Procedure for the glycation of bovine serum albumin with maltose

Bovine serum albumin (BSA, Sigma A7030, 70 mg) was dissolved in 3.3 mL of 50 mM KH2PO4/K2HPO4 pH 7.4, 100 mM NaCl buffer and mixed with 6.7 mL 276.4 mM maltose monohydrate (Fluka 63419). The solution was incubated for 11 days at 40 °C. After incubation the buffer was exchanged by ultrafiltration to ddH2O using Amicon (Sigma Aldrich Amicon Ultra-15, UFC9030) with a cutoff of 30 kDa. 100 µL of the obtained 1.5 mL solution was taken for MS/MS analysis. Subsequently, the remaining sample was lyophilized and then dissolved in 550 µL of a 7 M urea-d4 (Armar Chemicals 049500,3041) solution in D2O (Armar Chemicals 014400,0010) for NMR analysis. The pH* (uncorrected readout measured in D2O) was adjusted to 7.4 by adding 3% DCl in D2O (Armar Chemicals 042100.0035). To reduce the disulfide bonds DTT-d10 (Cambridge Isotope Laboratories DLM-2622-0) was added to a concentration of 67 mM. The sample was heated to 60 °C for 15 min. The final protein concentration was ~ 1.9 mM.

Sample preparation of abatacept for NMR analysis

The sample preparation was previously described (Hinterholzer et al. 2022). In brief, abatacept (ORENCIA®, Bristol Myers Squibb; Lot. OE61132, exp. 08/2012) 60 mg in 2.4 mL formulation buffer was dialyzed twice against 4 L ddH2O overnight using a SpectraPor membrane with a cut-off of 3.5 kDa. After lyophilization the sample was dissolved in 650 μL of a 7 M urea-d4 (98 atom%D, ARMAR Chemicals) solution in D2O. DTT-d10 (Cambridge Isotope Laboratories) was added to a concentration of 15 mM, and after an incubation for 15 min at 60 °C, the pH* was adjusted to 7.4 using NaOD (Armar Chemicals).

NMR spectroscopy

Spectra were measured on a 600 MHz Bruker Avance III HD spectrometer equipped with a 1H/13C/15N/31P quadruple-resonance room temperature probe at 298 K. All samples were measured in a standard 5 mm NMR tube (Armar, Type 5TA) with a volume of 500 µL or 550 µL. For assigning the resonances of the Amadori products, the following 2D experiments were recorded: 1H–13C HSQC, 1H–13C HMBC (hmbcgpndqf), 1H–1H TOCSY with mixing times of 100 ms and 12 ms, 1H–1H COSY (cosygpppqf), 1H–13C HSQC–TOCSY (hsqcdietgpsisp.2) with mixing times of 13 ms and 100 ms. More details of the experimental parameters are given in the Figure captions. The data was processed with Topspin 3.6.2 (Bruker) and analyzed with Sparky 1.470 (Lee et al. 2015).

Sample preparation for HPLC–MS/MS analysis

Ultrapure water was produced with a MilliQ® Integral 3 instrument (Millipore, Billerica, MA, USA). Triethylammonium bicarbonate buffer (TEAB, pH 8.50 ± 0.10, 1 mol L−1), sodium dodecyl sulfate (SDS, ≥ 99.5%), tris(2-carboxyethyl)phosphin-hydrochlorid (TCEP, ≥ 98.0%), iodoacetamide (IAA, ≥ 99.0%), formic acid (FA, 98.0–100%) and trifluoroacetic acid (TFA, ≥ 99.0%) were acquired from Sigma-Aldrich (Vienna, Austria). Methanol (MeOH, LiChrosolv®) and ortho-phosphoric acid (85%) were purchased from Merck (Darmstadt, Germany). Acetonitrile (ACN, LC–MS grade) was purchased from “VWR International” (Vienna, Austria). Trypsin (sequencing grade modified, porcine) was acquired from Promega (Madison, WI, USA). 74 µg of glycated BSA (after buffer exchange against ddH2O, pH 7.4, conc.: 7.4 mg.mL−1) were diluted to a concentration of 1.6 µg.µL−1 in 50 mmol L−1 TEAB (pH 8.50) buffer containing 5% (w/w) SDS and denatured by heating for 5 min at 95 °C. Next, disulfides were reduced by addition of TCEP to a concentration of 5.0 mmol L−1 and incubation at 55 °C for 15 min, followed by alkylation of the cysteine residues by addition of IAA to a concentration of 40 mmol L−1 and incubation at 22 °C in the dark for 10 min. Following, the protein was precipitated at a pH ≤ 1 with 12% (v/v) ortho-phosphoric acid and by adding 7:1 (v/v) of 100 mmol L−1 TEAB (pH 7.55) in 90% MeOH (v/v). Next, the proteins were purified by suspension trapping employing S-Trap mini columns (Protifi, Huntington, NY, USA) according to the manufacturer’s instructions, and digested to peptides employing trypsin at a protease/protein ratio of 1:10 (w/w) at 37 °C for 12 h. The obtained peptides were dried at 50 °C in a vacuum centrifuge and resuspended in 1% ACN + 0.10% FA to a concentration of 3.3 µg µL−1.

High-performance liquid chromatography coupled to MS/MS

Chromatographic separation of 1.0 µg peptides was carried out in five technical replicates employing reversed-phase HPLC on an UltiMate™ 3000 RSLCnano System (Thermo Fisher Scientific, Germering, Germany), on a DNV PepMap™ Neo column (150 × 0.075 mm i.d.) from Thermo Fisher Scientific, Germering, Germany. The mobile phases used for the separation were 0.10% aqueous FA (solvent A) and 0.10% FA in ACN (solvent B), pumped at a flow rate of 200 nL.min−1 in the following order: 1.0% B for 5.0 min, a linear gradient from 1.0 to 5.0% B in 5 min, a second linear gradient from 5.0 to 35.0% B in 60.0 min, and a third linear gradient from 35.0 to 45.0% B in 20.0 min. This was followed by flushing at 99.0% B for 10 min and column re-equilibration at 1.0% B for 35 min. The column temperature for the separation was kept constant at 50 °C. The nanoHPLC system was hyphenated to a Q Exactive™ Hybrid Quadrupole-Orbitrap™ mass spectrometer via a Nanospray Flex™ ion source (both from Thermo Fisher Scientific, Bremen, Germany). The source was equipped with a SilicaTip emitter with 360 µm o.d., 20 µm i.d. and a tip i.d. of 10 µm from CoAnn Technologies Inc. (Richland, WA, USA). The spray voltage was set to 1.5 kV, S-lens RF level to 60.0 and capillary temperature to 250 °C. Each scan cycle consisted of a full scan at a scan range of m/z 350–2000 and a resolution setting of 70,000 at m/z 200, followed by 5 data-dependent higher-energy collisional dissociation (HCD) scans in a 2.0 m/z isolation window at 28% normalized collision energy at a resolution setting of 17,500 at m/z 200. For the full scan, the automatic gain control (AGC) target was set to 3e6 charges with a maximum injection time of 100 ms, for the HCD scans the AGC target was 1e5 charges with a maximum injection time of 150 ms. Already fragmented precursor ions were excluded for 10 s. Data acquisition was conducted using Thermo Scientific™ Chromeleon™ 7.2 CDS (Thermo Fisher Scientific, Germering, Germany). For the identification of modification sites, as well as for sequence coverage mapping, Byonic 3.11.3 (Protein Metrics, Cupertino, CA, USA) was used with a precursor and a fragment mass tolerance of 10 ppm, applying a 1% false discovery rate. Relative quantification of the modified peptides was done using MaxQuant 2.0.1.0 (Cox and Mann 2008) with the setting Label-free quantification, applying a 1% false discovery rate.