Orally Administered Giardia duodenalis Extracts Enhance an Antigen-Specific Antibody Response

Abstract

We have identified novel adjuvant activity in specific cytosol fractions from trophozoites of Giardia isolate BRIS/95/HEPU/2041 (J. A. Upcroft, P. A. McDonnell, and P. Upcroft, Parasitol. Today, 14:281–284, 1998). Adjuvant activity was demonstrated in the systemic and mucosal compartments when Giardia extract was coadministered orally with antigen to mice. Enhanced antigen-specific serum antibody responses were demonstrated by enzyme-linked immunosorbent assay to be comparable to those generated by the “gold standard,” mucosal adjuvant cholera toxin. A source of adjuvant activity was localized to the cytosolic component of the parasite. Fractionation of the cytosol produced fraction pools, some of which, when coadministered with antigen, stimulated an enhanced antigen-specific serum response. The toxic component of conventional mucosal adjuvants is associated with adjuvant activity; therefore, in a similar way, the toxin-like attributes of BRIS/95/HEPU/2041 may be responsible for its adjuvanticity. Complete characterization of the adjuvant is under way.

Vaccination, conventionally by the parenteral route, has proven to be the most effective means of protecting a population and individuals from infectious disease. More recently, focus on the dangers involved with needle injection, including multiple use, hepatitis, and human immunodeficiency virus infection, and the expense and fear of needles, particularly in Third World countries, have dictated the need for alternative vaccination routes, such as the practical, noninvasive oral route.

Oral delivery requires that the vaccine preparation survive stomach acid and digestive enzymes to arrive intact at the desired site on the digestive tract and to overcome the phenomenon of oral tolerance. Successful delivery is frequently dependent on coadministration of an adjuvant for stimulation of an immune response. Traditionally, orally delivered or mucosal adjuvants have included bacterial ligands and toxins, such as Escherichia coli labile toxin (LT), cholera toxin (CT), and CT B subunit (CT-B) (22). These adjuvants have the ability to bind to intestinal epithelial cells and abrogate oral tolerance to coadministered or covalently coupled antigens (14). However, due to their toxicity, neither CT nor LT is suitable for human use. Alternative mutant toxin molecules (22), including recombinant forms (29) and fusion proteins (1, 30), have been derived, but are often not as effective mucosally when lacking the toxic A1 subunit, with maximal mucosal immunoglobulin A (IgA) responses achieved in the presence of whole CT (13, 22).

Additionally, a number of inert delivery systems, including gelatin capsules (21), microspheres or microparticles (15, 18), and bioadhesive preparations (25), improve vaccine efficacy by allowing intact delivery of the vaccine to the gut mucosa, but are generally without adjuvant activity. Cytokines (2, 8), recombinant bacterial and viral vectors (11, 12), oligodeoxynucleotides (17, 20), immune response-stimulatory complexes (ISCOMS) and lipid derivatives (3, 4, 26), alkyl-polyacrylate esters (16), and other candidate adjuvants and delivery mechanisms (35), including transgenic vegetables (28), have been assessed for mucosal delivery in animal models, with some advancing to human trials. However, the only adjuvants currently approved for human use are aluminum salts and MF59 (24), and the only vaccines routinely used for oral delivery are the live attenuated polio vaccine, the live attenuated Salmonella enterica serovar Typhi vaccine, and the tetravalent rotavirus vaccine, which is currently in doubt (34). Clearly there is much ongoing development of mucosal vaccination strategies. However, the toxicity of existing mucosal adjuvants and the limited range approved for human use, along with other issues (24), including effective dose requirements, stability, and economic measures, warrants additional investigation of putative adjuvants for human administration.

We have previously described an isolate of Giardia duodenalis (BRIS/95/HEPU/2041), referred to as 2041, established from a bird which died of an overwhelming infection with the parasite (31). Isolate 2041 chronically infects mice and produces higher peak parasite loads than the human Giardia isolate BRIS/83/HEPU/106 (referred to as 106) (31, 32). Mice infected with 2041 suffer weight gain impairment, with the most severe weight deficit occurring at the time of maximum parasite load (32). Total serum IgA levels of these mice are threefold higher than those from mice infected with 106 trophozoites, but their specific anti-Giardia serum IgA and IgM levels are significantly decreased (36). Weight loss is often associated with production of toxins by bacteria residing in or in transit through the gut. In general these toxins act by altering electrolyte transport across the intestinal mucosa, resulting in water loss and thus weight loss in animals (23). The toxin gene homologue overexpressed in parasites under drug pressure (10), our Giardia 2041 studies (31, 32, 36), and the body of literature equating some toxins with adjuvants (27, 13) have led to the current assessment of Giardia isolate 2041 adjuvanticity. The toxin gene homologue identified in a G. duodenalis laboratory isolate encodes a protein which has 57% homology with the gene encoding the precursor of the sarafotoxins, a group of snake toxins from the burrowing adder known to cause symptoms similar to those in humans acutely infected with Giardia (10). While identification of this gene in the 2041 isolate is still under investigation, these data demonstrate that Giardia isolates may be capable of producing toxins as part of their infective process, which in turn might display adjuvant activity.

This paper describes initial data on oral administration of antigen with various preparations of Giardia extracts and demonstrates antigen-specific mucosal and systemic adjuvant activity provided by the cytosolic component of Giardia isolate 2041 trophozoites.

MATERIALS AND METHODS

Mice.

Inbred C57BL/6J, hybrid C57BL/6J × CBA, and outbred Quackenbush 5- to 8-week-old female mice and pregnant Quackenbush female mice were obtained from the Animal Resources Centre (Perth, Australia) and housed under PC2 conditions within Queensland Institute of Medical Research. All protocols were approved by the Bancroft Centre Research Ethics committee, according to National Health and Medical Research Council guidelines.

Parasite culture.

Trophozoites of G. duodenalis isolates 2041 (31) and 106 (6) were cultured as previously described (7). Isolate 106 has arbitrarily been designated the standard laboratory isolate.

Parasitic infection.

Quackenbush 3-day-old neonates were each infected via intragastric tubing (7) with 5 × 10⁵ trophozoites from either Giardia 2041 or 106 in 50 μl of sterile phosphate-buffered saline (PBS) or with sterile PBS alone, and small intestines were collected at 2 and 3 weeks postinfection. The small intestine was excised from the duodenum to the cecum, and portions of approximately 10 mm in length were removed from the midpoint of this section, frozen in OCT (Tissuetech) embedding agent on dry ice, and then stored at −70°C.

Immunohistochemistry.

Frozen samples were sectioned in a microtome at −20°C, mounted on treated (1% 3-aminopropyltriethoxy–saline [Sigma] in acetone) glass microscope slides, then fixed in acetone at −20°C for 10 min, and stored desiccated at −20°C. Fixed sections were brought to room temperature (RT) and rehydrated in PBS for 5 min. Endogenous peroxidases were blocked with 0.8% hydrogen peroxide in methanol solution for 3 to 5 min, and the sections were washed in PBS and then blocked with 0.5% bovine serum albumin (BSA) in PBS for 1 h at RT. The sections were then incubated with either goat anti-mouse IgA-horseradish peroxidase (HRP), IgG-HRP, IgM-HRP, or rat anti-mouse IgE-HRP (Southern Biotechnology Associates Inc.), according to the manufacturer's instructions. The detecting antibody was removed with three washes in PBS, the color was developed using 3,3′diaminobenzidine (Sigma Fast DAB) for 3 to 5 min, and the reaction was stopped with PBS washes. The sections were counterstained with alcoholic eosin for 45 s, dehydrated with ethanol, cleared with three washes of xylene, and mounted in dpx (ProSciTech). Antibody-specific staining was identified by light microscopy.

Antigens and adjuvants.

Lyophilized keyhole limpet hemocyanin (KLH) (Sigma) was resuspended in either sodium bicarbonate (pH 9.0) or sterile PBS at the time of inoculation. Inactivated influenza virus (Flu) strain A/Texas/36/91 (3.25 mg ml⁻¹) was supplied by Lorena Brown (Department of Microbiology and Immunology, Melbourne University, Melbourne, Australia) and stored at 4°C. CT (Sigma) from Vibrio cholerae was reconstituted to 2 mg ml⁻¹ using sterile Milli-Q water and stored at 4°C.

Giardia trophozoites were harvested (36) from confluent cultures, counted, and aliquoted at a concentration of 10⁸ ml⁻¹ prior to snap freezing and storage at −70°C. Trophozoites were identified as different when a new stabilate was used to initiate the culture (2041, 2041B, 2041C, or 2041L). To prepare cytosolic extract, trophozoites were freeze-thawed two to three times in liquid nitrogen and sonicated on ice until no intact parasites were evident. This preparation was centrifuged at 140,000 × g for 40 min, and the supernatant (cytosolic extract) was removed, aliquoted, and snap frozen prior to storage at −70°C. The remaining pellet (membrane fraction) was resuspended in PBS, microcentrifuged, resuspended in an equivalent volume of PBS to that of the cytosolic extract, aliquoted, and snap frozen prior to storage at −70°C. To collect secretory product, 10⁷ trophozoites were harvested, washed three times with PBS, and resuspended in PBS for 1 to 5 h at 37°C. Following incubation, the trophozoites were removed by centrifugation (trophozoite harvest method), and the supernatant was collected and centrifuged at 1,000 × g for 10 min at 4°C. The supernatant (secreted product) was collected, aliquoted, and stored at −70°C. Cytosolic fractions were prepared by Isabel Roberts (CSL Limited, Melbourne, Australia) from cytosolic extract using a Q-Sepharose HPXK 16/5 column. Parasite protein was eluted, producing 72 2.5-ml fractions, which were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), using 12% polyacrylamide, and proteins were visualized by silver staining. There was a great diversity of clearly separated proteins seen between fractions, best exemplified by the 16 fractions shown in Fig. 1. Fractions were pooled into 12 groups as follows; pool 1, fractions 6 to 11; pool 2, fractions 12 to 15; pool 3, fraction 26; pool 4, fraction 27; pool 5, fraction 28; pool 6, fractions 29 to 31; pool 7, fractions 33 to 34; pool 8, fractions 36 to 38; pool 9, fractions 39 to 45; pool 10, fractions 47 to 54; pool 11, fractions 56 to 63; and pool 12, fractions 69 to 72. Pooled fractions were designated Fx (fraction pool) 1 to 12.

FIG. 1.

Giardia 2041 cytosolic fractions separated on SDS-PAGE gels (12% acrylamide) and silver stained demonstrate diverse protein profiles. M, molecular mass markers (in kilodaltons); lane 1, whole cytosolic extract; lane 2, fraction 7; lane 3, fraction 10; lane 4, fraction 12; lane 5, fraction 13; lane 6, fraction 15; lane 7, fraction 26; lane 8, fraction 27; lane 9, fraction 28; lane 10, fraction 29; lane 11, fraction 31; lane 12, fraction 33; lane 13, fraction 32; lane 14, fraction 36; lane 15, fraction 37; lane 16, fraction 39; lane 17, fraction 40.

Inoculation regimen.

KLH (0.5 to 5 mg per mouse) or Flu (100 μg per mouse) antigen was prepared alone or with CT (1 or 10 μg per mouse) or with Giardia preparations. In every case the amount of Giardia extract or secreted product administered to individual mice was derived from 10⁷ trophozoites. This value is the maximum number of isolate 106 trophozoites which colonize a mouse intestine in the standard laboratory mouse infection model (32). KLH prepared in combination with each of the 12 Giardia cytosolic fraction pools or KLH alone was dialyzed against PBS for 2 h at 4°C to produce a preparation suitable for administration to mice. Subsequently only the high-salt-containing preparations (KLH plus Fx12 and KLH plus Fx7) were dialyzed.

Inocula were prepared in sterile PBS or sodium bicarbonate (pH 9.0) to neutralize stomach acids, and 100 to 200 μl was administered per os (p.o.) to individual mice using Terumo 18-gauge drawing-up needles or via intragastric tubing (7). Mice inoculated via tubing were anesthetized intraperitoneally (i.p.) with 100 μl each of 20% pentobarbitone sodium (Rhône Mérieux) in sterile PBS. For each experiment, mice were dosed using one of three regimens: (i) day 0 and 1 week, (ii) day 0, 1 week, and 3 weeks; and (iii) day 0 and 1, 4, and 6 weeks.

Sample collection and processing.

Preimmune fecal and blood (tail vein) samples were collected from individual mice prior to inoculation. Fecal pellets (4 to 5) and blood samples (10 to 100 μl) were routinely collected from individual mice on a weekly or biweekly basis. Blood, saliva (10 to 100 μl), small intestine (gut) washings (2 ml), and whole small intestines (guts) were collected from each mouse on experiment termination.

Fecal pellets were collected into preweighed tubes on ice, either for individual mice or as a pooled group, and processed in 300 μl of 0.1-mg ml⁻¹ soybean trypsin inhibitor (STI) (Sigma) prepared in 50 mM EDTA for each 100 mg of sample. An equivalent volume of cold PBS was added, and a suspension was produced by vortexing. The samples were then centrifuged at 400 × g for 10 min at 4°C, the supernatant was removed, and 10 μl ml⁻¹ of 100 mM phenylmethylsulfonyl fluoride (PMSF) (Sigma) in 95% ethanol was added. The samples were again vortexed and centrifuged at 4°C. The supernatant was removed, snap frozen on dry ice, and stored at −70°C.

Blood was collected from the tail vein of each mouse or, on termination, from the axillary vessels, or by heart puncture, as individual samples or pooled to form a group sample and allowed to clot overnight at 4°C. Samples were then centrifuged at 4°C, and serum was collected, snap frozen on dry ice, and stored at −70°C. For initial experiments (Table 1, Fig. 2, and Fig. 4), blood collected from individual mice of each inoculation group was pooled and diluted to 5% with sterile PBS prior to storage at −70°C.

TABLE 1.

Giardia trophozoite adjuvanted KLH-specific antibody response^a

Inoculum	Titer
	Ser IgG	Ser IgA	Fec IgA
KLH	6500	320	92
KLH + 2041	20,000	10,240	400

^aGroups of four to six Quackenbush mice were inoculated with 5 mg of KLH alone or with 10⁷ 2041 trophozoites on days 0, 7, and 21. Pooled fecal (Fec) and serum (Ser) samples from each group were collected on day 31. 

FIG. 2.

Trophozoite-adjuvanted anti-KLH serum IgG titers vary relative to Giardia isolate. Groups of four Quackenbush mice were inoculated with 3 mg of KLH alone or with 10⁷ trophozoites on days 0 and 7. Pooled serum samples were collected from each group on days 14, 21, and 28, and anti-KLH serum IgG titers were detected by ELISA.

FIG. 4.

KLH-specific antibody titers from mice inoculated with antigen and crude Giardia extracts. Groups of six Quackenbush mice were inoculated with 5 mg of KLH alone or with Giardia extracts from 10⁷ 2041 trophozoites on days 0, 7, and 21. Serum and fecal samples were collected on days 0, 7, 31, and 35 and pooled for antibody detection by ELISA. Tr, trophozoites; Cyt, cytosolic extract; Mem, membrane extract.

For saliva collection, mice were injected i.p. with 30 to 60 μl of a 1.8 mM solution of pilocarpine (nitrate salt) (Sigma) prepared in sterile PBS to stimulate salivation. The mice salivated readily 4 to 5 min following treatment, and saliva was collected by pipette, snap frozen on dry ice, and stored at −70°C. Prior to assay, saliva samples were thawed on ice, 1 μl of 100 mM PMSF was added per tube, and samples were centrifuged at 4°C.

On experiment termination, the small intestine of each mouse was removed and washed in sterile PBS, and gut contents were flushed with 2 ml of STI (0.1 mg ml⁻¹) and stored on ice. To each 2-ml washing, 10 μl of 100 mM PMSF was added, and the sample was vortexed and centrifuged at 4°C. The supernatant was collected, snap frozen on dry ice, and stored at −70°C.

After removal of contents, guts were snap frozen on dry ice and stored at −70°C. For processing, frozen guts were placed in a mortar containing liquid nitrogen and crushed into a powder. Each sample was transferred into 2 ml of STI (0.1 mg ml⁻¹), vortexed, and centrifuged at 400 × g for 10 min at 4°C. The supernatant was transferred to tubes each containing 10 μl of 100 mM PMSF, then vortexed and centrifuged at 4°C. The supernatant was snap frozen on dry ice prior to storage at −70°C.

Antibody ELISA.

Enzyme-liked immunosorbent assay (ELISA) plates were coated overnight with 2 to 5 μg ml⁻¹ of antigen (KLH, Flu, or CT) or with a 1:100 dilution of Giardia 2041B cytosolic extract in 50 mM carbonate buffer (pH 9.0) (15 mM Na₂CO₃, 35 mM NaHCO₃, 3 mM NaN₃) at 4°C, then washed with PBS–0.05%Tween 20 (Sigma), and blocking solution containing 0.5% BSA (Sigma) in PBS was added for 1 h at RT. Plates were washed, and samples were added in dilution buffer (0.05% BSA, 0.05% Tween 20 in PBS) and serially diluted within the plate. Preimmune sera and fecal samples, and control saliva, gut washes, and gut homogenate samples collected from either Quackenbush or C57BL/6J × CBA control mice were included to determine cutoff values for antibody titers.

Plates were incubated at 37°C for 1 h in a humidified chamber or overnight at 4°C and then washed. Secondary antibodies, goat anti-mouse IgA-HRP and IgG-HRP (Southern Biotechnology Associates Inc.) and unlabeled rabbit anti-mouse IgA and IgG (Southern Biotechnology Associates Inc.), were added to the plate at a 1:2,000 dilution, according to the manufacturer's instructions, followed by a 1-h incubation at 37°C or a 6-h incubation at RT. To assess IgG subsets, HRP-conjugated sheep anti-mouse IgG1, IgG2a, IgG2b, or IgG3 antibodies (The Binding Site, Birmingham, United Kingdom) were added to the plate at a 1:3,000 dilution, according to the manufacturer's instructions, and incubated as described above. Plates were washed, and alkaline phosphatase-conjugated sheep anti-rabbit Ig (Southern Biotechnology Associates Inc.) was added, according to manufacturer's instructions, to plates containing unlabeled antibody for 1 h at 37°C.

3,3′,5,5′-Tetramethylbenzidine (TMB) peroxidase substrate (KPL Laboratories) was added to plates incubated with HRP-conjugated antibodies and allowed to develop for 1 to 5 min. The reaction was stopped with 100 μl per well of 0.5 M sulfuric acid, and optical density (OD) was read at 450 nm using an automated plate reader. Plates containing alkaline phosphatase-conjugated antibodies were washed, and p-nitrophenyl phosphate (1 mg ml⁻¹ in 10% diethanolamine [pH 9.8]) was added for 20 to 30 min at RT. The plates were then read at an OD of 415 nm using an automated plate reader.

Assays to calculate total IgA concentrations proceeded as follows. ELISA plates were coated with unlabeled goat anti-mouse IgA (Southern Biotechnology Associates Inc.) according to the manufacturer's instructions and incubated overnight at 4°C. Plates were washed and blocked as previously described. Mouse IgA isotype control, purified antibody (Southern Biotechnology Associates Inc.), was used as a standard and added to the plate in triplicate over a range of dilutions from 256 to 0.25 ng ml⁻¹. Gut homogenate samples were diluted 1:10,000 and added to the plate in triplicate. Following a 1-h incubation at 37°C, plates were washed, and goat anti-mouse IgA-HRP was added as previously described, and the assay was completed in the usual manner (see above). Total IgA was determined for average gut homogenate OD values using a curve generated from IgA standards, and results were recorded at a 1:10,000 dilution.

Calculation of titers.

Three times the background OD at 415 nm or 450 nm reading from control samples was used as a cutoff to determine antibody titers. For the Flu experiments, titers were calculated using the mean + three standard deviations of the OD₄₅₀ reading from preimmune samples at each dilution as a cutoff value. All serum titers are presented on a logarithmic scale and secretory titers on an arithmetic scale, which best demonstrate the range of titers from these extracts. All mean values have been calculated arithmetically.

RESULTS

Adjuvant activity of Giardia trophozoites.

To assess the adjuvant activity of Giardia trophozoites, we coadministered to mice whole Giardia 2041 or 106 trophozoites with KLH antigen using a two-dose regimen and observed an increase in anti-KLH serum IgG titers over those detected for mice inoculated with KLH alone (Fig. 2). By experiment day 28, anti-KLH serum IgG titers were twofold higher for the 2041 group (20,480) compared with the 106 group (10,240) (Fig. 2). When, using the same parameters, 2041B trophozoites were coadministered with a range of KLH concentrations, anti-KLH serum IgG titers measured on experiment day 28 diminished relative to antigen dose as follows; 5,120 at 3 mg/mouse, 2,560 at 1 mg/mouse, 1,280 at 0.5 mg/mouse, and 1,280 at 0.5 mg/mouse of nonadjuvanted KLH. Increased concentrations of KLH (up to 5 mg/mouse), coadministered with Giardia trophozoites, did not appreciably increase the anti-KLH serum IgG titers (data not shown).

When KLH was coadministered with 2041 trophozoites using a three-dose regimen, enhanced anti-KLH serum IgG and IgA and anti-KLH fecal IgA titers were detected (Table 1). The enhanced anti-KLH fecal IgA titers were indicative of adjuvant activity in the mucosal compartment. To further assess the effect of Giardia trophozoites on the secretion of mucosal antibodies, intestinal sections were taken from neonatal mice infected with 2041 or 106 trophozoites and stained for antibodies at 2 and 3 weeks postinfection. The density of IgA staining was greater at 2 weeks postinfection for trophozoite-infected mice over control mice (Fig. 3). In most sections from 2041-infected mice, the staining density appeared greater compared to the 106-infected mice. By 3 weeks postinfection, IgA staining was detected at the same level in all groups, including the control group (data not shown). There was no specific staining for IgE and IgM antibodies at each time point tested (data not shown). Mucosal IgG was not detected at 2 weeks postinfection but was demonstrable in sections from all inoculation groups at approximately the same density of staining at 3 weeks postinfection (data not shown).

FIG. 3.

Total mucosal IgA measured on the intestinal mucosa of Giardia trophozoite-infected mice. Three-day-old Quackenbush mice were each inoculated p.o. with one dose of either Giardia 2041 (C) or 106 (B) trophozoites (5 × 10⁵ per mouse) or with PBS (A) as a control. The small intestine was collected at 14 days postinfection, and IgA was bound with goat anti-mouse IgA-HRP antibody. The bar in each panel represents 100 μm.

Adjuvant activity of Giardia 2041 crude extracts.

Using a freeze-thaw and sonication technique, Giardia trophozoites were fractionated into membrane and cytosolic extracts. We coadministered 2041 trophozoites, cytosolic extract, or membrane fraction with KLH antigen using a three-dose regimen. Anti-KLH serum IgG titers peaked on experiment day 28 (Fig. 4A), with titers from the cytosolic extract and trophozoite groups (20,000) being greater than from the membrane fraction (2,000) and KLH alone (6,500) groups. An anti-KLH serum IgG1 and IgG2b response was detected, while there was no measurable anti-KLH serum IgG2a or IgG3. Anti-KLH serum IgG1 peaked by day 31 and anti-KLH serum IgG2b by day 35 in the trophozoite- and cytosolic extract-adjuvanted groups (data not shown).

An anti-KLH serum IgA response peaked on day 28, with titers from the cytosolic extract and trophozoite groups (10,240) being greater than those measured from the membrane fraction and KLH-alone groups (320) (Fig. 4B). The enhanced anti-KLH serum IgG and IgA responses remained constant until at least experiment day 35.

An anti-KLH fecal IgA response peaked by experiment day 31, with the highest titer measured from the 2041 cytosolic extract group (3,200) (Fig. 4C). Although the anti-KLH fecal IgA titer was lower in the trophozoite group (400), it was at least four times greater than the titer measured from the membrane fraction and KLH-alone groups (Fig. 4C). Comparatively, when 106 cytosolic extract was coadministered to mice with antigen, there was no enhancement of secretory IgA titers over those recorded for the antigen-alone group (data not shown).

Secreted product from 2041 and 106 trophozoite cultures was coadministered to mice with KLH antigen, but resulted in no significant increase in anti-KLH serum or fecal antibody titers over those from the group administered KLH alone (data not shown).

Reproducibility of Giardia 2041 adjuvant activity.

Our findings thus far have demonstrated that the cytosolic fraction of Giardia 2041 trophozoites, rather than the membrane fraction or secreted product, is the predominant source for adjuvant activity. We subsequently assessed the reproducibility of this activity through comparing antibody responses in mice administered antigen and different batches of 2041 cytosolic extract, measuring antibody titers from individual mice rather than from pooled groups, and comparing the level of Giardia adjuvant activity with the “gold standard” control adjuvant, CT.

Three batches of Giardia 2041 trophozoites were used to produce cytosolic extracts (2041C, 2041B, and 2041L), which were administered to mice with KLH using a two-dose regimen, and directly compared with CT for enhancement of anti-KLH serum IgG titers. By experiment day 20, anti-KLH serum IgG titers from the cytosolic extract-adjuvanted groups were on average five to six times greater (50,000 to 58,000) than those from the KLH alone group (≈10,000). The average anti-KLH serum IgG titer from the CT-adjuvanted group was three to four times greater (200,000) than the cytosolic extract-adjuvanted group's and 20 times greater than the KLH-alone group's. Earlier, on experiment day 13, a similar pattern of results was seen, but with some differences in response between the cytosolic extract-adjuvanted groups. Overall, the different batches of cytosolic extract produced similar results and demonstrated adjuvant activity.

Variation in Giardia adjuvant activity with antigen.

To confirm that the enhanced antibody responses seen with administration of KLH and Giardia crude extracts were not confined to one antigen, we repeated our studies using influenza virus (Flu). In contrast to KLH, there was no significant increase in anti-Flu serum IgG titers in the Giardia cytosolic extract- or CT-adjuvanted groups measured 1 week after three doses (data not shown). Flu is a strong antigen in its own right, and after four doses, by experiment day 58, average anti-Flu serum IgG titers had increased in all groups, with only the CT-adjuvanted group titer (45,900) above that of the Flu-alone group (24,400).

Anti-Flu serum IgA titers measured at 1 week after three doses were low in comparison with those measured for KLH, with the highest average anti-Flu serum IgA titer, 675, recorded for the CT-adjuvanted group. The nonadjuvanted Flu-alone group titer was 281. Subsequent to four doses, by experiment day 58, the average anti-Flu serum IgA titers measured from the CT-adjuvanted (1,505) and cytosolic extract-adjuvanted (1,370) groups were enhanced compared to the Flu-alone group (613) (Fig. 5A).

FIG. 5.

Giardia 2041 adjuvants a Flu-specific serum (A) and secretory (B) IgA response. Groups of 8 to 10 C57BL/6J mice were inoculated with 100 μg of Flu alone or with Giardia 2041 cytosolic extract (extracted from 10⁷ trophozoites) or 10 μg of CT on days 0, 7, 28, and 42. On day 58, individual serum samples were collected and gut homogenates were prepared from the excised small intestine of individual mice. Samples were assayed for the presence of Flu-specific antibodies using ELISA. Results are presented for individual mice, with the average group titer represented as a bar. Cyt, cytosolic extract.

To assess the Flu-specific secretory response, we measured anti-Flu secretory IgA in extracts of saliva, fecal pellets, gut washings, and gut homogenates at 1 to 2 weeks subsequent to dose 4. A variable number of responders were identified from each inoculation group when saliva, fecal extracts, and gut washing samples were assessed, with no convincing uniform increase in titers in the adjuvanted groups (data not shown). The highest anti-Flu secretory IgA titers were detected from gut homogenate extracts, with titers widely spread in the Giardia-adjuvanted and Flu-alone groups (Fig. 5B). The Giardia-adjuvanted group had the highest average titer of ≈13,000, compared to a titer of ≈8,500 for the Flu-alone group and ≈6,700 for the CT-adjuvanted group (Fig. 5B).

Levels of total IgA were assessed in the gut homogenate samples to ascertain a correlation between these results and the variable anti-Flu secretory IgA titers detected for individual mice within the same inoculation group. Although total IgA concentrations varied somewhat between individual mice, there was no apparent association between level of Flu-specific antibody titer and total IgA concentration (data not shown). In addition, the average (± standard deviation) concentration of total IgA (in a 1:10,000 dilution) was similar for each of the inoculation groups as follows: Flu, 35.6 ± 12.6 ng/ml; CT-adjuvanted group, 31.2 ± 16.4 ng/ml; and Giardia-adjuvanted group, 30.5 ± 10.4 ng/ml. Use of an inbred mouse strain (C57BL/6J) also did not increase the uniformity in response between individual mice from each group.

Adjuvant activity of fractions from Giardia 2041 cytosol.

To further investigate Giardia adjuvant activity, 2041 cytosolic extract was fractionated, and samples of pooled fractions 1 to 12 were prepared, which were subsequently coadministered with KLH to mice using a two-dose regimen. CT was also coadministered with KLH, producing a control adjuvant group.

By 2 to 3 weeks after dosing, pooled serum samples demonstrated an enhanced anti-KLH serum IgG response for 8 of the 12 pooled fraction groups tested, compared to the nonadjuvanted KLH-alone group (data not shown). Groups administered KLH and Fx 3, Fx 9, Fx 10, Fx 11, or Fx 12 had higher anti-KLH serum IgG titers by 3 weeks subsequent to dosing (≈200,000) than the CT-adjuvanted group (79,400). As seen previously with a two-dose regimen, there were no significantly high anti-KLH secretory IgA titers detected from fecal samples for any of the groups inoculated with KLH and the Giardia fractions (data not shown).

Giardia Fx 12, Fx 7, and Fx 2, which had shown measurable (Fx 12) and negligible (Fx 2 and Fx 7) anti-KLH serum IgG adjuvant activity, were further assessed for adjuvanticity using a three-dose inoculation regimen and measurement of anti-KLH serum antibody titers from individual mice. Peak anti-KLH serum IgG titers were detected from experiment days 30 to 35 (Fig. 6A), and a peak anti-KLH serum IgA response was detected on experiment day 30 (Fig. 6B).

FIG. 6.

Giardia 2041 pooled cytosolic fractions variously adjuvant a KLH-specific serum IgG (A) and serum IgA (B) response. Groups of five Quackenbush mice were inoculated with 3 mg of KLH alone or with Giardia 2041 pooled fractions, each extracted from the equivalent of ≈10⁷ trophozoites, or 1 μg of CT on days 0, 7, and 21. Individual mouse serum samples were collected on day 30 (B) and day 35 (A) for antibody analysis by ELISA. Results are presented for individual mice in each group, with the average group titer represented as a bar.

Groups administered CT or Fx 12 as adjuvants produced the strongest anti-KLH serum response (Fig. 6). When average titers of anti-KLH serum IgG were compared on experiment day 35, Fx 12 had increased adjuvant activity (94,000) over CT (27,000) (Fig. 6A). As recorded for 2041 trophozoites as adjuvant, the anti-KLH serum IgG response for the Giardia fraction- and CT-adjuvanted groups was strongest for the IgG1 and IgG2b isotypes, with no response detected for the IgG2a or IgG3 isotypes (data not shown). There was considerable spread among individual anti-KLH serum IgG titers within the Fx 2-adjuvanted and KLH-alone groups, with average titers for these groups (Fx 2, 2,600; KLH, 470) well below those for the Fx 12- and CT-adjuvanted groups (Fig. 6A).

Average anti-KLH serum IgA titers for the CT- and Fx 12-adjuvanted groups were similar (3,000 to 5,000) on experiment day 30 and considerably higher than those measured for the Fx 2-adjuvanted and KLH-alone groups (<1,000) (Fig. 6B). The Fx 7-adjuvanted group produced one responder mouse on day 30 with a positive anti-KLH serum IgA titer and one responder mouse on day 35 with a positive anti-KLH serum IgG titer. The mean anti-KLH serum IgG and serum IgA titer for all other mice in the Fx 7-adjuvanted group was similar to that recorded for the Fx 2-adjuvanted group.

On day 35, significant anti-CT serum IgG titers (65,000 and 6,500) were detected for two mice from the CT-adjuvanted group, which had the highest anti-KLH serum IgG response. However, no anti-Giardia serum IgG response was detected in mouse samples from the Fx 12-adjuvanted group (data not shown), suggesting that the Giardia fraction pool is capable of augmenting an antigen-specific response without inducing a detectable response to self, a desirable attribute for effective adjuvant activity.

DISCUSSION

Cytosolic components of Giardia isolate 2041 cultured trophozoites are capable of adjuvanting an antigen-specific serum and secretory antibody response to orally coadministered antigen in a mouse model. The Giardia-adjuvanted serum antibody response was comparable to that generated by the “gold standard” mucosal adjuvant, CT. Evidence of enhanced adjuvant activity induced by isolate 2041 over isolate 106 is consistent with previous observations of Giardia toxin-like properties (36, 10). We have also confirmed an increased production of total serum IgA without production of Giardia antigen-specific IgA in 2041-infected mice (36), an ideal attribute of adjuvant activity.

Similarly, our mouse infection model demonstrates Giardia trophozoite (particularly 2041)-induced secretion of total IgA in the gut epithelium of neonatal 17-day-old mice, which normally have an immature immune system in which the processes of antibody affinity maturation and isotype switching are in general poorly achieved (19). By 3 weeks of age the neonatal immune system has matured, and IgA is secreted within the gut epithelium regardless of incitement (data not shown).

The source of adjuvant activity was localized to the trophozoite cytosol rather than the membrane fraction or secreted product of the parasite. Although some activity from the secreted product might be expected, culture conditions are far removed from the environment experienced by the parasite in an animal host and therefore may not be conducive to secretion of toxins or other products. In earlier experiments (data not shown), proteinase K digestion and 60°C heat treatment of 2041 cytosol reduced adjuvant activity, indicating that a significant portion of the adjuvant molecule is protein. When different batches of cytosolic extract were used, limited variation was seen in antigen-specific serum IgG antibody titers. However, any variability between batches may result from variation in parasite growth phase, cell count, time of harvest, and variability in processing.

Both antigen and adjuvant doses play a role in production of an immune versus a tolerogenic immunological response to inoculation. Often a large dose of antigen with a strong mucosal adjuvant is required for administration via the oral route to overcome the phenomenon of oral tolerance (12). Inoculation quantities in experimental studies are usually on the order of 100 to 5,000 μg of antigen and at least 50 μg of adjuvant (22), although smaller quantities (1 to 10 μg) of purified CT (33, 39) and nanogram concentrations of LT (13) have successfully adjuvanted a mucosal response. In keeping with these concentrations, Giardia extracts were administered using the equivalent of 10⁷ trophozoites, which equates to approximately 500 μg (data not shown) of total protein. At this stage we have not identified what portion of this value can be equated with the adjuvant factor(s). The quantity of 10⁷ trophozoites was assessed as the maximum number of trophozoites which colonize a single mouse intestine in the standard laboratory mouse infection model (32). In an earlier experiment (data not shown), mice were administered a quarter dose of 2041 cytosol and KLH antigen, which produced anti-KLH serum IgG titers at 1.5 weeks subsequent to dosing of up to 200 times less than those measured with administration of a full dose (equivalent to 10⁷ trophozoites) of cytosol.

Giardia extracts were administered with a limited range of KLH concentrations, and a lower limit of 3 to 5 mg per mouse was determined to be the most suitable concentration for production of a strong antigen-specific antibody response comparable to those recorded in the literature (33, 38). When Flu was administered at a concentration of 100 μg per mouse, significant anti-Flu serum IgG titers were detected with only the CT-adjuvanted group and not with the Giardia adjuvants. Average anti-Flu serum IgA titers were enhanced from the CT- and Giardia cytosolic extract-adjuvanted groups. Anti-Flu secretory IgA titers varied with the type of secretory extract tested (fecal, saliva, gut washing, or gut homogenate). The highest and most consistent titers were measured in the gut homogenates of the cytosolic extract-adjuvanted groups. High IgA titers in gut homogenate samples may correlate with the content of the lamina propria, which is rich in IgA plasma cells and expected to produce higher antigen-specific titers than gut washings alone. It is difficult to compare adjuvant activity, on an equivalent basis, of the crude Giardia extracts with highly purified and potent CT. Crude Giardia adjuvant activity may not be of sufficient potency to produce a secretory response in the less concentrated mucosal samples (fecal and gut washing) or at distal mucosal sites, e.g., in saliva. The concept of a common mucosal system where inoculation of a preparation at one site stimulates a response at all other mucosal sites does not always follow and may vary with the type of adjuvant administered (5).

The source of adjuvant activity from Giardia 2041 trophozoites was localized to at least one pool of cytosolic fractions, fraction pool 12. This fraction pool demonstrated the strongest adjuvant activity throughout the 3-week postinoculation period and generated adjuvant activity at a similar level or higher than the titers recorded for CT. Giardia trophozoites, cytosolic extract, and fraction pool 12 produced antigen-specific serum IgG1 and IgG2b isotypes, indicative of a Th2 response. This is in keeping with the response generated by CT and its mutants, which primarily stimulate a Th2 response through increased production of interleukin-4 (IL-4), IL-5, and IL-10, with predominant production of IgG1 and IgE (22) and in some cases IgG2b (9) isotypes. This polarization in the T-cell response, however, varies with adjuvant type and is less pronounced with LT, which activates both Th1 and Th2 cells (22).

In comparison with other Giardia extracts, and regardless of dose regimen, we were unable to detect an enhanced antigen-specific secretory response with administration of the fraction pools. The ability to adjuvant a secretory response may be diminished with Giardia protein preparation purification, causing separation of potent adjuvant complexes. This is not totally unforeseen, as loss of adjuvant activity has been recorded with the preparation of mutant molecules of CT and LT, especially those in which the toxic A1 subunit has been removed (22). If, like CT, the Giardia adjuvant extracts bind to receptors on the gut epithelia (22), an alteration in adjuvant structure might affect its binding ability, therefore preventing transport to immune inductor sites for stimulation of an immune response. It is possible that more than one form of adjuvant is contained within the Giardia trophozoites, with different forms adjuvanting the immune response via a range of mechanisms, which differentially affect the systemic and mucosal compartments. We intend to further define the Giardia adjuvant molecule(s), including its physical properties, which will allow a more direct comparison by weight with CT and other available adjuvants.

For standard prophylactic immunization in healthy individuals, only adjuvants that induce minimal side effects will prove acceptable (24). The critical issue of vaccine safety has precluded many candidate adjuvants from further development due to their toxic form. Our studies have demonstrated that extracts produced from the pathogenic Giardia 2041 isolate function with adjuvant-like activity in an animal model. The same animals do not display overt behavioral changes synonymous with a toxic response to inoculation, such as reduced mobility, loss of appetite, and substantial weight loss. These results indicate that Giardia extracts are worthy of further investigation as a future source for candidate mucosal and systemic adjuvants.