FPD, flavodiiron protein; ROS, reactive oxygen species; MG132, proteasome inhibitor Z-Leu-Leu-Leu-al; Gtr, Giardia trophozoites; PBS, phosphate buffer saline; FlRd-red, E. coli NADH:flavorubredoxin oxidoreductase; Rd, truncated rubredoxin domain of E. coli flavorubredoxin; SI, small intestine; LI, large intestine; GSH, glutathione.
Giardia intestinalis is the microaerophilic protozoon causing giardiasis, a common infectious intestinal disease. Giardia possesses an O2-scavenging activity likely essential for survival in the host. We report that Giardia trophozoites express the O2-detoxifying flavodiiron protein (FDP), detected by immunoblotting, and are able to reduce O2 to H2O rapidly (∼3 μM O2 × min × 106 cells at 37 °C) and with high affinity (C50 = 3.4 ± 0.7 μM O2). Following a short-term (minutes) exposure to H2O2 ≥ 100 μM, the O2 consumption by the parasites is irreversibly impaired, and the FDP undergoes a degradation, prevented by the proteasome-inhibitor MG132. Instead, H2O2 does not cause degradation or inactivation of the isolated FDP. On the basis of the elevated susceptibility of Giardia to oxidative stress, we hypothesize that the parasite preferentially colonizes the small intestine since, compared with colon, it is characterized by a greater capacity for redox buffering and a lower propensity to oxidative stress. © 2011 IUBMB IUBMB Life, 63(1): 21–25, 2011
Giardia intestinalis, the amitochondriate protozoon causing human giardiasis (1), is highly vulnerable to both O2 and reactive oxygen species (ROS), due to: (i) the lack of conventional ROS-scavenging enzymes, such as catalase, superoxide dismutase, and glutathione (GSH) peroxidase; (ii) the expression of the ROS-generating NAD(P)H:menadione oxidoreductase [so-called DT-diaphorase (2)]; and (iii) the O2-lability of key metabolic enzymes, such as pyruvate-ferredoxin oxidoreductase.
Despite their O2-susceptibility, Giardia trophozoites (Gtr) inhabit the fairly aerobic mucosa of the human small intestine (SI) (3). To date, there is not yet a clear-cut explanation for this apparently paradoxical preferential localization of the parasite; however, Gtr are endowed with an efficient H2O-producing O2-scavenging function (4), likely essential for parasite survival in the host that has been attributed to NADH-oxidase (5) and, more recently, also to the flavodiiron protein (FDP) (6).
Gtr are killed by 2h-incubation with H2O2 ≥ 50 μM (7) and damaged by 1h-incubation with sublethal H2O2 concentrations (8), but the mechanisms underlying H2O2 toxicity are not fully understood yet. Lloyd et al. (9) showed that prolonged exposure of Gtr to O2 (from several tens of minutes to hours) induces intracellular accumulation of H2O2 and inhibition of the O2-consumption activity. Moreover, the same authors showed that addition of H2O2 (≥ 30 μM) to Gtr has an almost immediate inhibitory effect on the O2-consumption activity. This prompted us to address in the present study the issue of H2O2 toxicity and O2-detoxification in Giardia, in the light of the recent report that FDP is an efficient O2-scavenging enzyme (6).
Cell Cultures and H2O2 Stress
Trophozoites of the G. intestinalis strain WB clone C6 (ATCC Number 50803™) were cultured at 37 °C in bile- supplemented Diamond's TYI-S-33 medium. Typically, 60-mL medium was inoculated with 25 × 106 cells in 25-cm2 flasks. After 1 day, cells were harvested by chilling the flasks on ice for 10 min, followed by centrifugation (5 min at 300 g). Then, cells were resuspended into phosphate buffer saline (PBS) + 30 mM glucose at a density of 2 × 106 cells mL−1, followed by addition of an aliquot of either H2O2 or PBS for control experiments. After 10-min incubation at room temperature, cells were centrifuged, washed with PBS + 30 mM glucose, counted on an hemocytometer and immediately used for experimentation. After H2O2-treatment, the parasites appeared less motile but excluded trypan blue similarly to controls. To assess the effect of the proteasome-inhibitor Z-Leu-Leu-Leu-al (MG132), before the H2O2 addition, 1 × 106 cells mL−1 were incubated in the medium with 100 μM MG132 for 1 h at 37 °C, washed in the presence of MG132 and resuspended into PBS + 30 mM glucose.
Oxygraphic assays were carried out using a high-resolution respirometer (Oxygraph-2k, Oroboros Instruments). Measurements on intact cells were performed at 37 °C in PBS + 30 mM glucose, whereas those on the recombinant Giardia FDP, purified as described in (6), were carried out at 20 °C in 50 mM Tris, 18% glycerol, 20 μM EDTA, pH = 7.5. According to ref.6, the O2 consumption by 300 nM FDP, untreated or preincubated with H2O2, was measured using 2 mM NADH, 0.8 μM E. coli NADH:flavorubredoxin oxidoreductase (FlRd-red) and 4 μM truncated rubredoxin domain of E. coli flavorubredoxin (Rd) as the reducing system; the assay was run in the presence of 2.5 μg mL−1 catalase. The O2 concentration at which Giardia cells consumed O2 with half maximal rate (C50) was estimated from the analysis of cellular VO2 in the 0–10 μM O2 concentration range.
Whole-cell extracts were prepared by diluting cultured Gtr into NuPAGE® LDS sample buffer (Invitrogen). Typically, the equivalent of 1–2 × 106 cells was loaded per lane. Proteins were separated by SDS-PAGE and blotted on a nitrocellulose membrane. Blots were incubated with either rabbit polyclonal antibodies against Giardia FDP (Davids Biotechnologie GmbH) or monoclonal antibodies against porcine brain α-tubulin (Exbio, clone Tu-01), followed by incubation with alkaline phosphatase-conjugated secondary antibodies (Sigma) and then with 4-nitroblue-tetrazolium chloride and 5-bromo-4-chloro-3-indolylphosphate (Sigma). Densitometric analysis was carried out with the software Scion Image for Windows (Scion Corporation, Frederick, MD), and data normalized based on tubulin bands intensity.
As shown in Fig. 1 (dashed trace), Gtr have a remarkable O2-consuming activity (4, 10, 11); at 37 °C, 1.3 × 107 cells mL−1 consume more than 200 μM O2 within a few minutes. O2-consumption proceeds at an essentially constant rate (∼3μM O2 × min × 106 cells) from ∼170 μM down to ∼10 μM O2. Below this value, the rate decreases according to an apparent C50 = 3.4 ± 0.7 μM O2 (inset to Fig. 1). Compared with control cells, trophozoites pre-exposed for 10 min to H2O2 (≤ 500 μM, see “Experimental Procedures”) and carefully washed, progressively lose their activity, leading to a nonlinear time-course of O2 consumption (Fig. 1, solid line). After O2 exhaustion, the addition of catalase to H2O2-preincubated cells does not reverse the activity loss (confront rates after reoxygenation), rather revealing the accumulation of H2O2 (∼30 μM based on O2 evolution, solid line in Fig. 1). In these experiments, H2O2-treated cells are diluted 1,500-fold, which rules out that the H2O2 detected results from a contamination of the peroxide used to stress the cells before the assay. Thus, in line with previous reports (9, 12), we conclude that the H2O2 detected is produced endogenously by the H2O2-treated cells.
To gather information on the concentration-dependence of the H2O2 effect, Giardia cells were shortly exposed to increasing amounts of H2O2 (20, 200, or 500 μM for 10 min), washed, and their O2-consuming activity monitored for a prolonged time (40–50 min) by sequential buffer reoxygenation (Fig. 2A). According to these measurements, H2O2 treatment in a dose- dependent manner causes both a decrease in the initial O2- consumption rate and an activity decay with time faster than in control cells (Fig. 2B).
By protein determination, both H2O2-treated and -untreated cells were shown to release a small (< 7%), similar amount of cellular proteins into the medium. Consistently, in the oxygraphic assays, the addition of cell-impermeable NADH to H2O2-treated cells did not enhance O2 consumption, whereas the same addition to cell lysates caused a remarkable stimulation of the consumption (not shown). Hence, we conclude that the lower O2-consumption activity measured with H2O2-stressed cells, at this stage of damage, is not due to cell lysis and loss of substrates.
The protein level of FDP in control and H2O2-treated cells was assayed by immunoblotting. At [H2O2] ≥ 100 μM, the amount of the enzyme was significantly lower than in control cells (Fig. 3A). The proteasome-inhibitor MG132 prevented the H2O2-mediated protein degradation, though causing itself a slight decrease in the FDP level (Fig. 3B); the protease inhibitor cocktail (Sigma P8340) was instead ineffective (not shown). Compared with control cells, MG132-incubated trophozoites displayed only slightly reduced O2-consumption activity, but a severe activity impairment was observed in those cells following exposure to 500 μM H2O2 (data not shown).
To verify whether H2O2 directly inactivates or degrades the FDP, the purified recombinant FDP from Giardia was treated with H2O2 and thereafter assayed. Even after 20-min incubation with H2O2 at the highest concentration (500 μM), the protein showed unaffected O2-consumption activity and integrity (not shown).
Given its relatively high O2-susceptibility, survival of Giardia in the fairly aerobic environment of the proximal SI (3) likely relies on the ability of the trophozoites to safely reduce O2 to H2O, a function originally and uniquely assigned to NADH oxidase (5). In agreement with ref.13, herein, we show by immunoblotting that the recently characterized O2-detoxifying FDP is expressed at detectable levels in trophozoites grown under standard conditions. Consistently, by high-resolution respirometry, we assessed that these cells consume O2 at a relatively high rate (∼3 μM O2 × min × 106 cells) with an apparent C50 = 3.4 ± 0.7 μM O2. This value, within the experimental error including the O2 leak into the apparatus, is compatible with the C50 ≤ 2 μM measured with the isolated FDP (6). Both the Vmax and C50 values determined for Gtr are in fairly good agreement with those measured under steady state conditions by using open O2-electrode systems (10, 11). While less evident in the assays previously carried out on the isolated protein (6), in the present study, a clear deviation from the Michaelis–Menten kinetics was invariantly observed with Giardia cells (see Fig. 1); it remains to be assessed where such a kinetic behavior originates from.
We report that shortly exposing Gtr to H2O2 (≥ 100 μM, for 10 min) induces both a substantial degradation of the FDP and, in agreement with previous studies (9), the irreversible inactivation of cellular O2-consumption. The experiments with purified FDP show that the degradation is not due to the direct reaction of H2O2 with the protein; it rather involves a cellular proteolytic pathway sensitive to the proteasome-inhibitor MG132. At present, we do not know why incubation with MG132 induces a decrease in the amount of FDP detected in control cells; a possible explanation is that MG132 itself is known to promote oxidative stress via stimulation of intracellular ROS production (14), thereby, possibly favoring a partial degradation of the protein.
Taken together the observations suggest that, following short-term exposure of the cells to H2O2, among a number of detrimental effects particularly observed at the highest H2O2 concentrations, cellular O2-consumption is severely impaired, an MG132-sensitive proteolytic pathway is activated, and the FDP is rapidly degraded, so that the reaction of O2 with ROS-producing enzymes, such as the DT-diaphorase (2), is favored. This leads to enhanced endogenous production and hence toxicity of H2O2 eventually causing cell death.
H2O2 is expected to cause several detrimental effects, particularly in Giardia that lacks catalase. Thus, it is conceivable that impairment of O2-consumption does not result solely from the, herein, documented degradation of FDP. Other potential H2O2 targets, including the other O2-consuming enzyme NADH- oxidase, and the yet unknown proteins implicated in the electron transfer to FDP, can be envisaged. Thus, it is not surprising that at lower H2O2 concentrations (20 μM), not causing the FDP degradation, cellular O2-consumption activity is slightly affected (compare Figs. 2B and 3), whereas at higher H2O2 concentrations it is severely impaired, even when FDP degradation is prevented by MG132. Within these experimental limits, the effect of MG132 strongly suggests an involvement of the proteasome, without excluding that proteins other than FDP undergo a similar H2O2-induced degradation.
G. intestinalis was suggested to preferentially colonize the SI, below the Vater ampulla, because this tract of the intestine is rich in bile and nutrients necessary for survival of the parasite [see (1, 15) and references therein]. Particularly, biliary lipids are believed to sustain the in vivo proliferation of Gtr, as these cells are unable to carry out the de novo synthesis of most lipids. Consistently, in vitro the bile stimulates the growth of Gtr, and bile acids increase the uptake of phospholipids and cholesterol by the parasite. Moreover, conjugated bile acids reportedly protect Gtr from lysis by free fatty acids generated in the small intestinal lumen during digestion.
Over and above these reasons, and based on the high susceptibility of Giardia to H2O2, here, it is tempting to speculate that the peculiar localization of G. intestinalis in the SI is related also to the higher redox buffering capacity of this intestinal tract compared with the large intestine (LI) (16–18); interestingly, a similar argument was used to account for the widely observed phenomenon that SI is much less susceptible to cancer than LI (17, 18). The minor propensity to oxidative stress of SI versus LI might derive from the several antioxidants contained in the bile, including GSH, cysteine, and bilirubin. Consistently, biliary GSH was shown to be a major contributor to the luminal intestinal GSH pool (19), and, in animal models, the concentration of reduced GSH was reported to be higher in the SI than in the colon (20, 21), although some conflicting data were also reported (22). It is also intriguing that a clinical study on patients with obstructive jaundice showed that, compared with controls, these subjects have a duodenal mucosa presenting significantly higher levels of oxidative stress (i.e., lower GSH/GSSG ratio and higher lipid peroxidation) (23). All this information is consistent with the view that the bile contributes to make SI less prone to oxidative stress and, thus, more favorable to G. intestinalis proliferation.
Summing up, in the present study, we have shown that exposure of G. intestinalis to H2O2 causes impairment of the O2 consumption and prompt degradation of the FDP, presumably via activation of the proteasome. Also, we raised the hypothesis that this parasite, highly susceptible to oxidative stress, preferentially colonizes the proximal SI, because there, compared with the colon, the redox buffering capacity is higher, an intriguing pathogenetic hypothesis that requires to be experimentally validated.
The authors thank M. Teixeira (Instituto de Tecnologia Quimica e Biologica, Universidade Nova de Lisboa, Lisbon, Portugal) for kindly providing the E. coli FlRd-red and Rd proteins. The work was partially supported by Ministero dell'Istruzione, dell'Università e della Ricerca of Italy (PRIN 2008FJJHKM_ 002 to P.S. and FIRB RBFR08F41U_001 to A.G.), by the European Society of Clinical Microbiology and Infectious Diseases (Research grant 2009 to A.G.), and by Regione Autonoma della Sardegna (L.R. 7 agosto 2007, n.7 to P.L.F.).