Posttranslational-Modifications of Human-Serum-Albumin Analysis by a Top-Down Approach Validated by a Comprehensive Bottom-Up Analysis

Bibliographic info

  • Authors: Mohamad-Ali Rahali, Roy Lakis, François-Ludovic Sauvage, Emilie Pinault, Pierre Marquet, Franck Saint-Marcoux, Souleiman El Balkhi
  • Journal: Journal of Chromatography B (submitted 2022/2023; manuscript CHROMB-S-23-00284)
  • Institution: P&T UMR1248 INSERM / CHU Limoges

Key question

Do top-down (TD) LC-MS isoform identifications accurately reflect the true PTM status of HSA, and can a comprehensive bottom-up (BU) approach — including SWATH-MS library — confirm TD assignments and reveal additional modifications?

Methods

  • Sample type: Human serum — controls + liver dysfunction patients
  • Approach 1 (Top-down): LC-TOF-MS; multicharged spectrum → deconvolution → isoform identification by MW; trapezoidal integration for relative quantification
  • Approach 2 (Bottom-up):
    • Phase 1: Reduction ± alkylation → trypsin digestion → Q-TOF DDA → SWATH-MS spectral library (127 HSA peptides)
    • Phase 2: Same samples analyzed in DIA mode → confirms isoforms and PTM sites using the library
  • Cross-validation: compare TD isoform list against BU-confirmed modification sites

Main findings

  1. Top-down detected 15 isoforms: glycation, cysteinylation, nitrosylation, oxidation (di- and tri-)
  2. BU spectral library: 127 peptides; enabled site-level confirmation of isoforms and identification of additional PTMs
  3. BU-only PTMs: carbamylation, deamidation, amino acid substitution — not detected by TD due to resolution limitations at intact protein MW
  4. Critical validation result: all CLD-relevant HSA isoforms detected by TD were confirmed by BU — TD is not generating false positives for the clinically significant isoforms
  5. TD risks: some isoforms can be missed or misidentified by TD (overlapping masses), particularly low-abundance or closely-spaced modifications
  6. Reproducibility: intra- and inter-assay CV < 15% for all isoforms present at relative abundance > 2%
  7. Speed: 3-minute analysis runtime

PTMs characterized

PTMMethodModified site(s)Clinical relevance
Cysteinylation (+119 Da)TD + BUCys34Oxidative stress
Glycation (+162 Da)TD + BULys (multiple)Metabolic stress
Oxidation (di/tri, +32/+48 Da)TD + BUCys34Severe oxidation
NitrosylationTD + BUCys34 / TyrInflammation
CarbamylationBU onlyLysUremia, renal disease
DeamidationBU onlyAsn/GlnSpontaneous modification
AA substitutionBU onlyVariousGenetic variant
N-term truncation (-Asp)TD + BUN-terminusCLD marker
C-term truncation (-Leu)TD + BUC-terminus

Clinical context

  • Disease: Liver dysfunction patients vs controls
  • Implication: Establishes the TD approach as valid for CLD-relevant isoforms, while flagging its limitations for less abundant or novel PTMs
  • TD vs BU trade-off: TD = fast (3 min), clinically deployable, good for known isoforms; BU = comprehensive, slow, requires digestion, but can find novel PTMs

Limitations

  • BU library required substantial upfront effort (DDA library generation)
  • Carbamylation and deamidation found by BU not quantifiable by TD — missing important PTMs in renal/inflammatory disease contexts
  • Relative quantification only (TD) — absolute quantification requires lakis-2024 method

Connections

  • Top-down proteomics — primary approach validated here
  • Bottom-up proteomics — validation approach; SWATH/DIA library
  • HSA — sole protein; 15 isoforms by TD, 127 peptides by BU
  • CQFD-PTM pipeline — pipeline is built on TD approach validated in this paper
  • lakis-2024 — subsequent absolute quantification method (same group, same platform)
  • el-balkhi-2025 — ALBOM clinical application using the validated TD approach

My notes

  • This paper is the methodological foundation justifying the TD-only approach used in el-balkhi-2025 and lakis-2024. Its key message: TD is not perfect, but for the specific set of CLD-relevant isoforms, it’s valid and fast enough for clinical use.
  • Nitrosylation detected by TD — not quantified in ALBOM. Potential biomarker of inflammation in CLD?
  • Carbamylation by BU only: relevant for patients with comorbid renal disease (common in decompensated cirrhosis). This is a gap in the current ALBOM panel.
  • 127-peptide BU library is a valuable resource for future site-level resolution studies.