Anchorchip plate

The methylation reactions were performed using the unmodified H3K36 (aa 26-44), H3K36me1 (aa 26-44), or H1.5 (aa 160–176) peptide (4.5 μM) in methylation buffer (50 mM Tris/HCl, pH 9, 5 mM MgCl₂ and 1 mM DTT) supplemented with 1 mM unlabeled AdoMet (Sigma-Aldrich) and 6.7 μM NSD1 for 4 h at 37 °C. The reactions were halted by the addition of 0.1% trifluoroacetic acid (TFA). All the samples were cleaned using C18 tips (Agilent Technologies). The eluted samples were spotted onto an anchor chip plate (Bruker-Daltonics) followed by drying. Next, 1 μl of HCCA matrix (0.7 mg/ml α-cyano-4 hydroxycinnamic acid dissolved in 85% acetonitrile, 0.1% TFA, 1 mM ammonium dihydrogen phosphate) was added to the dried sample spots and allowed to dry again. Afterward, the dried spots on the anchor plate were analyzed using an Autoflex Speed MALDI-TOF mass spectrometer (Bruker-Daltonics). The mass spectra were collected using the Flex control software (Bruker-Daltonics). For calibration, the peptide calibration standard (Bruker-Daltonics) with peptides ranging from 700 to 3200 Da was used. The collected spectra were analyzed with Flex analysis software (Bruker-Daltonics).

The protein bands that in the western blot were recognized by the pool of positive sera were retrieved from the gels, sequenced and identified at the “Servicio de Proteómica del Centro de Biología Molecular Severo Ochoa (CBMSO)” in Madrid, Spain. The bands were excised manually and digested automatically in situ with a robot digester (Bruker, Billerica, Massachusetts, USA) using trypsin following a previously described protocol [17 (link)]. The digestion products (containing the peptides) were acidified with trifluoroacetic acid (0.1% final concentration), dried in a Speed Vac (Thermo Fisher Scientific, Waltham, Massachusetts, USA ) and finally resuspended in TA (0.1% trifluoroacetic acid, 33% acetonitrile). A 0.5 ml aliquot was placed on an anchor-chip plate (Bruker), using 2.5-dihydroxybenzoic acid (DHB) as a matrix, at a concentration of 5 g/l via the “fast evaporation” method. The samples on the plate were analysed using an Autoflex matrix assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometer (Bruker) equipped with a reflector. The mass spectra from these experiments were used as a peptide fingerprint to identify proteins using Mascot and Profound databases.

Obtained trypsin digests of IgA samples were enriched for glycopeptides by a two-step microtip cotton HILIC SPE, using cotton thread as the solid phase, as described before38 (link), with minor modifications in the used solution volumes and composition. Briefly, 15 μL of the digest was transferred to a 96 wells V-bottom plate, and adjusted to 70% ACN by addition of 35 μL ACN. The cotton thread was washed three times with 20 μL MQ, and conditioned three times with 20 μL 70% ACN, using a 12-channel micropipette (2–20 μL). The sample was then loaded on the cotton by pipetting up-and-down 20 times. Subsequently, the cotton was washed 3 times with 20 μL 70% ACN containing 1% TFA and three washes with 20 μL 70% ACN, followed by elution in 5 μL MQ. Twenty-five microliter ACN was then added to the tryptic digest to bring the sample to 80% ACN for a sequential HILIC enrichment with 80% instead of 70% ACN, to capture glycopeptides that were not detectable in the 70% HILIC eluates. The eluates from both HILIC experiments were mixed separately with 15 μL 4-chloro-α-cyanocinnamic acid matrix (Cl-CCA; 0.33 mg/mL in acetone:ethanol 1:2) and 1 μL of each eluate was spotted on a Bruker AnchorChip plate with 800 μm anchors. All glycopeptide samples were spotted in duplicate, resulting in 384 spots per 96 serum samples.