Obesity and its associated diseases are increasing worldwide. However, the mechanisms behind the development of obesity is not fully understood. There is evidence that intestinal bacteria may play a role in the development and perpetuation of obesity through regulation of energy and fat storage.

Bariatric surgery is currently the most effective modality for treating severe obesity with evidence to support long-term sustained weight loss and improvement in obesity-related comorbidities. The two most commonly performed bariatric surgical procedures are the Roux-en-Y gastric bypass (RYGB) and sleeve gastrectomy (SG). RYGB leads to greater weight loss than SG and improved diabetes control in patients following surgery. Despite the success of RYGB and SG in inducing weight loss and improving comorbidities, the underlying mechanisms leading to clinical improvement following these operations is not completely understood. Multiple factors are thought to play a role including reduced caloric intake, decreased nutrient absorption, increased satiety, release of hormones and shifts in bile acid metabolism.

Recent evidence has suggested that the gut bacteria mediates a number of the beneficial effects of bariatric surgery. Small studies have demonstrated changes in the composition and diversity of the gut microbiota after RYGB and SG in humans. One study also confirmed long-term microbial changes for RYGB. However, comparative trials have been small (less than 15 participants per treatment group) and important differences between specific bacterial populations have not been well elucidated. Furthermore, no human study has examined the differences in bacterial composition following RYGB and SG in relation to their metabolic consequences.

The aim of this study is to investigate and compare the metabolic and microbial changes that occur with RYGB, SG, and dietary controls. Specifically, the investigators aim to use a systems biology approach utilizing powerful analytic techniques including metagenomics, metabolomics, and multiplex immune profiling to define the combined microbial, metabolic and immunologic changes that occur after bariatric surgery.

HYPOTHESIS The investigators hypothesize that intestinal microbial dysbiosis and a decrease in diversity contributes to the development and perpetuation of obesity. The effect of dysbiosis is multifactorial and includes a decrease in intestinal barrier function and resultant local and systemic inflammation that incites the metabolic syndrome. The altered intestinal physiology following RYGB and SG will lead to identifiable changes in specific microbial populations and an increase in diversity. Specific beneficial microbial changes will subsequently result in weight loss, reduced inflammation, and a normalized metabolic profile.

METHODS Study population: Patients will be recruited from the Edmonton Adult Specialty Bariatric Clinic at the Royal Alexandra hospital.

Sample size: Each cohort of RYGB, SG and non-surgical dietary controls will have 30 patients (total n = 90). Previous studies have included 15 or less participants per arm.

Sample size calculation: Sample size calculation was designed to ensure the study would adequately capture microbial changes induced by surgery. In prior literature, an important short-chain fatty acid-producing bacterial species' (F. prausnitzii) relative abundance was lower in a post-RYGB group compared to non-operative controls (0.031 v. 0.053 σ 0.024). With an alpha of 0.05 and a Beta of 0.90, this would require 26 subjects per arm. Including a dropout rate of 10%, this increases to 30 subjects per arm.

Study design: For the intervention arm, subjects will be enrolled at the time they are scheduled for surgery. Fecal, urine, and blood samples will be collected in clinic 2-6 weeks prior to surgery. In the post-operative period, fecal collection will take place at scheduled three- and nine- month clinic visits. All pre-operative samples will be collected prior to subjects initiating a 2-4 week pre-operative liquid designed to reduce hepatomegaly and ease in the technical surgical aspects of the procedure.

Non-surgical controls will be patients who are treated with dietary and behavioural interventions for weight loss. This includes dietary and activity modifications and excludes meal replacement or pharmacologic interventions. For this cohort, subjects will have initial sampling (fecal, urine, blood) taken prior to initiating weight loss interventions. Further sampling will then occur at three months and nine months following initiation of the intervention.

Sample processing and immune analysis will take place at the Centre of Excellence for Gastrointestinal Inflammation and Immunity Research (CEGIIR) at the University of Alberta. Sequencing will be carried out by the Applied Genomics Center within CEGIIR and metabolomics will be done at the Metabolomic Innovation Center at the University of Alberta as a fee for service.

Fecal microbial analysis: Fecal sample collection will be driven by a previous developed protocol used by our group for diet studies in inflammatory bowel disease. Collection cups will be provided to patients and they will be instructed to collect a specimen the night prior or morning of their appointment. Subjects will be instructed to store the specimen in the fridge in the interim. Fecal samples will be analyzed for microbial composition, inflammatory signals, and fecal calprotectin.

The microbial community composition of fecal samples will be assessed using 16S rRNA gene analyses. DNA will be extracted from the fecal homogenates combining enzymatic and mechanical cell lysis with the QIAamp DNA Stool Mini Kit (Qiagen, Valencia, CA, USA). Fecal microbiota composition will be characterized by 16S rRNA tag sequencing using the MiSeq Illumina technology (pair-end), targeting the V3-V5 regions. Quality-controlled reads will be analyzed using 1) taxonomic-based approaches such as Global Alignment for Sequence Taxonomy (GAST)15 and the Ribosomal Database Project MultiClassifier tool and 2) non-taxonomic-based clustering algorithms for Operational Taxonomic Unit determination with the UPARSE pipeline. Alpha-diversity (observed species, Shannon, Simpson) and β-diversity indices (Bray-Curtis, binary Jaccard) will be calculated in QIIME and R (VEGAN package). Ordination plots for β-diversity metrics will be generated by non-parametric multidimensional scaling ordination in R. In order to assess the functional composition of the microbiome, gene content of the microbial community will be inferred using the PICRUSt algorithm16. PICRUSt uses information about gene content and 16S rRNA gene copy number from the IMG (integrated microbial genomes) database to predict which genes are present in organisms of the experimental samples. OTUs tables generated with QIIME/UPARSE will be normalized by 16S rRNA gene copy number and such normalized values are multiplied by the calculated abundance of gene families in each taxon during the gene content inference procedure performed with PICRUSt. The result is a table of gene family counts that is comparable to those generated by metagenome annotation pipelines such as HUMAnN and MG-RAST and which can be organized into metabolic pathways. Finally, the contribution of each OTU to a given gene function will be quantified. To identify microbial populations and metabolic pathways with differentiating abundance in the different groups, the LDA (Linear Discriminant Analysis) Effect Size (LEfSe) algorithm will be used with the online interface Galaxy (http://huttenhower.sph.harvard.edu/galaxy/root).

Urine and serum metabolomics: Urine and blood samples will be collected during the subject's clinic visit. Urine samples will be obtained in a standard urine collection jar containing sodium azide to prevent bacterial growth, and frozen after collection. Blood samples will be collected in heparinized collection tubes at the time of routine clinically indicated bloodwork, specifically six weeks pre-operatively and three and nine months post-operative. Serum will be isolated by centrifugation at 2 000 g for 10 minutes following collection and stored at -80⁰C. Urine and serum samples will be used for metabolomics profiling using NMR spectroscopy at each time point.

Metabolomic profiling will be done with NMR spectroscopy through the Metabolomics Innovation Center at the University of Alberta. Samples will be run on a 4-channel Varian INOVA 600 MHz NMR spectrometer. Standard Chenomx acquisition and processing parameters will be followed. 1 H-NMR analysis using Chenomx NMR Suite software will allow simultaneous identification of up to 300 small molecules. The resulting NMR spectra will be subjected to analysis using the technique of targeted profiling comparing spectra to a known reference database to identify metabolites.

Inflammatory cytokines and chemokines: Serum will be assessed for measurement of erythropoietin sedimentation rate (ESR) and C-reactive protein (CRP) as measurements of systemic inflammation, and LPS, as a measurement of bacterial translocation.

Both serum and tissue samples will be analyzed for inflammatory cytokines and chemokines. Tissue samples will be collected by a member of the study team at the time of surgery. A mucosal specimen from the stomach will be collected for SG patients and from both the stomach and jejunum for RYGB patients, flash frozen in the operating room using liquid nitrogen, and subsequently stored at -80⁰C. These specimen are removed as a standard part of the procedures. They will be analyzed for inflammatory cytokines. The investigators have established a good working relationship with operating room staff and surgeons around our cities in previous studies which will facilitate this process.

Host immune response will be assessed in samples by protein expression of cytokines and chemokines using the Meso Scale Discovery platform (MSD, Gaithersburg, Maryland USA). Using this multi-array technology will provide us with a dynamic range and the sensitivity to measure a large number of inflammatory and homeostatic signals simultaneously in a single sample. The investigators will initially focus on cytokine involved in the Farnesoid X receptor pathways, given the apparent relationship between bariatric surgery, weight loss, and bile acids17. Specifically, these cytokines include IL-1β, IL-6, IL-8, IL-12, TNFα, and MCP-1.

Analysis: A systems biology approach will be used to combine the metagenomics, metabolomics, and multiplex immune profiling to define the combined microbial, metabolic and immunologic changes that occur after bariatric surgery. Differences between continuous variables and outcome will be assessed by Wilcoxon rank sum test. Differences between categorical explanatory variables and the outcome will be assessed by the chi-square test, or Fisher's exact test when cell size is <5. Variables significant at the p < 0.10 level by likelihood ratio testing from univariate logistic regression will be entered into multivariable logistic regression. A backward stepwise selection procedure will be utilized and those variables with a likelihood ratio p-value <0.05 will be maintained in the multivariable model. QIIME (Quantitative Insights Into Microbial Ecology), MEGAN (MEtaGenome Analyzer), and Metastats, will be performed using the expertise developed at CEGIIR and in collaboration with Dr. Gane Wong, a systems biology expert.

LIMITATIONS

• Applicability of changes in bariatric surgery to weight loss in general
• Differentiating cause and effect of changes
• Assessing for effect of dietary changes following and preceding surgery

Inclusion Criteria:

• 30 severely obese controls: BMI > 35 kg/m2
• 30 severely obese patients scheduled for sleeve gastrectomy
• 30 severely obese patients scheduled for Roux-en-Y gastric bypass
• Cohorts will be BMI matched

Exclusion Criteria:

• Antibiotic, liraglutide, or methotrexate usage within two months preceding enrollment
• Meal replacement use within one month
• Previous bowel resection
• Inflammatory bowel disease
• Previous bariatric surgery