To the lay public, sugars are the villains behind expanding waistlines and rotting teeth, and until recently, the view from the lab bench was not much different. Sugars were considered so irrelevant, says Massachusetts Institute of Technology researcher Ram Sasisekharan, that biochemists "mainly developed tools to remove them from proteins they were studying."
Today, however, the mood is decidedly different. Sugars, or more properly, the complex sugars called glycans, are now recognized as critical mediators of cell-cell communication, playing roles in cancer, infection, immunity, and even the interactions between egg and sperm. Studies of knockout mice for glycan-related proteins suggest a prominent role in development, and a number of human glycan-related genetic disorders have been identified.12
On the technology front, improved analytical techniques are helping researchers see the depth and breadth of biologically relevant carbohydrates, while enhanced synthesis methods ease the development of tools to screen proteins for their sugar-binding specificities. Genome-sequencing projects, meanwhile, continue to pour in data on glycan-related genes and enzymes.
The resulting critical mass of information and techniques has spawned a new 'omic-level discipline. Carbohydrate researchers have coined the term "glycome" to describe not only the total cellular complement of ubiquitous, structurally complex, polymeric sugars, but also the many genes and gene products involved in glycan synthesis, binding, and regulation.
These are interesting biological issues, to be sure, but many "glycomicists" are motivated by more practical problems. Many biopharmaceuticals, including antibodies and some hormones, are glycoproteins in their native states,3 and proper glycosylation can be critical for these drugs' activities and bioavailability, and to prevent adverse immunological reactions, says Sarah Harcum, a chemical engineer at Clemson University. Harcum studies glycosylation pathways in the cultured mammalian cells commonly used to produce such drugs. A number of biotechnology companies have thus emerged to exploit the need for analysis, testing, and production of properly glycosylated protein therapeutic agents.
STRENGTH IN NUMBERS
In 2001, the National Institute of General Medical Sciences gave its ringing endorsement to the glycome concept by establishing the Consortium for Functional Glycomics, headquartered in La Jolla, Calif. According to James Paulson, consortium leader, the initial group of 40 participating investigators has grown in three short years to more than 110, with groups participating both in the United States and Europe. The resulting flood of data simultaneously thrills and sobers investigators. "The complexity is quite enormous," says Richard Cummings, a consortium participant and a founding member of the Oklahoma Center for Medical Glycobiology. "It's going to take a supercomputer to piece it all together."
Glycomics owes its complexity to the fact that polymeric sugars can be strung together using a choice of several different attachment points on each monomer. Besides leading to an astronomically large number of possible arrangements, multiple attachment points allow for branched sugars. The potential structural diversity of glycans far exceeds that of either DNA or proteins, whose monomers generally link together using a single attachment point at each end. This results in an "extremely large potential dataset," says Sasisekharan, who has reason for concern, as the leader of the consortium's information/bioinformatics core facility.
But he and Paulson express optimism. While they are "still exploring the true extent, it is a finite series," says Paulson, in part because biological systems evidently do not use every possible sugar combination. "This is a very manageable set of information," he asserts confidently.
Their optimism is supported by recent expansion in the glycomics field, indicating that many hands are available to the task. Cummings points out that the number of glycomics-related citations has been steadily increasing, from about 10,000 per year in 1990 to well over 60,000 per year today.
Much of this increase can be credited to improvements in existing analytical techniques. In particular, "the huge amount of progress in mass spectrometry over the past five or six years enables us to tackle the problems [in glycomics] much more effectively," says Sasisekharan. Cummings agrees, calling the technique a "godsend." While other techniques, such as nuclear magnetic resonance, can be used to investigate glycan structures, mass spectrometry has an advantage in that it can use very small samples. This is critical since many glycans are produced as complex mixtures, often containing only a very small amount of each individual molecular species.
GOTTA HAVE A CHIP
No self-respecting 'ome these days would be without a microarray, and the glycome is no exception. Improved synthesis methods have led to the development of glycan arrays, which are already facilitating faster screening of proteins, cells, and tissues for their carbohydrate-binding specificities.4
© 2004, GlycoFi Inc
This image of interferon-β (ball-and-stick model) with its high-mannose glycan modification (space-filling model) underscores how glycosylation can affect protein structure and function. Thanks to advances in mass spectrometry, scientists are beginning to catalog and enumerate these complex sugars, the first step in learning how to recognize – and possibly repair – aberrant patterns.
Members of the Consortium for Functional Genomics have developed an array chip spotted with 185 different glycans, many with complex branching patterns. Enzymes or other carbohydrate-binding proteins of unknown specificity are passed over the array, and binding to specific sugars is detected using fluorescent labels. This chip replaces labor-intensive binding assays performed in test tubes.
Bacterial or eukaryotic cells also can be passed over such arrays to test the specificities of carbohydrate-binding proteins located on their cell surfaces. Goverdhan Sachem's group at the Oklahoma Center for Medical Glycobiology, for instance, is using fluorescently labeled
Synthetic chemistry is especially critical in glycomics, because sugars are not readily amplifiable in the sense that DNA is amplified using PCR, nor are synthetic methods as mature as those for peptides, according to Milan Mrksich, University of Chicago, whose group studies ways to improve carbohydrate microarray production and analysis. In addition to glycan synthesis, improved chemistry methods are needed to attach sugars to chip substrates in a defined, bioactive orientation.
Mrksich's group has made substantial progress in this area. His team has also developed a method of using mass spectrometry to identify carbohydrates and proteins interacting on a chip directly, without resorting to biochemical intermediaries such as antibodies or lectins, which are carbohydrate-binding proteins of known specificities. This method should facilitate high-throughput screening for drug discovery, particularly since it requires no preconceived notions about the identities of the molecules present in a biological sample, says Mrksich.
BUILD IT AND THEY WILL COME
The critical importance of glycosylation in biopharmaceutical production, in combination with these improved methods, has led to an increasing number of biotechnology companies intent on filling a glycomics niche. Among the problems these companies are eager to address is the fact that biopharmaceuticals are typically produced in cells that lack human glycosylation pathways, such as bacteria, yeast, or cultured Chinese hamster ovary cells.
The protein drugs produced in these cells contain a mixture of glycosylation patterns, or glycoforms, that is not necessarily typical of the human form of the protein. Moreover, even human-produced glycoproteins are often mixtures of different glycoforms, some of which are more stable or bioactive than others. These companies are concentrating on methods to analyze carbohydrate structures more quickly and efficiently, to separate them from mixtures so they can be tested individually for efficacy and stability, and to control the final glycosylation pattern during synthesis so the drug can be produced more cost-effectively.
The rewards of this approach are potentially huge. Scientists at Amgen of Thousand Oaks, Calif., engineered two new glycosylation sites into human erythropoietin. Commonly used to treat anemia in patients with cancer and dialysis, erythropoietin represents an approximately $3 billion annual market. The new drug, called Aranesp, shows a greatly extended half-life in the human body.
Two European companies, UK-based Procognia, and Glycotope of Berlin, are capitalizing on their technological strengths to offer glycan-analysis services to larger pharmaceutical companies. Researchers at Procognia have developed a lectin microarray that can be used to rapidly characterize and quantitate the sugars in mixtures of glycosylated proteins. This lectin microarray produces a glycan "fingerprint" that can be used, for example, to rapidly screen culture conditions for their effects on the production of a desired glycoform. Glycotope offers glycan analysis using more traditional biochemical methods, such as HPLC and mass spectrometry. Researchers at this company are also working on engineering mammalian cell lines that will produce proteins with only selected forms of glycosylation.
GlycoFi, based in Lebanon, NH, has attacked the problem of finding and producing optimal glycoforms from another angle. As chief science officer Tillman Gerngross describes it, researchers at GlycoFi performed a "gut rehab" of the yeast glycosylation machinery. They removed the yeast-specific glycosylation genes and replaced them with their human counterparts to make a series of yeast strains, each of which produces only a single glycan for addition to proteins. These strains can then be used to produce glycoforms of therapeutic proteins individually for testing, eliminating costly and time-consuming separation steps. This system also minimizes the chances of adverse immunological effects "since we are playing only with the human repertoire" of glycans, according to Gerngross. Plus, yeast can produce much larger amounts of protein than mammalian cells, and their growth properties are well-known owing to years of use in food technology, he adds.
ON THEIR OWN
Sugars are not just the faithful sidekicks of polypeptides; they also have their own bioactivities. A few companies focus on using glycans as therapeutic molecules in their own right. Sasisekharan is a cofounder of Momenta Pharmaceuticals, Cambridge, Mass., which hopes to capitalize on his research group's many years of experience with heparins. Heparins are linear glycans belonging to a group known as the glycosaminoglycans, which "coat virtually all eukaryotic cell surfaces," he says. Heparins act as anticoagulants and are used to treat a variety of clotting disorders; they represent a $3 billion-a-year market.
Pro-Pharmaceuticals of Newton, Mass., develops glycans as an adjunct to improve the safety and efficacy of existing chemotherapeutics. Cancer cells contain specific sugar-recognition sites that are different from those of normal cells. Sugars that bind to these sites can be covalently attached to or administered with anticancer drugs. This modification improves targeting to cancer cells, while interfering with absorption in the liver, a major factor in toxicity. The company's compound Davanat is currently in Phase I clinical trials for combination therapy with 5-fluorouracil, one of the most frequently used and highly toxic anticancer agents.
Also in clinical trails is GCS-100, a sugar-based drug from GlycoGenesys of San Diego. GCS-100, which targets the animal lectin galectin-3, appears to mount a three-pronged attack against cancer. It causes cancer cells to apoptose, interferes with angiogenesis, and hinders metastasis by interfering with cellular adhesion.
Johanna Griffin of Procognia points out that the business opportunities in glycomics are not limited to analyzing and synthesizing glycans. Growing evidence suggests that changes in the patterns of glycans may be biomarkers for disease states, such as cancer or autoimmune disease. Israeli biotech Glycominds, for instance, has discovered a biomarker for multiple sclerosis (MS), an autoimmune disease. According to the company, this biomarker will allow faster, more cost-effective diagnosis of MS, including more rapid classification of the disease's severity in individual patients. The company's GlycoChip, a commercially available carbohydrate microarray, facilitated the marker's discovery.
As usual in biotechnology, many of these companies are spin-offs from academic research. One endeavor that has been particularly successful in this regard is the Oklahoma Center for Medical Glycobiology. According to Cummings, this center has produced more than 50 glycobiology-related patents and four biotechnology companies since its founding in 1999.
Research at the center covers a wide range, including the role of abnormal glycosylation patterns in the lungs of patients with cystic fibrosis, the carbohydrate-binding specificity of the influenza virus, and the impact of glycans such as hyaluronic acid on synovial fluid and joint-replacement devices. Cummings' own laboratory group studies the role of glycans in parasitic infections by worms and protozoans, including the organism responsible for malaria.
Vernon Reinhold, a longtime researcher in the field and who is credited with coining the term glycome, describes the future succinctly: "Now that glycosylation has taken hold as really important in biological function, it will start opening the doors into what these structures do." That future should provide the sweet smell of success for glycomics researchers.
Megan M. Stephan