Honokiol induces superoxide production by targeting mitochondrial respiratory chain complex I in Candida albicans

Background Honokiol, a compound extracted from Magnolia officinalis, has antifungal activities by inducing mitochondrial dysfunction and triggering apoptosis in Candida albicans. However, the mechanism of honokiol-induced oxidative stress is poorly understood. The present investigation was designed to determine the specific mitochondrial reactive oxygen species (ROS)-generation component. Methods/results We found that honokiol induced mitochondrial ROS accumulation, mainly superoxide anions (O2•−) measured by fluorescent staining method. The mitochondrial respiratory chain complex I (C I) inhibitor rotenone completely blocked O2•− production and provided the protection from the killing action of honokiol. Moreover, respiratory activity and the C I enzyme activity was significantly reduced after honokiol treatment. The differential gene-expression profile also showed that genes involved in oxidoreductase activity, electron transport, and oxidative phosphorylation were upregulated. Conclusions The present work shows that honokiol may bind to mitochondrial respiratory chain C I, leading to mitochondrial dysfunction, accompanied by increased cellular superoxide anion and oxidative stress. General significance This work not only provides insights on the mechanism by which honokiol interferes with fungal cell, demonstrating previously unknown effects on mitochondrial physiology, but also raises a note of caution on the use of M. officinalis as a Chinese medicine due to the toxic for mitochondria and suggests the possibility of using honokiol as chemosensitizer.


Introduction
Currently, disseminated invasive candidiasis has an approximate estimated mortality rate of 40%, even after being treated with antifungal drugs [1,2]. Candida albicans is the major cause of candidiasis and is the fourth most commonly reported nosocomial infection [3][4][5]. As everyone knows, the majority of living organisms need oxygen to survive. Mitochondrial respiration, dealing with transfer of unpaired electrons to oxygen (O 2 ), may produce reactive oxygen species (ROS) such as superoxide anion ( [6]. Within the mitochondria, the main sites of superoxide production have been localized to the electron transport chain (ETC) complexes I (C I) and Ⅲ [6][7][8][9]. The bulk of mitochondrial ROS typically arise because of electron leakage from forward electron transport onto O 2 during aerobic respiration as side products. It is driven by NADH-linked substrates, C I exhibits only minimal ROS production, but the addition of a ubiquinone-site inhibitor, such as rotenone, results in a significant increase in its rate [7][8][9]. The other mechanism by which ETC produces large amounts of O 2 •− is during reverse electron transport. During reverse electron transport, driven by succinate, ROS production by C I is increased significantly, and in this case, inhibited by the addition of rotenone [7,8]. In addition to C I, C Ⅲ is regarded as an important site of O 2 •− production, especially when mitochondrial respiration is suppressed by antimycin, an inhibitor of C Ⅲ [6]. O 2 •− is then dismutated by superoxide dismutases to H 2 O 2 that is reduced to H 2 O by catalase, peroxiredoxins, and glutathione peroxidases [9].When intracellular levels of ROS are high, ROS can have deleterious effects on cellular biomolecules including protein, lipid, RNA and DNA and cause subsequent cell death [9]. Honokiol, a neolignan isolated from the oriental medicine plant Magnolia officinalis, is an interesting compound exhibiting various pharmacological activities in preclinical experimental models [10][11][12][13][14]. Honokiol molecule contains two phenolic groups which can exhibit antioxidant properties similar to vitamin E or polyphenols. A paradox of honokiol is that it may have both pro-and antioxidant activities [15,16]. Previous mechanism studies, including those from our own laboratory, indicated that the apoptosis induction by honokiol in C. albicans was associated with production of ROS [15]. However, we are still lacking a detailed mechanistic knowledge of the architecture of mitochondrial ROS-producing systems induced by honokiol such as of C I or C Ⅲ and detailed insights on the mechanisms controlling their activities. The present study will make an attempt to clarify specific proposed mitochondrial ROS-producing components after honokiol treatemt.

Microorganisms
C. albicans wild type strain SC5314 was used in this study [17]. The strain SC5314 was cultured in YPD (yeast extract/ peptone/dextrose) broth. For agar plates, 2% (w/v) bacto agar (Difco, BD Biosciences) was added to the medium. The strain was stored as frozen stock with 15% (v/ v) glycerol at -80•C. Before each experiment, cells were freshly revived on YPD plate from the stock.

ROS determination
SC5314 cells were adjusted to 1×10 7 cells/mL in YPD medium and exposed to different concentration of honokiol at 37˚C for 4h. Intracellular ROS concentrations were determined in liquid cultures of C. albicans after honokiol treatment. H 2 O 2 and O 2 •− levels were detected by adding DCFH-DA and DHE to the culture, respectively. After staining with 10 μmol/L DCFH-DA or 5 μmol/L DHE at 37˚C for 30 min, the cells were collected and washed three times with PBS. The fluorescence intensities of the resuspended cells were measured by a flow cytometer using 488nm excitation and a 515nm band-pass filter for DCF detection and a filter >560nm for DHE detection (Becton-Dickinson Immunocytometry Systems, San Jose, CA).

Determination of mitochondrial C I activity
Extraction of mitochondrial proteins was performed as previously described [18].The enzyme activity assay of NADH CoQ reductase (mitochondrial C I) was carried out according to the instruction manual of the Mitochondrial C I Assay Kit (Genmed Scientifics, Inc.). Protein quantity was estimated by BCA protein assay kit (Beyotime, China). The C I activities were all normalized by the protein content in each sample and converted to the percentage of the control group.

Respiratory activity
The tetrazolium salt CTC is frequently used as indicator of microorganisms' respiration [19]. Reduction of CTC is an indication of respiratory enzyme activity. Respiratory activity was assessed by using CTC (5-cyano-2,3-ditolyl tetrazolium chloride), a monotetrazolium redox dye which produces a CTC-formazan (CTF) fluorescent complex (indicated by cells stained in red) when it is biologically reduced, indicating respiration (metabolic activity). Samples were stained with 2.5mM CTC for 30 min. The respiratory activity was determined by the intensity of fluorescence with flow cytometry. Images of cells were obtained using a fluorescence microscope.

Viability detection
To perform viability detection, the yeast cells were suspended at a concentration of 10 4 cells/ mL in 5 mL YPD medium with different concentrations of compounds. The tubes were incubated at 37˚C without shaking for 24 h, and then 100 μL aliquots were washed, diluted and plated for colony counts on YPD agar medium. The plates were incubated for 48 h at 37˚C and colonies were counted to determine CFU per mL.
RNA isolation, cDNA library construction and sequencing C. albicans SC5314 cells were grown overnight at 37˚C. The overnight culture was used to inoculate 20 ml of YPD to an initial 1×10 7 cells/mL, and incubated at 37˚C for an additional 4 h at 150 rpm. The cells were then harvested and RNA extraction was performed by using the hot acidic phenol method [20]. Then genomic DNA was removed from RNA sample using DNase (New England Biolabs). RNA purity was assessed using the Nanodrop-2000 (Thermo Scientific, USA). Each RNA sample had an A260:A280 ratio above 1.9 and A260:A230 ratio above 1.8. Total RNA integrity was then subsequently checked using an Agilent Technologies 2100 Bioanalyzer with an RNA Integrity Number (RIN) value greater than or equal to 8. Next, two sequencing libraries were constructed by TruSeq™RNA Sample Preparation Kit according to the product instruction (Illumina). Each library was sequenced using Illumina HiSeq2500 for 2 × 150 bp pair-end (PE) sequencing.

RNA-Seq data analysis and differentially expressed gene identification
Quality control of all raw reads was conducted by Fastqc (http://www.bioinformatics.babraham. ac.uk/projects/fastqc/) software. An initial filtering step was performed to exclude poor quality reads, including adaptor reads, ambiguous nucleotides and low-quality reads (reads having more than 50% bases with quality value). Then the clean reads were mapped onto the C. albicans SC5314 reference genome [21] (C_albicans_SC5314_version_A21-s02-m07-r10(http://www. candidagenome.org/download/sequence/C_albicans_SC5314/Assembly21/archive/) independently by TopHat [22] (v2.0.10). The '-G' option of Tophat together with the Gene Transfer Format (GTF) file of the Ensemble gene annotation was used for read mapping. The other parameters were set to default values. Based on the Tophat alignment BAM file, HTSeq [23] (v0.6.1) was used to estimate and quantify gene expression with default parameters, yielding raw read count for each genes. Gene expression was measured in reads per kilobase of exon per million reads mapped (RPKM). Finally, edgeR [24] was used to identify the differentially expressed genes (DEGs) by pairwise comparisons. The difference was considered significant if the FDR 0.05.

Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis
GO enrichment analysis provides all GO terms that are significantly enriched in DEGs, relative to the genome background, and filters the DEGs that correspond to a specific biological functions. This method firstly maps all DEGs to GO terms in the database (http:// www.geneontology.org/), calculating gene numbers for every term, and then uses the hyper geometric test to find significantly enriched GO terms in the input list of DEGs. GO: TermFinder (http://smd.stanford.edu/help/GO-TermFinder/GO_TermFinder_help. shtml) was used to enrichment the differentially expressed genes. The calculated p-value goes through Bonferroni Correction, taking corrected p-value 0.05 as a threshold. KEGG pathway enrichment analysis identifies significantly enriched metabolic pathways or signal transduction pathways in the whole genome background using KOBAS [25], taking the FDR 0.05 as a threshold to identify enrichment pathway.

Quantitative real-time PCR analysis
Quantitative PCR (qPCR) was performed as described previously [20]. Primer sequences used for the amplification of specific genes are shown in S1 Table. Statistical analysis All data were presented as means ± standard error of the mean (S.E.M.). Graphs were generated using Microsoft Excel (Microsoft Corp., Redmond, WA). Statistical analysis was performed using SPSS 12.0 (SPSS Inc., Chicage, USA). Differences between groups were determined using analysis of variance (ANOVA).A p value < 0.05 was considered significant.

Induction of oxidative stress after honokiol treatment on C. albicans
The term ROS encompasses oxygen free radicals, such as O 2 •− and H 2 O 2. The levels of O 2 •− and H 2 O 2 can be detected by DHE and DCFH-DA staining, respectively [26]. Mitochondria have long been established as a major source of ROS, which is generated from O 2 by electron leakage or reverse electron transport originating in the mitochondrial transport chain, during respiration [6]. Fig  with DCFH-DA or DHE. The quantification of ROS was performed by flow cytometry. 16μg/ mL honokiol induced DCF positive staining in 28.7%±3.1% cells, while most of cells (90.3 ±3.7%) were DHE positive staining after honokiol exposure. When addition of honokiol at 32μg/mL, the levels of ROS were further increased, 62.0%±3.6% for DCF positive cells and 99.1%±1.8% for DHE positive cells, respectively ( Fig 1B). These results concluded that honokiol stimulated ROS formation, mainly O 2 Effect of respiratory chain inhibitors on the oxidative burst and C. albicans killing induced by honokiol •− typically arises because of electron leakage or reverse electron transport from the electron transport chain onto O 2 during aerobic respiration [7]. To further investigate the sites of O 2 •− production in mitochondria, we first used rotenone (respiratory chain C I inhibitor) to observe the effect on C. albicans killing and ROS production by honokiol. Pre-cultured with rotenone (0.156mM or 0.31mM) eliminated the antifungal activities of honokiol when the concentration was 16μg/mL (Fig 2A). We even tested the fungicide concentration 32μg/mL of honokiol when combined with two different concentrations of rotenone, yet the results were the same. In addition, ROS production (both H 2 O 2 and O 2 •− ) by honokiol was also completely eliminated by rotenone (Fig 2B and 2C). Rotenone can completely block both the antifungal activity of honokiol and ROS production (Fig 2). Then, we used several other respiratory chain inhibitors to observe the effect on C. albicans killing and ROS production by honokiol. Precultured with thenoyltrifluoroacetone (TTFA), menadione, sodium azide (NaN3), and oligomycin showed almost no effect on honokiol-induced Candida killing and O 2 •− accumulation (Fig 3). Salicylhydroxamic acid (Sham), a cyanide-resistant alternative oxidase inhibitor, could slightly enhance the susceptibility to honokiol and raised the level of O 2 •− (Fig 3). These data indicated that the potential antifungal activity of honokiol was related to mitochondrial C I. Nevertheless, the mechanisms and consequences of such an interaction still require further investigation.
Respiratory activity in C. albicans after exposure to honokiol CTC staining represents an index of the respiratory activity of the cell at the time of observation [19].CTC is a soluble crystal that forms a nearly colorless non-fluorescent solution.
In the electron transport system, CTC, as an artificial redox partner, was reduced to a fluorescent formazan crystal (CTF) primarily by membrane-bound NADH-dehydrogenase [19]. The accumulation of CTF red fluorescence was clearly observed in control group, but honokiol induced a concentration-dependent decrease in respiratory activity ( Fig 4A). Almost no CTF accumulation was observed both in 16μg/mL and 32μg/mL of honokioltreated cells, indicating that the respiratory activity of C. albicans was blocked by honokiol. These results were also confirmed by flow cytometry (Fig 4B). We further investigated the C I enzyme activity in C. albicans after honokiol exposure. As shown in Fig 4C, the C I enzyme activities in C. albicans after honokiol treatment at 8μg/mL and 16μg/mL were reduced to 68.5% and 39.9%, respectively. In addition, 32μg/mL honokiol could almost completely inhibit the activity of C I enzyme. There was no significant difference between 4μg/mL honokiol treatment and the control group.
The effect of honokiol on mtDNA in C. albicans Staining with DAPI revealed that honokiol-treated cells were devoid of mtDNA, while the control group showed the punctuate staining for mtDNA when observed under fluorescence  (Fig 5A). In addition, we observed a significant induction of mitochondrialencoded subunit 1 of cytochrome c oxidase (COX1) gene and subunit 6 of the F0 sector of mitochondrial F1F0 ATP synthase (ATP6) gene, compared with control ( Fig 5B). Honokiol treatment increases the sensitivity to cell wall-perturbing agents Mitochondrial dysfunction in C. albicans is reported to be linked with loss of cell wall integrity [27][28][29]. Therefore, we compared the sensitivities of C. albicans to the cell-wall perturbing agents Congo red (4μg/mL) or SDS (0.01%) alone or in combination with honokiol (4μg/mL). As shown in Fig 6, honokiol significantly increased the sensitivity of the cell wall-perturbing agents against C. albicans.

GO assessment
Gene annotation and classification with GO terms for C. albicans was performed in Candida Database. The DEGs were mapped to GO ontologies of cellular components, molecular functions, and biological processes for GO enrichment analysis (Table 1). GO analysis revealed DEGs in honokiol treatment were enriched for 9 GO terms. The largest number of DEGs were

KEGG analysis
We also employed the KEGG pathway database to determine if the DEGs belong to particular signaling pathways. KEGG pathway mapping is the process to map molecular datasets, especially large-scale datasets, to the KEGG pathway maps for biological interpretation of higherlevel systemic functions. The signaling pathway hypothesized to be affected by honokiol treatment mainly included signaling pathways related to metabolism (oxidative phosphorylation, arginine and proline metabolism, alanine, aspartate and glutamate metabolism,glycolysis/gluconeogenesis, fatty acid degradation, nitrogen metabolism, pyruvate metabolism, carbon metabolism, alpha-linolenic acid metabolism, glutathione metabolism, degradation of aromatic compounds, and citrate cycle), genetic information processing (ribosome, ribosome biogenesis in eukaryotes, proteasome, DNA replication, and ubiquitin mediated proteolysis), cellular processes (cell cycle, peroxisome, endocytosis, and regulation of autophagy), and environmental information processing (ABC transporters and MAPK signaling pathway) (Fig 9).

Discussion
Candida species are a group of opportunistic fungal pathogens in humans, particularly among immunocompromised and hospitalized population [1][2][3]. Only few classes of antifungals such as azoles, flucytosine, polyenes, allylamines, and echinocandins are available for the treatment of Candida infections [30]. However, the rising drug resistance is an inevitable problem [31]. Therefore, it is necessary to discover new antifungal agents or other safer alternatives to improve the effectiveness of treatment of Candida infections [32,33]. Honokiol is an interesting small-molecule exhibiting pro-and antioxidant activities, similar with vitamin C and E [14]. Past studies have also demonstrated that honokiol has antifungal activities (MIC rang, 8-32 μg/mL) and when used in combination with fluconazole, can exhibit synergistic activity against clinical isolates of fluconazole-resistant C. albicans [20,34,35].To better understand how honokiol works on a cellular level, we sought to explore its antifungal mechanism and found honokiol caused an increase of intracellular ROS, lipid peroxidation, and protein oxidation in a dose-dependent manner in C. albicans [15]. This study will aim to clarify specific mitochondrial ROS-generation components that are involved in honokiol treatemt. Mitochondria have always been considered as a primary source of ROS like the superoxide anion radicals, which are converted from O 2 by electron leakage from the mitochondrial transport chain during cellular respiration [6,7]. In the current study, we demonstrated that superoxide radicals was mainly generated and accumulated in mitochondria after honokiol exposure by using fluorescent ROS probes DHE and DCFH-DA (Fig 1). The mitochondrial respiratory chain is the major source of intracellular ROS generation. About 1-2% of the molecular oxygen consumed during normal physiological respiration is converted into superoxide radicals who serve as the precursor of most ROS [6]. There is growing evidence that most of the superoxide radicals generated by mitochondria is produced by C I and C Ⅲ [7,8]. To obtain an insight into intra-mitochondrial site of honokiol-induced overproduction of superoxide radicals, we used specific respiratory chain inhibitors. Only rotenone can completely block superoxide radicals generation and deprived of fungicidal activity of honokiol (Figs 2 and 3). Since rotenone is a C I inhibitor, we hypothesis that honokiol may compete with rotenone at or close to the quinone binding sites of C I or the binding site of honokiol is upstream from the rotenone binding site. In addition, we also found mitochondrial NADH dehydrogenase activity was decreased after honokiol, indicating the C I activities were inhibited (Fig 4). Consistent with the above results, respiratory activity was decreased, which may be associated with the inhibition of C I (Fig 4). In view of the fact that O 2 •− production is completely inhibited by the addition of a specific C I ubiquinone-site inhibitor rotenone that prevents C I binding to CoQ, we concluded the mechanism by which honokiol produces large amounts of O 2 •− is specifically through respiratory C I reverse electron transport [36,37]. One important target of ROS is the mitochondrial DNA due to the close proximity to the electron transport chain, the major locus for free radical production, and the lack of protective histones [38]. DAPI staining revealed that honokiol treatment caused loss of mtDNA, especially after 32μg/mL honokiol treatment, the cells were completely devoid of mtDNA. However, we observed increased transcription of mitochondrial genes (ATP6 and COX2) after honokiol exposure (Fig 5). It is possible that cells after honokiol exposure attempts to overcome the ROS-induced mtDNA damage by increased mtDNA genes expression, which reflects a compensatory mechanism [39,40].
We identified a number of DEGs from honokiol treatment. After honokiol treatment, we observed increased expression of genes involved in mitochondrial respiratory electron transport chain, oxidoreductase activity, and ribosome biogenesis (Fig 7 and Table 1). C I is the first step in the respiratory chain and is located in the inner mitochondrial membrane [7]. The molecular mass of CI is around 800kDa and is composed of at least 39 subunit proteins [29]. To date, the CI structure in C. albicans remains somewhat unclear. C. albicans has 14 core CI genes of bacterial progenitor that are proton-pumping NADH: ubiquinone oxidoreductases, seven of which are encoded by mitochondrial DNA (NAD1, NAD2, NAD3, NAD4, NAD4L, NAD5, and NAD6), the other subunit genes are nuclear-encoded. Since methods to selectively mutate the mitochondrial genome in C. albicans are unavailable, the core CI genes encoded by mitochondrial DNA in C. albicans are not studied yet [29]. Our further work is required to clarify the interaction of honokiol with mitochondrial C I.
There are many lines of evidence to suggest that oxidative stress is associated with the action of honokiol [14,15]. Honokiol could trigger a multifaceted response (Fig 10). It activates stress response pathway to tolerate ROS toxicity, including increased expression of genes involved in the oxidative stress response (CAP1, GTT12, PRX1, MXR1, GPX1, TRX1, IFH1, and SOD6), proteolysis and protein degradation (UBC4, UBC15, UBC8, and UBP6), and autophagy (ATG8, ATG9, ATG27, orf19.3998, APG13, and PRB1) (Fig 8 and S2 Table). CAP1, an AP-1 bZIP transcription factor, was significantly upregulated after honokiol treatment. Cap1 regulates a variety of oxidative stress response including the cellular antioxidant defense system, carbohydrate metabolism and energy metabolism, protein degradation, ATP-dependent RNA helicase, and resistance pathways [41]. Redox homeostasis in cells is important for maintaining proper cellular functions [42]. Honokiol targets on mitochondrial respiratory chain C I, inducing ROS accumulation, disruption of intracellular redox homeostasis, irreversible oxidative modifications of lipid, protein or DNA, and activation of autophagy or apoptosis signaling pathway.
There are many lines of evidence to suggest that mitochondrial dysfunction in C. albicans is known to be linked with loss of cell wall integrity [27][28][29]. The fungal cell wall has vital roles in growth, survival, morphogenesis, and pathogenicity. Critical for the coordination of these activities is the dynamic nature of the cell wall, its ability to respond to external and internal stimuli [43]. Altered sensitivities to cell wall perturbing agents indicate modified cell wall properties [43,44]. In our study, honokiol treatment resulted in an increased sensitivity to Congo red and SDS (Fig 6). In addition, the genes for cell wall synthesis have three major changes in transcriptional levels (Fig 8): 1). chitin/mannan/glucan biosynthesis, such as upregulated CHS2 which encodes chitin synthase and BMT3 which encodes β-mannosyltransferase is required for β-mannan elongation either in phosphopeptidomannan or in phospholipomannan [44]. Reduced glucanase gene expression, for example, SKN1 encoding a N-glycosylated type II membrane protein with a predicted role in β-1,6-glucan synthesis was downregulated after honokiol exposure; 2). GPI anchor protein, such as PGA31, EXG2, PGA32, PGA57, and PGA37 were upregulated. GPI-anchored proteins are important cell surface proteins in C.albicans, and they have important impact on the adhesion,morphogenesis and cell wall synthesis [44]; 3). cell wall associated kinase. such as PSK1(PAS kinase), a putative serine/threonine protein kinase, was upregulated after honokiol exposure. It has been reported that inhibition of cell wall biogenesis by caspofungin also causes an increase in PSK1 expression [44]. Honokiol treatment caused mitochondrial dysfunction and then damaged the cell wall integrity, thus, increased sensitivity to cell wall perturbing agents.
The present work shows that honokiol may target mitochondrial respiratory chain C I, leading to mitochondrial dysfunction, accompanied by increased cellular superoxide anion and oxidative stress (Fig 10). Use of antifungal agents that are mitochondrial respiratory chain C I inhibitors can also be toxic to mammalian cells associated with aging and age-related diseases such as Parkinson's disease [45]. However, potential side effects of mitochondrial respiratory chain inhibitors as antifungal agents can be reduced if effective dosage can be diminished. Such a reduction can be achieved by use of honokiol as a chemosensitizer to azoles agents or echinocandins. This work not only provides insights on the mechanism by which honokiol interferes with fungal cell, demonstrating previously unknown effects on mitochondrial physiology, but also raises a note of caution on the use of M. officinalis as a Chinese medicine due to the toxic for mitochondria and suggests the possibility of using honokiol as chemosensitizer.
Supporting information S1 Table. Gene-specific primers used for real-time RT-PCR. (DOC)