Molecular analysis of antigenic variation in Giardia lamblia and influence of intestinal inflammatory reactions on a Giardia lamblia infection in mice

Inauguraldissertation der Philosophisch-naturwissenschaftlichen Fakultät der Universität Bern

vorgelegt von Nicole Eva von Allmen von Lauterbrunnen

Leiter der Arbeit: Prof. Dr. Norbert Müller Institut für Parasitologie der Veterinärmedizinischen und der Medizinischen Fakultät Universität Bern

Molecular analysis of antigenic variation in Giardia lamblia and influence of intestinal inflammatory reactions on a Giardia lamblia infection in mice

Inauguraldissertation der Philosophisch-naturwissenschaftlichen Fakultät der Universität Bern

vorgelegt von Nicole Eva von Allmen von Lauterbrunnen

Leiter der Arbeit: Prof. Dr. Norbert Müller Institut für Parasitologie der Veterinärmedizinischen und der Medizinischen Fakultät Universität Bern

Von der philosophisch-naturwissenschaftlichen Fakultät angenommen.

Bern, 24. November 2005 Der Dekan: Prof. Dr. P. Messerli von Allmen Nicole Eva - 1 -

1 Abstract

Giardia lamblia is a significant, environmentally transmitted, human and an amitochondriate protist, often hypothesized to be the most basal eukaryote. It is a major contributor to the enormous worldwide burden of human diarrheal diseases, yet the basic biology of this parasite is not well understood. Antigenic variation of G. lamblia is a widely investigated field of this protozoan parasite and has also driven our experimentations in the beginning of this study. Experimental infections in the mother-offspring mouse model were performed to investigate on one hand the antigenic variation process on the transcriptional level, and on the other to simulate the natural infection mode of the parasite. Our results demonstrated that antigenic switching of duodenal trophozoite and caecal cyst populations was accompanied with an obvious reduction in vsp H7 mRNA levels but without a simultaneous increase in transcripts of any of the analyzed subvariant vsp genes. The simulated natural transmission of the parasite revealed an antigenic reset of the excysted trophozoite population thereby the new generation essentially consisted of the original VSP H7 type. Investigations of antigenic variation on the molecular level were established by analysing giardial vsp RNA levels focusing on sense and complementary antisense vsp transcripts. The study analyzed not only vsp genes involved in antigenic variation but also a major gene of encystation, cwp 1 encoding cyst wall 1 (CWP 1). In the first case, we were able to demonstrate that sense vsp H7 RNA predominated in VSP H7-type trophozoites whereas sense RNA from only one of the 8 investigated subvariant vsp genes, vsp IVg, had increased in the subvariant trophozoite population. In both groups, H7-type and subvariant type trophozoites, similar relative distribution regarding vsp H7, or vsp IVg antisense RNA molecules could be found. Analogous to these findings were the results from sense and antisense RNA transcripts of cwp 1. Finally, these data demonstrated that giardial antisense RNA production was directly linked to complementary sense RNA production and made speculations about the RNA interference mechanism in Giardia lamblia doubtful.

The contribution of intestinal and host immune responses are still a well discussed subject in Giardia research. There exist controversial studies about the importance and immunological relevance of mast cells for example as well as for effector mechanisms like T-cell and -dependent immune responses. With the induction of a transient intestinal inflammation, realized by a co-infection experiment with Trichinella spiralis as inducer of the inflammation, we analyzed the G. lamblia parasite burden, the mucosal mast cells and associated IL-6 production of mast cells as well as the IgE

von Allmen Nicole Eva - 2 - production. Our findings raised the possibility that the inflammatory responses to an intestinal infection favours establishment and maintenance of a G. lamblia infection in mice.

von Allmen Nicole Eva - 3 -

Table of Contents

1 Abstract 1 2 Abbreviations 4 3 List of figures 5 4 Introduction 6 4.1 Giardia lamblia 6 4.2 Life cycle 7 4.2.1 Trophozoites 8 4.2.2 Cysts 9 4.2.3 Transmission 10 4.3 Epidemiology 12 4.4 Pathogenesis and clinical symptoms 12 4.5 Diagnosis and treatment 13 4.6 Encystation 14 4.7 Excystation 15 4.8 Characterization of the G. lamblia transcriptome 16 4.9 Surface antigenic variation in G. lamblia 17 4.10 Immunological host reactions against G. lamblia infections 20 4.11 Physiological host reaction against G. lamblia infections 22 4.12 Intestinal pathogenesis associated with G. lamblia infections 24 5 Aim of the present thesis 28 6 Summary of publications 30 I) N. von Allmen, M. Bienz, A. Hemphill and N. Müller. 2004. Experimental infections of neonatal mice with cysts of Giardia lamblia GS/M-83-H7 are associated with an antigenic reset of the parasite. Infection and 72: 4763-4771.

II) N. von Allmen, M. Bienz, A. Hemphill and N. Müller. 2005a. Quantitative assessment of sense and antisense transcripts from genes involved in antigenic variation (vsp genes) and encystation (cwp 1 gene) of Giardia lamblia clone GS/M-83-H7. Parasitology 130: 389-396.

III) N. von Allmen, S. Christen, U. Forster, B. Gottstein, M. Welle and N. Müller. 2005b. Acute trichinellosis increases susceptibility of Giardia lamblia infections in the mouse model. Submitted to Infection and Immunity.

IV) N. Müller, N. von Allmen. 2005. Recent insights into the mucosal reactions associated with Giardia lamblia infections. International Journal of Parasitology, in press (Review).

7 References 34 8 Publications 43 9 Acknowledgements 94 10 Curriculum Vitae 95

von Allmen Nicole Eva - 4 -

2 Abbreviations

B-cell Bone marrow derived cell B10 Mice strain exhibit a low IgG production against a variety of T-dependent CD8+ T Class I MHC-restricted, cytotoxic T cells CM Cyst membrane C-terminus Carboxy terminus of a peptide CWP Cyst wall protein cwp Gene encoding cyst wall protein EGF Epidermal growth factor ELISA Enzyme linked immunosorbent assay ESVs Encystation-specific vesicles IFN-γ Interferon gamma IgA IL Interleukin () iNOS Inducible nitric oxide synthase mAB Mb Million base pairs MMP Matrix metalloproteinase NO Nitric oxide N-terminus Amino terminus of a peptide PCR Polymerase chain reaction RNAi RNA interference RT-PCR Reverse transcription polymerase chain reaction SAGE Serial analysis of gene expression STAT Transcription factor that stimulates transcription of IL-4 and other TH2 cytokine genes T-cell Thymus derived cell VSP Variant surface protein vsp Gene, encoding variant surface protein Zn Zinc

von Allmen Nicole Eva - 5 -

3 List of figures

Fig. 1: Giardia lamblia trophozoite (Dönges, 1988; von Allmen, N. 2003) Fig. 2: Giardia lamblia cyst (Dönges, 1988; von Allmen, N. 2003) Fig. 3: Giardia lamblia life cycle (http://sprojects.mmi.mcgill.ca/tropmed/disease/giardia/life.htm Fig. 4: VSP H7-type Giardia lamblia trophozoites (von Allmen N., 2004)

von Allmen Nicole Eva - 6 -

4 Introduction

4.1 Giardia lamblia

Giardia lamblia (syn. Giardia intestinalis, Giardia duodenalis) is a protozoan flagellate which colonises the lining of the upper part of the small intestine. This early diverged extant organism lives anarobically in the gut and causes gastrointestinal infections in humans and various other mammalian hosts (Adam, 1991). It is the most commonly diagnosed intestinal parasite worldwide with a prevalence of 2-3% in industrial and 20- 30% in development countries. Giardiasis infection begins in most cases with the ingestion of waterborne cysts or food contaminated with cysts. In addition to being significant for the medical importance of G. lamblia, the Giardia species are of interest because they belong to the Diplomonadida, one of the most basal eukaryotic branches on a phylogenetic tree (Sogin et al., 1989). G. lamblia and the other Diplomonadida are also of interest because of their possession of two nuclei. The two nuclei replicate approximately at the same time (Wiesenhahn et al., 1984), are both transcriptionally active (Kabnick et al., 1990) and contain similar quantities of DNA (Bernander et al., 2001). G. lamblia was first described by van Leeuwenhoek in 1681 when he investigated his own diarrhoeic stool under a microscope made by glass lenses and set them into metal frames. The protozoan was initially named Cercomonas intestinalis by Vilem Lambl in 1859, because he thought the organism belonged to the genus Cercomonas. Subsequently, Kofoid and Christiansen renamed the organism G. lamblia in 1915 (Adam, 2001). Giardia is in the subphylum Sarcomastigophora, the super class Mastigophora and in the order Diplomonadida. It belongs to the family Hexamiticae which contains six genera, three of which, including Giardia, are exclusively parasitic. There are three major morphological subtypes of Giarda: G. agilis from amphibians, G. muris from mice and G. intestinalis from humans and some other vertebrates. These three types can be distinguished by the overall shape and dimensions of the trophozoite body and also by the distinctive shapes of the median bodies.

von Allmen Nicole Eva - 7 -

4.2 Life cycle (e.g. reviewed by Adam, 1991; Thompson, 2000)

The Giardia life cycle is direct and asexual and begins with uptake of the cyst by the host. The exposure of cysts to gastric acid and to proteases as they move through the stomach triggers the excystation process. Colonization involves three processes, namely excystation, attachment of trophozoites to the intestinal epithelium by means of its adhesive disk and multiplication in the duodenum. There is an alternation between the binucleated trophozoite stage and environmentally resistant, infectious cysts. The formation of cysts is triggered by the dehydration of gut contents moving through the large intestine (Keas, 1999; www.msu.edu/course/zol/316/glaminfect.). Cysts are characterized by a thick extracellular matrix forming a cyst wall and are resistant to adverse environmental conditions (Adam, 2001). Although the parasite life cycle is simple, encystation and excystation are not fully understood. In vivo excystation was first described in 1925, but it was not until 1979 that the process was produced under laboratory conditions (Isaac-Renton et al., 1992).

von Allmen Nicole Eva - 8 -

4.2.1 Trophozoites (reviewed by Adam, 1991)

The trophozoite is very easily identified because of its numerous distinguishing features. The teardrop shaped trophozoite has almost perfect bilateral symmetry with two nuclei at the anterior end (Fig. 1). The nuclei look like a set of eyes, starring up at the observer. The two nuclei are assumed to be identical. They replicate at arrangement like approximately the same time (Wiesenhahn et al., 1984) and are both transcriptionally active (Kabnick and Paeatti, 1990). The trophozoites acquire their tumbling leaf-like motility with the aid of four pairs of flagella distributed evenly on both sides of the organism. They also have axostyles. A cross section of an axostyle revealed a nine plus two microtubule most flagella and cilia. The trophozoites live upon the surface of the villi and attaches to intestinal epithelial cells with a unique sucking disc that is present on its ventral surface. The adhesive power of the sucking disc is strong enough to resist detachment during peristaltic contractions, therefore may also cause damage to the intestinal lining and leaves behind a circular indentation. Trophozoites can be found in jejunal aspirate and sometimes in the liquid stool of an infected person.

(Dönges, 1988) (von Allmen, N. 2003)

Fig. 1: Giardia lamblia trophozoite, for explanation see text.

von Allmen Nicole Eva - 9 -

4.2.2 Cysts

The cystic form of G. lamblia is 8-14µm x 5-10µm, slightly smaller than the trophozoite and egg shaped. The cyst has thick exterior wall that allows it to withstand harsh conditions like cold water, chlorination or low pH levels. An immature cyst has two nuclei and a mature cyst has four nuclei (Fig. 2). It survives in water or in hypotonic solutions, and at low temperatures (Nino et al., 2003). The cyst contains remnants of the axostyles and parabasal body. The cyst is the infectious form of G. lamblia and is encased in a resistant fibrillar extrecellular matrix composed of at least three cyst wall (CWPs) and glycans (GalNAc homopolymer) (Adam, 2001). Cyst wall proteins are expressed in a stage-specific manner and form a thick and resistant wall enclosing two trophozoites.

(Dönges, 1988) (von Allmen, N. 2003)

Fig.2: Giardia lamblia cyst; for explanation, see text.

von Allmen Nicole Eva - 10 -

4.2.3 Transmission

Infection occurs by ingestion of cysts (Adam, 1991), usually with contaminated drinking water (Fig. 3). Also person-to-person transmission and food-borne transmission are recognised routes of infection. Exposure of ingested cysts to gastric acid during passage through the host stomach triggers excystation (Adam, 2001). Excystation in the small intestine releases two trophozoites into the upper small bowel. After entry into the small intestine and stimulation by intestinal pH, bicarbonate, and proteases, the parasite emerges and divides into two identical binucleate flagellated trophozoites (Bernander et al., 2001). The vegetative form of the parasite colonises the human small intestine below the entrance of the common bile duct, where it attaches to enterocytes and mucus by a ventral sucking disk or swims in the intestinal fluid (Adam, 2001). Trophozoites feed on mucus secretion and multiply by longitudinal binary fission remaining in the lumen of the proximal small bowel. They reproduce and are swept down the intestine in the faecal stream. Trophozoites which are free in the lumen of the intestine begin to dehydrate in the large intestines so they start to encyst as the parasites are transported toward the colon. The new formed cysts are passed as accumulated piles in the faeces to the environment and may survive over several months under moisture and cool conditions and are responsible for transmission of giardiasis. Both trophozoites and cysts can be found in the excrement and serve for the diagnostic finding (Adam, 1991).

von Allmen Nicole Eva - 11 -

Fig. 3: Giardia lamblia life cycle (http://sprojects.mmi.mcgill.ca/tropmed/disease/giardia/life.htm)

von Allmen Nicole Eva - 12 -

4.3 Epidemiology (reviewed by Adam, 1991)

Giardia is found in soil, food, water, or surfaces that have been contaminated with the faeces from infected humans or animals. Several community-wide outbreaks of giardiasis have been linked to drinking municipal water or recreational water contaminated with Giardia. Giardiasis is more prevalent in children than in adults, possibly because many individuals seem to have a lasting immunity after infection. This organism is implicated in 25% of the cases of gastrointestinal disease and may be present asymptomatically. The overall incidence of infection in the United States is estimated at 2% of the population (Centre for food safety and applied nutrition; www.cfsan.fda.gov/~mow/chap22.html). This disease afflicts many homosexual men, both HIV-positive and HIV-negative individuals. This is presumed to be due to sexual transmission. The disease is also common in child day care centres, especially those in which diapering is done. The disease has prevalence in developing countries, where infections are associated with poor sanitary conditions and poor water quality control. Giardiasis affects three times more children than adults (Harms et al., 2002).

4.4 Pathogenesis and clinical symptoms

About 40% of those who are diagnosed with giardiasis demonstrate disaccharide intolerance during detectable infection and up to 6 months after the infection can no longer be detected. Lactose (i.e., milk sugar) intolerance is most frequently observed (Lindley et al., 2001). Some individuals (less than 4%) remain symptomatic more than 2 weeks; chronic infections lead to a malabsorption syndrome and severe weight loss. The immune status of the host appears to influence the susceptibility of the host to infection and the severity of clinical signs (Olson et al., 2000; Müller and von Alllmen, 2005). Chronic cases of giardiasis in immundeficient and normal individuals are frequently retractile to drug treatment. In some immune deficient individuals, giardiasis may contribute to a shortening of the life span. Giardia undergoes two pathogenetically important types of differentiation: encystation is required for survival outside the host, and excystation is necessary for infection. Giardia infection can cause a variety of intestinal symptoms, which include diarrhea, gas or flatulence, greasy stools that tend to float, stomach cramps, upset stomach or nausea. These symptoms may lead to weight loss and dehydration. Some people with giardiasis have no symptoms at all. But when symptoms appear they include diarrhea, loose or watery stool, stomach cramps, malabsorption and upset stomach (Gardner and Hill,

von Allmen Nicole Eva - 13 -

2001). In otherwise healthy persons, symptoms of giardiasis may last 2 to 6 weeks. Occasionally, symptoms last longer. The more chronic stage is associated with vitamin B12 malabsorption, disaccharidase deficiency and lactose intolerance. These symptoms may lead to weight loss and dehydration. The reasons for the varying clinical manifestations are unknown but may be due to variable virulence of the isolates (Nash et al., 1987).

4.5 Diagnosis and treatment (reviewed by Adam, 1991)

Giardia lamblia is frequently diagnosed by visualizing the organism, either the trophozoite (active reproducing form) or the cyst (the resting stage that is resistant to adverse environmental conditions) in stained preparations or unstained wet mounts with the aid of a microscope. Multiple examinations are often required to diagnose the infection because of the irregular shedding of parasites. A commercial fluorescent antibody kit is available to stain the organism. Organisms may be concentrated by sedimentation or flotation; however, these procedures reduce the number of recognizable organisms in the sample. An enzyme linked immunosorbant assay (ELISA) that detects excretory secretory products of the organism is also available (Vesy and Peterson, 1999). So far, the increased sensitivity of indirect serological detection has not been consistently demonstrated. Presently, there is no drug available to prevent giardiasis. Specifically precautions to avoid an infection with G. lamblia should be taken before visiting an area where the parasite may exist (Gardner and Hill, 2001). The drug of choice for the treatment of giardiasis remains Metronidazole (Flagyl), but Quinacrin hydrochloride and Furazolidone are also frequently used. However, drug resistance has been observed with each of these compounds. In addition, toxicity has restricted their use in women of child- bearing age and although less effective, Furazolidone has been used preferentially for children as it is can be administered as a suspension (Eckman and Gillin, 2001).

von Allmen Nicole Eva - 14 -

4.6 Encystation

The differentiation process from trophozoites to cysts begins when trophozoites confront low concentrations of cholesterol in the inferior parts of the intestine (Lujàn et al., 1996). When trophozoites leave the mucus layer, in response to unknown stimuli, it is exposed to lumenal content that triggers encystation. Encystation triggers are bile salts and a slightly alkaline pH (7.8) was subsequently found to be optimal inducers of encystation (Eichinger, 2001). Cysts possess a rigid extracellular wall composed of both protein and carbohydrate molecules, the wall also comprises an inner and outer membrane, which surround the periplasmic space; the outer part contains filaments of different lengths. In mature Giardia cysts original plasma membrane of the trophozoite becomes the outermost membrane of the cyst wall (CM1). The large vacuoles form a second membrane surrounding the cyst (CM2), and also form a third membrane (CM3), that becomes the new plasma membrane of the trophozoites (Chavez-Munguia et al., 2004). Encystation is accompanied by the formation of typical secretory granules called encystation-specific vesicles, which transport cyst wall proteins (CWPs) and carbohydrates to the cell surface for release and assembly into the protective fibrillar cyst wall. The amino-terminal overlapping domains of CWP1 (26kDa) and CWP2 (39kDa) are 61% identical in sequence; contain a cysteine-rich region and a five tandem copies of leucine-rich repeats (Touz et al., 2002). The cyst wall proteins are targeted to the secretory apparatus by their predicted hydrophobic leader sequence where they apparently associate covalently via disulfide bonds shortly after their synthesis. Proteins and carbohydrates destined for the cyst wall are transported by a novel class of as yet undefined vesicles termed encystation-specific vesicles (ESVs) (Hehl et al., 2000). The ESV begin to appear between 3 and 6 hr in trophozoites grown in the presence of bile and a pH of 7.8 , that is, in conditions that mimic the small intestinal environment. These encysting trophozoites loose the ability to attach to the culture vessel as they begin to round up to form cysts in vitro (Faubert et al., 1991). The ability to induce Giardia encystation in vitro makes this organism an excellent model to learn about the process involved in gene regulation and signal transduction (Lujàn et al., 1997). Gillin (1987) first described the in vitro encystations of G. lamblia trophozoites after the addition of primary bile salts to the TYI-S-33 medium. The development within encysting Giardia of encystation-specific secretory vesicles (ESVs), which arise during differentiation and shuttle proteins (e.g. cyst wall proteins) and carbohydrates used for cyst wall assembly to the plasma membrane. They also developed the in vitro method of inducing encystation by addition of a high concentration of bile salts after growing in pre-encystation medium to initiate encystation (Gillin, 1987). A better understanding of the process of encystation of trophozoites may lead to the development of strategies to control the transmission of giardiasis.

von Allmen Nicole Eva - 15 -

4.7 Excystation

During the excystation process G. lamblia cysts are exposed to the high H+ concentrations of the host stomach, and then to the complex mixture of secretions released from the common bile duct into the duodenum (Eichinger, 2001). This process of excystation is a fundamental step in the infection. Giardia trophozoites escape from the cyst and establish infection in the upper small intestine of the host. An increase in oxygen uptake has been observed during excystation. The increase begins immediately, as soon as the excystation is induced, and continues exponentially during the 30 min thereafter. These changes are linked to an increase in the rate of the metabolism. It is unclear whether there is de-novo synthesis of the related enzymes during excystation. A study by Hetsko et al. (1998 using the mRNA differential display method, revealed evidence of complex transcriptional changes, some transcripts appeared, disappeared or changed in abundance, during the excystation process (Nino et al., 2003). Excysted trophozoites express new cell-surface variant proteins (variant surface proteins, VSPs) as a result of a change in VSP gene expression during the encystaton process (Svärd et al., 1998). The excystation protease is the product of one of three cysteine protease genes present in the Giardia genome. In biochemical assays of parasite extracts, cysteine proteases were previously identified as the major proteolytic activity of Giardia (Werries et al., 1991) Sequence analysis places the Giardia cysteine protease gene family as the earliest known branch of the cathepsin B lineage, a major eukaryotic protease family thought to have evolved and diversified in parallel with emergence of the first eukaryotic cells (Berti and Storer, 1995).

von Allmen Nicole Eva - 16 -

4.8 Characterization of the G. lamblia transcriptome

The Giardia genome project database provides online genome sequence information for G. lamblia (WB strain, clone C6). The G. lamblia genome contains a haploid size of approximately 12 Mb pairs distributed onto five chromosomes (Gillin and Sogin, 2000). The best characterized isolate is WB and the complete genome of WB clone C6 is currently being sequenced (Morrison et al., 2005; McArthur et al., 2005; www.mbl.edu/giardia). The genome study is a collaborative effort between different laboratories involved in research on giardiasis. To monitor genome-wide levels of mRNA expression throughout the in vitro reproduced life-cycle SAGE (serial analysis of gene expression) were used to examine giardial gene expression. SAGE technique is a quick and efficient method to detect 15 base nucleotide sequences from every mRNA transcript present in a sampled population of cells (McArthur et al., 2005; www.mbl.edu/giardia). Pulse-field analysis of the WB isolate (Le Blancq and Adam, 1998; Adam, 2000) revealed that it contains five different chromosomes of approximate size 1.6 Mb (chromosome 1 and 2), 2.3 Mb (chromosome 3), 3.0 Mb (chromosome 4) and 3.8 Mb (chromosome 5). The now available Giardia genome project database provides an online resource for G. lamblia (WB strain, clone C6) genome sequence information. Its analysis promises to provide insights about the origins of nuclear genome organization, the metabolic pathways used by parasitic protists, and the cellular biology of host interaction and avoidance of host immune systems. Since the divergence of G. lamblia lies close to the transition between eukaryotes and prokaryotes in universal ribosomal RNA phylogenies, the study of complete genomes such as the Giardia genome will increase our understanding of the early origin of eukaryotes (Morrison et al., 2005; www.mbl.edu/giardia). Previous comparisons of several different gene families have demonstrated Giardia' s basal position in molecular phylogenies. In evolutionary terms, the divergence of this organism is at least twice as ancient as the common ancestor for yeast and man.

von Allmen Nicole Eva - 17 -

4.9 Surface antigenic variation in G. lamblia

The surface of trophozoites contain one of a set of variant surface proteins (VSPs). The VSPs form a unique family of cysteine-rich proteins that are extremely heterogeneous in size. At least some of these surface proteins are sparsely glycosylated (Papanastasiou et al., 1997a) and possess a molecular weight between 22.3-kDa to over 200-kDa. The cysteine content is usually about 11-12% and most commonly found in CXXC motifs. All VSPs contain a well conserved hydrophobic tail of about 38 amino acids terminated by the invariant hydrophilic amino acids, CRGKA (Nash, 2002). The prediction says that the hydrophobic tail spans the outer parasitic membrane and the terminal CRGKA is in the cytosol (Mowatt et al., 1991). Another potentially important conserved element is a unique Zn finger motif (Nash and Mowatt, 1993; Zhang et al., 1993). Each parasite contains ~150 different VSP genes, totalling ~2% of the parasite genome. Surface labelling and electron microscopy demonstrated that VSPs are the major surface proteins of the parasite (Singer et al., 2001) and form a dense coat on its entire surface including flagella (Pimenta et al., 1991). Several initial investigations showed that the antigenic surfaces of G. lamblia are basically variable among different isolates (Nash and Kaiser, 1985; Ungar and Nash, 1987) and that surface antigenic variation may occur within an isolate (Nash et al., 1988; Adam et al., 1988; Aggarwal and Nash, 1988; Aggarwal et al., 1989). About 2.4% of a sampling of the Giardia genome is vsp genes (Smith et al., 1998). The genes encoding these proteins are highly conserved at the 3` terminus, but frequently demonstrate little similarity in the remainder of the coding region (Yang and Adam, 1995). Change in expression from one VSP to another occurs approximately once every six to 13 generations from a repertoire estimated at approximately 100-150 vsp genes (Nash et al., 1990). The VSPs undergo spontaneous switching in vitro and in vivo and in response to selection by and intestinal proteases (Nash, 1992; Meng et al., 1993). Therefore the in vivo aspects of the process of antigenic variation of G. lamblia and the host`s against the parasite has best been studied in experimentally infected mice. For these approaches the G. lamblia clone GS/M-83-H7 originating from a human isolate GS (Nash et al., 1985) was mostly used as a model parasite (Müller and Gottstein, 1998). This clone expresses a major surface VSP H7 which represents one of the few members of the G. lamblia VSP family already well characterized at the biochemical (Nash and Mowatt, 1993), immunological (Gottstein and Nash, 1991; Stäger and Müller, 1997; Stäger et la., 1998) and molecular (Nash and Mowatt, 1992; Bienz et al., 2001; von Allmen et al., 2004; von Allmen et al., 2005). The surface antigen VSP H7 has an apparent Mr of 56 832 (Nash and Mowatt, 1992) and appears to be antigenically distinct from most other VSPs identified so far (Nash, 1992). This major protein consisted of two functionally and biochemically distinct set of ; one set of variant-specific

von Allmen Nicole Eva - 18 - and conformational epitopes, most likely stabilised by disulfide bonds and located within a 314 amino acid N-terminal region; and the other set of non-conformational ones on a 171 amino acid peptide sequence located within the 208 amino acid C-terminal region of the protein (Stäger and Müller, 1997). The complex process of antigenic variation is an interesting and challenging experimental part in investigations about G. lamblia. In an overall view of the entire panel of in vivo- and in vitro-derived parasite populations, Bienz et al. (2001) has demonstrated that this clone expresses at least 29 different vsp gene sequences. In following experimental infections of neonatal mice we were recently able to show that VSP H7-type trophozoites were dominated the initial intestinal parasite population (von Allmen et al., 2004). This antigenic reset was irrespective of the variant-type composition of the cyst inoculum applied for infection of the animals (von Allmen et al., 2004). But these findings revealed the question if such an antigenic reset after transmission of the parasite via cysts was due to the growth-selective process of the establishment of the trophozoite population in the intestinal murine habitat. This could be if a few VSP H7-type positive trophozoites had resided within the inoculum and then have overgrown the other subvariant types. Another question is addressed to the direct association of the antigenic reset to the excystation process. Surface antigen alterations related to en- and / or excystation of G. lamblia is a known phenomenon that was particularly demonstrated for clone WB C6 (Gottstein et al., 1990; Svärd et al., 1998). For this clone (WB D6) it could be demonstrated that in vitro stage conversion triggered switching of vsp gene transcription. Conversely, such a switching process under analogous experimental conditions could not be shown for clone GS/M-83-H7 (Svärd et al., 1998). Intentionally repeated trials to follow the same excystation protocol used by Svärd et al. (1998), we failed because the excystation efficiency of non-VSP H7-type cultures was too low for reliable detection (von Allmen et al., 2004). However, our data of this simulation of a natural infection course in the mouse model related the antigenic reset mechanism via cysts of G. lamblia GS/M-83-H7 to an integral part of transmission from one to another murine host individuum. The VSP H7 protein may have evolved as a predominant variant which achieves an optimal initiation of the parasite infection and / or maintenance of the parasite within an affected host population. The biological role of antigenic variation among organisms differs and, in some cases, may be very complex. Immunological escape in which antibodies produced by the host against the dominant antigen destroy those organisms bearing this antigen and therefore resulting in replacement by the organisms that possessed a variant form of the antigen (Singer et al., 2001) is the most common biological role of altering the surface coat.

von Allmen Nicole Eva - 19 -

A further biological function of VSPs is indicated by the observation that at least some of these surface proteins possess a peptide conformation which resists treatment with intestinal proteases (Nash et al., 1991; Papanastasiou et al., 1996). The resistance against the enzymatic attack within the intestinal habitat may also be an effect of the selection for the growth of highly infectious variants during the antigenic process (Nash, 1992). Certainly these specific surface proteins play a major role in the infection course of the parasite, because the relatively amount of the organism` s resources are devoted to it.

Further infection studies in different animal models have to be established to investigate the role of the specific surface coat of the parasite. Particularly interesting are analyses of infection courses in potential experimental (e.g., gerbils) or natural hosts (e.g., dogs) with the GS/M-83-H7 parasite and to demonstrate if the VSP H7-type is also able, not only in mice, to induce a successful G. lamblia infection in these animals.

In vitro analyses of antigenic variation are a further required tool to get more information to the organisation of vsp genes in the G. lamblia genome. In a previous study Elmendorf et al. (2001) have shown by random sampling of cDNAs that ~20% of cDNAs in the libraries are sterile transcripts and they conclude that this antisense RNA production represents a striking molecular biological feature of G. lamblia. Our investigations on sense and antisense transcripts was performed in two different trophozoite populations, one expressing VSP H7 on their surface, the other was a subvariant trophozoite culture (von Allmen et al., 2005a). We found that sense vsp H7 RNA predominated in VSP H7-type trophozoites while sense RNA from only one (vsp IVg) of 8 subvariant vsp genes predominated in subviariant-type trophozoites. Both trophozoite populations exhibited a similar relative distribution regarding vsp H7 and vsp IVg antisense RNA molecules. From this study we could conclude that antisense RNA production in G. lamblia does not occur in a constitutive manner but is directly linked to complementary sense RNA production (von Allmen et al., 2005a). It will be important to discover whether these antisense transcripts are involved in gene regulation and what unique selective pressures have driven Giardia to adopt such molecular mechanisms.

von Allmen Nicole Eva - 20 -

4.10 Immunological host reactions against G. lamblia infections

In the past decade, the immune response against G. lamblia has been especially investigated in terms of the parasite's ability to continuously change its surface antigen coat (e.g. reviewed by Müller and Gottstein, 1998; Nash, 2002). Antigenic variation is thought to play an important role in facilitating chronic as well as repeated infections in natural hosts. However, conclusive experimental evidence proving this hypothesis is still not available. Until now, antigenic variation of G. lamblia and the consequences of this phenomenon on the course of infection have nearly exclusively been studied in experimental hosts. In this respect, experimental G. lamblia clone GS/M-83-H7 infections in the mouse system is a successful model for investigating the different parameters associated with surface antigen alterations of the parasite (e.g. reviewed by Müller and Gottstein, 1998; Nash, 2002). Variant surface protein H7 from clone GS/M-83-H7, like various other VSPs investigated so far, induces strong serum (Gottstein et al., 1990; Müller et al., 1996; Stäger et al., 1997a; Stäger et al., 1997b), intestinal (Stäger et al., 1997b; Stäger and Müller, 1997; Müller and Stäger, 1999), and milk (Stäger et al., 1998; Müller and Stäger, 1999) antibody responses in G. lamblia-infected individuals. In the experimental murine host, the humoral immune response is predominantly directed against VSP, but in human patients antibodies against invariant antigens are also produced during a G. lamblia infection (Edson et al., 1986; Hassan et al., 2002; Palm et al., 2003; Weiland et al., 2003). As the establishment of a humoral immune response in both experimental and natural hosts coincides with the elimination of the original variant-type population, a function for antibodies in the process of antigenic variation of G. lamblia was proposed (e.g. reviewed by Müller and Gottstein, 1998). Various studies suggested a central role of the in determining the outcome of a Giardia infection but the knowledge on those mechanisms mediating immunity is still rather rudimentary (e.g. reviewed by Faubert, 2000; Eckmann, 2003; Müller and von Allmen, 2005). In the past, intense investigations on the immunobiology of giardiasis have been performed by using the Giardia muris/mouse model system. Since the results from these immunological studies have been major subjects of two recent reviews (Faubert, 2000; Eckmann, 2003) the present article focuses on experimental data generated in the G. lamblia/mouse model. In giardiasis, reinfections are common because acquired immunity against G. lamblia is not complete either due to insufficient immune defences or antigenic variation of the parasite. Many studies in natural and experimental rodent hosts addressed the question whether antibodies, and more specifically, local secretory immunoglobulin (Ig) A antibodies play a role in control of the parasite infection. Studies in patients with selective IgA-deficiency provided conflicting results in that some

von Allmen Nicole Eva - 21 - investigations demonstrated increased incidences of infections while others did not (Eckmann, 2003). Previous findings in the experimental mouse/G. lamblia GS/M-83-H7 model indicated that B-cell-deficient animals were unable to clear giardial infections (Stäger and Müller, 1997; Langford et al., 2002). Conversely, by comparing the courses of infection in IL-6-deficient and wild-type mice, no obvious correlation between intestinal IgA production and decrease of parasite load was observed at least during the acute infective stage (Bienz et al., 2003; Zhou et al., 2003). Data from an investigation based on the use of different transgenic mouse strains indicated that an as yet unknown T-cell-dependent mechanism is essential for controlling the acute phase of a G. lamblia infection (Singer and Nash, 2000a). In this study, both wild-type and B-cell-deficient mice eliminated the majority of intestinal parasites within 1 to 2 weeks. While T-cell depletion of wild-type mice with anti-CD4 antibodies prevented elimination of G. lamblia, interferon (IFN) γ-, interleukin (IL)-4-, IL4Rα-, and STAT-6-deficient mice were able to control infection in a manner similar to wild-type mice. From this observation, the authors concluded that (i) a T-cell-dependent mechanism is essential for controlling the acute phase of a G. lamblia infection and (ii) in the mouse model neither the Th1 nor the Th2 subset is absolutely essential for protection from G. lamblia. Several investigations in the past generated evidence that mast cells are substantially involved in intestinal elimination of G. lamblia (Mitchell et al., 1982; Erlich et al., 1983; Li et al., 2004). For example, Li et al. (2004) recently found that mast cell-deficient, or – depleted, C57BL/6 mice failed to control a G. lamblia clone GS/M-83-H7 infection. In this report, mast cells were considered to be a potential source for IL-6. Since in the abovementioned (Li et al., 2004) and other studies (Bienz et al., 2003; Zhou et al., 2003) murine IL-6-deficiency was associated with an increased susceptibility to a G. lamblia clone GS/M-83-H7 infection mast cell-derived IL-6 was suggested to be important for control of such an infection. Interestingly, the acute phase of giardiasis was found to be associated with a delayed recruitment of intestinal mast cells (Hardin et al., 1997). It is feasible that this phenomenon reflects a survival strategy of G. lamblia which retards the mast cell-dependent antigiardial effector mechanism during the initial phase of the parasite infection. Although the investigations listed above could not completely dissect the cellular network involved in antigiardial immune defence it is evident that antibody-independent immune effector mechanisms interfere directly, or indirectly, with maintenance and growth of the intestinal parasite population particularly during the acute phase of the infection (Singer and Nash, 2000a; Langford et al., 2002; Li et al., 2004). However, these results are inconsistent with findings in human patients suffering from acquired T-cell deficiencies (Janoff et al., 1988). Compared to the immunocompetent control group, the patient group

von Allmen Nicole Eva - 22 - was not found to be more susceptible to giardiasis. In contrast, there are indications that protection against giardial infections in humans may be associated with anti-Giardia antibody production. Such a correlation was found by investigating children with x-linked agammaglobulinaemia which exhibited a predisposition for severe and chronic giardiasis (LoGalbo et al., 1982). Considering these controversial findings new experimental strategies will have to be developed to further elucidate those effector mechanisms which mediate immunity against G. lamblia infections in natural and experimental hosts of the parasite.

4.11 Physiological host reaction against G. lamblia infections

While the immunological processes of the antigiardial host response has already been intensely investigated only little is known about non-immune defences (e.g. reviewed by Müller and von Allmen, 2005). For example, intestinal Paneth cell-derived defensins (Ouellette, 1999), also known as cryptdins, have been proven to display a cytotoxic effect on G. lamblia in vitro (Aley et al., 1994) but no conclusive data are available regarding the physiological significance of these antimicrobial peptides in vivo. G. lamblia seems not to be able to induce release of defensins in ex vivo-maintained small intestinal crypts (Ayabe et al., 2000). This finding questioned the relevance of defensins in the host defence directed against the parasite. However, a recent study provided preliminary evidence for the existence of an indirect antigiardial effect of these bioactive peptides (Eckmann, 2003). This was achieved by analysing the course of a G. muris infection in matrix metalloproteinase (MMP)-7-deficient mice which are incapable of producing Paneth cell- derived α-defensins. Interestingly, these mice exhibited a significantly lower initial infection load than normal control mice. According to the authors this finding might have indicated that the lack of α-defensins in MMP-7-deficient mice influenced the abundance and/or composition of the intestinal microbiota and thus generated a physiological environment which inhibited giardial colonization. In this context, another study focusing on the impact of the enteric flora on the possible parasitocidal function of Paneth cells in Giardia-infected mice also provided interesting results (El-Shewy and Eid, 2005). Here, transmission electron mircoscopical examinations of intestinal specimens from infected animals revealed that some of the trophozoites harboured bacterial endosymbionts and it seemed that only those parasites containing peripheral bacterial inclusions were killed in close proximity of activated Paneth cells. This finding indicated that Paneth cells, and more specifically Paneth cell-derived defensins, may interfere in intestinal growth of Giardia trophozoites (see also below). However, it certainly remains to be elucidated if the

von Allmen Nicole Eva - 23 - observed phenomenon was indicative for a causative link between bacterial invasion and subsequent Paneth cell-mediated killing of trophozoites, or rather reflected biologically irrelevant intracellular accumulation of bacteria in dead trophozoites. A further investigation addressed the putative function of epithelial nitric oxide (NO) as a antigiardial effector. Nitric oxide is produced enzymatically and in intestinal epithelial cells the most important pathway mediating this enzymatic reaction involves the inducible NO synthase (iNOS) (Salzmann et al., 1996). The existence of such an antigiardial effector mechanism was assumed because NO was revealed to inhibit in vitro growth of the parasite (Fernandes and Assreuy, 1997; Eckmann et al., 2000). However, the effectiveness of NO against a giardial infection was questioned by the observation that co- cultivation of G. lamblia trophozoites with human epithelial cells led to a remarkable suppression of the epithelial NO production. Eckmann et al. (2000) found that this suppression was the consequence of a depletion of arginine (a substrate for cellular NO synthesis) in the culture medium which was caused by a high affinity uptake of the compound by the parasite. According to the authors, it is feasible that this competitive effect represents a survival strategy which enables G. lamblia to counteract antiparasitic NO production within the intestinal habitat of the parasite. Besides epithelial cells, IFNγ- activated act as a source for NO. In experimental murine giardiasis, IFNγ seemed to contribute to the relative resistance of B10 mice against the parasite infection but according to the data generated by Venkatesan et al. (1997) NO production had not contribute to this relative resistance. This observation may also confirm the above assumption that Giardia has developed a strategy to inactivate the NO-mediated antigiardial attack of the host. However, the scenario outlined above is challenged by a study on G. lamblia infections in IFNγ-deficient mice (Singer and Nash, 2000a). Since these mice were able to control infection in a manner similar to wild-type mice (see also above) a substantial participation of IFNγ in the antigiardial immune effector mechanisms was excluded. Here, it certainly remains to be clarified whether these controversial findings reflected differences in the immunobiology of the G. muris and G. lamblia species or rather were the consequence of a differential set-up of the respective infection experiments. In the field of giardiasis, intestinal mucin secretion is also discussed as a putative physiological defence mechanism against the parasite. Mucins are glycoproteins that are secreted from intestinal goblet cells and they constitute the intestinal mucus layer (Deplancke and Gaskins, 2001). The presence of mucins was found to reduce adhesion of to mucosal surfaces and thus may be of particular importance in early life stages because the acquired immune system is not fully functional in the neonatal intestine (Cebra, 1999). Mucin secretion is known to be rapidly enhanced in response to

von Allmen Nicole Eva - 24 - enteric microorganisms (Moncada et al., 2003). Trapping of microorganisms by mucus and their subsequent removal from the intestine by peristalsis is thought to influence the intensity of a respective infection (Walker and Owen, 1990). In giardiasis, preliminary data in the gerbil infection model indicated that increased mucus secretion reduces epithelial attachment of G. lamblia trophozoites and lowers the infection intensity (Leitch et al., 1989). A participation of such a mechanism in anti-giardial defence was also suggested because commercially available mucins turned out to inhibit attachment of G. lamblia trophozoites to an artificial surface (Roskens and Erlandsen, 2002). However, in another study, an increased attachment was observed when G. lamblia trophozoites were exposed to a mucin fraction prepared from postmortem samples of human intestine (Zenian and Gillin, 1985). Furthermore, human intestinal mucus was shown to stimulate growth of the parasite in vitro (Gault et al., 1987). Based on the conflicting results from these in vitro studies it is not possible to decide whether mucins are physiological factors involved in antigiardial host defence or rather contribute to the successful colonisation of the parasite on the mucosal surface of its intestinal habitat. These two possibilities will have to be evaluated by using appropriate infection models which allow an eventual correlation between mucin secretion and the course of a G. lamblia infection in the respective experimental host.

4.12 Intestinal pathogenesis associated with G. lamblia infections

As outlined above, the most important clinical signs of giardiasis are diarrhoea and malabsorption. Although various intestinal abnormalities have been described, the pathophysiology associated with these symptoms is still incompletely understood (e.g. reviewed by Farthing, 1996; Farthing, 1997; Eckmann and Gillin 2001; Müller and von Allmen, 2005). In giardiasis, intestinal colonisation by the parasite seems to cause microvillus shortening (Erlandsen and Chase, 1974; Scott et al., 2000; Scott et al., 2004), villous flattening or atrophy (Williamson et al., 2000). These abnormalities possibly in combination with other pathological mechanisms such as reduction of intestinal disaccharidase (Daniels and Belosevic, 1992) and protease (Seow et al., 1993) activities may be a direct cause of diarrhoea in giardiasis. In experimental Giardia muris infections, inhibition of the intestinal disaccharidase (sucrase and maltase) activities during the acute infective stage, and loss of intestinal brush border surface area could be related to the function of infiltrating CD8+ T cells (Scott et al., 2004) and a possible involvement of several others than IL-6 has been reported (Scott et al., 2000). However, the particular host mechanisms mediating the T

von Allmen Nicole Eva - 25 - cell effects on the microvillus structure still remain to be investigated. Respective investigations in G. lamblia infections were for a long time hampered by the fact that no experimental animal model was available which developed pathological effects in response to the parasite infection. However, a few years ago Williamson et al. (2000) were able to isolate a bird G. lamblia strain (BRIS/95/HEPU/2041) which caused significant pathophysiological alterations to intestinal mucosa including villus atrophy, and an increase of goblet cell and vacuolated epithelial cell populations. In addition to host immunological components, parasite-derived factors also contribute to pathogenesis in giardiasis and investigations demonstrating this participation were often performed in vitro. For many years, co-incubation of intestinal pathogens with epithelial cell lines have been successfully used as in vitro models which helped to improve our knowledge on the mechanisms involved in pathogen-induced brush border damage and malabsoption. Several human and murine cell lines have been established and these exhibit characteristics of the normal intestinal epithelium such as polarisation, tight junction formation, ion transport and regulated synthesis of inflammatory mediators. Most of these cell lines are of colonic origin and thus they cannot completely simulate the epithelial surface of the duodenum which is the major habitat of G. lamblia trophozoites. However, duodenal and colonic epithelial cells are very similar from the physiological point of view. Considering this aspect, co-cultivation of G. lamblia trophozoites with such colonic epithelial cell lines was often chosen as an approach to explore various molecular and cellular aspects of the process of parasite-host cell interaction (see below). Attachment of trophozoites to epithelial cells is essential for giardial colonisation of the intestine and most likely also results in damage of the intestinal epithelium. In various studies it was demonstrated that surface lectins are involved in attachment of the parasite to epithelial cells (Pegado and de Souza, 1994; Katelaris et al., 1995; Céu Sousa et al., 2001). In mixed cultures, Giardia cells were found in various orientations to epithelial cells but were mostly observed ventral surface down. This indicated (i) that surface components may be involved in a primary and randomly oriented, giardial attachment but also (ii) that mechanical, equally oriented adhesion through the ventral disk plays the major role in attachment of the parasite to the mucosal surface of the duodenal site. As demonstrated by Buret et al. (2002) attached G. lamblia trophozoites were able to increase the permeability of non-transformed human epithelial cell layers by disrupting tight junctions. Interestingly, these effects were reversed by incubation of the mixed culture with epidermal growth factor (EGF) which reduced epithelial colonisation by the parasites. From these observations, the authors concluded that (i) altered epithelial permeability may contribute to intestinal pathophysiology in giardiasis and (ii) EGF may inhibit these alterations by preventing intestinal colonisation by Giardia, and/or by directly

von Allmen Nicole Eva - 26 - inhibiting its cytopathic effects. In the same co-cultivation system, Giardia-induced epithelial permeabilisation was demonstrated to be associated with distinct rearrangements in the cytoskeleton of the epithelial cells (Teoh et al., 2000). In addition to that increased epithelial permeability turned out to be linked to an apoptotic effect which disrupted the epithelial barrier function in a caspase-3-dependent manner (Chin et al., 2002). In the respective paper, the authors discussed the possibility that cytoskeletal disintegration upon exposure of epithelial cells to G. lamblia is the direct consequence of a caspase-mediated cleavage of cytoskeletal proteins. Interestingly, apoptosis was only observed with two of four G. lamblia strains included in the investigations. This finding suggested further studies aimed at the evaluation of a possible correlation of epithelial apoptosis in vitro and increased giardial pathogenicity in vivo. Mucosal inflammation may also be a possible factor of pathogenesis in giardiasis although many histories of natural symptomatic infections are apparently not associated with substantial intestinal inflammatory responses (Farthing, 1996; Eckmann and Gillin, 2001). For example, in experimentally infected goat kits, inflammatory infiltration in the lamina propria coincided with villus-shortening and crypt hyperplasia (Koudela and Vitovec, 1998). Furthermore, human symptomatic giardiasis seems to result occasionally (3-4% of patients with Giardia-positive biopsies) in formation of moderate polymorphonuclear and mononuclear cell (Oberhuber et al., 1997). In the mouse model, intestinal mucosal inflammation was particularly evaluated regarding its importance in resistance to giardiasis. By co-infecting mice with G. muris and the nematode Trichinella spiralis, a significant reduction of the G. muris infection was observed (Roberts-Thompson et al., 1976). This protective effect may have been causatively linked to the mucosal inflammatory response during the intestinal phase of trichinellosis. In contrast, a study on G. muris infections in a resistant versus susceptible mouse model did not evidence a correlation between resistance against infection and mucosal inflammation (Venkatesan et al. 1997). This finding is consistent with the results from our recent G. lamblia clone GS/M-83-H7/T. spiral co-infections in mice (von Allmen et al., 2005b). Here, intestinal inflammation initiated by a T. spiralis infection did not exhibit antigiardial activity but substantially promoted proliferation of duodenal G. lamblia trophozoites. The enteric flora is a further factor that can potentially interfere with the process of a G. lamblia infection. Co-cultivation of G. lamblia and human epithelial cells in presence of commensally lactobacilli constituted a growth environment which resulted in significant inhibition of giardial proliferation (Perez et al., 2001). This growth-inhibitory effect was due to (an) as yet unidentified factor(s) that was/were released by the bacteria included in the culture. Depending on the composition of the enteric flora in Giardia-infected individuals, such bacterial compounds may also have an influence on in vivo growth of the parasite.

von Allmen Nicole Eva - 27 -

As reported by Singer and Nash (2000b), the intensities of G. lamblia infections differed significantly within two isogenic mouse strains that originated from different commercial breeders. When these mice were housed together, resistance to infection was transferred to normally susceptible mice. By treatment with antibiotics normally resistant mice were “converted” into susceptible mice. This observation clearly demonstrated that the interaction of G. lamblia with the enteric flora represents a physiological factor which may modulate the course of the infection. This conclusion was further supported by data from a recent study which assessed the interaction between the bacterial flora and Giardia trophozoites in the context of the intestinal physiology (El-Shewy and Eid, 2005). Respective experimentation provided at least preliminary evidence that intestinal Giardia trophozoites were able to incorporate bacterial endosymbionts. and trophozoites harbouring peripheral bacteria apparently stimulated degranulation of intestinal Paneth cells. It seemed that this degranulation process resulted in a release of lytic peptides such as defensins and thus displayed a lytic effect on trophozoites in close proximity to the activated Paneth cells (see also above). Accordingly, the enteric flora may be indirectly involved in those physiological reactions that interfere with intestinal proliferation of the parasite. The intestinal co-habitation of Giardia and microbiota has also to be evaluated from the pathophysiological point of view. Very early studies on giardiasis indicated that in some human patients clinical manifestations of the disease were associated with the presence of increased numbers of aerobic and/or anaerobic bacteria in the upper part of the intestine (Tandon et al. 1977; Tomkins et al., 1978). It is feasible that in certain cases the pathological effects are not caused by the Giardia infection per se but are the consequence of an overgrowth of commensal bacteria. It was recently reported that under normal steady-state conditions commensal bacteria initiate a Toll-like receptor signal transduction pathway in epithelial cells which is crucial for the maintenance of intestinal homeostasis (Rakoff-Nahoum et al., 2004). However, under conditions of a bacterial overgrowth related to giardiasis, it cannot be excluded that these commensals trigger a pro-inflammatory response similar to that e.g. observed in patients with inflammatory bowel disease.

von Allmen Nicole Eva - 28 -

5 Aim of the present thesis

1.) In the first part of the presented study, we were interested in assessing the vsp gene expression profile associated with the surface antigen alterations of G. lamblia clone GS/M-83-H7 during transmission of the parasite from one to another host individual. Furthermore, we intended to find out if this antigenic diversification process follows a specific strategy which may contribute to the survival of the parasite within the experimental murine host. These questions were tackled by infecting neonatal mice with antigenically non-diversified (VSP H7-positive) and diversified (VSP H7-negative) cysts and subsequently testing by RT-PCR the emerging intestinal trophozoite populations regarding their vsp gene expression patterns. 2.) The genetic mechanisms controlling antigen switching are, as yet, only poorly understood. It is discussed that the RNAi mechanism controls expression of the vsp genes and that RNA-dependent RNA polymerase is involved in restricting expression of the vsp gene repertoire to a single gene at one time (unpublished data from Lujan, reviewed by Ullu et al. 2004). Elmendorf et al. (2001) found that G. lamblia trophozoites generate a substantial amount of antisense RNA molecules which correspond to approximately 20% of the total RNA content. These interesting observations have driven the next experimentation part where we decided to establish an RT-PCR-based quantitative assessment of sense and antisense transcripts from genes involved in both antigenic variation (vsp genes) and cyst formation (cwp 1 gene) of Giardia lamblia clone GS/M-83-H7. These investigations were focused on the possibility of a bidirectional vsp gene transcription and respective data were expected to provide an initial idea about the relevance of the RNA interference mechanism in gene regulation of G. lamblia. 3.) The particular processes participating in the interplay between G. lamblia and its host immune response and determining the outcome of the parasite infection are still incompletely understood. Recent data (Williamson et al., 2000; Jimenez et al., 2004; Li et al., 2004) indicated that a high priority will have to be given to those investigations which explore the influence of intestinal inflammatory reactions on the course of G. lamblia infections. The process of inflammation has been found to influence not only the immunological, but also the physiological, environment inside the intestinal habitat of the parasite. These previous findings were basis for the third experimentation part of my thesis project aimed at an investigation of giardial growth in an inflammatory intestinal environment of a murine host. In our approach, we decided to initiate inflammation by infecting mice with the nematode Trichinella spiralis. By choosing this experimental strategy, we were also able to address the question whether a (pre-) existing nematode

von Allmen Nicole Eva - 29 - infection is a factor which may contribute to increased susceptibility of a host to a G. lamblia infection.

von Allmen Nicole Eva - 30 -

6 Summary of publications

I) N. von Allmen, M. Bienz, A. Hemphill and N. Müller. 2004. Experimental infections of neonatal mice with cysts of Giardia lamblia clone GS/M-83-H7 are associated with an antigenic reset of the parasite. Infection and Immunity 72: 4763-4771.

Infection with the parasite Giardia lamblia occurs with peroral ingestion of cysts which release two trophozoites each in the small intestine. The successful establishment of the parasite in the host environment is still unknown and to investigate the parasite infection strategy in more detail the present study focused on the major surface antigen, VSP. The variant surface antigen is responsible for the antigenic variation of G. lamblia and is thought to be a mechanism of the parasite to escape the host immune responses. Our interest lied on the determination of the antigenic switch process on the molecular level during an infection course. For this purpose, the mother-offspring mouse model was used for these experimental infections and neonatal mice were infected with VSP H7-type trophozoites collected from the intestines of individual animals at different time points post infection. These G. lamblia clone GS/M-83-H7 trophozoites, as well as cysts inoculations were, without additional in vitro cultivation, previously analyzed for their vsp gene transcription patterns. With a quantitative 3` RACE reverse transcription-PCR (RT-PCR) variant-specific regions of different GS/M-83-H7 vsp sequences of intestinal trophozoite populations at different time points were analyzed regarding their content of vsp transcripts. In this study we were able to demonstrate that vsp H7 mRNA was predominant in VSP H7-type parasites. The initial antigen switch of clone GS/M-83-H7 showed a obvious reduction of vsp H7 mRNA levels but were not associated with a significant increase of mRNA levels of any other subvariant VSP-type. In one experimental part, we simulated the natural transmission of G. lamblia by intragastric inoculation of in vivo-derived cysts. This infection mode revealed a VSP H7-type trophozoite majority in the initial intestinal parasite population irrespective of the variant- type composition of the cyst inoculum. This reset to the major VSP H7 of the GS/M-83-H7 clone is probably a consequence of the excystation process rather than a selective process which favors colonization of a small number of VSP H7 type cysts in the inoculum.

von Allmen Nicole Eva - 31 -

II) N. von Allmen, M. Bienz, A. Hemphill and N. Müller. 2005a. Quantitative assessment of sense and antisense transcripts from genes involved in antigenic variation (vsp genes) and encystation (cwp 1 gene) of Giardia lamblia clone GS/M-83-H7. Parasitology 130: 389-396.

In a previous study it has been shown by random sampling of cDNAs of two Giardia strains that approximately 20% of cDNAs in the libraries represent polyadenylated sterile transcripts. This observation is interesting regarding the organization of antigenic variation on the molecular level. To investigate giardial vsp gene transcription, VSP H7-type and subvariant trophozoites vsp RNA levels were assessed by quantitative reverse transcription-(RT)-PCR. Antisense RNA production in Giardia is highly diversified and involves a large part of different genes as well as vsp genes. These transcripts are polyadenylated and consist of sterile RNA molecules that are not used in the translation process. We focused on the evaluation of a possible bidirectional vsp gene transcription in in vitro cultivated trophozoites and cysts, but not on possible gene-regulatory function of these antisense transcripts. VSP H7-type and subvariant-type GS/M-83-H7 trophozoite cultures were generated and subsequently analyzed by quantitative RT-PCR for their relative production of sense and antisense vsp H7 or subvariant vsp RNA. Investigations of the encystation gene cwp 1 (encoding cyst wall protein 1) revealed analogous sense versus antisense RNA pattern. The encystation gene cwp 1 is only transcribed and translated when the encystation process is induced by specific cultivation conditions. Our experimentations demonstrated that in vitro antigenic variation of G. lamblia is associated with alterations in sense and antisense vsp RNA production. But these alterations in sense and complementary antisense vsp RNA of VSP H7-type and subvariant-type trophozoites are not characterized by a reciprocal manner. This observation is contrary to the assumption of different research groups that G. lamblia dispose of RNA interference (RNAi) mechanism. Concluding clearness would bring the determination of the relative vsp H7 versus vsp IVg RNA composition on the level of the extremely fragmented and hardly detectable siRNA sequences. Perhaps, the detected antisense RNA production in G. lamblia is only a feature of the simplicity of this organism which cannot control unspecific antisense gene transcription eventually occurring as artifact of DNA unwinding. Further studies have to be made to answer the question if this antisense RNA production in G. lamblia is biologically relevant and, if this is the case, in which extend for the parasite.

von Allmen Nicole Eva - 32 -

III) N. von Allmen, S. Christen, U. Forster, B. Gottstein, M. Welle and N. Müller. 2005b. Acute trichinellosis increases susceptibility of Giardia lamblia infections in the mouse model. Submitted to Infection and Immunity.

As previously described, Giardia lamblia infections may cause diarrhea in humans and animals. To generate novel information to intestinal immune responses in the murine host we applied a double-infection model with the nematode Trichinella spiralis and Giardia lamblia. Our experimentations were in analogy to a former T. spiralis/Giardia muris co- infection study which demonstrated a suppression of G. muris trophozoite proliferation consequently to the preceding T. spiralis infection. Acute trichinellosis causes a transient intestinal inflammation which is associated with a massive mucosal mast cell infiltration. As recently reported intestinal mast cell infiltration seemed to involve resolution of the G. lamblia infection. The aim of this study was to investigate if intestinal mast cells stimulated during a concurrent T. spiralis infection are also able to modulate the infection course of G. lamblia and which effect a former infection may have on the outcome of a preceding Giardia infection course. In our project, mice pre-infected with T. spiralis demonstrated a transiently increased susceptibility to a G. lamblia infection an observation which is contrary to the previous study on the T. spiralis/G. muris double infection. These different observations may be due to the fact that two distinguishing Giardia species, G. muris and G. lamblia, were used for the infection mode but still not all reasons for this differentiation are answered. The early phase of T. spiralis infection was associated with a massive mucosal and mast cell infiltration, mast cell degranulation and IgE production. Our co-infection experiments revealed intestinal accumulation of mast cells for a much longer time than in the T. spiralis infection alone and this observation correlated with a transient increase of the intestinal G. lamblia trophozoite burden. Conclusively, our evaluated data indicated that a T. spiralis infection, perhaps in particular the intestinal inflammation, favors colonization and maintenance of a G. lamblia population in mice. The phenomenon of a pre-existing parasitic, bacterial and/or viral infection prior to a G. lamblia infection is an obvious indication in endemic areas of giardiasis and may therefore have a direct influence on the epidemiological relevance of the disease.

von Allmen Nicole Eva - 33 -

IV) N. Müller, N. von Allmen. 2005. Recent insights into the mucosal reactions associated with Giardia lamblia infections. International Journal of Parasitology, in press (Review).

The subjects concerning Giardia-host interactions and the effects on the host immune responses are a very busy discussed field in the Giardia research. Parts in this field are already proven and well-known facts like the importance of intestinal antibodies in the process of antigenic variation whereas their effect in the immune response to clear a Giardia infection is still controversial. Another unproven observation is the potential of inflammatory mast cells to directly or indirectly interfere in duodenal growth of G. lamblia trophozoites. Some reports do not underline this evidence of intestinal inflammation and resistance to infection. Moreover, it has been shown in in vitro co- cultivation experiments with G. lamblia trophozoites and epithelial cell lines that intestinal infiltration of inflammatory cells may have an effect on the pathogenesis of giardiasis. Abnormalities like loss of intestinal brush border surface area, villus flattening inhibition of disaccharidase activities seem to have influence on the course of the disease. Immunological processes of the antigiardial host defense have already been intensively analyzed while non-immune effects are still under investigations. Studies focusing on antiparasitic NO production have demonstrated that these defense mechanisms do not contribute to a relative resistance; even if the parasite has developed an effective strategy to inactivate the NO-mediated antigiardial attack of the host. Another example in this subject is the putative physiological role of intestinal mucus secretion against parasite colonization. The presence of mucins was found to reduce adhesion of pathogens to mucosal surfaces and thus may be of particular importance in early life stages because the acquired immune system is not yet fully functional in the neonatal intestine. In another study, an increased attachment of G. lamblia trophozoite was observed when the parasites were exposed to a mucin fraction prepared from postmortem samples of human intestine. These conflicting results must be further evaluated and the role of intestinal mucus has to be determined. The combination of novel information of the Giardia genome, host-parasite interactions like Giardia-induced intestinal abnormalities, host cellular immune mechanisms as well as new insights into the immunological and physiological interplay between the parasite and its host intestinal environment will once help to understand how the outcome of a Giardia infection is established.

von Allmen Nicole Eva - 34 -

7 References

Adam, R.D., 1991. The biology of Giardia spp. Microbiology Review 55: 706-732.

Adam, R.D., 2000. The Giardia lamblia genome. International Journal of Parasitology 30: 475-484.

Adam, R.D., 2001. Biology of Giardia lamblia. Clinical Micorbiology Reviews 14: 447-475.

Adam, R.D., Aggarwal, A., Lal, A.A., De La Cruz, V.F., Cutchan, T., Nash, T.E., 1988. Antigenic variation of cysteine-rich protein in Giardia lamblia. Journal of Experimental Medicine 167: 109-118.

Aggarwal, A. and Nash, T.E., 1988. Antigenic variation of Giardia lamblia in vivo. Infection and 56: 1420-1423.

Aggarwal, A., Meritt , J.W.Jr., Nash, T.E., 1989. Cysteine-rich variant surface proteins of Giardia lamblia. Molecular and Biochemical Parasitology 32: 39-48.

Aley, S.B., Zimmerman, M., Hetsko, M., Selsted, M.E., Gillin, F.D., 1994. Killing of Giardia lamblia by cryptdins and cationic peptides. Infection and Immunity 62: 5397- 5403.

Ayabe, T., Satchell, D.P., Wilson, C.L., Parks, W.C., Selsted, M.E., Ouellette, A.J., 2000. Secretion of microbicidal alpha-defensins by intestinal Paneth cells in response to bacteria. Nature Immunology 1: 113-118.

Bernander, R., Palm, J.E., Svärd S.G., 2001. Genome ploidy in different stages of the Giardia lamblia life cycle. Cellular Microbiology 3: 55-62.

Berti, P.J. and Storer, A.C., 1995. Alignment/phylogeny of the papain superfamily of cysteine proteases. Journal of Molecular Biology 246: 273-283.

Bienz, M., Siles-Lucas, M., Wittwer, P., Müller, N., 2001. vsp gene expression by Giardia lamblia clone GS/M-83-H7 during antigenic variation in vivo and in vitro. Infection and Immunity 69: 5278-5285.

Bienz, M., Dai, W.J., Welle, M., Gottstein, B., Müller, N., 2003. Interleukin-6-deficient mice are highly susceptible to Giardia lamblia infections but exhibit normal intestinal IgA responses against the parasite. Infection and Immunity 71: 1569-1573.

Buret, A.G., Mitchell, K., Muench, D.G, Scott, K.G., 2002. Giardia lamblia disrupts tight junctional ZO-1 and increases permeability in non-transformed human small intestinal epithelial monolayers: effects of epidermal growth factor. Parasitology 125: 11-19.

Cebra, J.J., 1999. Influences of microbiota on intestinal immune system development. American Journal of Clinical Nutrition 69: 1046-1051.

Centre for food safety and applied nutrition. Foodborne pathogenic microorganisms and natural toxins handbook. U.S. Food & Drug Administration. Giardia lamblia. www.cfsan.fda.gov/~mow/chap22.

Céu Sousa, M.C., Goncalves, C.A., Bairos, V.A., Poiares-Da-Silva, J., 2001. Adherence of Giardia lamblia trophozoites to Int-407 human intestinal cells. Clinical Diagnosis by Laboratory Immunity 8: 258-265.

von Allmen Nicole Eva - 35 -

Chavez-Munguia, B., Cedillo-Rivera, R., Martinez-Palomo, A., 2004. The ultrastructure of the cyst wall of Giardia lamblia. Eukaryotic Microbiology 51: 220-226.

Chin, A.C., Teoh, D.A., Scott, K.G., Meddings, J.B., Macnaughton, W.K., Buret, A.G., 2002. Strain-dependent induction of enterocyte apoptosis by Giardia lamblia disrupts epithelial barrier function in a caspase-3-dependent manner. Infection and Immunity 70: 3673-3680.

Daniels, C.W.and Belosevic, M., 1992. Disaccharidase activity in the small intestine of susceptible and resistant mice after primary and challenge infections with Giardia muris. American Journal of Tropical Medicine and Hygiene 46: 382-390.

Deplancke, B. and Gaskins, H.R., 2001. Microbial modulation of innate defense: goblet cells and the intestinal mucus layer. American Journal of Clinical Nutrition 73: 1131-1141.

Dönges, 1988. Journal of Parasitology, 2 Auflage- Georg Thieme Verlag Stuttgart: 76-77.

Eckmann, L., Laurent, F., Langford, T.D., Hetsko, M.L., Smith, J.R., Kagnoff, M.F., Gillin, F.D., 2000. Nitric oxide production by human intestinal epithelial cells and competition for arginine as potential determinants of host defense against the lumen-dwelling pathogen Giardia lamblia. Journal of Immunology 164: 1478-1487.

Eckmann, L. and Gillin, F.D., 2001. Microbes and microbial toxins: Paradigms for microbial- mucosal interactions I. Pathophysiological aspects of enteric infections with the lumen-dwelling protozoan pathogen Giardia lamblia. American Journal of Physiological, Gastrointestinal and Liver Physiology 280: G1-G6.

Eckmann, L., 2003. Mucosal defences against Giardia. Parasite Immunology 25: 259-270.

Edson, C.M., Farthing, M.J.G., Thorley-Lawson, D.A., Keusch, G.T., 1986. An 88,000-Mr Giardia lamblia surface protein which is immunogenic in humans. Infection and Immunity 54: 621-625.

Eichinger, D., 2001. Encystation in parasite . Current Opinions in Microbiology 4: 421-426.

Elmendorf, H.G., Singer, S.M., Nash, T.E., 2001. The abundance of sterile transcripts in Giardia lamblia. Nucleic Acids Research 29: 4674-4683.

El-Shewy, K.A, and Eid, R.A., 2005. In vivo killing of Giardia trophozoites harbouring bacterial endosymbionts by intestinal Paneth cells: an ultrastructural study. Parasitology 130: 269-274.

Erlandsen, S.L. and Chase, D.G., 1974. Morphological alterations in the microvillous border of villous epithelial cells produced by intestinal microorganisms. American Journal of Clinical Nutrition 27: 1277-1286.

Erlich, J.H., Anders, R.F., Roberts-Thomson, I.C., Schrader, J.W., Mitchell, G.F., 1983. An examination of differences in serum antibody specificities and reactions as contributing factors to chronic infection with the intestinal protozoan parasite, Giardia muris, in mice. Australian Journal of Experimental Biological and Medical Science 61: 599-615.

Farthing, M.J.G., 1993. Diarrhoeal disease: current concepts and future challenges. Transaction of the Royal Society of Tropical Medical Hygiene 87: 17-21.

Farthing, M.J., 1996. Giardiasis. Gastroenterology Clinics of North America 25: 493-515.

von Allmen Nicole Eva - 36 -

Farthing, M.J.G., 1997. The molecular pathogenesis of giardiasis. Journal of Pediatric Gastroenterology and Nutrition 24: 79-88.

Faubert, G., Reiner, D.S., Gillin, F.D., 1991. Giardia lamblia: Regulation of secretory vesicle formation and loss of ability to reattach during encystations in vitro. Experimental Parasitology 72: 345-354.

Faubert, G., 2000. Immune Response to Giardia duodenalis. Clinical Microbiology 13: 35- 54.

Fernandes, P.D. and Assreuy, J., 1997. Role of nitric oxide and superoxide in Giardia lamblia killing. Brazilian Journal of Medical and Biological Research 30: 93-99.

Gardner, T.B. and Hill, D.R., 2001. Treatment of giardiasis. Clinical Microbiology Reviews 14: 114-128.

Gault, M.J., Gillin, F.D., Zenian, A.J., 1987. Giardia lamblia: stimulation of growth by human intestinal mucus and epithelial cells in serumfree medium. Experimental Parasitology 64: 29-37.

Gillin, F.D., 1987. Encystation and expression of cyst antigens by Giardia lamblia in vitro. Science 235: 1040-1043.

Gillin, F.D. and Sogin, M.L., 2000. The Giardia genome project database. FEMS Microbiology Letters 189: 271-273.

Gottstein, B., Harriman, G.R., Conrad, H.T., Nash, T.E., 1990. Antigenic variation in Giardia lamblia: cellular and humoral immune response in a mouse model. Parasite Immunology 12: 659-673.

Gottstein, B. and Nash, T.E., 1991. Antigenic variation in Giardia lamblia: infection of congenitically athymic nude and scid mice. Parasite Immunology 13: 649-659.

Hardin, J.A., Buret, A.G., Olson, M.E., Kimm, M.H., Gall, D.G., 1997. Mast cell hyperplasia and increased macromolecular uptake in an animal model of giardiasis. Journal of Parasitology 83: 908-912.

Harms, G., Dörner, F., Bienzle, U., Stark, K., 2002. Infections and diseases after travelling to tropical and subtropical areas. Deutsche Medizinische Wochenschrift 127: 1748-1753.

Hassan, S.M., Maachee, M., Cordova, O.M., Diaz de la Guardia, R., Martins, M., Osuna, A., 2002. Human secretory immune response to fatty acid-binding protein fraction from Giardia lamblia. Infection and Immunity 70: 2226-2229.

Hehl, A.B., Marti, M., Köhler, P., 2000. Stage-specific expression and targeting of cyst wall protein –green fluorescent protein chimeras in Giardia. Molecular Biology of the Cell 11: 1789-1800.

Isaac-Renton, J.L., Shahriari, H., Bowie, W.R., 1992. Comparison of an in vitro method and an in vivo method of Giardia excystation. Applied and Environmental Microbiology 58: 1530-1533.

Janoff, E.N., Smith, P.D., Blaser, M.J., 1988. Acute antibody responses to Giardia lamblia are depressed in patients with AIDS. Journal of Infectious Disease 157: 798-804.

von Allmen Nicole Eva - 37 -

Jimenez, J.C., Fontaine, J., Grzych, J.M., Dei-Cas, E., Capron, M., 2004. Systemic and mucosal responses to oral administration of excretory and secretory antigens from Giardia intestinalis. Clinical Diagnostic Laboratory Immunology 11: 152-160.

Kabnick, K.S. and Peattie, D.A., 1990. In situ analyses reveal that the two nuclei of Giardia lamblia are equivalent. Journal of Cell Science 95: 353-360.

Katelaris, P.H., Naeem, A., Farthing, M.J.G., 1995. Attachment of Giardia lamblia trophozoites to a cultured human intestinal cell line. Gut 37: 512-518.

Keas, B.E., 1999. Life Cycle. www.msu.edu/course/zol/316/glamgut.htm

Koudela, B. and Vitovec, J., 1998. Experimental giardiasis in goat kids. Veterinary Parasitology 74: 9-18.

Langford, D.T., Housley, M.P., Boes, M., Chen, J., Kagnoff, M.F., Gillin F.D., Eckmann, L., 2002. Central importance of immunoglobulin A in host defense against Giardia spp. Infection and Immunity 70: 11-18.

Le Blancq, S.M. and Adam, R.D., 1998. Structural basis of karyotype heterogeneity of Giardia lamblia. Molecular and Biochemical Parasitology 97: 199-208.

Leitch, G.J., Visvesvara, G.S., Wahlquist, S.P., Harmon, C.T., 1989. Dietary fiber and giardiasis: dietary fiber reduces rate of intestinal infection by Giardia lamblia in the gerbil. American Journal of Tropical Medicine Hygiene 41: 512-520.

Li, E., Zhou, P., Petrin, Z., Singer, S., 2004. Mast cell-dependent control of Giardia lamblia infections in mice. Infection and Immunity 72: 6642-6649.

Lindley, K.J. and Macdonald, S., 2001. Malabsorption in children. Practitioner 245: 169- 170.

LoGalbo, P.R., Sampson, H.A., Buckley, R.H., 1982. Symptomatic giardiasis in three patients with x-linked agammaglobulinemia. Journal of Pediatric 101: 78-80.

Lujàn, H.D., Mowatt, M.R., Byrd, L.G., Nash, T.E., 1996. Cholesterol starvation induces differentiation of the intestinal parasite Giardia lamblia. Proceedings of the National Academy of Science of the USA 93: 7628-7633.

Lujàn, H.D., Mowatt, M.R., Nash, T.E., 1997. The mechanisms of Giardia lamblia differentiation into cysts. Microbiology Molecular and Biological Review 61: 294-304.

McArthur, A.G., Davids, B.J., Svärd, S.G., Elmendorf, H.G., Gillin, F.D., 2005. Characterization of the Giardia lamblia transcriptome. www.mbl.edu/Giardia.

Meng, T.C., Hetsko, M.L., Gillin, F.D., 1993. Antigenic switching of TSA 417, a trophozoite variable surface protein, following completion of the life cycle of Giardia lamblia. Infection and Immunology 61: 5394-5397.

Mitchell, G.F, Anders, R.F., Brown, G.V., Handman, E., Roberts-Thompson, I.C., Chapman, C.B., Forsyth, K.P., Kahl, L.P., Cruise, K.M., 1982. Analysis of infection characteristics and antiparasite immune responses in resistant compared with susceptible hosts. Immunological Review 61: 137-188.

Moncada, D.M., Kammanadiminti, S.J., Chadee, K., 2003. Mucin and Toll-like receptors in host defense against intestinal parasites. Trends in Parasitology 19: 305-311.

von Allmen Nicole Eva - 38 -

Morrison, H.G., McArthur, A.G., Adam, R.D., Aley, S.B., Gillin, F.G., Olsen, G.J., Sogin, M.L., 2005. Giardia lamblia – A model for ancient eukaryotic genome analysis. www.mbl.edu/Giardia.

Mowatt, M.R., Aggarwal, A., Nash, T.E., 1991. Carboxy-terminal sequence conservation among variant-specific surface proteins of Giardia lamblia. Molecular and Biochemical Parasitology 49: 215-227.

Müller, N., Stäger, S., Gottstein, B., 1996. Serological analysis of antigenic heterogeneity of Giardia lamblia variant surface proteins. Infection and Immunity 64: 1385-1390.

Müller, N. and Gottstein, B., 1998. Antigenic variation and the murine immune response to Giardia lamblia. International Journal of Parasitology 28: 1829-1839.

Müller, N. and Stäger, S., 1999. Periodic appearance of a predominant variant antigen type during a chronic Giardia lamblia infection in a mouse model. International Journal of Parasitology 29: 1917-1923.

Müller, N. and von Allmen, N., 2005. Recent insights into the mucosal reactions associated with Giardia lamblia infections. International Journal of Parasitology, in press (Review).

Nash, T.E., 1992. Surface antigen variability and variation in Giardia lamblia. Parasitology Today 8: 229-243.

Nash, T.E., 2002. Surface antigenic variation in Giardia lamblia. Molecular Micorobiology 45: 585-590.

Nash, T.E. and Kaiser D.B., 1985. Differences in excretory-secretary products and surface antigens among 19 isolates of Giardia. Journal of Infectious Disease 152: 1166-1171.

Nash, T.E., McCutchan, T., Keister, D., Dame, J.B., Conrad, J.D., Gillin, F.D., 1985. Restriction-endonuclease analysis of DNA from 15 Giardia isolates obtained from humans and animals. Journal of Infectious Disease 152: 64-73.

Nash, T.E., Herrington, D.A., Losonsky, G.A., Levine, M.M., 1987. Experimental human infections with Giardia lamblia. Journal of Infectious Diseases 156: 974-984.

Nash, T.E., Aggarwal, A., Adam, R.D., Conrad, T.J., Merrit, J.W., 1988. Antigenic variation in Giardia lamblia. Journal of Immunology 141: 636-641.

Nash, T.E., Banks, S.M., Alling, D.W., Merrit, J.W., Merrit, J.W.Jr,. Conrad, J.T., 1990. Frequency of variant antigens in Giardia lamblia. Experimental Parasitology 71: 415-421.

Nash, T.E., Meritt, J.W., Conrad, J.T., 1991. Isolate and variability in susceptibility of Giardia lamblia to intestinal proteases. Infection and Immunology 59: 1334-1340.

Nash, T.E. and Mowatt, M.R., 1992. Characterization of a Giardia lamblia variant-specific surface protein (VSP) gene from isolate GS/M and estimation of the VSP gene repertoire size. Molecular and Biochemical Parasitology 51: 219-228.

Nash, T.E. and Mowatt, M.R., 1993. Variant-specific surface proteins of Giardia lamblia are zinc-binding proteins. Proceedings of the National Academy of Science of USA 90: 5483-5489.

Nino, C.A. and Wassermann, M., 2003. Transcription of metabolic enzyme genes during the excystation of Giardia lamblia. International Parasitology 52: 291-298.

von Allmen Nicole Eva - 39 -

Oberhuber, G., Kastner, N., Stolte, M., 1997. Giardiasis: a histologic analysis of 567 cases. Scandinavian Journal of Gastroenterology 32: 48-51.

Olson, M.E., Ceri, H., Morck, D.W., 2000. Giardia vaccination. Parasitology Today 16: 213-217.

Ouellette, A.J., Darmoul, D., Tran, D., Huttner, K.M., Yuan, J., Selsted, M.E., 1999. Peptide localization and gene structure of cryptdin 4, a differentially expressed mouse paneth cell alpha-defensin. Infection and Immunity 67: 6643-6651.

Palm, J.E., Weiland, M.E., Griffiths, W.J., Ljungstrom, I., Svärd, S.G., 2003. Identification of immunoreactive proteins during acute human giardiasis. Journal of Infectious Disease 187, 1849-1859.

Papanastasiou, P., Hiltpold, A., Bommeli, C., Köhler, P., 1996. The release of the variant surface protein of Giardia to its soluble isoform is mediated by the selective cleavage of the conserved carboxy-terminal domain. Biochemistry 35: 10143-10148.

Papanastasiou, P., McConville, M.J., Ralton, J., Köhler, P., 1997a. The varaiant-specific surface protein of Giardia, VSP4A1, is a glycosylated and palmitoylated protein. Biochemical Journal 322: 49-56.

Pegado, M.G., de Souza, W., 1994. Role of surface components in the process of interaction of Giardia duodenalis with epithelial cells in vitro. Parasitology Research 80: 320-326.

Perez, P.F., Minnaard, J., Rouvet, M., Knabenhans, C., Brassart, D., De Antoni, G.L., Schiffrin, E.J., 2001. Inhibition of Giardia intestinalis by extracellular factors from Lactobacilli: an in vitro study. Applied and Environmental Microbiology 67: 5037-5042.

Pimenta, P.F., da Silva, P.P., Nash, T.E., 1991. Variant surface antigens of Giardia lamblia are associated with the presence of a thick cell coat: thin section and label fracture immunocytochemistry survey. Infection and Immunity 59: 3989-3996.

Rakoff-Nahoum, S., Paglino, J., Eslami-Varzaneh, F., Edberg, S., Medzhitov, R., 2004. Recognition of commensal microflora by toll-like receptors is required for intestinal homeostasis. Cell 118: 229-241.

Roberts-Thompson, I.C., Grove, D.I., Stevens, D.P., Warren, K.S., 1976. Suppression of giardiasis during the intestinal phase of trichinosis in the mouse. Gut 17: 953-958.

Roskens, H. and Erlandsen, S.L., 2002. Inhibition of in vitro attachment of Giardia trophozoites by mucin. Journal of Parasitology 88: 869-873.

Salzmann, A., Denenberg, A.G, Ueta, I., O`Connor, M., Linn, S.C., Szabo, C., 1996. Induction and activity of nitric oxide synthase in cultured human intestinal epithelial monolayers. American Journal of Physiology 270: 565-573.

Scott, K.G.E., Logan, M.R., Klammer, G.M., Teoh, D.A., Buret, A.G., 2000. Jejunal brush border microvillous alterations in Giardia-muris-infected mice: role of T lymphocytes and interleukin-6. Infection and Immunity 68: 3412-3418.

Scott, K.G.E., Yu, L.C.H., Buret, A.G., 2004. Role of CD8+ and CD4+ T lymphocytes in jejunal mucosal injury during murine giardiasis. Infection and Immunity 72: 3536-3542.

Seow, F., Katelaris, P., Ngu M., 1993. The effect of Giardia lamblia trophozoites on trypsin, chymotrypsin and amylase in vitro. Parasitology 106: 233-238.

von Allmen Nicole Eva - 40 -

Singer, S.M. and Nash, T.E., 2000a. T-cell-dependent control of acute Giardia lamblia infections in mice. Infection and Immunity 68: 170-175.

Singer, S.M. and Nash, T.E., 2000b. The role of normal flora in Giardia lamblia infections in mice. Journal of Infectious Disease 181: 1510-1512.

Singer, S.M., Elmendorf, H.G., Conrad, J.T., Nash, T.E., 2001. Biological selection of variant-specific surface proteins in Giardia lamblia. Journal of Infectious Diseases, 183: 119-124.

Smith, M.W., Aley, S.B., Sogin, M., Gillin, F.D., Evans, G.A., 1998. Sequence survey of the Giardia lamblia genome. Molecular and Biochemical Parasitology 95: 267-280.

Sogin, M.L., Gunderson, J.H., Elwood, H.J., Alonso, R.A., Peattie, D.A., 1989. Phylogenetic meaning of the kingdom concept: an unusual ribosomal RNA from Giardia lamblia. Science 243: 75-77.

Stäger, S., Felleisen, R., Gottstein, B., Müller, N., 1997a. Giardia lamblia variant surface protein H7 stimulates a heterogenous repertoire of antibodies displaying differential cytological effects on the parasite. Molecular and Biochemical Parasitology 85: 113-124.

Stäger, S., Gottstein, B., Müller, N., 1997b. Systemic and local antibody response in mice induced by a recombinant peptide fragment from Giardia lamblia variant surface protein (VSP) H7 produced by a Salmonella typhimurium vaccine strain. International Journal of Parasitology 27: 965-971.

Stäger, S. and Müller, N., 1997. Giardia lamblia infections in B-cell deficient transgenic mice. Infection and Immunity 65: 3944-3946.

Stäger, S., Gottstein, B., Sager, H., Jungi, T.W., Müller, N., 1998. Influence of antibodies in mother`s milk on antigenic variation of Giardia lamblia in the murine mother-offspring model of infection. Infection and Immunity 66: 1287-1292.

Svärd, S.G., Meng, T.C., Hetsko, M.L., McCaffery, J.M., Gillin, F.D., 1998. Differentation- associated surface antigen variation in the ancient eukaryote Giardia lamblia. Molecular Microbiology 30: 979-989.

Tandon, B.N., Tandon, R.K., Satpathy, B.K., Shriniwas, B.K., 1977. Mechanism of malabsorption in giardiasis: a study of bacterial flora and bile salt deconjugation in upper jejunum. Gut 18: 176-181.

Teoh, D.A., Kamieniecki, D., Pang, G., Buret, A.G., 2000. Giardia lamblia rearranges F- actin and alpha-actinin in human colonic and duodenal monolayers and reduces transepithelial electrical resistance. Journal of Parasitology 86: 800-806.

Thompson R.C.A., 2000. Giardiasis as a re-emerging infectious disease and its zoonotic potential. International Journal of Parasitology 30: 1259-1267.

Tomkins, A.M., Drasar, B.S., Bradley, A.K., Williamson, W.A., 1978. Bacterial colonization of jejunal mucosa in giardiasis. Transaction of the Royal Society of Tropical Medicine and Hygiene 72: 33-36.

Touz, M.C., Nores, M.J., Slavin, I., Piacenza, L., Acosta, D., Carmona, C., Lujàn, H.D., 2002. Membrane-associated dipeptidyl peptidase IV is involved in encystation-specific gene expression during Giardia differentiation. Biochemical Journal 364: 703-710.

von Allmen Nicole Eva - 41 -

Ullu, E., Tschudi, C., Chkrabortx, T., 2004. RNA interference in protozoan parasites. Cellular Microbiology 6: 509-519.

Ungar, B.L.P. and Nash, T.E., 1987. Cross-reactivity among different Giardia lamblia isolates using immunofluorescent antibody and enzyme immunoassay techniques. American Journal for Tropical Medical Hygiene 37: 283-289.

Venkatesan, P., Finch, R.G., Wakelin, D., 1997. A comparison of mucosal inflammatory responses to Giardia muris in resistant B10 and susceptible BALB/c mice. Parasite Immunology 19: 137-143.

Vesy, C.J. and Peterson, W.L., 1999. Review article: the management of giardiasis. Aliment Pharmacological Therapy 13: 843-850. von Allmen, N., Bienz, M., Hemphill, A., Müller, N., 2004. Experimental infections of neonatal mice with cysts of Giardia lamblia clone GS/M-83-H7 are associated with an antigenic reset of the parasite. Infection and Immunity 72: 4763-4771. von Allmen, N., Bienz, M., Hemphill, A., Müller, N., 2005a. Quantitative assessment of sense and antisense transcripts from genes involved in antigenic variation (vsp genes) and encystations (cwp 1 gene) of Giardia lamblia clone GS/M-83-H7. Parasitology 130: 389-396. von Allmen, N., Christen, S., Forster, U., Gottstein, B., Welle, M., Müller, N., 2005b. Acute trichinellosis increases susceptibility of Giardia lamblia infections in the mouse model. Submitted to Infection and Immunity.

Walker, R.I. and Owen, R.L., 1990. Intestinal barriers to bacteria and their toxins. Annual Review in Medicine 41: 393-400.

Weiland, M.E, Palm, J.E., Griffiths, W.J., McCaffery, J.M., Svärd, S.G., 2003. Characterisation of alpha-1 giardin: an immunodominant Giardia lamblia annexin with glycosaminoglycan-binding activity. International Journal of Parasitology 33: 1341-1351.

Werries, E., Franz, A., Hippe, H., Acyl, Y., 1991. Purification and substrate specifity of two cysteine proteases of Giardia lamblia. Journal of Protozoology 34: 378-383.

Wiesenhahn, G.P, Jarroll, E.L., Lindmark, D.G., Meyer, E.A., Hallick, L.M., 1984. Giardia lamblia: autoradiographic analysis of nuclear replication. Journal of Experimental Parasitology 58: 94-100.

Williamson, A.L., O`Donoghue, P.J., Upcroft, J.A., Upcroft, P., 2000. Immune and pathophysiological responses to different strains of Giardia duodenalis in neonatal mice. International Journal of Parasitology 30: 129-136.

Yang, Y. and Adam, R.D., 1995. A group of Giardia lamblia variant-specific surface protein (VSP) genes with nearly identical 5` regions. Molecular and Biochemical Parasitology 75: 69-74.

Zenian, A. and Gillin, F.D., 1985. Interactions of Giardia lamblia with human intestinal mucus: enhancement of trophozoite attachment to glass. Journal of Protozoology 32: 664- 668.

Zhang, Y.Y., Aley, S.B., Stanley, S.L., Gillin, F.D., 1993. Cysteine-dependent zinc binding by membrane proteins of Giardia lamblia. Infection and Immunology 61: 520-524.

von Allmen Nicole Eva - 42 -

Zhou, P., Li, E., Zhu, N., Robertson, J., Nash, T., Singer, S.M., 2003. Role of Interleukin-6 in the control of acute and chronic Giardia lamblia infections in mice. Infection and Immunity, 71: 1566-1568.

von Allmen Nicole Eva - 43 -

8 Publications

INFECTION AND IMMUNITY, Aug. 2004, p. 4763–4771 Vol. 72, No. 8 0019-9567/04/$08.00ϩ0 DOI: 10.1128/IAI.72.8.4763–4771.2004 Copyright © 2004, American Society for Microbiology. All Rights Reserved.

Experimental Infections of Neonatal Mice with Cysts of Giardia lamblia Clone GS/M-83-H7 Are Associated with an Antigenic Reset of the Parasite N. von Allmen,1 M. Bienz,2 A. Hemphill,1 and N. Mu¨ller1* Institutes of Parasitology1 and Hematology,2 University of Berne, CH-3001 Berne, Switzerland

Received 22 February 2004/Returned for modification 21 March 2004/Accepted 15 April 2004

Transmission of the protozoan parasite Giardia lamblia from one to another host individuum occurs through peroral ingestion of cysts which, following excystation in the small intestine, release two trophozoites each. Many studies have focused on the major surface antigen, VSP (for variant surface protein), which is respon- sible for the antigenic variability of the parasite. By using trophozoites of G. lamblia clone GS/M-83-H7 (expressing VSP H7) and the neonatal mouse model for experimental infections, we quantitatively assessed the process of antigenic variation of the parasite on the transcriptional level. In the present study, variant-specific regions identified on different GS/M-83-H7 vsp sequences served as targets for quantitative reverse transcrip- tion-PCR to monitor alterations in vsp mRNA levels during infection. Respective results demonstrated that antigenic switching of both the duodenal trophozoite and the cecal cyst populations was associated with a massive reduction in vsp H7 mRNA levels but not with a simultaneous increase in transcripts of any of the subvariant vsp genes analyzed. Most importantly, we also explored giardial variant-type formation and vsp mRNA levels after infection of mice with cysts. This infection mode led to an antigenic reset of the parasite in that a VSP H7-negative inoculum “converted” into a population of intestinal trophozoites that essentially consisted of the original VSP H7 type. This antigenic reset appears to be associated with excystation rather than with a selective process which favors expansion of a residual population of VSP H7 types within the antigenically diversified cyst inoculum. Based on these findings, the VSP H7 type has to be regarded as a predominant variant of G. lamblia clone GS/M-83-H7 which (re-)emerges during early-stage infection and may contribute to an optimal establishment of the parasite within the intestine of the experimental murine host.

Giardia lamblia is an intestinal protozoan parasite of hu- itoring vsp gene expression during in vitro en- and excystation mans and various animals. Manifestation of the infection varies of G. lamblia isolate WB, clone C6 (21). This investigation from asymptomatic carriage to severe diarrhea and malabsorp- demonstrated that major surface antigen VSP C6 predomi- tion. G. lamblia has a two-stage life cycle. The nonproliferat- nates in cysts but is downregulated during the process of ex- ing, quadrinucleated cysts are ingested through consumption cystation. Conversely, in vitro-maintained G. lamblia clone GS/ of contaminated freshwater or food. Passage of the cysts M-83-H7 (expressing VSP H7) did not change its surface through the stomach triggers excystation. This excystation pro- antigen properties during life cycle stage transition (21). cess continues in the small intestine, where cysts release two Antigenic variation has been extensively studied by using the binucleated trophozoites each. The life cycle is completed G. lamblia clone GS/M-83-H7 as a model parasite. The VSP when proliferating trophozoites convert into infective cysts. repertoire of this clone was estimated to be encoded by about Cysts are excreted in the feces and are able to persist for a long 60 to 80 distinct genes (13). Experimental infections in a com- time under humid conditions until they are ingested by a sus- bined mother-offspring mouse system revealed that a predom- ceptible host. inant VSP H7-type population can be replaced by a mixture of In the last few years, multiple studies have addressed the different new variant types during infection (11). This initial ability of G. lamblia to alter its surface antigen properties (10, antigenic switching of the parasite population inside the suck- 12). These studies have revealed that antigenic variation is ling offspring is under the growth-selective influence of lacto- associated with a unique family of surface antigens, named genic anti-VSP H7 antibodies maternally produced in response VSP (for variant surface protein). VSP has a potential function to the parasite infection (19). in antigenic variation to circumvent the host immune response, By using G. lamblia clone GS/M-83-H7 and the mother- and it may exhibit physiological functions important for the offspring mouse model for experimental infections, we recently intestinal survival of the parasite within the host (14). Surface investigated the process of antigenic variation of the parasite antigen alterations were observed within proliferating popula- on the molecular level (4). For this purpose, trophozoites col- tions of intestinal trophozoites (2, 4, 6, 7, 19), as well as among lected from the intestines of individual animals at different individual trophozoites upon the release from nonproliferative time points postinfection (p.i.) were analyzed directly for their cysts (9, 21). The latter phenomenon was investigated by mon- vsp gene transcription patterns, i.e., without cultivating the recovered parasites in vitro. This was done by using a com- bined 5Ј rapid amplification of cDNA ends (5ЈRACE) reverse * Corresponding author. Mailing address: Institute of Parasitology, P.O. Box 8466, CH-3001 Berne, Switzerland. Phone: (4131) 6312474. transcription-PCR (RT-PCR) approach which allowed detec- Fax: (4131) 6312622. E-mail: [email protected]. tion, and subsequent sequence analysis, of vsp gene transcripts

4763 4764 VON ALLMEN ET AL. INFECT.IMMUN.

FIG. 1. Schematic illustration of the 3ЈRACE RT-PCR-based quantification of vsp mRNA. First, mRNA (within total RNA preparations) from G. lamblia clone GS/M-83-H7 trophozoites or cysts was reverse transcribed into cDNA by using Moloney murine leukemia virus reverse transcriptase and the T-ANC primer. This primer contains a 3Ј-terminal oligonucleotide (dT) stretch (upper cases) which anneals to the poly(A) tail (lower case) of the mRNA, and a 5Јterminal anchor (ANC) sequence as indicated. The T-ANC primer consists of a mixture of molecules containing either base A, or C, or G (V ϭ A/C/G) at the very 3Ј-terminal position in order for the primer to anneal to the inner end of the poly(A) tail. The cDNA was taken as a template for a 3ЈRACE PCR which allowed specific amplification of vsp cDNA molecules. Amplification of vsp cDNA was achieved by performing the PCR with forward vsp primers listed in Fig. 2A or forward primer MM16 (targeted to a highly conserved 3Ј-terminal sequence of the GS/M-83-H7 vsp genes [see reference 15]) and a reverse anchor (ANC) primer as indicated. For amplification of cDNA from “housekeeping” gene gdh, forward gdh primer (see Materials and Methods) was used instead of a forward vsp primer.

after the generation of amplified cDNA analogues. The same by incubation for 20 min at 37°C. After two washes with 1 ml of prewarmed PCR approach was applied for analysis of vsp gene transcripts (37°C) PBS, the trophozoites were detached from the bottom of the well by a in variants obtained after negative selection of axenic tropho- 15-min incubation step on ice. These trophozoites were further processed (i.e., without prior expansion of the recovered parasite populations by in vitro culti- zoites by treatment with a cytotoxic, VSP H7-specific mono- vation) for infection of mice as previously described (20). clonal antibody. In an overall view on the entire panel of in The in vivo-derived cyst inoculi were prepared as follows. Cecal content from

vivo- and in vitro-derived parasite populations, the transcrip- six mice per experimental time point was pooled in 10 ml of distilled H2O, and tion of 29 different vsp genes was demonstrated. cysts were purified and concentrated by using sucrose flotation according to the method of Belosevic et al. (3). Briefly, this method is based on the principle that, In the present study, the immunofluorescence method and a 3 Ј after centrifugation in a highly concentrated sucrose solution (1.12 g/cm ), the quantitative 3 RACE RT-PCR approach were applied to in- intestinal content separates into a precipitate consisting of high-density fecal vestigate variant-type formation and levels of vsp gene tran- particles, suspended medium-density particles, and low-density particles such as scripts during the course of a G. lamblia GS/M-83-H7 infection giardial cysts floating on the surface of the solution. The floating cysts were in mice. The contents of vsp transcripts in intestinal tropho- collected, and residual trophozoites were removed from water-resistant cysts zoite populations were analyzed after experimental infection of through hypoosmotic lysis by repeated incubation and washing in distilled H2O. Finally, cysts were resuspended at a concentration of 5 ϫ 105 parasites per ml of mice with trophozoites. Respective analyses were performed ␮ ϫ 4 distilled H2O, and 50 l (corresponding to ca. 2.5 10 cysts) from this sus- by intragastric inoculation of mice with trophozoites but also pension was used as an inoculum. for the first time by intragastric inoculation of the animals with Total RNA from in vitro-cultivated GS-M-83-H7 trophozoites was isolated as cysts, thus approaching conditions that model a natural route previously described (4). For total RNA extraction (see below) from intestinal of infection. trophozoites, parasites were isolated as follows. Sections of about 1 cm from the upper part of the duodenum were slit longitudinally and subsequently incubated for 20 min in 1 ml of PBS on ice to detach trophozoites from the intestinal MATERIALS AND METHODS surface. Then, 0.5 ml of PBS supernatant containing detached trophozoites (ca. 4 5 Mice. Gravid 10- to 12-week-old outbred CD-1(ICR)BR mice were obtained 10 to 10 parasites) was transferred into a well of a microtiter plate (Cellstar from Charles River GmbH, Sulzfeld, Germany. Animals were kept according to TC-Plate), and viable parasites were allowed to adhere to the bottom of the well the Swiss regulations of animal experiments with free access to germfree food by incubation for 20 min at 37°C. After three washes with 1 ml of prewarmed ␮ and sterile water. (37°C) PBS, residual, adherent trophozoites were resuspended in 100 l of lysis ␤ Parasite. The origin, axenization, and cloning of G. lamblia clone GS/M-83-H7 buffer– -mercaptoethanol mixture from the StrataPrep Total RNA Microprep has been described by Aggarwal et al. (1). This clone expresses a major 72-kDa kit (Stratagene, La Jolla, Calif.). For total RNA extraction (see below) from ϫ 4 5 antigen (VSP H7) on its surface that is recognized by monoclonal antibody cysts, ca. 5 10 to 10 parasites (for concentrated cyst samples, see above) were (MAb) G10/4. Trophozoites from G. lamblia clone GS/M-83-H7 were cultivated resuspended in 1 ml of QIAzol lysis reagent from the RNeasy lipid minikit in modified TYI-S-33 medium with antibiotics as previously described (8). (Qiagen, Basel, Switzerland). The lysates from both trophozoites and cysts were Experimental parasite infection, sample collection, and total RNA extraction. processed for extraction of total RNA as instructed, including treatment with ␮ Three-day-old offspring and respective mothers were infected with 5 ϫ 104 in RNase-free DNase I. Finally, total RNA preparations (solubilized in 50 lof Ϫ vitro-cultivated G. lamblia clone GS/M-83-H7 trophozoites as previously de- RNase-free distilled H2O) were stored at 80°C until further use. scribed (20). Analysis of vsp mRNA by quantitative RT-PCR. cDNA was synthesized by RT The in vivo-derived trophozoite inoculum was prepared as follows. Sections of from total RNA, prepared from intestinal trophozoite and cyst populations by about 1 cm from the upper part of the duodenum were slit longitudinally and using 17.5 ␮M T-ANC primer (Invitrogen, Basel, Switzerland) (see Fig. 1), as subsequently incubated for 20 min in 1 ml of phosphate-buffered saline (PBS; well as Moloney murine leukemia virus reverse transcriptase (Promega, Madi- containing 0.15 M NaCl [pH 7.2]) on ice to detach trophozoites from the intes- son, Wis.) and other components as instructed by the manufacturer of the tinal surface. Then, 0.5 ml of PBS supernatant containing detached trophozoites reverse transcriptase. (ca. 5 ϫ 104 to 105 parasites) was transferred into a well of a microtiter plate Quantitative RT-PCR was carried out on a LightCycler (Roche Diagnostics, (Cellstar TC-Plate, 24 wells; Greiner Labortechnik GmbH, Frickenhausen, Swit- Rotkreuz, Switzerland) by using SYBR Green I as a double-stranded DNA zerland), and viable parasites were allowed to adhere to the bottom of the well (dsDNA)-specific fluorescent dye and continuous fluorescence monitoring as VOL. 72, 2004 ANTIGENIC RESET DURING TRANSMISSION OF GIARDIA LAMBLIA 4765 previously described (22). Amplifications of cDNA molecules representing ana- primer (see Fig. 1) in a quantitative PCR with an equimolar (1 nM) mixture of logues of mRNA from either different vsp genes or “housekeeping” gene gdh molecules (4-␮l aliquots) of previously generated amplification products from (coding for glutamate dehydrogenase) were performed by 3ЈRACE PCR (see the vsp sequences H7, V17, V12, and V18 as a template (see Fig. 2B). The Fig. 1). Amplification reaction mixes included forward primers either specific for specific character of the PCRs with the forward vsp primers H7a, V17, V12, or individual vsp genes and/or gene groups (indicated in Fig. 1 as forward vsp primer V18 was determined by demonstrating that each of the reactions exclusively and listed in Fig. 2A) or complementary to the MM16 vsp region (primer amplified the corresponding vsp gene sequence from the template mixture. The sequence, 5Ј-GGCTTCCTCTGCTGGTGGTTC-3Ј), which has been assumed to specificity of the different amplification products was assessed by comparative be shared by the entire GS/M-83-H7 vsp gene repertoire (15), or specific for gdh DNA melting-point analysis (see above). The above-mentioned template mixture (5Ј-CCTCAAGTTCCTCGGC-3Ј) and reverse anchor (ANC) primer as indi- was also used for a quantitative PCR with forward primer MM16 annealing to a cated in Fig. 1. Quantitative PCR was done with 4 ␮l of 1:100-diluted (cDNA highly conserved region close to the 3Ј terminus of the GS/M-83-H7 vsp genes prepared from trophozoites) or 1:10-diluted (cDNA prepared from cysts) cDNA (15) (see Fig. 2B). This amplification reaction differed from the PCRs listed (see above) by using the Quanti-Tect SYBR Green PCR kit (Qiagen) in 10 ␮lof above in that it was performed with a slightly modified temperature profile that standard reaction containing a 0.5 ␮M concentration of forward and reverse included measurement of the fluorescence signal at 72°C (instead of 82°C [see primers (Invitrogen). All PCRs containing cDNA were performed either in above]) after each annealing phase. triplicates (with cDNA from trophozoite and cyst pools, see Fig. 3) or in dupli- Sequence alignments. Alignment of the vsp nucleotide sequences derived from cates (with cDNA of trophozoites sampled from individual animals) (see Fig. 4 a previous study (4) and accessible in GenBank (for accession numbers, see Fig. and 6). Furthermore, a control PCR included RNA equivalents from samples 2A) was done by using MultAlin and ESPript1.9 computer software, which are that had not been reverse transcribed into cDNA (not shown) to confirm that no available at the ExPASy molecular biology server. DNA was amplified from any residual genomic DNA that might have resisted Immunofluorescence assays. The kinetics of expression of the major surface DNase I digestion (see above). PCR was started by initiating a “hot-start” Taq antigen on trophozoites isolated from the duodenum of experimentally infected DNA polymerase reaction at 95°C (15 min). Subsequent DNA amplification was mice (see below) were assessed for immunofluorescence by using MAb G10/4 as done in 50 cycles consisting of 7 cycles of denaturation (94°C for 15 s), annealing described previously (6). For staining of nuclei, trophozoites were incubated for with a stepwise temperature shift (2°C per cycle [first cycle at 48°C, second cycle 3 min in presence of the 1:300-diluted dsDNA-specific fluorescent dye Hoechst at 50°C, etc., to the sixth cycle at 60°C]; with each cycle lasting 30 s), and 33258 (Sigma, Steinheim, Germany) stock solution (1 mg/ml in distilled H2O) extension (72°C for 30 s) and an additional 43 cycles of denaturation (94°C for and subsequently washed twice in PBS and once in distilled H2O. 15 s), annealing (60°C for 30 s), and extension (72°C for 30 s). The temperature In order to perform the immunofluorescence assays with cysts, ca. 5 ϫ 104 to transition rates in all cycle steps were 20°C/s. Fluorescence was measured at 82°C 105 parasites (for discussion of concentrated cyst samples, see above) were fixed during the temperature shift after each annealing phase in the “single” mode in 100 ␮l of 3% paraformaldehyde solution (in PBS) for1hatroom tempera- with the channel setting F1. Fluorescence signals from the amplification products ture, followed by one wash in PBS, and a 3-h incubation with 100 mM glycine in were quantitatively assessed by applying the standard software (version 3.5.3) of PBS. Fixed cysts were permeabilized with 0.1% Triton X-100 in PBS for1hand the LightCycler. Quantification of PCR products was performed during the log blocked for2hin5%bovine serum albumin (BSA) in PBS. Fixed and perme- phase of the reaction and was achieved by using the secondary derivative max- abilized cells were incubated with MAb G10/4 diluted 1:100 in 5% BSA–0.1% imum mode for plotting of the fluorescence signals versus the cycle numbers. As Triton X-100 in PBS for 1 h. After three washes with ice-cold PBS, the cysts were external standards, serial 10-fold dilutions (4-␮l aliquots) of previously generated incubated for 1 h with goat anti-mouse -fluorescein isothio- amplification products from the different target sequences were included in the cyanate (Fc specific; Sigma). Cysts were then washed three times in ice-cold PBS quantitative PCR analyses. The standard curves from the different assays (vsp- and incubated in presence of 1:20 Texas red-conjugated mouse MAb A300-TR and gdh-PCRs) were run in duplicates and contained 4 log units within a linear (Waterborne, New Orleans, La.), an anti-cyst wall protein antibody. range that essentially covered the maximal and minimal concentrations of the Specimens containing trophozoites and cysts were inspected on a Nikon vsp- and gdh-cDNA sequences within the different samples. Linearity among the Eclipse E800 digital confocal fluorescence microscope at a 600-fold magnifica- standard reactions was reflected by the correlation coefficient that was calculated tion. Processing of the images was performed by using the Openlab 3.11 software by the computer program to be extremely high (between 0.97 and 1.0) for all (Improvision, Heidelberg, Germany). The percentages of VSP H7-positive ver- PCR assays applied. Furthermore, the efficiencies of the quantitative PCRs were sus negative parasites were determined by inspection of 300 parasites. Since the revealed to be nearly identical and exhibited high amplification rates that ranged immunofluorescence analyses of cysts included several washing steps, which between 1.81 (with forward vsp H7 primer [see below]) and 2.08 (with forward reduced the final recovery of parasitic material (Ͻ100 cysts per specimen), often vsp V12 primer [see below]) per cycle. only a few parasites (Ͻ10 cysts per microscopic field) were detected within an Lack of PCR-inhibitory effects and overall comparability of the different stan- individual specimen. In these cases, more than one specimen was inspected. dard and sample reactions were evidenced by the quasi-identity of the slopes from the amplification plots (monitoring amplification rates). Control experiments for identification of PCR products included a DNA RESULTS melting-point analysis (16; data not shown). The DNA melting profile assay was run after the final PCR cycle by gradually increasing the temperature to 95°Cat Immunofluorescence-based assessment of in vivo antigenic a transition rate of 0.1°C/s with continuous acquisition (determination of the switching of G. lamblia clone GS/M-83-H7 after infection of melting profile by measuring loss of fluorescence). The data from the DNA mice with in vitro-cultivated trophozoites. In order to study in melting profile assay were processed by using the standard software (version 3.5.3). In all PCR tests performed, identical melting temperatures of amplicons vivo antigenic switching of G. lamblia clone GS/M-83-H7, we from samples and respective standards indicated identical and specific amplifi- used a previously described protocol (19) for infection of cation reactions without unwanted primer-dimer formation (not shown). This 3-day-old murine offspring and respective mothers with tro- overall identity and specificity of reactions was confirmed by subsequent agarose phozoites. From these animals, we collected duodenal tropho- gel electrophoresis (2% gels) (17) that monitored the PCR products as single DNA bands of expected sizes (not shown). In the cases of the analyses of the zoites as well as cecal cysts from 6 animals at days 0 (repre- trophozoite and cyst pool samples shown in Fig. 3, positions H7 and V17, senting the original inoculum), 7 and 14 (representing parasite nucleotide sequence authenticity of the amplification products of the vsp H7- and populations at time points during the lactation phase of mice), vsp V17-specific PCRs was confirmed by automated sequencing through a com- and 21 (representing parasite populations at a time point after mercial sequencing service (Mycrosynth, Balgach, Switzerland). the lactation phase of the animals) p.i. Immunofluorescence In order to compensate for variations in input RNA amounts and efficiencies of RT, mRNA of the housekeeping gene gdh was quantitated. Respective mean analysis with VSP H7-specific MAb G10/4 (19) demonstrated values from triplicate (pool samples [see above]) or duplicate (samples from that both trophozoites from the inoculum (representing day individual animals [see above]) determinations were taken for the calculation of 0 p.i.) and the trophozoite populations isolated from the duo- the relative vsp mRNA levels (vsp-mRNA level/gdh mRNA level). denum at day 7 p.i. were mostly (ca. 92 to 95%) VSP H7 Specificity of forward vsp H7-specific primer H7a (amplifying partial se- quences of vsp gene H7, see Fig. 2A) and forward subvariant vsp-specific primers positive. In contrast, the duodenal parasite populations recov- (amplifying partial sequences of vsp genes V17, V12, and V18, see Fig. 2A) were ered after day 7 p.i., i.e., on days 14 and 21 p.i., only possessed assessed by applying each of these primers in combination with reverse ANC few (ca. 1% at day 14 p.i.) or no (0% at day 21 p.i.) detectable 4766 VON ALLMEN ET AL. INFECT.IMMUN.

FIG. 2. Determination of target sequence (A) and specificity (B) of forward vsp primers used for quantitative RT-PCR-based assessment of relative vsp mRNA levels in different G. lamblia GS/M-83-H7 populations (see Fig. 3, 4, and 6). (A) Primers specific for different vsp genes are targeted to a relatively variable vsp stretch as identified by alignment of nucleotide sequences from 3Ј-terminal regions of vsp H7 (indicated as H7) and subvariant GS/M-83-H7 vsp genes that had previously been described (see reference 4). The figure shows nucleotide (nt) positions of partial sequences from cDNA clones that had been isolated from either trophozoites representing parasite populations at days 0 (original inoculum) (clone H7, clone 0a), 7 (clones 7a to 7c), 14 (clones 14a to 14g), 21 (clones 21a to 21f), and 42 (clones 42a to 42e) p.i. or trophozoites representing an in vitro-switched parasite population (clones IVa to IVh). The asterisk labeling clone 14d in box V8 indicates that this clone shares 100% sequence identity with clones 21b and IVd that are both not included in the figure. Positions demonstrating a high (light gray type), medium (dark gray type) or low (boldface) degree of similarity among vsp H7 and the corresponding nt regions from the subvariant vsp genes identified, as well as the highly diversified target sequences (open boxes) of forward vsp primers (see Fig. 1) are indicated. Dashes within the sequences indicate alignment gaps. Individual vsp genes, or groups consisting of closely related vsp genes with identical forward vsp primer target sequences (H7 and V1 to V18), are indicated within boxes in front of the sequences. Behind the sequences, corresponding GenBank accession numbers are given. The specificity of forward vsp H7-specific primer H7a (amplifying a partial sequence of vsp gene H7) and forward subvariant vsp-specific primers (amplifying a partial sequence of vsp genes V17, V12, and V18) were assessed by applying each of these primers in combination with reverse ANC primer (see Fig. 1) in a quantitative PCR with an equimolar (1 nM) mixture of molecules (4-␮l aliquots) of previously generated amplification products from the vsp sequences H7, V17, V12, and V18 as a template (B). These PCRs turned out to be vsp gene- or gene group-specific because each of the reactions exclusively amplified the corresponding vsp gene sequence (vsp H7, V17, V12, or V18) from the template mixture. The template mixture was also included in PCRs with forward primer MM16 (targeted to a highly conserved 3Ј-terminal sequence of the GS/M-83-H7 vsp genes [see reference 15]) that was subsequently used to monitor total vsp mRNA levels [V(tot)] in G. lamblia GS/M-83-H7 populations (see Fig. 3). The PCR amplification rates are given as arbitrary amplification units and represent logarithmic mean values (plus standard deviations) from triplicate determinations. VOL. 72, 2004 ANTIGENIC RESET DURING TRANSMISSION OF GIARDIA LAMBLIA 4767

FIG. 3. Quantitative RT-PCR-based assessment of relative vsp mRNA levels (vsp mRNA level/gdh mRNA level) in duodenal trophozoite (A) and cecal cyst (B) pools composed of samples from six G. lamblia clone GS/M-83-H7-infected mice (infection with in vitro-cultivated trophozoites). Samples were taken at days 0 (inoculum), 7, 14, and 21 p.i. as indicated. The relative levels of mRNA from vsp genes or gene groups H7, V17, V12, and V18 (see Fig. 2A), as well as relative levels of total vsp mRNA [V(tot)] represent logarithmic mean values (plus standard deviations) from triplicate determinations.

VSP H7-type trophozoites. These results are in agreement with 3؅RACE RT-PCR-based quantification of vsp mRNA levels our previous findings (4), indicating that antigenic switching of associated with in vivo antigenic switching of G. lamblia clone duodenal GS/M-83-H7 trophozoite populations in offspring GS/M-83-H7 after infection of mice with in vitro-cultivated occurred between day 7 and 14 p.i. Immunofluorescence test- trophozoites. For 3ЈRACE RT-PCR-based quantitative assess- ing of cecal cyst populations in the present study revealed ca. ment of vsp mRNA levels in different G. lamblia clone GS/M- 81% VSP H7-type parasites at day 7 p.i. and no detectable 83-H7 populations during in vivo antigenic switching, duodenal parasites of this variant-type at days 14 and 21 p.i. This obser- trophozoite and cecal cyst pools were generated by combining vation demonstrated that in vivo antigenic switching of the respective parasite populations from six animals representing trophozoite and corresponding cyst populations occurred be- the experimental time points at days 7, 14, and 21 p.i. From the tween day 7 and 14 p.i. different sample groups, total RNA was prepared and then 4768 VON ALLMEN ET AL. INFECT.IMMUN.

PCR with an equimolar mixture of previously generated am- plification products from the vsp sequences H7a, V17, V12, and V18 as a template. The specificities of the primers were confirmed in that the individual PCRs amplified the homolo- gous, but not the three heterologous, vsp sequences within the template mixture (Fig. 2B). Respective amplification rates were ca. 5.0 (vsp V17-specific PCR) to 2.3 (vsp H18-specific PCR) times lower than the amplification rate achieved with forward primer MM16 (see above and reference 15) annealing to all four vsp sequences included in the template mixture. By applying different forward vsp primers and reverse ANC primer for the quantification of the individual vsp cDNAs, we assessed the relative amounts of vsp mRNA within intestinal trophozoite and cyst populations that had consecutively emerged in infected offspring (Fig. 3 and 4). Analyses of tro- phozoites were performed with both pool samples and samples separately collected from individual animals. Conversely, cysts were exclusively analyzed as pool samples because the ex- tremely low mRNA content of the cysts did not allow differ- ential RT-PCR-based testing of individual parasite popula- tions. As can be seen for both pool samples (Fig. 3) and samples from individual animals (Fig. 4), duodenal tropho- zoites at day 7 p.i. contained Ͼ500 times more vsp H7 mRNA than trophozoites at days 14 and 21 p.i. In an analogous assay of the corresponding cyst pools, samples taken at day 7 p.i. exhibited Ͼ250 more vsp H7 mRNA than those taken at days 14 and 21 p.i. These findings confirmed that antigenic switch- ing had occcured between day 7 and 14 p.i. on the transcrip- tional level. An overview on the non-vsp H7 gene transcription in both VSP H7- and non-VSP H7-type parasites (trophozoites and FIG. 4. Quantitative RT-PCR-based assessment of relative vsp cysts) revealed that mRNA derived from vsp genes and/or gene mRNA levels (vsp mRNA level/gdh mRNA level) in duodenal tropho- groups V17, V12, and V18 was synthesized on an extremely zoite samples taken from six G. lamblia clone GS/M-83-H7-infected mice (infection with in vitro cultivated trophozoites) at days 7 (A), 14 low level (Fig. 3). In this case, mRNA production was deter- (B), and 21 (C) p.i. The relative levels of mRNA from vsp genes or mined to be Ͼ200 (trophozoites) or Ͼ150 (cysts) times lower gene groups H7, V17, V12, and V18 (see Fig. 2A) represent logarith- than vsp H7 mRNA production in VSP H7-type populations. mic mean values from duplicate determinations. Within the pooled trophozoite and cyst samples, the mRNA levels derived from the other subvariant vsp genes and/or gene groups (V1 to V11 and V13 to V16 [see Fig. 2A]) included in used as a template for cDNA synthesis (Fig. 1). The cDNA the study (not shown) turned out to be consistently below the molecules were generated by applying an oligo(dT) primer mRNA levels derived from vsp genes or gene groups H7, V17, containing an anchor sequence at its 5Ј end. In order to estab- V12, and V18. As assessed by RT-PCR with a primer comple- lish a PCR system for quantitative and differential vsp gene mentary to the highly conserved vsp region MM16 (15), total Ј amplification, 3 regions of gene vsp H7 and subvariant vsp vsp mRNA levels (Vtot) were consistently high in all tropho- genes from clone GS/M-83-H7 were subjected to a nucleotide zoite (Fig. 3A) and cyst (Fig. 3B) populations investigated. sequence alignment (Fig. 2A). This comparative sequence Respective values were similar to those determined for the vsp analysis led to the identification of a relatively variable vsp H7 transcripts in VSP H7-type trophozoite or cyst populations. region, which allowed us to design forward primers that were Taken together, the above-mentioned results indicated that specific for either individual vsp genes or at least small groups downshift of the vsp H7 mRNA level during in vivo antigenic consisting of two to four closely related vsp genes. Each of switching of both trophozoites and cysts of G. lamblia clone these forward vsp primers in combination with universal re- GS/M-83-H7 was not associated with a significant upshift of verse ANC primer (see Fig. 1) were used for PCR-based spe- the mRNA levels derived from the 18 subvariant vsp genes, or cific quantification of corresponding vsp cDNA molecules gen- gene groups, tested. erated from the different trophozoite and cyst samples outlined Immunofluorescence-based assessment of in vivo antigenic above. In order to validate our approach, specificities of ex- switching of G. lamblia clone GS/M-83-H7 after infection of perimentally most relevant (see below) forward vsp primers mice with in vivo-derived cysts and trophozoites. In a further amplifying a partial sequence of either vsp H7 (sequence H7a; experiment, we infected 3-day-old murine offspring and re- see Fig. 2A) or subvariant vsp genes V17, V12, and V18 (see spective mothers with cecal cysts which had been sampled at Fig. 2A) were assessed. This was achieved by testing each of days 7 (identified as VSP H7-type cysts [see above]) and 21 the four primers plus the reverse ANC primer in a quantitative (identified as non-VSP H7-type cysts [see above]) p.i. from VOL. 72, 2004 ANTIGENIC RESET DURING TRANSMISSION OF GIARDIA LAMBLIA 4769

FIG. 5. Immunofluorescence-based assessment of VSP H7 (MAb G10/4) positivity of parasite populations related to an infection of mice with in vivo-derived cysts or trophozoites. Analyses of both inocula used for infection of mice (A) and duodenal trophozoites sampled from infected mice at day 7 p.i. (B) are shown. For infection, VSP H7-type cysts (isolated in the previous infection experiment at day 7 p.i.) (a), non-VSP H7-type cysts (isolated in the previous infection experiment at day 21 p.i.) (b), and non-VSP H7-type trophozoites (isolated in the previous infection experiment at day 21 p.i.) (c) were used as inocula. VSP H7-type parasites were immunostained by a two-step incubation with VSP H7-specific mouse MAb G10/4 and anti-mouse immunoglobulin G-fluorescein isothiocyanate visible as yellow-green (cysts) or green (trophozoites) stain. In order to localize the MAb G10/4 (VSP H7)-negative parasites, cysts were stained with a Texas red-conjugated anti-cyst wall protein MAb (red stain) and trophozoites were visualized by staining of nuclei with dsDNA-specific fluorescent dye Hoechst 33258 (blue stain). Approximate percentages of VSP H7-positive parasites within the inocula (indicated in box on the top of the figure) and approximate mean values (plus ranges) representing percentages of VSP H7-positive parasites within duodenal trophozoites populations from three infected mice (indicated in box at the bottom of the figure) are indicated.

mice that belonged to the previous G. lamblia GS/M-83-H7 switching of the trophozoite populations had occurred between infection experiment (see above). In parallel, a control infec- days 7 and 14 p.i (not shown). tion with non-VSP H7-type duodenal trophozoites sampled at Substantially different results were obtained by performing day 21 p.i. of the previous infection experiment was performed. immunofluorescence analyses with the duodenal trophozoite In order to exactly monitor the initial parasite populations populations sampled at days 7 (Fig. 5Bc) and 14 and 21 p.i. emerging during these infections, three animals per experi- (data not shown) from individual animals that had been in- mental group were sacrificed daily between days 1 and 7 p.i. fected with non-VSP H7-type trophozoites (sampled at day 21 and were assessed for the presence of duodenal trophozoites. during the previous infection [see above]) (Fig. 5Ac). Here, all For all three types of infection, this monitoring exhibited de- trophozoite populations including the one taken at day 7 p.i. tectable numbers of parasites starting from day 6 p.i. and exhibited a VSP H7-negative phenotype. substantial amounts of parasites (Ͼ104 trophozoites per 1 cm 3؅RACE RT-PCR-based quantification of vsp mRNA levels of duodenal section) at day 7 p.i. As can be seen in the immu- associated with in vivo antigenic switching of G. lamblia clone nofluorescence assay shown in Fig. 5, infection of mice with GS/M-83-H7 after infection of mice with in vivo-derived cysts VSP H7-type cysts (Fig. 5Aa) resulted in an initial duodenal and trophozoites. The immunofluorescence results shown in trophozoite population (at day 7 p.i.) that was essentially VSP Fig. 5 were essentially confirmed on the transcriptional level by H7 positive (91 to 94% positivity) (Fig. 5Ba). Surprisingly, a using the 3ЈRACE RT-PCR approach as outlined above (Fig. similar VSP H7-type positivity (90 to 92%) (Fig. 5Bb) was also 6). A highly abundant level of vsp H7 mRNA was detectable in observed within the initial duodenal trophozoite population initial duodenal trophozoite populations taken at day 7 p.i. that resulted from the infection with non-VSP H7-type cysts from mice that had been inoculated with either VSP H7-type (sampled at day 21 during the previous infection [see above]) (Fig. 6A) or non-VSP H7-type (Fig. 6B) cysts. Conversely, only (Fig. 5Ab). Immunofluorescence analyses with samples taken low vsp H7 mRNA levels were found in the initial trophozoite at later time points (days 14 and 21 p.i) after infection of mice population samples from mice that had received a non-VSP with cysts consistently demonstrated predominance (Ͼ99%) of H7-type trophozoite inoculum. non-VSP H7-type trophozoites. This indicated that antigenic Compared to intestinal trophozoite populations emerging at 4770 VON ALLMEN ET AL. INFECT.IMMUN.

DISCUSSION

G. lamblia clone GS/M-83-H7 expresses the well-character- ized surface antigen VSP H7 and represents one of the few members of G. lamblia isolates that develops an active infec- tion in the experimental murine host (e.g., reviewed in refer- ences 10 and 12). Especially the mother-offspring infection model had previously proven to be well suited for investigating those parameters that are related to antigenic variation of the parasite (e.g., reviewed in references 10 and 12). The goal of our study was to assess, in a quantitative manner, vsp mRNA levels in intestinal trophozoite populations emerging during an experimental G. lamblia clone GS/M-83-H7 infection in mice. By using a quantitative RT-PCR approach, we were able to quantitatively assess levels of mRNA representing both vsp H7 and a set of 18 subvariant vsp genes that had been identified in one of our recent studies (4). In particular, we demonstrated that vsp H7 mRNA was predominant in VSP H7-type para- sites. Our observations from this analysis indicated that initial antigenic switching of clone GS/M-83-H7 was associated with a downshift of the vsp H7 mRNA level but was not accompanied by a significant upshift of the mRNA levels representing any of the subvariant vsp genes tested. In this context, it is important to note that our study only assessed the mRNA levels from Ͻ30% of the entire GS/M-83-H7 vsp gene repertoire (18 of ca. 60 to 80 genes estimated to constitute the entire vsp repertoire [see reference 13]). The primer design of our present quanti- FIG. 6. Quantitative RT-PCR-based assessment of relative vsp mRNA levels (vsp mRNA level/gdh mRNA level) in mice upon infec- tative RT-PCR approach relied on the limited vsp gene se- tion of mice with cysts and trophozoites. Three mice were infected with quence information that had been generated in a previous in vivo-derived VSP H7-type (isolated in the previous infection exper- study which in a nonquantitative manner allowed detection of iment at day 7 p.i.,) (A, open circle) and non-VSP H7-type (isolated in vsp mRNA molecules in VSP H7- and non-VSP H7-type tro- the previous infection experiment at day 21 p.i.) (B, open circle) cysts, or with in vivo-derived non-VSP H7-type trophozoites (isolated in the phozoites (4). In that investigation, vsp gene transcripts had previous infection experiment at day 21 p.i.) (C, open circle). From been identified as cDNA analogues by performing a multistep these mice, duodenal trophozoites (closed circles) were sampled at day 5ЈRACE RT-PCR that included degenerate vsp primers for 7 p.i. and analyzed by quantitative RT-PCR for relative levels of amplification of vsp cDNA molecules. By applying this exper- mRNA from vsp genes or gene groups H7, V17, V12, and V18 (see Fig. 2A). Relative vsp mRNA levels are given as logarithmic mean values imental strategy, a selective process during the different am- from triplicate (cyst inocula) or duplicate (duodenal trophozoites iso- plification steps (e.g., caused by inefficient priming or lack of lated from infected animals) determinations. priming of degenerate vsp primers to a subset of vsp gene sequences) might have excluded detection of, and subsequent generation of respective sequence information on, cDNA an- day 7 upon infection with VSP H7-type trophozoites (Fig. 4A), alogues from predominant vsp mRNA molecules within the populations emerging upon infection with VSP H7-type and non-VSP H7-type parasite populations. Due to this possible non-VSP H7-type cysts (Fig. 6A and B) exhibited a reduced lack of important sequence information it is possible that our (ϳ4-fold reduction) content of transcripts concerning both vsp study was not able to demonstrate the entire composition of H7 and subvariant vsp (V17, V12, and V18) mRNA. This the non-vsp H7 gene transcripts emerging during antigenic finding suggested differences in steady-state levels of vsp switching of the parasite. This possibility has to be taken into mRNA within the first and the two latter parasite populations account because, in contrast to the vsp H7 mRNA levels, total mentioned above. However, the absence of significant differ- vsp mRNA levels turned out to be similar in VSP H7-type and ences in the ratios between vsp H7 and subvariant vsp (V17, non-VSP H7-type intestinal parasite populations (see Fig. 3). V12, and V18) mRNA levels did not indicate a differential vsp Fortunately, the above-mentioned experimental constraints mRNA composition within these populations. The ratios were did not substantially affect the outcome of those investigations, consistently high (Ͼ100) and indicated predominance of vsp which were focused on the primary goal of the present study. H7 mRNA within the parasites. Our major efforts were dedicated to a series of experimental In summary, infection of mice with cysts of G. lamblia clone infections which were aimed at the examination of eventual GS/M-83-H7 resulted in VSP H7-type-dominated initial tro- surface antigen alterations of G. lamblia clone GS/M-83-H7 in phozoite populations irrespective of the variant type(s) used as association with transmission of the parasite from one to an- the inoculum. In contrast, inoculation of animals with non- other host individuum. In one of our experiments, we simu- VSP-type trophozoites did not result in detectable VSP H7 lated the natural infection mode in that we applied a novel positivity of the intestinal parasite populations at any experi- protocol that allowed intragastric inoculation of in vivo-iso- mental time point of the infection. lated cysts into suckling mice. In this infection experiment, VOL. 72, 2004 ANTIGENIC RESET DURING TRANSMISSION OF GIARDIA LAMBLIA 4771

VSP H7-type trophozoites were revealed to dominate the ini- ACKNOWLEDGMENTS tial intestinal parasite population irrespective of the variant- We thank A. Hehl and M. Marti (Institute of Parasitology, Zu¨rich, type composition of the cyst inoculum applied for infection of Switzerland) and N. Keller (Institute of Parasitology, Berne, Switzer- the animals. As a consequence of these findings, we first ad- land) for technical support and T. E. Nash (National Institutes of dressed the question of whether such an antigenic reset after Health, Bethesda, Md.) for the gift of MAb G10/4 and G. lamblia clone GS/M-83-H7. transmission of the parasite via cysts was due to a growth- This study was financed through a grant obtained from the Swiss selective process that had occurred during establishment of the National Science Foundation (31-066795.01). trophozoite population inside the intestinal habitat of the mu- rine host. Such a selection could have been promoted through REFERENCES overgrowth of subvariant types by a few VSP H7 types that had 1. Aggarwal, A., J. W. Merritt, and T. E. Nash. 1989. Cysteine-rich variant eventually resided within the inoculum. However, this scenario surface proteins of Giardia lamblia. Mol. Biochem. Parasitol. 32:39–48. 2. Aggarwal, A., and T. E. Nash. 1988. Antigenic variation of Giardia lamblia in was rather unlikely because immunofluorescence analysis of vivo. Infect. Immun. 56:1420–1423. the inoculum in question did not detect VSP H7-type cysts that 3. Belosevic, M, G. M. Faubert, J. D. McLean, C. Law, and N. A. Croll. 1993. Giardia lamblia infections in Mongolian gerbils: an animal model. J. Infect. could have initiated such a growth-selective process after in Dis. 147:222–226. vivo excystation. The existence of such a process could be 4. Bienz, M., M. Siles-Lucas, P. Wittwer, and N. Mu¨ller. 2001. vsp gene ex- largely excluded by demonstrating that infection of suckling pression by Giardia lamblia clone GS/M-83-H7 during antigenic variation in vivo and in vitro. Infect. Immun. 69:5278–5285. mice with in vivo-derived non-VSP H7-type trophozoites did 5. Gilman, R. H., G. S. Marquis, E. Miranda, M. Vestegui, and H. Martinez. not exhibit the antigenic reset phenomenon. 1988. Rapid reinfection by Giardia lamblia after treatment in a hyperen- The findings described above raise the question of whether demic third world community. Lancet i:343–345. 6. Gottstein, B., G. R. Harriman, J. T. Conrad, and T. E. Nash. 1990. Antigenic the antigenic reset had occurred in direct association with in variation in Giardia lamblia: cellular and humoral immune response in a vivo excystation of the parasite. Surface antigen alterations mouse model. Parasite Immunol. 12:659–673. 7. Gottstein, B., and T. E. Nash. 1991. Antigenic variation in Giardia lamblia: related to en- and/or excystation of G. lamblia is a known infection of congenitally athymic nude and scid mice. Parasite Immunol. phenomenon that was particularly demonstrated for clone WB 13:649–659. C6 (9, 21). In the case of clone WB C6, in vitro stage conver- 8. Keister, D. B. 1983. Axenic culture of Giardia lamblia in TYI-S-33 medium supplemented with bile. Trans. R. Soc. Trop. Med. Hyg. 77:487–488. sion was shown to trigger switching of vsp gene transcription. 9. Meng, T. C., M. L. Hetsko, and F. D. Gillin. 1993. Antigenic switching of Conversely, testing of G. lamblia clone GS/M-83-H7 (express- TSA 417, a trophozoite variable surface protein, following completion of the ing VSP H7) under analogous experimental conditions did not life cycle of Giardia lamblia. Infect. Immun. 61:5394–5397. 10. Mu¨ller, N., and B. Gottstein. 1998. Antigenic variation and the murine exhibit such a switching effect (21). In the present study, we immune response to Giardia lamblia. Int. J. Parasitol. 28:1829–1839. repeatedly tried to follow the same protocol to achieve in vitro 11. Mu¨ller, N., S. Sta¨ger, and B. Gottstein. 1996. Serological analysis of the antigenic heterogeneity of Giardia lamblia variant surface proteins. Infect. en- and/or excystation with non-VSP H7-type cultures from Immun. 64:1385–1390. clone GS/M-83-H7 (not shown). We intended to do these in 12. Nash, T. E. 2002. Surface antigenic variation in Giardia lamblia. Mol. Mi- vitro experiments to definitively confirm that the antigenic crobiol. 45:585–590. 13. Nash, T. E., J. T. Conrad, and M. R. Mowatt. 1995. Giardia lamblia: iden- reset phenomenon is both temporally and causally linked to tification and characterization of a variant-specific surface protein gene fam- the process of excystation. Unfortunately, all of these trials ily. J. Eukaryot. Microbiol. 42:604–609. failed because, at least in our studies, the excystation efficiency 14. Nash, T. E., J. W. Merritt, Jr., and J. T. Conrad. 1991. Isolate and epitope variability in susceptibility of Giardia lamblia to intestinal proteases. Infect. of non-VSP H7-type cultures was extremely low. Immun. 59:1334–1340. Based on our data, the antigenic reset mechanism has to be 15. Nash, T. E., and M. R. Mowatt. 1992. Characterization of a Giardia lamblia variant-specific surface protein (VSP) gene from isolate GS/M and estima- regarded as an integral part of transmission (via cysts) of G. tion of the VSP gene repertoire size. Mol. Biochem. Parasitol. 51:219–228. lamblia clone GS/M-83-H7 from one to another murine host 16. Ririe, K. M., R. P. Rasmussen, and C. T. Wittwer. 1997. Product differen- individuum. Clone GS/M-83-H7 may have evolved the VSP tiation by analysis of DNA melting curve during the polymerase chain reac- tion. Anal. Biochem. 245:154–160. H7-type as a predominant variant that achieves an optimal 17. Sambrook, J., and D. W. Russel. 2001. Molecular cloning: a laboratory initiation of the parasite infection and/or maintenance of the manual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, parasite within an affected host population. Further infection N.Y. 18. Singer, S. M., H. G. Elmendorf, J. T. Conrad, and T. E. Nash. 2001. Bio- experiments in different animal systems will reveal whether the logical selection of variant-specific surface proteins in Giardia lamblia.J.In- VSP H7 type is able to establish a G. lamblia GS/M-83-H7 fect. Dis. 183:119–124. 19. Sta¨ger, S., B. Gottstein, H. Sager, T. W. Jungi, and N. Mu¨ller. 1998. Influ- infection not only in mice but also in other species representing ence of antibodies in mother’s milk on antigenic variation of Giardia lamblia potential experimental (e.g., gerbils) or natural hosts (e.g., in the murine mother-offspring model of infection. Infect. Immun. 66:1287– dogs) of the parasite. Investigations in these host systems will 1292. 20. Sta¨ger, S., R. Felleisen, B. Gottstein, and N. Mu¨ller. 1997. Giardia lamblia also address the question of whether the strategy of such an variant surface protein H7 stimulates a heterogeneous repertoire of anti- antigenic reset is compatible with both (i) a participation of bodies displaying differential cytological effects on the parasite. Mol. Bio- chem. Parasitol. 85:113–124. surface antigen alterations in an adaptive process facilitating 21. Sva¨rd, S. G., T. C. Meng, M. L. Hetsko, J. M. McCaffery, and F. D. Gillin. transmission of G. lamblia from one to another host species, as 1998. Differentiation-associated surface antigen variation in the ancient eu- considered by Singer et al. (18), and (ii) the occurrence of karyote Giardia lamblia. Mol. Microbiol. 30:979–989. 22. Wittwer, C. T., K. M. Ririe, R. V. Andrew, D. A. David, R. A. Gundry, and repeated G. lamblia infections in the same individuum, as U. J. Balis. 1997. The LightCycler௢: a microvolume multisample fluorimeter suggested by Gilman et al. (5). with rapid temperature control. BioTechniques 22:176–178.

Editor: W. A. Petri, Jr. 389 Quantitative assessment of sense and antisense transcripts from genes involved in antigenic variation (vsp genes) and encystation (cwp 1 gene) of Giardia lamblia clone GS/M-83-H7

N. VON ALLMEN1,M.BIENZ2,A.HEMPHILL1 and N. MU¨ LLER1* 1 Institute of Parasitology, University of Berne, La¨nggass-Strasse 122, CH-3012 Berne, Switzerland 2 Hematology, Inselspital (University Hospital), University of Berne, Switzerland

(Received 5 August 2004; revised 14 September 2004; accepted 14 September 2004)

SUMMARY

Antigenic variation of the intestinal protozoan parasite Giardia lamblia is caused by an exchange of the parasite’s variant surface protein (VSP) coat. Many investigations on antigenic variation were performed with G. lamblia clone GS/M-83-H7 which produces surface antigen VSP H7. To generate novel information on giardial vsp gene transcription, vsp RNA levels were assessed by quantitative reverse transcription-(RT)-PCR in both axenic VSP H7-type trophozoites and subvariants obtained after negative selection of GS/M-83-H7 trophozoites by treatment with a cytotoxic, VSP H7-specific monoclonal antibody. Our investigation was not restricted to the assessment of the sense vsp transcript levels but also included an approach aimed at the detection of complementary antisense vsp transcripts within the two trophozoite populations. We found that sense vsp H7 RNA predominated in VSP H7-type trophozoites while sense RNA from only one (vsp IVg)of 8 subvariant vsp genes totally analysed predominated in subvariant-type trophozoites. Interestingly, the two trophozoite populations exhibited a similar relative distribution regarding the vsp H7 and vsp IVg antisense RNA molecules. An analogous sense versus antisense RNA pattern was also observed when the transcripts of gene cwp 1 (encoding cyst wall protein 1) were investigated. Here, both types of RNA molecules only appeared after cwp 1 had been induced through in vitro encystation of the parasite. These findings for the first time demonstrated that giardial antisense RNA production did not occur in a constitutive manner but was directly linked to complementary sense RNA production after activation of the respective gene systems.

Key words: Giardia lamblia, antigenic variation, encystation, transcription, antisense RNA.

INTRODUCTION As extensively assessed in the in vitro cultivation system, stage conversion is associated with induction Giardia lamblia (Giardia duodenalis, Giardia in- of various genes, some of which are structurally in- testinalis) is a zoonotic protozoan parasite which volved in cyst formation of the parasite (Hehl et al. resides in the small intestine of human and various 2000; Sun et al. 2002; Davis-Hayman et al. 2003; mammalian hosts. In humans, G. lamblia is a common Lujan & Touz, 2003; Marti et al. 2003). Here, par- cause of endemic and epidemic diarrhoea throughout ticularly the cwp genes encoding the cyst wall pro- the whole world. Transmission of the parasite from teins are massively upregulated during encystation. one to another host individual occurs through per- Previous studies on cyst wall proteins (CWP) 1 (Hehl oral ingestion of cysts which, following excystation et al. 2000) and 2 (Davis-Hayman et al. 2003) in- in the small intestine, release 2 trophozoites each. dicated that control of the corresponding genes, cwp The life-cycle is completed when cysts are formed 1 and cwp 2, occurs on the transcriptional level and through encystation of proliferating trophozoites and involves cis-acting elements in the 5 flanking se- subsequently excreted in the feces. Both, en- and ex- k quence of the gene. Transcriptional activation of cystation can also be triggered in vitro by applying encystation-inducible genes involves a nuclear pro- in vitro growth conditions which – to a certain tein which is related to the Myb family transcription extent – simulate the intestinal (for encystation) and factors and initiates transcription by interacting with gastric (for excystation) milieu of the mammalian a specific binding site in the promoter region (Sun hosts. et al. 2002). As demonstrated for gene cwp 1, steady- state levels of cellular mRNA are modulated by a cis-element in 3k untranslated region which seems * Corresponding author: Institute of Parasitology, La¨nggass-Strasse 122, CH-3012 Berne, Switzerland. Tel: to be crucial for the processing of the transcripts +41 31 631 24 74. Fax: +41 31 631 26 22. E-mail: during the late stage of the encystation (Hehl et al. [email protected] 2000).

Parasitology (2005), 130, 389–396. f 2004 Cambridge University Press DOI: 10.1017/S0031182004006742 Printed in the United Kingdom N. von Allmen and others 390

In the past decade, G. lamblia has been especially repertoire to a single gene at one time. Interestingly, investigated in terms of the parasite’s ability to con- database mining provided clear evidence that the tinuously change its surface antigen coat (Mu¨ller & genome of G. lamblia encodes key functions of the Gottstein, 1998; Nash, 2002). These studies have RNAi pathway (Ullu et al. 2004). The functionality revealed that antigenic variation is associated with a of this pathway in G. lamblia has already been exper- unique family of surface antigens, named VSP imentally proven in that transfection of the parasite (variant surface protein). By using trophozoites of with a vector carrying an antisense sequence of a G. lamblia clone GS/M-83-H7 (expressing VSP H7) target gene was demonstrated to downregulate cor- and the neonatal mouse model for experimental in- responding gene expression (Touz et al. 2002). RNAi fections, we recently quantitatively assessed the is now basically considered to be an ideal molecular process of antigenic variation of the parasite on the biological tool to achieve targeted silencing of any transcriptional level (von Allmen et al. 2004). In this gene function in various protozoan parasites, in- study, variant-specific regions identified on different cluding G. lamblia (Ullu et al. 2004). GS/M-83-H7 vsp sequences served as targets for Elmendorf et al. (2001) found that G. lamblia quantitative reverse transcription (RT)-PCR to trophozoites generate a substantial amount of anti- monitor alterations in vsp mRNA levels during in- sense RNA molecules which correspond to approxi- fection. Respective results demonstrated that antigen mately 20% of the total RNA content. Antisense switching of both the duodenal trophozoite and the RNA production in G. lamblia is highly diversified caecal cyst populations was associated with a strong and involves vsp genes as well as many other consti- reduction in vsp H7 mRNA levels. The same study tutive and regulated genes, which e.g. are (possibly) also explored giardial variant-type formation and vsp relevant to developmental stage conversion, cell mRNA levels after infection of mice with anti- division, and gene transcription. The antisense genically diversified (VSP H7-negative) GS/M-83- transcripts are polyadenylated and they essentially H7 cysts. This infection mode led to an antigenic consist of sterile RNA molecules that do not par- reset of the parasite in that a VSP H7-negative in- ticipate in protein synthesis of the parasite. The oculum ‘converted’ into a population of intestinal study described by Elmendorf et al. (2001) clearly trophozoites that essentially consisted of the original demonstrated that antisense RNA production rep- VSP H7-type. resents a striking molecular biological feature of G. Antigenic variation and the VSPs have been ex- lamblia. However, this investigation was not focused tensively investigated on the molecular, biochemical on the possible gene-regulatory function of antisense and immunological level (Mu¨ller et al. 1998; Nash, transcripts and a preliminary Northern blot analysis 2002). Conversely, the genetic mechanisms control- included in the respective experimentation was not ling antigen switching are, as yet, poorly understood able to reveal a participation of these RNA molecules and only little information on this process is available in developmental gene regulation. in current literature. Nash et al. (2001) demonstrated The major aim of the present study was to evaluate that the shift of VSP expression coincides with tran- the possibility of a bidirectional vsp gene transcrip- sient appearence of few trophozoites which simul- tion in association with in vitro antigenic variation of taneously express at least two VSPs. Furthermore, G. lamblia clone GS/M-83-H7. For this purpose, preliminary data suggested that VSP expression may VSP H7-type and subvariant-type GS/M-83-H7 be regulated at the transcriptional level (Nash & trophozoite cultures were generated and sub- Mowatt, 1992). However, another study tackling the sequently analysed by quantitative reverse RT-PCR genome organization of a novel family of cysteine- for their relative production of sense and antisense rich proteins (CRP65, CRP136) in G. lamblia may vsp H7 or subvariant vsp RNA, respectively. The allow the consideration of a gene rearrangement ef- quantitative RT-PCR approach was also applied for fect as a possible mechanism responsible for antigen investigating the giardial sense and antisense tran- switching of the parasite (Cheng, Upcroft & Upcroft, scription in the encystation-inducible gene system 1996; Upcroft et al. 1997). Since Elmendorf, Singer cwp 1 encoding the major cyst component CWP 1 of & Nash (2001) recently found that G. lamblia con- the parasite. tains sterile antisense vsp RNA, reverse RNA tran- scription, and more specifically, RNA interference (RNAi) (Tijsterman & Plasterk, 2004) is currently MATERIALS AND METHODS also discussed as a potential mechanism that might Parasite and in vitro growth conditions be involved in vsp gene regulation (Ullu, Tschudi & Chkrabortx, 2004). In this context, the review of The origin, axenization and cloning of G. lamblia Ullu et al. (2004) refered to as yet unpublished clone GS/M-83-H7 has been described by Aggarwal findings from H. Lujan and colleagues, which sug- et al. (1989). This clone expresses a major 72 kDa gested that RNAi controls expression of the vsp genes antigen (VSP H7) on its surface which is recognized and that an RNA-dependent RNA polymerase is by MAb G10/4. Trophozoites from G. lamblia clone involved in restricting expression of the vsp gene GS/M-83-H7 were cultivated in modified TYI-S-33 Sense and antisense transcripts from Giardia lamblia genes vsp and cwp 1 391

sense vsp RNA 5' 3' 3' reverse vsp primer 5' cDNA of sense vsp RNA reverse transcription cDNA of antisense vsp RNA 5' forward vsp printer 1 3' 3' 5' antisense vsp RNA

forward vsp primer 2 PCR reverse vsp primer Fig. 1. Schematic illustration of the RT-PCR-based quantification of sense and antisense vsp RNA. First, sense and antisense RNA (within total RNA preparations) from VSP H7-type and subvariant-type G. lamblia clone GS/M-83-H7 trophozoites were reverse transcribed into cDNA using reverse vsp primer (synthesis of cDNA from sense vsp RNA) or forward vsp primer 1 (synthesis of cDNA from antisense RNA). The cDNAs were taken as template for quantitative PCRs which allowed specific amplification of different vsp cDNA molecules. Amplification of vsp cDNA was achieved by performing the PCR with forward vsp primers listed in Table 1 and the reverse vsp primer as indicated. medium with antibiotics as previously described 100 ml of lysis buffer-b-ME mixture from the (Keister, 1983). In vitro growth conditions for in- StrataPrepTM Total RNA Microprep Kit duction of encystation of the parasite were described (Stratagene, La Jolla, CA, USA). In parallel, an by Kane et al. (1991). analogous cellular lysate was prepared from about 106 VSP H7-type trophozoites that had not been treated with MAb G10/4. Both cellular lysates were Immunofluorescence assays then further processed for total RNA extraction as Expression of the major surface antigen VSP H7 in described above. Finally, total RNA preparations in vitro cultivated GS/M-83-H7 trophozoites was were solubilized in 30 ml of elution buffer and stored tested by immunofluorescence using VSP H7- at x80 xC until further used. specific MAb G10/4 as described (Gottstein et al. 1990). CWP 1 synthesis in encysting trophozoites Analysis of vsp mRNA by quantitative reverse was detected by incubation of cells with 1 : 20 Texas transcriptase (RT)-PCR Red-conjugated mouse MAb A300-TR (Water- borne, New Orleans, LA, USA), an anti-CWP 1 Complementary DNA (cDNA) was synthesized by antibody. For staining of nuclei, trophozoites were reverse transcription from total RNA, prepared from incubated for 3 min in the presence of 1 : 300-diluted VSP H7-type and subvariant-type trophozoites double strand (ds) DNA-specific fluorescent dye by using 17.5 mM of degenerate forward vsp primer Hoechst 33258 (Sigma, Steinheim, Germany) stock 1(5k-ACIAAYGGIGTITGYACIGC-3; positions solution (1 mg/ml in distilled H2O) and subsequently I=deoxyinosine, Y=C,T; Invitrogen, Basle, washed twice in PBS and once in distilled H2O. Switzerland) (reverse transcription of antisense vsp Specimens were inspected on a Nikon Eclipse E800 RNA) or reverse vsp primer (5k-GAACCACCAGC digital confocal fluorescence microscope at a 600-fold AGAGGAA-3k) (reverse transcription of sense vsp magnification. Processing of images was performed RNA) (see Fig. 1) as well as M-MLV reverse tran- using the Openlab 3.11 software (Improvision, scriptase (Promega, Madison, WI, USA) and other Heidelberg, Germany). Percentages of VSP H7- or components as instructed by the manufacturer of the CWP 1-positive parasites were determined by in- reverse transcriptase. The forward vsp primer 1 spection of approximately 400 parasites. covers a relatively conserved vsp region located be- tween nucleotides (nt) 1147 and 1166 of gene vsp H7. The reverse vsp primer is complementary to nt 1636 Antigen switching by GS/M-83-H7 trophozoites to 1653 of the vsp H7 encoding part of the highly in vitro, sample collection and total RNA extraction conserved transmembrane domain of the surface The procedure used for cytotoxic (VSP H7-specific) protein. The reverse transcription reaction for gen- MAb G10/4-mediated antigen switching of in vitro erating cwp 1 cDNA included forward cwp 1 primer cultivated G. lamblia clone GS/M-83-H7 tropho- 1(5k-CACCTGGACTGCAACCAGCTG-3k) (re- zoites from a VSP H7-positive to a – negative (sub- verse transcription of antisense cwp 1 RNA) and variant-type) parasite population was previously reverse cwp 1 primer (5k-AGTACTCTCCGCAGT described (Bienz et al. 2001). After confirming by CCGGATC-3k) (reverse transcription of sense cwp 1 immunofluorescence that the resulting MAb G10/4- RNA). The forward cwp 1 primer 1 corresponds resistant parasite population was essentially VSP to nt positions 6 to 26 and the reverse cwp 1 primer H7-negative (>99%), about 106 of these negatively- to nt positions 207 to 228 of a 3k terminal GS/M- selected variant trophozoites were resuspended in 83-H7 cwp1 gene segment (GeneBankTM Accession N. von Allmen and others 392

Table 1. Forward vsp primers used for quantification of Giardia lamblia GS/M-83-H7 vsp transcripts by RT-PCR

Nt position vsp GeneBankTM of primer on gene Accession no. vsp sequence Forward vsp primer sequence

H7 M80480 1282–1305# 5kCAAGATAAAGACAGCAATGGTTCA3k IVc AF354540 117–144* 5kCAAACAGCAGACAGTGGGACAGGATCC3k IVe AF354542 37–59* 5kCAACAGCCAGCTAGTGGTGTG3k IVf AF354543 146–165* 5kCAACAGCCAGCTAGTGGGGTC3k IVd AF354541 142–162* 5kCAAACATATGCTAATAACAAC3k IVh AF354545 142–162* 5kCAACAGCCAGCTAGTGGGGTC3k IVb AF354525 184–207* 5kCAAAGTGCAAATGGTCAAGGCGTG3k IVg AF354544 138–165* 5kTATGTCAAGCTCAGCAACGCTCAAACT3k IVa AF354539 46–66* 5kCAGAATCCTACTAATGGCAAT3k

# Nucleotide (nt) positions from complete vsp H7 sequence (Nash & Mowatt, 1992). * Nt positions from a partial sequence proximal to the 3k terminus of the vsp gene (von Allmen et al. 2004). number: AY676465). Reverse transcription of sense performed during the log phase of the reaction and RNA from gene gdh (encoding glutamate dehydro- was achieved by using the secondary derivative genase) was performed as previously described maximum mode for plotting of the fluorescence sig- (von Allmen et al. 2004). nals versus the cycle numbers. As external standards, Quantitative RT-PCR was carried out on a serial 10-fold dilutions (4 ml aliquots) of previously LightCyclerTM Instrument (Roche Diagnostics, generated amplification products from the different Rotkreuz, Switzerland) by using SYBRTM Green I as target sequences were included in the quantitative a dsDNA-specific fluorescent dye and continuous PCR analyses. The standard curves from the differ- fluorescence monitoring as described (Wittwer et al. ent assays (vsp- and cwp 1-PCRs) were run in 1997). Amplification reaction mixes for the vsp PCRs duplicates and contained 4 log units within a linear included forward primers specific for individual vsp range that essentially covered the maximal and genes (indicated in Fig. 1 as forward vsp primer 2 and minimal concentrations of the vsp- and cwp1-cDNA listed in Table 1), and reverse vsp primer. For the sequences within the different samples. Linearity cwp 1 PCRs, forward cwp 1 primer 2 (5k-GTCCC among the standard reactions was reflected by the AGTTGGCCTTATGACTCT-3k, corresponding correlation coefficient which was calculated by the to nt positions 36 to 58 of 3k terminal GS/M-83-H7 computer program to be extremely high (between cwp 1 gene segment) and reverse cwp 1 primer (see 0.99 and 1.0) for all PCR assays applied. above) were applied. Quantitative PCR was done Lack of PCR-inhibitory effects and overall com- with 4 ml of 1 : 100-diluted cDNA using the Quanti parability of the different standard and sample re- TectTM SYBR Green PCR Kit (Qiagen) in a 10 ml actions were evidenced by the quasi-identity of the standard reaction containing a 0.5 mM concentation of slopes from the amplification plots (monitoring forward and reverse primers (Invitrogen). All PCRs amplification rates). containing cDNA were performed in triplicates. Control experiments for identification of PCR Furthermore, a control PCR included RNA equiv- products included a DNA melting point analysis alents from samples that had not been reverse tran- (Ririe, Rasmussen & Wittwer, 1997) (not shown). scribed into cDNA (not shown) to confirm that no The DNA melting profile assay was run after the final DNA was amplified from any residual genomic DNA PCR cycle by gradually increasing the temperature that might have resisted DNase I digestion (see to 95 xC at a transition rate of 0.1 xC/s with con- above). PCR was started by initiating the ‘Hot-Start’ tinuous acquisition (determination of the melting Taq DNA-polymerase reaction at 95 xC (15 min). profile by measuring loss of fluorescence). Data from Subsequent DNA amplification was done in 50 the DNA melting profile assay were processed by cycles including denaturation (94 xC, 15 s), annealing using the standard software (version 3.5.3). In all (48 xC, 30 s), and extension (72 xC, 30 s); tempera- PCR tests performed, identical melting temperatures ture transition rates in all cycle steps were 20 xC/s. of amplicons from samples and respective standards Fluorescence was measured at 82 xC during the indicated identical and specific amplification reac- temperature shift after each annealing phase in the tions without unwanted primer-dimer formation ‘single’ mode with the channel setting F1. Fluor- (not shown). This overall identity and specificity of escence signals from the amplification products were reactions was confirmed by subsequent agarose gel quantitatively assessed by applying the standard electrophoresis (2% gels) (Sambrook & Russel, 2001) software (version 3.5.3) of the LightCyclerTM which monitored the PCR products as single DNA Instrument. Quantification of PCR products was bands of expected sizes (not shown). In the cases of Sense and antisense transcripts from Giardia lamblia genes vsp and cwp 1 393 the PCRs amplifying segments of vsp H7, vsp IVg, A and cwp 1, nucleotide sequence authenticity of the 1600 amplification products was confirmed by automated 1400 sense antisense sequencing through a commercial sequencing service 1200 (Microsynth, Balgach, Switzerland). 1000 In order to compensate for variations in input, RNA levels 800 RNA amounts and efficiencies of reverse transcrip- vsp 600 tion, sense RNA of the ‘housekeeping’ gene gdh was 400 quantitated as recently described (von Allmen et al. 2004). Respective mean values from triplicate deter- relative 200 0 minations were taken for the calculation of the rela- H7 IVa IVb IVc IVd IVe IVf IVg IVh tive sense and antisense vsp RNA levels (vsp RNA B level/gdh RNA level) or sense and antisense cwp 1 1600 RNA levels (cwp 1 RNA level/gdh RNA level), re- 1400 sense spectively. 1200 antisense 1000 Sequence alignments RNA levels 800

vsp Alignment of the vsp nucleotide sequences derived 600 from a previous study (Bienz et al. 2001) and ac- 400 TM cessible in GenBank (Accession numbers, see relative 200 TM Table 1), was done using the MultAlin and the 0 ESPript1.9TM computer software, available at the H7 IVa IVb IVc IVd IVe IVf IVg IVh TM ExPASy Molecular Biology Server. Fig. 2. Quantitative RT-PCR-based assessment of relative sense (black bars) and antisense (grey bars) vsp RNA levels (vsp RNA level/gdh RNA level) in VSP RESULTS H7-type (A) and subvariant-type (B) trophozoite cultures RT-PCR-based quantification of sense and antisense from Giardia lamblia clone GS/M-83-H7. The relative levels of mRNA from vsp genes H7 and IVa-IVh vsp RNA associated with in vitro antigen switching (see Table 1) represent mean values (plus standard of G. lamblia clone GS/M-83-H7 deviations) from triplicate determinations. To select for new variants within G. lamblia clone GS/M-83-H7 originally consisting of approximately (Bienz et al. 2001). For specific reverse transcription 95% VSP H7-type (MAb G10/4-positive) tropho- of antisense vsp RNA, we applied a previously de- zoites (not shown), cultivated parasites were treated scribed degenerate primer (forward vsp primer 1) twice with VSP H7-specific, cytotoxic MAb G10/4. corresponding to nt positions 1147 to 1166 of vsp H7 The efficacy of this selection was assessed by immu- and covering a relatively conserved region of GS/M- nofluorescence which demonstrated that the treated 83-H7 vsp genes (von Allmen et al. 2004). culture contained more than 99% VSP H7-negative To establish a PCR system for quantitative and (MAb G10/4-negative) trophozoites (not shown). differential amplification of cDNA synthesized from For RT-PCR-based quantitative assessment of sense and antisense vsp RNA, a nucleotide sequence sense and antisense vsp RNA levels in VSP H7-type alignment was performed with 3k regions of gene vsp and subvariant-type GS/M-83-H7 trophozoites, H7 and subvariant vsp genes that had previously total RNA was prepared and then used as a templete been identified in VSP H7-negative trophozoite cul- for synthesis of cDNA representing either sense or tures of G. lamblia clone GS/M-83-H7. This com- antisense vsp RNA (Fig. 1). The primers for reverse parative sequence analysis led to the identification transcription were designed on the basis of pre- of a relatively variable vsp region which allowed to viously generated GS/M-83-H7 vsp gene sequence design forward primers (forward vsp primer 2) that information and allowed synthesis of cDNAs from were specific for individual vsp genes (Table 1). Each sense and antisense RNA templates covering the 3k of these forward vsp primers in combination with a terminal region of the individual vsp genes (Fig. 1). reverse primer (reverse vsp primer) targeted to a For specific reverse transcription of sense RNA, a conserved vsp stretch close to the 3k terminus (Fig. 1) primer (reverse vsp primer) was used that was were used for PCR-based specific quantification complementary to nts 1636 to 1653 close to the 3k of vsp cDNA molecules generated from VSP H7- terminus of the vsp H7 encoding sequence. This se- type and subvariant-type trophozoite cultures out- quence encodes part of the transmembrane domain lined above. of the surface protein and had previously been As can be seen in Fig. 2, sense vsp H7 RNA revealed to be highly conserved within the vsp genes predominated in VSP H7-type trophozoites. of clone GS/M-83-H7 and other G. lamblia isolates Subvariant-type trophozoites, however, did not N. von Allmen and others 394 produce significant amounts of sense vsp H7 RNA specific reverse transcription of sense RNA, reverse but exhibited predominance of subvariant sense vsp cwp 1 primer was used that was complementary to IVg RNA. The low, but clearly detectable, amount nts 207 to 228 of a 3k terminal GS/M-83-H7 cwp 1 of sense vsp IVg RNA in the VSP H7-posititive gene segment. For specific reverse transcription of culture was indicative for the existence of a minor antisense vsp RNA, we applied forward cwp 1 primer subvariant-type population representing approxi- 2 corresponding to nt positions 6 to 26 of the same mately 5% of the total trophozoite population (see gene segment. above). Taken together, these finding were in con- RT-PCR-based quantification of cwp 1 RNA re- cordance with results from the immunofluoresence vealed that in vitro encystation of clone GS/M-83- analysis (not shown) indicating that in vitro antigen H7 induced simultaneous synthesis of sense and, to a switching ‘converted’ an essentially VSP H7-posi- lower extent, antisense cwp 1 RNA. While sense cwp 1 tive into a VSP H7-negative (subvariant-type) tro- RNA levels increased for at least 12 h during en- phozoite culture. Furthermore, data evidenced at cystation, antisense cwp 1 RNA already reached its least on the transcriptional level that the in vitro- maximal level at 6 h. At 24 h, a strong reduction in the switched trophozoite culture represented a relatively cellular contents of both sense and antisense cwp 1 homogeneous subvariant-type population which RNA was detected. Sequence analysis demonstrated preferentially expressed vsp IVg. full complementarity between the sense and anti- RT-PCR-based assessment of the antisense vsp sense cwp 1 RNAs, and sizes of the cwp 1 amplicons RNA contents in both VSP H7-type and subvariant- were determined to be 192 bp (not shown). type trophozoites provided analogous results although cellular concentrations of these molecules DISCUSSION were lower than those of corresponding sense RNAs. Antisense vsp H7 RNA predominated in VSP H7- The discovery of giardial antisense vsp transcription positive trophozoites and antisense vsp IVg RNA has implications for the current thinking about the in -negative trophozoites. Sequence analysis revealed as yet poorly understood genetic mechanism con- that sense and antisense RNAs from these two genes trolling surface antigen alterations in G. lamblia were fully complementary, and the corresponding (Ullu et al. 2004). Considering still unpublished data PCR products exhibited the expected sizes of 377 presented by H. Lujan and colleagues at the 2002 base pairs (bp) (sense and antisense vsp H7 RNA) Molecular Parasitology Meeting in Woods Hole and 361 bp (sense and antisense vsp IVg RNA), re- (MA, USA), particularly RNA interference (RNAi) spectively. (Tijsterman & Plasterk, 2004) has to be discussed as a potential mechanism regulating giardial vsp gene activity. Such a mechanism could favour transcrip- RT-PCR-based quantification of sense and tion of a primary vsp gene in that all other genes from antisense cwp 1 RNA associated with the vsp repertoire are silenced by overproduction of encystation of G. lamblia clone GS/M-83-H7 complementary antisense vsp RNA molecules, and To test whether the simultaneous production of small (20–26 nucleotide pairs) interfering RNA sense and complementary antisense RNA is unique (siRNA) (Tijsterman & Plasterk, 2004), respectively. to the vsp genes or rather a more general phenom- According to this scenario, antigen switching could enon of giardial gene transcription, we tested an occur through a spontaneous, or controlled, mech- encystation-inducible gene system, cwp 1, regarding anism which up-regulates antisense RNA from the its sense versus antisense transcription profile. Gene primary vsp gene and down-regulates antisense RNA cwp 1 encodes cyst wall protein (CWP) 1 which rep- from a secondary vsp gene. In the present study, we resents a major component of the giardial cyst demonstrated for the first time that in vitro antigen (Lujan et al. 1995; Mowatt et al. 1995). As assessed switching of G. lamblia is associated with alterations by immunofluorescence assay using an antibody in both sense and antisense vsp RNA production. In against CWP 1 for immunostaining of VSP H7- our analysis, we found that sense and complementary positive GS/M-83-H7 trophozoites, in vitro stage antisense vsp RNA production in both VSP H7-type conversion resulted in approximately 12% CWP and subvariant-type trophozoites occurred in a sim- 1-positive parasites after 24 h growth in encystation ultaneous, and not a reciprocal, manner. This finding medium (not shown). For RT-PCR-based quanti- may be taken as an argument against the above- tative determination of sense and antisense cwp 1 mentioned speculation suggesting that RNAi could RNA levels in trophozoites, corresponding cDNA act as a mechanism involved in regulation of vsp gene was generated from parasites sampled at 0, 6, 12 and expression during antigen switching of the parasite. 24 h of the encystation period. The RT-PCR ap- The solution of this problem relies on the deter- proach monitoring relative sense and antisense cwp 1 mination of the relative vsp H7 versus vsp IVg levels in encysting trophozoites was designed in RNA composition (see above) on the level of the analogy to the procedure applied for the assessment extremely fragmented siRNA sequences. Respective of sense and antisense vsp RNA (see above). For experimentation will be rather difficult because the Sense and antisense transcripts from Giardia lamblia genes vsp and cwp 1 395

100 with a dramatic decrease of the intracellular cwp 1 90 sense mRNA concentration. In addition, we observed that 80 antisense the antisense cwp 1 RNA levels started to decrease 70 earlier (before 12 h) during encystation than comp- 60 RNA levels lementary sense RNA levels (between 12 and 24 h). 50 40 According to data from Hehl et al. (2000), the re- cwp 1 30 duction in the sense cwp 1 RNA concentration may 20 be due to a post-transcriptional RNA degradation 10 process which is modulated by a cis-acting element in relative 0 0612 24 the 3k untranslated region of the RNA molecules. h after induction of encystation Since antisense cwp 1 RNA is supposed to lack an equivalent regulatory element, it may even undergo Fig. 3. Quantitative RT-PCR-based assessment of faster degradation than sense cwp 1 RNA. Such a dif- relative sense (black bars) and antisense (grey bars) cwp 1 ferential RNA degradation process may have caused RNA levels (cwp 1 RNA level/gdh RNA level) in Giardia lamblia clone GS/M-83-H7 trophozoites the observed bias in the relative sense and antisense sampled at 0, 6, 12, and 24 h after induction of cwp 1 RNA levels at advanced encystation stages of encystation. The relative levels of mRNA from cwp 1 the parasite. gene represent mean values (plus standard deviations) As assessed by determination of the nucleotide from triplicate determinations. sequences of RT-PCR amplification products, sense and antisense RNAs from vsp H7, vsp IVg, and cwp 1 exhibited complete sequence complementarity. standard RNA isolation protocols were not designed This suggested that the individual sense/antisense to efficiently recover small RNA pieces. Further- RNA pairs had been transcribed in cis from opposing more, methods for detection of small RNA molecules DNA strands at the same genetic locus. A significant are supposed to be relatively insensitive. amount of antisense RNA molecules from many As outlined in the Materials and Methods section, different genes including vsp genes had already been the individual vsp primers used for the quantitative detected in G. lamblia clones GS/M-83-H7 and WB/ RT-PCRs target to a sequence stretch that is located 1267, respectively (Elmendorf et al. 2001). In this in the 3k terminal region of the corresponding vsp previous investigation, the overall content of sense genes. In this region, the vsp H7 sequence is indis- RNA was estimated to be approximately four times tinguishable from the sequence of another vsp gene higher than the content of antisense RNA. Based on (vsp H7-1) previously identified in G. lamblia clone this finding, the authors concluded that this compar- GS/M-83-H7 (Nash, Conrad & Mowatt, 1995). ably low synthesis of antisense RNA was eventually Accordingly, our present RT-PCR approach was not driven by frequently occurring AT-rich sequences suitable to discriminate between transcripts from acting as cryptic promoters on the antisense DNA these two closely related genes. However, since vsp strand of the respective gene loci. This conclusion is H7-1 had previously been shown to be silent in clone compatible with our present data demonstrating GS/M-83-H7 (Nash et al. 1995), we concluded that that giardial antisense RNA contents were substan- our approach must have exclusively monitored tial, but unambigously lower than those of simul- transcripts that originated from vsp H7. taneously produced sense RNA molecules. By analysing cwp 1 transcription of G. lamblia It is certainly feasible that antisense RNA pro- clone GS/M-83-H7, we detected maximal levels duction in G. lamblia is only a tribute to the sim- of sense and antisense RNA between 6 h (antisense plicity of this organism which perhaps cannot control RNA) and 12 h (sense RNA) after encystation ‘unspecific’ antisense gene transcription eventually had been induced (see Fig. 3). Interestingly, the occurring as a side-effect of DNA unwinding during cwp 1 gene activation resulted in sense/antisense regular gene transcription. Nevertheless, it has to co-expression patterns that resembled those ob- be kept in mind that the parasite has evidently served for transcription of the vsp genes. Based on maintained abundant antisense RNA production this observation, a participation of RNAi in control during evolution although a high energy metabolism of the cwp 1 gene activity can still not be excluded but is needed for running this process. Accordingly, seems to be rather unlikely (see also above). This further studies will have to address the question assumption is compatible with data from previous whether antisense RNA production in G. lamblia is studies which demonstrated that the up-shift of the biologically significant in that it is relevant for the cell cwp 1 mRNA levels during early encystation is the biology or even pathogenicity of the parasite. consequence of the induction of the transcriptional activity of a classical promoter (Hehl et al. 2000; Sun We acknowledge A. Hehl, M. Marti (Institute of Parasitology, Zu¨rich, Switzerland), N. Keller (Institute et al. 2002). of Parasitology, Berne, Switzerland) and M. E. Weiland In agreement with previous data (Hehl et al. 2000), (Karolinska Institute, Stockholm, Sweden) for technical we found that late-stage encystation was associated support, and T. E. Nash (NIH, Bethesda, Maryland, N. von Allmen and others 396

USA) for his gift of MAb G10/4 and G. lamblia clone MOWATT, M. R., LUJAN, H. D., COTTEN, D. B., BOWERS, B., YEE, GS/M-83-H7. This work was financed through a grant J., NASH, T. E. & STIBBS, H. H. (1995). Developmentally obtained from the Swiss National Science Foundation regulated expression of a Giardia lamblia cyst wall (No. 31-066795.01). protein gene. Molecular Microbiology 15, 955–963. MU¨ LLER, N. & GOTTSTEIN, B. (1998). Antigenic variation and the murine immune response to Giardia lamblia. REFERENCES International Journal for Parasitology 28, 1829–1839. AGGARWAL, A., MERRITT, J. W. & NASH, T. E. (1989). NASH, T. E. (2002). Surface antigenic variation in Giardia Cysteine-rich variant surface proteins of Giardia lamblia. Molecular Microbiology 45, 585–590. lamblia. Molecular and Biochemical Parasitology 32, NASH, T. E., CONRAD, J. T. & MOWATT, M. R. (1995). 39–48. Giardia lamblia: identification and characterization BIENZ, M., SILES-LUCAS, M., WITTWER, P. &MU¨ LLER, N. of a variant-specific surface protein gene family. (2001). Vsp gene expression by Giardia lamblia clone Journal of Eukaryotic Microbiology 42, 604–619. GS/M-83-H7 during antigenic variation in vivo and NASH, T. E., LUJAN, H. T., MOWATT, M. R. & CONRAD, J. T. in vitro. Infection and Immunity 69, 5278–5285. (2001). Variant-specific surface protein switching in CHENG, N., UPCROFT, J. A. & UPCROFT, P. (1996). A new Giardia lamblia. Infection and Immunity 69, 1922–1923. cysteine-rich protein-encoding gene family in Giardia NASH, T. E. & MOWATT, M. R. (1992). Characterization of a duodenalis. Gene 169, 33–38. Giardia lamblia variant-specific surface protein (VSP) DAVIS-HAYMAN, S. R., HAYMAN, J. R. & NASH, T. E. (2003). gene from isolate GS/M and estimation of the VSP gene Encystation-specific regulation of the cyst wall protein 2 repertoire size. Molecular and Biochemical Parasitology gene in Giardia lamblia by multiple cis-acting elements. 51, 219–228. International Journal for Parasitology 33, 1005–1012. RIRIE, K. M., RASMUSSEN, R. P. & WITTWER, C. T. (1997). ELMENDORF, H. G., SINGER, S. M. & NASH, T. E. (2001). The Product differentiation by analysis of DNA melting abundance of sterile transcripts in Giardia lamblia. curve during the polymerase chain reaction. Analytical Nucleic Acids Research 29, 4674–4683. Biochemistry 245, 154–160. GOTTSTEIN, B., HARRIMAN, G. R., CONRAD, J. T. & NASH, T. E. SAMBROOK, J. & RUSSEL, D. W. (2001). Molecular Cloning: A (1990). Antigenic variation in Giardia lamblia: Cellular Laboratory Manual, 3rd Edn. Cold Spring Harbor and humoral immune response in a mouse model. Laboratory Press, Cold Spring Harbor, N.Y.

Parasite Immunology 12, 659–673. SUN, C. H., PALM, D., MCARTHUR, A. G., SVA¨ RD, S. G. & GILLIN, HEHL, A. B., MARTI, M. &KO¨ HLER, P. (2000). Stage-specific F. D. (2002). A novel Myb-related protein involved in expression and targeting of cyst wall protein-green transcriptional activation of encystation genes in Giardia fluorescent protein chimeras in Giardia. Molecular lamblia. Molecular Microbiology 46, 971–984. Biology of the Cell 11, 1789–1800. TIJSTERMAN, M. & PLASTERK, R. A. H. (2004). Dicer at KANE, A. V., WARD, H. D., KEUSCH, G. T. & PEREIRA, M. E. RISC: The mechanism of RNAi. Cell 117, 1–4. (1991). In vitro encystation of Giardia lamblia: large- TOUZ, M. C., GOTIG, N., NASH, T. E. & LUJAN, H. D. (2002). scale production of in vitro cysts and strain and clone Identification and characterization of a novel secretory differences in encystation efficiency. International granule calcium-binding protein from the early Journal for Parasitology 77, 974–981. branching eukaryote Giardia lamblia. Journal of KEISTER, D. B. (1983). Axenic culture of Giardia lamblia in Biological Chemistry 52, 50557–50563. TYI-S-33 medium supplemented with bile. ULLU, E., TSCHUDI, C. & CHKRABORTX, T. (2004). Transactions of the Royal Society of Tropical Medicine RNA interference in protozoan parasites. Cellular and Hygiene 77, 487–488. Microbiology 6, 509–519. LUJAN, H. D., MOWATT, M. R., CONRAD, J. T., BOWERS, B. & UPCROFT, P., CHENG, N. & UPCROFT, J. A. (1997). Telomeric NASH, T. E. (1995). Identification of a novel Giardia organization of a variable and inducible toxin gene lamblia cyst wall protein with leucine-rich repeats. family in the ancient eukaryote Giardia duodenalis. Implications for secretory granule formation and protein Genome Research 7, 37–46.

assembly into the cyst wall. Journal of Biological VON ALLMEN, N., BIENZ, M., HEMPHILL, A. &MU¨ LLER, N. Chemistry 270, 29307–29313. (2004). Experimental infections of neonatal mice with LUJAN, H. D. & TOUZ, M. C. (2003). Protein trafficking in cysts of Giardia lamblia clone GS/M-83-H7 are Giardia lamblia. Cellular Microbiology 5, 427–434. associated with an antigenic reset of the parasite. MARTI, M., REGOS, A., LI, Y., SCHRANER, E. M., WILD, P., Infection and Immunity 72, 4763–4771. MU¨ LLER, N., KNOPF, L. G. & HEHL, A. B. (2003). An WITTWER, C. T., RIRIE, K. M., ANDREW, R. V., DAVID, D. A., TM ancestral secretory apparatus in the protozoan parasite GUNDRY, R. A. & BALIS, U. J. (1997). The LightCycler : Giardia intestinalis. Journal of Biological Chemistry 278, a microvolume multisample fluorimeter with rapid 24837–24848. temperature control. Biotechniques 22, 176–178. 1

Acute trichinellosis increases susceptibility to Giardia lamblia infection

in the mouse model

N. VON ALLMEN1, S. CHRISTEN1, U. FORSTER2; B. GOTTSTEIN1, M. WELLE2* AND N. MÜLLER1*

Institutes of Parasitology1 and Veterinary Pathology2, University of Berne, Berne, Switzerland

Running title: Giardia lamblia/Trichinella spiralis co-infection in mice

*Corresponding authors. Mailing address: Institutes of Parasitology (NM) and Veterinary Pathology

(MW); Vetsuisse Faculty, University of Berne; P.O. Box 8466, CH-3001 Berne, Switzerland. Phone:

(4131) 6312474 (NM), (4131) 6312416 (MW); Fax: (4131) 6312477 (NM), (4131) 6312635 (MW)

2

The intestinal protozoan parasite Giardia lamblia causes diarrhea in humans and animals. In the present study, we used the C57BL/6 inbred mouse model to assess the impact of a nematode

(Trichinella spiralis) infection on the course of a G. lamblia (clone GS/M-83-H7) infection. Acute trichinellosis coincided with transient intestinal inflammation and generated an intestinal environment that strongly promoted growth of G. lamblia trophozoites although the local anti-Giardia immunoglobulin (Ig) A production was not affected. This increased G. lamblia infection intensity correlated with intestinal mast cell infiltration, mast cell degranulation, and total IgE production.

Furthermore, a G. lamblia single-infection investigated in parallel also resulted in intestinal mast cell accumulation but severe infiltration was found to be triggered in absence of IgE. Recently, intestinal mast cells emerging during a G. lamblia infection were reported to be involved in those immunological mechanisms that control intestinal proliferation of the parasite in mice. This antigiardial activity was assumed to be related to the capacity of mast cells to produce IL-6. However, this previous assumption was questioned by our present immunohistological findings indicating that murine intestinal mast cells, activated during a G. lamblia infection were IL-6-negative. In the present co- infection experiments, mast cells induced during acute trichinellosis were not able to control a concurrent G. lamblia infection. This observation makes it feasible that the T. spiralis infection created an immunological and physiological environment that superimposed the antigiardial effect of mast cells and thus favoured intestinal growth of G. lamblia trophozoites in double-infected mice. Furthermore, our findings raise the possibility that intestinal inflammation e.g. as a consequence of a “pre-existing” nematode infection is a factor which contributes to increased susceptibility of a host to a G. lamblia infection. The phenomenon of a “pre-existing” nematode infection prior to a G. lamblia infection is a frequent constellation in endemic areas of giardiasis and may therefore have a direct impact on the epidemiological situation of the disease. 3

Giardia lamblia (syn. Giardia duodenalis, Giardia intestinalis) is a common intestinal protozoan and causes diarrhea in humans and animals. In many individuals, the infection remains asymptomatic, whereas some patients exhibit severe symptoms such as abdominal pain, nausea, as well as servere watery diarrhea as a consequence of malabsorption (e.g. reviewed in reference 1). In many cases, spontaneous resolution of the infection occurs after a few weeks but the disease may also develop into chronic stage. The outcome of the infection is supposed to largely depend on the immunological status of the infected individuum but non-immunological factors are also involved in the interaction between the host and the parasite (e.g. reviewed in references 7, 9, and 32).

In giardiasis, reinfections are common because acquired immunity against G. lamblia is not complete either due to insufficient immune defences or antigenic variation of the parasite. Many studies in natural and experimental rodent hosts addressed the question whether antibodies, and more specifically, local secretory immunoglobulin (Ig) A antibodies play a role in control of the parasite infection (7). Previous findings in the experimental mouse/G. lamblia GS/M-83-H7 model indicated that

B-cell-deficient animals were unable to clear giardial infections (24, 39). Conversely, by comparing the courses of infection in IL-6-deficient and wild-type mice, no obvious correlation between intestinal IgA production and the elimination of the parasite in the duodenum was observed (3, 43). Data from an investigation based on the use of different transgenic mouse strains indicated that an as yet unknown

T-cell-dependent mechanism is essential for controlling the acute phase of a G. lamblia infection (37).

Although the investigations listed above could not completely elucidate the cellular network involved in antigiardial immune defence it became evident that antibody-independent immune effector mechanisms directly, or indirectly, interfere with maintenance and growth of the intestinal parasite population particularly during the acute phase of the infection (24, 26, 37). Several investigations in the past provided evidence that mast cells are substantially involved in intestinal elimination of Giardia

(8, 26, 30). Li et al. (26) recently found that mast cells are activated during a G. lamblia infection and that mast cell-deficient, or –depleted, C57BL/6 mice, failed to control such an infection. In this report, mast cells were considered to be a potential source for IL-6. Since in the abovementioned (26) and other studies (3, 43) murine IL-6-deficiency was associated with a much increased susceptibility to a

G. lamblia clone GS/M-83-H7 infection, mast cell-derived IL-6 was suggested to be important for the control of respective infection.

As outlined above, the immunological parameters of a G. lamblia infection have mostly been determined in experimental hosts. However, only very few of these investigations took into account 4 that the outcome of giardiasis may be influenced by other infections that had eventually occurred prior to the G. lamblia infection. Although not necessarily clinically relevant, such infections may alter the immunological and physiological intestinal environment and thus increase, or reduce, the susceptibility of a host to a subsequent G. lamblia infection. The existence of such a phenomenon was exemplified in a former investigation which assessed various parameters of a double infection of mice with the gut- dwelling nematode Trichinella spiralis and Giardia muris (35). As evidenced by multiple studies, T. spiralis infection is accompanied by Th2 cell-mediated eosinophilia and IgE production (11, 28, 31, 36,

42). During the early-infective, intestinal phase, T. spiralis triggers recruitment of inflammatory cells in the mucosa (13), and a subsequent worm expulsion is an immune-mediated process involving activation of Th2-cells (10, 11, 40) and mast cells (6, 19, 21). In the co-infection experiment performed by Roberts-Thompson et al. (35), T. spiralis-induced mucosal inflammation was associated with a strong reduction of the G. muris parasite load. This reduction in the giardial infection intensity turned out to be proportional to the mucosal inflammatory response that had emerged as a consequence of the T. spiralis infection.

Refering to these findings, we now performed a study to assess the impact of a murine T. spiralis infection on the course of a G. lamblia infection. For the infection experiments, we used the well-characterized G. lamblia clone GS/M-83-H7 (2). As expected, ”pre-infection” of mice with T. spiralis larvae resulted in a transient mucosal infiltration of inflammatory cells ( and mast cells). However, in contrast to the T. spiralis/G. muris co-infection experiment described by Roberts-

Thompson et al. (35), intestinal inflammation during a T. spiralis/G. lamblia co-infection was accompanied by a transient massive increase of the Giardia parasite load. These results indicated that intestinal inflammatory reactions, induced by a T. spiralis infection, and more specifically T. spiralis- induced mucosal mast cell functions, did not contribute to the elimination of a G. lamblia GS/M-83-H7 infection.

MATERIALS AND METHODS

Animals. C57BL/6 mice were obtained from Charles River GmbH (Germany). The study was performed with 4 week-old female animals that were kept under specific pathogen-free (SPF) conditions according to Swiss regulations governing animal experimentation and rules for animal protection that restrict the use of experimental animals to a minimum. 5

Parasite and experimental infection. The origin, axenization and cloning of G. lamblia

GS/M-83-H7 has been described by Aggarwal et al. (2). G. lamblia trophozoites were cultivated in TYI-

S-33 medium with antibiotics as previously described (23).

Experimental G. lamblia GS/M-83-H7 infections were done with 106 trophozoites (suspended in 200µl of a 0.3M NaHCO3 solution) of G. lamblia GS/M-83-H7 using a blunt-end needle for peroral inoculation. The course of the G. lamblia infection within mice was determined according to Gottstein et al. (18) by quantitating the parasite burden through microscopical examination of adherent trophozoites from intestinal washes.

T. spiralis (strain ISS 384) was maintained through consecutive passages in outbred CD®-

1(ICR)BR mice which were purchased from Charles River GmbH (Germany). Infectious larvae (L1) were prepared by artificial digestion of muscle tissue from infected carcasses with 1% Pepsin (Sigma-

Aldrich, St. Louis, MO) and 1% HCl and subsequent repeated washing in H2O as described by Friend et al. (12). The released larvae were microscopically counted and 300 larvae were resuspended in 50

l phosphate-buffered saline (PBS), pH7.2, and were then applied for peroral infection of mice using a blunt-ended needle. Infectivity of the larvae was confirmed by demonstrating that the means of the total carcass larval counts within the T. spiralis-infected and T.spiralis/G. lamblia double-infected animal groups was between 4’000 and 8’000 larvae per animal. The efficiency of the T. spiralis infection was determined at a time-point where muscle larvae were expected to be fully developed (7 weeks after T. spiralis infection).

Determination of intestinal mast cells and eosinophils. Mice were sacrificed by CO2 euthanasia. At necropsy, approximately 0.5cm of the small intestine was fixed in formalin and embedded in paraffin wax. Serial sections of 4m were cut from each tissue block. One section of each deparaffinized tissue sample was stained with the Luna`s method (34) to assess the number of eosinophilic .

The mast cell numbers were determined with an enzyme-histochemical reaction for the detection of chymase activity with a commercially available detection kit (Sigma, Catalogue No. 91-C) using Naphthol-AS-D-Chloroacetate as substrate. The only modification to the kit protocol was the use of Fast Blue BB (Base) instead of Fast Red Violet LB (Base).

Immunohistological staining of IL-6-positive intestinal mast cells. For the detection of IL-

6 produced in intestinal cells, an immunohistochemical staining using a polyclonal goat anti-rat IL-6 antibody (Santa Cruz Biotechnology, Santa Cruz, Ca, catalogue No. sc 1266) was performed. The 6 specificity of the immunohistochemical staining was demonstrated with a blocking peptide (Santa Cruz

Biotechnology, Santa Cruz, Ca, catalogue No. sc 1266P). In detail, slides were blocked (0.5% BSA-

PBS/TBS, 10min at room temperature) before the specimen was incubated with the primary antibody

(diluted 1:35 in PBS, 45min at room temperature). This step was followed by incubation with a mouse anti-goat horseradish peroxidase conjugate (1:40 in PBS, 20min at room temperature). Endogenous peroxidase was quenched by incubating the section in H2O2 (1:10 in methanol, 10min at room temperature). Staining was completed by incubation with AEC Substrate-Chromogen for approximately 20min at room temperature. Between each incubation step slides were washed twice for 5min in PBS. Positive staining for IL-6 resulted in a red colour indicating the antigen localization of

IL-6. Slides were counterstained with Ehrlich’s hematoxylin and mounted in GVA mounting solution

(Zymed Laboratories Inc. San Francisco CA, U.S.A.).

Determination of serum mast cell protease 1 and IgE levels. Serum concentrations of mouse mast cell protease 1 (MMCP-1) were determined as previously described by Li et al. (26).

Levels of serum IgE were determined by using the PharMingen IgE-capture ELISA (PharMingen, San

Diego, CA; catalogue number 553413) according to the instructions of the manufacturer. IgE concentrations were extrapolated from a graph of standard OD405nm-values versus concentrations.

Determination of intestinal anti-G. lamblia IgA levels. For determination of intestinal anti-

Giardia IgA concentrations, we applied the procedure described by Gottstein et al. (18), which is suitable for the extraction of IgA from intestinal epithelium and lamina propria. Immunoreactivity of intestinal IgA antibodies was tested as previously described (38), by using a soluble protein extract from G. lamblia clone GS/M-83-H7.

Statistical methods. The significance of the differences among the G. lamblia single-infected and the T. spiralis/G. lamblia double-infected animal groups was determined by the Student's t-test using the Microsoft Excel™ program. P values of <0.01 were considered statistically highly significant.

RESULTS

Intestinal Giardia lamblia proliferation in relation to inflammatory and local antibody reactions during primary Trichinella spiralis and secondary Giardia lamblia infection. To determine the impact of a T. spiralis infection on a subsequent G. lamblia GS/M-83-H7 infection, 4 week-old femal mice either remained uninfected, or were perorally infected with 300 T. spiralis larvae 7

(week –1, representing time-point at 1 week prior to G. lamblia infection). After one week (week 0, time-point at which animals were infected with G. lamblia), part of the previously uninfected, or T. spiralis-infected mice were perorally infected with 106 G. lamblia trophozoites. At different time-points corresponding to weeks –1, and 0, as well as weeks 1, 2, 3, and 6 post G. lamblia infection (p. G. i.), 6 animals per experimental group (uninfected, T. spiralis-infected, G. lamblia-infected, and T.spiralis/G. lamblia-double-infected animals) were sacrificed by CO2 euthanasia.

The efficiency of the T. spiralis infection was determined by assessing mean total carcass larval counts within the T. spiralis-infected and T.spiralis/G. lamblia double-infected animal groups at a time-point where muscle larvae were expected to be fully mature (7 weeks after T. spiralis infection).

Mean larval counts within the two experimental animal groups was between 4’000 and 8’000 per animal, respectively.

Regarding the course of the G. lamblia infection in T. spiralis-pre-infected or non-pre-infected animals, the following results were obtained: animals exclusively infected with G. lamblia, exhibited a low-intensity infection and trophozoites were only detectable at weeks 1 (mean: 4.6x104 parasites per cm duodenum) and 2 (mean: 0.2x104 parasites per cm duodenum) p. G. i. (Fig. 1A). Compared to these animals, T. spiralis-pre-infected animals exhibited an approximately 5-fold and 150-fold higher duodenal trophozoite burden at weeks 1 and 2 p. G. i., respectively. Furthermore, in double-infected animals residual trophozoites were still visible at weeks 3 (mean: 6.6x104 parasites per cm duodenum) and 6 (0.2x104 parasites per cm duodenum) p. G. i.. Accordingly, pre-infection of mice with T. spiralis transiently promoted the intestinal G. lamblia trophozoite proliferation and thus up-shifted both intensity and duration of the G. lamblia infection.

In addition, we analysed various parameters linked to the intestinal inflammatory status of the different experimental animal groups in relation to the course of the G. lamblia infection in single- and double-infected mice. In particular, we found that the T. spiralis infection resulted in a transient accumulation of eosinophils in the intestinal mucosa, which already reached its maximum around the time-point where the G. lamblia infection was initiated (week 0 p. G. i.) (Fig. 1B). Conversely, the G. lamblia infection alone did not lead to a significant eosinophil infiltration.

In T. spiralis-infected mice, mucosal mast cells started to accumulate around week 0 p. G. i. and reached a maximal density around week 3 p. G. i. (Fig. 1C). At week 6 p. G. i., only relatively small numbers of mast cells resided in the intestinal mucosa of these animals. In G. lamblia-single- infected mice, mucosal mast cell numbers were somewhat lower and peak levels were observed at 8 week 2 p. G. i. before respective cell numbers dropped to basal levels during week 3 p. G. i..

Interestingly, T. spiralis/G. lamblia double-infected mice exhibited mucosal mast cell numbers that at weeks 1 and 2 p. G. i. were significantly higher than those observed in T. spiralis or G. lamblia single- infected animals. Accordingly, consecutive infection of mice with T. spiralis and G. lamblia apparently had a synergistic effect on mast cell infiltration into the mucosa.

As assessed by Li et al. (26), the serum MMCP-1 concentration is a suitable marker to monitor the degranulation status of intestinal mast cells activated during a G. lamblia infection. In the present study, infection of animals with T. spiralis resulted in an immediate and strong induction of serum

MMCP-1 already detectable at week 0 p. G. i. (Fig. 1D). In both, T. spiralis single-infected and T. spiralis/G. lamblia double-infected mice, the serum MMCP-1 levels remained relatively high until week

3 p. G. i., and only slighty declined between week 3 and 6 p. G. i.. Compared to these two experimental animal groups, G. lamblia single-infected animals only produced approximately 50% of serum MMCP-1 between weeks 1 and 2 p. G. i., and MMCP-1 production essentially did not extend week 3 p. G. i..

As previously described, murine trichinellosis is associated with IgE production that is involved in mast cell homeostasis, and that partially triggers mast cell degranulation including the release of mediators such as MMCP-1 (20). An assay monitoring serum total IgE levels revealed that serum from

T. spiralis single-infected and T. spiralis/G. lamblia double-infected mice, but not serum from G. lamblia-infected and non-infected mice, contained detectable amounts of IgE (Fig. 1E). In the two positive groups, IgE production started around week 0 p. G. i. and extended the final time-point (week

6 p. G. i.) of the infection experiment. Interestingly, double-infected animals exhibited a marked rise in total IgE levels between weeks 0 and 1 p. G. i. whereas in T. spiralis-single-infected animals high-level serum IgE production started between week 1 and 2 p. G. i..

Since in current literature (e.g. reviewed in reference 7) intestinal IgA is discussed as a possible immunological effector involved in the control of giardiasis, we decided to examine the different experimental animal groups in view of their intestinal anti-G. lamblia IgA-production. For this purpose, intestinal IgA was extracted from intestinal tissue samples and IgA-antibody reactivities to soluble G. lamblia GS/M-83-H7 trophozoite antigen were determined by ELISA as described in

Materials and Methods (Fig. 1F). In this analysis, both G. lamblia single- and T. spiralis/G. lamblia double-infected animals elicited a significant intestinal IgA response to G. lamblia. Respective antibody reactions were already detectable between one and two weeks p. G. i.. 9

Intestinal Giardia lamblia proliferation in relation to inflammatory and local antibody reactions during primary Giardia lamblia and secondary Trichinella spiralis infection. To determine the impact of a T. spiralis infection on the course of a previous G. lamblia GS/M-83-H7 infection, 5-week-old femal mice were perorally infected with 106 G. lamblia trophozoites (week 0).

Part of these animals was sacrificed at time-points corresponding to weeks 1 (time-point at which T. spiralis infection had been carried out) and 3 p. G. i.. As shown in Fig. 2A, during acute-phase G. lamblia infection around week 1 p. G. i., exposure of the animals to a subsequent T. spiralis infection resulted in (re-)activation of the intestinal G. lamblia trophozoite growth within a following 2-week period. However, when the secondary T. spiralis infection was performed at later stages, no relapsing of an eventual residual G. lamblia population was observed (data not shown).

Comparative analyses of intestinal inflammation and local antibody production during T. spiralis-induced re-activation of a G. lamblia GS/M-83-H7 infection provided the following findings: while no massive intestinal accumulation of eosonophils was detected in any of the experimental animal groups (Fig. 2B), G. lamblia single- and G. lamblia/T. spiralis double-infected mice exhibited a significant increase of intestinal mast cells (Fig. 2C). As shown in Fig. 2D, this increase went along with a mast cell degranulation pattern. Thus, at week 3 p. G. i., G. lamblia/T. spiralis double-infected animals contained approximately 9-fold higher MMCP-1 serum levels than G. lamblia-single infected animals. Furthermore, elevated serum total IgE production was restricted to double-infected animals and was only observed at week 3 of the co-infection experiment (Fig. 2E). Conversely, anti-G. lamblia

IgA was detected in both G. lamblia-single and G. lamblia/T. spiralis-double-infected mice (Fig. 2F). In both animal groups, antibody concentrations reached similar levels and started around week 1 p. G. i..

Characterization of IL-6-producing intestinal cells in Trichinella spiralis/Giardia lamblia single- and double-infected mice. Intestinal mast cell-derived IL-6 may be important for control of a

G. lamblia clone GS/M-83-H7 infection in mice (26). In order to find out if a previous T. spiralis infection increases murine susceptibility to infection with G. lamblia clone GS/M-83-H7 and if this is associated with a reduced IL-6-positivity of intestinal mast cells, a combined histological/immunohistological investigation was carried out on formalin-fixed, paraffin-embedded intestinal tissue, representing the different experimental animal groups. In these histological sections, mast cells were visualized by chymase staining. In corresponding sections, IL-6-producing cells were detected with an IL-6-specific antibody and the specificity of the immunostaining was confirmed by demonstrating neutralization of the antibody reactivity with a blocking peptide. 10

Typical features from the different (immuno-)histological stainings as outlined above are shown in Fig. 3. In intestinal tissue from G. lamblia-single-infected (Fig. 3A), T. spiralis-single-infected (not shown), T.spiralis/G. lamblia-double-infected (not shown), and uninfected (Fig. 3B) mice no anti-IL-6 antibody staining patterns were observed that may have reflected the distribution of intestinal cellular infiltrates such mast cells. In contrast, epithelial cells exhibited a specific and consistent reactivity with the anti-IL-6-antibodies, and in all experimental animal groups epithelial immunostaining was visible at about the same intensity. Conclusively, these results indicated that neither in single- and double- infected mice, nor in uninfected control mice, intestinal mast cells (and other intestinal cellular infiltrates) synthesized detectable amounts of IL-6. Conversely, intestinal epithelial cells seemed to produce IL-6 irrespective of the infection status of the animals investigated.

DISCUSSION

As recently reported, G. lamblia (clone GS/M-83-H7) infections in mice were associated with intestinal mast cell infiltration and this process seemed to be involved in resolution of infection (26).

The aim of the present study was to find out if intestinal mast cells - stimulated during a concurrent T. spiralis infection - are also able to modulate the course of such a G. lamblia infection. Our investigation was performed in analogy to a former T. spiralis/G. muris co-infection experiment that demonstrated a suppression of G. muris trophozoite proliferation as consequence of a preceding T. spiralis infection (35). However, in contrast to the findings from this former study, our results revealed that a T. spiralis infection did not suppress, but rather stimulate, intestinal growth of a respective G. lamblia trophozoite population. The divergent outcome of these two analogous co-infection experiment may reflect differences in the (immuno-) biology of the G. lamblia and the G. muris species, but the particular reasons for these differential giardial growth patterns under conditions of a concomittent T. spiralis infection still remain to be elucidated.

In our study, C57BL/6 mice pre-infected with T. spiralis demonstrated a transiently increased susceptibility to a G. lamblia (clone GS/M-83-H7) infection. Furthermore, the intensity of a primary G. lamblia infection was enhanced by secondary infection of animals with T. spiralis. In concordance with previous data (20), the present co-infection experiment in mice revealed that early phase T. spiralis infection coincided with transient intestinal inflammation and was particularly associated with a massive eosinophil and mast cell infiltration into the mucosa, mast cell degranulation, and IgE 11 production. Intestinal eosinophil infiltration was visible only during a short period (approximately 1 week) during the initial stage of the T. spiralis infection and thus most likely did not have a substantial influence on the course of the secondary G. lamblia infection. Conversely, in our co-infection experiments, intestinal accumulation of mast cells was observed during a much longer period

(approximately 3 weeks) and this phenomenon strikingly coincided with a transient increase of the intestinal G. lamblia trophozoite burden. Thus, it became evident that mast cells emerging within the intestinal inflammatory environment during an acute T. spiralis infection were not able to control a concurrent G. lamblia infection. Rather, our findings suggested that these mast cells may have contributed to enhanced growth of the intestinal G. lamblia population. However, this hypothesis is challenged by data from a recent study demonstrating that mast cells induced by G. lamblia display an inhibitory effect on giardial proliferation in mice (26). In combination, the previous and present results outlined above may indicate that murine mast cells induced by either T. spiralis or G. lamblia were involved in functionally distinct immunological processes that either promoted, or suppressed, intestinal growth of G. lamblia trophozoites. Since both of these putatively distinct mast cell populations turned out to release MMCP-1 (see Figs 1 and 2 D), we concluded that the differential effect on giardial growth was probably not related to differential degranulation characteristics of these particular cells.

One possible mechanism for mast cell activation is dependent on IgE and Fcε-receptor I although other factors such as IL-10 are involved in the degranulation process as well (17). Activation of mast cells by IgE is supposed to elicit the release of an array of cytokines and chemokines which are involved in regulation of the acute inflammatory response (20). Interestingly, in our investigation the G. lamblia infection was found to trigger MMCP-1 release in absence of IgE, whereas in T. spiralis- infected mice this process occured in presence of IgE. This finding certainly needs further attention and future investigations e.g. based on the use of an IgE-deficient mouse model will have to explore if production of IgE has a negative effect on antigiardial immunity and increases susceptibility of the animals to the parasite infection. Respective results will be particularly interesting because a correlation between increased IgE production and susceptibility to symptomatic giardiasis has already been observed in an allergic versus non-allergic group of Venezuelan children (5).

As reported formerly, epithelial cells (29) but also mast cells (e.g. reviewed in reference 14) have the capacity to express IL-6. Furthermore, previous experimental G. lamblia infections in IL-6- deficient mice clearly evidenced a central role of IL-6 in antigiardial immunity (3, 43). These 12 observations are in agreement with recent data from Li et al. (26) suggesting that IL-6-producing mast cells may be key players within the intestinal immunological network that mediates immunity against a

G. lamblia infection in mice. However, this assumption is challenged by our present immunohistological data indicating that murine mast cells, induced by a G. lamblia infection, seemed not to produce IL-6 in a quantity above the level of IL-6 production of intestinal epithelial cells from uninfected, or G. lamblia-infected, mice. Our conclusion was also confirmed by results from a quantitative reverse transcription (RT-)-PCR approach, which did not demonstrate increased intestinal

IL-6 expression upon infection of mice with G. lamblia (data not shown). The conflicting results listed above make it evident that further extensive experimental work will be needed to elucidate the IL-6- and mast cell-dependent processes that contribute to murine antigiardial immunity.

The observed phenomenon of an increased G. lamblia trophozoite load in presence of a T. spiralis infection in mice has to be discussed under the following considerations. From current literature it is known that a T. spiralis infection directs the immune response towards a Th2 outcome and can affect the immune responsiveness to a second infection (27). Interestingly, our recent experimental G. lamblia infections in IL-6-deficient mice revealed that susceptibility to infection was accompanied by Th2-biased immune reactions towards the parasite (3). Taking these two aspects into account it is feasible that in the present co-infection experiment a Th2 bias due to the T. spiralis infection created an immunosuppressive environment which favoured intestinal growth of G. lamblia trophozoites. In the experimental mouse model, T. spiralis-induced immunosuppression is a well- known phenomenon that was e.g. demonstrated to affect Th1 cell-mediated processes involved in the delayed-type hypersensitivity reaction to the heterologous protein ovalbumin (4). In this particular case, T. spiralis-induced immunosuppression was supposed to be linked to a shift in the Th1/Th2 balance, relating to an increased IL-10 production in response to the worm infection.

Intestinal mucosal inflammation has already previously been evaluated regarding its importance in resistance to murine giardiasis (e.g. reviewed in reference 32). For example, a study on G. muris infections in a resistant versus susceptible mouse model did not evidence a correlation between resistance against infection and mucosal inflammation (41). Conversely, Li et al. (26) demonstrated that accumulation of inflammatory mast cells contributed to the elimination of a G. lamblia infection in the experimental murine host. Finally, our present data indicated that a T. spiralis infection, and perhaps particularly the intestinal inflammation induced by this infection, favours establishment and maintenance of a G. lamblia population in mice. It is obvious that in natural hosts, a G. lamblia 13 infection often occurs in a situation where the intestinal tissue is inflammed e.g. as consequence of a

“pre-existing“ nematode infection. Nematode infections are evidently widespread in most endemic areas of giardiasis (16, 33) and the phenomenon of a ”pre-existing” nematode infection prior to a G. lamblia infection may therefore have a direct influence on the epidemiological relevance of the disease.

As e.g. demonstrated for nematode infections, the inflammatory process involved in intestinal

(immuno-)pathogenesis can change both the immunological and physiological conditions inside the intestinal environment (15, 22, 25, 27). Accordingly, approaches investigating giardial growth in relation to an eventual intestinal inflammatory reaction will provide novel information on the immunological and physiological functions that promote either resistance or susceptibility to a G. lamblia infection.

ACKNOWLEDGMENTS

We acknowledge U. Brönnimann, H. Sager (Institute of Parasitology), M. Bozzo, E. Garchi, and

E. Rohner for technical support, and T. E. Nash (NIH, Bethesda, Maryland, U.S.A.) for his gift of G. lamblia clone GS/M-83-H7. This work was financed through a grant obtained from the Swiss National

Science Foundation (No. 31-066795.01). 14

REFERENCES

1. Adam, R.D. 1991. The biology of Giardia spp. Microb. Rev. 55:706-732.

2. Aggarwal, A., J. W. Merritt , and T. E. Nash. 1989. Cysteine-rich variant surface proteins of

Giardia lamblia. Mol. Biochem. Parasitol. 32:39-48.

3. Bienz, M., W.J. Dai, M. Welle, B. Gottstein, and N. Müller. 2003. Interleukin-6-deficient mice are

highly susceptible to Giardia lamblia infections but exhibit normal intestinal IgA responses against

the parasite. Infect. Immun. 71:1569-1573.

4. Boitelle, A., C. Di Lorenzo, H.E. Scales, E. Devaney, M.W. Kennedy, P. Garside, and C.E.

Lawrence. 2005. Contrasting effects of acute and chronic gastro-intestinal helminth infections on a

heterologous immune response in a transgenic adoptive transfer model. Int. J. Parasitol. 35:765-

775.

5. Di Prisco, M.C., I. Hagel, N.R. Lynch, R.M. Barrios, N. Alvarez, and R. Lopez. 1993. Possible

relationship between allergic disease and infection by Giardia lamblia. Ann. 70:210-213.

6. Donaldson, L.E., E. Schmitt, J.F. Huntley, G.F. Newlands, and R.K. Grencis. 1996. A critical

role for stem cell factor and c-kit in host protective immunity to an intestinal helminth. Int. Immunol.

8:559-567.

7. Eckmann, L. 2003. Mucosal defences against Giardia. Parasite Immunol. 25:259-270.

15

8. Erlich, J.H, R.F. Anders, I.C. Roberts-Thomson, J.W. Schrader, and G.F. Mitchell. 1983. An

examination of differences in serum antibody specificities and hypersensitivity reactions as

contributing factors to chronic infection with the intestinal protozoan parasite, Giardia muris, in

mice. Aust. J. Exp. Biol. Med. Sci. 61:599-615.

9. Faubert, G.M. 2000. Immune response to Giardia duodenalis. Clin. Microbiol. Rev. 13:35-54.

10. Faulkner, H., N. Humphreys, J.C. Renauld, J. Van Snick, and R. Grencis. 1997. Interleukin-9 is

involved in host protective immunity to intestinal nematode infection. Eur. J. Immunol. 27:2536-

2540.

11. Finkelman, F.D., T. Shea-Donohue, J. Goldhill, C.A. Sullivan, S.C. Morris, K.B. Madden, W.C.

Gause, and J.F. Jr. Urban. 1997. Cytokine regulation of host defense against parasitic

gastrointestinal nematodes: lessons from studies with rodent models. Annu. Rev. Immunol. 15:505-

533.

12. Friend, D.S., N. Ghildyal, K.F. Austen, M.F. Gurish, R. Matsumoto, and R.L. Stevens. 1996.

Mast cells that reside at different locations in the jejunum of mice infected with Trichinella spiralis

exhibit sequential changes in their granule ultrastructure and chymase phenotype. J. Cell Biol.

135:279-290.

13. Friend, D.S., M.F. Gurish, K.F. Austen, J. Hunt, and R.L. Stevens. 2000. Senescent jejunal

mast cells and eosinophils in the mouse preferentially translocate to the spleen and draining lymph

node, respectively, during the recovery phase of helminth infection. J. Immunol. 165:344-352.

14. Galli, S.J., S. Nakae, and M. Tsai. 2005. Mast cells in the development of adaptive immune

responses. Nat. Immunol. 6:135-142.

16

15. Garside, P., M.W. Kennedy, D. Wakelin, and C.E. Lawrence. 2000. Immunopathology of

intestinal helminth infection. Parasite Immunol. 22:605-612.

16. Gendrel, D., J.M. Treluyer, D. Richard-Lenoble. 2003. Parasitic diarrhea in normal and

malnourished children. Fundam. Clin. Pharmacol. 17:189-197.

17. Ghildyal, N., H.P. McNeil, S. Stchschulte, K.F. Austen, D. Silberstein, M.F. Gurish, L.L.

Sommerville, and R.L. Stevens. 1992. IL-10 induces transcription of the gene for mouse mast cell

protease-1, a serine protease preferentially expressed in mucosal mast cells of Trichinella spiralis-

infected mice. J. Immunol. 149:2123-2129.

18. Gottstein, B., P. Deplazes, and I. Tanner. 1993. In vitro synthesized immunoglobulin A from nu/+

and reconstituted nu/nu mice against a dominant surface antigen of Giardia lamblia. Parasitol. Res.

79:644-648.

19. Grencis, R.K., K.J. Else, J.F. Huntley, and S.I. Nishikawa. 1993. The in vivo role of stem cell

factor (c-kit ligand) on mastocytosis and host protective immunity to the intestinal nematode

Trichinella spiralis in mice. Parasite Immunol. 15:55-59.

20. Gurish M.F., P.J. Bryce, H. Tao, A.B. Kisselgof, E.M. Thornton, H.R. Miller, D.S. Friend and

H.C. Oettgen. 2004. IgE enhances parasite clearance and regulates mast cell responses in mice

infected with Trichinella spiralis. J. Immuol. 172: 1139-1145.

21. Kamiya, M., Y. Oku, H. Itayama, and M. Ohbayashi. 1985. Prolonged expulsion of adult

Trichinella spiralis and eosinophil infiltration in mast cell-deficient W/Wv mice. J. Helminthol.

59:233-239.

17

22. Khan, W.I., and S.M. Collins. 2004. Immune-mediated alteration in gut physiology and its role in

host defence in nematode infection. Parasite Immunol. 26:319-326.

23. Keister, D. B. 1983. Axenic culture of Giardia lamblia in TYI-S-33 medium supplemented with bile.

Trans. R. Soc. Trop. Med. Hyg. 77:487-488.

24. Langford, T.D., M.P. Housley, M. Boes, J. Chen, M.F. Kagnoff, F.D. Gillin, and L. Eckmann.

2002. Central importance of immunoglobulin A in host defense against Giardia spp. Infect. Immun.

70:11-18.

25. Lawrence, C.E. 2003. Is there a common mechanism of gastrointestinal nematode expulsion?

Parasite Immunol. 25:271-281.

26. Li, E., P. Zhou, Z. Petrin, and S. Singer. 2004. Mast cell-dependent control of Giardia lamblia

infections in mice. Infect. Immun. 72:6642-6649.

27. Maizels, R.M., A. Balic, N. Gomez-Excobar, M. Nair, M.D. Taylor, and J.E. Allen. 2004.

Helminth parasites-masters of regulation. Immunol. Rev. 201:89-116.

28. Matossian, R.M., I. Salti, and E. Stephan. 1977. Variation in serum immunoglobulin levels in

acute trichinosis. J. Helminthol. 51:1-4.

29. McGee, D.W., C.O. Elson, and J.R. McGee. 1993. Enhancing effect of cholera toxin on

interleukin-6 secretion by IEC-6 intestinal epithelial cells: mode of action and augmentation effect of

inflammatory cytokines. Infect. Immun. 61:970-978.

18

30. Mitchell, G.F, R.F. Anders, G.V. Brown, E. Handman, I.C. Roberts-Thompson, C.B.,

Chapman, K.P. Forsyth, L.P. Kahl, and K.M. Cruise. 1982. Analysis of infection characteristics

and antiparasite immune responses in resistant compared with susceptible hosts. Immunol. Rev.

61:137-188.

31. Morakote, N., K. Sukhavat, C. Khamboonruang, V. Siriprasert, S. Suphawitayanukul, and W.

Thamasonthi. 1992. Persistence of IgG, IgM, and IgE antibodies in human trichinosis. Trop. Med.

Parasitol. 43:167-169.

32. Müller, N., and N. von Allmen. 2005. Recent insights into the mucosal reactions associated with

Giardia lamblia infections. Int. J. Parasitol., in press.

33. Owen, I.L. 2005. Parasitic zoonoses in Papua New Guinea. J Helminthol. 79:1-14.

34. Prophet, E.D., B. Mills, J.B. Arrington, and L.H. Sobin. 1994. Laboratory methods in

histotechnology of the Armed Forces Institute of Pathology. 2nd ed. American Registry of

Pathology, p. 146.

35. Roberts-Thompson, I.C., D.I. Grove, D.P. Stevens, and K.S. Warren. 1976. Suppression of

giardiasis during the intestinal phase of trichinosis in the mouse. Gut 17:953-958.

36. Rosenberg, E.B., S.H. Polmar, and G.E. Whalen. 1971. Increased circulating IgE in trichinosis.

Ann. Intern. Med. 75:575-578.

37. Singer, S.M., and T.E. Nash. 2000. T-cell-dependent control of acute Giardia lamblia infections in

mice. Infect. Immun. 68:170-175.

19

38. Stäger, S., B. Gottstein, and N. Müller. 1997. Systemic and local antibody response in mice

induced by a recombinant peptide fragment from Giardia lamblia variant surface protein (VSP) H7

produced by a Salmonella typhimurium vaccine strain. Int. J. Parasitol. 27:965-971.

39. Stäger, S., and N. Müller. 1997. Giardia lamblia infections in B-cell-deficient transgenic mice.

Infect. Immun. 65:3944-3946.

40. Urban, J.F. Jr., L. Schopf, S.C. Morris, T. Orekhova, K.B. Madden, C.J. Betts, H.R. Gamble, C.

Byrd, D. Donaldson, K. Else, and F.D. Finkelman. 2000. Stat6 signaling promotes protective

immunity against Trichinella spiralis through a mast cell- and -dependent mechanism. J.

Immunol. 15:2046-2052.

41. Venkatesan, P., R.G. Finch, and D. Wakelin. 1997. A comparison of mucosal inflammatory

responses to Giardia muris in resistant B10 and susceptible BALB/c mice. Parasite Immunol.

19:137-143.

42. Watanabe, N., F. Bruschi, and M. Korenaga. 2005. IgE: a question of protective immunity in

Trichinella spiralis infection. Trends Parasitol. 21:175-178.

43. Zhou, P., E. Li, N. Zhu, J. Robertson, T. Nash, and S.M. Singer. 2003. Role of interleukin-6 in

the control of acute and chronic Giardia lamblia infections in mice. Infect. Immun. 71:1566-1568.

20

FIGURE LEGENDS

Fig. 1. Determination of intestinal G. lamblia trophozoite (A), intestinal eosinophil (B), and intestinal mast cell numbers (C), and serum MMCP-1 (D), serum total IgE (E), and intestinal anti-G. lamblia IgA levels in uninfected control (contr.), G. lamblia- (G.l.), T. spiralis- (T.s.), and T. spiralis/ G. lamblia-

(T.s./G.l.) infected mice. Samples were taken at week –1 (time-point of T. spiralis infection as indicated by an arrow with an open head), week 0 (time-point of G. lamblia infection as indicated by an arrow with a closed head), and at weeks 1, 2, 3, and 6 post G. lamblia infection. Values are given as means

(+ standard errors) and highly significant differences between the values representing G. lamblia- and

G. lamblia/T. spiralis-infected animal groups were assessed by using the Student's t-test (* P<0.01).

Fig. 2. Determination of intestinal G. lamblia trophozoite (A), intestinal eosinophil (B), and intestinal mast cell numbers (C), and serum MMCP-1 (D), serum total IgE (E), and intestinal anti-G. lamblia IgA levels in uninfected control (contr.), G. lamblia- (G.l.), and G. lamblia/T. spiralis- (G.l./T.s.) infected mice. Samples were taken at week 0 (time-point of G. lamblia infection as indicated by an arrow with a closed head), week 1 (time-point of T. spiralis infection as indicated by an arrow with an open head), and at weeks 2 and 3 post G. lamblia infection. Values are given as means (+ standard errors) and highly significant differences between the values representing G. lamblia- and G. lamblia/T. spiralis- infected animal groups were assessed by using the Student's t-test (* P<0.01).

Fig. 3. Detection of intestinal mast cells and IL-6 producing cells in G. lamblia infected (A) and uninfected control (B) mice. Enzyme-histochemical staining for mast cell chymase resulted in significantly higher numbers of mast cells infected (A, panel a) mice compared to the controls (B, panel a). Tissue areas with mast cell accumulations are indicated by arrows. Immunohistochemical staining using an anti-murine IL-6 antibody revealed a mild to moderate IL-6 expression in the epithelial cells of infected (A, panel b) and control (B, panel b) mice. No IL-6 expression was detected in the areas of mast cell accumulation. Note the absence of IL-6 positivity of the epithelial cells after neutralization of the IL-6 antibody (A and B, panels c). Areas representing epithelial cell layers are indicated by arrows (magnification x 200).

40 10 * * A contr. D* G.l. 8 * 30 * T.s. T.s./G.l. 6 * )/cm )/cm duodenum 4 * 20 4

troph. troph. (x10 10 * (ng/ml) serum MMCP-1 2

0 0 G. lamblia -1 0 1 2 3 4 5 6 -1 0 1 2 3 4 5 6 week week 400 600 B E 500

2 300 * 400 * 200 300 * * * 200 eosinophils/mm 100

serum serum total IgE (ng/ml) 100

0 0 -1 0 1 2 3 4 5 6 -1 0 1 2 3 4 5 6 week week 600 0.3

C IgA) F 500 * 0.25 2 400 0.2 G. lamblia 300 * 0.15

200 0.1 mastcells/mm (intest. (intest. anti- 100 * 0.05 * 405

0 OD 0 -1 0 1 2 3 4 5 6 -1 0 1 2 3 4 5 6 week week

Fig. 1 12 10 contr. 10 A D G.l. 8 G.l./T.s. * 8 * 6 )/cm )/cm duodenum 4 6 4 4 troph. troph. (x10

serum MMCP-1 (ng/ml) serum MMCP-1 2 2

0 0 G. lamblia 0 1 2 3 0 1 2 3 week week 400 600 B E 500 * 300 2 400

200 300

200 eosinophils/mm 100 serum serum total IgE (ng/ml) 100

0 0 0 1 2 3 0 1 2 3 week week 600 0.3

C IgA) F 500 0.25

2 400 0.2

* G. lamblia 300 0.15

200 0.1 mastcells/mm (intest. (intest. anti-

100 405 0.05 OD 0 0 0 1 2 3 0 1 2 3 week week

Fig. 2 a

b

c

A B

Fig. 3 DTD 5 ARTICLE IN PRESS

International Journal for Parasitology xx (2005) 1–9 www.elsevier.com/locate/ijpara

Invited Review Recent insights into the mucosal reactions associated with Giardia lamblia infections

N. Mu¨ller*, N. von Allmen

Institute of Parasitology, La¨nggass-Str. 122, CH-3012 Bern, Switzerland

Received 20 May 2005; received in revised form 22 July 2005; accepted 27 July 2005

Abstract

Giardia lamblia is an intestinal protozoan parasite infecting humans and various other mammalian hosts. The most important clinical signs of giardiasis are diarrhoea and malabsorption. Giardia lamblia is able to undergo continuous antigenic variation of its major surface antigen, named VSP (variant surface protein). While intestinal antibodies, and more specifically anti-VSP IgA antibodies, were proven to be involved in modulating antigenic variation of the parasite the participation of the local antibody response in control of the parasite infection is still controversial. Conversely, previous studies based on experimental infections in mice showed that cellular immune mechanisms are essential for elimination of the parasite from its intestinal habitat. Furthermore, recent data indicated that inflammatory mast cells have a potential to directly, or indirectly, interfere in duodenal growth of G. lamblia trophozoites. However, this finding was challenged by other reports, which did not find a correlation between intestinal inflammation and resistance to infection. Since intestinal infiltration of inflammatory cells and/or CD8CT-cells were demonstrated to coincide with villus-shortening and crypt hyperplasia immunological reactions were considered to be a potential factor of pathogenesis in giardiasis. The contribution of physiological factors to pathogenesis was essentially assessed in vitro by co-cultivation of G. lamblia trophozoites with epithelial cell lines. By using this in vitro model, molecular (through surface lectins) and mechanical (through ventral disk) adhesion of trophozoites to the epithelium was shown to be crucial for increased epithelial permeability. This phenomenon as well as other Giardia-induced intestinal abnormalties such as loss of intestinal brush border surface area, villus flattening, inhibition of disaccharidase activities, and eventually also overgrowth of the enteric bacterial flora seem to be involved in the pathophysiology of giardiasis. However, it remains to be elucidated whether at least part of these pathological effects are causatively linked to the clinical manifestation of the disease. q 2005 Published by Elsevier Ltd on behalf of Australian Society for Parasitology Inc.

Keywords: Giardia lamblia; Giardiasis; Immunity; Pathogenesis; Intestinal immune reactions; Intestinal physiological reactions; Intestinal inflammation; Antigenic variation

1. Giardia lamblia and giardiasis The life cycle of G. lamblia includes two major stages: the proliferative trophozoite (Fig. 1, panels a and b) and the Giardia lamblia (syn. Giardia duodenalis, Giardia non-proliferative, infectious cyst (Fig. 1, panel b). Infection intestinalis) is a common intestinal dwelling protozoan occurs upon peroral ingestion of cysts. Following excysta- and causes diarrhoea in humans and animals worldwide. tion in the upper part of the small intestine, cysts release an One of the major sources of infection in humans is excyzoite, which only represents a transient stage of the life contaminated water but there is also strong evidence that cycle and immediately divides into four trophozoites the parasite is transmitted by person-to-person contact or by (Bernander et al., 2001). During this initial proliferation contact with domestic and wild animals (e.g. reviewed by step, the trophozoites form an adhesive ventral disk (Palm et Adam, 1991; Thompson, 2000). al., 2005). Adherence of the trophozoites to the intestinal wall and, more specifically, to the microvillous brush border of enterocytes, basically occur through the function of the * Corresponding author. Tel.: C41 31 6312 474; fax: C41 31 6312 477. ventral disk but specific receptor–ligand interactions also E-mail address: [email protected] (N. Mu¨ller).

0020-7519/$30.00 q 2005 Published by Elsevier Ltd on behalf of Australian Society for Parasitology Inc. doi:10.1016/j.ijpara.2005.07.008 DTD 5 ARTICLE IN PRESS

2 N. Mu¨ller, N. von Allmen / International Journal for Parasitology xx (2005) 1–9

Fig. 1. Scanning electron micrographs showing ventral surface down-(panels a and b) and up-(panel a) oriented Giardia lamblia trophozoites (TR), and a cyst (CY) (panel b) associated with the intestinal mucosa from an infected mouse. The ventral disk (vd) and flagella (fl) are indicated (images kindly provided by Andrew Hemphill, Institute of Parasitology, Berne, Switzerland).

seem to be involved (Aley and Gillin, 1995; Elmendorf circumvent the host immune response and it may exhibit et al., 2003). physiological functions important for the intestinal survival The symptoms of human giardiasis are highly variable of the parasite within the host (Nash et al., 1991). Antigenic (e.g. reviewed by Adam, 1991). In many individuals, the variation has been extensively studied by using the infection remains asymptomatic whereas other patients G. lamblia clone GS/M-83-H7 as model parasite (e.g. exhibit severe symptoms. The most prominent clinical signs reviewed by Mu¨ller and Gottstein, 1998; Nash, 2002). of giardiasis are abdominal pain, nausea, followed by severe Antigenic variation is thought to play an important role in watery diarrhoea possibly as a consequence of malabsorp- facilitating chronic as well as repeated infections in natural tion. Chronic courses are characterised by recurrent brief or hosts. However, conclusive experimental evidence proving persistent episodes of diarrhoea. In many cases, spontaneous this hypothesis is still not available. Until now, antigenic resolution of the infection occurs after a few weeks but the variation of G. lamblia and the consequences of this disease may also develop into a chronic state. The outcome phenomenon on the course of infection have nearly of the infection is supposed to largely depend on the exclusively been studied in experimental hosts. In this immunological status of the infected individual but non- respect, experimental G. lamblia clone GS/M-83-H7 immunological factors are also involved in the interaction infections in the mouse system is a successful model for between the host and the parasite. Chronicity of the investigating the different parameters associated with infection may be causatively linked to the phenomenon of surface antigen alterations of the parasite (e.g. reviewed antigenic variation of the parasite. by Mu¨ller and Gottstein, 1998; Nash, 2002. Variant surface protein H7 from clone GS/M-83-H7, like various other VSPs investigated so far, induces strong serum (Gottstein et al., 1990; Mu¨ller et al., 1996; Sta¨ger et al., 2. Immunological host reactions against G. lamblia 1997a,b), intestinal (Sta¨ger et al., 1997b; Sta¨ger and Mu¨ller, infections 1997; Mu¨ller and Sta¨ger, 1999), and milk (Sta¨ger et al., 1998; ¨ ¨ G. lamblia In the past decade, the immune response against Muller and Stager, 1999) antibody responses in - G. lamblia has been especially investigated in terms of the infected individuals. In the experimental murine host, the parasite’s ability to continuously change its surface antigen humoral immune response is predominantly directed against coat (e.g. reviewed by Mu¨ller and Gottstein, 1998; Nash, VSP, but in human patients antibodies against invariant 2002). These studies have revealed that antigenic variation antigens are also produced during a G. lamblia infection is associated with a unique family of surface antigens, (Edson et al., 1986; Hassan et al., 2002; Palm et al., 2003; named VSP (variant surface protein). Surface antigen Weiland et al., 2003). As the establishment of a humoral alterations are observed within proliferating populations of immune response in both experimental and natural hosts intestinal trophozoites (Gottstein et al., 1991, 1993; coincides with the elimination of the original variant-type Gottstein and Nash, 1991; Byrd et al., 1994; Mu¨ller et al., population, a function for antibodies in the process of 1996; Mu¨ller and Gottstein, 1998; Mu¨ller and Sta¨ger, 1999, antigenic variation of G. lamblia was proposed (e.g. Nash et al., 2001; Sta¨ger and Mu¨ller, 1997a,b; Sta¨ger et al., reviewed by Mu¨ller and Gottstein, 1998). 1998; Bienz et al., 2001a,b; Von Allmen et al., 2005), as Various studies suggested a central role of the immune well as among individual trophozoites upon release from system in determining the outcome of a Giardia infection non-proliferating cysts (Meng et al., 1993; Sva¨rd et al., but the knowledge on those mechanisms mediating 1998; Von Allmen et al., 2004). The variant surface proteins immunity is still rather rudimentary (e.g. reviewed by have a potential function in antigenic variation to Faubert, 2000; Eckmann, 2003)(Table 1). In the past, DTD 5 ARTICLE IN PRESS

N. Mu¨ller, N. von Allmen / International Journal for Parasitology xx (2005) 1–9 3

Table 1 Possible immunological and physiological mechanisms involved in control of Giardia lamblia infection

Mechanism References (supportive)a References (not supportive)b Local antibody response (control of chronic Sta¨ger and Mu¨ller, 1997; Langford et al., 2002; Bienz et al., 2003; Zhou et al., 2003 infective phase) LoGalbo et al., 1982 T-cell response (control of acute infective Singer and Nash, 2000a Janoff et al., 1988 phase) Mast cell activation Mitchell et al., 1982c; Erlich et al., 1983; Li et al., 2004 Increased IL-6 production Bienz et al., 2003; Zhou et al., 2003; Li et al., 2004 Epithelial nitric oxide production Fernandes and Assreuy, 1997; Eckmann et al., Eckmann et al., 2000 2000 Defensin/cryptdin release by Paneth cells Aley et al., 1994; El Shewy and Eid, 2005 Ayabe et al., 2000 Increased mucin secretion by goblet cells Leitch et al., 1989; Roskens and Erlandsen, Zemian and Gillin, 1985; Gault et al., 1987 2002; (In)direct giardial growth inhibition by enteric Perez et al., 2001, Singer and Nash, 2000b; bacterial flora El-Shewy and Eid, 2005

a References describing data supporting the relevance of the mechanism. b References describing data not supporting the relevance of the mechanism. c Review article. intense investigations on the immunobiology of giardiasis essential for controlling the acute phase of a G. lamblia have been performed by using the Giardia muris/mouse infection; and (ii) in the mouse model neither the Th1 nor model system. Since the results from these immunological the Th2 subset is absolutely essential for protection from G. studies have been major subjects of two recent reviews lamblia. (Faubert, 2000; Eckmann, 2003) the present article focuses Several investigations in the past generated evidence that on experimental data generated in the G. lamblia/mouse mast cells are substantially involved in intestinal elimin- model. In giardiasis, reinfections are common because ation of G. lamblia (Mitchell et al., 1982; Erlich et al., 1983; acquired immunity against G. lamblia is not complete either Li et al., 2004). For example, Li et al. (2004) recently found due to insufficient immune defences or antigenic variation that mast cell-deficient, or -depleted, C57BL/6 mice failed of the parasite. Many studies in natural and experimental to control a G. lamblia clone GS/M-83-H7 infection. In this rodent hosts addressed the question whether antibodies and report, mast cells were considered to be a potential source more specifically, local secretory IgA antibodies, play a role for IL-6. Since in the abovementioned (Li et al., 2004) and in control of the parasite infection. Studies in patients with other studies (Bienz et al., 2003; Zhou et al., 2003) murine selective IgA-deficiency provided conflicting results in that IL-6-deficiency was associated with an increased suscepti- some investigations demonstrated increased incidences of bility to a G. lamblia clone GS/M-83-H7 infection mast cell- infections while others did not (Eckmann, 2003). Previous derived IL-6 was suggested to be important for control of findings in the experimental mouse/G. lamblia GS/M-83-H7 such an infection. Interestingly, the acute phase of giardiasis model indicated that B-cell-deficient animals were unable to was found to be associated with a delayed recruitment of clear Giardia infections (Sta¨ger and Mu¨ller, 1997; Langford intestinal mast cells (Hardin et al., 1997). It is feasible that et al., 2002). Conversely, by comparing the courses of this phenomenon reflects a survival strategy of G. lamblia, infection in IL-6-deficient and wild-type mice, no obvious which retards the mast cell-dependent antigiardial effector correlation between intestinal IgA production and decrease mechanism during the initial phase of the parasite infection. of parasite load was observed at least during the acute Although the investigations listed above could not infective stage (Bienz et al., 2003; Zhou et al., 2003). Data completely dissect the cellular network involved in from an investigation based on the use of different antigiardial immune defence it is evident that antibody- transgenic mouse strains indicated that an as yet unknown independent immune effector mechanisms interfere T-cell-dependent mechanism is essential for controlling the directly, or indirectly, with maintenance and growth of the acute phase of a G. lamblia infection (Singer and Nash, intestinal parasite population particularly during the acute 2000a). In this study, both wild-type and B-cell-deficient phase of the infection (Singer and Nash, 2000a; Langford mice eliminated the majority of intestinal parasites within et al., 2002; Li et al., 2004). However, these results are 1–2 weeks. While T-cell depletion of wild-type mice with inconsistent with findings in human patients suffering from anti-CD4 antibodies prevented elimination of G. lamblia, acquired T-cell deficiencies (Janoff et al., 1988). Compared interferon (IFN) g-, IL-4-, IL4Ra-, and STAT-6-deficient with the immunocompetent control group, the patient group mice were able to control infection in a manner similar to was not found to be more susceptible to giardiasis. In wild-type mice. From this observation, the authors contrast, there are indications that protection against concluded that: (i) a T-cell-dependent mechanism is giardial infections in humans may be associated with anti- DTD 5 ARTICLE IN PRESS

4 N. Mu¨ller, N. von Allmen / International Journal for Parasitology xx (2005) 1–9

Giardia antibody production. Such a correlation was found produced enzymatically and in intestinal epithelial cells the by investigating children with x-linked agammaglobulinae- most important pathway mediating this enzymatic reaction mia, which exhibited a predisposition for severe and chronic involves the inducible NO synthase (iNOS) (Salzmann et giardiasis (LoGalbo et al., 1982). Considering these al., 1996). The existence of such an antigiardial effector controversial findings new experimental strategies will mechanism was assumed because NO was revealed to have to be developed to further elucidate those effector inhibit in vitro growth of the parasite (Fernandes and mechanisms, which mediate immunity against G. lamblia Assreuy, 1997; Eckmann et al., 2000). However, the infections in natural and experimental hosts of the parasite. effectiveness of NO against a giardial infection was questioned by the observation that co-cultivation of G. lamblia trophozoites with human epithelial cells led to 3. Physiological host reaction against G. lamblia a remarkable suppression of the epithelial NO production. infections Eckmann et al. (2000) found that this suppression was the consequence of a depletion of arginine (a substrate for While the immunological processes of the antigiardial cellular NO synthesis) in the culture medium, which was host response has already been intensely investigated, little is caused by a high affinity uptake of the compound by the known about non-immune defences. For example, intestinal parasite. According to the authors, it is feasible that this Paneth cell-derived defensins (Ouellette, 1999), also known competitive effect represents a survival strategy, which as cryptdins, have been proven to display a cytotoxic effect enables G. lamblia to counteract host antiparasitic NO on G. lamblia in vitro (Aley et al., 1994) but no conclusive production within the intestinal habitat of the parasite. data are available regarding the physiological significance of Besides epithelial cells, IFNg-activated macrophages act as these antimicrobial peptides in vivo. G. lamblia seems not to a source for NO. In experimental murine giardiasis, IFNg be able to induce release of defensins in ex vivo-maintained seemed to contribute to the relative resistance of B10 mice small intestinal crypts (Ayabe et al., 2000). This finding against the parasite infection but according to the data questioned the relevance of defensins in the host defence generated by Venkatesan et al. (1997) NO production did directed against the parasite. However, a recent study not contribute to this relative resistance. This observation provided preliminary evidence for the existence of an may also confirm the above assumption that Giardia has indirect antigiardial effect of these bioactive peptides developed a strategy to inactivate the NO-mediated attack (Eckmann, 2003). This was achieved by analysing the course of the host. However, the scenario outlined above is of a G. muris infection in matrix metalloproteinase (MMP)- challenged by the results of a study on G. lamblia infections 7-deficient mice, which are incapable of producing Paneth in IFNg-deficient mice (Singer and Nash, 2000a). Since cell-derived a-defensins. Interestingly, these mice exhibited these mice were able to control infection in a manner similar a significantly lower initial infection load than normal to wild-type mice (see also above) a substantial partici- control mice. According to the authors this finding might pation of IFNg in the antigiardial immune effector have indicated that the lack of a-defensins in MMP-7- mechanisms was excluded. It remains to be clarified deficient mice influenced the abundance and/or composition whether these controversial findings reflected differences of the intestinal microbiota and thus generated a physiologi- in the immunobiology of the G. muris and G. lamblia cal environment, which inhibited Giardia colonization. In species or rather were the consequence of differential this context, another study focusing on the impact of the protocols of the respective infection experiments. enteric flora on the possible parasitocidal function of Paneth In the field of giardiasis research, intestinal mucin cells in Giardia-infected mice also provided interesting secretion is also discussed as a putative physiological results (El-Shewy and Eid, 2005). Here, TEM examinations defence mechanism against the parasite. Mucins are of intestinal specimens from infected animals revealed that glycoproteins that are secreted from intestinal goblet cells some of the trophozoites harboured bacterial endosymbionts and they constitute the intestinal mucus layer (Deplancke and it seemed that only those parasites containing peripheral and Gaskins, 2001). The presence of mucins was found to bacterial inclusions were killed in close proximity of reduce adhesion of pathogens to mucosal surfaces and thus activated Paneth cells. This finding indicated that Paneth may be of particular importance in early life stages because cells, and more specifically Paneth cell-derived defensins, the acquired immune system is not fully functional in the may interfere in intestinal growth of Giardia trophozoites neonatal intestine (Cebra, 1999). Mucin secretion is known (see also below). However, it remains to be elucidated if the to be rapidly, enhanced in response to enteric microorgan- observed phenomenon suggested a causative link between isms (Moncada et al., 2003). Trapping of microorganisms bacterial invasion and subsequent Paneth cell-mediated by mucus and their subsequent removal from the intestine killing of trophozoites, or, rather reflected biologically by peristalsis is thought to influence the intensity of a irrelevant intracellular accumulation of bacteria in dead respective infection (Walker and Owen, 1990). In giardiasis, trophozoites. preliminary data in the gerbil infection model indicated that A further investigation addressed the putative function of increased mucus secretion reduces epithelial attachment of epithelial nitric oxide (NO) as an antigiardial effector. NO is G. lamblia trophozoites and lowers the infection intensity DTD 5 ARTICLE IN PRESS

N. Mu¨ller, N. von Allmen / International Journal for Parasitology xx (2005) 1–9 5

(Leitch et al., 1989). The participation of such a mechanism infiltrating CD8CT-cells (Scott et al., 2004) and a possible in anti-giardial defence was also suggested because involvement of several cytokines others than IL-6 has been commercially available mucins also inhibited attachment reported (Scott et al., 2000). However, the particular host of G. lamblia trophozoites to an artificial surface (Roskens mechanisms mediating the T-cell effects on the microvillus and Erlandsen, 2002). However, in another study, an structure still remain to be investigated. Respective increased attachment was observed when G. lamblia investigations in G. lamblia infections were for a long trophozoites were exposed to a mucin fraction prepared time hampered by the fact that no experimental animal from postmortem samples of human intestine (Zenian and model was available which developed pathological effects Gillin, 1985). Furthermore, human intestinal mucus was in response to the parasite infection. However, a few years shown to stimulate growth of the parasite in vitro (Gault ago Williamson et al. (2000) were able to isolate a bird et al., 1987). Based on the conflicting results from these G. lamblia strain (BRIS/95/HEPU/2041), which caused in vitro studies it is not possible to decide whether mucins significant pathophysiological alterations to intestinal are physiological factors involved in antigiardial host mucosa including villus atrophy, and an increase of goblet defence or rather contribute to the successful colonisation cell and vacuolated epithelial cell populations. of the parasite on the mucosal surface of its intestinal In addition to host immunological components, parasite- habitat. These two possibilities will have to be evaluated by derived factors also contribute to pathogenesis in giardiasis using appropriate infection models, which allow an eventual and investigations demonstrating this participation were correlation between mucin secretion and the course of a often performed in vitro. For many years, co-incubation of G. lamblia infection in the respective experimental host. intestinal pathogens with epithelial cell lines have been successfully used as in vitro models which helped to improve our knowledge on the mechanisms involved in pathogen-induced brush border damage and malabsorption. 4. Intestinal pathogenesis associated with G. lamblia Several human and murine cell lines have been established infections and these exhibit characteristics of the normal intestinal As outlined above, the most important clinical signs of epithelium such as polarisation, tight junction formation, giardiasis are diarrhoea and malabsorption. Although ion transport and regulated synthesis of inflammatory various intestinal abnormalities have been described mediators. Most of these cell lines are of colonic origin (Table 2), the pathophysiology associated with these and thus they cannot completely simulate the epithelial symptoms is still incompletely understood (e.g. reviewed surface of the duodenum, which is the major habitat of by Farthing, 1996, 1997; Eckmann and Gillin, 2001). In G. lamblia trophozoites. However, duodenal and colonic giardiasis, intestinal colonisation by the parasite seems to epithelial cells are very similar from the physiological point cause microvillus shortening (Erlandsen and Chase, 1974; of view. Considering this aspect, co-cultivation of Scott et al., 2000, 2004), villous flattening or atrophy G. lamblia trophozoites with such colonic epithelial cell (Williamson et al., 2000). These abnormalities, possibly in lines was often chosen as an approach to explore various combination with other pathological mechanisms such as molecular and cellular aspects of the process of parasite– reduction of intestinal disaccharidase (Daniels and Belo- host cell interaction (see below). sevic, 1992) and protease (Seow et al., 1993) activities, may Attachment of trophozoites to epithelial cells is essential be a direct cause of diarrhoea in giardiasis. for Giardia colonisation of the intestine and, most likely, In experimental G. muris infections, inhibition of the also results in damage of the intestinal epithelium. In intestinal disaccharidase (sucrase and maltase) activities various studies, it has been demonstrated that surface lectins during the acute infective stage, and loss of intestinal brush are involved in attachment of the parasite to epithelial cells border surface area could be related to the function of (Pegado and de Souza, 1994; Katelaris et al., 1995; Ce´u

Table 2 Possible mechanisms of pathogenesis associated with giardiasis

Mechanism References Reduction in intestinal disaccharidase activity Daniels and Belosevic, 1992; Farthing, 1996a; Scott et al., 2004 Reduction in intestinal protease activities Seow et al., 1993 Disruption of microvillous brush border Farthing, 1996a; Farthing, 1997a; Scott et al., 2000; Scott et al., 2004 Villus shortening or atrophy Farthing, 1996a; Farthing, 1997a; Erlandsen and Chase, 1974; Williamson et al., 2000; Koudela et al., 1998 Crypt hyperplasia Farthing, 1996a; Farthing, 1997a; Koudela et al., 1998 Increased epithelial permeability Buret et al., 2002; Teoh et al., 2000; Chin et al., 2002 Mucosal inflammation Jimenez et al., 2004; Koudela et al., 1998; Oberhuber et al., 1997 Bacterial overgrowth Tandon et al., 1977; Tomkins et al., 1978

a Review article. DTD 5 ARTICLE IN PRESS

6 N. Mu¨ller, N. von Allmen / International Journal for Parasitology xx (2005) 1–9

Sousa et al., 2001). In mixed cultures, Giardia cells were resistant versus susceptible mouse model did not show found in various orientations to epithelial cells but were a correlation between resistance against infection and mostly observed ventral surface down. This indicated: (i) mucosal inflammation (Venkatesan et al., 1997). This finding that surface components may be involved in a primary and is consistent with the results from our recent G. lamblia clone randomly oriented, giardial attachment; but also (ii) that GS/M-83-H7/T. spiralis co-infections in mice (N. von mechanical, ventrally oriented adhesion through the ventral Allmen, N. Mu¨ller, unpublished observation). Here, intesti- disk plays the major role in attachment of the parasite to the nal inflammation initiated by a T. spiralis infection did not mucosal surface at the duodenal site. exhibit antigiardial activity but substantially promoted As demonstrated by Buret et al. (2002) attached proliferation of duodenal G. lamblia trophozoites. G. lamblia trophozoites were able to increase the In a recent study, Jimenez et al. (2004) found that permeability of non-transformed human epithelial cell excretory and secretory antigens (E/S Ags) from G. lamblia layers by disrupting tight junctions. Interestingly, these induced an intestinal pathogenesis, which coincided with effects were reversed by incubation of the mixed culture mucosal inflammation. In BALB/c mice, oral administration with epidermal growth factor (EGF), which reduced of the E/S Ags not only stimulated production of antibodies epithelial colonisation by the parasites. From these with parasitocidal activity but also resulted in histological observations, the authors concluded that: (i) altered alterations within the intestinal tissue that were comparable epithelial permeability may contribute to intestinal patho- to those observed in natural and experimental Giardia physiology in giardiasis; and (ii) EGF may inhibit these infections. These histological abnormalities included eosi- alterations by preventing intestinal colonisation by Giardia, nophilic infiltration, hypercellularity and enterocyte des- and/or by directly inhibiting its cytopathic effects. In the quamation. The authors speculated that these alterations in same co-cultivation system, Giardia-induced epithelial the intestinal tissue might have been promoted by the permeabilisation was demonstrated to be associated with activity of cysteine proteinases present in the E/S Ag distinct rearrangements in the cytoskeleton of the epithelial fraction. This scenario was suggested on the basis of the cells (Teoh et al., 2000). In addition, increased epithelial mode of the pathophysiological action of cysteine proteases permeability was linked to an apoptotic effect, which in amebic colitis. Here, parasite-derived cysteine protein- disrupted the epithelial barrier function in a caspase-3- ases not only enzymatically disrupt the intestinal epithelium dependent manner (Chin et al., 2002). In the respective but also damage the intestinal tissue by stimulating an paper, the authors discussed the possibility that cytoskeletal inflammatory response. The latter process was shown to be disintegration upon exposure of epithelial cells to initiated by the function of extracellular cysteine proteinases G. lamblia is the direct consequence of a caspase-mediated to convert pre-IL-1b (released from lysed epithelial cells) cleavage of cytoskeletal proteins. Interestingly, apoptosis into active IL-1b and thus to establish an IL-1b-mediated was only observed with two of four G. lamblia strains pro-inflammatory reaction in the epithelium (see below). included in the investigations. This finding suggested Based on data generated by Jimenez et al. (2004) it is further studies aimed at the evaluation of a possible feasible that a similar mechanism is involved in pathogen- correlation of epithelial apoptosis in vitro and increased esis associated with giardiasis, at least in patients with a giardial pathogenicity in vivo. history of mucosal inflammation. Mucosal inflammation may also be a factor in the The enteric flora is a further factor that can potentially pathogenesis of giardiasis although many observations of interfere with the process of a G. lamblia infection. Co- natural symptomatic infections are apparently not associated cultivation of G. lamblia and human epithelial cells in with substantial intestinal inflammatory responses (Farthing, presence of commensal lactobacilli constituted a growth 1996; Eckmann and Gillin, 2001). For example, in environment, which resulted in significant inhibition of experimentally infected goat kids, inflammatory infiltration Giardia proliferation (Perez et al., 2001). This growth- in the lamina propria coincided with villus-shortening and inhibitory effect was due to (an) as yet unidentified factor(s) crypt hyperplasia (Koudela and Vitovec, 1998). Further- released by the bacteria into the culture medium. Depending more, human symptomatic giardiasis seems to result on the composition of the enteric flora in Giardia-infected occasionally (3–4% of patients with Giardia-positive individuals, such bacterial compounds may also have an biopsies) in formation of moderate polymorphonuclear and influence on in vivo growth of the parasite. As reported by mononuclear cell (Oberhuber et al., 1997). In the mouse Singer and Nash (2000b), the intensities of G. lamblia model, intestinal mucosal inflammation was particularly infections differed significantly within two isogenic mouse evaluated regarding its importance in resistance to giardiasis. strains that originated from different commercial breeders. By co-infecting mice with G. muris and the nematode When these mice were housed together, resistance to Trichinella spiralis, a significant reduction of the G. muris infection was transferred to normally susceptible mice. By infection was observed (Roberts-Thompson et al., 1976). treatment with antibiotics normally resistant mice were This protective effect may have been causatively linked to the ‘converted’ into susceptible mice. These observations mucosal inflammatory response during the intestinal phase of clearly demonstrated that the interaction of G. lamblia trichinellosis. In contrast, a study on G. muris infections in a with the enteric flora represents a physiological factor, DTD 5 ARTICLE IN PRESS

N. Mu¨ller, N. von Allmen / International Journal for Parasitology xx (2005) 1–9 7 which may modulate the course of the infection. This of G. lamblia infections. The process of inflammation has conclusion was further supported by data from a recent been found to influence not only the immunological, but study, which assessed the interaction between the bacterial also the physiological, environment inside the intestinal flora and Giardia trophozoites in the context of the intestinal habitat of the parasite. Intestinal accumulation of inflam- physiology (El-Shewy and Eid, 2005). Respective exper- matory cells as a consequence of a G. lamblia infection may imentation provided at least preliminary evidence that either benefit (Li et al., 2004), or be detrimental to intestinal Giardia trophozoites were able to incorporate (Oberhuber et al., 1997; Koudela and Vitovec, 1998, bacterial endosymbionts and trophozoites harbouring Jimenez et al., 2004), the host. Accordingly, approaches peripheral bacteria apparently stimulated degranulation of investigating Giardia growth in an inflammatory intestinal intestinal Paneth cells. It seemed that this degranulation environment will provide novel information on the process resulted in a release of lytic peptides (such as immunological and physiological functions that are defensins) that lysed trophozoites in close proximity to the involved in pathogenicity and promote either resistance or activated Paneth cells (see also above). Accordingly, the susceptibility to the parasite infection. enteric flora may be indirectly involved in those physio- logical reactions that interfere with intestinal proliferation of the parasite. Acknowledgements The intestinal co-habitation of Giardia and microbiota has also to be evaluated from the pathophysiological point We thank Andrew Hemphill for careful reading of the of view. Very early studies on giardiasis indicated that in manuscript. Supported by a grant obtained from the Swiss some human patients clinical manifestations of the disease National Science Foundation (No. 31-066795.01) were associated with the presence of increased numbers of aerobic and/or anaerobic bacteria in the upper part of the intestine (Tandon et al., 1977; Tomkins et al., 1978). It is References feasible that in certain cases the pathological effects are not caused by the Giardia infection per se but are the Adam, R.D., 1991. The biology of Giardia spp. Microbiol. Rev. 55, consequence of an overgrowth of commensal bacteria. It 706–732. was recently reported that under normal steady-state Aley, S.B., Gillin, F.D., 1995. Specialized surface adaptations of Giardia conditions commensal bacteria initiate a Toll-like receptor lamblia. Infect. Agents Dis. 4, 161–166. Aley, S.B., Zimmerman, M., Hetsko, M., Selsted, M.E., Gillin, F.D., 1994. signal transduction pathway in epithelial cells, which is Killing of Giardia lamblia by cryptdins and cationic neutrophil crucial for the maintenance of intestinal homeostasis peptides. Infect. Immun. 62, 5397–5403. (Rakoff-Nahoum et al., 2004). However, under conditions Ayabe, T., Satchell, D.P., Wilson, C.L., Parks, W.C., Selsted, M.E., of a bacterial overgrowth related to giardiasis, we cannot Ouellette, A.J., 2000. Secretion of microbicidal alpha-defensins by exclude the possibility that these commensals trigger a pro- intestinal Paneth cells in response to bacteria. Nat. Immunol. 1, 113–118. inflammatory response similar to that observed in patients Bernander, R., Palm, J.E., Sva¨rd, S.G., 2001. Genome ploidy in different with inflammatory bowel disease. stages of the Giardia lamblia life cycle. Cell. Microbiol. 3, 55–62. Bienz, M., Wittwer, P., Zimmermann, V., Mu¨ller, N., 2001. Molecular characterisation of a predominant antigenic region of Giardia lamblia 5. Concluding remarks variant surface protein H7. Int. J. Parasitol. 31, 827–832. Bienz, M., Siles-Lucas, M., Wittwer, P., Mu¨ller, N., 2001. vsp gene expression by Giardia lamblia clone GS/M-83-H7 during antigenic Although scientists in the field of giardiasis have already variation in vivo and in vitro. Infect. Immun. 69, 5278–5285. investigated multiple aspects of the immunological and Bienz, M., Dai, W.J., Welle, M., Gottstein, B., Mu¨ller, N., 2003. physiological interplay between G. lamblia and its host Interleukin-6-deficient mice are highly susceptible to Giardia lamblia intestinal environment the particular processes that deter- infections but exhibit normal intestinal IgA responses against the parasite. Infect. Immun. 71, 1569–1573. mine the outcome of the parasite infection are still Buret, A.G., Mitchell, K., Muench, D.G., Scott, K.G., 2002. Giardia incompletely understood. In the past, most of the relevant lamblia disrupts tight junctional ZO-1 and increases permeability in data were generated either in vitro by co-cultivation of non-transformed human small intestinal epithelial monolayers: effects G. lamblia with different intestinal epithelial cell lines or of epidermal growth factor. Parasitology 125, 11–19. in vivo by analysing G. lamblia infections in natural and Byrd, L.G., Conrad, J.T., Nash, T.E., 1994. Giardia lamblia infections in adult mice. Infect. Immun. 62, 3583–3585. experimental hosts. Future studies will have to elucidate in Cebra, J.J., 1999. Influences of microbiota on intestinal immune system more detail those molecular and cellular mechanisms that development. Am. J. Clin. Nutr. 69, 1046–1051. stimulate the immune system or are involved in the physical Ce´u Sousa, M.C., Goncalves, C.A., Bairos, V.A., Poiares-Da-Silva, J., interaction between the parasite and the intestinal epi- 2001. Adherence of Giardia lamblia trophozoites to Int-407 human thelium. Furthermore, recent data (Williamson et al., 2000; intestinal cells. Clin. Diagn. Lab. Immunol. 8, 258–265. Chin, A.C., Teoh, D.A., Scott, K.G., Meddings, J.B., Macnaughton, W.K., Jimenez et al., 2004; Li et al., 2004) indicated, that a high Buret, A.G., 2002. Strain-dependent induction of enterocyte apoptosis priority should be given to investigations that explore the by Giardia lamblia disrupts epithelial barrier function in a caspase-3- influence of intestinal inflammatory reactions on the course dependent manner. Infect. Immun. 70, 3673–3680. DTD 5 ARTICLE IN PRESS

8 N. Mu¨ller, N. von Allmen / International Journal for Parasitology xx (2005) 1–9

Daniels, C.W., Belosevic, M., 1992. Disaccharidase activity in the small Jimenez, J.C., Fontaine, J., Grzych, J.M., Dei-Cas, E., Capron, M., 2004. intestine of susceptible and resistant mice after primary and Systemic and mucosal responses to oral administration of excretory and challenge infections with Giardia muris. Am. J. Trop. Med. Hyg. secretory antigens from Giardia intestinalis. Clin. Diagn. Lab. 46, 382–390. Immunol. 11, 152–160. Deplancke, B., Gaskins, H.R., 2001. Microbial modulation of innate Katelaris, P.H., Naeem, A., Farthing, M.J.G., 1995. Attachment of Giardia defense: goblet cells and the intestinal mucus layer. Am. J. Clin. Nutr. lamblia trophozoites to a cultured human intestinal cell line. Gut 37, 73, 1131–1141. 512–518. Eckmann, L., 2003. Mucosal defences against Giardia. Parasite Immunol. Koudela, B., Vitovec, J., 1998. Experimental giardiasis in goat kids. Vet. 25, 259–270. Parasitol. 74, 9–18. Eckmann, L., Gillin, F.D., 2001. Microbes and microbial toxins: paradigms Langford, T.D., Housley, M.P., Boes, M., Chen, J., Kagnoff, M.F., Gillin, for microbial-mucosal interactions. I. Pathophysiological aspects of F.D., Eckmann, L., 2002. Central importance of immunoglobulin A in enteric infections with the lumen-dwelling protozoan pathogen Giardia host defense against Giardia spp. Infect. Immun. 70, 11–18. lamblia. Am. J. Physiol. Gastrointest. Liver Physiol. 280, 1–6. Leitch, G.J., Visvesvara, G.S., Wahlquist, S.P., Harmon, C.T., 1989. Eckmann, L., Laurent, F., Langford, T.D., Hetsko, M.L., Smith, J.R., Dietary fiber and giardiasis: dietary fiber reduces rate of intestinal Kagnoff, M.F., Gillin, F.D., 2000. Nitric oxide production by human infection by Giardia lamblia in the gerbil. Am. J. Trop. Med. Hyg. 41, intestinal epithelial cells and competition for arginine as potential 512–520. determinants of host defense against the lumen-dwelling pathogen Li, E., Zhou, P., Petrin, Z., Singer, S., 2004. Mast cell-dependent control of Giardia lamblia. J. Immunol. 164, 1478–1487. Giardia lamblia infections in mice. Infect. Immun. 72, 6642–6649. Edson, C.M., Farthing, M.J.G., Thorley-Lawson, D.A., Keusch, G.T., 1986. LoGalbo, P.R., Sampson, H.A., Buckley, R.H., 1982. Symptomatic An 88,000-Mr Giardia lamblia surface protein which is immunogenic giardiasis in three patients with x-linked agammaglobulinemia. in humans. Infect. Immun. 54, 621–625. El-Shewy, K.A., Eid, R.A., 2005. In vivo killing of Giardia trophozoites J. Pediatr. 101, 78–80. harbouring bacterial endosymbionts by intestinal Paneth cells: an Meng, T.C., Hetsko, M.L., Gillin, F.D., 1993. Antigenic switching of TSA ultrastructural study. Parasitology 130, 269–274. 417, a trophozoite variable surface protein, following completion of the Elmendorf, H.G., Dawson, S.C., McCaffery, J.M., 2003. The cytoskeleton life cycle of Giardia lamblia. Infect. Immun. 61, 5394–5397. of Giardia lamblia. Int. J. Parasitol. 33, 3–28. Mitchell, G.F., Anders, R.F., Brown, G.V., Handman, E., Roberts- Erlich, J.H., Anders, R.F., Roberts-Thomson, I.C., Schrader, J.W., Mitchell, Thompson, I.C., Chapman, C.B., Forsyth, K.P., Kahl, L.P., Cruise, G.F., 1983. An examination of differences in serum antibody K.M., 1982. Analysis of infection characteristics and antiparasite specificities and hypersensitivity reactions as contributing factors to immune responses in resistant compared with susceptible hosts. chronic infection with the intestinal protozoan parasite, Giardia muris, Immunol. Rev. 61, 137–188. in mice. Aust. J. Exp. Biol. Med. Sci. 61, 599–615. Moncada, D.M., Kammanadiminti, S.J., Chadee, K., 2003. Mucin and Toll- Erlandsen, S.L., Chase, D.G., 1974. Morphological alterations in the like receptors in host defense against intestinal parasites. Trends microvillous border of villous epithelial cells produced by intestinal Parasitol. 19, 305–311. microorganisms. Am. J. Clin. Nutr. 27, 1277–1286. Mu¨ller, N., Gottstein, B., 1998. Antigenic variation and the murine immune Farthing, M.J., 1996. Giardiasis. Gastroenterol. Clin. North. Am. 25, 493– response to Giardia lamblia. Int. J. Parasitol. 28, 1829–1839. 515. Mu¨ller, N., Sta¨ger, S., 1999. Periodic appearance of a predominant variant Farthing, M.J.G., 1997. The molecular pathogenesis of giardiasis. antigen-type during a chronic Giardia lamblia infection in a mouse J. Pediatr. Gastroenterol. Nutr. 24, 79–88. model. Int. J. Parasitol. 29, 1917–1923. Faubert, G.M., 2000. Immune response to Giardia duodenalis. Clin. Mu¨ller, N., Sta¨ger, S., Gottstein, B., 1996. Serological analysis of the Microbiol. Rev. 13, 35–54. antigenic heterogeneity of Giardia lamblia variant surface proteins. Fernandes, P.D., Assreuy, J., 1997. Role of nitric oxide and superoxide in Infect. Immun. 64, 1385–1390. Giardia lamblia killing. Braz. J. Med. Biol. Res. 30, 93–99. Nash, T.E., 2002. Surface antigenic variation in Giardia lamblia. Mol. Gault, M.J., Gillin, F.D., Zenian, A.J., 1987. Giardia lamblia: stimulation Microbiol. 45, 585–590. of growth by human intestinal mucus and epithelial cells in serumfree Nash, T.E., Merritt Jr.., J.W., Conrad, J.T., 1991. Isolate and epitope medium. Exp. Parasitol. 64, 29–37. variability in susceptibility of Giardia lamblia to intestinal proteases. Gottstein, B., Nash, T.E., 1991. Antigenic variation in Giardia lamblia: Infect. Immun. 59, 1334–1340. infection of congenitally athymic nude and scid mice. Parasite Nash, T.E., Lujan, H.T., Mowatt, M.R., Conrad, J.T., 2001. Variant-specific Immunol. 13, 649–659. surface protein switching in Giardia lamblia. Infect. Immun. 69, 1922– Gottstein, B., Harriman, G.R., Conrad, J.T., Nash, T.E., 1990. Antigenic 1923. variation in Giardia lamblia: Cellular and humoral immune response in Oberhuber, G., Kastner, N., Stolte, M., 1997. Giardiasis: a histologic a mouse model. Parasite Immunol. 12, 659–673. analysis of 567 cases. Scand. J. Gastroenterol. 32, 48–51. Gottstein, B., Stocks, N.I., Shearer, G.M., Nash, T.E., 1991. Human cellular Ouellette, A.J., Darmoul, D., Tran, D., Huttner, K.M., Yuan, J., Selsted, immune response to Giardia lamblia. Infection 19, 421–426. M.E., 1999. Peptide localization and gene structure of cryptdin 4, a Gottstein, B., Deplazes, P., Tanner, I., 1993. In vitro synthesized differentially expressed mouse paneth cell alpha-defensin. Infect. immunoglobulin A from nu/C and reconstituted nu/nu mice against a dominant surface antigen of Giardia lamblia. Parasitol. Res. 79, 644– Immun. 67, 6643–6651. 648. Palm, J.E., Weiland, M.E., Griffiths, W.J., Ljungstrom, I., Sva¨rd, S.G., Hardin, J.A., Buret, A.G., Olson, M.E., Kimm, M.H., Gall, D.G., 1997. 2003. Identification of immunoreactive proteins during acute human Mast cell hyperplasia and increased macromolecular uptake in an giardiasis. J. Infect. Dis. 187, 1849–1859. animal model of giardiasis. J. Parasitol. 83, 908–912. Palm, D., Weiland, M., McArthur, A.G., Winiecka-Krusnell, J., Cipriano, Hassan, S.M., Maachee, M., Cordova, O.M., Diaz de la Guardia, R., M.J., Birkeland, S.R., Pacocha, S.E., Davids, B., Gillin, F., Linder, Martins, M., Osuna, A., 2002. Human secretory immune response to E., Svard, S., 2005. Developmental changes in the adhesive disk fatty acid-binding protein fraction from Giardia lamblia. Infect. during Giardia differentiation. Mol. Biochem. Parasitol. 141, Immun. 70, 2226–2229. 199–207. Janoff, E.N., Smith, P.D., Blaser, M.J., 1988. Acute antibody responses to Pegado, M.G., de Souza, W., 1994. Role of surface components in the Giardia lamblia are depressed in patients with AIDS. J. Infect. Dis. 157, process of interaction of Giardia duodenalis with epithelial cells 798–804. in vitro. Parasitol. Res. 80, 320–326. DTD 5 ARTICLE IN PRESS

N. Mu¨ller, N. von Allmen / International Journal for Parasitology xx (2005) 1–9 9

Perez, P.F., Minnaard, J., Rouvet, M., Knabenhans, C., Brassart, D., De Sva¨rd, S.G., Meng, T.C., Hetsko, M.L., McCaffery, J.M., Gillin, F.D., 1998. Antoni, G.L., Schiffrin, E.J., 2001. Inhibition of Giardia intestinalis by Differentiation-associated surface antigen variation in the ancient extracellular factors from Lactobacilli: an in vitro study. Appl. Environ. eukaryote Giardia lamblia. Mol. Microbiol. 30, 979–989. Microbiol. 67, 5037–5042. Tandon, B.N., Tandon, R.K., Satpathy, B.K., Shriniwas, B.K., 1977. Rakoff-Nahoum, S., Paglino, J., Eslami-Varzaneh, F., Edberg, S., Mechanism of malabsorption in giardiasis: a study of bacterial flora and Medzhitov, R., 2004. Recognition of commensal microflora by toll- bile salt deconjugation in upper jejunum. Gut 18, 176–181. like receptors is required for intestinal homeostasis. Cell 118, 229–241. Teoh, D.A., Kamieniecki, D., Pang, G., Buret, A.G., 2000. Giardia lamblia Roberts-Thompson, I.C., Grove, D.I., Stevens, D.P., Warren, K.S., 1976. rearranges F-actin and alpha-actinin in human colonic and duodenal Suppression of giardiasis during the intestinal phase of trichinosis in the monolayers and reduces transepithelial electrical resistance. mouse. Gut 17, 953–958. J. Parasitol. 86, 800–806. Roskens, H., Erlandsen, S.L., 2002. Inhibition of in vitro attachment of Thompson, R.C.A., 2000. Giardiasis as a re-emerging infectious disease Giardia trophozoites by mucin. J. Parasitol. 88, 869–873. and its zoonotic potential. Int. J. Parasitol. 30, 1259–1267. Salzmann, A., Denenberg, A.G., Ueta, I., O’Connor, M., Linn, S.C., Szabo, Tomkins, A.M., Drasar, B.S., Bradley, A.K., Williamson, W.A., 1978. C., 1996. Induction and activity of nitric oxide synthase in cultured Bacterial colonization of jejunal mucosa in giardiasis. Trans. R. Soc. human intestinal epithelial monolayers. Am. J. Physiol. 270, 565–573. Trop. Med. Hyg. 72, 33–36. Scott, K.G., Logan, M.R., Klammer, G.M., Teoh, D.A., Buret, A.G., 2000. Venkatesan, P., Finch, R.G., Wakelin, D., 1997. A comparison of mucosal Jejunal brush border microvillous alterations in Giardia muris-infected inflammatory responses to Giardia muris in resistant B10 and mice: role of T lymphocytes and interleukin-6. Infect. Immun. 68, susceptible BALB/c mice. Parasite Immunol. 19, 137–143. 3412–3418. Von Allmen, N., Bienz, M., Hemphill, A., Mu¨ller, N., 2004. Experimental Scott, K.G., Yu, L.C., Buret, A.G., 2004. Role of CD8C and CD4CT infections of neonatal mice with cysts of Giardia lamblia clone GS/M- lymphocytes in jejunal mucosal injury during murine giardiasis. Infect. 83-H7 are associated with an antigenic reset of the parasite. Infect. Immun. 72, 3536–3542. Immun. 72, 4763–4771. Seow, F., Katelaris, P., Ngu, M., 1993. The effect of Giardia lamblia Von Allmen, N., Bienz, M., Hemphill, A., Mu¨ller, N., 2005. Quantitative trophozoites on trypsin, chymotrypsin and amylase in vitro. Parasitol- assessment of sense and antisense transcripts from genes involved in ogy 106, 233–238. antigenic variation (vsp genes) and encystation (cwp 1 gene) of Giardia Singer, S.M., Nash, T.E., 2000a. T-cell-dependent control of acute Giardia lamblia clone GS/M-83-H7. Parasitology 130, 389–396. lamblia infections in mice. Infect. Immun. 68, 170–175. Singer, S.M., Nash, T.E., 2000b. The role of normal flora in Giardia Walker, R.I., Owen, R.L., 1990. Intestinal barriers to bacteria and their lamblia infections in mice. J. Infect. Dis. 181, 1510–1512. toxins. Annu. Rev. Med. 41, 393–400. Sta¨ger, S., Mu¨ller, N., 1997. Giardia lamblia infections in B-cell-deficient Weiland, M.E., Palm, J.E., Griffiths, W.J., McCaffery, J.M., Sva¨rd, S.G., transgenic mice. Infect. Immun. 65, 3944–3946. 2003. Characterisation of alpha-1 giardin: an immunodominant Giardia Sta¨ger, S., Felleisen, R., Gottstein, B., Mu¨ller, N., 1997. Giardia lamblia lamblia annexin with glycosaminoglycan-binding activity. Int. variant surface protein H7 stimulates a heterogeneous repertoire of J. Parasitol. 33, 1341–1351. antibodies displaying differential cytological effects on the parasite. Williamson, A.L., O’Donoghue, P.J., Upcroft, J.A., Upcroft, P., 2000. Mol. Biochem. Parasitol. 85, 113–124. Immune and pathophysiological responses to different Sta¨ger, S., Gottstein, B., Mu¨ller, N., 1997. Systemic and local antibody strains of Giardia duodenalis in neonatal mice. Int. J. Parasitol. 30, response in mice induced by a recombinant peptide fragment from 129–136. Giardia lamblia variant surface protein (VSP) H7 produced by a Zenian, A., Gillin, F.D., 1985. Interactions of Giardia lamblia with human Salmonella typhimurium vaccine strain. Int. J. Parasitol. 27, 965–971. intestinal mucus: enhancement of trophozoite attachment to glass. Sta¨ger, S., Gottstein, B., Sager, H., Jungi, T.W., Mu¨ller, N., 1998. Influence J. Protozool. 32, 664–668. of antibodies in mother’s milk on antigenic variation of Giardia lamblia Zhou, P., Li, E., Zhu, N., Robertson, J., Nash, T., Singer, S.M., 2003. Role in the murine mother-offspring model of infection. Infect. Immun. 66, of interleukin-6 in the control of acute and chronic Giardia lamblia 1287–1292. infections in mice. Infect. Immun. 71, 1566–1568. von Allmen Nicole Eva - 93 -

von Allmen Nicole Eva - 94 -

9 Acknowledgements

My Ph.D. time at the University of Bern is over and I would like to express my gratitude to the following persons who contributed in different ways to its successful completion.

My first thank goes to Norbert Müller for his intense and professional supervision. I furthermore appreciated very much the patient guidance and excellent technical advice of Marianne Bienz. Thanks a lot also to Selina Christen and Géraldine Rühle for friendly team work during their stay as diploma students. Very warm thanks go to Malin Weiland for her competent introduction into the in vitro encystation procedure, for many interesting discussions and for the funny time which we had together!

I am grateful to Prof. Bruno Gottstein for giving me the opportunity to do my diploma work at the Institute of Parasitology in Bern. Furthermore, I would like to thank Prof. Thomas Seebeck, as well as Prof. Staffan Svärd for taking the responsibilities as my examination experts. Many thanks to Monika Welle and Ursula Forster for their professional instructions and patient help in the immune histological examination and evaluation. I thank André Schneider for his engagement in Northern Blot analysis. Special thanks for technical instructions and constant help in the daily lab work go to Ursula Brönnimann, Elisabeth Frei, Christian Gianinazzi, Nadine Keller, Silvia Marthy, Arunasalam Naguleswaran, Sandra Núñez, Marc Schild, Philippe Stünzi, Nathalie Vonlaufen. Thanks also to Heinz Sager for his professional support in solving computer technical problems and evaluation of results, as well as to Andrew Hemphill for beautiful EM-images from Giardia.

Finally my warmest thanks to my whole family for their enthusiastic and patient support, and their advice and encouragement during all my study and Ph.D. time!

This project was supported by a grant of the Swiss National Science Foundation (Number: 31-006795.01).

von Allmen Nicole Eva - 95 -

10 Curriculum Vitae

PERSONAL DATA:

Name: Nicole Eva von Allmen Date of birth: 20.07.1975 Nationality: Swiss Working address: Institute of Parasitology, Faculty of Veterinary Medicine, University of Bern Länggassstrasse 122, 3012 Bern +41 31 631 24 80 [email protected]

EDUCATION: 1982-1991 Primary school, Zollikofen 1991-1993 Continuation course Muristalden, Bern 1993-1998 Teacher training Muristalden, Bern 1998-2002 Studies in Biology with emphasis on Microbiology / Immunology, University of Bern 2002-2003 Diploma work at the Institute of Parasitology, Faculty of Veterinary Medicine, University of Bern Supervisor: Prof. Dr. N. Müller 2003-2005 Ph.D. thesis at the Institute of Parasitology, Faculty of Veterinary Medicine, University of Bern Supervisor: Prof. Dr. N. Müller

DIPLOMA:

1998 Diploma for primary school teacher 2003 lic. phil. nat.