Alternative Oxidase Mediates Pathogen Resistance in Paracoccidioides brasiliensis Infection

Background Paracoccidioides brasiliensis is a human thermal dimorphic pathogenic fungus. Survival of P. brasiliensis inside the host depends on the adaptation of this fungal pathogen to different conditions, namely oxidative stress imposed by immune cells. Aims and Methodology In this study, we evaluated the role of alternative oxidase (AOX), an enzyme involved in the intracellular redox balancing, during host-P. brasiliensis interaction. We generated a mitotically stable P. brasiliensis AOX (PbAOX) antisense RNA (aRNA) strain with a 70% reduction in gene expression. We evaluated the relevance of PbAOX during interaction of conidia and yeast cells with IFN-γ activated alveolar macrophages and in a mouse model of infection. Additionally, we determined the fungal cell's viability and PbAOX in the presence of H2O2. Results Interaction with IFN-γ activated alveolar macrophages induced higher levels of PbAOX gene expression in PbWt conidia than PbWt yeast cells. PbAOX-aRNA conidia and yeast cells had decreased viability after interaction with macrophages. Moreover, in a mouse model of infection, we showed that absence of wild-type levels of PbAOX in P. brasiliensis results in a reduced fungal burden in lungs at weeks 8 and 24 post-challenge and an increased survival rate. In the presence of H2O2, we observed that PbWt yeast cells increased PbAOX expression and presented a higher viability in comparison with PbAOX-aRNA yeast cells. Conclusions These data further support the hypothesis that PbAOX is important in the fungal defense against oxidative stress imposed by immune cells and is relevant in the virulence of P. brasiliensis.


Introduction
An essential event during cell growth and host invasion of pathogenic fungi is to avoid the toxic effects of reactive oxygen species (ROS), such as superoxide (O 2 ), hydrogen peroxide (H 2 O 2 ) and hydroxyl (OH 2 ) radicals. ROS can either be produced by the fungal mitochondrial electron transport chain or derive from host immune cells (e.g., macrophages) during host-pathogen interaction [1,2,3]. Independently of its origin, these oxidizing agents may ultimately alter the bioenergetic status of the cell and affect essential metabolic pathways and/or its growth within host tissues, representing a toxic stimulus that decreases fungal survival [4,5].
Human pathogenic fungi such as Paracoccidioides brasiliensis, Histoplasma capsulatum, Candida albicans, among others, possess a defense mechanism against ROS that attempts to modulate the oxidative attack. These mechanisms include some enzymes such as superoxide dismutase which catalyze dismutation of O 2 2 producing H 2 O 2 and O 2 and catalases involved in decomposition of H 2 O 2 to H 2 O [6]. In addition, it has been demonstrated that Aspergillus fumigatus produces an alternative oxidase (AOX) that contributes to both reduction of ROS generated by mitochondria and regulation of energy production and metabolism [7]. Recently, Magnani et al (2007) suggested that AOX is required for the A. fumigatus pathogenicity, mainly for the survival of A. fumigatus conidia during host infection and reduction of ROS generated by activated macrophages [1].
P. brasiliensis, the causal agent of Paracoccidioidomycosis (PCM) [8], is a thermal dimorphic fungus that at environmental temperature grows as a mold producing conidia (the infectious particle). Once inhaled by the host, these fungal cells reach the lung alveoli were they interact with epithelial cells and alveolar macrophages [9]. At 37uC the transition to the parasitic yeast form occurs; however, disease development depends on both the virulence of the fungal strain and host-related factors [10]. Thus, fungal survival inside the host cells depends on the capacity to respond to external factors, ranging from temperature changes to oxidative stress. In fact, as a facultative intracellular pathogen, P. brasiliensis can persist within the macrophage phagolysosomes due to the existence of defense mechanisms that allows the fungus to survive under nutritionally poor environments and in the presence of ROS [11,12,13]. Also, P. brasiliensis has been shown to express a powerful antioxidant defense system in the presence of ROSmediated oxidative stress [14].
Recently, Maricato and co-workers (2010) showed that under oxidative stress induced by H 2 O 2 and in the presence of activated macrophages, P. brasiliensis yeast increases the expression of the flavoprotein monoxigenase family, an important group of antioxidative enzymes [2]. In addition, the analysis of the mitochondrial function of P. brasiliensis yeasts revealed the existence of an alternative respiratory chain (AOX), previously demonstrated to play an important role in the control of ROS and other oxidative molecules [7,15,16]. Campos et al (2005) described in P. brasiliensis several genes in an expressed sequence tag (EST) database encoding proteins involved in antioxidant defense, such as catalase, superoxide dismutase, peroxiredoxin, and cytochrome c peroxidase, among others [17]. More recent studies in P. brasiliensis have suggested the role of AOX gene in the mycelia to yeast differentiation process and in the maintenance of intracellular redox balancing [18].
In the present study, we have evaluated the role of an AOX in P. brasiliensis during host-pathogen interaction. To achieve this we used a P. brasiliensis AOX-antisense (PbAOX-aRNA) strain to elucidate the role of this enzyme during interaction with alveolar macrophages, considered one of the first lines of defense against this human pathogen [19,20], as well as in a mouse model of infection.

Methods
Microorganisms and culture media P. brasiliensis yeast cells (P. brasiliensis ATCC 60855) were maintained at 36uC by sub-culturing in Brain Heart Infusion (BHI) media supplemented with 1% glucose (Becton Dickinson and Company, Sparks, MD, USA). Unless indicated otherwise, yeast and mycelia cells were grown in BHI liquid medium at 36uC and 20uC respectively, with aeration on a mechanical shaker and were routinely collected during their exponential phase of growth (72-96 h) [21]. Morphological transition from yeast-to-mycelia was performed in BHI liquid medium at 20u [21]. P. brasiliensis conidia were produced as described by Restrepo et al (1986). For the assays, conidia were purified using the glass-wool filtration protocol [22] with cell quantification and viability being determined with Neubauer chamber and ethidium bromidefluorescence staining procedures, respectively [23]. To evaluate cell morphology, yeast cells were exponentially grown, collected, and fixed in a slide and visualized with an Axiostar Plus (Zeiss) microscope.
Agrobacterium tumefaciens strain LBA1100 was used as recipient for the binary vectors constructed in this study [24]. Bacterial cells were maintained at 28uC in Luria Bertani (LB) medium containing kanamycin (100 mg/ml). Escherichia coli XL-1-Blue strain was grown at 37uC in LB medium supplemented with appropriate antibiotics and was used as host for plasmid amplification and cloning [25].
Plasmid construction for aRNA gene repression and Agrobacterium tumefaciens-mediated transformation (ATMT) of P. brasiliensis were performed as previously described [26,27]. The amplified PbAOX-aRNA oligonucleotides were inserted into the pCR35 plasmid under the control of the calcium-binding protein (CBP-1) promoter region from H. capsulatum [28]. The pUR5750 plasmid was used as a parental binary vector to harbor the aRNA cassette within the transfer DNA (T-DNA). The constructed binary vectors were introduced into A. tumefaciens LBA1100 ultracompetent cells by electroporation [29] and isolated by kanamycin selection (100 mg/ml).
ATMT of P. brasiliensis yeast cells was performed using the A. tumefaciens cells harboring the desired binary vector as described by Almeida et al (2007). Selection of P. brasiliensis transformants was performed in BHI solid media with hygromycin B (Hyg; 50 mg/ ml) during a 15-day incubation period at 36uC. Randomly selected Hyg resistant transformants were tested for mitotic stability and decrease in PbAOX gene expression was analyzed after consecutive subculturing for 15, 30, 60, and 120 days. P. brasiliensis yeast cells were also transformed with the empty parental vector pUR5750 as a control during assays carried out in this study.

Author Summary
Thermally dimorphic pathogenic fungi are responsible for potentially life-threatening diseases of immunocompetent and immunocompromised individuals. These microorganisms grow as conidia-producing mycelia in the environment, which when inhaled by the host convert to the pathogenic yeast form at 37uC. During adaptation and growth, fungi interact with host immune cells and must cope with defense mechanisms such as imposed-oxidative stress (e.g., reactive oxygen species; ROS). Alternative oxidase (AOX) is an enzyme recently implicated in the reduction of ROS production by the mitochondria when triggered by external stimuli, such as temperature and ROS. During this work we have evaluated the relevance of AOX during infection with Paracoccidioides brasiliensis, the etiological agent of one of the most prevalent mycoses in Latin America, paracoccidioidomycosis. We show that PbAOX gene expression is stimulated after interaction with alveolar macrophages or in the presence of H 2 O 2 and is essential for survival against fungicidal activity of both the immune cells and the ROS compound. Moreover, decreasing PbAOX gene expression in P. brasiliensis led to increased survival of infected mice. Altogether, our data supports a relevant role for AOX in the virulence of P. brasiliensis.
Role of PbAOX during P. brasiliensis Infection www.plosntds.org activated) were washed once with RPMI 1640 culture medium and co-cultured with P. brasiliensis conidia or yeasts at a concentration of 4610 5 conidia per well and 8610 4 yeast cells per well (corresponding to a ratio of 1:1 for P. brasiliensis conidia:IFN-cg-activated alveolar macrophages and 1:5 for P. brasiliensis yeasts: IFN-c-activated alveolar macrophages), and incubated for 15 and 30 min, 1, 3, 6, 12, 24 and 48 h at 36uC, 5% CO 2 . After interaction, IFN-c-activated alveolar macrophages were lysed by adding H 2 O, then PbWt and PbAOX-aRNA cells were removed, and dilutions of these suspensions were plated on BHI plates supplemented with 0.5% glucose, 4% horse serum and EDTA 300 mM and incubated at 18uC as previously described by Kurita et al. [30]. These results were compared with the number of conidia or yeast cells added to each well. Percentage of viable cells was expressed as the number of CFUs obtained from each experimental well (P. brasiliensis conidia or yeast cells with IFN-cgactivated alveolar macrophages) divided by the number of CFUs in the controls (P. brasiliensis conidia or yeast cells alone).

Gene expression analysis
Total RNA from PbWt and PbAOX-aRNA fungal cells was obtained using TRIzol H (Invitrogen, Carlsbad, CA, USA). Total RNA was treated with DNase I (Invitrogen, Carlsbad, CA, USA) and tested using a conventional PCR with b-tubulin primers to confirm the absence of chromosomal DNA contamination [31]; cDNA was synthesized using 2 mg of total RNA with SuperScript III reverse transcriptase according to the manufacturer's instructions (Invitrogen, Carlsbad, CA, USA).
Real-time PCR was done using SuperScript TM III PlatinumH Two-Step qRT-PCR Kit with SYBRH Green, according to the manufacturer's instructions (Invitrogen, Carlsbad, CA, USA). The CFX96 Real-Time PCR Detection System (Bio-Rad, Headquarters Hercules, California, USA) was employed to measure gene expression levels. PbAOX gene expression was evaluated in both PbWt and PbAOX-aRNA conidia and yeast cells alone and after interaction with MH-S cell line at the different time points described above. Melting curve analysis was performed after the amplification phase of real time PCR assays to eliminate the possibility of non-specific amplification or primer-dimer formation. Fold changes in mRNA expression were calculated using the difference between the target gene and b-tubulin (house-keeping gene) using delta-delta-Ct [32]. Each experiment was done in triplicate and the expression level measured by triplicate. The following primers were used: PbAOX, L:agggctgggaaatattctttg and R:cttgggagcaagaggtgct); and b-tubulin, L:gtggaccaggtgatcgatgt and R:accctggaggcagtcaca).
To determine cytokine expression we used the same samples and RT-PCR procedures as described above, but in this case we used ubiquitin as the housekeeping gene. We evaluated the expression of interleukin (IL)-6, IL-10, IL-12p40 and tumor necrosis factor alpha (TNFa) at different time points (15 and 30 min and 1, 3, 6, 12, 24 and 48 h) of infection; as a control we used non-infected IFN-cg-activated alveolar macrophages. Each experiment was done thrice and the expression level was measured in triplicate. The following primers were used: IL-6, L:acacatgttctctgggaaatcgt and R:aagtgcatcatcgttgttcataca; IL-10, L:tttcaattccctgggggagaa and R:gctccactgccttgctcttatt; IL-12p40, L:caaattactccggacggttca and R:agagacgccattccacatgtc; TNFalpha, L:gccaccacgctcttctgtct and R:tgagggtctgggccatagaac; ubiquitin L:tggctattaattattcggtctgcat and R:gcaagtggctagagtgcagagtaa. Animals were infected intranasally (i.n.) with 4610 6 P. brasiliensis conidia cells suspended in PBS buffer from either PbWt or PbAOX-aRNA strains; conidia cell viability was always above 95%. Prior to infection, fungal cells were washed thrice with PBS and submitted to Neubauer counting procedures. Separate groups of ten mice were used to compare the percentage of survival over a 265-day period post-challenge.

Mouse model of infection
Groups of eight mice were challenged with PbWt and PbAOX-aRNA conidia and evaluated for fungal burden at 3 days and 8 and 24 weeks post-challenge. Groups of mice per time point were examined for colony-forming units (CFU) in lung homogenates. The homogenates were plated on Mycosel supplemented with 0.5% glucose, sealed, and incubated at 18uC; colonies were counted daily until no increase in counts was observed. The CFU counts/g of lung tissue is expressed on a log scale.

Statistic analysis
Data are expressed as average 6 standard error of the mean (SEM), and all assays were done at least three times. All statistical analysis, including analysis of variance (ANOVA), was performed using the SPSS 17.0 statistics program. A p value,0.05 was considered statistically significant. Survival rate from two independent infections in a mouse model (n = 10 mice for each P. brasiliensis strain) was analyzed using Kaplan Meyer and log rank test.

Results
Generation of a P. brasiliensis PbAOX-aRNA strain A P. brasiliensis PbAOX-aRNA strain was generated via ATMT by expressing an aRNA oligonucleotide targeting the coding sequence of the PbAOX gene in PbWt yeast cells (Fig. 1A). For control experiments, PbWt yeast cells were independently transformed with the empty vector (PbWt+EV). After confirming mitotic stability, 3 random PbAOX-aRNA and empty-vector transformants were selected for gene expression analysis of PbAOX. A 70% decrease was detected in all tested PbAOX-aRNA strains after 15, 30, 60, and 120 days of sub-culturing without selection of yeast cells, confirming knock-down of gene expression and stable genomic integration of the T-DNA (Fig. 1B). For the functional assays we used the PbAOX-aRNA transformant growing after 60 days of sub-culturing (resulting from insertion of the antisense construct AS2).

PbAOX is up-regulated during infection of alveolar macrophages
PbAOX expression was evaluated during infection of alveolar macrophages. To serve as a control for subsequent assays, we Role of PbAOX during P. brasiliensis Infection www.plosntds.org evaluated PbAOX gene expression of yeast and conidia of either PbWt or PbAOX-aRNA strains placed alone in macrophage culture medium (Fig. 2). No major alterations were detected in yeast cells of either strain; however, PbAOX expression in PbWt conidia cells increased slightly most likely due to the metabolic activation of conidia after being placed in the medium.
No increase in PbAOX gene expression was detected after interaction with non-activated IFN-g-activated alveolar macrophages. However, a continuous increase in PbAOX gene expression was observed over time in PbWt (yeast and conidia) during interaction with IFN-g-activated alveolar macrophages ( Fig. 2A  and 2B). PbAOX expression in PbAOX-aRNA yeast and conidia cells was kept at very low levels throughout all the assays (Fig. 2C  and 2D).

PbAOX expression during infection of alveolar macrophages is essential to maintain fungal viability
To elucidate the role of PbAOX during oxidative stress response induced by host alveolar macrophages, we determined cell viability upon infection of alveolar macrophages with yeast and  Role of PbAOX during P. brasiliensis Infection www.plosntds.org conidia cells of the PbWt and PbAOX-aRNA strains (Fig. 3). Fungal viability of either PbWt or PbAOX-aRNA in the absence of alveolar macrophages and after interaction with non-activated alveolar macrophages was unaltered during all evaluated periods. After 3 h of interaction with IFN-g-activated alveolar macrophages, we observed a decrease in cell viability of yeast and conidia of both strains. Nonetheless, PbAOX-aRNA yeast and conidia cells presented a significant decrease in viability when compared to the PbWt strain ( Fig. 3D and 3C).

PbAOX does not affect cytokine expression in IFN-gactivated alveolar macrophages
In order to evaluate whether PbAOX suppression could alter the cytokine profile in alveolar macrophages, we determined the gene expression of IL6, IL10, IL12p40, and TNF-a during infection of non-activated and IFN-g-activated alveolar macrophages. We did not observe expression for any of the tested cytokines in noninfected alveolar macrophages, either non-activated or activated with IFN-g (data no shown). However, PbWt and PbAOX-aRNA yeast and conidia cells were able to induce TNF-a gene expression in IFN-g-activated alveolar macrophages after 30 min of initial interaction (Fig. 4). Interestingly, higher levels of TNF-a were observed during either PbWt or PbAOX-aRNA conidia-alveolar macrophages interaction when compared to yeast cells after 30 min of infection of IFN-g-activated alveolar macrophages. Expression of IL6, IL10, and IL12p40 were below detection levels for these experiments.

Decrease in PbAOX expression decreases fungal burden in lungs of infected mice and increases survival rate
The relevance of PbAOX during host-pathogen interaction was further evaluated using a mouse model of infection challenged with conidia. We did not observe a significant difference in the colonyforming units (CFU) counts recovered from lungs of mice three days after infection with PbWt or PbAOX-aRNA conidia (Fig. 5A). However, 8 and 24 weeks post-challenge, lower CFU counts in lungs of mice infected with PbAOX-aRNA conidia were significant lower and similar to those recovered at 3 days post-challenge. Contrarily, fungal burden was significantly higher in the lungs of mice infected with PbWt conidia at the different evaluated periods (Fig. 5A).
Regarding survival, animals infected with the PbWt strain started to die at day 154, with an average survival of 167 days,

PbAOX is involved in the detoxification process exerted by exogenous H 2 O 2
Given the role AOXs can play to dampen ROS production by mitochondria [35], we evaluated the viability of P. brasiliensis yeast cells from PbWt and PbAOX-aRNA strains when exposed to different concentrations of H 2 O 2 . Although systematic differences were consistently registered at all concentrations of H 2 O 2 , they were significant only for concentrations higher than 0.0625 M (Fig. 6A). Importantly, PbWt yeasts cells highly expressed PbAOX when exposed to exogenous oxidative stress, peaking at 0.0625 M of H 2 O 2 , but decreasing at higher concentrations (Fig. 6B). Only a minimal but not significant increase in the PbAOX expression levels was detectable in PbAOX-aRNA yeast cells upon exposure to H 2 O 2 (at 0.0625 M).

Discussion
It is well documented that ROS (e.g., O 2 2 , H 2 O 2 OH 2 ) are produced as a part of host defense mechanisms against fungal pathogens, including P. brasiliensis [18,36], and that these molecules induce damage of mitochondrial DNA [37]. Pathogenic fungi possess mechanisms to attenuate the cell damage induced by ROS observed during host adaptation and growth through the production of detoxifying molecules [18,38]. Among these is alternative oxidase (AOX), an enzyme that is part of a non-protonmotive non-energy-conserving pathway of the mitochondrial respiratory process and plays a role in lowering ROS formation in fungi and plant mitochondria [39,40]. AOX couples oxidation of ubiquinone to the reduction of O 2 to H 2 O and its activation is thought to be triggered by diverse external stimuli, such as temperature and ROS [4,14,15,40]. In P. brasiliensis, AOX has been associated not only to the reduction of mitochondrial-generated ROS and maintenance of intracellular redox balance in yeast cells but also a putative role during the mycelia to yeast morphological switch [18]. In the present study, we have explored the involvement of PbAOX during interaction of P. brasiliensis yeast and conidia with host alveolar macrophages and in the presence of exogenous H 2 O 2 , as well as in a mouse model of infection. To support our data, we generated a P. brasiliensis strain with consistently reduced PbAOX gene expression using aRNA technology [26,27,41] to compare with the PbWt strain (Fig. 1). Role of PbAOX during P. brasiliensis Infection www.plosntds.org INF-g-activated alveolar macrophages are able to produce ROS that are important effectors against infecting fungi [11,20,42]. In vitro assays revealed that the decrease in PbAOX expression led to increased susceptibility against the fungicidal activity of INF-gactivated alveolar macrophages (Fig. 3). This susceptibility was even greater in conidia than in yeast cells revealing a differential response depending on the fungal morphotype. In addition, a reduction of PbAOX mRNA levels significantly decreased survival of P. brasiliensis yeast cells exposed to exogenous H 2 O 2 , further supporting a role of this enzyme against oxidative agents (Fig. 6). These results further implicate the PbAOX gene in the resistance of P. brasiliensis to the oxidative burst generated by alveolar macrophages and are in agreement with previous reports indicating that AOX is required for A. fumigatus pathogenicity [1]. As P. brasiliensis is an intracellular facultative pathogen, it will be important to understand if PbAoxp is also important as a defense mechanism against host-imposed oxidative stress or for persistence within the phagolysosome.
Interestingly, PbAOX gene expression was higher in PbWt conidia than in yeast cells and was only induced in the presence of INF-g-activated alveolar macrophages (Fig. 3). The fact that conidia had to cope not only with INF-g-activated alveolar macrophages but also with the temperature-induced morphological switch might explain the diminished survival. Taking in to account that conidia basal metabolic activity is much lower than yeast cells we hypothesize that exposure to 37uC and metabolic activation of the conidia-to-yeast (C-Y) morphological transition generates a higher quantity of ROS as by-products. However, conidia are not yet equipped with the enzymatic machinery to detoxify these molecules as yeast cells possibly explaining the difference in viability (either in the presence or absence of alveolar macrophages). In fact, recent work in our laboratory has shown that PbAOX plays an important role in the maintenance of cellular homeostasis during the C-Y transition (manuscript under preparation). Martins and colleagues (2010) have also demonstrated that PbAOX plays an important role during the morphological transition [18].
TNF-alpha plays an important role in modulation of immune response against P. brasiliensis [20,43]. Previous reports have shown that this pro-inflammatory cytokine activates macrophages, which exert a fungicidal mechanism independent of nitric oxide production [43,44]. In the present study, we observed that INFg-activated alveolar macrophages infected with either fungal morphotype induced expression of TNF-alpha (Fig. 4). Although this cytokine could be implicated in the fungicidal activity of INF-g-activated alveolar macrophages against P. brasiliensis through ROS production, we show that reduction of PbAOX expression did not affect the production of this cytokine. Interestingly, P. brasiliensis conidia induced a higher production of TNF-alpha. This difference could be attributed to the differential cell wall composition of conidia (mainly 1,3-beta glucans) and yeast (mainly 1,3-alpha glucans) cells. It has been demonstrated that 1,3-beta glucans are recognized by the dectin-1 receptor present in macrophages triggering cell activation, including cytokine production such as TNF-alpha [45,46]. Additionally, our data show that the reduction in expression of PbAOX leads to a significantly increased survival rate in a mouse model of infection (Fig. 5). Also, fungal burden in mice challenged with PbAOX-aRNA conidia were lower at all the evaluated periods, suggesting that PbAOX is important for the establishment of the infection, namely during the initial phase when fungal cells must face the temperature-induced morphological shift to the yeast form while simultaneously interacting with immune cells (e.g. alveolar macrophages) that are able to induce an oxidative stress response (e.g. ROS). The relevance of AOX during infection was previously reported in other human pathogenic fungi such as Cryptococcus neoformans and Aspergillus fumigatus [1,47]. This further suggest that this enzyme may play an important role in the adaptation of P. brasiliensis and development of the disease, possibly by improving survival within phagocytic cells [1,6,47].
Altogether, evidence presented throughout this work support the relevance of PbAOX in the virulence of P. brasiliensis. Taking into consideration that AOX has also been shown to cause decreased susceptibility of Candida albicans to the antifungal drug fluconazole [3], PbAOX may represent a putative target for treatment of paracoccidioidomycosis. Work in our laboratory is being finalized to further contribute to the elucidation of biochemical events associated to AOX's role in P. brasiliensis life cycle.