Abnormal gene expression in the intestinal mucosa along with an imbalance in the intestinal microflora is one of the main causes of colorectal disease, and several mechanisms by which intestinal microbes and abnormal gene expression affect the development of colonic tumors have been suggested 36. In this study, we integrated the results of RNA-seq, metagenomics, and sociomedical pattern analysis of ACRC, HRA, and NC samples to determine significant differences. We identified the diversity of the microbiome, showed a correlation between gene expression and the microbiome, and performed network analysis. We separated the signatures between the three groups and visualized distinct patterns. Our results provide a basis for manipulating the microbiome in treatment strategies for colorectal diseases.
Dysbiosis due to environmental factors such as dietary pattern or genetic variations can disrupt the immune system and may promote colorectal neoplasm 37–39. The gut microbiota change can alter the efficacy of CRC treatment by increasing the sensitivity of chemotherapeutic agents, radiotherapy, and immune checkpoints inhibitors and reduced the toxicity of these treatment modalities 40. Recent scientific evidence suggests that colorectal microbiota modification can inhibit ACRC progression and improve the treatment effect in ACRC 41. A literature survey revealed that changing the colorectal microbiota composition by probiotics, prebiotics, and diet protects ACRC patients from treatment-associated adverse effects 18, 41–43. This study provides insights into the association between colorectal microbiota and colorectal diseases (including ACRC and HRA) to provide innovative strategies for enhancing the safety and efficacy of ACRC and HRA therapy.
Many studies have examined specific gut bacterial species associated with colorectal diseases. A typical example is sulfidogenic bacteria. Hydrogen sulfide-producing bacteria such as Fusobacterium, Desulfovibrio, and B. wadsworthia are known to be involved in ACRC development through the production of DNA-damaging hydrogen sulfide 44–46. In addition, patients with beneficial gut microbiota, such as B. longum, Ruminococcaceae spp., E. faecium, Faecalibacterium spp., and C. aerofaciens, have superior systemic and antitumor immunity compared to patients with low strain diversity and relatively high abundance of Bacteroidetes 42, 47. This phenomenon suggests that intestinal microbiota can modulate immune function in the intestine and increase tumor immunity.
Most existing studies on gut microbiota in CRC have analyzed gut microbiota in feces. Examination of the feces is non-invasive and may be appropriate as a screening test, but there may be variables that affect the metagenomic results, such as the collection process, dietary pattern, and antibiotic administration 10, 12. In this study, the microbiome genome was profiled in the colonic mucosa using samples removed during colonoscopy. We were able to identify species with high relative abundance in ACRC- and HRA-derived mucosa (three Bacteroides and two Clostridium species) and extracted genes that were highly correlated with these bacterial species.
Understanding the molecular mechanisms of CRC development and progression is key to early diagnosis and the development of personalized medicines. Several previous studies have clarified the importance of the interaction between host cells and the microbiome in the pathogenesis of CRC 24, 31, 48, 49. To understand the role of these interactions in the adenoma-carcinoma sequence, we correlated host gene expression and mucosal microbiome genomic composition data using microbiota and RNA-seq data in HRA, ACRC, and NC samples. In the correlation and network analyses of mucosal-derived microorganisms and gene expression, T. nexilis was a hub species related to DEGs in HRA and NC samples and in ACRC and HRA samples. In addition, C. spiroforme was identified as a hub species related to differential gene expression when comparing ACRC and NC samples. C. spiroforme had a strong positive relationship with NBPF13P and a strong negative correlation with RMRP.
REG3A was found to be elevated in ACRC samples compared to NC samples. High REG3A levels are correlated with larger tumor size, poorer tumor differentiation, higher tumor stage, and lower survival rate 50. REG3A has been shown to have pro-tumorigenic effects, including promotion of cell proliferation, inhibition of cell apoptosis, and regulation of cancer cell migration by activating AKT and ERK1/2 pathways in gastric cancer cells 51. REG3A has also been considered to play a key role in inflammation-linked pancreatic carcinogenesis 52, 53. Therefore, REG3A may serve as a promising therapeutic target in ACRC. We are the first to identify a relationship between ACRC and NBPF13P. We revealed a relationship between the microbiome and NBPF13P, which could provide a new pathway for targeting in colorectal diseases.
The colon has the highest load of gut microbiota, with over 1011 bacteria per milliliter. Colonic symbionts can be classified according to their anatomical distribution as 1) luminal-resident bacteria, 2) mucous-resident bacteria, 3) epithelial-resident bacteria, and 4) lymphoid tissue-resident symbionts. Intestinal epithelial cells (IECs) play an important role in innate immunity by forming a physical barrier against environmental stimuli, including gut genetic toxins, and maintaining a balance between commensal bacteria and host cells 54, 55. Although this barrier is sterile, invasive bacteria, including adherent-invasive Escherichia coli, segmented filamentous bacilli, Enterococcus faecalis, Bacteroides fragilis, and Clostridium spp., can reside in and attach to IECs 56, 57. This can lead to chronic inflammation of the mucous membrane, which is one of the critical pathogeneses of inflammatory bowel disease and CRC and correlates with disease severity. Since our study analyzed the microbiota of the colonic mucosa, it is difficult to exclude the possibility that a large portion of the microbiota present in IECs might be included in the metagenomic analysis. Tyzzerella and Clostridium, which are correlated with CRC progression and differential gene expression, are known to reside in IECs.
Clostridium spp.,a representative epithelium-resident bacteria, forms endospores and has strong dissemination power, survival, and resistance to antibiotics 58. Spore-forming bacteria have the following characteristics: resistance to antibiotic treatment, strong binding properties, high permeability, and harmful spores 59. The role of sporobiota in the pathogenesis and progression of CRC remains unclear and needs to be elucidated. Our results suggest that gut sporobiota may be important in the pathogenesis of CRC. Understanding the mechanism of CRC pathogenesis is useful not only for the development of targeted therapeutics, which could potentially define markers and guide precision medicine, but also for the early detection and prevention of CRC. Identification of the exact role of sporobiota in colorectal tumorigenesis will help us understand the current limitations of gut microbiota modulations, such as antibiotic administration, diet modification, and probiotic administration, for CRC prevention and treatment and can suggest new target therapies for CRC.
In this study, we also investigated the association between the composition of the intestinal microflora and dietary patterns and other environmental factors. No definitive difference was observed in the gut mucosal microbiota diversity between HRA, ACRC, and NC samples. This result is not in line with those of previous studies using fecal microbiota analysis. This suggests that environmental factors, including dietary patterns and socioeconomic, psychiatric, and clinical factors may have less influence on the gut mucosal microbiota diversity compared to that of feces. Future large-scale studies are needed to clarify this.
Our study had several limitations. First, whether mucosal microbiota analysis reflects the effect of microbiota on the development and progression of CRC may be controversial. Since our colonic tissue was obtained during colonoscopy, it is possible that the results of the metagenomic analysis may have been affected by the bowel preparation process. Second, because of the small number of samples, differences in the mucosal microbiome and gene expression according to clinical characteristics of ACRC and HRA, such as tumor stage, were not fully analyzed. Third, our results have not been reproduced and proven using molecular and laboratory techniques. These results require further identification and validation studies.