Current Microbiology

, Volume 61, Issue 1, pp 69–78

Molecular Characterisation of the Faecal Microbiota in Patients with Type II Diabetes


  • Xiaokang Wu
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
  • Chaofeng Ma
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
    • Center for Disease Control and Prevention of Xi’an
  • Lei Han
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
  • Muhammad Nawaz
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
  • Fei Gao
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
  • Xuyan Zhang
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
  • Pengbo Yu
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
  • Chang’an Zhao
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
  • Lianchuan Li
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
  • Aiping Zhou
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
  • Juan Wang
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University
  • John E. Moore
    • Northern Ireland Public Health Laboratory, Department of BacteriologyBelfast City Hospital
    • School of Biomedical SciencesUniversity of Ulster
  • B. Cherie Millar
    • Northern Ireland Public Health Laboratory, Department of BacteriologyBelfast City Hospital
    • Department of Immunology and Pathogenic Biology, Molecular Bacteriology Laboratory, Key Laboratory of Environment and Genes Related to Diseases of Chinese Ministry of Education, School of MedicineXi’an Jiaotong University

DOI: 10.1007/s00284-010-9582-9

Cite this article as:
Wu, X., Ma, C., Han, L. et al. Curr Microbiol (2010) 61: 69. doi:10.1007/s00284-010-9582-9


The investigation provides molecular analyses of the faecal microbiota in type 2 diabetic patients. In order to characterise the gut microbiota in diabetic patients and to assess whether there are changes in the diversity and similarity of gut microbiota in diabetic patients when compared with healthy individuals, bacterial DNAs from 16 type 2 diabetic patients and 12 healthy individuals were extracted from faecal samples and characterised by PCR-denaturing gradient gel electrophoresis (DGGE) with primers specifically targeting V3 region of the 16S rRNA gene, as well as been sequenced for excised gel bands. The counts of Bacteroides vulgatus, Clostridium leptum subgroup and Bifidobacterium genus were assessed using quantitative PCR. By comparing species diversity profiles of two groups, we observed that there were no significant differences between diabetic and healthy group, although a few diabetic individuals (D6, D8) exhibited a remarkable decrease in species profiles. As for the similarity index, it was lower in inter-group than that in intra-group, which showed that the composition of gut microbiota in diabetic group might be changed due to diabetes status. Sequencing results also revealed that bacterial composition of diabetic group was different from that of the healthy group. B. vulgatus and Bifidobacterium genus were low represented in the microbiota of diabetic group, and the significant decrease was observed for Bifidobacterium by real-time PCR. Taken together, in this work we observed the characterisation of gut microbiota in diabetic patients, which suggestes that the gut microbiota of diabetes patients have some changes associated with occurrence and development of diabetes.


Gut microbiota may play an even more important role in maintaining human health than previously thought. Over the past 5 years, studies have highlighted some key aspects of the mammalian host–gut microbial relationship. This complex community (10 to 100 trillion bacterial cells) could now be considered as a “microbial organ” localised within the host in that it performs functions essential for our survival [1, 7, 29], such as defence against pathogens, immunity, the development of the intestinal microvilli, the degradation of nondigestible polysaccharides.

There has been growing consensus that the increased prevalence of obesity and type 2 diabetes cannot be attributed solely to changes in the human genome, nutritional habits, or the reduction of physical activity in our daily lives [7, 16]. The gut microbiota was recently proposed as an environmental factor responsible for the control of body weight and energy metabolism, which is closely linked to obesity [1, 19, 20] and metabolic disorders such as type 2 diabetes [6]. Several recent studies by of Gordon et al. (USA) highlighted that gut microbiota composition was involved in the regulation of energy homeostasis. Ley et al. showed that the relative abundance of two predominant bacterial phyla (Bacteroidetes and Firmicutes) in the distal gut was linked to obesity in mice and humans, noting that obese mice and humans had a reduction in the abundance of Bacteroidetes and a proportional increase in Firmicutes [19]. Whereas the ratio of Bacteroidetes to Firmicutes approached a lean type profile after 52 weeks of diet-induced weight loss [20]. Together, the results obtained in rodents and humans suggest that obesity alters the nature of the gut microbiota. However, we are unable to distinguish if changes in microbiota profiles are the cause of the obesity or a result thereof. Turnbaugh et al. transplanted caecal microbiota from lean and ob/ob mice to germ free wild-type recipients. They found that after only 2  weeks, mice harbouring the microbiota from obese mice had a modest fat gain, and extracted more calories from their food compared to the lean mice having received the gut microbiota from lean mouse donors. These data suggest that the characteristics of gut microbiota of obese mice participate per se to the accretion of fat and body weight gain [36]. At the same time, the role of gut microbiota has been increasingly highlighted in the development of high-fat diet-induced diabetes and inflammatory state [2, 3]. Wellen et al. revealed that the composition of the gut microbiota may be directly responsible for inducing a low-grade inflammatory state closely associated with type 2 diabetes in response to being fed a high-fat diet [15, 41]. Cani et al. [5] demonstrated that the alteration of gut microbiota composition, due to high-fat diet feeding, was associated with a significant increase in plasma lipopolysaccharide (LPS), fat mass, body weight gain, liver hepatic triglyceride accumulation, insulin resistance, type 2 diabetes and atherosclerosis by mechanisms dependent of the LPS and/or the fatty acids activation of the CD14/TLR4 receptor complex. Moreover, prebiotic dietary fibres were used to increase the content of gut bifidobacteria specifically in high-fat fed mice, and the results confirmed that in prebiotic treated-mice, Bifidobacterium spp. significantly and positively correlated with improved glucose-tolerance, glucose-induced insulin-secretion and normalised low-grade inflammation (decreased endotoxemia, plasma and adipose tissue proinflammatory cytokines) [4, 35]. Creely et al. [8] recently reinforced the hypothesis that metabolic endotoxemia might act as a gut microbiota related factor involved in the development of type 2 diabetes and obesity in humans.

All these studies reveal that the gut microbiota and its related factors/mechanisms exert a crucial role in the development of diet-induced obesity and diabetes. Nevertheless, progress in understanding the mechanisms by which the gut microbiota interact with the host, will provide new basis for putative pharmacological or dietary intervention [7]. Although the microbiota in obese mice and humans have been extensively studied, there is only very few data available on structural changes and compositional evolution of gut microbiota in type 2 diabetic patients. And some questions remain incompletely characterised, for example, does the ecological characterisation of the gut microbiota change in all diabetes patients associated with different body weight, age, or gender? Do some predominant bacteria and beneficial commensal bacteria (e.g. Bacteroides, Bifidobacterium) change in type 2 diabetes patients?

With the development of culture-independent molecular methods, e.g. fluorescence in situ hybridisation (FISH), denaturing gradient gel electrophoresis (DGGE), real-time quantitative PCR and 16S rRNA gene libraries, a considerable progress have been made in characterising complex microbial communities. Specifically, DGGE of PCR-amplified 16S rRNA genes, is routinely used now to assess the diversity of microbial communities [9], and provides a DNA fingerprint of each sample and permits subsequent identification of community members by sequence analysis to monitor their dynamics.

The aim of this investigation was to characterise faecal microbiota in diabetes patients and healthy individuals. PCR–DGGE was combined with image analysis to give insights into the microbial diversity and similarity of faecal samples, while UPGMA dendrogram construction and sequencing were done to test for disease-associated DGGE motifs and taxa. Bifidobacterium, Bacteroides fragilis group and Clostridium leptum subgroup were tested by real-time PCR to understand accurate changes of gut microbiota in diabetes patients.

Materials and Methods

Samples Collection and Processing

Sixteen faecal samples were collected from 16 inpatients with type 2 diabetes (10 male and 6 female) from January to June 2008. The patients were between 48 and 62 years old, 12 of whom were new cases with unstable blood sugar levels and four were treated cases with stable blood sugar levels. The patients were interviewed following a questionnaire assessing: age; gender; body length and weight; individual health status (including chronic or acute diseases and blood sugar levels); and life-style aspects (such as physical activity and dietary habits). Two patients were tracked over a 2-week period (samples taken on days 1, 5, 10 and 14) to observe the stability of intestinal microbiota.

Faecal samples, as well as the profiles of food habits, were also collected from 12 healthy volunteers (from 50- to 65-year old) as a control for this study. None of the patients had a history of gastrointestinal diseases nor received antibiotics, probiotics and prebiotics, within 30 days prior to sampling. All faecal samples were collected using sterile cups and were immediately stored at −80°C. The sampling process was performed with the approval of the local Ethics Committee and the informed consent of the patients.

Bacterial Strains

Three bacterial strains used in this study were obtained from Chinese Centre of Industrial Culture Collection (CICC, CHINA) and Culture of the Yakult Central Institute (Tokyo, Japan), including Bacteroides vulgatus (CICC 22938), Bifidobacteria longum (CICC 6186), Clostridium leptum (YIT 6169). These strains were cultured anaerobically in GAM broth (Nissui Seiyaku Co., Ltd., Tokyo) supplemented with 1.0% glucose at 37°C for 24 h.

Extraction of Total DNA from Stool Samples

Faecal samples were thawed on ice, and DNAs were extracted using the QIAGEN QIAamp MiniStool kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions, with an initial bead-beating step of 30 s at 5,000 rpm. Extracts were treated with DNase-free RNase (100 μg/ml) and the DNA concentration was determined using a NanoPhotometerTM (IMPLEN, Germany) [31].

PCR Amplification for DGGE

Community fingerprints were obtained for intestinal microbiota by using total faecal DNA as templates for PCR–DGGE. Primers targeting the variable V3 region of 16S rRNA (corresponding to positions 341–534 of Escherichia coli) were applied and whose nucleotide sequences are as follows: primer-F: 5′-GC clamp-CCT ACG GGA GGC AGC AG-3′, primer-R: 5′-ATT ACC GCG GCT GCT GG-3′. A GC clamp (CGC CCG CCG CGC GCG GCG GGC GGG GCG GGG GCA CGG GGG G) was attached to the 5′ end of the forward primer to allow detection of the corresponding PCR products with DGGE [27].

Each 50 μl PCR reaction mixture contained 20 pmol of each primer, 200 μM of each deoxynucleotide triphosphate (dNTP), 2.5 mM MgCl2, 2 U of Taq DNA polymerase (Promega, USA), 10× buffer and total faecal DNA (2 μl ≅ 120 ng approximately). PCR amplification was performed in an automated thermocycler (ABI2720, USA) using touchdown PCR program: Initial denaturation at 95°C for 5 min, followed by denaturation at 95°C for 1 min, annealing at 65°C for 1 min and extension at 72°C for 1 min. The annealing temperature was decreased by 1°C every second cycle until a touchdown 55°C, at which temperature 10 additional cycles were carried out, followed by final extension at 72°C for 7 min and holding at 4°C [27].

PCR products were quantified by NanoPhotometerTM (IMPLEN, Germany) and subsequently resolved by 2% [w/v] agarose gel electrophoresis (300–400 ng DNA per lane). Bands were visualised using ethidium bromide staining (5 μg/ml).

Denaturing Gradient Gel Electrophoresis

Denaturing gradient gel electrophoresis was conducted using the DCode™ Universal Mutation Detection System (Bio-Rad, Hercules, CA, USA) on 16 cm × 10 cm × 1 mm gels. The sequence-specific separation of the PCR products (the amplicons) was obtained in 10% (w/v) polyacrylamide (acrylamide-bis, 37.5:1) gels in 1 × Tris-acetate EDTA (TAE) buffer, containing 40–75% linear denaturant gradient. The 100% denaturing solution contained 7.0 M urea and 40% (vol/vol) deionised formamide. Electrophoresis was performed at a constant voltage of 100 V at 60°C for 12 h. The gels were stained with 5 µg/ml ethidium bromide solution for 30 min, then washed by deionised water and viewed by using BIO-RAD Gel Doc 2000.

Comparison of DGGE profiles in different gels was performed by employing a standard reference (DNA Marker: DL2000). The gel profiles were digitally normalised by comparison with a standard pattern using the BioNumerics software, version 2.50 (Applied Maths, St.-Martens-Latem, Belgium). All gels were run simultaneously under the same electrophoretic conditions to minimise experimental errors [38].

Statistical Analysis of DGGE Banding Patterns

The bacterial diversity of the diabetic group and the healthy group was evaluated by the number of bands and the band intensity of DGGE profiles using Quantity One software (Bio-Rad, USA). The Shannon–Weaver index of diversity (H′) was used to determine the diversity of taxa present in faecal microbiota from diabetic group and healthy group [11, 18]. The similarity score and cluster analysis of DGGE profiles were performed using the UPGMA method based on the Dice similarity coefficient (band-based) [10, 37]. As the data were nonuniformly distributed, a nonparametric statistical analysis using a Mann–Whitney U test was performed, where a probability value P < 0.05 was interpreted as statistically significant. The nonparametric statistical analysis was performed using SPSS (version 12) (Chicago, IL, USA). Similarities were displayed graphically as a dendrogram. The clustering algorithm used to calculate the dendrograms was an unweighted pair group method with arithmetic averages (UPGMA) [10, 33]. The Shannon–Weaver index was calculated by the following equation:
$$ {\text{Shannon}}{-}{\text{Weaver}}\,{\text{index}}\,(H\prime )\, = \, - \sum\limits_{i = 1}^{s} {(P_{i} )({\text{In}}\,P_{i} )} $$
where s is the number of species/bands in the sample and Pi is the proportion of species/bands for the ith species/band in the sample [31].

Recovery of Bands and Sequencing

DGGE bands (Fig. 1) were cut out from the polyacrylamide gels with a sterile scalpel under UV illumination, re-suspended in 100 μl of Tris–HCl (10 mmol, pH 8.0) and incubated overnight at 4°C. Being heated at 99°C for 30 min, a 4 μl aliquot was subsequently re-amplified by PCR using the original primers without GC clamp. After purification, PCR products were cloned into the p-GEM T Easy vector (Promega, USA) according to the manufacturer’s instructions. Competent E. coli DH5a cells were transformed and screened for plasmid insertions according to the manufacturer’s instructions. Final PCR products were purified using Qiaquick PCR purification kit (Qiagen, Valencia, CA, USA) and were sequenced by Shanghai Sangon Biological Engineering Technology Service Co. Ltd. (Shanghai, China). Sequences were analysed with Chromas v2.23 (Technelysium, Tewantin, Australia) and their similarity was checked against 16S rRNA sequences stored in GenBank and Ribosomal Database Project Release 10 ( by using BLAST and Seqmatch ( software [26].
Fig. 1

Clustering of DGGE profiles obtained with universal primer (V3) of 16 diabetics (D1–16) and 12 healthy group (C1–12) using Dice’s coefficient and UPGMA

Real-Time PCR

Real-time qPCR amplification and detection were performed in a Bio-Rad CFX96 real-time PCR detection system (Bio-Rad Laboratories, USA). Each 20 µl reaction mixture contained 10 µl of 2 × SYBR Green PCR Master Mix (TOYOBO, Osaka, Japan), 1 µl of each primer (5 µM) (Table 1), 2 µl of sample DNA and 6 µl sterilised ultra pure H2O, (20 µl). The amplification program consisted of one cycle of 95°C for 5 min, then 40 cycles of 95°C for 10 s, 55°C for 15 s and 72°C for 50 s, and finally one cycle of 95°C for 15 s. The fluorescent product was detected at the end of each cycle. Following amplification, melting temperature analysis of PCR products was performed to determine the specificity of the PCR. The melting curves were obtained by slow heating at 0.1°C/s increments from 65 to 95°C, with continuous fluorescence collection. For determination of the number of Bacteroides vulgatus, Clostridium leptum subgroup and Bifidobacterium genus present in each sample, fluorescent signals detected from six serial dilutions in the linear range of the assay were averaged and compared to a standard curve generated with standard DNA in the same experiment. Bacteroides vulgatus (CICC 22938), Bifidobacteria longum (CICC 6186) and Clostridium leptum (YIT 6169) were used as standard strains [14, 25].
Table 1

PCR primers used in this study

Target bacterial group


Sequence (5′–3′)

Size (bp)










Bacteroides vulgatus

Bact. F




Bact. R



Clostridium leptum subgroup









Dietary Aspects

Analysis of the participant’s dietary habits indicated similar consumption patterns of grain (rice, wheat), vegetables and a little meat in both groups. The only difference was that the daily amount of food in diabetes patient was limited.

DGGE Profiles Analysis of Diabetic Group and Healthy Group

PCR–DGGE analysis with universal primers targeting the V3 region of the 16S rRNA gene was used to analyze the dominant faecal bacterial microbiota of diabetic and healthy group (Fig. 1 right). The number, position and intensity of the bands were different among samples, which showed the complex fingerprints of intestinal microbiota. In general, most DGGE fingerprints were characterised by the presence of approximately 4–8 dominant bands with background of up to 30 distinct but less intense bands, some common bands (at the same position but in different lane) were also present in different samples.

To analyze the diversity of faecal microbiota in healthy group and diabetic group, the Mann–Whitney U test was used to compare the Shannon–Weaver indexes of diversity (H′) of the bands from DGGE profiles. The results revealed that there was no significant (P > 0.05) difference in diversity between the two groups (Table 2). Analysis of the DGGE fingerprints did not reveal association between community diversity and disease status.
Table 2

Microbiota diversity and similarity of type 2 diabetic and healthy group


Microbiota diversity

Microbiota similiary

DGGE bandsa (mean ± SD)

Shannon indexb (mean H′/H′ max ± SD)




14.2 ± 1.4

0.94 ± 0.037

47.1 ± 8.03

35.4 ± 9.89


13.7 ± 2.7

0.94 ± 0.050

49.7 ± 8.90

P value





Results which are significantly different (Mann–Whitney U test), with P < 0.05

aNumber of denaturing gel electrophoresis (DGGE) bands produced by each sample

bShannon diversity index (H′) as calculated using the relative intensities of all DGGE bands in each sample and expressed as a ratio of H′ to \( H_{\rm max}^\prime \), where \( H_{\rm max}^\prime \) is the maximum value of the Shannon index for a given sample

cDice similarity coefficients comparing DGGE band profiles within individual of a given group

dDice similarity coefficients comparing DGGE band profiles between members of diabetic and healthy group

The similarity level of all sample DGGE profiles was indicated by Dice similarity coefficient and UPGMA dendrogram (Fig. 1 left). The value of the band-based Dice similarity coefficient ranged from 26.8 to 70.1% for diabetic group, and from 28.1 to 66.2% in healthy group, with the mean similarity index of 49.7 ± 8.90% and 47.1 ± 8.03%, respectively (Table 2). When all the samples of both the diabetic group and the healthy group were compared, 192 Dice similarity coefficient values ranged from 7 to 60.2%. The mean similarity index between groups was 35.38 ± 9.89%, which was less than the intra-group similarity index of both groups. From the UPGMA clustering analysis of V3-DGGE (Fig. 1 left), we also found that the similarity of intra-groups similarity of both diabetic group and healthy group was significantly higher than that of the inter-groups (e.g.: D1 and D3, D4 and D13, D6 and D7, D9 and D10, C7 and C8, C4 and C6), which demonstrated statistically that the dominant microbiota of the diabetic group was different from that of the healthy group.

To assess the temporal stability of microbiota composition, two patients were tracked over a 2-week period, and DGGE analysis revealed high similarity of band patterns and host-specificity of intestinal microbiota (Dice coefficient of similarity ranging from 95 to 97%) (Figure not shown).

Sequence Analysis of Selected DGGE Amplicons

In order to identify these dominant bacteria, we selected as much as visible bands in both diabetic and healthy group. Also, bands in the same position but in different lanes were selected to certify the resolution capability of DGGE. Sequencing results of bands were identical, which indicates that some bacteria may have the same bands in the DGGE map.

Being separated by DGGE, the dominant bands of the V3 region in 16S rDNA were sequenced and 45 sequences were obtained. We analysed the sequences through BLAST and RDP.10 and submitted to genebank (Table 3). Most of these 16S rRNA gene sequences were corresponded to uncultured bacteria, thus we selected a taxon above the rank of species. The frequency of bacteria in Table 3 shows that the phylum Bacteroidetes and phylum Firmicutes, especially the phylum Bacteroidetes, were the dominant bacteria in both diabetic group and healthy group. Nevertheless, the difference between the diabetic group and healthy group existed in the distribution ratio of different genera within phylum Bacteroidetes. Sequencing results revealed that diabetic group was comprised of genus Bacteroides (frequency of 53.6%), genus Parabacteroides (10.7%), genus Prevotella (10.7%), genus Bifidobacterium (3.6%), genus Alistipes (7.1%), phylum Firmicutes (10.8%) and phylum Proteobacteria (3.6%). While bands presented in healthy group samples belonged to genus Prevotella (58.8%), genus Bacteroides (11.8%), genus Alistipes (5.9%), phylum Firmicutes (17.7%) and phylum Proteobacteria (5.9%). Bacteroides genus and Parabacteroides genus were most apparent associated with diabetic group. These results suggest that the dominant genera in the intestinal microbiota of diabetic group is different compared with that of healthy group.
Table 3

Sequences of PCR amplicons derived from DGGE gels and identities based on the BLAST database

Type bacteria genus (sequencing result of the bands)

Healthy groupa (n = 17)

Diabetic groupa (n = 28)

GenBank accession number

Phylum Bacteroidetes


 Genus Prevotella




 Genus Bacteroides




 Genus Parabacteroides




 Genus Alistipes




Phylum Actinobacteria 

 Genus Bifidobacterium




Phylum Firmicutes

 Genus Lachnospiraceae Incertae Sedis




 Genus Lactobacillus




 Genus Dialister








 Genus Anaerotruncus 




Phylum Proteobacteria

 Escherichia coli




 Genus Shigella




aRefers to the frequency (and percent) of each unique bacteria genus in healthy or diabetic group

Quantitative PCR Detection of Bacteroides vulgatus, Bifidobacterial, Clostridium leptum Subgroup

Real-time PCR analyses were performed to quantify Bacteroides vulgatus, Clostridium leptum subgroup and Bifidobacterium genus in faecal samples of diabetic group and healthy group. The standard plasmid with six serial dilutions was simultaneously used for each detection. The results indicated that the diabetic group had less copy numbers of Bacteroides vulgatus, Bifidobacterium in the faecal microbiota than healthy group, and the decrease of Bifidobacterium copy number in diabetic group was significant (P < 0.05) when compared with that of healthy group (Table 4). However, Clostridium leptum subgroup in diabetic group was higher in copy number than that in healthy group, though there was no significant difference (P = 0.88) between two groups. The data were presented as the means of triplicate determinations. The standard deviation of means varied by ≤20% (Fig. 2).
Table 4

Bacterial copy numbers in faecal samples analyzed quantitative real-time PCR


Copies/g of faecal microbiotaa

Healthy group (n = 12)

Diabetic group (n = 16)

Mann–Whitney U test (P value)

Bacteroides vulgatus

(1.18 ± 2.76) × 108

(0.36 ± 1.22) × 108


Clostridium leptum subgroup

(2.7 ± 2.05) × 107

(3.82 ± 4.62) × 107



(8.29 ± 1.11) × 105

(1.10 ± 1.13) × 105


Results which are significantly different (Mann–Whitney U test), with P < 0.05

aMeans and standard deviation
Fig. 2

Real-time PCR detection for Bacteroides vulgatus, Clostridium leptum subgroup, Bifidobacterium were presented to bacterial copy numbers in faecal microbiota of diabetic group and healthy group


Our study showed that the faecal microbial composition was different between the diabetic group and healthy group by molecular profiling, cloned sequencing and quantitative PCR (qPCR) analysis. More specifically, there was a significant difference in the similarity of bacterial community and the number of Bifidobacterium, measured by the Dice similarity coefficient and bacterial absolute copy number, in diabetic group compared to healthy group. Results of this study promote basic knowledge about the intestinal bacterial communities in diabetic group.

In this study, the bacterial diversity of the gut microbiota in diabetic group and healthy group was investigated by combining DGGE of the 16S rRNA gene with imaging and cluster analysis and sequencing of key PCR amplicons, together with statistical analyses. DGGE data indicated that the faecal samples were colonised by diverse bacterial communities. The amount of diversity, however, did not correlate with either health or diabetes, which means it is possible that obesity or diabetes are associated with a shift in the balance of the gut microbiota rather than the action of a single microbes or a simple increase in diversity. This may be contributed to understanding that no significant changes were observed in the bacterial diversity of gut microbiota in diabetic group. Ley et al. had analyzed 5088 bacterial 16S rRNA gene sequences from the distal intestinal microbiota of genetically obese ob/ob mice, lean ob/ + mother, all fed the same polysaccharide-rich diet. They found that microbial-community composition was inherited from mothers, and ob/ob animals have a 50% reduction in the abundance of Bacteroidetes and a proportional increase in Firmicutes compared with lean mice and regardless of kinship. Finally the conclusion indicated that obesity affects the diversity of the gut microbiota in this model [19]. However, this change was division-wide and not due to blooms or extinctions of specific bacterial species: bacterial diversity remained constant [20]. Besides, it should be recalled that only dominant populations are detected by using this PCR-based approach, so this is not a measure of the total diversity of the samples, but instead a measure of relative diversity for comparison between samples.

The similarity of faecal DGGE profiles in within-groups was significantly higher than that in between groups according to statistic analysis [32]. In other words, the gut microbiota composition is similar among diabetes individuals, so is the healthy group, and the difference could be clearly observed from DGGE profiling (Fig. 1), which suggests that the diabetic status may influence specific subpopulations of the gut microbial community. Sequence analysis of all dominant DGGE bands enabled the association of specific bacterial genotypes with health or diabetes to be investigated (Table 3), although this depends on the validity of community analysis and the accuracy of the sequence data derived from the technique. It has previously been verified that comigrating bands generally correspond to identical sequence[17, 18]. To verify the accuracy and reliability of the DGGE technique, the test was also detected in this study by excising bands of same migration positions and carried out sequence analysis to ensure that the same identification was obtained. Others have found that to narrow the range of the denaturing gradient may allow separation of such bands [28], and we had optimised the denaturing gradient for separation of co-migrating clones. Consistent with previous studies [12, 19], there is a dominance of the Bacteroidetes and Firmicutes in the distal intestines of diabetic group or healthy group, and the main difference lies in the proportion of genus–division bacteria within Bacteroidetes between diabetic group and healthy group. Prevotella genus was associated with healthy group, while Bacteroides genus was prevalent in diabetic group. The apparent association of Parabacteroides genus with diabetic group is noteworthy. Nevertheless, disease-related bacteria have not been found more, which probably attributes to that the relatively short sequences derived from DGGE did not generate a database match and presented uncultured bacterium clone, which affected bacterial determination. Even so, some significant basic characteristics of gut microbiota were obtained by combining similarity analysis with dominant PCR amplicons sequencing.

The diversity of the microbiota results in complex DGGE patterns with many bands in close proximity. Only organisms present in relatively high concentrations (~108 CFU/g stool) are represented on the gel [28]. Assessment of bacterial diversity using DGGE may be biased by variability in the efficiency of amplification of different sequences. However, the strength of DGGE will theoretically detect the presence and identity of any amplifiable target sequence above the detection threshold [28]. Despite these concerns, DGGE is currently one of the few techniques that allow reproducible visual comparisons of profiles from microbial communities to be derived and has been successfully applied to a wide variety of microbial ecosystems [18, 30, 34, 39].

It is important to note that DGGE is not a truly quantitative technique and that band density does not necessarily relate to target abundance. It is therefore possible that subtle associations between species abundance and diseases would not necessarily be identified [18]. Thus, we used quantitative detection to observe the changes of mircobiota. Real-time qPCR with species-specific primers can provide a precise and sensitive method for more accurate quantification of individual species as well as total bacteria, and will be a useful tool for studies on the ecology of gut microbiota [21].Takahiro et al. used 16S rRNA gene-targeted group-specific (or species-specific) primers to analyze the predominant bacteria in human faeces, and the result showed the distribution of the six bacterium in human gut microbiota and absolute quantitative of copy number of bacteria [23, 25]. Mariat et al. reported the comparative assessment of human faecal microbiota from three different age groups by qPCR, and demonstrated that the faecal microbiota composition evolves throughout life from early childhood to old age [22]. In this study, the remarkable result is that the level of Bifidobacterium spp. has been visibly reduced in diabetic group. Clostridium leptum subgroup and Bacteroides vulgatus increased and decreased, respectively, but there is no significant difference. Our results are consistent with the previous reports [5, 13], and the reduction of beneficial bacterium such as Bifidobacterium is one characteristic feature of diabetes. We suppose that diabetes status induced by high-fat feeding may alter the intestinal microbiota composition where Bifidobacterium spp. were reduced. Thus, it would be useful to develop specific strategies for modifying gut microbiota to favour specific gut microbiota (i.e. Bifidobacteria) to prevent the deleterious effect of high-fat or obesity-induced diabetes [7].

In order to observe the characteristics of gut microbiota in diseased conditions, we must exclude the influences which physiological factors have brought, such as age, gender, diet and genetic background. Thus, we attempted to minimise such uncorrelated variables, as much as possible, where diabetic and healthy group of the same age grouping, gender, with similar dietary habit and living in the same environment, were chosen. However, diets were not limited strictly between diabetic group and healthy group as diabetes are closely related with diet. The changes in eating habits have become a risk factor to type 2 diabetes. The previous data suggested hypothesis that high-fat feeding is associated with changes of gut microbiota and metabolic endotoxemia, and further metabolic endotoxemia involves in the development of type 2 diabetes and obesity in humans [8]. In brief, daily eating habits, which used to be of less concern to people, are possibly the primary cause of metabolic diseases such as diabetes, and intestinal microbiota is a part of this pathologic process.

The physiological influence of composition and diversity of the human gut microbiota on the human host requires more investigation to increase understanding of its functional role in normal and disordered states. In addition, because we can hardly collect serial samples from patients with diabetes before and after suffering, there is a one-time collection in this study and the microbiota-disease connection remains a query: it is not clear if microbial shifts actually cause disease or if they are simply a reflection of a diseased state [29]. Thus, more studies are needed to determine the succession process of gut microbiota correlated with the development of diabetes and the changes in diet habit and obtain a more detailed understanding of the role of the microbial population of the gut during diabetes. Besides, further study should be done to compare gut microbiota between poorly controlled diabetics and well-controlled diabetics according to blood sugar levels, trying to analyse the influences of the different severity of diabetes on gut microbiota.


This study shows that high levels of interindividual variation of gut bacterial communities are observed with no association between microbial diversity and diabetes. However, clear association between faecal microbiota composition and diabetes emerge during DGGE analysis. Through real-time PCR analysis for Bacteroides vulgatus, Bifidobacterial genus, Clostridium leptum subgroup, the result illustrates these three bacteria have undergone changes of different degree in diabetic group, and the copy number of Bifidobacterial genus is significantly declining.


This work was supported by the Ministry of Science and Technology, The People’s Republic of China with a 973 project (2004CB418503). We gratefully acknowledge Xuan Xie and Yang Jiao from Department of Endocrinology, the Second Affiliated Hospital, School of Medicine of Xi’an Jiaotong University for sample collection and the Center for Disease Control and Prevention of Xi’an, Shaanxi Province, for equipments and technical support. We gratefully acknowledge Ivan Yuan from St. Catherine’s College, University of Cambridge, UK for careful revision of the manuscript.

Copyright information

© Springer Science+Business Media, LLC 2010