Giardia duodenalis arginine deiminase modulates the phenotype and cytokine secretion of human dendritic cells by depletion of arginine and formation of ammonia

Abstract

Depletion of arginine is a recognized strategy that pathogens use to evade immune effector mechanisms. Depletion depends on microbial enzymes such as arginases, which are considered virulence factors. The effect is mostly interpreted as being a consequence of successful competition with host enzymes for the substrate. However, both arginases and arginine deiminases (ADI) have been associated with pathogen virulence. Both deplete arginine, but their reaction products differ. An ADI has been implicated in the virulence of Giardia duodenalis, an intestinal parasite that infects humans and animals, causing significant morbidity. Dendritic cells (DC) play a critical role in host defense and also in a murine G. duodenalis infection model. The functional properties of these innate immune cells depend on the milieu in which they are activated. Here, the dependence of the response of these cells on arginine was studied by using Giardia ADI and lipopolysaccharide-stimulated human monocyte-derived DC. Arginine depletion by ADI significantly increased tumor necrosis factor alpha and decreased interleukin-10 (IL-10) and IL-12p40 secretion. It also reduced the upregulation of surface CD83 and CD86 molecules, which are involved in cell-cell interactions. Arginine depletion also reduced the phosphorylation of S6 kinase in DC, suggesting the involvement of the mammalian target of rapamycin signaling pathway. The changes were due to arginine depletion and the formation of reaction products, in particular, ammonium ions. Comparison of NH(4)(+) and urea revealed distinct immunomodulatory activities of these products of deiminases and arginases, respectively. The data suggest that a better understanding of the role of arginine-depleting pathogen enzymes for immune evasion will have to take enzyme class and reaction products into consideration.

Fig 1

SDS-PAGE and Western blot analysis of purified recombinant G. duodenalis ADI. (A) SDS-PAGE and Coomassie staining of affinity-purified recombinant catalytically inactive ADI_C424A (lane 1), recombinant enzymatically active ADI (lane 2), and molecular mass standards (lane 3). (B) Antigenic identification of 0.5 μg affinity-purified recombinant ADI (lane 4), 0.5 μg catalytically inactive ADI_C424A (lane 5), and native ADI in 2.9 μg of G. duodenalis strain WB-C6 lysate (lane 6) by Western blotting with alpaca polyclonal antiserum raised against ADI. Preimmune serum, used as a control, did not react with any G. duodenalis protein (data not shown).

Fig 2

Enzymatic arginine depletion by ADI modulates cytokine secretion of LPS-activated human moDC. Immature moDC (5 × 10⁵ per sample) were exposed to the indicated amounts of recombinant ADI or, as a control, corresponding levels of ADI_C424A (only the largest amount is shown), and 1 μg/ml LPS was added. Citrulline content was determined as a measure of the cumulative ADI activity and arginine depletion over the 24-h assay time (A). DC cytokine secretion into supernatants was assessed by ELISA and is expressed as a percentage of the amount secreted by LPS-stimulated cells not exposed to any ADI protein (B). Bars in panel A represent the mean ± SD from experiments with DC prepared from seven different donors. Symbols in panel B represent values from individual donors, and means are indicated by horizontal lines. Differences between the amounts of cytokines secreted by cells exposed to mutant (control) or active ADI were analyzed by paired (two-tailed) t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Fig 3

Enzymatic arginine depletion by ADI reduces CD83 and CD86 surface marker induction by LPS activation on moDC. Cells were treated as described in the legend to Fig. 1, and all of the cells, except nonstimulated control cells, were treated with 1 μg/ml LPS. After incubation for 24 h at 37°C, moDC were harvested, stained with cell surface marker-specific antibodies, and analyzed by flow cytometry. (A) Representative dot plots for CD14/CD83 (top) and CD25/CD86 (bottom) expression with the respective mean fluorescence intensity (MFI) values for CD86- and CD83-positive populations (cells in the lower right and top left quadrants of panel A) for CD86 and CD83, respectively) are shown. wo, without. (B) Relative MFI values as percentages of those of control LPS-stimulated cells for CD83 (n = 3) and CD86 (n = 4) with moDC from different donors are shown. Horizontal lines correspond to mean values, and differences between the respective MFI values on cells exposed to mutant (control) and active ADI were analyzed by paired (two-tailed) t test. *, P < 0.05; **, P < 0.01.

Fig 4

Depletion of arginine and formation of ADI reaction products citrulline and ammonium ions modulate the moDC response to LPS activation. Immature moDC after harvesting were seeded into arginine-free growth medium. The medium was then supplemented as indicated with arginine (Arg), citrulline (Cit), and/or ammonium chloride (2 mM each; brackets indicate addition of ADI to arginine-replete medium to reflect formed, not supplemented products). Cells were activated with LPS and treated with ADI (●) or mutant ADI_C424A (○) as a control. Values correspond to percentages of the respective parameter determined for cells stimulated with LPS only in arginine-replete medium. After 24 h, supernatants were collected and cytokine concentrations were determined by ELISA. Horizontal lines correspond to mean values, and differences between the respective parameter determined with cells exposed to mutant (control) and active ADI were analyzed by paired (two-tailed) t test. *, P < 0.05; **, P < 0.01.

Fig 5

NH₄⁺ reduces IL-10 secretion by LPS-stimulated moDC. Immature moDC were prepared and stimulated with LPS as described in the legend to Fig. 4 in medium supplemented with either arginine alone (control) or arginine plus NH₄⁺Cl. After 24 h, supernatants were taken and cytokine concentrations were determined by ELISA. Cytokine concentrations are expressed as percent differences from the cytokine amounts produced by control cells. Each dot represents an independent experiment with moDC prepared from an individual donor. Significance was tested against the null hypothesis that addition of NH₄⁺ had no effect by paired (two-tailed) t test. P values were significant (P < 0.01) for IL-10 levels and nonsignificant (P > 0.05) for TNF-α levels.

Fig 6

Immunomodulation of LPS-activated moDC undergoing arginine starvation is different between NH₄⁺ and urea. MoDC were prepared as described in the legend to Fig. 4 and stimulated in arginine-free medium supplemented with either 2 mM ammonium chloride or urea. After 24 h, supernatants were collected and cytokine concentrations were determined by ELISA (A). The moDC were harvested, and surface marker proteins were analyzed by flow cytometry (B). Dots represent values from experiments with cells from six (A) or four (B) individual donors and are expressed as percentages of control values obtained with LPS-activated cells with no supplement. *, P < 0.05 (paired [two-tailed] t test); ns, nonsignificant.

Fig 7

ADI-mediated arginine depletion decreases mTOR signaling in LPS-stimulated moDC. Immature moDC (10⁶ in arginine [Arg]-free growth medium) were seeded into each well of a 12-well tissue culture plate. Arginine-free medium (ΔArg) was supplemented or not with 2 mM arginine, and cells were treated or not with 4 μg of ADI, 4 μg of ADI_C424A, or 2 μM rapamycin (Rapa). For activation, 1 μg/ml LPS was added to each sample after 2 h. (A) MoDC CD83 surface marker levels 24 h after stimulation and those of stimulated control cells are compared. Bars represent the mean ± SD of independent experiments with five different donors. *, P < 0.05 (paired [two-tailed] t test compared with the respective control). (B) In parallel, 30 min after LPS stimulation, cells were harvested, washed, and lysed. A 50-μg sample of cell extract was separated by SDS-PAGE, and p70-S6K, phospho-p70-S6K, and β-actin were detected by immunoblotting and quantified by image analysis. Phosphorylated p70-S6K levels were normalized relative to β-actin and compared to those of stimulated control cells. Levels of total p70-S6K were not significantly different between the different experimental groups. Bars represent the mean ± SD from five individual experiments with cells from different donors. *, P < 0.05; **, P < 0.01 (compared with the respective control by paired [two-tailed] t test).