TgMAPK1 is a Toxoplasma gondii MAP kinase that hijacks host MKK3 signals to regulate virulence and interferon-γ-mediated nitric oxide production

Abstract

The parasite Toxoplasma gondii controls tissue-specific nitric oxide (NO), thereby augmenting virulence and immunopathology through poorly-understood mechanisms. We now identify TgMAPK1, a Toxoplasma mitogen-activated protein kinase (MAPK), as a virulence factor regulating tissue-specific parasite burden by manipulating host interferon (IFN)-γ-mediated inducible nitric oxide synthase (iNOS). Toxoplasma with reduced TgMAPK1 expression (TgMAPK1(lo)) demonstrated that TgMAPK1 facilitates IFN-γ-driven p38 MAPK activation, reducing IFN-γ-generated NO in an MKK3-dependent manner, blunting IFN-γ-mediated parasite control. TgMAPK1(lo) infection in wild type mice produced ≥ten-fold lower parasite burden versus control parasites with normal TgMAPK1 expression (TgMAPK1(con)). Reduced parasite burdens persisted in IFN-γ KO mice, but equalized in normally iNOS-replete organs from iNOS KO mice. Parasite MAPKs are far less studied than other parasite kinases, but deserve additional attention as targets for immunotherapy and drug discovery.

Fig. 1

Construction of TgMAPK1^lo tachyzoites. Both sense (control) and antisense knockdown plasmids were constructed from pMiniHXGPRT and were stably transfected into parental (par) T. gondii Prugniaud strain deleted for HXGPRT (hypoxanthine-guanine phosphoribosyltransferase; PruΔHXGPRT). Filled black arrows indicate DNA orientation. The T. gondii α-tubulin (TUB1) and dihydrofolate reductase (DHFR) promoters are indicated by arrows with right angles. 5′ (TUB1) and 3′ (SAG1) untranslated regions are indicated by gray boxes. Clonally-derived sense (TgMAPK1^con; S1 and S2) and antisense (TgMAPK1^lo; AS1 – AS3) recombinant parasites were initially confirmed by genotypic analysis following PCR amplification of the 809 bp amplicon. These reactions were either digested with EcoRI (+) or remained undigested (−) prior to electrophoresis. DNA size is shown in base pairs (bp). Genotypes were ultimately verified by nucleotide sequencing (not shown). Genotype profiles of the parental strain (par.), one representative sense (S2) and antisense (AS3) clone are shown in the ethidium bromide stained gel shown in the inset.

Fig. 2

Phenotypic analysis of TgMAPK1^con (con) or TgMAPK1^lo (lo) tachyzoites. (a) human foreskin fibroblasts were infected with either TgMAPK1^con (con) or TgMAPK1^lo (lo) tachyzoites at a multiplicity of infection of 0.3. Intracellular tachyzoites were harvested seventy-two hours post-infection. Sodium dodecyl sulfate polyacrylamide gel electrophoresis and Western blotting was performed on 10⁷ tachyzoites, and TgMAPK1 was detected using a rabbit anti-TgMAPK1 antibody. T. gondii β-tubulin served as the loading control. (b) Tachyzoite growth in human foreskin fibroblasts was assessed in triplicate by [³H]uracil incorporation as a function of time.

Fig. 3

Reduced TgMAPK1 expression decreases parasite tissue burden independent of serum IFN-γ; (a) WT (n=10) or IFN-γ KO (n=5) mice were challenged intraperitoneally with 1,000 TgMAPK1^con (con) or TgMAPK1^lo (lo) tachyzoites and sacrificed one week later. Parasite burden by qPCR was compared using mixed-effects methods with the R package “nlme” (Pinheiro et al., 2008) adjusted for variations; (b) serum IFN-γ and IL-10 from mice in panel (a). Control serum from a naïve (uninf.) mouse is shown for comparison. Symbols represent individual mice.

Fig. 4

IFN-γ differentially regulates proliferation of TgMAPK1^lo versus TgMAPK1^con tachyzoites in vitro through an NO-dependent, IDO-independent mechanism. Mouse macrophages were infected with TgMAPK1^con (con) or TgMAPK1^lo (lo) tachyzoites at a multiplicity of infection (MOI) of 0.3. Tachyzoite proliferation was assessed in triplicate by [³H]uracil incorporation. 100% proliferation represents [³H]uracil incorporation without exogenous IFN-γ. One representative experiment of 2 – 4 experiments with similar results is shown; (a) tachyzoite proliferation in bone marrow-derived macrophages (BMDM) in the absence of exogenous IFN-γ as a function of time; (b) tachyzoite proliferation in BMDM assessed 52 h post-infection as a function of IFN-γ concentration (added 16 h post-infection). Assessments at 40 and 64 h yielded similar results (not shown); (c) tachyzoite proliferation in J774A.1 macrophages as a function of IFN-γ concentration (added 16 h post-infection). Indicated cells were treated with 1 mM N^G-monomethyl-L-arginine (L-NMMA) or 1-methyltryptophan (1-MT) two hours prior to infection. Proliferation was assessed 52 h post-infection. The mean ± standard error of the mean values is shown along with the P value for differences between proliferation curves by ANOVA; (d) tachyzoite proliferation in J774A.1 macrophages in the absence of exogenous IFN-γ as a function of time; (e) means of triplicate growth velocities between 40 and 52 hours for the experiment shown in panel (d), normalized to untreated TgMAPK1^lo tachyzoites (relative rate of proliferation = 1). Bars are standard error of the means. Differences compared by t-test.

Fig. 5

TgMAPK1-dependent IFN-γ-mediated NO regulation depends on host p38 MAPK activation; (a) bone marrow-derived macrophages (BMDM) from p38^fl/fl LysM-Cre⁻ (expressing WT p38 MAPK levels) or p38^fl/fl LysM-Cre⁺ mice (lacking p38 MAPK) were infected at a multiplicity of infection (MOI) of 0.3 with TgMAPK1^con (con) or TgMAPK1^lo (lo) tachyzoites and treated 16 h later with (+) or without (−) IFN-γ (3 ng/mL). Western blotting and enhanced chemiluminescence was performed 52 h post-infection, with one of two representative experiments being shown. Mouse α-tubulin (α-TUB) expression served as loading control. T. gondii proliferation was assessed by β-tubulin (β-TUB) expression. Densitometric ratios showing phospho-p38 induction (p-p38/total p38) and tachyzoite proliferation (T. gondii β-TUB/mouse α-TUB) are shown in Table 1; (b) RAW264.7 or J774A.1 macrophages were infected at an MOI of 0.3 and treated with IFN-γ 16 h later. NO was measured 52 h post-infection. The P value for comparison of curves by ANOVA is shown with the mean ± standard error of the mean; (c) BMDM from p38^fl/fl LysM-Cre⁻ or p38^fl/fl LysM-Cre⁺ mice were treated with 1 mM N^G-monomethyl-L-arginine (L-NMMA) two hours before being infected, and analyzed as described in panel (a). Densitometric ratios showing phospho-p38 induction (p-p38/total p38) and tachyzoite proliferation (T. gondii β-TUB/mouse α-TUB) are shown in Table 2.

Fig. 6

TgMAPK1 affects NO concentration without sensitizing tachyzoites to NO and iNOS deficiency equalizes parasite burden in tissues normally iNOS-replete; (a) bone marrow-derived macrophages (BMDM) from WT mice were infected at a multiplicity of infection (MOI) of 0.3 with TgMAPK1^con or TgMAPK1^lo tachyzoites and treated with 0 – 500 μM S-nitroso-N-acetylpenicillamine (SNAP); (a) tachyzoite proliferation by [³H]uracil incorporation was assessed as a function of delivered SNAP concentration at 52 h post-infection; (b) proliferation in panel (a) versus actual [NO₂⁻] in the culture supernatant. 100% proliferation represents [³H]uracil incorporation without SNAP. Means ± standard error of the means from triplicate determinations is shown. P values are comparisons of curves by ANOVA; (c) WT (n=9-10) or iNOS KO (n=5) mice were challenged with 1,000 TgMAPK1^con (con) or TgMAPK1^lo (lo) and sacrificed one week later. Parasite burden by qPCR was compared using mixed-effects methods with the R package “nlme” (Pinheiro et al., 2008) adjusted for variations. Symbols represent individual mice.

Fig. 7

TgMAPK1-mediated control of parasite proliferation is iNOS and MKK3-dependent; (a) bone marrow-derived macrophages (BMDM) from iNOS KO mice were infected with tachyzoites at a multiplicity of infection (MOI) of 0.3 and treated with IFN-γ 16 h later. Proliferation was assessed 52 h post-infection by [³H]uracil incorporation as a function of IFN-γ concentration. 100% proliferation represents [³H]uracil incorporation in the absence of IFN-γ. P value compares curves by ANOVA; (b) proliferation in the absence of exogenous IFN-γ over time. Means ± standard error of the means is shown. P value compares curves by ANOVA; (c) BMDM were infected with yellow fluorescent protein (YFP)⁺ TgMAPK1^con (con) or TgMAPK1^lo (lo) tachyzoites at a MOI of 0.3. CD11b⁺YFP⁺ (infected) cells were sorted 52 h post-infection, total RNA isolated, and quantitative RT-PCR for T. gondii HSP70 was performed, normalized to T. gondii GAPDH. The mean ± standard error of the mean is shown along with the P value (Student’s t-test); (d) BMDM from MKK3 KO mice were infected, treated, analyzed and presented as described in Fig. 5a, along with the densitometric ratios showing phospho-p38 induction (p-p38/total p38) in Table 3; (e) MKK3 KO BMDM were infected and treated with IFN-γ as described in panel (a); (f) proliferation in panel (e) in the absence of exogenous IFN-γ over time. Mean of triplicates ± standard errors of the mean is shown and P values compare curves by ANOVA.

Fig. 8

TgMAPK1-dependent virulence is IFN-γ, iNOS, and MKK3-dependent; (a) WT mice (n=10) were challenged with 50,000 TgMAPK1^con or TgMAPK1^lo tachyzoites and survival was assessed by the Kaplan-Meier method and compared by the log rank test. This inoculum was chosen for survival studies involving immunocompetent wild type mice because it consistently caused 100% mortality while demonstrating a significant difference between TgMAPK1^con or TgMAPK1^lo infection. Due to the fact that IFN-γ, iNOS, and MKK3 KO mice were highly susceptible to T. gondii infection, IFN-γ KO (n=8) (b), iNOS KO (n=5–6) (c), or MKK3 KO (n=6–8) (d) mice were challenged with a much lower inoculum (10,000 tachyzoites) and survival assessed as in panel (a). For panels (b) – (d), inocula as high as 50,000 tachyzoites were also examined but this caused very rapid mortality between 6 – 8 days post-infection without demonstrating any statistically significant differences between TgMAPK1^con or TgMAPK1^lo infection (not shown).