Hypopulvins, novel peptaibiotics from the polyporicolous fungus Hypocrea pulvinata, are produced during infection of its natural hosts

Abstract

In order to investigate the significance of antibiotics for the producing organism(s) in the natural habitat, we screened specimens of the polyporicolous fungus Hypocrea pulvinata growing on its natural hosts Piptoporus betulinus and Fomitopsis pinicola. Results showed that a particular group of nonribosomally biosynthesised antibiotic polypeptides, the peptaibiotics, which contain the nonproteinogenic marker amino acid α-aminoisobutyric acid (Aib), was produced in the natural habitat by the fungicolous producer and, consequently, released into the host. Using liquid chromatography coupled to electrospray high-resolution mass spectrometry we detected especially 19-, but also 11-, 18-, and 20-residue peptaibiotics in the five infected specimens analysed. Structures of peptaibiotics found were confirmed by analysing the peptaibiome of pure agar cultures obtained by single-ascospore isolation from the specimens. The 19-residue peptaibols were determined as deletion sequences of the trichosporins B lacking the Aib residue in position 6. Notably, 26 of the 28 peptaibiotics sequenced were novel; therefore the name ‘hypopulvins’ was introduced. Considering not only the ubiquity of both the two host species but also the highly specific association between H. pulvinata and P. betulinus/F. pinicola, and the abundance of this fungicolous species in north temperate regions of the world, a decisive role for the peptaibiotics detected in this study is predicted, which may act as mediators of the complex interactions between the basidiomycetous host and its fungicolous ascomycete ‘partner’. Structural analogies of the hypopulvins, particularly with other 18-, 19-, and 20-residue peptaibiotics, suggest that the hypopulvins are forming transmembrane ion channels and could thus support the hypothesis of a parasitic lifestyle of the fungicolous producer.

Introduction

Natural products of fungal origin

As of 2010 approximately half a million natural products were known of which 60 000–80 000 are estimated to be of microbial origin. Approximately half of the latter display some kind of biological activity (Bérdy 2012). The fungal kingdom, which currently comprises more than 98 000 validly described species (Kirk et al. 2008) contributes some 30 000 natural products of which 15 000–16 000 are bioactive. Of these, around 11 250 originate from microscopic fungi (Bérdy 2012). Peptide antibiotics constitute a considerable part of those metabolites, including therapeutically important β-lactam antibiotics (penicillins, cephalosporins), which account for more than 65 % of the world antibiotics market. However, the significance of other, nonantibiotic peptide drugs of fungal origin is comparable – the immunosuppressant market is still dominated by the nonribosomal biosynthesised cyclosporine A (Elander 2003; Demain & Sanchez 2009).

Peptaibiotics – nonribosomally biosynthesised fungal peptide antibiotics containing α,α-dialkyl α-amino acids

During the past two decades, a constantly growing group of peptide antibiotics, the peptaibiotics, have started to regain particular interest because of their unique bioactivities, resulting from their amphipathicity and helical conformations. Peptaibiotics are defined as nonribosomally biosynthesised, linear or cyclic polypeptide antibiotics of exclusively fungal origin, which (i) have a molecular weight between 500 and 2200 Da, thus containing 4–21 residues; (ii) show a high content of the marker α-aminoisobutyric acid (Aib) as well as other α,α-dialkylamino acids; (iii) are characterised by the presence of other nonproteinogenic amino acids and/or lipoamino acids; (iv) possess an acylated N-terminus, and (v) in the case of linear peptides, have a C-terminal residue that in large part consists of a free or acetylated amide-bonded β-amino alcohol. The C-terminus might also be an amine, amide, free amino acid, 2,5-diketopiperazine, or sugar alcohol. The majority of Aib-containing peptides carry a C-terminal residue representing a β-amino alcohol, and this subgroup is, therefore, referred to as peptaibols. As a result of the nonribosomal biosynthesis on large multifunctional peptide synthetases, highly complex mixtures of homologous and positionally isomeric peptides are produced. The latter phenomenon is commonly referred to as microheterogeneity, i.e., exchange of single or few amino acids or C-terminal amino alcohols, respectively (Degenkolb & Brückner 2008). Therefore, sophisticated, multidimensional analytical approaches such as HPLC coupled to tandem electrospray ionisation (ESI)-MSⁿ or Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF-MS) and chiral capillary gas-chromatography/selected ion monitoring electron impact MS became the methods of choice for structure elucidation of peptaibtiotics. To date, the presence of peptide-bound α,α-dialkyl α-amino acids, i.e., mostly Aib, but also d- and l-isovaline (Iva), l-α-ethylnorvaline (Etnva) as well as 1-aminocyclopropane-1-carboxylic acid (Acc) (for structures, see Fig 1), has been confirmed in acidic hydrolysates of more than 30 genera of fungi (Brückner et al. 2009). Among those the ubiquitous genus Trichoderma/Hypocrea, which currently contains approximately 200 validly described species (Jaklitsch 2009, 2011; Samuels & Ismaiel 2011; Jaklitsch & Voglmayr 2012; Jaklitsch et al. 2012; Samuels et al. 2012a, b) is generally regarded as the richest source of peptaibiotics (Degenkolb & Brückner 2008). Notably, d- and l-Iva may even occur in different positions of one and the same sequence, as was shown for clonostachin, neoefrapeptins, and integramides, for review see Degenkolb & Brückner (2008). Obviously these results contradict the still widespread belief that α,α-dialkyl α-amino acids do not or rarely occur in the biosphere and, if detected in the environment, are definitely of extraterrestrial origin. In fact fungal species, which contain peptaibiotic-producing strains, are ubiquitous and cosmopolitan, even occurring in marine, Arctic and Antarctic regions (Brückner et al. 2009). However, differentiation between biotically and abiotically synthesised Aib and homologues is possible by stable carbon and nitrogen isotopic composition (Elsila et al. 2011).

Fig 1. (A) Structures and configurations of α,α-dialkylamino acids found in peptaibiotics. (B) Building scheme of 20-residue SF1 peptaibiotics, produced by Hypocrea pulvinata. (C) General building scheme of 11-residue peptaibols found in SF4.

Bioactivities of peptaibiotics from Trichoderma/Hypocrea

In the past decade, many Trichoderma species, such as Trichoderma ovalisporum (Holmes et al. 2004), Trichoderma koningiopsis (Samuels et al. 2006a), and Trichoderma citrinoviride (Maddau et al. 2009) have gained increasing interest as potential biocontrol agents in plant protection. Further promising examples include Trichoderma paucisporum and Trichoderma theobromicola, displaying in vitro-activity against frosty pod rot of cocoa, Moniliophthora roreri (Samuels et al. 2006b), and Trichoderma martiale. In small-scale in situ field trials the latter proved highly effective against black pod rot of cocoa caused by Phytophthora palmivora (Hanada et al. 2009). The most successful story in cocoa biocontrol, however, is the recent approval of ‘Tricovab’ (Anonymous Novembro 2011/Fevereiro 2012), a formulation against Crinipellis (=Moniliophthora) perniciosa, the Witches’ broom pathogen, containing the hyperparasite Trichoderma stromaticum (Pomella et al. 2007). Although this species has been shown to produce five 18-residue peptaibols, trichostromaticins A-E (Degenkolb et al. 2006), the mechanism(s) contributing to its remarkable in vivo bioactivity have not been completely resolved yet. Peptaibiotics are prone to play a key role in the infection process of a host by a fungicolous species because of their unique ability of forming voltage-gated ion channels. This phenomenon is best described by the dipole flip-flop gating model (Menestrina et al. 1986). Their well-documented membrane activity, however, may also account for other striking bioactivities of peptaibiotics, such as neurolepsy (Berek et al. 2009), inhibition of amyloid β-peptide formation (Hosotani et al. 2007), inhibition of HIV-1 integrase (Singh et al. 2002), suppression of tumour cells, targeted Ca²⁺-mediated apoptosis and autophagy in human hepatocellular carcinoma cells (Shi et al. 2010), as well as induction of defence responses and systemic resistance in tobacco against tobacco mosaic virus (Luo et al. 2010), and programmed cell death in fungal plant pathogens (Shi et al. 2012).

In vivo-detection of peptaibiotics

Reports on the occurrence of peptaibiotics in the natural habitat of the producer(s) have sparsely been published so far. Most of almost 1000 individual sequences of peptaibiotics known to date have been sequenced in extracts of fungal cultures grown under artificial laboratory conditions. The first example of peptaibiotics obtained from natural specimens was the isolation of hypelcins A and B obtained from 2 kg of dried, crushed stromata of Hypocrea peltata (Fujita et al. 1984a; Matsuura et al. 1993, 1994). However, it appears that in recent years only a single contribution directly addressed the detection of peptaibols, viz. the 16-residue antiamoebins, under in vivo-conditions (Lehr et al. 2006) whereas other authors most probably have overlooked the presence of a fungicolous species in the examined material. Thus, the formation of peptaibiotics was accidentally ascribed to basidiomycetous hosts, which have rather been infested by an unrecognised fungicolous fungus, such as Sepedonium sp., for review see Degenkolb & Brückner (2008).

The present study is thus aimed at the question as to whether peptaibiotics formation under the conditions of the natural habitat of the producer(s) is either a rather infrequent or a more common phenomenon. If the latter hypothesis is true, peptaibiotics might, indeed, play a decisive role in the antibiotic-based colonisation and defence of the natural substrate, as was recently demonstrated for the antiamoebins from Stilbella fimetaria (Lehr et al. 2006). Given that this coprophilous species produces fungicidal amounts of peptaibols on herbivore dung, other highly specialised hypocrealean fungi occupying ecological niches must be considered as additional model systems for these investigations.

From an ecophysiological point of view, necrotrophic mycoparasites within the genus Trichoderma/Hypocrea are best suited for such study. This is due to the fact that a parallel formation and synergistic action of hydrolytic enzymes, above all as β-1,3-glucanase, and peptaibiotics have previously been attributed an important role in mycoparasitism between Trichoderma atroviride (originally misidentified as Trichoderma harzianum) and its fungal hosts such as Botrytis cinerea (Schirmböck et al. 1994; Lorito et al. 1996).

Choice of the model organism

Hypocrea pulvinata, the Ochre Cushion, a rather common fungus that occurs on decaying basidiomes of polypores in north temperate regions of the world (Jaklitsch 2011), was chosen as a first model organism. Notably, this polyporicolous species is assumed to display host specificity as it has only been unambiguously identified on the Birch Bracket Piptoporus betulinus and the Red-banded Bracket Fomitopsis pinicola so far. Records from other aphyllophoralean taxa such as Ganoderma sp. could not been confirmed in recent years; and the claimed association with Laetiporus sulphureus was recently shown to be based on misidentifications of the respective host (Jaklitsch 2011).

To prove our hypothesis of peptaibiotic production by members of the order Hypocreales under in vivo-conditions, basidiomes of P. betulinus and F. pinicola, infected with H. pulvinata, were collected from different locations. Cultures obtained from the respective teleomorphs were analysed using a peptaibiomics approach as described earlier (Krause et al. 2006; Degenkolb et al. 2012).

Materials and method

Chemicals

All solvents used, acetonitrile (MeCN), methanol (MeOH), CH₂Cl₂, and formic acid (FA), were of LC–MS grade from Sigma–Aldrich (Steinheim, Germany). Water was purified by a Merck-Millipore Milli-Q system (Schwalbach/Ts., Germany).

Isolation of pure agar cultures of anamorphs

Basidiomes of Piptoporus betulinus and Fomitopsis pinicola infected with Hypocrea pulvinata were collected from five different locations in Austria and Russia (I–V; Table 1). Pure agar cultures were obtained by single-ascospore isolations from the respective specimens as described by Jaklitsch (2009).

Table 1. Habitat and geographic origin of Hypocrea pulvinata isolates included in this study.

Isolate	Host	Collecting information	Culture
I	Old basidiome of Piptoporus betulinus	Austria, Styria, Liezen, Untertal S Schladming, bog near Gasthof Tetter, 1020 m s.m., grid square 8648/1, 12 Jun. 2011, H. Voglmayr & Greilhuber	CBS 133228
II	Basidiome of Piptoporus betulinus on Betula pendula	Austria, Lower Austria, Wien-Umgebung, Mauerbach, 31 Jul. 2011, W.M. Jaklitsch	CBS 133229
III	Basidiome of Piptoporus betulinus on Betula pendula	Austria, Lower Austria, Wien-Umgebung, Mauerbach, 31 Jul. 2011, W.M. Jaklitsch	CBS 133230
IV	Decaying basidiome of Piptoporus betulinus on Betula pendula	Russia, Republic of Buryatia, Bargusin district, Lake Baikal, ‘Holy Nose’ Peninsula; 53° 35'54.9〞 N, 108° 52'11.2〞 E; 24 Jul. 2011, A. Berg & H. Dörfelt
V	Decaying basidiome of Fomitopsis pinicola on Pinus sibirica	Russia, Republic of Buryatia, Severo-Baykalsky district, Nizhneangarsk; 55° 46'39.4〞 N, 109° 35'49.9〞 E; 24 Jul. 2011, A. Berg & H. Dörfelt

Extraction of specimens

Few milligram of teleomorph stromata were extracted with 40 ml of a mixture of CH₂Cl₂–MeOH (1:1v/v), centrifuged, the solvent was evaporated in vacuo (Rotavapor R-215, Büchi, Essen, Germany) or using a stream of nitrogen. For screening on the micrOTOF-Q II, the extracts were cleaned up over Sep-Pak Classic C₁₈ cartridges (Waters, Eschborn, Germany) as described by Krause et al. (2006) whereas the crude extracts (redissolved in MeOH) were analysed directly using the more sensitive maXis QTOF. Hymenophores of apparently uninfected basidiomes of P. betulinus were investigated as a control.

Cultivation and extraction of pure cultures

Cultures of specimens I, II, and III were grown on potato dextrose agar (PDA) (Becton, Dickinson & Co., Heidelberg, Germany) at 23 °C for 6 d in Petri dishes of 9 cm diameter. These subcultures were used for inoculation of the main cultures on PDA. After 10 d of cultivation at 23 °C in the dark, ten fully-grown Petri dishes were extracted, the extracts combined, evaporated to dryness in vacuo, and the residues analysed as described below.

LC-MS analysis

Two QTOF systems, both from Bruker Daltonic (Bremen, Germany) controlled by HyStar v. 3.2 were used. Both instruments were equipped with an orthogonal ESI source, and coupled to a Dionex UltiMate 3000 UHPLC (Dionex, Idstein, Germany).

System 1: high-resolution micrOTOF-Q II mass spectrometer. For separation, an Acclaim 120 C₈, 3 μm, 120 Å, 2.1 × 150 mm column (Dionex, Idstein, Germany) at a flow rate of 0.25 ml min⁻¹, and a temperature of 35 °C was used. Eluent A consisted of H₂O + 0.1 % FA, eluent B of 95 % MeCN + 0.1 % FA. Subsamples of 10 μl were injected. The column was held at 80 % A/20 % B for 5 min, then a gradient from 20 % B to 100 % over 55 min was applied. Thereafter, the column was held at 100 % B for 15 min, returned to the start conditions in 1 min, and finally equilibrated for 14 min.

Samples were screened for peptaibiotics in the positive ion mode using the following three-step routine procedure: A full scan was recorded from m/z 50 to 3000. This was followed by an in-source collision-induced dissociation (CID) scan from m/z 50 to 2000, recorded at energy of 150 eV. Finally, results of the in-source CID scan were verified by MS/MS experiments on selected precursor ions. For precursors <m/z 1000, a collision energy of 30 eV was applied, precursor ions in the m/z range from 1000 to 1500 were fragmented at a collision energy of 35 eV and precursor ions >m/z 1500 at a collision energy of 40 eV. The isolation width for MS/MS experiments was set to ±1 Da.

System 2: MaXis 3G QTOF mass spectrometer operated at a resolution of 40 000 full width at half maximum (FWHM). An Acquity UPLC^® BEH300 C₁₈, 1.7 μm, 2.1 × 150 mm column was used for separation, using H₂O + 0.1 % FA (eluent A) and B of 100 % MeCN + 0.1 % FA. The flow rate was set to 0.3 ml min⁻¹ and the temperature to 40 °C. The gradient started with 90 % A/10 % B at the time: 0 min and was changed to 50 % A/50 % B at time: 7 min, then to 30 % A/70 % B at time: 25 min, then raised to 100% B at time: 38 min and held at 100% B until time: 41 min before set to starting conditions from time 42 min to 46 min. Three microlitres were injected. MS were scanned in the m/z range 100–2000. Auto MS with precursor ion-dependent collision energy optimisation was used for fragmentation in the range of 10–65 eV.

Data interpretation was performed using the DataAnalysis v. 4.0 (Build 281) software (Bruker Daltonic, Bremen, Germany). Use of high-resolution ESI mass spectrometry allows the unequivocal sequencing of fragment ion series according to the Roepstorff/Fohlman/Biemann nomenclature. In cases where the isomeric amino acids (Leu/Ile and Val/Iva, respectively) or the corresponding amino alcohols (Leuol/Ileol) with the same elemental composition could not be distinguished, the abbreviations Lxx, Vxx, and Lxxol were used instead (Degenkolb et al. 2006, 2012; Krause et al. 2006).

Results

Screening of basidiomes infected by Hypocrea pulvinata

Biomass from the five infected samples investigated (Table 1) was shown to contain intense peaks of novel, especially 19-residue peptaibiotics, which we have named hypopulvins (HPVs). Analysis of two basidiomes from Piptoporus betulinus not infected by Hypocrea did not show any indications of peptaibiotics, thus verifying that they are produced by Hypocrea during infection of the basidiomes (data not shown). A comparison of base peak chromatograms (BPCs) is given in Fig 2A and B.

Fig 2. BPCs of (A) specimens I, IV, and V screened with the maxis TOF; (B) specimens I, II, and III screened with the micrOTOF-Q II; (C) plate cultures I, II, and III screened with the micrOTOF-Q II. † – no peptaibiotics; ‡ – coeluting peptaibiotics.

The peptaibiome of the teleomorph was dominated by 19-residue peptaibols (Tables 2a and 3a, Table S1a), as well as minor 18-residue deletion analogues (compounds 1 and 2), only detected in isolates IV and V.

Table 2a. Sequences of 18- and 19-residue peptaibiotics detected in specimens IeV of Hypocrea pulvinata.

No.	tR [min]		Residue
			1	2	3	4	5	–	6	7	8	9	10	11	12	13	14	15	16	17	18	19
1a	10.5	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	–
2a	11.2	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Ala	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	–
3	34.1–34.2	Ac	Aib	Ala	Ser	Ala	Aib	–	Gln	Aib	Lxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
4	34.7–34.9	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Ala	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
5	35.6–35.9	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Gly	Vxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
6	36.2	Ac	Aib	Ala	Ala	Ser	Aib	–	Gln	Aib	Lxx	Aib	Ala	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
7	36.3	Ac	Aib	Ala	Ala	Ala	Ala	–	Gln	Aib	Lxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
8	36.7–36.9	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Ala	Gly	Lxx	Aib	Pro	Vxx	Aib	Vxx	Gln	Gln	Pheol
9	37.2–37.4	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
10	37.3–37.5	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Lxxol
11	38.3–38.4	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Vxx	Gln	Gln	Pheol
12	38.7–38.9	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
13	39.2	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Ser	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
14	39.5–39.7	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Ala	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
15	39.9	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Ala	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Lxxol
16	40.4–40.5	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Ala	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
17	40.8–41.1	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Ala	Lxx	Aib	Pro	Vxx	Aib	Vxx	Gln	Gln	Pheol
18	41.3–41.6	Ac	Aib	Ala	Ala	Ala	Vxx	–	Gln	Aib	Lxx	Aib	Ala	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
No.	Compound identical or positionally isomeric with														References
1a	new (deletion sequence of 9)
2a	new (deletion sequence of 14, 15, and 16)
3	new (trichosporin B-Ia-[Aib]6)														Iida et al. 1990
4	new (trichosporin B-IIIa-[Aib]6, [Aib]9 → [Ala]9)														Iida et al. 1990
5	new (C-terminal pentadecapeptide resembles polysporin A with [Vxx]8 → [Lxx]8, N-terminal: trichosporin B-pentapeptide)														New et al. 1996
6	new (C-terminal pentadecapeptide resembles trichosporins B-Ia and B-IIIa with [Gly]10 → [Ala]10)														Iida et al. 1990
7	new (trichosporin B-IIIc-[Aib]6)														Iida et al. 1990
8	new (C-terminal pentadecapeptide resembles polysporin D with [Aib]9 → [Ala]9, N-terminal: trichosporin B-pentapeptide)														New et al. 1996
9	new (trichosporin B-IIIa-[Aib]6)														Iida et al. 1990
10	new (isomer of 9: [Pheol]19 → [Lxxol]19; C-terminal octapeptide found in some of the hypelcins A and stilbolflavins B)														Matsuura et al. 1993; Jaworski & Brückner 2001
11	new (trichosporin B-IVb-[Aib]6)														Iida et al. 1990
12	new (cf. 9; trichosporin B-IIIa-[Aib]6)														Iida et al. 1990
13	new (N-terminal hexadecapeptide: cf. trichosporins B IIIa/IVb; decapeptide [Aib]9-[Gln]18: cf. tricholongin B I)														Iida et al. 1990; Rebuffat et al. 1991
14	new (trichosporin B-IIIa-[Aib]6, [Gly]10 → [Ala]10)														Iida et al. 1990
15	new (cf. 14: [Pheol]19 → [Lxxol]19; cf. 10: [Gly]10 → [Ala]10; trichosporin B-IIIa-[Aib]6, [Gly]10 → [Ala]10, [Pheol]19 → [Lxxol]19)														Iida et al. 1990
16	new (positional isomer of 14)														Iida et al. 1990
17	new (positional isomer of 18)
18	new (cf. 14: [Aib]5 → [Vxx]5)

^aDetected in specimens IV and V, only.

Variable residues are underlined. Minor sequence variants are underlined in the individual sequences.

Table 3a. Comparison of the peptaibiotic pattern of specimens I–V.

	Compound
	1a	2a	3	4	5	6	7	8	9
tR [min]	10.5	11.2	34.1–34.2	34.7–34.9	35.6–35.9	36.2	36.3	36.7–36.9	37.2–37.4
[M + Na]+	n.d.	n.d.	1904.0666	1874.0586	1874.0624	1918.0835	1874.0788	1888.0757	1888.0711
[M + H]+	1732.9987	1747.0138	1882.0893	1852.0790	1852.0776	1896.0999	1852.0743	1866.0911	1866.1009
I	n.d.	n.d.	+	+	+	+	+	+	+
II	n.d.	n.d.	n.d.	(+)	(+)	+	(+)	+	+
III	n.d.	n.d.	n.d.	n.d.	n.d.	(+)	n.d.	+	+
IVb	(+)	(+)	(+)	(+)	+	(+)	+	n.d.	+
Vb	+	+	(+)	(+)	+	(+)	+	n.d.	+
	Compound
	10	11	12	13	14	15	16	17	18
tR [min]	37.3–37.5	38.3–38.4	38.7–38.9	39.2	39.5–39.7	39.9	40.4–40.5	40.8–41.1	41.3–41.6
[M + Na]+	1854.0799	1902.0918	1888.0783	1918.1024	1902.0898	1868.0893	1902.0974	1916.1049	1916.1051
[M + H]+	1832.1076	1880.1072	1866.0997	1896.1048	1880.1187	1846.1231	1880.1177	1894.1248	1894.1233
I	+	+	+	+	(+)	(+)	+	+	+
II	+	+	+	n.d.	n.d.	(+)	+	+	(+)
III	n.d.	+	+	n.d.	n.d.	n.d.	+	(+)	n.d.
IVb	n.d.	+	+	n.d.	+	+	+	(+)	(+)
Vb	+	+	+	+	+	+	+	(+)	n.d.

n.d.: not detected.

+: positive.

(+): weakly positive.

^aDetected by DTU maXis gradient, only.

^bInterpolated from DTU maXis gradient using higher injection volume.

All compounds display the typical characters of the peptaibol subfamily 1 (SF1: Table 4, Fig 1) the largest one of the nine subfamilies, as introduced by Chugh & Wallace (2001). Accordingly, in HPV’s, one Gln residue is found in position 6, and another two towards the C-terminus, in position 17 and 18. A highly conserved Pro residue is located in position 13 of the peptide chain. Many sequences have a Gly or Ala residue in position 10. Most of the HPV’s terminate in phenylalaninol (Pheol), only two in Lxxol. At least four, at most seven, residues are occupied by an Aib residue. Qualitative differences between the teleomorphs observed with the maXis and the micrOTOF-Q II might be explained by the different screening dates, considering that the peptaibiome of a living specimen is subjected to dynamic changes.

Table 4. Structural variation of 19-residue peptaibiotics produced by the five H. pulvinata isolates investigated.

	Residue
	1	2	3a	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
Ac	Aib	Ala	Ala	Ala	Aib	Gln	Aib	Lxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
			(Ser)b	(Ser)	(Ala)			(Vxx)	(Ala)	Ala	(Vxx)			(Lxx)		(Vxx)			(Lxxol)
					(Vxx)					(Ser)

^aVariable positions are underlined.

^bMinor sequence variations are parenthesised.

Screening of Hypocrea pulvinata plate cultures

Pure agar cultures were only obtained from single ascospores of specimens I, II, and III (Table 1), whereas specimens IV and V did not contain viable ascospores anymore. However, BPCs of the three plate cultures screened are highly similar (Fig 2C). Comparing the peptaibiotic pattern of specimens vs. plate culture (Tables 2b and 3b, Table S1b), 19-residue peptaibols were also found predominantly in the cultures although their microheterogeneity was less pronounced:

Table 2b. Sequences of 11-, 19-, and 20-residue peptaibiotics detected in plate cultures of Hypocrea pulvinata I–III.

No.	tR [min]		Residue
			1	2	3	4	5	–	6	7	8	9	10	11	12	13	14	15	16	17	18	19
19	34.0–34.2	Ac	Aib	Ala	Ala	Ser	Aib	–	Gln	Aib	Lxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
20a	37.0–37.2	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Vxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
20b	37.0–37.3	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Gly	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
21	38.9–39.0	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Ser	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
22	39.6–39.7	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Ala	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
23	40.2–40.5	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Ala	Lxx	Aib	Pro	Vxx	Aib	Aib	Gln	Gln	Pheol
24	41.0–41.1	Ac	Aib	Ala	Ala	Ala	Aib	–	Gln	Aib	Lxx	Aib	Ala	Lxx	Aib	Pro	Lxx	Aib	Aib	Gln	Gln	Pheol
			Residue
			1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20
25	41.3–41.6	Ac	Aib	Ala	Aib	Ala	Aib	Aib	Gln	Aib	Vxx	Aib	Gly	Lxx	Aib	Pro	Lxx	Aib	Aib	Gln	Gln	Pheol
26	41.5	Ac	Aib	Gln	Lxx	Lxx	Aib	Pro	Lxx	Lxx	Aib	Pro	Lxxol
27	42.3–42.5	Ac	Aib	Ala	Aib	Ala	Aib	Aib	Gln	Aib	Vxx	Aib	Gly	Lxx	Aib	Pro	Lxx	Aib	Aib	Glu	Gln	Pheol
28	42.9	Ac	Vxx	Gln	Lxx	Lxx	Aib	Pro	Lxx	Lxx	Aib	Pro	Lxxol
No.			Compound identical or positionally isomeric with													References
19			new (cf. 6: [Ala]10 → [Gly]10)
20a			new (trichosporin B-IIId -[Aib]6)													Iida et al. 1990
20b			new (cf. 9)
21			new (cf. 13)
22			new (cf. 14)
23			new (cf. 14)
24			new (cf. 22, [Vxx]14→ [Lxx]14)
25			new (trichosporin B-IIId: [Vxx]15 → [Lxx]15)													Iida et al. 1990
26			cf. trichorovins XIII, XIV; hypomurocins A-V, A-Va;													Iida et al. 1995; Wada et al. 1995; Becker et al. 1997
			trichorozin IV; trichobrachins III: I, J; trichobrachins III: 16a, 17, 18;													Brückner et al. 1993; Krause et al. 2007
			trichobrachins C-1, C-2; Tv29-11-V b													Ruiz et al. 2007; Mukherjee et al. 2011
27			new
28			cf. Trichofumin B, Tv29-11-VI, hypojecorin A-22 and A-23													Berg et al. 2003; Mukherjee et al. 2011
																Degenkolb et al. 2012

Variable residues are underlined in the table header. Minor sequence variants are underlined in the individual sequences.

Table 3b. Comparison of the peptaibiotic pattern of plate cultures I–III.

	Compound
	19	20a	20b	21	22	23	24	25	26	27	28
tR [min]	34.0–34.2	37.0–37.2	37.0–37.3	38.9–39.0	39.6–39.7	40.2–40.5	41.0–41.1	41.3–41.6	41.5	42.3–42.5	42.9
[M + Na]+	1904.0757	n.d.	1888.0737	n.d.	1902.0868	1902.0864	1916.1030	1987.1402	1189.7938	1988.1451	1203.8111
[M + H]+	1882.0812	1852.0721	1866.0909	1896.0874	1880.1063	1880.1065	1894.1222	1965.1576	1211.7767	1966.1576	1225.7929
I	+	(+)	+	(+)	+	+	+	n.d.	+	n.d.	+
II	+	(+)	+	(+)	+	+	+	+	+	+	+
III	+	(+)	+	(+)	+	+	+	n.d.	(+)	n.d.	(+)

The pattern of 19-residue peptaibols was less diverse, and amino acid exchanges were only found in positions 4, 8, 10, and 14 (Table 2b).

In contrast to what has been observed for the specimens, two new 20-residue minor compounds, (25) and (27), representing the building scheme of trichosporins B (Iida et al. 1990, 1993), were detected. Another two minor compounds, 11-residue peptaibols (26) and (28), display the structural characteristics of the growing subfamily 4 (SF4: (Chugh & Wallace 2001; Degenkolb et al. 2012)).

Considerable microheterogeneity, including positional isomerism, of long-chain peptaibiotics was observed in both stromata and cultures: Sequences (4), (5), (7), and (20a) represent positional isomers of m/z 1852.0721 [M + H]⁺, whereas sequences (8), (9), (12), and (20b) display m/z 1866.0909 ([M + H]⁺). Five sequences (11), (14), (16), (22), and (23) represent positional isomers of m/z 1880.1064 ([M + H]⁺); whereas another three (17), (18), and (24) are of m/z 1894.1222 ([M + H]⁺). Further, Ser-containing positional isomers of m/z 1882.085 [M + H]⁺ are represented by (3) and (19), whereas sequences (6), (13), and (21) display m/z 1896.0874 [M + H]⁺.

Discussion

General remarks

The most notable result of this investigation is the unequivocal confirmation of peptaibiotic biosynthesis in the natural habitat of a fungicolous fungus, indicating that these are required for infection. We here present the first example of in vivo-production of peptaibiotics by a fungicolous fungus growing on either of its natural hosts. Considering (i) the ubiquity of both the two host species (ii) the highly specific association between Hypocrea pulvinata and Piptoporus betulinus/Fomitopsis pinicola, and (iii) the abundance of this fungicolous species in north temperate regions of the world including Europe, North America, Japan, and Russia (Jaklitsch 2011), it seems reasonable to postulate a decisive role for the peptaibiotics detected in this study:

They are hypothesised to be one of the mediators of the complex interactions between the basidiomycetous host and its fungicolous ascomycete ‘partner’ (see below).

Twenty-four of the 26 new HPVs produced by H. pulvinata are 19-residue peptaibols, basically representing deletion sequences (Δ Aib⁶) of the 20-residue trichosporins B (Iida et al. 1990, 1993) and polysporins A–D (New et al. 1996), both isolated from strains of Trichoderma polysporum. This deletion, however, is predicted not to negatively influence the bioactivity of these long-chain peptaibols as all important structural features, which comply with the requirements for the formation of transmembrane ion channels in artificial lipid bilayer membranes, are still present. Generally, 18-, 19-, and 20-residue peptaibiotics display a higher membrane pore formation activity by several orders of magnitude in comparison to smaller peptaibols consisting of less than 17 residues (Grigoriev et al. 2003).

Importance of different structurally conserved amino acid residues for the bioactivity of peptaibiotics – general remarks

Aib and Iva residues strongly promote the formation of helical structures (α- or 3₁₀-helices, and even mixed forms) (Toniolo & Benedetti 1991), which is due to the steric constraints imposed by the geminal alkyl groups of the C^α-atom (Chugh & Wallace 2001).

The significance of glutamine residues

Gln residues in position 6 or 7 were postulated to play a key role for ion channel stabilisation (Fox & Richards 1982), as they are located in the pore lumen, lining it together with Gln^17–18 and Gly¹⁰ and Pro¹³ carbonyls of the adjoining transmembrane helices (Duclohier 2004). Molle et al. (1996) demonstrated that replacement of Gln⁷ by Ala⁷ resulted in complete loss of channel-forming activity, which supports this hypothesis. This phenomenon was explained by the inability of this nonpolar residue to form side-chain H-bonds. Interhelix H-bonds are thought to play a decisive role in the stabilisation of channel-forming helix bundles. This hypothesis was corroborated by the replacement of Gln⁷ by Asn⁷, restoring nearly initial functional properties. In contrast, replacement of Gln⁷ by Ser⁷ resulted in a significantly reduced voltage dependence and shorter channel lifetime. This was explained by the formation of much weaker H-bonds by the –CH₂OH group of Ser⁷ (Molle et al. 1996).

The decisive role of proline residues

Prolines in position 13 or 14 of the peptide chain create bends of the helical structure, thus affecting the (i) probability of ion channel formation, (ii) lifetime of the channel, and (iii) occurrence of multilevel conductance (Nagaoka et al. 1996).

For the 20-residue alamethicins (ALM) F30, proline in position 14 which corresponds to Pro¹³ of the 19-residue HPV’s was demonstrated to display optimal channel activity whereas synthetic analogues bearing Pro in positions 11, 12, 13, 15, 16, and 17 formed considerably less stable ion channels (Kaduk et al. 1997). A proline residue in position 14 of ALM F30 is essential for (i) haemolysis of human erythrocytes, (ii) stimulation of catecholamine secretion from bovine adrenal chromaffin cells, and (iii) induction of metabolic activity in bovine aortic endothelial cells (Dathe et al. 1998). Surprisingly, it was reported that Gln⁷, Gly¹¹, and Pro¹⁴ are not essential for channel formation but substitution of any of these residues reduced (i) the number of conductance levels and (ii) significantly decreased their lifetimes (Kaduk et al. 1998).

Role of the C-terminal substituents

Aromatic residues such as the C-terminal Pheol, which are located in the membrane interface between the hydrophilic lipid head groups and the hydrophobic fatty acid chains, have been postulated to stabilise the polypeptide transmembrane segments (Wallace 2000).

The two compounds (10) and (15) carrying a C-terminal Lxxol residue resemble some of the structural properties of two other 20-residue peptaibol mixtures – hypelcins A from stromata of Hypocrea peltata (Fujita et al. 1984a; Matsuura et al. 1993) and stilboflavins B from Stilbella flavipes CBS 146.81 (Jaworski & Brückner 2001). Hypelcins A were shown to (i) act as uncouplers of oxidative phosphorylation in rat liver mitochondria (Takaishi et al. 1980), (ii) exhibit antibacterial and antifungal activity, including growth inhibition of the shiitake mushroom Lentinus edodes, (iii) exert contractile action on guinea pig ileum (Fujita et al. 1984b), and (iv) induce ion channels in planar bilayer lipid membranes. Remarkably, the substitution of Pheol for Leuol and Ileol, respectively, resulted in prolonged open channel lifetime (Koide et al. 1997). Stilboflavins were reported to exhibit weak antibiotic activity against grampositive bacteria and cause haemolysis of sheep erythrocytes (Jaworski & Brückner 2001).

The two 18-residue sequences, compounds (1) and (2), exhibit a deletion of the C-terminal amino alcohol residue. Compared to the amount of the major 19-residue peptaibols, their possible contribution to the postulated bioactivity of the mixture of HPVs is regarded negligible. Despite this, truncated sequences of SF1 peptaibols have been reported before:

19-residue peptaibiotics, trichobrachins I (TB I) from Trichoderma ghanense CBS 936.69 (syn. Trichoderma parceramosum) lacking the C-terminal Pheol residue, were shown to originate from 20-residue trichobrachins II (TB II) by enzymatic degradation (Krause et al. 2007). Two minor desPheol-compounds F30 representing 1.3 % of the ALM mixture from Trichoderma arundinaceum CBS 123793 (formerly known as Trichoderma ‘viride’ NRRL 3199) have been detected by nonaqueous capillary electrophoresis (NACE) coupled to electrospray mass spectrometry (Psurek et al. 2006). Two further 17-residue minor compounds lacking the Gln residue in position 18 are predicted to be present; however, the intensity of their diagnostic CID-fragment ions was too low for sequencing.

Multiple bioactivities of the structurally related trichosporins B

The structurally similar trichosporins B (Fig 1, Tables 2a and b) were isolated from the culture filtrate of Trichoderma polysporum TMI 60146, a destructive mycoparasite of Lentinus edodes (Fujita et al. 1988; Iida et al. 1993). To date, a multitude of biological activities has been reported for trichosporins B, including uncoupling of the respiratory activity of rat liver mitochondria (Fujita et al. 1988; Okuda et al. 1994), Ca²⁺-dependent catecholamine secretion from bovine adrenal medullary chromaffin cells (Tachikawa et al. 1991, 1995, 1996), formation of voltage-gated ion channels (Nagaoka et al. 1995), and antitrypanosomal activity (Iwatsuki et al. 2010).

Minor, 11-residue peptaibols coproduced by Hypocrea pulvinata

The two 11-residue compounds (26) and (28) are typical representatives of the growing SF4 (Chugh & Wallace 2001; Degenkolb et al. 2012). Apparently, a complex mixture of homologues and positional isomers was detected in the plate culture of H. pulvinata I, whereas only trace amounts were found in isolates II and III. However, 11-residue peptaibols were absent in the teleomorphs. Typical signals, indicative of 11-residue peptaibols, viz. pseudomolecular ions m/z 1147.7/1169.7, 1161.7/1183.7, and 1175.8/1197.8 (all [M + H]⁺/[M + Na]⁺) were found. Due to the rather low intensity of these diagnostic ions, which is partly caused by their coelution with major 19-residue peptaibols, only two major peptaibols (26) and (28) could be sequenced (Table 2b, Table S1b). The huge number of positional isomeric SF4 11-residue sequences described to date (Brückner et al. 1993; Krause et al. 2007; Mukherjee et al. 2011; Degenkolb et al. 2012) would require sophisticated methods for chiral sequence analysis (Becker et al. 1997; Jaworski & Brückner 2001) in order to evaluate their novelty or recurrence, respectively. Basically, SF4 11-residue sequences may contribute to the bioactivity of H. pulvinata. The extent of this contribution cannot be evaluated; however, it is regarded as minor because of the comparatively low abundance of these 11-residue peptaibols. Biological activities reported for homologous and/or positional isomeric SF4 11-residue peptaibols comprise induction of voltage-gated ion channels in lipid bilayers (Iida et al. 1995; Wada et al. 1995, 1996), haemolysis of rat erythrocytes (Becker et al. 1997), cytotoxicity towards KB cancer cells (Ruiz et al. 2007), and inhibition of grampositive bacteria (Becker et al. 1997; Berg et al. 2003; Krause et al. 2007).

Thoughts about nonribosomal biosynthesis and module skipping

Only the single anamorphic isolate II has been shown to produce the 20-residue peptaibols (25) and (27), indicating the presence of a respective 20-module nonribosomal peptide synthetase (NRPS). It is thus surprising to detect both 19- and 20-residue peptaibols in this isolate, indicating that a 20-module NRPS may produce 19-residue peptaibols as major products. The exclusive 19-residue content of the other isolates could be interpreted with the deletion of module 6 to a 19-module multienzyme, or by module skipping during peptide elongation. Module skipping could be related to functional properties of this module with respect to binding or processing of Aib intermediates. Likewise Hypocrea pulvinata might harbour 19-module synthetases, while isolate II has a gain of function insertion mutagenesis to exclusively arrive at 20 modules. The 19-module NRPS detected by genomic sequencing in Hypocrea atroviridis had been associated with 19- and 20-residue atroviridins from Trichoderma atroviride (Komon-Zelakowska et al. 2007), but the strain used in that study has later been shown to produce only the 19-residue trichorzianins (Stoppacher et al. 2007), and this NRPS thus has been identified as trichorzianine synthetase. As in 19-residue HPVs, the first Gln of trichorzianins is found in position 6, clearly resulting from a deletion of the 6th module found in a 20-module NRPS synthetase (as paracelsin synthetase, HvD, unpubl. data). We cannot decide on this without knowledge of the respective gene structures of the anamorphs.

Conclusion

Taken these findings together, we dare to predict a mycoparasitic lifestyle of the host-specific polyporicolous² Hypocrea pulvinata:

It has been demonstrated by in vitro-studies, that chitinases and β-1,3-glucanases act synergistically with peptaibiotics in inhibiting spore germination and hyphal elongation of Botrytis cinerea. In the strain ATCC 36042 (=CBS 391.92), which was originally identified as Trichoderma harzianum (el Hajji et al. 1987) but later shown to belong to Trichoderma atroviride (Kuhls et al. 1996), parallel formation of hydrolytic enzymes and 19-residue antifungal trichorzianins A and B is triggered in the presence of cell walls of plant-pathogenic fungi (Schirmböck et al. 1994). Trichorzianins have previously been shown to form voltage-gated ion channels in planar lipid bilayers (Molle et al. 1987), modify the membrane permeability of liposomes, and they are active against Rhizoctonia solani and Phythophthora cactorum. Based on these results, a model of how peptaibiotics such as trichorzianins and hydrolases interact synergistically was proposed: First, the host cell wall is digested enzymatically; thereafter peptaibiotics will penetrate the cell membrane in order to form ion channels. Cell leakage reduces the ability of the host to effectively repair its cell wall. Eventually, inhibition of chitin and β-glucan synthesis further amplifies the destructive effect of chitinases and β-1,3-glucanases (Lorito et al. 1996). These mechanisms, however, may also account for the recently published induction of programmed cell death in plant fungal pathogens (Shi et al. 2012) caused by the 20-residue peptaibol trichokonin VI (=gliodeliquescin A: (Brückner & Przybylski 1984)), from Trichoderma koningii, Trichoderma pseudokoningii, and Trichoderma (syn. Gliocladium) deliquescens,³ the anamorph of Hypocrea lutea (Jaklitsch 2011). The presence of peptaibiotics was also shown to play a role in the induction of plant defence responses (Viterbo et al. 2007).

The data presented here (i) substantiate the hypothesis of a mycoparasitic lifestyle for H. pulvinata, (ii) show the potential for a possible application as a biological control agent, and (iii) indicate the presence of obviously different HPV synthetases (NRPSs) in isolates I, II, and III.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.funbio.2012.10.003.

Abbreviations

QTOF-MS

quadrupole time-of-flight mass spectrometry