Isolation of a new antibacterial peptide achromosin from Streptomyces achromogenes subsp. achromogenes based on genome mining

Lasso peptides are a class of ribosomally biosynthesized and post-translationally modified peptides with a common motif of knot structure in the molecule.¹ The amino group of the N-terminal amino acid forms a peptide bond with side chain carboxyl group of Asp or Glu in the eighth or the ninth position from the N-terminus, resulting in formation of a macrolactam ring. The macrolactam ring looks like a loop of a ‘lasso’ with a tail of the C-terminal linear peptide that normally locates through the ring. Regarding lasso peptides, a wide variety of biological activities such as anti-HIV,² antimycobacterial,³ endothelin type B receptor antagonist⁴ and prolyl endopeptidase inhibition⁵ were reported. In addition, lasso peptides normally show a stable property against proteolytic, thermal and chemical degradation, which makes lasso peptides attractive in terms of practical application as pharmaceutical reagents.

Lasso peptides derived from actinobacteria have been classified into three main classes on the basis of their N-terminal residues and the number of disulfide bridges.¹ The class I lasso peptides include siamycins I and II,² aborycin⁶ and sviceucin,⁷ which have an internal peptide linkage between β-carboxyl group residue of Asp9 (ninth amino acid residue from the N-terminus) and the amino residue of Cys1. These peptides commonly have additional two disulfide bridges between Cys1 and Cys13, and Cys7 and Cys19. The class II lasso peptides include anantin,⁸ lariatins,³ propeptin,⁵ RES-701-1,⁴ SRO15-2005⁹ and sungsanpin.¹⁰ These peptides have an internal peptide linkage between β-carboxyl residue of Asp8 or Asp9 and the amino residue of Gly1 without any disulfide bonds. The class III lasso peptide includes only one peptide named BI-32169.¹¹ The peptide BI-32169 has an internal peptide linkage between β-carboxyl residue of Asp9 and the amino residue of Gly1 with one disulfide bond between Cys6 and Cys19.

The lasso peptide microcin J25 was isolated from Escherichia coli, which is regarded as the archetype of lasso peptides.¹² Its biosynthetic gene cluster consists of four genes including a precursor peptide-coding gene: gene A (mcjA), two maturation enzymes including gene B (mcjB, cleavage of leader peptide) and gene C (mcjC, formation of macrolactam ring) and an ATP-binding cassette transporter-coding gene: gene D (mcjD).¹³ The protein McjC was reported to form the macrolactam ring, and the function of the protein McjB was assigned to cleave off the leader peptide from the precursor peptide by in vitro experiments.¹⁴ Normally lasso peptide biosynthetic genes in proteobacteria have a corresponding set of the genes, although the transporter gene is optional.¹ In actinobacteria, lasso peptide biosynthetic genes consist of a similar gene set, except that a maturation enzyme gene B has split-B genes (gene B1 and gene B2).^{1, 15} By genome mining, biosynthetic genes of a lasso peptide sviceucin were found on the genome of Streptomyces sviceus, and the lasso peptide was isolated and structure-determined by heterologous expression.⁷ The lasso peptide SRO15-2005 was identified by matrix-assisted laser desorption/ionization-time-of-flight tandem mass spectrometry (MALDI-TOF-MS/MS) from the extract of Streptomyces roseosporus, based on genome sequence data.⁹ On the basis of genome mining, a new lasso peptide chaxapeptin was also isolated as a lung cancer invasion inhibitor from Streptomyces leeuwenhoekii.¹⁶ These results prompted us to find a new lasso peptide from streptomycetes using genome sequence data. By genome search approach, we found new lasso peptide biosynthetic genes on the genome sequence of Streptomyces achromogenes subsp. achromogenes.¹⁷ The new antibacterial peptide was isolated by chromatographic separation from the culture of S. achromogenes subsp. achromogenes. Here, we describe isolation and structure determination of a new antibacterial peptide named achromosin.

In the genome sequence of Streptomyces achromogenes subsp. achromogenes,¹⁷ lasso peptide modification enzyme-coding genes (gene C named acrC: WP_063755122.1, acrB2: WP_037654156.1, acrB1: WP_037654159.1, shown in Figure 1a and Supplementary Table S1) were found by blastp similarity search. As the lasso precursor peptide-coding gene was not annotated, we searched for the lasso precursor peptide-coding gene in the close region to the modification enzyme-coding genes. Upstream of the gene acrC (WP_063755122.1), a new putative precursor peptide-coding gene for new peptide named achromosin (126 base pairs, 42 amino acids, Figure 1b) similar to chaxapeptin¹⁶ was found from position 72 827 to 72 952 bp in the genome sequence (GenBank accession number: NZ_JODT01000002.1). On the upstream of 9 residues of the precursor peptide-coding region (72827-72952), Shine–Dalgarno sequence (AGGAGGA) was present. As shown in Figure 1b, the expected peptide achromosin was deduced to have the amino acid sequence of GIGSQTWDTIWLWD (monoisotopic molecular weight: 1676.7 Da), after cleaving off the leader peptide at the same position after the conserved motif ‘GEFXEXTX’ as the biosynthesis of chaxapeptin¹⁶ (arrow in Figure 1b). The expected monoisotopic molecular weight of achromosin was calculated to be 1658.7 Da considering the loss of 18 Da, resulting in macrolactam formation of lasso peptide biosynthesis. The preliminary chemical investigation of S. achromogenes subsp. achromogenes NBRC12735^T indicated that the expected peptide was present in the methanol extract of aerial hyphae and spore cells by high-performance liquid chromatography (HPLC) and electrospray ionization mass spectrometry (ESI-MS). Thus, cultivation of S. achromogenes subsp. achromogenes was performed in a large scale to obtain enough amount of the peptide for structure determination. After 7 days of cultivation, cells of spore and aerial hyphae were harvested by a steel spatula. The cells were extracted with double volume of methanol (MeOH), followed by centrifugation. After condensation using rotary evaporation, the extract was subjected to open-column chromatography using hydrophobic resin (CHP-20P), eluted with 10%, 60% and 100% MeOH. The expected peptide achromosin was detected in 100% MeOH fraction by HPLC (Supplementary Figure S1) and ESI-MS analysis (Supplementary Figure S2). The ESI-MS analysis of the peptide gave an ion peak at m/z 1659.7 for [M+H]⁺. The 100% MeOH fraction was repeatedly subjected to HPLC purification to give pure achromosin.

Figure 1

(a) Gene cluster for biosynthesis of achromosin including four genes (acrA: structural gene, and modification genes: acrC, acrB1 and acrB2), (b) Alignment of amino acid sequences of achromosin and chaxapeptin precursor peptide genes (underlined letters: leader peptide, bold letter: conserved amino acid, arrow: cleavage position).

The molecular formula of achromosin was established to be C₇₉H₁₀₆N₁₈O₂₂ by accurate mass analysis using the ESI Fourier-transform ion cyclotron resonance mass spectrometry, as [M+2H]²⁺ was observed at m/z 830.3941 corresponding to C₇₉H₁₀₈N₁₈O₂₂ whose calculated value was 830.3937. The amino acid composition analysis was performed on achromosin following the reported method.¹⁸ The amino acid content analysis on achromosin afforded the relative molar ratios of the constituent amino acids (2 moles each of Asp/Asn, Gly, Ile and Thr, and 1 mole each of Glu/Gln, Leu and Ser), as shown in Supplementary Figure S3. Nuclear magnetic resonance analysis using dimethyl sulfoxide-d₆ as a solvent was not possible due to ambiguous broad peaks in the nuclear magnetic resonance spectrum. To obtain peptide sequence, MALDI-TOF-MS/MS analysis on achromosin was accomplished. As a result, the product ions from achromosin at m/z 1659 were of b-series peptides, b8-b13 (Figure 2a and Supplementary Table S2), which indicated that the sequence of TIWLWD was the C-terminus tail sequence. Macrolactam ring structure was reported not to give fragment ions,⁹ thus we proposed the structure of achromosin to be shown in Figure 2a, based on the amino acid sequence of precursor peptide gene. To confirm the amino acid sequence in the macrolactam ring, C-terminal peptide bonds of tryptophans were cleaved by BNPS-skatole. After BNPS-skatole reaction, the cleaved achromosin (BNPS-achromosin) was purified by HPLC separation. ESI-TOF-MS analysis on BNPS-achromosin gave an ion peak at m/z 1291.5 for [M+H]⁺ (Supplementary Figure S4). The molecular formula of BNPS-achromosin was clarified to be C₅₈H₇₈N₁₄O₂₀ by the accurate mass analysis. That is, [M+2H]²⁺ was observed at m/z 646.2832 corresponding to C₅₈H₈₀N₁₄O₂₀ whose calculated value was 646.2831. By the reaction of BNPS-skatole the Trp residue in a peptide is oxidized and transformed to 3-oxindole with a spirolactone, which increases the molecular weight due to the addition of two oxygens by 32 Da. As shown in Figure 2b, the MALDI-TOF-MS/MS of the cleaved achromosin gave the sequence of the peptide with one N-terminus and two C-terminal ends (Supplementary Table S3). The product ions of b1, b2 and b3 supported the sequence of DTIW* and b4 ion especially indicated that Trp at C-terminus was oxidized (indicated with asterisk, Figure 2b). The product ions of y2 to y7 supported the sequence of GIGSQTW* (Figure 2b). Above all, the structure of achromosin was proposed to be a peptide with the sequence of GIGSQTWDTIWLWD having one macrolactam ring which was formed by peptide bond between amino residue of Gly1 and β-carboxyl residue of Asp8 (Figure 2a). The structure of achromosin did not include any disulfide bridge in the molecule, which classified achromosin into class II lasso peptide.

Figure 2

(a) MALDI-TOF-MS/MS analysis of achromosin, (b) MALDI-TOF-MS/MS analysis of BNPS-skatole cleaved achromosin (the oxidized Trp residue is marked by an asterisk).

The antimicrobial activity of achromosin was tested using a paper disk agar-diffusion assay against microorganisms (bacterial strains including E. coli, Pseudomonas aeruginosa, Serratia marcescens, Bacillus subtilis, Staphylococcus aureus, Micrococcus luteus and Streptomyces antibioticus; Yeast strains including Saccharomyces cerevisiae, Schizosaccharomyces pombe, Kloeckera apiculata; fungi strains including Aspergillus niger, Aspergillus oryzae and Mucor hiemalis). At the dosage of 10 μg per disk, achromosin showed an inhibitory zone of 11 mm diameter against M. luteus (Supplementary Figure S5). On the other hand, achromosin did not show any inhibitory activity against the other testing microorganisms at the same dosage.

Biosynthetic gene clusters of lasso peptides of actinobacteria have been identified for lasso peptides including lariatin,¹⁹ SRO15-2005,⁹ lassomycin,²⁰ sviceucin,⁷ chaxapeptin¹⁶ and streptomonomicin.²¹ The biosynthetic gene cluster of chaxapeptin consisted of four genes including cptA, cptC, cptB1 and cptB2.¹⁶ Interestingly, the gene cluster of chaxapeptin lacked of transporter gene that often exists in the lasso peptide biosynthetic gene cluster. The gene cptA encoded chaxapeptin precursor peptide, and the three genes including cptC, cptB1 and cptB2 were proposed to be involved in macrolactam formation and leader peptide cleavage. The amino acid sequence of precursor peptide gene acrA which was found on the genome of S. achromogenes subsp. achromogenes¹⁷ showed high similarity with that of cptA (46% identity, 68% positive matches). By reference to chaxapeptin biosynthetic genes, we assigned the biosynthetic gene cluster for achromosin, which have four genes, acrA (annoted in this study, 42 aa), acrC (WP_063755122.1, 616 aa), acrB2 (WP_037654156.1, 150 aa) and acrB1 (WP_037654159.1, 95 aa) in this order with all the same direction (Figure 1a). Interestingly, there was no transport protein-coding genes near the gene cluster. The lack of transport gene was also reported in the chaxapeptin gene cluster.¹⁶ On the basis of the similarity of each gene, we proposed the functions of the genes as shown in Figure 1a. The gene acrA encoded the precursor of achromosin and the genes including acrC, acrB1 and acrB2 were proposed to be modification enzymes to give the mature lasso peptide. The gene acrC encoded putative asparagine synthase possibly responsible for formation of the Gly1–Glu8 amide bond, which showed high similarity to cptC by using a BLAST homology search (37% identity, 51% positive matches). The amino acid sequence of acrB2 showed high similarity to that of cptB2 by using a BLAST homology search (55% identity, 69% positive matches) and the amino acid sequence of acrB1 showed high similarity to that of cptB1 by using a BLAST homology search (40% identity, 54% positive matches). Above all, the biosynthetic genes of achromosin showed the similarity to those of chaxapeptin.

So far, no similar peptide has been found by the blastp search, which indicates the novelty of achromosin. As shown in Figure 1b, the amino acid sequence of core peptide is different even from that of chaxapeptin, the closest lasso peptide. The lasso peptide in class II were reported to have a wide variety of biological activities such as antimycobacterial,³ endothelin type B receptor antagonist⁴ and prolyl endopeptidase inhibition⁵. In this paper, the antimicrobial activity of achromosin was tested. Further bioactivity tests may lead to the discovery of additional activities of achromosin. In addition, the biosynthetic genes of achromosin were identified from the genome of S. achromogenes subsp. achromogenes, which will lead to genetic engineering using the gene cluster to create mutated lasso peptide based on achromosin by heterologous expression. The modified peptides with more potent antibacterial activity may be produced by altering the amino acid sequence of achromosin by further genetic engineering experiments.