMG5 was widely and differentially expressed between tumor and normal tissues by pan-cancer analysis, revealing the correlation between SMG5 expression and clinical prognosis. Our findings suggested that SMG5 had the potential to be an independent prognostic factor for many tumors and that SMG5 expression levels may vary in different types of tumors. In addition, SMG5 expression was significantly correlated with various aspects of immune and molecular subtypes of most tumors, immunotherapeutic biomarkers, immune scores of tumors, and expression of ICP-associated genes, especially in LGG, and we also examined it in terms of immune cell infiltration level of the tumor and drug sensitivity. This predicts that SMG5 may be a good indicator in cancer, which can reveal the occurrence of cancer in vivo from the side and support the diagnosis of tumor well. In particular, it can be used as a prognostic marker for LGG and may provide new clues for LGG immunotherapy. However, there are still some limitations to our study. We mainly focused on bioinformatics to analyze the effect of SMG5 expression on tumors and did not perform in vivo and in vitro validation. In the future, SMG5 can be continue to be investigated as a new target at cellular and molecular levels, which will probably help to find new and precise therapeutic options for LGG.

We are appreciated to all participants and the contributors in this study.

Ethical approval and consent were not required as this study was based on publicly available data.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

No funding was used for this research.

Leteng Yang and Tianbo Jin drafted the manuscript; Jie Wei performed the data analysis; Huan Zhang was responsible for the sample collection and information recording; Xiaoya Ma performed the manuscript modification; and Leteng Yang and Tianbo Jin conceived and supervised the study.

Our research data sets are available from corresponding authors upon reasonable request. The data set for this study can be found in the database listed in the article. There was no necessity to obtain Ethics Committee approval since all information was publicly available and open-access.