Structure and activity of a class II lanthipeptide from a thermophilic bacterium

Abstract

Lanthipeptides represent the largest group of ribosomally synthesized and post-translationally modified peptides (RiPPs). Lanthipeptides offer promising avenues for discovering new antibacterial and antifungal agents. Here, we identify and structurally analyze the product of the tla BGC, which encodes a class II lanthipeptide in the thermophilic bacterium Thermoactinomyces sp. DSM 45891. Heterologous co-expression of the lanthipeptide synthetase TlaM resulted in modification of the two precursor peptides TlaA1 and TlaA2, which share 58% identity. TlaA1 was dehydrated seven times and TlaA2 six times. In both peptides, four thioether rings were formed with two overlapping DL-(methyl)lanthionine rings at the C-terminus. Both peptides also contain two central and N-terminal non-overlapping DL-methyllanthionines. These findings demonstrate that these peptides deviate from the general rule of stereoselective LL-(methyl)lanthionine formation from a DhxDhxXxxXxxCys motif (Dhx = dehydroalanine or dehydrobutyrine). AspN-cleaved TlaM-modified TlaA1 displayed anti-microbial activity against a subset of bacteria including Gram-negative ESKAPE pathogens. We named the lantibiotic thermolanthin.

1. Introduction

Lanthipeptides represent a large group of ribosomally synthesized and post-translationally modified peptides (RiPPs) characterized by intramolecular thioether crosslinks called lanthionine and methyllanthionine.^[1] Their wide-ranging functions, including antimicrobial and antifungal activity, virulence functions, and signalling roles,^[2] and their promise for bioengineering,^[3] have motivated researchers to discover new members by genome mining exercises.^[4] Lanthipeptides with antibacterial activity have been termed lantibiotics.^[5] Lanthipeptide biosynthetic enzymes play a key role by introducing specific ring patterns that impart these activities.^[6] A large-scale bioinformatic genome mining effort focused on lanthipeptide biosynthetic gene clusters (BGCs) revealed high sequence diversity amongst precursor peptides across the lanthipeptide classes I-IV.^[7] A subset of the lanthipeptide BGCs detected in the bacterial genomes encode two precursor peptide genes within a single BGC.^[7] For two-component lanthipeptides that act synergistically to achieve antimicrobial activity,^[8] these two peptides typically have considerable sequence diversity and result in different ring patterns. However, in other cases, these precursor peptides are highly similar in sequence and result in a group of congeners.^[7] Notably, the individual peptides of two-component lantibiotics typically have little or no activity in the absence of their partner peptides. For instance, the two-component lantibiotics lacticin 3147 (Ltnα/Ltnβ), lichenicidin (Licα/Licβ), haloduracin (Halα/Halβ), cytolysin (CylL_L/CylL_S), plantaricin W (Plwα/Plwβ), and roseocin (Rosα/Rosβ) require both distinct peptides for physiological activity.^[9] The individual peptides of such two-component systems can be produced either by a single enzyme, capable of modifying both peptides,^[10] or by two distinct enzymes, each modifying one of the two precursor peptides.^{[9e, 9h, 11]}

Here we describe and characterize TlaM, a class II lanthipeptide synthetase from the thermophile Thermoactinomyces sp. DSM 45891 that modifies both TlaA1 and TlaA2 precursor peptides encoded in the tla BGC. The two precursor peptides share 57.9% sequence identity. TlaM is a member of the class II lanthipeptide synthetases that use an N-terminal domain to phosphorylate Ser and Thr residues and subsequently to eliminate the phosphate to generate dehydroalanine (Dha) and dehydrobutyrine (Dhb) residues, respectively.^[12] The C-terminal domain contains a Zn²⁺ in its active site that is important for catalyzing Michael-type additions of select Cys residues to the Dha and Dhb residues to form thioether crosslinks called lanthionine (Lan) and methyllanthionine (MeLan), respectively.^[13] These post-translational modifications (PTMs) take place in the C-terminal segment of the precursor peptide called the core peptide. An N-terminal leader peptide is removed after completion of the PTMs to release the final mature lanthipeptide.^[14]

Thermophilic bacteria represent a promising source of robust, adaptable biosynthetic machinery. Enzymes from thermophiles often exhibit elevated stability, enhanced tolerance to structural variation, and favorable biochemical resilience given their ability to function at elevated temperatures.^[15] Only a few examples of lanthipeptides have been reported that are produced by thermophilic bacteria.^[16] One such example, geobacillin I, is a class I lanthipeptide produced by Geobacillus thermodenitrificans NG80-2, which contains seven (methyl)lanthionines. Geobacillin I exhibits antimicrobial activity comparable to that of the commercial food preservative nisin A, but with greater thermal and pH stability.^[17] Geobacillin II, a class II lanthipeptide from the same organism that undergoes modification by the bifunctional enzyme GeoM, has a different ring topology than geobacillin I.^[17] In this study, we used Escherichia coli to heterologously produce two lanthipeptides from the tla BGC. We determined the ring pattern of modified TlaA1 and TlaA2 by nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry, and assessed their biological activity. Modified TlaA1 and TlaA2 (mTlaA1 and mTlaA2) have no structural homologs among known lanthipeptides.

2. Result and Discussion

2.1. Bioinformatic identification of the tla biosynthetic gene cluster

The tla BGC was chosen from a previously reported sequence similarity network of class II lanthipeptides with the expectation that it could be the first characterized two-component lanthipeptide from a thermophilic organism.^[7] Using the RODEO webtool 2.0,^[18] two precursor peptides, encoded by tlaA1 and tlaA2, were predicted to consist of 69 and 74 amino acid residues, respectively (Figure 1). No additional modifying enzymes were encoded near these peptides with the exception of a canonical class II lanthionine synthase, encoded by tlaM and a peptidase-containing ATP-binding cassette transporter (PCAT)^[19] ThtT predicted to remove the leader peptide from the modified precursor at a double Gly motif using its N-terminal C39 peptidase domain.^[20]

Figure 1.

The tla biosynthetic gene cluster from Thermoactinomyces sp. DSM 45891 and other orthologous BGCs (accession numbers in Table S1). A) Gene composition of the tla, bai, and cro BGCs. The genes encoding the precursor peptides are in black, LanMs in red, and PCATs in purple. B) Precursor peptide sequences from each BGC; Ser/Thr, and Cys are shown in blue and red, respectively, which were expected to be involved in post-translational modifications. Double glycine motifs are shown in purple. The amino acid residues after the double Gly cleavage site and before the first Cys or dehydrated amino acid are underlined and may be involved in a second proteolytic cleavage event.

TlaA1 and TlaA2 both contain eight Ser/Thr amino acid residues and four Cys residues in the predicted core peptides (Figure 1B). The constellation of these residues requires a product distinct from any previously characterized lanthipeptide. A BLAST search retrieved very few peptides with sequence homology to TlaA1 and TlaA2. Two putative precursor peptides BaiA and CroA from Baia soyae and Croceifilum oryzae, respectively, display sequence homology towards their C-termini but not at their N-termini (Figure 1B). Their BGCs differ from the tla BGC, as they encode additional genes, including an asparagine synthase and a clostripain C11 peptidase. The former may form lactam,^[21] amide^[22] or nitrile^[23] structures based on previous roles in RiPP biosynthesis, and the latter may be responsible for a second proteolytic cleavage event after PCAT processing to produce the mature product of the bai and cro cluster.^[14] Such second proteolytic cleavage events are common for two-component lantibiotics and typically involve the removal of six or seven amino acids. The TlaA1/A2, CroA, and BaiA peptides all contain a heptapeptide between the predicted double Gly cleavage site for a PCAT enzyme and the first Ser/Thr/Cys residue that is likely modified during maturation (Figure 1B). The presence of a second protease in the cro and bai BGCs provides further support for such a process. A sequence alignment of the lanthionine synthetases in the homologous BGCs showed relatively similar sequences. TlaM and BaiM share 62.4% identity, whereas TlaM and CroM have 62.6% identity (Figure S1). Hence, these lanthipeptide synthetases are hypothesized to produce similar ring patterns arising from their homologous C-terminal substrate sequences.

2.2. Heterologous production of modified TlaA1 and TlaA2 in E. coli

E. coli was used as heterologous host for co-expression of tlaA1 and tlaA2 with tlaM. The genes encoding TlaA1 and TlaA2 were separately inserted into pET-Duet vectors. The precursor peptides were fused at their N-termini to a histidine tag for purification and a small ubiquitin-like modifier (SUMO) tag to improve yield to produce His₆-SUMO-TlaA1 and His₆-SUMO-TlaA2. The lanthipeptide synthetase gene tlaM was cloned into multiple cloning site (MCS) 1 of a pRSF-Duet-1 vector, resulting in an N-terminal hexa-histidine-tag. After heterologous expression of His₆-SUMO-TlaA1 and purification by immobilized metal affinity chromatography (IMAC), both the His₆-SUMO and the putative leader peptide were removed in vitro with the purified C39 protease LahT150,^[24] an ortholog of the protease domain of TlaT. The expected peptide corresponding to the core peptide sequence was observed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) (Figure 2A). Coexpression of TlaM and SUMO-TlaA1 resulted in five, six, and seven dehydrations as determined by MALDI-TOF MS (Figure 2A). Extracted ion chromatograms (EICs) of mTlaA1 revealed peak area ratios of 1:10:10 for the 7-, 6-, and 5-fold dehydrated products, respectively (Figure S2A). Similarly, we co-expressed TlaM and His₆-SUMO-TlaA2, followed by cleavage with LahT150, and observed seven dehydrations as the major product (Figure 2B). EIC analysis confirmed that a peptide having undergone seven dehydrations was the major product formed (Figure S2B).

Figure 2.

MALDI-TOF mass spectra of the precursor peptides in the tla BGC with and without expression with TlaM, followed by in vitro cleavage with LahT150 to remove the leader peptide (LP). A) TlaA1, observed [M+H]⁺ m/z 3891.2; calculated 3894.7, the difference between calculated and observed values is likely due to partial formation of disulfide bonds from the four cysteine residues. After heterologous expression with TlaM, a mixture of dehydrations was observed. mTlaA1, observed [M+H–5 H₂O]⁺ m/z 3804.9; calculated 3804.7; [M+H–6 H₂O]⁺ m/z 3787.9; calculated 3786.7; [M+H–7 H₂O]⁺ m/z 3770.9; calculated 3668.7. B) MALDI-TOF MS of TlaA2, observed [M+H]⁺ m/z 4196.9; calculated 4202.0. After heterologous expression with TlaM, mTlaA2, observed [M+H–7 H₂O]⁺ m/z 4075.1; calculated 4075.9.

2.3. NEM and DTT assays

To determine whether all four Cys residues in TlaA1 and TlaA2 were cyclized and to assess whether any dehydro amino acids were formed after TlaM modification, the LahT150-cleaved TlaA products were subjected to chemical derivatization using N-ethylmaleimide (NEM) and dithiothreitol (DTT).^[1a] For mTlaA1 treatment with NEM did not result in the formation of an adduct, indicating the peptide was fully cyclized with all four Cys residues forming thioethers with four dehydroamino acids. The modified product was then subjected to a DTT assay. MALDI-TOF MS analysis demonstrated the presence of a mixture of products with one and two DTT adducts (Figure 3A). These observations were consistent with the five and six dehydrations and four cyclizations determined from the NEM assay. To determine the ring pattern and presence of Dha/Dhb residues, the modified TlaA1 was subjected to proteolytic cleavage. Modified SUMO-TlaA1 was treated with chymotrypsin, and two fragments were observed that encompass the core peptide (Figure 3B). Two and three dehydrations were observed in fragment 1, whereas fragment 2 was dehydrated three times. Along with the DTT and NEM assays described above, fragment 1 contains one ring that we term ring A, and fragment 2 contains three rings that we term rings B, C, and D. To determine the location of dehydration, mTlaA1 was also digested with endoproteinase GluC, followed by LysC resulting in two fragments A and B (Figure S3). Tandem MS analysis of fragment A demonstrated dehydration of Thr4, Thr8, Ser9, and Thr18, with a minor amount of dehydration of Ser3 (Figure S4). The C-terminal fragment B underwent two dehydrations, but their locations could not be determined because of a lack of fragmentation (Figure S5), which usually is the result of overlapping rings.^[25] The two dehydrations in fragment B would originate from Ser27, Ser28, or Thr31.

Figure 3. Mass spectrometric analysis of mTlaA1 and mTlaA2 after undergoing chemical modification and proteolytic cleavage.

A) MALDI-TOF mass spectra of NEM and DTT assays of mTlaA1. mTlaA1, observed [M+H–5 H₂O]⁺ m/z 3805.7; calculated 3804.7; [M+H-6H₂O]⁺ m/z 3787.7; calculated 3786.7; [M+H–7 H₂O]⁺ m/z 3770.9; calculated 3668.7. No NEM adduct was observed; however, one and two DTT adducts (+154 Da) were observed, [M+H–5 H₂O+1 DTT]⁺ m/z 3959.5; calculated 3958.7; [M+H–6 H₂O+1 DTT]⁺ m/z 3941.5; calculated 3940.7. A mass change of +308 Da for two DTT adducts, observed [M+H–6 H₂O+2 DTT]⁺ m/z 4095.5; calculated 4095.7. B) Proteolytic cleavage of mTlaA1 using chymotrypsin resulted in two fragments, 1 and 2. ESI-MS (positive mode) of fragment 1 showed a prominent ion at m/z [M+H-2 H₂O]³⁺ 1227.2370, calculated m/z 1226.9897; m/z [M+H–3 H₂O]³⁺ 1220.9037, calculated m/z 1220.9846. Fragment 2 showed m/z [M+H–3 H₂O]⁴⁺ 576.0306, calculated m/z 576.1945.C) MALDI-TOF mass spectra of NEM and DTT assays of mTlaA2. mTlaA2, observed [M+H–7 H₂O]⁺ m/z 4077.0; calculated 4075.9. No NEM adduct was observed; however, two and three DTT adducts were observed. Mass changes of +308 Da and +462 Da were observed, respectively, [M+H–7 H₂O+2 DTT]⁺ m/z 4385.3; calculated 4383.9; [M+H–7 H₂O+3 DTT]⁺ m/z 4538.2; calculated 4537.9. D) Proteolytic cleavage of mTlaA2 using LysC resulted in three fragments, 3, 4, and 5. ESI-MS (positive mode) of fragment 3 showed a m/z [M+H–4 H₂O]⁺ 1765.8692, calculated m/z 1765.8612; 4 m/z [M+H–H₂O]⁺ 1318.6101, calculated m/z 1318.6130. Fragment 5 showed m/z [M+H–2 H₂O]⁺ 1029.4960, calculated m/z 1029.4968. Lowercase letters represent the identified dehydroamino acids.

mTlaA2 cleaved with LahT150 was also subjected to both NEM and DTT assays followed by analysis by MALDI-TOF MS. Adducts were not observed in the NEM assay, indicating that all four cysteines were cyclized. Two and three DTT adducts were observed, suggesting that three dehydroamino acid residues are not involved in cyclization (Figure 3C). One of the dehydro amino acids is likely a Dhb, which reacts sluggishly with DTT, especially when located within a (methyl)lanthionine ring, thus explaining incomplete labelling by DTT.

Proteolytic cleavage with AspN and subsequent tandem MS analysis demonstrated that Thr8 and Thr9 were dehydrated in mTlaA2 and are not involved in cyclization (Figure S6). Additionally, Thr12 and Thr13 are dehydrated and likely part of a MeLan given the lack of fragmentation between former Thr12 and Cys 16 (Figure S6). Furthermore, Thr22 is likely dehydrated and part of a ring with Cys27. As observed for fragment B of mThrA1, the lack of fragmentation in the C-terminal segment of mThrA2 precluded definitive assignment of the final two dehydrations, which occur at Ser31, Ser21, or Thr35 (Figure S6). mTlaA2 was also digested with endoproteinase LysC, resulting in fragments 3, 4, and 5 (Figure 3D). Fragment 3 had been dehydrated four times and contains one ring and three dehydro amino acids. Fragment 4 had been dehydrated once, and product 5 twice (Figure 3D), suggesting that of the three Ser and Thr residues in this peptide (Ser27, Ser28, or Thr31; numbering based on core peptide, Figure 2), two are dehydrated.

2.4. Determination of ring pattern by NMR spectroscopy

Collectively, the proteolytic digests and tandem MS analysis established the presence of a MeLan ring between former Thr18 and Cys23 in mTlaA1 and between former Thr22 and Cys27 in mTlaA2 (rings B). Since no fragmentation was observed in the C-terminal region of both mTlaA1 and mTlaA2, overlapping ring patterns are likely formed from the sequence SSXATPCKRC in which two of the three Ser/Thr residues were dehydrated (Figure S5 and Figure S6). Because tandem MS analysis cannot deduce the ring pattern of these overlapping rings,^[25] fragment 5 was subjected to NMR analysis (Figure 4A). The ring pattern of the 10-amino acid peptide 5 was elucidated using 1D and 2D NMR experiments, including ¹H, ¹H-¹H TOCSY, ¹H-¹H NOESY, and ¹H-¹³C HSQC. The ¹H and ¹³C chemical shift assignments of the peptide in 90% H₂O and 10% D₂O at 25 °C are summarized in Table S2. A total of 7 amide protons were observed (Figure S7). The missing NH protons correspond to the N-terminal residue Ser1 (residue numbering based on the LysC product 5, Figure 4A), the residue at position 2 (formerly Ser2), and a proline at position 6. Both the ¹H and ¹³C chemical shift values of the former Ser2 residue and former Thr5 showed significant upfield shifts compared to typical Ser and Thr residues, suggesting their involvement in lanthionine and methyllanthionine formation (Figure S8). These changes include β-proton signals at 3.21 ppm and 3.04 ppm, and a β-carbon at 33.8 ppm for the former Ser2. The β-proton and β-carbon of Thr5 shifted to 3.64 ppm and 40.8 ppm, respectively, compared to the typical values of approximately 4.10 ppm (¹H) and 69.0 ppm (¹³C) observed when bonded to a more electronegative oxygen atom. Further analysis of the NOESY data suggested that a lanthionine was formed between Ala2 and Cys7, and a methyllanthionine was formed between Thr5 and Cys10. This assignment is supported by a set of NOE cross-peaks observed between these residues across the thioether bond. For example, a strong NOE was detected between the β-protons of former Cys7 and Ala2 (former Ser2) (Figure 4B). Similarly, NOEs were observed between the amide proton of former Cys10 and the β-proton of former Thr5, and between the amide proton of former Thr5 and the β-proton of former Cys10 (Figure 4C).

Figure 4. NMR analysis of fragment 5.

A) Sequence of fragment 5 and residue numbering used in the description of the NMR data. B) ¹H-¹H NOESY spectrum of fragment 5. Cross-peaks across the lanthionine bond are observed between the β-protons of Ala2 and former Cys7. C) ¹H-¹H NOESY spectrum of fragment 5. Cross-peaks across the methyllanthionine bond are observed between the amide proton of former Thr5 and the β-proton of former Cys10, and vice versa between former Cys10 and Thr5. D) Fragment 2 showing the ring patterns of rings B, C, and D. Fragment 5 is boxed.

As noted above, the amide proton of Ala2 (former Ser2) was absent in the TOCSY spectrum, and no NOE cross-peaks were observed between Leu3 and Ala2, likely due to line broadening. These missing peaks prevented an unambiguous distinction between ring formation between former Cys7 and either the first or second residue in fragment 5. To resolve this ambiguity, the longer 22-amino acid peptide, fragment 2 derived from mTlaA1, was used for further analysis (Table S3). In this peptide, the two serine residues in question are positioned at positions 13 and 14 and were clearly assigned (Figure S9). A distinct NOE cross-peak was observed between the amide protons of residue 14 and Leu15 (Figure S10, corresponding to Ala2 and Leu3 in fragment 5), as well as between the amide proton of residue 14 and the β-proton of Ser13 (Figure S11, corresponding to Ala2 and Ser1 in fragment 5). Therefore, the former Ser2 was assigned as the dehydrated serine in fragment 5. Collectively, the proteolytic cleavage fragments combined with NMR analysis confirmed the ring patterns in mTlaA1 and mTlaA2 for rings B, C, and D (Figure 4D).

2.5. Stereochemistry of the (methyl)lanthionines in mTlaA1 and mTlaA2

The stereochemical configurations of the rings in mTlaA1 and mTlaA2 were determined by Marfey analysis.^[26] mTlaA1 and mTlaA2 were hydrolyzed in acid followed by reaction with the advanced Marfey's reagent, Nα-(5-fluoro-2,4-dinitrophenyl)-L-leucinamide (L-FDLA).^[27] The resulting derivatized products were compared with authentic standards using LC-MS. The standard used for lanthionine analysis was CylLL, which contains DL-Lan, LL-Lan, and LL-MeLan.^[10a] Based on the retention times and co-injection with the standards, mTlaA1contains DL-lanthionine (Figure 5A). The standards for methyllanthionine were obtained from mCoiA1, which contains D-allo-L-MeLan, LL-MeLan, and DL-MeLan.^[28] The L-FDLA-derivatized MeLan from hydrolyzed mTlaA1 coeluted with DL-MeLan, which confirmed that the methyllanthionine rings all have DL-stereochemistry (Figure 5B). This observation was unexpected because the N-terminal sequence has a common motif (Dhx-Dhx-Xxx-Xxx-Cys; Dhx = Dha or Dhb) that resulted in LL-(Me)Lan in other lanthipeptides.^{[10a, 29]} Similar analysis of mTlaA2 indicated that it also contained DL-Lan and DL-MeLan (Figure 5C,D).

Figure 5.

Hydrolyzed peptides were derivatized with Marfey’s reagent and analyzed by LC–MS to determine the stereochemistry of thioether crosslinks. Extracted ion chromatograms (EICs) of L-FDLA derivatized mTlaA1, mTlaA2, MeLan standard (from mCoiA1), and Lan standard (from CylLl). A, B. EICs corresponding to m/z 795.2373 and 809.2530 are shown for Marfey-derivatized lanthionine (Lan) and methyllanthionine (MeLan), respectively, of modified TlaA1. C, D. EICs for modified TlaA2. (i) MeLan or Lan standard (ii) tested compound (iii) coinjection (tested compound + standard). E. The proposed structure of mTlaA1; Abu, 2-aminobutyric acid. F. The proposed structure of mTlaA2. Highlighted amino acid residues in red and blue are where PTMs occurred. The residues in pink and orange would be removed if a protease removes the N-terminal 6 or 7 residues as discussed in the text.

These data did not distinguish between a Lan or a MeLan as the A ring in mTlaA1 (involving Cys12 and either Thr8 or Ser9), nor whether the A-ring in mTlaA2 is formed between Cys16 and Thr12 or Thr13. To determine the residues involved in forming the N-terminal ring A, we also subjected fragment 1 derived from mTlaA1 to acidic hydrolysis and Marfey’s analysis. We confirmed that the MeLan A-ring in mTlaA1 has a DL configuration, and not an LL-configuration (Figure S12). But surprisingly, we also observed DL-Lan, albeit with a signal intensity that is about 5-fold smaller than that observed for MeLan. These findings suggest that in mTlaA1, both a MeLan is formed between Cys12 and Thr8 and a Lan is formed from Cys12 and Ser9. Lanthipeptide synthases usually do not form alternative ring patterns, and when other patterns are observed, they usually arise from non-enzymatic ring formation that competes with enzymatic ring formation. Because non-enzymatic MeLan formation is much slower than non-enzymatic Lan formation,^{[6, 30]} and because the stereochemistry of MeLan formation from the DhbDhaLeuValCys motif is opposite of that expected from non-enzymatic cyclization (and therefore is likely to require enforcement by the enzyme),^[31] we hypothesize that the small amount of Lan formed from Cys12 and Ser8 is non-enzymatic and non-physiological. Collectively, the data in this study suggest the structures shown in Figure 5E and F for the lanthipeptides from Thermoactinomyces sp. DSM 45891.

2.6. Antimicrobial activity

We next evaluated the antimicrobial activity of mTlaA1 and mTlaA2 cleaved with LahT150 and dissolved in 5% DMSO. The proteolytic product derived from mTlaA1 showed very weak antimicrobial activity against Bacillus subtilis and E. coli (Figure S13), whereas the product from mTlaA2 showed no detectable antimicrobial activity. When both modified lanthipeptides were combined to test potential synergistic activity, no antimicrobial activity was observed. In other two-component lantibiotics such as haloduracin,^[32] cytolysin,^[33] and lichenicidin,^[9f] a second proteolytic step is involved after removal of the leader peptide by the PCAT. In all three cases, the N-terminal sequence that is removed is a hexapeptide. For geobacillin II, a heptapeptide has been suggested to be removed in such a second proteolytic step.^[17] In some cases, the required second protease is encoded within the BGC, but in many other cases the second protease is encoded elsewhere.^{[14, 34]} Because the tla BGC does not encode a second protease, we used AspN endopeptidase cleavage of mTlaA1 to access a 31-mer 6 with 5 dehydrations and including rings A-D (Figure 6A). Peptide 6 demonstrated antibacterial activity against B. subtilis as well as the gram-negative ESKAPE pathogens Acinetobacter baumanii and Klebsiella pneumoniae (Figure 6B-D). Growth inhibition of gram-negative bacteria by lanthipeptides is a rare observation.^[35] These findings suggest that a second proteolytic step is likely required for bioactivity, although the precise site is currently not known and peptide 6 is unlikely to be the native peptide. Furthermore, the observed bioactivity with peptide 6 shows that the two additional N-terminal dehydrations in mTlaA1 are not required for bioactivity. We termed the active peptide 6, thermolanthin A, and anticipate that the natural peptide will likely have higher activity. Attempts to perform a second proteolytic step on mTlaA2 were unsuccessful. Treatment with endoproteinase GluC not only removed the N-terminal six amino acids but also resulted in cleavage after Glu20.

Figure 6.

ESI-MS characterization and bioactivity assessment of AspN-digested mTlaA1. A) ESI-MS spectrum showing the molecular mass of AspN-digested mTlaA1. Observed m/z [M+H-5H₂O]⁺ 3331.5065, calculated 3330.5097. B) Bioactivity assay against B. subtilis 2470 and two gram-negative ESKAPE pathogens. For each assay, 2 μL of AspN-cleaved mTlaA1 (~500 μM) was spotted.

3. Conclusion

This work identified two new lanthipeptides from the thermophile Thermoactinomyces sp. DSM 45891 with 5, 6, and 7 dehydrations. The thioether patterns of these peptides shown in Figure 6E and 6F have not been previously reported for known lanthipeptides. TlaM converted TlaA1 and TlaA2 to similar structures, with TlaA2 undergoing one more dehydration in the N-terminus at Thr8. The two peptides are very similar in structure and AspN-processed mTlaA1 displayed activity without the need for a partner peptide. We therefore do not have any evidence that the tla BGC generates a two-component lanthipeptide. Based on the current data, it is more likely that the two peptides are congeners. In the absence of information of the products formed by the producing organism, the precise proteolytic processing site is unresolved. It is likely that the products of the bai and cro BGCs will have a similar ring pattern to thermolanthin A with respect to rings B-D, which could be the bioactivity-inducing part of these molecules.

This study showed that TlaM catalyzes the stereoselective formation of both DL-MeLan and DL-Lan rings during post-translational modification of its precursor peptides. Based on previously observed stereochemical selectivity with Dhx-Dhx-Xxx-Xxx-Cys motifs in which LL-Lan and MeLan structures are usually generated as a result of an inherent preference imparted by the motif,^{[10a, 29a, 29b, 31]} the formation of a DL-MeLan A-ring is unique and suggests TlaM enforces the formation of a DL-MeLan. The mechanism by which it achieves this selectivity requires further investigation.

4. Experimental Section

4.1. General Experimental Procedures and Materials

Molecular cloning was performed using the primers listed in Supporting Information Table S4 to produce three different plasmids: His6-SUMO-TlaA1-pET15, His6-SUMO-TlaA2-pET15, and His6-TlaM-pRSFduet. Co-expression of His₆-SUMO-TlaA1 and His₆-TlaM-pRSFduet was carried out in E. coli BL21(DE3). Similarly, co-expression of His6-SUMO-TlaA2 and His6-TlaM-pRSFduet was done in the same expression strain. Purification by NiNTA affinity chromatography was performed after expression, followed by proteolytic cleavage with LahT150, and MALDI-TOF-MS analysis using a Bruker Ultraflex II. Cleaved peptide was desalted using a C18 ziptip and eluted in a saturated solution of super-DHB in 80% MeCN, which served as MALDI matrix. The cleaved peptide was stored at −20 °C pending further purification.

4.2. Co-expression using His6-SUMO-TlaA-pet15 and His6-TlaM-pRSFduet

For each TlaA/TlaM co-expression, 4 L flasks containing 1 L of LB medium with 50 μg/mL kanamycin and 100 μg/mL ampicillin were inoculated with 10 mL of an LB overnight culture of E. coli BL21(DE3) carrying the respective expression plasmids. Cells were grown at 37 °C until the cultures reached an optical density at 600 nm (OD₆₀₀) of ~0.75-0.95. Then, the culture flasks were placed on ice for 30 min, and the shaker temperature was lowered to 18 °C. Next, 0.5 mL of sterile stock solution of isopropyl β-D-1-thiogalactopyranoside (IPTG, 1 M) was added to each flask. After overnight incubation at 18 °C, cultures were harvested by centrifugation (6000×g, 4 °C, 10 min).

The cell pellets were resuspended in lysis buffer (20 mM Na₂HPO₄, 10 mM imidazole, 300 mM NaCl and 6 M guanidine hydrochloride, pH 7.5) and lysed using a sonicator (on 4 s and off 3 s) at 50 % amplitude. Afterwards, the lysate was cleared by centrifugation (10 000×g, 4 °C, 30 min) and filtered using a syringe filter (0.45 μm). Peptide isolation was performed by NiNTA affinity chromatography using pre-equilibrated 0.5 mL/ L of expression of NiNTA resin in a gravity-flow column. After applying the cell lysate, the column was washed with 3 mL of lysis buffer, followed by 3 mL of wash buffer 1 (20 mM Na₂HPO₄, 10 mM imidazole, 300 mM NaCl and 4 M guanidine hydrochloride, pH 7.5) and wash buffer 2 twice (20 mM Na₂HPO₄, 50 mM imidazole, 300 mM NaCl, pH 7.5). Next, the peptide was eluted from the resin by adding 3 mL of elution buffer (20 mM Na₂HPO₄, 1 M imidazole, 300 mM NaCl, pH 7.5).

4.3. Purification of His6-LahT150-pet15

Similar protein expressions using His6-LahT150-pet15 were carried out as described for the peptide co-expressions above. The cells were harvested by centrifugation (5 000×g, 4 °C, 15 min), and cell pellets were resuspended in protein Start buffer (20 mM Tris-HCl, 10 mM imidazole, 300 mM NaCl and 10% glycerol, pH 8.0) containing 4 mg/mL of lysozyme. This mixture was incubated at 4 °C, then further lysed via sonication (40% amplitude, 4.4 s on, 5.5 off) for 6 min. The lysate was centrifuged at 10,000 x g for 30 min. The supernatant was loaded onto an immobilized metal affinity chromatography (IMAC) column equilibrated with protein Start buffer, using 0.5 mL of Ni resin per gram of cell pellet. This step was followed by adding 2–3 column volumes (CV) of protein Start buffer to the column. The column was then washed with 2-3 CV of protein Wash buffer (20 mM Tris-HCl, 30 mM imidazole, 300 mM NaCl and 10% glycerol, pH 8.0) before eluting the protein with 2-3 CV of protein Elution buffer (20 mM Tris-HCl, 300 mM imidazole, 300 mM NaCl and 10% glycerol, pH 8.0). Each fraction from the wash and elution was collected for analysis by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) using a 4–20% Tris gel at 150 V. The elution fraction containing LahT150 was collected, and the buffer was exchanged to protein storage buffer (20 mM Tris-HCl, 500 mM NaCl, and 10% glycerol, pH 8.0) with a 3 kDa MWCO Amicon filter. Subsequently, the fraction was aliquoted and stored at −80 °C.

4.4. Protease Digestion

Proteolytic cleavages were performed using the following proteases: LahT150, chymotrypsin, LysC, and AspN endopeptidase under standard digestion conditions. Purified peptides (1-3 mg/mL) were dissolved in their appropriate buffer compatible with each protease (pH 7.8–8.0). Proteases were added at an enzyme-to-substrate ratio of 1:100. Digestions were carried out at 37 °C for all proteases except LahT150, which was carried out at RT for 16-18 h. Reactions were quenched by acidification with formic acid to a final concentration of 0.1–1% (v/v). The digested samples were analyzed directly by MALDI-TOF-MS to verify proteolytic cleavage. Next, the digested products were further purified by HPLC for further analysis.

4.5. HPLC purification of TlaA peptides

The peptides were further purified by preparative HPLC using an Agilent 1260 HPLC system. For separation, a Luna 10 μm C5 100 Å, LC Column 250 x 21.2 mm was used, employing a solvent system consisting of H₂O/0.1% trifluoroacetic acid (solvent A) and MeCN/0.1% trifluoroacetic acid (solvent B) and a linear gradient of 2-100% B in 45 min. Peptide fractions were identified by MALDI-TOF-MS, pooled, and the solvent was removed using a lyophilizer.

4.6. NEM Assay of mTlaA1 and mTlaA2

A 25-μL reaction was prepared with 25-50 μM peptide dissolved in 50 mM Tris-HCl, pH 7.5 (12.5 μL), along with 0.3 mM of fresh 1.0 mM TCEP stock solution (10 μL). The mixture was incubated at room temperature for 30 min before adding 2.5 μL of 500 mM NEM. The reaction was then incubated at room temperature for 3 h, desalted using a C18 ZipTip, and analyzed by MALDI-TOF MS. The formation of an adduct is indicated by a mass shift of +125 Da.

4.7. DTT Assay with mTLaA1 and mTlaA2

Modified TlaA1 and TlaA2 were dissolved in 50 μL of 100 mM Tris-HCl buffer, pH 7.5, to obtain a final concentration of 100 μM. N, N-diisopropylethylamine (DIPEA) and DTT were added to give final concentrations of 400 mM and 500 mM, respectively. The 50 μL reaction was incubated at room temperature for 24 h. After desalting with a C18 ziptip, the reaction was analyzed by MALDI-TOF MS.

4.8. Tandem Mass Spectrometry

The HPLC-purified TlaA cleaved precursor peptides were injected onto the Agilent 6545B Quadrupole Time-of-Flight mass spectrometer. Samples were run on a Phenomenex Cat. No. 00F-4497-AN Kinetex C800 column (2.6 μm, 100 Å, 150 x 2.1 mm) using a solvent system consisting of H₂O/0.1% formic acid (solvent A) and MeCN/0.1% formic acid (solvent B) and a linear gradient of 5-100% B over 20 min at a flow rate of 0.4 mL/min. The parameters used for MS(/MS) data acquisition included a resolution of 30,000 in positive mode, an isolation width of 2-3 Da, and a normalized collision energy of 20, 25, and 30 (MS/MS). Data analysis was performed using MassHunter Qualitative Analysis.

4.9. Marfey’s Analysis of mTlaA1 and mTlaA2

For the Marfey analysis,^[26] mCylL was used for the lanthionine standard and mCoiA1 for the methyllanthionine standard. The standards, along with modified TlaAs, were purified after heterologous expression in E. coli as described previously, followed by acidic hydrolysis, and the Lan and MeLan were derivatized. Dried peptides (100 μg) were dissolved in 0.8 mL of 6 M DCl in D₂O for hydrolysis using a long screw glass vial. N₂ gas was bubbled into the glass vials for 2 min. The reaction was heated to 120 °C and stirred for 20 h. Each reaction product was then dried using a rotary evaporator and placed in the lyophilizer for 45 min to remove any remaining DCl. After drying the samples, 600 μL of 0.8 M NaHCO₃ (H₂O) and 400 μL of 10 mg/mL of Nα-(5-fluoro-2,4-dinitrophenyl)-l-leucinamide (L-FDLA) in acetonitrile were added to the glass vials. The mixture was vortexed, and the glass vials were stirred in the dark at 67 °C for 3 h. Next, 100 μL of 6 M HCl was added to the derivatization product, and the mixture was vortexed. The mixture was placed in the lyophilizer to dry, then resuspended in 400 μL of acetonitrile and transferred to a 1.5 mL centrifuge tube. The resuspension mixture was centrifuged for 10 min, and the supernatant was collected and analyzed on an Agilent 6545 LC/Q-TOF instrument using a Kinetex 1.7 μm F5 100 Å, LC column (100 × 2.1 mm; Phenomenex; part no.00D-4722-AN) at a constant flow rate of 0.4 mL/min, and a gradient of 5-80 % of Solvent B (MeCN with 0.1% formic acid) in Solvent A (water with 0.1% formic acid). The following exact masses were extracted for MeLan (m/z 809.25299) and Lan (m/z 795.23734) from the chromatographic separation of each derivatized product and the co-injections.

4.10. NMR analysis

Fragments 2 and 5 were dissolved in 90% H₂O and 10% D₂O. Based on peak intensities, the estimated concentrations were ~0.23 mM for fragment 5 and ~0.20 mM for fragment 2. All NMR data were collected at 25 °C on a Bruker Avance NEO 600 MHz spectrometer equipped with a 5-mm curio BBO prodigy probe, including one-dimensional (1D) and two-dimensional (2D) homonuclear ¹H-¹H TOCSY (total correlation spectroscopy which reveals the correlation of protons in the same spin system) at 80 ms mixing time, 2D ¹H-¹H NOESY (Nuclear Overhauser Effect spectroscopy which reveals close spatial proximity between two protons) at 400 ms mixing time, as well as a 2D multiplicity edited ¹H-¹³C HSQC (heteronuclear single quantum coherence spectroscopy, revealing one-bond correlation between ¹H and ¹³C in which CH/CH₃ cross-peaks appear with opposite phase relative to CH₂ groups). All spectra were acquired using standard pulse programs in Bruker Topspin 4.1.4 software and processed and analyzed using Mnova (version 15.1.0.; Mestrelab Research).

4.11. Agar Well Diffusion Assays to Determine Antimicrobial Activity

Bioassays were performed using agar well-diffusion growth-inhibition assays with the protease-cleaved purified peptides. Peptides were dissolved in 50 mM Tris-HCl, 100 mM NaCl, pH 7.5 (initial assay peptide dissolved in 5 % DMSO) to achieve a starting concentration of ~0.5-1.8 mM. Agar plates were prepared by making a bottom layer with LB growth medium supplemented with 1.5% (w/v) agar (20–25 mL per 90 mm plate). The agar was allowed to solidify at room temperature. Next, the soft agar was prepared by mixing an overnight culture diluted to ~10⁶–10⁷ CFU/mL with molten LB agar containing 0.7% (w/v) agar. The mixture was poured over the solidified base layer and allowed to solidify. Then 1.5 μL of the positive controls, kanamycin (4.29 mM) or bacitracin (35 mM), was added to each plate as well as 2 μL of the tested compounds. The dried plates were incubated at 37 °C for 18 h, and the presence or absence of zones of growth inhibition was used to determine activity. The negative control was 50 mM Tris-HCl, 100 mM NaCl, pH 7.5.

Supplementary Material

The following files are available free of charge:

Supporting Information with additional figures, tables with primers and chemical shift assignments.

Data availability Statement

All data in this study are available at: Weir, Enleyona; Zhu, L; van der Donk, Wilfred (2026), “Data associated with "Structure and activity of a class II lanthipeptide from a thermophilic bacterium"”, Mendeley Data, V1, doi: 10.17632/jgbnmfyx44.1