Antimicrobial Natural Products from Plant Pathogenic Fungi

Isolates of a variety of fungal plant pathogens (Alternaria radicina ICMP 5619, Cercospora beticola ICMP 15907, Dactylonectria macrodidyma ICMP 16789, D. torresensis ICMP 20542, Ilyonectria europaea ICMP 16794, and I. liriodendra ICMP 16795) were screened for antimicrobial activity against the human pathogenic bacteria Acinetobacter baumannii, Pseudomonas aeruginosa, Escherichia coli, Mycobacterium abscessus, and M. marinum and were found to have some activity. Investigation of the secondary metabolites of these fungal isolates led to the isolation of ten natural products (1–10) of which one was novel, (E)-4,7-dihydroxyoct-2-enoic acid (1). Structure elucidation of all natural products was achieved by a combination of NMR spectroscopy and mass spectrometry. We also investigated the antimicrobial activity of a number of the isolated natural products. While we did not find (E)-4,7-dihydroxyoct-2-enoic acid (1) to have any activity against the bacteria and fungi in our assays, we did find that cercosporin (7) exhibited potent activity against Methicillin resistant Staphylococcus aureus (MRSA), dehydro-curvularin (6) and radicicol (10) exhibited antimycobacterial activity against M. marinum, and brefeldin A (8) and radicicol (10) exhibited antifungal activity against Candida albicans. Investigation of the cytotoxicity and haemolytic activities of these natural products (6–8 and 10) found that only one of the four active compounds, radicicol (10), was non-cytotoxic and non-haemolytic.


Introduction
Plant pathogenic fungi have evolved to invade and kill living plant tissue to derive nutrition and facilitate their reproduction. Entry of fungi into living plant tissue can be though natural openings such as leaf stomata [1] or using specialised fungal structures such as appressoria to punch holes though protective plant cell walls [2]. Once inside the plant cell, fungal pathogens must suppress or evade plant defensive responses typically though protein-protein interactions comprising fungal effectors and specific matching host plant resistant (R) proteins [3].
The role of secondary metabolites in this process is under explored but they can act as phytotoxins that enhance pathogenicity and virulence [4] and as mycotoxins that suppress endophytic and other competing phytopathogenic fungi [5]. In our ongoing investigation of the secondary metabolites of several necrotrophic plant pathogens from the International Collection of Microorganisms from Plants (ICMP), namely, Alternaria radicina, Cercospora beticola, Dactylonectria macrodidyma, D. torresensis, Ilyonectria europaea, and I. liriodendra, we isolated natural products 1-10, one of which was novel. Herein, we report the fermentation, isolation, and biological activities of these natural products.

Results
We screened several plant pathogens from the ICMP collection for antimicrobial activity against bioluminescent derivatives of Acinetobacter baumannii, Pseudomonas aeruginosa, Escherichia coli, Mycobacterium abscessus, and M. marinum. Antibacterial activity was measured as log reductions in light output compared to no-fungi controls and are presented as activity scores. Scores above 1 correspond to a >90% reduction in bacterial bioluminescence compared to the corresponding no-fungi control. Similarly, an activity score above 2 corresponds to a >99% reduction. ICMPs 5619, 15907, 16789, 16794, 16795, and 20542 were identified as hits against both M. abscessus with ICMPs 5619, 16794, and 16795 also exhibiting activity against M. marinum (Figure 1). In addition, ICMP 5619 was also found to be active against E. coli. Thus, the fungal isolates Alternaria radicina ICMP 5619, Cercospora beticola ICMP 15907, Dactylonectria macrodidyma ICMP 16789, D. torresensis ICMP 20542, Ilyonectria europaea ICMP 16794, and I. liriodendra ICMP 16795 were selected for further investigation. Data are presented as box and whisker plots of the activity scores. The solid line shown at 0 is the median control value while the dotted line at 1 is the activity threshold. Scores above 1 correspond to a >90% reduction in bacterial bioluminescence compared to the corresponding no-fungi control. Similarly, an activity score above 2 corresponds to a >99% reduction.
All compounds except for 9, due to insufficient sample, were evaluated for their antimicrobial activity against a panel of bacterial (A. baumannii, E. coli, Klebsiella pneumoniae, Methicillin-resistant Staphylococcus aureus (MRSA), and P. aeruginosa) and fungal (Candida albicans and Cryptococcus neoformans) pathogens (Table 1). Of all the compounds tested, cercosporin (7) was the most active against the bacterial strains with a MIC of ≤0.47 µM against MRSA. Intriguingly, the activity of ICMP 5619 we observed against E coli during initial screening was lost after purification, not an uncommon occurrence. Of note was the moderate antifungal activity exhibited by brefeldin A (8) and radicicol (10) against C. albicans with minimum inhibitory concentrations (MICs) of 57 and 44 µM, respectively. Neither of these compounds exhibited significant activity against any of the bacterial strains.
All compounds were also evaluated for their cytotoxicity against human embryonic kidney cells (HEK-293) and any haemolytic activity against human red blood cells. Compounds 6-8 were found to be cytotoxic with 50% cytotoxic concentrations (CC 50 ) of 1.27, 25.33, and 0.89 µM, respectively. Compound 10 was non-cytotoxic and non-haemolytic.
The antimycobacterial activity of compounds 1, 2, and 4-10 was also investigated using bioluminescent strains of M. abscessus and M. marinum ( Figure 4). Compound 3 was not investigated due to the lack of sample. In contrast to the observed antimycobacterial activity observed in the initial screening, none of the tested compounds showed activity against M. abscessus or M. marinum except for dehydro-curvularin (6) and radicicol (10), which showed activity against M. marinum (MIC of 32 and 64 µg/mL, respectively). shown at 0 is the median control value while the dotted line at 1 is the activity threshold. Scores above 1 correspond to a >90% reduction in bacterial bioluminescence compared to the corresponding no-fungi control. Similarly, an activity score above 2 corresponds to a >99% reduction.
The antimicrobial activities of radicinin (2), tetrahydropyrenophorin (4), curvularin (5), dehydro-curvularin (6), brefeldin A (8), and radicicol (10) have previously been reported ( Table 2). Radicinin (2) has been shown to exhibit moderate antifungal activity against Elymus repens and Mycotypha microspora [35] while tetrahydropyrenophorin (4) has been shown to exhibit moderate antibacterial activity against E. coli and Bacillus megaterium as well as antifungal activity against Microbotryum violaceum [36]. Interestingly, we could not replicate this anti-E. coli activity in the present study. This could be because of the differences in the methods used to measure antibacterial activity where Zhang et al. used an agar diffusion assay with an unspecified bacterial inoculum sprayed onto the agar plates [36] while we performed our assays in a liquid medium. It is well-known that factors such as inoculum size [37] can impact on the results of antimicrobial activity testing, so this may be the reason for the discrepancy. Betina and Mičeková [38] previously investigated the antimicrobial activities of 5, 8, and 10 against E. coli, Bacillus subtilis, C. albicans, Saccharomyces cerevisiae, and Botrytis cinerea in disk diffusion assays and found that none of the compounds were active against E. coli and only 10 was active against B. subtilis. Our findings agree with this. In addition, 8 has also been reported to exhibit activity against Aspergillus fumigatus and Microsporum gypseum while 10 has been reported to have activity against Aspergillus flavus [38][39][40]. Investigation of the antimicrobial activities of 5 and 6 against B. subtilis, S. aureus, S. cerevisiae, Sclerotinia sclerotiorum, and Mycobacteria tuberculosis showed that both compounds were inactive against all strains with the exception of 6, which was active against M. tuberculosis (MIC 40 µM) [9,26].
We tested the novel compound (E)-4,7-dihydroxyoct-2-enoic acid (1) from Alternaria radicina ICMP 5619 and while we found it to have no cytotoxic or haemolytic activity, it also possessed no antimicrobial activity against the bacterial and fungal strains we used. Of the other known natural products isolated, we found that compound 7 exhibited potent activity against MRSA and 6 and 10 had activity against M. marinum. This last finding is in keeping with the findings of Souza et al., who determined that compound 6 was active against another Mycobacterium species, M. tuberculosis [26]. Meanwhile, compounds 8 and 10 exhibited moderate antifungal activity against C. albicans. Unfortunately, compounds 6-8 were found to be cytotoxic but not haemolytic, while 10 exhibited no cytotoxicity or haemolytic activity. Table 2. Summary of natural products 1-10 isolated from fungal pathogens.

General Experimental Procedures
Mass spectra were acquired on a Bruker micrOTOF Q II spectrometer. Specific rotations were recorded on an Autopol IV polarimeter using a 1 dm cell (concentration units of g/100 mL). Melting points were recorded on an electrothermal melting point apparatus and were uncorrected. Electronic circular dichroism readings were obtained with a Chirascan circular dichroism spectrometer using a 1 mm cuvette (concentration units of molL −1 ). 1 H and 13 C NMR spectra were recorded at 298 K on a Bruker AVANCE 400 spectrometer at 400 and 100 MHz, respectively, using standard pulse sequences. Proto-deutero solvent signals were used as internal references (CD 3 OD: δ H 3.31, δ C 49.0; (CD 3 ) 2 CO: δ H 2.04, δ C 29.8; CDCl 3 : δ H 0.00 (TMS), δ C 77.16). For 1 H NMR, the data are quoted as position (δ), relative integral, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, p = pentet, m = multiplet, dd = doublet of doublets, ddd = doublet of doublets of doublets, dt = doublet of triplets, dq, = doublet of quartets, br = broad), coupling constant (J, Hz), and assignment to the atom. The 13 C NMR data are quoted as position (δ) and assignment to the atom. Flash column chromatography was carried out using Kieselgel silica gel (40-63 µm) or Merck diol bonded silica (40-63 µm), C 8 (Merck) reversed-phase (40-63 µm) solid support. Gel filtration flash chromatography was carried out on Sephadex LH-20 (Pharmacia). Thin layer chromatography was conducted on DC-plastikfolien Kieselgel 60 F254 plates. All solvents used were of analytical grade or better and/or purified according to the standard procedures.

ICMP 5619-Alternaria radicina
A. radicina is a pathogen of carrot (Daucus carota) causing the disease 'black rot'. It is a globally common fungus associated with carrot production [48]. Culture ICMP 5619 was isolated in February 1969 from a diseased carrot in Ohakune, the major carrot growing region of New Zealand. The identification of this culture is supported by GenBank sequence MW862781 [49].

ICMP 15907-Cercospora beticola
Eight PDA plates were inoculated with ICMP 15907 and incubated at room temperature for 4 weeks. Fully grown fungal plates were freeze-dried (4.62 g, dry weight) and extracted with MeOH (2 × 200 mL) for 4 h followed by CH 2 Cl 2 (200 mL) overnight. Combined organic extracts were concentrated under reduced pressure to afford a brown oil (0.543 g). The crude product was subjected to purification by C 8 reversed-phase column chromatography eluted with a gradient of H 2 O/MeOH to afford six fractions (F1-F6). F4 was subjected to purification by diol-bonded silica gel column chromatography, eluted with CH 2 Cl 2 /MeOH (gradient) to afford cercosporin (7)

ICMP 20542-Dactylonectria torresensis
Nineteen PDA plates were inoculated with ICMP 20542 and incubated at room temperature for 5 weeks. Fully grown fungal plates were freeze-dried (15.26 g, dry weight) and extracted with MeOH (2 × 400 mL) for 4 h followed by CH 2 Cl 2 (400 mL) overnight. Combined organic extracts were concentrated under reduced pressure to afford a brown oil (0.37 g). The crude product was subjected to C 8 reversed-phase column chromatography eluted with a gradient of H 2 O/MeOH to afford five fractions (F1-F5). F4 was subjected to purification by Sephadex LH-20 and eluted with MeOH/5% CH 2 Cl 2 to afford brefeldin A (8) (1.57 mg).

ICMP 16794-Ilyonectria europaea
Ten PDA plates were inoculated with ICMP 16794 and incubated at room temperature for 4 weeks. Fully grown fungal plates were freeze-dried (5.42 g, dry weight) and extracted with MeOH (2 × 200 mL) for 4 h followed by CH 2 Cl 2 (200 mL) overnight. Combined organic extracts were concentrated under reduced pressure to afford a brown oil (0.280 g).
The crude product was subjected to purification by C 8 reversed-phase column chromatography eluted with a gradient of H 2 O/MeOH to afford five fractions (F1-F5). Fraction F3 afforded radicicol (10)

ICMP 16795-Ilyonectria liriodendri
Four PDA plates were inoculated with ICMP 16795 and incubated at room temperature for 4 weeks. Fully grown fungal plates were freeze-dried (2.14 g, dry weight) and extracted with MeOH (2 × 200 mL) for 4 h followed by CH 2 Cl 2 (200 mL) overnight. Combined organic extracts were concentrated under reduced pressure to afford a brown oil (0.189 g). The crude product was subjected to purification by Sephadex LH-20, eluted with MeOH, to afford four fractions (F1-F4). Purification of F2 by silica gel column chromatography, eluted with n-hexane/EtOAc (1:1), afforded a brown solid that was subsequently triturated with CH 2 Cl 2 to afford radicicol (10) as a white solid (8.47 mg).

Antimicrobial Testing of Fungal Cultures
Fungal isolates were grown on PDA (Fort Richard, Auckland, New Zealand) prior to screening for antibacterial activity using a 24-well plate assay, using a modification of a protocol previously described [62]. Briefly, 0.5 mL aliquots of PDA agar were added to triplicate wells of a black 24-well plate (4titude, Millennium Science, Auckland, New Zealand) and allowed to set. With the aid of a sterile scalpel blade, fungal isolates grown on PDA were sectioned into cubes ≤5 mm in diameter and transferred to agar-filled wells of the 24-well plates, ensuring that each cube was placed fungus-side down and touching the agar. The inoculated 24-well screening plates were covered, sealed, and incubated at room temperature.
Fungal growth was monitored visually at regular intervals and recorded the time taken for them to either cover the entire well or to stop visibly growing. During this time, a 6 mm plug of agar was removed from each well using a biopsy punch twice. To screen for antimycobacterial activity, M. abscessus BSG301 [63] and M. marinum BSG101 [64] were resuspended in 0.8% Middlebrook 7H9 agar (Fort Richard, New Zealand) supplemented with 10% Middlebrook ADC enrichment media (Fort Richard, New Zealand) to a final concentration of approx. 10 7 colony forming units (CFU)/mL for M. abscessus and 10 8 CFU/mL for M. marinum. To screen for activity against A. baumannii, E. coli, and P. aeruginosa, bioluminescent derivatives of these bacteria were resuspended in 0.8% Mueller Hinton agar (Fort Richard, New Zealand) to achieve a final concentration of approx. 10 6 colony forming units (CFU)/mL. Thereafter, 50 µL of the bacterial-agar mixtures were pipetted into the cylindrical holes left after the removal of the fungal-agar plugs and allowed to set. The bacterial luminescence was measured at regular intervals using a Victor X-3 luminescence plate reader (PerkinElmer, Waltham, MA, USA) with an integration time of 1 s. Between measurements, plates were covered and incubated static at 28 • C for M. marinum and 37 • C for all of the other bacteria. Luminescence was also measured for bacteria inoculated into wells containing no fungus as the control.

Antimicrobial Assays of Pure Compounds
Antimicrobial evaluation of the pure compounds against A. baumannii ATCC 19606, Candida albicans ATCC 90028, Cryptococcus neoformans ATCC 208821, E. coli ATCC 25922, K. pneumoniae ATCC 700603, P. aeruginosa ATCC 27853, and S. aureus ATCC 43300 (MRSA) was undertaken at the Community for Open Antimicrobial Drug Discovery at The University of Queensland (St. Lucia, Queensland, Australia) according to standard protocols [65] as previously described [63,66,67].
Antimicrobial evaluation against M. abscessus and M. marinum was undertaken using in-house assays with the bioluminescent derivatives M. abscessus BSG301 [63] and M. marinum BSG101 [64]. Assays were performed as previously described [63,68]. Specifically, mycobacterial cultures were grown with shaking at 200 rpm in Middlebrook 7H9 broth (Fort Richard, Auckland, New Zealand) supplemented with 10% Middlebrook ADC enrichment media (Fort Richard, Auckland, New Zealand), 0.4% glycerol (Sigma-Aldrich, St. Louis, MO, USA), and 0.05% tyloxapol (Sigma-Aldrich, St. Louis, MO, USA). M. abscessus was grown at 37 • C and M. marinum at 28 • C. Cultures were grown until they reached the stationary phase (approximately 3-5 days for M. abscessus BSG301 and 7-10 days for M. marinum BSG101) and then diluted in MHB supplemented with 10% Middlebrook ADC enrichment media and 0.05% tyloxapol to give an optical density at 600 nm (OD 600 ) of 0.001, which is the equivalent of~10 6 bacteria per mL. Pure compounds were dissolved in DMSO and added to the wells of a black 96-well plate (Nunc, Thermo Scientific, Waltham, MA, USA) at doubling dilutions with a maximum concentration of 128 mg/mL. Then, 50 mL of diluted bacterial culture was added to each well of the compound containing plates giving final compound concentrations of 0-64 mg/mL and a cell density of~5 × 10 5 CFU/mL. Rifampicin (Sigma-Aldrich, St. Louis, MO, USA) was used as the positive control at 1000 mg/mL for M. abscessus and 10 mg/mL for M. marinum. Between measurements, plates were covered, placed in a plastic box lined with damp paper towels, and incubated with shaking at 100 rpm at 37 • C for M. abscessus and 28 • C for M. marinum. Bacterial luminescence was measured at regular intervals over 72 h using a Victor X-3 luminescence plate reader with an integration time of 1 s. We defined the MIC as causing a 1-log reduction in light production, as previously described [69]. Experiments were carried out in triplicate and repeated if there was sufficient compound.

Conclusions
Investigation of several pathogens from the ICMP collection, Alternaria radicina, Cercospora beticola, Dactylonectria macrodidyma, D. torresensis, Ilyonectria europaea, and I. liriodendra afforded ten secondary metabolites, one of which was novel. Of the isolated metabolites, dehydro-curvularin (6) and radicicol (10) exhibited good activity against M. marinum; cercosporin (7) exhibited potent activity against MRSA; while brefeldin A (8) and radicicol (10) exhibited moderate antifungal activity against C. albicans. Although three of the compounds, 6-8, were also found to be cytotoxic, 10 was non-cytotoxic and non-haemolytic, making it a promising candidate for further study.