Roux-en-Y gastric bypass surgery of morbidly obese patients induces swift and persistent changes of the individual gut microbiota

Altered microbial functions after RYGB

To characterize the changes in functional potential of microbes to adapt to the gut rearrangement after RYGB, we estimated the relative abundances for KEGG modules and pathways in each sample. We found 62 KEGG modules that changed in relative abundance between baseline and 3 months after RYGB (Wilcoxon signed-rank test, P??0.05, q??0.17, suggesting that up to 11 modules could be false positives; Additional file 1: Figure S8; Additional file 2: Table S3), and 63 KEGG modules that changed between baseline and 1 year (Wilcoxon signed-rank test, P??0.05, q??0.16, suggesting that up to ten could be false positives; Additional file 1: Figure S9; Additional file 2: Table S4), while we did not observe significant changes between 3 months and 1 year (q??0.86; Additional file 1: Figure S10). Most of these changes (53 out of 62 in the former and 56 out of 63 in the latter) reflected an increase in relative abundance over time (Fig. 4), which may reflect the increased species richness after RYGB. Of the 53 modules that increased their abundance within the first 3 months, 44 sustained it for a year. Thus, the functional changes mirrored the taxonomic changes, where most changes had occurred during the first 3 months and were merely maintained during the following 9 months.

Microbial functional changes following RYGB. Box plots represent fold changes (log2) in the 53 KEGG modules that increased in relative abundance between baseline and 3 months (3MO, left panel) and 44 among these that increased between baseline and 1 year (1Y, right panel) after RYGB. The different KEGG functional categories are represented by different colors and grouped together when possible with corresponding labels at the right side of the plot. PTS phosphotransferase systems, GABA gamma-aminobutyric acid

Among the KEGG modules that increased their abundance after RYGB, we observed several nutrient transport systems (Fig. 4). Morbidly obese patients who have undergone bariatric surgery have a smaller stomach pouch and their food intake skips the duodenum, which causes malabsorption of essential vitamins, minerals, and drugs [61]. To compensate for this malnutrition, patients follow a diet rich in proteins and take calcium, iron, and multivitamin supplements (see “Methods”). We observed an increased potential for microbial transport systems of thiamine, vitamin B12, manganese, iron, and zinc (Fig. 4), which could reflect the increased availability of these compounds that are also essential for microbes. Transport systems of phosphonates were also increased after RYGB. Some bacteria such as E. coli and Klebsiella spp. can utilize these compounds as an alternative source of phosphorus by breaking their C–P bonds [62].

There was also an increased abundance of transport systems for monosaccharides such as D-xylose, rhamnose, D-allose, and L-arabinose after RYGB. In E. coli, the presence of L-arabinose in the absence of glucose dynamically activates the operon that drives the catabolism of arabinose [63]. Arabinose is present in rice, wheat, beans, oats, or plant polysaccharides. Thus, our finding may reflect changes in diet, for example, a shift in food preferences towards lower-calorie-dense foods as reported after RYGB [64–66].

Other transport systems that increased in abundance after RYGB are the phosphotransferase systems (PTS; Fig. 4), which are only found in bacteria. PTS catalyze the transport and phosphorylation of numerous monosaccharides, disaccharides, amino sugars, polyols, and other sugar derivatives into the bacterial cell. Their increase could be attributed to an increased ability of microbes to assimilate all available sugars to compensate for the reduced dietary intake. Figure 4 also shows the increased potential of amino acid uptake, suggesting the utilization of amino acids as a source of energy, and an increased potential for beta-oxidation of fatty acids, indicating the utilization of these fatty acids as a source of energy.

The KEGG module for cytochrome c oxidase complex and the module for prokaryotic biosynthesis of ubiquinone also increased in abundance after RYGB. The former is the last enzyme of the electron transport chain in both bacteria and eukaryotic mitochondria. Ubiquinone, known as coenzyme Q₁₀, also plays a crucial role as an electron carrier in the electron transport chain. The increased relative abundance of these two modules together with the increase of facultative anaerobes such as E. coli, K. pneumoniae, E. faecalis, and Streptococcus spp. suggest a shift towards aerobic respiration among the facultative anaerobes to benefit from a higher presence of dissolved oxygen in the hindgut after RYGB [57]. Electron transport chains are major sites of premature electron leakage to oxygen, generating superoxide and potentially resulting in increased oxidative stress. Post-RYGB, we observed an increased abundance of a module encoding glutathione biosynthesis from glutamate and an increased abundance of transport systems of both glutamate and glutathione (Fig. 4). In bacteria, glutathione, in addition to its key role in maintaining the proper oxidation state of protein thiols, also protects the cell from oxidative and osmotic stress [67]. Thus, the increased capacity in glutathione biosynthesis and transport suggests that the gut microbes may be using glutathione to combat oxidative stress.

Intriguingly, following RYGB we observed an increase in abundance of two KEGG modules involved in putrescine transportation (Fig. 4). Although protein digestion is not impaired after RYGB [68], the increased potential for putrescine transportation might indicate a certain level of putrefaction in the colon, as other authors have hypothesized when studying the fecal metabolic profiles after RYGB in a non-obese rat model [36]. Fast pouch emptying and a delayed small intestinal transit time have been reported for RYGB patients [69]. Reduction of gastric acid secretion after surgery [70, 71], which plays a key role in protein digestion by activating proteolytic enzymes, might cause more incompletely digested proteins to reach the colon. Under these circumstances, a longer intestinal transit time could provide enough time for microbes to catabolize these proteins, resulting in the production of polyamines such as putrescine [36, 72], which is involved in key functions such as DNA and membrane stabilization but becomes toxic at high doses and can even produce carcinogenic nitrosamines [73]. Putrescine has been found in elevated concentrations in fecal samples post-RYGB in rats [36]. Enterobacteriaceae spp. such as E. coli or Klebsiella spp. can produce putrescine by decarboxylation of the amino acids ornithine and arginine [73], of which we also observed an increased transportation potential after RYGB (Fig. 4). The increased capacity to transport putrescine could also reflect the antioxidant and anti-inflammatory properties of this polyamine for microbes when oxidative stress increases [74–76]. Microbial processing of putrescine can produce gamma-aminobutyric acid (GABA) [77], which is an inhibitory neurotransmitter of the mammalian central nervous system and has been found increased in fecal samples after RYGB in a rat model [36]. This neurotransmitter is thought to stimulate the intestinal cells to release GLP-1 [78, 79]. Increased levels of GLP-1 observed after RYGB in our cohort (Fig. 1; Additional file 1: Figure S3) is consistent with this link. The increase in GLP-1 can, in turn, stimulate the biosynthesis of GABA via pancreatic beta-cells [80]. An increased capacity for GABA biosynthesis and GABA shunt (closed loop to produce and maintain the supply of GABA) pathways observed after RYGB (Fig. 4) provides further evidence for this metabolic path after RYGB.