Antifungal Activities of Volatile Secondary Metabolites of Four Diaporthe Strains Isolated from Catharanthus roseus

Four endophytic fungi were isolated from the medicinal plant, Catharanthus roseus, and were identified as Diaporthe spp. with partial translation elongation factor 1-alpha (TEF1), beta-tubulin (TUB), histone H3 (HIS), calmodulin (CAL) genes, and rDNA internal transcribed spacer (ITS) region (TEF1-TUB-HIS--CAL-ITS) multigene phylogeny suggested for species delimitation in the Diaporthe genus. Each fungus produces a unique mixture of volatile organic compounds (VOCs) with an abundant mixture of terpenoids analyzed by headspace solid-phase microextraction (SPME) fiber-GC/MS. These tentatively-detected terpenes included α-muurolene, β-phellandrene, γ-terpinene, and α-thujene, as well as other minor terpenoids, including caryophyllene, patchoulene, cedrene, 2-carene, and thujone. The volatile metabolites of each isolate showed antifungal properties against a wide range of plant pathogenic test fungi and oomycetes, including Alternaria alternata, Botrytis cinerea, Colletotrichum gloeosporioides, Fusarium graminearum, and Phytophthora cinnamomi. The growth inhibition of the pathogens varied between 10% and 60% within 72 h of exposure. To our knowledge, the endophytic Diaporthe-like strains are first reported from Catharanthus roseus. VOCs produced by each strain of the endophytic Diaporthe fungi were unique components with dominant monoterpenes comparing to known Diaporthe fungal VOCs. A discussion is presented on the inhibitive bioactivities of secondary metabolites among endophytic Diaporthe fungi and this medicinal plant.


Introduction
Many plants remain unexplored for their endophytic fungi and the potentially important products that they may produce [1]. Catharanthus roseus is known as a pharmaceutical plant containing rich anticancer alkaloids. The extracts of many organs of this plant also exhibit antimicrobial effects [2][3][4][5][6]. It turns out that Catharanthus roseus is host to a diverse group of endophytic fungi [7][8][9][10]. Some endopytic fungi were found to produce several metabolites biosynthesized by the host Catharanthus roseus. The endophytic fungi Curvularia sp. CATDLF5 and Choanephora infundibulifera CATDLF6 isolated from leaf issues were able to enhance leaf vindoline production content of C. roseus cv. Table 1. Access numbers for ITS, translation elongation factor 1-alpha (TEF1), beta-tubulin (TUB), histone H3 (HIS), calmodulin (CAL) genes region sequences of the four endophytic Diaporthe fungi in the GenBank of NCBI.

Sequence Alignment and Phylogenetic Analysis
In order to determine the phylogenetic locations of the four isolates within the Diaporthe genus, 143 reference taxa [26] (Table 2) together with the four isolates were used for building a phylogenetic a tree with Diaporthella corylina as a root outgroup species [23]. The evolutionary relationships were taken on a five-gene concatenated alignment of ITS, TEF1, CAL, HIS, and TUB regions by maximum likelihood (ML) and maximum parsimony (MP) phylogenetic analyses. Sequences were aligned using the MAFFTv.7 online program with default parameters [27]. A partition homogeneity test implemented in PAUP* v. 4.0 (Sinauer Associates, Sunderland, MA, USA) was applied to determine if the five sequence data could be combined. The best evolutionary model for the partitioning analysis was performed on the concatenated sequences by PartitionFinder 2.1.1 [28]. A concatenated alignment for the five gene regions was made from SequenceMatrix [29]. The inference methods of maximum likelihood and maximum parsimony in Mega 6.0 [30] were applied to estimate phylogeny for the concatenated sequences, with the evolutionary models GTR and AIC for ML and MP, respectively, with a bootstrap support of 1000 replicates. Evidence on the trees were visualized and edited by TreeGraph 2 [31].

Antifungal Activity Tests for Fungal VOCs
The antifungal activity of the VOCs was determined by the methods previously described [17,18,20]. The four endophytic fungal strains of Diaporthe and targeted plant pathogenic microorganisms were paired opposite to each other in Petri plates containing PDA with a diameter of 90 mm. The agar was divided into two halves by removing a 2 cm wide strip in the center. An endophytic test fungus was inoculated onto one half-moon of the agar and incubated at 25 • C for five days for optimum production of volatile compounds before the antagonism bioassay. A test pathogen was inoculated onto the opposite half-moon part of the agar at the fifth day. The plates were then wrapped with parafilm and incubated at 25 • C in dark for 72 h. Growth of filamentous pathogenic fungi were quantitatively assessed after 24 h, 48 h, and 72 h based on multiple measurements of growth relative to controls, as described previously [17,18]. The colony diameter was measured in an average of four diameters on hours 24, 48, and 72 h, disregarding the initial inoculum size. Percentage of growth inhibition was calculated as the formula: |(a − b/b)| × 100, a = mycelial colony diameter in control plate; b = mycelial colony diameter in the antagonism treatment plate. Statistical significance (p < 0.01) was evaluated by analysis of variance (ANOVA) followed by the Tukey 5% test. Antifungal activity of VOCs was tested against the plant pathogenic fungi Alternaria alternata, Botryosphaeria dothidea, Botrytis cinerea, Cercospora sp., Colletotrichum gloeosporioides, Fusarium graminearum, Sphaeropsis sapinea, and Valsa sordida, in addition to the oomycete Phytophthora cinnamomi. All tests were made in quintuplicate. Control cultures were obtained by growing each plant pathogen alone, under the same conditions.

Qualitative Analyses on Volatiles of the Four Endophytic Cultures
VOCs in the air space above the endophytic fungal colonies grown for five days at 25 ± 2 • C on PDA were analyzed using the solid phase microextraction (SPME) fiber technique according to previously described protocols [17,18,20]. Control PDA Petri plates not inoculated with the strain was used to subtract compounds contributed by the medium. All treatments and checks were done in triplicate. A fiber syringe of 50/30 divinylbenzene/carboxen on polydimethylsiloxane (Supelco, Bellefonte, PA, USA) was conditioned for 40 min at 200 • C, exposed to the vapor phase inside Petri during 40 min through a small hole (0.5 mm in diameter) drilled on the sides of the Petri plate. The fiber was directly inserted into the TRACE DSQ inlet (Thermo Electron Corporation, Beverly, MA, USA), at 200 • C, splitless mode. The desorption time was 40 s and the desorbed compounds were separated on a 30.0 m × 0.25 mm × 0.25 µm, HP-5MS capillary column, using the following GC oven temperature program: 2 min at 35 • C up to 220 • C at 7 • C/min. Helium was used as the carrier gas at a flow rate of 1 mL/min. The electronic ionization energy was 70 eV and the mass range scanned was 41-560 uma. The scan rate was 5 spec/s. Transfer line and ionization chamber temperatures were 250 • C and 200 • C respectively. Tentative identification of the volatile compounds produced by the four endophytic Diaporthe fungi was made via library comparison using the NIST database and all chemical compounds were described in this report following the NIST database chemical identity. Tentative compound identity was based on at least a 70% quality match with the NIST database information for each compound. Data acquisition and data processing were performed with the Hewlett Packard ChemStation software system (Version 2.0, Scientific Instrument Services, Inc., Ringoes, NJ, USA). Relative amounts of individual components of the treatments were determined and expressed as percentages of the peak area within the total peak area and as an average of the three replicates.

The Identification on the Four Endophytic Isolates within the Diaporthe Genus
Each of the four isolates falling within the genus Diaporthe were further defined using molecular analyses as they appeared different, morphologically ( Figure 1). For instance, strain FPYF3053 had compact mycelia with crenate margins, these colonies developed a brownish yellow pigmentation in the center on the underside having a growth rate of 18.3 mm day −1 (Figure 1a). On the other hand, strain FPYF3054 had aerial mycelium forming concentric rings with grey and dark pigmentation at the center showing a growth rate of 30.97 mm day −1 (Figure 1b). Strain FPYF3055 had vigorously-growing aerial hyphae near the margin, but loose hyphae scattered inside with aging, with a growth rate of 23 mm day −1 (Figure 1c). Finally, strain FPYF3056 had a compact mycelium with a crenate margin, but no pigmentation with a growth rate of 21.7 mm day −1 (Figure 1d). separated on a 30.0 m × 0.25 mm × 0.25 µm, HP-5MS capillary column, using the following GC oven temperature program: 2 min at 35 °C up to 220 °C at 7 °C/min. Helium was used as the carrier gas at a flow rate of 1 mL/min. The electronic ionization energy was 70 eV and the mass range scanned was 41-560 uma. The scan rate was 5 spec/s. Transfer line and ionization chamber temperatures were 250 °C and 200 °C respectively. Tentative identification of the volatile compounds produced by the four endophytic Diaporthe fungi was made via library comparison using the NIST database and all chemical compounds were described in this report following the NIST database chemical identity. Tentative compound identity was based on at least a 70% quality match with the NIST database information for each compound. Data acquisition and data processing were performed with the Hewlett Packard ChemStation software system (Version 2.0, Scientific Instrument Services, Inc., Ringoes, NJ, USA). Relative amounts of individual components of the treatments were determined and expressed as percentages of the peak area within the total peak area and as an average of the three replicates.

The Identification on the Four Endophytic Isolates within the Diaporthe Genus
Each of the four isolates falling within the genus Diaporthe were further defined using molecular analyses as they appeared different, morphologically ( Figure 1). For instance, strain FPYF3053 had compact mycelia with crenate margins, these colonies developed a brownish yellow pigmentation in the center on the underside having a growth rate of 18.3 mm day −1 (Figure 1a). On the other hand, strain FPYF3054 had aerial mycelium forming concentric rings with grey and dark pigmentation at the center showing a growth rate of 30.97 mm day −1 (Figure 1b). Strain FPYF3055 had vigorouslygrowing aerial hyphae near the margin, but loose hyphae scattered inside with aging, with a growth rate of 23 mm day −1 (Figure 1c). Finally, strain FPYF3056 had a compact mycelium with a crenate margin, but no pigmentation with a growth rate of 21.7 mm day −1 (Figure 1d).
A combined alignment of five loci ITS, TUB, TEF1, HIS, and CAL was used for ML and MP phylogenic analyses. Based on the multi-locus phylogeny (Figure 2), the four endophytic Diaporthe strains could not be placed in one species only because they are distinct from each other and from all reference species listed (Table 2, Figure 2). Strains FPYF3055 and FPYF3056 were clustered by giving a high bootstrap support (BS = 82) from MP inference ( Figure S1) while both separated from each other in ML inference ( Figure S2). The reference sequences used to construct the phylogenetic tree were listed in Table 2 with their Genbank accession numbers. The alignment was uploaded in Treebase assigned with SI 22757. A combined alignment of five loci ITS, TUB, TEF1, HIS, and CAL was used for ML and MP phylogenic analyses. Based on the multi-locus phylogeny (Figure 2), the four endophytic Diaporthe strains could not be placed in one species only because they are distinct from each other and from all reference species listed (Table 2, Figure 2). Strains FPYF3055 and FPYF3056 were clustered by giving a high bootstrap support (BS = 82) from MP inference ( Figure S1) while both separated from each other in ML inference ( Figure S2). The reference sequences used to construct the phylogenetic tree were listed in Table 2 with their Genbank accession numbers. The alignment was uploaded in Treebase assigned with SI 22757.   Figure 2. Phylogenetic tree based on combinedITS, CAL, TEF1, HIS, and TUB sequence alignment generated from a maximum parsimony and maximum likelihood analyses. Values near the branches represent parsimony/likelihood bootstrap support values (>70%), respectively. The tree is rooted with Diaporthella corylina. The four endophytic isolates were each named with strain ID marked green box. Compressed branches were used for saving space. The complete phylogenetic trees of MP and ML can be found in Figures S1 and S2, respectively.

The VOCs' Bioactivities of the Four Diaporthe Strains against Plant Fungal Pathogens
All of the four strains were observed to inhibit the growth of nine selected fungal pathogens by producing volatile compounds in the PDA medium ( Table 3). The nine pathogens, Alternaria alternata, Botryosphaeria dothidea, Botrytis cinerea, Cercospora asparagi, Colletotrichum gloeosporioides, Fusarium graminearum, Phytophthora cinnamomi, Sphaeropsis sapinea, and Valsa sordida, are important causal agents to major trees, such as poplars and pines, or agricultural crops in China and elsewhere. All FPYF strains showed different inhibitory activities along the measurements, an exception was observed for the case of strain FPYF3053, which promoted the growth of Phytophthora cinnamomi (Table 3). Furthermore, all selected pathogens, except V. sordid, achieved obvious growth inhibition over around 10% during the testing period. After 24 h, B. cinerea was the most sensitive to VOCs emitted by all endophytic strains, reaching percent inhibitions of more than 55% when dual cultured with each strain. B. dothidea and A. alternata were highly sensitive to VOCs of all the endophytic strains, getting percent inhibitions of more than 30% with an exception to 28% of A. alternata in VOCs of the strain FPYF3053. V. sordida had the least sensitive or insensitive performance in VOCs from all the strains, showing percent inhibitions around 3% when dual cultured with FPYF3056. The inhibitive intensity of FPYF strains' VOCs on growth of pathogens decreased in times to most duel cultures. The maximum drop of the intensity was by 31% in percent inhibition on the pathogen B. cinerea duel culturing with strain FPYF3056. The obvious increase in intensity occurred in the pathogen F. graminearum duel culturing with FPYF3055 and FPYF3056, increasing by around 10% during 72 h. Some pathogens grew fast without percent inhibition records after 24 h (V. sordida) or 72 h (B. dothidea and F. graminearum).  Notes: Data are averages of two cultures grown on the same medium with subtracting those from the control PDA plate. a The quality match is the % likelihood that the compound is identical to that which is listed on the table based on the NIST database. Compounds assigned as unknown with lower than 70% quality match. b The abundance figure presents the percentage amount of each compound in total area relative to all listed compounds detected for one strain.

Endophytic Diaporthe spp. from Catharanthus roseus
Four isolates of endophytes in the genus Diaporthe were obtained from the medicinal plant Catharanthus roseus growing in a conservation area of Southern China. In order to best distinguish these individual organisms they were subjected to a combined analysis of five-loci alignment of TEF1-TUB-CAL-HIS-ITS which gave a more robust isolate identification [23]. Adding our four endophytic isolates did not affect the congruency in each locus, partition homogeneity for the combination and the best evolutionary model for the five-locus concatenated alignment reported. Diaporthe fungi are one of the most common endophytic fungal communites found in plants [11]. However, the previous work on endophytic fungi from C. roseus [7,8,[32][33][34][35][36][37][38][39][40][41][42] did not record strains of the Diaporthe genus. Alternaria alternata was determined as the dominant endophytic species in leaf tissue of C. roseus along with associated fungi from the following genera, Aspergillus, Fusarium, Penicillium, and Helminthosporium [33]. In addition the endophytes of root tissue appeared including Colletotrichum sp., Macrophomina phaseolina, Nigrospora sphaerica, and Fusarium solani [7]. Other isolated endophytic fungi from this plant included Colletotrichum truncatum, Drechsclera sp., Cladosporium sp., and Myrothecium sp. [43]. To our four Diaporthe strains, no reproductive structures were obtained in the employed conditions. They were designated Diaporthe sp. strains (FPYF3053-3056) without spore characterization strictly using phylogenetic analysis. The strains seemed not to share a close phylogenetic relationship to any other species based on the five-locus alignment study (Figure 2, [12,23,26]). The robust inference on the strains will take place when fruits bodies appear combined with full species phylogeny in the genus Diaporthe.

VOCs Antifungal Effects of Endophytic Diaporthe spp. from Catharanthus roseus
Compounds extracted from Catharanthus roseus [4,5] and extracts from some endophytes of this plant [10,44] have been shown to have antimicrobial bioactivities to some human microbial pathogens and plant fungal pathogens, including Staphylococcus aureus, Pseudomonas aeruginosa, Bacillus subtilis, Escherichia coli, Aspergillus fumigatus, Candida albicans, etc. However, the VOCs or essential oils from Catharanthus roseus in the literature is scarce results on antimicrobial activities [45,46]. The previous work on the other endophytic fungi of this host plant did not consider that VOCs of the endophyte may have antimicrobial activities [7,8,[32][33][34][35][36][37][38][39][40][41][42]. However, this work shows that VOCs produced by four endophytic Diaporthe fungi from the plant are able to functionally inhibit the growth of a number of specifically-targeted fungal pathogens (Table 3).
In the past there have been three endophytic Diaporthe strains recorded with their VOCs [18,19,47]. Two of them were reported to be inhibitory to plant pathogens [18,19]. One strain PR4 was isolated from a medicinal plant growing in Kashimir, Himalayas [19]; the other strain EC-4 was isolated from Odontoglossum sp. in Northern Ecuador [18]. With our four strains, the volatile compounds from endophytic Diaporthe fungi varied in degrees of inhibition against selected pathogenic fungi and test timings depending on the endophytic strain (Tables 3 and 5). However, the maximal inhibition of fungal growth of Diaporthe was from strain PR4, which reduced growth of Rhizoctonia solani by 100%. FPYF strains' and EC-4 VOCs also appeared effective in the inhibition of growth of Botrytis cinerea by more than 30% with a maximal of 50.42 ± 1.8%. During the test course of 72 h, to most cases, FPYF strains' VOCs showed strong bioactivities in the first day and then decreased inhibition on the pathogens in following two days (Table 3). PR4 VOCs were effective in reducing radial growth of Pythium ultimatum by 13.3%; EC-4 VOCs were effective in reducing radial growth of Pythium ultimatum, Phytophthora cinnamomi, and Phytophthora palmivora by 59.1 ± 0.9%, 42.0 ± 0.5%, and 5.6 ± 0.5%, respectively. FPYF3054-3056's VOCs were effective against Phytophthora cinnamomi in a range of 25.21 ± 4.3~11.32 ± 4.2%. The alcohol compounds such as 1-propanol,2-methyl-and 1-butanol,3-methyl-might made the oomycete P. cinnamomi more sensitive to EC-4's VOCs [18], which were lack in VOCs of all FPYF strains ( Table 4). The two alcohol compounds had antimicrobial activities in VOCs of endophytic Phomopsis sp. strain EC-4 [18]. The sensitivity of the pathogen F. graminearum to VOCs from Diaprothe spp. might be analogous even though the VOCs components were not similar among Diaprothe strains. Two Diaporthe strains FPYF3053, 3055 (Table 2) and Diaporthe strain PR4 [19] had percent inhibition of F. graminearum growth of around 30% under their VOCs bioactivities. However, only beta-phellandrene was a common compound found in VOCs among them (Table 4, [19]). Contrast to cytochalasins as a predominantly common component in soluble secondary metabolites of Diaporthe strains [16], the genus-specific or predominant conserved components of fungal VOCs of genus Diaporthe should be proposed to illustrate further. The experimental data suggests that the VOCs of FPYF strains are both biologically active and biologically selective. Finally, isolate FPYF3053 were showed no effective inhibition of Phytophthora cinnamomi growth. In this study, we attempt to understand the VOCs inhibitory impacts from the endophytic Diaporthe strains without consideration of interaction between the strains and pathogenes. Future research is proposed to investigate the dual interaction in the VOCs' levels and other molecules between fungal interactions [48].
The headspace analyses of the four Diaporthe strains in potato dextrose medium revealed that three monoterpenes-β-phellandrene, biphenylene,1,2,3,6,7,8,8a,8b-octahydro-4,5-dimethyl and α-muurolene-seemed to be characteristic compounds of endophytic Diaporthe strains endophytic to Catharathus roseus. However, among all monoterpenes mentioned above, only 1-menthone can be found in volatile compounds of Catharathus roseus flowers, the essential oil of which is high in limonene and other monoterpenes [45,46]. Menthol and β-phellandrene were also found in VOCs of Diaporthe strain PR4 with very low relative amounts of less than 1.0% [19]. No chemicals were shared in VOCs between our FPYF strains and Phomopsis strain EC-4 (Table 4, [18]). Therefore, the antifungal VOCs from the four endophytic Diaporthe Chinese strains possesses unique VOC compositions compared with known Diaporthe VOCs. Although many fungi were reported to produce many terpene compounds in their VOCs [49], our Diaporthe fungi maybe of some interest as a source of some other monoterpenes, which often only have been thought to originate from specific plants. For instance, essential oils from many plants containing more or less such monoterpenes as α-thujene, β-phellandrene [50][51][52], γ-terpinene [53,54], l-menthone [55,56], cyclohexanol, α-muurolene, thujone, and caryophyllene have some antifungal activities. For example, γ-terpinene, singly or in mixtures with sabinene in oil from coastal redwood leaves, has strong antifungal activity on some endophytic fungi [53]. Therefore, it could be rational to infer the terpenes in FPYF strains synergistically played a main role in their inhibition pathogenic fungi growths. In addition, the high content of monoterpenes in the Diaporthe VOCs does have potential for the biofuel industry [18,20,57].