ɣS-crystallin is a major protein of the human lens and is highly modified with age and cataract due to a lack of lens protein turnover. Previous studies have identified some sites of isomerization and racemization of deamidated asparaginyl and aspartyl residues in ɣS but have been limited due to the complexity of isoforms and difficulty characterizing deamidation post-translational modifications. A total of 32 stable isotope labeled peptides were created for ɣS residues 7-18, 72-78, and 131-145 containing L-Asp, D-Asp, L-isoAsp, and D-isoAsp at D12, N14, N76, D77, and N143 to act as internal chromatography standards spiked into tryptic digests of nuclear insoluble protein of a cataractous human lens. High-resolution mass spectrometry was used to accurately assign deamidation status using the 19 mDa mass defect between isotopic peaks of deamidated and non-deamidated peptides. While peptides containing D-forms of Asp and isoAsp were assigned, the predominate isoforms contained L-isoAsp. High-resolution mass spectrometry using wide single ion monitoring- data-independent acquisition (WiSIM-DIA) also greatly improved the reliable identification of peptide deamidation states. These results will aid creation of ɣS using native chemical ligation to examine the role of isoAsp in crystallin aggregation and cataract.

Preparation of human lens samples. Eyes from human donors were obtained from the Lions Eye Bank of Oregon (approved through the Institutional Review Board at Oregon Health & Science University), lenses were dissected, imaged, then frozen at -70̊C until processed. Frozen lenses were cored by placing on a watch glass cooled on dry ice, a precooled 5 mm corneal trephine placed against a pole, and the cortical fibers along the equator removed using a razor blade. Approximately 1 mm of cortical fibers were then removed from each pole, and the still frozen cortical fibers and cylinder approximating the lens nucleus collected separately. These cortical and nuclear regions were then each dispersed using a glass Dounce homogenizer in 1.0 ml of homogenization buffer containing 20 mM phosphate (pH 7),150 mM NaCl,1 mM EDTA buffer and centrifuged at 20,000 x g for 30 min at 4̊ C. The supernatant was then removed, the pellet suspended by vigorous vortexing, the suspension centrifuged, and the two supernatants pooled. The combined supernatants were designated the soluble fraction. The pellet was dispersed uniformly in 1 ml of homogenization buffer by brief probe sonication while on ice, and this was designated the insoluble fraction. A lens from a 5-day-old human donor was similarly processed without separation into cortical or nuclear regions and the soluble fraction used as a control sample. BCA protein assays were performed on samples, and samples dried by vacuum centrifugation and stored at -70̊ C until used. Trypsinization of lens samples. One hundred µg portions of lens protein were dissolved in 50 µl of 8M deionized urea, 1.0 M Tris (pH 8.5), 8 mM CaCl2 and 0.2 M methylamine, 4 µl of 0.2 M dithiothreitol (DTT) added, samples heated at 50̊ C for 15 min, 4 µl of 0.5M iodoacetamide added, samples stored at room temperature in the dark for 30 min, and an additional 8 µl of 0.2 M DTT added. After another 15 min at room temperature, 294 µl of water was added, followed by 40 µl of 0.1 µg/µl MS grade Pierce trypsin added (Thermo Scientific). Samples were then incubated overnight at 37̊ C with shaking and 20 µl of formic acid added to stop digestion. Stable isotope labeled peptide synthesis. Synthetic peptides containing L-Asn, L-Asp and non-standard D-Asp, L-isoAsp,and D-isoAsp amino acids and C-terminally labeled 13C6, 15N4 Arg to introduce a 10 Da mass increase were synthesized using Fmoc protected amino acids on a Liberty Blue automated peptide synthesizer (CEM Corporation) using preloaded (13C6,99%; 15N4,99%) L-Arginine(PBF)-2-ClTrt Resin (Catalog # S101-001) (New England Peptide), and non-standard amino acids: N-Fmoc-L-aspartic acid 1-tert-butyl ester, 95%; N-Fmoc-D-aspartic acid 4-tert-butyl ester, 98%; and N-Fmoc-D-aspartic acid 1-tert-butyl ester, 95% (Catalog # H62355.06, H62356.06, and H62680.06, respectively, Thermo Scientific). Following cleavage, peptides were purified by preparative C18 reverse phase chromatography, masses determined using an LTQ ion trap mass spectrometer (Thermo Scientific), and purified fractions with correct masses were pooled and lyophilized. Prior to use, the peptides were dissolved in 20% ACN, their approximate concentrations determined by the absorbance at 214 nm, and from 0.2-0.5 pmoles added per µg of lens protein digest. To decrease the number of required LC/MS runs, a single isomeric form of each of the 3 synthetic peptide ɣS regions were often mixed prior to analysis. Nine isoforms of ɣS-crystallin peptide 7-18: ITFYEDKNFQGR (Table 1), 19 isoforms of ɣS-crystallin peptide 72-78: WMGLNDR (Table 2), and 4 isoforms of ɣS-crystallin peptide 131-145: VLEGVWIFYELPNYR (Table 3) were created (32 synthetic peptides in total). Mass spectrometric analysis. One µg of peptide digest of insoluble protein from an 88-year-old cataractous lens nucleus was first used to create a spectral library of assigned MS/MS scans using a data-dependent acquisition (DDA) run prior to the runs using the wiSIM-DIA method. The wiSIM-DIA method used 250 ng of peptide digests during the comparisons of 88-year-old lens soluble and insoluble protein and 5-day-old lens soluble protein as well as during the analysis of insoluble protein from the 88-year-old lens nucleus spiked with heavy labeled peptide standards. The localization of deamidation sites in ɣS peptide 7-18 used 1 µg of digest from the 88-year-old lens nucleus. All LC-MS runs used the same chromatographic method. Chromatographic method. Peptides were chromatographically separated using a Dionex NCS-3500RS UltiMate RSLCnano UPLC system and an Orbitrap Fusion Tribrid instrument with an EasySpray nano source for mass analysis (Thermo Scientific). Peptides were injected onto an Acclaim PepMap 100 μm x 2 cm NanoViper C18, 5 μm trap column at a 5 µl/min flow rate using a loading pump and 2% acetonitrile (ACN), 0.1% formic acid mobile phase. After 5 min of loading and washing, the trap column was connected using a switching valve to a PepMap RSLC C18, 2 μm, 75 μm x 25 cm EasySpray column (Thermo Scientific) maintained at 40̊ C flowing at 0.3 µl/min and equilibrated with 2% ACN, 0.1% formic acid. The ACN concentration in the mobile phase was then increased to 7.5% and peptides separated using a linear 7.5–30% ACN gradient over 205 min. The column was then washed by a 30-90% ACN gradient over 5 min, held at 90% ACN for 5 min, followed by a re-equilibration at 2% ACN for 15 min (total method length = 230 min). Data-dependent acquisition run used to create the spectral library. Survey MS scans from m/z = 400-1000 were collected using the instrument’s Orbitrap mass analyzer at a resolution of 120,000, with maximum injection time (MIT) = 50 ms, automatic gain control (AGC) target at 4x10^5, mass calibration using the m/z = 445.14 polysiloxane ion, and data collected in profile mode. MS2 peptide fragment spectra were collected using data-dependent acquisition with 3 sec cycle time between precursor scans, the MIPS filter activated, charge state filter to include +2-7 charge states, dynamic exclusion activated with a repeat count of 1, single charge state per precursor not activated, exclusion duration = 60 sec, with a signal intensity threshold of 5 x 103. The quadrupole isolation window was set to 1.6 Da, with CID fragmentation and MS/MS spectra generated by the ion trap using a collision energy of 35%, rapid scan rate, and AGC target = 1x 10^4. WiSIM-DIA method. While the wiSIM-DIA method usually uses wide isolation selected ion monitoring precursor scans at a 240,000 resolution setting [12, 19], we performed these scans at 500,000 resolution to better resolve the 19 mDa mass differences between isotopic peaks of deamidated vs non-deamidated peptides. The method using the Orbitrap Fusion instrument consisted of 6 data collections that were cycled throughout the chromatographic separation (tSIM, tMS2, tSIM, tMS2, tSIM, tMS2). The tSIM methods were all identical and specified a target of m/z =700 but used a 600 m/z isolation in the ion trap, precursor scan in the orbitrap at 500,000 resolution in profile mode, mass calibrant of polysiloxane at m/z = 445.12, maximum injection time = 50 ms, and AGC = 3 x10^4, resulting in m/z 400-1000 precursor scans lasting approximately 1 sec each. The 17 DIA tMS2 scans were performed in parallel during each 500,000 resolution Orbitrap tSIM precursor scan and used the ion trap at a normal scan rate so that the time required for these 17 DIA scans approximated the required time to complete the 500,000 resolution scan in the Orbitrap. Quadrupole isolation of 12 m/z was used with CID fragmentation, collision energy of 35%, scan range m/z = 250-1850, maximum injection time = 47 ms, and AGC target = 1 x 10^3. The first set of 17 MS2 scans specified targets starting at m/z 406 and sequentially increased by 12 m/z to a m/z = 598, effectively fragmenting all ions between m/z 400-604 in DIA mode. The second set of 17 MS2 scans specified targets starting at m/z 606, and sequentially increased by 12 m/z to a m/z of 798, effectively fragmenting all ions between m/z 600-804, and the third set of 17 MS2 scans specified targets starting at m/z 806, and sequentially increased by 12 m/z to a m/z of 998, effectively fragmenting all ions between m/z 800-1006. These lower resolution DIA MS2 scans in the ion trap were only used to confirm the identities of peaks in the precursor scans using fragment ions assigned in the spectral library. The full instrument settings are found in Supplementary Figure 1. Deamidation localization method. The method used the same 500,000 resolution Orbitrap MS scan for precursors masses but instead alternated with targeted MS2 scans using the Orbitrap to generate 240,000 resolution fragment ion spectra so that extracted ion chromatograms could be created that were specific for deamidation sites in peptides containing more than one amide using the 19 mDa mass differences between deamidated and non-deamidated fragment ions. The isolation used the quadrupole with a 1.6 Da isolation window, HCD activation, 30% collision energy, 150-1850 scan range, maximum injection time = 50 ms, and AGC target = 5 x 10^4. The lower 240,000 resolution in the Orbitrap was used for MS2, due to the smaller masses and lower charge states of fragment ions, and to allow approximately a dozen target MS2 scans in 8 sec time between precursor scans, providing approximately 8 data points across the typical 1 min peak elution times during chromatography. Data analysis The DDA results from the digest of insoluble proteins from the nucleus of the 88-year-old cataractous lens was searched against a human Swiss Prot database downloaded in July 2015 (20,207 sequences) using differential modifications of +15.995 on Met and +0.984 on Asn and Gln, with parent and daughter ion tolerances of 10 ppm and 1.0 Da, respectively, using Sequest within the Protein Discoverer 1.4 suite (Thermo Scientific). Peptide false discovery was controlled using Percolator software to include only peptide identifications with q values < 0.01. The resultant msf-format file was then imported into Skyline software (version 24.1.0.414) to create a spectral library. Deamidated and non-deamidated doubly charged forms of ɣS-crystallin peptides 7-18, 72-78, and 131-145 were added to the Skyline working list to create extracted ion chromatograms for both precursor and major fragment ions for the three peptides to allow comparison of peptide isoforms in the 88-year-old lens fractions and 5-day-old control lens. In samples containing spiked heavy labeled standards added to the digest of insoluble protein from an 88-year-old lens, deamidated versions of the three peptides were manually added to the working peptide list with the Asn residues changed from N to D instead of using the deamidated versions from the spectral library. This facilitated creating overlayed extracted ion chromatograms in different colors so that the peaks of non-deamidated endogenous, deamidated endogenous, and stable isotope labeled internal standard peptides could be more easily visualized using a control key to select the multiple versions of the peptides to display within a single chromatogram. Transition settings in Skyline included MS1 filtering of 3 peaks (M, M+1, and M+2) with an Orbitrap mass analyzer, resolving power of 500,000 at 400 m/z, MS/MS filtering set in DIA mode, and product mass analyzer set at QIT. A custom isolation scheme was created called WiderSIMDIA with “Use results data isolation targets” selected, isolation width set as fixed, a 12 m/z specified, deconvolution set at none, and the specified resolution of the QIT set at 0.7 m/z. Percent single deamidation within each peptide was determined using Skyline to measure the total area of all singly deamidated peaks and dividing this by the area of all non-deamidated + deamidated peaks X100. Calculation of the % area of each individual peptide isoform was performed using Qual Browser within Xcalibur software (Version 4.5.445.18, Thermo Scientific).

Tryptic digests from the nuclear soluble and insoluble fractions of a cataractous 88-year-old human donor lens and the soluble protein from a whole lens of a 5-day-old human donor. Spike in stable isotope labeled synthetic peptides matching residues 7-18, 72-78, and 131-148 of gamma S-crystallin containing L-Asn, L-Asp, L-isoAsp, D-Asp, and D-isoAsp at residues 12, 14, 76, 77, and 143 were added to identify modification sites found in vivo.