Peptide Analysis: Zwitterionic Chiral Ion-Exchangers as Complementary Option to HILIC and to Reversed-Phase Chromatography

Photo Credit: Daniel Diaz Santana / EyeEm/Getty Images

Tong Zhang¹, Emilie Holder¹, Pilar Franco¹, Michael Lämmerhofer², Adrian Sievers-Engler², Heike Gerhardt², Harald Gross², and Wolfgang Lindner³, ¹Chiral Technologies Europe, Illkirch, France,²Institute of Pharmaceutical Sciences, University of Tübingen, Tübingen, Germany, ³Lindner Consulting GmbH, Klosterneuburg, Austria.

Therapeutic peptides represent one of the fastest growing segments in the pharmaceutical market. To bring these products to the market in a consistent manner, high quality is a major concern and requires stringent quality control (QC) methods. The quality of peptide therapeutics cannot be comprehensively tested by a single method. Hence, a number of tools are necessary to meet the goals of QC in therapeutic peptides. This article discusses the potential of zwitterionic chiral ion-exchangers to support peptide analysis and quality control as a flexible complementary tool to monitor the stereochemical integrity and chemical modifications.

Peptides occupy important positions in therapeutic research and in healthcare, food, and cosmetic industries, as well as in many other fields. Appropriate analytical methods are required to check the purities in peptide synthesis and meet the regulatory challenges in R&D, manufacturing, and QC of therapeutic peptides.

In practice, multiple analytical aspects have to be considered, such as the chemical and enantiomeric/stereoisomeric purities of starting amino acid and small peptide building blocks, the monitoring and control of stereochemistry and impurities of intermediate molecules from synthetic process, as well as the assessment of the structural integrity (amino acid composition, sequence, chirality) of final peptide products (1–3).

To determine the enantiomeric or stereoisomeric purities of raw materials or for the assignment of the absolute configuration of constituent entities of peptides or other related biopharmaceutical preparations, pre‑column derivatization of the hydrolysates, followed by gas chromatography–mass spectrometry (GC–MS) or liquid chromatography (LC) coupled to MS analysis using capillaries or particulate packed columns has been the standard practice for decades.

Nevertheless, direct chiral analysis of amino acid or peptide fragments with no pre-column derivatization represents a more convenient and straightforward approach in terms of sample treatment, method simplicity, and speed of analysis cycle. High performance liquid chromatography (HPLC), often combined to MS detection, is undoubtedly the most attractive technique for this purpose.

A series of chiral selectors or stationary phases have been proven to be valuable for direct enantiomer or stereoisomer analysis of amino acids and small peptides by chromatography. The most important chromatographic supports to be cited in this regard include the chiral ligand exchangers (4–6), chiral crown ethers (7–11), macrocyclic antibiotics (12–15), and zwitterionic chiral ion‑exchangers (16–17). These chiral supports run analysis by chromatography on the basis of various enantio-selective or stereo-selective retention mechanisms.

Cinchona alkaloid-derived zwitterionic chiral ion‑exchangers with cyclohexane sulphonic acid moiety (Figure 1[a]) have proved to be effective for the direct analysis of enantiomers and stereoisomers of a wide range of ampholytes such as amino acids (18–21) and peptides (22–24,37). The chiral-recognition ability of these zwitterionic chiral ion-exchanger columns to a great variety of amino acids (18–21, 24–32) and other racemic ionic compounds (33–35) has been investigated and demonstrated.

Mechanistically, three distinct application modes can be distinguished: (i) For chiral acidic analytes the zwitterionic chiral ion-exchanger CSPs (chiral stationary phases) act like an enantioselective anion-exchanger. Thereby, the zwitterionic chiral ion-exchanger columns are typically operated under weakly acidic (polar organic or hydro‑organic) mobile phase conditions. Under such conditions the quinuclidine moiety is positively charged and attracts electrostatically acidic analytes (anion‑exchange mode). H-bond mediated ionic interaction combined with additional H-bonds (at carbamate moiety), π-π-interactions (with quinoline), dipole–dipole, and van der Waals forces or steric interactions support enantioselective complexation of a wide variety of chiral acids covering essentially the same application spectrum as the corresponding chiral anion-exchangers (with tert‑butyl moiety instead of the cyclohexane sulphonic acid residue at the carbamate group). Moreover, the sulphonic acid moiety of the chiral selector plays the role of an intramolecular counterion, which is present in equimolar concentrations with respect to the fixed ion-exchanger moiety. It leads to faster elution and allows the counterion concentration in the mobile phase to be reduced significantly, which is favourable for electrospray ionization (ESI) in LC–MS(–MS) applications; (ii) Vice versa, chiral basic compounds primarily interact at the chiral sulphonic acid moiety following the principles of cation-exchange (cation-exchange mode). In this case, the positively charged quinuclidinium represents the intramolecular counterion, which leads to accelerated separations compared to corresponding chiral cation-exchangers; (iii) For zwitterionic solutes, double ion-pairing at both cation- and anion-exchange sites may occur simultaneously (zwitterionic ion-exchange mode), giving zwitterionic‑type CSPs their unique character (see Figure 1[b]). Depending on which of the two domains provides higher affinity the analyte will primarily bind to this site with high strength while the other site may contribute by long-range electrostatic interactions. In any case, increased ionic strength (that is, higher acid/base additive concentrations) of the eluent leads to ionic shielding of charged groups and stronger analyte displacement from the ion-exchanger site, respectively, consequently decreasing retention.

In the current study, their performance in direct stereo‑selective separation of small peptides is further explored using LC–MS-compatible mobile phases. The second part of the report focuses on a case study where the zwitterionic chiral ion-exchanger columns investigated are deeply involved in the elucidation of the structural constitution and the stereochemistry of a lipopeptide, as well as to separate the peptide fragments generated from the linearized and the digested lipopeptide. In this case, the application of the zwitterionic chiral ion-exchanger columns is extended from analysis of enantiomers and stereoisomers of the hydrolysate amino acids, over determination of absolute configuration of the hydrolyzed fatty acid side chain, to the orthogonal properties of the zwitterionic chiral selectors with regard to other chromatographic modes such as hydrophilic interaction liquid chromatography (HILIC) and reversed-phase chromatography.

Experimental

Chemicals: Mobile phases for chromatography were prepared from HPLC-grade solvents. Methanol, acetonitrile, tetrahydrofuran (THF), and water were purchased from Carlo Erba Reagents. Formic acid (FA), diethylamine (DEA), ammonium hydroxide (NH4OH, 27–28%), and di- and tri-peptide samples were supplied by Sigma-Aldrich Chimie S.a.r.l. The compressed nitrogen (5.0, Messer France SAS) was used as the nebulizing gas for the evaporative light scattering detector (ELSD).

For Enantio- and Stereo-Selective Separation of the Peptides: The HPLC system used was an Agilent 1100 series apparatus optimized in terms of system void by using micro-flow cell (1.7 µL) and flow capillaries of 0.12‑mm i.d. An evaporative light scattering detector (ELSD 2000ES, Alltech) was hyphenated to a diode-array detector (Agilent) via an interface 35900E. The generic parameters of ELSD were: gas flow, 1.7 L/min; drift tube temperature, 70 °C; gain, 1; impactor, OFF.

The zwitterionic chiral ion-exchanger columns used for the study were 150 × 3 mm, 3-µm Chiralpak ZWIX(+) (based on trans-(1’’S,2’’S)-N-[[[(8S,9R)-6’-methoxycinchonan-9-yl]oxy]carbonyl]-2’’-aminocyclohexanesulphonic acid; quinine-derived) and 150 × 3 mm, 3-µm Chiralpak ZWIX(-) (based on trans-(1’’R,2’’R)-N-[[[(8R,9S)-6’-methoxycinchonan-9-yl]oxy]carbonyl]-2’’-aminocyclohexanesulphonic acid; quinidine‑derived) (Chiral Technologies Europe) (Figure 1[a]). The chemistries of these zwitterionic chiral ion-exchange columns will be referrred to genericaly as ZWIX(+) and ZWIX(–)] in this article. (The typical mobile phases used in the study were 50 mM FA + 25 mM DEA as additives in methanol/acetonitrile/H2O 49:49:2 v/v/v (MP‑I), methanol/THF/H2O 49:49:2 v/v/v (MP-II), methanol/H2O 98:2 v/v (MP-III), and methanol/H2O 90:10 v/v (MP-IV). The flow rate was set at 0.5 mL/min.

For Lipopeptide Analysis: Absolute configurations of amino acid constituents of the lipopeptide (after hydrolysis by 6N DCl in D2O for 24 h at 110 °C and evaporation) were assigned by enantioselective HPLC with a 150 × 4 mm, 3-µm ZWIX(+) column, methanol/H2O (98:2 v/v) containing 9.4 mM ammonium formate and 9.4 mM formic acid as mobile phase, flow rate of 0.7 mL/min and ESI with quadrupole time-of-flight (QTOF)-MS detection. For absolute configuration determination of the 3-hydroxy fatty acid side chain, this lipid was extracted by liquid–liquid extraction (LLE) with chloroform–water (1:1 v/v). The chloroform layer was evaporated and analyzed by 150 × 4 mm, 3-µm ZWIX(+) and 150 × 4 mm, 3-µm ZWIX(-) columns at 10 °C column temperature, acetonitrile/methanol/acetic acid (95/5/0.025 v/v/v) as mobile phase, flow rate of 0.3 mL/min, and MS detection.

Separations of cyclic and linear lipopeptide forms were performed using i) reversed-phase LC using octadecylsilica with a 100 × 2.1 mm, 5-µm ODS-Hypersil (Agilent) column using water/acetonitrile (40:60 v/v) with 0.1% (v/v) formic acid, ii) HILIC on a sulphobetaine stationary phase with a 150 × 2 mm, 3.5-µm SeQuant ZIC‑HILIC (Merck Millipore) column using water/acetonitrile (5:95 v/v) with 0.1% (v/v) formic acid, and iii) a 150 × 4 mm, 3-μm ZWIX(+) column with water/acetonitrile (65:35 v/v) containing 0.1% (v/v) formic acid. ESI-QTOF-MS was used as detector.

Complementarity plots were generated on chemical digests of the lipopeptide using reversed-phase LC–MS on a 100 × 3 mm, 2.6-µm core–shell octadecylsilica Kinetex C18 column (Phenomenex) with gradient elution using water (A) and acetonitrile (B) with 0.1% (v/v) formic acid and the following gradient profile: 20% B from 0–2.5 min, 20–64% B from 2.5–12.5 min, 64% B from 12.5–15 min, 64–80% B from 15–17 min, 80% B from 17–20 min, and 20% B from 20.1–23 min. Furthermore, the HILIC method made use of the sulphoalkylbetaine ZIC-HILIC column and the same eluents but inverted gradient: 80% B from 0–2.5 min, 80–36% B from 2.5–12.5 min, 36% B from 12.5–15 min, 36–20% B from 15–17 min, 20% B from 17–20 min, and 80% B from 20.1–23 min. Analysis for complementarity assessment in polar organic mode with the ZWIX(+) column was conducted in isocratic mode with methanol/acetonitrile/water (49:49:2 v/v/v) containing 25 mM ammonium formate and 25 mM formic acid.

Detection in LC–MS experiments was performed on a TripleTOF 5600+ (Sciex) QTOF MS instrument coupled via Duospray Ion Source (Sciex) and operated in ESI mode (curtain gas 30 psi, nebulizer and drying gas 60 psi, source temperature 400 °C) to an 1290 series UHPLC pump (Agilent) and column thermostat equipped with a CTC-PAL HTS autosampler (CTC Analytics). For amino acid analysis (free and derivatized with Sanger’s reagent) as well as peptide separations to elucidate complementarity profiles on a ZWIX(+) column in ESI(-) mode was used with an ion-source floating voltage of -4500 V, and declustering potential -100 V. Acquisition was performed in product ion-high sensitivity mode with -20 V collision energy for enhanced sensitivity and to generate confirmation fragment spectra. Peptide separations by reversed-phase LC and HILIC were performed in positive ESI(+) mode, ion-source floating voltage 5500 V, and declustering potential 100 V. Acquisition was performed as scheduled targeted product ion scans with collision energy of 25 V for generating confirmation fragment spectra.

Further details on experimental conditions can be found in reference (36).

Results and Discussion

Two zwitterionic chiral ion-exchange CSPs were investigated for this study. Chemically, they consist of cinchona alkaloid-derived quinine carbamate with sulphonic acid moiety (S,S)-trans-2-aminocyclohexanesulphonic acid, ZWIX(+), and cinchona alkaloid-derived quinidine carbamate with (R,R)-trans-2-aminocyclohexanesulphonic acid, ZWIX(-), (18) (see Figure 1[a]). Stereochemically, these two chiral selectors are diastereoisomers but behave most frequently as pseudo-enantiomers. Such an idiosyncrasy affords the possibility and convenience of reversing elution order of enantiomers by switching the ZWIX column from one to the other (24). The chiral recognition and stereo‑selective separation of ampholytic analytes such as free amino acids and peptides is primarily based on the synergistic double ion-pairing between the zwitterionic chiral ion-exchange selector and the zwitterionic analytes (zwitterionic ion‑exchange mode) (Figure 1[b]) and assisted by other weaker interactions such as hydrogen bonding, π-π stacking, and van der Waals forces, as outlined above. On account of the zwitterionic ion‑exchange mechanism, the co-presence of acidic and basic additives in an appropriate ratio is necessary to regulate the ionic interactions via displacement effects (16–17). The methanol-based mobile phase is recommended as the first choice for separations on the zwitterionic chiral ion-exchangers. The most useful mobile phase modifiers include water, acetonitrile, or tetrahydrofuran (THF). Among these solvents, methanol and water offer the most suitable solvation ability to all the ionized species involved in the ion-exchange equilibria. Exhibiting the strongest elution power, water is normally used at a low percentage (≤20% in volume). In contrast, the aprotic solvents acetonitrile and THF are weak eluting mobile phase components and are usually used up to 50%. Their presence in the mobile phase can efficiently contribute to the retention adjustment of fast-eluting analytes (21).

Stereo-Selective Separation of Di- and Tri-peptides: The experimental scheme or approach is essentially the same for amino acids and peptides on the zwitterionic chiral ion-exchanger columns investigated. For both a zwitterionic ion pairing/ion-exchange mode drives the retention and separation. Both the length of peptide, that is, the number of amino acid residues in the peptide chain, and the side chains modulate enantioselectivity. The first trials of separating enantiomers and stereoisomers of small peptides were attempted with easily accessible commercial small peptide standard samples. In total, 13 common di-peptides and six typical tri-peptides are involved in the first part of this study.

The experimental results are summarized in Table 1 for enantiomers of di- or tri-peptides bearing a single stereogenic centre. As indicated by the selectivity data (), each pair of enantiomers could be recognized on the ZWIX(+) column, with the highest resolution degree for Gly-dl-Trp and Gly-dl-Ser. It was observed that, even with extensive optimization of the chromatographic conditions, the achievement of full enantiomer resolution of dl-Ser proved to be challenging on these CSPs. With a glycine component attached to the N-terminus of serine (Gly‑dl‑Ser), however, a large and effortless separation of the enantiomers could be obtained. It should be noted that this does not represent a generic scenario and the enantio- or stereo-selectivity is highly structure-dependent.

When it comes to the peptides containing more than one stereogenic centre, the peak configuration in a single analysis becomes more complex. In this case, it would be of high importance to monitor the peak elution order while optimizing the enantio- and stereo-selectivities as well as the resolution degrees between the peaks. Unfortunately no information on elution order could be collected in our study because of the unavailability of enantiomer or diastereoisomer standards.

Again, the challenging configuration of multi-peaks in a single and short chromatographic analysis would require enhanced performance of the column. For the given chiral selector bonded on to the spherical silica matrix with well‑defined particle size, pore, and surface properties, the most viable options would be the variation in composition of bulk mobile phase, the reduction of column temperature, and the use of a longer column. While the first two approaches could be adopted in attempts to enlarge the selectivities (therefore the resolution degree), the third alternative would normally be more forthright by offering higher column efficiency without changing the capacity factors.

For separation of diastereoisomers of the di- or tri‑peptides containing two stereogenic centres (see Table 2), most of the data were acquired at 10 °C and represented an effective improvement in resolution degree of the adjacent peaks in regard to the conventional temperature at 25 °C. The effect of mobile phase on the resolution of dl-Leu-dl-Val is shown in Figure 2(e). Peak co-elution occurred while methanol/acetonitrile/H2O 49:49:2 (MP-I in [e-1]) or methanol/H2O 98:2 (MP-III in [e-3]) was in use as the bulk mobile phase. A full separation of the four stereoisomers could be achieved by simple replacement of acetonitrile in MP-I with THF (MP-II in [e-2]). The performance of the zwitterionic chiral ion-exchanger columns investigated can be further demonstrated with successful stereoisomeric separations under optimized conditions. For instance, the combination of low temperature and longer columns led to satisfactory separation of four stereoisomers for (a) dl-Leu-dl-Tyr, (b) dl-Ala-dl-Leu, and (d) Gly-dl-Leu-dl-Ala (Figure 2). As far as dl-Ala-dl-Leu-Gly is concerned, the complete resolution of the four stereoisomers could hardly be achieved on the ZWIX(+) column even with extensive optimization of the chromatographic conditions.

However, complete resolutions could be obtained with a ZWIX(-) column of the same column size (Figure 2[c]). This specifies the complementarity properties between ZWIX(+) and ZWIX(-) columns in terms of stereoselective recognition performance and resolution power.

Lipopeptide Analysis: The great utility of the cinchonan‑based zwitterionic chiral ion-exchanger columns investigated as complementary tools for peptide separation and characterization in quality control and drug discovery, taking benefit from its distinct selectivity profiles ranging from free and derivatized amino acid enantiomers, peptide enantiomer, epimers, diastereomers, and (minor) chemical modifications in peptides and peptide therapeutics, is illustrated in the following practical example from pharmaceutical biology (36). In this study ZWIX(+) was used as a basic tool for comprehensive structural elucidation of a lipopeptide (Figure 3[a]) isolated from the endophytic Pseudomonas poae strain RE*1-1-14, being supported by reversed-phase LC and HILIC. The cyclic lipopeptide poaeamide A constituted by 10 amino acids showed some bioactivity in terms of growth inhibition of the fungal pathogen Rhizoctonia solani. For comprehensive structural elucidation the lipopeptide was analyzed on three structural levels: in intact form, in fully hydrolyzed form, and at intermediate level in digested form (obtained by enzymatic and/or chemical digestion).

Amino acid sequence and constitution of fatty acid side chain were readily determined on the intact peptide by nuclear magnetic resonance (NMR) and MS after purification by reversed-phase LC. However, several structural features related to stereochemistry required careful elucidation by stereoselective methods. In a first step, the peptide was fully hydrolyzed by DCl/D2O to obtain free amino acid and 3-hydroxydecanoic acid constituents.

Absolute configurations of amino acids were determined by enantioselective chromatography on ZWIX(+) with mobile phases as specified in the experimental section (36). A hydrolyzed lipopeptide sample was analyzed without derivatization (free amino acids) in zwitterionic ion-exchange mode and after derivatization with Sanger’s reagent, which yielded 2,4-dinitrophenyl (DNP) derivatives of amino acids in anion-exchange mode. This was necessary because the current lipopeptide contained all the most challenging amino acids from a stereoisomer separation point of view (as a result of side chain chirality) (Ile/Leu/a-Ile as well as Thr/a-Thr). Chromatographic results for amino acids present in the lipopeptide sample are summarized in Table 3. Comparison of retention factors of amino acid standards and amino acids from the peptide sample revealed that the lipopeptide was constituted by d-Ser (determined as DNP-derivative), d-Glu (free, DNP), d-a-Thr (DNP), l-Leu (DNP), d-Leu (free), and l-Ile (free). Several options for verification and validation of the results exist: Analysis on ZWIX(-) is first choice because elution orders of enantiomers are reversed. Furthermore, other N-derivatives may be used and give distinct elution orders, retention factors, and separation factors allowing the unequivocal identification of the stereochemistry (34), or analysis of N-derivatized forms on tert-butylcarbamoylquinine or quinidine-based chiral anion exchangers might be considered as an appropriate choice for confirmation of results. In our example, we have also used GC–MS with a polysiloxane modified with N-(2-methylpropanoyl)-valyl-tert-butylamide column as the complementary analysis method, revealing that the stereochemistry was correctly identified (36).

A significant advantage of the zwitterionic chiral ion-exchanger columns investigated in this study was their broad applicability profile for various classes of compounds, ranging from free amino acids and derivatized amino acids over peptide stereoisomers, to chiral carboxylic acids. It is a peculiar feature of Pseudomonas lipopeptides that they contain a fatty acid side chain with at least one stereogenic centre. In the investigated lipopeptide it was identified as 3-hydroxydecanoic acid (36). Prior studies with racemic mixture showed that the enantiomers of 3-hydroxydecanoic acids can be resolved on ZWIX(+) in the anion-exchange mode when a mobile phase consisting of acetonitrile/methanol/acetic acid (95/5/0.025; v/v/v) was used and column temperature was 10 °C (36). Retention times on ZWIX(+) were 13.48 min for the first eluted enantiomer and 14.67 min for second eluted enantiomer, as well as 15.46 and 16.76 min on ZWIX(-). An R-enantiomer standard of 3-hydroxydecanoic acid isolated from rhamnolipid allowed the elution order to be pinpointed (R before S on ZWIX[+] and S before R on ZWIX[-]). Consequently, this enantioselective HPLC method was a suitable means to identify the stereochemistry in the fatty acid side chain that normally remains undetermined in research on lipopeptides. The chloroform extract was injected with and without a spike of racemic mixture, and it turned out that the absolute configuration of 3-hydroxydecanoic acid in poaeamide has R-configuration (36).

Absolute configurations of all constituents were determined at this point, however, the full structure was not yet identified. Positions of ring closure and the location of d-Leu in the peptide sequence were still unclear. A carboxylic acid derivatization procedure (using EDC and an amine label) gave rise to the conclusion that the C-terminus must be involved in the ring closure because the carboxylic acid moiety of the d-Glu residue was found by MS to be chemically modified by this derivatization step. On the other hand, a side product in the alkaline hydrolysis of the cyclic ester indicated that the a-Thr residue represented the alcoholic component for ester formation and ring closure, respectively (36).

A bit more challenging was to pinpoint the position of the d-Leu residue in the peptide sequence. One out of five Leu residues had d-configuration. Unfortunately, sequence information is lost as a result of hydrolysis of the lipopeptide sample prior to amino acid configuration analysis. Therefore, the determination of the position of the d-Leu amino acid residue in the peptide chain needs other strategies to fix this problem. If authentic standards of each possible stereoisomer were available, their analysis on the investigated zwitterionic chiral ion-exchanger columns had a great opportunity to reveal distinct retention times as a result of the good diastereomer selectivity of this column, as proven with several peptides. Unfortunately, authentic standards of each possible stereoisomer of this lipopeptide comprised of non-natural amino acids with lipid modification were too expensive and therefore a strategy to digest the lipopetide (enzymatic and chemical digestion) to produce smaller fragments followed by subsequent hydrolysis of isolated fractions and enantioselective analysis at the amino acid level was envisioned to solve the stereochemistry of poaeamide.

For enzymatic digestion with pepsin the cyclic lipopeptide was first hydrolyzed with 1N NaOH to ensure better digestion efficiency of the enzyme (36). Control of the reaction product of hydrolysis revealed another interesting feature of the zwitterionic chiral ion-exchanger columns studied. It appears to be sensitive for minor molecular changes (small structural modifications) in peptides, which are accompanied with changes in ionization state and effective charge, respectively. While the intact cyclic lipopeptide carries a single negatively charged functionality (in the form of a Glu side chain close to the N-terminal end) featuring an anion-exchange mode on zwitterionic chiral ion-exchanger column, it becomes two-fold negatively charged after hydrolysis in the linear form. Since the Gibbs free energy of binding is proportionally decreasing and the log k-values are directly proportionally increasing with the effective charge number of the ionic analyte, retention of the hydrolyzed linear form is significantly different in the prevailing anion-exchange mode on the zwitterionic chiral ion-exchanger column. It becomes evident from the chromatograms in Figure 3 that the zwitterionic chiral ion-exchanger column investigated exhibits better selectivity for distinction between the cyclic and linearized forms of the lipopeptide. While cyclic and linear forms essentially co-eluted both in reversed-phase LC (Figure 3[b]) and HILIC (Figure 3[c]), the zwitterionic quinine carbamate selector selectively recognized the hydrolyzed form. It was more strongly retained by additional interaction of the carboxylic acid C-terminus, which was available in free form only in the linear lipopeptide (Figure 3[d]). Similar situations may often exist in therapeutic peptides and their impurities in which such chemical modifications may occur because of degradation. Zwitterionic chiral ion-exchanger columns might therefore be valuable complementary tools for impurity profiling studies of therapeutic peptides.

For locating d-Leu within the peptide chain, the lipopeptide was digested under controlled conditions to small peptide fragments that contain only one or a few Leu residue(s). The peptide fragments were then separated by reversed-phase LC and individual peptide fragments containing a Leu residue were isolated micro-preparatively. Hydrolysis and subsequent amino acid enantiomer analysis of chromatographically isolated single peptide fragments allowed the stereochemistry of the peptide to be put together like a puzzle (36).

Pepsin is an endopeptidase that cleaves amide bonds preferentially at the carboxylic side of Phe, Leu, and partly Glu. D-configuration would block enzyme action. The linearized peptide was digested with pepsin at 37 °C for 90 min and subsequently Leu-containing peptides isolated by reversed-phase LC. An early eluted peak corresponding to LSL, SLL, or LSI tripeptides contained only l-Leu-, but no d-Leu-configurated peptides. Only 3 peaks, all with the peptide fragment 3-hydroxydecanoyl-(3-HDA)-LETLL (HDA-LETLL, HDA-LETLLSL, HDA-LETLLSLL), contained d-Leu. This allowed the possibilities for d-Leu positions to be narrowed down to the two Leu residues from the amino terminal side. Controlled chemical digestion (hydrolysis with 6N HCl at 110 °C for 20 min) produced several smaller peptides. Some potentially diagnostic peptides in terms of Leu stereochemistry were again collected by reversed-phase LC. Their amino acid enantiomer analysis revealed that the Leu residue of 3-HDA-L and 3-HDA-LE had L-configuration. In contrast, d-Leu was detected in the isolated peptides LETL and TLL. Consequently, the d-Leu residue was confirmed to be located in residue 4 from the N-terminus of the peptide.

Reversed-phase LC and HILIC showed some shortcomings in the above peptide separations, in particular for small peptides. Figure 3 has already demonstrated that ZWIX columns can generate complementary retention profiles to reversed-phase LC and HILIC in peptide separations. To demonstrate this more systematically on a wider scope, chemically digested poaeamide was analyzed by reversed-phase LC (on octadecylsilica core–shell column), HILIC (on silica-based sulphobetaine column), and chiral HPLC in polar organic mode (on ZWIX[+]) (36). Orthogonality plots in Figure 4 convincingly document evenly distributed data points in the two-dimensional design space (tR of peptide fragments on ZWIX[+] vs. reversed phase in Figure 4[a] and silica-based sulphobetaine column vs. the ZWIX[+] column in Figure 4[b]) and thus prove excellent complementarity of these chromatographic modes. A number of small peptides are well retained and resolved on the zwitterionic chiral ion-exchanger columns but co-eluted close to t0 in reversed-phase LC (Figure 4[a], indicated by dotted line). On top of this complementary retention profile of the zwitterionic chiral ion-exchanger, it provides additional selectivity for peptide stereoisomers (enantiomers, epimers, and diastereomers), which either do not exist for achiral columns, may be greatly enhanced, or be complementary and confirmatory on the zwitterionic chiral ion-exchanger phase. Furthermore, degradation and side product formation in therapeutic peptides may come along with alterations in the charge state of the peptide; for example, as a result of side chain deamidation (AsnAsp, GlnGlu), C-terminal deamidation, C-term amidation (COOHCONH2), isomerization (AspIsoAsp), and succinimide formation (Asn/AspSuc), which may be readily resolved by a zwitterionic chiral ion-exchanger column while it may be more challenging to resolve such structural modifications with common reversed-phase LC. Zwitterionic chiral ion-exchangers may therefore be regarded as a valuable tool for peptide separations and complement state-of-the-art reversed-phase LC and HILIC. Furthermore, its integration as one orthogonal dimension in 2D LC peptide separations could be another application of great interest and future potential.

Conclusions

Therapeutic peptides are a growing segment in pharmaceutical markets. In order to ensure the quality of such products, a large number of critical quality attributes have to be tested and controlled during production and release. Besides chemical integrity, stereochemical integrity is of central importance and its analytical quality assurance testing is challenging and not a trivial task. A common strategy for stereochemical impurity testing of peptides is to fully hydrolyze the sample and perform enantioselective amino acid analysis. GC–MS on a polysiloxane modified with N-(2-methylpropanoyl)-valyl-tert-butylamide column is one of the established and well accepted protocols for this purpose. Yet, this strategy also has some shortcomings. There are some critical proteinogenic-type amino acids like Cys, His, Asn, and Gln. Asn/Gln are hydrolyzed in the side chain and cannot be distinguished from corresponding acids Asp/Glu if they co-exist in the peptide. Cys and His (partly also Arg) are problematic in GC–MS amino acid analysis. Unfortunately, cysteine, histidine, and aspartic acid were reported to be susceptible to racemization in peptide synthesis and important amino acids to be controlled.

Furthermore, if individual amino acids are present more than once and are found to be present in d- and l-configuration, the correct stereochemistry of the peptide, be it the therapeutic peptide or an impurity, cannot be determined on the amino acid level. Hence, alternative support methods are required. We suggest performing peptide analysis systematically on three levels of structural integrity, namely the intact molecule, its fully hydrolyzed form (amino acid level), and on the intermediate level (partially digested form). On all levels, three complementary LC methods - reversed-phase LC, HILIC, and enantioselective or stereoselective HPLC with a zwitterionic chiral ion-exchanger column - should be used for a comprehensive structural elucidation, ideally in combination with complementary detection such as UV and MS. This would provide the maximum amount of information, allowing the chemical and stereochemical structure of the analyzed peptides to be pieced together. Although the zwitterionic chiral ion-exchanger columns studied were initially designed for enantiomer separation of very diversely structured natural and unnatural -amino acid but also for -and -amino acids, it represents a valuable tool in many other applications as well, in particular peptide separations. Zwitterionic chiral ion‑exchanger columns can provide enantioselectivity, epimer selectivity, diastereomer selectivity for peptides, enhanced selectivity for chemical modifications that are accompanied by changes in charge state, and last but not least it can be utilized for raw materials quality control (usually FMOC amino acids but also free -, -, -amino acids, and other amino acid derivatives).

Acknowledgements

M.L. acknowledges the support by the “Struktur- und Innovationsfonds Baden-Württemberg (SI-BW)” and by the German Research Foundation DFG for funding scientific equipment as part of the DFG’s Major Research Instrumentation Programme as per Art.91b GG (INST 37/821-1 FUGG).

References

1. I. Eggen, B. Gregg, H. Rode, A. Swietlow M. Verlander, and A. Szajek, Pharmaceutical Technology38(3), 27–4 (2014).
2. I. Eggen, B. Gregg, H. Rode, A. Swietlow M. Verlander, A. Szajek, BioPharm International38(4), 27–4 (2014).
3. B. Gregg, A. Swietlow, A.Y. Szajek, H. Rode, M. Verlander, and I. Eggen, BioPharm International38(5), (2014).
4. V.A. Davankov and S.V.Rogozhin, J. Chromatogr.60, 280–283 (1971).
5. M.H. Hyun, S.C. Han, C.W.Lee, and Y.K. Lee, J. Chromatogr. A950, 55–63 (2002).
6. B. Natalini, R. Sardella, A. Macchiarulo, and R. Pellicciari, J. Chromatogr. B405, 145–153 (2008).
7. T. Shinbo, T. Yamaguchi, K. Nishimura, and M. Sugiura, J. Chromatogr.405, 145–153 (1987).
8. A. Péter, G. Török, P. Csomós, M. Péter, G. Bernáth, and F. Fülöp, J. Chromatogr. A761, 103–113 (1997).
9. M.H. Hyun, J.S. Jin, and W. Lee, J. Chromatogr. A822, 155–161 (1998).
10. Y. Machida, H. Nishi, K. Nakamura, H. Nakai, and T. Sato, J. Chromatogr. A805, 85–92 (1998).
11. H.J. Choi and M.H. Hyun, J. Liq. Chrom. Rel. Tech.30, 853–875 (2007).
12. D.W. Armstrong, Y. Lui, and K.H. Ekborgott, Chirality7, 474–797 (1995).
13. A. Berthod, Y. Lui, C. Bagwill, and D.W. Armstrong, J. Chromatogr. A731, 123–137 (1996).
14. A. Péter, G. Török, and D.W. Armstrong, J. Chromatogr. A793, 283–296 (1998).
15. A. Péter, G. Török, D.W. Armstrong, G. Tóth, and D. Tourwé, J. Chromatogr. A904, 1–15 (2000).
16. C.V. Hoffmann, R. Pell, M. Lämmerhofer, and W. Lindner, Anal. Chem.80, 8780–8789 (2008).
17. C.V. Hoffmann, R. Reischl, N.M. Maier, M. Lämmerhofer, and W. Lindner, J. Chromatogr. A1216, 1157–1166 (2009).
18. T. Zhang, E. Holder, P. Franco, and W. Lindner, J. Chromatogr. A1355, 191–199 (2014).
19. M. Lämmerhofer, Anal. Bioanal. Chem.406, 6095–6103 (2014).
20. T. Fukushima, A. Sugiura. I. Furuta, S. Iwasa, H. Lizuka, H. Ichiba, M. Onozato, H. Hikawa, and Y. Yokoyam, Int. J. Trp. Res.8, 1–5 (2015).
21. I. Ilisz, Z. Gecse, Z. Pataj, F. Fülöp, C. Toth, W. Lindner, and A. Péter, J. Chromatogr. A1363, 169–177 (2014).
22. S. Wernisch and W. Lindner, J. Chromatogr. A1269, 279–307 (2012).
23. R.J. Reischl and W. Lindner, J. Pharm. Biomed. Anal.116, 123–130 (2015).
24. T. Zhang, E. Holder, P. Franco, and W. Lindner, J. Sep. Sci.37, 1237–1247 (2014).
25. I. Ilisz, N. Grecso, M. Palko, F. Fülöp, W. Lindner, and A. Péter, J. Pharm. Biochem. Anal.98, 130–139 (2014).
26. Z. Pataj,I. Ilisz,Z. Gecse, Z. Szakonyi, F. Fülöp, W. Lindner, and A. Péter, J. Sep. Sci.37, 1075–1082 (2014).
27. I. Ilisz, Z. Pataj, Z. Gecse, Z. Szakonyi, F. Fülöp, W. Lindner, and A. Péter, Chirality26, 385–393 (2014).
28. I. Ilisz, N. Grecso, A. Aranyi, P. Suchotin, D. Tymecka, B. Wilenska, A. Misicka, F. Fülöp, W. Lindner, and A. Péter, J. Chromatogr. A1334, 44–54 (2014).
29. I. Ilisz, N. Grecso, A. Misicka, D. Tymecka, L. Lazar, W. Lindner, and A. Péter, Molecules20, 70–87 (2015).
30. I. Ilisz, Z. Gecse, G. Lajko, M. Nonn, F. Fülöp, W. Lindner, and A. Péter, J. Chromatogr. A1384, 67–75 (2015).
31. I. Ilisz, Z. Gecse, G. Lajko, E. Forro, F. Fülöp, W. Lindner, and A. Péter, Chirality27, 563–570 (2015).
32. I. Ilisz, N. Grecso, R. Papousek, Z. Pataj, P. Bartak, L. Lazar, F. Fülöp, W. Lindner, and A. Péter, Amino Acids47, 2279–2291 (2015).
33. M.C.K. Geditz, W. Lindner, M. Lämmerhofer, G. Heinkele, R. Kerb, M. Ramharter, M. Schwab, and U. Hofmann, J. Chromatogr. B968, 32–39 (2014).
34. F. Ianni, Z. Pataj, H. Gross, R. Sardella, B. Natalini, W. Lindner, and M. Lämmerhofer, J. Chromatogr. A1363, 101–108 (2014).
35. F. Ianni, A. Carotti, M. Marinozzi, G. Marcelli, A. Di Michele, R. Sardella, W. Lindner, and B. Natalini, Anal. Chim. Acta885, 174–182 (2015).
36. H. Gerhardt, A. Sievers-Engler, G. Jahanshah, Z. Pataj, F. Ianni, H. Gross, W. Lindner, and M. Lämmerhofer, J. Chromatogr. A1428, 280–291 (2016).
37. F. Ianni, R. Sardella, A. Carotti, B. Natalini, W. Lindner, and M. Lämmerhofer, Chirality28, 5–16 (2016).

Tong Zhang received her Ph.D. degree in organic chemistry from University of Bordeaux I (France) in 1992. After three years postdoctoral work at Novartis (Switzerland) under the supervision of Eric Francotte, she joined Chiral Technologies Europe (Strasbourg, France). She is now the R&D Manager in the same company.

Emilie Holder graduated from the Institute of Technology of Strasbourg (France) in 2007. As an organic chemist she worked in the R&D team of Sunovion Pharmaceuticals Inc. (Marlborough, MA, USA) and then joined an organic chemistry laboratory at University of Strasbourg. She is now an R&D assistant at Chiral Technologies Europe in Strasbourg, France.

Pilar Franco received her Ph.D. degree in pharmacy from the University of Barcelona, Spain. After two years postdoctoral work in the University of Vienna, Austria, under the supervision of Wolfgang Lindner, she joined Chiral Technologies Europe (Strasbourg, France). She is now Manager of Technical Operations in the same company.

Michael Laemmerhofer is professor for pharmaceutical (bio-)analysis at the Institute of Pharmaceutical Sciences, University of Tübingen in Germany. His research interests include surface functionalization of micro- and nanoparticulate materials for separation and sample preparation, chiral separation, impurity profiling, bioanalysis (metabolomics and lipidomics), and biopharmaceuticals analysis.

Adrian Sievers-Engler is a PhD student in the pharmaceutical (bio-)analysis group at the University of Tübingen, Germany. He graduated in biochemistry in 2009 from the University of Greifswald, Germany. His research interests include high resolution mass spectrometry and separation techniques in bioanalysis and pharmaceutical bioanalysis.

Heike Gerhardt studied chemistry at the universities of Tübingen and Vienna. She finished her Ph.D. in May 2015 at the University of Tübingen under the supervision of Michael Lämmerhofer working on analysis of insect secretion by GC–MS. Nowadays, she works as COO at the company C.A.T. GmbH & Co KG in Tübingen, Germany.

Harald Gross is professor for pharmaceutical biology at the Pharmaceutical Institute, University of Tübingen, Germany. His research interests include genome mining in microbial genomes for secondary metabolites, biosynthesis research, isolation, 2D-/3D-structure elucidation (NMR, MS), and mode of action investigations of natural products.

Wolfgang Lindner studied chemistry at the KFU-Graz, Austria, where he worked for over 20 years at the Institute of Pharmaceutical Chemistry. In 1996 he was appointed a Chair of Analytical Chemistry at the University of Vienna, Austria, and became Emeritus in 2012. Throughout his career he has been strongly influenced by life sciences themes spanning from pharmaceutical analysis, metabolomics, and proteomics, and separation science methods such as HPLC, SFC, GC, CE/CEC, and LC–MS.