|Home | About | Journals | Submit | Contact Us | Français|
In this study, we investigated the role of Trypanosoma cruzi invasion and inflammatory processes in reactive oxygen species (ROS) production in mouse atrial cardiomyocyte line (HL-1) and primary adult rat ventricular cardiomyocytes. Cardiomyocytes were incubated with T. cruzi (Tc) trypomastigotes, Tc lysate (TcTL) or Tc secreted proteins (TcSP) for 0-72 h, and ROS measured by amplex red assay. Cardiomyocytes infected by T. cruzi (but not those incubated with TcTL or TcSP) exhibited a linear increase in ROS production during 2-48 h post-infection (max.18-fold increase) which was further enhanced by recombinant cytokines (IL-1β, TNF-α and IFN-γ). We observed no increase in NADPH oxidase, xanthine oxidase, and myeloperoxidase activities, and specific inhibitor of these enzymes did not block the increased rate of ROS production in infected cardiomyocytes. Instead, the mitochondrial membrane potential was perturbed, and resulted in inefficient electron transport chain (ETC) activity, and enhanced electron leakage and ROS formation in infected cardiomyocytes. HL-1 rho (ρ) cardiomyocytes lacked a functional ETC, and exhibited no increase in ROS formation in response to T. cruzi. Together, these results demonstrate that invasion by T. cruzi and inflammatory milieu affect mitochondrial integrity and contribute to electron transport chain inefficiency and ROS production in cardiomyocytes.
Chagasic cardiomyopathy, initiated by Trypanosoma cruzi, accounts for >50,000 deaths and the loss of 2.74 million disability-adjusted life years per year in Latin America. Approximately 30% of infected individuals develop dilated cardiomyopathy that may ultimately lead to heart failure and patients' death .
Several investigators have suggested that Chagas disease is an inflammatory pathology. This idea is supported by the detection of extensive inflammatory infiltrate in the heart, and enhanced expression of interleukin 1β (IL-1β), tumor necrosis factor α (TNF-α) and IL-6 cytokines and a relatively low level of interferon γ (IFN-γ) in the myocardium of T. cruzi-infected experimental models . The macrophages and neutrophils infiltrating the heart are the main source of IL-1β and TNF-α cytokines [4, 5] while CD4+ and CD8+ T cells contribute to IFN-γ production . Oxidative burst of activated phagocytes result in the release of reactive oxygen species (ROS), e.g., superoxide (O2•−), hydrogen peroxide (H2O2), and hydroxyl radical (•OH) via activation of NADPH oxidase (NOX) and/or myeloperoxidase (MPO) enzymes . The inflammatory cytokines and ROS are important for the control of T. cruzi, and may be cytotoxic to the host cellular components.
Many of the ROS are highly reactive and diffusible, and may be released in extracellular milieu. While intracellular ROS serve mainly for host defense against infectious agents, the extracellular release of ROS, when present in abundance, directly damage the surrounding tissues or promote inflammatory processes . We have shown an antioxidant/oxidant imbalance pursue in the heart of T. cruzi-infected mice, and was associated with enhanced oxidative modification of cellular and mitochondrial proteins and lipids . Subsequently, mitochondria isolated from infected murine hearts were shown to produce enhanced ROS , indicating that cardiomyocytes contribute to oxidative stress during T. cruzi infection.
Overall, the published literature has demonstrated that inflammatory and oxidative pathology pursue in the heart during T. cruzi infection. No studies have been conducted to demonstrate whether invasion by T. cruzi or inflammatory processes induce ROS production in cardiomyocytes and whether it is a mitochondrial process. In this study, we have used HL-1 cardiomyocyte cell line and primary adult rat cardiomyocytes to examine whether cardiomyocytes produce ROS in response to T. cruzi infection, or if it is a bystander effect in response to T. cruzi antigens. We treated cardiomyocytes with a variety of inhibitors and specific fluorescent probes to identify the source of ROS release. Since cardiomyocytes in vivo encounter a complex and varying mix of cytokines, we studied the additive and/or synergistic effect of these cytokines on T. cruzi-induced ROS production.
HL-1 cardiomyocytes were cultured in Claycomb media supplemented with 10% FBS, as described . HL-1 rho (ρ) cells were generated by culturing in presence of ethidium bromide (1 μg/ml) and uridine (50 μg/ml) for 59 passages. HL-1 ρ clones were screened for loss of mtDNA by Southern blot analysis. Briefly, total DNA (3-μg) was digested with BamHI, resolved on 10% agarose gels, and transferred to Zeta probe (Bio-Rad). Membranes were hybridized with [α32P]-labeled COX II probe (mtDNA-encoded) and 18S rDNA (nuclear DNA-encoded), and the images captured using a phosphorimager (Molecular Devices) . The expression level of mtDNA-encoded transcripts was determined by Northern blot analysis. For this, total RNA was isolated by guanidinium thiocyanate-phenol-chloroform method, and treated with DNase (Ambion) to remove contaminating DNA. Samples (2.5-μg total RNA) were denatured in 50% formamide/2 M formaldehyde, resolved on 1% agarose gel containing 2 M formaldehyde in MOPS buffer and transferred. Membranes were hybridized with [α32P]-labeled cDNA probes and images captured as above .
Primary cardiomyocytes were isolated from adult rats using the Langendorff apparatus for heart perfusion . Sprague Dawley rats (6-8 weeks old) were purchased from Harlan (Indianapolis, IN). After intra-peritoneal injection of heparin (1000 IU/kg) and administration of anesthesia, heart was removed in ice cold Tyrode's buffer, and sequentially perfused with Ca2+ free Tyrode solution for 6 min at 37° C (flow rate of 12 ml/min), and 1 mg/ml collagenase (Type II, Worthington Biochem) until the heart became flaccid (20-30 min). Finally, heart was minced, cell suspension rinsed in Tyrode's buffer with a gradual increase in Ca2+ concentration (up to 1 mM), and cells were plated on petri dishes (60 mm) previously coated with laminin (10 μg/ml). Cultures consisting of ≥ 95% viable cardiomyocytes, validated by trypan blue exclusion method, were used for all experiments. Animal experiments were performed according to the National Institutes of Health Guide for Care and Use of Experimental Animals and approved by the UTMB Animal Care and Use Committee.
T. cruzi trypomastigotes (SylvioX10/4) were propagated in C2C12 cells. T. cruzi trypomastigotes lysate (TcTL) was prepared by repeated freeze thaw method (1 × 109 trypomastigotes/ml PBS). T. cruzi secreted proteins (TcSP) were prepared as described . HL-1 and primary cardiomyocytes were seeded in 6-well (5 × 104/well), or 24-well (1 × 104/well) plates and infected with T. cruzi trypomastigote (cell: parasite ratio, 1:5) for 3 h. Plates were rinsed with RPMI to remove free parasites, and then incubated in complete medium at 37°C, 5% CO2 for 0-72 h. In some experiments, cells were incubated with TcTL or TcSP (equivalent to 1 × 106 trypomastigotes/ml), and treated with recombinant human cytokines IL-1β, IFN-γ, and TNF-α (100 ng/ml each cytokine, from BD Pharmingen). Cell viability, determined by Trypan blue exclusion and MTT cell growth assay (Millipore), was decreased by 10% in T. cruzi-infected cardiomyocytes compared to controls during the course of incubation (up to 48 h).
We measured H2O2 release by cardiomyocytes using amplex red assay. Briefly, supernatants (50-μl), collected at various time-points post-infection (pi), were added in triplicate to 96 well, black flat-bottomed plates, and reaction was started with addition of 100 μM 10-acetyl-3,7-dihydroxyphenoxazine (amplex red, Invitrogen) and 0.3 U/ml horseradish peroxidase (HRP). HRP-catalyzed amplex red oxidation by H2O2, resulting in fluorescent resorufin formation, was monitored at Ex563nm/Em587nm using a Spectra MaxR M2 microplate reader (Molecular Devices) . In some experiments, cardiomyocytes were incubated with 10-μM diphenyliodinium (DPI), 100 nM salicylhydroxamic acid (SHX), and 100 nM allopurinol, the specific inhibitors of NADPH oxidase (NOX), myeloperoxidase (MPO) and xanthine oxidase (XOD), respectively. To measure the mitochondrial ROS production, mitochondria were isolated by differential centrifugation, incubated with 5 mM pyruvate/5 mM malate (pyr/mal, provide electrons to complex I) or 5 mM succinate (provides electron to complex II), and substrate-stimulated H2O2 generation was monitored by amplex red assay. Standard curve was prepared with H2O2 (50 nM - 5 μM) .
HL-1 or primary cardiomyocytes (104/well) were cultured in NuncR Lab Tek II chamber slides, infected with T. cruzi (cell: parasite ratio, 1:5), and incubated for 48 h (± recombinant cytokines, as above). Cardiomyocytes were washed thrice, and labeled for 30 min with following fluorescent probes: 200 nM MitoTracker red (mitochondria-specific, Ex498nm/Em598nm), 200 nM MitoTracker green (mitochondria specific, Ex498nm/Em598nm), 5 μM MitoSOX red (specifically detects mitochondrial ROS, Ex498nm/Em598nm), or 10 μM H2DCF-DA (detects cellular ROS, Ex498nm/Em598nm). Fluorescence was detected on an Olympus BX-15 microscope and images captured using a mounted digital camera (magnification 40× and 60×).
Mitochondrial membrane potential (Ψm) in normal and infected cardiomyocytes was determined using 5, 5′, 6, 6′-tetrachloro-1, 1′, 3, 3′-tetraethylbenzimidazolylcarbocyanine iodide (JC-1). Briefly, cells were incubated with 10 μM JC-1 for 30 min. Cells were washed, and green monomers (Ex485nm/Em/535nm) and red J-aggregates (Ex560nm/Em595nm, a sensitive marker of ΔΨm) were monitored by fluorescence microscopy. The fluorescent intensity of J-monomers (green) and J-aggregates (red) was determined using a Spectra MaxR M2 microplate reader. All fluorescent probes were purchased from Invitrogen/Molecular Probes.
Cardiomyocytes were cultured in Lab Tek II chamber slides and infected with T. cruzi (cell: parasite ratio, 1:5). At 24 h and 48 h pi, cells were permeabilized with 50 μg/ml saponin for 5 min, and washed thrice. For complex I catalytic staining, slides were immersed in 1 mM Tris buffer (pH 7.4) containing 0.5 mg/ml NADH substrate and 12.5 mg/ml nitroblue tetrazolium (NBT) electron acceptor. Staining for complex II activity was performed in 1.5 mM phosphate pH 7.4, 5 mM EDTA, 10 mM KCN reaction buffer containing 84 mM succinate substrate, and 0.2 mM phenazine methosulfate and 10 mM NBT electron acceptors. For complex IV staining, permeabilized cells were incubated in reaction buffer (50 mM phosphate pH 7.4, 24 units/ml catalase, 75 mg/ml sucrose) containing 1 mg/ml reduced cytochrome c substrate and 1.0 mg/ml 3,3′-diaminobenzidine electron acceptor. The specificity of the reaction for complex activity was determined using specific inhibitors, i.e., 6.5 μM rotenone (complex I), 10 mM sodium malonate (complex II), 1 μM antimycin (complex III), and 1 mM KCN (complex IV). Spectrophotmetric evaluation of respiratory complex activities was performed in isolated mitochondria from infected and normal cardiomyocytes .
Data are expressed as means ± SD, and were derived from at least triplicate observations per experiment. All experiments were repeated at least twice. Results were analyzed for significant differences using ANOVA procedures and Student's t-tests. The level of significance was accepted at p < 0.01.
We employed amplex red fluorescent probe (detects H2O2) to monitor T. cruzi-induced ROS production in cardiomyocytes. HL-1 cells infected with T. cruzi exhibited an increase in ROS production beginning at 2 h pi (Fig. 1A). The ROS level in infected HL-1 cells was increased by 8- and 15-fold at 24 h and 48 h pi, respectively, as compared to that detected in normal controls (Fig. 1A). Likewise, primary cardiomyocytes infected by T. cruzi exhibited a 12-fold and 18-fold increase in ROS level at 24 h and 48 h pi, respectively (Fig. 1B). HL-1 cells (Fig 1C) and primary cardiomyocytes (data not shown) exhibited a marginal increase in H2O2 release when incubated with trypomastigote lysate (TcTL) or secreted (TcSP) proteins. The specificity of amplex red was confirmed by >80% decline in fluorescence when catalase (CAT, scavenges H2O2) was added to cardiomyocytes infected by T. cruzi or incubated with TcTL or TcSP (Fig. 1C). Together, these results suggest that active invasion by T. cruzi elicits ROS production in cardiomyocytes.
In an in vivo state, cardiomyocytes are exposed to proinflammatory cytokines (e.g., IL-1β, TNF-α) and IFN-γ produced by macrophages and T cells, respectively. The cytokine treatment of HL-1 cells infected by T. cruzi resulted in a 76.4% and 86.2% increase in H2O2 level at 24 h (data not shown) and 48 h pi (Fig. 2A), respectively, compared to that detected in infected/untreated cardiomyocytes. Likewise, primary adult rat cardiomyocytes treated with cytokines exhibited a 2-fold increase in T. cruzi-induced H2O2 release at 24 h and 48 h pi (Fig. 2B). Cytokine treatment did not enhance ROS release in HL-1 only or HL-1 cells incubated with TcTL or TcSP (Fig. 2A). These results suggest that proinflammatory cytokines augment the ROS release in cardiomyocytes infected by T. cruzi.
ROS are elicited as a host defense response against infectious agents through the activation of NOX, MPO, and XOD. To examine whether these enzymes are activated and contributed to increased ROS in infected cardiomyocytes, we first determined the effect of specific inhibitors on ROS level in infected cardiomyocytes. The H2O2 levels in T. cruzi-infected HL-1 and primary cardiomyocytes, measured at 24 h and 48 h pi, was not altered by DPI (NOX inhibitor), SHX (MPO inhibitor) and allopurinol (XOD inhibitor) (data not shown). Likewise, the inhibitors of NOX, XOD, and MPO had no effect on the cytokine-stimulated enhanced ROS production in T. cruzi-infected HL-1 and primary cardiomyocytes (data not shown). Further, HL-1 and primary cardiomyocytes infected by T. cruzi and incubated with or without recombinant cytokines exhibited no detectable level of NOX, MPO, and XOD activities at 24 and 48 h pi (data not shown). These results suggested that ROS release by cardiomyocytes is not a defense response against T. cruzi, and other processes contribute to ROS release in infected cardiomyocytes.
T. cruzi attachment to and invasion of host cells stimulates Ca+2 flux . Changes in Ca+2 may promote opening of mitochondrial membrane permeability transition pores (MPTP) and affect the ψm and electrochemical gradient, leading to electron leakage to O2 and O2•− generation. We determined mitochondrial Δψm in infected cardiomyocytes using JC-1 dye. JC-1 accumulates in coupled, healthy mitochondria as red aggregates. Upon collapse of Δψm, JC-1 remains in a monomeric form and fluoresces green. Normal HL-1 cells exhibited a high red fluorescence of J- aggregates (Fig. 3A, panels a, c) and very low green fluorescence of monomers (Fig. 3A, panel b). In comparison, T. cruzi-infected HL-1 cells exhibited a marked increase in green fluorescence (Fig. 3A, panels e, f) with a corresponding decline in red fluorescence (Fig. 3A, panel d). Likewise, primary adult rat cardiomyocytes exhibited a T. cruzi-induced decline in red J-aggregates and a substantial increase in green monomers (Fig. 3B, compare panels a, b with panels d, e, respectively). Fluorometric analysis showed that JC-1 ratio (green monomers/red aggregates) was increased by ≥4-fold in infected HL-1 and primary cardiomyocytes at 24 h and 48 h pi (Fig. 3C, data not shown). Incubation of cardiomyocytes with TcTL or TcSP had no effect on JC-1 ratio (data not shown). HL-1 cardiomyocytes incubated with H2O2 (20 μM) exhibited a similar increase in JC-1 ratio as was noted in T. cruzi-infected cardiomyocytes (Fig. 3C). Treatment of T. cruzi-infected or H2O2-treated cardiomyocytes with catalase prevented the increase in green fluorescence, and preserved the J-red aggregates and JC-1 ratio to control level (Figs. 3C). Together, these results demonstrate that ROS altered the mitochondrial Δψm in cardiomyocytes infected by T. cruzi.
We determined the mitochondrial ROS production in infected cardiomyocytes using specific fluorescence probes. Representative images of HL-1 cardiomyocytes stained with MitoTracker red (localizes to mitochondria) and H2DCF-DA (detects intracellular ROS, green fluorescence) at 48 h pi are shown in figure 4A. HL-1 cells, irrespective of infection status, exhibited punctuated mitochondrial staining with MitoTracker red (Fig. 4A, panels a, d). Normal HL-1 exhibited minimal-to-no DCF fluorescence (Fig. 4A, panel b). In comparison, T. cruzi-infected HL-1 cardiomyocytes exhibited bright, punctuated, ROS-specific DCF fluorescence (Fig. 4A, panel e). Overlay images (yellow color) showed that DCF fluorescence (green) co-localized with MitoTracker red signal (Fig. 4A, panel f), thus, suggesting that mitochondrial ROS level was increased in T. cruzi-infected HL-1 cardiomyocytes. To validate in primary cardiomyocytes, we used MitoTracker green (mitochondrial probe) and MitoSOX red (detects O2•− in mitochondria of live cells). In absence of mitochondrial ROS, MitoSOX accumulates in nucleus. Primary adult cardiomyocytes exhibited substantial level of MitoTracker green signal (Fig. 4B, panels a, d). MitoSOX red fluorescence was detected in the nucleus of normal primary cardiomyocytes (Fig. 4B, panel b), also evident in overlay image (Fig. 4B, panel c). T. cruzi-infected primary cardiomyocytes exhibited a punctuated mitochondrial localization of MitoSOX red fluorescence (Fig. 4, panel e), also evidenced by overlay image (yellow) showing co-localization of MitoSOX red and MitoTracker green signal in infected cardiomyocytes (Fig. 4, panel f). These results demonstrated that mitochondrial rate of ROS production is increased in HL-1 and primary cardiomyocytes infected by T. cruzi.
The effect of the disturbance in Δψm on electron transport chain efficiency was determined by monitoring the a) activity of the respiratory complexes, and b) the rate of electron leakage to O2 resulting in O2•− formation. Histochemical staining showed a substantial decline in complex I activity in HL-1 (Fig. 5A, panels a, c) and primary cardiomyocytes (data not shown) at 48 pi. Catalytic staining of complex II was not significantly decreased (Fig. 5A, panels d, f) while complex IV staining was decreased in >30% of HL-1 (Fig. 5A, panels g, i) and primary cardiomyocytes at 48 h pi. Specificity of the reaction was confirmed by a loss in staining when normal cardiomyocytes were incubated with rotenone (inhibits complex I, panel b), malonate (inhibits complex II, panel e), and KCN (inhibits complex IV, panel h). Spectrophotometric assays validated the histochemical results. Isolated mitochondria from infected HL-1 cardiomyocytes, as compared to normal controls, exhibited 63.6% and 92% decline in complex I activity, and 51% and 46.3% decline in complex III activity at 24 h and 48 h pi, respectively (Fig. 5B, data not shown).
Inefficient transport of electrons due to altered Δψ and a loss in complex activities can result in increased electron leakage and ROS formation. To validate, that enhanced ROS formation in infected cardiomyocytes is due to impaired mitochondrial ETC efficiency, we generated HL-1 ρ cardiomyocytes by ethidium bromide treatment that selectively inhibits mtDNA replication and transcription. Selected HL-1 ρ cultures, examined by Southern blot analysis using COX II probe, showed 50-63% decline in the 16-kb mtDNA compared to HL-1 wt cells (Fig. 6A & B). Northern blot analysis exhibited the mRNA level for mtDNA-encoded COX II was decreased by 65-84% in HL-1 ρ cells compared to wt controls (Fig. 6C & D). Similar level of signal for 18S rDNA in HL-1 wt and ρ cardiomyocytes demonstrated that ρ cells were not compromised for nuclear DNA (Fig. 6A & B). The clone 3 (ρ3), consisting maximal repression of mtDNA-encoded transcripts, exhibited 95% loss of complex I activity and >52% loss in complex III activity as compared to HL-1 wt cardiomyocytes (Fig. 6E). These data showed that HL-1 ρ cells contain partial activity of respiratory complexes.
Upon infection with T. cruzi, HL-1 ρ3 cardiomyocytes exhibited no significant difference in the number of parasites/cell than that noted in wild-type HL-1 cells at 24 h and 48 h pi. We isolated mitochondria from infected cardiomyocytes at 24 h and 48 h pi by differential centrifugation. The rate of pyr/mal- and succinate-supported ROS production in isolated mitochondria from T. cruzi-infected HL-1 cardiomyocytes was increased up to 121% and 157%, respectively, when compared to that noted in normal controls (Fig. 7). HL-1 ρ cells, due to a partial ETC activity, exhibited a substantially higher level of substrate-stimulated ROS production than the HL-1 wt, and did not respond by further increase in ROS when infected with T. cruzi (Fig. 7). A substantial increase in pyr/mal- and succinate-supported ROS generation in mitochondria of HL-1 wt in response to T. cruzi infection, but not in the HL-1ρ, confirms that respiratory chain inefficiency contributes to increased ROS production in cardiomyocytes infected by T. cruzi.
In this study, we demonstrate that HL-1 and primary adult rat cardiomyocytes elicit a strong response to invading T. cruzi by substantial ROS production that was augmented by proinflammatory cytokines. The increase in ROS production in cardiomyocytes was not a bystander effect caused by parasite antigens and was not an outcome of activation of classical ROS producers (NOX, MPO, and XOD). Instead, invasion by T. cruzi altered the mitochondrial ψm, and subsequently, the activities of respiratory complexes were compromised leading to increased electron leakage and superoxide production in cardiomyocytes. HL-1 ρ cardiomyocytes exhibited no ROS response to T. cruzi infection. This, to the best of our knowledge, is the first study documenting that invasion by T. cruzi initiates ROS production and mitochondrial dysfunction in cardiomyocytes and the presence of inflammatory milieu augments the T. cruzi-induced ROS production.
In pathologic conditions, diverse pathways result in excessive ROS production that if not efficiently scavenged by the antioxidant system, leads to oxidative stress. ROS can rapidly oxidize proteins, lipids and DNA, and thereby result in dysfunction of physiological processes, oxidative damage and cell/tissue death . Several studies have shown that oxidant/antioxidant imbalance and oxidative damage are increased in the myocardium of experimental animals infected by T. cruzi   and in peripheral blood of human chagasic patients [20, 21]. The detection of inflammatory infiltrate constituted of neutrophils and macrophages in cardiac biopsies from infected mice  and humans  has led to the suggestion that oxidative burst of immune cells might be the primary source of oxidative stress during Chagas disease. We have demonstrated that mitochondrial electron transport chain is compromised in chagasic hearts, and the resultant electron leakage to molecular oxygen contributes to oxidative stress during the course of T. cruzi infection and disease development [10, 11]. Because of the presence of multiple cell types in the heart, it was not clear whether cardiomyocytes produced ROS in response to T. cruzi infection or infiltrating immune cells were the only source of ROS in the heart. Our data in this study clarify this dilemma, and demonstrate that HL-1 and primary cardiomyocytes respond to T. cruzi by elicitation of ROS within 2 h pi, and ROS level increased exponentially up to 48 h pi (Fig. 1). ROS response was not a bystander effect caused by the presence of parasite antigens, and active invasion by T. cruzi was essential to elicit ROS production in cardiomyocytes. This is the first observation of T. cruzi-induced ROS production in isolated primary cardiomyocytes.
The participation of cytokines, peptide hormones, and immune regulators in cellular pathways is modulated by redox status . Conversely, cytokine themselves are mediators of oxidative stress, and have the potential to alter redox equilibrium in cells . Some reports have shown ROS generation in cardiomyocytes directly exposed to TNF-α , and cytotoxicity of TNF-α was enhanced by IL-1β and IFN-γ. In this study, we found no detectable production of ROS in normal cardiomyocytes exposed to recombinant cytokines (IL-1β, TNF-α and IFN-γ, individually or in combination). The recombinant cytokines did not enhance the low level of ROS produced in cardiomyocytes treated with parasite antigens (membrane or secreted). Instead, when cardiomyocytes were infected with T. cruzi, exposure to IL-1β, TNF-α, and IFN-γ cytokines together, not individually, resulted in 2-fold increase in parasite-induced ROS production (Fig. 2). Whether an interaction between TNF-α, IFN-γ and IL-1β enhanced the modulating activity of each cytokine or all three cytokines played an essential role in signaling increased ROS formation remain to be investigated in future studies. Yet, our data demonstrate that infection by T. cruzi and the inflammatory milieu crosstalk, and enhance the oxidative stress in cardiomyocytes.
The oxidases and oxygenases expressed in different cell types and organelles contribute to ROS formation and oxidative stress. For example, monoamine oxidase, present in mitochondrial outer membrane, converts O2 to H2O2 on the cytoplasmic face . Xanthine dehydrogenase and XOD produce H2O2 and O2•− during purine hypoxanthine degradation to uric acid . The NOX and MPO enzymes are the primary source of ROS in macrophages, endothelial cells, and neutrophils. Activation of NOX, MPO, and XOD is documented in the heart of T. cruzi-infected mice . Increased activity of XOD is also noted in the heart during acute ischemia-reperfusion injury . Yet, these oxidases and peroxidases were not the source of ROS in cardiomyocytes infected by T. cruzi. We detected no change in NOX, MPO and XOD activities in infected cardiomyocytes as compared to controls. Treatment of infected cardiomyocytes with DPI, an inhibitor of NOX and other mixed function oxidases; SHX, an inhibitor of MPO and other ROS-generating peroxidases; and allopurinol, a specific inhibitor of XOD; did not attenuate ROS generation in infected cardiomyocytes. We surmise that processes other than the classical ROS producers contributed to redox stress in the infected cardiomyocytes.
Several observations in this study indicate that mitochondrial electron transport chain was the primary source of enhanced ROS release in cardiomyocytes infected by T. cruzi. One, our data showed co-localization of mitochondria-specific MitoTracker probes with ROS-specific H2DCF-DA and MitoSOX red fluorescence in infected cardiomyocytes (Fig. 4). Second, isolated mitochondria from infected cardiomyocytes exhibited a significant increase in ROS generation when energized with complex I (pyr/mal) or complex II (succinate) substrates (Fig. 7). Third, HL-1 ρ cardiomyocytes that harbored a dysfunctional electron transport chain (Fig. 6) did not respond to T. cruzi infection and cytokine treatment by an increase in ROS production (Fig. 7). Our data provide validation to the reports demonstrating the defects of mitochondrial respiratory complex III resulted in an increased electron leakage to molecular O2 and O2•− formation in the heart of chagasic mice [10, 11], and suggest that these defects are primarily accrued in cardiomyocytes of the infected host.
The question then arises, how mitochondrial ROS is initiated in infected cardiomyocytes. Our observation that ROS was elicited in cardiomyocytes only when infected by live T. cruzi and incubation with parasite lysates or secreted proteins had marginal-to-no effect on ROS release (Fig. 1 & 2) provides clues to this query. It is documented that invasion by T. cruzi trypomastigotes elicit transient elevations in intracellular Ca+2 (reviewed in ). In mammalian cells, Ca+2 overload is known to induce the opening of MPTPs . An increase in MPTP opening affects several mitochondrial functions. One of these is the dissipation of protonmotive force that in itself serves as a negative modulator of respiratory chain activity . It is, thus, likely that T. cruzi-induced Ca+2 overload might be the primary signal in mitochondrial dysfunction. We have noted a substantial decline in JC-1 fluorescence during T. cruzi infection (Fig. 3), indicating that significant fall in mitochondrial ψm must have occurred, and subsequently, respiratory chain activity was decreased (Fig. 5). Our notion is supported by others indicating the deleterious effects of Ca+2 flux on mitochondrial respiratory complex activities in cardiac and retinal cells [35, 36]. Amplification of cellular damage due to Ca+2 signaling of ROS generating pathways in mitochondria is reported in numerous in vitro and in vivo cellular systems (reviewed in ). We, thus, propose that T. cruzi invasion-dependent Ca+2 flux is the initial event that results in mitochondrial depolarization, ETC inefficiency and ROS production in cardiomyocytes. Further, we found that immunofluorescence staining with antibody to 4-hydroxynonenal (HNE, lipid peroxidation marker) co-localized with MitoTracker (specific for mitochondria) and H2DCF-DA (specific for H2O2) fluorescent probes in infected cardiomyocytes (unpublished observation). A substantial increase in malonyldialdehyde and protein carbonyl contents in cardiac mitochondria of mice infected by T. cruzi is also documented . Treatment of infected cardiomyocytes with catalase preserved the mitochondrial ψm, while treatment of normal cardiomyocytes with H2O2 resulted in a loss in JC1-red aggregates staining (Fig. 3). Thus, ROS-induced oxidation of mitochondrial membranes may constitute a secondary signal affecting Δψm and respiratory chain efficiency.
The extent of mitochondrial ROS production and membrane alterations promotes cell death . Mild-to-moderate level of ROS resulting in mitochondrial depolarization and Δψm results in cytochrome c release that is an important catalytic activator of apoptotic pathway . Excessive ROS production switches the mode of cell death from apoptosis to oncosis due to severe oxidative damage and downstream effects . Cardiomyocytes undergoing apoptotic and necrotic death are reported in chagasic hearts [41, 42]. In our study, infected cardiomyocytes exhibited a 10% reduction in viability and a corresponding increase in Annexin V staining that is an indicator of early apoptotic events (data not shown). In vivo studies are required to delineate further the role of mtROS in cardiomyocyte death and progression of Chagas disease.
In summary, we have demonstrated that cardiomyocytes respond to T. cruzi infection by an exponential increase in ROS production that is augmented by the proinflammatory cytokines. The ROS production by cardiomyocytes was not a defense response against T. cruzi. Instead, invasion by T. cruzi induced cellular events that affected the mitochondrial ψm, and initiated a feedback cycle of electron transport chain inefficiency and increased electron leakage and ROS production in the cardiomyocytes. The ROS-induced signaling and injurious processes may contribute to cardiomyocyte death and chronic heart failure in chagasic patients, to be investigated in future studies.
This work was supported by a grant (AI054578) from the National Institutes of Health/National Institute of Allergy and Infectious Diseases to NJG.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.