Glycoproteomics-based discovery of cancer and stem cell biomarkers
Given the strong correlation of altered glycosylation patterns with malignancy, glycosylated proteins may be an information-rich subset of the proteome from which cancer biomarkers can be discovered. We employ metabolic labeling with azidosugars as a means to tag specific classes of glycoproteins for enrichment from human tissue samples and subsequent identification by mass spectrometry. A challenge in this endeavor is defining sites of glycosylation on peptide digests derived from such complex samples. To facilitate this effort, we are creating isotopic labeling strategies and computational methods that enable the detection of glycosylated peptides independent of the mass of the pendant glycan. Collectively, these tools allow us to profile changes in protein glycosylation associated with human prostate cancer progression and embryoic stem cell differentiation.
Toxigenic bacteria, which include some species of Escherichia, Shigella, Vibrio, and Clostridium, release protein toxins that alter essential host processes, including endocytic pathways, cell signaling and cytoskeletal reorganization. These toxins are key virulence factors that damage host tissues, and aid the spread of bacteria and evasion of immune clearance. Many bacterial toxins have been shown to bind host glycans. Glycan receptor specificity is critical for the pathogenic process, as it determines host susceptibility, tissue tropism, and the nature and spectrum of the resultant pathology. With contributions from the CFG, knowledge of the molecular and structural bases for toxin-glycan interactions is providing a rational framework for design of specific toxin inhibitors with considerable potential as anti-infective therapeutic agents.
Cellular expression of GBP and ligands
HA is expressed on the surface of influenza virus infected cells before being budded out into progeny virions. H3N2 viruses infect the respiratory tract of humans and birds; in birds they may also infect the gut epithelia. Using MAA and SNA lectins, the upper respiratory tract appears to have more 2-6 linked sialic acid while 2-3 sialic acid appears more abundant in the lungs, but relationship between the different specificities of H3N2 HAs and the cell types infected remains unclear . H3N2 viruses infect very few continuous cell lines. Madin-Darby canine kidney cells are most commonly used. Non-permissive cell lines may take up virus efficiently, replicate RNA and express HA on the cell surface but do not bud new virus particles .
Biosynthesis of ligands
Sialylated glycoproteins or glycolipids recognized by human influenza hemagglutinin H3 are synthesized by host cells. The H3 hemagglutinin shows considerable diversity in binding but with rare exceptions the sialic acid is attached 2-6 to the next sugar on structures that are mostly typical of N-linked glycans on proteins. The enzymes required for biosynthesis of the type 2 poly N-acetyllactosamine chains and modification with sialic acid or with sialic acid and fucose, have been defined ([poly N-acetyllactosamine extension biosynthesis]). The sialyltransferases that generate ligands for most H3 subtype hemagglutinins are ST6Gal1, ST6GalII, ST6GalNAc1, ST6GalNAcII, ST6GalNAcIV.
Biological roles of GBP-ligand interaction
Sialylated glycans on the surface of cells lining the respiratory tract serve to capture virus to initiate infection. Glycan array analyses have confirmed that human influenza viruses such as those carrying the H3 HA bind only to structures with NeuAcα2-6 and avian isolates bind only to structures containing NeuAcα2-3. The role of this GBP-glycan interaction in initiating endocytosis is still unclear, but in the low pH of the endosomal compartment, the HA undergoes a large conformational change  that brings about fusion of the viral membrane with the cell membrane so that viral nucleocapsids are released, enter the nucleus and initiate viral transcription and replication.
Capsid Virus GBPs
A significant number of non-enveloped capsid viruses use glycans as cell attachment receptors. In some cases, viruses exclusively engage glycan structures at the cell surface, while in other cases glycans serve as co-receptors, and viruses also bind to proteins. The carbohydrate structures recognized by capsid viruses are diverse, ranging from gangliosides and other glycolipids to protein-bound glycans. Changes in carbohydrate binding specificity among family members of the same virus can often lead to altered tropism and pathogenicity. Thus, the elucidation of parameters that guide the specificity of interactions with glycans are of critical importance for understanding the complex biology of capsid viruses. The three paradigms listed below represent virus families whose glycan binding properties have been successfully studied by CFG investigators.
Glycocalyx engineering toward probing cancer glycome evolution
While altered glycosylation patterns have long been identified as hallmarks of cancer, their functional significance with respect to tumor progression are not well understood. Two examples are overexpression of mucin glycoproteins¬ (densely glycosylated cell-surface molecules with unusual physical properties) and hypermodification of glycoproteins with the terminal sugar sialic acid. These glycosylation phenotypes are found on numerous cancer types with highly varied underlying driver mutations and their magnitude tends to correlate with tumor aggressiveness. To test hypotheses regarding the functional significance of cancer glycomes, we developed an approach to engineer the cell surface “glycocalyx” with chemically defined glycopolymers that emulate cancer-associated structures. Using living polymerization and chemoselective ligation chemistries, we synthesize glycopolymers functionalized with a biophysical probe on one end and a lipid capable of membrane insertion on the other. These biomimetic structures can be displayed on live cell membranes where they acquire functions analogous to natural mucin glycoproteins. With this glycocalyx engineering platform, we identified a role for hypersialylation in protection of tumor cells from innate immunosurveillance. Ongoing work focuses on the effects of mucin overexpression on cell adhesion and signaling.
In vivo imaging of glycans and other biomolecules
We have found bioorthogonal chemistry and other chemical approaches to be powerful allies in studies of glycobiology, a frontier area of biomedical science with limited experimental tools available for probing relationships of structure and function. Glycans decorate cell surfaces and secreted proteins and their structures change during dynamic physiological processes such as malignant transformation and stem cell differentiation. Yet, the precise biological functions of cell surface glycans, considered by many to be the “dark matter” of the cell, are largely unknown. Cancer-specific glycosylation patterns have been identified on tumor tissue ex vivo, prompting much excitement around their potential use as targets for molecular imaging and as disease biomarkers. To realize these possibilities, we developed a technique for glycan imaging that involves metabolic labeling with synthetic azidosugars followed by in vivo bioorthogonal reaction with optical or MRI contrast probes. We employ this chemical imaging platform to study glycomic changes associated with cancer, as well as during embryogenesis and microbial infection using zebrafish as a model organism.
Recently we pursued an analogous method for imaging bacterial peptidoglycan using azide- or alkyne-functionalized D-alanine analogs as metabolic labels. We are using this technique to probe spatiotemporal changes in peptidoglycan biosynthesis associated with infection by the intracellular pathogens such as Mycobacterium tuberculosis and Listeria monocytogenes. More broadly, we consider metabolic labeling with bioorthogonal “chemical reporters” as a general approach to interrogate biomolecules that might not be accessible to conventional genetic reporter strategies.
Bacterial Adhesins and Lectins
Infection by bacteria is generally initiated by the specific recognition of host epithelial surfaces by adhesins and lectins. These GBPs are therefore virulence factors that play a role in the first step of adhesion and invasion. The GBPs are called adhesins when they are part of organelles, such as fimbriae and flagella. They are referred to as lectins when they are soluble, and lectin-domain when attached to other proteins, which are often hydrolytic enzymes also involved in the infection process. The human targets for bacterial adhesins and lectins are mostly fucosylated human histo-blood group and/or sialylated epitopes. Defining the biological role of bacterial adhesins and lectins together with their structure and specificity is a prerequisite for the development of strategies for inhibiting their binding to human tissues.
The C-type lectin family consists of proteins with diverse overall organization that contain structurally related carbohydrate-recognition domains. Although they generally share a common mechanism for interacting with sugars through a bound calcium ion, the spectrum of ligands bound by different members of the family is diverse and can include both endogenous mammalian oligosaccharides as well a sugar-containing structures on pathogenic micro-organisms. The biological functions of the C-type lectins are correspondingly diverse, but many of the best understood examples are membrane receptors found on the surface of cells of the immune system, which mediate interactions of these cells with each other and with viruses, bacteria, fungi and parasites, while other members of the family are soluble mediators of innate immunity. Outside the immune system, members of this group participate in clearance of circulating glycoproteins.
- Galectins are a family of glycan-binding proteins that are expressed in all multicellular organisms, in virtually every cell and tissue, and that vary considerably in function. There are 15 overall genes encoding galectins in different animals and 11 are expressed in humans.
- All galectins share a consensus sequence of about 130 amino acids and a homologous carbohydrate recognition domain (CRD) that specifically binds many different types of glycans, including those containing b-galactosides and poly-N-acetyllactosamines, but also including blood group antigens, and sialic acid- or sulfate-containing structures found in O- and N-glycans.
- The jellyroll-like conformation of the CRD, the hallmark of the galectin family, is composed of two anti-parallel b-sheets that establish a b-sandwich Differences in ligand specificity among this family are determined by specific amino acids in the CRDs, allowing recognition of different modifications of galactose-containing glycans, thus defining the affinity of a particular galectin for specific glycoprotein or glycolipid receptors in a certain tissue or cell type. Galectins are synthesized in the cytoplasm and secreted via a non-classical secretion pathway, so that galectins are found in a variety of intracellular compartments, as well as in the extracellular milieu of almost every cell and tissue type.
- There are three structural subfamilies of galectins – the prototype, the chimeric, and the tandem repeat. The three paradigmatic galectins described below represent the three structural subfamilies. The prototypical galectin subfamily include galectins-1, -2, -7, -10, -13, and -14; the chimeric galectin subfamily has a single representative galectin-3; and the tandem repeat galectin subfamily includes galectins-4, -8, -9, and -12. Recognition of cell surface glycans by many of the galectins, which is associated with oligomerization and lattice formation of the receptors, induces signaling pathways that are being well defined in leukocytes and epithelial cells. In addition, galectins are important in innate immune responses and can directly recognize glycans on pathogens and provide protection independently of antibodies.
Both calreticulin and the cation-dependent mannose-6-phosphate receptor are glycan-binding proteins recognizing discrete and specific carbohydrates involved in intracellular trafficking of membrane and secretory glycoproteins. In the case of calreticulin, glycan binding is associated with the quality control apparatus ensuring that only properly folded proteins will exit the endoplasmic reticulum. In the case of the cation-dependent mannose-6-phosphate receptor, phosphorylated mannose residues are recognized, allowing lysosomal proteins to be properly targeted. These paradigms are important for understanding diseases of protein misfolding and many lysosomal storage disorders. The ficolins are a class of glycan-binding proteins with affinity for N-acetyl residues that have distinct fold from other known lectins yet mediate pathogen recognition. Polymorphisms in ficoloin genes may also have pathophysiological implications.
Bacteria, virus and other pathogens need an acidic environment and attaches themselves to sugar-rich carbohydrates to grow. Those with genes related to low immune system and poor metabolism of sugar and carbs are more affected.
Eat whole foods rich in fiber , less sugar , to fight cancer profileration.