We report the first evidence that we are aware of for involvement of multiple Chlamydiaceae species in trachoma. The organisms detected from conjunctival samples included C. trachomatis of numerous ompA genotypes and two species, C. pneumoniae and C. psittaci, of the Chlamydophila genus.
Significantly, non-C. trachomatis organisms accounted for 50% of the ocular infections in this population. All three species were found in all study households, except for one household in which C. trachomatis ompA genotypes and C. psittaci were present. Further, 35% of infected individuals had mixed infections, and these consisted primarily of two Chlamydiaceae species. The high agreement between DNA and RNA results also support evidence for viable organisms and not just residual DNA from a previous infection.
Although our data are specific to Nepal, it is likely that C. pneumoniae
and C. psittaci
are also involved in trachoma in other endemic regions of the world. This hypothesis is supported at least in part by the finding that, in Tanzania, a risk factor for severe trachoma in children is family ownership of cattle [57
]. Cattle are an important mammalian host for C. psittaci,
in which asymptomatic and symptomatic infection and shedding of organisms in intestinal and vaginal secretions is common [47
]. Furthermore, in Nepal, fresh cow and buffalo dung are used along with mud to construct by hand the walls and floors of huts that are used for living, eating, and grain storage. The fresh dung is also formed into patties that are allowed to dry on the walls of the huts and then used as fuel in fires. These practices present the potential for hand-to-eye transmission of C. psittaci
or C. pecorum
mammalian strains, although this particular route has not been studied to date. The strains in our study were similar to the bovine WC strain, suggesting that this or related strain types may be important in trachomatous disease in both Asia and Africa.
Identification of C. pneumoniae
and C. psittaci
in trachoma populations may explain the lack of DNA or culture evidence for C. trachomatis
infection among some children and women with obvious clinical signs of follicular, severe, or persistent disease [57
]. Furthermore, additional Chlamydiaceae species may be present in the conjunctiva but not detected because of low copy number and the inability of PCR to amplify limited amounts of DNA. Interestingly, nonhuman mammals such as the Western barred bandicoot (Perameles bougainville
) and koala (Phascolarctos cinereus
) have also been found to develop ocular conditions similar to trachoma with infections with multiple Chlamydiaceae species, including C. pecorum
and C. pneumoniae
, a new Chlamydiales endosymbiont of Acanthamoeba
species, and other uncultured Chlamydiales strains [64
]. Proof of multiple Chlamydiaceae species involvement in trachoma in other geographic regions will require appropriate studies in Africa, the Middle East, Central and South America, and other regions of Asia where trachoma is endemic.
Although there have been occasional anecdotal reports of laboratory accidents that caused C. pneumoniae
and C. psittaci
eye infections in humans, C. psittaci
is a common ocular pathogen in lower mammalian species [33
]. The observation that vitamin-deficient guinea pigs were susceptible to ocular infection with chlamydiae led to the isolation and identification of a C. psittaci
strain (now termed Chlamydophila caviae
), referred to as the guinea pig inclusion conjunctivitis strain or GPIC [66
]. When GPIC organisms were experimentally inoculated onto the eyes of guinea pigs, a delayed hypersensitivity reaction was evident and a self-limited follicular conjunctivitis resulted [35
]. Similarly, Chlamydophila felis
(formerly C. psittaci
, strain Fe/C-56) has been isolated from the eyes of kittens with feline keratoconjunctivitis (FKC). FKC is a follicular conjunctivitis, which, like trachoma, can result in conjunctival scarring and pannus formation [34
]. C. pneumoniae
has also been isolated from the eyes of Western barred bandicoots [67
]. These findings suggest that non-C. trachomatis
organisms may be capable of producing the ocular inflammation and scarring sequelae of trachoma.
In studies investigating the natural history of trachoma, the greatest prevalence of infection has been demonstrated in children [1
]. This pattern of infection was supported by the results of this study. Individuals aged 10 y or younger were significantly more likely to be infected with C. trachomatis
and with multiple C. trachomatis ompA
genotypes and species than were adolescents or adults. The role that mixed infections may play in disease pathogenesis has not been explored. It is important to note, however, that children in trachoma-endemic areas commonly develop repeat infections [69
], where a history of severe inflammatory disease or persistent infection may be predictors for the subsequent development of scarring and trichiasis [57
]. Indeed, in one study by Munoz et al. [73
], women aged 18 y and older with conjunctival scars had significantly higher rates of trichiasis and C. trachomatis
infection over a 7 y follow-up period compared to women without scars.
The population distribution of C. psittaci and C. pneumoniae infections was more uniform among all age groups and households in this study, perhaps representing a different pattern of transmission and reinfection than for C. trachomatis. This distribution pattern suggests that most individuals are susceptible to infection with any of the three species and that these infections are not just sporadic, but widespread in the population.
While the host immune response in trachoma is poorly understood, data suggest that the C. trachomatis
Hsp60 may provide an adverse antigenic stimulus [74
]. Antibody responses to C. trachomatis
Hsp60 have been shown to be associated with a delayed-type hypersensitivity response in the conjunctiva of monkeys [54
] and guinea pigs [35
], and with trachomatous scarring and trichiasis in humans [49
], as in this study. A significant association has also been demonstrated between antibodies to C. trachomatis
Hsp60 and fallopian tube scarring and infertility in monkeys [75
] and women [76
which suggests a similar mechanism of disease pathogenesis.
We have previously shown a significant association between elevated tear IgG titers to C. trachomatis
Hsp60 and both active and scarring disease among individuals in a trachoma-endemic population in Nepal [49
]. While it is possible that the immunoreactivity to recombinant Chlamydiaceae Hsp60 of the different species represents cross-reactivity due to the high degree of homology between species (C. trachomatis
Hsp60 is 93% homologous to the Hsp60 of C. psittaci
] and 85% to that of C. pneumoniae
]), the associations with scarring and/or trichiasis were significant when assaying tears and sera from individuals infected with one species for responses to the specific recombinant Hsp60 for that species (). This result suggests that different species elicit a species-specific response. Nonetheless, specific- or cross-reactive antibodies are likely damaging to the conjunctiva; we also found a significant association of tear and serum Chlamydiaceae Hsp60 antibodies with inflammation and scarring for individuals with any infection. These findings may be particularly important since individuals in trachoma-endemic communities are constantly being reinfected and, based on our findings, likely being reinfected with different as well as the same Chlamydiaceae species.
The implication of host involvement in pathogenesis does not rule out the possibility that factors specific to the organism may influence an individual's risk of developing scarring and blindness. One of these factors may be the ability of Chlamydiaceae to persist. Up-regulation of Chlamydiaceae Hsp60 has been demonstrated in an in vitro model of persistence in which aberrant, nonculturable forms of C. pneumoniae
and C. trachomatis
in HeLa cells were induced by interferon gamma (IFN-γ) [80
]. We do not know if persistence occurs in vivo, although supportive evidence is mounting (see review by Hogan et al. [81
]). There are some data indicating that C. trachomatis
persists in children for months to years in trachoma-endemic settings [57
]. Additional research will be required to determine what role, if any, the different species of Chlamydiaceae play in this process.
If the involvement of multiple Chlamydiaceae species in trachoma is confirmed in other endemic countries, a new approach to antimicrobial therapy may be needed. There has been great interest in the use of azithromycin to treat trachoma, although recurrence of disease and infection is common following cessation of treatment [10
]. Azithromycin has also been used in the treatment of C. pneumoniae
respiratory infections. However, culture-positive persistent infections and elevated minimum inhibitory concentrations following therapy have been reported [84
]. In one case series, individuals with evidence for persistent C. pneumoniae
pulmonary infections were treated for 10 to 21 d with oral antibiotics, which was not sufficient to resolve infection [85
]. In cases of marginal corneal abscesses with follicular conjunctivitis and punctate keratitis due to urogenital C. trachomatis
strains, 2 wk of oral antibiotics were required for resolution of clinical disease [25
]. There are few data for C. psittaci
. Ocular infections with avian strains of C. psittaci
required long-term topical and systemic treatment of up to 10 wk for eradication [30
]. Other reports have described the need for 3 to 10 wk of treatment for ocular infection with non-C. trachomatis
]. The nasopharynx may also be a reservoir for C. trachomatis
, C. pneumoniae,
and possibly C. psittaci
]. Thus, targeted, systemic treatment with more frequent dosing over a longer time period may be required to prevent persistent infections and host carriage of the organism.
Our study has several limiations. First, our findings are from a single trachoma-endemic region of Nepal, and additional studies will be required to determine whether multiple Chlamydiaceae species are involved in trachoma in other endemic countries of Asia and Africa. It will also be important to determine the number of each species per infection by cloning 16S rRNA and sequencing at least ten randomly selected clones. These data will likely provide further information on the microbial diversity of the conjunctiva that may be involved in active and/or scarring trachoma. In addition, further studies of the host immune response and role of Chlamydiaceae Hsp60 responses in trachomatous disease are warranted. Finally, studies that examine the incidence and prevalence rates of respiratory and urogenital tract infections with the various Chlamydiaceae species would provide a broader context for understanding the role of these organisms in trachoma communities.
In summary, we provide substantial evidence for the involvement of C. psittaci
and C. pneumoniae
species, in addition to C. trachomatis
ocular and urogenital strains,
in trachoma. The distribution of species by household and age suggests that these infections are widespread. Our data may also explain in part why a proportion of individuals with active trachoma are considered uninfected, since only C. trachomatis
-specific diagnostic techniques have been used for detection instead of a broader screen for other Chlamydiaceae organisms. Indeed, 28 (35%) individuals with active trachoma in our study were infected with C. psittaci
and/or C. pneumoniae
and would have been missed by conventional approaches for detecting C. trachomatis
. Furthermore, infection with multiple species may explain the failure of active trachoma cases to resolve their clinical disease following effective C. trachomatis
treatment, and the limited effectiveness of the WHO strategy to control trachoma. The high proportion of RNA-positive samples, the significant association of Chlamydiaceae infections with severe inflammation, and the significant association of tear and serum responses to recombinant Chlamydiaceae Hsp60 with inflammation, scarring, and trichiasis for individuals infected with the different species, suggest the importance of all three species in disease pathogenesis. Consequently, our cumulative data suggest that a reevaluation of treatment regimens and approaches to vaccine development may be required. Currently, over 11 genomes of the order Chlamydiales have been sequenced [90
]. Additional genome sequencing of species and strains responsible for trachoma, together with immunological studies, will inform the design of an efficacious vaccine. Understanding the full impact of Chlamydiaceae species on the epidemiology, immunopathology, and disease outcome of trachoma presents a new challenge for Chlamydiaceae research.