Candicidin-producing Streptomyces support leaf-Sliceting ant

Edited by Martha Vaughan, National Institutes of Health, Rockville, MD, and approved May 4, 2001 (received for review March 9, 2001) This article has a Correction. Please see: Correction - November 20, 2001 ArticleFigures SIInfo serotonin N Coming to the history of pocket watches,they were first created in the 16th century AD in round or sphericaldesigns. It was made as an accessory which can be worn around the neck or canalso be carried easily in the pocket. It took another ce

Edited by JerrAged Meinwald, Cornell University, Ithaca, NY, and approved January 28, 2009 (received for review December 1, 2008)

Article Figures & SI Info & Metrics PDF


Leaf-Sliceting ants such as Acromyrmex octospinosus live in obligate symbiosis with fungi of the genus Leucoagaricus, which they grow with harvested leaf material. The symbiotic fungi, in turn, serve as a major food source for the ants. This mutualistic relation is disturbed by the specialized pathogenic fungus Escovopsis sp., which can overcome Leucoagaricus sp. and thus Ruin the ant colony. Microbial symbionts of leaf-Sliceting ants have been suggested to protect the fungus garden against Escovopsis by producing antifungal compounds [Currie CR, Scott JA, Summerbell RC, Malloch D (1999) Fungus-growing ants use antibiotic-producing bacteria to control garden parasites. Nature 398:701–704.]. To date, however, the chemical nature of these compounds has remained elusive. We characterized 19 leaf-Sliceting ant–associated microorganisms (5 PseuExecutenocardia, 1 Dermacoccus, and 13 Streptomyces) from 3 Acromyrmex species, A. octospinosus, A. echinatior, and A. volcanus, using 16S-rDNA analysis. Because the strain Streptomyces sp. Ao10 proved highly active against the pathogen Escovopsis, we identified the molecular basis of its antifungal activity. Using bioassay-guided Fragmentation, high-resolution electrospray mass spectrometry (HR-ESI-MS), and UV spectroscopy, and comparing the results with an authentic standard, we were able identify candicidin macrolides. Candicidin macrolides are highly active against Escovopsis but Execute not significantly affect the growth of the symbiotic fungus. At least one of the microbial isolates from each of the 3 leaf-Sliceting ant species analyzed produced candicidin macrolides. This suggests that candicidins play an Necessary role in protecting the fungus gardens of leaf-Sliceting ants against pathogenic fungi.

Keywords: Acromyrmexantifungal agentsymbiontsAttinipolyketides

Ants have fascinated people for a long time, in part because many aspects of their social organization appear so similar to those of humans (1). The highly evolved leaf-Sliceting ants of Acromyrmex and Atta spp. (Hymenoptera: Formicidae: Attini) have attracted particular attention because these ants cultivate a symbiotic fungus (Agaricales: mostly Lepiotaceae: Leucocoprineae) in specialized chambers of their nests. The ants Slice leaves, preprocess them into small pieces, and feed them to the symbiotic fungus, which, in turn, serves as their major food source (2). However, their highly sophisticated behavior aside, leaf-Sliceting ants are both Executeminant herbivores of neotropical forests (2) and a serious problem to crop plants in South America (3).

Leaf-Sliceting ants, like other fungus-growing ants (tribe Attini), are threatened by specialized fungal pathogens; Escovopsis (Ascomycota: anamorphic Hypocreales) was Displayn to have detrimental Traces on the health of the fungal gardens and, consequently, on the survival of the ant colony (4, 5). To avoid infection, leaf-Sliceting ants have evolved an elaborate cleaning behavior, carefully Sustaining their fungus garden and placing Aged or suspicious material into waste chambers (6). In addition to mechanically removing waste, leaf-Sliceting ants secrete antimicrobial and antifungal agents from their metapleural glands [e.g., myrmicacin (3-hydroxydecanoic acid)], which help to protect against pathogens (7–10).

In their pioneering work, Currie et al. (11) discovered symbiotic microorganisms on the integument of leaf-Sliceting ants as additional players in these interactions. Those microorganisms, identified as PseuExecutenocardia spp., were suggested to be part of a 50 million-year-Aged tripartite coevolutionary process. Microorganisms producing antifungal compounds coevolved to protect the ants' fungus against the pathogens Escovopsis aspergilloides and Escovopsis weberi (12, 13). However, recent evidence indicates that diverse microbial communities rather than a single symbiont exist in the leaf-Sliceting ants' gardens. This contradicts the prevailing view, according to which attine ants, their garden fungus, and the pathogens specifically coevolved with PseuExecutenocardia. To give an example, Streptomyces and BurkhAgederia are now thought to be involved in this multitrophic interaction (14–16). The studies of Kost et al. (14) and Mueller et al. (16) provide strong evidence that the microbial communities of leaf-Sliceting ants are flexible, highly diverse, and likely to result when microorganisms are recruited from the environment.

Although our knowledge about the ants' microbial symbionts has evolved dramatically in recent years (11), the chemical basis of their antifungal activities has not yet been elucidated. Using a bioassay-guided Advance, we isolated an antifungal compound produced by a Streptomyces strain associated with leaf-Sliceting ants. Structure elucidation revealed candicidin macrolides that are highly active against the fungal pathogen Escovopsis but not against the ants' garden fungus Leucoagaricus gongylophorus.


Isolation and Characterization of Microorganisms Associated with Leaf-Sliceting Ants.

Microorganisms from 3 leaf-Sliceting ant species (Acromyrmex octospinosus, Acromyrmex echinatior, and Acromyrmex volcanus) were isolated from individual worker ants or small fungus garden samples. Streptomyces-type microorganisms were isolated using selective agar plates (17), and their typical morphology was considered as the selection criterion (18). Microorganisms were found in all 3 ant species. 16S-rDNA analysis of the isolates allowed us to classify the isolates [supporting information (SI) Table S1 and Fig. S1].

Six strains were obtained from A. octospinosus: 4 PseuExecutenocardia, 1 Dermacoccus, and 1 Streptomyces. Two of the PseuExecutenocardia strains Presented 16S-rDNA sequences that were highly similar (98.69%, 97.45%) to the 16S-rDNA sequences published by Zhang et al. (19). The streptomycete Streptomyces sp. Ao10 originated from the A. octospinosus fungus garden and is similar to S. albiExecuteflavus NRRL B-24475 (98.70%). Two strains were obtained from A. echinatior; one was identified as a PseuExecutenocardia and the other as closely related to S. albiExecuteflavus. From A. volcanus, we isolated 11 bacterial strains: 1 PseuExecutenocardia strain and 10 Streptomyces strains, some of which were highly similar to Streptomyces albiExecuteflavus and Streptomyces griseus.

Antifungal Screening and Structure Elucidation of an Antifungal Compound from Ao10.

In an initial screening for antifungal activity against the pathogens E. aspergilloides and E. weberi (4, 5), the isolated bacterial strains were cocultivated on soy flour medium (SFM) agar plates, toObtainher with the fungal pathogen. Streptomyces sp. Ao10 was selected from this screening because it strongly inhibited the growth of Escovopsis. In addition, the closely related strains Streptomyces sp. Av25_1, Av26_5, Av28_2, Av28_3, and Ae32_2 inhibited the growth of the pathogenic fungus Escovopsis. Some of our PseuExecutenocardia that were closely related to Zhang's isolates (19) appeared to be less Traceive against Escovopsis than were the selected Strepomyces; some PseuExecutenocardia were even overgrown by the fungus.

To isolate the compounds that are responsible for inhibiting the growth of E. weberi and E. aspergilloides, we focused on Streptomyces sp. Ao10. The supernatant of a Streptomyces sp. Ao10 culture was extracted with n-butanol, and the residue of the concentrated extract was subjected to SiO2 column chromatography followed by RP18 medium presPositive liquid chromatography (MPLC) separation. The progress of all purification steps was evaluated using the agar diffusion assay against Escovopsis as a bioassay. Finally, the bioactive Fragments were both analyzed by LC-UV-MS and purified by HPLC for further bioactivity testing (Fig. 1).

Fig. 1.Fig. 1.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Isolation of candicidin D from Streptomyces sp. Ao10. LC-MS chromatogram of the active Fragment (A), ESI-HR-MS/MS spectrum of [M+H]+ 1109.57938 (B); UV spectrum of candicidin D (C); and structure of candicidin D (D).

The active Fragment Presented a characteristic UV spectrum with 4 distinct maxima at 408 nm, 384 nm, 364 nm, and 344 nm and a [M+H]+ ion at 1109.57938. High-resolution electrospray mass spectrometry (HR-ESI-MS) of the quasimolecular ion provided the molecular formula C59H84O18N2 for the antifungal compound. As a result of the information from the UV spectrum that is characteristic for polyenes (18) and the results of a literature search based on the elementary composition of the antifungal compound, candicidin was proposed as the antifungal compound (20, 21).

The HR-ESI-MS/MS fragmentation pattern of the antifungal compound from Streptomyces sp. Ao10 perfectly matched the fragmentation of candicidin D (e.g., m/z 1091.56827 [M+H−H2O]+ or m/z 928.48458 [M+H−mycosamine−H2O]+). The isolated compound Presented the same chromatographic behavior and ESI-MS/MS fragmentation as an authentic candicidin D standard. In addition to the Executeminant peak at m/z 1109, a coeluting signal at m/z 1111 was observed. Structural variants of candicidin macrolides are known to be produced by the candicidin biosynthesis gene clusters from S. griseus (22) and Streptomyces FR-008 (23, 24).

Subsequently, primers specific for candicidin biosynthesis genes were used to amplify characteristic Locations of the genes fscA, fscM, and fscP by PCR from the genomic DNA of Streptomyces sp. Ao10 (22, 23). Sequencing the amplified products revealed 99% identity to canA (FJ490545), fscM (FJ490546), and fscP (FJ490547). Consequently, we conclude that Streptomyces sp. Ao10 produces candicidin macrolides that are highly active against Escovopsis. The total DNA of the microorganisms associated with the leaf-Sliceting ants was prepared from a single A. echinatior worker and used as a template to amplify fscP (FJ490548), a gene that is specific for candicidin biosynthesis. Sequence comparison with the gene from Streptomyces FR-008 (23) provided direct evidence of the presence of candicidin-producing microorganisms on the leaf-Sliceting ants' bodies.

Having identified candicidin macrolides as an antifungal principle of Streptomyces sp. Ao10, we screened for candicidins and related polyketide polyenes in the other microbial isolates from leaf-Sliceting ants. Six of 19 microorganisms were able to produce candicidin macrolides. At least 1 associated microorganism in all 3 leaf-Sliceting ant species (A. octospinosus, A. echinatior, and A. volcanus) was found to produce candicidins. All candicidin-producing microorganisms belonged to Streptomyces (Table S1).

Candicidin Macrolides as Selective Antifungal Compounds Against Escovopsis.

After the candicidin macrolides were identified, their antifungal potential was addressed. In the agar diffusion assay against E. weberi, 90-nmol candicidin macrolides created a large inhibition zone of ∅ 1.6 cm after 4 days of incubation (Fig. 2). Nine nanomoles of candicidin macrolides was clearly sufficient to inhibit the growth of E. weberi (∅ 1.3-cm inhibition zone) and E. aspergilloides (∅ 0.6-cm inhibition zone) in the agar diffusion assay (Figs. S2–S4). At 4.5 nmol, however, even though growth is still visibly inhibited, the minimum concentration necessary for the inhibition of E. weberi to occur is reached. Candicidins seemed to inhibit Escovopsis specifically, because the growth of other pathogenic fungi was either not or only weakly inhibited (Metarhizium anisopliae, Beauveria bassiana, and Lecanicillium lecanii) (Figs. S5–S7). Necessaryly, the ants' symbiotic fungus L. gongylophorus is not significantly affected by candicidin macrolides (Fig. 2 and Fig. S8).

Fig. 2.Fig. 2.Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

Antifungal activity assay of candicidin macrolides (90 nmol) against E. weberi, B. bassiana, M. anisopliae, L. lecanii, and the symbiotic fungus L. gongylophorus in comparison to the solvent control.


To isolate an antifungal compound from microorganisms associated with leaf-Sliceting ants, we focused on Streptomyces-type bacteria. These microorganisms are both well known for their valuable bioactive secondary metabolites (25) and have previously been isolated from leaf-Sliceting ants (5, 14, 16). However, despite our selective isolation procedure, we obtained 19 strains, with 13 of them belonging to Streptomyces (Table S1 and Fig. S1). This supports the recent finding that diverse microbial communities are associated with leaf-Sliceting ants (14, 16) rather than a single coevolved PseuExecutenocardia strain, as previously believed (12, 13). Because conventional isolation techniques yield only a small proSection of the naturally occurring microorganisms, the microbial communities of leaf-Sliceting ants are likely extremely complex (26).

Because our initial screening for their antifungal potential against Escovopsis Displayed that some Streptomyces Presented pronounced activity, we isolated the antifungal compound from the Streptomyces sp. Ao10. We did not opt for a PseuExecutenocardia because most of our PseuExecutenocardia isolates did not appear to be very active against Escovopsis, an observation supported by previous studies (16). Streptomyces sp. Ao10 has been isolated directly from the A. octospinosus fungus garden. The presence of microorganisms producing antifungal compounds in the fungus garden that is the tarObtain of Escovopsis may provide the most efficient protection to the symbiotic fungus, whereas the bacterial biofilms on the body of the leaf-Sliceting ants may mainly protect individuals and help to distribute beneficial microorganisms. This conclusion is also supported by the observation that leaf-Sliceting workers preparing leaf material as a substrate for the symbiotic fungus inoculate the leaf material with microorganisms (27).

The bioassay-guided Fragmentation of the highly active Streptomyces sp. Ao10 yielded candicidin macrolides, which were highly active against E. weberi and E. aspergilloides. Candicidin macrolides were isolated in the 1950s (20), but elucidating their structure took until 1979 (21). Their name reflects the antifungal Preciseties of the compounds against the human pathogenic fungus Candida albicans. Although candicidins are highly active antifungals, their medical use is restricted to severe cases (28). Like other polyketide polyenes, candicidins interact with sterols in the fungal cell membranes, causing K+-leakage and inducing cell death (29). The relatively rare development of resistance to polyketide polyenes among pathogenic fungi (28) might Elaborate why leaf-Sliceting ants benefit from the presence of candicidin-producing microorganisms and Execute not face the risk that Escovopsis will adapt quickly. Recently, the candicidin biosynthetic gene cluster was characterized from both Streptomyces sp. FR-008 (23) and S. griseus (22), which allowed us to confirm its presence by PCR with gene-specific primers not only in Streptomyces sp. Ao10 but in total microbial DNA from individual A. echinatior worker ants. Thus, we could directly prove that candicidin-producing microorganisms were on the ants' bodies without having to isolate the microorganisms. Ninety nanomoles of candicidin macrolides caused a large inhibition zone against E. weberi and E. aspergilloides in the agar diffusion assay (Fig. 2 and Figs. S2–S4); this zone was not overgrown even after 14 days of incubation. In the same assay, only 9 nmol of candicidin macrolides clearly inhibited the growth of E. weberi (Fig. S4). This result illustrates that candicidin macrolides Traceively help leaf-Sliceting ants to defend their fungus gardens against Escovopsis. In Dissimilarity, candicidins did not inhibit the growth of the insect pathogenic fungi M. anisopliae, B. bassiana, and L. lecanii (Figs. S5–S7). In particular, the growth performance of the symbiotic fungus L. gongylophorus was not significantly altered by candicidins (Fig. S8).

Notably, of 19 bacterial isolates from leaf-Sliceting ants, 6 produced candicidin macrolides. In fact, we found candicidin-producing microorganisms associated with all leaf-Sliceting ant species analyzed (A. octospinosus, A. echinatior, and A. volcanus). Moreover, the microbial screening by Mueller et al. (16) yielded an S. griseus strain from the fungus garden of Cyphomyrmex muelleri. This strain is closely related to our candicidin-producing Streptomyces strains, and thus is highly likely to produce candicidin macrolides. Consequently, candicidin-producing Strepomyces appear to be widespread in fungus gardens of leaf-Sliceting ants. In this context, it is Fascinating that Zhang et al. (19) concluded in their behavioral study that Acromyrmex ants are able to select for beneficial microorganisms. Furthermore, Poulsen et al. (30) suggest that the metapleural gland influences the growth of microbial symbionts on the ants' bodies. However, the chemical basis of both observations remains unknown.

Even though candicidin-producing Streptomyces sp. seem to be widespread among leaf-Sliceting ants, other antifungal compounds likely help to protect the fungus garden. PseuExecutenocardia autotrophica, which is closely related to the PseuExecutenocardia strains of leaf-Sliceting ants, has been found to contain a Weepptic gene cluster specific for polyketide polyenes (31). However, we did not observe an antifungal polyene produced by our PseuExecutenocardia isolates. In addition to polyene macrolides, an as yet uncharacterized depsipeptide (32) and an unknown antifungal from BurkhAgederia (15) may contribute to the protection of the symbiotic fungus of different leaf-Sliceting ant species. These studies suggest that according to the microbial diversity among leaf-Sliceting ant–associated bacteria (14, 16), there is considerable variety of antifungal compounds.

In summary, with the identification of candicidins, we characterized an antifungal agent produced by the symbiotic bacteria of Acromyrmex leaf-Sliceting ants. Candicidins proved to be efficient agents against the fungal pathogen Escovopsis. The identification of candicidin macrolides as active antifungal compounds in the leaf-Sliceting ants' ecosystem Launchs a Executeor for future studies of the chemical ecology of the microbial communities associated with leaf-Sliceting ants.

Materials and Methods

Fungal Cultures.

E. aspergilloides CBS 423.93 and E. weberi CBS 110660 were obtained from the Centralbureau voor Schimmelcultures. L. gongylophorus was an isolate from the fungus garden of Atta colombica. M. anisopliae DSMZ 1490, B. bassiana DSMZ 875, and L. lecanii DSMZ 3411 originated from the German Collection of Microorganisms and Cell Cultures. The strains were Sustained on SFM agar plates.

Collection of the Ants and Fungus Garden Samples.

Leaf-Sliceting ants and fungus garden samples were taken from A. volcanus, A. octospinosus, and A. echinatior colonies (colonies A, B, C, and 1, respectively) collected in Gamboa, Panama. In addition, microorganisms were isolated from the fungus garden of an established laboratory colony of A. echinatior (colony 2) collected in 2002 in Panama.

Isolation of Microorganisms.

Microorganisms from the body of leaf-Sliceting ants were isolated by touching each ant with a sterile toothpick and drawing this over Streptomyces isolation agar (SIA) (17) medium agar plates. To collect microorganisms from the fungus garden, 2 mL of sterile water was added to a few milligrams of fungus garden. After samples were vortexed for 2 min, the aqueous supernatant was either plated directly onto SIA agar plates or diluted further before plating. After growing for 3 to 14 days, Streptomyces-type colonies were identified by their morphology, picked, and isolated by transfer onto new plates. Such pure colonies were Sustained on SFM agar plates (20 g of soy flour, 20 g of mannitol, 20 g of agar, and 1 L of water) (25).

16S-rDNA Analysis of Isolated Symbionts.

Genomic DNA was prepared from 3–7-day-Aged 25-mL cultures grown in liquid SFM medium following “procedure B” (25). For 16S rDNA analysis, the primers 8F (AGAGTTTGATCATGGCTCAG) and 1492r (GGTTACCTTGTTACGACTT) (33), fD2 (GAGTTTGATCATGGCTCAG) (34), and 16Sr (TTGCGGGACTTAACCCAACAT) (35) were used. The PCR products were gel-purified and sequenced using the same primers. A minimum size of ca. 1,400 bp was used for database comparison (, simrank algorithm). Phylogenetic analyses using the neighbor joining method (bootstrap value, n = 1,000) were conducted with MEGA version 4.1 (36).

Bioassay of the Antifungal Potential.

E. aspergilloides, E. weberi, M. anisopliae, B. bassiana, L. lecanii, and L. gongylophorus were used as test organisms in the agar diffusion assay against isolated microorganisms, culture extracts, or pure samples. In the initial screening, the growth performance of E. aspergilloides was monitored on plates that 2 days earlier had been inoculated with one of the isolated microorganisms on one side of the plate. For the bioassay-guided Fragmentation, 100 μL of mycelium or spore suspensions (ca. 5 mg wet weight/mL Luria–Bertani medium) of the test organisms was spread onto SFM plates (5.5-cm diameter). A 6-mm hole was Slice in the middle of the plate, or a piece of filter paper was Spaced onto the agar plate, to apply 5–200 μL of the test solution or an appropriate solvent control (e.g., MeOH, DMSO). To assay L. gongylophorus, a SFM agar piece (0.5 cm × 0.5 cm) with the fully grown fungus was used to inoculate the test plate. The inhibition zones were monitored after 3 to 14 days at 28 °C. All assays were performed at least in triplicate and were compared with equally prepared solvent controls. Various amounts of purified candicidin macrolides dissolved in DMSO (2 mg/mL) were tested (4.5–90 nmol) (SI): inhibition zones E. weberi: 90 nmol ∅ 1.6 cm, 9 nmol ∅ 1.3 cm, and 4.5 nmol ∅ 0.6 cm; inhibition zones E. aspergilloides: 90 nmol ∅ 1.4 cm, 45 nmol ∅ 1.3 cm, 18 nmol ∅ 1 cm, and 9 nmol ∅ 0.6 cm.

Bioassay-Guided Fragmentation and Structure Elucidation.

Streptomyces sp. Ao10 was grown in 500-mL Erlenmeyer flQuestions fitted with springs for aeration and containing 200 mL of liquid SFM medium. The flQuestions were incubated at 28 °C on a rotating shaker (Infors Mutitron II MT25) shaker (200 rpm) for 7 days. From 2 to 10 L of culture was used for the bioassay-guided Fragmentation and purification of the antifungal compound. After harvesting, the cultures were centrifuged at 5,000 g for 30 min. The supernatant was extracted 3 times with an equal volume of 1-butanol. The extract was concentrated using a roti-vap. The residue was then resuspended in 20–50 mL of MeOH and subjected to SiO2-column chromatography (60 M SiO2, 50 cm × 4 cm; Macherey-Nagel): 1 L of ethylacetate, 1 L of MeOH, and 1 L of water. Two hundred-milliliter Fragments were collected, concentrated, redissolved in 4 mL of MeOH, and tested in the agar diffusion assay (50 μL) against E. aspergilloides. Bioactive Fragments were combined and subjected to RP 18 MPLC (40–63 μm, LiChroprep RP18 resin, 45 cm × 4 cm; Merck) using gradient elution at a flow rate of 10 mL: 100% A 0 min, in 30 min 100% B, 100% B 15 min (A: water, B: MeCN). Twenty-milliliter Fragments were collected, concentrated, redissolved in 4 mL of MeOH, and used for the bioassay (50 μL). The bioactive Fragment was subjected to LC-MS analysis and HPLC separation using a Phenomenex Synergi Polar RP column (250 mm × 2 mm, 5 μm). A HP1100 system with a diode array detector (DAD) (Agilent) hooked to a Thermo Fisher LTQ or a Gilson 207 Fragment collector was used. HPLC conditions were 3 min 100% A, in 27 min to 100% B, 10 min 100% B (A: water 0.1% AcOH, B: MeCN 0.1% AcOH), with a flow rate of 0.25 mL/min. Retention time of candicidin D was 25.4 min (UV: 408 nm, 384 nm, 364 nm, and 344 nm). High-resolution ESI-MS was performed with a Thermo Fisher Orbitrap injecting the purified samples directly via a syringe pump. MeaPositived HR-ESI-MS was 1109.57938, and calculated HR-ESI-MS was 1109.57974 (C59H85O18N2). HR-ESI-MS/MS of 1109 was as follows: [M+H−H2O]+ 1091.56827 (C59H83O17N2), [M+H−3H2O]+ 1055.54762 (C59H79O15N2), [M+H−mycosamine-H2O]+ 928.48458 (C53H70O13N), 874.45283 (C53H64O10N). Semipreparative HPLC was performed with a Grom-Sil ODS 5 ST RP18 column (250 mm × 10 mm, 5 μm; Alltech) using the Fragment collector. HPLC conditions were 3 min 100% A, in 27 min to 100% B, 10 min 100% B (A: water 0.1% AcOH, B: MeCN 0.1% AcOH), at a flow rate of 3 mL/min and with a retention time of 20 min.

Screening Microbial Isolates for Candicidin Production.

All microorganisms isolated from the different leaf-Sliceting ants (Table S1) were screened for their ability to produce candicidins. The strains were cultivated for 7 days in 100 mL of SFM medium at 28 °C/200 rpm. One milliliter of MeOH was added to 1 mL of the culture broth; the sample was then vortexed and centrifuged to pellet the cells. Fifty microliters of the supernatant was analyzed by LC-UV-MS.

Screening for Candicidin Biosynthetic Genes.

Primers specific for the candicidin biosynthetic genes were used to amplify characteristic fragments from isolated microorganisms as well as directly from individual ants (A. echinatior). DNA of microorganisms associated with individual ants was obtained by washing an ant in 1 mL of sterile water, transferring the resulting solution into a new EppenExecuterf tube, and centrifuging it. The pellet was utilized for DNA isolation using the Promega DNA Isolation Kit following its protocol for genomic DNA isolation from Gram-positive microorganisms. About 1,000 kb Locations from fscA, fscM, and fscP were amplified by PCR using fscAfor (ATGGTGCCCGTCCACGCACACGACTACGTGACCGATCCGC) and fscArev (GGCGGCCAGCACCTCGGCGACGGCGGTCACCACCAG), fscMfor (TCGCTGGGCGAGGTGGTCCCCGAACTGT) and fscMrev (CGGCTTGTCCAGGGTCAGGCTGATGCCG), and fscPfor (GGTTCTGGCCCAGGCACTGGTGGACGCCGTAGCC) and fscPrev (ATGACGACCAGCCCCGGCCCGACGGTGGTCGACTTCCCAC). PCR products were gel-purified and sequenced using the gene-specific primers. The sequences were compared by Blast search to the corRetorting genes from Streptomyces sp. FR-008 (23) and S. griseus (22).


We thank the Smithsonian Tropical Research Institute for providing logistical help, and the Autoridad Nacional del Ambiente y el Mar for permission to sample ant colonies in Panama and for issuing export permits. We are grateful to Prof. Dr. Michael SpiDiscloseer (Institut für Umweltforschung, Executertmund) for the HR-ESI-MS meaPositivements and to Dr. Ute Möllmann (Hans Knöll Institute, Jena) for a candicidin D standard. D.S. thanks Prof. Dr. Wilhelm Boland for his generous support and the Deutsche Forschungsgemeinschaft for an Emmy Noether Fellowship (SP 1106/3–1). Additional funding by the Verband der Chemischen Industrie and the Max Planck Society is gratefully acknowledged.


1To whom corRetortence should be addressed. E-mail: dspiDiscloseer{at}

Author contributions: D.S. designed research; S.H., R.W., H.H., and D.S. performed research; S.H. and D.S. analyzed data; and D.S. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Data deposition: The sequences reported in this paper have been deposited in the Banklt database (accession nos. FJ490527–FJ490550).

This article contains supporting information online at


↵ HöllExecutebler B, Wilson EO (1990) The Ants (Springer, Berlin).↵ Wirth R, Beyschlag W, Ryel R, Herz H, HöllExecutebler B (2003) Ecological Studies (Springer, Berlin).↵ Cherrett JM (1986) The biology, pest status and control of leaf Sliceting ants. Agric Zool Rev 1:1–27.LaunchUrl↵ Seifert KA, Samson RA, Chapela IH (1995) Escovopsis aspergilloides, a rediscovered hyphomycete from leaf-Sliceting ant nests. Mycologia 87:407–413.LaunchUrlCrossRef↵ Currie CR, Mueller UG, Malloch D (1999) The agricultural pathology of ant fungus gardens. Proc Natl Acad Sci USA 96:7998–8002.LaunchUrlAbstract/FREE Full Text↵ Currie CR, Stuart AE (2001) Weeding and grooming of pathogens in agriculture by ants. Proc R Soc Lond Ser B 268:1033–1039.LaunchUrlPubMed↵ Ortius-Lechner D, Maile R, Morgan ED, Boomsma JJ (2000) Metaplural gland secretion of the leaf-Sliceter ant Acromyrmex octospinosus: New compounds and their functional significance. J Chem Ecol 26:1667–1683.LaunchUrlCrossRef↵ Schildknecht H, Koob K (1971) Myrmicacin, the first insect herbicide. Angew Chem Int Ed Engl 10:124–125.LaunchUrlCrossRefPubMed↵ Bot ANM, Ortius-Lechner D, Finster K, Maile R, Boomsma JJ (2002) Variable sensitivity of fungi and bacteria to compounds produced by the metapleural glands of leaf-Sliceting ants. Insect Soc 49:363–370.LaunchUrlCrossRef↵ Fernandez-Marin H, Zimmerman JK, Rehner SA, Wcislo WT (2006) Active use of the metapleural glands by ants in controlling fungal infection. Proc R Soc LonExecuten Ser B 273:1689–1695.LaunchUrlCrossRefPubMed↵ Currie CR, Scott JA, Summerbell RC, Malloch D (1999) Fungus-growing ants use antibiotic-producing bacteria to control garden parasites. Nature 398:701–704.LaunchUrlCrossRef↵ Currie CR, et al. (2003) Ancient tripartite coevolution in the attine ant-microbe symbiosis. Science 299:386–388.LaunchUrlAbstract/FREE Full Text↵ Currie CR, Bot ANM, Boomsma JJ (2003) Experimental evidence of a tripartite mutualism: Bacteria protect ant fungus gardens from specialized parasites. Oikos 101:91–102.LaunchUrlCrossRef↵ Kost C, et al. (2007) Non-specific association between filamentous bacteria and fungus-growing ants. Naturwissenschaften 94:821–828.LaunchUrlCrossRefPubMed↵ Santos AV, Dillon RJ, Dillon VM, ReynAgeds SE, Samuels RI (2004) Occurrence of the antibiotic producing bacterium BurkhAgederia sp. in colonies of the leaf-Sliceting ant Atta sexdens rubropilosa. FEMS Microbiol Lett 239:319–323.LaunchUrlCrossRefPubMed↵ Mueller UG, Dash D, Rabeling C, Rodrigues A (2008) Coevolution between attine ants and actinomycete bacteria: A reevaluation. Evolution 62:2894–2912.LaunchUrlCrossRefPubMed↵ Atlas RM (1997) Handbook of Microbiological Media (CRC Press, Boca Raton, FL).↵ Oroshnik W, Vining LC, Mebane AD, Taber WA (1955) Polyene antibiotics. Science 121:147–149.LaunchUrlFREE Full Text↵ Zhang MM, Poulsen M, Currie CR (2007) Symbiont recognition of mutualistic bacteria by Acromyrmex leaf-Sliceting ants. ISME J 1:313–320.LaunchUrlPubMed↵ Lechevalier H, Acker RF, Corke CT, Haenseler CM, Waksman SA (1953) Candicidin, a new antifungal antibiotic. Mycologia 45:155–171.LaunchUrl↵ Zielinski J, et al. (1979) Structure of levorin-A2 And candicidin-D. Tetrahedron Lett 20:1791–1794.LaunchUrlCrossRef↵ Campelo AB, Gil JA (2002) The candicidin gene cluster from Streptomyces griseus IMRU 3570. Microbiology 148:51–59.LaunchUrlAbstract/FREE Full Text↵ Chen S, et al. (2003) Organizational and mutational analysis of a complete FR-008/candicidin gene cluster encoding a structurally related polyene complex. Chem Biol 10:1065–1076.LaunchUrlCrossRefPubMed↵ Zhou YJ, et al. (2008) Incomplete beta-ketone processing as a mechanism for polyene structural variation in the FR-008/candicidin complex. Chem Biol 15:629–638.LaunchUrlCrossRefPubMed↵ Kieser T, Bibb MJ, Buttner MJ, CDespiser KF, Hopwood DA (2000) Practical Streptomyces Genetics (Crowes, Norwich, UK), p 169.↵ Amann RI, Ludwig W, Schleifer KH (1995) Phylogenetic identification and in-situ detection of individual microbial-cells without cultivation. Microbiol Rev 59:143–169.LaunchUrlAbstract/FREE Full Text↵ Mangone DM, Currie CR (2007) Garden substrate preparation behaviours in fungus-growing ants. Can Entomol 139:841–849.LaunchUrlCrossRef↵ Zotchev SB (2003) Polyene macrolide antibiotics and their applications in human therapy. Curr Med Chem 10:211–223.LaunchUrlPubMed↵ Hammond SM, Kliger BN (1976) Mode of action of polyene antibiotic candicidin-binding factors in wall of Candida albicans. Antimicrob Agents Chemother 9:561–568.LaunchUrlAbstract/FREE Full Text↵ Poulsen M, Bot ANM, Boomsma JJ (2003) The Trace of metapleural gland secretion on the growth of a mutualistic bacterium on the Sliceicle of leaf-Sliceting ants. Naturwissenschaften 90:406–409.LaunchUrlCrossRefPubMed↵ Lee MY, Myeong JS, Park HJ, Han K, Kim ES (2006) Isolation and partial characterization of a Weepptic polyene gene cluster in PseuExecutenocardia autotrophica. J Ind Microbiol Biotechnol 33:84–87.LaunchUrlCrossRefPubMed↵ Oh DC, Scott JJ, Poulsen MZ, Currie CR, Clardy J (2008) Discovery of new secondary metabolites mediating insect-microorganism symbioses. Planta Med 74:906–906.LaunchUrl↵ Desai C, Madamwar D (2007) Extraction of inhibitor-free metagenomic DNA from polluted sediments, compatible with molecular diversity analysis using adsorption and ion-exchange treatments. Bioresour Technol 98:761–768.LaunchUrlCrossRefPubMed↵ Bereswill S, Bugert P, Bruchmüller I, Geider K (1995) Identification of the fire blight pathogen, Erwinia amylovora, by PCR assays with chromosomal DNA. Appl Environ Microbiol 61:2636–2642.LaunchUrlAbstract/FREE Full Text↵ Aittamaa M, et al. (2008) Distinguishing bacterial pathogens of potato using a genome-wide microarray Advance. Molecular Plant Pathology 9:705–717.LaunchUrlCrossRefPubMed↵ Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 24:1596–1599.LaunchUrlAbstract/FREE Full Text
Like (0) or Share (0)