Botrytis cinerea strain PPRI 7338 was acquired from the Agricultural Research Council (ARC) - Plant Health and Protection Institute in Pretoria, Mycology Division, South Africa. Strain PPRI 7338 was isolated from infected apples. Potato dextrose agar (PDA, NutriSelect^®Plus, Merck (Pty) Ltd., Johannesburg, South Africa) was used to culture the inoculum of B. cinerea PPRI 7338, and the plates were stored at 20°C for 5 days. Conidial stock suspensions were prepared and conserved as reported by Vallejo et al. (2002) with few modifications. Six mycelial plugs (0.432 cm²) from 5-day-old B. cinerea cultures maintained on PDA media were transferred to flasks (250 mL) with 100 mL of minimal salts medium (Supplementary Table 1) augmented with 1% Carboxymethylcellulose (CMC).

Conidial suspensions of B. cinerea were cultivated for 7 days at 20°C and agitated at 180 rpm using Orbital Shaker (OrbiShake, Labotec, Cape Town, South Africa). B. cinerea cultures supplemented with lemon and lemongrass essential oils individually (20 mg/L) and in combination (1:1, 20 mg/L) that were harvested at day 7 were used for enzymatic and proteomic analysis. Using Eppendorf microcentrifuges (5425 R, Stevenage, UK), treated and non-treated B. cinerea cultures were centrifuged for 10 min at 13 000 x g, and the mycelia stored at −80°C before other downstream analysis were conducted. This experiment was independently replicated in triplicate and three times (n = 9).

Mycelia harvested from treated and non-treated B. cinerea were grounded using liquid nitrogen into fine particle size. Mycelia (0.2 g) from each treatment was homogenized in 1 mL of 6% (w/v) trichloroacetic acid (TCA) for analysis of H₂O₂ content, and lipid peroxidation. For the measurement and detection of superoxide dismutase (SOD) and ascorbate peroxidase (APX) enzymatic activities, 1 mL of homogenizing PVP extraction buffer (40 mM K₂HPO₄ at pH 7.4; 1 mM EDTA; 5% PVP MW = 40,000; 5% glycerol in distilled H₂O). Concentrations of protein for each sample was quantified using the reducing agent and detergent compatible (RC DC™) Protein Assay Kit II (Bio-Rad Laboratories Ltd., Rosebank, Johannesburg, South Africa).

Freshly harvested mycelia (0.1 g) from each treatment were ground to a fine powder in liquid nitrogen and 0.5 g PVPP. Total soluble proteins were isolated using a method by Röhrig et al. (2006) with slight modifications previously described by Wang et al. (2006). Mycelia powder was homogenized in 10% (w/v) TCA/Acetone and centrifuged at 13 000 x g for 6 min. The obtained pellets were washed with 80% (v/v) methanol containing 0.1 M ammonium acetate, after the supernatants were discarded. This was followed by acetone wash with 80% (v/v). The pellets were allowed to air dry at room temperature for an hour and re-suspended in a 1:1 ratio of phenol and SDS buffer (Supplementary Table 1).

The re-suspended pellets were placed on ice for 6 min, and centrifuged at 13,000 x g for 6 min. The upper phenol phase of each sample was transferred to sterile 2 mL Eppendorf tubes and filled with the 80% (v/v) methanol containing 0.1 M ammonium acetate to allow for protein precipitation overnight at −20°C. The precipitates were centrifuged at 13,000 x g for 6 min and the supernatant discarded. The pellets were washed with 100% (v/v) methanol followed by an additional wash with 80% (v/v) acetone. Pellets were air-dried at room temperature for 1 h and re-suspended in IEF buffer (Supplementary Table 1) for One dimensional polyacrylamide gel electrophoresis (1 SDS-PAGE) analysis.

Cold-pressed extracts of lemon (Citrus limon, 1001714241), lemongrass (Cymbopogon citratus, 101400274) essential oils were obtained from Sigma-Adrich (St. Louis, MO, USA). The EOs were kept refrigerated (4°C) prior to gas chromatography-mass spectrometry (GC–MS) analysis. Each EO was diluted separately in hexane (1:10,000), and 1 μL of the mixture was thereafter injected into the column with a split ratio of 10:1. Chemical constituents of each EO were determined using Agilent 7890A gas chromatography (GC) system coupled with Agilent 5975C mass selective detector MSD (Agilent, Palo Alto, CA) and both equipped with a HP-5MS column (30 m × 0.25 mm i.d., 0.5 μm film thickness).

Helium at a constant flow rate of 1.3 mL/min was the carrier gas used in this measurement. Oven condition was set as follows: 50°C for 3 min, then ramped up to 80°C at the rate of 4°C/min for 3.5 min; and, finally at a rate of 10°C/min temperature was ramped up to 250°C, and held for 6 min. Mass spectra were analyzed in the scan mode over the range of 35 to 600 m/z. Compounds were identified qualitatively by their retention times (RT), retention index (RI) values and mass spectra names obtained from NIST MS library search and only match greater than 90% were accepted. Chemical constituents were identified based on individual external standards and the percentage composition determined under same chromatographic conditions mentioned above. All measurements were conducted in triplicate (n = 3).

The H₂O₂ content of the mycelia was measured as described by Velikova et al. (2000). The reaction mixture consisted of 50 μL TCA extract, 5 mM dipotassium phosphate (K₂HPO₄, pH 5.0) and 0.5 M potassium iodide in a final volume of 200 μL. Sample reactions were incubated in the dark at room temperature (25°C) for 20 min. Thereafter, absorbance values were measured using FLUOstar Omega UV-visible spectrophotometer (BMG LabTech GmbH, Ortenberg, Germany) at 390 nm. Using H₂O₂ standard curve approach based on known concentrations at absorbance (A390 nm), the B. cinerea mycelia H₂O₂ content of was calculated and expressed as nmol g^–1 fresh weight (FW) basis.

Malondialdehyde (MDA) is a known biomarker for oxidative stress and end-product of lipid peroxidation (reflective as MDA content). With slight modifications to the thiobarbituric acid (TBA) method reported by Egbichi et al. (2013), MDA content of each sample was measured. For this study, TBA extract (200 μL) was added to 400 μL of cold 20% (w/v) TCA. The mixture was homogenized and centrifuged under refrigerated condition (4°C) at 12,000 rpm for 30 min using Eppendorf microcentrifuges (5425 R, Stevenage, UK). An aliquot portion of the supernatant (100 μL) was mixed with 400 μL of 0.5% TBA (prepared in 20% TCA). This assay was then incubated in a water bath at 95°C for 20 min, thereafter, the reaction was cooled on ice for 5 min, and further centrifuged under refrigerated condition (4°C) at 12,000 rpm for 5 min. To correct for non-specific turbidity in the samples, absorbance of the supernatant was measured at 532 nm and 600 nm. The MDA content was calculated by using an extinction coefficient (ε = 155 mM^–1cm^–1) and reported as nmol g^–1 FW.

Ascorbate peroxidase (APX) activity was evaluated based on a modified approach reported by Asada (1992). Each reaction (with final volume of 200 μL) was incorporated with enzyme extract (10 μL) and topped-up with APX reaction buffer (180 μL) (Supplementary Table 1). The reaction was initiated with the addition of 10 μL H₂O₂ and the enzymatic activity was measured at 290 nm absorbance. To calculated APX activity, the extinction coefficient (ε = 2.8 mM^–1 cm^–1) was use, and results presented as mmol μg^–1.

To quantify superoxide dismutase activity, a modified method by Beyer and Fridovich (1987) was used. The reaction of the final mixture volume of 200 μL; containing enzyme extract (10 μL) and SOD (190 μL) reaction buffer (Supplementary Table 1) was initiated by adding riboflavin (2 μM), which was followed by exposure to light for 20 min or the observation of a change in color. Changes in the mixture complex by the addition of riboflavin was measured via absorbance readings at 560 nm. Under standard assay conditions, one unit of SOD activity represents the volume of enzyme required to reduce 50% of NBT to blue formazan.

A fraction from each protein pellet (10 μg) was denatured and fragmented using 12% SDS-PAGE electrophoresis gel ran for 90 min at 120 V. At the end of the electrophoresis run, the gel was stained for 30 min using Coomassie Brilliant Blue G-250, thereafter, de-stained for 2 h using 10% glacial acetic acid and 1% glycerol. The fractioned/fragmented protein bands on the gels were viewed and captured using the ENDUROTM GDS Gel Documentation System (Labnet International, Edison, NJ).

The left-over fraction of protein pellets was solubilized in solution A (100 mM Tris, 1% TritonX-100 and 4 M guanidine hydrochloride) followed by three cycles of 1-min sonication. After sonication, solution A was decanted and replaced with 1% formic acid (FA) in a new sterile tube (2 mL), and a one-minute sonication cycle was repeated three times. At the end of the repeated 1-min sonication step, FA was decanted and combined with solution A to perform a methanol-chloroform liquid-liquid analysis. Mixture of solution A-FA was blended with methanol (4 volumes). To the solution A-FA-methanol mixture was thoroughly mixed with 1 volume of chloroform. To facilitate partitioning of the mixture, three volumes of water were added, and the solution was centrifuged at 13,000 x g for 5 min. Four volumes of methanol were added after the upper phase of the solution A-FA-methanol-chloroform mixture was discarded. The reaction mixture was centrifuged at 13,000 x g for 5 min and the methanol-chloroform phase separated from the solution. The protein pellets were air-dried at room temperature and re-constituted in 50 mM Tris buffer containing 2% SDS and 4 M urea.

Solubilized proteins were re-constituted in triethyl ammonium bicarbonate (TEAB, 50 mM; Fluka™, Honeywell International Inc., Charlotte, North Carolina, US), followed by the addition of triscarboxyethyl phosphine (TCEP, 5 mM; Fluka™, Honeywell International Inc., Charlotte, North Carolina, US) and 100 mM TEAB. This mixture was stored under agitation for 1 h at room temperature. Cystein residues were thiomethylated for 30 min at room temperature, using 20 mM S-Methyl methanethiosulfonate (Sigma Aldrich, Johannesburg, South Africa) in 50 mM TEAB. After thiomethylation two-fold dilution of the samples was conducted using binding buffer (100 mM Ammonium acetate, 30% acetonitrile, pH 4.5).

According to manufacturer’s instructions the protein solution was added to MagResyn HILIC magnetic particles (Resyn Biosciences (Pty), Ltd., Gauteng, South Africa), and the mixture was incubated at 4°C overnight. After overnight incubation and binding, supernatant from the mixture was decanted and the magnetic particles were washed with washing buffer (100 mM Ammonium acetate, 15% acetonitrile, pH 4.5) twice. The washed magnetic particles were added into TEAB (100 mM) containing trypsin (New England Biolabs^®, Ipswish, UK) to a final ratio of 1:20 and incubated for 4 h at 37°C. At the end of the incubation period, peptides were extracted using water (50 μL) once followed by 50% acetonitrile (ACN). Samples obtained were dried down and re-constituted with 2% ACN:water and 0.1% FA (30 μL). Residue of the reagents used in the digestion step were removed using an in-house manufactured C₁₈ stage tip (Empore Octadecyl C₁₈ extraction discs; Supelco). The samples were loaded onto the stage tip after activating the C₁₈ membrane with methanol (30 μL) and equilibrated with 30 μL of 2% ACN: water and 0.05% trifluoroacetic acid (TFA). Before eluting with 30 μL 50% ACN:water and 0.05% TFA, the bound sample was washed with 30 μL of 2% ACN: water and 0.1% TFA. The eluate was dried and the dried up peptides were dissolved in 2% ACN:water; 0.1% FA for LC-MS analysis.

To fractionate the peptides before mass spectrometry analysis, a high-pressure liquid chromatography system running at dionex nano-flow rate was used. The procedure for LC-MS/MS analysis was adapted from Yasmeen et al. (2016). Liquid chromatography was performed using a Thermo Scientific Ultimate 3000 RSLC equipped with a 5 mm x 300 μm C₁₈ trap column (Thermo Scientific, USA) and an Asentic Express 15 cm x 75 μm of 2.7 μm size C₁₈ analytical column (Supelco). The solvent system used for loading were: 2% ACN:water; 0.1% FA; Solvent A: 2% ACN:water; 0.1% FA and Solvent B: 100% ACN: water. Using the loading solvent at a flow rate of 10 μL/min, samples were loaded onto the trap columns via a controlled autosampler set at 7°C. Loading was performed for 4 min before the sample was eluted onto the analytical column, and the flow rate was maintained at 350 nl/min. The following gradient were generated: 5-35% solvent B over 60 min; 35 - 50% solvent B from 60 to 80 min using Chromeleon non-linear gradient. Chromatography was performed at 40°C and the outflow delivered to the mass spectrometer through a stainless-steel nano-bore emitter.

Thermo Scientific Fusion mass spectrometer equipped with a Nanospray Flex ionization source was used for detection, and data were collected in a positive mode with spray voltage set to 1.8 kV and ion transfer capillary set to 280°C. Polysiloxane ions at m/z = 445.12003 and 371.10024 were used as internal calibration. MS1 scans were performed using the orbitrap detector set at 60,000 resolutions over the scan range 350-1,650 with AGC target at 5 E4 and maximum injection time of 40 min. Data was acquired in profile mode. MS2 acquisitions were performed using monoisotopic precursor selection for ion with charges + 2 to + 7 with error tolerance set to ± 10 ppm. Precursor ions were excluded from fragmentation once for a period of 60 s. Precursor ions were selected using the quadrupole mass analyzer at higher energy dissociation (HCD) mode with the HCD set to 30%. Fragment ions were detected in the orbitrap mass analyzer set to 15,000 resolutions. The AGC target and the maximum injection time were set to 5 E4 and 30 min, respectively. The data were acquired in centroid mode. The MS proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (Perez-Riverol et al., 2022) partner repository with the dataset identifier PXD038894 and 10.6019/PXD038894.

MS-generated files were transferred into Proteome Discoverer software (vr. 1.4, Thermo Scientific, USA). The files were processed using both Sequest and Amanda algorithms. Database interrogation was performed against a concatenated database created using the Uniprot Botrytis database concatenated with the cRAP contaminant protein database. Two missed cleavages were set as the allowed semi-tryptic cleavage, while the precursor mass tolerance and fragment mass tolerance was set to 0.05 Da and 10 ppm, respectively. Deamidation, oxidation, and acetylation of protein N-terminal was allowed as dynamic modifications and thiomethylation of C as static modification. Peptide validation was performed using the Target-Decoy PSM validator node set to search against a decoy database with strict FDR 1% and delta Cn of 0.1. Obtained results were transferred to Scaffold 1.4.4 and identified peptides validated using the X!Tandem, peptide and protein prophet search algorithm included in Scaffold. Peptide and Protein Prophet algorithms were used for peptide and protein validation, and the quantitation of protein was performed using Fischer’s Exact Test on the paired data with the Benjamini-Hochberg correction applied. Identified proteins were only considered acceptable if they were confirmed at greater than 95% probability and contained at least two unique identified peptides.

Experimental design used in this study was full factorial. Besides VOC analysis, all assays and proteomics experiment were independently replicated in triplicate and three times (n = 9). All statistical analyses for the enzymatic assays were performed using statistical software SAS (vr. 9.2, SAS Institute Inc., Cary, NC, USA). To test for normality, Shapiro-Wilk test was performed. Data obtained were analyzed using One-Way analysis of variance (ANOVA) at p ≤ 0.05, and mean values were tested according to Tukey’s multiple comparison test at p ≤ 0.05, with F-values indicated significant at p ≤ 0.01. An interactive tool, Venny 2.1.0, BioinfoGP was used for comparing lists of Venn’s diagrams (Oliveros, 2007-2015).

Based on GC–MS analysis, 17 volatile compounds were identified across the selected EOs. Lemon (Le) and lemongrass (Lg) EOs consisting of 7 and 10, respectively (Table 1). The major chemical constituent for Lg EO includes α-citral (50.1%), β-citral (35%), geraniol (4%), and geranyl acetate (3%), while other components (camphene, δ-limonene, linalool, α-terpineol, caryophyllene, and γ-cadinene) were 1% each. In contrast, the chemical constituents of Le EO were δ-limonene (71%), β-pinene (16%), and γ-terpinene (7%), with monoterpene hydrocarbons being the most abundance chemical class with 97% and oxygenated monoterpenes with 1.8%. Cumulatively for Lg EO, the oxygenated monoterpenes were the most abundant class (94%), while both monoterpene hydrocarbons and sesquiterpene hydrocarbon were 2%, respectively (Table 1).

Most abundant compounds for Le and Lg EOs in this study differ from those reported in other related studies (Table 1). For example, Lg EO was associated with the abundance of geranial, neral, geraniol, myrcene, and linalool (Munhuweyi et al., 2017; Moutassem et al., 2019). The discrepancies in relative percentage composition of VOCs in EOs could be associated with variations in the plant material parts extracted, the method of extraction and harvested location or agro-climatic regions (Kgang et al., 2022). Similarly, the composition of EOs identified in this study could play a crucial role in their antifungal efficacy (Shah and Mehta, 2018).

Hydrogen peroxide is an important biomarker for the membrane lipids peroxidation, and it has been demonstrated to have a direct antifungal effect (Qin et al., 2011; Wu et al., 2022). Results showed that exposure of B. cinerea to Le, Lg and Le + Lg EOs significantly reduced H₂O₂ content in the fungal mycelia (p < 0.005) at the end incubation day 7 (Figure 1A). The reduction in H₂O₂ content was most pronounced in the binary mixture involving the two essential oils (Le + Lg). In comparison to the untreated (control), H₂O₂ content was reduced by 21%, 58% and 67% under Le, Lg and (Le + Lg) combined treatments, respectively (Figure 1A).

The observation from this study is consistent with literature on the response of B. cinerea to EOs and other botanical fungicides (Yang et al., 2020a; Wu et al., 2022). For instance, Yang et al. (2020a) demonstrated that treatment of B. cinerea with cembratrien-diols induced oxidative stress, which resulted in injured membrane system with a significant reduction of H₂O₂ content in the mycelia. Similarly, H₂O₂ content was found to be significantly lower in B. cinerea mycelia exposed to juniper EO compared to the untreated control during 12 h incubation period (Wu et al., 2022). The balance between reactive oxygen species (ROS) production and scavenging capacity is crucial for any typical cellular processes. Over production of ROS (such H₂O₂) or inadequate antioxidant or scavenging enzyme activity could disrupt the cellular balance, which induces oxidative stress (Li et al., 2017; Yang et al., 2020b).

Both EOs (lemon and lemongrass) displayed abilities to induce scavenging as observed by the reduction of H₂O₂. The scavenging abilities of the essential oils may be ascribed to their chemical composition. Pino et al. (2018) reported the chemical composition of lemongrass whereby oxygenated monoterpenes and monoterpene hydrocarbons were the most represented class of compounds identified. According to Burits et al. (2001), monoterpene compounds identified in EOs may act as radical scavengers and EOs, which contain terpenes have greater antioxidative properties. As shown in this study in Figure 1, the content of H₂O₂ (one of the measured ROS) in EO treated B. cinerea was reduced in this study, although SOD was significantly higher in EOs treated B. cinerea mycelia compared to the control. In future studies, to fully understand the activity SOD, the level of total ROS should be determined. Overall, the results obtained suggests that the imbalance created by EOs, led to oxidative stress, and could be an important mechanism of antifungal activity.

Malondialdehyde is regarded as a end product of lipid peroxidation, and its excessive accumulation has been considered as a marker for oxidative stress (Mittler, 2002). The current study, showed that there were significant changes in the MDA content of essential oils treated B. cinerea when compared to the control (Figure 1B). Highest accumulation of MDA content was found in Lg-EO treated B. cinerea. The MDA content significantly increased in the Le-treatment (79%), Lg-treatment (278%), and in the combined Le + Lg (205%) in comparison to the untreated control at the end of day 7 (Figure 1B). Consistent with these current results, Khan et al. (2011) showed a 3-fold rise in MDA content in Candida albicans treated with eugenol, methyl eugenol and estragole, which are phenylpropanoid compounds found in EO.

Lipid peroxidation of cellular membrane is a complex process involving majorly, polyunsaturated fatty acids containing one or more methylene groups. These methylene groups are very sensitive to oxidizing agents resulting in the formation of peroxyl radicals, which can further cascade a sequence of reaction producing more free radicals and the accumulation of peroxidation by-products such as MDA (Avis et al., 2007). MDA can interact with DNA to form a propane adduct with 2’-deoxyguanosine. This complex has a significant impact on cell physiological activities this includes signaling, proliferation, differentiation, and death (Vasilaki and McMillan, 2011). Furthermore, it is reasonable to envisage the probable mode of action against B. cinerea involves oxidative stress as an antifungal activity of the EOs and their related compounds as shown in previous study. This activity could be associated with the activities of free radical scavenging agents, metal chelating and modification of cell signaling pathways (Khan et al., 2011). The increase in MDA content observed in our study in response to lemon and lemongrass EOs may thus be because of the EO constituents that bind to key enzymes on the membrane, causing membrane damage. This causes cell distortion and the loss of macromolecules from their interior which results in improper functioning of the cell membrane (Pramila et al., 2012). Therefore, EOs may induce damage to cellular protein, lipid, and nucleic acid content (Mani-López et al., 2021). Furthermore, each component of an EO contributes to the oil’s biological activity in its own way.

The major chemical compositions of EOs are terpenes, aldehydes, and phenols (Boubaker et al., 2016; Sharma and Kaur, 2022). Research have reported that EO’s with monoterpenes and sesquiterpenes exhibit antifungal activities against several fungi (Tian et al., 2015). Terpenes have been reported to induce membrane disruption which changes cell membrane permeability and fluidity and induces disturbances in membrane functions (Di Pasqua et al., 2007). Niu et al. (2022) also reported that the induction of MDA content of A. niger was associated with the inhibition of the ergosterol biosynthetic pathway, which led to plasma membrane damage and leakage of intracellular contents. In addition, ergosterol content was shown to be reduced in various fungal species treated with EOs (Mani-López et al., 2021). Ergosterol plays a role in regulating the activities of membrane-bound enzymes, membrane fluidity and permeability, as well as material transit by binding to phospholipids and stabilizing membrane structure (Krumpe et al., 2012). Tang et al. (2018) demonstrated that the geraniol and citral, which are major components of Lg EO effectively suppressed the growth of A. flavus and A. ochraceus by altering cell membrane permeability, causing electrolyte and cellular contents to leak, and interfered with key associated genes. These compounds are abundant in the Lg EO investigated in this study (Table 1).

In this study, total SOD activity of B. cinerea was assayed in response to the different EOs (Figure 1C). Given that SOD enzymatic activity catalyzes the conversion of O₂^– to produce H₂O₂, it is expected that changes in H₂O₂ content could be attributed to alteration in SOD activity (Figure 1C). In response to the reduction in H₂O₂ content observed in Figure 1A, this study suggests the role of EOs in modulating B. cinerea SOD activity. Both Le- and Lg- EOs increased SOD activity in B. cinerea by 46% and 137% respectively, and highest increase in SOD activity (463%) was observed in the combination Le + Lg EO treatment when compared to the control non-treated B. cinerea.

Superoxide dismutase (SOD) is one of the key enzymes responsible for defense against oxidative stress (Yaseen et al., 2014; Valenzuela-Cota et al., 2019). Antioxidants operate as free radical scavengers and prevent free radical mediated processes. Khan et al. (2011) reported that the presence of EO components elevated the levels of SOD activity in C. albicans. Therefore, it is possible that the increase in SOD activity presented in this study may be attributed to the phenolic content that these EOs display. Based on the increased SOD activity in response to essential oils, it is reasonable to assume that H₂O₂ content would increase, since SOD can directly scavenge O₂^– to produce H₂O₂. However, the current study showed that elevated levels of SOD activity was not associated with an increase in H₂O₂ content (Figure 1A).

Since EOs used in this study augmented SOD activity coupled with a reduction in H₂O₂ content, we investigated the effect of EOs on APX activity in B. cinerea. Given that APX catalyzes the conversion of H₂O₂ to water, we hypothesized that changes in APX activity would be highest in Lg treatment based on MDA content. This assumption was confirmed with APX activity in B. cinerea was highest in the Lg treatment with 206% increase followed the combination treatment (Le + Lg), with 73% in compared to the controls (Figure 1D). However, no significant difference in the APX enzymatic activity was observed in the Le treatment in comparison to the untreated control.

This work is the first to report APX activity in B. cinerea in response to Le and Lg EOs. Based on the results obtained from this work, we postulated that long progressive treatment of B. cinerea with Le and Lg EOs will reduce mycelia growth, and result in the development of differentially modulated biochemical responses. Furthermore, this study has shown that Lg EO is an efficient elicitor of B. cinerea response, as indicated by stress-induced oxidative damage (caused by enhanced MDA levels), which ultimately leads to membrane damage and cell death.

Identification of protein via 1D gel electrophoresis has been demonstrated to show the presence of fractionated proteins based on molecular weight, the distance migrated and band intensity. Protein bands from all treatments covered a molecular weight range of 30 to 150 kDa (Figure 2). Separation and visualization of B. cinerea demonstrated that the proteins are of high quality with no visible streaking or protein degradation. Bands of fragmented or fractionated proteins between the two EOs and their combination showed some similarity to the control based on their intensity and banding patterns. However, changes were observed in the protein profiles between treatments (indicated by arrows), with protein bands at ≈30, ≈40 and ≈90 kDa were shown to have higher band intensity suggesting possible over-expression or up regulation of proteins within these bands in the treated B. cinerea. Variation in protein intensity in this study could suggest that the protein expression in the non-treated control was down-regulated (Nsumpi et al., 2020).

The blank injection before the analysis showed no major contaminants judging by the total ion chromatogram (TIC; Supplementary Figure 1A). While the system suitability sample (a four-protein mixture) indicated successful digestion and demonstrated that peptides were both retained and eluted as expected judging by the TIC (Supplementary Figure 1B). As a critical component of bio-analysis was performed based on the data generated for the blank and the system suitability sample. All samples were analyzed and captured as individual TICs (Supplementary Figure 2). While the TICs indicated successful digestion on the system suitability sample (a four-protein mixture) as analyzed on the mass spectrometer, the combination treatment seemed to have less material or was more difficult to digest judging by the number of peaks and signal intensity (Supplementary Figure 2D).

Data base interrogation of the total ion chromatography for all samples against the Uniprot database for Botrytis was performed, and 110 proteins with 7 clusters was identified with an FDR of 0.95% (Figure 3A). To ensure correct identification of proteins, a selection threshold of 95% was set for the identified proteins. Furthermore, the protein exclusive unique peptide count should be ≥ 2 for the protein to be considered as a positive identification; and the identification probability should be ≥ 95%, and the percentage sequence coverage should > 0. Based on these selection and identification parameters, 42 proteins were confidently identified in this study, with 20 proteins identified in the control, 9 in Le-treatment, 35 in Lg-treatment, and 31 in combined treatment (Le + Lg)) as shown in Figure 3A and Table 2.

Protein accessions identified in this study (Table 2) were analyzed with the FunRich Multi analysis software (version 3.1.3), to distinguish proteins there were common and/or unique in each of the treatments (Figure 3B). Results showed that 13 of 42 proteins identified in this study were unique, with 0 – control; 1 – Le; 8 – Lg; and 4 – (Le + Lg) EOs (Figure 3B). On the other hand, 6 proteins (ACT BOTFU, GET3 BOTFB, CYNS BOTFU, RAS BOTFU, IF4A BOTFB and TBB BOTFU) were conserved in all treatments. Seven proteins were shared between untreated control and Le-treatments, while untreated control and Lg-treatment shared 18 proteins. For the comparison of control and combination treatment (Le + Lg) EO, 18 proteins were shared between untreated control and combined (Le + Lg) EOs treated B. cinerea (Figure 3B).

Proteomics plays a major role in the nature and uses of essential oils on fungi. The subcellular localization (Figure 4) helped in further determining key functional characteristic of protein-protein interaction. Our study showed that 72% of proteins identified were localized within the cellular anatomical entity, and 28% of proteins in the protein-containing complex (Figure 4A). Subcellular localization results showed that major proteins were widely distributed within organelles (27%) and cytoplasm (22%). Partitioning of the proteome between organelle and cytoplasm affects nearly every aspect of eukaryotic biology (Emanuelsson, 2002), and performs a wide range of metabolic functions, which play vital role in cellular integrity (Zybailov et al., 2008; Walker et al., 2018). This could be a major reason why most of the identified proteins were predominantly located in organelles and cytoplasm.

In response to chemical stress and defense, cellular integrity is essential for fungal adaptation and survival. The response to chemical stress causes fungal disruption which induces the efflux of cytoplasmic constituents (Liu et al., 2017). Notably, the subcellular location of approximately 16% of the identified proteins was in the membrane and some were identified in the integral component of the membrane (11%). The rest of the proteins were found in the extracellular region (6%), lumen (6%), supramolecular complex (5%), cytosol (3%), cell wall (2%), actin cortical patch (2%) and host intracellular parts (2%) (Figure 4B). Proteins localized in the protein-containing complex (Figure 4A) were further sub-localized in the ribonucleoprotein complex (43%), initiation factor 3 complex (25%) and the remaining proteins each had 8% belonging to its subclass or group (Figure 4C).

The subcellular localization (Figure 4) helped in further determining key functional characteristics of protein-protein interaction (Figures 4D–E). Since most proteins interact with other proteins for proper function, it is important that they should be investigated in concert (within the context of their interactions) to fully understand their biological functions. Similarly, to successfully assign functions to unknown proteins the complete interaction should be understood (Xenarios and Eisenberg, 2001; Kim et al., 2002). GO enrichment analysis was used to understand protein functions. The identified proteins were analyzed by GO annotation via the Uniprot database (http://www.uniprot.org) and categorized based on functionality into 11 different groups (Figure 4D). Majority of the enriched proteins were related with binding (24%) followed by proteins with unknown functions (17%), RNA helicase activity (14%), catalytic activity (12%) and translation initiation factor (12%). A few of the proteins identified were also involved in antioxidant activity (5%), structural molecules (5%), protein macromolecule adaptor activity (5%), G protein activity (2%), transmembrane transporter activity (2%) and microtubule motor activity (2%).

The current proteomics results also showed changes in the total proteins identified in the different EO-treatments for each category (Figure 4E). Treatment with Lg and its binary mixture (Le + Lg) increased the number of proteins involved in binding mechanisms, catalytic activity, and proteins with no known function. However, proteins involved in RNA Helicase activity were affected by treatment with Lg and the binary mixture (Le + Lg) as observed the numbers proteins involved in RNA Helicase activity were higher in Lg and (Le + Lg) EOs treated in B. cinerea mycelia compared to the Le EO treated and control samples. Furthermore, some proteins involved in translation initiation, antioxidant activity, protein macromolecule adaptor activity and microtubule motor activity were only identified in the Lg and combination treatment.