This article is mentioned in:

Introduction

Colorectal cancer (CRC) is one of the most common diseases globally, killing ~800,000 people each year (1). CRC has a complicated and varied etiology that is associated with risk factors, including environment, diet, living habits and genetic factors (2,3). In total, ≤90% of cancer cases are associated with lifestyle and the link between nutrition and CRC has received more attention (2).

The incidence of CRC is highest in economically developed countries and there is an increasing yearly trend in emerging countries, owing to increased consumption of high-calorie diets (4). Caloric restriction (CR) has been shown to suppress cell proliferation, increase apoptosis and lower the host inflammatory response; however, the underlying mechanism is uncertain (2,4).

The equilibrium of gut bacteria serves a key role in host physiological activities and dysbiosis of gut microbes can lead to conditions such as inflammatory bowel disease (IBD), irritable bowel syndrome, obesity, type 2 diabetes (5). The role of the microbiome in the development and progression of CRC has recently received increased attention; however, how the microbiota determines cancer susceptibility and progression remains unknown (6,7). Previous studies have investigated the connections between gut microbiota and CRC, in addition to the involvement of the microbiome in the development of CRC (8,9). Gut microbiota influence CRC susceptibility and advancement by influencing mechanisms such as inflammation and DNA damage, as well as excreting chemicals that promote or inhibit tumor formation (10).

The present study aimed to investigate the effects of CR on development of CRC and gut microbial diversity using a xenograft mouse model to determine the mechanisms by which CR suppresses tumorigenesis and the role of gut microbial dysbiosis, thus, providing potential approaches for the prevention and treatment of CRC.

Materials and methods

Animals

A total of 10 specific-pathogen-free (SPF) grade BALB/c male nude mice (age, 4 weeks; body weight, 18-20 g) were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd. Mice were kept in an SPF environment (temperature 22±1˚C; humidity 40-60%) with a 12/12-h cycle of light and darkness with free access to food and drink. All animal procedures were authorized by the Ningxia Medical University Ethics Committee (approval no. 2021-045).

Reagents

RPMI-1640 medium (cat. no. AG29714278) was obtained from Hyclone (Cytiva). FBS (cat. no. 11011-8611) was purchased from Zhejiang Tianhang Biotechnology Co., Ltd. Penicillin-streptomycin (cat. no. ST488) and trypsin cell digestion solution (0.05% trypsin; cat. no. C0202) were purchased from Beyotime Institute of Biotechnology. Anti-Bax (cat. no. ab32503) and anti-Bcl2 (cat. no. ab32124) antibodies were purchased from Abcam. Anti-Ki67 antibody (cat. no. GB111499) was purchased from Wuhan Servicebio Technology Co., Ltd. Rabbit two-step detection (cat. no. PV-9001) and DAB color development kit (cat. no. ZLI-9018) were purchased from Beijing Zhongshan Golden Bridge Biotechnology Co., Ltd. BSA (cat. no. A8020) was purchased from Beijing Solarbio Science & Technology Co., Ltd.

Cell line and culture

The HCT116 human colon cancer cell line was obtained from Wuhan Procell Life Science & Technology Co., Ltd. (cat. no. CL-0096). The cells were grown in RPMI-1640 complete media (100 µg/ml streptomycin, 100 U/ml penicillin and 10% FBS) at 37˚C in a constant temperature incubator with 5% CO2. The cells were trypsinized at 37˚C about 1 min after reaching 80-90% confluence and then passaged or used in the following experiments.

Xenograft mouse model and diet treatment

After one week of adaptive feeding, the 10 nude mice were subcutaneously implanted with 2x106/100 µl HCT116 cells on the right flank. Every other day, tumor size was measured by a vernier caliper and the tumor volume was calculated using the formula a2xbx0.5, where a is the shortest diameter and b is the diameter perpendicular to a (11). When the mean tumor volume of each nude mouse reached ~100 mm3 (12 days after subcutaneous tumor transplantation of mice), the mice were divided into the control group in nutrient-rich condition and the CR group in a nutrient-poor condition where subjects were fed with 70% of the usual food intake, with 5 animals in each group. The usual food intake for each group was calculated after monitoring for three consecutive days, using the random number table method (random.org) (12). Tumor size and health indicators, including body weight, feeding habits and locomotor activity were tracked every other day. The largest diameter of tumor size did not exceed 20 mm. Regardless of the size of the tumor, the mice were euthanized if they fulfilled the following prerequisites: i) Tumor position severely impaired usual body function; ii) tumor-associated pain or distress; iii) loss of >20% normal body weight; iv) ulceration or infection at tumor growth sites; and v) persistent self-mutilating behavior. None of the experimental animals reached the humane endpoints and were not euthanized before the end of the experiment. After 3 weeks, the mice were euthanized by inhalation of CO2 gas (20% of the euthanasia chamber volume/min as a controlled flow rate of CO2, which was increased to 100% of the euthanasia chamber volume/min once the mice were unconscious). Death was confirmed by cardiac and respiratory arrest, limb stiffness or dilated pupils. After mice were euthanized, the tumor masses were immediately dissected, weighed and fixed or preserved at -80˚C.

Immunohistochemistry (IHC) staining

After being fixed in 4% paraformaldehyde solution at 20-23˚C for 24 h and embedded in paraffin wax, tissue blocks were cut into 4 µm sections. These were deparaffinized in xylene and dehydrated in a series of ethanol concentrations (70, 80, 95 and 100% for 3 min each) at 20-23˚C. This was followed by citrate buffer antigen retrieval (after the solution was boiled, the microwave power was adjusted to 800 watts and continued heating for 2 min) and 5% BSA blocking at 20-23˚C for 30 min. Tissue sections were incubated with primary antibodies against Bax (1:250), Bcl-2 (1:250) and Ki67 (1:500) at 4˚C overnight. The following day, the tissue sections were returned to room temperature for 1 h followed by a 20 min incubation with secondary antibodies which were part of the Rabbit two-step detection kit at 37˚C. The tissue sections were stained using a DAB kit. The samples were dehydrated, made translucent, stained (20-23˚C) with hematoxylin (0.025%; 30 seconds) and sealed prior to light microscope (magnification, x40; Olympus Corporation; cat. no. BX53) observation for analysis. The staining standard was scored according to the intensity of cell staining as follows: i) No positive staining (negative, 0); ii) light yellow (weakly positive, 1); iii) brownish yellow (positive, 2) and iv) tan (strong positive, 3). The percentage of positive cells was divided into 4 grades: i) 1, ≤25%; ii) 2, 26-50%; iii) 3, 51-75% and iv) 4, >75%. The final score was obtained by multiplying the two scores (13,14).

16S ribosomal (r)RNA sequencing and data analysis

Intestinal contents of the mice (solid excreted feces or collected from the rectum and a small portion of semi-solid stool with relatively abundant water content that was at the end of the colon) were collected after mice were sacrificed. These samples were sent to Shanghai Personalbio Biotechnology Co., Ltd. for 16s rRNA sequencing and microbial community diversity composition analysis. DADA2 software (QIIME2 (version 2019.4)) was used for sequence denoising and the Vsearch software (version 2.13.4_linux_x86_64) was used for cluster analysis (15,16). After quality control which included the steps of primer removal, quality filtering, denoise, stitching, and removal of chimerism, the data were evaluated for bacterial species composition, α diversity (including Chao1, Observed, Shannon and Simpson indices) and β diversity (including the principal component analysis (PCA), principal coordinate analysis (PCoA) and non-metric multidimensional scaling analysis (NMDS)) and species differences using the QIIME2 (version 2019.4) gene cloud platform (https://www.genescloud.cn). Linear discriminant analysis effect size (LEfSe) analysis was performed using gene cloud platform (genescloud.cn). Receiver operating characteristic (ROC) curve analysis was performed to with the XianTao tool (https://www.xiantao.love/products). Microbial functions were predicted by Phylogenetic investigation of communities by reconstruction of unobserved states (PICRUSt2; version 2.5.1; github.com/picrust/picrust) upon Kyoto Encyclopedia of Genes and Genomes (KEGG; kegg.jp/) databases and Clusters of Orthologous Genes (COG; ncbi.nlm.nih.gov/COG/) databases.

Statistical analysis

Unpaired Student's t test was used to compare differences in tumor volume and mass, mouse body weight, staining scores and species composition using Prism v9.0 software (GraphPad Software, Inc.). Sequence data analysis was performed using QIIME2 (version 2019.4) and R packages (version 3.2.0) (17). Kruskal-Wallis rank-sum and Dunn's post hoc test were used to examine differences between different sample groups. P<0.05 was considered to indicate a statistically significant difference. At least three independent biological replicates, and unless otherwise noted, all associated data are presented as mean ± SD.

Results

CR suppresses in vivo proliferation of CRC cells by regulating apoptosis and proliferation

In the present investigation, 10 mice were employed to examine the impact of CR on the in vivo development of CRC cells. The mean maximum diameter of tumor in CR and control group were 15.63±0.840 and 16.68±1.360 mm, respectively. The tumor volume in the CR group was significantly lower compared with the control group at 32 days (Fig. 1A and D). Similarly, the tumor mass was significantly reduced in mice of CR group compared with the mice of control group (Fig. 1C). Notably, CR did not elicit a significant effect on the body weight of the mice (Fig. 1B).

Figure 1 CR suppresses in vivo growth of CRC tumor. Changes in (A) tumor volume, (B) body mass, (C) tumor weight and (D) resected tumor samples from CRC solid tumor model mice. Data are presented as mean ± SD, n=5. **P<0.01, ****P<0.0001 vs. control. CR, calorie restriction; CRC, colorectal cancer.

In the present study, subcutaneous xenografts were subjected to IHC examination. Bax expression in the CR group was significantly increased (Fig. 2A-C), while Bcl-2 and Ki67 expression levels were significantly decreased compared with the control (Fig. 2D-I). The aforementioned results implied that CR was able to slow the progression of CRC xenografts by promoting apoptosis and suppressing proliferation.

Figure 2 CR promotes apoptosis and inhibits proliferation of colorectal cancer cells. Immunohistochemical staining of genes associated with apoptosis or proliferation. Compared with (A) control group (A), the expression of Bax was increased in the CR group (B). (C) Staining score of Bax expression in each group, respectively. Compared with the control group (D), the expression of Bcl2 was decreased in the CR group (E). (F) Staining score of Bcl2 expression in each group, respectively. Compared with the control group (G), the expression of Ki67 was reduced in the CR group (H). (I) Staining score of Ki67 expression in each group, respectively. Magnification, x400. Data are displayed as mean ± SD, n=5. ***P<0.001, ****P<0.0001 vs. control. CR, calorie restriction.

CR increases the presence of gut microbiota in mice

Chao1, Observed, Shannon and Simpson indices were used to characterize bacterial species richness and diversity to thoroughly assess the changes in diversity of the microbial communities in CRC mice under CR circumstances (Fig. 3A). The CR group exhibited higher species abundance and community diversity than the control; however, this difference was not statistically significant.

Figure 3 CR increases the presence of gut microbiota in mice. (A) Comparison of α diversity by Chao1, Shannon, Simpson and observed species index-based genus level. Data were analyzed using Kruskal-Wallis test. (B) Rank abundance. (C) Rarefaction curve. The ordinate of the rank abundance curve is the abundance value/mean value of each OTU in the sample/group following Log2 logarithmic transformation. Data are displayed as mean ± SD, n=5. CR, calorie restriction; OTU, operational taxonomic unit.

The rank abundance curve revealed that the presence of gut microbiota in the CR group was not significantly different compared with the control (Fig. 3B). Rarefaction curve was used to determine whether the sequencing or sample volume was saturated (18). The rarefaction curves of all samples converged toward a plateau, indicating that all OTUs were sufficiently covered by the sequencing (Fig. 3C). The findings of rarefaction curve were in line with the α diversity index that CR increased the species diversity of the samples. The microbial abundance of the CR group was larger compared with the control at the same sequencing depth, showing that CR increased the abundance of gut microbiota in mice.

CR modifies the β diversity of gut microbiota

The PCA, PCoA and NMDS were used in the β diversity analysis. PCA, using analysis of variance, was performed to detect the differences between multiple groups of data on a two-dimensional coordinate graph (Fig. 4A). To distinguish both groups stereoscopically, 3D-PCA plots were employed (Fig. 4B). PCoA was used to assess similarities or differences in data (Fig. 4C). NMDS was used to compare differences in bacterial community composition between sample groups based on the Bary-Curtis distance (Fig. 4D). The majority of the samples from the CR group were clustered together, according to PCA, 3D-PCA, and PCoA, whereas they were divided from the control group. Similar results were obtained using NMDS analysis. The closer proximity between each point, the more similar the sample compositions are. The matrices of the two groups were separated, except PCA (Fig. 4A), suggesting that CR caused alterations in the structure of the gut microbiota in CRC mice.

Figure 4 CR modifies β diversity of the gut microbiome. (A) PCA, (B) 3D PCA, (C) PCoA and (D) NMDS plot. The close proximity indicates that the sample compositions are similar. The percentages show how much of the distance matrix sample difference data each axis can explain. CR, calorie restriction; PCA, principal component analysis; PCoA, principal coordinate analysis; NMDS, non-metric multidimensional scaling.

CR alters the composition of gut microbiota

To analyze phylogenetic or population genetics studies, the same mark (Sequences are divided into distinct OTUs according to a 97% similarity threshold, and each OTU is usually treated as a microbial species. A similarity of less than 97% was considered to belong to a different species, and a similarity of between 93 and 95% was considered to belong to a different genus.)is artificially set for a specific taxonomic unit (categories and groups, such as species, genus, family, order, class, phylum, and domain.), also known as the operational taxonomic unit (OTU) (19). OTU in each set were counted according to the grouping of samples and the Venn diagram was drawn to study which species were common and which were unique among different sample groups. Finally, a total of 18,626 OTUs were identified in the control group and 17,790 OTUs were found in the CR group, with the control and CR groups sharing a total of 2,886 OTUs (Fig. 5A).

Figure 5 CR alters composition of gut microbiota. (A) Operational taxonomic unit Venn diagram of the intestinal flora. Relative abundances of species (B), species with differences (C), and heat map of species composition (D) at the phylum level in control and CR groups. At the genus level, each group's relative abundances of species (E), species with differences (F), and heat maps of species composition (G) are shown. Data are presented as mean ± SD, n=5. *P<0.05 vs. control. CR, calorie restriction.

Following comparison of the species composition of taxa at the phylum and genus levels in the fecal samples of CRC mice, discrepancies were discovered between the species compositions of the CR group and the control group (Fig. 5B and C, E and F). Firmicutes, Bacteroidetes and Verrucomicrobia comprised the majority of the OTU at the phylum level; however, the differences in components between these two groups were not statistically significant. The OTU at the genus level was primarily composed of Lactobacillus, Oscillospira, Ruminococcus and Akkermansia. Moreover, the proportion of Lactobacillus in the CR group was significantly higher than that in the control. Subsequently, heat maps of species composition of mouse fecal samples at the phylum and genus levels were used to show species abundance in each group (Fig. 5D and G). Proteobacteria, Chloroflexi, Gemmatimonadetes, Chlorobi, Acidobacteria, Armatimonadetes, Spirochaetes, OD1, Firmicutes and WS3 were upregulated in the CR group compared with the control. Tenericutes, Deferribacteres, Actinobacteria, Fusobacteria, WPS-2, Verrucomicrobia, Nitrospirae, TM7, Bacteroidetes and Cyanobacteria were downregulated in the CR group at the phylum level (Fig. 5D). By contrast, at the genus level, Oscillospira, Streptococcus, Ruminococcus, Anaeroplasma, Dehalobacterium, Ruminococcus, Prevotella, AF12, cc115, Akkermansia, Coprococcus, Bifidobacterium and Adlercreutzia were downregulated in the CR group compared with the control group, whereas Parabacteroides, Bacteroides, Roseburia, Alistipes, rC4-4, Lactobacillus and Candidatus arthromitus were upregulated (Fig. 5G).

LEfSe was used to reveal species differences between samples within each group at all taxonomic levels to identify species with notable differences (20,21). Compared with the control group, the gut microbiota in the CR group had the highest abundance of Chloroflexi at the phylum levels and Saprospirae and Anaerolineae at the class levels (Fig. 6A and B). Saprospirales, Rhodobacterales and Pasteurellales accounted for a higher proportion in the CR group at the order level, which was not the case in the control group. Clostridiaceae, Peptococcaceae, Chitinophagaceae, Micrococcaceae, Porphyromonadaceae, Rhodobacteraceae, Pasteurellaceae and Streptomycetaceae were more prevalent at the family level in the CR group compared with the control. By contrast with the control group, the enteric microorganisms of the CR group with higher abundance at the genus level were Candidatus arthromitus, rc4-4, Aggregatibacter, Anaerostipes, Rothia, Rhodobacter, Parabacteroides, Streptomyces and Faecalibacterium.

Figure 6 LEfSe study revealed that gut microbiota of the CR and control group vary at every level. (A) LDA score showed notable differences in the intestinal flora of mice in each group. (B) Taxonomic cladogram using the LEfSe method to show the phylogenetic distribution of fecal microbes associated with groups. The larger the LDA score, the greater the influence of species abundance on differences between groups. White nodes in the taxonomic cladogram depict taxa that do not differ between groups, while nodes that are colored represent differential taxa. The letters indicate taxa with differences between groups. The abundance is larger in the grouped samples. CR, calorie restriction; LEfSE, linear discriminant analysis effect size; LDA, logarithmic discriminant analysis.

CR regulates key marker flora to interfere with CRC

Receiver operating characteristic (ROC) curve analysis is a widely used statistical analytic tool in medical research to assess the performance of diagnostic tests (22). In the present study, the species compositions of the control and the CR group were combined at the genus level to draw the ROC curve of bacteria and calculate the area under the curve (AUC) to determine which bacteria were regulated by CR (Fig. 7A and B). Analysis of the ROC curve revealed that Bacteroides (AUC=0.800), Lactobacillus (AUC=0.760) and Roseburia (AUC=0.720) had relatively high accuracy, indicating that CR may suppress CRC by controlling these bacteria.

Figure 7 Evaluation of the predictive ability of microbiota markers for CRC in CR mice. Receiver operating characteristic analysis of representative differential (A) up- and (B) downregulated gut microbiota in the CRC model mice. The anticipated FPR was higher than the observed FPR. CRC, colorectal cancer; CR, calorie restriction; TPR, true positive rate; FPR, false positive rate.

Gene function prediction of gut microbiota

The 16S sequencing data was analyzed to identify the roles of bacterial genes (microbial genes after eliminating the host components) to explore the association between dysbiosis and CRC. KEGG) database was used to annotate the mouse fecal sample genes and a total of six categories and 33 subcategories of functional gene were discovered (Fig. 8A). There were four gene annotations associated with ‘transport and catabolism’ under cellular processes. In the environment classification, the most annotated topics were ‘membrane transport’ and ‘signal transduction’ with three each. ‘Replication and repair’ received more gene annotations than other categories, such as ‘folding, sorting and degradation’, ‘Transcription’ and ‘Translation’ in the genetic section, totaling six. The majority of human disease-associated genes in the gut microbiota of CRC under CR were associated with ‘infectious diseases’, with seven genes. Compared with other metabolism-associated genes, ‘Carbohydrate metabolism’ and ‘xenobiotics biodegradation and metabolism’ obtained more annotations, 17 and 15, respectively in metabolism classification, showing that the CR group had robust carbohydrate metabolism and gut microbiota capable of breaking down and metabolizing of foreign chemicals. ‘Endocrine system’ genes had the highest level of annotation across all organismal systems classification, suggesting that CR may maintain the stability of endocrine function in mice with CRC.

Figure 8 Gene function prediction of gut flora. (A) KEGG and (B) COG functional pathways and counts of gut microbiota in the CR group. First-level pathways/classifications are mean abundance of all samples. Metabolic pathways with notable differences between groups in (C) KEGG and (D) COG pathway. The positive value of log2(fold-change) on the x-axis represents the up-regulation compared with the control group. The negative value is downregulation compared with the control group. CR, calorie restriction; KEGG, Kyoto Encyclopedia of Genes and Genomes; COG, Clusters of Orthologous Genes; AUC, area under the curve.

COG, a homologous protein annotation database established by National Center for Biotechnology Information (NCBI) to classify and assemble the encoded proteins of 21 entire genomes of bacteria, algae and eukaryotes, was used to analyze the associated COG pathways in the CR group (23). The COG database annotation results revealed that functional genes in CR mice fell into four primary categories and 25 subcategories (Fig. 8B). The findings revealed that 271 annotations were assigned to ‘amino acid transport and metabolism’. The gut microbiota in the CR group had 214 genes annotated in ‘Cell wall/membrane/envelope biogenesis’, indicating that the gut microbiota biofilm creation was a major function. ‘Defense mechanisms’ made up 107 annotations in the CR group, showing that CR may aid colon cancer by increasing resistance to environment hazards. The functional characteristics of 851 gene annotations remained unclear.

The most substantially divergent pathways were identified by calculating the abundance values of KEGG and COG databases. The most downregulated pathway from the KEGG database in the CR group, compared with the control, was cyanoamino acid metabolism (pathway ID: ko00460; Fig. 8C). Based on COG database analysis, ‘replication initiation protein REP (rolling circle plasmid replication) (pathway ID: COG5655), tRNA G10 N-methylase Trm11 (pathway ID: COG1041) and uncharacterized protein with cyclophilin fold, contains DUF369 domain’ (pathway ID: COG2164) were the most downregulated categories in the CR group (Fig. 8D).

Discussion

Controlling dietary intake has been hypothesized to relieve physical and mental burdens associated with obesity (24). However, in recent years, an increasing number of studies has shown that CR provides patients with numerous other benefits in addition to weight loss (25,26). Choi et al (27) stated that CR allows the body to eliminate harmful cells while generating healthy new cells, therefore alleviating multiple sclerosis symptoms. Studies have conducted more in-depth investigations on how CR impacts the ability to perform these functions that assist in protecting the body against disease and maintaining health (28-32). In addition to the above-mentioned studies, Sbierski-Kind et al (33) showed that the gut microbiome is shaped by CR by reducing the levels of effector memory CD8+ T cells and memory B cells in mice, possibly postponing immunological senescence. However, the precise mechanisms behind this change are still unclear.

The transformation and absorption of nutrients in the human body is affected by intestinal microecology. Previous studies involving the use of metagenomics and bioinformatics technology in microecology have analyzed the structure and characteristics of intestinal microecology in various populations, resulting in a body of research on the association between intestinal microecology and tumors (34,35). Gut microbiota crosstalk with innate and acquired immune cells has been shown to enhance the intermediate effects of innate immune cells, antitumor effect of acquired immune cells and tumor immunogenicity of cells, thereby reprogramming tumor microenvironment immunity and improving the immune checkpoint inhibitor response (36). Additional studies have investigated changes in the richness and percentage of flora to discover meaningful markers to aid clinical research (37-39).

The present study found that CR decreased the volume and weight of subcutaneous CRC xenografts in mice by promoting CRC apoptosis while also inhibiting proliferation. The analysis of 16s rRNA sequencing on feces revealed CR markedly enhanced the abundance of gut microbiota in mice. In the normal gut microbiome, Bacteroidetes and Firmicutes are the most prevalent phyla, accounting for >80% of the gut microbiota (40). In addition, the Firmicutes to Bacteroidetes (F/B) ratio is an important indicator of dysbiosis in the gut microbiome (41-43). However, increased proportion of Bacteroidetes is considered to be advantageous to host health (44). Stojanov et al (45) revealed that the F/B ratio was greater in obese patients and markedly lower in patients with IBD). It has been proposed that Firmicutes bacteria extract energy from food more efficiently than Bacteroidetes, resulting in more efficient calorie absorption and consequent increased weight gain (46). However, Firmicutes is negatively connected with gut immune factors and antimicrobial peptides and positively correlated with inflammatory proteins and oxidative stress parameters, as Xia et al (47). Magne et al (46) showed that Bacteroidetes produces mostly acetate and propionate, while Firmicutes produces more butyrate, a health-promoting molecule shown to optimize insulin sensitivity, exert anti-inflammatory activity, regulate energy metabolism and increase leptin gene expression (48). In the present study, Firmicutes was notably increased in the CR group at the phylum level, whereas Bacteroidetes was notably decreased; this increase in the F/B ratio in the CR group may indicate that CR led to changes in the mouse gut microenvironment that were not detrimental to the mice.

Akkermansia is the most pervasive Verrucomicrobia species observed in humans and high-fat and high-calorie meals enhance the presence of Akkermansia (49,50). Verrucomicrobia has also been shown to promote regulatory immunity (51), making it a target for gut microbial intervention to improve regulatory immunity. Wu et al (52) demonstrated that interleukin-6 knockout mice possess markedly changed gut microbiota diversity than wild-type C57BL/6J mice, which included a decrease in the presence of Firmicutes at the phylum level and Lactobacillus at the genus level but an increase in Verrucomicrobia at the phylum level and Akkermansia at the genus level. Despite the absence of statistical significance in the present study, it was noted that CR mice had lower levels of Verrucomicrobia at the phylum level and decreased Akkermansia at the genus level. Additional studies into the association between these aforementioned changes in the flora caused by CR and the immunity of the organism are still required.

The Lactobacillus genus is a group of microorganisms that live in the body and benefit the host health. Previous research has identified that oral preparations containing Lactobacillus strains restore intestinal barrier function and immune markers and decrease systemic inflammation and/or cancer progression (53,54). Lin et al (55) showed that probiotics, including Lactobacillus and Bifidobacterium, prevent CRC growth by decreasing inflammation and angiogenesis, as well as improving function of the intestinal barrier by secreting short-chain fatty acids. In the present research, a notable increase in Lactobacillus was discovered in CR mice, with AUC=0.760, indicating that CR may improve the intestinal barrier function of mice and effectively control the inflammatory response, thus inhibiting CRC growth.

In the present study, levels of Bacteroides and Roseburia in the CR group were higher compared with the control. Bacteroidetes species are constituents of the Bacteroidetes phylum, with the genus Bacteroidetes containing the most common Bacteroidetes species in the human gut (56). Bacteroidetes produce butyrate and induce regulatory T cell development, both of which decrease inflammation (57). Bacteroides levels are considerably lower in obese children and teenagers and are inversely associated with low-density lipoprotein cholesterol in the blood, waist circumference and BMI (58,59). Roseburia is a Gram-positive, anaerobic, butyrate-producing bacterium that was originally isolated from human excrement and belongs to the Firmicutes phylum (60). Roseburia is detected in low abundance in numerous intestinal disorders, implying that the bacteria serve a vital function in maintaining intestinal homeostasis, such as generating short-chain fatty acids (61). Furthermore, compared with the general population, the presence of Roseburia is negligible in patients with inflammatory bowel illness such as ulcerative colitis and Crohn's disease (62,63). The increase in Bacteroides and Roseburia in the CR group suggests that the ability of CRC mice to maintain intestinal homeostasis improved and CR may have altered the gut microbial environment of CRC mice, enhancing immunological function and exerting anti-tumor effects. However, our study used a right flank site xenograft tumor model rather than a CRC model in situ in the intestine, thus more studies are needed to corroborate our hypothesis.

Although microbiota genome information was not established, 16S sequencing data was analyzed using PICRUSt to identify the potential functions of the microbiota genes. With only the sequencing data of microbiota marker genes, the known microbial genome data was used to forecast the composition of microbiota genes or functional units for intestinal microbiota according to 16S rRNA sequencing results (64). KEGG and COG databases were used to identify the potential functions of associated metabolic pathways. The metabolism of cyanoamino acids leads to an increase in the metabolism and production of intracellular signaling molecules and proteins, as well as the creation of biofilms (65). In gastric cancer, cyanoamino acid metabolism is disturbed and disorganized, which primarily manifests as upregulation of glycine levels and the downregulation of alanine levels (66,67). The downregulation of cyanoamino acid metabolism in the CR group may contributes to a better understanding of the underlying mechanisms of CRC . In the present study, replication initiator protein REP (rolling circle plasmid replication), tRNA G10 N-methylase Trm11 and uncharacterized protein with cyclophilin fold proteins were all found to be downregulated in the CR group. Rolling circle replication (RCR) is a replication initiation mechanism used by plasmids of certain Gram-negative bacteria (68). tRNA G10 N-methylase Trm11 protein is ubiquitous in archaea and eukaryotes (69). The enrichment of these gene annotations contributes to understanding of the mechanism by which CR serves a role in suppressing CRC and regulating gut microbiota.

There are limitations to the present study. Previous studies have shown that gut dysbiosis serves a key role in the development, progression and response to the treatment of CRC (70-72). Therefore, it can be concluded that remodeling of gut microbiota contributes to the suppressive effect of CR on CRC development in the mice. However, a cause-effect investigation is required to determine the key functions and mechanisms of gut bacteria in the regulation of CRC by CR in future. In addition, intestinal mucosa is a dynamic environment where the host continually interacts with trillions of commensal microorganisms and sporadically interacts with pathogens (73). The present study failed to collect the microorganisms from the intestinal mucosa to analyze the intestinal transit bacteria due to technical limitations. It is hypothesized that examining the bacteria in the gut mucosa will support the present results.

In conclusion, the present study revealed that CR modified gut microbiota and inhibited CRC growth by regulating apoptosis and proliferation of CRC cells in a mouse model. CR increased the proportions of beneficial bacteria, such as Lactobacillus, which may provide a novel approach to treating CRC by CR-induced remodeling of gut microbiota. As studies on gut microbiota increase, it is anticipated that the development of a new food culture centered on low-calorie diet may assist in preventing and controlling the progression of CRC.

Acknowledgements

Not applicable.

Funding

Funding: The present study was supported by grants from the National Natural Science Foundation of China (grant no. 81860442), the Natural Science Foundation of Ningxia Province (grant nos. 2022AAC02027 and 2022AAC03475) and the Scientific Research Project of Ningxia Medical University (grant no. XZ2020006).

Availability of data and materials

The datasets generated and/or analyzed during the current study are available in the NCBI Sequence Read Archive repository (accession no. PRJNA890426; ncbi.nlm.nih.gov/bioproject/PRJNA890426).

Authors' contributions

XCD and YHZ wrote the draft of the manuscript. XCD, YHZ and YLH conducted the experiments. XCD, YHZ, YLH, YJF and XTW collected and analyzed the data. YJG and FW designed and supervised the study, and revised the manuscript. All authors have read and approved the final manuscript. YJG and FW confirm the authenticity of all the raw data.

Ethics approval and consent to participate

The Animal Experimental Ethics Committee of Ningxia Medical University (Yinchuan, China) approved this study (approval no. 2021-045).

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References