|Home | About | Journals | Submit | Contact Us | Français|
Infection with Giardia is the most commonly diagnosed parasitic cause of diarrhea in the developed world, yet no vaccine exists for human use and a commercially available veterinary vaccine is of limited utility. We have used the adult C57BL/6 mouse model of infection with Giardia duodenalis to better understand immunity to secondary infections with this parasite. Mice were primed by infection with either the GS or WB strains of Giardia and treated with metronidazole on day 7–12 to eliminate the primary infections. Challenge infections on day 21 or day 60 after the primary infections resulted in ~50-fold fewer parasites at day 5 than were found in unprimed mice that only received the challenge infection. Resistance to challenge infections was also observed in B cell deficient µMT mice and when primed mice were challenged with parasites of a different strain. While primed mice developed IgA, mast cell, and T cell responses against the parasite, no specific responses correlated with protection against challenge infections. Together these data suggest that development of an effective vaccine for giardiasis should be feasible since strong immunity can be developed against reinfection in the adult mouse model. Moreover we show that antibody responses are not essential for a protective vaccine and that protection is not parasite strain-specific.
Infection with the protozoan Giardia duodenalis (syn. G. lamblia, G. intestinalis) affects hundreds of millions of people in both the developed and developing worlds (reviewed in refs. 1-3). Infections result in a range of outcomes including severe diarrhea, nutrient malabsorption, cognitive and developmental defects as well as sub-clinical infections. Most infections are self-limiting although chronic and recurrent infections do occur. While morphologically indistinguishable, recent genetic analysis has identified 8 distinct assemblages of G. duodenalis (labeled A-H), 2 of which (A and B) are responsible for the vast majority of human infections.4,5
The immune response to primary Giardia infection has been well studied in humans and animal models (reviewed in refs. 1,2,6). Infections result in substantial production of parasite-specific IgA, activation of a CD4+ T cell response and recruitment of mast cells to the intestinal mucosa. While the role of each of these responses in eliminating primary infections has been examined, not as much is known about the immune mechanisms involved in resistance to secondary infections with Giardia. While resistance to secondary infection has been noted previously,7-10 the immunological mechanisms involved remained unclear. One study used G. lamblia of an undefined assemblage and found that mice produced IgA and exhibited resistance to reinfection.9 The other studies all used the G. muris model, and although one study used xid mice and adoptive transfers to suggest that antibodies did not contribute to resistance to reinfection.10 More recently, Rivero et al.11 used parasites genetically modified so that all variant-specific surface proteins (VSPs) are expressed and the normal process of antigenic variation is subverted to generate protection against challenge infections in the gerbil model. The authors suggested that protection in this model was due to antibodies, but did not perform a direct test of this hypothesis. By using the adult mouse model of infection we are now able to directly address the role of antibodies in the development of resistance to reinfection and to examine the ability of parasites from different genetic assemblages to produce cross-reactive responses.
Like most other protozoan infections, there is no vaccine licensed for use in humans. A veterinary vaccine, Giardiavax, has been used in cattle, cats, and dogs, although no studies have demonstrated the role of particular immune responses in contributing to protection.12 Studies in mice have examined responses to a subunit vaccine against a parasite cyst wall protein and shown that this can lead to reduced cyst shedding.13,14 Finally, data from human outbreaks indicates that infection with Giardia produces resistance to subsequent infection, suggesting the development of immunological memory.15,16 Together, these studies suggest that an effective human vaccine to prevent Giardia infection should be possible. A better understanding of immune mechanisms that contribute to protection against reinfection is necessary to help guide vaccine development.
In this study, we utilize infection followed by drug treatment in mice as a vaccine to demonstrate that a primary infection with Giardia leads to resistance to challenge infections. Furthermore, we show that this resistance does not require antibody and that a primary infection with parasites from one genetic assemblage can lead to protection against parasites of a different assemblage.
In order to determine the feasibility of vaccination to prevent infection with Giardia, we established a model in which primary infections were established in adult mice and then treated with metronidazole (MTZ) in order to generate immune responses. Protective immunity was then assessed using challenge infections 3–8 wk later using the identical or heterologous strains. Figure 1A outlines the infection protocol used for the majority of these studies. We began by testing infections with the assemblage B isolate GS(M)/H7 as this has been the best characterized model of infection in adult mice.17 Wild-type C57BL/6J mice that did not get a primary infection, but were only challenged with trophozoites had very high intestinal parasite burdens at 5 d post-infection. In contrast, mice that received the primary infection followed by MTZ treatment and were then challenged had few or no detectable parasites at this time (Fig. 1B). Similarly, mice that received a primary infection followed by MTZ treatment also had few or no detectable parasites. The presence of even a few parasites in the mice which were not given the challenge infection was surprising given that these animals were treated with MTZ for 5 d. Treatment failure was not observed in 2 other experimental trials in which primary infection protected mice from challenge infection, suggesting that successful metronidazole treatment was not actually essential for analysis of challenge infections.
Infections in human and mice generally induce large amounts of Giardia specific IgA.1 We therefore measured anti-parasite IgA levels in intestinal fluid obtained from animals after euthanasia. Figure 1C indicates that high levels of IgA were found in mice given both primary and challenge infections with Giardia, as well as in mice that only received the primary infection. Little IgA was seen in mice that were only given the challenge infection, probably because they were infected for only 5 d. To determine if these antibodies had a role in resistance to challenge infections we performed similar primary and challenge infections in µMT mice that are genetically deficient in B cells and produce no antibodies.18 As expected, these mice failed to produce parasite-specific IgA following infection (Fig. 1C). Control µMT mice that received only the challenge infection had large numbers of intestinal parasites. Indeed, µMT mice had more parasites at day 5 following primary infection than did wild-type controls, suggesting that antibodies do contribute to control of this infection. However, µMT mice that received a primary infection and MTZ treatment prior to challenge had ~100-fold fewer parasites in their small intestines, a reduction similar to what was observed in wild-type mice. Thus, while antibodies likely can contribute to parasite elimination in primary and challenge infections with Giardia, they are not absolutely required for protection against challenge infections in this model.
Since CD4+ T cell responses have been shown to be important for control of primary infections,3 we then sought to examine cytokine production by these cells. We have recently reported that MLN and spleen T cells produced similar patterns of cytokines following in vitro stimulation with parasite extracts.19 We therefore cultured spleen cells from mice in the presence of parasite extracts for 48 h. Supernatants were tested by ELISA for a panel of cytokines representative of Th1 (IFNγ and TNFα), Th2 (IL-4 and IL-6), Th17 (IL-17), and Treg (IL-10) cells. Figure 2 demonstrates that 5 d after a primary infection, splenocytes from mice that were not pre-infected (Challenge only) produced detectable levels of all the cytokines tested in response to parasite extracts. IFNγ, TNFα, IL-4, and IL-6, production was observed with as little as 1 µg/mL of extract while IL-10 and IL-17 production was observed with 10 or 100 µg/mL of extract. No cytokines were seen in the absence of antigen. Cytokine production was also abrogated by addition of blocking antibody against CD4.19 The amounts of IFNγ, TNFα, IL-6 and IL-10 were all in the ng/mL range, while the amounts of IL-4 and IL-17 were much lower, although still significantly greater than that produced by splenocytes from uninfected mice. Interestingly, splenocytes from mice that were primed and then challenged with live parasites had detectable in vitro production only of TNFα and IL-10, although the amounts of IL-10 produced were quite low (Fig. 3). Similarly, mice that were infected 3 wk prior to euthanasia and treated with MTZ also had reduced or absent production of cytokines in response to parasite extracts.
Giardia infections in mice and gerbils have also been shown to induce strong mast cell responses and blockade of mast cell responses using anti-c-kit antibodies or c-kit mutant mice leads to an absence of parasite elimination.3 We therefore examined mast cell responses following infection. Figure 4 shows that priming of wild-type mice leads to a significant accumulation of mast cells 26 d later. A challenge infection at day 26 led to an increase in the number of mast cells observed, but at 5 d post-infection very few mast cells could be seen. We also examined mast cell responses in the µMT mice and found that infection led to mast cell accumulation even in the absence of antibodies. Interestingly, an increase in goblet cell numbers was also noted in several mice following Giardia challenge (Fig. 4B and C). The reproducibility and importance of this finding, however, requires additional studies. Because mast cell enumeration by microscopy could potentially be skewed by observer bias, we also analyzed mast cell activation by measuring serum levels of mouse mast cell protease-1 (MMCP-1). While wild-type mice pre-infected with Giardia had <1 ng/mL MMCP-1 in the serum, primed and challenged mice had 198 ± 93 ng/mL MMCP-1. This shows that release of mast cell granule contents requires continued presence of the parasite. In addition, pre-infected and challenged µMT mice had 936 ± 183 ng/mL MMCP-1 in serum (P < 0.001) indicating that mast cell recruitment and activation can occur in the absence of antibody in this model.
Numerous genetic variants of Giardia have been isolated from humans and have been characterized as belonging to either of 2 major genetic assemblages, A and B.4 Sequence analysis of WB(C6), the prototype of assemblage A, and GS(M)/H7, the prototypical assemblage B strain, indicates that these strains are 78% identical at the amino acid level.20,21 To determine if protection against challenge infections was possible across the different assemblages we took advantage of our ability to infect mice using WB as well as GS. Figure 5A shows that primary infection with either WB or GS led to significant protection against challenge infections, either with the same strain or the strain from the other assemblage. Figure 5B indicates that primary infection resulted in significant production of IgA reactive with membrane proteins from GS, regardless of the strain used for primary or challenge infection. As expected, mice that were not pre-infected and only received the challenge infection did not produce detectable IgA in the 5-d time frame. This suggests that cross-reactive antibodies are readily generated across this genetic difference.
Analysis of spleen cell cytokine production in response to extracts from the GS strain indicated that cross-reactivity at the level of T cell epitopes was less common. Splenocytes from mice that were infected only once with GS (challenge only), exhibited significant production of IFNγ, IL-4, IL-10, and IL-17 after stimulation with GS extract (Fig. 5C–F). In contrast, mice infected with WB did not produce statistically significant levels of any of these cytokines when stimulated in vitro with GS extract. In other experiments, mice infected with WB and stimulated with extract from WB produce significant levels of cytokines, similar to what is seen with GS.19 Interestingly, mice that were primed with GS and then challenged with GS had significantly lower ex vivo cytokine responses whether the challenge infection was 21 d or 60 d after the initial infection. Finally splenocytes from mice that were pre-infected with GS and then challenged with WB produced significantly more cytokines than splenocytes from uninfected mice. Thus, while antibodies produced after infection with WB appear to cross-react with antigens from GS, cross-reactive T cell responses were not detected.
Infection with the GS strain generated robust mast cell responses. In contrast, infections with the WB strain produced a small, but statistically insignificant increase in mast cell numbers in the small intestines (Fig. 5G). As the mast cell response generally takes more than 1 wk to become apparent, analysis of the strain used for the initial round of infection is most informative. Thus, primary infection with WB followed by challenge infections with either WB or GS produced similar results. Primary infection with GS produced significant mast cell accumulation regardless of whether challenge infections were with WB or GS. Importantly, primary infection with GS followed by challenge with GS on day 60 also resulted in few mast cells in the small intestinal mucosa even though these mice were still protected against the challenge infection. Thus, protection against challenge infection did not correlate with the level of mast cell response observed.
The data presented clearly demonstrate that mice that were primed by infection with G. duodenalis are more resistant to a challenge infection than naïve mice. These data are similar to previous studies analyzing infections in wild-type mice with G. muris.7,8 Furthermore, we showed that following primary infection, wild-type mice produce significant amounts of anti-parasite IgA. However, since µMT mice which do not make any antibodies also have reduced parasite loads following challenge infections, antibodies are not necessary for protection against challenge infection. A previous study has examined secondary infections with G. muris in µMT mice.22 This study indicated a role for antibodies in control of both primary and secondary infections. These results differ from those shown here, possibly because of differences between G. muris and G. duodenalis or due to inoculation with trophozoites in our model compared with inoculation of cysts used in the G. muris studies. Importantly, the results of the present study do not indicate that antibodies cannot contribute to protection against challenge with G. duodenalis, only that in the absence of antibody other mechanisms are sufficient.
Our data also indicate that reduced parasite burdens following challenge infections does not depend on the genotype of parasite used for the primary infection. Thus, mice infected with the assemblage A strain WB were had reduced parasite burdens following challenge infections with the assemblage B strain GS and vice versa. Interestingly, while mice pre-infected with either WB or GS produced IgA antibodies which reacted with a protein preparation from GS, splenocytes from WB infected mice did not produce cytokines after stimulation with GS extracts. Given that these 2 strains share only 78% amino acid identity throughout their genomes,21 it is perhaps not surprising that there is so little cross-reactivity among T cell epitopes. However, this also suggests that antibodies may be mediating protection in this scenario.
In addition to antibody responses, we also show that infection leads to mast cell responses and cytokine production by splenocytes in response to parasite antigens. The role of mast cells in elimination of primary Giardia infections in mice and gerbils has been described previously.3,23 Interestingly, the number of mast cells in the mucosa returns to pre-infection levels by ~60 d after the primary infection, but the mice were significantly protected against challenge infections. This suggests that mast cells are not responsible for mediating protection against challenge infection. In addition, our data indicate that mast cells are recruited and activated in µMT mice, even though these mice make no antibodies. Thus, mast cell recruitment and activation can occur independently of IgE in this model.
CD4+ T cell responses are known to be essential in control of primary Giardia infections.2,24 However, the cytokines made by these cells in response to parasite antigens have not been well-defined. Our data indicate that IFNγ predominates 5 d after infection in C57BL/6 mice. This is consistent with reports of IFNγ production by peripheral blood and lamina propria lymphocytes from non-infected humans in response to Giardia antigen25 and by conA stimulated mesenteric lymph node cells from BALB/c or C57BL/10 mice infected with G. muris.26,27 It is also consistent with the elevated levels of IFNγ seen in sera from patients.28 Interestingly, significant amounts of IL-10 were also seen. Low levels of IL-17 and IL-4 were also produced in response to parasite antigens ex vivo. We have also found high levels of IL-13 produced by splenocytes stimulated with Giardia antigens.19 Thus, our results are consistent with a mixed T cell cytokine response following Giardia infection.
While cytokine responses from splenocytes were detected 5 d following challenge infections in mice that were not primed, they were generally not detected or were significantly diminished in challenged mice that had been primed. This is surprising in that the primed mice exhibit protection against reinfection. The absence of responses following challenge infection in primed mice could be due to homing of the antigen specific cells from the spleen to the intestinal mucosa following challenge infection or migration of memory cells to the bone marrow or other sites.29 It could also be due to clonal exhaustion of Giardia-specific T cells following challenge infection.30 Additional studies will be needed to clarify whether Giardia infection generates long-lived memory T cells, where these cells reside and if they can confer protection against reinfection. The production of IFNγ by T cells is consistent with a potential role for nitric oxide produced by NOS2 in parasite control. Nitric oxide has been shown to inhibit parasites in vitro31 and we have recently shown redundant roles for nitric oxide and α-defensins in Giardia immunity in vivo.32 The role of each of these responses in contributing to reduced parasite burdens following challenge infection remains to be determined, however.
These data represent an important advance in attempts to produce a protective vaccine for giardiasis. The use of mice unable to produce antibodies definitively demonstrates that protection can be achieved purely by stimulating cellular immunity. A second strength of this study is the direct analysis of T cell recall responses and mast cell responses in this system, rather than a focus purely on antibodies as has been done in almost all prior studies of Giardia vaccines. A third strength is the use of human parasites rather than the murine species, G. muris. This has allowed us to directly address the question of immunological reactivity between genetic assemblages. This study does have some limitations, however. While we have shown that antibodies are not required, the data do not exclude a contribution to protective immune responses. Furthermore, our analysis of antibody responses is limited to IgA. While this is the major antibody isotype in the mucosa, IgG and IgM are produced and can also be found in mucosal secretions. Finally, our data do not identify a specific immune parameter that correlates with protective vaccination. Future studies can help address these limitations.
Studies of giardiasis outbreaks in humans suggest that there may be development of immunity against reinfection.15,16 In contrast, there is also significant evidence of rapid reinfection, particularly among children in endemic areas.33 These disparate conclusions may reflect analysis of adults rather than children in the former studies, or may be due to the intensity of transmission in endemic areas. Our results using the adult mouse model of G. duodenalis infection allows us to examine cross-reactive responses among the genetic assemblages of Giardia which infect humans. The development of a useful vaccine for human giardiasis will be accelerated by understanding how memory immune responses against this pathogen are generated and how they contribute to protection from challenge infections. In this study, we use a simple model of challenge infections to demonstrate that resistance to challenge infections can be achieved in the absence of antibody. We also show that resistance to challenge infections can be generated that exhibits cross-reactivity between the major genetic assemblages of Giardia that infect humans.
Female C57BL/6J (wild-type) and B6.129S2Igh-6 < tm1Cgn > (B cell deficient; µMT) mice aged 6–8 wk were obtained from the Jackson Laboratory (Bar Harbor, ME). The GS(M)/H7 and WB(C6) strains of G. duodenalis were propagated in vitro using TYI-S-33 medium with bile and antibiotics as described.34,35
To facilitate infections, drinking water of the mice was supplemented with 1.0 mg/mL vancomycin (Hospira Inc.), 1.0 mg/mL Neomycin sulfate (Durvet), and 1.0 mg/mL ampicillin (Sigma-Aldrich) beginning 2 d before infection and throughout the infection periods.36 Mice were infected via gavage with 106 trophozoites in PBS. Mice were then treated with metronidazole (5.0 mg/kg; Sigma-Aldrich) by i.p. injection once per day for 5 d, beginning 7 d after infection. Mice were euthanized 5 d after challenge infections for analysis. Parasite burdens in the small intestine were determined as described.19 Briefly, 2 cm segments of duodenum were minced in 2 mL of ice cold PBS and parasites were enumerated using a hemocytometer.
Giardia strain GS trophozoites were grown to late-log phase and collected by chilling on ice and centrifugation. Parasites were washed extensively in PBS and aliquoted in cryovials at a concentration of 1–5 × 108/mL. Cells were frozen by placing vials in a −80° freezer for 10’ followed by 10’ in liquid nitrogen. Cells were thawed rapidly by placing in a 37° water bath. After 5 freeze-thaw cycles protein content was determined by Bio-Rad protein assay.
Spleen cells were suspended in PBS after removal of red cells by lysis in RBL buffer (155 mM NH4Cl, 10 mM KHCO3, 0.1mM EDTA). Spleen cells were adjusted to 5 million/mL in complete culture medium (RPMI-1640 with 10% Fetal Calf Serum, 1 mM L-Glutamine, 50 µM 2-mercaptoethanal, 100 units/mL penicillin G, 100 µg/mL streptomycin, and 2.5 µg/mL of amphotericin B. Cells were cultured in 24-well plates at 5 million cells/well in 1 mL medium. Cells were stimulated with Giardia duodenalis extract (doses of 0, 1, 10, or 100 µg/mL) and incubated for 48 h at 37 °C in a 5% CO2 atmosphere. The supernatants were harvested for analysis of cytokine content by ELISA.
Levels of mouse IFN-γ, IL-6, IL-10, TNFα (BD OptEIA), IL-17 (R and D system), and IL-4 (Southern Biotech) in culture supernatants were measured by ELISA using commercial kits. All supernatants were measured in duplicate.
Intestinal fluid was collected by flushing a 10 cm segment of jejunum with 1 mL cold PBS. To prepare a membrane-enriched fraction for ELISA, Giardia trophozoites were chilled and harvested by centrifugation. The cell pellet was resuspended in cold buffer (0.01 M TRIS-HCl pH 8.6, 0.25 M sucrose, 1 mM MgSO4, 0.001% PMSF, and 0.05% Trypsin inhibitor) and homogenized for 5 min using a probe-type sonicator on ice. Lysates were then centrifuged 10 min at 800 g at 4 °C to remove unlysed cells. Supernatants were then centrifuged for 20 min at 10000 g at 4 °C to remove nuclei. A final centrifugation of the supernatant for 1 h at 100000 g at 4 °C resulted in a pellet fraction enriched in plasma membrane antigens. The pellet was resuspended in buffer and protein concentration was measured using a Bio-Rad Protein Assay. For ELISAs, 96 well plates (Nunc Immunolon II) were coated overnight at 4 °C with 100 µl of 2 µg/mL Giardia membrane protein membrane in PBS. After blocking the plate with 1% BSA in PBS, 100 µl intestinal fluid diluted in PBS was added and the plate was incubated for 2 h at room temperature. 100 µl peroxidase-conjugated goat anti-mouse IgA (AbD Serotec) was followed by ABTS substrate (KPL) and IgA levels were determined by using a micro-plate reader at 405 nm.
Two cm segments of small intestine were fixed in 10% formalin, embedded in paraffin, and sectioned. Five µm tissue sections were stained for chloroacetate esterase activity and counterstained with hematoxylin as described.23 Mast cell abundance was evaluated in 4 microscopic fields per mouse using a 25× field of view (250× total magnification). Mast cell numbers were then determined by an observer who was unaware of the identities of the samples. Mouse mast cell protease-1 (MMCP-1) levels were determined in sera using a commercial ELISA (Moredun Scientific) as described.23
Means between groups were compared using student t tests. In analyses where the variance of groups were significantly different or where groups contained individual values beyond the limits of detection (e.g., no detectable parasites), non-parametric Mann–Whitney statistics were calculated. P < 0.05 was considered significant.
No potential conflicts of interest were disclosed.
This work was support by grants from the National Institutes of Health (AI-081033) to S.M.S and from the China Scholarship Council to M.L. Additional support was provided by the office of the Provost, Georgetown University. The Georgetown Lombardi Animal, Histopathology and Tissue Culture Shared Resources are partially supported by NIH/NCI grant P30-CA051008. These agencies had no role in the design or execution of these experiments.
Thanks to Shahram Solaymani-Mohammadi for helpful discussions and to Sara Higa for assistance with manuscript and figure preparation.