Intracellular parasitic protozoans

Parasitic protozoans that are mainly intracellular are specialized to evade and to manipulate the immune response. For this reason, these infections frequently lead to immunopathologies in which the progress of the disease will depend on the type and level of the immune response. In this scenario, the potential benefit of the human microbiota is to provide a more effective immune response against infections [3]. Although such correlations have been found, the effects of the microbiota on host immune response to protozoal infections have only been examined in a few examples, as discussed below.

Apicomplexans (phylum Apicomplexa) are obligatory intracellular parasites exhibiting an unusual type of motility and associated organelles necessary for host cell invasion. They also harbour a unique plastid-type of organelle (the apicoplast) and display asexual and sexual modes of reproduction with formation of spores. Coccidians are apicomplexan parasites that utilise intestinal epithelial cells of vertebrates (including humans) as a transient or final habitat. The impact of gut microbes on the infection of the coccidian parasites Cryptosporidium and Eimeria is variable and unclear. In Cryptosporidium parvum, an opportunistic intestinal coccidian parasite of humans, germ-free and immunodeficient mice develop heavy infections after a few weeks, which is in sharp contrast to the same immunodeficient mice harbouring a normal microbiota [11]. On the other hand, some species of Eimeria (intestinal coccidian parasites of many vertebrates) fail to infect germ-free animals as compared to conventionally reared and microbial-colonized animals [12,13].

T. gondii is a model organism among apicomplexan parasites. It causes toxoplasmosis, which, despite being clinically asymptomatic in a vast majority of cases, is the most prevalent human parasitic infection worldwide [14]. The very successful spread of T. gondii in human populations relies on its unusual biology and the way that humans respond to this infection, as described below. Like no other intracellular parasitic protozoan, T. gondii is extremely host generalist, being able to invade and multiply in any nucleated cell of any warm-blooded animal. In addition, T. gondii is easily transmitted to humans by ingestion of either environmentally-resistant spores released from cat faeces or pseudocysts present in infected meat from various sources. Lastly, but no less important, a protective and Th1-polarized adaptive immune response normally develops in immunocompetent individuals [15,16], and human infections become clinically silent in most cases. Gut commensal bacteria might largely contribute to the development of this protective immune response and the associated pathology [17], as discussed below (Fig 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Initiation of mucosal innate immune response via dendritic cells against Toxoplasma gondii infection in mice.

The toll-like receptor (TLR)-adaptor protein MyD88 is a key element to the protective response based on production of IL-12. Secretion of IL-12 will trigger an effective cellular-based immune response with production of INF-γ and activation of a Th1 T lymphocyte profile. (A) This innate response is mainly dependent on TLR11, which forms endolysosomal dimers with TLR12 that recognize profilin from T. gondii. This recognition is central to mucosal immunity triggering production of IL-12. (B) In the absence of TLR11, however, this response is still minimally and sufficiently compensated by indirect stimulation provided by the gut microbial commensals via TLR2, TLR4, and TLR9 [44]. In this case, infection-induced cell destruction and intestinal dysbiosis apparently trigger loss of tolerance to gut commensals. When the gut microbiota is severely reduced by prolonged antibiotic treatment, the following observations can be made: (C) Wild-type mice expressing TLR11 exhibit a reduced but not abolished IL-12 response. These animals can still build up Th1 immunity. (D) TLR11-knockout mice are unable to mount IL-12 responses against this parasite, and Th1 immunity is severely impaired. In conclusion, gut commensals serve as natural molecular adjuvants during T. gondii infection.

https://doi.org/10.1371/journal.pntd.0004176.g002

When T. gondii reaches the gut of an individual with a native and functional microbiota, an immune response is initiated at the level of the intestinal mucosa by activation of toll-like receptors (TLRs) in dendritic cells (DCs) (Fig 2). In vitro studies point to the involvement of various TLRs such as TLR2, TLR3, TLR4, TLR7, and TLR9 in response to T. gondii infection [18]. However, experiments in mice reveal that recognition of T. gondii profilin by the dimer TLR11/12 has a key role in the immune response against oral infection (Fig 2A and 2C) [18,19]. Profilin is the major parasite ligand that activates TLR11. Because profilin is an intracellular protein, it is either released by an unknown mechanism or taken up by DCs when they phagocytose dead parasites and debris. This recognition activates the transcription factor interferon (INF) regulatory factor 8, leading to the production of interleukin 12 (IL-12) [20,21]. IL-12 promotes a cellular-based immunity with production of INF-γ from natural killer cells and differentiation of Th1 T lymphocytes. As described below, these cells and molecules have a role in immunopathogenesis of T. gondii infections and are also required to initiate the adaptive phase of the immune response [18].

This IL-12 response in mice depends on the TLR-adaptor MyD88 and the myeloid differentiation factor 88 [22]. This molecule has a central role in the immune response against T. gondii oral infection (Fig 2). In MyD88 genetically ablated mice, IL-12 response is non-existent, and these knockout mice perish within two weeks after oral infection [23]. TLR-adaptor molecules, such as MyD88, can respond to many types of TLRs. MyD88 initiates various intracellular signalling pathways, but the resulting balance of co-stimulatory molecules depends on the context of the TLR ligands [24]. In the context of TLR11/12, it triggers the release of IFN-γ via the NF-κB pathway, which stimulates macrophages and CD8α^- DCs “priming” the immune response [18].

Importantly, TLR11^-/- mice are not completely impaired to respond to T. gondii as long as gut commensal bacteria are present (Fig 2B and 2D). In the presence of gut commensals (Fig 2B), partial resistance is retained with a decreased but existent IL-12 response that is sufficient to produce similar amounts of INF-γ as compared to wild-type mice. Therefore, TLR11^-/- mice survive the acute phase of infection. When TLR11^-/- mice are treated with antibiotics in order to eliminate gut commensal bacteria, this remaining protective IL-12 response is lost (Fig 2D). This protective response can be rescued with oral administration of bacterial lipopolysaccharide (LPS) [17]. Furthermore, TLR-knockout mice indicate the involvement of TLR2, TLR4, and TLR9 in the ability of mounting a minimally sufficient INF-γ response to the parasite. These TLRs are all MyD88-dependent but do not respond to T. gondii infection per se. LPS, for instance, is a bacterial agonist of TLR4. Therefore, gut commensal bacteria function as a natural molecular adjuvant during oral infection by T. gondii, providing an indirect stimulation of mucosal DCs. The natural community of gut microbes are apparently necessary since Bacteroides, an abundant genus of gut bacterium, cannot provide this stimulation alone [17].

Epithelial damage, resulting from initial replication of T. gondii inside intestinal epithelial cells, should allow translocation of lumen bacteria to the lamina propria. In addition, T. gondii oral infection leads to intestinal dysbiosis (i.e., bacterial imbalance) with deregulation or loss of Paneth cells [19]. These specialized intestinal epithelial cells produce antimicrobial peptides, which are important for maintaining bacterial homeostasis in the gut. T. gondii-induced dysbiosis is characterized by a decrease of gut bacterial diversity with a shift toward gram-negative bacteria, a key source of LPS [19]. Commensal-derived flagellin, the ubiquitous and main protein of a bacterial flagellum, could also be a source of indirect stimulation of T lymphocytes [25]. Altogether, DCs apparently trigger a loss of immunological tolerance against the gut bacterial flora.

This indirect immunological stimulation might be important to the immunopathology of T. gondii infection in mice. In the absence of TLR11, gut bacteria not only provide a sufficient Th1 response but also preclude the development of the immunopathology of toxoplasmosis in orally infected mice [17]. Humans lack a functional TLR11, which is found as a pseudogene, and an equivalent TLR sensor for profilin has not been found. In addition, it is unclear if TLR12 alone can accomplish the role of TLR11 [18]. The involvement of other TLRs and a TLR-independent mechanism for parasite recognition in human oral infections are not yet clear [18]. Meanwhile, it is tempting to speculate an interesting evolutionary scenario involving host, parasite, and bacterial microbiota. The co-evolution of T. gondii, humans, and their gut microbiota has had major impact on each other, favouring survival of both host and parasite. The co-evolution of this triad has resulted in preservation of the host, which is characterized by a general lack of life-threatening symptoms in infected humans, and has contributed to the successful spread of T. gondii in human populations worldwide.

Extracellular parasitic protozoans

Some parasitic protozoans of humans are exclusively extracellular. Examples are Entamoeba histolytica, Giardia lamblia, and Trichomonas vaginalis. Living outside of human cells and tissues, these parasites should physically interact with the native mucosal microbiota. By maintaining intimate and cooperative relationships among themselves and with the host, the native bacterial microbiota of the human gut and vaginal mucosa might represent a natural barrier to pathogen invasion. The interactions between extracellular parasitic protozoans and the human mucosal microbiota, their effects, and their mechanisms, are presented in this section.

E. histolytica is a human-specific extracellular parasite found in the lumen of the large intestine, the site that contains the largest bacterial population density in humans. The severity of ameobiasis is linked to the ability of this parasite to leave the intestinal lumen and to destroy the intestinal mucosa, promoting haemorrhagic dysentery and disseminating to other organs, a severe condition known as invasive or extraintestinal ameobiasis [26]. In early studies, the virulence of E. histolytica was shown to be significantly enhanced by intestinal bacteria in microbiota-controlled animal models [27,28]. Further insights into this microbial relationship have been gained under in vitro and laboratory-controlled experimental conditions [29,30].

Modulation of E. histolytica virulence by intestinal bacteria varies with time, species or strains of bacteria, and strains of the parasite [31–35]. The bacterial strain of Escherichia coli serotype O55 binds strongly to the surface Gal/GalNAc lectin of E. histolytica because of its natural surface carbohydrate composition rich in galactose and N-acetyl galactosamine. Recognition of E. coli O55 by E. histolytica via Gal/GalNAc, a central virulence factor of this amoeba, enhances its virulence in the first hour probably as a result of the activation of downstream pathways triggered by Gal/GalNAc recognition. However, in the course of the first month of monoaxenic cultivation, there is a significant reduction in virulence, which is restored after three months of further co-cultivation. This influence on the amoeba’s virulence was observed at different levels, such as regulation of Gal/GalNAc expression, phagocytosis, proteolysis, adhesion, and cytotoxicity. The interaction between this parasite and endogenous bacteria causes substantial changes in the amoeba’s gene expression, suggesting that some of these bacteria may also have a nutritional role that supports amoebic growth [31,34].

Other intestinal and non-intestinal pathogenic bacteria including E. coli, Shigella dysenteriae, and Staphylococcus aureus also promote substantial changes at both E. histolytica virulence and host response. Significant enhancement of virulence properties of E. histolytica was observed with an increase of Gal/GalNAc expression, proteolysis, adhesion, and cytotoxicity. Notably, this enhancement on the amoeba’s virulence was only seen with these pathogenic bacteria and not with non-pathogenic commensal E. coli. In addition, Entamoeba dispar (an intestinal non-pathogenic amoeba closely related to E. histolytica) was unable to promote these effects even when combined with the same pathogenic bacteria. Importantly, these pathogenic bacteria also altered the host response with respect to epithelial barrier function, chemoattraction of neutrophils, and inflammatory response [32,35]. Altogether, the aforementioned studies suggest that synergistic effects of some intestinal bacteria on the host and parasite responses may provide an environment more permissible for parasite invasion, which could contribute to disease development. Because the human gut microbiome is a complex system involving at least hundreds of bacterial species, these initial observations may still represent an oversimplification of what happens in intestinal infections by E. histolytica under natural conditions.

Giardia lamblia is another extracellular parasitic protozoan found in the lumen of the human intestine. However, in contrast to E. histolytica, G. lamblia is not invasive and resides in the small intestine, where microorganisms are not as abundant [36]. Individuals with G. lamblia infection (i.e., giardiasis) display alterations in the bacterial composition of the upper digestive tract [37,38]. The intestinal bacterial microbiota may potentially influence the outcomes of giardiasis, but it is unclear if these microbial alterations are cause or effect. Although lactobacilli can inhibit G. lamblia proliferation in vitro [39] or could help as a probiotic intervention in animal models [40–42], it is unclear if these bacteria are autochthonous or indigenous inhabitants of the site of G. lamblia infection in humans. An intriguing report has shown that natural resistance to G. lamblia infection between mice of similar genetic background but originating from different breeders can be transferred between these animals solely by housing animals together for few weeks [43]. Although antibiotic treatments and partial characterization of these microbial communities suggest a protective role of the gut microbiota, the mechanisms underlying resistance to G. lamblia infection in these animals remain elusive. The existing evidence does not yet clarify the role of the gut microbiota in giardiasis.

T. vaginalis is the only parasitic protozoan of the human genital tract. It causes trichomoniasis, the most common sexually transmitted infection of non-viral etiology worldwide [2]. In the vagina of women of childbearing age, this extracellular parasite invades the natural habitat of a dense and resilient community of bacteria. The vaginal microbiome of 396 asymptomatic North American women was recently characterized into five community state types (Fig 1, pie chart). Four of these are dominated by different species of Lactobacillus—L. iners, L. crispatus, L. gasseri, and L. jensenii—with a relatively high (and almost exclusive) taxon abundance. Except for L. iners, lactobacilli are poorly represented in the fifth community type, which is predominantly composed of anaerobic bacteria such as Atopobium vaginae, Prevotella bivia, Megasphaera sp., Sneathia sp., and Gardnerella vaginalis [9]. Eleven cases of asymptomatic T. vaginalis infections (i.e., 2.8%) were found in this group of women [44]. Observing the distribution of T. vaginalis infections amongst these bacterial community types, two important observations can be made. Firstly, there was a low abundance of lactobacilli in the vagina of 73% of the T. vaginalis- infected women (8/11), which is in agreement with previous clinical reports [45]. The exception, once again, was L. iners, which was found in a taxon abundance of >80% in the vagina of 18% of T. vaginalis-infected women (2/11). Secondly, those T. vaginalis infections lacking lactobacilli (8/11) were associated with the fifth community type, containing mostly anaerobic bacteria. Such microbial shift, with exclusion of lactobacilli and higher prevalence of these anaerobes, is typically associated with a common disease condition known as bacterial vaginosis (BV). L. iners has also been linked to this condition [46].

A note of caution, however, should be added. Firstly, this is a representation of an asymptomatic population displaying a relatively small number of T. vaginalis infections (11/396). Secondly, BV is a common condition in African-American populations [47], and this cohort was highly represented in this study [9,44]. Additional metagenomics studies comprising various ethnic groups may be needed to confirm unequivocally the existence of a strict association between T. vaginalis infection and a specific vaginal microbial community. If such a microbial association exists, two hypotheses might be drawn: (i) Lactobacilli and T. vaginalis are competitors interacting antagonistically in their natural environment; (ii) BV bacteria and T. vaginalis interact cooperatively, and the disease trichomoniasis might result from the interactions of T. vaginalis with one or more types of BV bacteria.

Using polymicrobial infection models in tissue culture, two recent studies give support to the hypotheses above [48,49]. Firstly, T. vaginalis was found to reduce numbers of Lactobacillus acidophilus, L. jensenii, and L. crispatus [48]. Secondly, T. vaginalis adhesion to human cells (a key aspect of its virulence) was significantly inhibited by L. gasseri in strain-specific and contact-dependent manners [49]. Finally, two common BV-associated bacteria, Atopobium and Gardnerella, were found to cause synergistic enhancement of T. vaginalis-induced chemokines [48]. In addition, clinical isolates of T. vaginalis often harbour mycoplasmas. Mycoplasma hominis, in particular, is frequently associated with enhanced inflammation during trichomoniasis [50]. M. hominis was shown to upregulate proinflammatory responses of human monocytes to T. vaginalis infections in vitro in a synergistic way [51].

Taken together, these recent studies suggest that microbial associations of this nature might influence significantly the outcomes of trichomoniasis by favouring both the growth of associated pathogenic bacteria and creating an inflammatory environment more permissible for disease development.