New Insights into the Structure and Mode of Action of Mo-CBP3, an Antifungal Chitin-Binding Protein of Moringa oleifera Seeds

Mo-CBP3 is a chitin-binding protein purified from Moringa oleifera Lam. seeds that displays inhibitory activity against phytopathogenic fungi. This study investigated the structural properties and the antifungal mode of action of this protein. To this end, circular dichroism spectroscopy, antifungal assays, measurements of the production of reactive oxygen species and microscopic analyses were utilized. Mo-CBP3 is composed of 30.3% α-helices, 16.3% β-sheets, 22.3% turns and 30.4% unordered forms. The Mo-CBP3 structure is highly stable and retains its antifungal activity regardless of temperature and pH. Fusarium solani was used as a model organism for studying the mechanisms by which this protein acts as an antifungal agent. Mo-CBP3 significantly inhibited spore germination and mycelial growth at 0.05 mg.mL−1. Mo-CBP3 has both fungistatic and fungicidal effects, depending on the concentration used. Binding of Mo-CBP3 to the fungal cell surface is achieved, at least in part, via electrostatic interactions, as salt was able to reduce its inhibitory effect. Mo-CBP3 induced the production of ROS and caused disorganization of both the cytoplasm and the plasma membrane in F. solani cells. Based on its high stability and specific toxicity, with broad-spectrum efficacy against important phytopathogenic fungi at low inhibitory concentrations but not to human cells, Mo-CBP3 has great potential for the development of new antifungal drugs or transgenic crops with enhanced resistance to phytopathogens.


Introduction
Plants use several strategies to overcome fungal attacks, including the production of antimicrobial peptides and proteins [1], [2]. Much effort has been dedicated to researching these bioactive constituents, particularly because the chemically-synthesized antifungal compounds used to prevent and contain these pathogens comprise a potential environmental threat [3], [4], [5]. In general, these defense-related proteins interfere with the fungal life cycle by either impairing growth or killing the pathogen [6], [7]. The antifungal properties of these proteins may be exploited for use in the development of transgenic crops that have enhanced resistance to phytopathogenic fungi [8].
Chitin-binding proteins (CBPs) represent a group of proteins also found in plants that often have a basic pI, a molecular mass ranging from 3.1 kDa up to 20 kDa, and high resistance to both extreme pH changes and proteolysis. Some CBPs have the ability to inhibit fungal growth [9], as they bind to and disrupt the proper function of chitin, a key component of the fungal cell wall [10]. It has been suggested that the binding of these proteins to chitin in filamentous fungi leads to the disruption of both cell wall biogenesis and cell polarity [11], [12].
Recently, our research group isolated a chitin-binding protein named Mo-CBP 3 from Moringa oleifera Lam. seeds [13]. This protein is a basic glycoprotein (18 kDa by SDS-PAGE) and does not display haemagglutination, chitinase or b-1,3-glucanase activity. Mo-CBP 3 presented potent antifungal activity against the phytopathogenic fungi Fusarium solani, F. oxysporum, Colletotrichum musae and C. gloeosporioides at a low concentration (0.05 mg.mL 21 ). The phytopathogenic effect of Mo-CBP 3 against fungi was observed even when the protein was heated at 100uC for 1 h or pre-treated with 150 mM N-acetyl-D-glucosamine.
As Mo-CBP 3 has a low molecular mass and is a protein with potent antifungal activity at low concentrations, it is a very promising bioactive candidate that may be explored to determine whether it can confer resistance against phytopathogenic fungi to economically and nutritionally important crops. To further test Mo-CBP 3 , it is essential to obtain additional knowledge about its structure and mode of action. Here, we report new structural features of Mo-CBP 3 that reveal a correlation between its structural stability and its antifungal activity. In addition, to know about the mechanisms by which this protein acts as an antifungal agent, its ability to induce the endogenous production of reactive oxygen species and to trigger morphologic and ultrastructural alterations were analyzed using F. solani as a model. F. solani is an easy-to-handle and fast-developing species, making it ideal for in vitro assays, and it holds relevance as a phytopathogenic fungus that attacks economically important crop plants. Furthermore, to have a preliminary clue whether Mo-CBP 3 , as a chemical agent against fungi, displays cytotoxicity, the level of lysis of the human red blood cells was examined.

Mo-CBP 3 preparation
A highly purified Mo-CBP 3 preparation devoid of contaminating proteins was obtained according to Gifoni et al. [13]. Mature seeds were ground in a coffee grinder, and the resulting flour was treated with n-hexane. Defatted flour was extracted with 50 mM Tris-HCl, pH 8.0, containing 150 mM NaCl (1:10 w/v), for 4 h at 4uC under constant stirring and was then filtered through cheesecloth. After centrifugation at 15,000 g at 4uC for 30 min, the supernatant was exhaustively dialyzed against Milli-Q grade water and centrifuged again under the same conditions. (NH 4 ) 2 SO 4 was added to the soluble material, denoted as the albumins, to yield 90% saturation. This protein fraction (F0-90%) was then dissolved in and dialyzed against the extracting buffer and applied to a chitin column that had been equilibrated with the same buffer. After elution with the starting buffer of the unbound proteins from the column, the chitin-bound proteins, named P NAG and P AC , were eluted with 100 mM N-acetyl-D-glucosamine (NAG) that was prepared in the extracting buffer and with 50 mM acetic acid (pH 5.0), respectively. The P NAG sample was dialyzed against 100 mM acetic acid and distilled water, lyophilized and applied to a cation-exchange matrix (Resource S) that had been previously equilibrated with 50 mM sodium acetate buffer, pH 5.2. Four major adsorbed protein peaks (Mo-CBP 2 , Mo-CBP 3 , Mo-CBP 4 , and Mo-CBP 5 ) were recovered after being selectively desorbed by stepwise elution with 400, 500, 600, and 700 mM NaCl, respectively, included in the equilibrium buffer. As Mo-CBP 3 was purified to homogeneity, had high yield and presented the highest activity against the phytopathogenic fungi Fusarium solani, Fusarium oxysporum, Colletotrichum musae and Colletotrichum gloesporioides, as previously reported by our research group [13], it was used for further analyses. The purity of Mo-CBP 3 was checked by denaturing gel electrophoresis [14]. The identity of Mo-CBP 3 was confirmed by N-terminal amino acid sequence analysis by Edman degradation (Shimadzu PPSQ-10A automated protein sequencer).

Protein concentration
The protein concentration was determined according to Bradford [15], using BSA as a standard. Absorbance at 280 nm was also used to monitor the protein elution profiles during chromatography.
Far-UV circular dichroism (CD) spectroscopy CD spectra measurements were made on a JASCO J-715 spectropolarimeter (Jasco Instruments, Tokyo, Japan) in an N 2 atmosphere at 25uC. Mo-CBP 3 (40 mg) was dissolved in 20 mM phosphate-buffered saline (PBS) at pH 7.0 and transferred to a rectangular quartz cuvette with a 0.1 cm path length. Eight scans were performed with a scan rate of 20 nm.min 21 and a 4 s response time. CD spectra were measured from 190 to 250 nm. The contributions of the secondary structural elements of Mo-CBP 3 were determined by CD spectra deconvolution analyses using the basis reference protein set SMP56 of the CDPro software [16] and applying three methods, CONTIN/LL, SELCON 3 and CDSSTR. CD spectroscopy was also used to assess Mo-CBP 3 thermal stability. To do this, Mo-CBP 3 (40 mg in PBS) was heated gradually in 10uC increments from 26 to 90uC in a TC-100 circulating water bath (Jasco). The samples were maintained at each temperature for 10 min, and spectra were recorded from 190 to 250 nm. To evaluate structural stability as a function of pH,  The filamentous fungus F. solani was grown in Petri dishes containing potato dextrose agar (PDA) medium for 12 days at room temperature (22uC). Fresh conidia suspensions were prepared by rinsing the surface of the 12-day-old sporulated cultures with sterile distilled water and the aid of a triangular Drigalski rod. Spore suspensions were filtered through cheesecloth in a laminar flux chamber under sterile conditions, and conidia were quantified using a Neubauer chamber under an optical microscope (Olympus System Microscope BX 60). Antifungal assays were conducted as described by Ji and Kuc [17]. For analyzing the changes in conidial germination as a function of pH, Mo-CBP 3 samples, at antifungal concentration (0.1 mg.mL 21 ), were dissolved in 20 mM sodium acetate/phosphate/borate buffer at different pH values (2.0, 4.0, 6.0, 8.0, 10.0 and 12.0) and incubated with 10 mL of the conidia suspension (2610 5 .mL 21 ) in reticulated plates. For the non-inhibitory controls, conidia were incubated in each buffer in the absence of Mo-CBP 3 . The plates were placed in a plastic box maintained near 100% relative humidity at 22uC in the dark for 24 h. After this time, 50 conidia were randomly selected from each treatment and evaluated for germination under an optical microscope. A conidium that had emitted a hyphae at least twice the length of the ungerminated conidium was considered to have successfully germinated. In parallel, to assess whether the ability of Mo-CBP 3 to inhibit spore germination was affected by heat treatment, Mo-CBP 3 was heated in a water bath at 100uC for 60 min, cooled on ice for 10 min, and assayed as described above. Each experiment was performed in triplicate, and images were taken with a digital camera (Sony, MCV-CD350 model, 14.2 megapixels). Effect of Mo-CBP 3 on the mycelial growth and conidial viability of F. solani A quantitative assay for fungal growth inhibition was performed following the protocol developed by Broekaert et al. [18]. A conidia suspension (2610 5 cells.mL 21 ) was incubated in 96-well flat microplates with 100 mL of potato dextrose broth in the absence of Mo-CBP 3 and allowed to germinate for up to 12 h in the dark at 37uC. Next, 100 mL of different concentrations of Mo-CBP 3 (0.05, 0.1, 0.5 and 1 mg.mL 21 ) were added. Cell growth was also determined without the addition of Mo-CBP 3 . Fungal growth was monitored by turbidimetry at 630 nm from 0 to 49 h using an automated microplate reader (Model Elx800, Bio-Tek Instruments). The absorbance values taken immediately after Mo-CBP 3 addition were recorded and established as zero and were discounted from every readings taken onwards. To evaluate the conidial viability of F. solani after treatment with different concentrations of Mo-CBP 3 , 150 mL aliquots were taken from the wells, transferred to Eppendorf tubes and centrifuged at 3,000 g for 1 min at 25uC and the supernatant discarded. The remaining conidia were washed with sterile distilled water to remove Mo-CBP 3 , reculturing in Petri dishes containing PDA medium and kept in an incubator at 27uC. Images of the mycelium growth were taken after 5 days. All experiments were carried out in triplicate. To evaluate the presence of electrostatic interactions between Mo-CBP 3 and the conidial plasma membrane, Mo-CBP 3 (0.1 mg.mL 21 ) was first dissolved in solutions with different NaCl concentrations (25, 75 and 150 mM) and antifungal assays were performed following the methodology described in Section 2.5. In the negative, non-inhibitory controls, conidia were incubated only in 25, 75 and 150 mM NaCl, all in the absence of Mo-CBP 3 . As a positive inhibitory control, Mo-CBP 3 was used at antifungal concentration (0.1 mg.mL 21 ). All experiments were carried out in triplicate.

Evaluation of reactive oxygen species (ROS) production by F. solani conidia after Mo-CBP 3 treatment
To evaluate the ability of Mo-CBP 3 to induce the endogenous production of ROS in F. solani conidia, the in situ assay described by Thordal-Christensen et al. [19], with some modifications [20], was conducted, using 3,39-diaminobenzidine (DAB).

Haemolytic assay
This was carried out using human red blood cells (hRBCs) collected from healthy donors in heparinized tubes [21]. hRBCs were separated from plasma by centrifugation (3,000 g, 10 min, 25uC) and washed three times with 100 mM sodium phosphate buffer, pH 7.4, containing 150 mM NaCl (PBS). A 1% (v/v)   suspension was prepared and incubated for 1 h at 37uC with serial dilutions of Mo-CBP 3 (from 280 to 0.137 mM) in PBS. After the incubation period, suspensions were centrifuged at 3,000 g for 10 min at 25uC, aliquots of the supernatants were transferred to Eppendorf tubes and the absorbances taken at 405 nm (spectrophotometer Novaspec II, Pharmacia) to monitor the release of haemoglobin. Triton X-100 and PBS were used as positive (100% haemolysis) and negative controls, respectively. The haemolysis percentage was calculated using the following equation

Mo-CBP 3 preparation
The Mo-CBP 3 preparations used in the present study were confirmed to be homogeneous and free of contaminants, as shown in a representative Figure 1. Mo-CBP 3 is an 18 kDa protein that inhibits the conidia germination of F. solani, F. oxysporum, C. musae and C. gloeosporioides, confirming the findings of Gifoni et al. [13]. The purity of Mo-CBP 3 was further proved by N-terminal sequencing analysis.

Structure-antifungal activity relationships
The far-UV CD spectra of native Mo-CBP 3 showed minima at approximately 208 and 222 nm (Figure 1). The deconvolution of the CD spectra performed using the CDPro package [16] revealed the following content of the secondary structure fraction: 30.3% ahelices, 16.3% b-sheets, 22.3% turns and 30.4% unordered forms. Therefore, Mo-CBP 3 can be classified as an alpha-beta protein [22].
To further characterize Mo-CBP 3 , the effects of temperature and pH on its secondary structure and antifungal activity were evaluated. No significant changes in the CD spectra of Mo-CBP 3 were observed after heat treatment at 90uC for 10 min (Figure 2A). After heat treatment at 100uC for 60 min, the CD spectra of Mo-CBP 3 demonstrated only a discrete alteration ( Figure 2B). Similarly, Mo-CBP 3 was still able to inhibit the spore germination of F. solani after heating at 100uC for 60 min.
Additionally, the CD spectral shape did not change from pH 2.0 to pH 12.0 ( Figure 3A), suggesting that the protein structure is maintained and that even the pH extremes were insufficient to alter the net charge of Mo-CBP 3 in a way to cause electrostatic repulsion, with later rupture of the hydrogen bonds. To correlate the structure of Mo-CBP 3 to its antifungal activity, Mo-CBP 3 was dissolved in 20 mM sodium acetate-borate-phosphate buffer at different pH values, and the antifungal activity on F. solani spore germination was tested. The inhibitory activity of Mo-CBP 3 was similar at all pH ranges tested (4.0, 6.0, 8.0, 10 and 12.0) ( Figure 3B). However, it was not possible to evaluate the behavior of the protein at pH 2.0, as spore germination did not occur even in the control, most likely because this pH is very acidic and has an adverse effect on the development of F. solani. These data demonstrate that Mo-CBP 3 exhibits high resistance to both temperature and pH changes, thus retaining its antifungal activity. The elevated structural and functional stability of Mo-CBP 3 can be attributed to the presence of cysteine residues in its structure. Mo-CBP 3 is a chitin-binding protein, and many proteins for which the amino acid sequences are known that possess this property share a common structural domain composed of 43 amino acids, with many cysteine and glycine residues in conserved positions [23]. In fact, of the 22 identified residues from the N-terminus of Mo-CBP 3, 27.3% were cysteines [13]. The presence of such residues can lead to formation of disulphide bridges, making these proteins more resistant to denaturation [9], [24]. In fact, the reduction of these disulphide bridges abolished the antifungal activity of Mo-CBP 3 . Several disulphide bonds are present in osmotins and thaumatin-like proteins, and it is thought that they contribute to the high structural stability of these proteins [25].

Effect of Mo-CBP 3 on the mycelial growth and conidial viability of F. solani
In addition to inhibiting spore germination, the ability of Mo-CBP 3 to affect fungal growth was analyzed. Mo-CBP 3 was inhibitory to the mycelial mass development of F. solani in comparison to the control incubated in the absence of Mo-CBP 3 ( Figure 4A). Mo-CBP 3 displayed a significant inhibitory effect on fungus growth at a concentration of only 0.05 mg.mL 21 , with close to 62% inhibition within 49 h. At higher concentrations (0.5 and 1.0 mg.mL 21 ), the antifungal effect was observed at earlier stages. For example, at 1.0 mg.mL 21 , Mo-CBP 3 inhibited 94% of the growth of F. solani 26 h post-incubation. In reality, Mo-CBP 3 behaves both as fungistatic and fungicidal protein, depending on its concentration and the stage of fungus development. Preincubation of F. solani spores with Mo-CBP 3 at concentrations up to 0.5 mg.mL 21 for 49 h followed by removal of the protein restored the mycelial growth capacity of the fungus, indicating the fungistatic effect of Mo-CBP 3 at low concentrations. In contrast, prior incubation of F. solani spores with 1.0 mg.mL 21 Mo-CBP 3 followed by removal of the protein abolished (100% inhibition) the fungus viability, as mycelial growth was inhibited ( Figure 4B), which characterizes Mo-CBP 3 fungicidal action. It is well known that several chitin-binding proteins have antifungal activity [26]. A chitin-binding lectin from Setcreasea purpurea (SPL) causes inhibition of Rhizoctonia solani, Penicillium italicum, Sclerotinia sclerotiorum, and Helminthosporium maydis at 1.51 mg.mL 21 [27]. However, it is remarkable that the inhibitory effects of Mo-CBP 3 on F. solani growth were more pronounced in this study than those observed by Gifoni et al. [13]. The differences observed are presumably due to the different protocols used. For example, the growth inhibition assays in the present study were performed in liquid medium, differing from the previous study, which was made on solid medium (Petri dish). In addition, it is worth noting that the cultivation medium has a great influence on the antifungal activity of Mo-CBP 3 , as F. solani growth inhibition could be detected only when yeast extract was not present in the medium composition. Based on these data, it is plausible to speculate that in the presence of yeast extract, which contains cell wall fragments and negatively charged mannan [28], the interaction of Mo-CBP 3 with the filamentous fungus would be compromised, affecting its antifungal activity.

Mode of action of Mo-CBP 3 upon fungal cell
Antimicrobial molecules possess several features to fulfill their role in plant defense mechanisms. For rapid killing, antimicrobial molecules often act at the cell surface rather than the cell interior [29]. Therefore, it was hypothesized that besides the binding of Mo-CBP 3 to the fungus chitin, since it is a chitin-binding protein as established by affinity chromatography [13], Mo-CBP 3 could also bind to F. solani cell membrane components, at least in part, via electrostatic interactions. This prediction is supported by the observation that NaCl at 25, 75 and 150 mM reduced the inhibitory effect of Mo-CBP 3 (0.1 mg.mL 21 ) on F. solani spore germination in comparison to the controls in the absence of Mo-CBP 3 and presence of the same above concentrations of NaCl ( Figure 5). Sensitivity to ionic strength with loss of antimicrobial activity has been described for other basic antimicrobial proteins and peptides that are thought to act via electrostatic interactions with negatively charged membrane components, as well as for those that bind to specific receptors [30], [31]. In the F. solani membrane, such electrostatic interactions of Mo-CBP 3 , which is a cationic protein, most likely occur with the negatively charged phospholipid phosphatidylinositol [32].
After this initial binding to components of the fungal membrane, secondary effects that are induced internally in the cell were investigated. Figure 6C shows that Mo-CBP 3 (0.1 mg.mL 21 ) induced ROS production as reveled by the presence of a reddish-brown pellet inside the F. solani spores, in contrast to the negative controls, H 2 O and BSA (0.1 mg.mL 21 ) ( Figures 6A and 6B, respectively). The ROS induction capacity of various antifungal peptides and proteins has been previously reported. Similar to Mo-CBP 3 , the defensin from Phaseolus vulgaris (PvD 1 ) causes ROS induction in F. solani cells at 0.1 mg.mL 21 [33]. Another example is the Raphanus sativus antifungal peptide 2 (Rs-AFP 2 ), which is able to stimulate ROS production in Candida albicans in a dose-dependent manner, but is unable to do so in an Rs-AFP 2 -resistant Dgcs C. albicans mutant that lacks the Rs-AFP 2 -binding site in its membranes [34]. This finding suggests that upstream binding of the macromolecule is needed for ROS production. An increase in the generation of ROS that exceeds the cellular neutralization capacity of the fungus promotes oxidative stress and may cause the hyperoxidation of proteins, lipids and nucleic acids and consequently cell death [35].
SEM was employed to allow visualization of any morphological changes promoted by Mo-CBP 3 on F. solani cells. Photomicrographs of F. solani conidia were taken 48 h after growth in the presence or absence of Mo-CBP 3 (0.05 mg.mL 21 ). Normal hyphal growth was observed in the control cells ( Figure 7A), but not in in the cells treated with Mo-CBP 3 ( Figure 7B). Closer examination of F. solani cells treated with Mo-CBP 3 revealed loss of asymmetry, deformations and wrinkles in comparison to control cells, as represented in Figures 7D and 7C, respectively. Similar alterations were detected in S. cerevisiae cells after incubation with a 2S albumin-homologous protein (0.1 mg.mL 21 ) from passion fruit seeds [36].
Ultrastructural analysis of F. solani cells also revealed alterations in the presence of Mo-CBP 3 (0.05 mg.mL 21 ). It was observed condensation of the cytosol content, vacuolation and shrinkage of the cell wall ( Figures 8B and C) when compared with control cells ( Figure 8A). Vacuoles serve as compartments either for storage of resources or for detoxification purposes [37]. Thus, possibly the increased vacuole formation in the fungus cell might be related to a defense response of F. solani to the toxic effects of Mo-CBP 3 . In addition to these above changes observed, notable accumulation of electrodense granular material was observed in the cytosol of the cells incubated with Mo-CBP 3 ( Figure 8C). It is plausible to speculate that the electrodense granular material observed might result from the electrostatic interactions of the cationic Mo-CBP 3 with negatively charged primary or secondary metabolites present into the fungus cell, based on the coagulant (flocculent) properties of this protein as previously reported [13]. These alterations in F. solani morphology as visualized by TEM are typically found in cells that have undergone apoptosis [38]. This is in agreement with the results shown above, as ROS are classical apoptotic markers [39]. Thus, these data together suggest that the antifungal properties of Mo-CBP 3 are triggered by alterations in the cell surface. Brul et al. [40] found that the filamentous fungi Penicillium roqueforti, Trichoderma harzianum, Paecilomyces variotii, Aspergillus niger, and A. nidulans allow molecules up to 150 kDa to cross the cell wall. Thus, it cannot be ruled out that Mo-CBP 3 may eventually pass through the cell wall barrier, interact with the cell membrane receptors and induce secondary effects internally in F. solani to promote cell death.

Evaluation of cytotoxicity effects of Mo-CBP 3
Many antimicrobial proteins also exhibit toxic potential on eukaryotic cells. In this context, the mechanical stability of the membrane of red blood cells is a good indicator to evaluate in vitro the effects of various compounds when screening for cytotoxicity [41]. These cells may undergo a loss of membrane integrity and die rapidly as a result of cell lysis. Thus, to evaluate whether Mo-CBP 3 causes cytotoxicity, haemolytic assay was utilized by measuring the release of haemoglobin at different Mo-CBP 3 concentrations. Mo-CBP 3 was compared with the detergent Triton X-100, whose relative haemoglobin release was set at 100%. For all concentrations tested (from 0.137 a 280 mM) Mo-CBP 3 did not show haemolytic activity (Figure 9), suggesting that the antifungal action of this protein occurs via a selective interaction with the fungal membrane. This result shows that although Mo-CBP 3 displayed remarkable antifungal activity against phytopathogenic fungi, it shows no haemolytic activity. Similar results were found for an antifungal peptide (AFP-J) purified from potato tubers (Solanum tuberosum cv. L Jopung) [42].
In conclusion, this study reinforces previous data on the antifungal properties of Mo-CBP 3 and reports new information about its structural features and mode of action. The CD spectral data from different temperature and pH conditions indicate that the high structural stability of Mo-CBP 3 results in the effectiveness of its antifungal activity through interactions with the cell membrane, which causes prominent morphological changes followed by the induction of oxidative stress, eventually leading to cell death. Considering its elevated stability and specific toxicity, with broad-spectrum efficacy against important phytopathogenic fungi at low inhibitory concentrations but not to human cells, Mo-CBP 3 has great potential for the development of new antifungal drugs or transgenic crops with enhanced resistance to phytopathogens.