• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of aemPermissionsJournals.ASM.orgJournalAEM ArticleJournal InfoAuthorsReviewers
Appl Environ Microbiol. Feb 2007; 73(3): 793–797.
Published online Nov 22, 2006. doi:  10.1128/AEM.01784-06
PMCID: PMC1800748

Rhizonin, the First Mycotoxin Isolated from the Zygomycota, Is Not a Fungal Metabolite but Is Produced by Bacterial Endosymbionts[down-pointing small open triangle]


Rhizonin is a hepatotoxic cyclopeptide isolated from cultures of a fungal Rhizopus microsporus strain that grew on moldy ground nuts in Mozambique. Reinvestigation of this fungal strain by a series of experiments unequivocally revealed that this “first mycotoxin from lower fungi” is actually not produced by the fungus. PCR experiments and phylogenetic studies based on 16S rRNA gene sequences revealed that the fungus is associated with bacteria belonging to the genus Burkholderia. By transmission electron microscopy, the bacteria were localized within the fungal cytosol. Toxin production and the presence of the endosymbionts were correlated by curing the fungus with an antibiotic, yielding a nonproducing, symbiont-free phenotype. The final evidence for a bacterial biogenesis of the toxin was obtained by the successful fermentation of the endosymbiotic bacteria in pure culture and isolation of rhizonin A from the broth. This finding is of particular interest since Rhizopus microsporus and related Rhizopus species are frequently used in food preparations such as tempeh and sufu.

The danger posed by toxins from fungi has a long history. The true causes of mycotoxicoses, however, have long been overlooked. Even so, the identification of mycotoxins responsible for mass intoxification of animals and humans was an epoch-making discovery in the last century (3). The epidemic Turkey “X” disease and the historical and modern cases of ergotism (St. Anthony's Fire) are infamous examples of mycotoxicoses, underscoring the global impact of food spoilage by fungi (3). Even low concentrations of mycotoxins in food can elicit deleterious effects such as chronic or acute toxic damage of liver and kidney. Since many health-threatening fungal metabolites are resistant to heat and degradation, it is imperative to identify the toxin producers and to control their growth in order to avoid the accumulation of harmful substances in the food chain.

The vast majority of known mycotoxin-producing fungi belong to the class Ascomycota. Only very few fungal species of the Zygomycota (“lower fungi”), particularly Rhizopus species, have been implicated in the spoilage of foods, which comprise fruits, nuts, and vegetables. Most importantly, Rhizopus spp. also ruin sorghum grain, the second most important nutrient in warm countries apart from rice (9). About two decades ago, the first mycotoxins from the Zygomycota, rhizonins A and B (1 and 2) (Fig. (Fig.1),1), were identified from a highly toxinogenic Rhizopus microsporus strain (MRC 303), which feasted on Mozambican groundnuts (15, 19). In animal tests, the cyclic heptapeptides exhibit severe nonspecific hepatotoxicity; they cause a wide range of hepatic lesions and induce acute and chronic failure of the liver, which resulted in 100% lethality (23). This alarming finding is of particular interest, not only because of the danger of intoxification through food spoilage but also because closely related Rhizopus species are used as enzymatic sources for the production of fermented foods such as tempeh and sorghum (5, 6, 9, 10, 12, 20) However, the detection of these mycotoxins in food by standard analytical methods is not a trivial issue. We thus aimed to elucidate the molecular basis for rhizonin biosynthesis in Rhizopus microsporus, which could strongly contribute to monitoring the presence of toxin producers. In this study, we demonstrate that rhizonin is in fact not produced by the fungus but is produced by bacteria that reside within the fungal cytosol.

FIG. 1.
Structures of hepatotoxic cyclopeptides 1 and 2 from R. microsporus.


Fungal strain and cultivation conditions.

Rhizopus microsporus van Tieghem (strain cbs112285) is the original strain isolated from moldy groundnuts in Mozambique (19). Liquid cultivation of this fungal strain was carried out for both extract analysis and DNA isolation in 500-ml Erlenmeyer flasks containing 100 ml of production medium composed of 1% corn starch, 0.5% glycerol, 1% gluten meal, 1% dried yeast, 1% corn steep liquor, and 1% CaCO3 at pH 6.5. Fermentations were performed for 2, 4, and up to 21 days at 30°C with orbital shaking at 80 rpm.

Analysis of metabolites.

Mycelium obtained from liquid fermentations was harvested and exhaustively extracted overnight with ethyl acetate. The organic phase was filtered, dried with sodium sulfate, and concentrated under reduced pressure. The residue was dissolved in methanol. High-performance liquid chromatography (HPLC) analysis was carried out as reported previously (17). Rhizonins A and B were monitored using a photodiode array detector and by electrospray ionization (ESI) in-source collision-induced dissociation tandem mass spectrometry (MSn) on a Finnigan LCQ bench-top mass spectrometer equipped with an electrospray ion source and ion trap mass analyzer.

DNA isolation, PCR, cloning, and sequencing.

Metagenomic DNA from R. microsporus (cbs112285) was obtained using a method reported previously (14). Bacterial DNA was recovered using a protocol described previously by Mutasa et al. (13). Plasmid DNA was isolated routinely by an alkaline lysis miniprep procedure (16). The 16S rRNA gene was amplified using universal primers as described previously (14). Amplification of nonribosomal peptide synthetase (NRPS) gene fragments coding for adenylation domains from both fungal and bacterial DNA was conducted using degenerate primers 6R (SGAGTGSCCSCCSAGCTCGAA) and 5F (ATCGAGCTSGGSGAGATCGAG) (7). Genomic DNA (50 to 250 ng) was used in all PCRs containing PCR buffer [10 mM Tris-Cl, KCl, (NH4)2SO4, 15 mM MgCl2, pH 8.7], forward and reverse primers (1 to 2 μM each), deoxynucleoside triphosphates, Q solution (QIAGEN), and Taq polymerase in a total volume of 50 μl. PCR conditions for primers 6R and 5F were 30 cycles at 94°C for 60 s, 65°C for 90 s, and 72°C for 30 s and a final extension step for 7 min at 72°C. Amplified DNA fragments of the expected size (about 1.5 kb for 16S rRNA and 0.5 kb for NRPS gene amplicons) were purified using GFX PCR DNA and a Gel Band Purification kit (Amersham Biosciences) and analyzed by agarose gel electrophoresis. The amplicons were individually cloned into the pGEM-T Easy vector (Promega) for sequencing. The sequencing was performed using an ABI PRISM 3100 Genetic Analyser.

Generation of a symbiont-free fungal strain.

For the generation of a symbiont-free fungal strain, R. microsporus was cultivated in the presence of ciprofloxacin (Bayer AG, Germany) at a concentration of 0.02 to 0.06 μg ml−1.

Localization of bacterial endosymbionts by electron microscopy.

A small mycelium pellet of a 3-day-old culture of R. microsporus van Tieghem cbs112285 was aseptically transferred into 0.5 ml of saline solution (0.85% NaCl). The pellet was washed once and then resuspended again in the saline buffer. Isolated bacterial endosymbionts (“Candidatus Burkholderia rhizoxina” B5) were cultivated in Bacto tryptic soy broth (TSB; Difco) for 2 days at 30°C and 150 rpm. The bacterial culture was concentrated by centrifugation and later washed and resuspended in saline solution. Samples were washed in phosphate-buffered saline (PBS) (pH 7.0) and fixed in 2% KMnO4-PBS (pH 7.2) for 2 h at 20°C. After a briefly washing in PBS, the sediment was embedded in 3% agar-PBS to facilitate handling. The postfixation step was performed with 1% OsO4-PBS in the dark at 4°C. The well-washed samples were dehydrated in an ascending water-acetone series and embedded in AGAR 100 epoxy resin (AGAR SCIENTIFIC). For polymerization, the resin capsules were kept at 60°C for 2 days. Sections were cut using an ultramicrotome (REICHERT ULTRACUT S; LEICA) and stained with uranyl acetate and lead citrate. The samples were examined with a transmission electron microscope (acceleration voltage of 80 kV) (CEM 902A; Zeiss).

Isolation of bacterial endosymbionts.

A small mycelial pellet was aseptically taken from a 2-day-old liquid fungal culture and submerged in 500 μl 0.85% NaCl. Using mechanical stress (pipetting), the mycelia were broken and then subjected to centrifugation for 30 min at 13,200 rpm. A loop of supernatant was plated onto nutrient agar plates and incubated at 30°C until the presence of bacteria and/or fungi could be observed. Bacterial colonies were then cultured in TSB medium and preserved in glycerol at −80°C.

Fermentation of bacterial endosymbiont strain B5.

From the B5 bacterial cell bank, an 800-ml seed culture was prepared in TSB medium. This preculture was used to inoculate 20 liters of sterile production medium composed of 1% corn starch, 0.5% glycerol, 1% gluten meal, 1% dried yeast, 1% corn steep liquor, and 1% CaCO3; the pH was adjusted to 6.5 after sterilization. The fermentation was carried out at 30°C for 36 h in a 30-liter aerated tank fermenter.

Rhizonin A isolation and characterization.

Broth and biomass of a 36-h-old fermentation of cultured symbiont B5 (20 liters) were exhaustively extracted with ethyl acetate. The crude extract (11.2 g) was subjected to open-column chromatography (ODS-A 60-S50, 5 by 24 cm) using an aqueous MeCN gradient. The enriched fractions were further purified by chromatography on a silica gel column (Kieselgel 60, 0.040 to 0.063 mm, 14.5 by 160 mm, dichloromethane-MeOH; MERCK) and by repeated reverse-phase HPLC on a Eurospher 100/5-C18 column (20 by 250 mm) using a 42% to 83% MeCN gradient at a flow rate of 10 ml min−1 and with UV detection (220 nm). The yield was 4.9 mg rhizonin A. High-resolution ESI-MS, m/z [M + Na]+ = 834.4679 (calculated C42H65O9O723Na 834.4741). All physicochemical data are in full agreement with those reported previously (Table (Table1)1) (19).

1H and 13C NMR data for rhizonin A in CDCl3 at 303 K

Nucleotide sequence accession number.

The nucleotide sequence has been deposited in the EMBL database under accession number AM420302.


To establish the molecular basis of rhizonin formation, the authentic R. microsporus van Tieghem strain from Mozambique (cbs112285) was reinvestigated. This fungus was reported previously by Jennessen et al. as being the only rhizonin-producing species out of 14 different Rhizopus spp. (four R. microsporus var. microsporus, four R. microsporus var. chinensis, and six R. microsporus var. oligosporus isolates) (9).

The formation of rhizonin A in the R. microsporus culture could be monitored by HPLC-MS albeit only after an initial purification of the broth extract (Fig. (Fig.2).2). Comparison with an authentic sample of rhizonin A and the MSn fragmentation pattern proved the presence of the toxin. The structure of rhizonin suggests that its biosynthesis affords an NRPS, similar to enzymes that are involved in the formation of structurally related fungal cyclopeptides, e.g., the immunosuppressive cyclosporine (8, 21, 22). Typically, NRPSs represent giant multidomain enzymes with a modular architecture. Each module is composed of adenylation (A) and condensation (C) domains and a peptidyl carrier protein. In particular, A and C domains contain sequence motifs that are sufficiently conserved to allow the design of specific degenerate primers and the detection of NRPS gene fragments by PCR (2, 18). However, we failed to detect any fungal NRPS genes in the Rhizopus, which immediately prompted us to doubt the fungal biosynthetic origin of rhizonin. Only recently were we able to demonstrate that the polyketide metabolite rhizoxin, the causal agent of rice seedling blight, is not produced by the plant-pathogenic fungus Rhizopus microsporus, as has been believed for over two decades. The true producers are in fact bacterial symbionts of the genus Burkholderia that reside within the mycelium (14). Is the cyclopeptide rhizonin, the first reported mycotoxin from the Zygomycota, also produced by bacterial endosymbionts? To test this hypothesis, a number of experiments addressing potential symbiotic bacteria were conducted.

FIG. 2.
HPLC profiles of extracts (detection at 220 nm) from (a) genuine rhizonin-positive R. microsporus van Tieghem (cbs112285), (b) the same strain treated with antibiotic (symbiont free), (c) cultured endosymbionts of R. microsporus van Tieghem (cbs112285), ...

First, the presence of bacteria in the rhizonin-positive Rhizopus strain was unequivocally proven by the detection of bacterial small-subunit (16S) rRNA gene sequences by PCR. Amplification of the 1.5-kb 16S rRNA gene region using universal primers revealed that the rhizonin-positive Rhizopus strain is associated with a single bacterial strain that belongs to the genus Burkholderia (Fig. (Fig.3).3). Sequencing and phylogenetic analyses showed that the symbiont is closely related to the previously identified Rhizopus endosymbionts, which all fall into a single distinct clade (Fig. (Fig.44).

FIG. 3.
Results of PCR experiments. Agarose (2%) gel electrophoresis of (left) 16S rRNA gene and (right) NRPS PCR products amplified with universal or NRPS primers, respectively. Lane 1, R. microsporus total DNA; lane 2, DNA from cured R. microsporus; lane 3, ...
FIG. 4.
Phylogenetic tree based on the 16S rRNA gene sequences of R. microsporus symbionts and selected closely related Burkholderia species. The neighbor-joining bootstrap values depicted represent the percent support of the nodes. The tree was rooted using ...

To localize the associated bacteria, the mycelium was fixed in a matrix, cut using an ultramicrotome, and stained with uranyl acetate and lead citrate. Examination of the samples with a transmission electron microscope unequivocally proved that the bacteria are true endosymbionts, as transverse sections of bacterial rods were clearly visible (Fig. (Fig.5).5). Bacteria residing within the cytosol could be distinguished from fungal organelles by their membranes, morphology, and reserve materials. To test whether these bacterial endosymbionts are involved in cyclopeptide biosynthesis, we aimed to cure the fungal strain. For this purpose, the fungal culture was treated with a series of antibiotics. Ciprofloxacin proved to be most efficient in generating a symbiont-free fungal strain. Comparisons of the metabolic profiles of untreated fungi with those grown in the presence of ciprofloxacin by HPLC gave a clear result. While cyclopeptide formation could be monitored in the wild type, rigorously no rhizonin could be detected in the treated culture (Fig. (Fig.3).3). The lack of 16S rRNA genes demonstrated by PCR leads to the conclusion that the presence of bacteria is essential for “mycotoxin” production.

FIG. 5.
Transmission electron micrograph of fungal mycelium of R. microsporus van Tieghem (cbs112285) containing endosymbionts and isolated endosymbionts in pure culture. (a) Sections of endosymbionts (*) residing in R. microsporus mycelium. (b) Longitudinal ...

The final evidence for a bacterial involvement in rhizonin biosynthesis was obtained by investigating the symbionts without their fungal hosts. By mechanic disruption of mycelium and centrifugation using our previously reported protocol, we succeeded in isolating the bacterial symbionts in pure culture (14, 17). 16S rRNA gene analysis established the identity of the isolated strain. HPLC-MS analysis of the extract from the pure symbiont culture and detection of rhizonin finally disclosed that the bacterial symbionts, and not the fungi, are the true producers of the “mycotoxin.” This result was supported by the isolation of pure rhizonin A (4.9 mg) from a scaled-up fermentation (20 liters) of the cultured endosymbiont strain. The identity of the cyclopeptide was fully established by MSn, high-resolution MS, and NMR analyses, which were in full agreement with previously published data (19) (Table (Table1).1). It should be noted that the quantity of toxin produced by the cultured symbiont in pure culture (245 μg liter−1) does not differ from the amount produced in symbiosis (100 to 300 μg liter−1).

Considering the bacterial origin of rhizonin, sets of primers that are specific for bacterial NRPS genes (7) were tested using total DNA isolated from the rhizonin-positive fungal strain as well as from the cultured symbiont. In fact, various NRPS gene fragments of the expected size (0.5 kb) could be amplified. Conversely, no NRPS genes could be detected in the symbiont-free R. microsporus strain. Although the cloned bacterial NRPS gene fragments represent orthologs of biosynthetic genes from related bacteria, such as the surfactin, plubactin, and syringopeptin NRPS genes, their role in rhizonin biosynthesis needs to be established. The identification of toxin biosynthesis genes by PCR would allow the detection of toxinogenic Rhizopus strains or, more precisely, the presence of bacterial endosymbionts in food fermentations (e.g., tempeh and sufu). Cloning of the biosynthesis genes and functional studies are currently under way.

In summary, we have demonstrated that the severe hepatotoxin rhizonin A is not a fungal metabolite but is produced by bacteria residing within the fungal cytosol. These endofungal bacteria were localized by transmission electron microscopy and could even be isolated and grown in pure culture. Phylogenetic analyses revealed that these symbionts belong to the genus Burkholderia. The strongest evidence for their involvement in toxin production was provided by PCR experiments and HPLC-MS investigation of the fungal wild-type strain, a symbiont-free nonproducer, which has been cured with an antibiotic, and the isolated bacterial symbionts. In addition, pure rhizonin was isolated from an upscaled fermentation of the cultured endosymbiont. Only very few examples of endofungal bacteria have been described so far (1, 4, 11). This finding is only the second example of a bacterium-fungus symbiosis in which a fungus harbors endosymbionts for toxin production, and this is the first study involving a toxic cyclopeptide. While the phytopathogenic alliance of Rhizopus and rhizoxin-producing endosymbionts has a clear function as a virulence factor, the ecological role of rhizonin formation is not yet clear. Nonetheless, knowing the true producer of the hepatotoxic cyclopeptides is undoubtedly of great importance for human health and the food industry.


We thank D. Schwartz and C. Müller for providing cosmid AU52 and for helpful discussions, G.-M. Schwinger for strain maintenance, M. Cyrulies and M. Steinacker for assistance in large-scale fermentation, and F. A. Gollmick and A. Perner for NMR and MS measurements, respectively.


[down-pointing small open triangle]Published ahead of print on 22 November 2006.


1. Boer, W., L. B. Folman, R. C. Summerbell, and L. Boddy. 2005. Living in a fungal world: impact of fungi on soil bacterial niche development. FEMS Microbiol. Lett. 29:795-811. [PubMed]
2. Finking, R., and M. A. Marahiel. 2004. Biosynthesis of nonribosomal peptides. Annu. Rev. Microbiol. 58:453-488. [PubMed]
3. Franck, B. 1984. Mycotoxins from mold fungi—weapons of uninvited fellow-borders of man and animal: structures, biological activity, biosynthesis and precaution. Angew Chem. Int. Ed. Engl. 23:493-505.
4. Frey-Klett, P., and J. Garbaye. 2005. Mycorrhiza helper bacteria: a promising model for the genomic analysis of fungal-bacterial interactions. New Phytol. 168:4-8. [PubMed]
5. Guermani, L., C. Villaume, H. W. Bau, V. Chandrasiri, J. P. Nicolas, and L. Mejean. 1992. Composition and nutritional value of okara fermented by Rhizopus oligosporus. Sci. Aliments 12:441-451.
6. Han, B. Z., F. M. Rombouts, and M. J. R. Nout. 2001. A Chinese fermented soybean food. Int. J. Food Microbiol. 65:1-10. [PubMed]
7. Heinzelman, E., S. Berger, O. Puk, B. Reichenstein, W. Wohlleben, and D. Schwartz. 2003. A glutamate mutase is involved in the biosynthesis of the lipopeptide antibiotic friulimicin in Actinoplanes friulensis. Antimicrob. Agents Chemother. 47:447-457. [PMC free article] [PubMed]
8. Hoppert, M., C. Gentzsch, and K. Schorgendorfer. 2001. Structure and localization of cyclosporin synthetase, the key enzyme of cyclosporin biosynthesis in Tolypocladium inflatum. Arch. Microbiol. 176:285-293. [PubMed]
9. Jennessen, J., K. F. Nielsen, J. Houbraken, E. K. Lyhne, J. Schnürer, J. C. Frisvad, and R. A. Samson. 2005. Secondary metabolite and mycotoxin production by the Rhizopus microsporus group. J. Agric. Food Chem. 53:1833-1840. [PubMed]
10. Keuth, S., and B. Bisping. 1994. Vitamin B12 production by Citrobacter freundii or Klebsiella pneumoniae during tempeh fermentation and proof of enterotoxin absence by PCR. Appl. Environ. Microbiol. 60:1495-1499. [PMC free article] [PubMed]
11. Lumini, E., S. Ghignone, V. Bianciotto, and P. Bonfante. 2006. Endobacteria or bacterial endosymbionts? To be or not to be. New Phytol. 170:205-208. [PubMed]
12. Mohammed, S. I., L. R. Steenson, and A. W. Kirleis. 1991. Isolation and characterization of microorganisms associated with the traditional sorghum fermentation for production of Sudanese kisrat. Appl. Environ. Microbiol. 57:2529-2533. [PMC free article] [PubMed]
13. Mutasa, E. S., A. M. Tymon, B. Gottgens, F. M. Mellon, P. F. R. Little, and L. A. Casselton. 1990. Molecular organization of an A-mating type factor of the basidiomycete fungus Coprinus cinereus. Curr. Genet. 18:223-229.
14. Partida-Martinez, L. P., and C. Hertweck. 2005. Pathogenic fungus harbours endosymbiotic bacteria for toxin production. Nature 437:884-888. [PubMed]
15. Potgieter, M., P. S. Steyn, F. R. van Heerden, P. H. van Rooyen, and P. L. Wessels. 1989. Conformational analysis of the cyclic peptide rhizonin A in solution and crystalline state. Tetrahedron 45:2337-2350.
16. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
17. Scherlach, K., L. P. Partida-Martinez, H.-M. Dahse, and C. Hertweck. 2006. Antimitotic rhizoxin derivatives from cultured symbionts of the rice pathogenic fungus Rhizopus microsporus. J. Am. Chem. Soc. 128:11529-11536. [PubMed]
18. Schwarzer, D., R. Finking, and M. A. Marahiel. 2003. Nonribosomal peptides: from genes to products. Nat. Prod. Rep. 20:275-287. [PubMed]
19. Steyn, P. S., A. A. Tuinman, F. R. van Heerden, P. H. van Rooyen, P. L. Wessels, and C. J. Rabie. 1983. The isolation, structure, and absolute configuration of the mycotoxin rhizonin A, a novel cyclic heptapeptide containing N-methyl-3-(furyl)alanine, produced by Rhizopus microsporus. J. Chem. Soc. Chem. Commun. 1983:47-49.
20. Tunçel, G., M. J. R. Nout, L. Brimer, and D. Goektan. 1990. Toxicological, nutritional and microbiological evaluation of tempe fermentation with Rhizopus oligosporus of bitter and sweet apricot seeds. Int. J. Food Microbiol. 11:337-344. [PubMed]
21. Weber, G., and E. Leitner. 1994. Disruption of the cyclosporin synthetase gene of Tolypocladium niveum. Curr. Genet. 26:461-467. [PubMed]
22. Weber, G., K. Schorgendorfer, E. Schneider-Scherzer, and E. Leitner. 1994. The peptide synthetase catalyzing cyclosporine production in Tolypocladium niveum is encoded by a giant 45.8-kilobase open reading frame. Curr. Genet. 26:120-125. [PubMed]
23. Wilson, T., C. J. Rabie, J. E. Fincham, P. S. Steyn, and M. A. Schipper. 1984. Toxicity of rhizonin A, isolated from Rhizopus microsporus, in laboratory animals. Food Chem. Toxicol. 22:275-281. [PubMed]

Articles from Applied and Environmental Microbiology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...