The Role of Endosymbiotic Wolbachia Bacteria in the Pathogenesis of River Blindness

See allHide authors and affiliations

Science  08 Mar 2002:
Vol. 295, Issue 5561, pp. 1892-1895
DOI: 10.1126/science.1068732


Parasitic filarial nematodes infect more than 200 million individuals worldwide, causing debilitating inflammatory diseases such as river blindness and lymphatic filariasis. Using a murine model for river blindness in which soluble extracts of filarial nematodes were injected into the corneal stroma, we demonstrated that the predominant inflammatory response in the cornea was due to species of endosymbiotic Wolbachia bacteria. In addition, the inflammatory response induced by these bacteria was dependent on expression of functional Toll-like receptor 4 (TLR4) on host cells.

Wolbachia bacteria are essential symbionts of the major pathogenic filarial nematode parasites of humans, includingBrugia malayi and Wuchereria bancrofti, which cause lymphatic filariasis, and Onchocerca volvulus, which causes river blindness (1). Wolbachia spp. are abundant in all developmental stages of filarial nematodes, including the hypodermis and reproductive tissue of adult parasites (1). In contrast to their relatives in arthropods,Wolbachia spp. in filarial nematodes appear to have evolved as an essential endosymbiont. Antibiotic therapy in humans and experimental filarial infection has shown that embryogenesis is completely dependent on the presence of Wolbachia(2, 3). Furthermore, parasites recovered from tetracycline-treated animals are stunted, and larval development is attenuated (2). In O. volvulus–infected individuals, adult worms survive for up to 14 years in subcutaneous nodules in the human host and release millions of microfilariae over this time (4). Microfilariae migrate through the skin and can enter the posterior and anterior regions of the eye. While alive, the microfilariae appear to cause little or no inflammation; however, when they die, either by natural attrition or after chemotherapy, the host response to degenerating worms can result in ocular inflammation that causes progressive loss of vision and ultimately leads to blindness (4, 5).

Following the discovery of endosymbiont-derived endotoxin-like activity of B. malayi and O. volvulus(6, 7), we sought to determine the role of Wolbachia in the pathogenesis of ocular onchocerciasis. We used a murine model of corneal inflammation (keratitis) in which parasite extracts were injected directly into the corneal stroma, and corneas were examined by scanning confocal microscopy using confocal microscopy through focusing software (CMTF), (8, 9). This method measures stromal thickness and stromal haze, which are indicators of corneal edema and opacity, respectively, and can measure corneal inflammatory responses that cannot be detected by slit lamp microscopy. To determine the role ofWolbachia in O. volvulus keratitis, we compared soluble extracts of worms recovered from doxycycline-treated individuals with worms from untreated individuals (3). The effect of doxycycline on reducing Wolbachia numbers in the worms was demonstrated by immunohistochemistry with antibodies to bacterial heat shock protein 60 (hsp60) (3). Furthermore, semiquantitative polymerase chain reaction with Wolbachia16S ribosomal DNA and filarial 5S DNA showed that the 16S/5S index in doxycycline–treated worms was lower than that in untreated worms by a factor of 32 (10).

Extracts from doxycycline-treated worms induced significantly lower stromal thickness, stromal haze, and neutrophil infiltration than extracts from untreated worms (Fig. 1A). Although minor compared with untreated worms, extracts from treated worms also induced inflammatory responses that were significantly higher than those from naı̈ve corneas or corneas injected with saline. Similarly, extracts from the rodent filariaAcanthocheilonema viteae, which do not harborWolbachia (1), induced significantly less stromal thickness, stromal haze, and neutrophil infiltration than extracts fromBrugia malayi (Fig. 1B). Together, these data indicate that endosymbiotic Wolbachia bacteria in filarial parasites have a major role in the development of corneal pathology.

Figure 1

(A) O. volvulus worms containing Wolbachia were recovered from infected individuals in Ghana who were either untreated (OvAg) or had been treated with doxycycline (OvAg/doxy) (3). Soluble extracts were prepared from these worms as described (6), and 1 μg was injected into the corneal stroma of C57Bl/6 mice. Stromal thickness and haze were measured by CMTF (8, 9), and the number of neutrophils per 5-μm corneal section was determined after immunostaining (14, 27). For controls, corneas were either untreated (naı̈ve) or were injected with Hanks' balanced salt solution (HBSS). Data points represent individual corneas from a single experiment. Significant differences were found for all parameters between extracts from doxycycline-treated and from untreated worms (P < 0.0001), and between doxycycline-treated and HBSS-treated worms (P < 0.05). (B) B. malayi andA. viteae worms were recovered from gerbils, andsoluble extracts were prepared as de- scribed (7). Parasite extract (1 μg) was injected into C57Bl/6 mice, and stromal thickness, stromal haze, and neutrophil infiltration were determined as described in (A). Data points represent individual corneas from a single, representative experiment.P < 0.0001 for B. malayi versus A. viteae for all parameters, and P < 0.05 for A. viteae versus HBSS for all parameters. All experiments were repeated at least three times with similar results.

Because TLR4 is essential for the host cell response to bacterial lipopolysaccharide (LPS) (11, 12), we determined if Wolbachia mediate corneal pathology by activating TLR4.O. volvulus extracts (containing Wolbachia) were injected into corneas of C3H/HeJ mice, which are hyporesponsive to LPS owing to a single point mutation in the Tlr4 gene (13), and immune responses to these extracts were compared with those from congenic, LPS-responsive C3H/HeN mice. Stromal thickness, stromal haze, and neutrophil recruitment were significantly diminished in C3H/HeJ mice compared with C3H/HeN mice (Fig. 2A), indicating that TLR4 regulates the development of O. volvulus keratitis. Expression of platelet endothelial cell adhesion molecule–1 (PECAM-1), macrophage inflammatory protein–2 (MIP-2), and KC, which are essential for neutrophil recruitment to the cornea (14, 15), was significantly diminished in C3H/HeJ mice compared with C3H/HeN mice (Fig. 2B), indicating that TLR4 regulates O. volvuluskeratitis by modulating expression of PECAM-1, MIP-2, and KC in the cornea.

Figure 2

(A) O. volvulus extract (containing Wolbachia) was injected into corneas of LPS-hyporesponsive C3H/HeJ mice or congenic C3H/HeN mice. Stromal thickness, haze, and neutrophil numbers were determined as described inFig. 1. Data points represent individual corneas from a representative experiment. P < 0.0001 for C3H/HeJ versus C3H/HeN mice for all parameters measured. (B) C3H/HeJ and C3H/HeN mice were injected intrastromally with O. volvulus extracts containing Wolbachia bacteria, and PECAM-1 expression on limbal vessels and MIP-2 and KC production were determined as described (27, 28). Data points represent individual corneas from a representative experiment. P < 0.0001 for C3H/HeN versus C3H/HeJ mice for PECAM-1, MIP-2, and KC.

Previous studies showed an important role for T cells, parasite-specific antibodies, and formation of immune complexes inO. volvulus keratitis (16,17). Here, we sought to determine whether TLR4 also regulates keratitis in the presence of an adaptive immune response. C3H/HeN and C3H/HeJ mice were immunized subcutaneously and injected intrastromally with a soluble extract from untreated O. volvulus worms (containing Wolbachia) (18, 19). Stromal thickness, stromal haze, and neutrophil infiltration were assessed at 24 and 72 hours, and eosinophil infiltration was examined after 72 hours, when there is a pronounced eosinophil infiltration in immunocompetent mice (14, 16). C3H/HeJ mice had significantly less stromal thickness, stromal haze, and neutrophils than C3H/HeN mice (Fig. 3, A and B). However, there was no significant difference in eosinophils between C3H/HeJ and C3H/HeN mice at 72 hours (Fig. 3C), indicating that in contrast to neutrophils, eosinophil recruitment to the corneal stroma is not regulated by TLR4.

Figure 3

C3H/HeJ and C3H/HeN mice were immunized subcutaneously and injected intrastromally with extracts from O. volvulus worms containing Wolbachia bacteria (19). After 24 and 72 hours, stromal thickness and stromal haze were assessed by CMTF (8), and neutrophils and eosinophils per high-power field (×200) were determined by immunohistochemistry (A) 24 hours or (B andC) 72 hours after intrastromal injection. Data points represent individual corneas from a single experiment.P < 0.05 for immunized C3H/HeN versus C3H/HeJ mice for neutrophils, stromal thickness, and stromal haze at both time points. There was no significant difference in eosinophil numbers beween C3H/HeN versus C3H/HeJ mice.

When taken together, these findings demonstrate an essential role for Wolbachia and TLR4 in corneal pathology, even in the presence of an adaptive immune response. These results also indicate that the innate immune system plays a critical role in the inflammatory response associated with the pathogenesis of river blindness. In infected individuals, the innate inflammatory response is therefore likely to be initiated by release of Wolbachiaendotoxin-like molecules from dead and degenerating worms, which then activate TLR4 on resident corneal epithelial cells and keratocytes. Because corneal cells express functional CD14 and TLR4 that mediate endotoxin-induced interleukin-6 (IL-6) and IL-8 production (20), it is likely that TLR4 activation results in production of chemotactic cytokines in the corneal stroma and PECAM-1 expression on limbal vessels, which facilitate recruitment of neutrophils to the corneal stroma, neutrophil activation, and subsequent loss of corneal function.

Although the role of Wolbachia and eosinophils has yet to be determined in chronic infection, and further studies are needed to determine the basis for the Wolbachia and TLR4-independent responses, several observations support an important role for Wolbachia and the innate response in individuals infected with filarial nematodes. First, fewer neutrophils are detected in nodules of doxycycline-treated individuals compared with untreated individuals (21), indicating that Wolbachiamediate neutrophil recruitment in human onchocerciasis. Second, local diethylcarbamazine (DEC) application to the skin of infected individuals causes abscess formation that is associated with an intense neutrophil infiltrate compared with infiltration by eosinophils (22). Third, in human lymphatic filariasis, release ofWolbachia into the blood following anti-filarial drug treatment coincides with the onset of severe systemic inflammation (23).

Wolbachia have emerged as the only target for a chemotherapy that results in long-term sterility of the worms in human onchocerciasis (24, 25), a priority research objective for the World Health Organization for the last 20 years (26). Wolbachia depletion by effective treatment with antibiotics such as doxycycline leads to reduction of microfilariae in the skin and to a drastic improvement of skin lesions in hyperergic onchocerciasis (sowda) (24,25). The results presented here demonstrate that in addition to targeting Wolbachia for sustained anti-filarial effects, clearance of Wolbachia by antibiotic treatment may also reduce and prevent ocular onchocerciasis.

  • * To whom correspondence should be addressed. E-mail: exp2{at}


View Abstract

Navigate This Article