Sequences of Acretocins, Peptaibiotics Containing the Rare 1‐Aminocyclopropanecarboxylic Acid, from Acremonium crotocinigenum CBS 217.70

Seven non-ribosomal polypeptide antibiotics, named acretocins (ACRs), were obtained from Acremonium crotocinigenum strain CBS 217.70. The microheterogeneous peptide mixture was isolated from the culture broth by column chromatography on XAD-2 and Sephadex LH-20. Sequences were determined by ESI-MSn and GC-MS. Besides Gly, Leu, and Ala, the peptides contain the non-proteinogenic amino acids Acc (1-aminocyclopropane-1-carboxylic acid), Aib (α-aminoisobutyric acid), Iva (isovaline), Pip (pipecolic acid), β-Ala (β-alanine) and a C-terminal heterocyclic residue N-peptido-1-isobutyl-2-(2,3,4,6,7,8-hexahydro-1-pyrrolo[1,2-a]pyrimidinio)ethylamine (abbreviated X). After vigorous acidic total hydrolysis, release of L-Iva besides D-Iva was established by chiral GC-MS. ACRs show in part sequence identity with neoefrapeptins. Two sequences, ACR-5(6), are new: Ac-L-Pip-Aib-L-Pip-D-Iva-D-Iva(Aib)-L-Leu-β-Ala-Gly-Acc-Aib-L-Pip-Aib-L-Ala-L-Leu-L-Iva-X. The taxonomy of fungal producers of acretocins and neoefrapeptins is compared.


Peptaibiotics -Fungal Peptide Antibiotics Containing α-Aminoisobutyric Acid
About 30 genera of cosmopolitan filamentous fungi, mostly belonging to the order of Hypocreales, have been recognized as prolific sources of a particular group of non-ribosomally synthesized bioactive peptides containing the non-proteinogenic, eponymous α-aminoisobutyric acid (2-methylalanine, Aib). For this group of peptides, the acronym peptaibiotics became established. Peptaibiotics include the large sub-group of peptaibols that are defined as linear, N-acetylated peptides containing Aib and a C-terminal 1,2-amino alcohol. In peptaibiotics, the C-terminal amino alcohol is formally replaced by a structurally diverse substituent, including α-amino acids, polyamines, sugar alcohols, or heterocyclic residues. Besides Aib, this group of unique peptides frequently contains additional nonproteinogenic amino acids including those with quaternary C-atoms, in particular isovaline (2-ethylalanine, Iva), or, as shown here for two groups of peptaibiotics, 1-aminocyclopropane-1-carboxylic acid (Acc).
In the Peptaibol Database, first released in 1997 by Birkbeck College, School of Crystallography, London, UK, [1] 317 peptaibol sequences are compiled. [2] Due to a shortage of external resources, this database has not been updated for more than 13 years. In view of the constantly growing number of peptaibiotics reported in literature, the open domain Comprehensive Peptaibiotics Database became established in May 2013, which comprised the sequences of approximately 1000 peptaibiotics known by December 2012. Based on Microsoft (MS) Access, it could easily be installed and operated on any computer offering a Windows XP/7 environment. Finally, the Peptaibiotics Database was launched in May 2015 as a comprehensive online resource. [3,4] At the time of release, 1297 sequences of peptaibiotics were included. Due to basic infrastructural changes in spring of 2018, this database, which hosted between 1400 and 1500 individual sequences of peptaibiotics in 2017, is currently not available containing 1 cyclotetrapeptaibiotic, [24] whereas the Ivaresidue in FR235222 from Acremonium cf. murorum No. 27082 (= Acremonium sp. FERM BP-6539) was assigned the L-configuration. [25] Later on, the same strain was shown to produce the cyclotetrapeptaibiotic AS1387392, which also contains one L-Iva residue. [26] The atypical, sponge-associated marine fungus Acremonium sp. strain 021172cKZ was a prolific producer of five efrapeptins E, F, G, Eα, and H. [27,28] In addition to those strains discussed in the last paragraph, several other species and strains of Acremonium have been found to produce the marker amino acid Aib. [14] Therefore, it is reasonable to assume that further prolific producers of peptaibiotics will be found in this particular genus. It should be kept in mind, however, that the taxon Acremonium is nowadays rather considered as a generic concept that is still under evaluation using state-of-the-art phylogenetic approaches (see Discussion section).
Furthermore, production of efrapeptins by many species and strains of Tolypocladium was established as reported in the literature. [30,32 -35] Notably, for the first time, peptaibiotic sequences distinguished from efrapeptins in particular by formal substitution of Aib 9 by the cyclic side-chain analog Acc, presence of D-Iva as well as L-Iva, and replacement of Pip by 3-methylproline  in another series were discovered and, therefore, named neoefrapeptins. [36] The fungal producer, however, was assigned as Geotrichum candidum (for taxonomic considerations, see Discussion section). To date, only two more peptaibiotics containing the rare residue X are known, namely adenopeptin from Chrysosporium sp. 2 PF1201 [37] and acremopeptin from Acremonium sp. PF1450, [38] the Nterminus of which is substituted by AcPro instead of AcPip in both cases. A possible biosynthetic pathway of the residue X was proposed by Uma et al., [39] providing evidence that spermidine serves as a linear precursor of the 1,5-diazabicylo[4 : 3 : 0]non-5-ene ring system.

Natural Occurrence of Acc in Peptides and Other Natural Products
Acc in the free and conjugated state occurs in small amounts in fruit such as pears and apples and serves as a biosynthetic precursor of the plant hormone ethylene. [40] It is rarely found as a constituent of microbial metabolites. Cytotrienins A and B are two triene-ansamycins from a taxonomically unidentified Streptomyces sp. RK95-74 (= FERM P-15904), in which a hydroxy group of a 21-membered cyclolactam ring is esterified with N-acylated Acc. [41] From Streptomyces sp. TC 1190, two other triene-ansamycins were obtained, which are structurally closely related to cytotrienins. [42] A third, closely related mixture of triene-ansamycins, UCF116, contains one minor component A, which carries an N-acylated Acc residue. [43] Examples from the fungal kingdom include the two cycloheptapeptaibiotics serinocyclins A and B produced by the entomopathogenic fungus Metarhizium anisopliae var. anisopliae ARSEF #2575 (Hypocreales, Clavicipitaceae). [44] The nonadepsipeptide BZR-cotoxin II is biosynthesized by the ascomycetous fungus Bipolaris zeicola (Pleosporales, Pleosporaceae) race 3, the northern corn leaf spot pathogen. [45] The alkaloid cottoquinazoline A was obtained from a marine strain of Aspergillus versicolor (Eurotiales, Aspergillaceae) MST-MF495. [46] Synthetic Approaches towards Neoefrapeptins and Analogs The first total synthesis of efrapeptin C, notably distinguished from neoefrapeptin N only by Aib 9 in place of Acc 9 , has been described. Notably, the challenging synthesis of the C-terminal X 16 residue has also been achieved. [47] Continuation of this approach using combined solid-and solution-phase synthesis of peptide segments provided efrapeptins D -G and analogs, which also contain L-Iva. [48][49][50] In a previous contribution, we briefly reported on the components and sequence determination of ACRs and assigned the configuration of Iva by chiral GC-MS in hydrolysates (6 M HCl, 110°C, 24 h) . [16] Here, we would like to give a detailed account on the laboratory-scale fermentation, isolation, and determination of the amino acid sequences of ACRs using LC/ ESI-MS and chiral GC/MS. The chiral sequences of ACRs determined in this work are compared with those of neoefrapeptins, and taxonomic relationships of their fungal producers are discussed.

Isolation and Purification of ACRs from the Culture Broth of Fermentations
Fermentation of strain CBS 217.70 was carried out in malt extract medium in mechanically agitated Erlenmeyer flasks. The optimum of the production of peptaibols was monitored by TLC. After 8 days of large-scale fermentation (see Experimental Section), intensive production of ACRs was observed. ACR provided one single spot of R f 0.69 on TLC as revealed by spraying with water and TDM-reagent (Figures 1ad). The mixture of peptides was isolated from the culture broth using XAD-and Sephadex LH-20 chromatography as described in the Experimental Section.

Composition and Chirality of ACR Constituents
The N-trifluoroacetyl-O-2-propyl esters, which are prepared in a two-step derivatization of hydrolysates,were analyzed on Chirasil-L-Val TM (see Experimental Section, GC/MS, Instrument A). This analysis revealed the presence of Gly, L-Ala, and L-Leu as well as L-Pip, β-Ala, Aib, Acc, and DL-Iva (not resolved on this column). The 2 From a chemotaxonomic point of view, it appears that the genus Chrysosporium (Onygenales, Onygenaceae) is not related to any of the 'classical' peptaibiotic-producing genera of fungi. Unfortunately, Hayakawa et al. [37] did not comment on the procedure how strain PF1201 was identified.
absence of Ile in the isolated peptide mixture was also proven by GC/MS. The presence of Acc in ACR hydrolysates was confirmed by GC/MS and diagnostic ions by comparison with an Acc standard ( Figure 2). Release of the C-terminal X-residue was not recognized by GC/ MS but its presence and chemical nature could be deduced from ESI-MS by the diagnostic fragment ion m/z 325.4 ([y 2 + 2H] + , see Table 1) and, in particular, from a neutral loss of 101 Da from the y 6 ion m/z 703.9 and the y 14 ion m/z 1422.6, respectively (see Figure 7 and Discussion section). Notably, only D-Iva was found when the ACR peptide mixture was hydrolyzed under standard conditions (6 M

HPLC and ESI-MS of ACRs
The analytical HPLC elution profile of the mixture of ACRs is shown in Figure 4; the total ion current of the isolated ACR peptide mixture in positive ion mode ESI-MS is displayed in Figure 5. Relative quantities of peptides in the ACR mixture were calculated from peak heights at λ = 205 nm and included in Table 2. Retention times of peptides increase with increasing molecular weight. The increasing hydrophobicity of peptides of the same molecular weight in dependence of positions of their amino acid methylene homologs (� 14 Da) is also reflected in the HPLC elution profile. In peptides of identical molecular weight, which are distinguished by the positions of Aib and Iva (m/z 1618 and m/z 1632), a C-terminal Iva residue increases the hydrophobicity more in comparison to C-terminal Aib (ACR-3 vice ACR1a, and ACR-4 vice ACR-2). Replacement of Gly 13 by Ala 13 increases hydrophobicity of otherwise identical sequences (ACR-5 vice ACR-3 and ACR-6 vice ACR-4). Furthermore, formal replacement of Iva 5 and Gly 13 by Aib 5 and Ala 13 increases the hydrophobicity of the latter despite identical molecular weights of m/z 1646 (ACR-5 vice ACR-4). Finally, replacement of Aib 5 by Iva 5 results in the most hydrophobic peptide of the series (ACR-6 vice ACR-5); for sequences see Table 2.

Sequencing of ACR Peptides
In the following, we comprehensively discuss molecular and fragment ions compiled in Table 1, and sequences derived thereof that are presented in Table 2. Diagnostic differences (Da) between fragment ions of the same series are attributed to amino acid residues as follows: Ac (43), Pip (111), Iva (99), Aib (85), Leu (113), β-Ala/Ala (71), Gly (57), Acc (83), and X (224). Note that exchange of β-Ala by Ala in the same position has not yet been reported in peptaibiotics, and it is not likely to occur in the non-ribosomal biosynthesis pathways of this class of peptides. Consequently, this particular exchange has not been taken into account. Because Ile was not detected by GC/MS in ACR hydrolysates, exchange of Leu/Ile was not considered here.
The corresponding [M + Na] 2 + ions were also observed. Notably, the intensive ion m/z 325.4 is in agreement with [y 2 + 2H] + fragment ions expected to be generated from the C-terminal sequences Iva 15 -X 16 . The corresponding ion m/z 309, expected to be generated from the C-terminal fragment Aib 15 -X 16 , was not recorded. The fragment ion m/z 281.3 observed in the HPLC-ESI-MS spectrum might result from the b 3 fragment ion (m/z 350.9) that loses acetyl (43 Da) and carbonyl (28 Da; see Figure 6).
All peptides provided identical series of b 1 to b 3 fragment ions. Consequently, the N-terminal sequence Ac-Pip 1 -Aib 2 -Pip 3 was deduced. A difference of 14 Da between fragment ions b 4 and b 5 established the presence of either Iva or Aib in position 4 and 5 of the ACR 2 -6. Sequences of the isobaric peptides ACR 1a and 1b were deduced from the difference of b 5 -b 3 , corresponding to 183.4 (Iva-Aib) or 169.6 (Aib-Aib), respectively. These positions were corroborated by y 12 and y 13 fragment ions, also differing by 14 Da. Notably, almost complete regular series of y 7 -y 15 fragment ions were observed. Release of fragment ions y 7 -y 11 also confirmed the identical sequence Leu 6 -β-Ala 7 -Gly 8 -Acc 9 -Aib 10 for all ACR peptides.
As a general rule, intensive y 6 and y 14 fragment ions were observed, resulting from cleavage of the Aib 2 -  Table 1 and Figure 7). The neutral loss of 101 Da corresponds to the molecular formula C 6 H 15 N that is attributed to the release of the corresponding dialkylamine residue (ethylbutylamine or dipropylamine) from the C-terminal residue X (m/z 224) (see Figure 8). This observation definitely proves the structure of the C-terminal residue X.
For all ACR peptides, almost regular series of y 3 -y 15 fragment ions could be observed, with the exceptions of y 9 (resistance of β-Ala 7 -Gly 8 cleavage) and y 5 in minor peptides ACR 1a and 1b. Accordingly, cleavage of the β-Ala 7 -Gly 8 bond, yielding the b 7 fragment ions, was either not detected or provided ions of very low intensity (ACR-1a and ACR-3).
Fragment ions of the series y 3 -y 11 provided identical sequence domains Leu 6 -β-Ala 7 -Gly 8 -Acc 9 -Aib 10 -Pip 11 -Aib 12 -Gly 13 for ACR 1-4 peptides and presence of Ala 14 in ACR peptides 5 and 6. The positions of Aib and Iva in the C-terminal sequence Leu 14 -Aib 15 /Iva 15 -X 16 were deduced from y 2 and y 3 fragment ions. Exchange of Gly 13 by Ala 13 in ACR-5 and ACR-6 was concluded from the series of y 2 -y 6 fragment ions.
Detailed sequencing of ACR peptides is illustrated in the following with the new acretocin peptide 6. The + ESI-MS total ion current (TIC) of the isolated ACR peptide mixture is depicted in Figure 5, with ACR-6 eluting at 23.93 -24.7 min. Diagnostic fragment ions b 1 -b 11 and y 6 -y 15 are assigned in a CID-MS, depicted in Figure 9. ESI-CID-MS and ESI-MS 2 using m/z 1660 as parent ion provided the series b 5 -b 15 and y 4 -y 15 fragment ions as well as diagnostic fragment ions at y 6 -101 Da and y 14 -101 Da (Figures 7 and 9).
Owing to the presence of Ala 13 in ACR-5 and ACR-6, these peptides represent new sequences. The sequences of ACR-1b, 3, and 4 correspond to neoefrapeptins (NEF)-D, A, and B, respectively (see Table 2). Since about 20 % L-Iva -besides abundant D-Iva -is released under vigorous hydrolytic conditions (see Figure 3), the presence of L-Iva 15 in ACR-peptides is assigned by analogy with neoefrapeptins.

Configuration of Amino Acids and Unusual Hydrolytic Stability of the C-Terminal Iva 15 -X 16 Bond in Acretocins and Structurally Related Peptides
Presence and configuration of chiral amino acids in ACRs were determined by GC/MS on Chirasil-L-Val column. Since only D-Iva was detected in acidic standard hydrolysates, exclusively D-Iva was assigned to positions 4 and 5. Based on analytical data, sequence identity of ACR 1b, 3, and 4 with neoefrapeptins D, A, and B, respectively, was recognized, the major difference being detection of exclusively D-Iva in ACR hydrolysates. However, in the course of a  Chem. Biodiversity 2019, 16, e1900276 discussion on neoefrapeptins we became informed [52] that in acidic total hydrolysates of the intact neoefrapeptin A under standard conditions (6 M HCl, 110°C, 24 or 48 h, respectively) the ratio D/L was not 50:50, as expected, but about 85:15. In order to clarify this ambiguity, enzymatic cleavage of neoefrapeptins  using papain was performed providing peptides which could be separated by HPLC. On total hydrolysis of isolated peptide exclusively L-Iva (if present) was released from the isolated peptides Leu 13 -Iva 14 -X 15 and β-Ala 7 -Gly 8 -Acc 9 -L-Iva 10 -Pip 11 -Aib 12 -Gly 13 , but only D-Iva from N-terminal peptides AcPip 1 -Aib 2 -Pip 3 -Iva 4 -Iva 5 -Leu 6 . Thus, in a laborious approach, the location of Ivaenantiomers in neoefrapeptins could ultimately be assigned. The decelerated release of C-terminal L-Iva from the intact peptide was explained by the positively charged X 16 residue. [36] Bullough et al., [31] however, determined L-Iva 15 in hydrolysates of efrapeptin D (6 M HCl, 105°C, 48 h) using GC/MS on a 'chiral column' without providing further specifications. Since only a single C-terminal Iva occurs in efrapeptin D, even partial release would provide evidence for L-Iva. Resistance of the 14-residue peptaibiotic adenopeptin, which also carries the C-terminal dipeptide Iva 13 -X 14 , to acidic hydrolysis was briefly mentioned whilst commenting on amino acid analysis. [37] These authors confirmed previous observations by Gupta et al. [30] who noted that the C-terminal capping group of efrapeptins was rather difficult to hydrolyze.
Such an unexpected resistance to standard hydrolysis conditions (6 M HCl, 110°C, 24 or 48 h), preventing release of the C-terminal 1,2-diamino-4-methylpentane unit from Ile or Leu, was also recognized in cicadapeptins I and II, peptaibiotics from the entomopathogenic fungus Cordyceps heteropoda ARSEF #1880 (� Ophiocordyceps heteropoda: Hypocreales, Cordycipitaceae). [35] Taking all data together, this indicates -at least for peptaibiotics -general resistance towards acidic hydrolysis of C-terminal residues from peptides of the structure -Aaa-X (Aaa, αamino acid including Aib and Iva; X, amide-bound, positively charged capping group), preventing or hampering release of the preceding C-terminal amino acid.
This information caused us to subject a sample of the ACR mixture to severe hydrolysis conditions (37 % HCl, 135°C, 72 h). Indeed, L-Iva was released as revealed by analyzing the N-acetyl-Iva-O-2-propyl esters on Chirasil-L-Val (see Figure 3). Notably, release of 20 % L-Iva from the ACR mixture was observed. From that, the presence of terminal L-Iva 15 residues is concluded for the corresponding sequences. Our observation is in analogy to neoefrapeptins and the closely related efrapeptins. [29,36] An acid-catalyzed partial enantiomerization of D-Iva, located in positions 4 and 5 of ACRs, is excluded because treatment of enantiomerically pure D-Iva under these drastic con-  [53] Chiral resistance of Iva against alkaline hydrolysis (2 N NaOH, 16 h, 100°C) has already been reported by Fischer and von Grävenitz. [54] To summarize, based on i) the release of L-Iva from ACR on vigorous acidic hydrolysis, ii) analogy of sequences of some ACR peptides with neoefrapeptins, and iii) presence of a C-terminal L-Iva also in efrapeptins, the configuration of D-Iva 4,5 and L-Iva 15 as depicted in Figure 3 and Table 2 is assumed. The new acretocins 5 and 6 contain Ala in positions 13, whereas this position is occupied by Gly in all neoefrapeptins. In contrast to the neoefrapeptin producer DSM 15068, the ACR producer CBS 217.70 does not biosynthesize peptides containing 3-MePro. Thus, the architecture of the nonribosomal peptide synthetase of the acretocin producer differs to some extent from that one of the neoefrapeptin producing strain.

Taxonomic Considerations of the Neoefrapeptin and Acretocin Producers
Acremonium crotocinigenum has originally been described as Cephalosporium crotocinigenum, [55] which was later transferred to the genus Acremonium. [17] Based on morphological characteristics, the species A. crotocinigenum was considered as an intermediate between the genera Acremonium and Cylindrocarpon. In 2011, Summerbell et al. [56] published a phylogenetic revision of Acremonium and the genera closely related to it, i. e., Gliomastix, Sarocladium, and Trichothecium. Molecular sequencing of the nuclear ribosomal large subunit (nucLSU) and second largest subunit (SLU) revealed a close relationship of A. crotocinigenum CBS 129.64 (= type culture of A. crotocinigenum isolated from the white rot fungus Trametes versicolor) with 'Trichothecium indicum' CBS 123.78, T. roseum DAOM 208997, and T. sympodiale ATCC 36477, which formed a separate 'Trichothecium clade'. Based on these results, the authors recombined A. crotocinigenum and Spicellum roseum in Trichothecium, irrespective of different morphological characters and different modes of conidiogenesis. The question as to whether the producer of acretocins, A. crotocinigenum CBS 217.70, has to be included in the genus Trichothecium cannot be resolved for the time being because no detailed phylogenetic analysis of this particular strain has been performed up to now.
To continue, neoefrapeptins A -N have been reported from a filamentous fungus originally identified and deposited as Geotrichum candidum SID 22780 (= DSM 15068). Given that this strain would have been correctly identified, this would have made it the first peptaibiotic-producing yeast-like organism (Saccharomycetales, Dipodascaceae) reported in literature. Careful light-microscopic examination of plate cultures of DSM 15068 revealed that no yeast-like organism was present (see Experimental Section). The strain formed a thin white mycelium, rich in septations, but no reproductive structures such as conidiophores and conidia could be observed. Consequently, a taxonomic reinvestigation based on marker genes of DSM 15068 was ordered by us (H.B. and T.D.) at DSMZ (Braunschweig, Germany). Sequencing of the large subunit rRNA gene (LSU rRNA = nucLSU) revealed that DSM 15068 is, in fact, closely related to Trichothecium indicum (syn. Leucosphaerina indica) and Trichothecium crotocinigenum (syn. Acremonium crotocinigenum, type strain CBS 129.64). In contrast, little similarity was observed in  Table 1).
Chem. Biodiversity 2019, 16, e1900276 internal transcribed spacer (ITS) 3 sequences of T. crotocinigenum type strain CBS 129.64 and DSM 15068; and no ITS sequences are publicly available for T. indicum. Consequently, there is strong evidence that the neoefrapeptin-producer 'Geotrichum candidum' DSM 15068 has been misidentified and does not belong to the genus Geotrichum. Like the acretocin producer A. crotocinigenum CBS 217.70, it may also represent a new species. In conclusion, since efrapeptins biosynthesized by numerous species of the fungal genus Tolypocladium (Hypocreales, Ophiocordycipitaceae), comparison of the respective peptide sequences with those of neoefrapeptins and acretocins from a chemotaxonomic point of view is of limited value. Non-ribosomal biosynthesis of acretocins and neoefrapeptins is, of course, independent of changing taxonomic considerations of their fungal producers.
TLC. For analytical TLC, glass plates pre-coated with silica gel 60 F 254 (10 × 20 cm, 0.25 mm layer thickness, Merck, Darmstadt, Germany) were used. TLC was performed in glass chambers (Desaga, Wiesloch, Germany) coated with filter paper to saturate the atmosphere. The distance from start to solvent front was usually 7.5 -8 cm. The mobile phase was composed of CHCl 3 /MeOH 8:2 (v/v). ACRs provided white spots of R f 0.69 by spraying with water (see Figure 1a and 1c). After nearly drying the plates, treatment with chlorine for 20 min (generated in a desiccator from KMnO 4 and conc. HCl) was performed, followed by cold blow-drying to remove excess of chlorine and spraying with the TDM-reagent, dark-blue spots were furnished indicating peptides (see Figure 1b and 1d).

Cultivation, Laboratory-Scale Fermentation, and Product Monitoring of Strain CBS 217.70
The fungus was obtained as a freeze-dried culture. This strain is still available under this accession number from Westerdijk Fungal Biodiversity Institute, CBS-KNAW, Fungal Culture Collection, Utrecht, The Netherlands. The herbarium specimen of this strain is deposited at CBS with accession number CBS H-8135. After conditioning of the inoculum in sterile 0.9 % NaCl solution, it was grown in Petri dishes on malt extract agar (MEA), consisting of [g/l] malt extract 30 (Servabacter light, No. 28397, Serva, Heidelberg, Germany); soy peptone 3 (Oxoid Unipath, Wesel, Germany), and agar-agar 15 g/l (No. 1614, Merck, Darmstadt, Germany). Liquid malt extract media (MEM) was prepared without agar and adjusted to pH 5.6 + /À 0.1 prior to sterilization. Vigorous growth of the fungus on MEA was observed after 12 days at 25°C, and the color of the aerial mycelium changed from white to beige. Subcultures were prepared by inoculating two baffled 2-l Erlenmeyer flasks, each containing 400 ml of MEM, with sterile discs of 1.5 cm diameter. After 8 days of gentle shaking at room temperature, the production of ACRs was monitored daily by TLC as follows. For solid phase extraction (SPE) of culture broths, commercial Sep-Pak C18 cartridges (Waters/Millipore, Eschborn, Germany) or laboratory-made cartridges (20 × 10 mm) filled with LiChroprep RP-8, particle size 40 -63 μm (Merck, Darmstadt, Germany) were used. Cartridges were connected to a syringe with Luer-lock TM tip and conditioned with Me 2 CO, MeOH, and water (20 ml each). After that, 20 ml aliquots of culture filtrates resulting from fermentations were applied. After washing with 20 ml of water, the adsorbed peptides were eluted with 5 ml MeOH. After evaporation to dryness, the remaining residues were dissolved in 500 μl MeOH and aliquots of 1 -10 μl analyzed by TLC or used for other analytical procedures.
After 13 days, ACR production was observed in the pre-cultures, of which 20 ml aliquots were used for inoculating 12 2-l baffled Erlenmeyer flasks, each one containing 400 ml of MEM. Vigorous fungal growth was observed already after 24 h. Production of ACRs was monitored again by TLC. After eight days of fermentation, intensive production of ACRs was detected. Hence, the mycelium was removed by filtration using a nylon cloth followed by vacuum-assisted filtration using filter paper and final centrifugation at 3500 rpm. A total of 4.7 l of clear culture broth and 215 g of wet mycelium were obtained. The mycelium was not analyzed further for ACRs.

Cultivation and Light-Microscopic Examination of Strain DSM 15068
The fungus was obtained from DSMZ as a living culture on HA agar upon the rights granted by patent law. [59] Upon arrival, it was streaked out on HA agar plates and grown for five days at 25°C. The actively growing strain was transferred to new HA agar plates and grown up to five weeks at 25°C. The fungus was also cultivated on TSM-37 agar plates for up to 19 days at 25°C. [59] Plate cultures growing on HA or TSM-37, respectively, were regularly (every seven days) examined under the light microscope after staining with lactic acid-cotton blue (glycerol, 20 ml; lactic acid, 20 ml; aqua dist., 20 ml; cotton blue, 50 mg). [60]

XAD-Medium Pressure Liquid Chromatography
A Labomatic MPLC System (Sinsheim, Germany) comprising pump, glass columns, and fraction collector was used. The filtrate (4.7 l) was pumped through a heavy-wall glass column (38 × 3.7 cm i.d.) filled with XAD-2 resin at a flow rate 20 ml/min). Peptides were eluted with a gradient from 40 % to 100 % MeOH at a flow rate of 5 ml/min. Fractions of 25 ml were collected, and the elution of peptides was monitored by TLC, applying 10 μl samples (see Figure 1a and 1bd). ACRs, displaying R f 0.69, were abundant in fractions 50 -59. They were combined, yielding 615 mg of solid material.

Sephadex LH-20 Chromatography
The crude peptide mixture (615 mg) obtained by XAD chromatography was dissolved in 10 ml MeOH and subjected to Sephadex LH-20 gravity-flow column chromatography (glass column, 100 cm × 3.1 cm i.d.). Two samples of 5 ml were applied onto the column. MeOH was used as an eluent at a flow rate of 2.5 ml/ min. Elution of peptides was monitored by TLC. The first 150 ml of the eluate were discarded; afterwards 15 ml fractions were collected. Fractions 8 -10 from first and fractions 7 -10 from second run (see Figure 1c and 1d) were combined. After evaporation to dryness, 184 mg of ACR peptides were obtained. The instrument was run in total-ion current (TIC) and in selected-ion monitoring mode (SIM), respectively, using electron impact (EI) ionization at 70 eV. The Ntrifluoroacetyl-amino acid 2-propyl esters were prepared as follows: to the dry residue resulting from the total hydrolysate of 0.2 mg ACR peptide, 500 μl AcCl in 2-PrOH (8 : 2; v/v) were added and esterified in a closed reaction vessel for 1 h at 100°C. After evaporation to dryness in a stream of nitrogen, 300 μl CH 2 Cl 2 and 50 μl TFAA were added for trifluoroacetylation. This solution was heated for 20 min at 100°C. After evaporation to dryness in a stream of cold N 2 , 100 μl CH 2 Cl 2 were added and 1 μl aliquots injected into the GC in split mode (ca. 1 : 30). The temperature of injector and interface was set at 250°C. The temperature program was 70 -100°C (2.5°C/min); 100-135°C (3.5°C/min); 135 -150°C (5°C/min); 150 -190°C (20°C/ min). Inlet pressure of helium was 5 kPa for 1 min, increased at 0.2 kPa/min to 7 kPa, then 0.3 kPa/min to 11 kPa, then 1.6 kPa/min to 15.0 kPa. Diagnostic ions [m/z] resulting either from C α or ester cleavage (the latter in parentheses) of N-trifluoroacetyl-amino acid-O-2-propyl esters were: Aib (154); D/L Iva, not resolved (168); L-Ala 140; Gly 126 (154); β-Ala 140 (168); L-Pip (180); Acc 152 (179/180); L-Leu (182). For GC and MS fragmentation scheme of Acc used for identification see Figure 2.
GC/MS (Instrument B). The chirality of D-and L-Iva in ACR hydrolysates, which resulted from vigorous hydrolysis using fuming or concentrated HCl, was determined using a 6890N gas chromatograph with mass-selective-detector MSD 5973 injector 7673 (all from Agilent Technologies, Waldbronn, Germany). The GC was equipped with a Chirasil-L-Val capillary column (25 m × 0.25 mm i.d., film thickness 0.12 μm), (from Varian-Chrompack, Darmstadt, Germany). The instrument was run in the electron impact ionization mode at 70 eV. The MSD was set at mass range m/z 35 -350 using the full scan mode. Helium was used as carrier gas at 6.5 kPa inlet pressure, inlet temperature 220°C, quadrupole temperature 150°C, ion source temperature 230°C. The N-acetyl-DL-Iva-2-propyl esters were prepared as follows: to the dry hydrolysate resulting from vigorous total hydrolysis was esterified as described before. For subsequent acetylation, 300 μl CH 2 Cl 2 and 50 μl Ac 2 O were added, and the mixture was heated for 20 min at 100°C. After evaporation to dryness in a stream of nitrogen, 100 μl CH 2 Cl 2 were added and 1 μl aliquots injected into the GC in split mode (ca. 1 : 30). The temperature of the oven was kept at 85°C for 10 min, then ramped to 200°C in 25 min, and held for 5 min. For GC/MS, see Figure 3.