Antifungal Activity against Botrytis cinerea of 2,6-Dimethoxy-4-(phenylimino)cyclohexa-2,5-dienone Derivatives

In this work the enzyme laccase from Trametes versicolor was used to synthetize 2,6-dimethoxy-4-(phenylimino)cyclohexa-2,5-dienone derivatives. Ten products with different substitutions in the aromatic ring were synthetized and characterized using 1H- and 13C-NMR and mass spectrometry. The 3,5-dichlorinated compound showed highest antifungal activity against the phytopathogen Botrytis cinerea, while the p-methoxylated compound had the lowest activity; however, the antifungal activity of the products was higher than the activity of the substrates of the reactions. Finally, the results suggested that these compounds produced damage in the fungal cell wall.


Introduction
Botrytis cinerea is a phytopathogenic fungus promoted by the presence of free surface water or high relative humidity and causing significant crop losses in a wide variety of plant species [1]. Regarding the control, methods aiming to reduce humidity can be combined to help decrease this disease, in addition to chemical fungicides or biocontrol treatments [1]. Chemical control is the most common way to manage B. cinerea, mainly using synthetic compounds [1]. The restriction of this type of control becomes necessary to reduce the impact on the environment [2] and to avoid the acquired resistance to botrycides [3][4][5][6][7]. For this reason, the development of new antifungal compounds is essential. Natural products can be a good alternative to commercial fungicides [8,9]. For instance, phenolic compounds, terpenoids, nitrogen-containing compounds, and aliphatic compounds isolated from plants have shown antifungal activities [10][11][12]. Additionally, new antifungal compounds against B. cinerea derived of natural products have been synthesized, such as derivatives of natural stilbene resveratrol [13], chlorophenyl derivatives [14], or different clovanes [15].
Several phenolic metabolites found in grape pomace have shown low antifungal activity against B. cinerea [16], therefore, it is possible to increase the biological activity of phenolic compounds using the enzyme laccase [17]. These enzymes (benzenediol: oxygen oxidoreductase, EC 1.10.3.2) belong to the oxidase group, and they are also used for cleaner industrial application [18]. Laccases are also known as multicopper oxidases, they belong to the family of copper-containing phenol oxidases [19] and can oxidize a diversity of compounds, e.g., phenolic and nonphenolic compounds [18]. Aromatic compounds can produce reactive radical intermediates, which undergo self-coupling reactions, thus forming different dimers and trimers [20][21][22][23][24]. This enzyme has been previously used to improve the activity of antibiotics [25,26]. On the other hand, the synthesis of a heterodimeric compound (2,6-dimethoxy-4-(phenylimino)cyclohexa-2,5-dienone) by the laccase-mediated coupling reaction between syringic acid and aniline was reported, this compound showed an antifungal effect against B. cinerea with an EC 50 value of 0.14 mM [27].
This work aimed to determine the antifungal activity against B. cinerea and the effect on the cell wall integrity of ten 2,6-dimethoxy-4-(phenylimino)cyclohexa-2,5-dienone derivatives (3a-j) obtained by reaction of syringic acid (1) with substituted anilines (2a-j). To analyze the effect of the carboxylic group in these laccase-catalyzed reactions, syringaldehyde was used instead syringic acid and the reaction product was characterized.
Molecules 2019, 24 FOR PEER REVIEW 2 Several phenolic metabolites found in grape pomace have shown low antifungal activity against B. cinerea [16], therefore, it is possible to increase the biological activity of phenolic compounds using the enzyme laccase [17]. These enzymes (benzenediol: oxygen oxidoreductase, EC 1.10.3.2) belong to the oxidase group, and they are also used for cleaner industrial application [18]. Laccases are also known as multicopper oxidases, they belong to the family of copper-containing phenol oxidases [19] and can oxidize a diversity of compounds, e.g., phenolic and nonphenolic compounds [18]. Aromatic compounds can produce reactive radical intermediates, which undergo self-coupling reactions, thus forming different dimers and trimers [20][21][22][23][24]. This enzyme has been previously used to improve the activity of antibiotics [25,26]. On the other hand, the synthesis of a heterodimeric compound (2,6-dimethoxy-4-(phenylimino)cyclohexa-2,5-dienone) by the laccase-mediated coupling reaction between syringic acid and aniline was reported, this compound showed an antifungal effect against B. cinerea with an EC50 value of 0.14 mM [27].
This work aimed to determine the antifungal activity against B. cinerea and the effect on the cell wall integrity of ten 2,6-dimethoxy-4-(phenylimino)cyclohexa-2,5-dienone derivatives (3a-j) obtained by reaction of syringic acid (1) with substituted anilines (2a-j). To analyze the effect of the carboxylic group in these laccase-catalyzed reactions, syringaldehyde was used instead syringic acid and the reaction product was characterized.

Laccase-Mediated Synthesis of 2,6-Dimethoxy-4-(phenylimino)cyclohexa-2,5-dienone Derivatives
In this work, laccase catalyzed reactions using 1 and 2a-j were carried out. It has been previously reported that using laccases from different fungal sources (Trametes sp. and Rhizoctonia praticola), catalyze reactions between phenolic compounds and anilines, heterodimeric compounds are formed [25,26,31,32], similarly found in this work (Scheme 1). To determine the reaction yields in the formation of the products, different substrate ratios were analyzed. Excluding reactions 1, 7 and 10, most reactions reached higher yields using ratio 1:1 (syringic acid:aniline) (Table 1). Moreover, when aniline was used as substrate, the same result was reported [27], indicating that the increase of the concentration of one of them decrease the yield of the obtained compounds.

Compounds
Highest yields were obtained using 3-chloroaniline and 3,5-dichloroaniline as substrates (2b and 2h) ( Table 1). This high yield could be explained because the oxidation by laccase (from Trametes versicolor) of 3-chloroaniline does not occur [33]. On the other hand, using methoxyanilines (2c and 2d) low yields were obtained, due to a high amount of side products (data not shown).
On the other hand, yield did not increase when the enzyme concentration was increased (data not shown). Bollag et al. [31] showed that the prolonged incubation or higher enzyme amounts caused further polymerization reaction decreasing cross-coupling formation. Furthermore, Itoh et al. [34] concluded that reactivity of laccase mediated reaction between phenolic acids and chlorophenols is due to the substrate specificity of the laccase rather than the chemical property of the substrates, which could explain the lack of relations among electron donating and withdrawing groups and yield of the reactions. The ten synthetized compounds were purified using semipreparative chromatography and were identified ( Figure 1) using 1 H-NMR and 13 C-NMR spectra and mass spectrometry ( Figure S1-S30, Supplementary Materials). Compound 3b showed two aliphatic proton signals (δ 3.670 (s, 3H, H8) and δ 3.874 (s, 3H, H7)) and two aliphatic carbon signals (δ 56.212 and δ 56.321) that determined the presence of two methoxy moieties. Two olefin hydrogen signals at higher fields (δ 6.010 (d, 1H, H3 J = 1.9 Hz) and δ 6.368 (d, 1H, H5 J = 1.9 Hz)), the olefin carbon signals (δ 98.583 (C3) and δ 111.717 (C5)) and one carbon signal at δ 176.633 (C1) indicated the quinonoid character of the products. Table 2 presents the NMR data (chemical shift assignments for short and long-range heteronuclear coupling) of compound 3b.
The spectra of compounds 3c-j ( Figures S8, S9, S11, S12, S14, S15, S17, S18, S20, S21, S23, S24, S26, S27, S29 and S30) only showed differences in the aromatic region; the assignment of the 1 H and 13 C-NMR spectra can be found in the spectroscopic data section (3.4). Figure 1 shows the structures of the ten synthetized compounds in this work. To our knowledge compounds 3a and 3f were previously synthesized [31], 1 H-NMR spectra for compounds 3a and 3f (spectroscopic data Section 3.4) have the same number of signals and comparable chemical shifts and coupling constants like those found by Bollag et al. [31]; furthermore, the mass spectra of 3a and 3f showed a base peak with m/z 277 and 311, respectively, corresponding to the molecular ions, equivalent to the previously described data [31]. Therefore, the other eight compounds (3b, 3c, 3d, 3e, 3g, 3h, 3i and 3j) have not been previously reported.  The spectra of compounds 3c-j ( Figures S8, S9, S11, S12, S14, S15, S17, S18, S20, S21, S23, S24, S26, S27, S29 and S30) only showed differences in the aromatic region; the assignment of the 1 H and 13 C-NMR spectra can be found in the spectroscopic data section (Section 3.4). Figure 1 shows the structures of the ten synthetized compounds in this work. To our knowledge compounds 3a and 3f were previously synthesized [31], 1 H-NMR spectra for compounds 3a and 3f (spectroscopic data Section 3.4) have the same number of signals and comparable chemical shifts and coupling constants like those found by Bollag et al. [31]; furthermore, the mass spectra of 3a and 3f showed a base peak with m/z 277 and 311, respectively, corresponding to the molecular ions, equivalent to the previously described data [31]. Therefore, the other eight compounds (3b, 3c, 3d, 3e, 3g, 3h, 3i and 3j) have not been previously reported.
Interesting, compound 3a was also obtained using syringaldehyde instead of syringic acid in the reaction with 4-chloroaniline. This could be explained with an extra step when using syringaldehyde, an oxidation of the aldehyde to a carboxylic acid (syringic acid), similar oxidations has been previously described using several aromatic aldehydes with laccase, yielding carboxylic acids [35]. Hence, syringaldehyde is oxidized to syringic acid and then the same product (compound 3a) could be found in both reactions, starting with syringaldehyde or with syringic acid. However, this synthesis had a very low yield (data not shown).

Antifungal Activity
Antifungal activity of compounds 1 and 2a-j and compounds 3a-j against B. cinerea was measured on mycelial growth in solid media and the EC 50 were calculated using the mycelial growth (Tables 3 and 4). The most active compound was the 3,5-dichloro-substituted product (compound 3h), while compound 3c had the lowest activity. It has been reported that the substituent affects the antifungal activity of a molecule [36], for instance, the position of the chlorine atom in the aromatic ring is important for the antifungal activity against B. cinerea since para-substituted compound (3a) and ortho-substituted compound (3j) were more active than the meta-substituted compound and unsubstituted compound (EC 50 = 0.14 ± 0.02) [27], while activity of meta-substituted compound (compound 3b) and unsubstituted compound are similar [27]. The number of chlorine atoms in the aromatic ring is also important, both dichlorinated compounds 3f and 3h showed higher antifungal activity than mono chlorinated compounds 3a, 3b, and 3j, however, dichlorinated compound 3i showed an antifungal activity comparable to the monochlorinated compounds, therefore, the number and position of chlorine atoms in the aromatic ring seems to be important for the antifungal activity of these compounds.  Furthermore, the methoxy derivative compounds (3c and 3d) were less active against the fungus than the other compounds, even the nonsubstituted compound 3 [27], the same effect was observed for aspirin derivatives, where the methoxy para-substituted derivative showed almost 30% less antifungal activity against B. cinerea than the chlorinated para-substituted compound [37]. Similar behavior was previously reported for oxadiazole derivatives when tested the activity of the methoxy meta-substituted oxadiazole derivative against B. cinerea, and its activity was less than half compared to the nonsubstituted compound [38]. Usually, the chloro-substituted compounds have higher antifungal activity in commercial fungicides, for example, chlorine compounds like boscalid, chlorothalonil, and iprodione have been used to control B. cinerea [6]. The antifungal activity of iprodione has been tested against this strain of B. cinerea, showing an EC 50 of 0.015 ± 0.003 mM [27], this antifungal activity is in the same order of magnitude than the most active compound obtained in this work (3h). Additionally, p-nitro and p-trifluoromethyl compounds (3e and 3g) were tested against this fungus, 3e showed no antifungal activity, probably because of the low solubility of this molecule, for this reason 3e was not used in further assays. Compound 3g only showed an intermediate antifungal activity compared to the rest of the synthetized molecules in this work. Lastly, most of the substrates used in the reactions (i. e., 1 and 2a-j) showed lower antifungal activity than the products (Table 4), only 2f was more active than 3f.

Effect on the Cell Wall Integrity of B. cinerea
To analyze the effect of the compounds on the cell wall integrity, the dye calcofluor white (CFW) was used. This dye binds to β-1,3 and β-1,4 polysaccharides, for example chitin, which is a primary component of the cell wall in fungi, and fluorescence of the hyphae can be detected [39]. Figure 2 shows the effect of compound 3a on the cell wall of B. cinerea. Treatment with this compound showed lower fluorescence intensity than the negative control (acetone), indicating that this compound can damage the cell wall of this fungus. The same assay was performed using compounds 3b-j. The ten synthesized compounds caused a decrease of the fluorescence intensity compared to the control; relative fluorescence intensity is observed in Figure 3. This result could be attributed to the toxicity of quinones, which could be connected to the production of reactive oxygen species (ROS) which cause oxidation of cell molecules [40]. Quinone derivative N-acetyl-p-benzoquinone imine (NAPQI) can react with nucleophiles such as thiol groups of proteins or glutathione [41,42]; this last molecule is an important antioxidant molecule in fungi [43]. On the other hand, some aromatic antifungal compounds have shown effects on cell wall synthesis [6] by inhibiting chitin and glucan synthases [44], enzymes that catalyze the synthesis of the main polymers of the cell wall in fungi.    Fluorecense relative intensity (%) Figure 3. Effect of the compounds on the cell wall integrity of B. cinerea. The effect was measured as relative fluorescence intensity compared to maximum fluorescence (negative control).

General Experimental Procedures
The NMR spectra of 3a-j were acquired using a Bruker Avance 400 MHz spectrometer (Bruker, Billerica, MA, USA) (400,133 MHz for 1 H, 100.624 MHz for 13 C). Measurements were done in CDCl 3 at 27 • C. Chemical shifts were calibrated to solvent signal: CHCl 3 7.26 ppm (residual signal solvent) and 77.16 ppm for 1 H and 13 C, respectively, and informed relative to Me 4 Si. Thin-layer chromatography was done with a Merck Kiesegel 60 F 254 , 0.2 mm thick and semipreparative thin layer chromatography on Merck Kieselgel 60 F 254 0.25 mm thick. A Thermo Scientific GC-MS system (GC: model: Trace 1300 and MS: model TSQ8000Evo) (Waltham, Massachusetts, USA) was used to analyze the sample. The separation was performed on a 60 m × 0.25 mm internal diameter fused silica capillary column coated with 0.25 µm film Rtx-5MS. The oven temperature was maintained at 40 • C for 5 min, then it was programmed from 40 to 80 at 5 • C/min for 1 min, then from 80 to 300 at 10 • C/min and finally maintained at 300 • C for 10 min. The mode used was splitless injection, helium was used as carrier gas, and flow-rate was 1.2 mL/min. Mass spectra were recorded over a range of 40 to 400 atomic mass units at 0.2 s/scan. Solvent cut time was 11 min. Ionization energy was 70 eV.
For the first reaction, syringic acid (1) and 4-chloroaniline (2a) were dissolved in 1 mL ethyl acetate and laccase was dissolved in 1 mL sodium acetate buffer (20 mM, pH 4.5). Both solutions were mixed and stirred at 100 rpm for 180 min at 22 • C. Afterwards, the solvent was evaporated at 40 • C using a rotary evaporator. The synthetized compounds were purified by using semipreparative thin layer chromatography with hexane: ethyl acetate (1:1) as an eluent system. Same procedure was carried out using a different substituted aniline (2b-j).
Alternatively, compound 3a was also found when using syringaldehyde and 4-chloroaniline under the same conditions described above.

Effect on Mycelial Growth
The antifungal activity of the compounds was evaluated in vitro as described by Caruso et al. [13]. Compounds were dissolved in acetone and then added to Petri dishes along with malt yeast agar medium. Inhibition percentages were calculated after 72 h of incubation. Antifungal activity was expressed as the concentration that reduced mycelial growth by 50% (EC 50 ), calculated by regressing the antifungal activity percentage against compound concentration. These experiments were done at least in triplicate.

Effect on the Cell Wall Integrity of B. cinerea
The effect of compounds 3a-j on cell wall integrity was evaluated using the method described by Mendoza et al. [27]. Compounds 3a-j were tested at 0.16 mM. To measure the effect of these compounds on the cell wall, fluorescence intensity was quantified using ImageJ (v1.80), an outline was drawn around each hypha, and mean fluorescence was measured, along with several adjacent background readings. Mean fluorescence was compared to the negative control (maximum fluorescence).

Conclusions
Ten compounds were synthetized; two of them (compounds 3a and 3f) have been previously described. All the products showed higher antifungal activity than the substrates. Chloro-substituted compounds showed the highest antifungal effect against B. cinerea being the 3,5-dichlorinated product 3h the most active. Synthesis using syringic acid or syringaldehyde with p-chloroaniline yield the same main product (compound 3a). Finally, regarding the inhibition mechanism of these compounds, the results suggest that these compounds damage the cell wall.
Funding: This research was funded by FONDECYT, grant number 3170478, and by Universidad de Santiago de Chile DICYT 021640VG_POSTDOC.