The dynamic modification of specific serine and threonine residues of intracellular proteins by O-linked N-acetyl-β-D-glucosamine (O-GlcNAc) mitigates injury and promotes cytoprotection in a variety of stress models. The O-GlcNAc transferase (OGT) and the O-GlcNAcase (OGA) are the sole enzymes that add and remove O-GlcNAc, respectively, from thousands of substrates. It remains unclear how just two enzymes can be specifically controlled to affect glycosylation of target proteins and signaling pathways both basally and in response to stress. Several lines of evidence suggest that protein interactors regulate these responses by affecting OGT and OGA activity, localization and substrate specificity. To provide insight into the mechanisms by which OGT function is controlled, we have utilized quantitative proteomics to define OGT’s basal and stress-induced interactomes. OGT and its interaction partners were immunoprecipitated from OGT wild-type, null and hydrogen peroxide treated cell lysates that had been isotopically labeled with light, medium and heavy lysine and arginine (stable isotopic labeling of amino acids in cell culture [SILAC]). In total, more than 130 proteins were found to interact with OGT, many of which change their association upon hydrogen peroxide stress. These proteins include the major OGT cleavage and glycosylation substrate, host cell factor 1, which demonstrated a time-dependent dissociation following stress. To validate less-well characterized interactors, such as glyceraldehyde 3-phosphate dehydrogenase and histone deacetylase 1, we turned to parallel reaction monitoring, which recapitulated our discovery-based SILAC approach. Although the majority of proteins identified are novel OGT interactors, 64% of them are previously characterized glycosylation targets that contain varied domain architecture and function. Together these data demonstrate that OGT interacts with unique and specific interactors in a stress-responsive manner.

PRM was used to confirm the changes in co-immunoprecipitation of several identified OGT interactors under stress conditions . The immunoprecipitation of OGT was repeated as described above; however, OGT and interactors were eluted with 4% w/v SDS in 50 mM Tris-HCl pH 7.6. Samples were reduced with 5 mM dithiothreitol in 100 mM ammonium bicarbonate (ABC) (1 h, 56°C) and alkylated with 10 mM iodoacetamide in 100 mM ABC (30 min, room temperature). The protein content was precipitated with acetone and 10% v/v trichloroacetic acid (16 h, -20°C). After centrifugation at 10,000 x g (4oC, 15 min) the supernatant was removed, and the pellet rinsed with cold acetone. The pellet was reconstituted in 50 µL of 80 mM ABC with 20% v/v acetonitrile and 3 µg of Trypsin/LysC (Promega) were added for digestion (16 h, 37°C). Three biological replicates in the WT, stress and null were analyzed in triplicate by targeted LC-MS/MS on an Fusion Orbitrap Lumos mass spectrometer interfaced with Easy-nLC 1200 system. Previously identified high intensity peptides from that are unique to OGT, HCF1, CARM1, histone-lysine N-methyltransferase SETD1A (SETD1A), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), histone deacetylase 1 (HDAC1), eukaryotic translation initiation factor 3 subunit B (EIF3B), and heat shock factor protein 1 (HSF1) were included for PRM analysis using a 0.4 m/z isolation window by higher-energy collisional dissociation (HCD) for 8 min surrounding the observed elution time (Table S4). Peptides were then separated over a 60 min gradient with a precursor ion scan from 350 to 2000 m/z was acquired at 120,000 resolution every 3 sec followed by fragmentation scans acquired at 60,000 resolution.

Mouse embryonic fibroblasts (MEF) were treated with 2.5 mM hydrogen peroxide for 1.5 hours. MEFs where the O-GlcNAc transferase can be inducibly deleted were used as a negative control. OGT was immunoprecipitated and eluted with 4% SDS.