The unexhausted potential of E. coli

1In 1884, the German microbiologist and pediatrician Theodor Escherich began a study of infant gut microbes and their role in digestion and disease. During this study, he discovered a fast-growing bacterium that he called Bacterium coli commune, but which is now known as the biological rock star that is Escherichia coli. E. coli's meteoric rise and exalted status in biology stem from how easy it is to find and work with. Hardy, non-pathogenic, and versatile strains that grow quickly on many different nutrients can be isolated from virtually any human. These traits made E. coli a mainstay in microbiology teaching lab collections. Consequently, when early 20th century microbiologists cast about for a model organism, E. coli was one of the most widely available choices.

2Those who chose to work with E. coli included,,, and,, who between them performed groundbreaking studies on bacterial physiology, viruses, and genetics. By the 1940s, its use in many foundational studies firmly established E. coli as the bacterial model organism of choice, making it the obvious organism to work with at the onset of the molecular biology revolution in the 1950s. As a result, it became the organism in which the most basic aspects of life, including the genetic code, transcription, translation, and replication, were first worked out (;; see for an excellent history of early molecular biology and E. coli's role in it). The resulting knowledge and molecular methods for investigating and manipulating its biology have since led to E. coli's prominence in academic and commercial genetic engineering, pharmaceutical production, and experimental microbial evolution (see Box 1 for a glossary of specialist terms used in this article ), not to mention the biotechnology industry, which contributed $500 billion to the global economy in 2011. It is not hyperbole to say that E. coli is now the most important model organism in biology (; see Box 2).

3For all of its importance, E. coli is quite nondescript. It is a fairly typical Gram-negative bacillus (see ‘Glossary’), measuring only about 1 μm long by 0.35 μm wide, although this can vary considerably depending on the strain and its conditions. Even at high magnification it looks like nothing more than a tiny sausage (Figure 1A). It may have whip-like flagella that it uses to move about its environment, or hair-like pili that allow it to attach to surfaces or to other cells (Figure 1B). Physiologically, it is a facultative aerobe, meaning that it can grow happily with or without oxygen, but it cannot grow at extremes of temperature or pH nor can it degrade dangerous pollutants, photosynthesize, or do a variety of other things that interest microbiologists. Phylogenetically, it is a member of the Enterobacteriaceae, and is closely related to such pathogens as Salmonella, Klebsiella, Serratia, and the infamous Yersinia pestis, which causes plague.

4In nature, E. coli is principally a constituent of the mammalian gut microbiome (see ‘Glossary’), but it is also found, albeit less commonly, in the gut microbiomes of birds, reptiles and fish, as well as in soil, water, plants, and food. Its mammalian abode is why E. coli can metabolize lactose, the control of which was the subject of seminal studies of gene regulation. E. coli is typically the most common aerobe in the lower intestine of mammals. However, the gut is primarily an anoxic environment, and the extremely large, highly diverse (500+ taxa) gut microbial community is dominated by obligate anaerobes, such as members of Bacteroides and Firmicutes, which alone comprise ∼90% or more of the total gut microbial population. By contrast, E. coli typically constitutes only 0.1–5% of the community, which partly reflects the fact that its niche is to be found in the relatively thin layer of mucus that lines the gut. In the mucus layer, E. coli grows in a complex, multi-species biofilm (see ‘Glossary’) in which it competes for an array of nutrients—one of the origins of its broad diet. This rich food supply enables E. coli to maintain population densities of 106–109 cells per gram of fecal matter despite unavoidable and regular bulk losses. In the human gut, the E. coli population typically includes a set of long-term residential strains, and also short-term transients that vary with diet, health, and with exposure to antibiotics. The human gut E.coli population also varies due to its constant, largely unknown interactions with the broader microbiome, and with the host and the host's vast virome (see ‘Glossary’), against which it defends itself with restriction enzymes, which have become a key tool in molecular biology (see ‘Glossary’).

5A host organism carefully regulates its microbiome in poorly understood ways that have an impact on its health. For instance, the human gut secretes immunoglobulin A, which appears to facilitate the formation of E. coli biofilms on the intestinal mucosa, suggesting that their presence is welcome. While long considered to have a commensal relationship with its host, in which E. coli secures food and a nice warm home while contributing little in return, it is increasingly clear that the host-E. coli relationship is really a mutualism. Indeed, E. coli benefits its host in a number of ways. It produces vitamin K and vitamin B12, both of which are required by mammalian hosts. E. coli also maintains a friendly environment for its anaerobic neighbors by consuming oxygen that enters the gut. Perhaps most importantly, E. coli competitively excludes pathogens from its niche in the gut, rather like how friendly barbarian tribes, settled by the Roman Empire on its frontiers, helped to keep out the more dangerous tribes.

6E. coli's relationship with a host literally begins at birth. Newborns are typically inoculated with maternal E. coli through exposure to her fecal matter during birth and from subsequent handling. Although perhaps disconcerting to ponder, this inoculation seems to be quite important. Indeed, E. coli becomes more abundant in the mother's microbiome during pregnancy, increasing the chances of her newborn's inoculation. The colonizing strains typically have secretion systems and pili that allow them to attach to and interact with the infant's gut epithelium. This newly established and rapidly growing E. coli population then changes the structure and function of the epithelial cells in ways that appear crucial for healthy microbiome development. It is therefore concerning that early human infant colonization by E. coli has been declining in the US and in other Western nations as rates of caesarean delivery have increased and hospital hygiene has continued to improve. Indeed, this decrease has been accompanied by broader microbiome changes, including increased infant gut colonization by Staphylococcus aureus, which is linked to an increased risk of developing a variety of disorders, including asthma, obesity, and diabetes. The long-term consequences of disrupting humanity's long association with E. coli are under intensive investigation.

7An inevitable consequence of being a gut microbe is to be regularly excreted into the external world. The mucus lining of the gut is constantly sloughed off and excreted in fecal matter, so cells of a resident E. coli population are shed almost as soon as that population is established. E. coli's long-term life cycle is hence biphasic, and despite being exquisitely adapted to the good life inside of a host, E.coli must also be adapted to successfully acclimate to a harsher life outside the host. This is a remarkable feat. Whereas life on the inside is easy and stable, every aspect of the external environment, be it nutrition, temperature, oxygen, moisture, pH, and/or the surrounding microbial community, can fluctuate wildly. It is likely that the hardiness, metabolic flexibility, and substrate breadth that have made E. coli such a valuable model organism evolved in part to permit it to survive this hostility and variability long enough to make it back to a host.

8Another interesting trait that is almost certainly relevant to E. coli's survival of its environmental phase is the production of persister variants. First observed in Staphylococcus during experiments with penicillin, persisters are rare, highly antibiotic-tolerant phenotypic variants that arise at random in bacterial populations. Persisters are not an adaptation specifically to antibiotics; their hallmark antibiotic tolerance is attributable to their metabolic inactivity, which is triggered by several redundant pathways, including those that govern stress response. Persistence therefore appears to be a general adaptation that permits small numbers of dormant cells to survive a variety of environmental fluctuations. E. coli has been, as with so many other phenomena, a model in which to study bacterial persistence, and it seems likely that this capacity to enter a dormant state plays a significant role in surviving the considerable fluctuations it encounters in its external environment.

9The external world was long thought to be so harsh as to preclude E. coli's growth outside of its host. While a tiny minority might eventually reach a new host, most cells were expected to eventually die. This is the basal assumption behind using the presence of E. coli as an indicator of fecal contamination. However, recent studies have shown that E. coli can, in fact, establish itself as a member of microbial soil, water, and plant-associated communities. Moreover, genomic and phylogenetic analyses of collections of E. coli strains have identified divergent lineages that appear to be adapted to a primarily non-host lifestyle. What adaptations are required for E. coli to make such radical ecological shifts, what niches it fills in its new communities, how stable its presence in those communities might be, and what impact its adaptation to these new niches might have on its capacity to return to a host remain outstanding questions that must be addressed (see Box 3).

10It is possible that these environmental E. coli populations will help resolve the interesting problem of swarming motility. It has long been observed that groups of E. coli cells on water-restricted surfaces will congregate and engage in social, coordinated movement over the surface, a behavior also seen in other bacteria (;; Partridge and Harshey, 2013a,b). However, swarming has principally been observed and studied on agar plates in the lab (Partridge and Harshey, 2013a). What function it might serve for E. coli in nature has been unclear. The gut generally lacks the sorts of surfaces on which swarming works. Given that ∼216 E. coli genes are specifically involved in swarming motility, it is a costly trait that would not be maintained if it did not confer some selective benefit. It is as yet unclear if swarming plays any role in the initial colonization of the gut, but it is associated with pathogenesis, likely due to its improved colonization of tissue surfaces that facilitate opportunistic infection (see the next section) (; Partridge and Harshey, 2013a). Given that surfaces, and interfaces between surfaces, are regularly encountered by environmental bacteria, swarming may well prove to be an adaptation that enables the exploration and colonization of viable habitats by E. coli upon their excretion into the external environment.

11E. coli's presence in the environment is a cause for concern because its relationship with humans is not entirely benign. Indeed, E. coli is a major cause of diarrheal diseases, peritonitis, colitis, bacteremia, infant mortality, and urinary tract infections that world-wide cost billions of dollars to treat and kill roughly 2 million humans each year. Some strains may even cause cancer. Some opportunistic E.coli infections are caused by normally harmless or beneficial strains when introduced to sick hosts or to parts of a host's body outside of the gut. However, there are also pathogenic strains that produce virulence factors and can cause illness in even the healthiest host. These strains are classified by where and how they cause disease into groups called pathotypes (see ‘Glossary’), which include enteroaggregative, enterohemorrhagic, enteropathogenic, enterotoxigenic, uropathogenic, meningitis-associated, and septicemic-associated E. coli (see ‘Glossary’).

12The most notorious of these is E. coli O157:H7, an enterohemorrhagic strain that produces a shiga-like toxin. This toxin (see ‘Glossary’) attacks small blood vessels, killing intestinal cells, and causing bloody diarrhea and severe abdominal pain, as well as hemolytic uremic syndrome (HUS), a potentially deadly condition that can involve widespread clots in capillaries and hemolytic anemia, thrombocytopenia, and renal failure (see ‘Glossary’;; ). Treatment can be difficult because antibiotics increase the risk of HUS. As a result, treatment is generally limited to the provision of fluids, adequate nutrition, medication for pain and fever, and blood transfusions when necessary.

13O157:H7 is particularly dangerous because it can easily contaminate human food supplies. It resides asymptomatically in cattle and in other livestock, and can be transferred to humans via the fecal contamination of meat during its butchering and packaging. It can also contaminate vegetables via fertilizers and water, and through contact with live-stock-associated birds. Through these points of entry into the human food chain, O157 has caused numerous outbreaks of illness. In the US alone, such outbreaks have annually affected ∼63,000 individuals, killing 20, and costing around $405 million in healthcare and in lost productivity. When E. coli gets bad press, O157:H7 is almost always the culprit, and rightfully so.

14The many pathogenic strains of E. coli testify to the diversity of this single microbial ‘species’, the full extent of which was only revealed by the advent of whole-genome sequencing. For example, genome sequences have firmly placed all Shigella strains within the broader E. coli clade.

15More importantly, sequencing has uncovered the remarkable plasticity and dynamism of the E. coli genome that contribute to its genetic and phenotypic diversity. In 2002, reported that three strains of E. coli, the popular lab strain K-12, O157:H7, and the uropathogenic strain CFT073, share only 39.2% of their genes. Subsequent sequencing of more strains has reduced this core genome to less than 20% of the more than 16,000 genes in the E. coli pan-genome (see ‘Glossary’;; ).

16The remaining genes constitute the flexible genome (see ‘Glossary’), a vast pool of ‘plug and play’ genetic variation that can be acquired via horizontal gene transfer. This variation includes prophages, transposable elements, and accessory genes (see ‘Glossary’). These genes encode functions that can: improve fitness in particular niches; increase metabolic flexibility; and affect pathogenicity. Flexible genomic elements are often large and integrate into the genome at select insertion hotspots. This capacity to mix and match accessory genetic elements means that new E.coli strains with novel combinations of traits can arise very quickly. Another consequence is that the size of the E. coli genome can vary greatly between strains. While standard lab strains have genomes of ∼4.5 million base pairs and 4000 genes, pathogenic strains can have genomes of over 5.9 million base pairs and 5500 genes.

17This extensive genetic plasticity poses major questions for understanding how E. coli evolves in the wild over the long-term. For instance, how does the rate of genomic evolution in a given E. coli lineage by mutation compare to that by horizontal gene transfer? Moreover, how much of the E. coli flexible genome ultimately derives from other organisms?

18E. coli has been tremendously valuable as a de-contextualized, abstracted model organism, but it could be of even greater value were we to gain a better understanding of its ecology and natural history. As I have discussed, it is a highly diverse and broadly distributed species that occupies an expansive, generalized niche in which it experiences a vast range of environmental and ecological conditions. This breadth presents two unique and synergistic opportunities. First, the tools and techniques developed for, and the knowledge derived from, the study of E. coli's lab strains can be applied to studying its wild relatives in far greater detail than is possible for any other microbe. This capacity promises to yield new and profound insights into the biology of other microbes that experience similar conditions, as well as the discovery and identification of new microbiological phenomena. Second, any increase in our knowledge of E. coli's natural history expands the range of biological phenomena for which its strains can be used as models to study. In other words, the study of E. coli's natural history can reveal more than is possible with most organisms, and an increased understanding of its natural history in turn expands its potential as a model organism.

19An example of this dynamic can be seen in studies of E. coli biofilms, which have revealed much about how harmless and pathogenic strains colonize the gut and persist in the environment. These studies have, in turn, led to the development of E. coli as a model for studying the formation, genetics, physiology, and consequences of biofilms, generating important findings on this most common of microbial lifestyles. Experimental evolutionary studies of E. coli biofilms have also led to findings of far-reaching consequence, such as how biofilms impact microbial evolution and how they facilitate the evolution of antibiotic resistance even when antibiotics are not present.

20There are many unanswered questions about E. coli's natural history (see Box 3). How does E. coli adapt to non-host environments? What role does it play in non-host communities? How does it adapt to life in the soil? Just how fluid is the E. coli genome? How do environmental and ecological conditions affect this fluidity? How does the pan-genome evolve? How much interaction is there between E. coli's pan-genome and that of other microbes? How tightly adapted are hosts to their E. coli populations? How is E. coli adapting to changes in the human diet and lifestyle? Each unanswered question presents opportunities for novel research into unexplored corners of E. coli's natural history, and the subsequent expansion of its potential as a model. Improved appreciation and interest in E. coli's natural history can only uncover more questions, and increase its potential even more. If kept in mind by researchers, this dynamic will guarantee that the most important model organism of the 20th century will continue to be one of the most important model organisms of the 21st century and beyond.