Pyrrolnitrin from Burkholderia cepacia - Wiley Online Library

Journal of Applied Microbiology 1998, 85, 69–78

Pyrrolnitrin from Burkholderia cepacia: antibiotic activity against fungi and novel activities against streptomycetes N. El-Banna and G. Winkelmann Microbiology and Biotechnology, University of Tu¨bingen, Tu¨bingen, Germany 6413/09/97: received 22 September 1997, revised 5 December 1997 and accepted 10 December 1997 N . E L- B AN NA A ND G. W IN KE L MA NN . 1998. A bacterial strain identified as Burkholderia cepacia NB-1 was isolated from water ponds in the botanical garden in Tu¨bingen, Germany, and was found to produce a broad spectrum phenylpyrrole antimicrobial substance active against filamentous fungi, yeasts and Gram-positive bacteria. In batch culture containing glycerol and L-glutamic acid, the isolate NB-1 produced the antibiotic optimally late in the growth phase and accumulated a main portion in their cells. Isolation and purification of the antibiotic from Burkholderia (Pseudomonas) cepacia NB-1 by acetone extraction, gel filtration on Sephadex LH-20 and preparative HPLC yielded 0·54 mg l−1 of a pure substance. Spectroscopic data (HPLC, MS and NMR) confirmed that the compound was pyrrolnitrin [3-chloro-4-(2?-nitro-3?-chloro-phenyl) pyrrole]. Pyrrolnitrin has an inhibitory effect on the electron transport system, as demonstrated by isolated mitochondria from Neurospora crassa 74 A. This inhibition was relieved by N,N,N?,N?-tetramethyl-p-phenylenediamine dihydrochloride (TMPD), indicating that pyrrolnitrin blocked the electron transfer between the dehydrogenases and the cytochrome components of the respiratory chain. Among Gram-positive bacteria, pyrrolnitrin was most active against certain Streptomyces species, especially S. antibioticus, which has not previously been described in the literature. In the presence of pyrrolnitrin, aerial mycelium and spore formation of Strep. antibioticus was suppressed, although growth continued via substrate mycelium. The new findings of inhibition of streptomycetes and their secondary metabolism by pyrrolnitrin may contribute to the fact that Pseudomonas species predominate in soil and compete even with antibiotic-producing Streptomyces.

INTRODUCTION

Strains of Pseudomonas species are aggressive colonizers of the rhizosphere of various crops and have a broad spectrum of antagonistic activity against plant pathogens, such as antibiosis (the production of inhibitory compounds) (Upadhyay et al. 1991), siderophore production (iron-sequestering compounds) (Winkelmann and Drechsel 1997) and nutrition or site competition (Bull et al. 1991). Some species of Pseudomonas can also produce levels of HCN that are toxic to certain pathogenic fungi (David and O’Gara 1994). These characteristics make Pseudomonas species good candidates for Correspondence to: Prof. Dr G. Winkelmann, Microbiology and Biotechnology, University of Tu¨bingen, Auf der Morgenstelle 28, 72076 Tu¨bingen, Germany. © 1998 The Society for Applied Microbiology

use as seed inoculants and root dips for biological control of soil-borne plant pathogens. Among Pseudomonas species, Ps. cepacia, which has been renamed Burkholderia cepacia (Yabuuchi et al. 1992), is a ubiquitous bacterium with potential for biological control against fungal pathogens (Janisiewicz and Roitman 1988). The production of an antifungal compound by B. cepacia has been regarded as one of the mechanisms involved in antagonism (Upadhyay et al. 1991). It is known that B. cepacia produces several antibiotics. Most of them have antifungal activity, such as cepaciamide A (Jiao et al. 1996), cepacidine A (Lee et al. 1994) and xylocandin complex (Meyers et al. 1987). However, compounds such as cepacin A and cepacin B (Parker et al. 1984) exhibited only antibacterial activity, whereas pyrrolnitrin was effective against fungi, yeasts and Gram-positive bacteria (Arima et al. 1965). Pyrrolnitrin, a

70 N . E L- B AN NA A ND G. W IN KE L MA NN

chlorine-containing phenylpyrrole derivative with antifungal activity, is produced by a number of Pseudomonas strains (Arima et al. 1964 ; Hamill et al. 1967), a Myxococcus strain (Gerth et al. 1982) and Enterobacter agglomerans (Chernin et al. 1996). Its biosynthesis originates in several steps from tryptophan (van Peé et al. 1980). It has been shown that pyrrolnitrin can provide biocontrol of several fungal diseases, such as damping-off in radish (Homma and Suzui 1989) and blue mould and grey mould in apple (Janisiewicz and Roitman 1988). During studies on screening for antifungal substances among non-fluorescent pseudomonads, an isolate was found to produce pyrrolnitrin, possessing potent antifungal and antistreptomycetal activity. Therefore, the aim of this investigation was : i) to screen bacteria of the non-fluorescent Pseudomonas group for production of antifungal antibiotics; ii) to optimize the production conditions under which the antibiotic can be produced, to elaborate isolation methods by using extraction procedures, column separation and PHPLC, and to establish an HPLC system for identification and quantitative determination of the antibiotic; and iii) to determine the actual target of the antibiotic by measuring the inhibition of respiration of isolated fungal mitochondria by the oxygen electrode method, and to investigate the activity against species of Streptomyces.

MATERIALS AND METHODS Strains and growth conditions

Both antagonistic and non-antagonistic strains were isolated from ponds in the botanical garden, Tu¨bingen, Germany. All bacterial isolates were initially screened for antifungal activity by using the Petri plate assay on NGYMT plates, containing (g l−1) nutrient broth 8, glucose 4, yeast extract 5, malt extract 10, tryptone 10, pH 6·8, using Trichoderma pseudokonigii as a preliminary test strain. Single colonies were selected and patched along the perimeters of plates on which 30 ml of the suspension of T. pseudokonigii (107 spore ml−1) was placed at the centre and spread over the entire surface of the plate. The plates were incubated at 27 °C for 48 h, and the antifungal activity was determined by measuring zones of inhibition of fungal growth. The isolates were identified with API 20 NE diagnostic strips (bioMerieux Marcy L’Etoile, France). All experiments dealing with the optimization of antibiotic production were carried out in 500 ml Erlenmeyer flasks containing 100 ml of medium inoculated with 2 ml of a 24 h preculture. Inoculated flasks were incubated at 27 °C on a rotary shaker (Pilot-Shaker, LAB-shaker, Braun, Melsungen, Germany) at 110 rev min−1 for 5 d. Bacterial growth of B. cepacia was measured spectrophotometrically as an increase of optical density at 600 nm.

Fermentation and pyrrolnitrin isolation

Fermentation of pyrrolnitrin was performed in a 100 litre fermenter (Pilot–fermenter type P 100 Bioengineering, Wald, Switzerland) equipped with a flat-blade turbine stirrer. The production medium contained (g l−1): Ca(NO3)2, 3 ; KH2PO4, 21·8 ; Na2HPO4, 5·7 ; MgSO4, 0·5 ; ZnSO4, 0·05 ; FeSO4 . 7H2O, 0·5 ; glycerol, 35 ; monosodium glutamate, 10. Two litres of a 24 h old preculture grown in the same medium in a 5 litre Erlenmeyer flask with shaking (110 rev min−1) at 27 °C was used for inoculation. The fermentation was run for 5 d at 27 °C with an aeration rate of 0·5 vvm (volume air per volume fluid per minute) and stirrer speed of 150 rev min−1. The start pH of the medium was adjusted to 7·0. At the end of the fermentation, the cells were collected and extracted with acetone. The acetone extract was then concentrated under vacuum at ³50 °C, and the dry residue was dissolved in methanol and purified by gel filtration (Sephadex LH-20) using methanol as the eluent. Pure pyrrolnitrin was obtained by reverse phase HPLC on a Nucleosil C 18 column (5 mm, 250×8 mm, Grom, Herrenberg, Germany). A gradient of acetonitrile–water (20 to 100%) at a flow rate of 1·6 ml min−1 was used. Absorbance of the effluent was monitored at 220 nm. Major peaks were assayed for antifungal activity. The fraction that displayed antifungal activity was collected and stored at −20 °C. Detection and quantitative determination of pyrrolnitrin was made by analytical HPLC (Shimadzu, Europe GmbH, Duisburg, Germany) using a Nucleosil C 18 column (5 mm, 250×4·6 mm, Grom) and a gradient of acetonitrile–water (20 to 100%) at a flow rate of 1 ml min−1. Absorbance of the effluent was monitored at 200 nm. Physical and chemical characterization of pyrrolnitrin

Mass spectrum. Pneumatically-assisted electrospray mass

spectrum was recorded on a Sciex API II triple quadrupole mass spectrometer with a 2400 Da mass range equipped with an IonsprayTM source (Sciex, Toronto, Canada). A solution of the antimicrobial substance in methanol was injected at a continuous flow of 5 ml min−1 and measured in negative mode. 1

H-NMR spectrum. A one-dimensional NMR spectrum was recorded on a Bruker AMX 250 instrument at 250 MHz (1H) at 305 K (Bruker, Karlsruhe, Germany) in CDCl3. C-NMR spectrum. The 13C-NMR spectrum was recorded

13

as a gated spin echo on a Bruker AMX 250 instrument at 62·5 MHz (250 MHz 1H) at 305 K in deuteriochloroform. Antimicrobial spectrum of pyrrolnitrin. The antimicrobial

© 1998 The Society for Applied Microbiology, Journal of Applied Microbiology 85, 69–78

P YR RO L NI TR I N F RO M BU RK H OL DE R IA CE P AC IA 71

activity of the isolated substance against fungi and streptomycetes was determined by the agar diffusion test. A 10 ml aliquot of the antibiotic solution was put on a filter disc. After sterilization by microwave, a filter disc was placed on the surface of an agar test plate medium. Minimum inhibitory concentration (MIC) of the antibiotic against bacteria and yeasts was additionally determined by microtitre plates (Dynatech Laboratories, Denkendorf, Germany). Test plates with spore-forming fungi or streptomycetes were prepared as follows. Flasks with YM agar, which contained (g l−1) yeast extract 4, malt extract 10, glucose 4, pH 5·5, were inoculated with fungi or Streptomyces and incubated at 27 °C for 10 d. After sporulation, the spores were harvested using Tween-80-saline (0·1% Tween-80, 0·9% NaCl). The spores were then washed and resuspended in normal saline ; 200 ml of test medium (soft agar) was inoculated with 1 ml of a spore suspension (107 spore ml−1). Inhibition of Neurospora crassa 74 A mitochondria

Two 5 litre Erlenmeyer flasks containing 1 litre of medium were inoculated with spores of Neurospora crassa 74 A (108 spore ml−1). The inoculated flasks were incubated at 27 °C on a rotary shaker and rotated at 110 rev min−1 for 18–24 h. The culture was filtered and the residue (fungal mycelia) was washed with normal saline at 4 °C. The mycelia were suspended in 100 ml of homogenization medium (sucrose, 0·5 mol l−1 ; EDTA, 1 mmol l−1 ; TRIS, 50 mmol l−1 ; mercaptoethanol, 4 mmol l−1 ; pH 7·4 ; 4 °C), ground in mortar and pestle with quartz sand (40–100 mesh), and centrifuged at 2000 g for 10 min to remove unbroken cells and large cell fragments. The supernatant fraction was centrifuged at 16 000 g for 20 min to obtain a pellet ; the pellet was washed twice with a washing solution (sucrose, 0·5 mmol l−1 ; EDTA, 1 mmol l−1 ; TRIS, 50 mmol l−1 ; pH 7·4 ; 4 °C), and finally suspended in electrode medium (sucrose, 0·5 mmol l−1 ; TRIS, 50 mmol l−1 ; MgSO4, 2 mmol l−1 ; KH2PO4, 5 mmol l−1 ; BSA, 0·1% ; pH 7·4) (O.D.578 0·5 ; 1 : 10) (Wong et al. 1971). Electrode medium (400 ml) and 100 ml of mitochondrial suspension were placed in the measuring cell of the O2-electrode recording device. After 5 s, 20 ml NADH (10 mmol l−1 NADH in 300 mmol l−1 NaHCO3) were added. After a few more seconds, 20 ml of pyrrolnitrin (in methanol) at different concentrations were added.

lected from different ponds in the botanical garden, Tu¨bingen, yielded a number of non-fluorescent Pseudomonas isolates. The antifungal activity of these bacterial isolates was tested by the Petri plate assay against T. pseudokonigii. Those strains that inhibited the growth of T. pseudokonigii were then tested against other known phytopathogens. A zone of inhibition indicated antifungal activity. The most active strain (NB-1) was selected and used for further study. The isolate NB-1 was identified as B. cepacia with API 20 NE diagnostics strips. To determine and monitor the time course for the production of the antimicrobial compound under batch conditions, agar diffusion tests were employed. The antimicrobial activity was first detected after 24 h of incubation, corresponding to the late exponential phase, and continued to increase during stationary phase to reach maximal activity at 120 h (Fig. 1). Antibiotic production was greatly influenced by nutritional conditions. The carbon source has an important role in this respect. The 14 carbon sources used in this study were added to the production medium at a final concentration of 2% (w/v). There was a high degree of variation in the level of antifungal activity when different carbon sources were tested in the medium (Fig. 2). Glycerol was most effective in

Fig. 1 Time course of pyrrolnitrin production (Ž, inhibition zone ; ž, optical density)

RESULTS Screening of water samples for pseudomonads producing antifungal compounds

In the present study, emphasis was placed on the isolation of antifungal compounds from non-fluorescent pseudomonads. Screening of antimicrobial compounds from bacteria col-

Fig. 2 Effect of carbon sources on the antifungal activity of Burkholderia cepacia NB-1



increasing the antimicrobial activity, followed by glucose, mannose, sucrose and fructose. Ribose, maltose, arabinose and xylose were least effective. No antimicrobial activity was observed for rhamnose, starch, galactose, lactose and citrate. Several amino acids were added to the production medium (Table 1). Maximum antifungal activity was observed by addition of L-glutamic acid, followed by L-arginine, L-histidine and L-aspartic acid. The activity was substantially decreased by the addition of L-tryptophan, L-valine, L-serine, L-phenylalanine and L-cysteine. The isolation procedure of the antibiotic from cells grown in liquid cultures yielded 0·54 mg l−1 (Fig. 3). High performance liquid chromatography (HPLC) analysis of the active fraction yielded a peak with a retention time of approximately 27·5 min and similar to the standard pyrrolnitrin (Fig. 4).

Structure verification of pyrrolnitrin

The mass of the active antifungal fraction as determined by electrospray mass spectroscopy (negative ion spectrum) was 256 (Fig. 5). This molecular mass is consistent with the formula C10H6Cl2N2O2. The isotope ratio pointed to a chlorine-containing compound. The 1H-NMR spectrum (Fig. 6) showed distinct and well resolved signal groups in the aromatic range, which could be assigned to three aromatic protons and two pyrrole protons, and an additional broad N-H signal. The 13C-NMR spectrum (Fig. 7) was taken as a gated spin echo experiment showing upward CH and CH3, and downward CH2 and Cq signals. The spectrum revealed complete identity with the data reported by Martin et al. (1972) for pyrrolnitrin. Therefore, all spectroscopic data confirmed that the compound isolated from the strain B. cepacia NB-1 was indeed pyrrolnitrin (Fig. 8).

Table 1 Effect of amino acids on the production of pyrrolnitrin

of Burkholderia cepacia NB-1 expressed as antifungal activity — ––––––––––––––––––––––––––––––––––––––––––––––––––––– Amino acid (10 mmol l−1) Inhibition zone (mm) — ––––––––––––––––––––––––––––––––––––––––––––––––––––– L-Glutamic acid 30·5 L-Tyrosine 16 L-Serine 18·8 L-Cysteine 14·7 L-Valine 21·6 L-Aspartic acid 29·8 L-Arginine 30 L-Histidine 30 L-Tryptophan 23 L-Phenylalanine 15·8 — –––––––––––––––––––––––––––––––––––––––––––––––––––––

Fig. 3 Isolation process of pyrrolnitrin

Antimicrobial spectrum of pyrrolnitrin

Pyrrolnitrin produced by B. cepacia NB-1 exhibited a broad spectrum of activity against many fungi and bacteria, even at relatively low concentrations, depending on the organism tested (Table 2). The antimicrobial effect of pyrrolnitrin (agar diffusion test) was most pronounced against Streptomyces antibioticus, S. violaceoruber, Paecilomyces variotii and Penicillium puberulum. Some other species, such as S. prasinus, S. ramulosus, Aspergillus proliferans and A. terreus, were comparatively tolerant to pyrrolnitrin. The antimicrobial effect of pyrrolnitrin, tested in a microtitre plate assay, was most pronounced against Ustilago maydis and Candida albicans. Hansenula anomala, Arthrobacter oxidans, Bacillus coagulans, B. lichenifernis, B. subtilis and B. thuringiesis were inhibited at medium concentrations of pyrrolnitrin (6·25 mg ml−1), whereas B. megaterium, B. polymyxa, B. pumilus, Corynebacterium glutamicum, Micrococcus luteus, Staphylococcus aureas, Staph. saprophyticus and Streptococcus faecalis were inhibited at higher concentrations. All tested Gram-negative bacteria were resistant to the doses tested except Proteus vulgaris which was inhibited at 25 mg ml−1 pyrrolnitrin. Effect of pyrrolnitrin on the electron transport chain

To confirm the effect of pyrrolnitrin on the respiratory chain, several experiments with mitochondria from Neurospora



revealed that these colonies represent possible mutant strains of S. antibioticus with altered growth characteristics which made them unable to form either aerial mycelia or spores. DISCUSSION

Fig. 4 HPLC analysis of (a) isolated pyrrolnitrin from

Burkholderia cepacia NB-1 and (b) standard pyrrolnitrin (gift from Prof. K.-H. van Peé)

crassa 74 A were carried out. Addition of inhibitors of the electron transport chain inhibits oxygen consumption and its concentration remains constant. Figure 9 shows the effect of pyrrolnitrin on oxygen consumption by mitochondria of N. crassa 74 A and the resumed respiration after addition of N,N,N?,N?-tetramethyl-p-phenylenediamine dihydrochloride (TMPD). Pyrrolnitrin at a concentration of 0·11 mg ml−1 inhibited the electron transport chain (oxygen consumption was stopped), and this inhibition was relieved by TMPD, which acts as a shunt between the initial segment of the electron transport chain and cytochrome C1 (oxygen consumption started again). Effect of pyrrolnitrin on growth of Streptomyces antibioticus

Agar diffusion tests showed that the minimal concentration of pyrrolnitrin needed to inhibit the growth of S. antibioticus was 0·2 mg ml−1. The test plates were then incubated further. After 5 d of incubation, some colonies appeared to grow around the filter disc within the inhibition zone but inside the agar (not on the surface). Microscopic examination

In recent years, there has been considerable interest in using B. cepacia as a biocontrol agent because of its ability to antagonize and repress soil-borne plant pathogens (Janisiewicz and Roitman 1988). Other biotypes of B. cepacia have been implicated in human diseases (Thomassen et al. 1985), plant disease (Burkholder 1950) and biodegration of pesticides (Karns et al. 1983). Various factors responsible for the antagonistic properties of B. cepacia have been suggested (Weller 1988). The mechanisms by which B. cepacia restricts the growth of phytopathogenic fungi are most likely to be the production of antifungal agents (Homma and Suzui 1989). Furthermore, the production of siderophores (Winkelmann and Drechsel 1997), and aggressive root colonization by organisms that displace or exclude deleterious rhizosphere micro-organisms, seem to be involved in biocontrol (Homma and Suzui 1989). Variations in the fermentation environment often result in an alteration in antibiotic production. The alteration involves changes in both yield and composition of the compound (Upadhyay et al. 1991). Roitman et al. (1990) reported that by varying the conditions under which B. cepacia is grown, the yield and composition of the phenylpyrrole metabolites could be changed. Antibiotic production of B. cepacia NB-1 was greatly influenced by nutritional and environmental factors. Glycerol strongly enhanced the antifungal activity, whereas glucose, mannose and fructose decreased it. Galactose, lactose, rhamnose and starch repressed the production of pyrrolnitrin. Frequently, antibiotics are produced only after completion of the growth phase. The synthesis of antibiotics is often repressed by substances that favour rapid cellular growth, such as glucose (catabolite repression) and ammonium ions (nitrogen repression). When the level of these nutrients is low, the rate of cell growth is slowed and antibiotic synthesis is derepressed (Chater and Bibb 1997). With single amino acids, there was wide variation in the antifungal activity of B. cepacia NB-1. On the basis of their effect on antagonism, L-glutamic acid had the strongest impact of all amino acids added to the fermentation medium on the antifungal activity of B. cepacia NB-1, followed by Larginine, L-histidine and L-aspartatic acid ; L-tryptophan had no effect in increasing pyrrolnitrin production.D-tryptophan is a specific and direct precursor of pyrrolnitrin. Its addition to the culture medium increased pyrrolnitrin production. Ltryptophan enters the cell very quickly and is used extensively for protein synthesis before antibiotic production starts, whereas more of the D-isomer, which enters the cell more slowly, is still available during the period of pyrrolnitrin



Fig. 5 Mass spectrum of pyrrolnitrin

Fig. 6 1H-NMR spectrum of pyrrolnitrin

synthesis (Hamill et al. 1967). The addition of L-tryptophan did not increase production of pyrrolnitrin (Lively et al. 1967). The addition of a specific D-amino acid usually inhibited the synthesis of antibiotic, e.g. D-tryptophan in

actinomycin (Albertini et al. 1996). The inhibition was specifically reversed by addition of L-tryptophan (Lively et al. 1967). Burkholderia cepacia NB-1 seem to accumulate pyrrolnitrin



Fig. 7

13

C-NMR spectrum of pyrrolnitrin

Fig. 8 Structure of pyrrolnitrin

within their cells. Under the test conditions used, pyrrolnitrin accumulated in the stationary phase in laboratory media, reaching a maximum at 120 h. Roitman et al. (1990) reported that 98% of phenylpyrroles are contained in the cell extracts and that the broth contains only 1% of the pyrroles produced by the cells during fermentation. The present data clearly establish that the production of pyrrolnitrin by B. cepacia NB-1 is not only under genetic regulation, but is also greatly influenced by physiological and environmental conditions. HPLC profile of the isolated pyrrolnitrin from B. cepacia NB-1 revealed a retention time (27·5 min) identical to that of a standard sample of pyrrolnitrin. The molecular ion at m/z 256, determined by electrospray mass spectroscopy (negative mode), is consistent with the formula C10H6Cl2N2O2 of pyrrolnitrin. The isotopic pattern of the signal cluster with a distinct peak at [M-H]− (255) indicated chlorine. Simulation of different assumed formulae with one or more chlorine atoms was most similar for two atoms of chlorine. As well as the pyrrole NH group at high

ppm values, distinct and well developed signal groups in the aromatic range were seen in the 1H-NMR spectrum. One unresolved doublet at 6·8 ppm could be assigned to the pyrrole protons ; the others have been assigned to the three aromatic protons. 1H-NMR analysis was characterized by chemical shifts as described previously by Martin et al. (1972). The 13C-NMR spectrum was recorded as a gated spin echo experiment. The three signals at high ppm correspond to the three aromatic CH atoms, while the smaller ones could be assigned to the pyrrole CH atoms. Four quaternary signals are seen downward oriented. The 13C-NMR spectrum exposed data similar to those reported for pyrrolnitrin (Martin et al. 1972). According to the HPLC, MS and NMR data of the present study, the antibiotic produced by B. cepacia NB-1 is pyrrolnitrin, 3-chloro-4-(2?-nitro-3?-chloro-phenyl) pyrrole. Pyrrolnitrin, first isolated from Pseudomonas pyrrocinia by Arima et al. (1964) has a 33 year history of recorded antifungal activity. After the original report on the isolation of pyrrolnitrin, numerous reports appeared on pyrrolnitrin production by other pseudomonads (Pseudomonas aeruginosa, Ps. cepacia and Ps. fluorescens) and demonstrated that pyrrolnitrin was biosynthesized from tryptophan. Gerth et al. (1982) and Chernin et al. (1996) reported evidence that pyrrolnitrin production is not restricted to the genus Pseudomonas, but is also produced by Myxococcus fulvus and E. agglomerans, respectively. The antimicrobial spectrum of pyrrolnitrin exhibited a broad spectrum of activity against phytopathogenic fungi, yeasts and Gram-positive bacteria, with a particular activity against Streptomyces, whereas Gram-negative bacteria, except Proteus vulgaris, were resistant. Orig-



Table 2 Antimicrobial spectrum of pyrrolnitrin

— ––––––––––––––––––––––––––––––––––––––––––––––––––––– MIC Test micro-organism (mg ml−1)* — ––––––––––––––––––––––––––––––––––––––––––––––––––––– Aspergillus proliferans stock of the Institute w.a.† Aspergillus terreus Tu¨ 155 w.a. Mucor miehei Tu¨ 284 6 Paecilomyces variotii Tu¨ 137 2 Penicillium notatum Tu¨ 136 6 Penicillium puberulum stock of the Institute 2 Pythium ultimum stock of the Institute 6 Trichoderma pseudokonigii CBS 931.96 10 Candida albicans Tu¨ 164 3·12 Hansenula anomala Tu¨ 279 6·25 Saccharomyces cerevisiae Tu¨ 125 12·5 Ustilago maydis 001 Prof. S.A. Leong, Madison, 0·78 USA Arthrobacter oxydans ATCC 14358 6·25 Bacillus coagulans ATCC 7050 6·25 Bacillus licheniformis ATCC 14580 6·25 Bacillus megaterium ATCC 14581 25 Bacillus polymyxa ATCC 842 25 Bacillus pumilus ATCC 7061 25 Bacillus subtilis ATCC 6051 6·25 Bacillus thuringiensis ATCC 10792 6·25 Corynebacterium glutamicum ATCC 13745 12·5 Micrococcus luteus ATCC 381 12·5 Rhodococcus fasciens ATCC 12975 12·5 Staphylococcus aureus ATCC 381 12·5 Staphylococcus saprophyticus ATCC 15305 12·5 Streptococcus faecalis ATCC 29212 12·5 Enterobacter cloacae ATCC 13047 n.a.‡ Escherichia coli K 12 n.a. Proteus vulgaris ATCC 13315 25 Pseudomonas fluorescens ATCC 13525 n.a. Streptomyces antibioticus Tu¨ 1998 0·2 Streptomyces fradiae Tu¨ 11 200 Streptomyces griseoflavus Tu¨ 2038 0.6 Streptomyces olivaceus Tu¨ 2108 0·6 Streptomyces prasinus Tu¨ 30 200 Streptomyces ramulosus Tu¨ 34 200 Streptomyces violaceoruber Tu¨ 22 0·2 Streptomyces viridochromogenes Tu¨ 57 0·3 — ––––––––––––––––––––––––––––––––––––––––––––––––––––– * Agar diffusion test (against fungi and streptomycetes). Microtitre plate (against bacteria and yeasts). † Weak activity without halo. ‡ No activity.

inally, it was suggested that pyrrolnitrin may be useful against dermatophytic fungi, especially Trichophyton and Microsporum, and had low toxicity (Wong et al. 1971). It also exhibited activity against Aspergillus, B. subtilis and Proteus vulgaris (Arima et al. 1965). Compounds related to pyrrol-

Fig. 9 Oxygen electrode recording showing the effect of

pyrrolnitrin on O2 consumption by mitochondria isolated from a wild type strain of Neurospora crassa 74 A

nitrin, namely isopyrrolnitrin, oxypyrrolnitrin and monochloropyrrolnitrin, were reported to have lower antifungal activity than pyrrolnitrin (Leisinger and Margraff 1979). In initial experiments with N. crassa mitochondria, pyrrolnitrin had an inhibitory effect on the electron transport system. This inhibition was relieved by TMPD, which acts as a shunt between the initial segment of electron transport chain and cytochrome C1. The inhibitors of NADH dehydrogenase, succinate dehydrogenase and cytochrome C reductase were circumvented by TMPD, which means that pyrrolnitrin probably blocked electron transfer between the dehydrogenases and the cytochrome components of the respiratory chain. Pyrrolnitrin was also thought to inhibit bacterial growth by complexing with phospholipids of cell membranes. Although pyrrolnitrin inhibited the respiration of intact cells, the oxidative phosphorylation of mitochondria isolated from Candida utilis was not inhibited (Nose and Arima 1969). Tripathi and Gottlieb (1969) concluded that inhibition of electron transfer in a yeast was the site of action of pyrrolnitrin as an antibiotic. The latter conclusion was supported by a study by Wong and Airall (1970) who demonstrated similar inhibition by pyrrolnitrin on the mammalian respiratory system. The present study clearly shows that pyrrolnitrin inhibits respiration of fungal mitochondria. This is similar to the results of Tripathi and Gottlieb (1969), Wong and Airall (1970) and Lambowitz and Slayman (1972). Lambowitz and Slayman (1972) considered the primary effect of pyrrolnitrin on Neurospora mitochondria to be the uncoupling of oxidative phosphorylation. Complete inhibition of electron transport requires a higher concentration. It is confirmed here that the site of action of pyrrolnitrin is in the respiratory electron transport system.



In this investigation, Streptomyces species, including S. antibioticus and S. violaceoruber have been shown to be inhibited by pyrrolnitrin at a very low concentration. This is thought to be the first report which describes activity against Streptomyces species. During the activity tests of pyrrolnitrin against streptomycetes, resistant variants of S. antibioticus could be observed. It is assumed that these variants possess altered respiratory chain components which are insensitive to pyrrolnitrin and prevent the formation of aerial mycelia and spores but still allow growth of substrate mycelia. Inhibition of A-factor production by pyrrolnitrin might also be involved. Streptomycetes are known to produce derivatives of g-butyrolactone known as A-factor [2-(6?-methylheptanoyl)-3-hydroxymethyl-4-butanolide] (Khokhlov 1982). A-factor is needed as an endogenous regulatory molecule for normal cytodifferentiation of some strains of S. griseus (Chater and Bibb 1997). Therefore, mutant strains incapable of producing A-factor do not form both aerial mycelium and antibiotics such as streptomycin or daunomycin, but respond to the addition of a low amount of these autoregulators to the growth medium with full reconstitution of sporulation and/or secondary metabolism. A-factor and its relatives expresses its regulatory activities through binding with A-factor-binding protein. A mutant strain of S. griseus that requires no Afactor for streptomycin production or sporulation was found to have a defect in the A-factor-binding protein. This observation implied that the A-factor-binding protein, in the absence of A-factor, repressed both phenotypes in the wild type strain. Reversal of the defect in the A-factor-binding protein of this mutant led to the simultaneous loss of streptomycin production and sporulation (Miyake et al. 1990). These data suggest that the A-factor-binding protein plays a role in repressing both streptomycin production and sporulation, and that the binding of A-factor to the protein releases its repression. The new findings of inhibition of streptomyces and their secondary metabolism by pyrrolnitrin may contribute to the fact that Pseudomonas species predominate in soil and compete even with antibiotic-producing Streptomyces. ACKNOWLEDGEMENTS

The authors thank Prof. K.-H. van Peé for providing a sample of pyrrolnitrin. Thanks are also due to Dr H. Drechsel for his help in the structure elucidation of isolated pyrrolnitrin. This work was supported by the German Academic Exchange Service to N.E-B. REFERENCES Albertini, A., Tiboni, O. and Giferri, O. (1966) Relation between tryptophan configuration and actinomycin biosynthesis. Journal of Labelled Compounds 2, 90–101.

Arima, K., Imanaka, H., Kousaka, M., Fukuta, A. and Tamura, G. (1964) Pyrrolnitrin, a new antibiotic substance, produced by Pseudomonas. Agricultural and Biological Chemistry 28, 275–276. Arima, K., Imanaka, H., Kousaka, M., Fukuda, A. and Tamura, G. (1965) Studies on pyrrolnitrin, a new antibiotic. I. Isolation and properties of pyrrolnitrin. Journal of Antibiotics 18, 201–204. Bull, C.T., Weller, D.M. and Thomashow, L.S. (1991) Relationship between root colonization and suppression of Gaeumannomyces graminis var. tritici by Pseudomonas fluorescens strain 2-79. Phytopathology 81, 950–959. Burkholder, W.H. (1950) Sour skin, a bacterial rot of onion bulbs. Phytopathology 40, 115–117. Chater, K.F. and Bibb, M.J. (1997) Regulation of bacterial antibiotic production. In Biotechnology 2nd edn Vol. 7, Products of Secondary Metabolism, ed. Rhem, H.-J. and Reed, G., pp. 57–105. Weinheim : VCH. Chernin, L., Brandis, A., Ismailov, Z. and Chet, I. (1996) Pyrrolnitrin production by an Enterobacter agglomerans strain with a broad spectrum of antagonistic activity towards fungal and bacterial phytopathogens. Current Microbiology 32, 205–212. David, N.D. and O’Gara, F. (1994) Metabolites of Pseudomonas involved in the biocontrol of plant disease. TIBTECH, April 12, 133–141. Gerth, K., Trowitzsch, W., Wray, V., Ho¨fle, G., Irschik, H. and Reichenbach, H. (1982) Pyrrolnitrin from Myxococcus fulvus (Myxobacterales). Journal of Antibiotics 35, 1101–1103. Hamill, R., Elander, R., Mabe, J. and Gorman, M. (1967) Metabolism of tryptophan by Pseudomonas aureofaciens. V. Conversion of tryptophan to pyrrolnitrin. Antimicrobial Agents and Chemotherapy 1967, 388–396. Homma, Y. and Suzui, T. (1989) Role of antibiotic production in suppression of radish damping-off by seed bacterization with Pseudomonas cepacia. Annuals of the Phytopathological Society Japan 55, 643–652. Janisiewicz, W. and Roitman, J. (1988) Biological control of blue mold and gray mold on apple and pear with Pseudomonas cepacia. Phytopathology 78, 1697–1700. Jiao, Y., Yoshihara, T., Ishikuri, S., Uchino, H. and Ichihara, A. (1996) Structural identification of cepaciamide A, a novel fungitoxic compound from Pseudomonas cepacia D-202. Tetrahedron Letters 37, 1039–1042. Karns, J.S., Dattagupta, S. and Chakrabarty, A.M. (1983) Regulation of 2,4,5-trichloro-phenoxyacetic acid and chlorophenyl metabolism in Pseudomonas cepacia AC 1100. Applied and Environmental Microbiology 46, 1182–1186. Khokhlov, A.S. (1982) Low molecular weight microbial bioregulators of secondary metabolism. In Overproduction of Microbial Products, ed. Krumphanzl, V. et al., pp. 96–109. New York : Academic Press. Lambowitz, A.M. and Slayman, C.W. (1972) Effect of pyrrolnitrin on electron transport and oxidative phosphorylation in mitochondria isolated from Neurospora crassa. Journal of Bacteriology 112, 1020–1022. Lee, G.H., Kim, S., Hyun, B. and Suh, J.-W. (1994) Cepacidine A, a novel antifungal antibiotic produced by Pseudomonas cepacia. I. Taxonomy, production, isolation and biological activity. Journal of Antibiotics 47, 1402–1405.



Leisinger, T. and Margraff, R. (1979) Secondary metabolites of the fluorescent pseudomonads. Microbiological Reviews 43, 422–442. Lively, D., Gorman, M., Haney, M. and Mabe, J. (1967) Metabolism of tryptophans by Pseudomonas aureofaciens. I. Biosynthesis of pyrrolnitrin. Antimicrobial Agents and Chemotherapy 1966, 462– 469. Martin, L.L., Chang, C.-J., Floss, H.G., Mabe, J.A., Hagman, E.W. and Wenkert, E. (1972) A 13C nuclear magnetic resonance study on the biosynthesis of pyrrolnitrin from tryptophan by Pseudomonas. Journal of the American Chemical Society 94, 8942–8944. Meyers, E., Bisacchi, G.S., Dean, L. et al. (1987) Xylocandin : a new complex of antifungal peptide. I. Taxonomy, isolation and biological activity. Journal of Antibiotics 21, 1515–1519. Miyake, K., Kuzuyama, T., Horinouchi, S. and Beppu, T. (1990) The A-factor-binding protein of Streptomyces griseus negatively controls streptomycin production and sporulation. Journal of Bacteriology 172, 3003–3008. Nose, M. and Arima, K. (1969) On the mode of action of a new antibiotic, pyrrolnitrin. Journal of Antibiotics 22, 135–143. Parker, W.L., Rathnum, M.L., Seiner, V., Trejo, W.H., Principe, P.A. and Sykes, R.B. (1984) Cepacin A and cepacin B, two antibiotics produced by Pseudomonas cepacia. Journal of Antibiotics 37, 431–440. Roitman, J.N., Mahoney, N.E. and Janisiewicz, W.J. (1990) Production and composition of phenylpyrrole metabolites produced by Pseudomonas cepacia. Applied Microbiology and Biotechnology 34, 381–386. Thomassen, M.J., Denko, C.A., Klinger, J.D. and Stern, R.C. (1985) Pseudomonas cepacia colonization among patients with cys-

tic fibrosis. A new opportunist. American Review of Respiratory Diseases 113, 791–796. Tripathi, R. and Gottlieb, D. (1969) Mechanism of action of the antifungal antibiotic pyrrolnitrin. Journal of Bacteriology 100, 310–318. Upadhyay, R., Visintin, L. and Jayaswal, R. (1991) Environmental factor affecting antagonisms of Pseudomonas cepacia against Trichoderma viride. Canadian Journal of Microbiology 37, 880–884. van Peé, K.-H., Salcher, O. and Lingens, F. (1980) Bildung von Pyrrolnitrin and 3-(2-Amino-3-chlorphenyl) pyrrol aus 7-chlortryptophan. Angewandte Chemie 92, 855–856. Weller, D.M. (1988) Biological control of soilborne pathogens in the rhizosphere with bacteria. Annual Review of Phytopathology 62, 379–407. Winkelmann, G. and Drechsel, H. (1997) Microbial siderophores. In Biotechnology 2nd edn Vol. 7, Products of Secondary Metabolism, ed. Rehm, J.-J. and Reed, G., pp. 199–246. Weinheim : VCH. Wong, D.T. and Airall, J.M. (1970) The mode of action of antifungal agents : effect of pyrrolnitrin on mitochondrial electron transport. Journal of Antibiotics 23, 55–62. Wong, D.T., Horng, J.-S. and Gordee, R.S. (1971) Respiratory chain of a pathogenic fungus, Microsporum gypseum : effect of the antifungal agent pyrrolnitrin. Journal of Bacteriology 106, 168– 173. Yabuuchi, E., Kosako, Y., Oyaizu, H. et al. (1992) Proposal of Burkholderia gen. nov. and transfer of seven species of the genus Pseudomonas homology group II to a new genus, with the type species Burkholderia cepacia (Palleroni and Homes 1981) comb. nov. Microbiology and Immunology 36, 1251–1275.
