In contrast to the relatively rare harmful encounters with pathogens, indigenous human-microbe relationships in which either microbe or host benefits without causing harm (commensal relationships) and relationships in which both benefit (mutualistic relationships) are the dominant forms of interaction and are fundamentally important to human biology. Coevolution, co-adaptation, and co-dependency are features of our relationships with our indigenous microbiota.¹ The human microbiota facilitates nutrient acquisition and energy extraction from food, promotes terminal (postnatal) differentiation of mucosal structure and function, and stimulates both the innate and adaptive immune systems. By so doing, it helps to maintain epithelial boundary function and integrity, as well as to “educate” our innate immune defenses. It also provides “colonization resistance” against pathogen invasion, regulates intermediary metabolism, processes ingested chemicals, and provides small amounts of human accessory growth factors.^2,3 The rules and features of microbial community assembly are fundamentally important but, so far, are poorly understood.⁴ In the neonatal period, the community assembly process is especially dynamic and is influenced by early environmental (in particular, maternal) exposures and stochastic effects. The composition and functional capabilities of the indigenous microbiota evolve in a generally orderly fashion, as diet, hormonal environment, other environmental factors, and occasional ecologic disturbances play out their effects on a distinct, albeit diverse, human genetic background.⁵

Bacterial diversity in the indigenous communities of the human body is striking in its richness of distinct species and strains but also noteworthy for the limited number of phyla commonly found. Despite exposure to more than 100 bacterial phyla in the surrounding environment, members of the phyla Firmicutes, Bacteroidetes, Actinobacteria, Proteobacteria, and Fusobacteria dominate human body sites, suggesting a role for strong selective forces and microbial diversification over hundreds of thousands of years of coevolution with their host. Within the domain Archaea, diversity in the human body is apparently limited to a handful of methanogen species: Methanobrevibacter smithii is commonly found in the healthy distal gut, and Methanobrevibacter-related species are found in the inflamed subgingival crevice in some patients with moderate or severe chronic periodontitis. Of interest, patterns of bacterial diversity in humans display individual-specific features. The distinctness of an individual’s microbiota is less evident when viewed in terms of the overall functional capabilities of the community, rather than in terms of the names and relatedness of the strains and species⁶; this difference probably reflects the functional redundancy of strains and species within the human microbiota, which, in turn, may contribute to stability of this ecologic system. Yet, differences in the capability of strains may explain variation among individuals in the metabolism of drugs such as digoxin and other exogenous chemicals.⁷ Differences in the capability of strains to tolerate normal inflammation may also influence the composition of the microbiota. Although there is evidence for shared functional capabilities among the intestinal microbial communities of different healthy humans, host genetics is a source of variation in the makeup of the human indigenous microbiota.⁸

Infection (or colonization) is simply the establishment of a microorganism on or within a host; it may be short lived, as in our encounters with “transients” (Table 1-1), or be persistent and may result in only low gain or harm to either participant. The term infectious disease applies when an interaction with a microbe causes damage to the host and the associated damage or altered physiology results in clinical signs and symptoms of disease. A pathogen is usually defined as any microorganism that has the capacity to cause disease. It is a medical definition; it is not a biologic definition, and certainly, not all pathogens have an equal probability of causing clinically apparent disease. Virulence provides a quantitative measure of pathogenicity or the likelihood of causing disease. For example, encapsulated pneumococci are more virulent than nonencapsulated pneumococci, and Escherichia coli strains that express Shiga-like toxins are more virulent than those that do not express these toxins. Virulence factors refer to the properties (e.g., gene products) that enable a microorganism to establish itself and replicate on or within a specific host species and that enhance the microbe’s potential to cause overt pathology. In many ways, what we refer to as virulence factors are in the biologic sense colonization factors that permit replication in the host and subsequent transmission to a new susceptible host.

Thus, it is useful to distinguish pathogens that regularly cause disease in some proportion of susceptible individuals with apparently intact defense systems from other potentially pathogenic microorganisms, such as Pseudomonas aeruginosa. This microorganism does not usually cause disease in individuals with intact host defense systems, yet it causes devastating disease in many immunocompromised patients. Many microorganisms with a capacity for sustained multiplication in humans, including members of the indigenous microbiota, cause disease more readily in individuals with underlying chronic disease or in those who are otherwise compromised. The common term opportunist suits this category of pathogen well (see Table 1-1).

An emerging concept of microbial disease causation, with origins in the field of ecology, is the notion of “community as pathogen,” in which a conserved broad feature of the microbial community contributes to pathology, rather than any one specific member or component. This concept may be relevant to a wide variety of chronic inflammatory processes of skin and mucosa, including inflammatory bowel disease and chronic periodontitis. It suggests that studies of pathogenesis consider general properties of microbial communities, such as resilience, or conserved functional interactions, such as syntrophy (cross-feeding), rather than the role of single microbes in isolation, especially for development of novel approaches for maintaining or restoring health.

The difficulty, therefore, is that the distinctions between a commensal, an opportunist, and a pathogen can be blurred at times, in part because some commensals cause disease, albeit usually in immunocompromised hosts, and some of the most feared pathogens can persist in humans for a lifetime without causing disease symptoms. In addition, microbial pathogenesis involves synergies between organisms, as well as between gene products, each of which may be insufficient alone in causing disease. For example, several members of the human health-associated nasopharyngeal microbiota, including Streptococcus pneumoniae, Neisseria meningitidis, and Streptococcus pyogenes, regularly cause well-defined, well-known human diseases. Immunization against the first two of these microbes not only protects against disease but also prevents, in an antigen-specific fashion, their ability to colonize the host. Such commensal pathogens persist in a significant proportion of the human population, the vast majority being asymptomatic carriers. Are they pathogens, or are these organisms members of the indigenous microbiota that have evolved to compete with other members of the community and live in a perilous location—associated with respiratory tract lymphatic tissue, where they regularly come into contact with elements of an immune system that hold them at bay most of the time but occasionally fail to do so, resulting in disease?

An understanding of the definition of a pathogen is not required when a clinician is faced with an infected patient who needs treatment. However, if we are to understand disease-associated microbes and discover effective therapies, we will also need to appreciate their fundamental biology and ecologic niche. And it is important to be reminded that antibiotics do not always, and are increasingly less likely to, work against many pathogens; that antibiotics incur a cost, in terms of resistance and collateral damage to commensals⁹; and that we still lack effective vaccines against a multitude of infectious agents that are encountered in everyday medical practice.

Thus, the capacity of certain microorganisms to cause disease in healthy, uncompromised human hosts on a regular basis should reflect fundamental biologic differences in their virulence capabilities from those of opportunists and commensal species that rarely, if ever, cause disease. In the following sections, we address this issue and discuss how insight into pathogenesis has been applied to the practice of contemporary infectious disease medicine.

Attributes of Microbial Pathogens

What are the distinguishing characteristics of microbes that live in humans? A successful pathogen or commensal must do the following: (1) enter the human host; (2) become established, which includes successful competition with indigenous microbes; (3) acquire nutrients; (4) avoid or circumvent the host’s innate defenses and a powerful immune system; (5) above all, replicate; (6) disseminate if necessary to a preferred site; and (7) eventually be transmitted to a new susceptible host.

Whether a pathogen or a commensal, a microorganism must also possess an interactive group of complementary genetic properties, sometimes coregulated, that promote its interaction with the human host. For a given microorganism, the genetic traits define unique attributes that enable it to follow a common sequence of steps used in establishing infection or, in some cases, subsequent disease.^10,11

Elegant molecular and genetic techniques now permit the identification, isolation, and characterization of many of these genes and their products. We now also possess the complete genome sequences of virtually every major pathogenic bacterial species. This information provides important clues and insight into the potential of a microorganism for causing disease and facilitates new experimental strategies for understanding pathogens and commensals alike.^12,13 The availability of the host (e.g., human) genome sequence also enables multiple synergistic approaches for understanding virulence, including the identification of host susceptibility traits, genome-wide assessments of host response, and clues about the mechanisms of host defense and pathogen counterdefense.¹⁴ It is important to recognize that pathogenicity can only be understood in the context of a specific host.

These genomic analyses have lent credence to the working hypothesis of almost a half-century of research—that the distinguishing characteristic of microorganisms that regularly cause disease is a set of special genetic traits that provide them with the capacity to breach intact host anatomic, cellular, or biochemical barriers that ordinarily prevent entry by other microorganisms into sterile tissue sites. Thus, pathogens “go where other microbes dare not.” In addition, many pathogens, such as Mycobacterium tuberculosis, Treponema pallidum, Chlamydia trachomatis, and Salmonella typhi, have the capacity to establish persistent (usually asymptomatic) infection in the human host and have evolved the extraordinary capacity to live in the inner sanctums of our innate and adaptive immune defenses or, in general, to compete well in the face of otherwise hostile host conditions. For example, Salmonella profits from the inflammatory response that it provokes in the gut by using the oxidized form of a locally produced host factor for a selective growth advantage against commensals.¹⁵ A distinction, then, between a primary pathogen and opportunist is that the pathogen has an inherent ability to breach the host barriers that ordinarily restrict other microbes, whereas the opportunist requires some underlying defect or alteration in the host’s defenses, whether it be genetic, ecologic (altered microbiota), or caused by underlying disease, to establish itself in a usually privileged host niche. Clearly, the nature of the host plays as important a role as the pathogen in determining outcome.¹⁶

An initial step required of a pathogen is to gain access to the host in sufficient numbers. Such access requires that the microorganism not only make contact with an appropriate surface but also then reach its unique niche or microenvironment on or within the host. This requirement is not trivial. Some pathogens must survive for varying periods in the external environment. Others have evolved an effective and efficient means of transmission. To accomplish this goal, the infecting microbe may make use of motility, chemotactic properties, and adhesive structures (or adhesins) that mediate binding to specific eukaryotic cell receptors or to other microorganisms.^17,18 Pathogens that persist at the surface of skin or mucosa usually rely upon multiple redundant adhesins and adherence mechanisms. If the adhesin is immunogenic, expression is usually regulated; in addition, antigenic variants may arise (see “Regulation of Bacterial Pathogenicity”). Preexisting microorganisms, the indigenous microbiota, provide competition against establishment of the newcomer; furthermore, the latter must adapt, at least temporarily, to the particular nutrient environment in which it now finds itself.

Normal inherent host defense mechanisms pose the most difficult set of obstacles for pathogens and commensals in establishing themselves in a host. For any set of specific host defenses, an individual pathogen will have a unique and distinctive counterstrategy. Some of the best-known mechanisms that pathogenic microbes use for countering host defenses include the use of an antiphagocytic capsule and the elaboration of toxins and microbial enzymes that act on host immune cells and/or destroy anatomic barriers. Microorganisms also use subtle biochemical mechanisms to avoid, subvert, or, as we now increasingly understand, manipulate host defenses. These strategies include the elaboration of immunoglobulin-specific proteases, iron sequestration mechanisms, coating themselves with host proteins to confuse the immune surveillance system, or causing host cells to signal inappropriately, leading to dysregulation of host defenses or even host cell death. Examples of these mechanisms include the production of immunoglobulin A1 protease by the meningococci, the use of receptors for iron-saturated human transferrin and lactoferrin by N. gonorrhoeae, and the coating of T. pallidum with human soluble fibronectin. Yersinia, Mycobacterium, and Bordetella induce host cell production of interleukin-10, which is a potent immunosuppressive cytokine, thereby downregulating important elements of the innate immune defense. Antigenic variation and intracellular invasion are other common strategies used by successful pathogens to avoid immune detection.^19,20 The intimacy of the relationships between viral pathogens and host is reflected in the frequency with which these pathogens co-opt host molecules and pathways for subverting host defenses (see “Subversion of Host Cellular Processes and Immune Defenses”).^19,21–23

The ability to multiply is a characteristic of all living organisms. Whether the pathogen’s habitat in the relevant host is intracellular or extracellular, mucosal or submucosal, within the bloodstream or within another privileged anatomic site, pathogens have evolved a distinct set of biochemical tactics to achieve this goal. The ultimate success of a pathogen, indeed, of any microorganism, is measured by the degree to which it can multiply. The pathogen must not only replicate sufficiently to establish itself in a host on reaching its specific niche, but it also must replicate sufficiently at some point in its life cycle to ensure its potential transmission to a new susceptible host. The rate of pathogen multiplication is appreciated by a clinician in terms of a characteristic incubation period spanning the time of exposure to the appearance of signs and symptoms of disease.

Infectious disease, in one sense, is simply a byproduct of the method and site chosen by pathogens for replication and persistence; disease per se is not a measure of microbial success. Disease, in part, reflects the status of the host as much as it does the virulence characteristics of the offending microorganism. Death of the host is fortunately a rare event and one that must be viewed with the dispassion of biology as being detrimental to both parties involved! The usual rules of host-pathogen engagement most often produce a tie: sufficient multiplication of the pathogen to ensure its establishment within the host (transient or long-term infection) and to ensure its successful transmission to a new susceptible host, while at the same time, no more than is tolerated by the host as it gains immunity from further incursion by the same and even related pathogens.

Why do some pathogens cause disease more readily than others? The strategy used for multiplication on or within the host (i.e., its ability to overcome host barriers) often defines fundamental differences between pathogens that commonly cause acute disease symptoms and those that do not. An organism that can reach and multiply in privileged anatomic sites away from the competitive environment of skin and mucosal surfaces is more likely to disrupt homeostasis in the host and cause disease than one that chooses a different strategy. If a microorganism has evolved a means to nullify or destroy phagocytic cells to multiply successfully, it is more likely to be found in deeper compartments and associated with acute disease. Commensal or mutualistic organisms are restrained so that they multiply just enough, in the midst of competing microbiota, to persist but not damage the host’s self-preserving homeostatic and innate immunity mechanisms. It is important to emphasize that a microorganism equipped to multiply efficiently in a human may be exceptional in the biologic sense but unexceptional as a pathogen in the medical sense, and only infrequently, if ever, a cause of clinically manifested disease.

Some organisms, such as P. aeruginosa, are “only” opportunists in humans despite their impressive array of virulence factors. These virulence factors work well in some plant hosts and in predators it encounters in the environment. However, for Pseudomonas, these same pathogenic determinants usually fail to overcome the average human’s defenses. For opportunistic pathogens, the state of the host is the major determinant of whether disease is the outcome of their interaction with the host. Commensals and mutualists, for example, which are the usual cause of opportunistic infections, may be very adept at colonization, but because of their preferred growth locale (e.g., at the mucosal surface) and preferred growth conditions (e.g., a microaerobic environment), they may have limited growth opportunities outside their restricted niche in an unimpaired individual. Innate immune factors are difficult to overcome for the vast majority of commensals. Little more than 50 years ago, there was a prevalent view that pathogens had undergone retrograde evolution and caused disease because they were little more than parasites. Pathogens were then viewed as organisms often unadapted to their hosts and that elaborated potent toxins or other powerful aggressive factors causing the signs and symptoms of disease. However, bacterial pathogenicity has been redefined over the past quarter century by using the tools of molecular genetics, genomics, and cell biology. We can now directly address the question: why are some bacteria pathogenic for humans, whereas other (closely related) bacteria are not?

We understand now that pathogenic bacteria are indeed often exquisitely adapted microbes that use sophisticated biochemical strategies to interfere with, or manipulate for their own benefit, the normal function(s) of the host cell. They are impressive human cell biologists! It is now also quite clear that these sophisticated biochemical properties that distinguish pathogens from their nonpathogenic brethren derive from specialized genes possessed by pathogens but absent from nonpathogens. The driving force for the inheritance of pathogenic traits is not slow adaptation to the host but, rather, a more dynamic process of horizontal (lateral) gene transfer via mobile genetic elements. Hence, the genes for many specialized “bacterial” products, such as toxins and adhesins, actually reside on transposons (“jumping genes”) and bacterial viruses (bacteriophages) (Table 1-2). Larger packets of information have also been shared among bacteria by genetic transfer. The lateral inheritance of large blocks of genes, called pathogenicity islands, is often the key to the expression of pathogenicity in bacteria. Many of these virulence determinants acquired by lateral gene transfer have several features that are apparent by simple inspection of genome sequences, including distinct chromosomal nucleotide composition and association with plasmids or phages, suggesting that their ancestry derives from an unrelated microbe. One surprising finding is that the amount of acquired DNA associated with virulence and adaptation to a host habitat in many bacteria can be substantial. For example, uropathogenic, enterohemorrhagic, and extraintestinal types of E. coli all display mosaic genome structure, with hundreds of gene islands distinct to each type, comprising as much as 40% of the overall gene content in each of these strains.²⁴ Each pathotype is as distinct from others as each is from a nonpathogenic laboratory strain of E. coli. Conversely, no more than half of the combined gene set is common to all E. coli strains. From this and other similar findings arises the concept of the “pan-genome” or the complete set of genes for a species. E. coli has a relatively “open” pan-genome, in that with every new genome sequence, a new set of approximately 300 unique genes is discovered, suggesting ongoing evolution of this species by gene acquisition.²⁵ Bacillus anthracis and other accidental human pathogens with a restricted environmental habitat or a different host preference instead display a relatively “closed” pan-genome and a much greater fraction of shared genes.

Hence, we can conclude that, in most cases, human-adapted pathogens have virulence genes not present in nonpathogenic relatives, and the distribution of these genes suggests that bacteria evolve to become pathogens by acquiring virulence determinants and not by the gradual loss of genes. This is not to say that, over time, some pathogens do not dispense with some genes that are no longer useful for a pathogenic lifestyle. Indeed, the study of host adaptation suggests that gene loss or gene inactivation is often associated with the adaptation of a particular pathogen to a particular host. For example, S. typhi, compared with Salmonella typhimurium, has lost or inactivated a large number of genes. Yet, it also has acquired by lateral transfer a unique surface determinant, Vi, and a unique toxin. Thus, the fundamental evolutionary push to pathogenicity results from gene acquisition. This is not simply a mechanism that microbes use to become pathogenic but, instead, a general strategy for microbial specialization and success in some environmental niches that are highly competitive. Why bacteria have adapted this tactic to maximize their diversity and to increase their opportunity for continuing evolution is most likely a reflection of their haploid state and their need to conserve fundamental characteristics, such as the ability to live on a mucosal surface, while still being able to try new combinations of genes. The sharing of genes among seemingly disparate microorganisms occupying the same niche in a sense provides these microbes with an endless number of combinations of genes for evolutionary experimentation, as it were, within a habitat such as the human intestinal tract.²⁶ Overall, across the bacterial world, the number of such successful experiments resulting in the emergence of a pathogen appears to have been quite rare (see “Clonal Nature of Bacterial Pathogens”). Yet, when successful, these experiments are surprisingly efficient, at least from the perspective of the microbe, yet manageable, from the standpoint of the host.

Some infectious diseases occur predominantly in dramatic epidemic form, which argues against the establishment of a balanced host-parasite relationship; however, in many such epidemics, mitigating circumstances involving herd immunity and other underlying social, economic, and political issues impinge on this relationship. So-called emerging infectious diseases reflect various aspects of imbalance in the relationships between host, pathogen, and environment.²⁷ Many of the most serious and feared infectious diseases occur when humans are infected by microorganisms (accidental pathogens) that prefer and are better adapted to another mammalian host. In fact, most emerging infectious diseases in humans are of zoonotic origin.²⁸ As seen in many zoonotic diseases, the rules of engagement between the host and the pathogen are blurred, often to the detriment of both the host and the microbe. It is often an evolutionary dead end for both parties.

Given the increasingly frequent and unexpected emergence of previously unrecognized pathogens, it is appropriate to question how well we appreciate the true diversity and distribution of extant microorganisms capable of causing human disease. Although most emerging pathogens are zoonotic agents and already adapted to a different host, the question also concerns a more basic uncertainty about how often, in what phylogenetic backgrounds, and through what mechanisms virulence for humans among microbes can arise. Pathogenicity appears to have arisen on multiple occasions throughout the domain Bacteria but only in a small fraction of the overall phyla, that is, those whose members typically colonize humans (see “Diversity of Human-Microbe Relationships”). Although there are currently no known traditional pathogens within the domain Archaea, the methanogens, through a synergistic interaction with other microbes, known as syntrophy, may contribute to pathology in certain clinical settings.^29–31 For example, in chronic periodontitis, methanogens in the subgingival crevice may enhance the growth of fermentative, would-be pathogenic bacteria by consuming the hydrogen produced by the latter.

Finally, before considering several facets of pathogenicity in more detail, three further points should be considered: (1) pathogen detection and identification remain suboptimal, in part because of continuing dependence on cultivation methods, and therefore a number of novel pathogens may have gone undetected³²; (2) some potential pathogens may not have had adequate contact with humans to have made themselves known (yet)³³; and (3) dominant ideas of microbial disease causation (e.g., a single pathogenic agent in a susceptible host) may be too restrictive. As mentioned earlier, some microbial diseases may require a consortium of agents (e.g., intra-abdominal abscess), thereby posing challenges for pathogen identification. If we define success for a microbial pathogen as disease without a requirement for long-term survival, a much larger number of organisms may qualify, in being able to cause devastating human misery but only over a limited number of generations. These matters have obvious relevance to the troubling issue of bioterrorism and the potential malevolent use and genetic manipulation of microorganisms.

Clonal Nature of Bacterial Pathogens

As noted previously, pathogenicity is not a microbial trait that has become fixed by chance. Instead, particular microbial strains and species adapted to a particular host have evolved to carry very specific arrays of virulence-associated genes. By examining the genetic organization of pathogens, opportunists, and nonpathogenic bacteria, one can begin to understand the origins of pathogenicity and why some pathogens are more pathogenic or more successful than their peers.

Techniques used in the study of genetic relatedness include primary protein or nucleic acid sequence comparisons and DNA hybridization methods, including DNA microarray-based approaches.¹² Some genetic sequences, such as those of the small- and large-subunit ribosomal RNAs, have been used as reliable evolutionary clocks.^34,35 Comparative analysis of these sequences allows one to infer phylogenetic relationships among all known cellular life, but these sequences provide only limited resolution between strains and limited insight into organismal biology and function. The increasing ease with which primary genomic sequence information can be acquired, differences quantified, and these data shared has led to more precise methods of strain characterization, such as multilocus sequence typing³⁶ and whole-genome sequencing. Today, full-genome sequencing and genome-wide single nucleotide polymorphism analysis are feasible on a large-scale basis and offer the greatest insight into evolutionary relationships and population biology of microbes.¹³ All of these sequence-based approaches avoid the pitfalls of classic comparisons of phenotypes (i.e., gross observable characteristics of a microbe), which can be unreliable. When these sequence-based techniques are used, a consistent finding emerges concerning the population structure of microorganisms: most natural populations of microorganisms consist of a number of discrete clonal lineages.³⁷

A clonal population structure implies that the rates of recombination of chromosomal genes between different strains of the same species and between different bacterial species are very low. Clonal organization has been substantiated by the concordance between evolutionary trees derived from unrelated chromosomal sequences.³⁸ Even though bacteria possess well-established, naturally occurring genetic exchange mechanisms, they retain their individuality, just as the human- and other animal-associated bacterial communities demonstrate stable membership over long periods of time in an individual. We might have thought that in light of unmistakable gene shuffling among and between bacteria, we might see homogenization of bacterial species and little specialization. In fact, the opposite is true. Bacterial species have remained discrete and distinct taxonomic entities³⁹ because the bacterial chromosome is a highly integrated and co-adapted entity that has, in general, resisted rearrangement. The same may be true of the overall structure of the indigenous bacterial communities of humans.

It is intriguing that analysis of natural populations of microorganisms with pathogenic potential has revealed the prominent representation of a relatively few clones. In fact, most cases of serious disease may be caused by only a few of the many extant clones that constitute a pathogenic bacterial species. For example, one sees this in meningococcal disease, where there is a clear predominance of a particular clone in large areas worldwide. In contrast, in the case of the typhoid bacillus, there is only one major clone worldwide, although antibiotic resistance may be forcing diversity.⁴⁰ Indeed, in some extreme cases, all members of a species, such as Shigella sonnei or Bordetella pertussis, belong to the same clonal type or small group of closely related types. Not all pathogenic bacterial species reveal this pattern of clonal organization. Two notable exceptions are N. gonorrhoeae and Helicobacter pylori, which appear to use chromosomal recombination to increase their genetic diversity. It may be that organisms such as N. gonorrhoeae and H. pylori, which are quite specialized in their preference for discrete human anatomic sites where they rarely encounter related species, must resort to constant recombination and genetic reassortment by using DNA transformation as a means of sharing the well-adapted alleles that accumulate within their populations. Thus, genetic variability among gonococcal and H. pylori isolates from discrete geographic locations suggests that these organisms are essentially and appropriately sexual.

Clonal analysis has generated other important conclusions concerning the evolution of bacterial species and pathogenic strains in particular. Study of E. coli populations in the human intestinal tract indicates that only a small number of clonal lineages persist, whereas numerous unrelated cell lines appear and disappear.³⁷ E. coli urinary tract pathogens that cause symptomatic disease in humans may be even less genetically diverse than E. coli strains found in the intestinal microbiota or those that cause asymptomatic urinary tract colonization.⁴¹ Perhaps the evolution of these E. coli strains to live in a more specialized epithelial niche results in constraints on recombination that preserve their added degree of specialization. At the same time, specialization for one body site may not preclude fitness for another site: in some individuals with recurrent urinary tract infections, there can be a simultaneous and identical shift in the dominant E. coli population of the bladder and distal gut between one episode and the next.⁴² Pathogens have even taught us about human prehistory: sequence-based population analysis of human-restricted and human-adapted bacterial pathogens, such as H. pylori, has clarified important aspects of human migration and human population structure.⁴³