LC-MS and Peptide Purity: What the Analytical Literature Reports

Electrospray ionization mass spectrometry has been described in the methods literature as a rapid approach for verifying that a synthetic peptide has the expected mass and for identifying most synthesis by-products, with mass analysis used at three stages: confirming the target product in the crude mixture, guiding purification, and confirming the mass and purity of the isolated final product [1]. That review noted that the liquid-flow characteristics of the electrospray source are a practical advantage when LC-MS is required, and that fragment ions generated on relatively simple instruments can supply structural information about synthetic peptides [1]. A recurring theme in the chromatographic literature is that mass detection alone does not resolve every impurity class. In a two-dimensional LC-MS study of pharmaceutical peptides, the authors emphasised that compounds sharing a mass-to-charge ratio — such as peptide isomers and diastereomers — are not readily differentiated by mass spectrometry and must therefore be separated chromatographically [2]. Thirty column and mobile-phase combinations were evaluated using forcibly degraded peptide samples and synthetic diastereomer mixtures, and ranking of more than 300 UV and MS chromatograms led the authors to propose that screening four columns and four volatile mobile phases is adequate to cover the selectivity space [2]. For second-dimension separation of peptide isomers below roughly 10 kDa, a C8/C18 column without ionic functionality and an acetic acid/ammonium acetate mobile phase buffered at pH 5 gave good selectivity at 25 °C in that dataset, while trifluoroacetic acid was found to reduce selectivity differences between columns, likely through ion pairing [2]. A separate methods-development study combined automated LC screening instrumentation with in silico retention modelling to build MS-compatible purity and purification methods for peptides, proteins, and small molecules, reporting agreement between experimental and predicted retention times within 7% [3]. Sample handling upstream of the instrument has also been treated as a purity variable. A bottom-up proteomics protocol noted that the sensitivity and resolving power of current instruments allow detection of peptides at low femtomole quantities, but that this performance demands high sample purity to limit artefacts, and described protein extraction, proteolysis, and C18 polarity-based fractionation at the peptide level ahead of nanoscale LC and quadrupole-Orbitrap analysis [4]. Several metrology studies have addressed the harder question of assigning a numerical purity value to a peptide reference material. Work on a candidate angiotensin II certified reference material combined isotope dilution LC-MS/MS amino acid analysis after hydrolysis, quantitative 1H-NMR on the intact peptide, and a mass balance approach, using a validated 19F-qNMR measurement of the trifluoroacetic acid counter-ion, which accounted for nearly 25% of the material by mass [5]. A Bayesian treatment combining the corrected results produced a final assigned purity of 691 ± 9 mg/g (k = 2) [5]. Another group evaluated several isotope dilution strategies for naturally abundant and 13C-labelled angiotensin I standards, using microwave-assisted hydrolysis completed in 150 minutes and validating the procedure against certified peptide reference materials by both LC-MS/MS and GC-MS/MS [6]. For parathyroid hormone in a high-purity material, LC-MS was first used to confirm molecular weight and amino acid sequence, after which amino acid isotope dilution mass spectrometry and peptide-based isotope dilution using the tryptic peptide ADVNVLTK gave concordant results, with a reported mass fraction of 0.718 ± 0.022 g/g [7]. Orthogonal techniques have been benchmarked against mass spectrometric purity values. Gas chromatography–isotope dilution infrared spectrometry applied to [Glu1]-fibrinopeptide B returned an average purity of 0.755 ± 0.017 g/g across six sub-samples, in agreement with 0.754 ± 0.012 g/g obtained by isotope dilution mass spectrometry, with repeatability of 2.2% versus 1.7%, though the infrared route had higher limits of detection and quantitation [8]. A related gas chromatography–isotope dilution mass spectrometry study of the same model peptide identified and corrected for the primary impurity and reported a purity of 0.715 ± 0.012 g g-1 with 0.5% repeatability [9]. Work on hepcidin-25 illustrated a further limitation: reversed-phase separation of copper-bound and copper-free forms required a basic mobile phase containing 0.1% ammonia, and the authors concluded that metal complexes, isoforms, and isomers — the latter usually undetected by mass spectrometry — mean complementary methods such as qNMR, ICP-MS, ion-mobility spectrometry, and chiral amino acid analysis are needed to characterise a peptide calibrant comprehensively [10]. In biopharmaceutical quality control, LC-MS peptide mapping has been developed as a multi-attribute method intended to consolidate several conventional HPLC- and capillary-electrophoresis-based purity assays; a qualified new-peak-detection workflow validated according to ICH Q2 guidelines was reported to recognise relevant peptide species below 1% relative abundance without reporting false positive peaks in drug substance and drug product case studies [11]. An earlier LC-MS approach to a prototype bispecific antibody used intact deglycosylated mass measurement together with Lys-C peptide mapping, and detected spiked homodimer impurities at 2% and minor C-terminal truncation species down to 0.6% [12]. In an anti-VEGFR-2 monoclonal antibody, reduced CE-SDS showed heavy chain size heterogeneity that mass spectrometry indicated was not due to peptide bond cleavage; glycosylation profiling and glycosidase treatment attributed the heterogeneity to differing galactosylation levels [13]. Quantitative precision has been characterised for related peptide ratio measurements. Using six pairs of proteotypic peptides and same-sequence stable isotope labelled standards, MALDI-TOF measurements gave an average coefficient of variation of 2.2% across a 100-fold dilution curve at 55 fmol applied per spot and 1.0% at near-equivalence, and the authors described antibody capture enrichment of a target peptide to near purity prior to analysis [14]. Detection of chemically modified peptides raises its own identification issues; a systematic study of collision-induced dissociation spectra of biotinylated peptides catalogued signature fragment ions for biotin, heavy biotin, and biotin-XX-phenol adducts, noting that biotin enables enrichment and purity not easily achieved with other reagents but that localisation of the modified residue requires understanding these fragmentation patterns [15]. Applied peptide studies routinely report LC-MS purity and identity data alongside their primary experiments. A study of collagen extracted from sea bass scales reported a citric-acid extraction purity of 92.35 ± 1.12% and used nano-LC-MS/MS to identify 19 collagen-derived peptides in the most active chromatographic fraction, including the sequence GPPGSPGLPGPPGPS [16]. In a preclinical imaging study, HPLC purity of two d-amino acid bombesin conjugates was reported as greater than 95% with LC-MS (ESI+) used to confirm identity, before evaluation in murine xenograft glioma models [17]. A review of peptides in diagnostics stated that assay specificity depends on the purity of the biomolecule used as a probe and listed LC-MS/MS among the non-imaging platforms used for direct analysis of biofluids [18]. Outside peptide chemistry proper, LC-MS/MS with a 34S-labelled sulfur source was used to establish regiochemistry and greater than 99% isotopic enrichment in a glutathione trisulfide preparation reported at greater than 95% purity [19], and LC-MS together with MALDI-ToF and NMR was used to characterise sequence-defined star-shaped macromolecules above 11 kDa in an approach the authors describe as translatable to peptides and peptoids [20].