ALDOB is a prognostic biomarker and a potential immunotherapy target for clear cell renal cell carcinoma

Data source

The Cancer Genome Atlas (TCGA), a publicly accessible database containing comprehensive data from a large-scale cancer genome project, provides researchers and academicians with clinicopathological data on 33 distinct cancer types. By using the TCGA browser, we compiled RNA-Seq data and matched clinicopathological information from 271 individuals with ccRCC exhibiting elevated ALDOB level, and 270 patients with low levels of ALDOB expression. Additionally, we evaluated their correlation based on the clinicopathological characteristics of the individuals. This study was approved by the Ethics Committees of The Fifth Affiliated Hospital, Southern Medical University, with the IRB approval number: 2023-MNWK-K-003.

External validation from the GEO and HPA databases

To verify the differential expression of ALDOB mRNA between ccRCC tissues and adjacent normal tissues, we downloaded the expression profiling data of GSE66271 and GSE53757 from the GEO datasets and extracted the expression data of ALDOB, which were grouped into tumor and normal groups. Then, independent-sample t-test was performed for data analysis, and R package ggplot2 [3.4.4] was used for data visualization. Additionally, protein immunohistochemistry (IHC) data are available for tumors and normal tissues through the Human Protein Atlas (HPA). Our investigation focused on analyzing the ALDOB protein levels in both human cancerous and normal tissues.

Clinicopathological analysis of ALDOB

To address how ALDOB expression was found to be associated with clinical parameters in ccRCC, we obtained and organized RNAseq data and clinical data of the TCGA-KIRC project from the TCGA database, specifically extracting TPM format data. After removing normal samples and those lacking clinical information, the data were processed using log2(value+1). The Kruskal-Wallis test, supported by R packages stats[4.2.1] and car[3.1-0], was then applied to assess the associations between ALDOB mRNA expression and clinical-pathological parameters, including pathological tumor (T) stage, lymph node (N) stage, metastasis (M) stage, and histological grading in ccRCC cases. Data visualization was performed using the ggplot2 [3.4.4] package.

Development and verification of nomograms

The prognostic features encompassing OS, progression-free interval (PFI), and DSS were evaluated utilizing Cox regression and Kaplan–Meier (KM) methodologies within the clinical significance module of the Xiantao platform (https://www.xiantaozi.com). Our computation of the truncation value for the high and low ALDOB expressions was grounded in the median number. The association between clinicopathological characteristics and ALDOB was determined via the Wilcoxon signed rank sum test and logistic regression analysis. Through the utilization of multivariate Cox analyses, we examined survival rates and various clinical characteristics. A P < 0.05 was established as the significance criterion. The multivariate analysis yielded independent prognostic markers, which were subsequently employed to predict the chances of survival over 1, 3, and 5 years. The 45° line represents the best prediction based on comparing predicted probabilities to observed events.

Functional enrichment analysis

The functional enrichment of ALDOB-associated genes was conducted using the clusterProfiler package in R (v3.6.3), including Gene Ontology (GO) terms: biological process (BP), cellular component (CC), and molecular function (MF). GO analysis was performed with a minimum gene count >3, enrichment factor >1.5, and P < 0.01. In addition, Gene Set Enrichment Analysis (GSEA) was performed to identify pathways significantly associated with ALDOB expression. Genes were ranked based on their log2 fold change between high and low ALDOB expression groups. Statistical significance in GSEA was assessed using a permutation test with 1,000 permutations, which is a commonly used default. In this procedure, sample or gene labels are randomly shuffled 1,000 times to generate a null distribution of enrichment scores (ES), against which the observed ES is compared to calculate the P-value. More permutations lead to more accurate P-value estimation but also increase computation time. Pathways were considered significantly enriched with an adjusted P < 0.05 and false discovery rate (FDR) <0.25. Normalized enrichment scores (NES), adjusted P-values, and FDR were used to report enrichment significance. GSEA enrichment analysis and visualization were conducted utilizing the Cluster Profiler package (Yu et al., 2012).

Immune infiltration analysis

Bindea G’s study (Bindea et al., 2013) identified marker genes for 24 different kinds of immune cells, which were subsequently analyzed for infiltration into tumors using ssGSEA. Using Spearman correlation analysis, we studied the immune cell infiltration in cohorts exhibiting elevated and reduced ALDOB levels and examined the association of ALDOB with the 24 immune cell types. We applied the immune infiltration module of the “Xiantao tool” (https://www.xiantaozi.com; accessed on 1 January 2023) to complete the above analysis.

TIMER database analysis

Tumor Immune Estimation Resource (TIMER2.0) database was utilized for further analysis of immune infiltration in different cancer types, offering a user-friendly online platform for such investigations (https://timer.cistrome.org/) (Li et al., 2020). A total of 10,897 specimens from the TCGA database comprising 32 types of cancer were evaluated for immune internal infiltrates. Moreover, the database is capable of accurately estimating tumor purity as well.

Tissue microarray and immunohistochemistry

ALDOB expression was measured using microarrays comprising 80 ccRCC tissues and 80 adjacent normal tissues (Zhuoli Biotechnology Co, Shanghai, China). In summary, after using a normal serum to block the slides, they were incubated with an anti-ALDOB antibody from Proteintech and left at 4 °C for the night. After that, peroxidase-conjugated avidin-biotin complex and biotinylated secondary antibodies were added (Zhuoli Biotechnology Co, Jiaozhu, China). 3,3-diaminobenzidine (Zhuoli Biotechnology Co., Jiaozhu, China) was utilized as a chromogen for visualizing the immunostaining, with hematoxylin serving as a counterstain. The H-Score method was employed to measure the intensity of ALDOB staining, computed as H-Score = ∑ (pi × i) = (% of weak intensity × 1) + (% of moderate intensity × 2) + (% of strong intensity ×3). The collection and use of pathological tissues were obtained with written informed consent from the patients.

Tumor samples have lower levels of ALDOB expression than normal tissues

The TCGA dataset was utilized to determine whether ALDOB’s low expression is widespread in cancer. Cancers such as cholangiocarcinoma (CHOL), esophageal carcinoma (ESCA), and kidney renal papillary cell carcinomas (KIRP) display down-regulated levels of ALDOB expression, whereas colon adenocarcinoma (COAD) and rectum adenocarcinoma (READ) produce higher levels (Figs. 1A–B). TCGA databases were used to predict ALDOB mRNA levels in 539 samples of ccRCC and 72 samples of normal tissue. CcRCC samples expressed markedly lower levels of ALDOB mRNA than normal tissue (Fig. 1C, P < 0.001). It was also found that ALDOB expression is diminished in adjacent ccRCC samples relative to GTEx samples (Fig. 1D, P < 0.001). A notable decrease in ALDOB expression was noted in 72 ccRCC samples relative to adjacent matched samples (Fig. 1E, P < 0.001). A receiver operating characteristic (ROC) curve was executed to estimate the diagnostic value of ALDOB levels. The AUC for ALDOB levels was 0.836 (CI = 0.777−0.895), showing promising diagnostic potential (Fig. 1F). Further, normal tissues showed an elevation in ALDOB protein levels when compared to ccRCC tissues (Fig. 1G).

Additionally, the ALDOB gene levels were verified through examination of the GEO datasets (GSE66271 and GSE53757). The results were consistent with the data from TCGA, the ALDOB level was diminished in ccRCC tissues versus adjacent normal tissues (Figs. 2A, 2B). HPA data also showed that ccRCCs exhibited a decrease in ALDOB expression compared to normal tissues (Fig. 2C). Based on the above data, ALDOB appears to be an important diagnostic biomarker in ccRCC tissues.

Clinical characteristics associated with ALDOB expression

We assessed whether ALDOB was correlated with different clinical-pathological features, which include pathologic T stage (T1, T2, T3 and T4), pathologic N stage (N0 and N1), pathologic M stage (M0 and M1), histologic grade (grade 1, 2, 3, and 4), and pathologic stage (stages I, II, III, and IV) by applying TCGA dataset. Statistical evaluation through logistic regression confirmed the link between ALDOB and the clinical-pathological features of individuals with ccRCC utilizing the TCGA-KIRC cohort. The findings demonstrated that individuals with ccRCC who were upgraded in the T stage, N stage, M stage, pathologic stage, and histologic grade exhibited decreasing ALDOB levels. ALDOB was found to be overexpressed in the patients with T1 stage (Fig. 3A), no lymph node metastases (Fig. 3B) or no distant metastases (Fig. 3C), pathologic stage I (Fig. 3D), and histologic grade 1 (Fig. 3E), respectively. As a result of analyzing the expression of ALDOB in male and female ccRCC patients, it was determined that males had a relatively lower expression of ALDOB than females (Fig. 3F).

In addition, the Fisher exact test and the chi-square test yielded consistent results (Table 1). Subsequently, several clinical variables were markedly correlated with ALDOB expression according to the univariate logistic regression analysis, which include pathologic stage (odds ratio (OR) = 0.662 [0.464−0.658], P = 0.023), histologic grade (OR = 0.629 [0.447−0.886], P = 0.008), gender (OR = 0.645 (0.452−0.922), P = 0.016), and pathologic stage (OR = 0.668 [0.471−0.948], P = 0.024) (Table 2). Nonetheless, no marked variations were observed in pathologic N stage (OR = 0.368 (0.115−1.174), P = 0.091), and pathologic M stage [OR = 0.623 (0.383−1.015), P = 0.057] (Table 2). According to these findings, ALDOB might contribute to tumor development and progression in individuals with ccRCC.

Prognostic value of ALDOB expression in ccRCC

Subsequently, we generated an OS heatmap for the ALDOB gene. The analysis revealed a significant association between low ALDOB expression and overall survival (OS) in multiple tumor types, including COAD, KIRC, and LGG (Fig. 4A). Data from the TCGA database was utilized to examine the link between ALDOB level and prognostic outcomes (OS, DSS, PFI). Low ALDOB expression was linked to unfavorable OS, DSS, and PFI in the study population, with hazard ratios (HR) of 0.49 (95% CI [0.36–0.66]), 0.37 (95% CI [0.24–0.55]), and 0.47 (95% CI [0.34–0.66]), respectively, all with a p-value <0.001 (Figs. 4B–4D). A further investigation of ALDOB expression and subgroups was conducted in this study. Lower expression of ALDOB was detected in the T3-T4 stage (HR = 0.49 (0.33–0.73), P < 0.001), pathologic stage III–IV (HR = 0.51 (0.35–0.74), P < 0.001), histologic grade G3-G4 (HR = 0.48 (0.33–0.68), P < 0.001) (Fig. 4F). We developed a clinical prognostic risk score using the stages T, N, M, pathologic stage, histologic grade, and ALDOB expression in ccRCC (Fig. 4E). Utilizing Cox logistic regression analyses outcomes, these prognostic indicators were integrated to construct nomograms predicting 1-year, 3-year, and 5-year OS for ccRCC patients within TCGA-KIRC dataset (Fig. 4G). Furthermore, a calibration chart was employed to evaluate the model’s predictive accuracy (Fig. 4H). According to the findings, ALDOB expression levels could better predict patient survival at 3- and 5-year intervals.

Subsequently, the univariate and multivariate Cox regression analyses were executed to further ascertain the independent predictors of OS, DSS, and progression-free survival (PFS) in individuals with ccRCC. The forest plots indicated that ALDOB level exhibited marked links to OS (HR = 2.065, 95% CI [1.330–3.205], P = 0.001) (Fig. 5A), DSS (HR = 2.352, 95% CI [1.340–4.129], P = 0.003) (Fig. 5B), and PFS (HR =2.003, 95% CI [1.268–3.163], P = 0.003) (Fig. 5C) in ccRCC. Further analysis of OS outcomes among patients with different phenotypes between low and high ALDOB expression subgroups was conducted. Moreover, the analysis illustrated that ALDOB downregulation was correlated with worse OS outcomes in patients with T3 stage (p = 0.001), histologic grade 3 (p = 0.028), N0 stage (p = 0.001), M0 stage (p = 0.006), and M1 stage (p = 0.010) (Fig. 6). According to these findings, ALDOB expression levels are associated with prognosis in ccRCC.

GO/GSEA enrichment analysis related to ALDOB gene expression in ccRCC tissue

A gene expression profiling analysis was undertaken to understand ALDOB’s biological significance in ccRCC. The analysis identified 2,256 downregulated and 482 upregulated genes that were substantially connected with ALDOB level (Padj < 0.05 and logFC > 1) (Fig. 7A). Additionally, GO and KEGG pathway examinations of ALDOB and its associated genes were executed to explore ALDOB’s functional implications in ccRCC. The findings indicated that within ccRCC, ALDOB primarily functioned in the basal part of the cell (cellular component), serine-type peptidase activity (molecular function), and neuroactive ligand–receptor interaction (KEGG pathway) (Table 3).

We performed GSEA to find functional and biological pathways between low and high ALDOB levels. According to the normalized enrichment scores (NESs), the pathway with the highest enriched signaling was selected based on the expression of the ALDOB gene (Fig. 7B). GSEA indicated that the phenotype associated with low ALDOB expression was primarily enriched in KEGG_PPAR_SIGNALING_PATHWAY (Fig. 7C, NES = 2.599, P < 0.001). WikiPathways_PPAR_SIGNALING_PATHWAY (Fig. 7D, NES = 2.599, P < 0.001). And KEGG_FATTY_ACID_METABOLISM (Fig. 7E, NES = 2.625, P < 0.001).

Association of immune infiltration with ALDOB expression

Subsequently, 24 different immune cells were evaluated for correlation with ALDOB levels in ccRCC. The expression of ALDOB demonstrated significant correlations with neutrophils (r = 0.295), and Th17 cell (r = 0.238), while exhibiting adverse associations with Treg (r =  − 0.160), and macrophages (r =  − 0.133), with all p-values < 0.001 (Figs. 8A–8B). Subsequent analysis revealed notable variations in ALDOB expression levels across various infiltrating immune cell types, such as neutrophils, TReg, and TH17 cells (Figs. 8C–8E). These findings suggest a pivotal role for ALDOB in immune infiltration within ccRCC.

Expression level of ALDOB in tissue microarray samples

We further obtained 80 pairs of ccRCC clinical tissues from tissue microarray (cancerous tissues and paracancerous tissues). IHC showed that the ALDOB level in ccRCC samples was notably decreased relative to that in the adjacent normal tissues, which was confirmed using the corresponding IHC Score (p < 0.05, Figs. 9A–9B). Additionally, we also examined the difference of ALDOB expression in tissues of renal clear cell carcinoma with different pathological stages, and found that the ALDOB expression in tumor tissues of stage II/III was markedly lower than that of stage I/II (p < 0.001, Fig. 9C). ALDOB expression in stage T2 tumors was markedly lower than that in stage T1 tumors (p < 0.05, Fig. 9D). It was found that the expression of ALDOB in AJCC stage II tumors was markedly diminished relative to the AJCC stage I (p < 0.05, Fig. 9E).