• Users Online: 139
  • Home
  • Print this page
  • Email this page
Home About us Editorial board Ahead of print Current issue Search Archives Submit article Instructions Contacts Login 


 
 Table of Contents  
ORIGINAL ARTICLE
Year : 2017  |  Volume : 2  |  Issue : 2  |  Page : 45-54

Genetically and dietary induced obesity associate differently with gut microbiota in a murine intestinal tumorigenesis model


1 Department of Chemistry, Biotechnology and Food Sciences, Norwegian University of Life Sciences, NO-1432 1432 Ås, Norway
2 Department of Toxicology and Risk Assessment, Division for Infection Control and Environmental Health, Norwegian Institute of Public Health, NO-0403 Oslo, Norway

Date of Submission11-Mar-2017
Date of Acceptance26-May-2017
Date of Web Publication30-Jun-2017

Correspondence Address:
Inger-Lise Steffensen
Department of Toxicology and Risk Assessment, Division for Infection Control and Environmental Health, Norwegian Institute of Public Health, PO Box 4404 Nydalen, NO-0403 Oslo
Norway
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/ed.ed_5_17

Rights and Permissions
  Abstract 

Background: Overweight and obesity are risk factors for human colorectal cancer. Growing evidence suggests that the gut microbiome affects both obesity and cancer. In this study, we examined how the murine microbiota composition correlated with obesity, intestinal tumorigenesis, glucose regulation, and inflammation.
Materials and Methods: We used 16S ribosomal RNA gene analyses of feces and data obtained from a double-mutant mouse model; multiple intestinal neoplasia (Min), mice, which spontaneously develop intestinal tumors, crossed with obesity (ob), mice, which develop obesity, fed 10% or 45% fat diet.
Results: We found that diet and genotypes imposed a major impact on the gut microbiota composition. Likewise, we found strong associations of the microbiota composition with obesity, number of small intestinal tumors, and blood glucose levels. Specifically, bacteria related to Clostridium perfringens and Lactobacillus showed strong positive associations with both dietary induced and genetically induced obesity, while Bacteroidales showed strong negative associations. Representatives of Lachnospiraceae and Peptostreptococcaceae only showed significant negative associations with genetically induced obesity and no associations with dietary induced obesity.
Conclusions: We found complex associations between the microbiota and genetic background, diet, obesity, glucose levels, inflammation, and intestinal tumorigenesis. This could contribute to the lack of consensus between the results in previous studies regarding correlations of microbiota with obesity and cancer.

Keywords: Blood glucose levels, diet-induced obesity, inflammation, intestinal tumorigenesis, microbiota, multiple intestinal neoplasia mouse, obese mouse, obesity


How to cite this article:
Rudi K, Ludvigsen J, Dirven H, Steffensen IL. Genetically and dietary induced obesity associate differently with gut microbiota in a murine intestinal tumorigenesis model. Environ Dis 2017;2:45-54

How to cite this URL:
Rudi K, Ludvigsen J, Dirven H, Steffensen IL. Genetically and dietary induced obesity associate differently with gut microbiota in a murine intestinal tumorigenesis model. Environ Dis [serial online] 2017 [cited 2022 Jan 20];2:45-54. Available from: http://www.environmentmed.org/text.asp?2017/2/2/45/209265


  Introduction Top


Overweight and obesity are risk factors for human colorectal cancer,[1],[2] but the causal relationship is still controversial.[3],[4] There is growing evidence suggesting that the gut microbiota is established early in life and affects not only the function of the gut but also systemic health throughout adult life.[5],[6],[7] Diet-related variations in the gut microbiota have been linked to a variety of noncommunicable chronic diseases, including colorectal cancer, obesity, diabetes, gut inflammatory and autoimmune conditions, central nervous system function, and cardiovascular disease.[8],[9],[10],[11],[12] However, the causal relationship between diet, gut microbiota, and human diseases such as obesity or cancer is still poorly understood.

Obesity alters the physiology of the whole organism, and therefore, animal models are useful to study the effects of increased adiposity. In addition, models that integrate lifestyle factors such as diet and genetic factors in a single model system provide a physiologically intact system valuable for studies of complex relationships. In a previous study,[13] we examined the relationship between obesity and intestinal tumorigenesis in a double-mutant mouse model obtained by crossing the multiple intestinal neoplasia (Min) mice, which spontaneously develop intestinal tumors, and the obesity (ob) mice, which develop obesity. In this experiment, both genetic- and diet-induced obesity (DIO) was compared in mice within the same experiment, and for DIO, comparable high- and low-fat diets were used. All the littermates with the various genotypes served as each other's internal controls. We found that ob/ob mice had significantly increased body weight and number of spontaneous small intestinal tumors (in Min/+ mice) versus ob/wt and wt/wt mice. A 45% fat diet further increased body weight and spontaneous tumor numbers versus 10% fat. We also examined disturbed glucose regulation and inflammation as mechanisms involved in the association between obesity and intestinal tumorigenesis. The ob/ob mice had hyperglucosemia and insulinemia compared with ob/wt and wt/wt mice. A 45% fat diet further increased glucose but not insulin. Increased tumor necrosis factor-α (TNFα) levels, indicating inflammation, were also seen in ob/ob mice.

The murine gut microbiota model has been used to uncover basic mechanistic insight about the association between the gut microbiota and the host in health and disease. However, although major discoveries have been made in these fields, knowledge about how factors such as genetic background, gender, and diet contribute to shaping the gut microbiota is still limited. In addition, potential confounding factors such as cage and litter relationships should be taken into account in the interpretation of such studies.

The aim of this work was to examine how the experimental variables such as intestinal tumorigenesis, obesity and mechanistic data on blood glucose regulation and inflammation from a mouse experiment comprising different genotypes, diets, gender and litters correlated with the murine microbiota composition. We used fecal material from the colons and data from our previous study.[13] The gut microbiota composition was determined by Illumina 16S ribosomal RNA gene sequencing of the V3 and V4 regions.[14] A high-resolution approach for error correction and taxonomic assignment was used,[15] to account for the distinctness of the murine microbiota.[16]


  Materials and Methods Top


Mice

The mice used in this study were double mutants obtained by crossing through two generations the Min mice, which spontaneously develop intestinal tumors (adenomas), and the ob mice, which develop obesity. The results obtained from these mice on intestinal tumorigenesis and body weight, as well as on the two main hypotheses for the relationship between intestinal tumorigenesis and obesity, i.e., disturbed blood glucose regulation or increased inflammation, were recently published.[13] From this experiment, 6 to 16 mice from 6 to 14 different litters per experimental group, minimizing maternal effects, were included in this study [Table 1]. The breeding, housing, and termination of the mice were described in detail.[13] The animal experiment was performed in strict accordance with the law and regulations for animal experiments in Norway. The protocols were approved by the Norwegian Animal Research Authority (permit numbers 1357 and 4856). Cardiac puncture and cervical dislocation were performed under ZRF cocktail anesthesia, before dissection of organs after death. Every effort was made to minimize suffering. The littermates of all genotypes of the same gender were housed together after weaning, minimizing potential cage effects.
Table 1: Genotypes, gender, diets, number of mice and number of litters per experimental group

Click here to view


The C57BL/6J-ApcMin/+ mouse is a well-established model for sporadic colorectal cancer in humans and the inherited disorder familial adenomatous polyposis.[17],[18] It is heterozygous for the germ line nonsense Min mutation in the tumor suppressor gene Adenomatous polyposis coli (Apc) leading to a truncated nonfunctional APCprotein and therefore develops numerous spontaneous intestinal adenomas in the small intestine and, to a much lesser degree, in the colon.[19],[20]Apc is a key component in the Wingless-related integration site signaling pathway.[21],[22] The C57BL/6J-ob/+ mouse is heterozygous for the obese (ob) mutation in the leptin (lep) gene and becomes obese when having a homozygous mutation (ob/ob) and therefore lacking functional leptin hormone.[23],[24],[25] Leptin regulates food intake and energy expenditure and has effects on immune functions, including inflammation and reproduction.[25],[26]

In the present study, a subset of the mice from our previous study [13] was used. Four different double genotypes (Apc+/+ × Lepwt/wt, Apc+/+ × Lepob/ob, ApcMin/+ × Lepwt/wt, or ApcMin/+ × Lepob/ob) were included, being genotyped as previously described for both Min[27] and ob[13] status by allele-specific polymerase chain reaction (PCR) analysis. To separate the wild-type (normal, nonmutated) allele of the ob gene from the wild-type allele of the Apc gene, throughout this paper the wild-type allele of the ob gene is designated “wt” whereas the wild-type allele of Apc is designated “+.” Both genders were included. By studying obesity induced genetically in the ob mouse, we could examine the effect of obesity independent of diet. In addition, we studied environmentally induced obesity, i.e., DIO, by giving the mice comparable diets, either a high-fat diet or a low-fat diet from weaning and until termination at 11 weeks. To make sure the effects observed were due to higher fat at the expense of carbohydrates only and not decrease in protein, vitamins, or minerals, commercial diets were chosen that added fat as % Atwater fuel energy (AFE) with isocaloric exchange with carbohydrate. The high-fat diet used was 45% AFE fat diet, code 824053 (45% kcal from fat, 20% kcal from proteins, 35% kcal from carbohydrate, and 4.54 kcal AFE/g), for comparison with the normal-fat diet 10% AFE fat diet, code 824050 (10% kcal from fat, 20% kcal from proteins, 70% kcal from carbohydrate, and 3.68 kcal AFE/g), both from SDS Special Diets Services (Essex, UK). Feed intake for the experimental groups was previously reported.[13] All mice had been given a single subcutaneous injection of 0.9% NaCl on day 3 to 6 after birth as vehicle control to carcinogen-treated mice not included in this work. Thus, in this paper, only spontaneous intestinal tumors were included. The microflora composition was examined in fecal material collected from the colons of the mice terminated at 11 weeks of age and stored at −80°C until analysis.

Experimental data variables

In this work, we have examined how intestinal tumorigenesis, body weight, blood glucose levels, and plasma levels of insulin and the pro-inflammatory cytokines interleukin-6 (IL-6) and TNFα correlated with microbiota composition. Further, we have studied the influence of mouse genotypes (Apc and ob), diet (10% fat diet or 45% fat diet), gender, and litter on the composition of gut microbiota.

The number and diameter of tumors in the small intestine and colon were scored by transillumination in an inverse light microscope after fixation in 10% neutral buffered formalin and staining with 0.2% methylene blue (Sigma-Aldrich Norway AS, Oslo, Norway). Body weight was evaluated in three ways; as area under the curve (AUC) for body weight recorded from 3 to 11 weeks of age, calculated by the macro in SigmaPlot 12.3 (Systat Software Inc., San Jose, CA, USA), as terminal body weight or terminal body mass index (BMI) at 11 weeks. Nonfasted blood glucose levels were measured with a glucometer FreeStyle Freedom Lite (Abbott Diabetes Care Inc., Alameda, CA, USA) at both age 6 and 11 weeks. Insulin was measured by ELISA from MyBioSource Inc. (San Diego, CA, USA) and IL-6 and TNFα by BD Cytometric Bead Arrays from BD Biosciences (San Jose, CA, USA) in a flow cytometer, all from plasma obtained at termination at 11 weeks of age. For further details on these experimental data variables, please see our previous paper.[13]

DNA extraction, library construction, and sequencing

DNA from 1 or 2 fecal pellets per mice (n = 183) was extracted using a combined mechanical and chemical lysis/extraction method. Samples were mechanically lysed in 2 mL plastic tubes (Sarstedt AG and Co., Nümbrecht, Germany) with ~0.25 g <106 μM acid-washed glass beads (Sigma-Aldrich Chemie GmbH, Schnelldorf, Germany) and 400 μL S.T.A.R. buffer (Roche Diagnostics, Basel, Switzerland) using MagNALyzer (Roche) with 2 × 20 sec at 6500 rpm and 1 min rest at 4°C between runs. The MagLGC™ Total Nucleic Acid Isolation kit (LGC Genomic, Berlin, Germany) for blood samples was used in combination with the KingFisher Flex robot (Thermo Fisher Scientific, MA, USA) for an automated DNA extraction protocol. Extracted DNA was quantified with Qubit dsDNA HS assay kit (Thermo Fisher Scientific).

The microflora composition was determined by Illumina sequencing of the V3–V4 regions of the 16S ribosomal RNA gene using the primer pair PRK341F/PRK806R (amplicon = 466bp)[28] in a nested approach with the same primers but modified by addition of Illumina-specific adapters (TruSeq LT barcodes). Each PCR reaction contained: 1 x HOT FIREPol PCR mix (Solis BioDyne, Tartu, Estonia); 200 nM forward and reverse primers, uniquely tagged for the nested approach; 1 μL of sample DNA and water. The thermal cycling conditions were 95°C for 15 min and 25 cycles (10 cycles in the nested approach) of 95°C for 30 s, 50°C for 1 min, and 72°C for 45 s. The PCR products were cleaned with Agencourt AMPure XP (Beckman Coulter, CA, USA) between PCRs using one:one ratio. Samples were normalized, pooled, and diluted to a concentration of 4 nM; then, the sequencing setup was performed according to Illumina Metagenomic 16S sequencing setup recommendations. Sequencing was performed on the Illumina MiSeq (Illumina, San Diego, CA, USA) using Version 3 sequencing chemistry with 300 bp paired-end reads. Illumina 16S ribosomal RNA gene amplicon data were quality filtered using USEARCH 8,[15] and sequences were clustered into operational taxonomic units (OTUs) with 97% similarity threshold using the UPARSE pipeline.[15] Taxonomic assignment against the Greengenes 13.8 database [29] was performed using QIIME 1.8[30] (with UCLUST default parameters), and alpha- and beta-diversity measures were analyzed at 4000 sequences per sample.

Statistical analyses

The effects of genetic background and diet on the microbiota were tested using ANOVA-simultaneous component analysis (ASCA).[31] The direct associations between the experimental data variables, i.e., number and diameter of intestinal tumors, body weight, blood glucose levels, insulin and cytokine levels, and the microbiota OTUs, were analyzed with Kruskal–Wallis Analysis of Variance on Ranks on Ranks (Kruskal–Wallis test). The correlations between the experimental data variables and the microbiota were measured with Spearman rank-order correlation (r). The software used was MATLAB version R2015a (The MathWorks Inc., Natick, MA, USA). P < 0.05 was considered statistically significant.


  Results Top


General library and sequence characteristics

The total number of sequencing reads passing the Illumina quality control filter was 8.89 million, with a total sequence yield of 6.3 G bp with 80.3% of the positions above Q30. The samples were rarified to 4000 as a tradeoff between number of samples and sequences.

Phylum-level associations and diversity

At the phylum level, there was an overall dominance of Clostridia, in addition to high levels of Bacteroidia, Erysipelotrichi, and Bacilli for all experimental variables. Diet was the only variable that showed a pronounced effect at the phylum level, with Bacteroidetes showing a positive association with the 10% fat diet (median of 25.4%), as opposed to the 45% fat diet (median of 16.6%) (P = 0.015, Kruskal–Wallis test).

Beta-diversity analyses using Unifrac metrics supported that diet was the main variable contributing to differences in microbiota composition [Figure 1]a. Rarefaction analyses for species richness showed that the 45% fat diet led to a reduced richness compared to the 10% fat diet [Figure 1]b. Simpson's 1-D also showed a decreased alpha-diversity (1-D = 0.932) of microbiota composition for the 45% fat diet compared with 1-D = 0.945 for the 10% fat diet (P = 0.001, Kruskal–Wallis test).
Figure 1: Beta- (a) and alpha- (b) diversity with respect to diet. (a) Beta-diversity with respect to diet, shown by weighted Unifrac analyses colored by diet. (b) Alpha-diversity presented as rarefaction curve for observed species with respect to diet

Click here to view


Associations between the experimental data variables and the microbiota

We directly associated the experimental data variables obtained previously with the OTUs using ASCA analyses. These analyses showed that all the variables had significant effects on the microbiota composition, with the ob genotype and diet being the most significant [Figure 2]. We also found significant interactions of the experimental data variables, where the ob genotype both showed a significant association with gender (P = 0.0018, ASCA) and diet (P = 0.0007, ASCA). The statistically most significant interaction, however, was between gender and diet (P = 0.0001, ASCA). All the interactions are summarized in [Table 2].
Figure 2: Operational taxonomic unit-level association of design variables (a) Min genotype, (b) ob genotype, (c) gender and (d) diet with the gut microbiota. The associations were investigated using ANOVA-simultaneous component analysis. The box plots in panels a-d represent the ANOVA-simultaneous component analysis score value. Significant P values are marked in italics. The box plots show the median, and the first (25%) and third (75%) quartile. The whiskers span the values that are not considered outliers

Click here to view
Table 2: Interactions in the ANOVA-simultaneous component analyses

Click here to view


With respect to the importance of OTUs, the pattern was complex, with closely related OTUs having opposite associations [Figure 3]. OTU 8 being classified as Clostridium perfringens showed the strongest associations with both the Apc Min/+ genotype (median of 3.1%) as opposed to 1.6% for the Apc wild-type (+/+) (P = 0.002, Kruskal–Wallis test) and the ob/ob genotype (median of 3.7%) as opposed to 1.3% for the ob wild-type (wt/wt) (P < 0.0005, Kruskal–Wallis test). OTU 7 classified as Clostridiales showed the strongest associations both to the Min/+ matched wild-types (+/+) (P = 0.023, Kruskal–Wallis test) and the ob/ob matched wild-types (wt/wt) (P < 0.0005, Kruskal–Wallis test), with respective medians of 4.5% and 5.2% for the wild-types and 3.6% and 2.9% for the mutant genotypes. For gender, OTU 1 and OTU 3, both belonging to the genus Allobaculum, showed the strongest but still not significant (P > 0.05, Kruskal–Wallis test) association with males for each of these OTUs alone, while the sum of these two OTUs showed a highly significant association (P < 0.0005, Kruskal–Wallis test), with a median of 16.7% for males and 10.0% for females. For the diet, OTU 2 classified as Lactobacillus sp. showed a strong association with the 45% fat diet with a median percentage of 8.2%, as opposed to 3.3% for the 10% fat diet (P < 0.0005, Kruskal–Wallis test). The sum of the two Allobaculum OTUs (OUT 1 and 3) showed a strong association with the 10% fat diet, with a median of 15.1% for the 10% fat diet and 8.6% for the 45% fat diet (P < 0.0005, Kruskal–Wallis test).
Figure 3: Phylogeny and experimental associations of operational taxonomic units. A neighbor-joining phylogenetic tree was constructed based on signature sequences for each of the operational taxonomic units. The color code for the inner circles represents the importance of the respective variables

Click here to view


The OTU 2 association with 45% fat diet was dependent on gender. For females, the 45% fat diet showed a level of 14.1%, while the 10% fat diet showed a level of 2.4% (P < 0.0005, Kruskal–Wallis test), while there were no 45% fat diet associations for males. Allobaculum (sum of OTU 1 and OTU 3) also showed a gender dependence for the diet association. Allobaculum was reduced with a median of 6.1% for females on a 45% fat diet compared with males on a 45% fat diet or both genders on a 10% fat diet, which all had medians of approximately 15% (P < 0.0005, Kruskal–Wallis test).

There was also diet-dependence for the OTU 8 association with Min/+ mice. The median for the 45% fat diet was 7.6%, while it was 1.3% for the 10% fat diet (P < 0.0005, Kruskal–Wallis test). Finally, there was also diet-dependence for the OTU 8 association with ob/ob mice. The 45% fat diet showed a median of 7.9%, while the 10% fat diet showed a median of 3.1% (P = 0.02, Kruskal–Wallis test).

Effect of family association

There were in total 24 sibling pairs across the experimental groups in our data. Siblings showed a higher correlation in microbiota composition (median Pearson ρ = 0.73) compared to nonsiblings (median Pearson ρ = 0.60) (P < 0.0005, Kruskal–Wallis test). Removing the siblings, however, had no effect on the associations between the experimental variables and the microbiota (Pearson ρ > 0.96 for pairwise comparisons of loadings between OTU level models with and without siblings). The P values, however, were slightly higher for the nonsibling model, probably due to fewer samples in this model.

Correlation of microbiota with intestinal tumorigenesis, body weight, and other experimental data

We determined the correlation of the previously obtained experimental data variables; number and diameter of small intestinal and colonic tumors, body weight, levels of blood glucose at week 6 or 11, levels of insulin, and inflammatory cytokines IL-6 and TNFα at termination [13] with the microbiota composition. This was done by investigating the direct Spearman rank-order correlation between the data variables and the OTUs. Overall, we found 555 significant correlations (P < 0.05) [Supplementary Table S1 [Additional file 1] ]. The OTU showing the largest number of significant correlations (10 out of 12 experimental variables) was OTU 31 classified as Desulfovibrionaceae, being negatively correlated with all experimental variables except IL-6 and colonic tumor size.

Hierarchical clustering of experimental data variable-OTU correlations showed three main clusters [Figure 4]. Cluster I consisting of blood glucose levels at week 6 and week 11, body weight (measured in three ways), and number of small intestinal tumors showed the overall strongest associations with the microbiota, with four OTUs (OTU 8 C. perfringens, OTU 370 Clostridiales, OTU 16 Lactobacillus sp., and OTU 62 Bacteroidales S24-7) showing strong positive correlations (Spearman ρ > 0.3, P < 0.05) and four OTUs (OTU 7 Clostridiales, OTU 65 Lachnospiraceae, OTU 40 Bacteroidales S24-7, and OTU 142 Peptococcaceae) with strong negative correlations (Spearman ρ < −0.3, P < 0.05). OTUs within Firmicutes and Bacteroidetes both showed positive and negative correlations, even with OTUs within the same Bacteroidetes family S24-7 showing opposite directions (OTU 62 positive and OTU 40 negative). Cluster II comprising levels of IL-6 and TNFα and number of colonic tumors only showed one main, positive, correlation (Spearman ρ > 0.3, P < 0.05) to OTU 239 belonging to the Clostridiales. Finally, Cluster III comprising insulin levels and colonic and small intestinal tumor diameters showed positive correlation with OTU 17 within the Bacteroidales and negative correlation with OTU 31 belonging to the Desulfovibrionaceae. Unexpectedly, glucose levels and insulin levels were associated with different clusters of OTUs, indicating separate influence of microbiota on these two related factors. This is consistent with the finding that whereas genetically induced obesity caused both hyperglucosemia and insulinemia in ob/ob mice, a 45% fat diet further increased glucose but not insulin [Supplementary Table 1 [Additional file 2] ].[13]
Figure 4: Clustering of measured variables based on correlations with the microbiota. The clustering was based on Spearman correlation between operational taxonomic units and measured variables. The bars are placed on the branches they describe the relationships for (with arbitrary horizontal locations) and their heights represent average ρ values for the Spearman correlation coefficient between 0.3 and 0.4 with P < 0.05 [Supplementary Table 1]

Click here to view


Separate comparisons of dietary induced and genetically induced obesity associations

Since increased body weight, in addition to blood glucose levels and number of small intestinal tumors, showed the largest number of associations with OTUs (four positive and four negative), we investigated whether these OTUs were directly correlated with diet and/or genotype. The separate effects of a high-fat diet and obesity as such on the microbiota are unclear. A separate comparison of the OTUs was performed between ApcMin/+ × Lepwt/wt and ApcMin/+ × Lepob/ob mice on a 10% fat diet (measuring the separate effects of genetically induced obesity) and between ApcMin/+ × Lepwt/wt mice on either 10% or 45% fat diets (measuring the separate effects of high-fat diet) [Figure 5]. In addition, comparison of the OTUs was performed between ApcMin/+ × Lepob/ob mice on either 10% or 45% fat diet (measuring the combined effects of genetically induced obesity and a high-fat diet).
Figure 5: Associations between the microbiota operational taxonomic units and obesity. Bacteria with positive correlations (upper row) and negative correlations (lower row) to (a) dietary induced obesity (45% vs. 10% fat diets) for Lepwt/wt mice, (b) genetically induced obesity (genotype Lepob/ob vs. Lepwt/wt) on 10% fat diet, and (c) with combined genetically- and dietary induced obesity (Lepob/ob on 45% and 10% fat diets) are shown. Statistically significant P values are shown in italic. The box plots show the median, and the first (25%) and third (75%) quartile. The whiskers span the values that are not considered outliers

Click here to view


For the OTUs positively correlating with increased body weight and increased blood glucose levels (Cluster I), we found that OTU 8 C. perfringens was positively associated with obesity irrespective of whether the obesity was genetically or dietary induced as well as with their interaction. OTU 16 Lactobacillus was significantly associated with both dietary and genetically induced obesity while not with the interaction between genetically and dietary induced obesity. OTU 370 Clostridiales was only associated with dietary induced obesity and OTU 62 Bacteroidales S24-7 only with genetically induced obesity [Figure 5]. For the negatively correlating OTUs, OTU 40 Bacteroidales S24-7 showed a strong association irrespective of induction, while OTU 65 Lachnospiraceae and OTU 142 Peptococcaceae only showed significant associations with genetically induced obesity, and OTU 7 Clostridiales was only significantly associated with the combined dietary and genetically induced obesity [Figure 5].

The OTUs not directly correlated with obesity were also affected by diet. Interestingly, both OTUs associated with small intestinal and colonic tumor diameter and insulin levels (Cluster III) [Figure 4] were both promoted by low-fat diet; OTU 17 Bacteroidales S24-7 with median 10% fat diet 1.3% and median 45% fat diet 0.88% (P = 0.001, Kruskal–Wallis test) and OTU 31 Desulfovibrionaceae with median 10% fat diet 0.73% and 45% fat diet 0.54% (P = 0.002, Kruskal–Wallis test).


  Discussion Top


In concordance with previous studies,[9],[32] we found strong effects on the microbiota composition of diet as well as of genetic background (Apc, ob). In addition, we found strong correlations of the microbiota composition with both intestinal tumorigenesis and body weight. OTU 8 related to C. perfringens was strongly associated with the 45% fat diet, in addition to both the Min/+ and ob/ob genotypes. This OTU also showed strong correlations with small intestinal tumor numbers, body weight, and blood glucose levels. C. perfringens belongs to Clostridium Cluster I, which is associated with several pathogenic and toxigenic Clostridia.[33] From these considerations, a pathogenic phenotype of OTU 8 could be suggested. However, OTU 7 showing opposite relations to OTU 8 belongs to Clostridium Cluster XI, which is also a pathogen-associated cluster, with C. difficile as one of the members. Due to the enormous diversity in Clostridia, it is difficult to deduce functional characteristics based on taxonomic relatedness. This notion is also shown for the Bacteroides, with family S24-7, containing OTUs with strong opposite correlations with i.a. number of small intestinal tumors, body weight, and blood glucose levels.

Specific foods as well as certain dietary patterns can contribute to the development of obesity as well as to the composition of the gut microbiota.[32],[34] A dominant role of diet has been found in murine models in shaping interindividual variations in host-associated microbial communities in murine models.[35] Individuals in the general population with a low gene count (LGC) in metagenomic analysis of the microbiota have a domination of Bacteroides and a higher incidence of obesity and metabolic syndrome compared to high gene count (HGC) individuals.[36],[37] In our data, we found an association between high-fat diet and reduced diversity, but this reduced diversity was due to increase in Firmicutes rather than Bacteroidetes. Some studies, but not all, have found an alteration in the ratio of Firmicutes to Bacteroidetes both in mice and humans, as well as decreased numbers of Akkermansia muciniphila, associated with either dietary or genetically induced obesity.[10],[12] The obese ob/ob mice had a 50% reduction in levels of Bacteroidetes and a proportional increase in Firmicutes.[38] In a study similar to our study, Pfalzer et al. examined the effects on the microbiota in the colon and cecum contents of another mouse model for intestinal tumorigenesis (Apc1638N) and genetically induced obesity (Leprdb/db) and of a 60% fat diet versus a 10% fat diet.[39] They found that taxa enriched after the high-fat diet were mostly Firmicutes (class Clostridia), while those enriched in the genetically induced obese db/db mice, having obesity caused by mutated leptin receptors, were split between Firmicutes and Proteobacteria (class Gamma Proteobacteria).[39] Therefore, the high-fat diet and the genetically induced obesity altered the microbiota differently. Our study, however, showed a more complex pattern, with even closely related OTUs responding differently to dietary induced and genetically induced obesity. OTUs related to both C. perfringens and Lactobacillus showed the strongest positive associations with both genetically and dietary induced obesity. Others have also found a correlation between certain Lactobacillus species and BMI in humans.[40]

It is hypothesized that obesity- or diet-associated changes in the gut microbiota composition may facilitate increased fermentation of otherwise indigestible substrates, which result in increased energy harvest from the diet as well as production of metabolites that may affect the metabolic phenotype and the disease risk of the host.[11],[12] In addition, these changes may impair immune function and increase gut permeability, thereby increasing bacterial translocation, which further induce metabolic changes in peripheral tissues such as increased lipogenesis, gluconeogenesis, adipogenesis, inflammation, and insulin resistance.[11],[12] However, there is still uncertain whether changes in composition of gut microbiota are a direct cause of obesity and other metabolic disorders or whether the changes in the microbiota composition are an adaptation to the changes in the host's diet. The reduced diversity observed with the high-fat diet, however, may indicate a dietary induced change in the composition of gut microbiota. A Western diet high in fat and carbohydrates was previously associated with marked reduced diversity in mice.[41]

The gut microbiota may influence susceptibility to cancer and its progression by mechanisms such as modulation of inflammation or influence on the genomic stability of the host's cells. Other potential mechanisms are production of metabolites that function as histone deacetylase inhibitors to epigenetically regulate host gene expression or production of carcinogenic metabolites from dietary components, affecting the production of short-chain fatty acids or affecting the immune system.[8],[12] The bacteria species Fusobacterium nucleatum has been associated with colorectal cancer both in humans and mice, and although it is not known for certain that the relationship is causal, there are plausible mechanisms involving inflammation and upregulated oncogenes.[10] A depletion of the anti-inflammatory species Parabacteroides distasonis was found in Apc1638N mice with small intestinal tumors.[39] In the K-rasG12Dint mouse, a high-fat diet promoted tumor progression in the small intestine independent on obesity.[42] In our data, small intestinal tumor size was also negatively associated with high-fat diet, possibly because of induction of new tumors and therefore smaller mean tumor size, while we found that the low-fat diet promoted OTU 17 Bacteroidales S24-7, which is associated with the size of both colonic and small intestinal tumors. Low-fat diet, however, also promoted OTU 31 Desulfovibrionaceae, which has an opposite association with tumor size and insulin levels to that of OTU 17. This may indicate that the tumorigenic outcome of dietary intervention can partly be dependent on the individual gut microbiota.

Sulfation of mucin has been associated with a healthy state of the human gut.[43] Therefore, our observation of a negative correlation of OTU 31 Desulfovibrionaceae with the numbers and size of intestinal tumors was surprising. It would be expected that these bacteria could potentially have an indirect increasing effect on intestinal tumorigenesis, and therefore, a positive correlation, through their reduction of levels of sulfated mucin, which normally is protecting the mucosa from damage from luminal substances leading to inflammation. However, this relationship may be dependent of the overall levels of Desulfovibrionaceae, which were rather low in all our samples. In addition, it has also been shown that hydrogen sulfide may have anti-inflammatory and anticancer effects.[44] Therefore, it could also be that strains within the Desulfovibrionaceae, through their production of hydrogen sulfide, could have protective effects toward cancer.

Regarding the relevance of the murine microbiota to human microbiota, there are obvious differences between mice and humans in anatomy of the intestinal tract, genetic variation, and influence of external environmental factors, such as housing, maternal microbiota, stress, diet, or antibiotic use, that may influence the microbiota. For instance, diet could explain 81% of the intestinal microbiota composition in mice, whereas in humans, this is thought to be less than 50%.[11] However, both the human and murine gut microbiota are dominated by two major phyla, Bacteroidetes and Firmicutes, whereas there are differences in abundance at the level of specific genus or species.[38],[45] The influence of conditions such as obesity on the gut microbiota has been shown to a large extent to be similar between mice and humans, and therefore, murine models are useful and relevant for humans in microbiota research. The use of murine models makes possible experiments that could not be performed on humans and are offering better control over the confounding factors. Obviously, absolute comparisons cannot be made.


  Conclusions Top


We have shown that complex associations exist between the microbiota and genetic background, diet, obesity, blood glucose regulation, inflammation, and intestinal tumorigenesis. These complex associations could be a contributing reason for the lack of consensus between the results in previous studies regarding correlations of microbiota with obesity and cancer. The complexity of the system and the opposing effects found between closely related species also points to caution in using therapeutic interventions to modulate microbiota until a more complete understanding of factors affecting the microbiota is obtained.

Financial support and sponsorship

The mouse study generating the fecal material and experimental data used in this study was supported by the Research Council of Norway (project no. 196112/H10 to I-LS).

Conflicts of interest

There are no conflicts of interest.

 
  References Top

1.
World Cancer Research Fund/American Institute for Cancer Research. Food, Nutrition, Physical Activity and the Prevention of Cancer: A Global Perspective. Washington, DC, USA; 2007. Available from: http://www.aicr.org/research/research_science_expert_report.html. [Last accessed on 2017 Feb 22].  Back to cited text no. 1
    
2.
World Cancer Research Fund/American Institute for Cancer Research. Continuous Update Project Report. Food, Nutrition, Physical Activity, and the Prevention of Colorectal Cancer. Washington, DC, USA; 2011. Available from: http://www.dietandcancerreport.org/cancer_resource_center/downloads/cu/Colorectal-Cancer-2011-Report.pdf. [Last accessed on 2017 Feb 22].  Back to cited text no. 2
    
3.
Renehan AG, Zwahlen M, Egger M. Adiposity and cancer risk: new mechanistic insights from epidemiology. Nat Rev Cancer 2015;15:484-98.  Back to cited text no. 3
    
4.
Ungefroren H, Gieseler F, Fliedner S, Lehnert H. Obesity and cancer. Horm Mol Biol Clin Investig 2015;21:5-15.  Back to cited text no. 4
    
5.
Arrieta M-C, Stiemsma LT, Amenyogbe N, Brown EM, Finlay B. The intestinal M-C in early life: health and disease. Front Immunol 2014;5:427.  Back to cited text no. 5
    
6.
Joyce SA, Gahan CGM. The gut microbiota and the metabolic health of the host. CGM 2014;30:120-7.  Back to cited text no. 6
    
7.
Rodríguez JM, Murphy K, Stanton C, Ross RP, et al. The composition of the gut microbiota throughout life, with an emphasis on early life. Microb Ecol Health Dis 2015;26:26050.  Back to cited text no. 7
    
8.
Bultman SJ. Emerging roles of the microbiome in cancer. Carcinogenesis 2014;35:249-55.  Back to cited text no. 8
    
9.
Power SE, O'Toole PW, Stanton C, Ross RP, Fitzgerald GF. Intestinal microbiota, diet and health. Br J Nutr 2014;111:387-402.  Back to cited text no. 9
    
10.
Underwood MA. Intestinal dysbiosis: Novel mechanisms by which gut microbes trigger and prevent disease. Prev Med 2014;65:133-7.  Back to cited text no. 10
    
11.
Bakker GJ, Zhao J, Herrema H, Nieuwdorp M. Gut microbiota and energy expenditure in health and obesity. J Clin Gastroenterol 2015;49 Suppl 1:S13-9.  Back to cited text no. 11
    
12.
Carding S, Verbeke K, Vipond DT, Corfe BM, et al. Dysbiosis of the gut microbiota in disease. Microb Ecol Health Dis 2015;26:26191.  Back to cited text no. 12
    
13.
Ngo HT, Hetland RB, Nygaard UC, Steffensen I-L. Genetic and diet-induced obesity increased intestinal tumorigenesis in the double mutant mouse model multiple intestinal neoplasia X obese via disturbed glucose regulation and inflammation. J Obes 2015;343479.  Back to cited text no. 13
    
14.
Ravi A, Avershina E, Foley SL, Ludvigsen J, Storrø O, Øien T, et al. The commensal infant gut meta-mobilome as a potential reservoir for persistent multidrug resistance integrons. Sci Rep 2015;5:15317.  Back to cited text no. 14
    
15.
Edgar RC. UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nat Methods 2013;10:996-8.  Back to cited text no. 15
    
16.
Xiao L, Feng Q, Liang S, Sonne SB, Xia Z, et al.A catalog of the mouse gut metagenome. Nat Biotechnol 2015;33:1103-8.  Back to cited text no. 16
    
17.
Fearon ER. Molecular genetics of colorectal cancer. Annu Rev Pathol Mech Dis 2011;6:479-507.  Back to cited text no. 17
    
18.
Al-Sohaily S, Biankin A, Leong R, Kohonen-Corish M, Warusavitarne J. Molecular pathways in colorectal cancer. J Gastroenterol Hepatol 2012;27:1423-31.  Back to cited text no. 18
    
19.
Moser AR, Pitot HC, Dove WF. A dominant mutation that predisposes to multiple intestinal neoplasia in the mouse. Science 1990;247:322-4.  Back to cited text no. 19
    
20.
Su L-K, Kinzler KW, Vogelstein B, Preisinger AC, Moser AR, Luongo C, et al. Multiple intestinal neoplasia caused by a mutation in the murine homolog of the APC gene. Science 1992;256:668-70.  Back to cited text no. 20
    
21.
Boman BM, Fields JZ. An APC: WNT counter-current-like mechanism regulates cell division along the human colonic crypt axis: a mechanism that explains how APC mutations induce proliferative abnormalities that drive colon cancer development. Front Oncol 2013;3:244.  Back to cited text no. 21
    
22.
Nelson S, Näthke IS. Interactions and functions of the adenomatous polyposis coli (APC) protein at a glance. J Cell Sci 2013;126(Pt 4):873-7.  Back to cited text no. 22
    
23.
Ingalls AM, Dickie MM, Snell GD. Obese, a new mutation in the house mouse. J Hered 1950;41:317-8.  Back to cited text no. 23
    
24.
Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM. Positional cloning of the mouse obese gene and its human homologue. Nature 1994;372:425-32.  Back to cited text no. 24
    
25.
Chehab FF, Qiu J, Ogus S. The use of animal models to dissect the biology of leptin. Recent Prog Horm Res 2004;59:245-66.  Back to cited text no. 25
    
26.
Friedman JM, Halaas JL. Leptin and the regulation of body weight in mammals. Nature 1998;395:763-70.  Back to cited text no. 26
    
27.
Ngo HT, Hetland RB, Sabaredzovic A, Haug LS, Steffensen I-L. In utero exposure to perfluorooctanoate (PFOA) or perfluorooctane sulfonate (PFOS) did not increase body weight or intestinal tumorigenesis in multiple intestinal neoplasia (Min/+) mice. Environ Res 2014;132:251-63.  Back to cited text no. 27
    
28.
Yu Y, Lee C, Kim J, Hwang S. Group-specific primer and probe sets to detect methanogenic communities using quantitative real-time polymerase chain reaction. Biotechnol Bioeng 2005;89:670-9.  Back to cited text no. 28
    
29.
DeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL, Keller K, et al. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microbiol 2006;72:5069-72.  Back to cited text no. 29
    
30.
Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods 2010;7:335-6.  Back to cited text no. 30
    
31.
Smilde AK, Jansen JJ, Hoefsloot HCJ, Lamers R-JAN, van der Greef J, Timmerman ME. ANOVA-simultaneous component analysis (ASCA): a new tool for analyzing designed metabolomics data. Bioinformatics 2005;21:3043-8.  Back to cited text no. 31
    
32.
Graf D, Di Cagno R, Fåk F, Flint HJ, Nyman M, Saarela M, et al. Contribution of diet to the composition of the human gut microbiota. Microb Ecol Health Dis 2015;26:26164.  Back to cited text no. 32
    
33.
Gupta RS, Gao B. Phylogenomic analyses of clostridia and identification of novel protein signatures that are specific to the genus Clostridium sensu stricto (cluster I). Int J Syst Evol Microbiol 2009;59(Pt 2):285-94.  Back to cited text no. 33
    
34.
Kong LC, Holmes BA, Cotillard A, Habi-Rachedi F, Brazeilles R, Gougis S, et al. Dietary patterns differently associate with inflammation and gut microbiota in overweight and obese subjects. PLoS One 2014;9:e109434.  Back to cited text no. 34
    
35.
Carmody RN, Gerber GK, Luevano JM Jr., Gatti DM, Somes L, Svenson KL, et al. Diet dominates host genotype in shaping the murine gut microbiota. Cell Host Microbe 2015;17:72-84.  Back to cited text no. 35
    
36.
Le Chatelier E, Nielsen T, Qin J, Prifti E, Hildebrand F, Falony G, et al. Richness of human gut microbiome correlates with metabolic markers. Nature 2013;500:541-6.  Back to cited text no. 36
    
37.
Cotillard A, Kennedy SP, Kong LC, Prifti E, Pons N, Le Chatelier E, et al. Dietary intervention impact on gut microbial gene richness. Nature 2013;500:585-8.  Back to cited text no. 37
    
38.
Ley RE, Bäckhed F, Turnbaugh P, Lozupone CA, Knight RD, Gordon JI. Obesity alters gut microbial ecology. Proc Natl Acad Sci U S A 2005;102:11070-5.  Back to cited text no. 38
    
39.
Pfalzer AC, Nesbeth P-DC, Parnell LD, Iyer LK, Liu Z, Kane AV, et al. Diet- and genetically-induced obesity differentially affect the fecal microbiome and metabolome in Apc1638N mice. PLoS One 2015;10:e0135758.  Back to cited text no. 39
    
40.
Million M, Angelakis E, Maraninchi M, Henry M, Giorgi R, Valero R, et al. Correlation between body mass index and gut concentrations of Lactobacillus reuteri, Bifidobacterium animalis, Methanobrevibacter smithii and Escherichia coli. Int J Obes 2013;37:1460-6.  Back to cited text no. 40
    
41.
Turnbaugh PJ, Bäckhed F, Fulton L, Gordon JI. Diet-induced obesity is linked to marked but reversible alterations in the mouse distal gut microbiome. Cell Host Microbe 2008;3:213-23.  Back to cited text no. 41
    
42.
Schulz MD, Atay C, Heringer J, Romrig FK, Schwitalla S, Aydin B, et al. High-fat-diet-mediated dysbiosis promotes intestinal carcinogenesis independently of obesity. Nature 2014;514:508-12.  Back to cited text no. 42
    
43.
Lennon G, Balfe Á, Bambury N, Lavelle A, Maguire A, Docherty NG, et al. Correlations between colonic crypt mucin chemotype, inflammatory grade and Desulfovibrio species in ulcerative colitis. Colorectal Dis 2014;16:O161-9.  Back to cited text no. 43
    
44.
Elsheikh W, Blackler RW, Flannigan KL, Wallace JL. Enhanced chemopreventive effects of a hydrogen sulfide-releasing anti-inflammatory drug (ATB-346) in experimental colorectal cancer. Nitric Oxide 2014;41:131-7.  Back to cited text no. 44
    
45.
Nguyen TLA, Vieira-Silva S, Liston A, Raes J. How informative is the mouse for human gut microbiota research? Dis Model Mech 2015;8:1-16.  Back to cited text no. 45
    


    Figures

  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]
 
 
    Tables

  [Table 1], [Table 2]



 

Top
 
 
  Search
 
Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

 
  In this article
Abstract
Introduction
Materials and Me...
Results
Discussion
Conclusions
References
Article Figures
Article Tables

 Article Access Statistics
    Viewed1822    
    Printed175    
    Emailed0    
    PDF Downloaded279    
    Comments [Add]    

Recommend this journal


[TAG2]
[TAG3]
[TAG4]