The future of protein-based therapeutics is based on technologies to provide accurate and robust analysis, however protein profiling remains a major challenge in drug discovery and development. Protein characterization is among the topics to be discussed at the upcoming Cambridge Healthtech “PepTalk Conference” in San Diego.
Researchers at Wyeth (www.wyeth.com) are using an integrated approach to characterize glycoproteins. By combining top-down mass spectrometry with a bottom-up approach, “we can get even more specific information about what the amino acid sequence is, as well as do a site-specific determination of where post-translational modifications are,” explains Joseph McClennan, Ph.D., department of characterization and analytical development. The top-down approach involves looking at an intact protein or a sub-unit of a protein and provides information on heterogeneity and whether there are modifications. The bottom-up approach includes peptide mapping with MS identification of the peptide. “When you couple these two things, it gives you confidence that you are characterizing and observing a lot of different things about the proteins,” he adds.
This integrated approach differs from the traditional approach, which uses SDS- PAGE, chromatography, and some type of cation exchange analysis of the protein.
Dr. McClennan says that, traditionally, these techniques would drive all characterization. However, his group is now doing top-down and bottom-up analysis first. “We’re using it to drive what type of traditional analytics are done and how much work is going into characterizing certain peaks observed in SDS-PAGE. So, it’s much more of a flow, integrating all these things to get the covalent glycoprotein structure.”
Overall, this translates into time and cost savings. It enables quick screens without the full array of traditional analytical techniques. If integrated into the whole development structure, says Dr. McClennan, it provides faster results, which in turn, allows faster development of the cell culture-purification process. His group has shown this method also works well for antibodies, and they have started using this strategy with several commercial products that are more complex recombinant glycoproteins.
Protein Silencing with Phylomers
Antibodies are not well-suited for target validation because of poor tissue penetration, difficulty delivering to cells, and high production costs. This has increased demand for an alternative to validate protein complexes both inside and outside cells. Phylomers®, developed by Phylogica (www.phylogica.com), are a new class of peptides derived from protein subdomains small enough to synthesize easily and deliver into cells. They are derived from natural proteins and are therefore easily soluble, enabling potential oral delivery, according to the company.
“We realized that nature has come up with an enormous amount of structural diversity within the few thousand classes of scaffolds found within natural proteins,” says Paul Watt, Ph.D., vp, technology development, drug discovery.
The company has created large libraries with more than 260 million different parts of proteins from biodiverse sources, which are screened for blockers of protein interaction. “We find we get encouraging hit rates—better than ones obtained by other approaches to blocking protein interactions.”
The reason, he says, may lie in the structural diversity, not the numerical diversity.
The source of Phylomers is fully sequenced microbial genomes, which allows easy analysis of hits and also enhances their potency against human protein targets. Their small size (15–50 amino acids) also reduces the probability of immunogenicity issues.
Dr. Watt says the company is currently screening Phylomers against factors involved in inflammatory diseases, and that one of the most interesting applications is for intracellular targets. “We’ve had success screening against transcription factors and have achieved in vivo efficacy. This is exciting because you can target a whole new range of targets inside cells, which are difficult to access with antibodies.”
Characterizing Ligand Interactions
Drugs have been discovered from many proteins by determining their inhibitory effects on a specific, measurable activity of the protein. However, many new protein targets have no direct activities to measure, as is the case for many membrane-bound receptors.
Light scattering and analytical ultracentrifugation provide information about molecular interactions. These techniques can accurately measure the size of a protein, says John Doran, Ph.D., research scientist, protein biochemistry at Vertex Pharmaceuticals (www.vpharm.com).
Light scattering relies on the fact that the size of a protein influences the intensity of scattered light from an incident light wave, whereas with analytical ultracentrifugation, the protein size affects its behavior in a centrifugal field. Equations may be derived from fundamental first principles that permit the accurate determination of the protein size.
Once the size is accurately known, it is possible to define the concentration of all the species in solution, and in the case of analytical ultracentrifugation, associative interactions between species can be described. Analytical ultracentrifugation quantitates dissociation constants, which describe how tightly a ligand, such as a drug, binds to a protein. “It is desirable to quantitate these constants so that when comparing a series of compounds, one can correlate the structural variations of the compounds with tightness of the binding,” explains Dr. Doran.
Multiplex Protein Biomarker Profiling
Multiplex Biosciences (www.multiplexbiosciences.com) recently merged with Rules-Based Medicine (www.rbmmaps.com) and plans to expand its immunoassay panel for testing biomarkers in plasma. “Rules-Based Medicine can test up to 130 human analytes and wants to expand that up to 200 new assays over the next 18 months,” says Dominic Eisinger, Ph.D., president, Multiplex Biosciences. The protein biomarkers are from various sources, hormones, cytokines, or anything that provides a good indication of response to disease, disease diagnosis, and drug response.
“Multiplexing is much more cost-effective than single assays,” explains Dr. Eisinger. “Mass spectrometry is a great technique for discovery, but it doesn’t have the sensitivity to detect low abundant proteins.”
However, he says for plasma protein biomarker analysis, the quantitative sensitivity of the miniaturized multiplex immunoassays is in the single femtomole range from just a few microliters of raw unfractionated plasma. This high sensitivity is required for low-abundant tissue leakage factors, a promising biomarker candidate.
The company has reagents that prevent immunoassay interference. “There can be a lot of interference when you multiplex and have around 40 antibodies in a single well all working in concert. We work our permutations to make the assay as sensitive as possible,” says Dr. Eisinger. The company’s focus over the next year will be to build more multiplex miniaturized immunoassays for screening services. There have already been several novel biomarker discoveries made out of the company’s 130-analyte panel.
Current technologies allow for profiling thousands of proteins in plasma and discovering ones modulated by disease or drugs. Researchers at Caprion Pharmaceuticals (www.caprion.com) have developed CellCarta®, a platform for profiling proteins in plasma, other bodily fluids, or tissue.
“We use a combination of global proteomics and the reductionist approach,” says Paul Kearney, Ph.D., executive director, bioinformatics. “We try not to reduce the data set of proteins modulated by disease or drugs, but do try to keep all the modulated proteins for several reasons. We can perform a global assessment of not just the protein being modulated, but the biological processes that are modulated by disease or a drug. This allows you to study mechanism of action.”
Furthermore, he says, this enables a robust panel of biomarkers. “The global approach allows a much deeper understanding of what’s happening biologically to your proteins.”
It also allows quantification of disease severity and drug effects and provides information on pharmacodynamic questions like dose response and compound efficacy, as well as eliminates the need to develop an ELISA test. “You get answers to these questions immediately, without going into the additional 15 months of ELISA development and validation and taking the risk of failure,” adds Dr. Kearney.
The company has its own internal plasma biomarker discovery program on human plasma clinical samples in Alzheimer’s disease and hypertension. “Our proteomic assay correlates highly with the current diagnostic test used by most clinics to assess Alzheimer’s, the Mini Mental State Exam.”
Using MS-based techniques, researchers at Centocor Research and Development (www.centocor.com) are characterizing antigen structure early in the development process. “Understanding our antigens has helped to drive programs and make decisions faster and more accurately than we have normally been able to do,” says Jennifer Nemeth, Ph.D., senior research scientist, discovery research.
“Drug companies are going to characterize their antibodies, but not all will characterize their antigens. This is where the potential novelty exists.” She adds that most products being developed now are antibodies or antibody-based. “These are easy to develop because their structure is similar, and production and purification lessons can be applied from project to project.”
The group’s antigen-characterization approach started because of an increased interest in understanding the molecules they were working with. According to the company, the potential time savings using this approach could equal up to a year.
Some standard quality control analyses in biology include SDS-PAGE gels, SEC (size-exclusion chromatography) for purity, and N-terminal sequencing to look at the first 15 residues of the protein. However, Dr. Nemeth says advances in high-resolution mass spectrometry have enabled antigen characterization, along with crystallography studies.
“You want to know the tertiary structure of your antigen and determine where the epitope is located. You can get that from crystallography studies. We’re starting to routinely do crystal structures on each antigen and then the antibody complex.
“Some of this characterization work is being done earlier—three to four years before an IND filing. We’re investing more up-front with the idea that in the long run, we’ll spend less money and develop molecules faster. We’re applying this methodology across all our programs. This is now our standard paradigm,” notes Dr. Nemeth.