Carbohydrates comprise one of the largest and most diverse collections of biologically active molecules, and it is increasingly clear that they participate in nearly every aspect of biology, including diseases such as cancer and neurodegenerative disorders. However, compared to their macromolecular peers (i.e., proteins and nucleic acids), carbohydrates remain relatively unexplored, and their structure-function relationships are still poorly understood.
There are several fundamental challenges inherent in studying carbohydrates, including (1) the chemical complexity of carbohydrates, (2) the lack of efficient and sensitive analytical methods for their detection and quantification, and (3) the lack of reliable and scalable methods for their synthesis.
Research in my laboratory uses the principles and tools of chemistry to study the roles of carbohydrates and their associated proteins in normal and disease settings. We use chemistry to overcome the fundamental challenges described above and to answer questions that cannot be addressed by the use of biochemistry or genetics. In this way, we have provided new insights into fundamental biological processes and defined exciting new roles for carbohydrates in neurobiology and cancer.
Chemical Tools to Study Dynamic O-GlcNAc Glycosylation
O-GlcNAc glycosylation is the covalent modification of serine and threonine residues of cytoplasmic and nuclear proteins by β-N-acetylglucosamine. This dynamic and reversible modification is emerging as a key regulator of many cellular processes, including signal transduction, transcription, and proteasomal degradation. Perturbations in O-GlcNAc glycosylation levels have been associated with diseases such as diabetes, neurodegenerative diseases, and cancer. O-GlcNAc has several features that make it more difficult to study than other post-translational modifications. It serves as a regulatory modification (e.g., dynamic, labile, low cellular abundance), which requires more sensitive, specialized tools for its detection and study. Moreover, whereas phosphorylation has only three major forms (phosphoserine, phosphothreonine, and phosphotyrosine), the GlcNAc sugar is present in hundreds of different cellular glycans, necessitating the development of selective methods for O-GlcNAc detection. Finally, genetic approaches to studying O-GlcNAc glycosylation are confounded by the presence of a single O-GlcNAc transferase (OGT) gene, whose deletion can induce pleiotropic effects and embryonic lethality.
We have developed a set of chemical tools to overcome these challenges and advance a molecular-level understanding of O-GlcNAc glycosylation. One such approach is a chemoenzymatic strategy that enables the rapid and sensitive detection of O-GlcNAc–glycosylated proteins. We exploit an engineered galactosyltransferase to transfer a synthetic ketone- or azide-containing substrate onto O-GlcNAc–modified proteins. Once installed, the ketone or azide moiety can be selectively reacted with radiolabels, affinity tags, or chemiluminescent probes. This approach solves a major challenge—sensitive detection—and has enabled the identification of more than 200 new O-GlcNAc–glycosylated proteins from the mammalian brain, including proteins involved in gene expression, neuronal signaling, and synaptic plasticity. These results significantly expand the number of proteins known to possess the modification and suggest that O-GlcNAc glycosylation plays important roles in regulating neuronal functions.
A key unresolved question is how O-GlcNAc glycosylation is dynamically regulated within cells. We have developed methods to study the dynamics of O-GlcNAc glycosylation in vivo on a proteome-wide scale and to map O-GlcNAc sites using electron-transfer dissociation mass spectrometry. These studies have revealed that O-GlcNAc glycosylation is reversible in neurons and is rapidly cycled at certain sites within proteins. Only a subset of the O-GlcNAc sites changed in response to a given stimulus, providing strong evidence for complex regulation of the modifying enzymes. We also discovered that O-GlcNAc glycosylation is induced in vivo by neuronal activity via activation of calcium-dependent signaling pathways. Thus, the modification is dynamically regulated by important neuronal stimuli and not merely controlled passively by cellular glucose concentrations.
We have also developed tools to interrogate the stoichiometry of O-GlcNAc glycosylation on a dynamic timescale, which provides invaluable information about its role in a particular process. Specifically, we have developed a facile method to quantify glycosylation stoichiometries in vivo by tagging the O-GlcNAc moiety with a polyethylene glycol mass tag. The mass tag shifts the molecular weight of the glycoproteins, which are then visualized by immunoblotting. This approach enables one to establish whether proteins are mono-, di-, or multiply glycosylated, to quantify glycosylation stoichiometries and sugar occupancies at specific sites, and to monitor the dynamics of O-GlcNAc glycosylation on proteins of interest—all without the need for protein purification, radiolabels, or advanced instrumentation.
New Roles for O-GlcNAc Glycosylation
An emerging theme from our studies is that O-GlcNAc glycosylation may play critical roles in the regulation of gene expression. We have identified the O-GlcNAc modification on numerous transcription factors and transcriptional regulatory proteins (e.g., CREB, ATF-2, Sox2, ΔFosB, CBP) involved in regulating processes such as drug addiction, hormone receptor activation, and memory consolidation. We have also discovered that the modification extends beyond transcription factors to new classes of proteins, including transcriptional corepressors, coactivators, and chromatin-remodeling enzymes. These findings suggest that O-GlcNAc glycosylation may serve as a general mechanism for the control of transcription.
Indeed, we have found that O-GlcNAc glycosylation of the transcription factor CREB (cyclic AMP response element–binding protein) at a specific site, Ser-40, represses basal and activity-induced gene expression in neurons and has important functional consequences for axonal and dendritic growth, as well as long-term memory formation. These studies establish a role for O-GlcNAc glycosylation in coupling neuronal activity to transcription, and they provide the first demonstration that the O-GlcNAc modification can influence higher-order brain functions.
Regulatory networks of post-translational modifications are an emerging area of study in the context of histones and many regulatory proteins. Studies have shown the potential for O-GlcNAc and phosphorylation to act reciprocally (i.e., yin-yang) or engage in combinatorial crosstalk at the level of signaling cascades. Our chemical tools have enabled us to distinguish subpopulations of post-translationally modified proteins and to study the kinetics of the interplay between the two modifications. We have observed a close coupling and complex interplay of O-GlcNAc glycosylation and phosphorylation on neuronal proteins. For instance, we found that glycosylation is induced specifically on the phosphorylated subpopulation of CREB, suggesting activation of a specific sequence of modifications in response to neuronal depolarization. Whereas phosphorylation stimulates CREB-mediated transcription, glycosylation represses CREB-mediated transcription, and together, the two modifications may finely tune the expression of genes.
Current studies in our laboratory seek to determine how O-GlcNAc glycosylation is dynamically regulated at synapses and its implications for dendritic protein synthesis and synaptic plasticity. We are also studying the roles of the modification in the context of cancer and neurodegenerative diseases.
Chemical Tools for Studying Glycosaminoglycans
Glycosaminoglycans (GAGs) are polysaccharides composed of repeating disaccharide units that undergo regioselective sulfation to give rise to diverse sulfation patterns in vivo. Genetic, structural, and biochemical studies have demonstrated their importance in regulating processes such as development, cancer metastasis, and spinal cord injury. However, efforts to elucidate the functional importance of specific sulfation motifs have been hampered by the complex chemical structure and biosynthesis of GAGs. For instance, genetic approaches that target a specific sulfotransferase gene often perturb other sulfation patterns within the polysaccharide chain and disrupt multiple proteoglycan core proteins, while biochemical methods typically afford a mixture of heterogeneously sulfated compounds.
To solve this fundamental problem, we harnessed the power of synthetic chemistry to access well-defined oligosaccharides, in which sulfate groups are installed at precise positions along the carbohydrate chain. Over the years, we have established synthetic routes to all of the known sulfation motifs of chondroitin sulfate (CS) in the mammalian brain. We have also created synthetic glycopolymers with programmable sulfation sequences and tunable chemical and biological properties. These molecules provide powerful tools for studying the role of specific motifs and for manipulating the functions of GAGs at the molecular level. The ability to synthesize defined molecules has also enabled the generation of monoclonal antibodies selective for particular sulfation motifs. These antibodies have been used to interrogate the roles of sulfation in vivo and to evaluate CS motifs as potential targets for therapeutic intervention.
A defining feature of GAGs is their ability to interact with proteins and assemble protein-protein signaling complexes. However, few methods existed for the rapid identification and study of GAG-protein interactions. To this end, we have devised methods to immobilize synthetic oligosaccharides, glycopolymers, and polysaccharides onto surfaces. We have demonstrated that these GAG microarrays are valuable tools to identify novel GAG-binding proteins, probe the specificities of proteins for specific motifs and GAG subclasses, and study the assembly of multimeric GAG-protein complexes. In parallel, we have developed computational methods to obtain structural insights into GAG-binding sites in proteins.
New Roles for Chondroitin Sulfate Glycosaminoglycans
Understanding the barriers to axon regeneration is critical to the development of new treatments for neuronal injury and neurodegenerative diseases. We have focused on CS GAGs, which play important roles in brain development and form a major barrier to axon regeneration after central nervous system injury. Using well-defined GAG tools, we have shown that a specific sulfation motif, known as CS-E, exerts both growth-promoting and growth-inhibiting effects on neurite outgrowth in vitro. These context-dependent effects involve the interaction of CS-E with specific protein partners, such as the neurotrophin family of growth factors and the protein phosphatase PTPσ. Blocking the CS-E motif with a monoclonal antibody reversed the neurite inhibition mediated by CS proteoglycans and stimulated optic axon regeneration in vivo. Thus, antibodies or other reagents specific for the CS-E epitope may provide a promising new strategy to promote axon regeneration and neural plasticity after injury.
More broadly, our results provide support for the concept of a "sulfation code," whereby the sulfation patterns of GAGs encode functional information in a structure-specific manner analogous to DNA, RNA, and proteins. We are investigating the functional roles and spatiotemporal expression of CS-E and other sulfation sequences in the brain. In addition, we are exploring the therapeutic potential of the CS-E antibody to promote axon regeneration after spinal cord injury.
This work is also supported by grants from the National Institutes of Health, the Christopher and Dana Reeve Foundation, and the Tobacco-Related Disease Research Program.
As of May 30, 2012