Parachlamydia acanthamoebae is a strict intracellular bacterium which naturally infects free-living amoebae. Like other members of the
Chlamydiales order, it exhibits a two-stage developmental cycle with infectious elementary bodies and metabolically active replicating reticulate bodies (
25). Several pieces of evidence support the role of
P. acanthamoebae as a new agent of lower respiratory tract infection (reviewed in references
18 and
32). The first hint was provided by the recovery of
P. acanthamoebae strain Hall's coccus from the water of a humidifier associated with an outbreak of fever and the presence of anti-
Parachlamydia antibodies among exposed individuals (
2). Additional serological studies demonstrated a higher seropositivity rate among patients with pneumonia than among controls (
20,
33). Furthermore, parachlamydial DNA was detected by PCR in mononuclear cells of a patient with bronchitis and in sputa and bronchoalveolar lavage samples from patients with lower respiratory tract infections (
11,
12,
19,
37). Moreover,
Parachlamydia infects human pneumocytes and macrophages
in vitro (
7,
23,
24). Finally, we recently developed an experimental model in which mice injected with living
Parachlamydia showed signs of severe pneumonia and bacterial localization in cells that were likely pneumocytes and macrophages (
9,
10).
The sensing of invasive pathogens by innate immune cells relies on their capacity to sense microbial molecular motifs through pattern recognition receptors. Toll-like receptors (TLRs) expressed on the surface or in the endosomes of immune cells allow the detection of microbially derived molecular structures such as lipids, proteins, and nucleic acids. TLR4 is an obligate partner for the host response to bacterial lipopolysaccharide (LPS) (endotoxin) and most Gram-negative bacteria (
1). TLR2, which generates heterodimers in combination with either TLR1 or TLR6, has been reported to recognize a broad range of microbial compounds, among which are lipopeptides, lipoproteins, peptidoglycan subcomponents, and β-glucans (
29). The activation of the intracellular signaling pathways upon microbial recognition by TLRs engages several adaptor molecules. Myeloid differentiation primary response gene 88 (MyD88) specifies most TLRs and the TLR4 MyD88-dependent signaling pathway, which is involved in the activation of mitogen-activated protein kinases and nuclear factor κB and in the generation of proinflammatory cytokines and immune-related genes (
35).
Although
Parachlamydia may represent an emerging agent of pneumonia (
18,
26), very little is known about its recognition by innate immune cells. Unfortunately, the mechanisms involved in the sensing of
Chlamydia are most likely irrelevant for
Parachlamydia given the specificities of this bacterium, such as its ability to replicate within amoebae (
22) and its genome size, which is more than twice that of
Chlamydia (
21). In addition, the bioinformatics-based annotation of the genome of the
Parachlamydia-related symbiont UWE25 indicates that it likely possesses a truncated LPS and lacks most immunogenic outer membrane proteins present in the
Chlamydiaceae (
27). Considering the central role played by TLRs in microbial sensing, we studied the contributions of MyD88, TLR2, and TLR4 in the recognition of
P. acanthamoebae by macrophages
in vitro and in the outcome of parachlamydial infection in an experimental model of pneumonia in mice. Our results show that living
P. acanthamoebae enters but does not multiply in macrophages and that heat-inactivated
P. acanthamoebae stimulates cytokine production by macrophages in a MyD88/TLR4-dependent manner and to a lesser extend through TLR2. Furthermore, MyD88
−/−, TLR4
−/−, and TLR2
−/− mice are resistant to
P. acanthamoebae infection. Taken together, these data indicate that
P. acanthamoebae weakly stimulates the innate immune system, which may allow the bacterium to survive in infected cells despite its low replication capacity.
DISCUSSION
Clinical and experimental data suggest that
P. acanthamoebae represents a new agent of lower respiratory tract infection (
18). The characterization of the innate immune receptors engaged by
P. acanthamoebae is essential for improving our understanding of the pathogenesis of
P. acanthamoebae infection. Analyses of the response of macrophages to
P. acanthamoebae revealed a critical role for the MyD88/TLR4 pathway, and to a lesser extent TLR2, in the production of proinflammatory cytokines. Yet, heat-inactivated bacteria were much more immunogenic than live bacteria. In line with these results, mice deficient in either MyD88, TLR4, or TLR2 were resistant to lung infection with
P. acanthamoebae.
P. acanthamoebae quickly entered BMDMs.
P. acanthamoebae internalization was not affected in macrophages deficient in either TLR2 or TLR4, corroborating the notion that TLRs do not function primarily as phagocytic receptors (
46). Of note, the initial rate of infection of BMDMs (around 10 bacteria per BMDM) was similar to that of human macrophages, A549 pneumocytes, and HEL lung fibroblasts (two to four bacteria per infected cell [
7,
24]). Nonetheless, upon internalization,
P. acanthamoebae did not multiply inside BMDMs, whereas it grows poorly in human macrophages, A549, and HEL cells (
7,
24). Overall, these data suggest that
P. acanthamoebae has a rather limited replication capacity in innate immune cells compared to amoebae, where an exponential growth of 2 log units was observed (
22).
Unexpectedly, heat-inactivated
P. acanthamoebae induced the production of substantial levels of cytokines by macrophages, whereas living bacteria only weakly stimulated cytokine production. Most likely, heat inactivation exposed an immunogenic determinant, probably a protein given its partial susceptibility to trypsin, hidden in living bacteria. An obvious candidate is a heat shock protein (HSP). Indeed, the genome of
P. acanthamoebae contains several genes encoding putative HSPs (
21), and HSP60 from
Chlamydia pneumoniae and
Chlamydia trachomatis have been reported to strongly stimulate innate immune cells (
4,
13,
31). Further investigations will be required to characterize the structure of the microbially associated molecular pattern(s) of
P. acanthamoebae that stimulates innate immune cells.
The recognition of heat-inactivated
P. acanthamoebae by macrophages occurred mainly through MyD88/TLR4. Yet, TNF and IL-6 production by TLR2
−/− BMDMs was also reduced to some extent. Interestingly, the MyD88/TLR4 pathway has previously been reported to be involved in the sensing of
C. pneumoniae and
C. pneumoniae HSP60 (
4,
42,
43) and to mediate the acute pulmonary inflammation in mice following the intratracheal instillation of
C. pneumoniae HSP60 (
5). Furthermore, the TLR4 pathway was implicated in
C. trachomatis HSP60-induced trophoblast apoptosis (
16). Besides TLR4, TLR2 was shown to play a primary role in the activation of dendritic cells by
C. pneumoniae and
C. pneumoniae HSP60 (
13,
38) and to mediate
C. pneumoniae-induced macrophage foam cell formation (
6). Moreover, TLR2 was implicated in oviduct pathology resulting from genital tract infection with
C. trachomatis (
14) and triggered the early immune response to
C. pneumoniae pulmonary infection (
39). Further substantiating a role for TLRs in the sensing of
Chlamydiaceae,
C. pneumoniae-infected MyD88
−/− mice showed impaired production of cytokines and chemokines and delayed recruitment of T cells, and they ultimately died from severe chronic lung infection (
34). Altogether, these data point toward a fundamental role of MyD88, TLR2, and/or TLR4 in detecting
Chlamydia spp., in particular
Chlamydia HSPs. In the future, it might be interesting to study the role of intracellular sensors of bacteria such as nucleotide-binding and oligomerization domain-like receptors (NLRs), which have been shown to mediate a cell response to
C. pneumoniae,
C. trachomatis, and
Chlamydia muridarum (
3,
36,
44,
47).
Models of lung infection have underscored the vital role of the early inflammatory response in mediating bacterial clearance and conditioning survival. As mentioned above, mice deficient in MyD88, TLR4, and TLR2 ultimately developed a severe and chronic lung inflammation involved in morbidity and mortality after
C. pneumoniae infection (
34,
39,
44). These observations demonstrate that TLRs exert a two-sided role in
Chlamydia infections, acting on the one hand as key initiators of the host anti-
Chlamydia defense responses and on the other hand as important contributors to tissue damage and
Chlamydia pathogenicity. Astonishingly at first sight, MyD88, TLR4, and TLR2 deficient mice were fully resistant to infection with
P. acanthamoebae. Whereas these data could argue for a minor pathogenic role of
P. acanthamoebae in healthy individuals, it is worthwhile to remember that most documented cases of probable infection with
P. acanthamoebae were reported in immunocompromised HIV-infected or grafted patients (
19,
33). Thus, it would be of interest to further analyze the pathogenicity of
P. acanthamoebae in severely immunocompromised mice. Taking into account that
P. acanthamoebae modestly replicated in infected mammalian cells
in vitro (
7,
24) and that the bacterial load decreased with time (10-fold in 2 days and 1,000-fold in 10 days) in mice infected with
P. acanthamoebae (
9),
P. acanthamoebae may compensate for its low replication capacity by escaping innate immune surveillance. Importantly, there are significant differences in the replication capacities of
Parachlamydia in murine and human macrophages (no versus limited replication capacity) as well as in cytokine production by these cells following exposure to heat-inactivated bacteria (presence and absence of proinflammatory cytokine secretion). Thus, the mouse model may partially reflect what occurs in human infections.
In summary, we report the first insight into the mechanisms of recognition of
P. acanthamoebae by the innate immune system. MyD88 or TLR4 deficiencies do not impair mouse survival after
P. acanthamoebae infection, whereas they did so after
C. pneumoniae infection, indicating that this
Chlamydia-related bacterium has evolved a different strategy to interact with the host. It will be interesting to investigate the innate immune response to other
Chlamydia-related bacteria such as
Waddlia chondrophila, a potential agent of miscarriage and lower respiratory tract infection which was shown to replicate in human macrophages (
17). Finally, further work will be devoted to the identification of the molecular pattern(s) of
P. acanthamoebae that stimulates innate immune cells, as this may help in development of improved serologic diagnostic tests for infection and better understanding of the pathogenesis of
Parachlamydia infection.