Recent Insights Into Innate and Adaptive Immune Responses to Giardia

Abstract

Infection with Giardia produces a wide range of clinical outcomes. Acutely infected patients may have no overt symptoms or suffer from severe cramps, diarrhea, nausea and even urticaria. Recently, post-infectious irritable bowel syndrome and chronic fatigue syndrome have been identified as long-term sequelae of giardiasis. Frequently, recurrent and chronic Giardia infection is considered a major contributor to stunting in children from low and middle income countries. Perhaps the most unusual outcome of infection with Giardia is the apparent reduced risk of developing moderate-to-severe diarrhea due to other enteric infections which has been noted in several recent studies. The goal of understanding immune responses against Giardia is therefore to identify protective mechanisms which could become targets for vaccine development, but also to identify mechanisms whereby infections lead to these other diverse outcomes.

Giardia induces a robust adaptive immune response in both humans and animals. It has been known for many years that there is production of large amounts of parasite-specific IgA following infection and that CD4⁺ T cell responses contribute to this IgA production and control of the infection. In the past decade, there have been advances in our understanding of the non-antibody effector mechanisms used by the host to fight Giardia infections, in particular the importance of the cytokine interleukin (IL)-17 in orchestrating these responses. There have also been major advances in understanding how the innate response to Giardia infection is initiated and how it contributes to the development of adaptive immunity. Finally, there here have been significant increases in our knowledge of how the resident microbial community influences the immune response and how these responses contribute to the development of some of the symptoms of giardiasis. In this article, we will focus on data generated in the last 10 years and how it has advanced our knowledge about this important parasitic disease.

1. Protective Immune Responses against Giardia

1.1. Interleukin 17

The most notable change in our understanding of the immune response to Giardia in the last decade has been the discovery of the role of interleukin 17 (IL-17). IL-17 has been shown to be an important component of mucosal immunity to fungal pathogens such as Candida albicans and to promote neutrophil responses and anti-microbial expression by epithelial cells (Li et al., 2018a). IL-17 production was first reported in giardiasis by Solaymani-Mohammadi and Singer in supernatants of Giardia extract-stimulated splenocytes obtained from mice seven days following G. duodenalis infection (Solaymani-Mohammadi and Singer, 2011). In contrast, a microarray analysis in calves found a down-regulation of IL-17 transcripts in the jejunum three weeks post-infection with G. duodenalis (Dreesen et al., 2012). Peroxisome proliferator-activated receptors alpha and gamma (PPAR-α and PPAR-γ) were shown to be increased in this same infection, which the authors hypothesized could lead to downregulation of NF-κB and activator protein 1 (AP1) and thus, a decrease in cytokine production.

In contrast to the calves infected with G. duodenalis, IL-17 seems to play an important role in the mouse response to G. muris (Dreesen et al., 2014; Dreesen et al., 2012). Infected mice had significantly increased levels of IL-17 mRNA in the small intestine three weeks post infection (Dreesen et al., 2014). Importantly, mice lacking the IL-17A receptor mice exhibited increased cyst shedding in feces compared to wildtype mice (Dreesen et al., 2014). Interestingly, PPARα knockout mice did not seem to have significantly altered cyst counts or IL-17 levels, although the data for IL-17 were not shown (Dreesen et al., 2014). This report clearly showed that IL-17A was important for elimination of G. muris infections.

In a similar study of mice infected with G. muris, mucosal IL-17A transcripts were elevated in the first week of infection, reached a maximum level by week 2, and decreased in week 3 (Dann et al., 2015). Tissue levels of cytokines were also significantly elevated at 2 weeks in this study. IL-17A was also shown to be expressed following G. duodenalis infection in wild-type mice, and IL-17A deficient mice had a defect in clearance of G. duodenalis as well as G. muris. Trophozoite numbers in the small intestine were greater in IL-17A deficient mice than wild-type controls at 2, 3 and 7 weeks post-infection with G. muris. Parasite numbers appeared to be decreasing slowly with time, even in the absence of IL-17A, consistent with a multi-factorial control of Giardia.

Several immune cell types may be contributing to IL-17 production during a Giardia infection. Analysis of infections in bone marrow chimeric mice indicated that hematopoietic cells are the important source of IL-17A in this model (Dann et al., 2015). Flow cytometry demonstrated an upregulation of CD4⁺ T cells producing IL-17 in the lamina propria of G. muris infected mice, which likely represent Th17 cells (Dann et al., 2015). This is consistent with increased detection of the transcription factor RORγt reported by Dreesen et al. (2014). Dann et al. also reported an increase in IL-17 producing intraepithelial lymphocytes in G. muris infected mice (Dann et al., 2015). Interestingly, CD4^−/− and Rag2^−/− mice still exhibited elevated numbers of IL-17 producing cells and IL-17 mRNA within the epithelial layer, indicating that Th17 cells are not the only cells responsible for the production of IL-17, and suggesting an innate lymphocyte population might be involved (Dann et al., 2015). Further work is needed to determine if these are type 3 innate lymphoid cells (ILC3s), γδ17 cells or another cell type. The role of this innate production of IL-17 is also unclear since both CD4^−/− and SCID mice exhibit prolonged infections with Giardia (Dann et al., 2015; Fink and Singer, 2017; Singer and Nash, 2000b).

The importance of IL-17 in human giardiasis has also been examined. Peripheral blood mononuclear cells (PMBCs) isolated from humans were stimulated with G. duodenalis and cytokine expression from CD4⁺CD197⁻CD45RA⁻ cells (markers for effector memory T cells) was analyzed using flow cytometry. These studies revealed an upregulation of IL-17A and a combination of IL-17A and TNF-α in samples from individuals previously exposed to Giardia (Saghaug et al., 2016). Interestingly, stratification of patients into those whose infections resolved in fewer than 8 weeks and those who required longer than 8 weeks to eliminate their infections showed that more rapid parasite elimination was associated with a more pronounced IL-17A and TNF-α expression profile (Saghaug et al., 2016). These data are consistent with IL-17 having a protective role in human giardiasis as well as in mouse models.

As might be expected, given the importance of IL-17A, mice lacking the cognate IL-17A receptor (IL-17RA) also exhibited a defect in parasite elimination (Dann et al., 2015). Interestingly, bone marrow chimera studies demonstrated that IL-17RA expression by bone marrow derived cells, but not from epithelial cells or other radiation-resistant cells, was necessary and sufficient for parasite elimination (Dann et al., 2015). IL-17A is known to recruit neutrophils and induce anti-microbial peptide expression in the mucosa (Li et al., 2018a). IL-17A deficient mice were shown to have a near complete absence of fecal IgA, although serum IgA levels were normal in these mice (Dann et al., 2015). These data suggest that transport of IgA is lacking in IL-17A deficient mice, consistent with a role for IL-17 in enhancing expression of the poly-immunoglobulin receptor (polyIgR) on epithelial cells (Cao et al., 2012; Paerewijck et al., 2017). Given the outcome of infections in bone marrow chimeras, however, this regulation may be indirect as IL-17RA on epithelial cells was not needed for parasite control. Other changes seen in IL-17RA deficient mice infected with G. muris included reduced expression of genes for numerous anti-microbial peptides and the mannose binding lectin (Dann et al., 2015).

1.2. Immunoglobulin A

Immunoglobulin A (IgA) is the primary antibody in the intestine/mucosa, making it a popular candidate for giardiasis research. Studies focusing on the role of IgA in controlling infection have shown different effects for G. duodenalis and G. muris infections. Mice deficient in IgA and infected with G. muris showed significantly delayed clearance of primary and secondary infections. A similar observation is made for mice lacking pIgR (Davids et al., 2006). In contrast, lack of pIgR had no significant effect on G. duodenalis infection. Mice with a deletion in the μ exons of the immunoglobulin heavy chain locus (μMT mice) lack mature B cells. μMT mice primed with G. duodenalis and then re-infected showed much lower parasite burdens 5 days post-infection compared to mice which were not primed against the parasite, despite the absence of B cells (Li et al., 2014). These data do not rule out a potential role for antibodies in protection against re-infection with Giardia, but they do indicate that antibodies are not essential and that other mechanisms are sufficient for protection in this model.

Several studies in the last decade have further documented that infection with Giardia does induce a robust IgA response. IgA levels were found to be increased in the saliva and serum of individuals infected with Giardia (El-Gebaly et al., 2012; Zarebavani et al., 2012). Studies have looked at IgA stimulation by excretory/secretory products (ESP) derived from Giardia and soluble Giardia extract, as well as the response to variant surface specific proteins on the parasite surface (Hjøllo et al., 2018; Jimenez et al., 2014). Mice given weekly oral administration of Giardia ESP exhibited increased anti-Giardia antibodies in the serum, including IgA (Jimenez et al., 2014). IgA in the feces was also elevated, although not until 100 days from the start of the inoculations. Another target of anti-Giardia IgA is the family of variant-specific surface proteins (VSPs) on the surface of trophozoites. VSPs have a variable region, a semi-conserved region, and a region anchoring them to the parasite (Hjøllo et al., 2018). Hjøllo et al. followed IgA levels in patients’ sera against two semi-conserved VSP regions (VSP3 and VSP5) prior to treatment and then at 6 weeks, 6 months, and 12 months post-treatment (Hjøllo et al., 2018). They noted that IgA levels were high prior to treatment but fell dramatically by the follow up 6 weeks later (Hjøllo et al., 2018). The rapid fall in IgA levels against conserved portions of the VSP could contribute to the abundance of reinfections seen in individuals residing in endemic areas.

1.3. Mast cells

Mast cells have an important role in controlling infections with both G. muris and G. duodenalis (Li et al., 2007; Li et al., 2004). Upon infection, mast cells are recruited to the intestine where they degranulate. In addition to releasing histamines and mouse mast cell protease-1 (MMCP-1), previous work has shown that they may interact with cholecystokinin (CCK), which in turn leads to increased intestinal contractility. Treating tissues with compounds inhibiting mast cell degranulation (ketotifen) or depleting them of granule contents (compound 48/80) removes the effects of CCK (Li et al., 2007).

One potential mechanism of mast cell recruitment is via activation of complement via the lectin pathway. Mannose-binding lectin (MBL) has been shown to bind to Giardia duodenalis in vitro (Li et al., 2016). This relationship was disrupted by adding soluble N-acetylglucosamine (GlcNAc), suggesting that MBL binds to GlcNAc on the surface of the parasite (Li et al., 2016). Complement activation is known to contribute to recruitment of immune cells, specifically mast cells, and indeed mast cell recruitment was severely diminished in MBL2 and C3aR knockout mice models (Li et al., 2016). As noted above, increased intestinal expression of MBL2 was shown to depend on IL-17 following Giardia infection (Paerewijck et al., 2017). Interestingly, unlike Tako et al. (2013) who identified significant upregulation of mast cell transcripts in the mucosa 10 days following G. duodenalis infection, Paerewijck et al., (2017) did not identify any mast cell related transcripts in their transcriptomic analysis of G. muris infected mice, 21 days post-infection. This might reflect differences between the parasite species or the different time points analyzed in the two studies.

Mast cells may also contribute to resistance to reinfection with Giardia. Mice that were infected with Giardia, treated with metronidazole, and then reinfected exhibited a rapid recruitment of mast cells to the mucosa along with greatly reduced parasite loads compared with mice challenged with Giardia but not pre-infected (Li et al., 2014). Interestingly, mast cell recruitment and activation were not dependent on IgE as serum levels of the mast cell protease MMCP-1 were similarly elevated in wild-type and μMT mice following infection (Li et al., 2014). Thus, mast cells clearly contribute protection against Giardia, but may also be involved in its pathophysiology, especially defects in intestinal motility such as the severe cramps noted in some patients (Solaymani-Mohammadi and Singer, 2010).

1.4. Nitric Oxide and Anti-Microbial Peptides

In 2000, Eckman showed that nitric oxide (NO) inhibited replication of G. duodenalis and that the parasite production of arginine deiminase (ADI) may reduce the amount of arginine available for intestinal epithelial cells, thus reducing their ability to produce NO (Eckmann et al., 2000). Interestingly, the depletion of arginine may not just block production of NO, but may also have an effect on the proliferation of intestinal epithelial cells (Stadelmann et al., 2012). Growing CaCo2 cells in arginine-free media decreased the number of cells when compared to media supplemented with arginine or with citrulline (Stadelmann et al., 2012). Co-culturing Giardia and CaCo2 cells similarly reduced epithelial cell counts, a phenotype which could be partially restored by supplementing with additional arginine (Stadelmann et al., 2012). WB strain Giardia overexpressing ADI further decreased cell counts in co-culture experiments (Stadelmann et al., 2012). Host macrophages expressing arginase (Arg1) have recently been shown to accumulate in the lamina propria of mice infected with G. duodenalis (Maloney et al., 2015). Thus, at least three enzymes, ADI from the parasite and arginase (Arg1) and inducible nitric oxide synthase (NOS2) from the host, may all compete for available arginine during in vivo infections.

During G. duodenalis infection in mice, Nitric oxide, produced by NOS2, functions in a redundant pathway alongside α-defensins to help eliminate infections (Tako et al., 2013). Tako and colleagues found elevated mRNA expression for the matrix metalloprotease MMP7 in the mucosa of Giardia infected mice. MMP7 is involved in activating α-defensins during their synthesis and release from Paneth cells. Mice lacking MMP7 had slightly increased parasite counts on day 7 and on day 13 compared with wild-type mice (Tako et al., 2013). In contrast, mice with a deletion in the NOS2 gene or wild-type mice treated with N-iminoethyl-L-Lysine (L-NIL), a NOS2-selective inhibitor, had parasite levels indistinguishable from control mice following infection (Tako et al., 2013). Double deficient, MMP7^−/− NOS2^−/− mice, however, had significantly higher parasite counts both on day 5 and on day 13 (Tako et al., 2013). Furthermore, MMP7^−/− mice treated with L-NIL also had significantly increased parasite counts on 10 days post-infection compared to untreated controls (Tako et al., 2013). Together these data suggest that NO and α-defensins may be functioning in redundant pathways to contribute to Giardia control.

The importance of antimicrobial peptides during giardiasis has been the subject of investigation for many years. In 1994, Aley et al. showed that culturing G. duodenalis with defensins and other antimicrobial peptides was destructive to cell morphology and decreased the viability of trophozoites (Aley et al., 1994). Human neutrophil peptide (HNP)-1, rabbit neutrophil peptide NP-2, bovine indolicidin, and the murine Paneth cell α-defensins cryptdins 2 and 3 all significantly decreased trophozoite viability in vitro (Aley et al., 1994). As alluded to above, cryptdins in mice are synthesized as propeptides that require proteolytic cleavage by MMP-7 in order to mature into active molecules and data from MMP7 deficient mice indicate that cryptdins contribute to parasite control in vivo (Tako et al., 2013). As noted in the discussion of the importance of IL-17, mice lacking IL-17A or its receptor have been shown to express less mRNA for several members of the β-defensin family (Dann et al., 2015). However, direct demonstrations that these peptides can kill Giardia in vitro or that they are important in vivo is lacking. Recently, lactoferrin and lactoferricin have also been shown to affect parasite growth in vitro and promote the encystation to futile cysts (Frontera et al., 2018). The process seems to require the endocytosis of lactoferrin and lactoferricin into the parasite through a G. duodenalis low-density lipoprotein receptor (GlLPR) (Frontera et al., 2018). It seems likely that the full range of host effector mechanisms involved in controlling Giardia infections has not been identified.

Similar to its ability to inhibit production of NO, Giardia appears to protect itself from the actions of antimicrobial peptides. The dose-dependent cleavage of Human β-Defensin 1 (β-HD1) and α-Human Defensin 6 (α-HD6) by Giardia-released cysteine proteases has recently been demonstrated in vitro (Liu et al., 2019). Specifically, β-HD1 is cleaved by the parasite proteases CP16160, CP14019, and CP16779. α-HD6 is cleaved by CP14019 and CP16779 but not CP16160 (Liu et al., 2019).

Interestingly, antimicrobial peptides might play a part in the protective role of Giardia during co-infection with Citrobacter rodentium (Manko et al., 2017). C. rodentium is a commonly used murine model for enteropathogenic E. coli (EPEC), and co-infection with Giardia alleviated some of the symptoms seen during C. rodentium infection alone, including weight loss and bacterial translocation. Co-infection also resulted in increased expression of murine β-defensin 3 (MBD-3) and Trefoil factor 3 (TFF3) (Manko et al., 2017). A similar result was also observed in vitro in CaCo2 cells where co-incubation was associated with an increase in human β-defensin 2 (HBD-2) and TFF3 (Manko et al., 2017). The increased expression of these antimicrobial peptides was blocked using inhibitors of cysteine proteases, suggesting that parasite proteases might be involved in this response (Manko et al., 2017). Given the large numbers of antimicrobial peptides present in the intestinal mucosa, obtaining solid in vivo evidence for their role during infections and co-infections will be a significant challenge. However, given reports of reduced severe disease during Giardia coinfections with other enteropathogens in humans, further studies exploring how Giardia infections interact during co-infections are clearly warranted.

1.5. Mucous

Anti-microbial peptides and IgA are thought to be embedded within the mucous layer which covers the epithelium. Recent work has shown that Giardia cysteine proteases have effects on host mucins and that mucous helps protect the host during infection (Amat et al., 2017). Incubation of biopsied human colon tissue with G. duodenalis resulted in a decrease of the mucous-positive area per crypt as indicated by histology and staining of the mucous (Amat et al., 2017). The thickness of the mucous layer was also decreased during mouse infections with the GS strain of G. duodenalis (Amat et al., 2017). Furthermore, incubation of a goblet cell-like cell line (LS174T) with parasites reduced the amount of mucin stored in cells, and inhibition of cysteine protease, protein kinase C and phosphoinositide-3 kinase activities all could block this effect (Amat et al., 2017). At the same time, mRNA for Muc2 was increased in LS174T cells exposed to parasites. Induction of Muc2 and Muc5Ac mRNA levels was observed in the colons of mice infected with GS on day 7, although only Muc5Ac was induced in the small intestine (Amat et al., 2017). The importance of mucous in responding to the infection was seen in Muc2^−/− mice. Mice lacking this mucin exhibited weight loss when infected with GS, while wild-type control mice did not. Mice lacking Muc2 also had thinner colonic mucous layers than wild-type mice and also had significantly higher parasite burdens in the small intestine at day 7 following infection (Amat et al., 2017). Whether immune responses such as IL-13 production (see below) contribute to goblet cell production of mucous during infections has not been studied.

2. Activation of the Immune Response During Giardiasis

2.1. The Complement System

One fundamental challenge for the field of Giardia immunology is identifying the specific initial interactions with innate immune cells and the mechanisms that are used to activate downstream effector cells, such as T cells. Indeed, various reports involving early activation and protective mechanisms have begun to reveal how anti-Giardia immunity is initiated. In the past, the role of the complement system has been somewhat overlooked as both an effector mechanism as well as an initiator of immunity. This chemical death squad plays a role in not only microbial killing, but also contains molecules that serve as recruiters of immune cells. A role for the complement system in protective Giardia immunity was established by Hill et al., when trophozoite killing was observed following exposure of trophozoites to human sera that had no history of prior Giardia infection. Killing was lost when complement components were blocked, suggesting parasite killing was complement-dependent. Furthermore, this killing was dependent on structures found on the parasite’s surface, as trophozoites treated with neuraminidase or trypsin did not activate the complement pathway (Hill et al., 1984). At this point in time, the authors credited the activation of complement to the alternative pathway, however the lectin pathway had not been identified and several more recent reports have described a role for this pathway in Giardia protection.

The lectin pathway is initiated through the binding of mannose binding lectin (MBL), a C-type lectin, to carbohydrates such as mannose and N-acetylglucosamine (GlcNAc), the latter being commonly found on Giardia trophozoites (Ratner et al., 2008; Ward et al., 1988). It was shown in vitro that MBL is able to bind to trophozoites and cause cell lysis through the complement system (Evans-Osses et al., 2010). Moving forward, the Singer group reported that MBL was expressed in the murine small intestine during Giardia infection (Tako et al., 2013) and that MBL-deficient mice exhibited a delay in the clearance of Giardia parasites and recruitment of mast cells to the small intestine (Li et al., 2016). This defect in clearance and recruitment was also dependent on C3a signaling via C3a receptor as mice lacking C3aR exhibited the same delayed-clearance phenotype as infected MBL-deficient mice. Moreover, significantly reduced T-cell cytokine responses against parasite antigens following infection were observed in C3aR deficient mice, although IgA levels within the small intestine were unaffected (Li et al., 2016). These results have been generally confirmed using the G. muris infection model (Paerewijck et al., 2017) suggesting that this pathway may also be relevant to human giardiasis. Despite the evidence provided, it is still unclear if other canonical cellular recognition strategies leading to the activation of the immune system are in play during Giardia infections.

2.2. Toll-Like Receptors, Macrophages, and Dendritic Cells

The lamina propria of the small intestine is considered one of the largest sites of immune cell localization in the human body; this layer contains not only myeloid cells that are primarily responsible for initiating downstream effector immune mechanisms, but also T cells that mediate protection (Singer and Nash, 2000b) and pathology (Keselman et al., 2016; Scott et al., 2004) during Giardia infection. Several mechanisms are known to allow macrophages and dendritic cells (DCs) in the lamina propria to interact with antigens in the intestinal lumen (e.g. M cell transport or dendritic cell “periscopes”). Giardia is well-known to induce epithelial barrier breakdown and to promote microbial translocation out of the lumen (Chen et al., 2013; Scott et al., 2002; Zhou et al., 2007). Thus, macrophages and DCs could act as sentinels during giardiasis and interact with conserved microbial-associated molecular patterns (e.g., lipopolysaccharide [LPS] from gram⁻ bacteria) as well as with antigens from Giardia.

The role of DCs in the immune system is thought mainly to be involved in the activation of the adaptive immune response through presentation of antigen:MHC complexes to T cells, expression of co-stimulatory molecules such as CD80 and CD86, and the provision of cytokines that help direct the differentiation of naïve T cells into mature effector cells. Several studies have reported that the DC cytokine response is able to be modulated through a combined treatment of Giardia protein extract and lipopolysaccharide (LPS) which leads to the inhibition of pro-inflammatory cytokine production (Kamda and Singer, 2009; Obendorf et al., 2013; Summan et al., 2018). Murine bone marrow derived DCs (BMDCs) exposed to Giardia alone led to poor cytokine induction and only slightly increasing expression of the costimulatory molecules, CD80 and CD86; yet, the combination of Toll-like Receptor (TLR) agonists and parasite extract led to the reduction of IL-6, IL-12, and TNFα while enhancing IL-10 production when compared to LPS controls (Kamda and Singer, 2009). This report further showed that blocking IL-10R and phosphoinositide-3 kinase (PI-3K) restores production of IL-12, which suggests that Giardia may activate PI-3K in DCs and that Giardia may restrict development of Th1 cells via IL-12 inhibition (Kamda and Singer, 2009). This reduction of pro-inflammatory cytokines was also recapitulated in TLR4-stimulated cell cultures using human monocyte-derived DCs (mo-DCs) and the report also suggested a reduction in IL-12/23p40, IL-12p70, and IL-23 (Obendorf et al., 2013; Summan et al., 2018). However, when these cells were stimulated with parasite antigen plus a TLR2 agonist, PAM3CSK4, an increase in IL-12/23p40 and IL-23 production occurred (Obendorf et al., 2013). Human mo-DC survival was also reduced when exposed to Giardia extract or live parasites, and the number of apoptotic DCs increased when compared to controls. Moreover, this effect was Giardia specific as mo-DC cultures solely exposed to helminth products did not lead to a reduction in viability, yet co-incubation with both Giardia and helminth extract did (Summan et al., 2018).

Members of the TLR family are often the main receptors used for detecting microbial patterns and activating innate immune responses. Currently, only two reports have suggested that TLRs may play a role in the recognition of Giardia; the HSP70 immunoglobulin binding protein, BiP, from Giardia was found to be a TLR4 agonist using both TLR4 reporter HEK 293 cells and primary cells from TLR4-deficient BALB/c mice (Lee et al., 2014). Immature murine (BMDCs) treated with anti-TLR4 blocking antibody had reduced TNF-α cytokine production in response to recombinant BIP (rBIP) when compared to controls. Furthermore, when TLR4-expressing HEK 293 cells were treated with rBIP, a NF-κB-dependent response was induced, similar to what was seen with for LPS-stimulated controls. Using BMDCs, it was found that rBIP was able to induce greater production of canonical DC maturation and expression of surface molecules, MHCII, CD80, and CD86, to levels that resembled LPS-induced stimulation. The production of cytokines in rBIP-activated BMDCs was also measured using ELISAs; IL-12, TNF-α, IL-6, and IL-4 concentrations were all increased in response to rBIP. However, the production of both IL-6 and IL-4 was decreased at the highest concentration of rBIP. BMDCs cultured from TLR4-deficent mice were unable to reproduce this cytokine response and most cytokines had drastically reduced concentrations when compared to wild-type mice (Lee et al., 2014).

A role for TLR2 has also recently been described in recognition of Giardia and, interestingly, it suggests that TLR2 may have a detrimental role for protective immunity (Li et al., 2017). Using C57BL/6J peritoneal macrophages, quantitative reverse transcription polymerase chain reaction (qRT-PCR) revealed that TLR2 gene expression was greatly increased in these cells in response to Giardia WB trophozoites. Immunohistochemical staining of duodenum tissues from mice infected with WB cysts also indicated increased levels of TLR2. The production of IL-6, IL-12p40, and TNF-α in response to live trophozoites was increased in peritoneal macrophages that were treated with TLR2 blocking antibody or if the macrophages were obtained from mice deficient for TLR2 and compared to wild-type controls. Moreover, the concentration of IFN-γ was also reduced in TLR2-deficient cells. Western blot analysis of macrophages exposed to trophozoites revealed the phosphorylation of both p38 and ERK (MAP kinases) along with AKT; phosphorylation of p38, ERK, and AKT as well as the production of cytokines that were reduced in TLR2-deficient macrophages. These data suggest that TLR-2 signaling in response to Giardia activates AKT, p38 and ERK and results in enhanced expression of pro-inflammatory cytokines. However, inhibition of AKT via chemical or siRNA methods instead increased the production of IL-6, IL-12p40, and TNF-α, suggesting that the signaling mechanisms involved are not straightforward. In murine infections using WB cysts, parasite burden and typical Giardia–associated pathologies were found to be drastically decreased in both TLR2-knockout and AKT-inhibited mice when compared to wild-type or vehicle-treated controls; however, this infection model relied on the use of antibiotics for stable infection which may play a role in the presentation of pathologies and immune cell response (Keselman et al., 2016). Additionally, it is unclear if littermate control mice were used in the infection model comparisons as different intestinal microbiota may provide protection during Giardia infections (Singer and Nash, 2000a). Taken together these data suggest that Giardia mediates the reduction of cytokines via TLR2 by activating AKT, p38, and ERK (Li et al., 2017).

Downstream of the TLRs are cytosolic adaptor proteins that enable the transmission of TLR signals that eventually reach the nucleus; MyD88, along with TRIF, are the main link between the TLRs and downstream signaling proteins to initiate an immune response. However, at this point, it is still not fully understood if MyD88 or TRIF truly plays a meaningful biological role in protective Giardia immunity. BMDCs derived from MyD88-deficient mice stimulated with rBIP were unable to produce cytokines when compared to wild-type controls, suggesting a role for a TLR-MyD88 interaction leading to cytokine production (Lee et al., 2014). G. muris infections in MyD88^−/−,IL-10^−/− double knockout mice cleared parasites equally as well as IL-10^−/− controls (and similar to wild-type controls) suggesting the TLR pathway is dispensable for protection (Dann et al., 2018). Lastly, TLRs are not solely found on antigen presenting cells and are also found on intestinal epithelial cells (Abreu, 2010); thus, it is unclear if receptors found on these cells have a role in protection or coordinate any immune responses during infection.

The interaction of DCs with Giardia has been mainly studied in vitro and only a few reports have analyzed their function during in vivo infections. IL-6 deficient mice have a defect in parasite control and DCs have been shown to be a significant source of IL-6 during infection (Bienz et al., 2003; Kamda et al., 2012; Zhou et al., 2003). Interleukin-6 helps to drive the differentiation of IL-17 producing (Th17) T cells in vitro (Weaver et al., 2006), and as such, mice that lack IL-6 ought to have reduced IL-17 expression. However, this has not been directly demonstrated during Giardia infection.

Studies of macrophage responses during giardiasis are also sparse. The Singer lab recently reported an increase in a population of intestinal macrophages, following infection, that expresses both nitric oxide synthase 2 (NOS2) and arginase 1 (ARG1) (Maloney et al., 2015). During G. muris infection in IL-10-deficient mice, an expansion of CD11b⁺, CD11c⁻ macrophages was observed in the colon, yet not in the small intestine. This expansion led to colonic inflammation in IL-10-deficient mice, although no such inflammation was seen in wild-type mice. This suggests a regulatory role for these cells in limiting inflammatory signals (Dann et al., 2018). Peritoneal macrophages also were reported to release extracellular traps in response to Giardia trophozoites to ensnare parasites. However these macrophages were collected, and consequently activated, using thioglycollate which ultimately may not describe the true biological macrophage function (Li et al., 2018b).

Overall, it is clear that the cytokine production of DCs and macrophages is able to be directed by Giardia towards a less inflammatory state within the gut; yet this shift may require other accessory signals that are commonly found in the lumen of the intestine – highlighting the importance of maintaining the integrity of the intestinal epithelial barrier. While cytokines are able to dictate the differentiation of activated T-cells and other immune responses, it is still unclear Giardia infection stimulates the immune system to produce a robust immune response while simultaneously limiting inflammation. Understanding the mechanisms whereby Giardia limits inflammation could be valuable for developing therapeutics for chronic intestinal inflammation and for understanding how environmental enteropathy leads to poor vaccine efficacy.

3. Giardia and the Intestinal Microbiome

Since the turn of the millennium, there has been a rapid increase in the number of reports regarding the interactions between the intestinal microbiome and Giardia (Fink and Singer, 2017). However, the foundation for these reports was first reported in 1994, when the Nash group identified that Assemblage A strains, such as WB, could not colonize adult mice while Assemblage B strains like GS could (Byrd et al., 1994). Singer and Nash followed up on this observation and next reported that infection outcomes in the adult mouse model of giardiasis were dependent on the vendor source of the animals. Mice purchased from Taconic Farms were resistant to Giardia infections while mice of the same immunodeficient strain from Jackson Laboratories were susceptible. Treatment with antibiotics made mice from either vendor susceptible to infection, while cohousing animals made them all resistant, suggesting that the composition of microbes within the gut determined susceptibility to this parasitic infection (Singer and Nash, 2000a). The same year, a group from Brazil reported that duodenal microbes were required to stimulate Giardia pathogenicity during infection. Conventional mice, germ-free mice, and germ-free mice reconstituted with microbiota derived from symptomatic giardiasis patients, were all infected with Giardia trophozoites. Infected conventional mice showed the greatest signs of intestinal pathology among the three groups, with the reconstituted germ-free mice showing moderate pathology. However, the most intriguing observation from this report was that infected germ-free mice did not develop intestinal pathology as compared to the other groups, suggesting a need for intestinal microbes to stimulate Giardia-induced pathology (Torres et al., 2000).

The interactions between intestinal microbiota and Giardia are not unexpected, as both reside in the same environmental niche along the lumen of the epithelial barrier during Giardia colonization (Adam, 2001). However, it is quite interesting that this parasite is able to affect the composition of the intestinal microbiota and the virulence of other microbes, while also being susceptible to various commensal bacterial secretions.

The Singer and Dawson groups first described the influence of Giardia on the architecture and abundance of the intestinal microbiota during in vivo murine infections (Barash et al., 2017). Significant shifts in microbial ecology were observed following G. duodenalis strain GS infection in mice pre-treated with antibiotics, as well as in non-antibiotic treated infected mice. Decreases in overall microbial diversity and the abundance of the obligate anaerobic Firmicutes occurred, while an increase in the abundance of aerobic taxa such as Rhodocyclaceae, Moraxellaceae, Flavobacteriales, and Comamonadaceae was reported (Barash et al., 2017). These enriched taxa are all considered to be metabolically flexible, and prosper with increased oxygen tension, lipid availability, and competition for arginine.

During the same year, Bartelt et al. (2017) also reported an alteration of the intestinal microbiota in mice fed a protein-deficient diet, prior to and continuously throughout, an infection with commercially obtained Giardia H3 cysts (Assemblage B). Quantitative PCR analysis of the 16S rRNA revealed that infected mice had greater duodenal bacterial abundance and an increase in the Firmicutes:Bacteroidetes ratio, with Giardia infections leading to increases in the population of Firmicutes while reducing the number of Bacteroidetes. Moreover, stool qPCR also revealed a large increase in the relative abundance of Enterobacteriaceae 14 days post infection, which was considered a better predictor of growth impairment than Giardia burden in stool or duodenum. This shift in the duodenal microbial composition of mice fed a protein-deficient diet also coincided with the persistence of the parasite within the gut rather than clearance (Bartelt et al., 2017). Interestingly, antibiotic treatment of mice with ampicillin, vancomycin, and neomycin prevented Giardia-induced growth impairment which suggests that Giardia interaction with intestinal microbes is a critical factor for the determination of growth outcome. Growth impairment was also exacerbated in mice that were co-infected with a pathogen commonly isolated from malnourished children, enteroaggregative Escherichia coli (EAEC) (Bartelt et al., 2017).

In a human study of the intestinal microbiome of children from Medellin, Columbia the diversity and composition of the microbiome was compared between uninfected children and groups infected with either Cryptosporidium alone, Giardia alone, or co-infected with both helminths and protozoans (Toro-Londono et al., 2019). Using 16S rDNA sequencing, Giardia-only infected children (mixed-gender, 5 years or younger) had an enrichment of Prevotella along with a drastic reduction in Bacteroides when compared to uninfected controls, and this alteration was also seen in children co-infected with both helminths and protozoans. The most abundant groups observed in Giardia-only infected children were Ruminococcus of the phylum Firmicutes followed by Prevotella. The authors suggest that in this case, Giardia is able to shift the intestinal microbiome from a Bacteroides-dominant composition (type 1 enterotype), which was associated with the uninfected control children group, to a Prevotella-dominant composition (type 2 enterotype), which was seen in both Giardia-only and co-infected children (Toro-Londono et al., 2019). Although a Giardia-mediated shift in the composition of the young human intestinal microbiome is clearly evident, it must be noted that different regions of the world may not express the same type 1 enterotype as seen in uninfected controls, as diet is clearly a significant factor determining the composition of the microbiome. Thus, more human clinical studies examining microbiome composition and changes in giardiasis will be needed in the future to elucidate the role and effect of Giardia in mediating shifts of the human intestinal microbiome. Nevertheless, these data suggest that Giardia colonization leads to a compositional and metabolic shift among the intestinal microbial community, which could contribute to both microbial-mediated protection and variation seen in the clinical manifestations of giardiasis.

The Buret group has also described several studies that further suggest the ability of Giardia infections to induce damage in infected host may result from changes in the resident microbiota. When Giardia was exposed to laboratory strains of Escherichia coli or human intestinal microbiota each bacterial group were converted to a toxic state that was deadly to Caenorhabditis elegans (Gerbaba et al., 2015). Moreover, the Buret lab has also demonstrated that Giardia not only induced microbial dysbiosis in human mucosal biofilms, but also increased the virulence of commensal microbes towards human epithelial cells in vitro. An increase in the levels of inflammatory markers, TLR4 and CXCL-8, were also reported in germ-free mice that were reconstituted with human microbes that had been exposed to Giardia products, yet was not induced during exposure to the living parasite. It was found that Giardia cysteine proteases [described in greater detail in the next section] were required for the induction of intestinal permeability that also mediated the disruption of microbial biofilms (Beatty et al., 2017).

While Giardia clearly has effects on the commensal microbiota, it is also clear that the microbiota impacts the ability of the parasite to colonize the host. Several microbial probiotic therapies involving various strains of Lactobacilli have been examined to shield the host from the detrimental effects mediated by Giardia during infection. The proliferative ability of the WB strain (assemblage A) of Giardia was inhibited when trophozoites were exposed to supernatants of Lactobacillus johnsonii culture medium (Perez et al., 2001), and this effect was confirmed in vivo using a gerbil model that evaded WB colonization when treated with L. johnsonii (Humen et al., 2005). The cause for this toxicity to Giardia trophozoites was attributed to the generation of deconjugated bile salts produced by L. johnsonii (Travers et al., 2016). Three Bile-Salt-Hydrolase-like (BSH) genes were recently identified in this probiotic microbe and were cloned into E. coli in order to assess the ability to be used as an anti-giardial agent (Allain et al., 2018). Two recombinant BSH proteins, rBSH47 and rBSH56, were purified from transgenic E. coli cultures and incubated with Giardia WB trophozoites. A dose-dependent inhibition of Giardia growth was reported when compared to non-BSH controls with concentrations of 1 μg/mL or greater resulting in 100% parasite killing. Scanning electron microscopy revealed significant structural damage to the parasite surface with several protrusions and perforations seen along the parasite plasma membrane. Using a murine infection model, the protective effects of rBSH47 were tested in vivo with Giardia-infected OF1 suckling mice. After infection with 1×10⁵ trophozoites, groups of mice were administered different (but increasing) dosages of rBSH47 for 5 consecutive days post-infection. Parasite burden analysis revealed an increase in parasite numbers in PBS-treated controls (1×10⁶ parasites/intestine); rBSH47-treated groups contained lesser numbers of parasites with the highest dosage group having the fewest parasites in the intestine (Allain et al., 2018).

In murine models used by the Shukla and Goyal group, Lactobacillus casei was able to reduce parasite-mediated mucosal damage and lessen both trophozoite burden and cyst shedding (Shukla et al., 2008). Giardia-mediated immunopathologies were also reduced with Lactobacillus rhamnosus treatments as infected mice treated with L. rhamnosus had increased antioxidants and brush-border disaccharidases levels, while oxidant concentrations were decreased in the small intestine (Goyal et al., 2013). Using BALB/c mice, prior treatment with L. rhamnosus led to a significant increase in secretory IgA antibody concentration and in the percentage of CD4⁺ T cells in the small intestine following Giardia infection. Moreover, they also observed a decrease in the percentage of CD8⁺ T cells (Goyal and Shukla, 2013), which are required to mediate the hallmark pathology seen during infection (Keselman et al., 2016). This effect was recapitulated in C57BL6/J mice where probiotic administration of Enterococcus faecium lessened parasite burdens and induced a stronger IgA antibody immune response (Benyacoub et al., 2005). Recently, a reduction in Giardia-mediated pathology, decreased duration of parasite burden, and an increase in anti-Giardia IgA and nitric oxide were seen in infected mice that were treated with killed probiotics or probiotic proteins (Shukla et al., 2019). Whether manipulation of the microbiome presents a viable strategy for reducing the impact of giardiasis clinically remains to be determined.

4. Giardia Proteases and Host Chemokines

The interface between the intestinal lumen and the epithelial barrier is the site of attachment for Giardia trophozoites (Adam, 2001); thus, it is a critical area for initiating both parasite colonization and host immunity. At this point, it can be agreed upon that Giardia induces a breakage in this barrier which allows luminal contents to exit into the intestinal lamina propria to activate immune cells (Figure 1), however the actual mechanisms leading to barrier breakdown are still uncertain. Early studies examining the ability of Giardia to colonize the intestinal lumen revealed that trophozoites use a variety of excretory-secretory products (ESPs) to aid in the colonization of the intestine (Nash et al., 1983; Nash and Keister, 1985; Samra et al., 1988). The Svärd lab was able to identify the major ESPs from both WB and GS strains of Giardia using proteomic methods, and revealed that both serine (alanyl dipeptidyl peptidases) and cysteine (cathepsins B and L) proteases are the among the most abundant proteins secreted by either strain of this parasite (Ma’ayeh et al., 2017). Similarly, the Tyler lab reported that tenascins were abundant in the ESP from Giardia (Dubourg et al., 2018). Giardia-secreted cysteine proteases have been extensively studied throughout the history of Giardia research (de Carvalho et al., 2008; Jiménez et al., 2000; Rodriguez-Fuentes et al., 2006), yet most importantly, it seems that this protease class are critical mediators in Giardia pathogenicity.

Figure 1.

Model of the interactions among Giardia, the microbiota and the host that can result in activation of protective, IL-17 dependent immune responses and also pathogenic CD8⁺ T cell dependent responses. Reprinted with permission from Fink, M.Y., Singer, S.M., 2017. The intersection of immune responses, microbiota, and pathogenesis in giardiasis. Trends Parasitol. 33, 901-913.

In vitro studies suggest that cysteine proteases mediate the degradation or rearrangement of tight junction proteins such as ZO-1, occludin, and claudin-1 following co-culture with Giardia (Bhargava et al., 2015; Chin et al., 2002; Liu et al., 2018; Ma’ayeh et al., 2017; Ortega-Pierres et al., 2018; Scott et al., 2002) and infections in humans have been shown to produce increases in the lactulose:mannitol ratio, a common marker of epithelial permeability and degradation of claudin-1 (Troeger et al., 2007). Furthermore, in animal models, increased rates of microbial translocation and decreased transepithelial electrical resistance have also been observed, which provides further evidence that cysteine proteases induce breakages in the intestinal epithelial barrier (Chen et al., 2013; Ortega-Pierres et al., 2018; Scott et al., 2002; Zhou et al., 2007).

Interestingly, the Ortega-Pierres group has characterized two proteases in cell culture with rat epithelial cell (IEC6) monolayers and Madin-Darby canine kidney (MDCK) cells (Cabrera-Licona et al., 2017; Ortega-Pierres et al., 2018). A variable surface protein, VSP9B10A, was identified during a screen of the WB Giardia secretome and bioinformatic analysis identified a cysteine protease-like domain within this VSP. When IEC6 and MDCK cell cultures were exposed to Giardia trophozoites that constitutively expressed VSP9B10A, greater disruption of the monolayer and atypical morphology of cells were reported when compared to parasites transfected with empty vector or to non-transfected Giardia. Moreover, this effect was lost when polyclonal antibodies specific for VSP9B10A were added to the transfected Giardia co-cultures. However, trophozoite supernatants could not mediate this proteolytic effect, suggesting that VSP9B10A is generated during the initial stages of contact with epithelial cells (Cabrera-Licona et al., 2017). More recently, a secreted protease that structurally resembles a papain-like cysteine protease was characterized and named giardipain-1 (Ortega-Pierres et al., 2018). IEC6 and MDCK monolayers were both severely damaged and had altered morphology when exposed to purified giardipain-1 and giardipain-1 expressing trophozoites; these changes were reversed when cell cultures were incubated with trophozoites that expressed antisense RNA complementary to giardipain-1 or in the presence of the protease inhibitor, E-64. Additionally, induction of Caspase-3 mediated apoptosis and degradation of occludin-1 and claudin-1 were reported in IEC6 and MDCK cell cultures, respectively, when exposed to purified giardipain-1. Together, these results describe possible mechanisms for the pathology seen in human biopsies (Troeger et al., 2007) and identify two candidate proteins that should be further studied to prevent Giardia-induced intestinal breakdown.

Following disruption of the epithelial barrier, chemokines are released by intestinal epithelial cells to recruit immune cells to the barrier and to initiate protective immune responses. Microarray analysis revealed that exposure of human colon carcinoma, CaCo2 cells, to Giardia parasites resulted in increased expression of numerous chemokines, including CXCL1–3, CCL2, and CCL20 (Ma’ayeh et al., 2018; Roxstrom-Lindquist et al., 2005). However, the immunomodulatory effects of Giardia ESPs were seen in the degradation of these chemokine signals as the Buret group reported that cathepsin B protease degrades CXCL8 (IL-8) in both cell culture and ex vivo models (Cotton et al., 2014a). Chemotaxis of neutrophils was reduced when supernatants were used from CaCo2 cells incubated with Giardia, but this downregulation was alleviated upon addition of a cysteine protease inhibitor (Cotton et al., 2014b). Since IL-8 is a chemoattractant for neutrophils, these data suggest that degradation of chemokines by Giardia can limit neutrophil recruitment. Interestingly, the effects of the parasite on IL-8 and neutrophil chemotaxis were often dose- and Giardia strain-dependent (Cotton et al., 2014b). Furthermore, the Svärd group also reported that CXCL1–3, CCL2, and CCL20 protein can also be degraded when exposed to Giardia-secreted cysteine proteases (Liu et al., 2018; Ma’ayeh et al., 2018). Given the large number of proteases present within the ESP of Giardia, additional studies of their roles in modulation of the epithelial barrier and induction of immune responses are definitely required.

5. Vaccines

A major goal of understanding immune responses in Giardia is to develop a vaccine. GiardiaVAX is approved for veterinary use, but does not have great efficacy, despite being able to induce good IgA responses against the vaccine. Several Giardia antigens have been examined for their potential to establish an immune response, especially as this pertains to developing a vaccine. Potential immunogens include variant-specific surface proteins (VSPs), the immunoglobulin binding protein (BIP), excretory-secretory products (ESPs), the annexin homolog α1-giardin, and cell wall protein 2 (CWP-2)(Feng et al., 2016; Jimenez et al., 2014; Lee et al., 2014; Lopez-Romero et al., 2017; Quintero et al., 2013; Quintero et al., 2017; Rivero et al., 2010; Serradell et al., 2018; Serradell et al., 2016).

Variant-specific proteins (VSPs) are the dominant protein expressed on the surface of Giardia parasites. Normally, each parasite will only express one type of VSP at a time. Trophozoites periodically undergo antigenic variation and express one of the other ~150 VSP genes in the genome to evade immune responses (Prucca et al., 2011). Monoclonal antibodies (mAb) against specific VSPs often exhibit cytotoxicity against parasites expressing that VSP, prompting Rivero et al. to suggest that generating an antibody response against a diverse set of VSPs may allow for protection (Rivero et al., 2010). The Lujan group developed an siRNA-based approach to generate strain WB trophozoites that simultaneously expressed multiple VSPs (Prucca et al., 2008). They also used a combination of limiting dilution and mAb screening to isolate several trophozoite populations that expressed unique VSP molecules on the trophozoite surface. Gerbils infected with Giardia that simultaneously expressed multiple VSPs had fewer excreted cysts upon a secondary infection (Rivero et al., 2010). Gerbils were also resistant to secondary infection when the same single-VSP expressing Giardia clone was used in both primary and secondary infections; however resistance was lost when a different single-VSP clone was used for the secondary infection. Taken together, these infection studies suggest that a primary infection with Giardia expressing multiple VSPs may be protective against secondary infections with Giardia expressing individuals VSPs (Rivero et al., 2010). Furthermore, culturing parasites with serum containing antibodies from gerbils infected with the multi-VSP expressing parasite revealed that the serum antibodies can cause agglutination of a number of different VSP-expressing clones (Rivero et al., 2010). In contrast, gerbils that were initially infected with parasites expressing a single VSP were then only protected against re-infection with parasites expressing the same VSP (Rivero et al., 2010). These data suggest that mounting a response to a collection of VSPs may be protective against future exposure to parasites with diverse VSP profiles (Rivero et al., 2010).

In simulating a potential vaccine, Rivero et al. administered to gerbils VSPs purified from trophozoites simultaneously expressing multiple VSPs and then challenged them with live parasites two months later (Rivero et al., 2010). Animals receiving the immunization had a lower fecal cyst count (Rivero et al., 2010). Later studies indicated that immunization with purified VSPs in gerbils did not induce the same increase in size and number of Peyer’s patches as an actual infection did; immunization also maintained the weight of the mesenteric lymph nodes which normally increased by day 4 post-infection before going down (Serradell et al., 2018). Serum IgG and fecal IgA also increased post-immunization and were high in infection with different VSP clones (Serradell et al., 2018).

In addition to gerbils, 6–8 week-old dogs and 5–8 week-old cats were similarly protected from a secondary infection regardless of the specific VSP being expressed if they were initially infected with Giardia expressing the full VSP repertoire or received an immunization with VSPs purified from multi-VSP expressing parasites (Serradell et al., 2016). Following this study, cats and dogs from a community in Cordoba, Argentina where cats, dogs, and children are almost universally infected were vaccinated and followed over 2 years (Serradell et al., 2016). Giardia was not observed in any vaccinated animals over this time period (Serradell et al., 2016). Together, these data suggest that the VSPs are a promising target for vaccine development in giardiasis.

Quintero et al. characterized a 71 kDa potential immunogen (5G8) which was seen intracellularly and, to a lesser extent, on the cell surface (Quintero et al., 2013). Trophozoites expressing 5G8 were agglutinated after treatment with antibodies for 5G8 (Quintero et al., 2013). Further studies suggested that 5G8 was a member of the VSP family, but not the VSP-H7 nor the immunoglobulin binding protein (BIP) which are both similar in size (Lopez-Romero et al., 2017; Quintero et al., 2017).

The G. duodenalis immunoglobulin binding protein (GlBIP) has been shown to have strong immunogenic properties (Lee et al., 2014). As noted earlier, GlBiP activates DCs through TLR4. GlBIP itself is composed of two domains; the Hsp70 domain seems to be the immunoreactive one (Lee et al., 2014). The isolated Hsp70 domain was able to induce production of IL-12 and TNF-α, while the other domain was not successful at doing so (Lee et al., 2014). Similarly, dendritic cells stimulated with the Hsp70 domain were then able to promote IL-2 and IFN-γ production from CD4⁺ T cells, which was not observed with the other domain (Lee et al., 2014). On the other hand, using C3H/HeN mice, it was reported that rBIP was not a potent stimulator of proliferation in cell culture using the splenocytes of infected mice, yet this variance could be attributed to the differing mouse strains used. However, rBIP did seem to be immunogenic as the serum of C3H/HeN mice, which were re-infected with GS Giardia trophozoites 6-week post infection, had greater binding to rBIP protein in vitro when compared to non-re-infected samples. Furthermore, when mice were infected and re-infected with a culture of GS trophozoites containing greater numbers of parasites that expressed the immunodominant antigen 5G8 (Quintero et al., 2013), IgG binding to rBIP protein was even greater than infections using standard GS cultures (Lopez-Romero et al., 2017).

The effects of ESP and their proteases were described earlier in this chapter. Jimenez et al. showed that after infection with G. duodenalis in mice, serum IgG1, IgG2, IgA, IgE and secretory IgA against ESP and Giardia extract were increased, although the increase was only significant for some of the antibodies (Jimenez et al., 2014). Furthermore, restimulation of spleens from infected mice with ESP led to an increase in spleen cell proliferation and an increase in the amount of cytokines produced (Jimenez et al., 2014). ESP restimulation increased IL-4 responses at day 7 post-infection, while increased IFN-γ was seen 14 days post-infection, increased IL-5 21 days post-infection, and increased IL-10 at 21 and 28 days post-infection. Overall, this research supports the immunogenicity of E/S products (Jimenez et al., 2014).

Finally, α1-giardin and cyst wall proteins (CWPs) have also been examined as potential antigens to generate vaccines (Feng et al., 2016). A bivalent DNA vaccine with both α1-giardin and CWP2 DNA leads to both decreased trophozoite and fecal cyst numbers (79% reduction in trophozoites and 93% reduction in cysts) (Feng et al., 2016). In contrast, univalent α1-giardin vaccines generated a strong trophozoite decrease (74.2%) but a lower cyst output decrease (28%), while a univalent CWP2 vaccine generated decreased cyst output only (89%) (Feng et al., 2016). A combined vaccine allowed for dual targeting of the parasite, both its trophozoite and cyst form (Feng et al., 2016). Additionally, the bivalent vaccine led to an increase of both anti-α1-giardin IgG2 and anti-CWP2 IgG2, as well as an increase in both anti-α1-giardin IgA and anti-CWP2 IgA (Feng et al., 2016).

6. Pathogenic Responses

While the perfect immune response would succeed in eliminating infections without producing concomitant damage in the host, it is clear that most actual immune responses produce significant ancillary pathology. Early work from our lab and others showed that intestinal transit rates were accelerated in wild-type mice infected with Giardia, but not in SCID mice (Andersen et al., 2006; Li et al., 2006), that smooth muscle hypercontractility was dependent on mast cell function (Li et al., 2007; Li et al., 2004) and that disaccharidase deficiency involved CD8⁺ T cells (Keselman et al., 2016; Scott et al., 2004). More recent work has focused on understanding regulation of these immune responses in order to help understand the variability seen in the outcomes of infections in humans.

6.1. IFN-γ

Previous work has indicated that neither Th1 (IFN-γ) nor Th2 (IL-4, IL-5) response are required for control of Giardia infections (Singer and Nash, 2000b). IFN-γ^−/− mice cleared the infection as well as wild-type controls. Recently, Babaei et al. (2006) measured IFN-γ levels in symptomatic and asymptomatic patients infected with Giardia, along with uninfected controls (Babaei et al., 2016). IFN-γ levels were increased in the symptomatic patient group compared to the uninfected controls. The difference in IFN-γ levels between symptomatic and asymptomatic patients was not statistically significant, although there was a trend toward more IFN-γ in the symptomatic group. (Babaei et al., 2016). Studies with larger sample sizes should be conducted in order to further test this hypothesis.

Additional information regarding the role of IFN-γ has recently come from a study of children in Ecuador infected with the soil-transmitted helminth, Ascaris lumbrocoides, G. duodenalis, or both (Weatherhead et al., 2017). Children who were co-infected with Ascaris and Giardia had reduced serum levels of IL-2, IL-12, and TNF-α compared to children infected with only Giardia (Weatherhead et al., 2017). While each of these, is associate with Th1 responses, the level of IFNγ was not found to be reduced. Interestingly, the IL-10/IFN-γ ratio was higher in children with co-infections compared to children infected with Giardia alone (Weatherhead et al., 2017). Levels of the Th2 cytokines IL-4, IL-5 and IL-13 were also not different among groups in this study. Whether children with co-infection exhibited differences in acute symptoms like diarrhea or chronic symptoms including stunting was not investigated. In a different cohort of children in Venezuela, Giardia infection in children with a light concurrent Ascaris infection (<5,000 eggs/g feces) exhibited increased serum IFN-γ levels compared to children with similar Ascaris-only infections (Hagel et al., 2011). However, in children where the Ascaris infection was more intense, the levels of IFN-γ between Giardia co-infected and Ascaris-only infected individuals were not significantly different (Hagel et al., 2011). Given that polyparasitism is common in children in low and middle income countries, additional analyses of co-infections are clearly warranted.

6.2. Other cytokines

In addition to IFN-γ and IL-17, other cytokines play important roles during a Giardia infection. While IL-17 appears to be the most important cytokine for controlling infections, it was actually among the least abundant cytokines produced when splenocytes or mesenteric lymph node cells from infected mice were stimulated in vitro with parasite extracts (Solaymani-Mohammadi and Singer, 2011). Data suggest that IL-4, IL-5, IL-6, IL-10, IL-13, IL-17, IL-22, IFN-γ and TNF-α are all produced following Giardia infections in mice (Bartelt et al., 2017; Dann et al., 2015; Li et al., 2014; Solaymani-Mohammadi and Singer, 2011; Zhou et al., 2003). Recent data from gerbils infected with G. duodenalis strain WB1267 also indicated a mixed cytokine response to infection. Messenger RNA for almost all of the cytokines queried were elevated in mesenteric lymph nodes at day 14. IL-5 mRNA was not elevated at day 14, although it was seen at day 28. (Serradell et al., 2018). The changes in cytokines observed over time likely reflect evolution of the immune response as the majority of parasites are eliminated between days 7 and 14 in this model.

The abundant production of IL-13 was reported for spleen cells from mice infected with either GS or WB strain Giardia after in vitro stimulation with parasite extracts (Solaymani-Mohammadi and Singer, 2011). Bartelt et al., also reported elevated IL-13 mRNA in mice infected with cysts of strain H3 (Bartelt et al., 2017). While the role of this cytokine has not been tested directly, it is intriguing to speculate that it may be involved in the goblet cell responses, including increased expression of Muc2, TFF3 and RELMβ that has been reported (Amat et al., 2017; Paerewijck et al., 2017).

Interleukin-10 (IL-10) is an anti-inflammatory cytokine involved in limiting chronic inflammation to commensal microbial antigens and self-proteins. Mucosal IL-10 mRNA levels were shown to increase between weeks 2–4 post infection in mice infected with G. muris (Dann et al., 2018). IL-10 mRNA was also increased in mesenteric lymph nodes of gerbils infected with G. duodenalis (Serradell et al., 2018) and spleen cells from mice infected with G. duodenalis produced IL-10 after in vitro stimulation with parasite extracts (Solaymani-Mohammadi and Singer, 2011). In the G. muris model, IL-10 mRNA levels peaked in the mucosa at week 3 corresponding with the major decrease in trophozoites in the small intestine (Dann et al., 2018). However, IL-10^−/− mice were not deficient in clearing the parasite, indicating that this cytokine is not required for protective immunity.

Intriguingly, mice lacking IL-10 developed more severe symptoms from the infection including diarrhea and colonic inflammation (Dann et al., 2018). In IL-10^−/− mice, the colon which is not the major site for Giardia infection showed increased inflammation as early as 1 week post infection, with increasing crypt depth and lamina propria cell accumulation in weeks 4 and 7 post infection (Dann et al., 2018). Neutrophil recruitment was also increased in the IL-10^−/− infected mice (Dann et al., 2018). The mRNA levels of the CXCL5 neutrophil chemokine were high on days 5 and 12 post-infection; MPO mRNA levels increased by day 12 (Dann et al., 2018). Dann et al. concluded that these symptoms resemble the colitis IL-10^−/− mice develop naturally as they age (Dann et al., 2018). TCRα^−/− mice similarly develop colitis-like symptoms as they age; G. muris infection in TCRα^−/− mice led to similarly high histological scores, increased crypt depth, and lamina propria cell accumulation, although unlike the IL-10 deficient mice, these mice failed to control trophozoite counts (Dann et al., 2018).

T cells seem to play a role in developing post-infectious colitis in IL10^−/− mice (Dann et al., 2018). Infected IL10^−/−Rag2^−/− mice had a reduced histological score compared to infected IL10^−/− mice, although there was significantly more inflammation than in infected wild-type mice or Rag2^−/− mice (Dann et al., 2018). This suggests that both T cells and non-T cells contribute to colitis in this model. Indeed, the authors demonstrated dependence on CD11c⁺ cells (most likely dendritic cells), the intestinal microbiota and the MyD88 adaptor protein. These data are consistent with a model in which Giardia infection leads to epithelial barrier defects and microbial translocation that initiates an inflammatory response that is normally controlled by IL-10 dependent homeostatic mechanisms.

T cells are clearly not the only source of cytokines during a Giardia (or any) infection. Interleukin-6 was the first cytokine demonstrated to have a significant role in parasite clearance, since IL-6^−/− mice take significantly longer to clear the infection compared to wild-type controls (Bienz et al., 2003; Zhou et al., 2003). While many cell types can produce IL-6, bone marrow chimera experiments have demonstrated that IL-6 from bone marrow derived cells is necessary and sufficient for control of Giardia (Kamda et al., 2012). Furthermore, transferring spleen T cells from IL-6^−/− donors into TCRβ^−/− mice allowed these mice to control infections, suggesting that T cell production of IL-6 is not crucial (Kamda et al., 2012). Interestingly, dendritic cells were identified as a potent source of IL-6 (Kamda et al., 2012). Adoptive transfer of IL-6 competent dendritic cells, pulsed with Giardia antigens prior to transfer into IL-6^−/− mice, reduced parasite counts following infections compared to transfers of just extract or dendritic cells which were not pulsed with antigens (Kamda et al., 2012). Furthermore, since IL-6 in combination with TGF-β is able to drive differentiation of naïve T cells into IL-17 producing cells (Weaver et al., 2006), these data suggest that IL-6 from dendritic cells helps drive development of IL-17 producing CD4⁺ T cells, leading to successful control of infections. Direct demonstration of a role for IL-6 in Th17 development during giardiasis remains to be demonstrated.

Lee et al. looked at cytokine expression from human PBMCs and the monocytic cell line THP-1 stimulated with GS and WB (Lee et al., 2012). PMBCs from naïve donors produced TNF-α, IL-6, IL-8, IL-1β and IFN-γ, with GS stimulation facilitating higher levels of cytokine production than WB (Lee et al., 2012). THP-1 cells were also more responsive to GS, but only secreted TNF- α, IL-1β, IL-8, and IFN-γ in response (Lee et al., 2012). A similar experiment was performed with HT-29 cells which were physically separated from the live Giardia, so the effects of excretory/secretory antigens could be assessed (Lee et al., 2012). While stimulation of HT-29 cells with live G. duodenalis increased production of TNF-α, IL-6, and IL-1β, IL-8 responses decreased with increasing numbers of Giardia trophozoites (Lee et al., 2012). A potential explanation is that IL-8 is cleaved/degraded by cysteine proteases released from the parasite (Liu et al., 2019; Ma’ayeh et al., 2018). Additional data indicated that IL-8 transcription is dependent on ERK1/2 and p38 signaling and activation of the transcription factors NF-κB and AP-1 (Lee et al., 2012). However, the receptors involved in recognition of parasite antigens were not identified and the role of macrophage responses in contributing to the outcomes of infection remain unclear.

6.3. CD8⁺ T Cells and Disaccharidase Deficiency

As we mentioned during discussion of the role of the microbiome in giardiasis, CD8⁺ T cells have been implicated in mediating some of the epithelial damage observed in giardiasis. Specifically, upon adoptive transfer of either total mesenteric lymph node cells or purified CD8⁺ T cells from mice infected with G. muris to athymic mice, shortening of the microvilli, loss of sucrase activity, and increase of the crypt/villus ratio are observed in the mucosa (Scott et al., 2004). The same effects are not observed in an adoptive transfer of CD4⁺ T cells from infected mice (Scott et al., 2004). These observations were confirmed by analysis of mice with a deletion in the gene for β₂-microglobulin, a component of all class I MHC molecules. These mice lack class I MHC molecules, do not develop CD8⁺ T cell populations and do not exhibit sucrase defects following Giardia infection, although they eliminate the infections as well as wild-type mice (Solaymani-Mohammadi and Singer, 2011).

Given that Giardia is an extracellular pathogen, it is unclear if or how CD8⁺ T cells, which normally respond to viral antigens or other peptides found in host cell cytosol, would become activated. Keselman used flow cytometry to demonstrate that CD8⁺ T cells in the lamina propria of Giardia infected mice exhibit an activated phenotype, i.e. high expression of CD44 and CD69. Interestingly, antibiotic treatment of mice ameliorated the sucrase deficiency normally seen following Giardia infection. It also prevented the increase in CD8⁺ T cells with an activated phenotype (Keselman et al., 2016). In contrast, antibiotic treatment did not seem to have an effect on the increased accumulation of CD4⁺ T cells in the lamina propria following infection (Keselman et al., 2016). Activation of certain intraepithelial lymphocyte populations was also observed following Giardia infection in this study, although antibiotic treatment did not seem to affect this response. These data suggest that intestinal microbiota somehow contribute to activation of CD8⁺ T cells and the reduction in disaccharidase activity that is often observed in giardiasis.

Mucosal integrity can be severely affected by a Giardia infection, especially if coupled with malnourishment. Shukla et al. measured sucrase, lactase, and maltase activity, as well as alkaline phosphatase in wildtype and malnourished mice infected with Giardia duodenalis (Shukla et al., 2013). Giardia decreased disaccharide activity and alkaline phosphatase at 4, 9, and 10 days of infection (Shukla et al., 2013). Malnutrition and probiotic treatment at different time points modified the effect of the parasite on disaccharide activity, although the specifics varied based on the disaccharidase being analyzed (Shukla et al., 2013). Nevertheless, these data are consistent with the model that intestinal microbiota are key players in the development of the disaccharidase deficiency phenotype and that they act by modulating the activation of CD8⁺ T cells. The exact role of the microbiota in activation of these cells remains to be determined.

6.4. Mast Cells and Irritable Bowel Syndrome

Giardiasis has been linked to a number of long-term sequelae, including chronic fatigue and irritable bowel syndromes. Several studies followed a population in Norway after a Giardia infection outbreak there in 2004. Hanevik et al. considered a group of patients for post-infectious chronic fatigue syndrome (PI-CFS), post-infectious functional gastrointestinal disorders (PI-FGID), as well as a potential co-morbidity relationship between the two (Hanevik et al., 2012). Five years post-infection the population of CD8⁺ cells in PI-FGID-reporting patients was increased, while PI-CFS-reporting patients had a lower count of natural killer cells (Hanevik et al., 2012). Later work in the same population specifically focused on PBMCs from PI-CFS patients stimulated with antigens from Giardia assemblage A or assemblage B parasites. No statistical differences were seen comparing activation of CD4⁺ T cell responses between patients with PI-CFS and those who had been infected with Giardia but did not develop CFS (Hanevik et al., 2017). Similarly, the amount of cytokines produced including IL-4, IL-6, IL-10, IFN-γ, TNF-α, and IL-2 were also similar (Hanevik et al., 2017). There was significantly more sCD40L in supernatants of stimulated PBMCs obtained from PI-CFS patients, although the authors concluded that further studies will be needed to determine if this is specifically a marker of post-Giardia infectious CFS (Hanevik et al., 2017).

In terms of irritable bowel syndrome (IBS), work in mice has shown that tight junctions in the jejunum are impaired during Giardia infection, specifically that occludin is cleaved on day 7 post-infection (peak of giardiasis) and day 35 post-infection (after resolution) (Chen et al., 2013). Permeability is also increased in jejunal tissues at both time points and in colon tissues at the later time point (Chen et al., 2013). Neutrophil counts, as well as myeloperoxidase activity, in the mucosa are higher both at day 7 and day 35 after oral gavage with G. duodenalis (Chen et al., 2013). IFN-γ levels are also increased on day 7 and day 35, while TNF-α and IL-1β levels are only increased on day 35 (Chen et al., 2013). Overall, the authors hypothesized that a Giardia infection leads to disrupted epithelial integrity, followed by an immune response to the microbes in the intestine that are now able to traffic across the damaged layer (Chen et al., 2013).

Fisher et al set up an in vitro system to model the interactions between G. duodenalis, the epithelial layer, and macrophages residing under the epithelium. The purpose of this model was to allow for study of prolonged giardiasis in vitro (Fisher et al., 2013). Similar to earlier work using either a different cell line or patient biopsies (Chin et al., 2002; Troeger et al., 2007), they showed that Giardia led to an increase in the activation of caspase-3 in epithelial cells, indicating apoptosis of these cells (Fisher et al., 2013). Furthermore, when culturing CaCo2 cells and macrophages together, the authors saw an increase in production of the chemokines GRO and IL-8, both of which are involved in neutrophil recruitment. Interestingly, this response was diminished when Giardia was cultured with the CaCo2 cells and macrophages (Fisher et al., 2013), consistent with results described above where parasite proteases can degrade IL-8.

In addition to acute symptomatic infections, Giardia often causes sub-clinical infections which can become chronic or recurrent. It is thought that chronic and recurrent Giardia infections contribute to growth faltering in children. Although studies have often been contradictory and inconclusive (Bartelt and Sartor, 2015), recent data from the very large MAL-ED cohort study suggests that Giardia is one of the four main contributors to stunting in children (Bartelt and Sartor, 2015; Rogawski et al., 2017; Rogawski et al., 2018). Bartelt et al. (2017) developed a chronic giardiasis mouse model utilizing H3 G. duodenalis cysts in under-nourished mice to investigate the interaction of Giardia infection and malnutrition. Similar parasite counts were observed at 8 days post infection and at 64 days post infection in the duodenum, indicating a chronic infection (Bartelt et al., 2013). Additionally, increased crypt depth and eosinophil counts in both the villi and crypts were observed on day 64 (Bartelt et al., 2013). Mice with Giardia had reduced weight gain compared to uninfected mice on the low-protein diet; additionally, villus height was also shown to decrease with malnourishment prior to infection (Bartelt et al., 2013). In varying the length of malnourishment prior to infection, the authors discovered that a 15-day low-protein diet expedited weight loss compared to an 8-day diet (Bartelt et al., 2013). Furthermore, while in nourished infected mice, there was an influx of neutrophils and an increase in IL-4 and IL-5 mRNA, these responses were decreased in malnourished infected mice (Bartelt et al., 2013). These data indicate that nutrition and infection interact with immune responses in complicated ways and provide a valuable model for environmental enteropathy in humans.

7. Remaining Questions

While there has been significant progress in several areas, there remain numerous areas important for further study. In terms of the innate response to Giardia, it remains unclear which pattern recognition receptors recognize antigens from Giardia and contribute to the activation of macrophages and dendritic cells. Importantly, we do not know which innate cells provide the signals that help drive differentiation of the IL-17 producing CD4⁺ T cells that are required for control of the infection. While several responses downstream of IL-17 have been investigated, it remains unclear which effector mechanisms are most important for control of the infection. Similarly, it is unclear how the CD8⁺ T cell response that leads to shortening of the microvilli and reduced levels of disaccharidases in the mucosa are activated. If these CD8⁺ T cells target parasite antigens, it is unclear how peptides from an extracellular pathogen would be presented in class I MHC molecules that normally present antigens from the host cell cytosol. Finally, the contribution of immune responses to the development of acute symptoms of giardiasis in humans has not been thoroughly studied, nor is the role of immune responses in long-term sequelae (e.g., stunting, irritable bowel syndrome or chronic fatigue syndrome) of the infection understood.