Josh P. Roberts has an M.A. in the history and philosophy of science, and he also went through the Ph.D. program in molecular, cellular, developmental biology, and genetics at the University of Minnesota, with dissertation research in ocular immunology.
In the world of biomedical science, perhaps no enterprise is riskier—and more lucrative—than drug discovery. Find the magic formula, and you can count on a billion-dollar blockbuster for years to come. Get that formula wrong, and tens of millions of dollars go down the drain.
In part, drugs fail because of side effects, the unanticipated interplay of pharmaceutical and protein. Increasingly, drug developers are turning to chemical proteomics to minimize such events.
Chemical proteomics is just what it sounds like, sort of. Although researchers may differ on the specifics of the history, scope or tools, chemical proteomics (CP, also called chemoproteomics) basically comes down to how small molecules interact with a collection of proteins. “’Chemical’ just means using some kind of small molecule to focus your dataset,” explains Matthew Bogyo, a Stanford University professor of pathology whose lab uses CP to identify proteins that have been modified by the lipid palmitate. “So instead of looking at every single protein a cell makes, you’re looking at a subset of proteins that bind a specific small molecule.”
The term itself was coined in the late ‘90s, Bogyo says. But, broadly speaking, CP covers things that have been done far longer—for example, using radiolabeled drugs to look for targets. “The technique hasn’t really changed that much. It’s just that we’ve come up with better ways to tag and modify the molecules and then to identify things that we’re able to pull out.”
CP simultaneously studies both intended- and off-target interactions with bioactive molecules, for example, and enables screening of enzyme agonists and inhibitors. Says Bernhard Kuster, chair of proteomics and bioanalytics at Technische Universität München, CP is a field of science, “whereas, say, [mass spectrometry] and hot-ATP assays are techniques.”
The main approaches
Because drug development fundamentally is about how drugs do and do not interact with proteomes, CP is particularly attractive in drug discovery, Kuster says. But academic labs use CP, too, for instance to find novel targets for existing drugs or to understand how cell-signaling pathways are regulated.
One of the principal approaches used in CP is activity- or affinity-based protein profiling (ABPP) [1]. Small-molecule probes, typically consisting of a reactive group, linker and tag, are used to bind to and covalently label proteins of interest at their active site. Activity-based probes seek out enzymes with similar structural features and catalytic mechanics, often binding directly to a nucleophilic residue in the active site. Affinity-based probes, after recognizing the active site, covalently attach to the protein at a separate site through a nonspecific mechanism, such as photochemical crosslinking or click chemistry. A known inhibitor, for example, can often be used as an affinity-based probe by joining it to a reactive moiety.
A second approach, sometimes called compound-centric chemical proteomics (CCCP), is more akin to affinity chromatography. Here the small molecule, typically immobilized onto a matrix, is used to fish proteins from cell lysates or tissue extracts.
MS for CP
Especially when the purified proteins are not known ahead of time, researchers typically will identify and quantify them using mass spectrometry (MS). “Most of the big proteomics labs now have these really high-resolution, high mass-accuracy instruments that allow you to make definitive determinations of what your proteins are with very small amount of [peptide] coverage,” says Bogyo.
Kuster concurs, explaining that with up to 5,000 to 10,000 proteins across four to six logs of dynamic range in a given sample, the use of “high-end instrumentation” such as a quadrupole time-of-flight system or Thermo Scientific Orbitrap generally is required to discriminate fragments of interest from background in an MS proteomics experiment.
Other (mostly antibody-based) technologies, such as protein microarrays, 2D Western blots or even ELISA can at least partially tackle the same tasks. “The principal difference being that when you use an antibody-based readout, you already have to know what you’re looking for,” Kuster says. For example, a Luminex bead array can be configured to measure 200 phosphoproteins, in hundreds of samples, relatively quickly and inexpensively. “But if you want to look more broadly and want to do discovery-type of experiments, then currently nothing gets you around using MS.”
One recent study by Benjamin Cravatt’s group at The Scripps Research Institute used limiting concentrations of a electrophilic alkylating probe to label cysteine residues in several mammalian proteomes, quantifying the results using MS [2]. The team reasoned that a difference in alkylation would indicate differences in reactivity rather than abundance. “That allowed us to rank all the cysteines in the human proteome by their reactivity,” says first author Eranthie Weerapana, now an assistant professor of chemistry at Boston College. Cysteines previously annotated as functional—playing a role in catalysis or protein regulation, for example, or subject to oxidative post-translational modification—were enriched in the subset of reactive cysteines. “Extending that, we were then able to identify novel functional cysteines. We could pick a cysteine from our list that we found to be hyper-reactive but was completely unannotated, and we assumed the reactivity meant that it was functional. So we then did more biochemical experiments to actually assign a function to these unannotated cysteines.”
The question for drug developers and researchers alike is where and how to incorporate such data into their workflows. Cravatt and colleague Raymond Moellering argue that CP’s pharmacology-like methodology should be (re)integrated into the earliest stages of the discovery process when possible, providing “proof of relevance” that compounds have both the potency and selectivity to inhibit or agonize a protein of interest before moving them forward [3]. Genetic disruptions may not mirror all the effects of a pharmacological intervention, they note. In fact, many proteins, unrelated by sequence, structure or even function, can nonetheless interact with the same small molecules, and it is vital to the drug-discovery process that such interactions be uncovered early on.
The result, CP practitioners hope, is that drug development itself will become more efficient and on-target—good news for pharmaceutical companies and their potential clients alike.
References
[1] Miao, Q, et al., “Chemical proteomics and its impact on the drug discovery process,” Expert Rev Proteomics, 9:281–91, 2012. [PubMed ID: 22809207]
[2] Weerapana, W, et al., “Quantitative reactivity profiling predicts functional cysteines in proteomes,” Nature, 469:790–5, 2010. [PubMed ID: 21085121]
[3] Moellering, RE, Cravatt, BF, “How chemoproteomics can enable drug discovery and development,” Chemistry and Biology, 19:11–22, 2012. [PubMed ID: 22284350]
Image: iStockPhoto