Advertisement

Bioinformatic prediction of putative conveyers of O-GlcNAc transferase intellectual disability

Open AccessPublished:July 18, 2022DOI:https://doi.org/10.1016/j.jbc.2022.102276
      Protein O-GlcNAcylation is a dynamic posttranslational modification that is catalyzed by the enzyme O-GlcNAc transferase (OGT) and is essential for neurodevelopment and postnatal neuronal function. Missense mutations in OGT segregate with a novel X-linked intellectual disability syndrome, the OGT congenital disorder of glycosylation (OGT-CDG). One hypothesis for the etiology of OGT-CDG is that loss of OGT activity leads to hypo-O-GlcNAcylation of as yet unidentified, specific neuronal proteins, affecting essential embryonic, and postnatal neurodevelopmental processes; however, the identity of these O-GlcNAcylated proteins is not known. Here, we used bioinformatic techniques to integrate sequence conservation, structural data, clinical data, and the available literature to identify 22 candidate proteins that convey OGT-CDG. We found using gene ontology and PANTHER database data that these candidate proteins are involved in diverse processes including Ras/MAPK signaling, translational repression, cytoskeletal dynamics, and chromatin remodeling. We also identify pathogenic missense variants at O-GlcNAcylation sites that segregate with intellectual disability. This work establishes a preliminary platform for the mechanistic dissection of the links between protein O-GlcNAcylation and neurodevelopment in OGT-CDG.

      Keywords

      Abbreviations:

      AHCY (adenosyl homocysteinase), CDG (congenital disorder of glycosylation), CTE (C-terminal extension), DD (developmental delay), EESC (embryonic stem cell), GO (gene ontology), ID (intellectual disability), mESC (mouse embryonic stem cell), OGT (O-GlcNAc transferase), PDB (Protein Data Bank), PTM (posttranslational modification), TPR (tetratricopeptide)
      Pathogenic missense variants in O-GlcNAc transferase (OGT) give rise to OGT congenital disorder of glycosylation (OGT-CDG), a rare, clinically heterogenous intellectual disability (ID) syndrome where the underlying molecular etiology is unknown. By filtering the O-GlcNAc proteome for putatively functional O-GlcNAc sites and O-GlcNAc sites mutated in ID patients, we have identified a subset of the O-GlcNAcome, the hypo-O-GlcNAcylation of which may give rise to OGT-CDG.
      O-GlcNAcylation, the covalent attachment of monosaccharide units of β-linked GlcNAc to the hydroxyl groups of protein serine/threonine (Ser/Thr) residues, is a dynamic and highly conserved posttranslational modification (PTM) occurring on thousands of nucleocytoplasmic and mitochondrial proteins (
      • Yang X.
      • Qian K.
      Protein O-GlcNAcylation: emerging mechanisms and functions.
      ). This dynamic PTM is catalyzed by the OGT (
      • Holt G.D.
      • Hart G.W.
      The subcellular distribution of terminal N-acetylglucosamine moieties. Localization of a novel protein-saccharide linkage, O-linked GlcNAc.
      ,
      • Haltiwanger R.S.
      • Blomberg M.A.
      • Hart G.W.
      Glycosylation of nuclear and cytoplasmic proteins. Purification and characterization of a uridine diphospho-N-acetylglucosamine:polypeptide beta-N-acetylglucosaminyltransferase.
      ). OGT is a type B glycosyltransferase, composed of 13.5 N-terminal tetratricopeptide (TPR) repeats and a C-terminal glycosyltransferase domain (
      • Martinez-Fleites C.
      • He Y.
      • Davies G.J.
      Structural analyses of enzymes involved in the O-GlcNAc modification.
      ). The N-terminal TPRs confer both substrate specificity and serve as a molecular hub for the assembly of multiprotein complexes (
      • Stephen H.M.
      • Praissman J.L.
      • Wells L.
      Generation of an interactome for the tetratricopeptide repeat domain of O-GlcNAc transferase indicates a role for the enzyme in intellectual disability.
      ). The C-terminal glycosyltransferase domain possesses both sugar transfer activity and a noncanonical protease activity, the latter being required for proteolytic maturation of the cell cycle regulator HCF-1 (
      • Capotosti F.
      • Guernier S.
      • Lammers F.
      • Waridel P.
      • Cai Y.
      • Jin J.
      • et al.
      O-GlcNAc transferase catalyzes site-specific proteolysis of HCF-1.
      ). The antagonistic enzyme to OGT, O-GlcNAc hydrolase, removes the O-GlcNAc modification (
      • Vocadlo D.J.
      O-GlcNAc processing enzymes: Catalytic mechanisms, substrate specificity, and enzyme regulation.
      ).
      OGT is most abundant in the brain (
      • Okuyama R.
      • Marshall S.
      UDP-N-acetylglucosaminyl transferase (OGT) in brain tissue: temperature sensitivity and subcellular distribution of cytosolic and nuclear enzyme.
      ), where O-GlcNAc mass spectrometric studies have identified thousands of O-GlcNAc-modified proteins involved in synaptic plasticity, excitatory neurotransmission, and neuronal activity–dependent gene transcription (
      • Vosseller K.
      • Trinidad J.C.
      • Chalkley R.J.
      • Specht C.G.
      • Thalhammer A.
      • Lynn A.J.
      • et al.
      O-linked N-acetylglucosamine proteomics of postsynaptic density preparations using lectin weak affinity chromatography and mass spectrometry.
      ,
      • Alfaro J.F.
      • Gong C.X.
      • Monroe M.E.
      • Aldrich J.T.
      • Clauss T.R.
      • Purvine S.O.
      • et al.
      Tandem mass spectrometry identifies many mouse brain O-GlcNAcylated proteins including EGF domain-specific O-GlcNAc transferase targets.
      ,
      • Trinidad J.C.
      • Barkan D.T.
      • Gulledge B.F.
      • Thalhammer A.
      • Sali A.
      • Schoepfer R.
      • et al.
      Global identification and characterization of both O-GlcNAcylation and phosphorylation at the murine synapse.
      ). However, the site-specific roles of O-GlcNAc in regulating neurodevelopmental and postnatal neuronal processes remains poorly understood, with the functional consequences of O-GlcNAcylation understood for only a handful of neuronal OGT substrates. Despite this, previous work has established the role of O-GlcNAc in modulating dozens of neurodevelopmental processes, including axonal growth (
      • Rexach J.E.
      • Clark P.M.
      • Mason D.E.
      • Neve R.L.
      • Peters E.C.
      • Hsieh-Wilson L.C.
      Dynamic O-GlcNAc modification regulates CREB-mediated gene expression and memory formation.
      ), dendritic spine maturation (
      • Lagerlöf O.
      • Hart G.W.
      • Huganir R.L.
      O-GlcNAc transferase regulates excitatory synapse maturity.
      ), and synaptic vesicle dynamics (
      • Skorobogatko Y.
      • Landicho A.
      • Chalkley R.J.
      • Kossenkov A.V.
      • Gallo G.
      • Vosseller K.
      O-linked β-N-acetylglucosamine (O-GlcNAc) site thr-87 regulates synapsin I localization to synapses and size of the reserve pool of synaptic vesicles.
      ). Furthermore, hippocampal-specific depletion of OGT abrogates long-term potentiation and excitatory synapse maturation, key processes underpinning learning and memory, potentially due to disrupted α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor trafficking (
      • Lagerlöf O.
      • Hart G.W.
      • Huganir R.L.
      O-GlcNAc transferase regulates excitatory synapse maturity.
      ). Additionally, O-GlcNAcylation is essential for dopaminergic neuron survival, with loss of OGT activity leading to massive apoptosis and near-abolition of dopamine release as a second order consequence (
      • Lee B.E.
      • Kim H.Y.
      • Kim H.J.
      • Jeong H.
      • Kim B.G.
      • Lee H.E.
      • et al.
      O-GlcNAcylation regulates dopamine neuron function, survival and degeneration in Parkinson disease.
      ). At the organismal level, O-GlcNAcylation is protective against neurodegeneration and age-related decline in learning and memory (
      • Wheatley E.G.
      • Albarran E.
      • White C.W.
      • Bieri G.
      • Sanchez-Diaz C.
      • Pratt K.
      • et al.
      Neuronal O-GlcNAcylation improves cognitive function in the aged mouse brain.
      ). Strikingly, pan-neuronal KO of OGT results in perinatal lethality in mice, which is immediately preceded by locomotor dysfunction and hyperphosphorylation of the Alzheimer’s disease associated protein Tau (
      • O'Donnell N.
      • Zachara N.E.
      • Hart G.W.
      • Marth J.D.
      Ogt-dependent X-chromosome-linked protein glycosylation is a requisite modification in somatic cell function and embryo viability.
      ). Taken together, O-GlcNAcylation is an essential PTM for neurodevelopment and neuronal function, though a comprehensive mechanistic understanding of how O-GlcNAc functions in the brain and developing central nervous system is lacking.
      Recently, several groups have independently reported missense variants in the OGT gene that segregate with ID (
      • Pravata V.M.
      • Muha V.
      • Gundogdu M.
      • Ferenbach A.T.
      • Kakade P.S.
      • Vandadi V.
      • et al.
      Catalytic deficiency of O-GlcNAc transferase leads to X-linked intellectual disability.
      ,
      • Pravata V.M.
      • Gundogdu M.
      • Bartual S.G.
      • Ferenbach A.T.
      • Stavridis M.
      • Õunap K.
      • et al.
      A missense mutation in the catalytic domain of O-GlcNAc transferase links perturbations in protein O-GlcNAcylation to X-linked intellectual disability.
      ,
      • Vaidyanathan K.
      • Niranjan T.
      • Selvan N.
      • Teo C.F.
      • May M.
      • Patel S.
      • et al.
      Identification and characterization of a missense mutation in the O-linked β-N-acetylglucosamine (O-GlcNAc) transferase gene that segregates with X-linked intellectual disability.
      ,
      • Willems A.P.
      • Gundogdu M.
      • Kempers M.J.E.
      • Giltay J.C.
      • Pfundt R.
      • Elferink M.
      • et al.
      Mutations in N-acetylglucosamine (O-GlcNAc) transferase in patients with X-linked intellectual disability.
      ). IDs are a group of clinically and genetically heterogenous neurodevelopmental disorders, collectively characterized by an intelligence quotient of less than 70, as well as compromised adaptive and social behavior (
      • Vissers L.E.
      • Gilissen C.
      • Veltman J.A.
      Genetic studies in intellectual disability and related disorders.
      ). ID can result from mutations in one of over 700 genes and typically co-occurs with failure to reach key milestones in social, motor, or language development (known as developmental delay) (
      • Vissers L.E.
      • Gilissen C.
      • Veltman J.A.
      Genetic studies in intellectual disability and related disorders.
      ). The ID syndrome which segregates with OGT missense variants, recently named the OGT-CDG, is symptomatically highly heterogenous (
      • Pravata V.M.
      • Omelková M.
      • Stavridis M.P.
      • Desbiens C.M.
      • Stephen H.M.
      • Lefeber D.J.
      • et al.
      An intellectual disability syndrome with single-nucleotide variants in O-GlcNAc transferase.
      ). Patients typically present with ID, developmental delay, hypotonia, hearing/visual impairments, facial dysmorphia, long-thin fingers, clinodactyly, short stature, and eye abnormalities (
      • Pravata V.M.
      • Omelková M.
      • Stavridis M.P.
      • Desbiens C.M.
      • Stephen H.M.
      • Lefeber D.J.
      • et al.
      An intellectual disability syndrome with single-nucleotide variants in O-GlcNAc transferase.
      ). However, additional conditions such as epilepsy and microcephaly are comorbid with OGT-CDG. This clinical heterogeneity strongly suggests that multiple signaling cascades and developmental processes are disrupted in OGT-CDG, further complicating therapeutic strategies and dissection of molecular etiology. Moreover, exactly how missense mutations in OGT give rise to OGT-CDG is not clear. Pathogenic OGT mutations have been reported in both the substrate specificity conferring TPRs (
      • Vaidyanathan K.
      • Niranjan T.
      • Selvan N.
      • Teo C.F.
      • May M.
      • Patel S.
      • et al.
      Identification and characterization of a missense mutation in the O-linked β-N-acetylglucosamine (O-GlcNAc) transferase gene that segregates with X-linked intellectual disability.
      ,
      • Willems A.P.
      • Gundogdu M.
      • Kempers M.J.E.
      • Giltay J.C.
      • Pfundt R.
      • Elferink M.
      • et al.
      Mutations in N-acetylglucosamine (O-GlcNAc) transferase in patients with X-linked intellectual disability.
      ) and the glycosyltransferase domain, leading to divergent hypotheses regarding molecular etiology. For example, it has been hypothesized that OGT missense variants abrogate proteolytic maturation of HCF1, preventing normal cell cycle progression (
      • Pravata V.M.
      • Omelková M.
      • Stavridis M.P.
      • Desbiens C.M.
      • Stephen H.M.
      • Lefeber D.J.
      • et al.
      An intellectual disability syndrome with single-nucleotide variants in O-GlcNAc transferase.
      ). However, OGT modifies >4000 substrates, and reduced OGT activity toward a single substrate cannot explain the clinical heterogeneity observed in OGT-CDG patients. Alternatively, it has been hypothesized that the missense variants reduce OGT stability, indirectly leading to reduced protein O-GlcNAcylation. While this has been observed for the OGTL254F variant, where the TPR domain displays reduced thermal stability in vitro (
      • Gundogdu M.
      • Llabrés S.
      • Gorelik A.
      • Ferenbach A.T.
      • Zachariae U.
      • van Aalten D.M.F.
      The O-GlcNAc transferase intellectual disability mutation L254F distorts the TPR helix.
      ), this has not been reported for other OGT-CDG mutations. The majority of reported OGT missense variants map to the TPRs (
      • Vaidyanathan K.
      • Niranjan T.
      • Selvan N.
      • Teo C.F.
      • May M.
      • Patel S.
      • et al.
      Identification and characterization of a missense mutation in the O-linked β-N-acetylglucosamine (O-GlcNAc) transferase gene that segregates with X-linked intellectual disability.
      ,
      • Willems A.P.
      • Gundogdu M.
      • Kempers M.J.E.
      • Giltay J.C.
      • Pfundt R.
      • Elferink M.
      • et al.
      Mutations in N-acetylglucosamine (O-GlcNAc) transferase in patients with X-linked intellectual disability.
      ,
      • Pravata V.M.
      • Omelková M.
      • Stavridis M.P.
      • Desbiens C.M.
      • Stephen H.M.
      • Lefeber D.J.
      • et al.
      An intellectual disability syndrome with single-nucleotide variants in O-GlcNAc transferase.
      ), leading to the hypothesis that selective loss of OGT interactors may give rise to OGT-CDG. However, no data have surfaced to date which substantiates this hypothesis. Instead, characterization of the OGT-CDG N648Y and N567K variants revealed that mutations in the OGT glycosyltransferase domain lead to lower global O-GlcNAcylation levels in mouse embryonic stem cells (mESCs) and Drosophila melanogaster, impair proteolytic maturation of HCF1, and strongly attenuate O-GlcNAcylation of the physiological OGT substrate Table 1 in vitro (
      • Pravata V.M.
      • Muha V.
      • Gundogdu M.
      • Ferenbach A.T.
      • Kakade P.S.
      • Vandadi V.
      • et al.
      Catalytic deficiency of O-GlcNAc transferase leads to X-linked intellectual disability.
      ,
      • Pravata V.M.
      • Gundogdu M.
      • Bartual S.G.
      • Ferenbach A.T.
      • Stavridis M.
      • Õunap K.
      • et al.
      A missense mutation in the catalytic domain of O-GlcNAc transferase links perturbations in protein O-GlcNAcylation to X-linked intellectual disability.
      ). Collectively, these data indicate that hypo-O-GlcNAcylation of OGT substrates at particular developmental time points, due to catalytic deficiency of OGT, may give rise to OGT-CDG. The identity of these OGT substrates, referred to in this manuscript as “phenotypic conveyers,” is unknown. Identification of these phenotypic conveyers would allow elucidation of the molecular etiology of OGT-CDG and further develop an understanding of the roles of O-GlcNAcylation in neurodevelopment and neuronal function through dissection of the site-specific roles of O-GlcNAc on these proteins.
      Table 1Missense variants at conserved sites of O-GlcNAcylation identified in patients with ID/DD
      GeneVariantPatient conditionPathogenicityPolyPhen score
      PUM1S802FSpinocerebellar ataxia 47Uncertain significance0.921
      TUBB2BS172PCortical dysplasia, complex, with other brain malformationsPathogenic
      TUBB2BS172LCortical dysplasia, complex, with other brain malformationsLikely pathogenic
      DOCK7S190NEpileptic encephalopathy, early infantile, 23Uncertain significance0.995
      DOCK7S190GEpileptic encephalopathy, early infantile, 23Uncertain significance0.989
      ATRXS594CAlpha thalassemia-X-linked intellectual disabilityUncertain significance0.993
      HCF1T556MMental retardation 3, X-linkedUncertain significance0.959
      KMT2AS1858IWiedemann–Steiner syndromeUncertain significance0.855
      The ClinVar database was screened for pathogenic missense mutations at O-GlcNAc sites reported in a recently published O-GlcNAc database (see (
      • Tarrant M.K.
      • Rho H.S.
      • Xie Z.
      • Jiang Y.L.
      • Gross C.
      • Culhane J.C.
      • et al.
      Regulation of CK2 by phosphorylation and O-GlcNAcylation revealed by semisynthesis.
      )). Patient condition and the associated pathogenicity of the missense mutation are summarized. Where the significance of the missense mutation is uncertain, PolyPhen HumVar scores are provided, where default cut-offs of 0.15 for “likely benign” and 0.85 for “probably damaging” are applied.
      One of the challenges in identifying phenotypic conveyers of the OGT-CDG phenotype is the lack of depth of information provided by O-GlcNAc site-mapping studies. O-GlcNAc proteomic studies have expanded the size of the O-GlcNAcome dramatically in the past 15 years (
      • Vosseller K.
      • Trinidad J.C.
      • Chalkley R.J.
      • Specht C.G.
      • Thalhammer A.
      • Lynn A.J.
      • et al.
      O-linked N-acetylglucosamine proteomics of postsynaptic density preparations using lectin weak affinity chromatography and mass spectrometry.
      ,
      • Alfaro J.F.
      • Gong C.X.
      • Monroe M.E.
      • Aldrich J.T.
      • Clauss T.R.
      • Purvine S.O.
      • et al.
      Tandem mass spectrometry identifies many mouse brain O-GlcNAcylated proteins including EGF domain-specific O-GlcNAc transferase targets.
      ,
      • Trinidad J.C.
      • Barkan D.T.
      • Gulledge B.F.
      • Thalhammer A.
      • Sali A.
      • Schoepfer R.
      • et al.
      Global identification and characterization of both O-GlcNAcylation and phosphorylation at the murine synapse.
      ,
      • Hahne H.
      • Sobotzki N.
      • Nyberg T.
      • Helm D.
      • Borodkin V.S.
      • van Aalten D.M.
      • et al.
      Proteome wide purification and identification of O-GlcNAc-modified proteins using click chemistry and mass spectrometry.
      ,
      • Woo C.M.
      • Lund P.J.
      • Huang A.C.
      • Davis M.M.
      • Bertozzi C.R.
      • Pitteri S.J.
      Mapping and quantification of over 2000 O-linked glycopeptides in activated human T cells with isotope-targeted glycoproteomics (isotag).
      ). However, these same studies typically identify O-GlcNAc sites without subsequently predicting or attempting to ascribe functions to identified O-GlcNAc sites. This is crucial, as the presence of O-GlcNAc on a protein does not necessarily imply a functional consequence. For example, of the 11 O-GlcNAc sites identified on Keap1, only one—S104—modulates downstream NRF2 signaling in response to stress (
      • Chen P.H.
      • Smith T.J.
      • Wu J.
      • Siesser P.F.
      • Bisnett B.J.
      • Khan F.
      • et al.
      Glycosylation of KEAP1 links nutrient sensing to redox stress signaling.
      ). Similarly, the O-GlcNAcylation stoichiometry of the methionine cycle regulator adenosyl homocysteinase (AHCY) is predominantly contributed by a single O-GlcNAc site, whereas the remaining three sites make negligible contributions to AHCY O-GlcNAcylation stoichiometry and lack an identifiable function (
      • Zhu Q.
      • Cheng X.
      • Cheng Y.
      • Chen J.
      • Xu H.
      • Gao Y.
      • et al.
      O-GlcNAcylation regulates the methionine cycle to promote pluripotency of stem cells.
      ). Consequently, although the O-GlcNAcome stands at over 4000 proteins, it is difficult to predict if and how hypo-O-GlcNAcylation of specific O-GlcNAcylated neuronal proteins would alter protein function and subsequently contribute to the OGT-CDG phenotypes.
      In this study, a filter-based bioinformatics approach was used to predict the function of O-GlcNAc sites on neuronal proteins that have already been reported to be linked to ID. Utilizing O-GlcNAc site-mapping data in combination with sequence conservation, structural data, clinical databases, and literature review, we predict the existence of 38 functional O-GlcNAc sites across 22 neuronal proteins. In most cases, the role of the identified O-GlcNAc site is unknown. As these proteins are pathogenically mutated in reported ID/developmental delay (DD) syndromes and harbor putatively functional O-GlcNAc sites, their hypo-O-GlcNAcylation may contribute to OGT-CDG etiology. Additionally, by screening the ClinVar database against a recently compiled O-GlcNAc site database, we identify eight missense variants at O-GlcNAcylation sites in ID/DD patients, further suggesting a convergence of the O-GlcNAc proteome and ID etiology.

      Results and discussion

      A filter-based approach identifies OGT-CDG candidate conveyers

      To identify candidate conveyers of the OGT-CDG phenotypes, we used a filter-based approach integrating O-GlcNAc site-mapping data, available structural data, clinical databases, and the literature. O-GlcNAc site-mapping data were pooled manually from 14 studies that reported the O-GlcNAcomes of various cell lines (HEK293 (
      • Hahne H.
      • Sobotzki N.
      • Nyberg T.
      • Helm D.
      • Borodkin V.S.
      • van Aalten D.M.
      • et al.
      Proteome wide purification and identification of O-GlcNAc-modified proteins using click chemistry and mass spectrometry.
      ), mESCs (
      • Myers S.A.
      • Panning B.
      • Burlingame A.L.
      Polycomb repressive complex 2 is necessary for the normal site-specific O-GlcNAc distribution in mouse embryonic stem cells.
      ), HeLa (
      • Nandi A.
      • Sprung R.
      • Barma D.K.
      • Zhao Y.
      • Kim S.C.
      • Falck J.R.
      • et al.
      Global identification of O-GlcNAc-modified proteins.
      ), T-lymphocytes (
      • Woo C.M.
      • Lund P.J.
      • Huang A.C.
      • Davis M.M.
      • Bertozzi C.R.
      • Pitteri S.J.
      Mapping and quantification of over 2000 O-linked glycopeptides in activated human T cells with isotope-targeted glycoproteomics (isotag).
      )) and tissue extracts (rat brain (
      • Khidekel N.
      • Ficarro S.B.
      • Peters E.C.
      • Hsieh-Wilson L.C.
      Exploring the O-GlcNAc proteome: direct identification of O-GlcNAc-modified proteins from the brain.
      ), mouse brain (
      • Rexach J.E.
      • Rogers C.J.
      • Yu S.H.
      • Tao J.
      • Sun Y.E.
      • Hsieh-Wilson L.C.
      Quantification of O-glycosylation stoichiometry and dynamics using resolvable mass tags.
      ), and mouse synaptosomes (
      • Vosseller K.
      • Trinidad J.C.
      • Chalkley R.J.
      • Specht C.G.
      • Thalhammer A.
      • Lynn A.J.
      • et al.
      O-linked N-acetylglucosamine proteomics of postsynaptic density preparations using lectin weak affinity chromatography and mass spectrometry.
      ,
      • Trinidad J.C.
      • Barkan D.T.
      • Gulledge B.F.
      • Thalhammer A.
      • Sali A.
      • Schoepfer R.
      • et al.
      Global identification and characterization of both O-GlcNAcylation and phosphorylation at the murine synapse.
      ,
      • Chalkley R.J.
      • Thalhammer A.
      • Schoepfer R.
      • Burlingame A.L.
      Identification of protein O-GlcNAcylation sites using electron transfer dissociation mass spectrometry on native peptides.
      ,
      • Zaro B.W.
      • Yang Y.Y.
      • Hang H.C.
      • Pratt M.R.
      Chemical reporters for fluorescent detection and identification of O-GlcNAc-modified proteins reveal glycosylation of the ubiquitin ligase NEDD4-1.
      )). Next, we extracted O-GlcNAcylated proteins, where missense mutations in the corresponding coding gene were reported to give rise to ID/DD, using the list of ID/DD-associated genes compiled by Vissers et al. (
      • Vissers L.E.
      • Gilissen C.
      • Veltman J.A.
      Genetic studies in intellectual disability and related disorders.
      ) and the DECIPHER database (
      • Firth H.V.
      • Richards S.M.
      • Bevan A.P.
      • Clayton S.
      • Corpas M.
      • Rajan D.
      • et al.
      DECIPHER: database of chromosomal imbalance and phenotype in humans using ensembl resources.
      ) as reference points. The DECIPHER database reports on the pathogenicity and associated phenotype of variants in a given gene, and proteins passed this filter if their corresponding gene was the sole pathogenic variant in a patient with ID or DD. About 105 O-GlcNAcylated proteins, segregating with X-linked ID passed this first filter (see File S1 for list of O-GlcNAcylated X-linked ID proteins). Sequence conservation of O-GlcNAc sites was assessed from sequence alignments of higher eukaryotic orthologues (human, mouse, rat, rabbit, and sheep), leading to rejection of candidates where the O-GlcNAcylated Ser/Thr was poorly conserved (absent in ≥1 species). Subsequently, the available literature was used to investigate whether such conserved O-GlcNAc sites resided in functionally important domains or motifs of the protein. Examples of such functionally important domains or motifs include protein:protein interaction domains, protein:nucleic acid interaction domains, enzymatic active sites, or metal-binding sites. These filters reduced the number OGT-CDG candidate conveyers in the whole O-GlcNAcome to 22 candidate proteins (Fig. 1).
      Figure thumbnail gr1
      Figure 1Illustration of the rational, filter-based approach for identification of candidate conveyers of the OGT-CDG phenotypes. Site-mapping data for O-GlcNAcylated proteins identified in 14 O-GlcNAc mass spectrometric (O-GlcNAcomic) studies of mouse, rat, and human CNS tissue samples and cell lines were pooled and applied to the above illustrated, filter-based approach (see ). The relevant filters applied at each level are annotated. GRID2 (∗) was predicted to harbor both cotranslational and posttranslational O-GlcNAc sites on the basis of IUpred disorder prediction. CDG, congenital disorder of glycosylation; CNS, central nervous system; OGT, O-GlcNAc transferase.
      We further segregated O-GlcNAc sites on our 22 candidate proteins into two groups. Crystallographic studies of OGT in complex with substrate peptides have suggested that the active site and TPR domain of OGT accommodate disordered regions of proteins adopting an extended conformation (
      • Lazarus M.B.
      • Jiang J.
      • Kapuria V.
      • Bhuiyan T.
      • Janetzko J.
      • Zandberg W.F.
      • et al.
      HCF-1 is cleaved in the active site of O-GlcNAc transferase.
      ,
      • Rafie K.
      • Raimi O.
      • Ferenbach A.T.
      • Borodkin V.S.
      • Kapuria V.
      • van Aalten D.M.F.
      Recognition of a glycosylation substrate by the O-GlcNAc transferase TPR repeats.
      ,
      • Pathak S.
      • Alonso J.
      • Schimpl M.
      • Rafie K.
      • Blair D.E.
      • Borodkin V.S.
      • et al.
      The active site of O-GlcNAc transferase imposes constraints on substrate sequence.
      ),. However, several examples exist of O-GlcNAc sites residing in ordered regions of the protein fold and these studies do not yet explain why (
      • Trinidad J.C.
      • Barkan D.T.
      • Gulledge B.F.
      • Thalhammer A.
      • Sali A.
      • Schoepfer R.
      • et al.
      Global identification and characterization of both O-GlcNAcylation and phosphorylation at the murine synapse.
      ,
      • Zhu Q.
      • Cheng X.
      • Cheng Y.
      • Chen J.
      • Xu H.
      • Gao Y.
      • et al.
      O-GlcNAcylation regulates the methionine cycle to promote pluripotency of stem cells.
      ,
      • Lazarus M.B.
      • Jiang J.
      • Kapuria V.
      • Bhuiyan T.
      • Janetzko J.
      • Zandberg W.F.
      • et al.
      HCF-1 is cleaved in the active site of O-GlcNAc transferase.
      ). Importantly, Ser/Thr residues in ordered secondary structures would be inaccessible to the OGT active site. An example of this is AHCY, which is activated by O-GlcNAcylation at Thr136, despite this site residing in an α-helix that is theoretically inaccessible to the OGT active site (
      • Zhu Q.
      • Cheng X.
      • Cheng Y.
      • Chen J.
      • Xu H.
      • Gao Y.
      • et al.
      O-GlcNAcylation regulates the methionine cycle to promote pluripotency of stem cells.
      ). Recently, it has been discovered that O-GlcNAcylation can also occur cotranslationally on disordered peptidyl-tRNAs, as they emerge from the ribosome (
      • Zhu Y.
      • Liu T.W.
      • Cecioni S.
      • Eskandari R.
      • Zandberg W.F.
      • Vocadlo D.J.
      O-GlcNAc occurs cotranslationally to stabilize nascent polypeptide chains.
      ). It is possible that O-GlcNAc sites in inaccessible, ordered regions of the protein fold represent a subpopulation of O-GlcNAc sites that are cotranslationally modified by OGT. We therefore further segregated the list of candidate proteins/sites into those with accessible (in surface-exposed, disordered protein regions) and inaccessible (within sequestered, ordered regions of the protein fold) O-GlcNAc sites. To achieve this, we used available Protein Data Bank (PDB) structures of proteins or where such data were unavailable, the disorder predictor IUpred2A (
      • Erdos G.
      • Dosztanyi Z.
      Analyzing protein disorder with IUPred2A.
      ). Overall, our filter-based approach (illustrated in Fig. 1) suggests that of the 22 candidate conveyors of the OGT-CDG phenotypes, 13 are posttranslationally O-GlcNAcylated, eight cotranslationally O-GlcNAcylated, and one candidate (GRID2) harbors both cotranslational and posttranslational O-GlcNAc sites. For summaries of each OGT-CDG candidate and details regarding conserved O-GlcNAc site(s) of interest, see Table S1.
      Of the 39 candidate O-GlcNAc sites, the majority (∼58%) were predicted to be functional based on their localization to a protein:protein binding interface. Four candidate O-GlcNAc sites were either reported as being phosphorylated or SUMOylated or as proximal (±10 residues) to a functional phosphorylation/SUMOylation site (Fig. 2A). The remaining sites were either located in enzyme active sites (RSK-2, GNAS), a regulatory domain/motif (KAT6A, SMAD4) or a protein:nucleic acid interface (ADNP, ARID1A; Fig. 2A). Classification of OGT-CDG candidate conveyers by protein class, using PANTHER (Protein ANalysis THrough Evolutionary Relationships) (42), showed a substantial portion of candidates were cytoskeletal or microtubule-associated proteins (19%; TUBB2B; ACTB; ACTG1; DYNC1H1; Fig. 2B). Other classes of protein present in our candidate list were transcriptional regulators (SOX2, SMAD4), chromatin binding or regulatory proteins (P300, KAT6A), and metabolic enzymes (ACO2; SUCLA2). Transmembrane signaling receptors (GRID2), protein modifying enzymes (RSK-2), and nucleic acid metabolism enzymes (MECP2) were also identified (Fig. 2B). Seven candidates were not annotated as belonging to a single protein class. Overall, our OGT-CDG candidate conveyers represent a diverse pool of proteins.
      Figure thumbnail gr2
      Figure 2Classification of OGT-CDG candidate conveyers by protein class, positioning of O-GlcNAc site(s), and convergent signaling pathways and biological processes. A, percentage of O-GlcNAc sites on OGT-CDG candidate conveyers, which were predicted to be functional based on either: (I) present on a protein:protein interface; (II) adjacent (±10 residues) from another PTM site, which has been validated as functional in previous work; (III) within or directly abutting against an enzyme active site; and (IV) present within a known regulatory region. B, stratification of OGT-CDG candidate conveyers by class of protein using PANTHER, UniProt accessions of human homologs were used (see ). C, classification of biological processes involving OGT-CDG candidate conveyers using the Gene Ontology (GO) resource. GO terms were clustered manually into common groups and highlighted. D, signaling pathways involving OGT-CDG candidate conveyers were extracted using PANTHER analysis of Gene IDs, which were manually retrieved from NCBI. CDG, congenital disorder of glycosylation; OGT, O-GlcNAc transferase.
      In addition to conserved O-GlcNAc sites residing in functional domains or regions of ID/DD-associated proteins, we also investigated whether poorly conserved O-GlcNAc sites on ID/DD-associated proteins may reside in interesting regions of a given protein. Toward this end, we analyzed whether the 105 ID/DD-associated O-GlcNAc proteins identified after the first filter harbored poorly conserved O-GlcNAc sites (O-GlcNAc Ser/Thr not conserved in ≥ 1 species) in functional domains/regions of the protein. Four hits (NDUFS3, aprataxin, PDHA1, and ankyrin G) were identified due to O-GlcNAc sites either localizing to mitochondrial targeting sequences (PDHA1, NDUFS3), the interaction site for PARP-1 (aprataxin), or an isoform-specific domain associated with regulation of protein localization (ankyrin G). These lower confidence hits are summarized in Table S2.
      We subsequently used gene ontology (GO) analysis to assess possible convergence of candidate conveyers on a small set of biological processes and signaling pathways. Most of the candidates were annotated as being involved broadly in development and neurogenesis (Fig. 2C). Intriguingly, we also observed that some of our candidate conveyers were annotated as forming part of the stress response (DDX3X, DYNC1H1; see discussion on DDX3X later), a process known to be regulated by O-GlcNAcylation (
      • Ohn T.
      • Kedersha N.
      • Hickman T.
      • Tisdale S.
      • Anderson P.
      A functional RNAi screen links O-GlcNAc modification of ribosomal proteins to stress granule and processing body assembly.
      ). Additional biological processes included cell cycle regulation, chromatin organization, regulation of gene expression, and protein modification (acetylation/ubiquitination, Fig. 2C). For all these biological processes, previous work has strongly suggested a functional role for O-GlcNAcylation in their regulation (
      • Capotosti F.
      • Guernier S.
      • Lammers F.
      • Waridel P.
      • Cai Y.
      • Jin J.
      • et al.
      O-GlcNAc transferase catalyzes site-specific proteolysis of HCF-1.
      ,
      • Chen P.H.
      • Smith T.J.
      • Wu J.
      • Siesser P.F.
      • Bisnett B.J.
      • Khan F.
      • et al.
      Glycosylation of KEAP1 links nutrient sensing to redox stress signaling.
      ,
      • Zhu Y.
      • Liu T.W.
      • Cecioni S.
      • Eskandari R.
      • Zandberg W.F.
      • Vocadlo D.J.
      O-GlcNAc occurs cotranslationally to stabilize nascent polypeptide chains.
      ,
      • Ohn T.
      • Kedersha N.
      • Hickman T.
      • Tisdale S.
      • Anderson P.
      A functional RNAi screen links O-GlcNAc modification of ribosomal proteins to stress granule and processing body assembly.
      ,
      • Sakabe K.
      • Wang Z.
      • Hart G.W.
      Beta-N-acetylglucosamine (O-GlcNAc) is part of the histone code.
      ,
      • Ranuncolo S.M.
      • Ghosh S.
      • Hanover J.A.
      • Hart G.W.
      • Lewis B.A.
      Evidence of the involvement of O-GlcNAc-modified human RNA polymerase II CTD in transcription in vitro and in vivo.
      ). We further carried out pathway analysis using PANTHER and found that the candidate conveyers were involved in a number of signaling pathways. Of particular interest, we noted that Dpp signaling, Wnt signaling, and synaptic vesicle trafficking were hits (Fig. 2D). In D. melanogaster, Dpp (BMP4 in vertebrates) signaling is negatively regulated at the level of the bone morphogenic protein (BMP) receptor Sax by O-GlcNAcylation (
      • Moulton M.J.
      • Humphreys G.B.
      • Kim A.
      • Letsou A.
      O-GlcNAcylation dampens Dpp/BMP signaling to ensure proper Drosophila embryonic development.
      ), and the stability of Wnt signaling nodes CK2 and β-catenin are negatively and positively regulated by O-GlcNAc, respectively (
      • Olivier-Van Stichelen S.
      • Dehennaut V.
      • Buzy A.
      • Zachayus J.L.
      • Guinez C.
      • Mir A.M.
      • et al.
      O-GlcNAcylation stabilizes β-catenin through direct competition with phosphorylation at threonine 41.
      ,
      • Tarrant M.K.
      • Rho H.S.
      • Xie Z.
      • Jiang Y.L.
      • Gross C.
      • Culhane J.C.
      • et al.
      Regulation of CK2 by phosphorylation and O-GlcNAcylation revealed by semisynthesis.
      ). O-GlcNAcylation of Synapsin I, which is also identified as a candidate conveyer in this study (Table S1), regulates presynaptic vesicle dynamics and the size of the reserve pool of synaptic vesicles (
      • Skorobogatko Y.
      • Landicho A.
      • Chalkley R.J.
      • Kossenkov A.V.
      • Gallo G.
      • Vosseller K.
      O-linked β-N-acetylglucosamine (O-GlcNAc) site thr-87 regulates synapsin I localization to synapses and size of the reserve pool of synaptic vesicles.
      ). Thus, the OGT-CDG candidate conveyers identified here are involved in diverse biological processes and signaling pathways, many of which are already known to be regulated by O-GlcNAcylation. Future work, characterizing the function of O-GlcNAc on these candidate proteins, guided by the hypotheses we have presented, may reveal additional levels of O-GlcNAc-dependent regulation in these processes. Later, we discuss selected examples of candidate conveyers and their associated O-GlcNAc sites of interest.

      O-GlcNAcylation of cytoskeletal proteins suggests that dysregulated cytoskeletal dynamics may contribute to OGT-CDG pathology

      Precise regulation of cytoskeletal dynamics is essential for the ability of initially spherical cell soma to give rise to distinct somatodendritic and axonal compartments (
      • Flynn K.C.
      • Bradke F.
      Role of the cytoskeleton and membrane trafficking in axon–dendrite morphogenesis.
      ). This process, termed neuronal polarization, is essential for the correct flow of electrochemical impulses and the trafficking of RNA, ribosomes, and ion channels to distal synapses (
      • Craig A.M.
      • Banker G.
      Neuronal polarity.
      ). Indeed, disruption of cytoskeletal dynamics leads to neurodevelopmental disorders such as lissencephaly and cortical dysplasia (
      • Jaglin X.H.
      • Poirier K.
      • Saillour Y.
      • Buhler E.
      • Tian G.
      • Bahi-Buisson N.
      • et al.
      Mutations in the beta-tubulin gene TUBB2B result in asymmetrical polymicrogyria.
      ). Regarding cytoskeletal proteins, our bioinformatic analysis predicted O-GlcNAc sites on both ß-tubulin (isoform TUBB2B) and actin (isoforms ACTB1 and ACTG1) to be important for regulating their respective proteins’ functions (see later).
      In our bioinformatic analysis, we identified S172 as a highly conserved ß-tubulin O-GlcNAc site. S172L and S172P mutations segregate with cortical dysplasia, complex, with other brain malformations, seven in several ID patients (
      • Jaglin X.H.
      • Poirier K.
      • Saillour Y.
      • Buhler E.
      • Tian G.
      • Bahi-Buisson N.
      • et al.
      Mutations in the beta-tubulin gene TUBB2B result in asymmetrical polymicrogyria.
      ) (CDCBM7; See section “Identification of pathogenic missense variants at O-GlcNAcylation sites” and File S2). Additionally, S172 resides on a loop that forms part of the GTPase active site (Fig. 3A) (
      • Fourest-Lieuvin A.
      • Peris L.
      • Gache V.
      • Garcia-Saez I.
      • Juillan-Binard C.
      • Lantez V.
      • et al.
      Microtubule regulation in mitosis: tubulin phosphorylation by the cyclin-dependent kinase Cdk1.
      ,
      • Löwe J.
      • Li H.
      • Downing K.H.
      • Nogales E.
      Refined structure of alpha beta-tubulin at 3.5 A resolution.
      ). Phosphorylation of S172 by CDK1, during mitosis, prevents GTP/GDP exchange, presumed to be due to hydrogen bonding interactions between the phosphate and GDP preventing GDP dissociation from the active site (
      • Fourest-Lieuvin A.
      • Peris L.
      • Gache V.
      • Garcia-Saez I.
      • Juillan-Binard C.
      • Lantez V.
      • et al.
      Microtubule regulation in mitosis: tubulin phosphorylation by the cyclin-dependent kinase Cdk1.
      ). The presence of a bulky GlcNAc moiety in the GTPase active site would likely sterically hinder GTP binding, thus reducing the pool of GTP-bound tubulin and inhibiting tubulin polymerization. Interestingly, Ji. et al.(
      • Ji S.
      • Kang J.G.
      • Park S.Y.
      • Lee J.
      • Oh Y.J.
      • Cho J.W.
      O-GlcNAcylation of tubulin inhibits its polymerization.
      ) previously observed that O-GlcNAcylated tubulin is exclusively monomeric, supporting the hypothesis that tubulin O-GlcNAcylation prevents microtubule assembly, though the functional sites of tubulin O-GlcNAcylation were not identified in this study. If S172 O-GlcNAcylation does regulate tubulin polymerization indirectly, through modulation of GTP binding affinity to tubulin, then hypo-O-GlcNAcylation of tubulin in the context of OGT-CDG could lead to increased rates of tubulin polymerization. This increased rate of tubulin polymerization, in-turn, could deplete the pool of monomeric tubulin, altering the structural plasticity of the microtubules in neurons and potentially preventing morphological changes in the dendrites and axons that are crucial for neuronal polarization (
      • Flynn K.C.
      • Bradke F.
      Role of the cytoskeleton and membrane trafficking in axon–dendrite morphogenesis.
      ). This hypothesis could be investigated by an in vitro polymerization assay, comparing TUBB2BWT and TUBB2BS172A polymerization rates (
      • Shelanski M.L.
      • Gaskin F.
      • Cantor C.R.
      Microtubule assembly in the absence of added nucleotides.
      ).
      Figure thumbnail gr3
      Figure 3The β-tubulin O-GlcNAc site S172 and associated backbone angles. A, crystal structure of β-tubulin from Lowe et al. (PDB entry: 1JFF (
      • Löwe J.
      • Li H.
      • Downing K.H.
      • Nogales E.
      Refined structure of alpha beta-tubulin at 3.5 A resolution.
      ) with S172 and bound GDP highlighted as ball and stick depictions. Note how the S172 hydroxyl points directly toward the bound GDP in the GTPase active site. B, measured dihedral angles for S172 (Phi = −57 °, Psi = 147 °). P171, S172, and P173 are highlighted as ball and stick depictions.
      In addition to the microtubule networks in neurons, the actin cytoskeleton plays an essential role in regulating neuronal polarization, and fine-tuning actin cytoskeleton dynamics and function severely impacts both axon differentiation (
      • Bradke F.
      • Dotti C.G.
      The role of local actin instability in axon formation.
      ) and growth cone motility (
      • Gomez T.M.
      • Letourneau P.C.
      Actin dynamics in growth cone motility and navigation.
      ) prior to synaptogenesis. Pharmacological inhibition of actin polymerization induces spontaneous axon differentiation from immature neurites (
      • Bradke F.
      • Dotti C.G.
      The role of local actin instability in axon formation.
      ). Furthermore, heterozygous actin loss-of-function variants have been reported in patients displaying ID, developmental delay, internal organ malformation, and growth retardation (
      • Cuvertino S.
      • Stuart H.M.
      • Chandler K.E.
      • Roberts N.A.
      • Armstrong R.
      • Bernardini L.
      • et al.
      ACTB loss-of-function mutations result in a pleiotropic developmental disorder.
      ). The highly conserved actin O-GlcNAc site S365 lies on the interface between actin and its regulatory protein profilin (
      • Schutt C.E.
      • Myslik J.C.
      • Rozycki M.D.
      • Goonesekere N.C.
      • Lindberg U.
      The structure of crystalline profilin-beta-actin.
      ). The S365 hydroxyl group points toward a E364-K126 salt bridge on the actin:profilin interface, with the S365 hydroxyl and E364 carboxylate only 3.3 Å apart (see PDB entry 2BTF (
      • Schutt C.E.
      • Myslik J.C.
      • Rozycki M.D.
      • Goonesekere N.C.
      • Lindberg U.
      The structure of crystalline profilin-beta-actin.
      )). A bulky O-GlcNAc moiety on S365 could therefore disrupt formation of this salt bridge to inhibit actin:profilin binding. Profilins stimulate the exchange of GDP for GTP on actin, thus regulating the size of the pool of polymerizable GTP-actin monomers (
      • Davey R.J.
      • Moens P.D.
      Profilin: many facets of a small protein.
      ). Knockdown of profilin IIa in mice leads to increased neurite outgrowth, ostensibly due to reduced actin polymerization and reduced F-actin density (
      • Da Silva J.S.
      • Medina M.
      • Zuliani C.
      • Di Nardo A.
      • Witke W.
      • Dotti C.G.
      RhoA/ROCK regulation of neuritogenesis via profilin IIa-mediated control of actin stability.
      ). Whether actin S365 O-GlcNAcylation inhibits actin polymerization through reduced affinity of actin for profilin and whether hypo-O-GlcNAcylation of actin at S365 alters actin polymerization in the context of OGT-CDG, will require further investigation.
      In addition to S365, our bioinformatic analysis of the O-GlcNAcome identified S265 as a previously reported, conserved O-GlcNAc site on actin, which resides in a hydrophobic loop (residues 264–273). In the Holmes model of F-actin, this hydrophobic loop inserts into a pocket formed by an adjacent F-actin protomer in a manner analogous to a “socket” and a “plug,” with the loop constituting the plug and the socket formed by hydrophobic residues in the adjacent promoter (
      • Holmes K.C.
      • Popp D.
      • Gebhard W.
      • Kabsch W.
      Atomic model of the actin filament.
      ). Holmes et al. (
      • Holmes K.C.
      • Popp D.
      • Gebhard W.
      • Kabsch W.
      Atomic model of the actin filament.
      ) hypothesized that loss of this interaction between adjacent actin protomers would likely severely disrupt actin polymerization, as other electrostatic interactions along the length of F-actin may not be sufficient to compensate for loss of the hydrophobic interactions contributed by residues 264 to 273. The presence of O-GlcNAc on S265 therefore presents an interesting target for study and suggests that O-GlcNAc may regulate actin polymerization both directly (through disruption of actin protomer assembly) and indirectly (through reduced affinity for profilin).

      O-GlcNAcylation of DDX3X implicates dysregulated RNA metabolism in OGT-CDG etiology

      The filter approach identified a member of the DEAD box RNA helicase family—DEAD box RNA helicase 3 X-linked (DDX3X)—as a putative conveyer of the OGT-CDG phenotype. DDX3X is an ATP-dependent RNA helicase, which assembles into trimers capable of unwinding structured mRNA 5′ UTRs (
      • Floor S.N.
      • Condon K.J.
      • Sharma D.
      • Jankowsky E.
      • Doudna J.A.
      Autoinhibitory interdomain interactions and subfamily-specific extensions redefine the catalytic core of the human DEAD-box protein DDX3.
      ,
      • Putnam A.A.
      • Gao Z.
      • Liu F.
      • Jia H.
      • Yang Q.
      • Jankowsky E.
      Division of labor in an oligomer of the DEAD-box RNA helicase Ded1p.
      ). This local strand separation is essential for docking and scanning of mRNA for translation initiation sites, and DDX3X-mediated RNA unwinding is crucial for the translation of cell cycle progression genes (
      • Lai M.C.
      • Chang W.C.
      • Shieh S.Y.
      • Tarn W.Y.
      DDX3 regulates cell growth through translational control of cyclin E1.
      ) and components of the Rac1-PKA signaling axis required for neurite outgrowth (See Fig. 4) (
      • Chen H.H.
      • Yu H.I.
      • Tarn W.Y.
      DDX3 modulates neurite development via translationally activating an RNA regulon involved in Rac1 activation.
      ). Furthermore, DDX3X knockdown in bone marrow–derived macrophages reduces stress granule formation in response to oxidative stress by ∼50% (
      • Samir P.
      • Kesavardhana S.
      • Patmore D.M.
      • Gingras S.
      • Malireddi R.K.S.
      • Karki R.
      • et al.
      DDX3X acts as a live-or-die checkpoint in stressed cells by regulating NLRP3 inflammasome.
      ). Stress granules are translationally repressive aggregates of ribonucleoproteins formed in response to cellular stresses such as glucose deprivation, osmotic stress, or oxidative stress to protect untranslated mRNAs from damage (
      • Protter D.S.W.
      • Parker R.
      Principles and properties of stress granules.
      ). Curiously, OGT knockdown in U2OS cells also abolishes stress granule formation following arsenite-induced osmotic stress (
      • Ohn T.
      • Kedersha N.
      • Hickman T.
      • Tisdale S.
      • Anderson P.
      A functional RNAi screen links O-GlcNAc modification of ribosomal proteins to stress granule and processing body assembly.
      ), and mass spectrometric studies have identified at least nine DEAD box RNA helicase as being O-GlcNAcylated (
      • Woo C.M.
      • Lund P.J.
      • Huang A.C.
      • Davis M.M.
      • Bertozzi C.R.
      • Pitteri S.J.
      Mapping and quantification of over 2000 O-linked glycopeptides in activated human T cells with isotope-targeted glycoproteomics (isotag).
      ). However, exactly how OGT and stress granule formation are mechanistically linked is presently unknown (see Fig. 4 for illustrated hypothesis). Furthermore, missense mutations in the ddx3x gene are linked with a clinically heterogenous ID syndrome comprising of, though not limited to, developmental delay, hypotonia, corpus callosum hypoplasia, epilepsy, behavioral problems, microcephaly, and abnormal gait (
      • Snijders Blok L.
      • Madsen E.
      • Juusola J.
      • Gilissen C.
      • Baralle D.
      • Reijnders M.R.
      • et al.
      Mutations in DDX3X are a common cause of unexplained intellectual disability with gender-specific effects on Wnt signaling.
      ).
      Figure thumbnail gr4
      Figure 4The DDX3X O-GlcNAc sites S584 and S588, and hypothesized functions of DDX3X O-GlcNAcylation in RNA metabolism and neurodevelopment. Illustrated summary of hypothesized function(s) of DDX3X O-GlcNAcylation. Briefly, DDX3X O-GlcNAcylation in the CTE at residues S584 and S588 may promote formation of functional DDX3X trimers. Functional DDX3X trimers subsequently promote the translation of signaling nodes in the Rac1/PKA and cyclin E1/CDK5 signaling axes. Additionally, in response to cellular stress, DDX3X and at least nine other DEAD-box RNA helicases (∗) may be O-GlcNAcylated as part of the stress response. CTE, C-terminal extension.
      We identified DDX3X as a candidate conveyer of OGT-CDG based on the presence of two highly conserved O-GlcNAc sites in the DDX3X C-terminal extension (CTE)—S584 and S588. The CTE is a highly conserved 26-residue stretch that is essential for DDX3X RNA unwinding activity (
      • Floor S.N.
      • Condon K.J.
      • Sharma D.
      • Jankowsky E.
      • Doudna J.A.
      Autoinhibitory interdomain interactions and subfamily-specific extensions redefine the catalytic core of the human DEAD-box protein DDX3.
      ). Deletion of this same region in the Saccharomyces cerevisiae ortholog Ded1p abolishes Ded1p trimerization (
      • Putnam A.A.
      • Gao Z.
      • Liu F.
      • Jia H.
      • Yang Q.
      • Jankowsky E.
      Division of labor in an oligomer of the DEAD-box RNA helicase Ded1p.
      ), suggesting that the DDX3X CTE is required for association of DDX3X into functional trimers. Additionally, the CTE forms the majority of the region required for interaction with the mRNA export receptor TAP (residues 582–631) (
      • Lai M.C.
      • Lee Y.H.
      • Tarn W.Y.
      The DEAD-box RNA helicase DDX3 associates with export messenger ribonucleoproteins as well as tip-associated protein and participates in translational control.
      ). The presence of two highly conserved O-GlcNAc sites in the CTE suggests that O-GlcNAcylation of DDX3X could modulate DDX3X trimerization and/or RNA unwinding activity. Alternatively, DDX3X O-GlcNAcylation could modulate nuclear mRNA export by altering the affinity of DDX3X for TAP. Both S584, S588, and the surrounding sequence are highly conserved in eukaryotes despite the region being intrinsically disordered (see File S1 for IUPred2A scores), further suggesting these residues are important for regulating DDX3X function.

      Hypo-O-GlcNAcylation of transcription factors may lead to transcriptomic dysregulation in OGT-CDG

      The link between O-GlcNAcylation and regulation of eukaryotic gene expression has been a source of considerable interest, with several components of the transcription/translation machinery being O-GlcNAc modified (
      • Ranuncolo S.M.
      • Ghosh S.
      • Hanover J.A.
      • Hart G.W.
      • Lewis B.A.
      Evidence of the involvement of O-GlcNAc-modified human RNA polymerase II CTD in transcription in vitro and in vivo.
      ,
      • Zeidan Q.
      • Wang Z.
      • De Maio A.
      • Hart G.W.
      O-GlcNAc cycling enzymes associate with the translational machinery and modify core ribosomal proteins.
      ,
      • Zhang X.
      • Shu X.E.
      • Qian S.B.
      O-GlcNAc modification of eIF4GI acts as a translational switch in heat shock response.
      ). For example, O-GlcNAcylation of RNA polymerase II and eIF4G modulates transcription and translation, respectively (
      • Ranuncolo S.M.
      • Ghosh S.
      • Hanover J.A.
      • Hart G.W.
      • Lewis B.A.
      Evidence of the involvement of O-GlcNAc-modified human RNA polymerase II CTD in transcription in vitro and in vivo.
      ,
      • Zhang X.
      • Shu X.E.
      • Qian S.B.
      O-GlcNAc modification of eIF4GI acts as a translational switch in heat shock response.
      ). Of the candidates annotated as signaling nodes in neurogenesis (see GO analysis, Fig. 2C), the transcriptional regulator Methyl CpG binding protein 2 (MECP2), is a particularly interesting hit. MECP2 mutations are the predominant cause of Rett’s syndrome, an autistic spectrum disorder characterized by postnatal developmental regression, loss of motor skills, dystonia, loss of language/communication skills, and stereotypic hand movements (
      • Gold W.A.
      • Krishnarajy R.
      • Ellaway C.
      • Christodoulou J.
      Rett syndrome: a genetic update and clinical review focusing on comorbidities.
      ). In resting neurons, MECP2 represses transcription at loci containing methylated CpG dinucleotides. This transcriptional repression is mediated via the scaffolding function of the MECP2 transcriptional repression domain, which recruits mSin3A, N-CoR, and c-Ski to silence transcription (
      • Bellini E.
      • Pavesi G.
      • Barbiero I.
      • Bergo A.
      • Chandola C.
      • Nawaz M.S.
      • et al.
      MeCP2 post-translational modifications: a mechanism to control its involvement in synaptic plasticity and homeostasis?.
      ). Following neuronal depolarization, an unknown kinase phosphorylates MECP2 at S423 and S426, leading to dissociation of MECP2 from methylated loci and up-regulated expression of the neurotrophin BDNF (
      • Li H.
      • Zhong X.
      • Chau K.F.
      • Williams E.C.
      • Chang Q.
      Loss of activity-induced phosphorylation of MeCP2 enhances synaptogenesis, LTP and spatial memory.
      ). S426 is also a conserved site of O-GlcNAcylation on MECP2. Mice expressing a MECP2S421A;S424A (S423 and S426 in human) knock-in display enhanced hippocampal-dependent learning and memory in a Morris water maze test and altered gene expression patterns (
      • Li H.
      • Zhong X.
      • Chau K.F.
      • Williams E.C.
      • Chang Q.
      Loss of activity-induced phosphorylation of MeCP2 enhances synaptogenesis, LTP and spatial memory.
      ). MECP2 O-GlcNAcylation has already been reported to be antagonistic to phosphorylation, with glucosamine supplementation elevating O-GlcNAcylation stoichiometry and reducing total MECP2 phosphorylation (
      • Rexach J.E.
      • Rogers C.J.
      • Yu S.H.
      • Tao J.
      • Sun Y.E.
      • Hsieh-Wilson L.C.
      Quantification of O-glycosylation stoichiometry and dynamics using resolvable mass tags.
      ). However, the extent to which MECP2 S426 O-GlcNAcylation antagonizes neuronal depolarization-dependent changes in MECP2 phosphorylation and BDNF expression has yet to be investigated (See Fig. 5 for an illustrated summary of the previous discussion). Furthermore, it remains to be investigated whether hypo-O-GlcNAcylation of S426 in mature neurons dysregulates neuronal depolarization-dependent gene expression in the context of OGT-CDG.
      Figure thumbnail gr5
      Figure 5Summary of the proposed functions of O-GlcNAc on selected examples of OGT-CDG candidate conveyers. Hypothesized functions of O-GlcNAc on the OGT-CDG candidate conveyers are illustrated here. OGT-CDG candidate conveyers are annotated with a blue square, highlighting their O-GlcNAcylation status. Where expression of specific genes is expected to be affected by candidate O-GlcNAcylation status, these same genes are listed vertically in italics along with the expected change in expression following O-GlcNAcylation of the corresponding candidate protein. CDG, congenital disorder of glycosylation; OGT, O-GlcNAc transferase.
      In addition to the genetic landscape of differentiated neurons, disrupted gene expression patterns early in development could also contribute to OGT-CDG etiology. SOX2, a key pluripotency marker in embryonic stem cells (ESCs), was identified as a candidate conveyer of the OGT-CDG phenotype on account of the proximity of the O-GlcNAc site S246 to the SUMOylation site K247. SUMOylation of K247 reduces SOX2-dependent gene expression from the Fgf4 promoter due to reduced SOX2 DNA binding (
      • Tsuruzoe S.
      • Ishihara K.
      • Uchimura Y.
      • Watanabe S.
      • Sekita Y.
      • Aoto T.
      • et al.
      Inhibition of DNA binding of Sox2 by the SUMO conjugation.
      ). Intriguingly, work by Myers et al.(
      • Myers S.A.
      • Peddada S.
      • Chatterjee N.
      • Friedrich T.
      • Tomoda K.
      • Krings G.
      • et al.
      SOX2 O-GlcNAcylation alters its protein-protein interactions and genomic occupancy to modulate gene expression in pluripotent cells.
      ) found that mutagenesis of S246 to alanine in mESCs altered SOX2 genomic occupancy and increased SOX2 interaction with PARP1 (see later). Importantly, SOX2 loss-of-function variants give rise to an ID syndrome characterized by anophthalmia/microphthalmia, as well as developmental delay and increased risk of intracranial teratomas, indicating that hypo-O-GlcNAcylation of this essential pluripotency factor could contribute to OGT-CDG (
      • Blackburn P.R.
      • Chacon-Camacho O.F.
      • Ortiz-González X.R.
      • Reyes M.
      • Lopez-Uriarte G.A.
      • Zarei S.
      • et al.
      Extension of the mutational and clinical spectrum of SOX2 related disorders: description of six new cases and a novel association with suprasellar teratoma.
      ).
      During retinoic acid–induced ESC differentiation, Erk1/2 phosphorylates PARP1 (
      • Kauppinen T.M.
      • Chan W.Y.
      • Suh S.W.
      • Wiggins A.K.
      • Huang E.J.
      • Swanson R.A.
      Direct phosphorylation and regulation of poly(ADP-ribose) polymerase-1 by extracellular signal-regulated kinases 1/2.
      ). This phosphorylation event stimulates auto-PARylation of PARP1 and promotes PARP1–SOX2 complex formation (
      • Lai Y.S.
      • Chang C.W.
      • Pawlik K.M.
      • Zhou D.
      • Renfrow M.B.
      • Townes T.M.
      SRY (sex determining region Y)-box2 (Sox2)/poly ADP-ribose polymerase 1 (Parp1) complexes regulate pluripotency.
      ). In this complex, SOX2 is unable to bind Oct4-Sox enhancers and upregulate expression of key pluripotency genes including Nanog, Oct4, Fgf4, and Sox2, thus stimulating differentiation (
      • Lai Y.S.
      • Chang C.W.
      • Pawlik K.M.
      • Zhou D.
      • Renfrow M.B.
      • Townes T.M.
      SRY (sex determining region Y)-box2 (Sox2)/poly ADP-ribose polymerase 1 (Parp1) complexes regulate pluripotency.
      ). These data are consistent with reports that SOX2 O-GlcNAcylation decreases with differentiation, and loss of O-GlcNAc on SOX2 increases PARP1 interaction, leading to reduced expression of genes required for maintenance of pluripotency (
      • Myers S.A.
      • Peddada S.
      • Chatterjee N.
      • Friedrich T.
      • Tomoda K.
      • Krings G.
      • et al.
      SOX2 O-GlcNAcylation alters its protein-protein interactions and genomic occupancy to modulate gene expression in pluripotent cells.
      ). Whether PARP1–SOX2 complex formation is reduced during early stages of embryogenesis in OGT-CDG patients or in previously generated ESC models of the N567K and N648Y OGT mutations, has yet to be investigated. If complex stoichiometry is increased prematurely during embryogenesis, then premature ESC differentiation and dysregulated SOX2-dependent gene expression could contribute to the OGT-CDG phenotype.
      In addition to transcription factors, a member of the histone acetyl transferase family of transcriptional coregulators—P300—was identified as a candidate conveyer of OGT-CDG in our analysis. P300 acetylates lysine residues on histones to facilitate chromatin unwinding and upregulate gene expression. During DNA damage, P300 is recruited to P53, where P300 upregulates P53-dependent transcription via histone acetylation and acetylation of five C-terminal lysine residues on P53 (
      • Gu W.
      • Roeder R.G.
      Activation of p53 sequence-specific DNA binding by acetylation of the p53 C-terminal domain.
      ). The latter event activates the DNA-binding activity of P53 (
      • Gu W.
      • Roeder R.G.
      Activation of p53 sequence-specific DNA binding by acetylation of the p53 C-terminal domain.
      ). P300 mutations are the second most common cause of Rubinstein-Taybi syndrome, which manifests in mild-to-moderate ID, growth delay, skeletal abnormalities, and facial dysmorphia (
      • Negri G.
      • Magini P.
      • Milani D.
      • Colapietro P.
      • Rusconi D.
      • Scarano E.
      • et al.
      From whole gene deletion to point mutations of EP300-positive rubinstein-taybi patients: new insights into the mutational spectrum and peculiar clinical hallmarks.
      ). This study identified S1734 as a conserved O-GlcNAc site residing on the interface of the P300 Taz2 domain with the P53 N-terminal transactivation domain (TAD) (
      • Miller Jenkins L.M.
      • Feng H.
      • Durell S.R.
      • Tagad H.D.
      • Mazur S.J.
      • Tropea J.E.
      • et al.
      Characterization of the p300 Taz2-p53 TAD2 complex and comparison with the p300 Taz2-p53 TAD1 complex.
      ). The P300 Taz2–P53 TAD interaction is predominantly mediated by hydrophobic contacts, though electrostatic and polar interactions between several arginine, glutamine, and serine residues also contribute to complex formation (
      • Miller Jenkins L.M.
      • Feng H.
      • Durell S.R.
      • Tagad H.D.
      • Mazur S.J.
      • Tropea J.E.
      • et al.
      Characterization of the p300 Taz2-p53 TAD2 complex and comparison with the p300 Taz2-p53 TAD1 complex.
      ). S1734 forms hydrogen bonds to both T18 and Q17 of P53, stabilizing the complex. Phosphorylation of T18 on P53 increases the affinity of the Taz2–TAD complex 11-fold, indicating this residue is particularly important for modulating the affinity of P53 for P300 (
      • Miller Jenkins L.M.
      • Feng H.
      • Durell S.R.
      • Tagad H.D.
      • Mazur S.J.
      • Tropea J.E.
      • et al.
      Characterization of the p300 Taz2-p53 TAD2 complex and comparison with the p300 Taz2-p53 TAD1 complex.
      ). The presence of a bulky GlcNAc moiety could be hypothesized to act antagonistically to T18 phosphorylation, by inducing a steric clash between S1734 and T18/Q17 on the Taz2–TAD interface. This in turn would likely exert negative downstream effects on P53-dependent transcriptional activation due to reduced P300 recruitment and subsequent reductions in the levels of histone acetylation. Overall, P300 O-GlcNAcylation at S1734 may negatively regulate P300–P53 complex formation, as indicated by prior structural studies showing a close proximity of the S1734 hydroxyl group to the regulatory P53 phospho-site T18 and Q17 (See Figure 5 for an illustrated summary of the previous discussion). Hypo-O-GlcNAcylation of this site in the context of OGT-CDG could potentially lead to dysregulated P53 activity, which is essential for cell cycle arrest and apoptosis in response to DNA damage. To investigate the functional effects of O-GlcNAc on P300, in vitro acetylation of P53 by S1734 O-GlcNAcylated P300 and GlcNAc-deficient P300 could be analyzed as described previously (
      • Gu W.
      • Roeder R.G.
      Activation of p53 sequence-specific DNA binding by acetylation of the p53 C-terminal domain.
      ).

      Identification of pathogenic missense variants at O-GlcNAcylation sites

      So far, this study has focused on predicting functional O-GlcNAc sites on ID/DD-associated proteins using the filter-based approach outlined in Figure 1. Because mutations in these proteins segregate with ID/DD, it is also possible that hypo-O-GlcNAcylation (i.e., mutations of Ser/Thr to other amino acids) of functional O-GlcNAc sites on these proteins may contribute to OGT-CDG pathology. We explored this possibility across the entire O-GlcNAcome (i.e., not limited to the candidates in Fig. 1).
      To identify mutations at O-GlcNAc sites that segregate with ID/DD, the ClinVar database (
      • Landrum M.J.
      • Lee J.M.
      • Benson M.
      • Brown G.R.
      • Chao C.
      • Chitipiralla S.
      • et al.
      ClinVar: Improving access to variant interpretations and supporting evidence.
      ) was screened for missense variants using the O-GlcNAc database (
      • Wulff-Fuentes E.
      • Berendt R.R.
      • Massman L.
      • Danner L.
      • Malard F.
      • Vora J.
      • et al.
      The human O-GlcNAcome database and meta-analysis.
      ), a recently compiled database of O-GlcNAc sites reported across multiple eukaryotic species. We also screened for missense variants ±5 residues from the O-GlcNAc site, as mutations proximal to the O-GlcNAc site may affect the affinity of OGT for a given substrate. The ClinVar screen identified 125 missense variants on O-GlcNAc sites, eight of which segregated with ID/DD (See File S2 for the list of ClinVar hits). The remaining 117 O-GlcNAc site missense variants were either pathogenic for a non-ID/DD syndrome or scored as likely benign or possibly damaging by PolyPhen scoring (PolyPhen HumVar score < 0.85) Additionally, we identified 34 pathogenic and likely pathogenic variants proximal (±5 residues) to the O-GlcNAc site, which segregate with ID/DD (See File S2). Two O-GlcNAcylation site mutations (TUBB2BS172P and TUBB2BS172L) were pathogenic or likely pathogenic in patients with cortical dysplasia, complex, with other brain malformations (CDCBM7) (
      • Jaglin X.H.
      • Poirier K.
      • Saillour Y.
      • Buhler E.
      • Tian G.
      • Bahi-Buisson N.
      • et al.
      Mutations in the beta-tubulin gene TUBB2B result in asymmetrical polymicrogyria.
      ). Six O-GlcNAcylation site missense variants (PUM1S802F, DOCK7S190N, DOCK7S190G, KMT2AS1858I, ATRXS594C, and HCF1T556M) were of uncertain significance, although PolyPhen scoring predicted all six variants as probably pathogenic (PolyPhen HumVar score > 0.85). For summaries of ID/DD missense variants and associated pathogenicity/PolyPhen scores, see File S2.
      Of the hits identified by the ClinVar search, the pathogenic TUBB2BS172P mutation was particularly interesting. The TUBB2BS172P variant segregates with CDCBM7 and displays impaired heterodimerization with α-tubulin and is not capable of polymerizing into microtubules (
      • Jaglin X.H.
      • Poirier K.
      • Saillour Y.
      • Buhler E.
      • Tian G.
      • Bahi-Buisson N.
      • et al.
      Mutations in the beta-tubulin gene TUBB2B result in asymmetrical polymicrogyria.
      ). A similar effect would be predicted from the presence of a bulky O-GlcNAc moiety at S172 due to steric occlusion of the GTP/GDP-binding site (see earlier discussion and Fig. 3A), further substantiating the proposed role of this site in regulating tubulin polymerization. Additionally, an S172L variant was also reported for the CDCBM7 syndrome (Table 1); however, the effects of this mutation on tubulin function have not been investigated biochemically or structurally. It is of course possible that these mutations affect tubulin function independent of O-GlcNAcylation. Indeed, mutagenesis from Ser to Pro (a “helix breaker” with restricted backbone conformations) may disrupt tubulin function independently of the loss of O-GlcNAc. However, the backbone conformation of S172 is compatible with that preferred by a proline (Fig. 3B). Future work could explore the effects of proline mutagenesis and loss of S172 O-GlcNAcylation on tubulin function in ID/DD.
      From the six missense variants predicted to be pathogenic ID/DD mutations by PolyPhen scoring, the HCF1T566M variant is of particular interest. The HCF1T566M variant is present in a patient with mental retardation 3, X-linked. T566 was not identified as a candidate O-GlcNAc site in our initial filter-based approach as it is not in a (known) functional region/domain, although it is proximal to the binding site for OGT (
      • Wysocka J.
      • Myers M.P.
      • Laherty C.D.
      • Eisenman R.N.
      • Herr W.
      Human Sin3 deacetylase and trithorax-related Set1/Ash2 histone H3-K4 methyltransferase are tethered together selectively by the cell-proliferation factor HCF-1.
      ). It is possible, however, that T566 O-GlcNAcylation stabilizes HCF1; HCF1 is cotranslationally O-GlcNAcylated (
      • Zhu Y.
      • Willems L.I.
      • Salas D.
      • Cecioni S.
      • Wu W.B.
      • Foster L.J.
      • et al.
      Tandem bioorthogonal labeling uncovers endogenous cotranslationally O-GlcNAc modified nascent proteins.
      ), and the sole reported function of cotranslational O-GlcNAcylation is stabilization of nascent protein chains during translation (
      • Zhu Y.
      • Liu T.W.
      • Cecioni S.
      • Eskandari R.
      • Zandberg W.F.
      • Vocadlo D.J.
      O-GlcNAc occurs cotranslationally to stabilize nascent polypeptide chains.
      ). Thus, the HCF1T566M variant may represent a destabilized hypomorph due to loss of T566 O-GlcNAcylation. Whether this is the case and whether HCF1 hypo-O-GlcNAcylation at T566 contributes to OGT-CDG etiology, are avenues for future research.

      Conclusions

      We have identified 22 ID/DD-associated proteins, the hypo-O-GlcNAcylation of which may contribute to OGT-CDG pathology. We arrived at these candidates by inferring functionality for conserved O-GlcNAc sites on these ID/DD-associated proteins through a filter-based approach integrating inputs from proteomic, clinical, structural, and literature databases. Additionally, we identified several conserved O-GlcNAc sites, which are pathogenically mutated in syndromic ID patients, adding further impetus to study the function of O-GlcNAc on these neuronal proteins. We anticipate that the results of this study will further highlight the convergence between the O-GlcNAc proteome and ID and galvanize further research into specific O-GlcNAc signaling nodes and how hypo-O-GlcNAcylation of these same proteins may give rise to OGT-CDG.

      Experimental procedures

      Filter-based approach for the identification of OGT-CDG candidate conveyers

      O-GlcNAc proteomic (O-GlcNAcomic) data were pooled manually from 14 publications into a Microsoft Excel spreadsheet (version 16.46) (
      • Vosseller K.
      • Trinidad J.C.
      • Chalkley R.J.
      • Specht C.G.
      • Thalhammer A.
      • Lynn A.J.
      • et al.
      O-linked N-acetylglucosamine proteomics of postsynaptic density preparations using lectin weak affinity chromatography and mass spectrometry.
      ,
      • Alfaro J.F.
      • Gong C.X.
      • Monroe M.E.
      • Aldrich J.T.
      • Clauss T.R.
      • Purvine S.O.
      • et al.
      Tandem mass spectrometry identifies many mouse brain O-GlcNAcylated proteins including EGF domain-specific O-GlcNAc transferase targets.
      ,
      • Trinidad J.C.
      • Barkan D.T.
      • Gulledge B.F.
      • Thalhammer A.
      • Sali A.
      • Schoepfer R.
      • et al.
      Global identification and characterization of both O-GlcNAcylation and phosphorylation at the murine synapse.
      ,
      • Hahne H.
      • Sobotzki N.
      • Nyberg T.
      • Helm D.
      • Borodkin V.S.
      • van Aalten D.M.
      • et al.
      Proteome wide purification and identification of O-GlcNAc-modified proteins using click chemistry and mass spectrometry.
      ,
      • Woo C.M.
      • Lund P.J.
      • Huang A.C.
      • Davis M.M.
      • Bertozzi C.R.
      • Pitteri S.J.
      Mapping and quantification of over 2000 O-linked glycopeptides in activated human T cells with isotope-targeted glycoproteomics (isotag).
      ,
      • Myers S.A.
      • Panning B.
      • Burlingame A.L.
      Polycomb repressive complex 2 is necessary for the normal site-specific O-GlcNAc distribution in mouse embryonic stem cells.
      ,
      • Nandi A.
      • Sprung R.
      • Barma D.K.
      • Zhao Y.
      • Kim S.C.
      • Falck J.R.
      • et al.
      Global identification of O-GlcNAc-modified proteins.
      ,
      • Khidekel N.
      • Ficarro S.B.
      • Peters E.C.
      • Hsieh-Wilson L.C.
      Exploring the O-GlcNAc proteome: direct identification of O-GlcNAc-modified proteins from the brain.
      ,
      • Rexach J.E.
      • Rogers C.J.
      • Yu S.H.
      • Tao J.
      • Sun Y.E.
      • Hsieh-Wilson L.C.
      Quantification of O-glycosylation stoichiometry and dynamics using resolvable mass tags.
      ,
      • Chalkley R.J.
      • Thalhammer A.
      • Schoepfer R.
      • Burlingame A.L.
      Identification of protein O-GlcNAcylation sites using electron transfer dissociation mass spectrometry on native peptides.
      ,
      • Zaro B.W.
      • Yang Y.Y.
      • Hang H.C.
      • Pratt M.R.
      Chemical reporters for fluorescent detection and identification of O-GlcNAc-modified proteins reveal glycosylation of the ubiquitin ligase NEDD4-1.
      ,
      • Wang Z.
      • Gucek M.
      • Hart G.W.
      Cross-talk between GlcNAcylation and phosphorylation: site-specific phosphorylation dynamics in response to globally elevated O-GlcNAc.
      ,
      • Zhao P.
      • Viner R.
      • Teo C.F.
      • Boons G.J.
      • Horn D.
      • Wells L.
      Combining high-energy C-trap dissociation and electron transfer dissociation for protein O-GlcNAc modification site assignment.
      ,
      • Dias W.B.
      • Cheung W.D.
      • Hart G.W.
      O-GlcNAcylation of kinases.
      ). As a quality control step, O-GlcNAc sites were excluded if the reported O-GlcNAc site mapped to asparagine (N-linked glycosylation), if the corresponding peptide was misannotated (reported O-GlcNAcylated peptide absent from the UniProt FASTA sequence of the full-length protein) or if the protein was exclusively extracellular (determined by manual inspection of UniProt entries for protein localization). To assess whether an O-GlcNAcylated protein was ID/DD associated, the list of site-mapped O-GlcNAcylated proteins was then compared to a recently reported list of all known ID/DD-associated genes (
      • Vissers L.E.
      • Gilissen C.
      • Veltman J.A.
      Genetic studies in intellectual disability and related disorders.
      ) and to the DECIPHER database of clinical genomic variation (
      • Firth H.V.
      • Richards S.M.
      • Bevan A.P.
      • Clayton S.
      • Corpas M.
      • Rajan D.
      • et al.
      DECIPHER: database of chromosomal imbalance and phenotype in humans using ensembl resources.
      ). If the gene encoding an O-GlcNAcylated protein was reported as being the sole “pathogenic” variant in a patient with “ID” or “developmental delay,” the O-GlcNAcylated protein passed this filter. Subsequently, to assess whether O-GlcNAc sites were conserved in vertebrates, amino acid sequences in FASTA format for human, mouse, rat, rabbit, and sheep homologs of a given O-GlcNAcylated protein were extracted from the UniProt database (see File S1 for UniProt IDs used). Clustal Omega was used to align amino acid sequences, with O-GlcNAc sites not conserved in ≥1 species excluded. In 3.84% of cases, one of rat, rabbit, or sheep sequences were absent for a given protein. In such cases, bovine, porcine, or chimpanzee sequences (in descending order of preference) were used instead, due to their availability in the UniProt database. Structural inspection of O-GlcNAc sites, to assess whether O-GlcNAc sites were surface exposed and thus accessible to the OGT active site, was carried out on available PDB structures using PyMOL (https://pymol.org/2/) v 1.7.4.5. Where structural information was not available, the IUPred2A (https://iupred2a.elte.hu) disorder predictor was used to predict whether an O-GlcNAc site resided in a disordered (and therefore accessible) region of the protein, using the long disorder (default) setting (
      • Erdos G.
      • Dosztanyi Z.
      Analyzing protein disorder with IUPred2A.
      ). O-GlcNAc sites were classified as ordered if IUPred scores were higher than 0.5. Disorder predictions were carried out using FASTA sequences from human homologs. IUPred scores for individual O-GlcNAc sites are detailed in File S1.

      GO and PANTHER pathway analysis

      GO analysis for biological processes involving OGT-CDG candidate conveyers was carried out using the Gene Ontology resource (http://geneontology.org) (
      • Ashburner M.
      • Ball C.A.
      • Blake J.A.
      • Botstein D.
      • Butler H.
      • Cherry J.M.
      • et al.
      Gene ontology: tool for the unification of biology. The gene ontology consortium.
      ). For investigating pathways in which OGT-CDG candidate conveyers are signaling nodes and classifying candidate proteins into classes, PANTHER version 16.0 was used (
      • Mi H.
      • Ebert D.
      • Muruganujan A.
      • Mills C.
      • Albou L.P.
      • Mushayamaha T.
      • et al.
      PANTHER version 16: a revised family classification, tree-based classification tool, enhancer regions and extensive API.
      ). For both GO and PANTHER pathway analysis, UniProt identifiers for human homologs were used (see File S1). Outputs from classification of shared “biological processes,” “protein class,” and “pathways” were exported as .txt files into Microsoft Excel and used to generate the associated figures.

      ClinVar database screening for missense mutations in ID/DD-associated proteins

      To retrieve variants in genes encoding O-GlcNAc modified proteins, the ClinVar short variants dataset (
      • Landrum M.J.
      • Lee J.M.
      • Benson M.
      • Brown G.R.
      • Chao C.
      • Chitipiralla S.
      • et al.
      ClinVar: Improving access to variant interpretations and supporting evidence.
      ) (https://ftp.ncbi.nlm.nih.gov/pub/clinvar/vcf_GRCh38/clinvar.vcf.gz, accessed 03/01/2022) was parsed using the PyVCF library and was cross compared with a recently published O-GlcNAcome (
      • Wulff-Fuentes E.
      • Berendt R.R.
      • Massman L.
      • Danner L.
      • Malard F.
      • Vora J.
      • et al.
      The human O-GlcNAcome database and meta-analysis.
      ) and an independent list of 146 conserved O-GlcNAc sites (File S1) parsed using the csv library (
      • Van Rossum G.
      ). Identified variants were filtered to only include missense mutations. Amino acid substitutions for each variant were retrieved using the requests library (
      • Chandra R.V.
      • Varanasi B.S.
      Python Requests Essentials.
      ) to interface with the esummary ClinVar API, parsed using the json library. Missense mutations affecting O-GlcNAcylated residues or sites ±5 residues were then filtered based on pathogenicity and retained for manual follow up. Code can be accessed at https://github.com/IggyCz/ClinVar_O-GlcNAcome_search/blob/main/clinvar_searcher_v1.py.

      PolyPhen scoring

      PolyPhen v2 (
      • Adzhubei I.A.
      • Schmidt S.
      • Peshkin L.
      • Ramensky V.E.
      • Gerasimova A.
      • Bork P.
      • et al.
      A method and server for predicting damaging missense mutations.
      ) was used to predict the pathogenicity of missense variants at conserved sites of O-GlcNAcylation. UniProt FASTA sequences for human homologs of a given candidate conveyer were manually entered into the search bar. HumVar scores, which factor in the presence of single nucleotide polymorphisms at a given site in healthy individuals, were used, along with default cut-offs of 0.85 for “probably damaging” and 0.15 for “likely benign.”

      Data availability

      All data are contained within the article.

      Supporting information

      This article contains supporting information (
      • Skorobogatko Y.
      • Landicho A.
      • Chalkley R.J.
      • Kossenkov A.V.
      • Gallo G.
      • Vosseller K.
      O-linked β-N-acetylglucosamine (O-GlcNAc) site thr-87 regulates synapsin I localization to synapses and size of the reserve pool of synaptic vesicles.
      ,
      • Jaglin X.H.
      • Poirier K.
      • Saillour Y.
      • Buhler E.
      • Tian G.
      • Bahi-Buisson N.
      • et al.
      Mutations in the beta-tubulin gene TUBB2B result in asymmetrical polymicrogyria.
      ,
      • Fourest-Lieuvin A.
      • Peris L.
      • Gache V.
      • Garcia-Saez I.
      • Juillan-Binard C.
      • Lantez V.
      • et al.
      Microtubule regulation in mitosis: tubulin phosphorylation by the cyclin-dependent kinase Cdk1.
      ,
      • Cuvertino S.
      • Stuart H.M.
      • Chandler K.E.
      • Roberts N.A.
      • Armstrong R.
      • Bernardini L.
      • et al.
      ACTB loss-of-function mutations result in a pleiotropic developmental disorder.
      ,
      • Da Silva J.S.
      • Medina M.
      • Zuliani C.
      • Di Nardo A.
      • Witke W.
      • Dotti C.G.
      RhoA/ROCK regulation of neuritogenesis via profilin IIa-mediated control of actin stability.
      ,
      • Holmes K.C.
      • Popp D.
      • Gebhard W.
      • Kabsch W.
      Atomic model of the actin filament.
      ,
      • Floor S.N.
      • Condon K.J.
      • Sharma D.
      • Jankowsky E.
      • Doudna J.A.
      Autoinhibitory interdomain interactions and subfamily-specific extensions redefine the catalytic core of the human DEAD-box protein DDX3.
      ,
      • Lai M.C.
      • Lee Y.H.
      • Tarn W.Y.
      The DEAD-box RNA helicase DDX3 associates with export messenger ribonucleoproteins as well as tip-associated protein and participates in translational control.
      ,
      • Bellini E.
      • Pavesi G.
      • Barbiero I.
      • Bergo A.
      • Chandola C.
      • Nawaz M.S.
      • et al.
      MeCP2 post-translational modifications: a mechanism to control its involvement in synaptic plasticity and homeostasis?.
      ,
      • Tsuruzoe S.
      • Ishihara K.
      • Uchimura Y.
      • Watanabe S.
      • Sekita Y.
      • Aoto T.
      • et al.
      Inhibition of DNA binding of Sox2 by the SUMO conjugation.
      ,
      • Myers S.A.
      • Peddada S.
      • Chatterjee N.
      • Friedrich T.
      • Tomoda K.
      • Krings G.
      • et al.
      SOX2 O-GlcNAcylation alters its protein-protein interactions and genomic occupancy to modulate gene expression in pluripotent cells.
      ,
      • Miller Jenkins L.M.
      • Feng H.
      • Durell S.R.
      • Tagad H.D.
      • Mazur S.J.
      • Tropea J.E.
      • et al.
      Characterization of the p300 Taz2-p53 TAD2 complex and comparison with the p300 Taz2-p53 TAD1 complex.
      ,
      • Wysocka J.
      • Myers M.P.
      • Laherty C.D.
      • Eisenman R.N.
      • Herr W.
      Human Sin3 deacetylase and trithorax-related Set1/Ash2 histone H3-K4 methyltransferase are tethered together selectively by the cell-proliferation factor HCF-1.
      ,
      • Kneussel M.
      • Haverkamp S.
      • Fuhrmann J.C.
      • Wang H.
      • Wässle H.
      • Olsen R.W.
      • et al.
      The gamma-aminobutyric acid type A receptor (GABAAR)-associated protein GABARAP interacts with gephyrin but is not involved in receptor anchoring at the synapse.
      ,
      • Yap C.C.
      • Murate M.
      • Kishigami S.
      • Muto Y.
      • Kishida H.
      • Hashikawa T.
      • et al.
      Adaptor protein complex-4 (AP-4) is expressed in the central nervous system neurons and interacts with glutamate receptor delta2.
      ,
      • Malakhova M.
      • Tereshko V.
      • Lee S-.Y.
      • Yao K.
      • Cho Y-.Y.
      • Bode A.
      • et al.
      Structural basis for activation of the autoinhibitory C-terminal kinase domain of p90 RSK2.
      ,
      • de Caestecker M.P.
      • Hemmati P.
      • Larisch-Bloch S.
      • Ajmera R.
      • Roberts A.B.
      • Lechleider R.J.
      Characterization of functional domains within Smad4/DPC4.
      ,
      • Rokudai S.
      • Laptenko O.
      • Arnal S.M.
      • Taya Y.
      • Kitabayashi I.
      • Prives C.
      MOZ increases p53 acetylation and premature senescence through its complex formation with PML.
      ,
      • Jaberi E.
      • Chitsazian F.
      • Shahidi G.A.
      • Rohani M.
      • Sina F.
      • Safari I.
      • et al.
      The novel mutation p.Asp251Asn in the β-subunit of succinate-CoA ligase causes encephalomyopathy and elevated succinylcarnitine.
      ,
      • Nishiyama M.
      • Oshikawa K.
      • Tsukada Y.
      • Nakagawa T.
      • Iemura S.
      • Natsume T.
      • et al.
      CHD8 suppresses p53-mediated apoptosis through histone H1 recruitment during early embryogenesis.
      ,
      • Mandel S.
      • Rechavi G.
      • Gozes I.
      Activity-dependent neuroprotective protein (ADNP) differentially interacts with chromatin to regulate genes essential for embryogenesis.
      ,
      • Itoh G.
      • Kanno S.
      • Uchida K.S.K.
      • Chiba S.
      • Sugino S.
      • Watanabe K.
      • et al.
      CAMP (C13orf8, ZNF828) is a novel regulator of kinetochore-microtubule attachment.
      ,
      • Dallas P.B.
      • Pacchione S.
      • Wilsker D.
      • Bowrin V.
      • Kobayashi R.
      • Moran E.
      The human SWI-SNF complex protein p270 is an ARID family member with non-sequence-specific DNA binding activity.
      ,
      • Beinert H.
      • Kennedy M.C.
      Aconitase, a two-faced protein: enzyme and iron regulatory factor.
      ,
      • Hoang H.T.
      • Schlager M.A.
      • Carter A.P.
      • Bullock S.L.
      DYNC1H1 mutations associated with neurological diseases compromise processivity of dynein-dynactin-cargo adaptor complexes.
      ,
      • Mou T-C.
      • Gille A.
      • Fancy D.A.
      • Seifert R.
      • Sprang S.R.
      Structural basis for the inhibition of mammalian membrane adenylyl cyclase by 2’(3’)-O-(N-Methylanthraniloyl)-guanosine 5’-triphosphate.
      ,
      • Kordeli E.
      • Lambert S.
      • Bennett V.
      AnkyrinG. A new ankyrin gene with neural-specific isoforms localized at the axonal initial segment and node of Ranvier.
      ,
      • Harris J.L.
      • Jakob B.
      • Taucher-Scholz G.
      • Dianov G.L.
      • Becherel O.J.
      • Lavin M.F.
      Aprataxin, poly-ADP ribose polymerase 1 (PARP-1) and apurinic endonuclease 1 (APE1) function together to protect the genome against oxidative damage.
      ,
      • Xu G.
      • Shin S.B.
      • Jaffrey S.R.
      Global profiling of protease cleavage sites by chemoselective labeling of protein N-termini.
      ,
      • Takakubo F.
      • Cartwright P.
      • Hoogenraad N.
      • Thorburn D.R.
      • Collins F.
      • Lithgow T.
      • et al.
      An amino acid substitution in the pyruvate dehydrogenase E1 alpha gene, affecting mitochondrial import of the precursor protein.
      ).

      Conflict of interest

      The authors declare that they have no conflicts of interest with the contents of this article.

      Acknowledgments

      We would like to thank Dr Sergio Galan Bartual and Dr Florence Authier for insightful discussions and feedback. Conor Mitchell was funded by the BBSRC EASTBIO Doctoral Training Program. Ignacy Czajewski was funded by the National Centre for the Replacement, Refinement and Reduction of Animals in Research (NC3Rs), T001682

      Author contributions

      C. W. M. and D. M. F. v. A. conceptualization; C. W. M. and I. C. methodology; I. C. software; I. C. validation; C. W. M. and I. C. investigation; C. W. M. and I. C. data curation; C. W. M. writing–original draft; C. W. M., I. C., and D. M. F. v. A. writing–review & editing; C. W. M. visualization; D. M. F. v. A. supervision; D. M. F. v. A. funding acquisition.

      Funding and additional information

      This work was funded by a Wellcome Trust Investigator Award (grant no. 110061 to D. M. F. v. A) and a Novo Nordisk Foundation Laureate Award (to D. M. F. v. A).

      References

        • Yang X.
        • Qian K.
        Protein O-GlcNAcylation: emerging mechanisms and functions.
        Nat. Rev. Mol. Cell Biol. 2017; 18: 452-465
        • Holt G.D.
        • Hart G.W.
        The subcellular distribution of terminal N-acetylglucosamine moieties. Localization of a novel protein-saccharide linkage, O-linked GlcNAc.
        J. Biol. Chem. 1986; 261: 8049-8057
        • Haltiwanger R.S.
        • Blomberg M.A.
        • Hart G.W.
        Glycosylation of nuclear and cytoplasmic proteins. Purification and characterization of a uridine diphospho-N-acetylglucosamine:polypeptide beta-N-acetylglucosaminyltransferase.
        J. Biol. Chem. 1992; 267: 9005-9013
        • Martinez-Fleites C.
        • He Y.
        • Davies G.J.
        Structural analyses of enzymes involved in the O-GlcNAc modification.
        Biochim. Biophys. Acta. 2010; 1800: 122-133
        • Stephen H.M.
        • Praissman J.L.
        • Wells L.
        Generation of an interactome for the tetratricopeptide repeat domain of O-GlcNAc transferase indicates a role for the enzyme in intellectual disability.
        J. Proteome Res. 2021; 20: 1229-1242
        • Capotosti F.
        • Guernier S.
        • Lammers F.
        • Waridel P.
        • Cai Y.
        • Jin J.
        • et al.
        O-GlcNAc transferase catalyzes site-specific proteolysis of HCF-1.
        Cell. 2011; 144: 376-388
        • Vocadlo D.J.
        O-GlcNAc processing enzymes: Catalytic mechanisms, substrate specificity, and enzyme regulation.
        Curr. Opin. Chem. Biol. 2012; 16: 488-497
        • Okuyama R.
        • Marshall S.
        UDP-N-acetylglucosaminyl transferase (OGT) in brain tissue: temperature sensitivity and subcellular distribution of cytosolic and nuclear enzyme.
        J. Neurochem. 2003; 86: 1271-1280
        • Vosseller K.
        • Trinidad J.C.
        • Chalkley R.J.
        • Specht C.G.
        • Thalhammer A.
        • Lynn A.J.
        • et al.
        O-linked N-acetylglucosamine proteomics of postsynaptic density preparations using lectin weak affinity chromatography and mass spectrometry.
        Mol. Cell Proteomics. 2006; 5: 923-934
        • Alfaro J.F.
        • Gong C.X.
        • Monroe M.E.
        • Aldrich J.T.
        • Clauss T.R.
        • Purvine S.O.
        • et al.
        Tandem mass spectrometry identifies many mouse brain O-GlcNAcylated proteins including EGF domain-specific O-GlcNAc transferase targets.
        Proc. Natl. Acad. Sci. U. S. A. 2012; 109: 7280-7285
        • Trinidad J.C.
        • Barkan D.T.
        • Gulledge B.F.
        • Thalhammer A.
        • Sali A.
        • Schoepfer R.
        • et al.
        Global identification and characterization of both O-GlcNAcylation and phosphorylation at the murine synapse.
        Mol. Cell Proteomics. 2012; 11: 215-229
        • Rexach J.E.
        • Clark P.M.
        • Mason D.E.
        • Neve R.L.
        • Peters E.C.
        • Hsieh-Wilson L.C.
        Dynamic O-GlcNAc modification regulates CREB-mediated gene expression and memory formation.
        Nat. Chem. Biol. 2012; 8: 253-261
        • Lagerlöf O.
        • Hart G.W.
        • Huganir R.L.
        O-GlcNAc transferase regulates excitatory synapse maturity.
        Proc. Natl. Acad. Sci. U. S. A. 2017; 114: 1684-1689
        • Skorobogatko Y.
        • Landicho A.
        • Chalkley R.J.
        • Kossenkov A.V.
        • Gallo G.
        • Vosseller K.
        O-linked β-N-acetylglucosamine (O-GlcNAc) site thr-87 regulates synapsin I localization to synapses and size of the reserve pool of synaptic vesicles.
        J. Biol. Chem. 2014; 289: 3602-3612
        • Lee B.E.
        • Kim H.Y.
        • Kim H.J.
        • Jeong H.
        • Kim B.G.
        • Lee H.E.
        • et al.
        O-GlcNAcylation regulates dopamine neuron function, survival and degeneration in Parkinson disease.
        Brain. 2020; 143: 3699-3716
        • Wheatley E.G.
        • Albarran E.
        • White C.W.
        • Bieri G.
        • Sanchez-Diaz C.
        • Pratt K.
        • et al.
        Neuronal O-GlcNAcylation improves cognitive function in the aged mouse brain.
        Curr. Biol. 2019; 29: 3359-3369.e4
        • O'Donnell N.
        • Zachara N.E.
        • Hart G.W.
        • Marth J.D.
        Ogt-dependent X-chromosome-linked protein glycosylation is a requisite modification in somatic cell function and embryo viability.
        Mol. Cell Biol. 2004; 24: 1680-1690
        • Pravata V.M.
        • Muha V.
        • Gundogdu M.
        • Ferenbach A.T.
        • Kakade P.S.
        • Vandadi V.
        • et al.
        Catalytic deficiency of O-GlcNAc transferase leads to X-linked intellectual disability.
        Proc. Natl. Acad. Sci. U. S. A. 2019; 116: 14961-14970
        • Pravata V.M.
        • Gundogdu M.
        • Bartual S.G.
        • Ferenbach A.T.
        • Stavridis M.
        • Õunap K.
        • et al.
        A missense mutation in the catalytic domain of O-GlcNAc transferase links perturbations in protein O-GlcNAcylation to X-linked intellectual disability.
        FEBS Lett. 2020; 594: 717-727
        • Vaidyanathan K.
        • Niranjan T.
        • Selvan N.
        • Teo C.F.
        • May M.
        • Patel S.
        • et al.
        Identification and characterization of a missense mutation in the O-linked β-N-acetylglucosamine (O-GlcNAc) transferase gene that segregates with X-linked intellectual disability.
        J. Biol. Chem. 2017; 292: 8948-8963
        • Willems A.P.
        • Gundogdu M.
        • Kempers M.J.E.
        • Giltay J.C.
        • Pfundt R.
        • Elferink M.
        • et al.
        Mutations in N-acetylglucosamine (O-GlcNAc) transferase in patients with X-linked intellectual disability.
        J. Biol. Chem. 2017; 292: 12621-12631
        • Vissers L.E.
        • Gilissen C.
        • Veltman J.A.
        Genetic studies in intellectual disability and related disorders.
        Nat. Rev. Genet. 2016; 17: 9-18
        • Pravata V.M.
        • Omelková M.
        • Stavridis M.P.
        • Desbiens C.M.
        • Stephen H.M.
        • Lefeber D.J.
        • et al.
        An intellectual disability syndrome with single-nucleotide variants in O-GlcNAc transferase.
        Eur. J. Hum. Genet. 2020; 28: 706-714
        • Gundogdu M.
        • Llabrés S.
        • Gorelik A.
        • Ferenbach A.T.
        • Zachariae U.
        • van Aalten D.M.F.
        The O-GlcNAc transferase intellectual disability mutation L254F distorts the TPR helix.
        Cell Chem. Biol. 2018; 25: 513-518.e4
        • Hahne H.
        • Sobotzki N.
        • Nyberg T.
        • Helm D.
        • Borodkin V.S.
        • van Aalten D.M.
        • et al.
        Proteome wide purification and identification of O-GlcNAc-modified proteins using click chemistry and mass spectrometry.
        J. Proteome Res. 2013; 12: 927-936
        • Woo C.M.
        • Lund P.J.
        • Huang A.C.
        • Davis M.M.
        • Bertozzi C.R.
        • Pitteri S.J.
        Mapping and quantification of over 2000 O-linked glycopeptides in activated human T cells with isotope-targeted glycoproteomics (isotag).
        Mol. Cell Proteomics. 2018; 17: 764-775
        • Chen P.H.
        • Smith T.J.
        • Wu J.
        • Siesser P.F.
        • Bisnett B.J.
        • Khan F.
        • et al.
        Glycosylation of KEAP1 links nutrient sensing to redox stress signaling.
        EMBO J. 2017; 36: 2233-2250
        • Zhu Q.
        • Cheng X.
        • Cheng Y.
        • Chen J.
        • Xu H.
        • Gao Y.
        • et al.
        O-GlcNAcylation regulates the methionine cycle to promote pluripotency of stem cells.
        Proc. Natl. Acad. Sci. U. S. A. 2020; 117: 7755-7763
        • Myers S.A.
        • Panning B.
        • Burlingame A.L.
        Polycomb repressive complex 2 is necessary for the normal site-specific O-GlcNAc distribution in mouse embryonic stem cells.
        Proc. Natl. Acad. Sci. U. S. A. 2011; 108: 9490-9495
        • Nandi A.
        • Sprung R.
        • Barma D.K.
        • Zhao Y.
        • Kim S.C.
        • Falck J.R.
        • et al.
        Global identification of O-GlcNAc-modified proteins.
        Anal. Chem. 2006; 78: 452-458
        • Khidekel N.
        • Ficarro S.B.
        • Peters E.C.
        • Hsieh-Wilson L.C.
        Exploring the O-GlcNAc proteome: direct identification of O-GlcNAc-modified proteins from the brain.
        Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 13132-13137
        • Rexach J.E.
        • Rogers C.J.
        • Yu S.H.
        • Tao J.
        • Sun Y.E.
        • Hsieh-Wilson L.C.
        Quantification of O-glycosylation stoichiometry and dynamics using resolvable mass tags.
        Nat. Chem. Biol. 2010; 6: 645-651
        • Chalkley R.J.
        • Thalhammer A.
        • Schoepfer R.
        • Burlingame A.L.
        Identification of protein O-GlcNAcylation sites using electron transfer dissociation mass spectrometry on native peptides.
        Proc. Natl. Acad. Sci. U. S. A. 2009; 106: 8894-8899
        • Zaro B.W.
        • Yang Y.Y.
        • Hang H.C.
        • Pratt M.R.
        Chemical reporters for fluorescent detection and identification of O-GlcNAc-modified proteins reveal glycosylation of the ubiquitin ligase NEDD4-1.
        Proc. Natl. Acad. Sci. U. S. A. 2011; 108: 8146-8151
        • Firth H.V.
        • Richards S.M.
        • Bevan A.P.
        • Clayton S.
        • Corpas M.
        • Rajan D.
        • et al.
        DECIPHER: database of chromosomal imbalance and phenotype in humans using ensembl resources.
        Am. J. Hum. Genet. 2009; 84: 524-533
        • Lazarus M.B.
        • Jiang J.
        • Kapuria V.
        • Bhuiyan T.
        • Janetzko J.
        • Zandberg W.F.
        • et al.
        HCF-1 is cleaved in the active site of O-GlcNAc transferase.
        Science. 2013; 342: 1235-1239
        • Rafie K.
        • Raimi O.
        • Ferenbach A.T.
        • Borodkin V.S.
        • Kapuria V.
        • van Aalten D.M.F.
        Recognition of a glycosylation substrate by the O-GlcNAc transferase TPR repeats.
        Open Biol. 2017; 7
        • Pathak S.
        • Alonso J.
        • Schimpl M.
        • Rafie K.
        • Blair D.E.
        • Borodkin V.S.
        • et al.
        The active site of O-GlcNAc transferase imposes constraints on substrate sequence.
        Nat. Struct. Mol. Biol. 2015; 22: 744-750
        • Zhu Y.
        • Liu T.W.
        • Cecioni S.
        • Eskandari R.
        • Zandberg W.F.
        • Vocadlo D.J.
        O-GlcNAc occurs cotranslationally to stabilize nascent polypeptide chains.
        Nat. Chem. Biol. 2015; 11: 319-325
        • Erdos G.
        • Dosztanyi Z.
        Analyzing protein disorder with IUPred2A.
        Curr. Protoc. Bioinform. 2020; 70: e99
        • Ohn T.
        • Kedersha N.
        • Hickman T.
        • Tisdale S.
        • Anderson P.
        A functional RNAi screen links O-GlcNAc modification of ribosomal proteins to stress granule and processing body assembly.
        Nat. Cell Biol. 2008; 10: 1224-1231
        • Sakabe K.
        • Wang Z.
        • Hart G.W.
        Beta-N-acetylglucosamine (O-GlcNAc) is part of the histone code.
        Proc. Natl. Acad. Sci. U. S. A. 2010; 107: 19915-19920
        • Ranuncolo S.M.
        • Ghosh S.
        • Hanover J.A.
        • Hart G.W.
        • Lewis B.A.
        Evidence of the involvement of O-GlcNAc-modified human RNA polymerase II CTD in transcription in vitro and in vivo.
        J. Biol. Chem. 2012; 287: 23549-23561
        • Moulton M.J.
        • Humphreys G.B.
        • Kim A.
        • Letsou A.
        O-GlcNAcylation dampens Dpp/BMP signaling to ensure proper Drosophila embryonic development.
        Dev. Cell. 2020; 53: 330-343.e3
        • Olivier-Van Stichelen S.
        • Dehennaut V.
        • Buzy A.
        • Zachayus J.L.
        • Guinez C.
        • Mir A.M.
        • et al.
        O-GlcNAcylation stabilizes β-catenin through direct competition with phosphorylation at threonine 41.
        FASEB J. 2014; 28: 3325-3328
        • Tarrant M.K.
        • Rho H.S.
        • Xie Z.
        • Jiang Y.L.
        • Gross C.
        • Culhane J.C.
        • et al.
        Regulation of CK2 by phosphorylation and O-GlcNAcylation revealed by semisynthesis.
        Nat. Chem. Biol. 2012; 8: 262-269
        • Flynn K.C.
        • Bradke F.
        Role of the cytoskeleton and membrane trafficking in axon–dendrite morphogenesis.
        in: Cellular Migration and Formation of Axons and Dendrites. Elsevier, Amsterdam, Netherlands2020: 21-56
        • Craig A.M.
        • Banker G.
        Neuronal polarity.
        Annu. Rev. Neurosci. 1994; 17: 267-310
        • Jaglin X.H.
        • Poirier K.
        • Saillour Y.
        • Buhler E.
        • Tian G.
        • Bahi-Buisson N.
        • et al.
        Mutations in the beta-tubulin gene TUBB2B result in asymmetrical polymicrogyria.
        Nat. Genet. 2009; 41: 746-752
        • Fourest-Lieuvin A.
        • Peris L.
        • Gache V.
        • Garcia-Saez I.
        • Juillan-Binard C.
        • Lantez V.
        • et al.
        Microtubule regulation in mitosis: tubulin phosphorylation by the cyclin-dependent kinase Cdk1.
        Mol. Biol. Cell. 2006; 17: 1041-1050
        • Löwe J.
        • Li H.
        • Downing K.H.
        • Nogales E.
        Refined structure of alpha beta-tubulin at 3.5 A resolution.
        J. Mol. Biol. 2001; 313: 1045-1057
        • Ji S.
        • Kang J.G.
        • Park S.Y.
        • Lee J.
        • Oh Y.J.
        • Cho J.W.
        O-GlcNAcylation of tubulin inhibits its polymerization.
        Amino Acids. 2011; 40: 809-818
        • Shelanski M.L.
        • Gaskin F.
        • Cantor C.R.
        Microtubule assembly in the absence of added nucleotides.
        Proc. Natl. Acad. Sci. U. S. A. 1973; 70: 765-768
        • Bradke F.
        • Dotti C.G.
        The role of local actin instability in axon formation.
        Science. 1999; 283: 1931-1934
        • Gomez T.M.
        • Letourneau P.C.
        Actin dynamics in growth cone motility and navigation.
        J. Neurochem. 2014; 129: 221-234
        • Cuvertino S.
        • Stuart H.M.
        • Chandler K.E.
        • Roberts N.A.
        • Armstrong R.
        • Bernardini L.
        • et al.
        ACTB loss-of-function mutations result in a pleiotropic developmental disorder.
        Am. J. Hum. Genet. 2017; 101: 1021-1033
        • Schutt C.E.
        • Myslik J.C.
        • Rozycki M.D.
        • Goonesekere N.C.
        • Lindberg U.
        The structure of crystalline profilin-beta-actin.
        Nature. 1993; 365: 810-816
        • Davey R.J.
        • Moens P.D.
        Profilin: many facets of a small protein.
        Biophys. Rev. 2020; 12: 827-849
        • Da Silva J.S.
        • Medina M.
        • Zuliani C.
        • Di Nardo A.
        • Witke W.
        • Dotti C.G.
        RhoA/ROCK regulation of neuritogenesis via profilin IIa-mediated control of actin stability.
        J. Cell Biol. 2003; 162: 1267-1279
        • Holmes K.C.
        • Popp D.
        • Gebhard W.
        • Kabsch W.
        Atomic model of the actin filament.
        Nature. 1990; 347: 44-49
        • Floor S.N.
        • Condon K.J.
        • Sharma D.
        • Jankowsky E.
        • Doudna J.A.
        Autoinhibitory interdomain interactions and subfamily-specific extensions redefine the catalytic core of the human DEAD-box protein DDX3.
        J. Biol. Chem. 2016; 291: 2412-2421
        • Putnam A.A.
        • Gao Z.
        • Liu F.
        • Jia H.
        • Yang Q.
        • Jankowsky E.
        Division of labor in an oligomer of the DEAD-box RNA helicase Ded1p.
        Mol. Cell. 2015; 59: 541-552
        • Lai M.C.
        • Chang W.C.
        • Shieh S.Y.
        • Tarn W.Y.
        DDX3 regulates cell growth through translational control of cyclin E1.
        Mol. Cell Biol. 2010; 30: 5444-5453
        • Chen H.H.
        • Yu H.I.
        • Tarn W.Y.
        DDX3 modulates neurite development via translationally activating an RNA regulon involved in Rac1 activation.
        J. Neurosci. 2016; 36: 9792-9804
        • Samir P.
        • Kesavardhana S.
        • Patmore D.M.
        • Gingras S.
        • Malireddi R.K.S.
        • Karki R.
        • et al.
        DDX3X acts as a live-or-die checkpoint in stressed cells by regulating NLRP3 inflammasome.
        Nature. 2019; 573: 590-594
        • Protter D.S.W.
        • Parker R.
        Principles and properties of stress granules.
        Trends Cell Biol. 2016; 26: 668-679
        • Snijders Blok L.
        • Madsen E.
        • Juusola J.
        • Gilissen C.
        • Baralle D.
        • Reijnders M.R.
        • et al.
        Mutations in DDX3X are a common cause of unexplained intellectual disability with gender-specific effects on Wnt signaling.
        Am. J. Hum. Genet. 2015; 97: 343-352
        • Lai M.C.
        • Lee Y.H.
        • Tarn W.Y.
        The DEAD-box RNA helicase DDX3 associates with export messenger ribonucleoproteins as well as tip-associated protein and participates in translational control.
        Mol. Biol. Cell. 2008; 19: 3847-3858
        • Zeidan Q.
        • Wang Z.
        • De Maio A.
        • Hart G.W.
        O-GlcNAc cycling enzymes associate with the translational machinery and modify core ribosomal proteins.
        Mol. Biol. Cell. 2010; 21: 1922-1936
        • Zhang X.
        • Shu X.E.
        • Qian S.B.
        O-GlcNAc modification of eIF4GI acts as a translational switch in heat shock response.
        Nat. Chem. Biol. 2018; 14: 909-916
        • Gold W.A.
        • Krishnarajy R.
        • Ellaway C.
        • Christodoulou J.
        Rett syndrome: a genetic update and clinical review focusing on comorbidities.
        ACS Chem. Neurosci. 2018; 9: 167-176
        • Bellini E.
        • Pavesi G.
        • Barbiero I.
        • Bergo A.
        • Chandola C.
        • Nawaz M.S.
        • et al.
        MeCP2 post-translational modifications: a mechanism to control its involvement in synaptic plasticity and homeostasis?.
        Front. Cell Neurosci. 2014; 8: 236
        • Li H.
        • Zhong X.
        • Chau K.F.
        • Williams E.C.
        • Chang Q.
        Loss of activity-induced phosphorylation of MeCP2 enhances synaptogenesis, LTP and spatial memory.
        Nat. Neurosci. 2011; 14: 1001-1008
        • Tsuruzoe S.
        • Ishihara K.
        • Uchimura Y.
        • Watanabe S.
        • Sekita Y.
        • Aoto T.
        • et al.
        Inhibition of DNA binding of Sox2 by the SUMO conjugation.
        Biochem. Biophys. Res. Commun. 2006; 351: 920-926
        • Myers S.A.
        • Peddada S.
        • Chatterjee N.
        • Friedrich T.
        • Tomoda K.
        • Krings G.
        • et al.
        SOX2 O-GlcNAcylation alters its protein-protein interactions and genomic occupancy to modulate gene expression in pluripotent cells.
        Elife. 2016; 5e10647
        • Blackburn P.R.
        • Chacon-Camacho O.F.
        • Ortiz-González X.R.
        • Reyes M.
        • Lopez-Uriarte G.A.
        • Zarei S.
        • et al.
        Extension of the mutational and clinical spectrum of SOX2 related disorders: description of six new cases and a novel association with suprasellar teratoma.
        Am. J. Med. Genet. A. 2018; 176: 2710-2719
        • Kauppinen T.M.
        • Chan W.Y.
        • Suh S.W.
        • Wiggins A.K.
        • Huang E.J.
        • Swanson R.A.
        Direct phosphorylation and regulation of poly(ADP-ribose) polymerase-1 by extracellular signal-regulated kinases 1/2.
        Proc. Natl. Acad. Sci. U. S. A. 2006; 103: 7136-7141
        • Lai Y.S.
        • Chang C.W.
        • Pawlik K.M.
        • Zhou D.
        • Renfrow M.B.
        • Townes T.M.
        SRY (sex determining region Y)-box2 (Sox2)/poly ADP-ribose polymerase 1 (Parp1) complexes regulate pluripotency.
        Proc. Natl. Acad. Sci. U. S. A. 2012; 109: 3772-3777
        • Gu W.
        • Roeder R.G.
        Activation of p53 sequence-specific DNA binding by acetylation of the p53 C-terminal domain.
        Cell. 1997; 90: 595-606
        • Negri G.
        • Magini P.
        • Milani D.
        • Colapietro P.
        • Rusconi D.
        • Scarano E.
        • et al.
        From whole gene deletion to point mutations of EP300-positive rubinstein-taybi patients: new insights into the mutational spectrum and peculiar clinical hallmarks.
        Hum. Mutat. 2016; 37: 175-183
        • Miller Jenkins L.M.
        • Feng H.
        • Durell S.R.
        • Tagad H.D.
        • Mazur S.J.
        • Tropea J.E.
        • et al.
        Characterization of the p300 Taz2-p53 TAD2 complex and comparison with the p300 Taz2-p53 TAD1 complex.
        Biochemistry. 2015; 54: 2001-2010
        • Landrum M.J.
        • Lee J.M.
        • Benson M.
        • Brown G.R.
        • Chao C.
        • Chitipiralla S.
        • et al.
        ClinVar: Improving access to variant interpretations and supporting evidence.
        Nucl. Acids Res. 2018; 46: D1062-D1067
        • Wulff-Fuentes E.
        • Berendt R.R.
        • Massman L.
        • Danner L.
        • Malard F.
        • Vora J.
        • et al.
        The human O-GlcNAcome database and meta-analysis.
        Sci. Data. 2021; 8: 25
        • Wysocka J.
        • Myers M.P.
        • Laherty C.D.
        • Eisenman R.N.
        • Herr W.
        Human Sin3 deacetylase and trithorax-related Set1/Ash2 histone H3-K4 methyltransferase are tethered together selectively by the cell-proliferation factor HCF-1.
        Genes Dev. 2003; 17: 896-911
        • Zhu Y.
        • Willems L.I.
        • Salas D.
        • Cecioni S.
        • Wu W.B.
        • Foster L.J.
        • et al.
        Tandem bioorthogonal labeling uncovers endogenous cotranslationally O-GlcNAc modified nascent proteins.
        J. Am. Chem. Soc. 2020; 142: 15729-15739
        • Wang Z.
        • Gucek M.
        • Hart G.W.
        Cross-talk between GlcNAcylation and phosphorylation: site-specific phosphorylation dynamics in response to globally elevated O-GlcNAc.
        Proc. Natl. Acad. Sci. U. S. A. 2008; 105: 13793-13798
        • Zhao P.
        • Viner R.
        • Teo C.F.
        • Boons G.J.
        • Horn D.
        • Wells L.
        Combining high-energy C-trap dissociation and electron transfer dissociation for protein O-GlcNAc modification site assignment.
        J. Proteome Res. 2011; 10: 4088-4104
        • Dias W.B.
        • Cheung W.D.
        • Hart G.W.
        O-GlcNAcylation of kinases.
        Biochem. Biophys. Res. Commun. 2012; 422: 224-228
        • Ashburner M.
        • Ball C.A.
        • Blake J.A.
        • Botstein D.
        • Butler H.
        • Cherry J.M.
        • et al.
        Gene ontology: tool for the unification of biology. The gene ontology consortium.
        Nat. Genet. 2000; 25: 25-29
        • Mi H.
        • Ebert D.
        • Muruganujan A.
        • Mills C.
        • Albou L.P.
        • Mushayamaha T.
        • et al.
        PANTHER version 16: a revised family classification, tree-based classification tool, enhancer regions and extensive API.
        Nucl. Acids Res. 2021; 49: D394-D403
        • Van Rossum G.
        The Python Library Reference, release. 3.8.2. Python Software Foundation., Beaverton, OR2020
        • Chandra R.V.
        • Varanasi B.S.
        Python Requests Essentials.
        Packt Publishing Ltd, Birmingham, UK2015
        • Adzhubei I.A.
        • Schmidt S.
        • Peshkin L.
        • Ramensky V.E.
        • Gerasimova A.
        • Bork P.
        • et al.
        A method and server for predicting damaging missense mutations.
        Nat Met. 2010; 7: 248-249
        • Kneussel M.
        • Haverkamp S.
        • Fuhrmann J.C.
        • Wang H.
        • Wässle H.
        • Olsen R.W.
        • et al.
        The gamma-aminobutyric acid type A receptor (GABAAR)-associated protein GABARAP interacts with gephyrin but is not involved in receptor anchoring at the synapse.
        Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 8594-8599
        • Yap C.C.
        • Murate M.
        • Kishigami S.
        • Muto Y.
        • Kishida H.
        • Hashikawa T.
        • et al.
        Adaptor protein complex-4 (AP-4) is expressed in the central nervous system neurons and interacts with glutamate receptor delta2.
        Mol. Cell. Neurosci. 2003; 24: 283-295
        • Malakhova M.
        • Tereshko V.
        • Lee S-.Y.
        • Yao K.
        • Cho Y-.Y.
        • Bode A.
        • et al.
        Structural basis for activation of the autoinhibitory C-terminal kinase domain of p90 RSK2.
        Nat. Struct. Mol. Biol. 2008; 15: 112-113
        • de Caestecker M.P.
        • Hemmati P.
        • Larisch-Bloch S.
        • Ajmera R.
        • Roberts A.B.
        • Lechleider R.J.
        Characterization of functional domains within Smad4/DPC4.
        J. Biol. Chem. 1997; 272: 13690-13696
        • Rokudai S.
        • Laptenko O.
        • Arnal S.M.
        • Taya Y.
        • Kitabayashi I.
        • Prives C.
        MOZ increases p53 acetylation and premature senescence through its complex formation with PML.
        Proc. Natl. Acad. Sci. U. S. A. 2013; 110: 3895-3900
        • Jaberi E.
        • Chitsazian F.
        • Shahidi G.A.
        • Rohani M.
        • Sina F.
        • Safari I.
        • et al.
        The novel mutation p.Asp251Asn in the β-subunit of succinate-CoA ligase causes encephalomyopathy and elevated succinylcarnitine.
        J. Hum. Genet. 2013; 58: 526-530
        • Nishiyama M.
        • Oshikawa K.
        • Tsukada Y.
        • Nakagawa T.
        • Iemura S.
        • Natsume T.
        • et al.
        CHD8 suppresses p53-mediated apoptosis through histone H1 recruitment during early embryogenesis.
        Nat. Cell. Biol. 2009; 11: 172-182
        • Mandel S.
        • Rechavi G.
        • Gozes I.
        Activity-dependent neuroprotective protein (ADNP) differentially interacts with chromatin to regulate genes essential for embryogenesis.
        Dev. Biol. 2007; 303: 814-824
        • Itoh G.
        • Kanno S.
        • Uchida K.S.K.
        • Chiba S.
        • Sugino S.
        • Watanabe K.
        • et al.
        CAMP (C13orf8, ZNF828) is a novel regulator of kinetochore-microtubule attachment.
        EMBO J. 2011; 30: 130-144
        • Dallas P.B.
        • Pacchione S.
        • Wilsker D.
        • Bowrin V.
        • Kobayashi R.
        • Moran E.
        The human SWI-SNF complex protein p270 is an ARID family member with non-sequence-specific DNA binding activity.
        Mol. Cell. Biol. 2000; 20: 3137-3146
        • Beinert H.
        • Kennedy M.C.
        Aconitase, a two-faced protein: enzyme and iron regulatory factor.
        FASEB J. 1993; 7: 1442-1449
        • Hoang H.T.
        • Schlager M.A.
        • Carter A.P.
        • Bullock S.L.
        DYNC1H1 mutations associated with neurological diseases compromise processivity of dynein-dynactin-cargo adaptor complexes.
        Proc. Natl. Acad. Sci. U. S. A. 2017; 114: E1597-E1606
        • Mou T-C.
        • Gille A.
        • Fancy D.A.
        • Seifert R.
        • Sprang S.R.
        Structural basis for the inhibition of mammalian membrane adenylyl cyclase by 2’(3’)-O-(N-Methylanthraniloyl)-guanosine 5’-triphosphate.
        J. Biol. Chem. 2005; 280: 7253-7261
        • Kordeli E.
        • Lambert S.
        • Bennett V.
        AnkyrinG. A new ankyrin gene with neural-specific isoforms localized at the axonal initial segment and node of Ranvier.
        J. Biol. Chem. 1995; 270: 2352-2359
        • Harris J.L.
        • Jakob B.
        • Taucher-Scholz G.
        • Dianov G.L.
        • Becherel O.J.
        • Lavin M.F.
        Aprataxin, poly-ADP ribose polymerase 1 (PARP-1) and apurinic endonuclease 1 (APE1) function together to protect the genome against oxidative damage.
        Hum. Mol. Genet. 2009; 18: 4102-4117
        • Xu G.
        • Shin S.B.
        • Jaffrey S.R.
        Global profiling of protease cleavage sites by chemoselective labeling of protein N-termini.
        Proc. Natl. Acad. Sci. U. S. A. 2009; 106: 19310-19315
        • Takakubo F.
        • Cartwright P.
        • Hoogenraad N.
        • Thorburn D.R.
        • Collins F.
        • Lithgow T.
        • et al.
        An amino acid substitution in the pyruvate dehydrogenase E1 alpha gene, affecting mitochondrial import of the precursor protein.
        Am. J. Hum. Genet. 1995; 57: 772-780