Mast Cell Coupling to the Kallikrein-Kinin System Fuels Intracardiac Parasitism and Worsens Heart Pathology in Experimental Chagas Disease

Abstract

During the course of Chagas disease, infectious forms of Trypanosoma cruzi are occasionally liberated from parasitized heart cells. Studies performed with tissue culture trypomastigotes (TCTs, Dm28c strain) demonstrated that these parasites evoke neutrophil/CXCR2-dependent microvascular leakage by activating innate sentinel cells via toll-like receptor 2 (TLR2). Upon plasma extravasation, proteolytically derived kinins and C5a stimulate immunoprotective Th1 responses via cross-talk between bradykinin B2 receptors (B2Rs) and C5aR. Awareness that TCTs invade cardiovascular cells in vitro via interdependent activation of B2R and endothelin receptors [endothelin A receptor (ET_AR)/endothelin B receptor (ET_BR)] led us to hypothesize that T. cruzi might reciprocally benefit from the formation of infection-associated edema via activation of kallikrein-kinin system (KKS). Using intravital microscopy, here we first examined the functional interplay between mast cells (MCs) and the KKS by topically exposing the hamster cheek pouch (HCP) tissues to dextran sulfate (DXS), a potent "contact" activator of the KKS. Surprisingly, although DXS was inert for at least 30 min, a subtle MC-driven leakage resulted in factor XII (FXII)-dependent activation of the KKS, which then amplified inflammation via generation of bradykinin (BK). Guided by this mechanistic insight, we next exposed TCTs to "leaky" HCP-forged by low dose histamine application-and found that the proinflammatory phenotype of TCTs was boosted by BK generated via the MC/KKS pathway. Measurements of footpad edema in MC-deficient mice linked TCT-evoked inflammation to MC degranulation (upstream) and FXII-mediated generation of BK (downstream). We then inoculated TCTs intracardiacally in mice and found a striking decrease of parasite DNA (quantitative polymerase chain reaction; 3 d.p.i.) in the heart of MC-deficient mutant mice. Moreover, the intracardiac parasite load was significantly reduced in WT mice pretreated with (i) cromoglycate (MC stabilizer) (ii) infestin-4, a specific inhibitor of FXIIa (iii) HOE-140 (specific antagonist of B2R), and (iv) bosentan, a non-selective antagonist of ET_AR/ET_BR. Notably, histopathology of heart tissues from mice pretreated with these G protein-coupled receptors blockers revealed that myocarditis and heart fibrosis (30 d.p.i.) was markedly and redundantly attenuated. Collectively, our study suggests that inflammatory edema propagated via activation of the MC/KKS pathway fuels intracardiac parasitism by generating infection-stimulatory peptides (BK and endothelins) in the edematous heart tissues.

Keywords: Chagas disease; G protein-coupled receptors; Trypanosoma cruzi; bradykinin; endothelin; kallikrein; mast cells.

Figure 1

Bradykinin-induced microvascular leakage is propagated via coupling of mast cell and contact phase factors. (A) Enzyme assays performed with high molecular weight kininogen mimetic substrate detect plasma kallikrein (PKa) activity in fresh hamster plasma (citrated) incubated with the contact activator dextran sulfate (DXS) (20 nM). The kinetics of hydrolysis was performed in the presence/absence of the PKa inhibitor PKSI-527. Substrate hydrolysis was monitored by the increase of fluorescence with time (mean ± SD). Data are representative of three independent experiments run in duplicates. (B–E) Microvascular leakage in the hamster cheek pouch (HCP) sensitized with DXS. (B) HCPs were superfused with DXS (4 µM) for 45 min and fluorescence was visualized at different time-points (representative images, n = 6). Where indicated, the hamsters were pretreated with cromoglycate. (C) Synergism between histamine and DXS was studied after 30 min tissue superfusion, with medium supplemented with DXS 0.4 µM, in the presence or absence of captopril, followed by 5 min of histamine application, i.e., without any interruption of superfusion., DXS/Hist (n = 5), DXS/Capto/Hist (n = 5). Where indicated, the hamsters were pretreated i.p. with cromoglycate (n = 2, and n = 4, in the presence and absence of captopril, respectively). Histamine-induced leakage (n = 12) was inserted (internal controls) for comparison. (D) Evidence that synergism between DXS/histamine leads to leakage via activation of the kallikrein–kinin system was sought in HCP superfused for 10 min with DXS (2 µM) and captopril, followed by 5 min of superfusion with DXS/histamine application (n = 9). Prior to the application of histamine, the DXS-treated HCPs were either treated (superfusion interrupted) with the contact phase inhibitor PdSP15a (n = 6) or with PKa inhibitor PKSI-527 (n = 5) for 5 min. HOE-140 (n = 4), the B2R antagonist (B2RA) was applied together with DXS + Captopril. (E) For plasma reconstitution experiments, HCPs were superfused with DXS + Captopril for 10 min, followed by topical application (superfusion interrupted) of heparinized hamster plasma for 7 min. Plasma (n = 5), and DXS/plasma (n = 8). HOE-140 (n = 4) was applied along with DXS + Captopril before addition of plasma. Data are expressed as: (C–E) Relative fluorescence units (RFUs) (mean ± SD). Statistical analyzes were done by analysis of variance and t-test. In (C), *P < 0.05 DXS/Capto/Hist versus DXS/Hist and ⁺P < 0.05 for DXS/Hist versus DXS/Crom/Hist and D, *P < 0.05 DXS/PdSP15a/Hist versus DXS/Hist.

Figure 2

A subtle destabilization of the endothelial barrier potentiates infection-associated inflammation. (A–C) Hamster cheek pouches (HCPs) were superfused with captopril for 10 min, followed by application of histamine (4 µM) along with tissue culture trypomastigotes (TCTs) (3 × 10⁷) during 10 min of interrupted superfusion (n = 12). TCT control group (n = 8) involved HCP treatment with captopril in the absence of histamine. Prior to TCT/histamine application, pharmacological interventions were performed with cromoglycate (pretreatment, n = 6); HOE-140 (B2RA; n = 5) and PKSI-527 (n = 5), applied for 5 min. In A at 60 min after TCT application histamine 4 µM (n = 8) was applied as an internal control and these values were then added to the TCT values (real sum – green curve) and used for the calculation of a synergistic effect of histamine and TCT. Results of histamine 4 µM (n = 10) applied to HCPs in steady state were inserted as a secondary control. Data are expressed as follows: (A) relative fluorescence units (RFUs) over time (mean ± SD), (B) area under the curve (AUC ± SD) measured during 30 min. Statistical analysis were done by analysis of variance (ANOVA) and Wilcoxon. In (A) *P < 0.05 (TCT/Hist versus TCT) and ⁺P < 0.05 (TCT/Hist versus real sum). In (B) *P < 0.05 (TCT/Hist versus TCT or TCT/Hist/inhibitor/antagonist) (C) HCPs were superfused with captopril for 10 min, followed by topical application of different doses of TCTs (60, 600, 6,000, and 60,000 TCTs/μl) and histamine (n = 4) during 10 min of interrupted superfusion (dark gray). At 60 min after TCT application, histamine was applied during 5 min as a control (light gray). Data are expressed as maximal increase in RFU (mean ± SD). Statistical analyzes were done by ANOVA and Wilcoxon (*P < 0.05, TCT/Capto/Hist versus histamine; TCT/Capto/Hist versus TCT; ⁺P < 0.05, TCT 60/Capto/Hist versus TCT 6000/Capto/Hist).

Figure 3

Tissue culture trypomastigotes (TCTs) evoke paw edema via the mast cell (MC)/kallikrein–kinin system pathway. (A) WT (B6-Kit^+/+) or MC-deficient mice (B6-kitW-sh/W-sh) were infected with TCTs (10⁶). The contralateral paw received PBS. When indicated, mice were pretreated (−1 h) with the B2R antagonist (B2RA) (HOE-140). (B,C) Infected BALB/c mice were pretreated (−1 h) with HOE-140, cromoglycate (MC stabilizer), histamine H1 receptor (H1R) antagonist (mepyramine) or factor XII (FXII) inhibitor (infestin-4). Captopril was used (i.p.) to inhibit kinin degradation. Edema was measured at 3 and 24 h and results are expressed as difference in volume (μl) between infected and contralateral paw (mean ± SD). Data are representative of three independent experiments (five mice/group). Statistical analyzes were done by analysis of variance with Bonferroni post test (*P < 0.05; **P < 0.01, ***P < 0.001).

Figure 4

Activation of the kallikrein–kinin system fuels heart parasitism. Measurements of Trypanosoma cruzi DNA (3 d.p.i.) in the heart of C57BL/6 infected [10⁶ tissue culture trypomastigotes (TCTs)] intracardiacally (A,C,D), WT (B6-Kit^+/+) mice or mast cell deficient (B6-kitW-sh/W-sh) mice (B). Where indicated, the mice were pretreated with cromoglycate, B2R antagonist (B2RA) (HOE-140), endothelin A receptor (ET_AR)/endothelin B receptor (ET_BR) double antagonist (bosentan) or factor XII (FXII) inhibitor (infestin-4). Results (mean ± SD) of T. cruzi DNA by quantitative polymerase chain reaction (qPCR) are representative of two independent experiments (five mice/group). Statistical analyzes were done by analysis of variance with Bonferroni post test (*P < 0.05; **P < 0.01, ***P < 0.001).

Figure 5

Early targeting of BRs and ETRs inhibit myocarditis/fibrosis. C57BL/6 mice were infected intracardiacally with tissue culture trypomastigotes (TCTs) (10⁶) or PBS (mock). Where indicated, mice were pretreated with B2R antagonist (B2RA) (HOE-140), or endothelin A receptor (ET_AR)/endothelin B receptor (ET_BR) antagonist (bosentan). At 30 d.p.i., heart sections were stained with (A) hematoxylin and eosin for cellular infiltration, (B) Safranin for mast cell detection, or (C) Picrosirius Red for collagen deposition. Bars, 50 µm. Data (mean ± SD) are representative of four independent experiments (five mice/group). Statistical analyzes were done by analysis of variance (*P < 0.05; **P < 0.01, ***P < 0.001).

Figure 6

Endothelial barrier breakdown via the mast cell (MC)/kallikrein–kinin system (KKS) pathway fuels Trypanosoma cruzi infectivity: working hypothesis. (1) Tissue culture trypomastigotes liberated from ruptured pseudocysts initiate inflammation upon sensing by toll-like receptor 2 (TLR2)-expressing tissue macrophages. (2) Secreted CXC chemokines evoke a subtle plasma leakage following CXCR2-dependent activation of endothelium/neutrophils (29, 38); although not explored in this infection model, it has been reported that TNF-α (57) is critically involved in neutrophil-evoked plasma leakage. (3) The diffusion of plasma proteins, including native C5 and kininogens, leads to cruzipain-mediated release of kinins and C5a anaphylatoxin in the parasite-laden tissues (31). (4) Results presented in the current work suggest that cardiac MCs activated by C5a and/or endothelin-1 (not shown in this part of the illustration) release histamine/polyphosphates (PolyP) (contact system activators) from their granules. (5) Activation of histamine H1 receptor (H1R) leads to the exposure of plasma-borne contact factors [factor XII (FXII), high molecular weight kininogen (HK) and plasma prekallikrein] to heparin/PolyP. Further downstream, the activated contact factor plasma kallikrein (PK) releases bradykinin (BK) from HK. (6) BK propagates the leakage response temporally and spatially via iterative cycles of endothelial [bradykinin B2 receptor (B2R)] activation (BK/B2R pathway), plasma leakage, MC degranulation, and contact system activation (MC/KKS pathway). (7) Acting jointly with endothelin-1 (ET-1), the released kinins fuel T. cruzi infectivity by signaling heart cells that naturally overexpress B2R (13) and ET_AR/endothelin B receptor (ET_BR) (33). (8) After endocytic internalization, the flagellated trypomastigotes transform into amastigotes which undergo multiple cycles of binary division in the host cell cytoplasm before transforming into infective trypomastigotes.