Article Text

Original article
Divergent metabolic outcomes arising from targeted manipulation of the gut microbiota in diet-induced obesity
  1. Eileen F Murphy1,2,
  2. Paul D Cotter1,3,
  3. Aileen Hogan1,
  4. Orla O'Sullivan3,
  5. Andy Joyce2,
  6. Fiona Fouhy3,4,
  7. Siobhan F Clarke3,4,
  8. Tatiana M Marques1,3,
  9. Paul W O'Toole1,4,
  10. Catherine Stanton1,3,
  11. Eamonn M M Quigley1,5,
  12. Charlie Daly1,
  13. Paul R Ross1,3,
  14. Robert M O'Doherty1,6,7,
  15. Fergus Shanahan1,5
  1. 1Alimentary Pharmabiotic Centre, University College Cork, Cork, Ireland
  2. 2Alimentary Health Ltd., Cork, Ireland
  3. 3Teagasc Food Research Centre, Moorepark Fermoy, County Cork, Ireland
  4. 4Department of Microbiology, University College Cork, Cork, Ireland
  5. 5Department of Medicine, University College Cork, Cork, Ireland
  6. 6Department of Medicine, Division of Endocrinology and Metabolism, University of Pittsburgh, Pittsburgh, Pennsylvania, USA
  7. 7Department of Microbiology and Molecular Genetics, University of Pittsburgh, Pittsburgh, Pennsylvania, USA
  1. Correspondence to Professor Fergus Shanahan, Department of Medicine and Alimentary Pharmabiotic Centre, University College Cork, National University of Ireland, Cork, Ireland; f.shanahan{at}ucc.ie Professor Robert O'Doherty, Department of Medicine, Division of Endocrinology and Metabolism and Department of Microbiology and Molecular Genetics, University of Pittsburgh, Pittsburgh, PA 15261, USA; rmo1{at}pitt.edu

Abstract

Objective The gut microbiota is an environmental regulator of fat storage and adiposity. Whether the microbiota represents a realistic therapeutic target for improving metabolic health is unclear. This study explored two antimicrobial strategies for their impact on metabolic abnormalities in murine diet-induced obesity: oral vancomycin and a bacteriocin-producing probiotic (Lactobacillus salivarius UCC118 Bac+).

Design Male (7-week-old) C57BL/J6 mice (9–10/group) were fed a low-fat (lean) or a high-fat diet for 20 weeks with/without vancomycin by gavage at 2 mg/day, or with L salivarius UCC118Bac+ or the bacteriocin-negative derivative L salivarius UCC118Bac (each at a dose of 1×109 cfu/day by gavage). Compositional analysis of the microbiota was by 16S rDNA amplicon pyrosequencing.

Results Analysis of the gut microbiota showed that vancomycin treatment led to significant reductions in the proportions of Firmicutes and Bacteroidetes and a dramatic increase in Proteobacteria, with no change in Actinobacteria. Vancomycin-treated high-fat-fed mice gained less weight over the intervention period despite similar caloric intake, and had lower fasting blood glucose, plasma TNFα and triglyceride levels compared with diet-induced obese controls. The bacteriocin-producing probiotic had no significant impact on the proportions of Firmicutes but resulted in a relative increase in Bacteroidetes and Proteobacteria and a decrease in Actinobacteria compared with the non-bacteriocin-producing control. No improvement in metabolic profiles was observed in probiotic-fed diet-induced obese mice.

Conclusion Both vancomycin and the bacteriocin-producing probiotic altered the gut microbiota in diet-induced obese mice, but in distinct ways. Only vancomycin treatment resulted in an improvement in the metabolic abnormalities associated with obesity thereby establishing that while the gut microbiota is a realistic therapeutic target, the specificity of the antimicrobial agent employed is critical.

  • Antimicrobials
  • diet-induced obesity
  • gut microbiota
  • mouse models
View Full Text

Statistics from Altmetric.com

Significance of this study

What is already known about this subject?

  • The gut microbiota is an environmental regulator of fat storage and adiposity.

  • We and others have shown the primacy of diet in influencing the microbiota in obesity.

  • The composition of the gut microbiota is significantly altered in obesity and diabetes in both animal and human studies.

  • It has been suggested that alteration in microbial composition increases the risk of obesity because of enhanced energy harvest from dietary intake.

What are the new findings?

  • Vancomycin and the bacteriocin-producing probiotic produced distinctive modifications in the gut microbiota in diet-induced obese mice at the phylum, family and genus levels.

  • To our knowledge, this is the first report to establish that a bacteriocin produced by a probiotic can substantially alter the composition of the gut microbiota in vivo.

  • However, only vancomycin treatment resulted in an improvement in the metabolic abnormalities associated with obesity.

  • Our findings provide further confirmation for the role of the microbiota in metabolic dysregulation and a supporting rationale for altering the microbiota as a prophylactic strategy using antimicrobial agents, including bacteriocins, but specificity of action will be crucial.

How might it impact on clinical practice in the foreseeable future?

  • These findings indicate that therapeutic manipulation of the microbiota may be a useful strategy in the prevention or management of obesity and metabolic disorders.

The basis of the modern obesity epidemic in developed societies is complex and involves a contribution from genetic susceptibility and, more importantly, changes in diet and other lifestyle elements. In addition, the molecular events underlying the pathogenesis of the metabolic abnormalities of obesity are incompletely understood and interventions to treat them are poorly developed. A body of work has recently implicated alterations of the gut microbiota as a contributory factor to obesity-related metabolic dysregulation.1 ,2

The composition of the gut microbiota is significantly altered in obesity and diabetes in both animal and human studies and is characterised by reduced diversity.3–8 It has been suggested that this alteration in microbial composition increases the risk of obesity because of enhanced energy harvest from dietary intake.6–8 We and others have demonstrated that, in murine models, changes in the gut microbiota in response to diet and obesity are dissociated from markers of energy harvest over time, suggesting that mechanisms other than energy harvest may contribute to microbiota-induced susceptibility to obesity and metabolic diseases.9–11 In this regard, the gut microbiota and associated products such as lipopolysaccharides and short-chain fatty acids have been reported to regulate gene expression, and thereby alter energy expenditure and storage through host-related mechanisms.12–15 However, it is unclear if the microbiota represents a realistic therapeutic target for improving metabolic health.

The composition of the microbiome is dynamic and adaptable, and a number of strategies including antibiotics, prebiotics and probiotics have the potential to influence host metabolism favourably by targeting the gut microbiota. In this study, we explored two antimicrobial strategies for their impact on metabolic abnormalities in murine diet-induced obesity: oral vancomycin and a bacteriocin-producing probiotic lactobacillus. Vancomycin exhibits anti-Firmicutes activity and has limited systemic effects,16–18 while Lactobacillus salivarius UCC118 is a genetically well characterised probiotic strain that produces a broad-spectrum class II bacteriocin, Abp118, and has previously been shown to protect mice from infection by Listeria monocytogenes.19–21 The availability of a bacteriocin-negative (Bac) derivative of L salivarius UCC118 allowed for the direct assessment of the effect of the bacteriocin on the gut microbiota and its impact on metabolic dysregulation. The results show that the antimicrobial agents modulated the gut microbiota in different ways, but only vancomycin treatment resulted in an improvement in metabolic markers in diet-induced obesity. These findings indicate that therapeutic manipulation of the microbiota may be a useful strategy in the prevention or management of obesity and metabolic disorders.

Materials and methods

Animals

C57BL/6J male mice, aged 3–4 weeks, were obtained from Harlan (Oxon, UK) and housed under barrier-maintained conditions within the biological services unit, University College Cork (UCC). Mice were allowed to acclimatise for 3–4 weeks before the start of the study. All experiments were approved by the UCC Animal Ethics Committee and experimental procedures were conducted under licence from the Irish government.

Experimental design

To assess the impact of the two antimicrobial strategies on metabolic abnormalities in murine diet-induced obesity, 7-week-old male C57BL/J6 mice (9–10 per group) were fed either a low-fat diet (control; 10% calories from fat; Research Diets, New Jersey, USA; #D12450B), a high-fat diet (diet-induced obesity; 45% calories from fat; Research Diets; #D12451) for 20 weeks or a high-fat diet for 20 weeks including an 8-week oral treatment from weeks 12 to 20 with either 2 mg/day of vancomycin (Sigma Aldrich, UK), the bacteriocin-producing probiotic L salivarius UCC118Bac+ at 1×109 cfu/day, or L salivarius UCC118Bac at 1×109 cfu/day. All mice were gavaged and phosphate-buffered saline was used as the vehicle control for the vancomycin-treated group. An initial study was undertaken to establish that vancomycin at a dose of 2 mg/day did not alter the health status of mice and was effective at altering the components of the gut microbiota in mice, using traditional plating methods, as outlined in the supplementary materials and methods (available online only).

Probiotic production

The bacteriocin-negative derivative L salivarius UCC118Bac was generated as described by Corr et al.19 Both the L salivarius UCC118Bac+ and the L salivarius UCC118Bac were grown in MRS broth media for 24 h at 37°C, harvested by centrifuge and freeze dried. The powders were resuspended in water for delivery to the mice.

Weights and tissue sampling

Body weight and food intake were assessed weekly. At the end of the study, fat and lean body mass were measured using a Minispec mq benchtop NMR spectrometer (Bruker Instruments, Rheistetten, Germany), after which they were killed and necropsied. Internal organs (liver and spleen) and fat pads (reproductive, renal, mesenteric and inguinal) were removed, weighed and stored at −80°C.

DNA extractions and amplicon sequencing

Total metagenomic DNA was extracted from individual faecal samples and 16S ribosomal RNA gene tags of the V4 region were amplified, sequenced and subjected to in-silico analysis as described in the supplementary materials and methods (available online only) and by Murphy et al.11 Real-time quantitative PCR was used to determine total bacterial numbers (16S rRNA gene copies), and hierarchical clustering was used to provide an overview of the data, as outlined in the supplementary materials and methods (available online only).

Metabolic markers

Mice were fasted for 5–6 h and blood glucose was determined using a Coutour glucose meter (Bayer, UK) on 1 μl of blood collected from the tip of the tail vein. Blood was collected by cardiac puncture from fasted mice and plasma obtained. Tumour necrosis factor alpha (TNFα) was measured in 25 μl of plasma using an ultrasensitive kit from MesoScale Discovery (Gaithersburg, Maryland, USA), and plasma insulin concentrations were determined in 5 μl of plasma using an ELISA kit (Mercodia, Uppsala, Sweden) as per the manufacturer's instructions. Plasma free fatty acids were determined using a Wako kit (Wako, Germany) and plasma triglyceride levels were determined on 3 μl of plasma using infinity triglyceride liquid stable reagent (Thermo Scientific, Middletown, Virginia, USA). The lipids from 50 mg of frozen liver were extracted according to the Folch method,22 and triglyceride levels were determined using infinity triglyceride liquid stable reagent (Thermo Scientific).

Statistical analysis

Data for all variables were normally distributed and allowed for parametric tests of significance. Data are presented as mean values with their SEM. Statistical analysis was performed by analysis of variance and Student's t test (Graph-Pad Software, San Diego, CA, USA).

Results

Diet-induced obesity alters the composition of the gut microbiota

As expected, diet-induced obese mice gained significantly more body weight compared with lean controls over the 20-week feeding period, and this increase in body weight (41.75±1.12 vs 33.22±0.92 g; p<0.001) was attributable to an increase in fat mass alone (16.03±0.77 vs 6.27±0.61; p<0.001). Diet-induced obese mice consumed significantly more calories than lean controls, as measured by the cumulative caloric intake over the 20-week period of the study (7212±35 vs 6136±27 kJ/mouse; p<0.001).

At the end of the 20-week study, the composition of the gut microbiota of individual mice was investigated by DNA sequencing (Roche-454 titanium, Roche Diagnostics Ltd, West Sussex, UK) of 16S rRNA (V4) amplicons generated from total faecal DNA. A total of 212 655 sequence reads was generated, averaging 4340 reads per faecal sample. Species richness, coverage and diversity estimations were calculated for each dataset (see supplementary figure 1, available online only). Rarefaction curves for each group indicated that the total bacterial diversity present was well represented. Of the reads, 173 444 (82%) were assigned at the phylum level, 100 282 (47%) at the family level and 80 899 (38%) at the genus level. In agreement with previous studies,4 ,6 ,8 the mouse faecal microbiota was dominated by Firmicutes and Bacteroidetes at the phylum level (table 1). Actinobacteria, Proteobacteria and Deferribacteria were also detected but at lower proportions. Consistent with the high levels of Firmicutes and Bacteroidetes detected, the most dominant bacteria at the genus level were Clostridium, Lactobacillus and Bacteroides (table 1).

Table 1

Vancomycin treatment of diet-induced obese mice results in a major disturbance in the gut microbiota

A comparison of the composition of the gut microbiota of lean and diet-induced obese mice showed that high-fat feeding for 20 weeks was associated with an increase in the relative proportions of Firmicutes (p<0.05) and a decrease in Bacteroidetes proportions (p<0.05) compared with lean controls (table 1 and supplementary figure 2, available online only). At the family level, diet-induced obesity in mice was associated with a relative increase in the proportions of the Firmicutes class Lactobacillaceae (p<0.01) and a decrease in the Bacteroidetes class Bacteroidaceae (p<0.05), while the abundance of Streptococcaceae (p<0.05) and Alcaligenaceae (p<0.05) increased and decreased, respectively (table 1 and supplementary figure 2, available online only). At the genus level, the proportions of Lactobacillus (p<0.01) and Lactococcus spp. (p<0.05) increased, while Bacteroides (p<0.001), Odoribacter (p<0.05) and Sutterella (p<0.05) decreased in abundance (table 1 and supplementary figure 2, available online only).

Vancomycin does not alter the health status of chow-fed mice

A preliminary investigation (outlined in the supplementary materials and methods, available online only) was undertaken to determine if vancomycin treatment was associated with any adverse health effects in healthy mice. The results confirmed that vancomycin did not negatively affect body weight, appearance or consistency of the stool in chow-fed C57BL/6 male mice at a dose of 2 mg/day over 4 weeks. In addition, culture-based microbial analysis of the faeces of healthy mice confirmed that, at this dose, vancomycin impacted specific microbiota elements, with a decrease in Enterococcus spp. (phylum Firmicutes), an increase in total Enterobacteriaceae (phylum Proteobacteria) and no change in either total anaerobes or Lactobacillus spp. (phylum Firmicutes; see supplementary figure 3, available online only).

Vancomycin treatment alters the composition of the gut microbiota in diet-induced obese mice

DNA sequencing-based analysis of the microbiota revealed that treatment of diet-induced obese mice for 8 weeks with vancomycin at 2 mg/day resulted in a significant reduction in the relative proportions of Firmicutes (p<0.01), Bacteroidetes (p<0.001) and Deferribacteres (p<0.05), and a dramatic increase in the proportions of Proteobacteria (p<0.001) compared with diet-induced obese mice (table 1 and supplementary figure 2, available online only). The increased proportion of Proteobacteria was largely accounted for by an increased percentage of Enterobacteriaceae (p<0.001), while the decrease in Firmicutes and Bacteroidetes corresponded to a decrease in the proportions of Clostridiaceae (p<0.001) and Bacteroidaceae (p<0.001), respectively, at the family level (table 1 and supplementary figure 2, available online only). Vancomycin treatment of diet-induced obese mice also resulted in changes among less common families, such as a relative decrease in the proportions of Porphyromonadaceae (p<0.001) and increases in the proportions of Streptococcaceae (p<0.01), Desulfovibrionaceae (p<0.01) and Alcaligenaceae (p<0.001). At the genus level, the proportions of Lactococcus (p<0.001), Sutterella (p<0.001) and Desulfovibrio spp. (p<0.01) increased, while Bacteroides (p<0.001), Clostridium (p<0.001) and Odoribacter spp. (p<0.05) decreased (table 1 and supplementary figure 2, available online only). The overall proportions of Lactobacillus spp. were unchanged by vancomycin treatment of diet-induced obese mice. Clustering of the data showed that diet-induced obese mice treated with vancomycin grouped together at the phylum level and separated well from lean and diet-induced obese mice (see supplementary figure 4, available online only). Better separation of lean and diet-induced obese mice was observed at the family and genus level (data not shown) than at the phylum level.

L salivarius UCC118 alters the composition of the gut microbiota in diet-induced obese mice by a mechanism involving bacteriocin production

Comparison of the L salivarius UCC118 Bac+ with the non-bacteriocin-producing strain, L salivarius UCC118Bac, showed that the production of the antimicrobial agent altered the composition of the gut microbiota in diet-induced obese mice. While the proportions of Firmicutes did not change, feeding of L salivarius UCC118Bac+ for 8 weeks resulted in a significant increase in the relative proportions of Bacteroidetes and Proteobacteria, and a decrease in the proportions of Actinobacteria compared with the bacteriocin-negative derivative, L salivarius UCC118Bac (table 2 and supplementary figure 5, available online only). Consistent with these results, the increase in Bacteroidetes corresponded to an increase in the Bacteroidaceae family and the genus Bacteroides (table 2 and supplementary figure 5, available online only). The decrease in Actinobacteria corresponded to a decrease in Bifidobacteriaceae at the family level and Bifidobacteria at the genus level (table 2 and supplementary figure 5, available online only). Hierarchical clustering showed that, consistent with small changes, separation was not apparent between L salivarius UCC118 Bac+ and the non-bacteriocin-producing strain, L salivarius UCC118 Bac, using this technique (see supplementary figure 6, available online only).

Table 2

L salivarius UCC118 alters the composition of the gut microbiota in diet-induced obese mice by a mechanism involving bacteriocin production

Total bacterial counts were altered by diet-induced obesity and vancomycin treatment, but not by bacteriocin-production by L salivarius UCC118, in diet-induced obese mice

Quantitative PCR analysis revealed that total bacterial counts (16S rRNA gene copies/g of stool) were significantly lower in the faeces of diet-induced obese mice compared with lean controls (p<0.001; figure 1A). Treatment of diet-induced obese mice with vancomycin resulted in a further large decrease in absolute faecal bacterial numbers compared with their diet-induced obese counterparts (p<0.001; figure 1A). In contrast, comparison of treatment with L salivarius UCC118Bac+ and its bacteriocin-negative derivative showed that bacteriocin production did not alter the total bacterial numbers in the faeces of diet-induced obese mice (figure 1B).

Figure 1

Numbers of 16s rRNA gene copies/g stool in lean, diet-induced obese (DIO) and vancomycin (vanco)-treated diet-induced obese mice (A) and in diet-induced obese mice treated with the bacteriocin-producing probiotic strain L salivarius UCC118 Bac+ and a non-bacteriocin-producing derivative L salivarius UCC118 Bac (B). Data represented as mean±SEM, n=9–10, ***p<0.001.

Antimicrobial strategies alter weight gain in diet-induced obese mice

Treatment of diet-induced obese mice with vancomycin resulted in a significant reduction in weight gain (p<0.05; figure 2A) compared with diet-induced obese mice, despite a large increase in caecum weight (p<0.0001) relative to the caeca of diet-induced obese controls (see supplementary table 1, available online only). However, a recovery in the rate of body weight gain in the vancomycin-treated diet-induced obese mice was evident after day 28 of the intervention. For probiotic-fed mice, a statistically significant reduction in weight gain was observed on treatment with L salivarius UCC118bac+ in diet-induced obese mice compared with L salivarius UCC118bac at days 14, 21 and 28 of the 8-week intervention period, but this effect did not persist over time (figure 2B). In addition, calorie intake was not altered by treatment with vancomycin, L salivarius UCC118Bac+ or L salivarius UCC118Bac in diet-induced obese mice (figure 2C).

Figure 2

Weight gain (g) (A and B) and cumulative energy intake (kJ/mouse) (C) over the 8-week intervention period in lean, diet-induced obese (DIO) and vancomycin (vanco)-treated diet-induced obese mice and in diet-induced obese mice treated with the bacteriocin-producing probiotic strain L salivarius UCC118 Bac+ (1×109 cfu/day) and a non-bacteriocin-producing derivative L salivarius UCC118 Bac (1×109 cfu/day). Data represented as mean±SEM, n=9–10. *p<0.05.

Vancomycin treatment, but not bacteriocin-production by L salivarius UCC118, improves metabolic markers in diet-induced obese mice

We next investigated the metabolic consequences of perturbing the gut microbiota using the two antimicrobial strategies. Comparison of lean and diet-induced obese mice after 20 weeks of feeding showed that diet-induced obese mice had elevated fasting blood glucose (p<0.001) and plasma insulin (p<0.05), while there was no change in plasma triglycerides compared with lean controls (see supplementary table 1, available online only). Treatment of diet-induced obese mice with vancomycin for 8 weeks resulted in an improvement in fasting blood glucose levels (p<0.05) and plasma triglyceride levels (p=0.06) compared with diet-induced obese mice, while there was no change in fasting insulin levels (see supplementary table 1, available online only). Diet-induced obese mice were characterised by elevated liver weight (p<0.01) and liver triglyceride levels (p<0.001) compared with lean controls. Interestingly, there was a trend towards a reduction in liver weight (p=0.06) in vancomycin-treated diet-induced obese mice relative to diet-induced obese mice. However, liver triglyceride levels were unaltered by vancomycin in diet-induced obese mice. In contrast, comparison of L salivarius UCC118Bac+ with the bacteriocin-negative derivative showed that the bacteriocin-induced alterations in the gut microbiota did not alter metabolic parameters in diet-induced obese mice (see supplementary table 2, available online only).

Vancomycin treatment is associated with reduced plasma TNFα levels and TNFα mRNA levels in liver and visceral adipose tissues of diet-induced obese mice

To investigate further the effect of vancomycin on metabolic health, the plasma levels of TNFα and the gene expression of TNFα, monocyte chemoattractant protein 1 (MCP-1) and the macrophage differentiation marker (F4+80) were assessed in the visceral adipose tissue and liver as markers of inflammation. Diet-induced obesity in mice was associated with an increase in plasma TNFα (p<0.05) and an increase in the gene expression of F4+80 (p<0.001), MCP-1 (p<0.001) and TNFα (p<0.05) in visceral adipose but not liver tissue (figure 3A,B). Treatment of diet-induced obese mice with vancomycin resulted in a decrease in plasma TNFα levels (p<0.05) compared with diet-induced obese mice (figure 3A), and this was associated with a tendency towards a reduction in the gene expression of TNFα in both adipose (p=0.06) and liver (p=0.09) tissue. The gene expression levels of F4+80 and MCP-1 in adipose or liver tissues were unchanged in vancomycin-treated diet-induced obese mice compared with diet-induced obese controls (figure 3B).

Figure 3

Vancomycin (vanco) treatment improves the inflammatory tone of diet-induced obese (DIO) mice. Plasma tumour necrosis factor alpha (TNFα) (pg/ml) (A) and gene expression levels of the macrophage differentiation marker (F4+80), monocyte chemoattractant protein 1 (MCP-1) and TNFα in visceral adipose tissue and liver (B) in lean and diet-induced obese mice after 20 weeks of feeding and in vancomycin-treated diet-induced obese mice (12 weeks of high-fat feeding followed by 8-week intervention with vancomycin (2 mg/day)). Data represented as mean±SEM, n=9–10. *p<0.05; ***p<0.001. Values for gene expression are fold change ±SEM, n=9–10. Expression is relative to β-actin.

Discussion

The findings show that although vancomycin and the bacteriocin-producing probiotic, L salivarius UCC118, altered the gut microbiota in diet-induced obese mice in distinct ways, only vancomycin treatment resulted in an improvement in the metabolic abnormalities associated with diet-induced obesity.

In the present study, vancomycin was chosen for its ability to target the Gram-positive component of the gut microbiota and its limited systemic impact.16–18 In addition, vancomycin exposure has been shown to alter significantly the host metabolome in healthy mice18 and the main components of the gut microbiota in a human distal colon model.23 In the current study, vancomycin treatment of diet-induced obese mice resulted in a major alteration in the composition of the gut microbiota whereby the respective proportions of the three dominant phyla were altered dramatically with respect to each other, with a large reduction in Firmicutes, and in particular the Bacteroidetes, and a dramatic increase in Proteobacteria. These alterations were associated with a reduction in body weight gain and an improvement in inflammatory and metabolic health of the host. In particular, plasma TNFα levels were reduced in vancomycin-treated diet-induced obese mice compared with diet-induced obese controls, and this corresponded to a trend towards a reduction in the gene expression of TNFα levels in the liver and visceral adipose tissues. In both in-vitro and in animal models, an increase in TNFα has been linked to tissue insulin resistance.18 ,24–26 Other studies have shown that modulation of the microbiota by broad-spectrum antibiotics results in a reduction in metabolic endotoxaemia in both high-fat-fed and ob/ob mice, and is associated with improvements in inflammation, glucose tolerance and hepatic steatosis, possibly through a mechanism involving Toll-like receptors.4 ,13 ,27 However, whether the effect on weight gain is sustained or is overcome by microbial compensatory adjustments is unclear, and more long-term studies in animal models and humans are required. These data suggest that the ability of the gut microbiota to regulate inflammatory responses in diet-induced obesity is important in the interaction between gut microbes and obesity-related metabolic dysfunction.

To our knowledge, this is the first report to establish that a bacteriocin produced by a probiotic can substantially alter the composition of the gut microbiota in vivo. We have also recently shown that the bacteriocin produced by L salivarius UCC118 alters the microbiota of pigs and healthy chow-fed mice (Bisson and O'Toole, manuscript in preparation). Bacteriocin production is thought to confer a competitive advantage on the producing strain, enabling it to dominate complex microbial populations.28–30 In this study, the bacteriocin produced by L salivarius UCC118 significantly altered the gut microbiota in diet-induced obese mice by increasing the relative proportions of Bacteroidetes and Proteobacteria and decreasing Actinobacteria compared with a bacteriocin-negative derivative. While these observations suggest that bacteriocins can play a significant role in determining the composition of gut bacterial populations in vivo, the alterations did not confer beneficial effects on metabolic health. Interestingly, the bacteriocin produced by L salivarius UCC118 reduced the proportions of Bifidobacteria in diet-induced obese mice. A partial inhibition of Bifidobacteria by L salivarius UCC118 has previously been observed in in-vitro studies.31 Bifidobacteria have been shown to be positively correlated with improved glucose tolerance and normalised inflammatory tone in high-fat-fed mice.14 These results suggest that while the gut microbiota is a realistic target for addressing obesity-related metabolic dysfunction, the specificity of the antimicrobial agent employed may be critical. Indeed, distinct clusters or enterotypes in the human microbiome have been described32 and support the use of targeted strategies.

While the proximate microbiota-related biomarkers of risk for obesity and metabolic dysregulation remain to be determined, recent reports have suggested an association between lactobacilli and the development of obesity.33–35 Indeed, Lactobacillus populations have been shown to be elevated in obese subjects34 and subjects with diabetes.3 In addition, vancomycin treatment of patients with infective endocarditis was associated with a significant increase in weight gain,36 leading the authors to speculate that this may be due to the selection of Lactobacillus spp. by vancomyin in the gut. However, in the present study, the introduction of L salivarius UCC118 did not contribute to weight gain in diet-induced obese mice. In addition, although higher levels of Lactobacillus spp. were detected in diet-induced obese compared with lean mice, treatment of diet-induced obese mice with vancomycin reduced the rate of weight gain, without significantly affecting the proportions of Lactobacillus spp. Other studies using lactobacilli as probiotics have shown beneficial effects on metabolic health in animals37–39 and humans.40 ,41 These observations suggest that lactobacilli are not related to the risk of obesity and strain-specific effects need to be taken into account.

Recent reports have shown that the relationship between obesity, gut microbiota and the risk of obesity is more complex than previously considered.9 ,11 ,42 ,43 In the present study, in agreement with others4 ,5 ,8 diet-induced obesity was characterised by a relative increase in the proportions of Firmicutes and a decrease in Bacteroidetes, and was also associated with a decrease in total bacterial numbers. These results suggest that in addition to altering the composition of the gut microbiota, obesity and diet may also alter the total intestinal microbial load or ‘density’. Indeed, diet has been shown to alter diversity and induce large and rapid changes in the gut microbiota.44 ,45 Furthermore, vancomycin treatment of diet-induced obese mice resulted in a further large decrease in total bacterial numbers, while the bacteriocin produced by L salivarius UCC118 did not alter microbial loads in diet-induced obese mice. Further work is required to understand the role of variations in the total microbial load and diversity in obesity and related conditions and the significance of changes in phylum proportions when the total bacterial load also varies.

In conclusion, our data demonstrate that while vancomycin and the bacteriocin-producing probiotic produced distinctive modifications in the gut microbiota in diet-induced obese mice, vancomycin treatment alone resulted in an improvement in the metabolic abnormalities associated with obesity. Our findings provide further confirmation for the role of the microbiota in metabolic dysregulation, and a supporting rationale for altering the microbiota as a prophylactic strategy using antimicrobial agents, including bacteriocins, but specificity of action will be crucial.

Acknowledgments

The authors would like to thank Talia Huffe, Dr Paul Kenneally and Dr David O'Sullivan for their technical assistance.

References

View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Funding The authors are supported in part by Teagasc (an agency of the Irish government Department of Agriculture, Fisheries and Food), by Science Foundation Ireland (in the form of a research centre grant to the Alimentary Pharmabiotic Centre and a PI award to PWOT and PC), by grant NIH RO1 DK058855 (to ROD) and by Alimentary Health Ltd.

  • Correction notice This article has been corrected since it was published Online First. An additional corresponding author, Professor Robert O'Doherty, has been added.

  • Competing interests None.

  • Ethics approval All experiments were approved by the University College Cork Animal Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.