Chemical Composition and In Vitro Activity of Plant Extracts from Ferula communis and Dittrichia viscosa against Postharvest Fungi

F. communis and D. viscosa are perennial Mediterranean weeds that have been used for different therapeutic purposes in traditional pharmacopeia. Plant extracts were obtained from air dried D. viscosa young shoots (DvA) and F. communis aerial part (FcA) and roots (FcR) with n-hexane. The chemical compositions of the extracts were analyzed by HPLC-DAD, LC-MS (ESI) and LC-Q-TOF techniques. Two sesquiterpene lactones (inuviscolide, tomentosin) and three sesquiterpene acids (costic acid, hydroxycostic acid, ilicic acid) were identified from the D. viscosa extract, while in F. communis extracts three daucane sesquiterpenes (acetoxyferutinin, oxojaeskeanadioyl anisate, fertidin) and one coumarin (ferulenol) derivates were found. Biological activities of plant extracts were studied in in vitro experiments on the colonies and conidia of Botryotinia fuckeliana, Penicillium digitatum, P. expansum, Monilinia laxa, M. fructigena and Aspergillus spp. Extracts showed varying degree of antifungal activities on colony growth and conidia germination. The extract from FcA showed the least effect, while DvA extract had the strongest fungitoxic effects. FcRextract presented a fungitoxic effect on the colony growth, but it was not able to inhibit the conidia germination. These distinctions can be attributed to the differences in chemical composition of plant extracts.


Introduction
Decay diseases caused by fungi create great postharvest losses of fruits and vegetables. Due to their attack they make the crop unfit for consumption and additionally some postharvest pathogens are able to produce mycotoxins [1][2]. Synthetic fungicides are the main measure used in the management of postharvest decay [3]. Pesticide residues [4] and development of resistant strains of fungi from the continuous application [5,6] are the major constraints to their use. Plant derived compounds are generally assumed to be more acceptable and less harmful than synthetic ones. These compounds possess a high potential for pest management since most of them are not phytotoxic, easily biodegradable and sometimes stimulatory to the host metabolism [7]. Therefore, exploitation of the new alternatives to synthetic fungicides should be considered.
D. viscosa belongs to the family Compositae and different chemical investigations have reported the presence of flavonoids [21], triterpenoids [22,23], sesquiterpene lactones and acids [24][25][26]. The biological activity of this plant has been subject of different investigations. Formulated leaf extracts of D. viscosa were tested for their nematicidial activity in field trials, where only a slight effect was obtained, even though the formulated extract was effective in pot experiments [27]. D. viscosa has also been subject of investigation against insects [28] and mites [29]. A sesquiterpene lactone, tomentosin, has been isolated and identified from D. viscosa and it was shown to be active under in vitro conditions against the dermatophytes Microsporum canis, M. gypseum and Trichophyton mentagrophytes [30]. The same authors observed a greater antifungal activity of leaf extracts, due to the high levels of the sesquiterpene carboxyeudesmadiene, against dermatophytes and Candida spp. [31]. D. viscosa extract causes a decline in chitin content, a very important constituent of fungal cell wall, which probably explains the antimycotic activity of the plant extract against dermatophytes. The extract caused dramatic changes in the hyphae and spore morphology due to severe damage in the fungal cell coat [32,33]. Qasem et al. [34] reported that, D. viscosa incorporated as shoot dried material or crude water extract in a culture medium, showed antifungal effects on the mycelial growth of Fusarium oxysporum f.sp. lycopersici. Abou-Jawdah et al. [35] found out that D. viscosa has high activity against spore germination, but only moderate activity against mycelial growth of different fungi. Under in vivo conditions preventive sprays of D. viscosa on squash and cucumber seedlings gave efficient protection against Botrytis cinerea and Sphaerotheca cucurbitae, but failed to control green mold of citrus fruits caused by Penicillium sp. [35]. Extracts of D. viscosa made with organic solvents were effective controlling downy mildew in grape [36] and late blight in potato and tomato, downy mildew in cucumber, powdery mildew in wheat and rust in sunflower, under controlled conditions [37]. Cohen et al. [38] using an emulsified concentrate formulation of the oily paste extracts provided very good control against downy mildew of grapes caused by Plasmopara viticola. The major inhibitory compounds were identified as tomentosin and costic acid [38]. Costic acid has also been sustained to represent one of the principle nematicidal ingredients of D. viscosa against Meloidogyne sp. [39]. Although the antifungal activities of D. viscosa have been studied its possible application in postharvest fungi control remains unexplored.
The aim of this research was to test the in vitro antifungal activity of plant extracts from F. communis and D. viscosa against postharvest fungi of fruits where limited or inexistent information was available and to characterize plant extracts by HPLC-DAD, LC-MS (ESI) and LC-Q-TOF analysis.

Extracted Yield
Extracted yields obtained from air dried plant parts of D. viscosa young shoots (DvA), F. communis aerial part (FcA) and root (FcR) are presented in Table 1. The highest yields of 79.5 and 76.4 g/kg were obtained from FcR and DvA, respectively, while the yield from FcA with 25.1 g/kg, was much lower. These differences are attributed to the fact that plant secondary metabolites often accumulate in specific plant parts [40] and their quantity is related to the plant's phenological stage [41].

Chemical Composition
The HPLC-DAD and LC-MS (ESI) analysis of the DvA n-hexane soluble fraction allowed us to indentify five compounds (Figures 1 and 2; Table 2). Figure 1 shows the HPLC profiles at 206 and 254 nm of the DvA extract under the optimized conditions of analysis, whereas Figure 2 shows the LC-MS (ESI) profile in positive and negative modes. Table 2 lists the compounds identified with their UV absorption maxima, molecular weights and characteristic adducts. The chromatograms in Figure 1 at 206 nm show the presence of two main peaks tomentosin (2) and costic acid (4). Furthermore, three minor peak compounds 1, 3, 5 were found. These peaks could be identified as inuviscolide, ilicic acid and 3α-hydroxycostic acid, respectively.    4 ] + adducts. The ppm difference between an observed ion mass/charge and exact ion mass/charge was below 10.03 ( Figure 5, Table 3).
Ilicic acid had been reported as one of the active anti-inflammatory principles of D. viscosa [42][43]. Costic acid (peak 4) presents a λ max at 210 nm in ESI + mode at m/z 257 [M+Na] + while its hydroxylderivative 3α hydroxycostic acid (peak 5) gives a signal at m/z 252 [M+H] + ( Figure 2). D. viscosa has been repeatedly studied from the chemical point of view and several sesquiterpene lactones, sesquiterpene acids and flavonoids have been isolated from specimens of diverse origins [14][15][16][17][18][19].  The chromatographic analysis of FcA and FcR n-hexane dissolved fraction allowed to identify three daucanes and one coumarin compound (Figures 3 and 4; Table 4). positive and negative ESI-MS spectra no spectra were recorded because of that we could not be certain that is lapiferin.  Figure 4) and was identified as ferulenol. The high resolution mass spectrometry analysis of pure standard showed only [M+H] + and [M+NH 4 ] + adducts ( Figure 5). The ppm difference between an observed ion mass/charge and exact ion mass/charge was 1.36 (Table 3).
In the FcA extract it is evident lower signals intensity attributed to identified compound in the respect to FcR extract ( Figure 4). The differences in the chemical composition between leaves, steam, roots and latex have been reported previously [15]. The components of FcA and FcR extracts were identified by comparing their HPLC retention time, UV absorption maxima and positive and negative mode fragmentation patterns to the corresponding data reported by Arnoldi et al. [14].    The quantity of inuviscolide and tomentosin from D. viscosa and ferulenol from F. communis extracts expressed as mg/g of extracts are presented in Table 5.

Antifungal Activities
Data of ED 50 s of the extracts on colony growth and conidial germination inhibitions are presented in Table 6. Table 6. ED 50 (mg/L) and MIC (mg/L, between the brackets) of the extracts on colony growth inhibitions after two (B. fuckeliana), six (P. digitatum), seven (M. laxa and M. fructigena), nine (Aspergillus spp. and P. expansum) days and conidial germination inhibitions after 10-12 h of incubations. M: colony growth; C: conidial germination; nd: not determined due to lack of inhibition; ─ not tested.
DvA was found to inhibit the colony growth of the fungi in a dose dependent manner. The highest antifungal activity of this extract was found on M. laxa and M. fructigena with ED 50 25.4 and 29 mg/L respectively, followed by B. fuckeliana (ED 50 153 mg/L) and P. digitatum (ED 50 154 mg/L). The lowest antifungal activity of this extract was found on Aspergillus spp. and on P. expansum and calculated ED 50 s were higher than the range of the used concentrations. DvA extract showed inhibition of conidial germination of all tested fungi. The ED 50 s of 371, 192, 201, 83, 106 and 94mg/L were calculated for Aspergillus spp, B. fuckeliana, P. expansum, P. digitatum, M. laxa and M. fructigena, respectively. DvA demonstrated relatively high activity on conidia germination but very low activity on colony growth inhibition of Aspergillus spp. and P. expansum. These results confirm the results found by Abou-Jawdah et al. [35] where extracts of D. viscosa showed high activity against spore germination but only moderate activity against mycelial growth. There are different publications that demonstrate the activity of D. viscosa extracts against dermatophytes [30][31][32][33] and fungal pathogens [35,[45][46] under in vitro conditions. However, there are no data that reports activities of D. viscosa extract on M. laxa and M. fructigena, even though in our experimental conditions it showed the highest effects. FcR show very low activity on the colonies of Aspergillus spp., P. expansum and P. digitatum whereas, on colony growth of M. fructigena, B. fuckeliana and M. laxa it was found to have a good fungitoxic effect with ED 50 of 75, 135 and 202 mg/L, respectively. No inhibitory activity of FcR was found on conidial germination in all tested fungi. FcA extract showed neglected inhibitory activity on tested fungi regarding colony growth inhibition under our experimental conditions. Therefore, the activity of this extract was not tested against the conidial germination. MICs values in most cases were of 400 mg/L except for M. laxa and M. fructigena of 200 and 300 mg/L for mycelia and conidia, respectively. In the case of Aspergillus spp. at 400 mg/L only 90% of mycelia growth inhibition was archived and MIC was not determined.
Principal component analysis (PCA) on the colony growth inhibitions and cluster analysis separate extracts into four different groups ( Figure 6). Figure 6 shows that the principal components PC1 and PC2 accounted for 98.51% of the variation. Based on PCA and cluster analysis, I group contains DvA 3,4, FcR extracts, II DvA 2, 6, FcR 2,3, III DvA 1, FcA 2,4, FcR 1,6 and IV DvA 5, FcA 3, 5, 6, FcR 5 ( Figure 6). The activity of plant extracts on conidial germination clearly distinguishes three different groups (Figure 7). Each one occupied very different ordinal spaces, indicating that inhibition differed substantially among the treatments. Figure 7 shows that the principal components PC1 and PC2 accounted for 96.15% of the variation. Based on PCA and cluster analysis, there are three different groups: I DvA 6 (main inhibition), II DvA 1 to 5 (intermediate) and III FcR 1 to 6 (minor inhibition) (Figure 7). The effectiveness on colony growth and conidial germination varied among the extracts and tested fungi based on PCA and cluster analysis.

Plant Material and Extraction
Plants were collected from rural areas in Apulia Region, south Italy. F. communis (L.) (Giant fennel) was collected during the flowering season and D. viscosa (L.) W. Greuter (False yellow head), during the leaf-rosette phenological phase. A voucher specimen was deposited at the Herbarium of the Botany Department, University of Bari. All samples were ground by a blade mill to a fine powder before extraction. Oleoresins were obtained by n-hexane extraction in a Soxhlet apparatus. Five grams of samples were extracted with 120 mL n-hexane for 6 hours at 68 ºC with 20 min/cycle. At the end of extraction process, solvent in the flask was separated from the oleoresins by evaporation at 40 ºC overnight. The content of oleoresin was corrected according to the dry weight of samples at 105 ºC for 24 hours (till constant weight). Extracts were stored in a freezer at -20 °C until use.

HPLC-DAD Analysis
The Ultimate 3000 HPLC system (Dionex, Germering, Germany) was equipped with a photodiode array detector, fluorescence detector (FR 2000), Ultimate 3000 low pressure pump pump, Rheodyne (Rheodyne, USA) 20 μL injector loop, Acclaim C18 reverse phase column (150 × 4.6 mm; 3 μm) and Acclaim C18 reverse phase precolumn (10 × 4.6 mm; 5 μm) and a column oven. The HPLC was controlled by and data were elaborated using the Chromeleon Software vs 6.8 (Dionex, Germering, Germany). The gradient profile for the separation of F. communis and D. viscosa extracts was as follows: starting from acetonitrile/water (45:55, v/v) for 5 min, then linear gradient to 15% water in 40 min, maintained for 5 min at acetonitrile/water (85:15, v/v) and equilibrated for 5 min with acetonitrile/water (45:55, v/v) at a flow rate of 1 mL/min. Analyses were performed at UV wavelengths of 197, 254 nm for F. communis extracts and 206, 257 nm for D. viscosa extract and PDA detection range was 190-400 nm. The external calibrations were provided, and the standard calibration curves were constructed by plotting concentration against peak area. Good linearities were achieved for inuviscolide, tomentosin and ferulenol between 0.01 and 5 mg/L, with a correlation coefficient of 0.9999.

HPLC/ESI-MS/MS Analysis
A Varian tandem mass spectrometer (Palo Alto, CA, USA) consisting of a ProStar 410 autosampler, two ProStar 210 pumps and a 1200 L triple quadrupole mass spectrometer equipped with an electrospray ionization source was used to analyze the plant extracts. Varian MS workstation, version 6.7 software was used for data acquisition and processing. Chromatographic separation was performed on an XDB column (2.1 × 250 mm I.D., particle size 5 μm, Milford, MA). The mobile phase consisted of (A) acetonitrile and (B) bidistilled water with 0.1% of formic acid. Elution started with acetonitrile/water (10:90, v/v) reaching acetonitrile/water (100:0, v/v) in 20 min and hold to 25 min than equilibration time for 10 min till acetonitrile/water (10:90, v/v). The mobile phase, previously degassed with high-purity helium, was pumped at a flow rate of 0.3 mL/min and the injection volume was 10 μL. The electrospray ionization-mass spectrometer was operated in the positive ion mode. The electrospray capillary potential was set to 50 V, while the shield was at 225 V. Nitrogen at 49 mTorr was used as a drying gas for solvent evaporation. The atmospheric pressure ionization (API) housing and drying gas temperatures were kept at 50 and 380 °C, respectively. The scan time was 1 s, and the detector multiplier voltage was set to 1500 V, with an isolation width of m/z 1.2 for quadrupole 1 and m/z 2.0 for quadrupole 3. ESI mass spectra were acquired modes by scanning over the 50-700 mass range.

HPLC-MS Q-TOF Analysis
The extract was analyzed by reverse phase HPLC on an Agilent 1200 series HPLC system fitted with microchip technology column (Agilent, Zorbax 300 SB-C18 5 μm, 43 mm75 μm). The HPLC conditions were as follows: flow rate, 0.4 μL/min; solvent A, 0.1% formic acid in water; solvent B, methanol; gradient, solvent B 20-100% over 10 min and kept at 100% for 5 min. Then 2 μL of the extract dissolved in methanol-water (80:20, v/v), was analyzed by ESI in positive mode using an Agilent 6520 time-of-flight (TOF) MS. Mass spectral data were acquired in the range m/z 100-1000, with an acquisition rate of 1.35 spectra/s, averaging 10,000 transients. The source parameters were adjusted as follows: drying gas temperature 250 C, drying gas flow rate 5 L/min, nebulizer pressure 45 psi, and fragmentor voltage 150 V. Data acquisition and processing were done using Agilent Mass Hunter Workstation Acquisition v. B.02.00 software.

Colony Growth Inhibition Test
The effect of plant extracts postharvest fungi was tested using a food poison technique. Aliquots of plant extracts diluted in methanol were added to the autoclaved culture medium PDA, when the temperature of the medium was 50 ºC, to reach final concentration of 5, 10, 50, 100, 200, 300 and 400 mg/L. Final concentration of methanol didn't exceed 0.5%. To increase the dispersion of extracts 0.1% of Tween 80 was added. Control in this test was PDA with 0.5% of MeOH and 0.1% of Tween 80. 13 mL of testing and control solutions were added to Petri dishes. All treatments were performed in five replicates. Colony growth inhibition tests were performed by placing 6 mm mycelia agar disks, cut from the margin of expanding fungal colonies, in the centre of Petri dishes. The colony diameters were measured after different periods of incubation in the incubation chamber at 23 ºC in the dark. The first reading was taken when the colony diameter in control reached the border of Petri dish. The first readings were taken after two (B. fuckeliana), six (P. digitatum), seven (M. laxa and M. fructigena), nine (Aspergillus spp. and P. expansum) days of incubation because the hyphal growth and colony formation of fungi depend of fungal species.

Conidial Germination Test
Aliquots of plant extracts diluted in methanol were added to the autoclaved culture medium potato dextrose broth (PDB). The conidia concentration of each fungus 100 µL (6 × 10 5 mL) with 900 µL of PDB (200 g of potato, 10 g of dextrose in 1 L distilled water) were mixed thoroughly to reach final concentration of 10, 50, 100, 200, 300 and 400 mg/L of plant extracts and a final concentration of conidia 6 × 10 4 mL. Final concentration of methanol didn't exceed 0.3%. To increase the dispersion of extracts 0.1% of Tween 80 was added. Control in this test was PDB with 0.3% of methanol and 0.1% of Tween 80. To carry out the test, Petri dishes of 90 mm were used, a sterile filter paper was placed in each plate dripping 2 mL of sterile distilled water. A slide was fixed with the help of plastic supporter in the Petri dish. For each treatment two Petri dishes were prepared. In each slide two drops of 20 µL for each concentration were placed. After 10-12 hours at 23 ± 1 °C of incubation at dark conditions, the percentage of spore germination was determined microscopically at x40 magnification by counting 100 spores in each drop. Conidia were considered to have germinated when germ tube length was greater than the width of the conidium.

Statistical Analysis
One-way analysis of variance (ANOVA) of yielded extracts was used to determine the statistical significance; P-values< 0.01 were considered significant. The means were separated by Duncan's multiple range test. The data were statistically analyzed using the software package STATISTICA 6, StatSoft Inc, Tulsa, USA. Also a principal component analysis (PCA) and cluster analysis was used to discriminate the main groups of extracts on the different fungi. The probit-analysis method was used to analyze the mean data of percentage inhibition of colony growth and conidia germination relative to concentration of each plant extracts. Percentage inhibition data were transformed to probit inhibition data. The regression lines of probit inhibition against log concentration were obtained and the effective concentration 50 (ED 50 ) was calculated in mg/L according to linear model. Means were plotted against log 10 values of the extract concentrations. MIC (minimum inhibitory concentration) was considered as the least concentration of plant extracts that does permit any visible growth of the inoculated mycelia or conidia during the incubation.

Conclusions
Results indicate that D. viscose possesses higher antifungal activity in comparison with that of F. communis, where information on its antifungal activity is limited or inexistent. F. communis root extract only affected colony growth of B. fuckeliana, M. laxa and M. fructigena and no inhibitory effects were observed on conidial germination under applied experimental concentrations. F. communis aerial part extract showed a lack of activity on colony growth and was not tested on conidia.
As expected, D. viscosa extract demonstrated high activity on colony growth of most tested fungi, except Aspergillus spp. and P. expansum. In addition it posses the best effect on conidia germination of M. laxa and M. fructigena where data are still lacking. The extract of D. viscosa could be a potential alternative for the control of fungal rotting of fruits but further in vivo studies should be done.
Different degrees of antifungal activities can probably be explained by the varied chemical composition of the tested plant extracts. It seems that sesquiterpene lactones and acids from D. viscosa were more active than the daucane sesquitepenes and coumarin from F. communis against postharvest fungi. Antifungal activity can be associated with the compounds that occurred in extracts but the activities of pure compounds should be examined.