Streptopeptolin, a Cyanopeptolin-Type Peptide from Streptomyces olivochromogenes.

Cyanopeptolin-type peptides are cyclic depsipeptides that commonly have 3-amino-6-hydroxy-2-piperidone (Ahp) unit in the molecules. So far, cyanopeptolin-type peptides have been isolated as protease inhibitors from a wide variety of cyanobacteria. In the course of screening for new peptides, a new peptide streptopeptolin, which had the similar structure to cyanopeptolin, was isolated from the extract of Streptomyces olivochromogenes NBRC 3561. Streptopeptolin is the first cyanopeptolin-type peptide isolated from actinobacteria. The structure of streptopeptolin was determined by the analysis of electrospray ionization mass spectrometry and NMR to be cyclic depsipeptide containing unusual amino acids, Ahp, and N-methyl tyrosine. As a result of protease inhibition test, streptopeptolin showed inhibitory activity against chymotrypsin. The whole genome sequence data of S. olivochromogenes revealed the biosynthetic gene cluster for streptopeptolin, which encoded a nonribosomal peptide synthetase. We proposed a biosynthetic pathway of streptopeptolin based on bioinformatics analysis.

Introduction

Cyanobacteria in the genera Microcystis, Anabaena, Planktothrix, Nostoc, Lyngbya, and Schizothrix are known to produce a wide variety of bioactive peptides.[1] According to the characteristics of chemical structure, the peptides are classified into the classes of cyanopeptolins, aeruginosins, microginins, anabaenopeptins, and microviridins.[2] Cyanopeptolin-type peptides are cyclic depsipeptides, which commonly have 3-amino-6-hydroxy-2-piperidone (Ahp) unit in the molecules.[3−8] Cyanopeptolins A–D were first isolated from Microcystis sp. PCC 7806.[3] From then, cyanopeptolin-type peptides have been isolated mostly as protease inhibitors from a wide variety of cyanobacteria with different names (micropeptins[9−11] and aeruginopeptins[12] from Microcystis aeruginosa; nostopeptins[13] from Nostoc minutum; nostocyclin[14] from Nostoc sp.; oscillapeptins[15] from Planktothrix agardhii; and somamides A and B[16] from a cyanobacterial assemblage of Lyngbya majuscula and Schizothrix species). Occurrences of cyanopeptolin-type peptides were also reported in water blooms all over the world;[7,8,10,17] however, their function in the natural environment is not clear. Cyanopeptolin-type peptides have been reported to be biosynthesized by nonribosomal peptide synthetase (NRPS) system.[18−20] The biosynthetic gene cluster of anabaenopeptilide was reported to contain three NRPS genes including apdA, apdB, and apdD.[21] Nishizawa et al. reported that a cyanopeptolin class peptide micropeptin K-19 was also biosynthesized via NRPS system.[19] In addition, similar NRPS genes were reported to be conserved over many cyanobacterial strains in the genera Planktothrix, Microcystis, and Anabaena.[18,19] The cyanopeptolin analogous peptide named crocapeptin was isolated from the myxobacterium Chondromyces crocatus Cm c5 and the biosynthetic gene cluster of crocapeptin was indicated to include one large NRPS protein (CpnD) and tailoring enzymes CpnE and CpnF, which convert a proline residue into Ahp.[22,23]

Actinomycetes are soil bacteria, which produce a wide variety of secondary metabolites including peptides. In actinomycetes, biosynthesis via NRPS is involved in the production of pharmaceutically important bioactive peptides such as daptomycin,[24] vancomycin,[25] and bleomycin.[26] In the course of chemical screening for new peptides using high-performance liquid chromatography (HPLC) coupled with diode array detection and electrospray ionization mass spectrometry (ESI-MS), we found a new cyanopeptolin-type peptide streptopeptolin from Streptomyces olivochromogenes NBRC 3561. S. olivochromogenes is an important strain that produces xylose isomerase in food industry.[27] To the best of our knowledge, this is the first report for the isolation of cyanopeptolin-type peptide from actinobacteria. We found the biosynthetic gene cluster encoding a NRPS for streptopeptolin from whole genome data of S. olivochromogenes NBRC 3561.[28] Here, we describe isolation and structure determination of streptopeptolin (1) from S. olivochromogenes NBRC 3561.

Results and Discussion

The new peptide streptopeptolin (1) was isolated from the extract of culture of S. olivochromogenes NBRC 3561. The molecular formula of 1 was established to be C46H61N9O13 by accurate ESI-MS analysis, as the ion corresponding to [M + H – H2O]+ (the calculated m/z value, 930.4361) was observed at m/z 930.4395. To determine the structure, the NMR spectra of 1 including 1H, 13C, DEPT-135, double-quantum-filtered correlation spectroscopy (DQF-COSY), total correlation spectroscopy (TOCSY), nuclear overhauser effect spectroscopy (NOESY), rotating-frame overhauser effect spectroscopy (ROESY), heteronuclear multiple bond correlation (HMBC), and heteronuclear single quantum coherence (HSQC) were obtained using the solvent (0.5 mL, MeCN-d3/DMSO-d6; 4:1). In the 1H NMR spectrum, seven α-protons of amino acid residues (3.71, 4.18, 4.37, 4.55, 4.62, 4.86, and 4.93 ppm) and five amide protons (7.21, 7.54, 7.65, 7.69, and 8.19 ppm) were observed. The assignments of the constituent seven amino acids including Gln1, Thr2, Gln3, Ahp4, Phe5, N-methyl tyrosine6 (N-Me-Tyr6), and Ala7 were completed using spin system identification (Figure and Table ). The construction of Ahp moiety was accomplished mainly by the TOCSY and HMBC spectra. The TOCSY spectrum indicated the spin system from amide proton to H-δ (bold line in Figure ) in the Ahp unit. The HMBC correlations from H-α (δH 3.71) and H-δ (δH 5.14) to C=O (δC 170.4) confirmed the cyclic structure in the Ahp unit. The chemical shifts of Ahp4 in 1 were similar to those of micropeptins. The two amino acid sequences (Gln1–Thr2–Gln3–Ahp4 and N-Me-Tyr6–Ala7) were determined by HMBC correlations from α-protons (H-α) and amide protons to carbonyl carbon (C=O) in adjacent amino acids (arrow in Figure ). The amide proton of Phe was not observed, and the HMBC correlations (H-δ of Ahp4/C-α of Phe5, H-α of Phe5/C-δ of Ahp4, and H-α of Phe5/C=O of Ahp4) indicated the presence of a tertiary amide in Phe and a connection between Ahp4 and Phe5. The HMBC correlation from methyl residue (δH 2.74) to C-α (δC 62.4) confirmed the presence of N-Me-Tyr6. The HMBC correlation from methyl residue (δH 2.74) to C=O (δC 171.9) indicated the connection between N-Me-Tyr6 and Phe5. The HMBC correlation from H-β (δH 5.35) in Thr to C=O (δC 174.8) in Ala indicated the lactone structure. The TOCSY and DQF-COSY correlations indicated the 1H spin system between H-3 (δH 6.42) and H-4 (δH 1.72) in methyl-2-butenoic acid (Mba). The HMBC correlations from H-2-Me (δH 1.79) to C-2 (δC 132.7), C-3 (δC 131.6), and C=O (δC 170.1) confirmed the presence of methyl-2-butenoic acid. The HMBC correlation from amide proton (δH 7.54) in Gln1 to C=O (δC 170.1) in Mba indicated the connection between Gln1 and Mba at the N-terminal of the peptide. The chemical shifts of C-4 (δC 14.2) and C-2-Me (δC 12.8) in Mba were compared with those of tiglic acid (E-methyl-2-butenoic acid, δC 13.8 and δC 10.9) and angelic acid (Z-methyl-2-butenoic acid, δC 15.9 and δC 20.2).[29,30] In addition, no correlation in NOESY and ROESY experiments was observed between H-3 and H-2-Me. The orientation of the double bond in Mba was assigned as E, considering the similarity of chemical shifts to tiglic acid.

Figure 1

Key two-dimensional (2D) NMR correlations of streptopeptolin (1).

Table 1

NMR Chemical Shift Values of 1 in MeCN-d3/DMSO-d6 (4:1)

unit	position	δH (J = Hz)	δC
Mba	CO		170.1
	2		132.7
	3	6.42 (m)	131.6
	4	1.72 (d, 6.9)	14.2
	2-Me	1.79 (s)	12.8
Gln1	CO		173.8
	NH	7.54 (d, 7.0)
	α	4.37 (q, 7.0)	54.2
	β	1.96 (m)	28.2
	γ	2.26 (m)	32.4
	δ		176.1
	δ-NH2	6.26 (brs), 7.02 (brs)
Thr2	CO		170.8
	NH	7.65 (d, 8.8)
	α	4.55 (m)	56.4
	β	5.35 (m)	73.3
	γ	1.23 (t, 6.8)	18.3
Gln3	CO		171.2
	NH	8.19 (d, 8.4)
	α	4.18 (m)	53.6
	β	1.65 (m), 2.18 (m)	27.2
	γ	2.09 (m)	32.5
	δ		175.5
	δ-NH2	6.10 (brs), 6.76 (brs)
Ahp4	CO		170.4
	NH	7.21 (d, 9.0)
	α	3.71 (m)	50.0
	β	1.65 (m), 2.35 (m)	22.9
	γ	1.64 (m), 1.75 (m)	30.4
	δ	5.14 (brs)	75.3
	OH	5.81 (brs)
Phe5	CO		171.9
	α	4.86 (dd, 11.6, 4.3)	52.0
	β	1.89 (m), 2.87 (m)	36.3
	γ		138.0
	δ	6.87 (d, 7.3)	130.5
	ε	7.20 (dd, 7.3, 7.1)	128.9
	ζ	7.15 (d, 7.1)	127.3
N-Me-Tyr6	CO		170.0
	N-Me	2.74 (s)	30.7
	α	4.93 (dd, 11.5, 3.2)	62.4
	β	2.70 (m), 3.18 (m)	33.6
	γ		129.1
	δ	7.04 (d, 8.4)	131.5
	ε	6.81 (d, 8.4)	116.6
	ζ		157.5
	OH	NDa
Ala7	CO		174.8
	NH	7.69 (d, 8.5)
	α	4.62 (m)	48.2
	β	1.24 (t, 6.8)	18.6

ND: not detected.

To elucidate the absolute stereochemistries of Gln, Thr, Ala, Phe, and N-Me-Tyr in 1, the modified Marfey’s method was applied.[31] The hydrolysate of 1 was derivatized with N-α-(5-fluoro-2,4-dinitrophenyl)-l-leucinamide (l-FDLA). The derivative was subjected to HPLC analysis to compare with standard amino acid derivatives. As a result, the stereochemistries of Gln, Thr, Ala, Phe, and N-Me-Tyr in 1 were determined to be L. Regarding the relative stereochemistry of Ahp, Ahp configuration of 1 was determined to be 3S, 6R or 3R, 6S by ROESY correlations (Figure ). Considering the biosynthesis of streptopeptolin by NRPS (described in Results and Discussion), which was similar to that of crocapeptin, the stereochemistry of Ahp in 1 was proposed to be 3S, 6R.

Figure 2

ROESY correlations in Ahp.

Because cyanopeptolin-type peptides were reported to have inhibitory activity against serine proteases such as trypsin and chymotrypsin,[11,15] the inhibitory activity of 1 was tested against trypsin and chymotrypsin. As a result, compound 1 inhibited chymotrypsin with IC50 of 5.0 μg/mL and did not inhibit trypsin at the concentration of 50 μg/mL. Crystal structures of the cyanopeptolin-type peptides in a complex with serine proteases indicate that inhibition is based on a substrate-like binding mode.[32,33] Streptopeptolin contains a hydrophobic amino acid Phe, which is a substrate for chymotrypsin. However, no basic amino acid is present in the molecule. Considering the inhibition mechanism,[32,33] the structural characteristic of 1 may explain the specific inhibitory activity against chymotrypsin.

Cyanopeptolin-type peptides are biosynthesized via NRPS pathways.[18,19] As we recently conducted whole genome sequencing of S. olivochromogenes NBRC 3561,[28] we searched the gene cluster for streptopeptolin synthesis in the draft genome sequence. The genome encodes four potential NRPS gene clusters,[28] among which three clusters in BDQI01000038, BDQI01000046, and BDQI01000077 each contain only two modules at most. The remaining cluster in BDQI01000045 encodes an NRPS comprising seven modules, as shown in Table S1 and Figure , consistent with that of amino acid residues in streptopeptolin (1). Analysis using anti-SMASH[34,35] suggested that substrates of adenylation (A) domains were as follows: 2nd residue (Thr), 4th residue (Pro), 5th residue (Phe), 6th residue (Tyr), and 7th residue (Val). The 1st and 3rd residues could not be predicted by the program. Although 7th residue did not match (predicted residue Val for Ala in 1), the predicted residues at 2nd, 4th, 5th, and 6th matched with the determined chemical structure of streptopeptolin (1), considering Ahp was proposed to be biosynthesized from Pro.[22] The two genes (spnB and spnC) in the gene cluster were orthologues to cpnF and cpnE, which possibly convert Pro to Ahp in crocapeptin biosynthesis (Figure ). The predicted peptide sequence had Pro at the 4th position from N-terminus, which corresponded to Ahp residue in the determined structure of streptopeptolin (1). These similarities with crocapeptin biosynthesis strongly suggested that the biosynthesis of streptopeptolin (1) was via NRPS. To eliminate the possibility that streptopeptolin belongs to ribosomally synthesized and post-translatinally modified peptides, the peptide that had the sequence of Gln–Thr–Gln–Pro–Phe–Tyr–Ala was searched in the genome of S. olivochromogenes NBRC 3561 using antiSMASH and BLASTp, and no possible peptide was found by the search. Therefore, we propose the cluster to be responsible for the synthesis of streptopeptolin.

Figure 3

Proposed biosynthetic pathway of streptopeptolin (1). Capital letters A, C, MT, and TE represent adenylation, condensation, methyltransferase, and thioesterase domains, respectively. Black circle represents peptidyl carrier protein (PCP) domain. Locus tags are as follows: spnA: SO3561_09657, spnB: SO3561_09655, and spnC: SO3561_09654.

Next, we propose a biosynthetic pathway via NRPS (spnA: SO3561_09657), as shown in Figure . Each amino acid as a building block is converted to aminoacyl adenylate by each A domain and transferred on the adjacent peptidyl carrier protein (PCP) domain within each module to form the corresponding aminoacyl thioesters. At module 1, Mba is bound to Gln by the condensation (C) domain. The starter Mba–Gln is then elongated by the incorporation of Thr, Gln, Pro, Phe, Tyr, and Ala by successive N-acylation by C domains of modules from 2 to 7. Tyr loaded onto module 6 undergoes N-methylation by the methyltransferase (MT) domain. The thioesterase (TE) domain at the C-terminal of the NRPS (SpnA) finally releases the elongated linear heptapeptidyl thioester from the PCP domain of module 7 by intermolecular lactonization via the hydroxyl group of Thr and carboxyl terminus of Ala. The lactonized compound is hydroxylated by cytochrome P450 (spnB: SO3561_09655, cpnF orthologue) at the Pro residue, which is then transformed to Ahp residue by cpnE orthologue (spnC: SO3561_09654) as reported in crocapeptin biosynthesis.[22] At present, the biosynthetic mechanism of Mba in streptomycetes is not clear. Tiglic acid is a defensive compound in certain beetles and biosynthesized from isoleucine via 2-methylbutyric acid.[36,37] In S. olivochromogenes, acyl-CoA dehydrogenase (SO3561_09667) and/or enoyl-CoA hydratase (SO3561_09669) may be involved in the formation of double bond in Mba (Table S1).

Cyanobacterial cyanopeptolin biosynthetic gene clusters such as mcnA–D encode three or four NRPSs to form seven modules.[18,19] Although the NRPS for streptopeptolin comprises seven modules and the domain organization is similar to those for cyanopeptolins, all the modules are present in a single polypeptide. Our recent database search suggested that many actinomycete strains such as S. olivochromogenes DSM 40451T, S. aureofaciens ATCC 10762T, S. viridifaciens DSM 40239T, Streptomyces sp. MM5, Streptomyces sp. MM100,[38]Streptomyces sp. CB03911, Streptomyces sp. CB02056, Streptomyces sp. CB01249, and Streptomyces sp. CB02460[39] harbor multiple NRPSs for streptopeptolin-like compounds and each NRPS contains seven modules without exceptions. Although cyanopeptolin and streptopeptolin show similar structures, the peptide backbones of such compounds are likely synthesized by a single NRPS in actinomycetes unlike cyanobacteria. The biosynthetic gene of crocapeptin also includes a single NRPS, so the biosynthesis of streptopeptolin seems to be similar to that of crocapeptin.[22] In cyanopeptolin biosynthetic gene clusters, specific domains such as halogenation, sulfotransferase, and glyceric acid loading domains are often present.[18] However, such domains are not observed in actinomycetes.

Conclusions

In this report, we isolated and determined the structure of a cyanopeptolin-type peptide streptopeptolin from S. olivochromogenes. To the best of our knowledge, this is the first report on the discovery of a cyanopeptolin-type peptide from actinomycetes, and we successfully identified the biosynthetic gene cluster. This study provides useful information to elucidate distribution, evolution, and biological significance of the cyanopeptolin-type peptides produced by taxonomically distant bacteria.

Methods

Bacterial Strains

S. olivochromogenes NBRC 3561 was obtained from the NBRC culture collection (Biological Resource Center, National Institute of Technology and Evaluation, Chiba, Japan).

Isolation of Streptopeptolin

S. olivochromogenes NBRC 3561 was cultured on ISP2 agar medium (2 L) at 30 °C for 5 days. Spores and aerial hyphae were harvested by a steel spatula after the cultivation. Double volume of methanol was added to the harvested cell for extraction, followed by filtration using filter paper (Whatman No. 1, GE Healthcare Life Sciences, Little Chalfont, U.K.). The filtrate was evaporated by a rotary evaporator and the concentrated extract was subjected to open-column chromatography (styrene–divinylbenzene resin, CHP-20P, Mitsubishi Chemical Corp., Tokyo, Japan) and eluted with 10% MeOH, 60% MeOH, and 100% MeOH. The 100% MeOH fraction was concentrated using a rotary evaporator, subjected to HPLC separation using an octadecyl-silica (ODS) column (4.6 × 250 mm2, Wakopak Handy ODS, Wako) and an UV detector set at 220 nm, and eluted by 22% MeCN containing 0.05% trifluoroacetic acid (TFA) at the flow rate of 1 mL/min to yield 5.0 mg of 1 (retention time of 1: 21.8 min).

NMR Experiments

NMR sample was prepared by dissolving 1 in 500 μL of mix solvent (MeCN-d3/DMSO-d6, 4:1). One-dimensional, 1H, 13C, DEPT-135, and all 2D NMR spectra were obtained on Bruker Avance 800 spectrometer with quadrature detection following the previous report.[40]

MS Experiments

ESI-MS analyses were performed using a JEOL JMS-T100LP mass spectrometer. For accurate MS analysis, polyethylene glycol was used as an internal standard.

Modified Marfey’s Method

The modified Marfey’s method was applied to 1, following previous report.[40] The reagents including N-α-(5-fluoro-2,4-dinitrophenyl)-l-leucinamide (l-FDLA, Tokyo Chemical Industry Co., LTD) and N-α-(5-fluoro-2,4-dinitrophenyl)-l-leucinamide (l-FDLA, Tokyo Chemical Industry Co., LTD, Tokyo, Japan) were used for derivatization. The standard amino acids including l-Glu, l-Thr, l-allo-Thr, l-Ala, l-Phe, and N-Me-l-Tyr were purchased from Wako Pure Chemical Co., Osaka, Japan. The HPLC analysis was performed for each derivative at a flow rate of 1 mL/min using solvent A (distilled water containing 0.05% TFA) and solvent B (MeCN containing 0.05% TFA) with a linear gradient mode from 0 to 70 min, increasing the percentage of solvent B from 25 to 60. The retention times (min) of l-FDLA- or d-FDLA-derivatized amino acids in this HPLC condition were as follows; l-Glu-d-FDLA (30.50 min), l-Glu-l-FDLA (28.46 min), l-Ala-d-FDLA (39.58 min), l-Ala-l-FDLA (32.42 min), l-Thr-d-FDLA (32.78 min), l-Thr-l-FDLA (23.72 min), l-allo-Thr-d-FDLA (28.90 min), l-allo-Thr-l-FDLA (25.73 min), l-N-Me-Tyr-d-FDLA (68.15 min), l-N-Me-Tyr-l-FDLA (65.62 min), l-Phe-d-FDLA (56.51 min), and l-Phe-l-FDLA (46.00 min).

Enzyme Inhibitory Assay

The trypsin inhibition assay was carried out at 37 °C by incubating 200 μL of trypsin solution (150 U/mL in 50 mM Tris–HCl, pH 7.6) with 120 μL of assay buffer (0.4 M Tris–HCl, pH 7.6) in the presence of 80 μL of different concentrations of 1 (prepared in DMSO) or the solvent control. After 10 min incubation, 400 μL of 1 mM substrate N-α-benzoyl-d, l-arginine-4-nitroanilide (prepared as 100 mM stock solution in DMSO and further diluted to 1 mM in the buffer 50 mM Tris–HCl, pH 7.6) was added. The reaction was monitored by measuring the absorbance at 405 nm at 0 and 30 min from the starting time of the reaction. The chymotrypsin inhibition assay was carried out at 37 °C by incubating 200 μL of trypsin solution (18 U/mL in 50 mM Tris–HCl, pH 7.6) with 120 μL of assay buffer (0.4 M Tris–HCl, pH 7.6) in the presence of 80 μL of different concentrations of 1 (prepared in DMSO) or the solvent control. After 10 min incubation, 400 μL of 1 mM substrate N-succinyl-phenylalanine-p-nitroanilide (2.5 mM in the buffer 50 mM Tris–HCl, pH 7.6) was added. The reaction was monitored by measuring the absorbance at 405 nm at 0 and 30 min from the starting time of the reaction (Scheme ).

Scheme 1

Chemical Structure of Streptopeptolin (1)