Prognostic value of sirtuin family members and experimental verification identify SIRT5 as diagnostic biomarkers in clear cell renal cell carcinoma

Timer

TIMER is a comprehensive resource for systematical analysis of immune infiltrates across diverse cancer types (Li et al., 2020). We used this database to get the scatterplots that show the correlation of SIRTs with six types of infiltrated immune cells and their signature markers.

The human protein atlas

The Human Protein Atlas (HPA) is a useful database based on proteomic, transcriptomic and systems biology data, it can map tissues, cells, and organs (Thul & Lindskog, 2018). Not only tumor tissues, but also the protein expression of normal tissues is included. The survival curves of tumor patients can also be viewed. We used this database to obtain the immunohistochemistry images of SIRTs in ccRCC and in normal tissues.

Immunohistochemistry

We purchased a tissue microarray containing slides of 90 ccRCC tumor tissues and adjacent normal kidney tissue from Outdo Biotech (HKidE180Su03; Outdo Biotech, Shanghai, China) with patients’ clinical records, including histopathological diagnoses and prognostic information. We used SIRT5 (Proteintech, #15122–1-AP, 1:200) as primary antibodies, and incubated with it overnight at 4 °C. The sections were then incubated with the secondary antibodies and DAB regents for staining (Zhong Shan Golden Bridge Biotechnology, Beijing, China). The results of IHC were scored by two independent pathologists respectively according to both staining intensity that contain negative, weak, moderate and strong, corresponding to 0, 1, 2, and 3, respectively and percentage of immunoreactive tumor cells (<5% = 0, 5–25% = 1, 26–50% =2, 51–75% = 3, >75% =4). The final expression level score was obtained by multiplying the above two scores (ranging from 0 to 12). The samples with score <3 were considered as negative expression, and ≥ 3 were considered as positive expression.

Data analysis

Abnormally-increased expression of SIRT family members in patients with ccRCC

Then, we used the Human Protein Atlas database to evaluate the protein expression levels of the seven SIRT family members in normal and ccRCC tissues. As shown in Fig. 1, the immunohistochemistry (IHC) staining data from the THPA database showed increased protein expression levels of SIRT1, SIRT2, SIRT3, SIRT6, and SIRT7, whereas the expression of SIRT4 and SIRT5 was significantly lower (particularly SIRT5), in ccRCC tissues. Thus, we speculated that SIRT5 may be an important tumor suppressor gene.

Clinical-pathological characterizations of SIRT5 and other SIRT family members with ccRCC

Next, we want to explore the influence of SIRT5 and other SIRT family members on the clinical–pathological characteristics of patients with ccRCC.

To understand disease prognosis and outcomes, the UALCAN database was used to investigate the protein expression of SIRT5 (Fig. 2A) and the relationship between it and patients’ age, ccRCC individual tumor stages, and grades (Fig. 2B). The results showed that the protein expression of SIRT5 was significantly lower in tumor tissue compared to normal tissue and was negatively related to the age of the patient ccRCC individual tumor stages, and grades. To further verify it, we performed IHC on a tissue microarray containing slides with patients’ clinical records (including histopathological diagnoses and prognostic information). And in 90 human ccRCC sample, strong IHC staining expression of SIRT5 was displayed in adjacent normal tissue than in tumor tissues (Fig. 2C showed the representative images). Based on the IHC staining results, we made univariate (Table 1) and multivariate (Table 2) analysis results between SIRT5 protein expression and clinicopathological features. By using univariate analysis, aberrant expression of SIRT5 in ccRCC was closely related with pathological grading and cancer size with significant statistically difference (p < 0.05), among them, the p-value of 0.0614 was greater than 0.05 under the chi-square test for pathological grading index, but this study was based on a small sample, so the p-value of 0.0487 from Fisher’s exact test was adopted. By using Logistic regression test, we found that age (p < 0.005), pathological grading (p < 0.005), T stage (p < 0.005) and SIRT5 expression (p < 0.05) were four independent risk factors (visualization chart is shown in Fig. 2D). Consequently, we believe that SIRT5 plays important role in tumor suppression, and closely related to better prognosis of patients with ccRCC.

Further, we explored the function of SIRT1, SIRT2, SIRT3, SIRT4, SIRT6, and SIRT7 of patients with ccRCC. Studies have shown that the expression of SIRT2, SIRT3, SIRT6, and SIRT7 is increased in tumor-stage 1-4 subgroups, SIRT1 expression is increased in tumor-stage 1, and the expression of SIRT4 is decreased (Fig. 3). The expression of SIRTs with different tumor grades was the same. These results indicate that SIRT4 may play important roles in tumor suppression, whereas the other family members could be involved in driving the progression of ccRCC.

Association between prognosis and SIRTs in patients with ccRCC

Next, we used the Kaplan–Meier database to generate overall survival (OS) curves and recurrence free survival (RFS) curves to explore the prognostic value of the SIRT family for patients with ccRCC (Figs. 4A and 4B). The OS and interactive correlations suggested that for patients with ccRCC, high SIRT4 and SIRT5 expression was significantly related to a better OS (P < 0.05). However, high SIRT6 and SIRT7 expression was shown to be related to worse OS (P < 0.05). The RFS results suggested that high SIRT3 expression was related to worse RFS (P < 0.05), whereas high SIRT5 levels were related to better RFS (P < 0.05). Therefore, SIRT4 and SIRT5 potentially inhibit the occurrence and development of ccRCC, whereas high expression of SIRT3, SIRT6, and SIRT7 might have an important tumor-promoting role.

SIRT DNA methylation levels in patients with ccRCC

Methylation is an important modification of proteins and nucleic acids, which regulates the expression and silencing of genes and is closely related to various diseases such as cancer. The study of methylation could be helpful for finding suitable drug targets for treatment. To explore new mechanisms to target for ccRCC treatment, we used the DiseaseMeth database to further investigate the levels of DNA methylation of SIRTs in patients with ccRCC. As shown in Fig. 5, the methylation of SIRT4 and SIRT5 was increased, whereas that of the SIRT6 and SIRT7 was decreased. Combined with SIRT expression (mRNA and protein) and the prognostic value of SIRTs (mentioned previously herein), we speculate that SIRT4 and SIRT5 might be possible tumor suppressor markers, whereas SIRT6 and SIRT7 could play an important role in promoting tumor development.

Functional enrichment analysis of the SIRTs-associated coexpressed genes

To explore the mechanism underlying the functions of SIRTs in ccRCC, we used the cBioPortal and Metascape databases to perform functional analysis. More than 20000 SIRT-correlated genes were downloaded. According to a —log ratio—>1.4 and P <0.05, 216 of these were further screened (Table S1). By using the STRING database and Cytoscape, the protein–protein interaction (Fig. 6A) was constructed; Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis was then conducted by using the WebGestalt database. We found that 10 potential signaling pathways were enriched, including collecting duct acid secretion, renin-angiotensin, pantothenate and CoA biosynthesis, cocaine addiction, Vibrio cholerae infection, glutathione metabolism, drug metabolism, amphetamine addiction, tight junction, and metabolic pathways. Among them, the identification of the drug metabolism pathway suggested that the SIRT family has the potential to be a drug target for ccRCC therapy (Fig. 6B). Further, the Metascape and WebGestalt databases were used to analyze Gene Ontology (GO) term enrichment (Fig. 6C). Results shown in Fig. 6C suggested that more than half of the genes were enriched in the membrane. We found that biological regulation was the most enriched biological process category, then was the metabolic process, response to stimulus, and multicellular organismal process.

Immune cell infiltration related to SIRTs

In recent years, the immune system has been considered a crucial aspect in understanding oncogenesis and cancer progression (Qu et al., 2022). Studies targeting SIRTs have shown that several members of this family are significantly associated with the immune response.

Combined with our speculation on the relationship between SIRTs and the immune system, the TIMER database was used to identify any correlations between tumor-infiltrating immune cells and the SIRT family (Figs. 7A–7G). We found that six SIRT family members (all except SIRT4) were related to several common immune cells, including B cells, CD8⁺ T cells, CD4⁺ T cells, macrophages, neutrophils, and dendritic cells. The expression of SIRT1 was positively correlated with all six immune cells. SIRT2 levels were positively correlated with CD8⁺ T cells (Cor = 0.148, P = 1.84e−03), and CD4⁺ T cells (Cor = 0.146, p =1.68e−03). Figure 7C shows that SIRT3 expression was negatively correlated with B cells (Cor = −0.159, P = 6.38e−04), CD4⁺ T cells (Cor = −0.102, P = 2.86e−02), macrophages (Cor = −0.339, P = 1.56e−13), neutrophils (Cor = −0.215, P = 3.57e−06), and dendritic cells (Cor = −0.258, P = 7.23e−04). SIRT5 levels were found to be positively correlated with macrophages (Cor = 0.214, P = 5.05e−06) and neutrophils (Cor = 0.096, P = 3.95e−02), as shown in Fig. 7E. SIRT6 expression was found to be negatively correlated with B cells (Cor = −0.173, P = 2.03e−04), macrophages (Cor = −0.249, P = 9.27e−08), neutrophils (Cor = −0.129, P = 5.86e−03), and dendritic cells (Cor = −0.111, P = 1.77e−02) but positively correlated with CD4+ T cells (Cor = 0.145, P = 1.82e−03). The expression of SIRT7 was positively correlated with CD4+ T cells (Cor = 0.294, P = 1.36e−10) and neutrophils (Cor = 0.138, P = 3.14e−03). These data suggest that SIRTs might have an important role in ccRCC through their influence on the infiltration of different immune cells by some certain mechanism. For example, the upregulation of SIRT3 and SIRT6 promoted ccRCC may be mediated by with the increasing immune cell infiltration levels, including B cells, CD4+ T cells, macrophages, neutrophils, and dendritic cells. The downregulation of SIRT5 may be related to the infiltration of macrophages and neutrophils.