Facile Method for Determining Lanthipeptide Stereochemistry

Lanthipeptides make up a large group of natural products that belong to the ribosomally synthesized and post-translationally modified peptides (RiPPs). Lanthipeptides contain lanthionine and methyllanthionine bis-amino acids that have varying stereochemistry. The stereochemistry of new lanthipeptides is often not determined because current methods require equipment that is not standard in most laboratories. In this study, we developed a facile, efficient, and user-friendly method for detecting lanthipeptide stereochemistry, utilizing advanced Marfey’s analysis with detection by liquid chromatography coupled with mass spectrometry (LC-MS). Under optimized conditions, 0.05 mg of peptide is sufficient to characterize the stereochemistry of five (methyl)lanthionines of different stereochemistry using a simple liquid chromatography setup, which is a much lower detection limit than current methods. In addition, we describe methods to readily access standards of the three different methyllanthionine stereoisomers and two different lanthionine stereoisomers that have been reported in known lanthipeptides. The developed workflow uses a commonly used nonchiral column system and offers a scalable platform to assist antimicrobial discovery. We illustrate its utility with an example of a lanthipeptide discovered by genome mining.

−4 Lanthipeptides are conformationally restrained to recognize their biological targets through thioether cross-links called lanthionine (Lan) and methyllanthionine (MeLan).These structures are formed by the enzymatic dehydration of Ser and Thr residues followed by the Michael-type addition of the thiols of Cys residues to the dehydrated amino acids (Figure 1). 5 Lanthipeptides owe their various bioactivities to the intricate stereochemistry embedded in their three-dimensional geometry.
−10 These isomers have been named DL-and LL-Lan, and DL-, LL-, and D-allo-L-MeLan, respectively (Figure 1).Their configurations were determined by first hydrolyzing the peptides to their individual amino acids and then chemically manipulating them to obtain volatile derivatives.These compounds were then analyzed by gas chromatography (GC) monitored by mass spectrometry using a column with a chiral stationary phase (CP-chirasil-L-Val) and compared to synthetic standards obtained by a multistep process. 8,11These methods have been successful but require relatively large amounts of material (∼1−2 mg for a typical lanthipeptide), a gas chromatograph, a chiral column, and chemical synthesis of standards of defined stereochemistry.Furthermore, in our hands, the chemical derivatization of the amino acids and subsequent injection onto the GC column damages the column that has a considerable price tag (∼$1400) such that its use is severely shortened (typically about 1 year of use).An alternative method that has been used is chemical desulfurization of Lan/MeLan residues and determination of the stereochemistry of the resulting Ala and 2-aminobutyric acid, 12 but this strategy will not report on the stereochemistry on carbon 3 of MeLan.
Altering the stereochemistry of the (Me)Lan residues in the lanthipeptide lacticin 481 lead to the abolishment of its bioactivity. 13This observation underscores the critical role of stereochemistry in the biological function of lanthipeptides.−18 Investigating the structural diversity and the interplay of stereochemistry in lanthipeptides holds significant importance in understanding their mechanism of action and exploring their application.However, the stereochemistry of the Lan and MeLan residues of the great majority of newly reported lanthipeptides is not determined.Therefore, as more and more lanthipeptides are discovered by genome mining, 19−46 a convenient method for determining their stereochemistry that is accessible to most laboratories would be valuable.We report here such a method including access to standards that use biochemical approaches that are complementary to chemical synthesis and methods available in most if not all laboratories studying lanthipeptides or RiPPs.
Marfey's reagent, 1-fluoro-2,4-dinitrophenyl-5-L-alanine amide (L-FDAA, Figure 1C), has been widely used in peptide structure determination due to its ability to provide a robust and reliable method for analyzing the stereochemistry of amino acids in complex mixtures. 47,48The process involves hydrolysis of the peptide sample under acidic conditions followed by derivatization of the amino groups with L-FDAA to form diastereomers. Subsequently, these diastereomers are separated by high-performance liquid chromatography (HPLC).By analysis of the elution order and retention time and comparison with synthetic standards, the absolute stereochemistry of the hydrolyzed fragments in the peptide sample can be determined.
The increase in the number of possible derivatizations originating from multiple reactive groups such as amino, hydroxyl, and thiol groups in an amino acid analog can pose a challenge to Marfey's analysis. 48Given that Lan and MeLan are bis(amino acids), they contain two sites of derivatization (Figure 1).The resulting problem was exemplified in the determination of the stereochemistry of D-allo-L methyllanthionine. 10In this case, bis-derivatization of the two amino groups with L-FDAA resulted in the overlap of the peaks of derivatized DL-MeLan and D-allo-L-MeLan in extracted ion chromatography (EIC). 10One possible explanation for this overlap is the limited interaction of FDAA with diastereomers that differ only in the stereochemistry at position 3 (Figure 1).We hypothesized that increasing the steric effect on the derivatization reagent might lead to full separation of the peaks of such diastereomers.In this regard, 1-fluoro-2,4-dinitrophenyl-5-L-leucine amide (L-FDLA, Figure 1C) has been used as a more robust reagent for amino acid characterization. 49Use of FDLA is often referred to as advanced Marfey's analysis.We show herein that FDLA can distinguish the five known stereoisomers of (Me)Lan, and we describe how standards of these isomers can be obtained using publicly available resources.We illustrate the use of the methods with a lanthipeptide discovered by genome mining.

■ EXPERIMENTAL SECTION
Lan and MeLan Standard Preparation.Nisin.Nisin was obtained by extraction from a Sigma-Aldrich product that contains about 5% nisin from Lactococcus lactis.The process involved suspending 100 mg of the commercial product in 10 mL of extraction solvent consisting of 80% acetonitrile (MeCN)/20% H 2 O + 0.1% trifluoroacetic acid (TFA) in a 50 mL conical tube, followed by vigorous vortexing.After centrifuging at 4000g for 10 min, the supernatant was collected, and the extraction of the insoluble material was repeated twice with 10 mL of the same solvent.The combined supernatant was then filtered using a 10 K Amicon Ultra-15 Centrifugal Filter, and the resulting flowthrough was collected.The solution was frozen with liquid nitrogen and lyophilized.The resulting dry solid was redissolved in 10% MeCN/90% H 2 O + 0.1% TFA, filtered, and injected onto an Agilent preparative HPLC system for purification using a Phenomenex Luna C5 column (10 μm, 250 × 10 mm 2 ).The following gradient conditions were used for the purification process: flow rate of 4 mL/min; solvent A H  method. 9 For the preparation of lanthionine standard, 1 L of expression of mCylL L in LB medium was typically sufficient.
The collected fractions were lyophilized to dryness and then redissolved in 10% MeCN/90% H 2 O + 0.1% TFA before injection onto an Agilent analytical HPLC system for purification using a Thermo Scientific C8 Hypersil GOLD column (5 μm, 4.6 × 250 mm 2 ).The following gradient conditions were used for the purification process: flow rate of 1.0 mL/min, solvent A H 2 O + 0.1% TFA, solvent B MeCN + 0.1% TFA.The gradient profile was as follows: 0−2 min with 2% B, 2−10 min using a gradient of 2−30% B, 10−15 min using a gradient of 30−40% B, and 15−20 min using a gradient of 40−100% B. mCylL L eluted at 11−12 min, corresponding to 32−35% B. To ensure purity, only fractions showing significant absorption at wavelengths of 254 and 280 nm on a diode-array detector were collected and lyophilized for further usage.This method is also applicable to the expression and purification of mCylL S (plasmid available at Addgene ID #208760).
mBuvA.Plasmids were constructed and used to transform E. coli using a previously reported method. 30Single colonies were selected and amplified in 5 mL of starter cultures in LB containing 50 μg/mL of kanamycin overnight.The starter culture was introduced into 1 L of LB medium containing 50 μg/mL of kanamycin and incubated at 37 °C with continuous shaking at 220 rpm until it reached an OD 600 of approximately 0.8.Subsequently, the cultures were cooled on ice for approximately 15 min before the addition of isopropylthio-βgalactoside (IPTG) to a final concentration of 0.5 mM.Following this, the cultures were shaken overnight at 180 rpm at 18 °C.Cells were harvested by centrifugation at 6000g for 15 min at 4 °C.The supernatant was discarded, and the pellet was stored at −80 °C before purification.
Peptide Hydrolysis and Derivatization of Amino Acids.The peptide sample (0.02−0.1 mg) was added to PYREX 25 mL screw cap culture tubes with phenolic caps (20 mm × 125 mm, No. 9825-20, Fisher Scientific), followed by 0.8 mL of 6 M DCl in D 2 O. N 2 was bubbled through the solution for 1 min, and then the tube was sealed with the cap, and the mixture was stirred at 120 °C for 20 h or 155 °C for 3 h (see Figure S1).The high walls of the tube serve to condense the liquid, keeping it at reflux.For an alternative hydrolysis procedure that uses a heat block and no stirring, see the Supporting Information.The resulting mixture was dried using a rotary evaporator with the water bath at 80 °C, subsequently redissolved in deionized water, and dried twice more.Next, 0.6 mL of 0.8 M NaHCO 3 (in H 2 O) and 0.4 mL of 10 mg/mL L-FDLA or D-FDLA (in MeCN) were added.Following stirring in the dark at 67 °C for 3 h, 0.1 mL of 6 M HCl was gently added, and the mixture was vortexed vigorously.The mixture was then frozen by using liquid nitrogen and lyophilized in the dark until dry.The dried solid was redissolved in 1 mL of MeCN and vortexed vigorously.The suspension was carefully transferred into 1.5 mL Eppendorf tubes using a long glass pipet and centrifuged at 16,000g for 15 min.Afterward, the resulting supernatant was transferred to screw vials for LC-MS analysis.For samples with a low peak intensity of (methyl)lanthionines, the extraction solution was further concentrated up to 5-fold before reanalysis.
LC-MS Analysis.LC-MS analysis was performed on an Agilent 6545 LC/Q-TOF instrument.A Kinetex F5 Core− Shell HPLC column (1.7 μm F5 100 Å, LC Column 100 mm × 2.1 mm) was used to analyze reactions of Lan/MeLan with L-FDLA and D-FDLA.During the analysis, the temperature of all columns was maintained at 45 °C.HPLC parameters were as follows: flow rate 0.40 mL/min and the use of two mobile phases (A: H 2 O+0.1% formic acid; B: acetonitrile).The gradient was set as 2%−30% B over 0−2.5 min, 30%−80% B over 2.5−10 min, 80%−100% B over 10−10.5 min, 100%−2% B over 10.5−11 min, and then 2% B re-equilibrium for 2.5 min.The mass spectrometer was set to ion polarity (negative mode), dual AJS ESI (gas temperature 325 °C, drying gas 10 L/min, nebulizer 35 psi, sheath gas temperature 375 °C, sheath gas flow rate 11 L/min, VCap 3500 V, nozzle voltage 0 V), MS TOF (Fragmenter 125 V, skimmer 65 V, Oct 1 TF Vpp 750 V), and acquisition parameters (automatic MS/MS mode, mass range 50−1700 m/z).A volume of 2−5 μL per injection is recommended.To avoid potential contamination of the ion source by large amounts of unreacted L-FDLA, the liquid chromatography stream was injected into the mass spectrometry starting at 7.0 min as L-FDLA eluted around 6.5 min.Any remaining LC eluent was then directed into the waste without further MS analysis.
■ RESULTS AND DISCUSSION Advanced Marfey's Analysis of Lanthionine.−52 It contains one DL-Lan and three DL-MeLan. 6,11We obtained a commercial nisin extract from L. lactis and purified it using HPLC.After acidic hydrolysis, the results of FDLA derivatization showed the expected product peaks by EIC that were observed in significantly higher intensity (approximately 5−10-fold) compared to the FDAA treatment of the same sample (Figure 2A).
Next, we chose the lanthipeptide cytolysin L (CylL L ″), 53 a compound that can be produced in E. coli, 9 and for which we deposited the plasmid at Addgene (ID #208759).Modified cytolysin L precursor peptide (mCylL L ) comprises a three (Me)Lan system including one DL-Lan, one LL-Lan, and one LL-MeLan (Figure 3A). 9 Production of mCylL L in E. coli as a Histagged peptide, purification using nickel affinity chromatography, and acid hydrolysis provided a mixture of amino acids.L-FDAA derivatization did not lead to full separation of the two peaks of DL-and LL-Lan, while L-FDLA successfully produced two well-separated peaks of DL-and LL-Lan-FDLA adducts with an evenly distributed peak area (Figure 3B).This observation indicates the advantages of L-FDLA over standard Marfey's analysis and gives the possibility for quantitative analysis of (methyl)lanthionine.Using nisin's single DL-Lan as a standard, we elucidated the order of elution of the two diastereomers of derivatized DL-and LL-Lan (Figure 3B).
Unlike most amino acids, Lan and MeLan have multiple chiral centers and therefore have a higher chance of kinetic resolution during derivatization.Derivatization with L-FDLA did not result in such resolution.The choice of MeCN/H 2 O and NaHCO 3 as a base in the derivatization solution proved critical to obtaining the optimal intensity of the derivatized Lan/MeLan.Other derivatization conditions commonly used (e.g., acetone-triethylamine) resulted in lower peak intensities.To reduce the occurrence of the [M + Na] + form, which would require integration over multiple EICs, negative mode ionization was chosen, which also resulted in the intensity of the relevant ions being consistently about 5−10 times higher compared to the analysis of [M + H] + /[M + Na] + in the same sample.
Advanced Marfey's Analysis of Methyllanthionine.Next, we analyzed the three previously established natural diastereomers of MeLan: DL-, LL-, and D-allo-L-MeLan.As the MeLan standard, we selected mCoiA 1 , a modified precursor peptide that contains all three known MeLan diastereomers (Figure 4A). 18Once again, the standard was prepared by expression of the His-tagged peptide in E. coli using plasmids  deposited at Addgene (ID #208761 and #208762).The mCoiA 1 was purified by nickel affinity chromatography, and the peptide was hydrolyzed in acid.After derivatization with L-FDLA, LC mass spectrometry analysis showed three separate peaks (Figure 4B).To assign the stereochemistry of each peak, we introduced two site-directed mutations, mCoiA 1 -T43S (to remove LL-MeLan) and mCoiA 1 -T50S (to remove DL-MeLan). 10After applying identical treatment and analysis, the two mutations eliminated one of the three peaks each, thus assigning the stereochemistry of the corresponding MeLan and establishing the elution order of the three derivatized diastereomers.Therefore, the full-length mCoiA 1 peptide proved to be an effective standard for characterizing all three methyllanthionine isomers (Figure 4).
Optimization and Determination of the Detection Limit.Hydrolysis of peptides in acid at 120 °C for 20 h is usually the standard condition for Marfey's analysis, but a previous study showed that increasing the hydrolysis temperature to 150 °C for 3 h improved efficiency without leading to breakdown or oxidation of specific residues such as Trp and Cys. 54Therefore, we tested these hydrolysis conditions with 0.05 mg of nisin, mCylL L , and mCoiA1 and obtained similar results to those observed after hydrolysis at 120 °C for 20 h (Figure S1), thus significantly decreasing the time required for the analysis.
We next focused on determining the detection limit of the advanced Marfey method for determining the Lan and MeLan stereochemistry.We used the modified precursor peptide of cytolysin S (mCylL s ) as a model sample.This wellcharacterized 8 kDa, 78 amino acid sequence contains two rings, one LL-MeLan and one DL-Lan, and serves as a representative structure of a lanthipeptide genome mining exercise in that its leader peptide is still attached to the modified core peptide after expression in E. coli. 55xperimentally, we analyzed the limit of detection of mCylLs by serial dilution starting from 0.2 mg.The results suggested that 0.026 mg of the sample was sufficient to produce distinct LC-MS peaks with minimal background noise (Figure S2).This sensitivity is a major improvement over the amount of material required by using the GC-MS method.We also evaluated commercially available D-FDLA as a derivatization agent instead of L-FDLA.The separation of the diastereomers formed upon the reaction of D-FDLA with LL and DL-Lan was better than with L-FDLA, but the diastereomers formed upon the reaction of D-FDLA with LL-, DL-, and allo-DL-MeLan provided a mixture in which two of the isomers coeluted   S1.
(Figure S3).Hence, L-FDLA is the reagent of choice for the analysis of lanthipeptides of unknown stereochemistry.
To enhance the general applicability of our approach, we also conducted hydrolysis of the mCylL L and mCoiA 1 standards using a commonly used dry/bath block heater.LC-MS results showed no significant differences when comparing samples heated using a hot plate stirrer and those using the dry bath heater (Figure S4).
Application of the Method to a New Lanthipeptide.A previously established system known as FAST-RiPPs (Fast, Automated, Scalable, high-Throughput pipeline for RiPPs discovery) combines RODEO, 56 a tool for mining genomes to identify specific RiPP biosynthetic gene clusters (BGCs) of interest, with an automated pathway refactoring platform 57 and employs E. coli for heterologous RiPP expression. 30FAST-RiPPs was successful in producing a large number of novel lanthipeptides, but the stereochemistry was determined for only a subset of them because of the aforementioned need of 1−2 mg of purified final product, which could not always be obtained.In this study, we chose one such compound for which the stereochemistry is not known, LanII-21.This compound is the product of the buv BGC identified in the genome of Butyrivibrio sp.VCD2006, butyrate-producing bacteria inhabiting the rumens of ruminant animals.The buv BGC encodes multiple proteins, including an rSAM BuvX (WP_026528692.1), a class II lanthipeptide synthetase BuvM (WP_081674439.1), a cyclase BuvC (WP_081674440.1), and a precursor peptide BuvA (WP_026528694.1).These genes were incorporated into a pET28a plasmid backbone for heterologous expression and post-translational modification in E. coli (Figure 5A).After purification, 0.05 mg of mBuvA was hydrolyzed at 150 °C for 3 h, followed by derivatization with L- FDLA and LC-MS analysis.The result shows a single DL-Lan peak compared to the Lan-standard obtained from mCylL L , as confirmed by coelution upon coinjection with the standard (Figure 5B).No significant MeLan (L-DLA) 2 peak was obtained.This observation suggests that mBuvA forms a 5membered DL-Lan ring at the C-terminus since that is the only location to form a Lan.To further confirm this structure, we digested mBuvA with AspN and performed tandem ESI-MS analysis, which resulted in no fragments being observed within the proposed lanthionine ring (Figure 5C), consistent with our assignment.

■ CONCLUSIONS
In this study, we used L-FDLA as a derivatization reagent for establishing the stereochemistry of the bis-amino acids lanthionine and methyllanthionine.We used two sets of plasmids that we deposited in a publicly accessible plasmid collection to make two samples that contain Lan and MeLan with all five previously reported configurations.Using these standards, we report the elution order of all isomers bisderivatized with L-FDLA and D-FDLA using LC-MS.We show that as little as 0.05 mg of a lanthipeptide with its leader peptide still attached is sufficient for determining the stereochemistry of its Lan and MeLan, which is a considerable improvement over the GC-MS method that we and others have used.This approach is advantageous for cases in which limited material is available and for laboratories that have molecular biology expertise but not synthetic chemistry capabilities.As such, this work provides a method that is complementary to GC-MS with synthetic standards.

Figure 1 .
Figure 1.(A) Lanthipeptide synthases catalyze the formation of lanthionines and methyllanthionines by dehydrating serine/threonine residues to generate dehydroalanine (Dha) and dehydrobutyrine (Dhb), followed by the thiol−Michael addition of L-cysteine onto the Dha/Dhb moieties.(B) Theoretical diastereomers of two lanthionines (Lan) and four methyllanthionines (MeLan).Five bolded diastereomers have been detected in lanthipeptides; L-allo-L-MeLan has thus far not been reported in natural lanthipeptides.(C) Structures of L- FDAA and L-FDLA.

Figure 5 .
Figure 5. (A) buv BGC.The producing organism, gene diagram, and precursor peptide sequence with the predicted structure based on the result of stereochemical and LC-MS/MS analysis are listed.Leader peptide with a 6× His-tag sequence: MHHHHHHAMTIKDFYDKAAGDDAMKK-KIAEMTKAGKSLEDFIAEFGIEGTVDDLKAYAETLSQEGKLDKSQL.(B) L-FDLA derivatization of hydrolyzed mBuvA using the standard treatment procedure used herein.(C) Tandem ESI-MS of mBuvA after proteolysis with AspN (resulting in the modified peptide shown in bold font in panel (A)).For theoretical and observed m/z, see TableS1.