Exploring Structural Signatures of the Lanthipeptide Prochlorosin 2.8 using Tandem Mass Spectrometry and Trapped Ion Mobility-Mass Spectrometry

Abstract

Lanthipeptides are a family of ribosomally synthesized and post-translationally modified peptides (RiPPs) characterized by intramolecular thioether cross-links formed between a dehydrated serine/threonine (dSer/dThr) and a cysteine residue. Prochlorosin 2.8 (Pcn2.8) is a class II lanthipeptide that exhibits a non-overlapping thioether ring pattern, for which no biological activity has been reported yet. The variant Pcn2.8[16RGD] has been shown to bind tightly to the αvβ3 integrin receptor. In the present work, tandem mass spectrometry, using collision induced dissociation (CID) and electron capture dissociation (ECD), and trapped ion mobility spectrometry – mass spectrometry (TIMS-MS), were used to investigate structural signatures for the non-overlapping thioether ring pattern of Pcn2.8. CID experiments on Pcn2.8 yielded b_i and y_j fragments between the thioether cross-links, evidencing the presence of a non-overlapping thioether ring pattern. ECD experiments of Pcn2.8 showed a significant increase of hydrogen migration events near the residues involved in the thioether rings with a more pronounced effect at the dehydrated residues as compared to the cysteine residues. The high-resolution mobility analysis, aided by site-directed mutagenesis ([P8A], [P11A], [P12A], [P8A/P11A], [P8A/P12A], [P11A/P12A] and [P8A/P11A/P12A] variants), demonstrated that Pcn2.8 adopts cis/trans-conformations at Pro8, Pro11 and Pro12 residues. These observations were found complementary to recent NMR findings, for which only the Pro8 residue was evidenced to adopt cis/trans-orientations. This study highlights the analytical power of the TIMS-MS/MS workflow for the structural characterization of lanthipeptides and could be a useful tool in our understanding of the biologically important structural elements that drive the thioether cyclization process.

Graphical Abstract

1. INTRODUCTION

Lanthipeptides are a structurally unique class of ribosomally synthesized and post-translationally modified peptides (RiPPs) that display various biological activities, including antimicrobial, antifungal, morphogenetic, antiviral, antinociceptive, and antiallodynic properties [1,2]. Lanthipeptides are characterized by intramolecular β-thioether cross-links, called (methyl)lanthionines ((Me)Lan), formed between a dehydrated serine (dSer) or threonine (dThr) residue and a cysteine residue through the action of lanthipeptide synthetases (Figure 1a) [1–3]. The thioether cross-links are generated by a 1,4-Michael addition between the thiol group of the cysteine residue and the β-carbon of the dehydrated residue (Figure 1a) [2]. The lanthipeptide family is divided into five classes depending on the dehydratation process (Figure 1b) [2–6]. Many lanthipeptides were discovered through genome-mining approaches and isolation of new lanthipeptide structures as well as generation of lanthipeptide libraries are an active area of research [7–16]. The utility of β-thioether cross-links for stabilizing lanthipeptides, both natural and designed, together with their panel of biological activities, makes them a promising scaffold for next-generation drug design [17–23]. However, altered lanthipeptide synthetases can result in different thioether ring patterns (Figure S1) [24,25], making their characterization not readily accessible using traditional structural biology tools. Thus, the discovery and design of new lanthipeptides as well as generation of lanthipeptide libraries as potential drug candidates require analytical tools capable of unambiguously and rapidly characterizing the lanthipeptide ring topologies.

Figure 1.

(a) Mechanisms of the lanthipeptide biosynthesis, starting with a dehydratation of Ser/Thr residues, followed by a cyclization through nucleophilic attack of Cys thiol groups. (b) General classification of lanthipeptide processing enzymes. The details of the recently reported class V lanthipeptides have not yet been worked out.

Prochlorosins are class II lanthipeptides produced by Prochlorococcus MIT 9313, which harbors multiple genes encoding ProcA precursor peptides but only a single gene encoding the lanthipeptide synthetase ProcM [2,26,27]. Among them, prochlorosin 2.8 (Pcn2.8) is a 19 residue lanthipeptide characterized by a non-overlapping thioether ring pattern (Figure 2) [28]. The intramolecular thioether cross-links are generated between Cys3 (highlighted in orange in Figure 2) and dSer9 (highlighted in green in Figure 2) residues as well as between dSer13 (highlighted in green in Figure 2) and Cys19 (highlighted in orange in Figure 2) residues. To date, no biological activity has been reported for Pcn2.8. However, an engineered variant, Pcn2.8[16RGD], has been demonstrated to tightly bind to the αvβ3 integrin receptor [28].

Figure 2.

Sequence, schematic and 3D representation (pdb: 6VLJ [29]) of the lanthipeptide prochlorosin 2.8. The Cys and dSer residues, involved in the thioether cross-link formation, are colored in orange and green, respectively. The Pro residues are highlighted in blue.

Modern nuclear magnetic resonance (NMR) spectroscopy is a method of choice for the comprehensive characterization of lanthipeptides and assignment of the ring patterns, as illustrated by several reported structures [2,4,27,29–39]. NMR studies on Pcn2.8 revealed a non-overlapping ring pattern as well as the presence of both cis- and trans-orientations at the Pro8 residue, with the major structure a cis-conformation, while Pro11 and Pro12 residues only exhibited nuclear Overhauser effects (NOE) characteristic of trans-conformations [29]. Due to the relatively large amount of sample required for NMR studies, the ring patterns of most other prochlorosins have only been predicted based on tandem mass spectrometry (MS/MS) using collision induced dissociation (CID) [25,28,40]. As the presence of a thioether cross-link prohibits cleavage inside the ring, this MS/MS approach is useful for lanthipeptide ring assignment since different fragmentation patterns would be observed for non-overlapping and overlapping ring topologies [25]. Pcn2.8 generates fragments between dSer9 (highlighted in green in Figure 2) and dSer13 (highlighted in green in Figure 2) residues, evidencing a non-overlapping ring pattern using CID [25,28]. Other complementary MS-based techniques, including electron capture dissociation (ECD) and ion mobility spectrometry – mass spectrometry (IMS-MS), were demonstrated to be very efficient alternate strategies for the structural characterization and pattern assignment for other RiPPs, such as lasso peptides [41–46].

In the present work, the lanthipeptide Pcn2.8 (Figure 2) was investigated using tandem mass spectrometry (CID and ECD) and trapped ion mobility spectrometry – mass spectrometry (TIMS-MS). The CID and ECD fragmentation patterns as well as mobility profiles were probed to evidence a non-overlapping ring pattern for Pcn2.8. In addition, the cis/trans-proline orientations at Pro8, Pro11 and Pro12 residues in Pcn2.8 (highlighted in blue in Figure 2) were studied using CID, ECD and TIMS-MS in combination with site-directed mutagenesis, involving [P8A], [P11A], [P12A], [P8A/P11A], [P8A/P12A], [P11A/P12A] and [P8A/P11A/P12A] variants, for direct comparison with recently reported NMR studies.

2. EXPERIMENTAL SECTION

2.1. Peptides and Sample Preparation.

Details on production of Pcn2.8 and its variants have been previously described [28]. Briefly, we employed a small scale expression protocol using E. coli BL21(DE3) carrying a pRSF Duet plasmid that encodes His₆-tagged ProcA2.8 (or variants thereof) and untagged ProcM. Production was accomplished by expression in 100 mL of lysogeny broth (LB) medium using 50 μg/mL kanamycin as selection marker. Expression cultures were inoculated 1:100 with a 37 °C LB overnight culture and then grown at 37 °C. When the optical density at 600 nm (OD600) reached ~0.4–0.5, the culture flasks were moved to another culture shaker at 18 °C, shaken for another 1 h, and then induced by adding 20 μL of a 0.5 M stock solution of isopropyl β-D-1-thiogalactopyranoside (IPTG, final concentration 0.1 mM) at an OD600 of ~0.5–0.7. Expression was carried out overnight at 18 °C. The next morning, 15 mL aliquots of the cultures were harvested by centrifugation. The modified His₆-ProcA2.8 precursor and variants thereof were isolated by resuspending cell pellets in 1 mL of guanidinium chloride lysis buffer (500 mM NaCl, 20 mM phosphate, 6 M guanidinium chloride, 0.5 mM imidazole, pH 7.5), lysing them at RT using a Vibra Cell sonicator (Sonics & Materials) with a microtip (settings: 40% amplitude, 30 s total sonication time alternating between 2 s on- and 2 s off-pulse), and then performing microscale NiNTA chromatography. For the latter, the lysates were cleared by centrifugation (RT, 20 min, 15700g) and transferred to fresh 2 mL microreaction tubes. Next, 180 μL of NiNTA resin (prequilibrated by washing twice with the guanidinium chloride lysis buffer) were added, and samples were incubated at RT for 30 min while slowly shaking. Samples were then loaded onto empty microbiospin chromatography columns (BioRad) and centrifuged at 270g until all liquid had passed through. The retained resin was washed twice. First by resuspending the resin with 800 μL of guanidinium chloride wash buffer (500 mM NaCl, 20 mM phosphate, 4 M guanidinium chloride, 30 mM imidazole, pH 7.5). Then, after again removing the liquid by centrifugation at 270g, by resuspending the resin with 800 μL of a PBS wash buffer (300 mM NaCl, 20 mM phosphate, 30 mM imidazole, pH 7.5). After centrifugation at 270g until all liquid was removed, the columns were transferred to fresh 2 mL microreaction tubes and eluted with 500 μL of a PBS elution buffer (300 mM NaCl, 20 mM phosphate, 500 mM imidazole, 1 mM tris(2-carboxyethyl)phosphine, pH 7.5). For the elution, the resin was resuspended on the column in the elution buffer and incubated for 5 min at RT, before centrifugation at 270g until all liquid passed into the collection tube. For release of the modified core peptide, the elution fraction were incubated with LahT(150) as described previously [28,47]. The [P8A], [P11A], [P12A], [P8A/P11A], [P8A/P12A], [P11A/P12A] and [P8A/P11A/P12A] variants of Pcn2.8 were produced by site-directed ligase-independent mutagenesis (SLIM) on the expression plasmid his₆-procA2.8(G−1K)[MCS1]:procM[MCS2] pRSF following standard protocols [48,49]. The lanthipeptides were subjected to C₁₈ Zip Tip (Millipore, Burlington, MA) purification, resulting in solutions at final concentrations of ~5 μM in 50:50 methanol/water containing 0.1% formic acid. The instrument was externally calibrated using the Tuning Mix (Agilent, Santa Clara, CA).

2.2. ECD Experiments.

ECD experiments were carried out on a Solarix 7 T FT-ICR mass spectrometer (Bruker, Billerica, MA) equipped with an Infinity cell and a nanoESI source operated in the positive ion mode. NanoESI emitters were pulled from quartz capillaries (O.D. = 1.0 mm and I.D. = 0.70 mm) with the use of a Sutter Instruments Co. P2000 laser puller (Sutter Instruments, Novato, CA). Sample aliquots (10 μL of 5 μM solution) were loaded in a pulled-tip capillary, housed in a mounted custom built XYZ stage in front of the MS inlet, and voltages were applied via a tungsten wire inserted inside the nESI emitters. The nESI high voltage, capillary exit, and skimmers I and II were set to 1300 V, 200 V, 30 V and 4 V, respectively. Precursor ions were isolated in the instrument quadrupole with a mass window of 5 Da, accumulated for 0.5 s in the collision cell, and further injected into the ICR cell. ECD experiments were performed on the mass-selected [M+3H]³⁺ ions with a heated hollow cathode operating at a current of 1.5 A. Electrons emitted during 0.2 s were injected into the ICR cell with a 2.2 V bias and 10 V ECD lens. A total of 150 scans (m/z range 100–2000) were co-added with a data acquisition size of 512 K words.

2.3. TIMS-MS and CID Experiments.

Ion mobility experiments were performed on a custom built nanoESI-TIMS coupled to an Impact Q-TOF mass spectrometer (Bruker, Billerica, MA, Figure S2) [50]. The TIMS unit is controlled using a custom software in LabView (National Instruments) synchronized with the MS platform controls [51]. NanoESI emitters were pulled from quartz capillaries (O.D. = 1.0 mm and I.D. = 0.70 mm) using a Sutter Instrument Co. P2000 laser puller (Sutter Instruments, Novato, CA). The general fundamentals of TIMS as well as the calibration procedure have been described previously [52–55]. Peptide sample solutions were loaded in a pulled-tip capillary, housed in a mounted custom built XYZ stage in front of the MS inlet, and sprayed at 950 V via a tungsten wire inserted inside the nESI emitters. TIMS-MS experiments were carried out using nitrogen (N₂) as buffer gas at room temperature. The gas velocity was kept constant between the funnel entrance (P1 = 2.6 mbar) and exit (P2 = 0.8 mbar, Figure S2). An rf voltage of 250 V_pp at 880 kHz was applied to all electrodes. A voltage ramp (V_ramp) of −190 to −75 V, deflector voltage (V_def) of 60 V and base voltage (V_out) of 60 V were used for the mobility separations. Changes in the mobility profiles were not observed for the systems reported over 100–500 ms time range after desolvation. Collision induced dissociation (CID) experiments were performed in the collision cell located after the TIMS analyzer (Figure S2). The mass-selected [M+3H]³⁺ ions were fragmented using nitrogen as collision gas at a collision energy of 20–23 V. The mobility resolving power (R) values were determined as R = CCS/ΔCCS, where ΔCCS is the full peak width at half maximum (FWHM) of the IMS band. A gaussian peak fitting algorithm with non-linear least squares functions (Levenberg–Marquardt algorithm) using OriginPro 2016 was used to evaluate the FWHM of each IMS band.

2.4. Correction of Pro substituted collision cross section (CCS).

A direct comparison of the mobility profiles between the lanthipeptide Pcn2.8 and the [P8A], [P11A], [P12A], [P8A/P11A], [P8A/P12A], [P11A/P12A] and [P8A/P11A/P12A] variants was done by adjusting the CCS profiles based on methodology previously described [56–59]. Differences in CCS between N₂ and He were adjusted using the CCS_N2 = 1.0857 (CCS_He) + 81.459 [Ų] equation [60,61]. This resulted in Pro to Ala substitutions to be corrected by 2.68 Ų in N₂.

3. RESULTS AND DISCUSSION

3.1. Fragmentation Pattern of Pcn2.8 in CID.

The mass spectrometry analysis of the lanthipeptide Pcn2.8 resulted in the observation of doubly ([M + 2H]²⁺, m/z 1025.9) and triply ([M + 3H]³⁺, m/z 684.2) protonated species (Figure S3). The CID analysis of the [M + 3H]³⁺ ions of Pcn2.8 (Figure 3a) exhibited multiple fragmentation patterns due to the presence of the thioether cross-links that prevent the formation of classic b_i/y_j fragment ions inside the ring. However, cross-linked product ions from b_i and y_j ions, denoted as [(b_i)−(y_j)], were observed. These fragments result from two bond cleavages in the thioether ring, yielding internal fragments denoted as (b_iy_j)_n, where the rest of the peptide remains covalently linked to the backbone through the thioether bond (Figure 3a). The main cross-linked product ions observed for Pcn2.8 were [(b₁₄)−(y₁)]²⁺, [(b₁₄)−(y₃)]²⁺, [(b₁₅)−(y₂)]²⁺, and [(b₁₅)−(y₃)]²⁺, associated with their complementary internal fragment ions (b₁₈y₅)₄, (b₁₆y₅)₂, (b₁₇y₄)₂, and (b₁₆y₄)₁, respectively. In addition, classic b_i/y_j series were observed, corresponding to fragmentations outside of the thioether rings (Figure 3a). The main product ions observed for Pcn2.8 were b₉²⁺, b₁₀²⁺, b₁₁²⁺ and b₁₂²⁺ series, together with their complementary y₁₀, y₉, y₈ and y₇ series. These fragments are located between the two thioether rings clearly evidencing the presence of a non-overlapping lanthipeptide ring pattern, as previously described [25,28]. In fact, b_i/y_j series between the two thioether cross-links cannot be observed for the case of an overlapping lanthipeptide ring pattern because these linker residues are involved in the thioether ring (Figure S1). The cross-linked product ions were not observed for the previously reported MS/MS of Pcn2.8 using the [M + 2H]²⁺ ions [25,28]. This observation demonstrates that choosing the [M + 3H]³⁺ charge state as the precursor ion in CID allowed the detection of cross-linked fragment species and hence facilitates sequence coverage of lanthipeptides by delineating the residues present inside the thioether ring. A similar behavior has also been reported for the class I/III lasso peptides containing disulfide bonds [43,44].

Figure 3.

Typical tandem MS/MS spectra of the [M + 3H]³⁺ ions of the lanthipeptide Pcn2.8 (m/z 684.2) using (a) CID and (b) ECD. Typical cross-linked/internal fragments and hydrogen migration events are highlighted in purple and red, respectively, and labeled on the peptide cartoons (right of each panel). (c) Bar plot showing the hydrogen migration events of Pcn2.8 obtained by comparison between the experimental and theoretical isotopic patterns. The residues involved in the thioether cross-link are colored in orange (former Cys) and green (former Ser), respectively. The Pro residues are highlighted in blue.

3.2. Fragmentation Pattern of Pcn2.8 in ECD.

We have previously illustrated the utility of ETD/ECD for the study of lasso peptides containing disulfide bonds as an alternative fragmentation approach to overcome potential limited structural information provided by CID [43,44]. The ETD/ECD processes involve the transfer/capture of a single electron that can induce dissociation in the backbone as well as in disulfide bonds [62–65]. ECD experiments were performed on the [M + 3H]³⁺ ions of Pcn2.8 yielding the spectrum presented in Figure 3b. The charge-reduced [M + 3H]^•2+ ions (m/z 1025.9) were detected as the most abundant species. In addition, c_i′/z_j^• series were observed that involve the cleavage of the two thioether cross-links concomitant with backbone cleavages at each extremity of the thioether rings (Figure 3b). Here, we assume that the cleavage mechanism of thioether rings is similar to disulfide bonds. In fact, the cleavage of the N–C_α bond at the backbone region within the thioether bond occurs first, generating an intermediate radical that contains a radical site in the C_α atom, then followed by radical-driven reactions, where the C_α radical attacks the thioether bond, inducing cleavage at the S–C bond [62,66,65]. This results in the coverage of almost the entire lanthipeptide sequence and proves to be a more efficient approach than CID for the amino acid sequence elucidation. The cross-linked fragment ions generated by CID will probably not be detected when the size of the thioether ring becomes smaller, as observed for the lasso peptides [44]. Furthermore, ECD experiments of Pcn2.8 exhibited different hydrogen migration events when comparing the backbone residues with the residues involved as well as in close proximity of the thioether cross-links (Figure 3c). The c₈^′, c₁₂^′ and c₁₃^′ product ions were mass shifted by ~1 Da by the loss of H•, giving the c₈^•, c₁₂^• and c₁₃^• fragment ions, respectively (highlighted in red in Figure 3b). Conversely, the z₆^•, z₇^•, z₁₁^•, z₁₅^• and z₁₆^•2+ fragments were mass shifted by the capture of H•, giving the z₆^′, z₇^′, z₁₁^′, z₁₅^′ and z₁₆^′2+ product ions, respectively (highlighted in red in Figure 3b). In addition, hydrogen migration events occurred less frequently when residues are located further away of the thioether cross-links (Figure 3c). Furthermore, larger extents of hydrogen migration were observed at the former Ser residues (highlighted in green) as compared to the former Cys residues (highlighted in orange). This suggests that ECD can assign which dSer/dThr residues are involved in thioether cross-links and therefore likely differentiate them from dSer/dThr residues not involved in a thioether for lanthipeptides where both are present. However, the assessment of the thioether ring connectivity remained ambiguous in ECD as compared to CID due to the cleavage of the thioether cross-links. As a consequence, the combination of both CID and ECD methods enabled to fully sequence Pcn2.8 as well as to firmly assess the lanthipeptide ring pattern. This analytical approach can be easily translated to other lanthipeptides and shows, like previously for lasso peptides [44,43], how CID and ECD of macrocyclic RiPPs complement each other.

3.3. TIMS-MS Analysis of Pcn2.8.

TIMS-MS proved to be a very effective approach for the characterization of lasso peptides by providing additional information on the conformational diversity due to the high mobility resolving power of the TIMS analyzer [41,43,67]. TIMS spectra corresponding to the doubly and triply protonated species of Pcn2.8 are illustrated in Figures 4a and 4b, respectively. TIMS analysis for the [M + 2H]²⁺ and [M + 3H]³⁺ molecular species of Pcn2.8 resulted in the identification of a large number of IMS bands, with an apparent resolvin power of R ~ 185 and 155 using a Sr = 0.23 V/ms, respectively. This observation suggests that the thioether cross-links probably prevent the lanthipeptide from completely collapsing toward compact structures, as reflected by the broad CCS distribution (> 50 Ų). In addition, the observation of a large conformational diversity suggests that the Pcn2.8 structure can be stabilized using several combinations of intramolecular interactions. For example, the absence of strong basic residues (e.g. Arg and Lys) can induce a competition in the protonation schemes, leading to different charge driven intramolecular interactions that will define a particular lanthipeptide structure. For example, one protonation scheme could preserve a rigid fold (compact structure) through charge solvation, possibly including salt bridges, bringing the residues involved in the two thioether cross-links in proximity and stabilized by intramolecular hydrogen bonding interactions. An alternative scheme could include protonation that will pull apart the two thioether cross-links (extended structure) induced by Coulombic repulsions. This hypothesis is consistent with the recently reported electrostatic surface map of Pcn2.8 from its NMR structure, showing that the lanthipeptide neither has a preferred fold nor a specific charge organization on the surface of the peptide [29]. In addition, a particularity of Pcn2.8 is the presence of several Pro residues (Pro8, Pro11 and Pro12). This composition suggests that the conformational diversity can also be induced by cis/trans-isomerization of Pro residues, as previously observed for the lasso peptide microcin J25 [68]. In fact, the recent NMR study showed that Pcn2.8 can adopt two stable conformations, involving a cis/trans-equilibrium of the Ala7-Pro8 peptide bond while the Pro11 and Pro12 residues occupy only the trans-configurations [29].

Figure 4.

Typical TIMS spectra of the (a) doubly and (b) triply charged species of Pcn2.8. The former Cys and former Ser (Ala) residues that are involved in the thioether cross-links are colored in orange and green, respectively. The Pro residues are highlighted in blue.

3.4. Evidence of cis/trans-Conformation at Pro8/Pro11/Pro12 Residues.

The Pro8, Pro11 and Pro12 residues of Pcn2.8 were substituted with Ala residues through site-directed mutagenesis to assess the influence of possible cis-conformations at Pro residues. These mutagenesis experiments resulted in the generation of the [P8A], [P11A], [P12A], [P8A/P11A], [P8A/P12A], [P11A/P12A] and [P8A/P11A/P12A] Pcn2.8 variants. The CID and ECD experiments using the [M + 3H]³⁺ precursor ions of Pcn2.8 variants were consistent with the altered sequences for each variant (Figures S4 and S5). In addition, b_i/y_j product ions were observed by CID between the two thioether cross-links, demonstrating that the substitution by Ala residues does not affect the non-overlapping lanthipeptide ring pattern of Pcn2.8 (Figure S4). The ECD experiments also illustrated an alteration in the environment of the residues in close proximity of the thioether cross-links, as reflected by the changes observed in the hydrogen migration events (Figure S5). These observations suggest that each Pro residue plays an important role in the Pcn2.8 lanthipeptide structure.

TIMS-MS analysis of the investigated Pcn2.8 variants allowed the identification of the cis/trans-orientations of the proline state in wild type (WT) Pcn2.8 by direct comparison with the corrected mobility profiles (Table S1). The conservation or absence of an IMS band in the Pcn2.8 variants will then permit the assignment of the proline orientation in WT Pcn2.8, as previously reported for the lasso peptide microcin J25 [68]. High resolution TIMS analysis enabled the identification of fourteen IMS bands for the doubly protonated species of WT Pcn2.8 (Figure 5). In addition, distinct TIMS profiles were observed for WT Pcn2.8 and the Pcn2.8 variants, suggesting that the conformational diversity of WT Pcn2.8 involves cis/trans-conformations at proline residues (Table S1). TIMS distribution of the [P8A/P11A/P12A] variant, for which only trans-conformations are expected, indicated that the IMS bands 1, 3, 4, 5 and 7 correspond to trans-Pro8/trans-Pro11/trans-Pro12 (blue dashed lines in Figure 5). Note that these structures appeared to be the most compact structures in WT Pcn2.8, with the cis-conformation at Pro residues inducing then an unfolding event. The IMS 2 band (magenta dashed line in Figure 5) of WT Pcn2.8 was found common with only the [P8A/P11A] and [P8A/P12A] variants, indicating that the IMS 2 band probably involves only one proline (Pro11 or Pro12) in a cis-configuration, while the Pro8 residues adopts a trans-configuration. Thus, these observations suggest the presence of two possible structures with similar CCS for the IMS 2 band, involving a trans-Pro8/cis-Pro11/trans-Pro12 and/or a trans-Pro8/trans-Pro11/cis-Pro12. The IMS 6 band (red dashed line in Figure 5) was only observed for the [P8A] variant concerning the single point mutation. This finding suggests the presence of a trans-Pro8/cis-Pro11/cis-Pro12 structure for the IMS 6 band. However, the IMS 6 band was also present for the [P8A/P11A], [P8A/P12A] and [P11A/P12A] variants, suggesting that structures including at least one of the prolines in a cis-configurations are observed. The IMS 8 band had a similar behavior as that of the IMS 6 band, for which the double mutations indicated that at least one of the prolines is in a cis-orientation, while the single point mutations pointed toward the presence of a cis-Pro8/trans-Pro11/trans-Pro12. The IMS 9, 10, 11 and 12 bands were generally observed in all Pcn2.8 variants, except for the [P8A/P11A/P12A] compound, for which only trans-configurations are observed, suggesting that these bands correspond to conformers in which at least one of the proline is in a cis-orientation (red dashed lines in Figure 5). The IMS 13 (purple dashed line) and 14 (green dashed line) bands were found unique for the [P11A] and [P12A] variants, respectively (Figure 5). This observation suggests that the IMS 13 band is specific for cis-Pro8/trans-Pro11/cis-Pro12, while the IMS 14 band is specific for cis-Pro8/cis-Pro11/trans-Pro12 (Table S1).

Figure 5.

Typical TIMS spectra for the [M + 2H]²⁺ ions of Pcn2.8 as well as the [P8A], [P11A] and [P12A] single variants, the [P8A/P11A], [P8A/P12A] and [P11A/P12A] double variants and the [P8A/P11A/P12A] triple variant. The CCS profiles of the mutants have been corrected for a direct IMS band comparison. The former Cys and former Ser residues that are involved in the thioether cross-links are colored in orange and green, respectively. The Pro residues are highlighted in blue.

These TIMS results are in good agrement with recently reported NMR experiments [29], for which two populations involving a cis/trans-equilibrium of the Ala7-Pro8 peptide bond were observed. These NMR experiments were not able to decipher the Pro11 and Pro12 residues in a cis-orientation, probably because the trans-configuration at Pro11 and Pro12 are largely predominant in solution.

4. CONCLUSION

Mass spectrometry based experiments were used to study the Pcn2.8 lanthipeptide structure and its potential ring patterns. The CID MS/MS experiments were found effective for the assignment of a non-overlapping lanthipeptide ring pattern by specifically yielding b_i and y_j fragments between the two thioether cross-links. The ECD MS/MS was effective to fully sequence the lanthipeptide by cleaving inside the thioether cross-links and by specifically showing larger extent of hydrogen migration near the residues involved in the thioether rings. The use of high-resolution mobility – mass spectrometry (TIMS-MS), in combination with site-directed mutagenesis ([P8A], [P11A], [P12A], [P8A/P11A], [P8A/P12A], [P11A/P12A] and [P8A/P11A/P12A] variants), permitted the identification of fourteen IMS bands, with varying prolines (Pro8, Pro11 and Pro12) in cis/trans-configurations. These observations provided complementary information to recent NMR findings, for which only the Pro8 residue was shown to adopt cis/trans-orientations. This study highlights the analytical power of the tandem MS/MS and TIMS-MS workflows for the structural characterization of lanthipeptides as a useful tool for a better understanding of the intramolecular forces that define the lanthipeptides folding motifs as well as the structural elements that drive the thioether cyclization process.