Of Microbes and Men

In this issue of Diabetes Care, Simon et al. (1) report the results of a study testing the hypothesis that the administration of Lactobacillus reuteri in subjects without diabetes improves insulin sensitivity. A quick review of the background may help us put this work in context.

Our digestive tract hosts some 2 kg of microbes, made up of ∼100 trillion microorganisms. Present in sparse colonies in the upper intestine, this indigenous microbial community (microbiota) jams into the cecum, colon, and rectum; the appendix is the ultimate shelter when antibiotics wage a mass killing. The taxonomy lists hundreds of species of bacteria (and minority groups of yeasts and viruses), whose collective genome is called microbiome (ninefold larger than our genome). Having coevolved with Homo sapiens, these microbes are so adapted to the human gut environment that they do not grow in culture as easily as do other germs and are therefore identified by their genetic material that is usually retrieved from feces.

Microbiota have long been known to train our immune system, process indigestible foodstuffs, manufacture vitamins, and break down toxins and medications. In the past decade, variants of gut microbiome have been associated with diverse human diseases ranging from enterocolitis to psychiatric disorders. Of special interest to us is the association with obesity (2) and diabetes, both type 2 (3) and type 1 (4). While learning that the ratios of Bacteroidetes to Firmicutes and of Bacteroides-Prevotella to Clostridium coccoides-Eubacterium rectale correlate with plasma glucose concentration (3) is neither immediately telling nor particularly exciting, the general questions, what are the mechanisms underlying these associations and whether the gut microbiota may be a vehicle of treatment, are relevant. In obesity, for example, the microbiome is enriched with genes encoding enzymes that break down indigestible dietary polysaccharides, thereby extracting more energy from nutrients; this heightened efficiency of caloric retention may contribute to obesity (2). Additional mechanisms involve changes in intestinal permeability to endotoxins (e.g., lipopolysaccharide [5]) and bile acid signaling (6). Whether the obese microbiome is a cause or consequence of obesity remains to be decided.

In a heroic (for the participants) proof-of-concept study, Vrieze et al. (7) infused fecal microbiota from lean, insulin-sensitive donors into the duodenum of obese subjects with metabolic syndrome and showed that this supplementation improved their insulin sensitivity. This approach, which has been used clinically to manage recurrent Clostridium difficile infection (8), is a less palatable application of a time-honored general theory, that of probiotics. Over a century ago, Metchnikoff (9) postulated that the ingestion of some microorganisms had beneficial effects for a variety of human ailments essentially by replacing pathogenic intestinal flora. Sufficient to convince my mother—who in my childhood would suavely but firmly dispense a preparation of Lactobacillus following enteritis or a course of antibiotics—the evidence supporting the theory is still unconvincing on systematic scrutiny (10). However, it is interesting that Lactobacillus—a genus of gram-positive, acid-producing bacteria with over 60 different species—has been consistently popular as a nutraceutical product. Perhaps this preference is some carryover from the circumstance that Lactobacilli are the first colonizers of germ-free neonates as they pass through the birth canal and ingest the first maternal milk.

Simon et al. (1) conducted a placebo-controlled randomized trial of L. reuteri—two oral doses of 10¹⁰ Lactobacilli per day—administered for 4 weeks to participants without diabetes who were either lean (BMI 19–25 kg/m²) or obese (BMI 30–45 kg/m²). A number of physiological end points were measured: energy expenditure and substrate oxidation rates (by indirect calorimetry), gastric emptying (by [¹³C]octanoic acid breath test), oral glucose tolerance (by the oral glucose tolerance test), incretin effect (by the isoglycemic intravenous glucose infusion), intrahepatic and intramyocellular lipids (by ¹H-magnetic resonance spectroscopy), insulin sensitivity (by the euglycemic-hyperinsulinemic clamp), endogenous glucose production (by tracer glucose), and a panel of circulating cytokines and reactive oxygen species. By analysis of fecal microbiota composition, treatment added measurable amounts of L. reuteri (∼100 DNA copies) to the constituent ecosystem (∼10¹⁴ DNA copies) without changing its composition or stability. Among the end points, the only significant differences between active-treatment and placebo groups were increases in plasma total glucagon-like peptide 1 (GLP-1) (76% on average) and GLP-2 (43%) concentrations, with the attendant changes in plasma insulin and C-peptide levels. The absolute increase in GLP-1 release was somewhat stronger than that observed with metformin (11) but was attenuated when assessed as an incretin effect with the isoglycemic protocol. No symptoms or signs of intestinal discomfort emerged during 4 weeks of treatment, but none were present before treatment. No change in glucose tolerance was observed, but subjects did not have impaired glucose tolerance to begin with.

This trial was impeccable in design, the range of the outcomes covering virtually all aspects of the hypothesis (Fig. 1); the use of state-of-the-art methods suggests high expectations by the investigators. Against this intense experimental effort, the harvest was rather meager (and perhaps transient), almost much ado about nothing. As the authors acknowledge, the study presents several limitations: the sample size was small, the trial duration short, and only one dose was tested. However, the lack of any trend or signal in the data, while it may spare other similar studies in humans, casts reasonable doubt on the hypothesis. Perhaps this probiotic only redresses an altered intestinal flora, helps recovery from an acute perturbation, or acts in concert with some other intervention. Or else, L. reuteri is the wrong probiotic for the problem at hand, i.e., simple obesity.

How valuable is the observed increase in GLP-1/GLP-2 release? It may be interesting to investigate how this stimulation came about, whether through the bacterial production of substances that enhance the expression of the preproglucagon gene in L cells (12) or by a degree of acidification or by nesting of the microbes in physical contact with L cells. Whatever the mechanism(s), the consequences (i.e., enhancement of insulin release) were insufficient to affect glucose tolerance, insulin sensitivity, subclinical inflammation, or anything else. For the time being and under the circumstances, therefore, we must halfheartedly conclude that supplements of this microbe may benefit the manufacturer but are unlikely to help either the frustrated obese patients or their hopeful doctors.