Gut Microbiota Alterations as Potential Biomarkers for Renal Cell Carcinoma: Insights from 16S rRNA Sequencing

2.1 Study design and ethical approval

This was a case–control study designed to investigate differences in gut microbiota composition between patients with RCC and healthy individuals. All procedures were conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Shanghai Huiyuan Hospital (Approval No. 2023–0528). Written informed consent was obtained from all participants prior to enrollment.

2.2 Study population

All patients in the RCC group were pathologically confirmed to have RCC following surgical resection. Healthy controls were recruited from Shanghai Huiyuan Hospital between July 2024 and March 2025. Demographic and clinical data, including age, sex, comorbidities, radiological findings, and postoperative histopathological results, were collected for all participants. All subjects were of Han Chinese ethnicity to minimize confounding effects related to diet and lifestyle.

Inclusion Criteria

Participants were eligible if they:(1) were aged ≥18 years;(2) had a confirmed postoperative histopathological diagnosis of RCC (case group) or no evidence of RCC or other diseases (control group);(3) had no infectious, neurological, or upper gastrointestinal diseases, renal dysfunction, or other active malignancies; and(4) had not used antibiotics, probiotics, vitamins, minerals, nonsteroidal anti-inflammatory drugs, hormones, or gastric acid–related medications (including bismuth, H₂ receptor antagonists, omeprazole, sulfasalazine, or misoprostol) within four weeks prior to enrollment. Participants with a history of gastric or duodenal ulcers or major gastrointestinal surgery were excluded.

Exclusion criteria included: Participants were excluded if they(1) had gastrointestinal discomfort (e.g., acid reflux, nausea), diabetes, or severe systemic disease;(2) had HIV infection or hepatitis B surface antigen positivity;(3) had heart failure, unstable angina, or recent myocardial infarction (within 6 months);(4) had a history of organ transplantation, radiotherapy, or chemotherapy; (5) were lactation; (6) had a chronic diarrhea or constipation; (7) had a inflammatory bowel disease;(8) had prior gastrointestinal surgery; (9) had gastrointestinal endoscopy within the past year; and(10) had irritable bowel syndrome.

2.3 Sample size determination

Sample size was estimated using the formula n = [(Uα + U2β)s/δ], where α and β represent Type I and Type II error probabilities, respectively, and Uα, U2α, and U2β are the corresponding t-values. Based on preliminary pilot data, a minimum of 41 participants per group was required. Ultimately, 47 fecal samples from RCC patients and 51 from healthy controls were included. Baseline characteristics of both groups are summarized in.

2.4 Collection and processing of fecal samples

Fresh fecal samples (approximately 500 mg) were collected from each subject using sterile collection tubes. Participants were instructed to avoid contamination from urine or toilet water. Samples were immediately frozen on dry ice, transported to the laboratory within 2 hours, and stored at −80 °C until DNA extraction. All samples were processed under aseptic conditions to minimize environmental contamination.

2.5 DNA extraction and quantification

Microbial genomic DNA was extracted using the DNeasy PowerSoil Pro Kit (Qiagen, USA) following the manufacturer’s protocol. DNA concentration and purity were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher, USA), and integrity was verified by 1% agarose gel electrophoresis. High-quality DNA samples were stored at −20 °C for downstream analysis.

2.6 16S rRNA gene amplification and library preparation

The V3–V4 hypervariable regions of the bacterial 16S rRNA gene were amplified using specific primers (341F: 5′-CCTACGGGNGGCTGCAG-3′; 805R: 5′-GACTACHVGGGTATCTAATCC-3′). PCR was performed in a 25 μL reaction mixture containing 25 ng of template DNA, 5× Phusion Buffer, 2 mM dNTPs, 0.5 U Phusion DNA polymerase, and 1 μM of each primer. The cycling conditions were: initial denaturation at 95 °C for 3 min; 27 cycles of 95 °C for 30 s, 55 °C for 30 s, 72 °C for 45 s; and a final extension at 72 °C for 5 min. PCR products were purified using magnetic beads, and sequencing libraries were prepared with dual-index adapters according to Illumina protocols. Library quality and fragment size were verified using a Qubit 4.0 fluorometer (Thermo Fisher) and agarose gel electrophoresis.

2.7 High-throughput sequencing

Sequencing was performed on the Illumina MiSeq PE300 platform (Illumina, USA). Raw paired-end reads were subjected to quality control using fastp to trim low-quality bases and remove ambiguous reads. High-quality paired reads were merged using FLASH with a minimum overlap of 10 bp. Chimeric sequences were detected and removed using UCHIME, and valid reads were clustered into operational taxonomic units (OTUs) at 97% similarity using Vsearch. Representative sequences from each OTU were taxonomically classified against the SILVA 138 database using the RDP classifier with a confidence threshold of 0.7.

2.8 Bioinformatics and statistical analysis

Alpha-diversity indices (Chao1, Shannon, Simpson, and Goods coverage) were calculated to assess microbial richness and evenness. Beta-diversity was analyzed using Principal Coordinates Analysis (PCoA) and Non-metric Multidimensional Scaling (NMDS) based on Bray–Curtis distances. Differences in community composition between groups were assessed using PERMANOVA. Differential taxa were identified using Linear Discriminant Analysis Effect Size (LEfSe), and key microbial markers were screened using random-forest modeling. Statistical analyses were conducted in R (v4.3.1), and a p-value <0.05 was considered statistically significant.

3.1 Participant characteristics

A total of 98 participants were enrolled, including 47 patients with RCC and 51 healthy controls. No significant differences were observed between the two groups in terms of age, sex, body mass index (BMI), or renal function parameters (p > 0.05), indicating comparability of baseline characteristics. Detailed demographic and clinical data are summarized in Table 1.

3.2 Differences in α diversity of gut microbiota between the RCC group and healthy control group

To assess the within-sample diversity (α-diversity) of the gut microbiota, five indices were calculated: the Chao1, Simpson, Shannon, Goods_coverage, and Observed_species indices ( Figure 1A–E). Overall, RCC patients exhibited significantly reduced microbial diversity and richness compared with healthy controls. The Chao1 index, reflecting species richness, was markedly lower in the RCC group than in the control group (***P < 0.001), indicating a substantial reduction in the number of bacterial taxa. Similarly, the Shannon index, which accounts for both richness and evenness, was significantly decreased in RCC patients (**P < 0.05), suggesting reduced diversity and compositional complexity. In contrast, the Simpson index, which inversely reflects diversity, was significantly higher in RCC patients (*P < 0.01), consistent with decreased microbial evenness and dominance of a limited number of taxa. Although the Observed_species index showed a decreasing trend in the RCC group, the difference was not statistically significant (P > 0.05). The Goods_coverage index was slightly higher in the RCC group (***P < 0.001), confirming sufficient sequencing depth and coverage for all samples.

Figure 1. α-diversity analysis of gut microbiota in RCC patients and healthy controls.

** > Violin plots showing differences in α-diversity indices between the RCC (blue) and control (red) groups: (A) Chao1, (B) Simpson, (C) Shannon, (D) Observed species, and (E) Goods coverage indices. Statistical significance was determined using the Wilcoxon rank-sum test. ns, not significant; P < 0.05 (*), P < 0.01 (**), P < 0.001 (***). The results indicate a significant reduction in microbial richness and diversity in RCC patients, suggesting gut microbial dysbiosis associated with disease development.

Collectively, these findings demonstrate that the gut microbiota of RCC patients displays significantly lower species richness and diversity compared with healthy individuals, indicating a state of microbial dysbiosis potentially associated with RCC development.

3.3 Differences in β diversity between the RCC and healthy control groups

To investigate differences in overall gut microbial community structure between patients with RCC and healthy controls, β-diversity was assessed using Principal Coordinates Analysis (PCoA) and Non-metric Multidimensional Scaling (NMDS) based on Bray–Curtis dissimilarity ( Figure 2A, 2B). Both analyses provide a visual representation of compositional variation among samples, where closer proximity of data points indicates greater microbial similarity.

Figure 2. β-diversity analysis of gut microbiota in RCC patients and healthy controls.

(A) Principal Coordinate Analysis (PCoA) based on Bray–Curtis distance shows distinct clustering between RCC (blue) and control (red) groups, indicating significant differences in overall microbial community composition. (B) Non-metric Multidimensional Scaling (NMDS) analysis further demonstrates the separation between the two groups (Stress = 0.106), confirming alterations in β-diversity. Each point represents one sample; ellipses denote 95% confidence intervals for each group. The results indicate that RCC patients harbor distinct gut microbial community structures compared with healthy individuals.

The PCoA plot revealed distinct clustering of the RCC and control samples along the principal coordinates, suggesting a clear separation in microbial community composition between the two groups. Similarly, NMDS analysis produced consistent results, with samples forming two independent clusters corresponding to RCC and healthy individuals (Stress = 0.106), indicating an acceptable model fit (Stress <0.2).

In summary，these β-diversity analyses demonstrate that the gut microbial community structure of RCC patients differs significantly from that of healthy controls, reflecting a marked alteration in microbial composition and confirming the presence of disease-associated dysbiosis.

3.4 Abundance and Structural variation of gut microbiota at different taxonomic levels

To further characterize differences in gut microbial composition between RCC patients and healthy controls, we compared the relative abundances of bacterial taxa at both the phylum and genus levels.

At the phylum level, the four dominant phyla in both groups were Bacteroidetes, Firmicutes, Proteobacteria, and Actinobacteria ( Figure 3A). However, the relative abundance of several less dominant phyla differed significantly between groups ( Figure 3B). Specifically, the abundances of Spirochaetes, Fibrobacteres, Zixibacteria, and WPS-2 were significantly higher in the RCC group (P < 0.05), whereas Verrucomicrobia, Firmicutes, Chloroflexi, and Elusimicrobia were significantly reduced compared with the control group (P < 0.05).

Figure 3. Gut microbial community composition and differential abundance between RCC patients and healthy controls.

(A) Stacked bar chart showing relative abundances of bacterial taxa at the phylum level in RCC (blue) and control (red) groups. (B) Box plots comparing major phyla between groups, showing significant differences in Firmicutes, Bacteroidetes, Verrucomicrobia, and Chloroflexi. (C) Relative abundance of major bacterial genera across the two groups, indicating distinct compositional profiles. (D) Differential analysis of predominant genera reveals decreased Faecalibacterium, Roseburia, and Blautia, and increased Streptococcus, Escherichia–Shigella, and Parabacteroides in RCC patients. Statistical significance was assessed using the Wilcoxon rank-sum test. P < 0.05 (*), P < 0.01 (**), P < 0.001 (***). These findings indicate significant alterations in gut microbial composition and the loss of beneficial commensal bacteria in RCC patients, consistent with disease-associated microbial dysbiosis.

At the genus level, both groups shared the same top four dominant genera—Bacteroides, Faecalibacterium, Prevotella_9, and Escherichia–Shigella ( Figure 3C). Among the top ten most abundant genera, Streptococcus and Parasutterella were significantly enriched in RCC patients, while beneficial genera such as Faecalibacterium, Roseburia, Blautia, and others were markedly decreased (P < 0.05; Figure 3D).

Overall，these findings demonstrate pronounced alterations in gut microbial composition in RCC patients, characterized by a loss of short-chain fatty acid–producing commensals and an increase in potential pathogenic or pro-inflammatory taxa. Such shifts in microbial structure are indicative of gut dysbiosis associated with renal cell carcinoma.

3.5 LEfSe analysis of differentially abundant taxa

To identify bacterial taxa that significantly contributed to the compositional differences between groups, Linear Discriminant Analysis Effect Size (LEfSe) analysis was performed ( Figure 4A, 4B). This method detects taxa with statistically significant differences in relative abundance across groups and estimates their effect size using Linear Discriminant Analysis (LDA).

Figure 4. Differential gut microbial taxa between RCC patients and healthy controls identified by LEfSe analysis.

(A) LEfSe-generated cladogram showing taxa with significantly different abundances between the RCC (green) and control (red) groups. Each circle represents a taxonomic level from phylum (outer ring) to genus (inner ring); the diameter of each circle is proportional to its relative abundance. (B) Histogram of linear discriminant analysis (LDA) scores (log10) for taxa with significant differences (LDA > 2.0, P < 0.05). Green bars indicate taxa enriched in RCC, while red bars indicate taxa enriched in healthy controls. The RCC group exhibited enrichment of Streptococcus, Streptococcaceae, and Christensenellaceae, whereas the control group showed higher abundances of Faecalibacterium, Roseburia, Lachnospiraceae, and Clostridia. These results highlight distinct microbial biomarkers that discriminate RCC patients from healthy individuals and suggest a loss of beneficial commensal bacteria associated with renal cell carcinoma–related gut dysbiosis.

The cladogram generated by LEfSe revealed distinct taxonomic distributions between the RCC and control groups ( Figure 4A). At various taxonomic levels, Streptococcus, Streptococcaceae, Christensenellaceae, and Christensenellaceae_R-7_group were significantly enriched in the RCC group. In contrast, the control group exhibited higher relative abundances of Subdoligranulum, Escherichia–Shigella, Faecalibacterium, Roseburia, Lachnospiraceae, Firmicutes, Clostridiales, and Clostridia (LDA score > 2.0, P < 0.05).

The LDA bar chart further quantified these differences, demonstrating that taxa enriched in healthy controls predominantly belonged to the phylum Firmicutes, particularly the families Lachnospiraceae and Clostridiaceae, which are known producers of short-chain fatty acids (SCFAs) involved in maintaining intestinal and systemic immune homeostasis. Conversely, enrichment of Streptococcus and Christensenellaceae in RCC patients may indicate a shift toward pro-inflammatory or metabolically altered microbial populations.

Cumulatively, these results highlight distinct bacterial biomarkers that differentiate RCC patients from healthy individuals, characterized by a loss of beneficial commensal SCFA-producing bacteria and an increase in opportunistic or inflammation-associated taxa, supporting the presence of RCC-related gut dysbiosis.

3.6 Correlation and model prediction analysis

To explore interrelationships among dominant bacterial genera and identify potential microbial biomarkers for RCC, we performed correlation analysis, indicator species analysis, and random forest modeling using the top 30 most abundant genera.

The Spearman correlation heatmap revealed strong positive correlations among several short-chain fatty acid (SCFA)-producing genera, including Blautia, Subdoligranulum, Anaerostipes, Roseburia, Faecalibacterium, and Fusicatenibacter ( Figure 5A). Similarly, a strong co-occurrence pattern was observed among Barnesiella, Ruminococcaceae_UCG − 014, [Eubacterium]_coprostanoligenes_group, Alistipes, and Ruminococcaceae_UCG − 002, whereas Bifidobacterium, Haemophilus, and Streptococcus also showed positive associations. These findings indicate that beneficial commensals tend to coexist and may be jointly depleted in RCC-associated dysbiosis, while potentially pathogenic taxa cluster together.

Figure 5. Correlation network and key bacterial biomarkers associated with RCC.

(A) Spearman correlation heatmap of the top 30 bacterial genera. Red indicates positive correlations, and blue indicates negative correlations. SCFA-producing genera (Faecalibacterium, Roseburia, Blautia) were positively correlated with each other but negatively correlated with potential pathogens such as Streptococcus and Escherichia–Shigella. (B) Comparative bar and dot plots showing differential abundance of major genera between RCC (blue) and control (red) groups. (C) Random forest analysis identifying the top discriminatory bacterial genera ranked by Mean Decrease Gini value, with Roseburia, Sphingomonas, and Escherichia–Shigella emerging as key taxa differentiating RCC from healthy controls. Collectively, these findings reveal distinct co-occurrence patterns and identify specific bacterial taxa that may serve as noninvasive biomarkers for RCC-associated gut dysbiosis.

Indicator species analysis identified Faecalibacterium and Roseburia (both belonging to Firmicutes) as genera significantly enriched in healthy controls, suggesting their role as potential protective taxa. In contrast, Bacteroides (phylum Bacteroidetes) was found at higher relative abundance in the RCC group and could serve as a disease-associated indicator genus ( Figure 5B).

Random forest analysis further identified the top five genera contributing most to group discrimination, with Roseburia, Sphingomonas, Escherichia–Shigella, Christensenellaceae_R-7_group, and Parasutterella showing the highest Mean Decrease Gini values (> 2.0) ( Figure 5C). These taxa exhibited strong discriminatory power between RCC patients and healthy controls, suggesting their potential as noninvasive microbial biomarkers for RCC detection.

Taken together, these analyses reveal distinct microbial co-occurrence patterns and highlight specific bacterial genera that may serve as diagnostic or prognostic indicators of RCC-associated gut microbiota dysbiosis.