The Influence of Ecological Factors on the Distribution and the Genetic Structure of Escherichia coli

David M. Gordon

[Section Editor: Monica Riley]

Posted February 27, 2004

School of Botany and Zoology, Australian National University, Canberra, ACT 0200, Australia

Phone: 61 2 6125 3552, Fax: 61 2 6125 5573, E-mail: This e-mail address is being protected from spambots. You need JavaScript enabled to view it

INTRODUCTION

Since the first print version of Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology in 1987, the development of new molecular techniques and the decline in cost of most molecular methods have resulted in a rapidly expanding database covering the genetic structures of bacterial populations. Consequently, while in 1987 a single chapter sufficed to describe the genetic structure of Escherichia coli, our knowledge base has grown to such an extent that several chapters are now required.

This chapter focuses on recent data concerning the ecological factors that determine the distribution of E. coli and the genetic structures of naturally occurring E. coli populations. Much of the information presented is descriptive. It will become clear that ecological factors play a significant role in shaping the genetic structures of E. coli populations. However, we have little understanding of the precise nature of the selective forces involved and even less understanding of the natures of the adaptations that enable E. coli to persist in the variety of habitats in which it is found.

The characteristics and evolution of the various pathogenic lineages of E. coli are described elsewhere in the EcoSal website. However, some references to the pathogenic E. coli isolates are made in this chapter because the ecological factors that have been found to affect the distribution and genetic structure of commensal E. coli are relevant to our understanding of the epidemiologies of the various E. coli pathotypes. Consequently, this chapter summarizes some of the older literature concerning the dynamics of E. coli populations within a host and poses some questions that arise from our more recently acquired understanding of the factors affecting the genetic structures of E. coli populations.

WHERE IS E. COLI FOUND?

Gordon and Cowling (19) reported the results of a survey of >2,300 nondomesticated vertebrates living in Australia. The authors defined the prevalence of E. coli as the fraction of hosts in whose feces E. coli could be detected. The method of isolation was such that E. coli would be detected if it represented, on average, at least 1% of the bacterial-cell population capable of growth on a MacConkey agar plate.

E. coli was found in 56% of the mammalian hosts examined. Mammal species live in different climates, and their body masses (log₁₀) and diets differ. Each of these factors was found to be a statistically significant predictor of the presence of E. coli in an individual host. Overall, E. coli was unlikely to be isolated from hosts living in the desert, while hosts living in the tropics were less likely to harbor E. coli than hosts living in temperate or grassland environments. In general, E. coli is less likely to be isolated from carnivores than from omnivores or herbivores. The likelihood of isolating E. coli from a mammal increases with the body mass (log₁₀) of the host (Fig. 1).

Fig. 1Relationship between prevalence of E. coli in carnivorous Australian mammals and body mass (log ₁₀ grams) of the host species. The solid line depicts the relationship between body mass and prevalence as determined from the data for 209 individuals belonging to the order Dasyuromorphia. The points represent the observed prevalences for carnivorous species, where n is >5. The circles indicate species in the order Chiroptera, the squares indicate species in the order Dasyuromorpha, and the triangles indicate species in the order Monotremata.

E. coli was detected in 23% of the birds examined. A significant number of the birds were collected from localities where they lived in close association with humans, while others were collected from localities without significant levels of human habitation. Diet varies substantially among species, and although not as extensive as in mammals, interspecies variation in body mass is considerable. Human association, diet, and log₁₀ body mass were found to be significant predictors of the prevalence of E. coli in birds. Birds living in close association with humans were approximately twice as likely to harbor E. coli than birds living away from human habitation. The probability of a host harboring E. coli increased with its body mass. The prevalence of E. coli was lowest in exclusively seed- or fruit-eating species and higher in those species that include nectar, insects, or vertebrates in their diet.

In general, E. coli was unlikely to be detected in ectothermic vertebrates. E. coli was isolated from 10% of the fish and from 12% of the frogs examined. In reptiles, E. coli was isolated from 33% of the crocodiles, 4% of the turtles, 2% of the snakes, and 10% of the lizards examined. As with birds, association with humans also appeared to explain some of the variation in the prevalence of E. coli among frog and lizard species. E. coli was unusually common in individuals of two species of lizards living in inner Sydney, New South Wales, and two species of tree frog living in the suburbs of Cairns, Queensland.

Climate and the Prevalence of E. coli

E. coli is less likely to be isolated from hosts living in tropical or desert climates than from hosts living in grassland or temperate climates. This pattern of variation was consistent across host orders and does not appear to be a host effect, as the prevalence of E. coli in species of the same genus differed depending on the climatic zone. Why E. coli is less prevalent in mammals living in the tropics is unknown. Maximum summer temperatures in the desert regions of central Australia are typically in excess of 35°C, the relative humidity is <25%, and median rainfall is <20 mm a month. It seems likely that these conditions would adversely affect the survival of E. coli in the external environment and may, in turn, limit the transmission of E. coli among individual hosts.

Alternatively, the near absence of E. coli in the feces of desert-dwelling mammals may be a consequence of host biology. Mammals living in arid regions have adaptations that enable them to conserve water. While water loss through defecation represents a small fraction of an animal’s water balance, morphological modifications in the colon to maximize water retention do occur (38, 60, 65). It may be that the lower water content of material in the distal colon adversely affects the survival of E. coli cells in this region, and as a result, cell densities are reduced to such an extent that E. coli is unlikely to be detected in feces.

Host Biology and the Prevalence of E. coli

The likelihood of a host harboring E. coli depends primarily on three factors. The first is the frequency with which a host individual is exposed to E. coli. The second is the probability that an exposure event will result in the establishment of a population. The third is the average length of time the E. coli population can persist in the host, or the rate at which the host loses its population of E. coli.

The observation that some hosts are more likely to harbor E. coli if they are commensally associated with humans than if they are not indicates that exposure rates vary among hosts. Many species of birds, such as house sparrows (Passer domesticus) and starlings (Turdus merula), are fed on household food scraps or forage in compost heaps that are likely to contain E. coli. In addition, these higher exposure levels may also be due to elevated levels of environmental contamination by E. coli due to the feces of domestic pets and livestock, hosts in which E. coli is prevalent. Elevated levels of environmental contamination may also lead to increased carriage of E. coli by invertebrates, which would explain the atypically high prevalence of E. coli in frogs and lizards living near human habitation.

It is unlikely that the extent to which a host is exposed to E. coli is the sole factor responsible for the variation in prevalence observed among different species. The variation may also be a consequence of host characteristics that affect the ability of E. coli to establish a population in the gastrointestinal tract.

Carnivorous mammals have relatively simple guts, because animals represent an energy-rich and easily digested food source. The digestive tract of carnivores is dominated by the small intestine, while the colon is short and often poorly differentiated from the small intestine. Among Australian carnivores, a cecum is usually lacking or poorly developed (25). Short intestines, coupled with high-quality food, result in rapid gut transit times. For example, in the carnivorous marsupials (Dasyuridae), gut transit times increase with body size and vary from about 1 h for the 18-g Sminthopsis crassicaudata to 13 h for the 1,000-g Dasyurus viverrinus (25).

By contrast, plant tissue represents an energy-poor food source that is difficult to digest. Microbial fermentation enables mammals consuming plant tissue to increase the nutritional value of the material they ingest, but it requires the presence of a fermentation chamber. In kangaroos and wallabies (Macropodidae), as well as the rat kangaroos (Potoroidae), the foregut provides the primary site of microbial fermentation, although these species also have a cecum (25). The hindgut is the site of microbial fermentation in the balance of Australian herbivores(25). In the majority of small (<10-kg) herbivores, the cecum forms the fermentation chamber, while in the larger species, fermentation takes place in the colon (25). Retention times are significantly longer in herbivores than in carnivores and, as in carnivores, increase with body mass. For example, food transit times are about 35 h in the 1-kg common ringtail possum (Pseudocheirus peregrinus) and about 100 h in the 10-kg koala (Phascolarctos cinereus) (25).

The digestive tracts of Australian omnivores are intermediate between those of carnivores and herbivores (25). In the small, primarily insectivorous mountain pygmy possum (Burramys parvus), the cecum is very small and the colon is short, while the sugar glider (Petaurus breviceps) has a well-developed cecum. The bandicoots have a modest-sized cecum and colon (25). All Australian rodents possess a cecum (9). In omnivores, food transit rates also depend on body mass and fall between those of carnivores and herbivores (25).

Several mathematical models have been developed with the purpose of exploring the dynamics of bacterial populations in the gastrointestinal tract. These models have investigated the process of bacterial establishment using chemostat models (2, 3, 16, 57, 59, 61) and plug flow reactor models (1, 2, 32). The chemostat models describe a physical system most closely resembling that of a cecum, while the plug flow reactor models describe a typical carnivore gut. These models have shown that establishment is more likely when the bacterial population exhibits wall growth than if no wall growth can occur. Nutrient concentration is also a factor, and there is a threshold concentration of nutrients entering the model gut below which the bacterial population cannot establish itself and above which the population will be established and persist. Further, the rate at which material moves through the gut (transit time) can also determine whether a bacterial population can become established: decreasing transit times lead to populations being less likely to colonize.

The quantity and type of nutrients entering the lower gastrointestinal tract will vary with the host diet. Gut turnover rates and transit times will also vary with diet, gut morphology, and body mass. The theoretical models suggest that these factors may well influence the likelihood of E. coli being able to establish a population in a host and hence determine the prevalence of E. coli in a particular host species. Further experimental and theoretical work is required in order to determine if one of the factors—host diet, gut morphology, or body mass (food transit rate)—is more likely to dominate the dynamics of bacterial colonization in the gastrointestinal tract.

To what extent the residence time of an E. coli population varies with diet, gut morphology, or body mass is unknown. There have been few studies that have examined the persistence time of an E. coli strain in a host, and these studies have largely been restricted to humans (see below). The results of these studies indicate that a strain may persist for months, if not years. However, the data also suggest that, inevitably, the resident strain is eventually lost. The factors responsible for the turnover of strains in a host are unclear. Long-term longitudinal studies of the persistence of E. coli in different host species are required.

The results of the survey of the distribution of E. coli in Australian vertebrates (19) indicate that, as a species, E. coli is adapted to mammals with hindgut microbial fermentation chambers, or in the absence of a hindgut fermentation chamber, the host must be "large." The contrast between rodents and carnivores highlights the apparent importance of a hindgut fermentation chamber. E. coli is common in most rodent species, all of which posses a cecum. E. coli is uncommon in bats and many marsupial carnivores, all of which lack a cecum. Yet, many species of rodent are at least as small as bats and smaller than many of the species of marsupial carnivore in which E. coli is uncommon. The observation that the prevalence of E. coli in carnivores increases with host body mass indicates that a large body mass can compensate for the lack of a hindgut fermentation chamber.

THE GENETIC STRUCTURE OF E. COLI

Multilocus enzyme electrophoresis (MLEE) studies indicate that E. coli, relative to other members of the family Enterobacteriaceae, exhibits a moderate degree of genetic diversity. For example, in a study of isolates from native Australian mammals, E. coli exhibited less MLEE diversity than Citrobacter freundii, Enterobacter cloacae, Hafnia alvei, or Klebsiella oxytoca and a level of diversity comparable to that of Klebsiella pneumoniae (21, 41). However, such comparisons are fraught with difficulties, as the degree of genetic variation observed can depend very much on the precise nature of the samples being compared due to host and geographic structures (21, 41).

Asexual reproduction (vertical gene transmission) gives rise to bacterial lineages that are related by descent (58). New lineages arise as a consequence of mutation. Bacteria also have the ability to exchange genetic material with cells of different strains or species through the processes of conjugation, transformation, and transduction (58). Gene transfer results in genetic diversification through the introduction of new genetic material into a lineage, but gene transfer can also lead to genetic homogenization as genetic material becomes distributed among lineages. Thus, the genetic structure of a bacterial population or species is the net result of the processes of mutation and gene transfer coupled with the outcomes of selection. When mutation is the predominant process, bacterial populations are highly clonal, as evidenced by significant levels of linkage disequilibrium (the nonrandom association of alleles among loci). By contrast, when gene transfer predominates (10 to 20 times the mutation rate [35]), the genetic structure of the population is said to be panmictic and linkage disequilibrium is not observed.

It is now well established that there is extensive variation in genetic structure among bacterial species (15). Species like Neisseria gonorrhoeae and Helicobacter pylori show little indication of clonality. Neisseria meningitidis and Streptococcus pneumoniae are species that are considered to be weakly clonal. By contrast, species such as Salmonella enterica are highly clonal, and mutation is the dominant force shaping this species’ genetic structure. E. coli also has a clonal genetic structure, although not to the extent exhibited by S. enterica (58).

Early MLEE studies demonstrated the existence of "subspecific" structure in E. coli (64). More extensive MLEE analysis, together with other techniques of genetic analysis, has confirmed the existence of this subspecific structure (12, 24, 33, 39). At present, there are four recognized subspecies of E. coli, which have been designated A, B1, B2, and D. A and B1 are considered to be sister groups, and group B2 is considered to represent the "ancestral lineage" of E. coli (33). It appears that the great majority of E. coli strains isolated from feces can be assigned to one of these four groups. Recently, a PCR-based method has been developed that enables an E. coli isolate to be assigned to one of the four E. coli groups by using a dichotomous-key approach based on the presence or absence of two genes (chuA and yjaA) and an anonymous DNA fragment (10).

The existence of subspecific structure in E. coli has for the most part been determined by assessing genetic variation that is thought to be largely neutral in its effects on the fitness of a strain (54). However, strains of the four groups of E. coli also differ in their phenotypic characteristics, including their abilities to exploit different sugars, their antibiotic resistance profiles, and their growth rate-temperature relationships (D. M. Gordon, unpublished).Genome size varies among the four E. coli groups, with A and B1 strains having smaller genomes than B2 or D strains (4). The distributions (presence or absence) of a range of putative virulence factors thought to be involved in the ability of a strain to cause extraintestinal disease also vary among the four E. coli groups (30).

Host Biology and the Distribution of A, B1, B2, and D Strains in Vertebrates

Host diet and the climate where the host was collected were found to account for a significant amount of the variation in the distribution of A, B1, B2, and D strains among nondomesticated mammalian hosts (19). Host body mass also explained a portion of the variation, but the nature of the relationship between body mass and the distribution of the four E. coli groups depended upon the host diet. In birds, no statistically significant effects of host body mass, diet, or human association on the distribution of A, B1, B2, and D strains could be detected, and too few strains were isolated from ectothermic vertebrates for any meaningful statistical analysis (19).

In mammalian herbivores, the frequency of group A strains declined with increasing body mass (Fig. 2). B2 strains predominated in herbivores weighing <6 kg, but as body size increased beyond 6 kg, there was a sharp decline in the frequency of B2 and a concomitant increase in the frequency of B1 strains. In the omnivorous mammals, the frequency of B2 strains increased with increasing host body mass and the frequencies of all other groups declined (Fig. 2). In carnivores, the frequency of group A strains increased with host body mass, and there was a decline in the frequencies of all other groups (Fig. 2). However, the significant effect of carnivore body mass was entirely due to a single species, the Tasmanian devil (Sarcophilus harrisii). Although Tasmanian devils represented only 37% of the carnivores from which E. coli were recovered, 76% of all group A strains isolated from carnivores were from this species. Thus, compared to other carnivores, group A strains were significantly overrepresented in Tasmanian devils.

Fig. 2Predicted change in frequency of strains of the E. coli groups A, B1, B2, and D with body mass (log ₁₀) for carnivorous, omnivorous, and herbivorous mammalian hosts. Red, E. coli reference (ECOR) group A; blue, ECOR group B1; pink, ECOR group B2; green, ECOR group D.

The relative frequencies of A, B1, B2, and D strains in the different vertebrate host groups are depicted in Fig. 3 (19). The relative abundances of the four groups were similar in mammalian herbivores and omnivores, but the patterns seen in herbivores and omnivores differed from that found in carnivores. The distributions of the four groups in birds and carnivorous mammals were similar. Ectothermic vertebrates had a distribution of the four E. coli groups very different from the patterns seen in birds and carnivorous mammals or from the distribution observed in mammalian omnivores or herbivores.

Fig. 3Observed frequencies of group A, B1, B2, and D strains of E. coli in different vertebrate host groups. The bars represent group A, B1, B2, and D strains, as indicated.

In ectotherms, birds, and carnivorous mammals, hosts which all have relatively simple tube-like gastrointestinal tracts, group B1 strains predominate. In hosts with a cecum, the omnivorous and herbivorous mammals, group B2 strains predominate, while in the mammals, the relative frequencies of A, B1, B2, and D strains vary with the host body mass and diet. Gordon and Cowling (19) argue that these observations support the conclusion that the observed interspecies variation in the prevalence of E. coli is due in part to the interaction between gut dynamics and bacterial establishment and is not solely a consequence of interspecies differences in exposure to E. coli.

Geography and the Genetic Structures of E. coli Populations

The results of MLEE studies suggest that there is little geographic differentiation among E. coli populations. Geographic differentiation accounted for about 2% of the allelic variation observed in a collection of E. coli isolates from humans living on four continents (64). In E. coli isolates from nondomesticated animals collected throughout Australia, 5% of the variation could be attributed to geographic effects (21), while 1% of the variation observed in E. coli populations isolated from humans living in different North American cities could be attributed to geographic differentiation (8). Similarly, locality accounted for only 2% of the variation observed in E. coli isolates from house mice living in two populations 15 km apart (17).

By contrast, large differences in the relative frequencies of A, B1, B2, and D strains have been observed in E. coli populations isolated from humans living in different regions (13; Gordon, unpublished). The frequencies of D strains in human feces appear to be relatively constant in different regions, but there are large differences in the relative abundances of group A, B1, and B2 strains (Fig. 4). The reasons for the differences are unknown, although climate may play some role. In a study of E. coli isolates from Australian mammals, the relative abundances of A, B1, B2, and D strains were found to vary significantly with the climate of the localities where the hosts were collected (19). Does diet matter? Certainly there are substantial differences among the diets of a typical French person, Australian, Croatian, or member of the Dogon people in Mali.

Fig. 4Observed frequencies of group A, B1, B2, and D strains of E. coli isolated from feces from people living in different countries ( ; Gordon, unpublished). The bars represent group A, B1, B2, and D strains, as indicated.

Primary versus Secondary Habitats of E. coli

Savageau (49) observed that the life cycle of E. coli involves a transition between two distinct environments. He considered the primary habitat of E. coli to be the gastrointestinal tract, while soil, water, and sediment represent the species’ secondary habitats. It has been estimated that a typical E. coli cell spends one-half of its life in an environment external to the host (23, 49). Two hypotheses have been proposed to describe the manner in which E. coli responds to the transition from its primary to secondary habitats. Savageau (49) suggested that E. coli cells cope with the transition by possessing a dual regulation system, in which genes with products in high demand are under positive control while genes with products in low demand are under negative control. The control systems are hypothesized to alternate depending on the specific demands imposed by the primary and secondary environments. Savageau (49) presented numerous examples of physiological functions in enteric bacteria that are consistent with the demand theory, and it has been demonstrated that in E. coli some genes are preferentially expressed in the external environment (14).

Whittam (63) suggested that selection is the dominant force acting on E. coli during the transition from primary to secondary habitats. That is, there are strains of E. coli primarily adapted to the host environment, while other strains are better adapted to the external environment. Whittam based this conclusion on the results of a study of E. coli isolated from domestic birds and the litter on which these birds were raised. Of the 113 distinct MLEE types identified, only 10% were recovered from both habitats. Further, the data revealed two genetically distinct groups of strains: the majority of isolates taken from the birds made up one cluster, while the majority of strains isolated from the environment made up a second cluster. Gordon et al. (22) observed a similar outcome. E. coli was sampled from two two-person households and their associated septic tanks. The only fecal inputs to the septic tanks were from the household members. In one of the households, a single MLEE type represented >90 % of the isolates recovered from the septic tank, and this strain could not be detected in either household member. Further, the growth rate-temperature responses of strains isolated from this household’s septic tank and strains isolated from the household members differed. The septic tank isolates grew better at low temperatures, but more slowly at high temperatures, than the strains isolated from humans.

The consequences for E. coli of the transition between its primary and secondary habitats are of considerable practical significance for water quality assessment (18) and disease transmission (47). Although the nature of the external environment in which E. coli finds itself influences its distribution and genetic structure, we have a very poor understanding of how this affects the transmission dynamics of E. coli or how transmission dynamics vary among strains of the different groups of E. coli.

Disease; the Distribution of A, B1, B2, and D Strains; and the within-Host Dynamics of E. coli

E. coli causes a significant fraction of human bacterial disease (56) and is responsible for two main types of disease in humans and domestic animals: diarrheal disease and extraintestinal infections. The strains responsible for diarrheal disease have been divided into different pathotypes: enterotoxigenic, enteropathogenic, enterohemorrhagic, enteroinvasive, enteroaggregative, and diffusely adherent E. coli and the various "Shigella" pathotypes. E. coli causes extraintestinal disease when it invades a normally sterile body site, and the range of clinical syndromes caused by E. coli is largely related to the part of the body that has been invaded. The kinds of extraintestinal disease caused by E. coli include cystitis, pyelonephritis, septicemia, and meningitis.

The great majority of strains responsible for extraintestinal infection are group B2 strains, and to a lesser extent group D strains (30, 31, 33, 34, 43). By contrast, strains from groups A and B1 are seldom responsible for extraintestinal disease (42, 43). Strains responsible for intestinal disease are most often derived from groups A and B1, and to a lesser extent group D, and do not appear to be derived from group B2 strains (44, 45, 48).

There is little understanding of the epidemiological factors that lead to infections that cause disease. The E. coli cells responsible for urinary tract infection are thought to originate in the intestinal E. coli community of the infected host. One view is that strains responsible for urinary tract infections are those that are numerically dominant in the intestine at the time of infection (37). Another view is that urinary tract infections are caused by a subset of strains present in the gastrointestinal tract which express a particular suite of virulence traits (37). As previously stated, strains of the B2 group of E. coli are far more frequently found to be the cause of extraintestinal infection than strains of the other E. coli groups. Group B2 strains are also more likely to possess those traits thought be responsible for a strain’s ability to cause extraintestinal disease (6, 28, 31). However, the frequency with which group B2 strains represent the dominant strain in a human host varies considerably among populations (Fig. 4). Do populations with a high frequency of B2 strains suffer from an elevated risk of urinary tract infections?

Many of the E. coli pathotypes responsible for intestinal disease can cause disease at very low infective doses, yet not all hosts exposed to these pathotypes develop disease. The factors responsible for this variation are largely unknown. However, there is a growing body of evidence that some bacterial species can protect the host from intestinal disease, presumably by reducing the likelihood that the pathogenic strain will successfully establish a population in the gastrointestinal tract of the host (62). There is some evidence to suggest that E. coli can limit the establishment of other members of the Enterobacteriaceae. Gordon and FitzGibbon (20) presented data showing that there was a negative association between the presence of E. coli in a host and the presence of most other species of Enterobacteriaceae. Are there commensal strains of E. coli that can limit the establishment of E. coli diarrheal pathotypes?

In a human host, E. coli typically occurs at a density of 10⁷ to 10⁸ cells per g of feces (23). When multiple isolates from a single human fecal sample are examined, the majority of the isolates are identical (7, 8, 22, 50). However, from 1 to 14 electrophoretic types can be recovered from a single sample, and the relative abundances of the electrophoretic types can vary substantially. For example, the probability of choosing two isolates and finding them to be different can vary from 0 to 0.91. Similar results have been seen when multiple isolates from individual nonhuman hosts have been characterized (36, 46). The most abundant clone present in a fecal sample has been called the dominant clone. However, the precise definition of what constitutes a dominant clone varies between studies. For example, Schlager et al. (50) defined the dominant clone as the clone which represented >50% of the isolates, while Gordon et al. (22) defined the dominant clone as any clone that could be recovered through nonselective plating. Is there a relationship between the clonal diversity observed in a host and the E. coli group represented by the dominant strain? For example, is clonal diversity typically lower in hosts in which a B2 strain is dominant and higher in hosts in which a group A strain predominates?

Even when a relatively large number (e.g., 100) of isolates are selected from a sample plated on nonselective agar, it is unlikely that a variant that represents <1% of the total cell population will be detected. Other techniques are required to detect variants occurring at lower frequencies (22, 29). Strains occurring at low frequencies have been designated minor, or minority, variants, although again, the definition varies among studies. Schlager (50) defined any variant that represented <50% of the isolates as a minor variant, while Gordon et al. (22) defined a minority variant as any variant recovered through selective plating. In the study of Gordon et al. (22), minority strains were recovered from both male and female human hosts. However, analysis of the MLEE data showed that in female human hosts, minority strains represented a population of strains distinct from the dominant strains. In the two males, the minority strains recovered were not distinct from the dominant strains found in these hosts. Are strains present at low frequencies in a host more likely to be members of one of the four groups of E. coli?

Few studies have examined the manner in which the clonal composition of an individual host changes with time (7, 26, 51, 52, 53, 55). Overall, variation through time accounts for about 25% of the observed variation in the clonal composition of a host (8, 22, 63). However, the magnitudes of temporal effects can vary considerably among individuals (22). For example, in four hosts examined twice, at either 7- or 11-day intervals, between-sample variation account for 0 and 14% of the observed clonal diversity in the two adult males examined, while temporal effects accounted for 26 and 85% of the observed clonal diversity in the two adult females. Variants recovered repeatedly over several samples have been called resident strains, while those variants recovered only once or from a few consecutive samples have been called transient strains. Does the typical residence time of a strain vary depending on whether the strain is a member of group A, B1, B2, or D? Does the rate at which transient strains appear at detectable frequencies over time vary depending on the E. coli group membership of the resident strain? If this is the case, can these strains prevent or limit the establishment of strains responsible for diarrheal disease?

The theoretical and empirical evidence suggests that the dynamics of the gastrointestinal tract influence both the probability of isolating E. coli from a host and the type of E. coli found in the host. The dynamics of the gastrointestinal tract differ between human males and females. For example, gut transit times are about 50% longer in females than in males (11). Although there is very limited evidence, there is some suggestion in the available data (22) that the clonal composition and dynamics of E. coli populations in the gut differ between human males and females. If this is in fact true, are the traits that enable a strain it to inhabit a female’s gut also traits which might "predispose" the strain to cause urinary tract infection?

The Ecological Niches of A, B1, B2, and D Strains—More Questions than Answers

The observed distribution of strains from the different E. coli genetic groups indicates that they have different life history tactics and ecological niches. A and B1 strains appear to be generalists, as they can be recovered from any vertebrate group. Group B2 and D strains appear to be more specialized, as they are largely restricted to endothermic vertebrates. At least in nonhuman mammals, group B2 strains appear to be the most specialized, as they predominate in hosts with a hindgut fermentation chamber. It is likely that there are many factors responsible for the different distributions of E. coli genetic groups among host groups. The observed distribution of A, B1, B2, and D strains may reflect intergroup differences in the ability to establish a population in the gut, as appears to be the case for B2 and D strains in ectotherms. Competitive interactions among strains within a host may also play a role. B1 strains can be isolated from hosts of any vertebrate group but predominate in hosts with a simple hindgut, while B2 strains predominate in hosts with hindgut fermentation. Does this outcome occur because B1 strains are primarily adapted to vertebrates with a simple hindgut, or are they competitively inferior to group B2 strains in hosts with hindgut fermentation? What are the attributes of D strains that apparently restrict them to homeotherms, and are they the same traits that restrict B2 strains to homeotherms? Why do D strains typically represent only 15 to 20% of the E. coli strains found in homeotherms? Why does the relative abundance of D strains remain the same regardless of the host species or geographic locality while the relative abundances of A, B1, and B2 strains vary substantially?

To what extent do the transmission biologies of A, B1, B2, and D strains differ? B2 strains are predominantly responsible for extraintestinal infection, and particularly in the case of urinary tract infections, the strains responsible for infection are thought to originate from the intestinal E. coli community of the infected host. By contrast, food and water are thought to be the main sources of strains responsible for intestinal infection, and most of these diarrheal pathotypes are derived from A, B1, and D strains. Are B2 strains primarily adapted for host-to-host transmission while strains of the other groups are more likely to be transmitted from host to environment to host?

CONCLUSIONS

Although E. coli can be isolated from a wide range of organisms, the data suggest that the preferred niche of E. coli is a mammal with a hindgut modified to facilitate microbial fermentation or, in the absence of a fermentation chamber, a large host. The morphology of the gastrointestinal tract, food transit time, and the nature of the nutrients available to E. coli resulting from different host diets are all interrelated factors. Further theoretical and empirical studies are needed if we are to determine whether one of the factors—gut structure, transit time, or nutrients—dominates the colonization dynamics of E. coli in a host.

E. coli exhibits subspecific genetic structure, and it appears that each of the subspecies has its own preferred ecological niche. The exact natures of the niches of A, B1, B2, and D strains need to be determined. Further studies are also required in order to discover the extent to which the within-host dynamics of each of the subspecies differ among different host groups, as well as the extent to which each of the subspecies differs in its transmission dynamics. In addition to defining the ecological niche and epidemiology of each of the subspecies of E. coli, studies to identify the natures of the adaptations responsible for these differences among subspecies are required. It is likely that acquiring this knowledge will enhance efforts to manage diseases caused by the various pathotypes of E. coli in humans and domestic animals.