Progressive visceral leishmaniasis is driven by dominant parasite-induced STAT6 activation and STAT6-dependent host arginase 1 expression

Abstract

The clinicopathological features of the hamster model of visceral leishmaniasis (VL) closely mimic active human disease. Studies in humans and hamsters indicate that the inability to control parasite replication in VL could be related to ineffective classical macrophage activation. Therefore, we hypothesized that the pathogenesis of VL might be driven by a program of alternative macrophage activation. Indeed, the infected hamster spleen showed low NOS2 but high arg1 enzyme activity and protein and mRNA expression (p<0.001) and increased polyamine synthesis (p<0.05). Increased arginase activity was also evident in macrophages isolated from the spleens of infected hamsters (p<0.05), and arg1 expression was induced by L. donovani in primary hamster peritoneal macrophages (p<0.001) and fibroblasts (p<0.01), and in a hamster fibroblast cell line (p<0.05), without synthesis of endogenous IL-4 or IL-13 or exposure to exogenous cytokines. miRNAi-mediated selective knockdown of hamster arginase 1 (arg1) in BHK cells led to increased generation of nitric oxide and reduced parasite burden (p<0.005). Since many of the genes involved in alternative macrophage activation are regulated by Signal Transducer and Activator of Transcription-6 (STAT6), and because the parasite-induced expression of arg1 occurred in the absence of exogenous IL-4, we considered the possibility that L. donovani was directly activating STAT6. Indeed, exposure of hamster fibroblasts or macrophages to L. donovani resulted in dose-dependent STAT6 activation, even without the addition of exogenous cytokines. Knockdown of hamster STAT6 in BHK cells with miRNAi resulted in reduced arg1 mRNA expression and enhanced control of parasite replication (p<0.0001). Collectively these data indicate that L. donovani infection induces macrophage STAT6 activation and STAT6-dependent arg1 expression, which do not require but are amplified by type 2 cytokines, and which contribute to impaired control of infection.

Figure 1. Dominant expression of arg1 in spleen tissue during progressive visceral leishmaniasis.

A) Parasite burden in the spleens of hamsters and mice (n = 6 per group) infected with L. donovani-Luc for 6 weeks. The mean and standard deviation (SD) of the parasite burden, determined by luminometry and interpolation from a standard curve generated by enumeration and luciferase assay of amastigotes from the same tissue, is shown from a single experiment that is representative of 3 independent experiments. B) Expression of NOS2 mRNA in spleens of uninfected (Un) and 6-week L. donovani infected (Inf) mice and hamsters (n = 6 per group). The mean and standard error of the mean (SEM) of the fold-increase of NOS2 mRNA relative to BHK cells, determined by real time RT-PCR in 2 experiments is shown. C) Arginase activity in spleens of uninfected (Un) and 6-week L. donovani infected (Inf) mice and hamsters (n = 6 per group). The mean and standard deviation (error bars) of the tissue arginase activity, determined by assay of urea production, is shown from a single experiment that is representative of 3 independent experiments. D) Expression of arg1 mRNA in spleens of uninfected (Un) and 4-week L. donovani infected (Inf) mice and hamsters. The mean and standard deviation (error bars) of the fold-increase of arg1 mRNA relative to BHK cells, determined by real time RT-PCR in groups of 7 animals, is shown from a single experiment that is representative of 2 independent experiments. E) Expression of L. donovani arginase mRNA in spleens of 4-week L. donovani infected mice and hamsters. The mean and standard deviation (error bars) of the fold-increase of L. donovani arg1 mRNA relative to BHK cells, determined by real time RT-PCR in groups of 7 animals, is shown from a single experiment that is representative of 2 independent experiments. F) Log₁₀ ratio of hamster arg1 mRNA to L. donovani arginase mRNA (open bars) and log₁₀ ratio of hamster arg1 mRNA to hamster NOS2 mRNA (filled bars) in the spleens of 4-week L. donovani infected mice and hamsters. The mRNA expression was determined by real time RT-PCR in groups of 7 animals and used to calculate the mean and SD of the log ratios. The ratios were calculated from the raw data shown in Figs. 1B, 1D, and 1E. The statistical significance of differences in each of the panels is identified by asterisks (** p<0.01; ***, p<0.001).

Figure 2. Kinetics of arginase expression in spleen tissue during progressive visceral leishmaniasis. A

) Parasite burden in the spleens of hamsters (n = 6 per group) infected with L. donovani for 14, 28, 42, and 56 days. The mean and SEM (error bars) of the parasite burden, determined by luminometry and interpolation from an amastigote standard curve, is shown from a single experiment that is representative of 2 independent experiments. B) Arginase activity in the serum (open bars) and spleen tissue (filled bars) of uninfected hamsters (Day 0) and hamsters infected with L. donovani (n = 6 per group) for 7, 14, 28, 42, and 56 days. The mean and SD of the tissue arginase activity, determined by assay of urea production, is shown. C) Hamster arg1 protein expression determined by western blot in spleens of control hamsters (0 days post-infection) and hamsters infected with L. donovani for 7, 14, and 28 days. The expression of GAPDH is shown as a control for protein loading. Each lane contains splenic lysate from a single hamster, with two lanes per time point. The anti-arginase antibody did not react with parasite arginase by immunoblot. D) Time course of expression of hamster arg1 mRNA (filled bars), L. donovani arginase mRNA (empty bars), hamster NOS2 (iNOS) mRNA (hatched bars), and hamster arg2 mRNA (hatched bars) in spleens of control hamsters (0 days post-infection) and hamsters infected with L. donovani for 7, 14, 28, 42, and 56 days. The mean and standard deviation (error bars) of the fold-increase of arginase mRNA relative to BHK cells, determined by real time RT-PCR in groups of 6 animals, is shown from a single experiment that is representative of 3 independent experiments. The statistical significance of differences in each of the panels is identified by asterisks (*, p<0.05; **, p<0.01; ***, p<0.001).

Figure 3. Polyamine content in spleen and liver tissue in L. donovani infected hamsters.

The concentration of polyamines in spleen (A) and liver (B) tissue samples from groups of 6 uninfected hamsters (open bars) and 56 day infected hamsters (filled bars) is expressed as the mean and SD (error bars) of nmol polyamine per mg protein. The data shown are from a single experiment that is representative of 2 independent experiments. The statistical significance of differences in each of the panels is identified by an asterisk (*, p<0.05).

Figure 4. Kinetics of cytokine expression in spleen tissue during progressive visceral leishmaniasis.

Expression of hamster IL-4 (A), IL-10 (B), IL-13 (C), and IL-21 (D) mRNAs in the spleens of control hamsters (0 days post-infection) and hamsters infected with L. donovani for 7, 14, 28, 42, and 56 days. The mean and standard deviation (error bars) of the fold-increase of cytokine mRNA relative to BHK cells, determined by real time RT-PCR in groups of 6 animals, is shown from a single experiment that is representative of 2 independent experiments. The statistical significance of differences in each of the panels is identified by asterisks (*, p<0.05; **, p<0.01; ***, p<0.001).

Figure 5. Arg1 expression and polyamine content in L. donovani infected macrophages.

A) Arginase activity in macrophages isolated by adherence from single-cell suspensions from the spleens of control hamsters (0 days post-infection) and hamsters infected with L. donovani for 7, 14, 21, 42, and 56 days. The mean and SD (error bars) of the arginase activity in 100,000 cells, determined by assay of urea production, is shown from a single experiment that is representative of 2 independent experiments. B) Log₁₀ ratio of host arg1 to NOS2 mRNA in splenic macrophages of 4-week L. donovani infected mice and hamsters. The mRNA expression was determined by real time RT-PCR in groups of 5 animals and used to calculate the log ratios, shown as the mean and SD (error bars) from a single experiment that is representative of 2 independent experiments. The ratios were determined from the following raw data: Arg1 mRNA fold-increase with reference to the BHK cell calibrator (mean ± SD): uninfected mice, 5.7±5.9; uninfected hamster, 3,367±2,858, infected mice, 8.5±10; infected hamster, 131,984±56,407; iNOS mRNA fold-increase with reference to BHK cell calibrator (mean ± SD): uninfected mice, 133.34±182.4; uninfected hamster, 1.03±0.3, infected mice, 2,984±4,535; infected hamster, 19.56±10.87). C) Arginase activity in peritoneal macrophages isolated from mice and hamsters that were uninfected (open bars) or infected in vitro with L. donovani for 48 hrs (filled bars). The mean and SD (error bars) of the arginase activity, determined by assay of urea production, is shown from a single experiment that is representative of 2 independent experiments. D) Arginase activity in hamster peritoneal macrophages that were uninfected (Un) or infected in vitro with L. donovani (1∶5 macrophage:parasite ratio). Parasites were either unopsonized (None), opsonized with normal fresh complement-containing hamster serum (Comp), or opsonized with freeze-thawed hamster serum containing anti-Leishmania antibody (Ab). The mean and standard deviation (error bars) of the arginase activity in 200,000 cells of 6 different samples determined by assay of urea production, is shown from a single experiment that is representative of 2 experiments. Statistical comparisons are made to the control group. E) Expression of hamster arginase mRNA in hamster peritoneal macrophages that were uninfected (Un) or infected in vitro with L. donovani stationary-phase promastigotes for 24 hrs (Inf). The mean and SEM (error bars) of the fold-increase of arg1 mRNA relative to BHK cells, determined by real time RT-PCR in groups of 6 samples per experiment, is shown as data pooled from 4 independent experiments. F) The concentration of polyamines in uninfected (open bars) and 48-hour in vitro infected hamster peritoneal macrophages (filled bars) (n = 6 per group) is expressed as the mean and SD (error bars) of nmol polyamine per mg protein. The data shown are from a single experiment that is representative of 2 independent experiments. The statistical significance of differences in each of the panels is identified by asterisks (*, p<0.05; **, p<0.01; ***, p<0.001).

Figure 6. Parasite-induced host arg1 expression impairs macrophage anti-leishmanial activity.

A) Expression of hamster arginase mRNA in primary hamster splenic fibroblasts that were uninfected (Un) or infected in vitro with L. donovani stationary-phase promastigotes for 24 hrs (Inf). The mean and standard deviation (error bars) of the fold-increase of arg1 mRNA relative to normal BHK cells, determined by real time RT-PCR in groups of 6 animals, is shown from data pooled from 3 independent experiments. B) Arginase activity and NO production in BHK fibroblasts. Arginase activity was determined stimulated with IL-4 or IFN- γ respectively. Arginase activity was determined in BHK cells that were unstimulated (Uns) or stimulated for 48h with IL-4 (10% v/v), LPS (1 µg/mL), or both. The mean and standard deviation (error bars) of the arginase activity of 6 different stimulated samples compared to that of non stimulated cells determined by assay of urea production, is shown from a single experiment that is representative of 2 experiments. NO production was estimated by the measurement of nitrites + nitrates in supernatants of unstimulated cells (Uns) and cells stimulated with IFN-γ (10% v/v of hamster recombinant IFN-γ supernatants) plus 1 µg/mL LPS for 48h. The mean and standard deviation (error bars) of the nitrites/nitrates of 6 different samples determined by Griess assay is shown from a single experiment that is representative of 2 experiments. C) Expression of hamster arginase mRNA in hamster BHK cells that were uninfected (Un) or infected in vitro with L. donovani stationary-phase promastigotes for 24 hrs (Inf) and cultured with or without recombinant hamster IL-4 (10% v/v supernatant) or sham supernatant. The mean and standard deviation (error bars) of the fold-increase of arg1 mRNA relative to BHK cells, determined by real time RT-PCR in groups of 6 animals, is shown from data pooled from 3 independent experiments. D) Expression of hamster arginase mRNA in hamster BHK fibroblasts that were uninfected (Un) or infected in vitro with L. donovani tissue-derived amastigotes for 24 hrs (Inf). The mean and standard deviation (error bars) of the fold-increase of arg1 mRNA relative to BHK cells, determined by real time RT-PCR in groups of 6 animals, is shown from data pooled from 2 independent experiments. E) Effect of cycloheximide (CHX) on parasite-induced hamster arg1 mRNA expression in BHK cells was determined by exposing cells to CHX (20 µg/mL) early (4–12 hrs) and/or late (12–24 hrs) during a 24-hr L. donovani infection. The mean and SD (error bars) of the fold-increase of arg1 mRNA relative to BHK cells, determined by real time RT-PCR in groups of 6 samples, is shown from data from a single experiment that is representative of 2 independent experiments. F) Induction of expression of hamster arg1 in peritoneal macrophages by soluble parasite factors was determined by culturing hamster peritoneal macrophages with stationary phase promastigotes (1∶10 ratio) separated by a 0.4 µ pore size membrane (Falcon). The mean and SEM (error bars) of the fold-increase of arg1 mRNA relative to BHK cells, determined by real time RT-PCR in groups of 6 animals, is shown from data pooled from 3 independent experiments. G) miRNAi-mediated arg1 knockdown in BHK cells. Arg1 protein in BHK cells stably transfected with a miRNAi vector targeting arg1 (Arg-1 KD) or stably transfected with a miRNAi vector coding a control sequence (Control) determined by Western blot using a specific polyclonal antibody raised against hamster arg1. Representative of 2 different blots. H) Effect of miRNAi-mediated knockdown on parasite burden was determined in BHK cells that were transfected with a non-targeting miRNAi vector (control) or a vector specific to hamster arg1. The transfected cells were infected with L. donovani metacyclic promastigotes and the mean and standard deviation (error bars) of the parasite burden at 4, 24, 48 and 72 hrs post-infection, determined by luminometry, is shown from a single experiment that is representative of 2 independent experiments. There was no difference in parasite burden between the two groups at 4 and 24 hrs post-infection. I) Arg1 knockdown did not induce NO production in BHK cells infected with L. donovani. NO production in BHK cells stably transfected with a miRNAi vector targeting Arg1 (Arg1 knockdown) or transfected with an irrelevant miRNAi vector (Non targeted). The mean and standard deviation (error bars) of the nitrites/nitrates released in the supernatant of cells after 48h infection with L. donovani was determined by Griess assay. Data are from a single experiment that is representative of 3 independent experiments. (**p<0.001). The statistical significance of differences in each of the panels is identified by asterisks (*, p<0.05; **, p<0.01; ***, p<0.001).

Figure 7. Role of parasite-induced STAT6 activation in host arg1 expression and L. donovani infection.

A) Phosphorylated STAT6 in spleens of hamsters infected with L. donovani determined by immunoprecipitation followed by western blot of whole splenic lysates. The samples (before immunoprecipation) were probed with anti-STAT6 anti-GAPDH antibodies to confirm equivalent protein loading. The blot shown is from tissue from a single animal, representative of 4 independent experiments. B) Dose-dependent induction of STAT6 activation by L. donovani in BHK cells transfected with a STAT6-luciferase reported vector. Data are presented as the mean and standard deviation (error bars) of the relative light units in uninfected (0 parasites) cells and cells exposed to 1–40 parasites per cell over 48 hrs of culture. Shown is data from a single experiment that is representative of 2 independent experiments. C-D) Phosphorylation of STAT6 in hamster peritoneal macrophages exposed to L. donovani (10 parasites per cell) or IL-4 (10% v/v supernatants) measured as fold-increase of mean fluorescence intensity (MFI) (C) or percent positive cells (D) by flow cytometry. Data are presented as the mean and standard deviation (error bars) and are from a single experiment that is representative of 2 independent experiments. E–F) Phosphorylation of STAT6 in hamster BHK cells exposed to L. donovani (10 parasites per cell) measured as fold-increase of mean fluorescence intensity (MFI) (E) or percent positive cells (F) by flow cytometry. Data are presented as the mean and standard deviation (error bars) and are from a single experiment that is representative of 2 independent experiments. G) Effect of miRNAi-mediated STAT6 knockdown on arg1 and STAT6 mRNA expression. BHK cells were transfected with a non-targeting miRNAi vector (control) or a vector specific to hamster STAT6 and the expression of arg1 and STAT6 mRNA determined by real time RT-PCR in uninfected (open bars) and 24-hr infected cells (filled bars). The data are presented as the mean and standard deviation (error bars) of the fold-increase of arg1 mRNA relative to BHK cells, determined by real time RT-PCR from a single experiment, representative of 2 independent experiments. H) miRNAi-mediated STAT-6 knockdown in BHK cells. The expression of STAT-6 protein in BHK cells stably transfected with a miRNAi vector targeting STAT-6 (STAT-6 KD) or transfected with a miRNAi vector coding a control sequence (Control) was determined by western blot using an anti-STAT6 antibody. An anti-GAPDH antibody was used to confirm equivalent protein loading. Data are from a single blot representative of 2 independent experiments. I) Effect of miRNAi-mediated STAT6 knockdown on parasite burden was determined in BHK cells that were transfected with a non-targeting miRNAi vector (control) or a vector specific to hamster STAT6. The transfected cells were infected with L. donovani metacyclic promastigotes and the mean and standard deviation (error bars) of the parasite burden at 4, 24, 48, and 72 hrs post-infection, determined by luminometry, is shown as data pooled from 2 independent experiments. There was no difference in parasite burden between the two groups at 4 hrs post-infection. The statistical significance of differences between groups in each of the panels is identified by asterisks (*, p<0.05; **, p<0.01; ***, p<0.001).