The Thin Line Between Gut Commensal and Pathogen

See allHide authors and affiliations

Science  28 Mar 2003:
Vol. 299, Issue 5615, pp. 1999-2002
DOI: 10.1126/science.1083534

Like any other organ of the human body, the gut flora of higher animals—including humans—is a complex association of cells that collectively performs essential functions (1). Not only does the highly evolved gut flora community extend the processing of undigested food to the benefit of the host, but it also contributes to host defense by limiting colonization of the gastrointestinal (GI) tract by pathogens. The GI tract consortium is, however, unique among organs in important ways. It contains more cells than the rest of the human body. The cells of the gut flora consortium are diverse, consisting of at least 500 different microbial species (most of which have yet to be cultured in vitro), and its composition varies with the organism's age, diet, and health status (1). In this week's issue, Xu et al. on page 2074 (2) and Paulsen et al. on page 2071 (3) report the genome sequences of two notable members of our gut flora: a symbiotic species, Bacteriodes thetaiotaomicron (2), and a rogue commensal-turned-pathogen, vancomycin-resistant Enterococcus faecalis (3). These studies provide molecular insights into the cooperation between human and commensal flora, and also provide clues as to how beneficial, or at least benign microbes can acquire new traits and turn against their human hosts.

Like other organs, the gut flora consortium communicates with surrounding tissues to stimulate development of a host environment that fosters implantation of the flora and nurtures its existence. However, quantitative studies of the interrelationship between human and GI tract flora have been hindered by the complexity, variability, and inability to culture many gut bacteria. Comparisons of conventionally raised and germ-free animals illustrate the contribution of the GI tract flora to animal health. Germ-free rats lacking a GI tract consortium require 30% more calories to maintain body mass than normal rats (4). Evidence suggests that the bacteria of the GI tract consortium contribute to host nutrition by liberating and generating simplified carbohydrates, amino acids, and vitamins (1). Additionally, GI tract bacteria influence the host environment by stimulating vascularization and development of intestinal villi (5), and by promoting host epithelial cell production of fucosylated glycans on which the GI tract bacterium, B. thetaiotaomicron, feeds (6). Thus, both host and flora help to create an environment that fosters the presence of the GI tract consortium, and optimizes the host's ability to capitalize on its presence.

B. thetaiotaomicron is one of the most abundant organisms in the human GI tract (7). Its genome is among the largest of bacteria thus far examined, with a large portion devoted to regulating the microenvironment surrounding this bacterium in the highly competitive GI tract (2). Because the GI tract consortium feeds on the unabsorbed remains of the diet, the B. thetaiotaomicron genome is replete with utilization pathways for breaking down undigestible carbohydrates (such as xylan and pullulan). As was predicted from previous studies, this organism's genome also encodes additional enzymes for metabolizing host-derived carbohydrates, including mucin, chondroitin sulfate, hyaluronic acid, and heparin (2). Interestingly, a majority of the glycohydrolase enzymes that break down carbohydrates appear to be localized on or in the surface layer of the bacterial cell wall (see the figure immediately below). This finding implies that B. thetaiotaomicron may liberate sugars not only for its own use, but perhaps also for the benefit of the host as well as other organisms in the immediate microenvironment (2).

Is a microbe man's best friend?

Undigested complex carbohydrate polymers are bound to the surface of the gut bacterium B. thetaiotaomicron. This organism devotes a large portion of its genome to producing surface proteins that bind to carbohydrates in the gut and surface-localized glycohydrolases that liberate simple sugars (blue dots) from complex sugar polymers. These simple sugars can be used by B. thetaiotaomicron itself, a neighboring member of the GI tract consortium (purple spheres), or be absorbed through the intestinal villus into the host bloodstream.


Enterococci, such as E. faecalis, are also prominent members of the GI tract microbial consortium. However, enterococci have gained notoriety because they can cause infections, primarily among hospitalized patients, that are extremely difficult to treat owing to antibiotic resistance (8, 9). Strains of enterococci from GI tracts of healthy humans rarely carry genetic elements conferring additional antibiotic resistance or overt virulence (8, 9). The most problematic enterococcal isolates from infected patients, namely those harboring genes for resistance to multiple antibiotics, appear to constitute a rogue subgroup of the species that, in addition to antibiotic resistance, has also acquired a number of genes conferring infectivity and virulence. Paulsen et al. (3) report the genome sequence of such an infection-derived isolate of E. faecalis, strain V583, which caused the first vancomycin-resistant enterococcal infection reported in the United States (10).

As a multiple antibiotic-resistant clinical isolate, this strain was found to be replete with mobile DNA elements, many of implied foreign origin. These mobile DNA elements include a pathogenicity island—a large mobile genetic element consisting of a number of virulence-associated genes (11)—a transposon carrying the complex of genes that mediate vancomycin resistance, three plasmids conferring resistance to other antibiotics, and a host of insertion sequences (3). These mobile elements constitute over a quarter of the genome of this strain. The occurrence of the pathogenicity island on the side of the chromosome opposite that of the vancomycin resistance transposon, as well as its occurrence in strains that predate the acquisition of vancomycin resistance (11), strongly suggest independent selection for antibiotic resistance and virulence traits.

Developing an understanding of how multiple antibiotic-resistant enterococcal infections occur is of particular importance in devising new strategies for their prevention and treatment. Evidence suggests that within days of admission to a hospital, patients acquire multiple antibiotic-resistant enterococcal strains in their GI tract consortium from contaminated objects and surfaces (8, 9). This is often not the result of direct antibiotic elimination of the indigenous commensal enterococci and selection for resistant strains, because many antibiotics that predispose patients to subsequent enterococcal infection have relatively little anti-enterococcal activity. Antibiotics may, however, destabilize the highly evolved GI tract consortium by eliminating susceptible organisms of other species, opening new niches for colonization by resistant enterococcal strains (see the figure below). It is then a question of whether colonization of the GI tract by antibiotic-resistant strains is competitive with the indigenous, less- resistant commensal enterococci. Because of the numerical advantage of indigenous enterococci within the healthy consortium, colonization of the GI tract by a few resistant cells would most efficiently result from a noncompetitive process.

Friend turned foe.

(A) Commensal strains of enterococci lacking overt antibiotic resistance plasmids and the pathogenicity island are prominent members of the GI tract flora. (B) Antibiotic use destabilizes commensal flora by killing susceptible strains. (C) Multiple antibiotic-resistant enterococcal strains (magenta), such as E. faecalis strain V583, are able to colonize new niches that may not be available to commensal strains, because they have acquired surface protein adhesins, bile acid hydrolase, carbohydrate utilization pathways, and other properties that confer fitness for a particular habitat. The GI tract then serves as a staging ground for infecting the bloodstream, urinary tract, and surgical wounds.


What new traits would an enterococcus require to enable it to colonize the gut in a manner that is noncompetitive? Traits predicted to influence selection of a habitat by a microorganism include new pathways for carbohydrate metabolism, new surface proteins that enable attachment to different host epithelial cell types, production of antimicrobial molecules such a bacteriocins, and development of activities that limit host defenses. These are precisely the new activities that are encoded by the enterococcal pathogenicity island (3, 11). This pathogenicity island encodes a cytolysin that has antibacterial properties, several surface adhesins, several new carbohydrate utilization pathways, and enzymes such as a bile acid hydrolase, which may permit colonization of areas of the intestine closer to the bile duct. From a successfully colonized GI tract, multiple antibiotic-resistant enterococci are well positioned to cause the urinary tract, bloodstream, and surgical-wound infections frequently observed in hospitalized patients.

Human and bacterial genome studies are beginning to reveal the underlying basis for the highly coevolved relationship between humans and microbes. On the basis of emerging information, testable models can be advanced for how the antibiotic-resistant pathogens of the 21st century are able to infiltrate the GI tract consortium and use it as a staging ground for causing infections that are increasingly difficult to treat. With more than 500 microbial species present in the GI tract consortium, publication of the genome sequences for a leading commensal strain (2) and a “commensal gone bad” (3) represents only a beginning. The insights gained into the biology of this important “organ” are a clear testament to the fertility of this line of pursuit.


View Abstract

Stay Connected to Science

Navigate This Article