OUP user menu

Testing the nematophagous biological control strain Paecilomyces lilacinus 251 for paecilotoxin production

Alamgir Khan, Keith Williams, Helena Nevalainen
DOI: http://dx.doi.org/10.1016/S0378-1097(03)00654-2 107-111 First published online: 1 October 2003

Abstract

Paecilomyces lilacinus is a nematophagous fungus currently developed as a biological control agent. In order to evaluate potential toxin production, culture extract and concentrated culture supernatant of P. lilacinus strain 251 were tested against Gram-negative and Gram-positive bacteria. High-performance liquid chromatography analysis was carried out to compare the chromatograms of P. lilacinus strain 251 with the chromatogram of known paecilotoxin. It was found that the 251 strain of P. lilacinus did not produce detectable levels of paecilotoxin or other toxins with antimicrobial activity.

Keywords
  • Paecilotoxin
  • Antibacterial
  • Paecilomyces lilacinus

1 Introduction

The use of fungi in biological control of pests has been adopted recently in agriculture. Paecilomyces lilacinus, a common soil hyphomycete is well known as an egg parasite of plant parasitic nematodes [13] and is currently developed as biocontrol agent [4]. In addition to infecting plant parasitic nematodes, P. lilacinus has been reported to infect humans [5,6] and animals [7]. Therefore, potential production of toxins such as paecilotoxin by isolates applied for biological control presents a safety risk and should be assessed carefully. Even though not compulsory for product registration, testing for toxin production is recommended by regulatory authorities in some countries such as Australia.

Toxins produced by microorganisms are typically secondary metabolites featuring peptides, polypeptides and non-peptide antibiotics [810]. Mycotoxin production and toxic effects vary according to the fungal strain, culture medium and target organism. For example, culture filtrates of P. lilacinus grown on a medium containing malt, tested against 17 species of nematodes, were shown to be toxic only against Meloidogyne and Heterodera spp. The toxic effect of the unknown toxic metabolite in nematodes was neurotropic [11]; however, it is not known whether the culture filtrates had any antibacterial activity.

Several P. lilacinus isolates from Japan, some of which are of clinical origin, have been shown to produce paecilotoxins, known as leucinostatins [12]. These paecilotoxins are neutral straight peptides that contain an unsaturated fatty acid and an amine residue in their N-terminus and C-terminus, respectively. They exhibited uncoupling activity against rat liver mitochondria [13] and antimicrobial activity against Gram-positive bacteria [12]. It was also demonstrated that the paecilotoxins caused oral toxicity and formation of mycoses in humans [13] and in the infection of insects and nematodes [12]. Since P. lilacinus kills nematodes by invading nematode eggs, paecilotoxins may not be involved in the infection process and therefore their absence is not likely to affect the overall ability of the fungus to control nematodes.

Paecilomyces marquandii (Massee) has been reported to produce different forms of paecilotoxins [14], which have been found to be identical with paecilotoxins produced by P. lilacinus [12]. Apart from paecilotoxins, there are no reports describing characterization of other toxic metabolites from Paecilomyces spp.

The Japanese finding that paecilotoxin isolated from the different strains of P. lilacinus showed antimicrobial activity against Gram-positive bacteria [12] would provide a simple test for the screening of other strains of Paecilomyces spp. for paecilotoxin production. The aim of this study was to assess the ability of P. lilacinus strain 251, a potential biological agent against plant parasitic nematodes, to produce paecilotoxin. This strain was selected for further development from 15 other P. lilacinus strains based on its nematophagous ability and resistance to UV [14]. The testing is part of the registration process and addresses an occupational health risk to workers exposed to the strain.

2 Materials and methods

2.1 Fungal strain and cultivation conditions

Two liquid culture media were used to induce toxin production by P. lilacinus strain 251 (deposited at Australian Government Analytical Laboratory, accession number 89/030550). The first medium (CM1) contained sucrose 5% (w/v), bactopeptone (Oxoid) 0.5% (w/v), yeast extract (Difco) 0.5% (w/v), Na2CO3 1% (w/v) K2HPO4 0.1% (w/v), and MgSO4·7H2O, 0.02% (w/v) at pH 8.5 [12]. The second medium (CM2) was a standard potato dextrose broth (Difco). Spores of P. lilacinus were harvested in sterile water (about 10 ml per plate) from a potato dextrose agar (PDA) plate grown for 10 days and 150 ml of medium in 500 ml Erlenmeyer flasks was inoculated with the spores at 2.75×106 ml−1. The flasks were incubated at 27°C for 2 weeks on an orbital shaker at 125 rpm.

2.2 Extraction of secreted metabolites

Potential toxins were extracted using the method described earlier [12] with minor modifications. Culture supernatants (100 ml) were adjusted to pH 3.0 with 1 N HCl and extracted with the same volume of ethyl acetate. The extracts were washed with 5% (w/v) NaHCO3 and vacuum-concentrated. Crude metabolite fraction prepared in this way was dissolved in a small amount (300–400 µl) of methanol. Each sample was passed through a 0.2-µm pore size teflon filter (Advantec MFS, CA, USA). The resulting extracted potential toxin (EPT) was analyzed by high-performance liquid chromatography (HPLC) and assayed against Gram-positive and -negative bacteria to reveal possible antibacterial activities. Concentrated culture supernatant (CCS) was also tested in the assay.

2.3 Detection of antimicrobial activity against bacteria

Two Gram-positive bacteria, Bacillus subtilis (strain CL062) and Micrococcus luteus [15], and the Gram-negative Escherichia coli (strain JM109) were prepared as follows. Standard medium (SM) plates (1.2% (w/v) agar, 1% (w/v) bactopeptone, 1% (w/v) glucose, 0.1% (w/v) yeast extract, 0.1% (w/v) MgSO4·7H2O, 0.22% (w/v) KH2PO4) were inoculated with 100 µl of a culture previously grown overnight in axenic broth [16] at 37°C.

Antibacterial activity was tested with the following methods. Method 1: a 3-mm hole was cut aseptically in the center of each 24-h-old culture plate of B. subtilis and E. coli and 10 µl of either EPT or CCS was pipetted into each well with four replications. Plates were then incubated at either at 33 or 37°C up to 48 h. Method 2: 7.5µl of an overnight culture of E. coli, B. subtilis or M. luteus was spread on SM plates with three replications. The bacteria were allowed to dry on the plate, after which 100 µl of either EPT or CCS was added on top and the plates were incubated as above. Antimicrobial activity was assessed by counting the number of colony forming units (cfu) and observing the diameter of bacterial colonies by eye. Method 3: 6-h-old liquid cultures of E. coli, B. subtilis and M. luteus were mixed with CCS from fungal cultivation (bacterial culture:CCS ratio was 5:2) for 18 h at 37°C. After incubation, 100 µl aliquots of the mixtures were plated on SM plates. Bacterial growth inhibition was assessed by counting the number of cfu and comparing the numbers to those from the treatment without CCS.

2.4 Comparison of antimicrobial activity of EPT with a crystallized paecilotoxin

This method involved a culture of M. luteus and paper disks. The crystallized toxin from P. lilacinus Odashima strain [12] was kindly donated by Dr. Y. Mikami, Research Center for Pathogenic Fungi and Microbial Toxicoses, Chiba University, Japan. Sterile paper disks (7 mm) were soaked either in EPT (T1) crystallized toxin in methanol (T2), chloramphenicol (5 µg ml−1) in 70% (v/v) methanol (T3), and methanol only (T4). The disks were then air-dried and placed on a 1-day-old bacterial lawn on SM. The plates containing paper disks with three replications were incubated at 26°C for 7 days. Toxic effect was shown by formation of clearing zones around the disks on the bacterial lawn.

2.5 HPLC analysis

HPLC was employed to compare the chromatograms of EPT from P. lilacinus strain 251 and the crystallized paecilotoxin from Odashima strain [12]. All chromatograms were performed using a C8 column (Pharmacia, Sweden). The solvent used was methanol:2-propanol:water:acetonitrile:diethylamine in a ratio 40:30:20:10:0.1. The flow rate was 1 ml min−1 and total running time was 20 min for each run.

3 Results and discussion

3.1 Effects of EPT and CCS on plate assays

The EPT and CCS of P. lilacinus 251 applied in a hole of a B. subtilis plate (method 1) showed almost no effect on the growth of bacteria (Table 1). A very small effect on the growth was observed by the application of CCS at 33°C, showing a 1-mm clearing zone around the hole. However, the clearing zone disappeared under bacterial growth when the plates were checked 48 h after incubation. There was no effect against E. coli.

View this table:
Table 1

Effect of EPT and CCS from P. lilacinus on the growth of B. subtilis on an agar plate

Incubation temperature (°C)Extract24 h after incubation48 h after incubation
33EPTno effectno effect
33CCS1 mm clearing around holeno effect
37EPTno effectno effect
37CCSno effectno effect
  • EPT or CCS were applied on a central hole cut in agar.

EPT and CCS applied on the dried E. coli, B. subtilis and M. luteus growth on SM plates (method 2) had little or no effect on the growth of bacteria (data not shown). A minor effect by CCS was observed 24 h after incubation on B. subtilis and M. luteus but this effect was minimal after 48 h. When CCS of P. lilacinus was mixed with 6-h-old liquid cultures of E. coli, B. subtilis and M. luteus and the mixture incubated for 18 h prior to plating on SM plates (method 3), no significant inhibition of bacterial growth was observed based on the cfu numbers (data not shown).

The negative results obtained suggest that P. lilacinus strain 251 does not produce detectable levels of antibacterial toxins, paecilotoxin included, or other metabolites that may be toxic to Gram-positive and Gram-negative bacteria. Using a similar test with B. subtilis, it was found that 19 of the 20 P. lilacinus isolates obtained mainly from humans but also including isolates from nematodes, insects and soil, produced toxins [12]. The toxic effect was observed only in Gram-positive bacteria.

Production of toxins seems to depend on a particular fungal isolate. This has shown to be true with a range of fungal species, for example, Aspergillus flavus in relation to production of aflatoxins, cycleopiazonic acid and aftatrem [17]. Also, it has been demonstrated that 28 of 42 isolates of Fusarium graminearum produced deoxynivalenol (DON) [18]. DON is one of the most common contaminating mycotoxins in food.

3.2 Production of potential paecilotoxin

In order to verify the production or non-production of paecilotoxin by the strain P. lilacinus 251, paper disks soaked with EPT (T1), crystallized paecilotoxin (T2) obtained from Mikami and an antibiotic chloramphenicol were placed on 1-day-old culture plates of M. luteus. A clearing zone around the bacterial culture indicating inhibition was observed only when using crystallized paecilotoxin (T2) and chloramphenicol (T3). No inhibition of M. luteus was observed by the EPT of P. lilacinus strain 251 (T1) or methanol only (T4) after incubation of the plates for 7 days (Fig. 1). In the case of T2 and T3, the clearing zone was significantly visible from the second day of incubation onwards. Inhibition of bacterial growth was not observed either using EPT (T1 in Fig. 1) or CSS of P. lilacinus in the same conditions (data not shown). Again, the results indicate that P. lilacinus strain 251 does not produce metabolites that are toxic to Gram-positive bacteria. Mikami et al. [12] observed toxic activity against another Gram-positive bacterium, B. subtilis, with the same crystallized paecilotoxin used in this experiment.

Figure 1

Effect of EPT of Paecilomyces lilacinus 251 (T1), crystallized toxin from Odashima strain by Mikami (T2), chloramphenicol (T3) and methanol (T4) on growth of M. luteus. Paper disks containing a particular compound are seen as white in the middle of the photograph. Plates were incubated at 26°C for 7 days. Clearing zone, a dark circle around the paper disk is seen only in T2 and T3. Shadow of the paper disk appears on T1 and T4.

3.3 HPLC analysis for evidence of toxin produced by P. lilacinus

HPLC chromatograms of the EPT of P. lilacinus strain 251 were compared with the chromatogram published for the crystallized paecilotoxin from the Odashima strain [12]. The chromatograms presented in Fig. 2B,C were obtained from the EPT of P. lilacinus strain 251 grown using conditions known to promote toxin production [12] and from a culture in PD broth, extracted according to the published method [12]. The chromatogram published for crystallized paecilotoxin is presented in Fig. 2A for comparison. The two large peaks appearing at 2 min after the start on all three chromatograms provide a cross-reference to the work of Mikami et al. [12], who showed that these peaks did not represent toxins. The three paecilotoxin peaks, labeled as a, b and c (Fig. 2A) appearing between 3.75 and 7.0 min, were absent from the chromatograms obtained from the P. lilacinus strain 251 using two different culture media (Fig. 2B,C). This indicates that P. lilacinus strain 251 does not produce detectable levels of paecilotoxin.

Figure 2

Chromatograms of the EPT of P. lilacinus strain 251 and crystallized paecilotoxin. A: Crystallized paecilotoxin from the Odashima strain; B: EPT of P. lilacinus strain 251 cultured under conditions promoting paecilotoxin synthesis; C: EPT of P. lilacinus strain 251 cultured in PD broth. Paecilotoxin peaks (a, b and c in panel A) are not produced by P. lilacinus strain 251 under the culture conditions applied.

On the basis of the evidence obtained in this study, P. lilacinus strain 251 does not produce paecilotoxin that would be detectable by HPLC analysis and bioassays. The crude extract (EPT) or CCS had no or only a minimal effect on growth of the Gram-negative E. coli and the Gram-positive bacteria B. subtilis and M. luteus. Even though some clinical isolates of Paecilomyces have caused oculomycosis disease in humans [5,12], it is unlikely that the strain 251 of P. lilacinus would cause a mycotic disease, as this strain does not grow at 37°C and does not survive more than 48 h at 37°C (unpublished observations). A study by Garcia [19] indicated that isolates of P. lilacinus differ genetically with only those isolates capable of growing at 37°C were shown to be infectious to humans. Therefore, it is reasonable to conclude that P. lilacinus strain 251 can be applied as a biocontrol agent without any hazard to humans and without interfering too much with the growth of other microorganisms in the soil.

Acknowledgments

We thank Warren Kett, Macquarie University Center for Analytical Biochemistry (MUCAB), for helping with HPLC and Rita Holland for assistance in collecting data. This work was funded by an Australian Post-graduate Award to A.K. The work was carried out when A.K. and K.W. were at the Department of Biological Sciences at Macquarie University in Sydney.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
View Abstract