Cell-mediated immunity to Toxoplasma gondii develops primarily by local Th-1 host immune responses in the absence of parasite replication¹

Experimental animals

Adult 6–8 week old C57BL/6, C57BL/6 B cell deficient (µMT), IL-15−/− mice (B6), or age matched BALB/c and BALB/c Ifn-γ −/− (Ifn-γ KO) were obtained from Jackson Labs. Mice were maintained in Tecniplast Seal Safe mouse cages on vent racks at the Dartmouth-Hitchcock Medical Center mouse facility (24). µMT mice were maintained under sterile food conditions. All mice were cared for and handled according to the Animal Care and Use Program of Dartmouth College using NIH approved institutional animal care and use committee guidelines.

Parasites, parasite infections, immunizations and preparations

Tachyzoites of the virulent Type I strain RH or attenuated Type I strain cps1-1 were obtained by serial passage in human foreskin fibroblast (HFF) cell monolayers in EMEM supplemented with 1% FBS and 250nM Uracil (cps1-1 only) (1). Viability of tachyzoites (30–50% infectious) was tested by plaque assay. Mice were immunized once i.v. or twice i.p. two weeks apart with 1 × 10⁶ cps1-1 tachyzoites. One month after final immunization, mice were challenged i.p. with either low 1 × 10² or high (1 × 10³ or 1 × 10⁵) doses of viable RH (25). Total tachyzoite lysate antigen was prepared from in vitro HFF cultures of RH as described (26, 27). For cytokine assays mice were infected i.p. with either 1 × 10⁶ cps1-1 or 1 × 10³ RH tachyzoites.

Ex vivo cell infections

Peritoneal exudate cells (PECs), peritoneal resident macrophages, and dendritic cells (DCs) were obtained from naïve C57Bl/6 mice. DCs were obtained from the spleens of WT mice and purified using EasySep CD11c positive selection as per manufacturers instructions (StemCell Technologies Inc.). PECs were obtained by peritoneal lavage with 5–7mls of ice cold sterile PBS and peritoneal resident macrophages were obtained by plating a portion of total PECs at 4 × 10⁶ cells/ml in 24 well plates for 4 hrs at 37°C in DMEM supplemented with 10% FBS. Non-adherent cells were removed by gently washing and purity of macrophages was verified (>90% CD11b+) (28). The DCs, PECs, and PEC resident macrophages were plated at 2 × 10⁶ cells/ml in infection medium consisting of EMEM supplemented with 1% FBS and 250 nM uracil (SIGMA). Purified cps1-1 tachyzoites were inoculated into individual wells at 5 × 10⁵ parasites/ml for 12 hours at 37°C. Infected cells were examined by light microscopy and contained 1–4 cps1-1 tachyzoites. Remaining extracellular tachyzoites were removed and mice were inoculated once with either 1 × 10⁶ cps1-1 tachyzoites or 5 × 10⁵ of infected DC, PEC, or PEC resident macrophages.

Flow cytometric differential cell analysis

Total PECs were obtained at indicated times post infection and were depleted of erythrocytes using buffered ammonium chloride and enumerated by trypan blue exclusion (29). PECS were fixed (1 % paraformaldehyde) and stained using standard procedures in stain wash buffer (SWB) containing 2% FBS in 1 X PBS, 5% normal mouse serum (Sigma), and Mouse FC Block (BD biosciences). Fixed PECs were singly or doubly stained with PE conjugated anti-mouse CD3 (17A2), PE conjugated anti-mouse GR1 (RB6-8C5), FITC conjugated anti-mouse CD45R/B220 (RA3-6B2), FITC conjugated anti-mouse CD11b (M1/70), FITC conjugated anti-mouse CD19, anti-mouse F4/80, or the intracellular stain FITC conjugated anti-mouse CD68. All antibodies were from BD biosciences except antibodies to F4/80 and CD68 both from Serotec. For the intracellular stain (CD68), fixed PECs were permeabilized using SWB 0.5% saponin. Non-specific Ig isotype specific antibodies were used as negative controls for all samples (BD biosciences). Flow cytometric data was acquired and analyzed using the FACS Calibur flow cytometer and CellQuest software (BD Immunocytometry Systems. Absolute numbers are the product of percentage of total events collected from 5 × 10⁴ total events multiplied by total cell number per mouse

Cytokine assays

Serum was obtained from cps1-1 or RH infected mice at indicated times from whole blood incubated at RT for 2 hours then centrifuged for 10 minutes at 14,000 rpm at 4°C. Total PECs and splenocytes from infected mice were harvested then depleted of erythrocytes, counted , and seeded in 24 well plates with DMEM supplemented with 10% FBS. PECs were plated at 1 × 10⁶ cells/well and splenocytes at 5 × 10⁶ cells/well and cultured for 24 hours without antigen restimulaiton (16). Cell cultures were lysed by freeze-thawing 3 times. Debris was removed from the supernatants by centrifugation at 14,000 rpm at 4°C for 10 minutes. Concentrations of the mouse cytokines Ifn-γ, IL-12p40, and IL-12p70 in serum (1:4 dilution) and cell supernatants were then determined using OptEIA ELISA kits (BD Biosciences) (11). ELISA was developed with TMB substrate (Pierce) and stopped using 0.2 N H₂SO₄ (30).

Tg lysate preparation and serum IgG assay

Anti-toxoplasma specific serum IgG responses were measured in cps1-1 immunized mice by coating each well of a 96 well plate (Nunc) with 100 µl of TLA at a concentration of 1µg/ml in 0.1M Na₂HPO₄ pH 9.0 overnight at 4°C. Plates were washed three times with 1 X PBS with 0.05% tween-20 (PT) then blocked with PT/1% BSA. Seras were serially diluted beginning with 1:100 in PT 0.1% BSA and applied to each well in triplicate and incubated overnight at 4°C. Plates were washed with PT then mouse anti-Toxoplasma Igs were detected with either anti-mouse IgG H+L, anti-mouse IgG1, or anti-mouse IgG2a (Jackson ImmunoResearch, Inc.). All secondary antibodies were HRP conjugated and developed as described for cytokine ELISA.

Adoptive transfer and T lymphocyte antibody depletion

All procedures were performed 1 month after final cps1-1 immunization. Spleens were harvested and after disruption and RBC lysis, CD19⁺ B cells or CD8⁺ T cells were purified using EasySep Mouse positive selection kits (StemCell Technologies Inc.) (29). Naïve WT mice received either 1 × 10⁷ CD8⁺ T cells, 5.0 × 10⁶ CD19⁺ B cells, 4 × 10⁷ immune or naive splenocytes (4 × 10⁷ splenocytes contained approximately 2 × 10⁶ CD8⁺ T cells and 4 × 10⁶ CD4⁺ T cells) via tail vein injection. 24 hours after cell transfer mice were i.p. challenged with 1 × 10³ RH tachyzoites (31). T cell subpopulations were depleted by i.p. injection of 500 µg of either control Rat IgG, anti-CD4 antibody (GK1.5), anti-CD8 antibody (TIB210), or both anti-CD4/CD8 antibody on days -3, -2, -1, 0 then every other day until day 21 post challenge (32). On Day 0 of antibody treatment T cell depletion was confirmed and mice were i.p. challenged with 1 × 10³ RH.

Quantitative real-time PCR for parasite equivalents

Spleens and brains from infected animals were harvested and DNA was extracted from entire organ using a Qiagen DNeasy Tissue Kit (QIAGEN Sciences, Inc.) and pooled. Amplification of parasite DNA from 400 ng of purified tissue DNA was performed using primers specific for the Toxoplasma gondii B1 gene (Forward primer GGAACTGCATCCGTTCATG and Reverse primer TCTTTAAAGCGTTCGTGGTC) at 10 pMol of each per reaction (12, 33) (Integrated DNA Technologies) and amplified by real-time flourogenic PCR using SMartMix HM (Cepheid, Sunnyvale) on a Cepheid Smart Cycler. Each reaction contained 1 lyophilized SMartMix HM bead, and 1:20,000 SYBR Green I (Cambrex Bio Science Rockland Inc.). Parasite equivalents were determined by extrapolation from a standard curve.

Statistical analysis

The Kaplan-Meier product limit test was used to measure significant differences between survival curves of i.v. and ex vivo loaded cell type route experiments (GraphPad PRISM software). All other means comparisons were subject to a Students t-test and are represented as the mean ± SEM.

Cps1-1 tachyzoites induce protective immunity capable of clearing a high dose lethal Type I RH challenge

A single inoculation of live-attenuated cps1-1 tachyzoites protect Type II T. gondii resistant BALB/C mice against a low dose lethal Type I challenge (1). Other studies report that a single inoculation of irradiation inactivated cps1-1 induces immunity in Tyk2−/− and C57Bl/6 mice in a lethal challenge model (34). We sought to determine the mechanism of immune protection and efficacy of cps1-1 induced immunity to clear a challenge infection in the highly sensitive C57BL/6 mouse background. Cps1-1 immunized C57Bl/6 mice were completely protected (Fig. 1A) and completely clear lethal Type I challenge (Fig. 1B) whereas all naïve mice uniformly succumbed to infection by day 10 post challenge and had high parasite burdens. All cps1-1 immunized mice regardless of challenge dose were continuously monitored for 18 months and uniformly survived challenge infection to old age showing that cps1-1 induces long-lasting immunity (data not shown).

Intraveinous immunization with cps1-1 tachyzoites alone or ex-vivo cps1-1 infected DCs or PECs induce long lasting protective immunity

In determining vaccine effectiveness numerous studies have investigated different routes and cell specific requirements for successful immunization (30, 35, 36). To explore the importance of route and cell type specific effects on the ability of cps1-1 vaccination to elicit long lasting immunity, mice were immunized i.v. once with cps1-1 alone, ex vivo cps1-1 infected DCs, PECs or macrophages derived from resident PECs and groups of mice were challenged at 2 (Fig. 2A) or 6 months (Fig. 2B) post immunization with a lethal dose of RH tachyzoites (LD100) and monitored for survival. Immunization with ex vivo cps1-1 infected DCs, PECS, or PEC derived macrophages resulted in nearly complete survival of RH challenged mice at 2 months post immunization and the afforded protection was similar to mice immunized with tachyzoites of cps1-1 (Fig. 2A). In contrast, when lethal challenge was administered at 6 months (Fig. 2B) significant differences in percent survival of mice immunized with ex vivo cps1-1 infected DCs (83% survival), PECs (50% survival), and PEC derived macrophages (33% survival) were observed. Percent survival of mice immunized with ex vivo cps1-1 infected DCs or PECs was not statistically different from mice immunized i.v. with cps1-1 tachyzoites alone. While peritoneal macrophages loaded with cps1-1 were effective at inducing short-term protection at 2 months, they were significantly less effective at inducing long lasting immunity (6 Month p=0.02) (Figs. 2A and B). Short-term protection to lethal RH challenge (LD1000) was also effectively induced following cps1-1 immunization of IL-15−/− mice (Fig. 2C)

Long lasting immunity elicited by cps1-1 immunization is primarily Th-1 biased, CD8⁺ T cell mediated and can be adoptively transferred

The Th-1 biased inflammation produced by infection with T. gondii leads to primarily CD8⁺ T cell mediated immunity and control of chronic infection (23, 26, 37, 38). To examine if cps1-1 induced immunity is also CD8⁺ T cell mediated, we utilized antibody depletion of CD4⁺ and CD8⁺ T cell subsets from immunized mice and measured T cell depleted mouse survival following lethal Type I RH challenge. In addition, we measured if B lymphocytes were also involved as effectors in cps1-1 induced immunity by immunizing B cell deficient mice (µMT) then measuring their survival against lethal challenge. We observed that all of the anti-CD4⁺ and control Ig treated mice survived whereas only 25% of either CD8⁺ or CD8⁺ and CD4⁺ depleted mice survived the challenge infection (Fig. 3A). Although cps1-1 immunized µMT mice survived longer than non-immunized naïve µMT mice, all immunized µMT mice died by day 27 post challenge (Fig. 3B). These data are consistent with previous results indicating that B cells may be required as effector cells or are necessary for the development or recall of an effective memory CD8⁺ T cell population (39–41).

To confirm that cps1-1 induced CD8⁺ T cells were the primary effector mechanism of immune protection we adoptively transferred naïve total splenocytes or purified CD8⁺ T cells, CD19⁺ B cells, or total splenocytes from cps1-1 immunized mice into naïve recipients and measured their survival following lethal Type I RH challenge. Mice receiving cps1-1 immune splenocytes, purified CD19⁺ B cells, or naïve splenocytes succumbed by day 10 post RH challenge. By contrast, all naïve mice receiving purified CD8⁺ T cells from cps1-1 immunized mice survived challenge infection (Fig. 3C). To determine if cps1-1 immunization induced a Th-1 biased response, we measured the anti-toxoplasma serum titre of IgG2a as an indicator of Th-1 biased isotype switching (42, 43). Consistent with reports from other T. gondii vaccine models, high levels of IgG2a were measured (Table I) (30, 44, 45). Our results demonstrate that a Th-1 biased CD8⁺ T cell mediated immune response is the major mechanism required for the protective immunity induced by immunization with cps1-1.

Inflammatory cells are recruited early into the peritoneum with a dynamic pattern in the absence of replicating and disseminating T. gondii infection

During T. gondii infection specific inflammatory cells infiltrate into the site of infection and can indicate the type of immune response occurring and provide mechanisms of direct or indirect control of active infection (4, 16, 17, 46, 47). We examined the total PEC number and kinetics of specific inflammatory cell infiltration by flow cytometry in response to non-replicating cps1-1 infection compared RH infection (Fig. 4). After i.p. infection, a significant (p=0.005) increase in total PEC numbers occurs by day 2 post cps1-1 inoculation followed by a steady increase (p<0.001) through day 8 as compared to day 0 naïve controls (Fig. 4B). By contrast, in RH infection an unexpected early decrease was followed by a greater after number of inflammatory cells infiltrating into the peritoneum (1.5 fold over cps1-1 (p=0.002), by day 6 post infection) (Fig. 4B).

Inflammatory cells such as granulocytes (PMNs), macrophages, and B and T lymphocytes are important for the development of protective immunity and for direct control of T. gondii infection (1, 13, 16, 47). To determine what cell types may be contributing to the development of immunity we used flow cytometry based differential cell analysis of PECs from mice infected with either cps1-1 or RH. We characterized three general cell populations composing total PECs (Fig 4A). In cps1-1 infection we observed a significant increase in the absolute number of GR1⁺ CD68⁻ granulocytes (p=0.03) by day 2 post inoculation with a 9-fold increase after which GR1⁺ CD68⁻ granulocytes return to uninfected control levels. By contrast, during RH infection the absolute numbers of GR1⁺ CD68⁻ granulocytes did not rise until day 4 (d0 vs. d4 6.8-fold, p=0.001) and reached maximal numbers at day 6 with a 6.0 fold increase from day 4 ( p=0.001) (Fig. 4C). Measurement of percent of total PECs and direct examination of Diff-quik stained cytospins confirmed these results for both cps1-1 (day 2) and RH (day 4–6) infections and the latter reveal by nuclear morphology that these GR1⁺ CD68⁻ cells are neutrophils (data not shown).

We measured the kinetics of CD68⁺ cells (resident peritoneal macrophages and monocytes) and GR1⁺ CD68⁺ inflammatory monocytes infiltrating after cps1-1 or RH infection. The absolute numbers of CD68⁺ cells infiltrating into the peritoneum with cps1-1 infection increases 1.5-fold (p<0.001) by day 4 and a 2.1 fold (p=0.001) from day 0 controls by day 6 post infection (Fig. 4D). Even though early after RH infection the absolute numbers of CD68⁺ cells decreased their numbers significantly increased through day 6 (d4, p<0.001, d6, p=0.03) post infection (Fig.4D). However, the rate of influx measured as CD68⁺ cells as a percentage of total PECs for either cps1-1 or RH infection did not significantly change over time (data not shown). Analysis of the CD68⁺ population revealed an interesting pattern of cellular infiltration. Two days after cps1-1 infection we observe a 168-fold increase (p<0.001) in the absolute numbers of GR1⁺ CD68⁺ inflammatory monocytes (Fig. 4E). This rapid increase was followed by a 2.5 fold decrease (p<0.001) by day 4 then a second stage of GR1⁺ CD68⁺ monocytes infiltrated into the site of cps1-1 inoculation at day 6 (2.5-fold increase, p=0.002) (Fig. 4E). In RH infection, GR1⁺ CD68⁺ inflammatory monocyte infiltration is delayed until day 4 (p<0.001) and increases continuously through day 8 post infection (Fig. 4E). The > 99 % replacement of total CD68⁺ macrophages (Fig. 4D and E) suggests some cells may express GR1⁺ in response to infection, however, in vitro infection of peritoneal derived macrophages with cps1-1 under replicating or non-replicating conditions, did not elicit the expression of GR1 (data not shown).

Analysis of CD19⁺ B lymphocytes reveals a similar pattern of flux for either cps1-1 or RH inoculation only through day 4 post infection after which the pattern is distinctly different. The absolute number of CD19⁺ B cells increases 2 fold (p=0.001) by day 6 and is maintained post cps1-1 infection while by day 6 with RH infection, a steady reduction in absolute numbers occurs (day 4, p=0.01). This reduction continues to near undetectable levels by day 8 post RH infection (Fig. 4F).

When measuring total T cells by CD3⁺ as well as CD3⁺ CD4⁺ and CD3⁺ CD8⁺ T cell flux, our analyses revealed that the absolute number of total CD3⁺ T cells after cps1-1 infection significantly increases early in the peritoneum 1.9 fold (p=0.02) by day 2 post infection and continues to increase through day 8 post infection (Fig. 4G). By contrast, total CD3⁺ T cells did not increase significantly until day 4 post RH infection (Fig. 4G). We observe that despite lower total PEC numbers after cps1-1 infection the absolute numbers of CD3⁺ T cells were significantly greater (d2 p=0.004, d4 p<0.001, d6 p<0.001, d8 p<0.001) than during RH infection each time point (Fig. 4G). The absolute numbers of CD3⁺ CD4⁺ T cells markedly increase in numbers at day 6 post cps1-1 infection (Fig. 4H). Interestingly, CD3⁺ CD8⁺ T cells increase in absolute numbers earlier than CD3⁺ CD4⁺ T cells at day 2 (2.8-fold, p=0.007) and day 4 (2.2-fold, p=0.01) post cps1-1 infection after which absolute numbers do not significantly change (Fig. 4I). By contrast, during RH infection, CD3⁺ CD8⁺ T cells have the greatest increase in absolute numbers at the site of infection while CD3⁺ CD4⁺ T cells remain nearly unchanged (Fig. 4H and I).

Systemic production of Ifn-γ is low and transient after cps1-1 inoculation

Previous studies of immunity employing replicating and disseminating T. gondii strains revealed high level systemic production of Ifn-γ, IL-12p40 and variable levels of IL-12p70 production that is dependent on parasite genotype (Type I, II, II) (5, 11, 47). Replication and dissemination during acute infection is likely to enhance the lethal overproduction of inflammatory cytokines (48, 49). Therefore, the level of pro-inflammatory cytokines necessary for the development of immunity is unclear. We measured systemic (serum), levels of pro-inflammatory Th-1 cytokines IL-12p40, IL-12p70, and Ifn-γ by ELISA in mice infected with cps1-1 or RH at days 0, 2, 4, 6, and 8 post infection. Infection with replicating and disseminating T. gondii is known to result in enhanced inflammation arising from host cell lysis and tissue damage during acute infection (11, 48, 50). As shown in Figure 5A and C, systemic Ifn-γ and IL-12p40 in serum of mice infected with RH did not significantly increase until day 4 post infection and quickly rose to extremely high levels by day 6 and 8. Confirming previous reports, systemic IL-12p70 was not significantly produced in response to RH over the course of infection (Fig. 5B) (11). By contrast, systemic Ifn-γ is detected at a very low level only early (D2, p=0.03) (Fig. 5A), and is not detected systemically after D2 post cps1-1 infection. and through day 8 post cps1-1 infection. IL-12p70 production was induced early (day 2) and continued to increase through day 6 (p=0.04) post cps1-1 infection (Fig. 5B) and was significantly higher than in response to RH (p<0.001) for the time points of day 4, 6, and 8, post infection. Interestingly, essentially all systemic IL-12 produced in response to cps1-1 was in the form of bioactive IL-12p70 (Fig. 5B and C). By day 6 and 8 post RH infection little bioactive IL-12p70 was present even though a large excess of IL-12p40 was detected systemically (Fig. 5B and C). Additionally, while systemic Ifn-γ was only detectable at a low level day 2 post infection, Ifn-γ is essential for immunity in this vaccine model (Fig. 5D).

Ifn-γ and IL-12p70 production is primarily localized to the site of cps1-1 inoculation

Because systemic production of Ifn-γ was low and transient after cps1-1 infection we measured the kinetics of local (PEC) and peripheral (spleen) Ifn-γ and IL-12p70 production induced by cps1-1 or RH infection. PECs from mice infected with RH produced moderate Ifn-γ detectable (p<0.001) at day 4 which was maintained through day 8 post infection (Fig. 6A). Moderate levels of IL-12p70 were produced by PECs from RH infected mice beginning on day 4 then decreasing through day 8 (Fig. 6C). Splenocytes from RH infected mice produced Ifn-γ at many fold greater levels than PECs (Fig. 6A and B) as previously observed (16, 48). Ifn-γ production by splenocytes significantly increased by day 4 post infection (d4 vs. d2, p=0.001) and continued to increase through day 8 (Fig. 6B). Minimal splenocyte production of IL-12p70 was observed on day 4 and day 6 post RH infection (Fig. 6D).

After cps1-1 infection, a novel and primarily local pattern of cytokine production is observed (Fig. 6). PEC Ifn-γ production was detectable as early as day 2 and was maximal by day 6 (d4 vs. d6, p=0.04) followed by a rapid decrease by day 8 after cps1-1 infection. Interestingly, PECs from cps1-1 infected mice produced significantly greater Ifn-γ (p=0.005) than PECs from RH infected mice by day 6 post infection (Fig. 6A). In Figure 6C we reveal that PECs from cps1-1 infected mice rapidly produce significantly greater levels of IL-12p70 by day 2 post infection as compared to RH (p=0.007). These IL-12p70 levels remained consistently high (days 2–8) after cps1-1 inoculation, while IL-12p70 production from PECs after RH infection was delayed to day 4 and significantly lower by day 6 and 8 with p=0.012 and p=0.001, respectively (Fig. 6C). PECs cytokine production, on a per cell basis, of Ifn-γ (Fig. 6A and B) and IL-12p70 (Fig. 6C and D) was higher than that of splenocytes from cps1-1 infected mice. In contrast to hypervirulent RH infection, Th-1 cytokines including IL-12p70 were produced more rapidly and at greater levels locally in response to cps1-1 infection.