Microbial Ecology

, Volume 66, Issue 2, pp 462–470

Dysbiosis Signature of Fecal Microbiota in Colorectal Cancer Patients

Authors

  • Na Wu
    • CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of MicrobiologyChinese Academy of Sciences
  • Xi Yang
    • CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of MicrobiologyChinese Academy of Sciences
  • Ruifen Zhang
    • CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of MicrobiologyChinese Academy of Sciences
  • Jun Li
    • Department of Abdominal Surgical Oncology, Cancer Institute and HospitalChinese Academy of Medical Sciences
  • Xue Xiao
    • CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of MicrobiologyChinese Academy of Sciences
  • Yongfei Hu
    • CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of MicrobiologyChinese Academy of Sciences
  • Yanfei Chen
    • State Key Laboratory for Diagnosis and Treatment of Infectious Disease, The First Affiliated HospitalZhejiang University
  • Fengling Yang
    • State Key Laboratory for Diagnosis and Treatment of Infectious Disease, The First Affiliated HospitalZhejiang University
  • Na Lu
    • CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of MicrobiologyChinese Academy of Sciences
  • Zhiyun Wang
    • CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of MicrobiologyChinese Academy of Sciences
  • Chunguang Luan
    • CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of MicrobiologyChinese Academy of Sciences
  • Yulan Liu
    • Department of GastroenterologyPeking University People’s Hospital
  • Baohong Wang
    • State Key Laboratory for Diagnosis and Treatment of Infectious Disease, The First Affiliated HospitalZhejiang University
  • Charlie Xiang
    • State Key Laboratory for Diagnosis and Treatment of Infectious Disease, The First Affiliated HospitalZhejiang University
  • Yuezhu Wang
    • Chinese National Human Genome Center at Shanghai
  • Fangqing Zhao
    • Beijing Institutes of Life ScienceChinese Academy of Sciences
  • George F. Gao
    • CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of MicrobiologyChinese Academy of Sciences
  • Shengyue Wang
    • Chinese National Human Genome Center at Shanghai
  • Lanjuan Li
    • State Key Laboratory for Diagnosis and Treatment of Infectious Disease, The First Affiliated HospitalZhejiang University
    • Department of Abdominal Surgical Oncology, Cancer Institute and HospitalChinese Academy of Medical Sciences
    • CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of MicrobiologyChinese Academy of Sciences
Host Microbe Interactions

DOI: 10.1007/s00248-013-0245-9

Cite this article as:
Wu, N., Yang, X., Zhang, R. et al. Microb Ecol (2013) 66: 462. doi:10.1007/s00248-013-0245-9

Abstract

The human gut microbiota is a complex system that is essential to the health of the host. Increasing evidence suggests that the gut microbiota may play an important role in the pathogenesis of colorectal cancer (CRC). In this study, we used pyrosequencing of the 16S rRNA gene V3 region to characterize the fecal microbiota of 19 patients with CRC and 20 healthy control subjects. The results revealed striking differences in fecal microbial population patterns between these two groups. Partial least-squares discriminant analysis showed that 17 phylotypes closely related to Bacteroides were enriched in the gut microbiota of CRC patients, whereas nine operational taxonomic units, represented by the butyrate-producing genera Faecalibacterium and Roseburia, were significantly less abundant. A positive correlation was observed between the abundance of Bacteroides species and CRC disease status (R = 0.462, P = 0.046 < 0.5). In addition, 16 genera were significantly more abundant in CRC samples than in controls, including potentially pathogenic Fusobacterium and Campylobacter species at genus level. The dysbiosis of fecal microbiota, characterized by the enrichment of potential pathogens and the decrease in butyrate-producing members, may therefore represent a specific microbial signature of CRC. A greater understanding of the dynamics of the fecal microbiota may assist in the development of novel fecal microbiome-related diagnostic tools for CRC.

Introduction

The human intestinal tract contains about 1014 bacteria [1, 2]. A highly diverse and dense gut microbiota plays a crucial role in intestinal health, through digestion of food, protection of mucosal surfaces, and crosstalk with the host immune system [3, 4]. Dysbiosis of the normally stable gut microbiota might adversely affect the health status of the host and has been associated with diseases such as obesity, diabetes, inflammatory bowel disease, and colorectal cancer (CRC) [58].

CRC is the third most common cancer and the fourth leading cause of cancer death worldwide [9]. Some specific gut bacteria have been associated with the pathogenesis of CRC [10]. Previous studies using both culture-dependent and culture-independent techniques have helped to elucidate the association of one or more microbial species with CRC. Fifteen bacterial taxa, including Bacteroides species, were implicated in CRC through cultivation of the fecal microbiota [11]. Using DNA fingerprinting, Scanlan et al. demonstrated that the Clostridium leptum and Clostridium coccoides subgroups were specific to CRC and polyposis [6]. Increased Fusobacterium species were also reported in CRC tumor tissues compared with controls [12, 13].

These previous studies have revealed the dysbiosis of gut microbiota in CRC patients. However, the precise microbial species present in the fecal microbiota involved in CRC remain unclear. Next generation sequencing technologies allow the study of fecal microbiota at a level of detail that was previously not available [14]. In this study, we used pyrosequencing of 16S rRNA gene V3 region to profile the composition of the microbiota of fecal samples from 19 CRC patients and 20 healthy subjects. Our results showed striking differences in fecal microbial population patterns between the two groups and suggested that fecal microbiota composition may be a useful marker for CRC.

Materials and Methods

Patients and Sample Collection

All the 20 CRC patients were from the Cancer Institute and Hospital, Chinese Academy of Medical Sciences, Beijing, China. The patients were selected based on the following criteria: no complicating diseases (such as diabetes or hypertension), no family history of CRC, and no antibiotic use within 3 months prior to the stool sample (Table S1). The 20 healthy subjects were selected based on matched sex, age, and body mass index; no gastrointestinal disorders; and no antibiotic use during the 3-month period prior to sample collection (Table S2). During the stool collection and DNA extraction phase, one of the CRC fecal samples (C2) was scant, and the DNA extracted from that sample did not pass quality control, which reduced the number of CRC samples from 20 to 19. There were 10 males and 9 females in the CRC group, and 11 males and 9 females in the healthy group. The mean (±SD) age of the subjects was 58.3 ± 8.7 years for the CRC group and 53.2 ± 5.4 years for the healthy group. All 19 of the CRC patients were categorized according to their clinical data following surgery, using the TNM classification system for malignant tumors (Table S1). One stool sample was collected from each patient and each healthy subject. Stool samples were frozen immediately after collection and stored at −80 °C until DNA extraction. This research project was approved by the Research Ethics Committee of the Institute of Microbiology, Chinese Academy of Sciences and informed consent forms were signed by all of the subjects prior to participation.

Pyrosequencing

Two hundred milligrams from each stool sample was used for DNA extraction using a QIAamp DNA Stool Mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The quality of the extracted fecal DNA was measured using a NanoDrop 1000 (NanoDrop Technologies, Wilmington, DE, USA). The 16S rRNA gene V3 region was amplified using the universal primers 341F (5′-NNNNNNNNCCTACGGGAGGCAGCAG-3′) and 534R (5′-NNNNNNNNATTACCGCGGCTGCTGG-3′). NNNNNNNN represents the sample-unique 8-bp barcode used to tag each polymerase chain reaction (PCR) amplicon. PCR amplification was performed using TakaRa Taq Hot start version DNA polymerase according to the manufacturer’s instructions (TakaRa, Dalian, China). PCR products were purified from a 0.8 % agarose gel using an AxyPrep DNA Gel Extraction kit (AxyGen, Tewksbury, MA, USA). One microgram of the gel-purified DNA was quantified using the NanoDrop 1000 and was added to a master pool of DNA for pyrosequencing using the Genome Sequencer FLX System (Roche, Branford, CT, USA).

Raw sequence reads were trimmed according to the following criteria: (1) a perfect match to at least one end of the barcode and the 16S rRNA gene primer, (2) a length of at least 120 nt, and (3) no more than one N in the sequence read. More than 106 raw sequence reads were obtained from the 39 gut microbiota samples. A set of 722,366 sequence reads satisfied the above criteria for the trimmed sequence reads and was filtered, accounting for 67.6 % of the raw reads produced.

Bioinformatics

Taxonomy definitions for the pyrosequencing reads were assigned using the Ribosomal Database Project (RDP) Classifier 2.3, with a confidence threshold of 50 % [15, 16]. Automation of the workflow was implemented by a set of customized Perl scripts. The relative abundance of the different phyla, families, and genera in each sample was analyzed. The data were further analyzed by principal coordinate analysis (PCoA), based on the unweighted UniFrac distance metric [17], to visualize the difference between the two groups. This method accounts for the degree of divergence between different sequences.

The 722,366 trimmed reads were clustered into operational taxonomic units (OTUs) at 97 % identity using MOTHUR [18], which allows users to use a single piece of software to analyze community sequence data. A comparison of the bacterial richness and diversity of the samples was performed by the Chao1 and Shannon indexes. The coverage analysis was performed using Good’s coverage estimator. OTUs present in 50 % or more of the gut samples were identified as core phylotypes, according to the definition published by Tap et al [19]. The 244 core OTUs were selected to perform the partial least-squares discriminant analysis (PLS-DA), using SIMCA-P + (version 12.0). A variable importance plot (VIP) parameter greater than 1 was then used to choose the variables with the most significant contribution to the separation of the groups. To determine the phylogeny of the OTUs, the longest sequence from each OTU was BLAST searched against the Greengenes database [20], which harvests the well-annotated full-length 16S rRNA gene sequences.

Real-Time Quantitative PCR

To determine the amounts of Fusobacterium, fecal DNA from each sample was assayed by real-time quantitative PCR (RT-qPCR) detection of 16S rRNA genes. Quantitative PCR was performed using Power SYBR Green PCR Master Mix (Applied Biosystems). The primer sequences were as follows: fuso-F (5′-GGATTTATTGGGCGTAAAGC-3′) and fuso-R (5′-GGCATTCCTACAAATATCTACGAA-3′) [21]. Amplifications were performed using the following cycling protocol: one cycle at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s, and 60 °C for 60 s. Denaturation curves were carried out from 60 to 95 °C for quality assurance. Plasmid DNA containing the respective amplicon was diluted in 10-fold increments (108–101 copies) and used as quantification standards for RT-qPCR. The copy number of target DNA was determined by comparison with the plasmid dilution series standard curves. Bacterial quantity was expressed as log10 bacteria per gram of stool.

Statistical Analysis

The Mann–Whitney test was used to evaluate the difference in bacterial populations between the CRC and healthy groups. Correlation between variables was computed using the Spearman rank correlation. The Mann–Whitney test and Spearman rank correlation were conducted in SPSS software 17.0.

Accession Number

The sequence data in this study have been deposited in the GenBank Sequence Read Archive under BioProject accession PRJNA168081.

Results

Sequencing Data

We analyzed the host-associated microbiota from fecal samples collected from 19 CRC patients and 20 healthy subjects based on the V3 region of 16S rRNA gene (Table S1 and Table S2). After applying strict trimming criteria to exclude low-quality reads, 722,366 sequences of 16S with acceptable quality were obtained, with an average of 18,522 reads per sample. Only one CRC patient sample and three healthy subject samples had fewer than 10,000 reads (Table S3). To investigate whether there was any effect of the discrepancy in sequence reads between different samples, we carried out statistical analysis and determined that the difference between the sequence reads of the two groups was not significant (P > 0.05). In addition, the coefficient of variation values for the sequence numbers was 34.7 and 32.0 % for the CRC and the healthy groups, respectively, which was acceptable for further analysis.

The total number of OTUs at 97 % identity was 25,891, and 244 OTUs were identified as core phylotypes [19]. The pyrosequencing results for the subjects had a mean Good’s coverage score of 94 %. The Chao1 index was used to estimate the microbial richness, and the Shannon index was used to assess the diversity of the fecal microbiota in the CRC and healthy groups. No significant differences in either the richness or the biodiversity were found.

To gain an overview of the microbial community structure, we performed a PCoA based on the unweighted UniFrac metric. The healthy subjects and the CRC patients had a clear separation on the first two principal component axes, which accounted for 14.02 and 10.15 % of the total variations (Fig. 1a). A multivariate analysis method, PLS-DA, was then used to confirm the differentiation of the two groups based on the relative abundance of the 244 core OTUs. The analysis showed a sharp clustering of the microbiome sequence data in all 39 samples (Fig. 1b). The model of the dataset was R2(Y) = 0.903 and Q2 = 0.382, suggesting a good performance for disease discrimination and prediction.
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-013-0245-9/MediaObjects/248_2013_245_Fig1_HTML.gif
Figure 1

Overall structure of the gut microbiota of all samples. a Plot of the principal coordinate analysis (PCoA) based on the unweighted UniFrac metric. b PLS-DA score plots based on the relative abundance of 244 core OTUs (97 % similarity level). Blue squares represent the CRC patients, and red circles represent the healthy subjects

Differences in Fecal Microbial Communities Between Healthy and CRC Groups Based on Taxonomic Comparisons

At the phylum level, 10 phyla were present in all of the samples, with Bacteroidetes and Firmicutes being predominant, as expected. The relative abundance of Bacteroidetes species was 53.9 and 56.5 % in the healthy and CRC groups, respectively, while the relative abundance of Firmicutes was 38.9 and 36.6 %, respectively. The remaining phyla were Actinobacteria, Proteobacteria, Fusobacteria, Lentisphaerae, Synergistetes, TM7, Tenericutes, and Verrucomicrobia. No significant differences were observed between the healthy subjects and the CRC patients for most of the phyla, with the exception of Fusobacteria. The CRC group had a marked increase in the relative abundance of Fusobacteria compared with the control group (mean 1.02 vs. 0.47 %, P < 0.01).

CRC clearly affected the gut microbial community at the family level (Fig. 2). The ratio of the three families Eubacteriaceae (P = 0.037), Clostridiales Family XI. Incertae sedis (P = 0.004) within the order Clostridiales, and Staphylococcaceae (P = 0.011) within the order Bacillales was significantly higher in the CRC group than in the healthy group. The family Enterococcaceae was also more prevalent in the CRC group, although the change was not significant (P = 0.062). Two other families, Campylobacteraceae (P = 0.014) and Porphyromonadaceae (P = 0.001), were significantly less prevalent in the healthy group. Fusobacteriaceae, the main component of the phylum Fusobacteria, were significantly enriched in the CRC group (P = 0.001).
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-013-0245-9/MediaObjects/248_2013_245_Fig2_HTML.gif
Figure 2

Comparison of the relative abundance at the family level between the CRC patients and healthy subjects. Subjects are indicated by squares (CRC group, n = 19) and triangles (healthy subjects, n = 20). Mean values are denoted by crosses. The Mann–Whitney test was used to evaluate the two groups. *P < 0.05; **P < 0.01

CRC samples were associated with genus-level changes; 16 enriched genera were significantly higher in CRC patients (Fig. 3), including Fusobacterium and Campylobacter. Remarkably, Fusobacterium had a higher relative abundance in the CRC group (P = 0.001). These results were further confirmed by RT-qPCR analysis (Table 1). To better characterize the composition of the genus Fusobacterium in human feces, we aligned the 16S sequences with the Greengenes database and selected the species definition with the most significant BLAST results. The results showed that several Fusobacterium species were present in different CRC patients: Fusobacterium nucleatum, Fusobacterium periodonticum, Fusobacterium necrophorum, Fusobacterium ulcerans, Fusobacterium varium, and Fusobacterium gonidiaformans (Table S4).
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-013-0245-9/MediaObjects/248_2013_245_Fig3_HTML.gif
Figure 3

Cladogram of CRC and healthy microbiota. Branches are colored based on phylum level assignments. Blue-colored nodes represent genera significantly overrepresented in the CRC group, and red-colored nodes represent genera significantly more abundant in the healthy group

Table 1

Real-time qPCR quantitation of the Fusobacterium 16S rRNA gene in the feces of CRC patients and healthy subjects

 

CRC patients

Healthy subjects

Log10 of Fusobacterium 16S rRNA gene copy number (per gram of wet weight)

Mean ± SD

7.5 ± 3.7

5.6 ± 3.6

Range

0–11.0

0–10.9

P value

0.014

CRC colorectal cancer, SD standard deviation

Identification of Key OTUs Responsible for Differentiation Between Healthy and CRC Groups

The PLS-DA method was performed to select 69 key phylotypes with VIP >1 that differentiated the two groups (Fig. 1b). We then clustered the samples according to the relative abundance of the 69 OTUs (Fig. 4). Forty-five OTUs were enriched in the CRC microbiota samples, with 17 OTUs belonging to the genus Bacteroides, 23 OTUs belonging to nine different genera (Alistipes, Blautia, Dorea, Escherichia/Shigella, Fastidiosipila, Odoribacter, Oscillibacter, Phascolarctobacterium, and Subdoligranulum), and 5 OTUs that were closely related to Ruminococcus. The remaining 24 OTUs of the 69 key phylotypes were enriched in the healthy microbiota, some of which even disappeared in CRC patients. Most of the OTUs related to the butyrate-producing genera Faecalibacterium and Roseburia were enriched in healthy subjects. Three OTUs were related to Faecalibacterium prausnitzii, while six OTUs belonged to the genus Roseburia.
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-013-0245-9/MediaObjects/248_2013_245_Fig4_HTML.gif
Figure 4

Community structure and composition in the gut microbiota of 69 key OTUs. Each vertical lane corresponds to a subject, and the colored squares in each row indicate the relative abundance of the OTU among the 39 subjects

The relative abundance of five OTUs out of the 17 Bacteroides-related OTUs was significantly enriched in the CRC microbiota. Bacteroides sequences were overrepresented in CRC fecal samples. Therefore, we then assessed the correlation between Bacteroides abundance and the disease status (TNM classification) of the CRC (Fig. 5). Remarkably, the relative abundance of the genus Bacteroides was positively associated with the disease status (R = 0.462, P = 0.046 < 0.5). However, there was no significant correlation between a host property (TNM classification) and the abundance of other bacterial groups (data not shown).
https://static-content.springer.com/image/art%3A10.1007%2Fs00248-013-0245-9/MediaObjects/248_2013_245_Fig5_HTML.gif
Figure 5

Correlation between the TNM classification and the relative abundance of the Bacteroides genus. The Spearman rank correlation (R) and probability (P) were used to evaluate the statistical importance

Discussion

In this study, we compared the composition of the human intestinal microbiota between CRC patients and healthy subjects using culture-independent pyrosequencing and RT-qPCR methods. In particular, we observed significant elevation of several bacterial groups, such as Bacteroides and Fusobacterium species, in the CRC group. Moreover, the high prevalence of Fusobacterium in fecal microbiota of CRC could be used as a possible fecal marker for the prediagnosis of CRC.

Pyrosequencing of 16S rRNA genes allows for in-depth characterization of complex microbial communities. Selection of a hypervariable region of the 16S rRNA gene is important to obtain representative characterization of complex microbial communities [22], owing to the biases associated with different regions. In our study, we selected the V3 region of the 16S rRNA gene for pyrosequencing. Pyrosequencing of V3 and V6 region amplicons has been used to describe the gut microbial community structures, and the findings largely are equivalent to results obtained using the longer Sanger sequences at the phylum, class, order, and genus levels [23, 24]. In addition, the V3 region has been widely used, while allowing a significantly greater depth of coverage than is possible with Sanger sequencing [25]. A total of 722,366 high-quality sequences were obtained and used for analysis.

It was reported that members of the genus Bacteroides have higher colonization rates in CRC patients [10, 36], which could be related to the production of fragilysin, an oncogenic bacterial toxin [26]. Fragilysin stimulates intestinal epithelial cells to secrete interleukin-8, resulting in an inflammatory response and increasing intestinal fluid secretion [27, 28]. In addition to inflammatory effects, fragilysin was reported to induce colonic epithelial cell proliferation, as well as expression of the oncogene c-Myc [29]. Enterotoxigenic Bacteroides fragilis was also shown to induce colon tumors in multiple intestinal neoplasia mice [30]. Here, we found that Bacteroides-related OTUs were more abundant in the CRC group than the control group, with a positive correlation between Bacteroides prevalence and CRC disease status (TNM classification). These results suggest that it is possible that the elevated Bacteroides species in colon play a role in exacerbating the disease. We hypothesize that the involvement of specific Bacteroides members can contribute to CRC development through the production potentially carcinogenic compounds and inflammatory mechanisms. In agreement with our findings, Sobhani et al. reported that Bacteroides species appears higher in 60 CRC patients than in 119 controls with normal colonoscopy in a proportion linked with IL17 overproduction [10], which might promote development of sporadic CRC [31, 32].

The role of infectious and inflammatory processes in colon carcinogenesis is of considerable interest [30]. Recently, F. nucleatum has been related to gastrointestinal disease [8, 33]. Kostic et al. [13] and Castellarin et al. [12] made tentative steps towards establishing an infectious cause of CRC, finding increased loads of Fusobacterium in colon cancer tissue. Our observations further confirm a significant overrepresentation of Fusobacterium in the fecal microbiota of CRC patients compared with healthy controls. Enriched amounts of Fusobacterium in the fecal samples of CRC patients were quantitatively confirmed by RT-qPCR assay. At the species level, we observed that F. nucleatum was prevalent in the gut microbiota of 13 CRC patients (Table S4). This bacterium is a very interesting species that shows prognostic associations with CRC, and an overabundance of F. nucleatum sequences in CRC tumors has also been positively associated with lymph node metastasis [12]. Therefore, the prevalence of F. nucleatum in CRC may be related to its invasive and inflammatory properties. More work is needed to confirm the increased prevalence of Fusobacterium in an independent cohort of CRC patients. This is currently under investigation and may lead to the development of a molecular method for the early detection of CRC.

Our findings also showed a greater abundance of other specific potential pathogens in CRC patients compared with the controls, for example, Enterococcaceae and Campylobacter species. The increased occurrence of Enterococcus within the family Enterococcaceae has been well described in the fecal samples of patients with Crohn’s disease [34] and CRC [35, 36]. A potential mechanism for the Enterococcus–CRC association could be related to the induction of extracellular inflammatory compounds that damage colonic epithelial cell DNA [37, 38]. In addition, Campylobacter species may be involved in the pathogenesis of intestinal disease through attachment to and invasion of intestinal epithelial cells, damage to intestinal barrier integrity, and toxin secretion [39, 40]. It should be noted that although we found that these bacterial groups were associated with CRC, the exact causal relationships between bacterial dysbiosis and CRC need to be further investigated.

Another feature of the dysbiosis in CRC microbiota was the significant reduction of butyrate-producing bacteria in the gut microbiota of CRC patients. These bacteria metabolize fiber and resistant starch to short chain fatty acids, such as butyrate, which has been suggested to strengthen and maintain mucosal barrier function through production of mucin, antimicrobial peptides, and tight-junction proteins [41]. A reduced concentration of butyrate-producing bacteria is implicated in the inflammatory cascade of the intestinal tract [42]. In our study, three key OTUs related to F. prausnitzii and six OTUs in the genus Roseburia, both of which supply butyrate in vivo, were significantly underrepresented in CRC patients. This is consistent with the decreased number of Roseburia species in fecal microbiota of CRC patients observed by Wang et al. [36]. A previous report also showed that F. prausnitzii numbers decreased approximately fourfold in CRC patients compared with healthy subjects [35]. Our current findings, together with these previous studies, support the assumption that the butyrate-producing bacteria F. prausnitzii and Roseburia species may play an important role in the prevention of CRC.

In summary, we have identified a dysbiosis signature of the gut microbiota in CRC patients using a case–control study. Whether these dysbiosis associations are involved in CRC pathogenesis, or simply the result of bacterial exploitation of an ecological niche because of a competitive advantage in the tumor microenvironment, remains to be tested. However, the observed abundance and detection rate of species belonging to the genus Fusobacterium provides a potential fecal marker for early detection of CRC. Further data on the relationship between these diverse organisms and CRC etiology may help monitor an individual’s microbiota for early detection of CRC.

Acknowledgments

This work was supported by funding from the National Basic Research Program of China (973 Program: 2009CB522605), the National Natural Science Foundation of China (NSFC, grant no. 81021003), and The Capital Health Research and Development of Special (grant no. 2011-4022-06).

Supplementary material

248_2013_245_MOESM1_ESM.pdf (159 kb)
ESM 1(PDF 159 kb)

Copyright information

© Springer Science+Business Media New York 2013