Immune Destruction of Larval Taenia crassiceps in Mice

Abstract

Immune destruction of larval Taenia crassiceps was examined by first injecting BALB/cJ mice subcutaneously with larval buds and 30 to 60 days later challenging the mice with larvae injected into the peritoneal cavity. The larvae injected intraperitoneally (i.p.) secondarily are killed by host cells that completely encase the larvae in a thick sheath. The peritoneal exudate cells and the cytokines they produced were characterized by flow cytometry, enzyme-linked immunosorbent assays (ELISAs), and reverse transcription PCR (RT-PCR). No changes in percentage of CD4⁺ T cells, CD8⁺ T cells, B1 cells, or macrophages were detected in the peritoneal cavities of mice that were killing larvae compared to mice with a primary 7-day infection i.p. Both RT-PCR and ELISA demonstrated a decrease in cytokines including gamma interferon (IFN-γ), interleukin-4 (IL-4), and IL-10 in mice that were killing the larvae compared to control mice infected for 30 to 60 days i.p. alone, although there was little difference compared to mice infected for 7 days i.p. alone. Serum cytokine levels in mice that were killing the larvae showed a decrease in IFN-γ and IL-4, an increase in IL-10 when compared to mice infected for 30 to 60 days i.p. alone, and increases in all cytokines compared to mice infected for 7 days i.p. alone. Inhibition of nitric oxide production did not significantly affect the number or the viability of larvae in the peritoneal cavity of mice that were killing larvae during secondary infection.

Human cysticercosis is a disease caused by the larval stage of the cestode parasite Taenia solium. Humans acquire the larvae by ingestion of eggs released from the adult tapeworm. Human cysticercosis is common in Mexico and underdeveloped countries and is increasing in prevalence in North America (5, 10, 11). Infection of the central nervous system leads to neurocysticercosis. An excellent model system for the study of this disease is infection of BALB/c mice with another taeniid parasite, Taenia crassiceps. The rodent is the natural intermediate host for this parasite, and the definitive host is a canine. The cysts multiply in the peritoneal cavity of the mouse by budding asexually in a seemingly uncontrolled manner, making the BALB/c mice extremely susceptible to infection. Analysis of the immune response that ensues during infection has shown that it is a mixed Th1/Th2 phenotype (15) that is ineffective in controlling parasite growth.

To better determine immune responses that have the capacity to kill cestode larvae, a model system in which T. crassiceps larvae are immunologically killed in vivo was developed by a method similar to one reported earlier (12). In our work, it was found that a primary subcutaneous (SQ) infection of T. crassiceps larvae induces killing of larvae during a secondary intraperitoneal (i.p.) infection. In the present study, we report the conditions for inducing larval destruction, the effect of host responses to larvae SQ and i.p. as observed by scanning electron microscopy (EM), and the cell populations and cytokine production present i.p. as determined by enzyme-linked immunosorbent assay (ELISA) and reverse transcription-PCR (RT-PCR). SQ followed by i.p. infection provides an in vivo model system in which the complete killing of larval T. crassiceps, and the immune mechanisms required for this killing can be examined.

MATERIALS AND METHODS

Mice and infections.

Six-week-old female BALB/cJ mice were purchased from The Jackson Laboratory (Bar Harbor, Maine). The ORF strain of T. crassiceps was used for infections (6). Parasites were obtained from the peritoneal cavity of BALB/cJ mice that were infected for 3 to 4 months and were washed three times with an equal volume of phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 8.1 mM Na₂HPO₄, 1.47 mM KH₂PO₄ [pH 7.2]) under sterile conditions. Mice were infected by injection of 10 small (<2-mm) larvae in 0.5 ml of PBS either i.p. or SQ, using a 20-gauge needle. Unless specified otherwise, mice designated as having SQ infections will be those infected SQ for 30 to 60 days. Mice infected for 30 to 60 days SQ followed by i.p. infection are also referred to as challenged mice.

Preparation of larvae and EM.

Larvae were removed from the peritoneal cavity of mice by peritoneal lavage with 10 ml of PBS, fixed in 10% formalin for at least 24 h, and then kept at 4°C until use. The specimens were then covered in 0.1 M Sorenson's physiological solution containing 2% glutaraldehyde (0.2 M Sorenson's buffer [4 ml] [6.41 g of NaH₂PO₄, 41.3 g of Na₂HPO₄ · 7H₂O, pH 7.2, distilled water to 1.0 liter], 8% glutaraldehyde [2 ml], water [2 ml]) and left overnight at 0°C. They were then rinsed with four changes (30 min each) of 0.1 M Sorenson's buffer, covered with 2% OsO₄ in Sorenson's buffer for 1 h, and taken through a dehydration series (50% through 100% ethanol). The specimens were dried in a Pelco CPD-2 critical point dryer (Ted Pella, Inc., Redding, Calif.), coated with gold in a Pelco SC-4 sputter coater (Ted Pella, Inc.), and visualized using an Amray (Bedford, Mass.) model 1810 scanning electron microscope, all according to the manufacturer's instructions. Photographs were taken with positive/negative film (Polaroid Corp., Cambridge, Mass.).

Peritoneal and spleen cell preparations.

Cells were cultured in RPMI 1640 with l-glutamine (Mediatech, Herndon, Va.) supplemented with heat-inactivated fetal bovine serum (Atlanta Biologicals, Norcross, Ga.) and antibiotics (5 U of penicillin, 5 μg of streptomycin, and 10 μg of neomycin per ml; Sigma Chemical Co., St. Louis, Mo.). This medium is referred to as RPMI-C.

Spleens were macerated in 5 ml of ice-cold RPMI-C, and single-cell suspensions were obtained by drawing the cells up through a 23-gauge needle and expelling them through a 26-gauge needle. Cells were spun at 500 × g, and red blood cells were then lysed by hypotonic shock. Cells in suspension were collected, spun at 500 × g, and resuspended in 5 ml of RPMI-C; cell viability was determined by trypan blue exclusion. Cell viability was consistently >95%.

Peritoneal cells (PECs) were obtained by peritoneal lavage using 10 ml of ice-cold PBS. Red blood cells were lysed as above, remaining cells were removed, spun at 500 × g, and resuspended in 5 ml of RPMI-C.

Flow cytometry.

PECs were examined for the percentages of CD4⁺, CD8⁺, CD5⁺, B220⁺, and F4/80⁺ cells. Phycoerythrin-labeled anti-CD4 (H129.19) and anti-CD8 (53.67) antibodies were used for fluorescent labeling. Nonspecific binding was blocked with FcBlock, and the isotype control was phycoerythrin-labeled rat immunoglobulin G2a. All of these reagents were purchased from PharMingen (San Diego, Calif.) except for the F4/80 antibody, which was purchased from Biosource International (Camarillo, Calif.). One million cells were washed in 1 ml of staining buffer (PBS, 1% bovine serum albumin, 0.1% NaN₃ [pH 7.5]) and then stained with 1 μg of the appropriate antibody in 100 μl of staining buffer (30 min, 4°C). Incubation of the unlabeled antibody against the F4/80 macrophage marker was followed by incubation with 1 μg of a fluorescein isothiocyanate-labeled anti-rat antibody in 100 μl of staining buffer (30 min, 4°C). The stained cells were fixed in 500 μl of Ortho Permeafix (Ortho Diagnostics Inc., Raritan, N.J.) and analyzed by flow cytometry (Coulter Epics XL) at the Wake Forest University School of Medicine. Analysis was completed using WinList (Verity Software House, Inc., Topsham, Maine).

Preparation of soluble larval antigen preparation (SLAP).

Larvae were removed from the peritoneal cavity from mice that were infected for at least 4 months. The larvae were washed three times with an equal volume of ice-cold PBS, and then all excess PBS was removed. Packed larvae were then sonicated until no large particles were apparent. To keep the solution cold, sonication was paused every 30 s and the solution was placed on ice. The preparation was then homogenized with a smooth pestle tissue grinder to disrupt any remaining aggregates. Insoluble materials were removed by centrifugation (20,000 × g, 1 h, 4°C). Supernatants were filter sterilized, and the protein concentration was determined by the Bradford method (Bio-Rad, Hercules, Calif.). The solution was brought to a final protein concentration of 1 μg/ml with sterile PBS and then stored at −20°C until use.

Spleen cell proliferation and stimulation for cytokine production.

Spleen cells were plated in RPMI-C in flat-bottom, polystyrene 96-well plates (Corning Glass Works, Corning, N.Y.) at 5 × 10⁵ cells/well in triplicate and were stimulated with indicated concentrations of concanavalin A (ConA; Sigma, St. Louis, Mo.) or SLAP in a final volume of 250 μl/well. After 60 h of incubation, 1 μCi of [³H]thymidine (ICN Pharmaceuticals, Irvine, Calif.) diluted in 10 μl of RPMI-C was added to each well. Thymidine incorporation was measured by liquid scintillation spectroscopy (LS-1801; Beckman Instruments Inc., Fullerton, Calif.).

PECs were cultured at 2 × 10⁵ cells/well in medium alone, with ConA (5 μg/ml) or SLAP (10 μg/ml), in a final volume of 250 μl. Approximately 10 wells of each treatment were run. After 24 h, 150 μl of supernatant was removed and frozen at −20°C until needed.

Cytokine ELISAs.

Capture and detection antibodies were purchased from PharMingen, and ELISAs were run according to the manufacturer's protocol. The concentrations of gamma interferon (IFN-γ), interleukin-4 (IL-4), and IL-10 in each sample were determined in triplicate, and the mean of each sample was calculated. Sensitivities of the ELISA for IFN-γ, IL-10, and IL-4 were 10, 2, and 2 pg/ml, respectively.

RT-PCR.

RNA was extracted from 10⁷ PECs as previously described (2). RNA was analyzed spectrophotometrically (Beckman DU-64) by measuring concentration at 260 nm and testing for purity using the A₂₆₀/A₂₈₀ ratio. All RNA samples had an A₂₆₀/A₂₈₀ ratio of between 1.5 and 2.0. Two micrograms of RNA was reverse transcribed in 40-μl reactions using Moloney murine leukemia virus reverse transcriptase (Amersham Life Science, Inc., Cleveland, Ohio) at 200 U/μl and oligo(dT)_12–18 (Pharmacia, Piscataway, N.J.) at 1 μg/ml. cDNAs were precipitated and resuspended in deionized water, and then cytokine-specific cDNAs were amplified by PCR using the following conditions and primers: 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 0.11 mM each deoxynucleoside triphosphate, 2.3 pmol of each cytokine-specific primer (1) or β-actin-specific primer (3′ primer, 5′CTCTTTGATGTCACGCACGATTTTC3′; 5′ primer, 5′GACGAGGCGCAGAGCAAGAGAGG3′) per μl, 1.15 mM MgCl₂, and 1.3 U of AmpliTaq DNA polymerase (Perkin-Elmer, Branchburg, N.J.) in a 50-μl reaction in Thermolyne oil-free tubes (Barnstead Thermolyne Corp., Dubuque, Iowa). Cycle conditions were as follows: premelting at 95°C for 2 min, annealing at 52°C for 1 min, elongation at 72°C for 1 min, and denaturation at 95°C for 50 s (35 cycles [25 cycles for β-actin]), final anneal at 52°C for 2 min, and final elongation at 72°C for 2 min. PCR products were run on a 1.5% agarose gel in Tris acetate buffer containing ethidium bromide. Gels were imaged and analyzed on an AlphaImager 2000 (Alpha Innotech Corporation, San Leandro, Calif.) using spot densitometry. Ratios of the cytokine PCR product to the corresponding β-actin PCR product were compared.

Detection of nitric oxide.

Nitric oxide production by PECs was detected using the Greiss reaction as previously described (7). Briefly, 50 μl of 1% sulfanilimide solution and 50 μl of 1% naphthylethylenediamine dihydrochloride were added to 50-μl samples in a 96-well plate. Absorbance was measured at 550 nm.

Nitric oxide inhibition.

BALB/cJ mice were injected SQ with 10 T. crassiceps larvae and then 30 to 60 days later injected i.p. with 10 larvae. Immediately thereafter, one half of the mice were injected i.p. with aminoguanidine (50 μg/kg of body weight in 500 μl of sterile PBS), and this injection was continued twice daily using the same concentration of aminoguanidine. Control mice received injections of PBS (500 μl). After 7 days, the mice were killed by cervical dislocation, larvae were counted, and serum and PECs were obtained as described.

Statistical analyses.

Data were analyzed for statistical significance using Student's t test. Unless otherwise stated, data are presented as means ± standard deviation and were considered significantly different when P was <0.05.

RESULTS

Primary SQ infection prevents secondary i.p. infection.

To determine if a primary infection of larvae would prevent the establishment of a secondary infection, mice were infected SQ for 100 days and then i.p. for 30 days. Mice that received an initial SQ injection of larvae destroyed all larvae (n = 3) in the peritoneal cavity, whereas mice receiving larvae only i.p. had progressing infections (27.0 ± 8.5; n = 3).

To examine the development of antilarval immunity, groups of mice were injected SQ with 10 buds and then 2, 4, 6, or 8 weeks later injected i.p. with 10 buds. Mice were killed 1 week after the i.p. infection, and the larvae were counted; there were no significant differences in numbers of larvae or viable larvae recovered from the peritoneal cavity between the groups of mice (data not shown). Regardless of duration of SQ infection, the larvae injected i.p. were killed. Based on these observations, in subsequent experiments mice were infected SQ for 30 to 60 days followed by i.p. infection. When this approach was followed, mice infected SQ for 30 to 60 days and then infected i.p. for 7 days showed a high level of larval destruction (Fig. 1A). It is clear from these data that the immune responses able to kill the larvae were developing effectiveness by day 7. Because it is important to examine killing responses as they are occurring instead of at the culmination of rejection, it was necessary to demonstrate that after the 7-day time point, the mice continued to kill the larvae present in the peritoneal cavity. Repeating the experiment but allowing the i.p. infection to progress for 14 days, we recovered a maximum of one larva from the peritoneal cavity of any mouse, confirming that at day 7 larvae in the peritoneal cavity were in the process of immunological destruction, and this continued for at least the next 7 days (Fig. 1B).

FIG. 1.

(A) Numbers of larvae in mice 7 days after i.p. infection. Mice (n = 21) were injected SQ and then 30 to 60 days later were injected i.p. Control mice (n = 11) received only the i.p. infection. Mice were killed 7 days after i.p. infection, and total larvae (white bars) and viable larvae (solid bars) in the peritoneal cavity were counted. (B) Numbers of larvae in mice 14 days after i.p. infection. Mice (n = 3) were injected SQ and then 30 to 60 days later were injected i.p. Control mice (n = 3) received only the i.p. injection. Mice were killed 14 days after i.p. infection, and total numbers of larvae (white bars) and viable larvae (cross-hatched bars) in the peritoneal cavity were counted. ∗, significant difference between values within the group of total larvae or viable larvae.

Condition of larvae recovered from the peritoneal cavity.

Larvae removed from both the SQ and i.p. locations were prepared for scanning EM. Figures 2A and B show scanning electron micrographs of larvae removed from mice infected i.p. alone for 7 days. Figure 2A is representative of larvae removed from the peritoneal cavity. Due to the dehydration and critical point drying process, some damage to the tegument of the larvae occurred. Otherwise, the larvae were intact and undamaged when removed from the peritoneal cavity. Figure 2B shows the surface of a larva under greater magnification; the microtrichs are plainly visible.

FIG. 2.

Scanning EM of larvae removed from the peritoneal cavity. (A and B) Larvae removed from mice infected 7 day i.p. only. The tegument is visible, and the larvae appear undamaged by the host immune system. (C and D) Larvae removed from mice infected SQ followed by i.p. infection for 7 days. Tegument is encased in host cells with underlying matrix. Original magnifications: A and C, ×60; B and D, ×650.

Figures 2C and D show larvae removed from the peritoneal cavity of mice first infected SQ and then infected i.p. Figure 2C shows a layer of host cells that has separated from the larva so that the tegument of the larva is visible. Figure 2D is a magnified portion of this layer of host cells, showing that encapsulating sheath is composed of multilayers of host cells overlaying a fibrous matrix which seems to have been deposited on the tegument of the larva.

Larvae removed from the peritoneal cavity of mice with existing SQ infections were assessed for viability by eosin exclusion (8) and by observing motility. These two viability tests gave the same results, showing that some larvae were viable and some were nonviable when removed from the peritoneal cavity. Figure 3 shows electron micrographs comparing the surfaces of larvae that were viable (Fig. 3A and B) to those that were dead (Fig. 3C and D). Note the absence of cells and the dense fibrous matrix that is present on dead larvae. Dead larvae had fewer adherent cells attached and a very dense matrix on the tegumental surface compared to viable larvae.

FIG. 3.

Scanning EM of surface of viable and nonviable larvae. Viable and nonviable larvae were separated after removal from the peritoneal cavity of mice infected SQ followed by i.p. infection for 7 days. (A and B) Viable larvae; (C and D) nonviable larvae. Note the less dense layer of cells and thick matrix of fibrous material that is present on the nonviable larvae. Magnifications range between ×300 and ×500.

Composition of PECs.

PECs from mice infected i.p. for 7 days and mice infected SQ followed by infection i.p. were stained for the presence of CD4⁺ T cells, CD8⁺ T cells, B220⁺ cells, CD5⁺ cells, and macrophages. Flow cytometry did not reveal any significant differences in the composition of PECs between the two groups of mice (data not shown).

Proliferation of spleen cells.

To estimate the immune response of splenocytes to ConA and to larval antigens in culture, splenocytes (5 × 10⁵/well) were stimulated with different concentrations of ConA or SLAP. When stimulated with ConA, splenocytes from SQ-infected mice proliferated significantly less than all other splenocytes from all other groups of mice (Fig. 4A). Proliferation of splenocytes from mice infected for 7 days i.p., mice infected for 30 to 60 days i.p., and mice infected SQ followed by infection i.p. did not show significant differences at a ConA concentration of 5 μg/ml or less. At 10 μg of ConA per ml, splenocytes from mice infected for 7 days i.p. and mice infected SQ followed by i.p. infection showed significantly less proliferation than both normal mice and mice infected for 30 to 60 days i.p. along. Interestingly, splenocytes from mice infected SQ alone proliferated significantly less in response to all concentrations of ConA than spleen cells from other groups of mice, infected or uninfected (Fig. 4A).

FIG. 4.

Splenocytes were stimulated in culture with indicated concentrations of ConA (A) or SLAP (B). Mean [³H]thymidine incorporation is presented. Data at one particular concentration of mitogen are significantly different if bars do not share the same asterisk (∗) or individual group of asterisks (∗∗, ∗∗∗).

In response to SLAP, spleen cells from mice infected SQ followed by i.p. infection demonstrated greater proliferation at all concentrations of SLAP than splenocytes from uninfected mice or from mice infected i.p. alone for 7 days (Fig. 4B). However, splenocytes from mice only infected i.p. for 30 to 60 days proliferated more than spleen cells from all other groups of mice at all concentrations of SLAP. Splenocytes from mice infected SQ alone also proliferated more than those of all other groups at all concentrations of SLAP with the exception of 5 μg/ml, which was not significantly greater than the mice infected SQ followed by i.p. infection (Fig. 4B). The extensive proliferation of splenocytes from SQ infected mice in response to SLAP did not correspond with the results seen with ConA, in which proliferation of the splenocytes was lower compared to all other stimulatory preparations (Fig. 4A).

Cytokine production in PECs.

PECs were cultured for 24 h in RPMI-C alone, with ConA (5.0 μg/ml), or with SLAP (10 μg/ml), and cytokine production was measured by ELISA. Unstimulated or SLAP-stimulated PECs from mice infected SQ followed by i.p. infection produced significantly less IFN-γ compared to mice infected i.p. alone for 30 to 60 days (Fig. 5A and C). IFN-γ production by PECs from mice infected SQ followed by i.p. infection was not significantly different from the production by PECs of mice infected i.p. alone for 7 days in ConA-stimulated cultures but was significantly lower in unstimulated and SLAP-stimulated cultures (Fig. 5).

FIG. 5.

PEC production of IFN-γ. PECs were incubated with RPMI-C only, ConA (5.0 μg/ml), or SLAP (10 μg/ml). Supernatants were harvested 24 h later, and IFN-γ production was measured by ELISA. Cytokine data are significantly different if bars do not share the same asterisk (∗) or group of asterisks (∗∗, ∗∗∗).

Production of IL-4 was similarly reduced in PECs from mice infected SQ followed by i.p. infection compared to mice infected i.p. alone for 30 to 60 days in unstimulated cultures and with ConA stimulation (Fig. 6A and B). When stimulated with SLAP, PECs from mice infected SQ followed by i.p. infection produced slightly but not significantly more IL-4 than PECs from mice infected i.p. alone for 30 to 60 days (Fig. 6C). Compared to mice infected i.p. alone for 7 days, ConA- and SLAP-induced IL-4 production was significantly higher in mice infected SQ followed by i.p. infection (Fig. 6B).

FIG. 6.

PEC production of IL-4. PECs were incubated with RPMI-C only, ConA (5.0 μg/ml), or SLAP (10 μg/ml). Supernatants were harvested 24 h later, and IL-4 production was measured by ELISA. Cytokine data are significantly different if bars do not share the same asterisk (∗) or group of asterisks (∗∗, ∗∗∗).

IL-10 production by PECs from mice infected SQ followed by i.p. infection was also downregulated compared to mice infected i.p. alone for 30 to 60 days in unstimulated cultures and in ConA-stimulated cultures but not in SLAP-stimulated cultures (Fig. 7). Compared to mice infected i.p. alone for 7 days, unstimulated, ConA-stimulated, and SLAP-stimulated production of IL-10 was significantly lower in PECs from mice infected SQ followed by i.p. infection. IL-10 production by PECs from mice infected for 7 days i.p. alone was also greater than that of PECs from mice infected for 30 to 60 days i.p. alone in all cultures (Fig. 7).

FIG. 7.

PEC production of IL-10. PECs were incubated with RPMI-C only, ConA (5.0 μg/ml), or SLAP (10 μg/ml). Supernatants were harvested 24 h later, and IL-10 production was measured by ELISA. Cytokine data are significantly different if bars do not share the same asterisk (∗) or group of asterisks (∗∗, ∗∗∗).

RT-PCR of cytokine-specific mRNA from PECs.

RNA was harvested from PECs ex vivo, and levels of mRNA for IFN-γ, IL-4, IL-10, and IL-2 were determined as ratios against β-actin. RT-PCR confirmed results from ELISAs that IFN-γ, IL-4, and IL-10 production were all downregulated in PECs from mice infected SQ followed by i.p. infection compared to mice infected i.p. alone for 30 to 60 days and mice infected i.p. alone for 7 days (data not shown). Production of IL-2 was not measured by ELISA, but RT-PCR indicated that IL-2 was upregulated in mice infected SQ followed by i.p. infection compared to mice infected i.p. alone for 30 to 60 days and in mice infected i.p. alone for 7 days (data not shown).

Serum cytokine levels.

IFN-γ levels were significantly higher in mice infected SQ followed by i.p. infection than in uninfected controls, mice infected i.p. alone for 7 days, and mice infected only SQ. However, compared to mice infected i.p. alone for 30 to 60 days, IFN-γ in mice infected SQ followed by i.p. infection was significantly lower (Fig. 8A). Levels of IL-4 in the serum of mice infected SQ followed by i.p. infection was significantly higher than in uninfected mice, mice infected i.p. alone for 7 days, and mice infected only SQ but were significantly lower than in mice infected i.p. alone for 30 to 60 days (Fig. 8B). Serum levels of IL-10 in mice infected SQ followed by i.p. infection were significantly higher than in all other groups of mice tested (Fig. 8C).

FIG. 8.

Serum cytokine levels. Serum samples were diluted 1/5 in RPMI-C, and then levels of IFN-γ (A), IL-4 (B), and IL-10 (C) were measured by ELISA. Values are significantly different if bars do not share the same asterisk (∗) or group of asterisk (∗∗, ∗∗∗).

Production of nitric oxide by PECs.

In challenged mice, nitric oxide production by PECs in response to ConA stimulation was measured on day 7 of i.p. infection alone or on day 7 in challenge infections. The NO level was higher in challenged mice than in mice infected i.p. alone for 7 days (Fig. 9B). However, when cells were unstimulated or stimulated with SLAP, NO production by PECs from challenged mice was not different from that in mice infected i.p. alone for 7 days (Fig. 9A and C). NO production by PECs from mice infected for 30 to 60 days i.p. only was significantly higher than that of mice infected for 7 days i.p. only and of challenged mice in unstimulated, ConA-stimulated, and SLAP-stimulated conditions (Fig. 9).

FIG. 9.

Nitric oxide production by PECs in response to RPMI-C only, ConA (5 μg/ml), or SLAP (10 μg/ml). Values are significantly different if bars do not share the same asterisk (∗) or group of asterisks (∗∗, ∗∗∗).

Effect of nitric oxide inhibitors on the killing of larvae.

Mice infected SQ followed by i.p. infection were treated with aminoguanidine, a nitric oxide inhibitor, to determine if nitric oxide played a role in the killing of larvae. Mice received twice-daily i.p. injections of 50 μg of aminoguanidine/kg of body weight in PBS or twice-daily injections of PBS alone beginning the day larvae were injected i.p. Treatment with the nitric oxide inhibitor did not significantly affect the total number of larvae recovered (treated, 9 ± 6; controls, 3.7 ± 0.58) or the numbers of viable larvae (treated, 2.3 ± 3.2; controls, 0.66 ± 1.2) that were recovered from the peritoneal cavity after 7 days of i.p. infection.

Primary i.p. infection prevents secondary SQ infection.

It was necessary to determine if the effect of primary infection and secondary challenge of larvae was similar if the sites of first injection were reversed. Groups of mice were injected i.p. for 30 days followed by SQ injection and with SQ only for 7 days. In these experiments, the secondary SQ larvae were killed (0.5 ± 1.0; n = 6), and the primary i.p. larvae (123 ± 16; n = 3) were apparently unharmed (for SQ-only controls, the values were 6 ± 1.0 [n = 3]). This demonstrated that the ability to immunologically reject larvae apparently is not limited to the i.p. site of infection and that killing occurs to those larvae in the challenge infection.

It has been reported previously that mice chronically infected with larval T. crassiceps are severely immunosuppressed (16; S. A. Toenjes, R. J. Spolski, M. A. Alexander-Miller, K. A. Mooney, and R. E. Kuhn, submitted for publication). To determine if this immunosuppression yielded mice ineffective at killing a secondary challenge of larvae, mice chronically infected i.p. (4 months) were injected SQ with larvae and killed 7 days later. The number of secondary SQ larvae that were recovered from challenged mice was not statistically different from that for mice infected SQ alone for 7 days, suggesting that the immunosuppression in chronically infected mice inhibits immune rejection of larvae in the challenge infection (7-day SQ-only controls, 6.0 ± 1.0 [n = 3]; chronic i.p. infection followed by SQ infection, 6.7 ± 1.0 [n = 3]).

DISCUSSION

Previous studies have shown that a primary SQ infection of T. crassiceps larvae can induce killing of a challenge infection of larvae (12–14). In none of these studies, however, was the killing response as immediate or effective as observed in our experiments (12). The differences in the results between our studies and the experiments of Siebert and Good (12) may be due to the fact that they used only 3 larvae per injection in both SQ and i.p. infections, whereas we used 10 larvae per injection. The larger dose of larvae SQ may have induced a stronger, more rapid, and more effective killing response in our experiments. In the experiments by Siebert and Good (12), however, there were also fewer larvae (3 versus 10) to be killed, yet they did not observe complete rejection of larvae.

The killing response observed in the present study was coincident with adherence of a sheath of host cells to the tegument of the larvae. Characteristics of the sheath of cells and fibrous material that surrounded larvae appeared different between viable and nonviable larvae. Viable larvae had a very dense layer of cells around them, whereas dead larvae had fewer cells and a dense, fibrous matrix surrounding them (Fig. 3). This difference may reveal different phases of the host cell attack against the larvae. Continued adherence of host cells to the surface of larvae may cease once larvae have died. Because there are fewer cells on the larval surface, the dense matrix that has been deposited on the larval surface may be more visible by scanning EM. Also, the density of the matrix may increase as the destructive activities progress; i.e., a denser matrix would indicate a more advanced stage of the killing response and would be more apparent on the tegument. The source and characterization of this matrix material are under investigation.

Flow cytometry of PECs from mice infected i.p. alone for 7 days and from challenged mice did not indicate any significant changes in the percentages of CD4⁺ T cells, CD8⁺ T cells, B220⁺ cells, CD5⁺ cells, B220 CD5⁺ cells, or macrophages. This was unexpected, as it might be predicted that there would be an increase or a decrease in some cell populations during the killing response. It is possible that the infiltrating cell populations differed but that the crucial effector populations were bound to the larvae. Flow cytometry was not performed on the cell layers adhered to the larvae, but in the future this may reveal some interesting information.

Despite the intense killing response in the peritoneal cavity of challenged mice, the proliferation of spleen cells from these mice in response to ConA was not significantly higher than that of spleen cells from mice infected i.p. alone for 7 days, mice infected i.p. alone for 30 to 60 days, or uninfected mice (Fig. 4A). It is also notable that when stimulated with SLAP, the mice infected SQ alone and mice infected i.p. alone for 30 to 60 days showed greater proliferation than challenged mice, which was not anticipated (Fig. 4B). Also, the proliferation of the spleen cells from these mice in response to ConA and to SLAP was comparatively low. It may be that the proliferative capacity of the spleen cells is unrelated to the killing response that is induced by SQ infection, making the splenocyte blast assay a less useful measure of immunity.

IFN-γ was downregulated in challenged mice compared to mice infected i.p. alone for 30 to 60 days in unstimulated, ConA-stimulated, and SLAP-stimulated cultures (Fig. 5). IFN-γ was also downregulated compared to mice infected i.p. alone for 7 days in unstimulated and SLAP-stimulated cultures. Considering the intense inflammatory response that was occurring in the peritoneal cavity, significant production of IFN-γ would be anticipated. Serum levels of IFN-γ were high in mice infected SQ followed by i.p. infection, indicating that there was significant systemic production of IFN-γ (Fig. 8A); however, ELISAs and RT-PCR showed that the PECs were not producing this cytokine. The highest levels of IFN-γ in serum were in the mice infected i.p. alone for 30 to 60 days. IFN-γ production, then, is highest in mice with progressing larval infection and is lower in the serum and in production by PECs of mice that are killing the larvae. This argues against the possibility that IFN-γ plays the prominent role by inducing inflammation and Th1-associated responses in PECs during the killing of larvae. Other work on larval T. crassiceps infection has shown that IFN-γ levels in the serum consistently rise over the course of infection, indicating again that the presence of this cytokine is associated with heavy infection (15). It is also of note that the production of IFN-γ in response to SLAP by PECs from mice infected i.p. only for 7 days, and for 30 to 60 days i.p. alone, produce high levels of IFN-γ, but that PECs from challenged mice do not.

When unstimulated, stimulated with ConA, or stimulated with SLAP, PECs from mice infected for 7 days i.p. alone produced levels of IL-10 that were significantly higher than levels produced by PECs from mice infected for 30 to 60 days i.p. alone and from challenged mice. This early production of high levels of IL-10 during larval T. crassiceps infection in BALB/c mice has been demonstrated previously (15). IL-10 is known as a downregulatory cytokine (3, 4), and the early production of high levels of IL-10 may inhibit the immune response from destroying the larvae (15). These observations are consistent with the finding that in challenged mice that are killing the larvae, production of IL-10 by PECs is very low (Fig. 7A). This provides more evidence that production of IL-10 by PECs early in infection is conducive to heavy parasitism but does not coincide with serum levels of IL-10, because in challenged mice levels of IL-10 in serum were approximately six times higher than in mice infected 7 days i.p. alone.

Inhibition of NO production did not significantly affect the number or viability of parasites recovered from the peritoneal cavity of challenged mice. NO has been implicated as playing a role in immunity in a number of parasitic diseases and as being able to decrease the burden of both intracellular and extracellular parasites (9, 17). NO production by PECs from challenged mice was always lower than that of mice infected for 30 to 60 days i.p. alone. Compared to mice infected for 7 days i.p. alone, production of NO by PECs stimulated with ConA was higher in challenged mice (Fig. 9B). Despite inhibition of NO production in vivo in challenged mice, there was no significant change in the number of parasites recovered from the peritoneal cavity.

It is of interest to determine why larvae in a secondary infection are killed while the primary infection of larvae remains unaffected. By reversing the injections of the larvae to a primary i.p. infection followed by a SQ challenge, the SQ larvae were killed and the i.p. larvae were apparently unharmed. This suggests that the parasites that are killed are restricted to the secondary infection, and that the parasites in the primary infection are unharmed.