Comparative Analysis of Microbiological Profiles and Antibiotic Resistance Genes in Subjects with Colorectal Cancer and Healthy Individuals

The gut microbiota (GM) comprises various populations of bacteria and is a metabolically active ecosystem that plays vital functions in maintaining human health (Sánchez-Alcoholado et al. 2020). Alterations to the GM composition produce changing levels of related metabolites, including short-chain fatty acids, polyphenols, vitamins, and polyamines. Over the decades, variations in the GM composition have been found to be related to diverse human pathologies, such as obesity (Davis 2016; Al-Assal et al. 2018), autoimmune diseases (Xu et al. 2019; Zhang et al. 2020b) and several types of cancers (Vivarelli et al. 2019; Cheng et al. 2020). Colorectal cancer (CRC) is considered to be the fourth cause of human deaths resulting from cancer globally (Brennan and Garrett 2016), accounting for about 10% of all new cancer cases (Wong and Yu 2019). Emerging evidence has demonstrated that GM dysbiosis may be involved in the development of CRC by producing metabolites that have important actions on the host-immune system that trigger the release of genotoxic compounds (Brennan and Garrett 2016; Sánchez-Sobhani et al. 2019; Wong and Yu 2019; Alcoholado et al. 2020). Microbial toxins induce DNA damage, promoting mutations. In addition, microbial activation of pathways such as NF-κB sustains chronic inflammation, fostering tumor initiation and progression.

Gut bacteria, including Fusobacterium nucleatum, form biofilms on tumor surfaces, enhancing invasiveness by disrupting epithelial barriers and facilitating microbial and cancer cell translocation. Microbes modulate immune responses in the tumor microenvironment (TME). F. nucleatum, for instance, inhibits antitumor immune activity by suppressing T-cell responses, enabling tumor survival and spread (Sakamoto et al. 2021). Microbial components and metabolites can prepare distant tissues for metastasis by creating an immunosuppressive and pro-inflammatory environment (Liu et al. 2023). Metagenomic studies in humans have identified several bacterial species that are associated with CRC progression (Brennan and Garrett 2016; Wong and Yu 2019).

Viruses and fungi are emerging contributors to CRC through mechanisms involving immune modulation, chronic inflammation, and microbial ecosystem alterations, although less extensively studied in CRC. These microorganisms influence tumor initiation, TME, and possibly therapy responses. In addition, studies have examined alterations in urinary microbiota in CRC patients and found significant shifts that may relate to cancer presence and progression. For example, one study showed a distinct gut microbiome in patients with CRC’s urine extracellular vesicles (EVs) compared to healthy controls. The study demonstrated that microbial signatures from EVs in urine could be potential biomarkers for diagnosing CRC (Yoon et al. 2023).

Obtaining the pure culture of microbiota is very efficient for understanding the intestinal microbial metabolites and their effects on host physiology (Dorrestein et al. 2014; Rowland et al. 2018). Actinomy-cetes play complex and diverse roles in tumors, including CRC. Due to their unique metabolic capabilities, these roles span from potential tumor-promoting effects to therapeutic applications. Actinomycetes can contribute to dysbiosis (imbalanced GM), a recognized factor in CRC development (Xu et al. 2022). Some Actinomycetes species are implicated in chronic inflammatory responses, creating a pro-carcinogenic environment (Sharma et al. 2023). Actinomycetes produce various metabolites that can affect host cells. While many are therapeutic, some may inadvertently promote cancer through genotoxic effects or by modifying the TME.

In contrast, Actinomycetes, especially Streptomyces, are prolific producers of bioactive secondary metabolites with anticancer properties. Some Actinomycetes species exert protective effects in the gut by modulating the microbiome (De Simeis and Serra 2021), improving epithelial barrier integrity, and reducing inflammation, potentially lowering the risk of CRC. In addition, Actinomycetes-derived compounds, such as immunomodulators, can stimulate anti-tumor immune responses, enhancing cancer therapy outcomes (Olano et al. 2009). However, the methods of isolating and pure culture of Actinomycetes from fecal samples have not been well established. Furthermore, the long developmental cycles, guanine-cytosine-rich genome, and abundant repeat regions impede routine molecular and genome sequencing technique analysis (Prudence et al. 2020). Traditional microbiology and genome sequencing use cultivated clonal cultures of bacteria. However, these approaches ignore the overwhelming microbial biodiversity, and clearly, metagenomics provides important insights into the rich and diverse microbial world (Quince et al. 2017; Ghosh et al. 2019). Addressing microbial biodiversity using traditional clonal cultures presents significant challenges, such as cultivation bias, loss of ecological context, low throughput and underrepresentation of rare species, particularly when investigating complex environments like the GM in CRC. These limitations have been addressed through advances in metagenomics, which provide a more comprehensive understanding of microbial communities and their impact on CRC.

Profiling ARG is crucial for understanding microbial dynamics, treatment challenges and disease progression, particularly in contexts such as CRC. Resistance profiling identifies reservoirs of resistance genes within the GM. The presence of ARG indicates the selective pressures exerted by the host environment, including inflammation, dietary factors, and therapeutic interventions. Antibiotic resistance genes linked to the production of genotoxic agents are key to understanding microbial roles in DNA damage and CRC initiation. Antibiotic resistance genes may influence the immune landscape by promoting microbial evasion or suppressing anti-tumor immune responses.

In this pilot study, we hypothesized that GM of CRC patients exhibits distinct differences in microbial relative abundance, diversity, metabolic functions, and antibiotic resistance genes compared to healthy controls. These differences can distinguish CRC patients from healthy individuals and may offer insights into the role of GM in CRC progression, treatment resistance, and overall disease pathogenesis. We aimed to compare the microbial relative abundance and diversity, analyze the metabolic functions, identify and profile antibiotic resistance genes, and apply multivariate analysis approaches to the gut microbiomes of CRC patients and healthy controls.

The present study compared the GM composition, metabolic functions, and ARGs between CRC patients and healthy controls using a next-generation high-throughput sequencing technique (HiSeq X™10). Our study demonstrated that two bacterial genera and seven species significantly differed between healthy individuals and CRC patients. Previous research has shown that Bacteroides were found to be enhanced in GM of colon cancer patients (Sobhani et al. 2011) and Alistipes spp. was associated with colon cancer (Moschen et al. 2016). Both taxa were enhanced in GM of CRC patients in the present study. Additionally, two metabolic pathways – benzoate and xylene degradation – were notably reduced in CRC patients. Furthermore, an increase in tetracycline resistance-encoding genes was observed in the CRC cohort.

Most importantly, our multivariate analysis (PCA-LDA) exhibited a promising classification accuracy in distinguishing healthy individuals from CRC patients. The recent advances in metagenomics provide valuable insights into gut microbiome alterations in CRC. With the high-dimensional data obtained from metagenomic analysis, a classification and prediction model could be developed to facilitate early CRC diagnosis. Current research on GM in CRC focuses heavily on the link between an imbalanced GM (dysbiosis) and the development, progression, and treatment response of CRC patients. Our study demonstrated that multivariate analysis can successfully distinguish CRC patients from healthy controls with a satisfactory accuracy (84.52% for metabolic function and 77.38% for resistant genes). This opens potential avenues for using GM as a biomarker for CRC diagnosis and as a target for therapeutic interventions such as dietary modifications and probiotic supplementation.

ANOSIM results revealed that both metabolic functions and antibiotic resistance genes were similar within or between CRC patients and healthy controls (p> 0.05). However, multivariate analysis (PCA-LDA) accurately classified these two groups. PCA plotting was used to analyze GM diversity in the stool and colon samples and revealed a clear difference by the taxonomic classification method (Mas-Lloret et al. 2020). When metagenome analysis was used to study bacterial infection of patients with acute surgical abdomen, PCA can roughly separate the bacteria of ascites into three groups, but failed to separate the variation in the sample collected from blood (Wu et al. 2018). PCA aims to understand the variance in the first components and reduce data complexity by evaluating combined related variables. In the present relatively small cohort of patients, LDA (Dinsdale et al. 2013) misclassified one or two patients into the wrong groups, decreasing the accuracy. We are confident the analysis will be improved by studying a larger sample of patients. More importantly, our pilot study shows the feasibility of multivariate analysis on metagenomic data as a potential biomarker for diagnosing CRC.

In a previous study, Actinomycetes relative abundance was increased after CRC patients had received chemotherapy, indicating that Actinomycetes may be intimately involved in treating CRC (Li et al. 2020). Hence, we compared the relative abundance of Actinobacteria at the family, genus, phylum, class, order, and species levels. At both the phylum and class level, the relative abundance of Actinomycetes was apparently decreased in CRC patients compared to the healthy control group, but statistical significance was not reached. At the order (Fig. 5A), families (Fig. 5B), and genus (Fig. 5C) level, the relative abundance was not found to be significantly different between the two groups. At the species level, 19 species of Actinomycetes were found in the two groups. Although no significant difference was found between CRC patients and healthy controls, the species of Bifidobacterium bifidum and Bifidobacterium dentium in CRC patients exhibited a higher relative abundance than in the healthy control group. In contrast, the species of Bifidobacterium breve were less abundant in the CRC patient group compared to the healthy control group (p = 0.053), indicating that bifidobacteria might be an indicator of CRC (Yoon et al. 2021). However, a larger sample size is necessary to verify this assumption further. In addition, all of our enrolled patients were required to be a first-time diagnosis and were surgically unresectable, so these cases had no previous chemotherapy regimens or chemotherapy treatments.

Fig. 5.

Comparison of Actinobacteria at the A) order, B) family, C) genus and D) species level in CRC patients and the healthy control group. CRC – colorectal cancer

A key limitation of the present study was the relatively small sample size, which restricts the generalizability of the findings to a broader population and potentially impacts the robustness of the conclusions drawn. Due to the limited number of participants, the results may not accurately reflect the broader phenomenon being investigated, potentially leading to skewed data and a higher risk of random variability that could obscure true effects. Therefore, further research with a larger sample size will be necessary to confirm the observed trends and establish more substantial evidence for the investigated relationships of GM in CRC patients. Our study was also limited to functional and taxonomic mapping but without the inclusion of metagenome-assembled genomes.

In conclusion, the results have provided valuable insights into the GM of CRC patients by analyzing the relative abundance, diversity, metabolic functions and resistance-encoding genes of GMs using metagenomic sequencing followed by in silico analyses. The research identified significant differences in the distribution of two bacterial genera and seven species between CRC patients and healthy controls. Furthermore, the study demonstrated that multivariate analyses of metabolic functions and resistance-encoding gene data can effectively classify subjects into CRC and healthy groups.

These findings underscore the potential of the GM as a diagnostic tool for CRC, offering a promising avenue for non-invasive screening and risk assessment. The identification of specific microbial signatures, particularly those linked to metabolic functions and resistance traits, could open new doors for understanding the microbiome’s role in CRC progression and treatment resistance.

A more in-depth exploration of the functional mechanisms underlying these microbial shifts will be essential for future research. Longitudinal studies investigating the causal relationships between microbial composition, CRC development, and treatment outcomes should further clarify the role of GM in cancer. Additionally, integrating host-microbiome interactions, including immune modulation and metabolic pathway alterations, could provide a more holistic understanding of CRC pathogenesis, to guide future therapeutic strategies.