Sage Journals: Discover world-class research

Abstract

Koch’s postulates have shaped our understanding of infectious diseases; however, one of the tangential consequences of them has been the emergence of a predominantly monomicrobial perspective concerning disease aetiology. This orthodoxy has been undermined by the growing recognition that some important infectious diseases have a polymicrobial aetiology. A significant new development in our understanding of polymicrobial infections is the recognition that they represent functional ecosystems and that to understand such systems and the outcome and impact of therapeutic interventions requires an understanding of how these communities arise and develop. Therefore, it is timely to explore what we can learn from other fields. In particular, ecological theory may inform our understanding of how polymicrobial communities assemble their structure and their dynamics over time. Such work may also offer insights into how such communities move from stable to unstable states, as well as the role of invasive pathogens in the progression of the disease. Ecological theory offers a theoretical framework around which testable hypotheses can be developed to clarify the polymicrobial nature and dynamics of such infections in the face of environmental change and therapeutic interventions.

Keywords

Polymicrobial ecology cystic fibrosis bacterial vaginosis neutral assembly

Introduction

In 1890 Robert Koch laid out a definitive set of criteria for the identification of a pathogenic microorganism. Although subtly reworded and extended with a fourth postulate over the century that has passed,¹ the postulates can be summarised as follows:

1. the bacterium must be present in all diseased individuals (but should not be isolated in healthy individuals);

2. the bacterium must be isolated from a diseased individual and grown in pure culture;

3. the bacterium must cause the specific disease when inoculated into a healthy individual;

4. the same bacterium must be re-isolated from the experimentally diseased individual.

These postulates have shaped our understanding of medical microbiology as the vast majority of microbial diseases conform to them. However, there are notable problems in satisfying all the postulates, as was recognised by Koch himself when he reformulated the first postulate in recognition of the colonisation of healthy carriers by Vibrio cholerae, the cause of cholera.¹ Likewise, there are notable instances where a causative bacterial agent cannot yet be cultured by routine microbiology, including Treponema pallidum, the causative agent of syphilis, and Mycobacterium leprae, the causative agent of leprosy. It can be argued that this is a result of shortcomings in the culture techniques employed and not that the organism itself is truly unculturable. Developments in molecular biology led Falkow² to propose modifying Koch’s postulates to identify the causal relationship between specific genes and disease.

One of the tangential consequences of Koch’s postulates has been the emergence of the concept of ‘one microbe, one infectious disease’ (the ‘monomicrobial’ approach). However, this orthodoxy has been undermined by the growing recognition in recent years that some important infectious diseases have a polymicrobial aetiology,^3–5 where several microorganisms are involved in causing a single disease. Such polymicrobial disease states include abscesses,^6–9 bacterial vaginosis,^10,11 cystic fibrosis (CF),^12–17 possibly inflammatory bowel diseases,^18–21 oral diseases,²² otitis media,²³ rhinosinusitis²⁴ and thoracic empyema.^25,26 Moreover, an immunopathogenic interplay between Plasmodium falciparum and Epstein-Barr virus may be a significant trigger for endemic Burkitt’s lymphoma, the most prevalent childhood cancer in the African ‘malaria belt’.^27,28

The pathogenesis of polymicrobial infections can be considered from a number of perspectives. Firstly, polymicrobial infections can be categorised according to the nature of the interaction between types of microbes (bacteria, fungi, viruses).⁵ Well-characterised examples of virtually every imaginable microbial interaction exist. For example, bacteriocin production can be seen as a spiteful behaviour because it has a negative effect on the producer and the recipient.²⁹ Bacteriocin-producing strains of Pseudomonas aeruginosa have been isolated from the CF lung and it is thought that with a greater understanding of microbial interaction the frequency of bacteriocin-producing strains in the population could be manipulated in CF to reduce bacterial virulence.³⁰ Secondly, there is the nature of the interactions. In some cases, the presence of multiple microbes may be ‘additive’, for example in mixed abscesses and bloodstream infections, or genuinely synergistic, as observed between HIV and Mycobacterium tuberculosis.³¹ Where microbes are acting synergistically it is also important to consider how the microbial community may change over time. Several polymicrobial diseases are recognisably ‘microbial-shift’ diseases, where the community diversity changes over time, for example bacterial vaginosis (see below). It is relevant to consider whether different microbes are present and interacting together or whether (a) microorganism(s) succeeds another as a consequence of the earlier microbe(s) predisposing the host to infection by a secondary pathogen(s). For example, succession can be seen in the relationships between prior viral infections and bacterial otitis media²³ or exacerbations of chronic obstructive pulmonary disease.³² Important light has been shed on these ‘microbial-shift’ diseases through the advent of 16S rRNA-based molecular techniques for microbial community analysis. There are many 16SrRNA-based techniques employed which can identify a host of bacteria present in a disease site, such as denaturing gradient gel electrophoresis,³³ 16S rRNA clone libraries^34,35 and, most recently, microarray^36,37 and metagenomic sequencing.³⁸ The application of these novel technologies has and will continue to play an important role in the reassessment of the monomicrobial versus polymicrobial theories of microbial disease. Other methodological advances are also likely to be significant in this respect, notably the development of appropriate animal models for polymicrobial diseases.^4,18

A significant new development in our understanding of polymicrobial infections is the recognition that they represent functional ecosystems; to understand the nature, outcomes and the impact of therapeutic interventions on such infections requires an understanding of assembly, community structure and ecosystem function relationships. There is a growing recognition that the theoretical tools devised in general ecology may be applied usefully to polymicrobial systems, enabling investigation using an established theoretical framework to inform the development of testable hypotheses. It also offers an insight into whether polymicrobial community ecology mirrors that of macroorganisms.

Bacterial vaginosis as a representative microbial shift disease

Some polymicrobial diseases are caused by a shift in diversity where certain taxa become more abundant in the community associated with disease but pathology is not caused by a single bacterial species. Bacterial vaginosis (BV) is a very common condition of women of reproductive age that is characterised by malodour and discharge.¹¹ In addition to its direct morbidity, BV is also associated with an increased risk of HIV acquisition, pelvic inflammatory disease and adverse pregnancy outcomes.^11,39–42 From an epidemiological perspective, it has long been argued whether BV can be considered a sexually transmitted infection—a situation exacerbated by the longstanding failure to correlate BV with any single microbial cause that can satisfy Koch’s postulates.⁴³ Instead, a recent change in emphasis suggests that BV should be best considered as a sexually enhanced disease.^11,44,45 Thus, the contemporary view of BV is that it represents a shift from the normal vaginal flora (most commonly but not always dominated by lactobacilli),⁴⁶ through intermediate stages to a ‘BV-associated’ flora containing a ‘pathologic core’ of bacteria.⁴⁷ This is particularly evident in consideration of how BV pathogenesis links to the principal diagnostic criteria for BV:¹⁰ loss of lactobacilli is associated with a decline in lactic acid production and a diagnostically meaningful increase in vaginal pH above pH 4.5; production of biogenic amines (such as trimethylamine) by the BV-associated anaerobes contributes to both the fishy odour characteristic of BV and the diagnostic KOH amine test; formation of diagnostic ‘Clue cells’ (i.e. vaginal epithelial cells with adherent, possibly biofilm-organised bacteria of diverse non-Lactobacillus morphologies) reflects greater microbial colonisation of the epithelium following mucinolysis; and the homogenous discharge may reflect the cytotoxic activity of vaginolysin and other damage to the vaginal epithelium.

This shift in the vaginal flora correlates with the simple Gram-staining criteria (e.g. Nugent’s test) that can be used to grade the microflora in BV diagnosis.^48,49 Historically, Gardnerella vaginalis has been implicated as one of the key players in the BV flora, along with Prevotella spp. and other anaerobes, including Mobiluncus and Peptostreptococcus spp.^10,46,50 More recently, the application of culture-independent methods has extended understanding of the pathologic core further still and implicated increased diversity and numbers of fastidious bacteria such as Atopobium vaginae, Clostridiales spp., Leptotrichia/Sneathia spp. and Megasphaera spp. as contributing to the disturbed vaginal flora.^{43,49,51–57} Molecular techniques have been utilised to examine bacterial community structure from vaginal biopsy samples and have found that dense bacterial biofilms adherent to the vaginal epithelium containing G. vaginalis are specific to patients with BV.⁵⁸ Furthermore, G. vaginalis has been found to be more virulent than other BV associated anaerobes.⁵⁹ This is a result of G. vaginalis having the ability to adhere to the vaginal epithelium, to form biofilms and to lyse epithelial cells.⁵⁹

However, bacterial community heterogeneity between patients is still observed for the BV positive vaginal flora which has led to the suggestion that BV should be reclassified as a syndrome owing to the variety of symptoms and phenotypic outcomes.⁴⁷ Such heterogeneity may reflect variations the genetic background of the host which may be responsible for differences in the susceptibility to BV observed between individuals.⁶⁰

Cystic fibrosis (CF) and inherited disease

CF is primarily a genetic disease, whereby mutations in the CF transmembrane conductance regulator (CFTR) cause abnormal mucous clearance in the lungs which leads to chronic broncho-pulmonary infection. This gives rise to a vicious cycle of infection and inflammation causing tissue damage to the lungs.⁶¹ Culture-based analyses demonstrate that Haemophilus influenzae and Staphylococcus aureus can be isolated from CF sputum during childhood and that these are replaced by P. aeruginosa and members of the Burkholderia cepacia complex during adolescence. However, many other bacterial and fungal isolates are commonly identified but are disregarded as being non-pathogenic or contaminating oral or nasopharyngeal flora. This finding is perplexing because although H. influenzae is a pathogen in other situations, the contribution of this organism towards the pathology in CF remains undefined. Recently, molecular studies on bronchoalveolar lavage fluid (BALF) from individuals as young as one year of age have demonstrated colonisation with up to eight different species of bacteria,^34,35 including P. aeruginosa, S. aureus, Stenotrophomonas maltophila, Achromobacter xyloxosidans, B. cepacia complex and H. influenzae.⁶² Similarly, a study examining the bacterial diversity in CF patients using oropharyngeal swabs demonstrated huge diversity (206–1329 species per person) with many of these taxa commonly found in the gut or oral flora,^36,37 suggesting it is likely that the oropharynx acts as a reservoir of microbes that can easily invade the lungs owing to cessation of mucociliary clearance.

Moreover, evidence for the CFTR genotype affecting the microbial community in the CF lung has been demonstrated by Klepac-Ceraj et al.³⁷ CF patients in this study were assigned to three groups based upon their genotype: ΔF508 homozygous, ΔF508 heterozygous and patients without a ΔF508 allele. This analysis showed that patients who were either ΔF508 homozygous or heterozygous had more similar communities than those who had no ΔF508 allele. Likewise, CFTR genotype has also been implicated in susceptibility to fungal infections, with ΔF508 homo- and heterozygous patients being more readily colonised by Geosmithia argillacea.⁶³ Furthermore, antibiotic therapy also affects the microbial diversity of the CF lung.³⁷ Patients who were receiving long-term antibiotic therapy and who were colonised with P. aeruginosa had reduced phylogenetic diversity, as well as having a more distinct microbial community composition from each other. The commonly isolated respiratory pathogens may, therefore, represent a pathologic core, analogous to that observed in BV. However, a very recent study has partitioned CF taxa into core and satellite species that indicated only P. aeruginosa was a core species whereas the other recognised CF pathogens were satellite taxa that are transient.⁶⁴ This suggests that the periodic exacerbations that are significant causes of morbidity and reduction of lung function may represent a shift state of the microbial flora and the exacerbation may be a manifestation of transient presence of key the pathogenic bacteria in the CF airway.⁶⁵ However, the cause and definition of exacerbations is still unresolved and it may be that they are caused by viruses promoting pathogenicity in other components of the microbial flora. For example, in a study of infants with CF, respiratory viruses were identified in 52% of cases that required hospitalisation.^66,67 Alternatively it could be that the acquisition of a new bacterial species facilitates exacerbations in CF patients whereas treatment of members of the Streptococcus milleri group, which would normally be regarded as oral flora, resulted in alleviation of pulmonary exacerbation. Notably, these collective CF data are replicated in chronic obstructive pulmonary disease where viral infections as precipitants of exacerbations are recognised,⁶⁸ a resident flora has been noted in (as subset of) apparently stable patients⁶⁹ and where a change in strain or predominant species is associated with exacerbations.⁷⁰ Thus, mechanistic and biological plausibility for the theory of a core microbiota has been established in more than one chronic respiratory disease.

What can we learn from other fields?

The nature of polymicrobial infections pose some important questions about how such communities become established and their role in disease. Ecological theory offers two theories that have been widely applied in macroorganismal systems that may resolve such questions. The first, neutral assembly, assumes all species to be ecologically equivalent and that patterns of taxa diversity, relative abundance and composition are determined by the size of the meta-community, dispersal and speciation rates.⁷¹ The second theory is niche or deterministic assembly, which assumes that different taxa co-existing in the same environment can do so because they have specific traits that represent evolutionary adaptations to the physical and biological milieu. This enables them to adapt and trade off traits leading to species sorting along environmental gradients.⁷² These two theories predict different outcomes to community assembly, the invasion by new taxa and temporal effects. Consequently, they represent powerful tools to inform experimental design that will test polymicrobial infection’s assembly and dynamics.

Neutral community assembly

Neutral assembly predicts random community patterns between two similar environments. Originally articulated to explain the community assembly on oceanic islands, MacArthur and Wilson^73,74 hypothesised that as the size of an island increases it would contain a greater number of species than a comparable island with a smaller area: the taxa-area relationship. They proposed that stochastic immigration and extinction by random dispersal from a meta-community of organisms drove this neutral community assembly.

The elegant simplicity of neutral theory makes it an attractive hypothesis that has been explored in a number of microbial systems. For example, Van der Gast⁷⁵ demonstrated that an analogous taxa-volume (rather than area) relationship is applicable to bacteria inhabiting membrane bioreactors. The membrane bioreactors behaved as islands where, with increasing volume, there was increased microbial diversity. Similarly, the community structure of bacteria from the lower respiratory tract of asthmatic patients was shown to follow the predictions of a neutral model.⁷⁶ Consistent with the neutral assembly model, there are a number of other studies which indicate that the flora from CF sputum samples, although affected by CFTR genotype, is patient specific.^15,34,35,37 Likewise, the vaginal tract flora differs between individuals and such differences may account for the susceptibility of some women to BV.⁷⁷

However, the simplicity of neutral theory is problematic in addressing a variety of important questions.⁷¹ For example, the impact of environmental changes, such as those precipitated by therapeutic intervention eliciting temporal shifts in populations. There is also the question of the ‘invasive’ species that are postulated to cause exacerbations in CF and other chronic pulmonary diseases.¹⁶ Finally, there are the functional roles of the species themselves in shaping communities through signalling and trophic interaction. Because neutral theory regards all species as functionally equivalent it is unable to offer predictions about the role of individual taxa.⁷¹

Niche-based community assembly

Niche-based assembly predicts that similar communities will be present at a given time from similar initial starting environmental conditions.⁷⁸ Recent work on the infant gut microbiota suggested a largely deterministic (at least at the phylum level) community assembly process.⁷⁹ Similarly, it is widely held that the normal vaginal flora plays a role in keeping pathogenic bacterial numbers low by exerting environmental gradients.^11,80 The normal vaginal flora is dominated by Lactobacillus or other taxa that produce lactic acid that keeps the pH low (i.e. <4.5), suggesting functional conservation of this feature;⁴⁶ the normal flora competes for nutrients, produces hydrogen peroxide and bacteriocins which directly antagonise the growth of competitor species, and stimulates the host innate immunity.⁸¹ Thus, it has recently been proposed that individual variation in the community structure of the normal vaginal flora can be grouped into relatively few community types and that the assembly of the normal vaginal microbiome may follow a ‘driver-passenger’ model, driven by lactic acid-producing taxa.⁴⁶ In the disease state, it would be hypothesised that ecological interactions occur between bacteria in polymicrobial infections leading to niche differentiation. In BV, for example, some studies indicate that once growth of G. vaginalis is established a likely significant virulence factor is its cytolysin, vaginolysin, which is specific for the lysis of human cells.⁸² Furthermore, sialidase and glycosulfatase production by Prevotella, Bacteroides and Gardnerella species⁸³ likely leads to mucin lysis. Although produced by different species within the community, the effect of these virulence determinants is likely to manifest at the general community level by facilitating access to the epithelium and, thus, increases the adherence of BV ‘pathologic’ core organisms.^84,85 Similarly, it has been shown that the gene expression patterns of P. aeruginosa can be altered when present in co-infection with a member of the oropharyngeal flora, significantly increasing lung damage without increasing P. aeruginosa load. The genes that showed increased expression included several virulence factors, such as lasB, which encodes elastase.⁸⁶

In the case of invasion of an established community by a new species, neutral and niche theory, again, make different predictions. Niche theory predicts resident species will strongly inhibit establishment and growth of species most similar to them because they have similar resource requirements. In contrast, neutral theory predicts a random assembly independent of species traits⁸⁷ but explaining the emergence and persistence of neutral phenotypes is more challenging. It may be that neutral phenotypes persist because several ecologically equivalent mutants arise at the same time, drive their ancestor extinct and then persist with nearly neutral dynamics.⁸⁸

There is some evidence from macro-ecological studies that taxa less related to native species are more invasive, indicating that phylogenetic relatedness of an invader to the native community provides a predictive tool for invasiveness.⁸⁹ In the context of polymicrobial infections there may be significant clinical consequences resulting from such events. For example, the mechanisms underlying exacerbations observed in chronic pulmonary infections are poorly understood but may be caused by invasion of a particular bacterial species, such as those of the S. milleri group.¹⁶ In contrast, it was demonstrated in the mouse gut microbiome that the presence of closely related species could increase the chance of invasion by incoming taxa⁹⁰ and a similar scenario was postulated in the airway of older CF patients in which diversity was decreased and the clonal expansion of colonising Pseudomonadaceae resulted in their predominance.^36,91 Hypermutable strains of P. aeruginosa have been identified in unusually high numbers in CF lung disease⁹² and these mutator strains increase in number as chronic respiratory infection progresses.⁹³ The strains displaying a mutator phenotype have been found to have greater resistance to antibiotics and reduced virulence than non-mutator strains.⁹² It is these adaptations to the harsh CF respiratory environment that may cause dominance of P. aeruginosa in older patients. In BV, individuals who suffer from recurrent infections appear to be re-invaded by BV associated bacteria after antibiotic treatment.⁵⁶

Utilising community assembly theories to aid hypothesis generation and experimental design

It is probably the case that many ecological communities are neither neutral nor strictly hierarchical and competitive.⁹⁴ Ruokolainen et al.⁷⁸ attempted to distinguish between neutral and non-neutral processes in community dynamics by using models run under identical environmental conditions where interspecific competition was assumed to be either niche-based or neutral, yet these models manifested identical patterns.⁷⁸ In other model microbial systems both niche-differentiation and neutral mechanisms could be demonstrated to be significant in maintaining phenotypic diversity within a community.⁹⁵ Such observations have led to attempts to reconcile neutral and niche theory.⁹⁶ Furthermore, a study examining the species of fish inhabiting the Bristol Channel recommends that core taxa should be examined under a niche model because they are specifically adapted to the environment being studied.⁹⁷

The application of niche- or neutral-based theory will predict different outcomes over time for communities in similar environments. Niche or deterministic community assembly will lead to single stable equilibria as species sorting occurs within the environmental gradients present. In contrast, a neutral community assembly model will lead to multiple stable equilibria as species arrive in a random way, with each species being competitively equal and divergent communities assembling in similar environments.⁸⁷ Bacterial communities seem to follow similar species–time relationships to those observed in plant and invertebrate communities, namely that, with increasing time the number of taxa increases and taxa turnover decreases as selective pressure increases, suggesting a switch from stochastic (neutral) to more deterministic (niche) community assembly.^98,99 This echoes macro-ecological studies in which environmental stress was demonstrated to generate communities with a more predictable community composition under similar environmental conditions typical of niche assembly.¹⁰⁰

The impact of the assembly history on the microbial communities of polymicrobial infections has not been extensively studied. The role of stochastic processes in producing communities that are comparatively different and dominated by a few taxa has also been observed in soil microbial communities.⁹⁵ Therefore, large differences in ecosystem functioning can be caused by small differences in species immigration history during community assembly. This has been shown to enable direct manipulation of early immigration, producing a threefold difference in fungal species richness and composition, and equivalent differences in rates of substrate decomposition.¹⁰¹ Recent studies of BV suggest that a normal flora dominated by some Lactobacillus species (e.g. Lactobacillus crispatus) may produce a more stable flora than those dominated by Lactobacillus gasseri and/or Lactobacillus iners.¹⁰² Culture-based studies suggested that the CF lung developed a microbial community with a single stable equilibrium. However, the application of molecular techniques suggests that there are multiple stable equilibria. In effect, communities whose composition reflects their assembly history (even when environmental conditions appear similar and these communities may be qualitatively and quantitatively very different) are frequently dominated by a single taxa, but this is often different between patients.^34,35

Temporal dynamics of microbial communities

The second dimension to temporal effects, that of the community dynamics over time, has received some attention in a number of microbial environments. Bacterial succession in biofilms on inert surfaces showed diversity was unimodal. Initial colonization gave a rapid increase in diversity that rapidly declined as some colonizers are unsuccessful. However, as the biofilm aged new resources and habitats, for example anoxic environments, emerged and diversity increased again.¹⁰³

Temporal effects in CF have been illustrated by a number of studies that have shown that an initial increase in diversity in younger patients is followed by a decrease in diversity as the patients age.^36,37 In addition, the older patient group has more phylogenetically related communities, which may result from disturbance caused by repeated antibiotic treatments or repeated invasions of P. aeruginosa.¹⁰⁴ The first of these demonstrated, in a stratified study, that younger CF patients had a more diverse microbial community than older CF patients and that this was, in part, a result of the younger individuals being culture-negative for P. aeruginosa.³⁷ This may also reflect a temporal switch to more deterministic community assembly mechanisms away from stochastic processes, as has been observed in other bacterial communities,^98,99 which echoes macro-ecological studies in which environmental stress was demonstrated to generate communities with a more predictable community composition under similar environmental conditions typical of niche assembly. The community variance over time during stable clinical conditions has indicated that for the gut microbiota 89–94% community remains stable but over a period of months rather than years.¹⁰⁵ In disease states, a key question is whether there are triggers that lead to the microbial shift from a normal to a pathological flora. For example in BV, what leads to the decline of the Lactobacillus-dominated flora? Various triggers have been proposed, including the temporary alkalinisation of the vagina that can follow sexual activity, which may relieve the microbial antagonism of low pH, and changes associated with the menstrual cycle. It is notable that a vaginal pH > 4.5 is one of the diagnostic criteria for BV and subtle increases in pH may lead to a ‘tipping point’ after which Lactobacillus numbers decline and those of BV-associated bacteria increase. The production of neutralising amines and ammonia by anaerobes may further contribute to this alkalinisation.⁸⁰ Other speculative triggers include the activities of Lactobacillus-specific bacteriophages, although recent data suggest that active lytic phage may be selected against in situ.¹⁰⁶ However, recent data suggest that the onset of BV is a very gradual process, as changes in the vaginal flora can precede the onset of clinical BV by weeks or months.⁴⁵

Conclusion

Microbial systems seem to follow the same species-area and species-time relationships that are observed in macrobiological communities⁹⁹ and, as a result, ecological theory is likely to offer insights into the structure diversity and dynamics of microbial systems. With the emergence of the concept of polymicrobial infections, these studies have proven timely and offer an opportunity to apply ecological theory to the design and interpretation of studies seeking to understand the contribution that polymicrobial communities make to the disease state. Our current technological limitations in characterising microbial diversity are problematic in determining the ‘deep’ diversity of many microbial systems. However, rapid advances and the application of high throughput sequencing and chip technologies means that we are beginning to get an insight into these environments.^36,37

As illustrated here, both BV and CF provide examples of how ecological theory can provide insights into the processes underlying polymicrobial diseases. In clinical situations this may offer opportunities to develop ‘ecological therapies’ that directly disrupt polymicrobial communities,¹⁰⁷ or indirect approaches. For example, probiotic supplements have been shown to reduce pulmonary deterioration in CF patients¹⁰⁸ and may offer the potential to manipulate the polymicrobial communities involved in the disease state via a gut microbiome-mediated impact on the patients’ immune response. Likewise, probiotic therapies may have some utility in the prevention of recurrent BV, particularly if combined with an effective antibiotic therapy that will destabilise the pathologic flora, although larger-scale clinical trials are needed.¹⁰⁹

Clearly Koch’s postulates have provided a benchmark for studies of microbial pathogenicity and their relevance has survived over a century of scientific progress.¹ Indeed, it is possible organisms remain to be discovered that satisfy these postulates with regard to the aetiology of some apparently polymicrobial diseases.²⁰ Nevertheless, there is now substantial evidence to support the proposal that some diseases are genuinely polymicrobial. In these cases, illustrated here by BV and CF, there is a need for a greater understanding and application of ecological theory and, potentially, the formulation of a set of ‘ecological Koch’s postulates’ for these diseases. Indeed, it has recently been proposed that to improve the applicability of Koch’s postulates to the study of oral polymicrobial diseases, the postulates could be reformulated by substituting ‘the bacterium’ with ‘the community’.¹¹⁰ Clearly, we can now agree with Brogden and Guthmiller³ that polymicrobial diseases are a “concept whose time has come”.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Fredericks

Relman

. Sequence-based identification of microbial pathogens: a reconsideration of Koch’s postulates. Clin Microbiol Rev 1996; 9: 18–33.

Falkow

. Molecular Koch’s postulates applied to microbial pathogenicity. Rev Infect Dis 10(Suppl. 2): S274–S276.

Brogden

Guthmiller

. Polymicrobial diseases, a concept whose time has come. ASM News 2003; 69: 69–73.

Bakaletz

. Developing animal models for polymicrobial diseases. Nat Rev Microbiol 2004; 2: 552–568.

Brogden

Guthmiller

Taylor

. Human polymicrobial infections. Lancet 2005; 365: 253–255.

Brook

. Microbiology of polymicrobial abscesses and implications for therapy. J Antimicrob Chemother 2002; 50: 805–110.

Brook

. The role of anaerobic bacteria in cutaneous and soft tissue abscesses and infected cysts. Anaerobe 2007; 13: 171–177.

Robertson

Smith

. The microbiology of the acute dental abscess. J Med Microbiol 2009; 58: 155–162.

Keller

Rampini

Bloemberg

. Detection of a mixed infection in a culture-negative brain abscess by broad-spectrum bacterial 16S rRNA gene PCR. J Clin Microbiol 2010; 48: 2250–2252.

10.

Spiegel

. Bacterial vaginosis. Rev Med Microbiol 2002; 13: 43–51.

11.

Verstraelen

. Bacterial vaginosis: a sexually enhanced disease. Int J STD AIDS 2008; 19: 575–576.

12.

Van Belkum

Renders

Smith

Overbeek

Verbrugh

. Comparison of conventional and molecular methods for the detection of bacterial pathogens in sputum samples from cystic fibrosis patients. FEMS Immunol Med Microbiol 2000; 27: 51–57.

13.

Rogers

Hart

Mason

Hughes

Walshaw

Bruce

. Bacterial Diversity in cases of lung infection in cystic fibrosis patients: 16S ribosomal DNA (rDNA) length heterogeneity PCR and 16S rDNA terminal restriction fragment length polymorphism profiling. J Clin Microbiol 2003; 41: 3548–3558.

14.

Rogers

Carroll

Serisier

Hockey

Jones

Bruce

. characterization of bacterial community diversity in cystic fibrosis lung infections by use of 16s ribosomal DNA terminal restriction fragment length polymorphism profiling. J Clin Microbiol 2004; 42: 5176–5183.

15.

Rogers

Carroll

Serisier

Hockey

Kehagia

Jones

. Bacterial activity in cystic fibrosis lung infections. Respir Res 2005; 6: 49–49.

16.

Sibley

Parkins

Rabin

Duan

Norgaard

Surette

. A polymicrobial perspective of pulmonary infections exposes an enigmatic pathogen in cystic fibrosis patients. Proc Natl Acad Sci USA 2008; 105: 15070–15075.

17.

Tunney

Field

Moriarty

Patrick

Doering

Muhlebach

. Detection of anaerobic bacteria in high numbers in sputum from patients with cystic fibrosis. Am J Respir Crit Care Med 2008; 177: 995–1001.

18.

Peterson

Frank

Pace

Gordon

. Metagenomic approaches for defining the pathogenesis of inflammatory bowel diseases. Cell Host Microbe 2008; 3: 417–427.

19.

Sartor

. Therapeutic correction of bacterial dysbiosis discovered by molecular techniques. Proc Natl Acad Sci USA 2008; 105: 16413–16414.

20.

Tannock

. Molecular analysis of the intestinal microflora in IBD. Mucosal Immunol 2008; 1(Suppl. 1): S15–S18.

21.

Sokol

Lay

Seksik

Tannock

. Analysis of bacterial bowel communities of IBD patients: what has it revealed? Inflamm Bowel Dis 2008; 14: 858–867.

22.

Darveau

. Periodontitis: a polymicrobial disruption of host homeostasis. Nat Rev Microbiol 2010; 8: 481–490.

23.

Bakaletz

. Immunopathogenesis of polymicrobial otitis media. J Leukoc Biol 2010; 87: 213–222.

24.

Stephenson

Mfuna

Dowd

Wolcott

Barbeau

Poisson

. Molecular characterization of the polymicrobial flora in chronic rhinosinusitis. J Otolaryngol Head Neck Surg 2010; 39: 182–187.

25.

Chin

Lim

. Controlled trial of intrapleural streptokinase in the treatment of pleural empyema and complicated parapneumonic effusions. Chest 1997; 111: 275–279.

26.

Maskell

Batt

Hedley

Davies

CWH

Gillespie

Davies

RJO

. The bacteriology of pleural infection by genetic and standard methods and its mortality significance. Am J Respir Crit Care Med 2006; 174: 817–823.

27.

Rochford

Feuer

Orem

Banura

Katongole-Mbidde

Mwanda

. Strategies to overcome myelotoxic therapy for the treatment of Burkitt’s and AIDS-related non-Hodgkin's lymphoma. East Afr Med J 2005; 82(Suppl. 9): S155–S160.

28.

Chene

Donati

Orem

Mbidde

Kironde

Wahlgren

. Endemic Burkitt’s lymphoma as a polymicrobial disease: new insights on the interaction between Plasmodium falciparum and Epstein-Barr virus. Semin Cancer Biol 2009; 19: 411–420.

29.

Gardner

West

Buckling

. Bacteriocins, spite and virulence. Proc Biol Sci 2004; 271: 1529–1535.

30.

Inglis

Gardner

Cornelis

Buckling

. Spite and virulence in the bacterium Pseudomonas aeruginosa. Proc Natl Acad Sci USA 2009; 106: 5703–5707.

31.

Lawn

. AIDS in Africa: the impact of coinfections on the pathogenesis of HIV-1 infection. J Infect 2004; 48: 1–12.

32.

Mallia

Message

Kebadze

Parker

Kon

Johnston

. An experimental model of rhinovirus induced chronic obstructive pulmonary disease exacerbations: a pilot study. Respir Res 2006; 7: 116–116.

33.

Nelson

De Soyza

Bourke

Perry

Cummings

. Assessment of sample handling practices on microbial activity in sputum samples from patients with cystic fibrosis. Lett Appl Microbiol 2010; 51: 272–277.

34.

Harris

De Groote

Sagel

Zemanick

Kapsner

Penvari

. Molecular identification of bacteria in bronchoalveolar lavage fluid from children with cystic fibrosis. Proc Natl Acad Sci USA 2007; 104: 20529–20533.

35.

Bittar

Richet

Dubus

J-C

Reynaud-Gaubert

Stremler

Sarles

. Molecular detection of multiple emerging pathogens in sputa from cystic fibrosis patients. PloS one 2008; 3: e2908–e2908.

36.

Cox

Allgaier

Taylor

Baek

Huang

Daly

. Airway microbiota and pathogen abundance in age-stratified cystic fibrosis patients. PloS one 2010; 5: e11044–e11044.

37.

Klepac-Ceraj

Lemon

Martin

Allgaier

Kembel

Knapp

. Relationship between cystic fibrosis respiratory tract bacterial communities and age, genotype, antibiotics and Pseudomonas aeruginosa. Environ Microbiol 2010; 12: 1293–1303.

38.

Armougom

Bittar

Stremler

Rolain

J-M

Robert

Dubus

J-C

. Microbial diversity in the sputum of a cystic fibrosis patient studied with 16S rDNA pyrosequencing. Eur J Clin Microbiol Infect Dis 2009; 28: 1151–1154.

39.

Hillier

Nugent

Eschenbach

Krohn

Gibbs

Martin

. Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal Infections and Prematurity Study Group. N Engl J Med 1995; 333: 1737–1742.

40.

Leitich

Bodner-Adler

Brunbauer

Kaider

Egarter

Husslein

. Bacterial vaginosis as a risk factor for preterm delivery: a meta-analysis. Am J Obstet Gynecol 2003; 189: 139–147.

41.

Myer

Kuhn

Stein

Wright

Jr Denny

. Intravaginal practices, bacterial vaginosis, and women’s susceptibility to HIV infection: epidemiological evidence and biological mechanisms. Lancet Infect Dis 2005; 5: 786–794.

42.

Ness

Kip

Hillier

Soper

Stamm

Sweet

. A cluster analysis of bacterial vaginosis-associated microflora and pelvic inflammatory disease. Am J Epidemiol 2005; 162: 585–590.

43.

Fredricks

Fiedler

Thomas

Mitchell

Marrazzo

. Changes in vaginal bacterial concentrations with intravaginal metronidazole therapy for bacterial vaginosis as assessed by quantitative PCR. J Clin Microbiol 2009; 47: 721–726.

44.

Verstraelen

Verhelst

. Bacterial vaginosis: an update on diagnosis and treatment. Expert Rev Anti Infect Ther 2009; 7: 1109–1124.

45.

Marrazzo

Thomas

Fiedler

Ringwood

Fredricks

. Risks for acquisition of bacterial vaginosis among women who report sex with women: a cohort study. PloS one 2010; 5: e11139–e11139.

46.

Ravel

Gajer

Abdo

Schneider

Koenig

SSK

McCulle

. Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci USA 2010; 108: 4680–4687.

47.

Lamont

Sobel

Akins

Hassan

Chaiworapongsa

Kusanovic

. The vaginal microbiome: new information about genital tract flora using molecular based techniques. BJOG 2011; 118: 533–549.

48.

Rosenstein

Morgan

Sheehan

Lamont

Taylor-Robinson

. Bacterial vaginosis in pregnancy: Distribution of bacterial species in different gram-stain categories of the vaginal flora. J Med Microbiol 1996; 45: 120–126.

49.

Verhelst

Verstraelen

Claeys

Verschraegen

Van Simaey

De Ganck

. Comparison between Gram stain and culture for the characterization of vaginal microflora: definition of a distinct grade that resembles grade I microflora and revised categorization of grade I microflora. BMC Microbiol 2005; 5: 61–61.

50.

Forsum

Holst

Larsson

Vasquez

Jakobsson

Mattsby-Baltzer

. Bacterial vaginosis—a microbiological and immunological enigma. APMIS 2005; 113: 81–90.

51.

Fredricks

Fiedler

Marrazzo

. Molecular identification of bacteria associated with bacterial vaginosis. N Engl J Med 2005; 353: 1899–1911.

52.

Bradshaw

Tabrizi

Fairley

Morton

Rudland

Garland

. The association of Atopobium vaginae and Gardnerella vaginalis with bacterial vaginosis and recurrence after oral metronidazole therapy. J Infect Dis 2006; 194: 828–836.

53.

Ferris

Norori

Zozaya-Hinchliffe

Martin

. Cultivation-independent analysis of changes in bacterial vaginosis flora following metronidazole treatment. J Clin Microbiol 2007; 45: 1016–1018.

54.

Fredricks

Fiedler

Thomas

Oakley

Marrazzo

. Targeted PCR for detection of vaginal bacteria associated with bacterial vaginosis. J Clin Microbiol 2007; 45: 3270–3276.

55.

Oakley

Fiedler

Marrazzo

Fredricks

. Diversity of human vaginal bacterial communities and associations with clinically defined bacterial vaginosis. Appl Environ Microbiol 2008; 74: 4898–4909.

56.

Srinivasan

Liu

Mitchell

Fiedler

Thomas

Agnew

. Temporal variability of human vaginal bacteria and relationship with bacterial vaginosis. PloS one 2010; 5: e10197–e10197.

57.

Zozaya-Hinchliffe

Lillis

Martin

Ferris

. Quantitative PCR assessments of bacterial species in women with and without bacterial vaginosis. J Clin Microbiol 2010; 48: 1812–1819.

58.

Swidsinski

Mendling

Loening-Baucke

Ladhoff

Swidsinski

Hale

. Adherent biofilms in bacterial vaginosis. Obstet Gynecol 2005; 106: 1013–1023.

59.

Patterson

Stull-Lane

Girerd

Jefferson

. Analysis of adherence, biofilm formation and cytotoxicity suggests a greater virulence potential of Gardnerella vaginalis relative to other bacterial-vaginosis-associated anaerobes. Microbiology 2010; 156: 392–399.

60.

Macones

Parry

Elkousy

Clothier

Ural

Strauss

. A polymorphism in the promoter region of TNF and bacterial vaginosis: preliminary evidence of gene-environment interaction in the etiology of spontaneous preterm birth. Am J Obstet Gynecol 2004; 190: 1504–1508.

61.

Cantón

del Campo

. Cystic fibrosis: deciphering the complexity. Clin Microbiol Infect 2010; 16: 793–797.

62.

Foweraker

. Recent advances in the microbiology of respiratory tract infection in cystic fibrosis. Br Med Bull 2009; 89: 93–110.

63.

Giraud

Pihet

Razafimandimby

Carrère

Degand

Mely

. Geosmithia argillacea: an emerging pathogen in patients with cystic fibrosis. J Clin Microbiol 2010; 48: 2381–2386.

64.

Van Der Gast

Walker

Stressmann

Rogers

Scott

Daniels

. Partitioning core and satellite taxa from within cystic fibrosis lung bacterial communities. ISME J 2010; 5: 780–791.

65.

Rogers

Stressmann

Walker

Carroll

Bruce

. Lung infections in cystic fibrosis: deriving clinical insight from microbial complexity. Expert Rev Mol Diagn 2010; 10: 187–196.

66.

Armstrong

Grimwood

Carlin

Carzino

Hull

Olinsky

. Severe viral respiratory infections in infants with cystic fibrosis. Pediatr Pulmonol 1998; 26: 371–379.

67.

Hiatt

Grace

Kozinetz

Raboudi

Treece

Taber

. Effects of viral lower respiratory tract infection on lung function in infants with cystic fibrosis. Pediatrics 1999; 103: 619–626.

68.

Sikkel

Quint

Mallia

Wedzicha

Johnston

. Respiratory syncytial virus persistence in chronic obstructive pulmonary disease. Pediatr Infect Dis J 2008; 27(Suppl. 10): S63–S70.

69.

Hilty

Burke

Pedro

Cardenas

Bush

Bossley

. Disordered microbial communities in asthmatic airways. PloS one 2010; 5: e8578–e8578.

70.

Sethi

Evans

Grant

BJB

Murphy

. New strains of bacteria and exacerbations of chronic obstructive pulmonary disease. N Engl J Med 2002; 347: 465–471.

71.

Chase

. Towards a really unified theory for metacommunities. Funct Ecol 2005; 19: 182–186.

72.

Harpole

Tilman

. Non-neutral patterns of species abundance in grassland communities. Ecol Lett 2006; 9: 15–23.

73.

MacArthur

Wilson

. An equilibrium theory of insular zoogeography. Evolution 1963; 17: 373–387.

74.

MacArthur

Wilson

. The Theory of Island Biogeography , Princeton, NJ: Princeton University Press, 1967.

75.

Van Der Gast

Jefferson

Reid

Robinson

Bailey

Judd

. Bacterial diversity is determined by volume in membrane bioreactors. Environ Microbiol 2006; 8: 1048–1055.

76.

Sloan

Lunn

Woodcock

Head

Nee

Curtis

. Quantifying the roles of immigration and chance in shaping prokaryote community structure. Environ Microbiol 2006; 8: 732–740.

77.

Zhou

Brown

Abdo

Davis

Hansmann

Joyce

. Differences in the composition of vaginal microbial communities found in healthy Caucasian and black women. ISME J 2007; 1: 121–133.

78.

Ruokolainen

Ranta

Kaitala

Fowler

. When can we distinguish between neutral and non-neutral processes in community dynamics under ecological drift? Ecol Lett 2009; 12: 909–919.

79.

Trosvik

Stenseth

Rudi

. Convergent temporal dynamics of the human infant gut microbiota. ISME J 2010; 4: 151–158.

80.

Pybus

Onderdonk

. Microbial interactions in the vaginal ecosystem, with emphasis on the pathogenesis of bacterial vaginosis. Microbes Infect 1999; 1: 285–292.

81.

Valore

Wiley

Ganz

. Reversible deficiency of antimicrobial polypeptides in bacterial vaginosis. Infect Immun 2006; 74: 5693–5702.

82.

Gelber

Aguilar

Lewis

KLT

Ratner

. Functional and phylogenetic characterization of Vaginolysin, the human-specific cytolysin from Gardnerella vaginalis. J Bacteriol 2008; 190: 3896–3903.

83.

Briselden

Moncla

Stevens

Hillier

. Sialidases (neuraminidases) in bacterial vaginosis and bacterial vaginosis-associated microflora. J Clin Microbiol 1992; 30: 663–666.

84.

Cauci

Driussi

Monte

Lanzafame

Pitzus

Quadrifoglio

. Immunoglobulin A response against Gardnerella vaginalis hemolysin and sialidase activity in bacterial vaginosis. Am J Obstet Gynecol 1998; 178: 511–515.

85.

Roberton

Wiggins

Horner

Greenwood

Crowley

Fernandes

. A novel bacterial mucinase, glycosulfatase, is associated with bacterial vaginosis. J Clin Microbiol 2005; 43: 5504–5508.

86.

Duan

Dammel

Stein

Rabin

Surette

. Modulation of Pseudomonas aeruginosa gene expression by host microflora through interspecies communication. Mol Microbiol 2003; 50: 1477–1491.

87.

Fargione

Brown

Tilman

. Community assembly and invasion: an experimental test of neutral versus niche processes. Proc Natl Acad Sci USA 2003; 100: 8916–8920.

88.

Zhang

Q-G

Buckling

Godfray

HCJ

. Quantifying the relative importance of niches and neutrality for coexistence in a model microbial system. Funct Ecol 2009; 23: 1139–1147.

89.

Strauss

Webb

Salamin

. Exotic taxa less related to native species are more invasive. Proc Natl Acad Sci USA 2006; 103: 5841–5845.

90.

Stecher

Chaffron

Käppeli

Hapfelmeier

Freedrich

Weber

. Like will to like: abundances of closely related species can predict susceptibility to intestinal colonization by pathogenic and commensal bacteria. PLoS Pathog 2010; 6: e1000711–e1000711.

91.

Zemanick

Harris

Conway

Konstan

Marshall

Quittner

. Measuring and improving respiratory outcomes in cystic fibrosis lung disease: opportunities and challenges to therapy. J Cyst Fibros 2010; 9: 1–16.

92.

Oliver

Cantón

Campo

Baquero

Blázquez

. High frequency of hypermutable Pseudomonas aeruginosa in cystic fibrosis lung infection. Science 2000; 288: 1251–1254.

93.

Ciofu

Riis

Pressler

Poulsen

Høiby

. Occurrence of hypermutable Pseudomonas aeruginosa in cystic fibrosis patients is associated with the oxidative stress caused by chronic lung inflammation. Antimicrob Agents Chemother 2005; 49: 2276–2282.

94.

Alonso

Etienne

McKane

. The merits of neutral theory. Trends Ecol Evol 2006; 21: 451–457.

95.

Dumbrell

Nelson

Helgason

Dytham

Fitter

. Relative roles of niche and neutral processes in structuring a soil microbial community. ISME J 2010; 4: 337–345.

96.

Tilman

. Niche tradeoffs, neutrality, and community structure: a stochastic theory of resource competition, invasion, and community assembly. Proc Natl Acad Sci USA 2004; 101: 10854–10861.

97.

Magurran

. Species abundance distributions over time. Ecol lett 2007; 10: 347–354.

98.

Van Der Gast

Ager

Lilley

. Temporal scaling of bacterial taxa is influenced by both stochastic and deterministic ecological factors. Environ Microbiol 2008; 10: 1411–1418.

99.

Redford

Fierer

. Bacterial succession on the leaf surface: a novel system for studying successional dynamics. Microb Ecol 2009; 58: 189–198.

100.

Chase

. Drought mediates the importance of stochastic community assembly. Proc Natl Acad Sci USA 2007; 104: 17430–17434.

101.

Fukami

Dickie

Paula Wilkie

Paulus

Park

Roberts

. Assembly history dictates ecosystem functioning: evidence from wood decomposer communities. Ecol Lett 2010; 13: 675–684.

102.

Verstraelen

Verhelst

Claeys

De Backer

Temmerman

Vaneechoutte

. Longitudinal analysis of the vaginal microflora in pregnancy suggests that L. crispatus promotes the stability of the normal vaginal microflora and that L. gasseri and/or L. iners are more conducive to the occurrence of abnormal vaginal microflora. BMC Microbiol 2009; 9: 116–116.

103.

Jackson

. Changes in community properties during microbial succession. OIKOS 2003; 101: 444–448.

104.

Döring

Parameswaran

Murphy

. Differential adaptation of microbial pathogens to airways of patients with cystic fibrosis and chronic obstructive pulmonary disease. FEMS Microbiol Rev 2010; 35: 124–146.

105.

Donskey

Hujer

Das

Pultz

Bonomo

Rice

. Use of denaturing gradient gel electrophoresis for analysis of the stool microbiota of hospitalized patients. J Microbiol Methods 2003; 54: 249–256.

106.

Martin

Soberon

Escobedo

Suarez

. Bacteriophage induction versus vaginal homeostasis: role of H₂O₂ in the selection of Lactobacillus defective prophages. Int Microbiol 2009; 12: 131–136.

107.

Harrison F. Microbial ecology of the cystic fibrosis lung. Microbiology 2007; 153(Pt 4): 917–923.

108.

Weiss

Bujanover

Yahav

Vilozni

Fireman

Efrati

. Probiotic supplementation affects pulmonary exacerbations in patients with cystic fibrosis: a pilot study. Pediatr Pulmonol 2010; 45: 536–540.

109.

Senok

Verstraelen

Temmerman

Botta

. Probiotics for the treatment of bacterial vaginosis. Cochrane Database Syst Rev 2009; CD006289–CD006289.

110.

Kolenbrander

Palmer

Periasamy

Jakubovics

. Oral multispecies biofilm development and the key role of cell-cell distance. Nat Rev Microbiol 2010; 8: 471–480.

Polymicrobial challenges to Koch’s postulates: Ecological lessons from the bacterial vaginosis and cystic fibrosis microbiomes

Abstract

Keywords

Introduction

Bacterial vaginosis as a representative microbial shift disease

Cystic fibrosis (CF) and inherited disease

What can we learn from other fields?

Neutral community assembly

Niche-based community assembly

Utilising community assembly theories to aid hypothesis generation and experimental design

Temporal dynamics of microbial communities

Conclusion

Footnotes

Funding

References