New proteomics tools enable researchers to dive deeply into signaling networks, allowing them to tease out interactions among key molecules. But this comes with a new challenge of increased complexity. Can cell signaling scientists balance the bewildering complexity that comes with the discovery power of proteomics technology?
Proteomics seems a natural discipline toward which cell signaling researchers might gravitate. Most signaling molecules studied today are proteins, after all. But proteomics presents much greater complexity than, say genomics, which many cell signaling scientists grew up learning. “Proteins are very complex, often functioning under only a specific set of conditions, which makes multiplexing and parallel analysis difficult,” says Chris Hebel, vice president of business development at LC Sciences. Add to this the multifold interactions between these complex molecules, and you have an explosion of possible questions to ask and data to be gathered. Alterations to proteins in the form of posttranslational modifications (PTMs) such as phosphorylation, acetylation, myristoylation, glycosylation, and more, can further change the functions of these proteins, their binding partners, and their signaling properties, perhaps even initiating a cascade of other biochemical changes across the cell.
Uncovering signaling pathways with mass spectrometry
Cell signaling research has recently benefitted from a combination of serial enrichment methods and advances in mass spectrometry technology. A prime example is “affinity-based methods for selective enrichment of PTMs,” says Steven Carr, the director of the Proteomics Platform at the Broad Instituteof Harvard University and the Massachusetts Institute of Technology. For example, recent refinements of metal-affinity enrichment methods have helped optimize samples for serine/threonine and tyrosine phosphorylation detection. When researchers use such pre-enriched samples as their starting sample for liquid chromatography/mass spectrometry (LC/MS), they “can get much, much deeper into these samples,” says Carr. “It’s very common now to get 20,000–30,000 phosphorylation sites identified, which is really quite deep.”
The MS technologies mainly used by Carr’s lab are the new Q Exactive systems from Thermo Fisher Scientific. “Previously it was Orbitraps, but now we’re heavily focused on using Q Exactive Plus and Q Exactive HF instruments, because the price per performance point is very good. We run a high throughput laboratory with many concurrent projects so having many high-performance instruments is essential,” says Carr. Other high-performance MS systems are commercially available from other vendors, including AB SCIEX and Waters.
“These new generations of hybrid instruments are extremely sensitive while maintaining very high performance in terms of mass accuracy and high resolution—we take that for granted nowadays,” says Carr. Eight to 10 years ago, researchers had to pay a price. “Either you gave up sensitivity for the performance factors, or you maintained sensitivity but gave up mass accuracy and high resolution. Today, you don’t have to make those compromises.”
And MS systems continue to be refined. “Not all of them have what’s called ion funnels on the front end,” says Carr. An ion funnel is a device that makes the MS system more sensitive. Waters, Agilent Technologies, and Thermo Fisher Scientific offer ion funnels on certain MS systems. “Another thing that will change is more widespread availability of ion mobility in MS systems,” says Carr. Ion mobility helps to separate peptides and proteins from one another in the gas phase to make the sample less complex, which Carr likens to gas phase chromatography. “This also improves sensitivity and can increase speed,” he says.
Other new developments in MS systems include the ETD option on Thermo’s Orbitrap Fusion. ETD stands for electron transfer dissociation, valuable as a “type of fragmentation that lets you map PTMs to your protein backbone,” says Andreas Huhmer, director of 'omics marketing for life science mass spectrometry at Thermo Fisher Scientific. Thermo also offers kinase probes for pulling out kinases. The Thermo Scientific Pierce Kinase Enrichment Kit uses ActivX ATP or ADP Probes to label kinases’ active ATPase sites. The probes contain a desthiobiotin tag that allows subsequent enrichment of labeled kinases.
Mark Knepper, senior investigator in the Epithelial Systems Biology Laboratory at the National Heart Lung and Blood Institute (NHLBI), part of the National Institutes of Health, also uses recent MS advances to study phosphoproteomics. “Progress in the development of mass spectrometers allows more spectra to be collected with greater mass resolution,” he says. “This results in remarkable increases in sensitivity.”
Knepper’s lab builds models of signaling networks by identifying specific protein kinases and phosphatases. Using the CRISPR-Cas9 genome editing system, the group can delete a particular kinase or phosphatase gene, and then study the subsequent engineered clones. “Phosphoproteomics in CRISPR clones can potentially allow conclusions about the role of specific gene products in specific signaling pathways,” says Knepper
The power of good antibodies
At their best, antibodies are indispensable for molecular tagging, identification, and affinity purification. But antibodies come with a downside. “There’s just not enough of them,” says Carr. “Most of them don’t work in all contexts, and some of them don’t work at all. So reliance on antibodies remains very challenging.” As with MS, recent improvements in antibody technology have helped advance cell signaling and proteomics research. Yet according to Carr, “mass spectrometry has played a major role in unraveling the off-target effects of antibodies.” Not all antibodies are created equal—some work well in some experiments, such as Western blotting or immunohistochemistry, but not for others, like immunoprecipitation.
Antibody technology has improved greatly for targets with PTMs, led mainly by “motif antibodies” from Cell Signaling Technologies (CST). “CST developed a class of antibodies that were designed around a particular PTM sequence motif as opposed to a site-specific PTM epitope,” says Jeffrey Silva, KinomeView and PTMScan proteomics service manager at CST. Examples include phosphorylated serine/threonine targets and phosphorylated tyrosine targets. Although it does not make a motif antibody for glycosylation, CST does have antibodies against acetyl-lysine, methyl-arginine, methyl-lysine, succinyl-lysine, and phospho-histadine in the works
Silva explains that CST developed the PTMScan Direct reagents—combinations of CST’s motif antibodies—in response to customers’ requests for tools to pinpoint six different signaling areas: tyrosine kinases, serine/threonine kinases, apoptosis, cell cycle and DNA damage, AKT and PI3 kinase signaling, and a Multipathway Reagent that enables the identification of approximately 19 major signaling pathways. This lets researchers sample key nodes from a number of different critical signaling pathways in a multiplex LCMS assay. “It’s an efficient way to ask just which key pathways are affected by the stimulus—then they can quickly focus their efforts on the affected modulated proteins,” says Silva. All in all, the Multipathway Reagent allows researchers to identify and quantify 409 proteins and 1,006 unique phosphopeptides. CST has also developed new PTM antibodies for ubiquitination, acetylation, and cleaved caspase substrates.
Antibodies to G-protein-coupled receptors (GPCRs), which are important in multiple signaling pathways, are in high demand, but historically they have been difficult to generate. The biggest challenge is “finding antibodies of sufficiently high affinity to distinguish between closely related receptors and receptor subtypes,” according to Lora Tebbetts, product manager at Enzo Life Sciences, which offers over 120 antibodies for GPCRs and associated proteins. Tebbetts says that more specific antibodies are made using “peptides from the more sequence-diverse areas of N- and C-terminal regions [of GPCRs], or the third intracellular loop combined with carrier proteins as immunogens.
Deciphering signaling networks with microchips
One of the advantages of using microchips for cell signaling research is that only small amounts of reagents and samples are needed. LC Sciences’ new phosphopeptide microarray is designed on a microchip to assess changes in expression levels of key signaling proteins in different pathways—all at once. It can “map tyrosine-phosphoproteome interaction networks by detecting the expression level of proteins containing [the] corresponding phosphoprotein-binding domains,” says Hebel. This is unique in that researchers can see at a glance where their experimental manipulations are affecting signaling pathways.
Hebel says that there are multiple reasons to use phospho-binding domains on a custom microarray instead of the more traditional antibody binding. “The probe density is much higher, which translates into higher specificity,” he explains, and the technique does not depend “on the availability of high-affinity antibodies [or] require many different antibodies with different affinities and optimal binding conditions—which would require complicated redevelopment and validation if any design change is made.”
Phosphoproteins are also the topic of study for James Heath, professor of chemistry and director of the NanoSystems Biology Cancer Center at the California Institute of Technology. Heath uses single cell proteomics to study phosphoprotein signaling cascades, with glioblastoma cells from brain tumors as a model system. Like many types of cancer, these tumors are not made of one kind of cancer cell, but rather a heterogeneous population. This heterogeneity is thought to explain why single cancer drugs that target one signaling protein often fail. Heath studies the signaling cascades of glioblastoma cells to learn how to combine cancer drugs to make effective therapies.
Heath’s group uses homemade microfluidic chips to study individual cells, which are positioned in their own tiny chambers inside the microchip. Each cell is separately lysed, so that its contents are captured by the waiting array of antibodies that lie within each chamber. The team then uses quantitative sandwich ELISAs that are calibrated to measure the copy number of proteins per cell. The array includes antibodies against proteins known to participate in glioblastoma signaling pathways, such as phosphoAKT, phosphoERK, phosphoSRC, and phosphoEGFR. Heath uses “targeted drugs to hit signaling pathways, like those driven by EGFR, that maintain the tumor and help it grow,” he says.
Treating a tumor with a cancer drug may cause it to stop growing, or even to shrink. But in all cases of glioblastoma, says Heath, the tumor soon develops resistance to the drug and starts growing aggressively. Yet for some tumors, this isn’t the result of Darwinian selection; Heath found that the resistant cells were not simply survivors—they were adapters.
“The same cells that were responding to the drug, actually developed resistance to the drug,” says Heath, by activating particular signaling pathways—the same interactions that Heath saw when analyzing single cells before and after drug treatment. “If you can identify those other pathways that are activated by the drug, that tells you in principle what combination therapies you would use to treat that resistance,” says Heath. In other words, if you can find a second drug to stop the adaptation response, then you can kill the tumor.
Heath’s results hold promise for cancer therapy. For example, if he treats a tumor with an ERK inhibitor, or an EGFR inhibitor, the results are unimpressive. But if he treats a tumor with two inhibitors that can work together to prevent drug resistance, he says the tumor “completely shuts down.”
From single cells to tissues
Studying single cells also has great value for better understanding tissues since they are communities of cells, according to Garry Nolan, director of the NHLBI Proteomics Center for Systems Immunology, and professor in the Department of Microbiology and Immunology at Stanford University. “They are interacting and living in the context of each other,” he says. “Their individual biology is really about what makes us normal or dysfunctional.” He sees cancer as an example of a tissue that takes on a life of its own and creates its own context or environment. “If we are to hope to understand the complexity of [the environment], and how drugs might act on that, then we need to understand how this community of players are interacting with each other,” Nolan says. “That means you have to assay as many things as you deem relevant without drowning yourself in information.”
Nolan pioneered a technique called mass cytometry to measure the complement of proteins expressed in single cells simultaneously. Using antibodies against multiple proteins of interest labeled with isotopic mass tags—rather than, say, fluorescent tags—he can multiplex to a greater extent than with microscopy because spectral overlap is not an issue. When an individual cell enters a mass cytometer, it vaporizes and the isotopic tags are read by a mass spectrometer. Using this system, Nolan’s lab was able to generate snapshots of signaling networks using 35–40 markers. Today they can measure about 50 parameters.
Nolan’s lab uses a mass cytometry system called CyTOF 2 from Fluidigm. Recent improvements to the CyTOF 2 make it easier to “do experiments to help better understand the signaling properties of the proteins in different cell types,” says Olga Ornatsky, principal scientist at Fluidigm. Fluidigm is adopting the CyTOF 2 system for imaging of immunohistochemistry-type stained fixed-tissue sections. While conventional immunohistochemistry assays measure three or four fluorescently labeled markers at a time, the CyTOF version can measure more than 30 markers at once. “This will widen the scope of questions that can be asked,” says Ornatsky.
Increasingly, proteomics researchers are using such technologies to study more than just changes in protein levels, as they can now look into the complex interactions of signaling molecules and networks. How PTMs regulate and alter such pathways brings an additional layer to this research, with scientists having to ask: “How do thescome modifications affect one another?” says Carr. And as progressively more powerful proteomics tools bee available, such as better antibodies and MS technologies, researchers will be able to dive even deeper into such questions.
Newly offered instrumentation, apparatus, and laboratory materials of interest to researchers in all disciplines in academic, industrial, and governmental organizations are featured in this space. Emphasis is given to purpose, chief characteristics, and availability of products and materials. Endorsement by Science or AAAS of any products or materials mentioned is not implied. Additional information may be obtained from the manufacturer or supplier.