February 15, 2018 (Vol. 38, No. 4)

Kathy Liszewski

When the Proteome Is Subjected to Analysis, It May Reveal Meaningful Patterns

Like the bubbling contents of a cauldron, the proteome may bring about toil and trouble instead of clear and reliable prophecy. The proteome, however, is no mere witches’ brew of proteins. When the proteome is subjected to analysis, it may reveal meaningful patterns.

For example, in clinical applications, proteomic patterns may support predictions about health and disease. Proteomic analyses—whether they are performed on samples consisting of cells, tissues, or body fluids—may yield trustworthy visions, including clear, comprehensive views of biological systems, as well as dependable representations of large patient cohorts.

The mainstream proteomic approach, termed bottom-up or shotgun proteomics, employs mass spectrometry (MS) technologies. These technologies continue to evolve as investigators seek more speed and sensitivity in their proteomic analyses, as well as the ability to carry out multiplex assays.

MS advances relevant to translational research and clinical practice were discussed at the Twelfth International Symposium on Mass Spectrometry in the Health and Life Sciences: Molecular and Cellular Proteomics. This event, which was organized by the Mass Spectrometry Facility of the University of California, San Francisco, emphasized chemical proteomics, methodologies enabling new insights into biology, and emerging challenges in medicine and biology.

Several presenters noted that clinical proteomics is on the rise because MS can create unbiased, real-time snapshots of the cell proteome and identify pathways of therapeutic resistance. Other presenters predicted that proteomics would gain new powers through microfluidics and nanotechnology.

Reading Obscure Mechanisms of Resistance

Despite the emergence of new anticancer therapies, many patients are resistant to therapy and face poorer prognoses. At the MS symposium, this issue was addressed by Tamar Geiger, Ph.D., professor, Sackler Faculty of Medicine, Tel Aviv University. She described how her laboratory is pursuing a clinical proteomic approach to identify cellular mechanisms causing resistance to anticancer therapy. To identify proteins that increase upon cancer development and remain highly expressed despite anticancer treatments, Dr. Geiger’s team employs a quantitative MS approach (Figure 1).

Dr. Geiger recalled one of her studies as follows: “We assembled two patient cohorts, used macro-dissection of formalin-fixed paraffin-embedded slides to enrich for cancer cells, and performed deep proteomic analysis. Extensive computational examination revealed proteins that changed between patient groups and allowed us to elucidate resistance mechanisms.”

In the breast cancer cohort, the group identified a metabolic pathway that increases cancer cell resistance to chemotherapy. “Inhibition of this pathway,” Dr. Geiger proposed, “could increase therapeutic response of breast cancer, irrespective of tumor subtype.”

“In the second study of melanoma response to immunotherapy,” she continued, “the proteomic analysis revealed pathways that increase antigen presentation, thereby likely affecting melanoma recognition by T cells within the tumor microenvironment.”

Dr. Geiger also commented on the importance of addressing key proteomic challenges: “First, analyzing clinical samples rather than cell lines and animal models requires a much larger research scope, due to the high variability between patients. Working with patient samples, however, offers the great advantage of direct clinical relevance. Second, proteomic expression levels of clinical samples are still much less examined than the genomic levels. Fortunately, several studies in the last few years have shown the potential of proteomics to reveal disease mechanisms that cannot be identified on the genomic levels.”

“I believe that proteomic technology has matured,” she concluded. “It has finally reached a level that allows us to address critical clinical questions.”


Figure 1. Proteomic pipeline for the study of anticancer drug response. Clinical proteomics workflow is initiated with the assembly of patient cohorts, extraction of proteins from formalin-fixed paraffin-embedded (FFPE) blocks, and high-resolution mass spectrometric analysis of the tumor proteomes. Computation analysis of the quantitative information reveals the protein networks that are associated with response to drugs. These proteins may then be targeted to increase treatment efficacy. They may also serve as novel biomarkers of response. [Courtesy Tamar Geiger, Ph.D.]

Taking Proteomic Snapshots

Like the pedals that accelerate or stop a vehicle, certain proteins activate go/no-go mechanisms. For example, activator proteins and repressor proteins may increase or decrease gene expression, respectively. If the relative abundances of such proteins could be captured in a proteomic snapshot, cellular gene-expression patterns could be revealed. Although the proteome is notoriously camera shy, its fleeting expressions are being snapped by scientists who are using a new labeling technique.

“We developed a method to capture and identify the nascent proteome in vivo across many different cell types without disturbing normal growth conditions,” declared Craig Forester, M.D., Ph.D., assistant professor of pediatrics, University of California, San Francisco.

Dr. Forester and colleagues employ cell-permeable O-propargyl puromycin (OPP), an alkyne analog that becomes incorporated into nascent elongating peptides. “Subsequently, labeled peptides are subjected to click chemistry and conjugated to biotin-azide, and then they are captured by streptavidin beads,” Dr. Forester explained. “After enzymatic digestion, samples are analyzed with liquid chromatography-tandem mass spectrometry.”

Dr. Forester’s team illustrated the technique by assessing the response to a potent mTOR inhibitor, MLN128. “We identified more than 2,100 proteins and could follow protein networks of early erythroid progenitor and differentiation states that were not amenable before by alternative approaches,” Dr. Forester asserted. “We see this method as a means to quickly (within a short two-hour pulse of OPP) and quantitatively identify proteomes within many biology contexts while preserving the subtleties elaborated by signaling cascades in their native cellular environment.”

The team is planning animal studies as well as more complex experiments, including investigations that may reveal the nuances of cellular responses to anticancer therapeutics.

Unbiased Discovery of Genome Regulators

The gold standard for surveying DNA-protein interactions and identifying regulatory elements is chromatin immunoprecipitation (ChIP). This technology, however, has several limitations. For example, it requires antibodies that can bind to proteins of interest.

An alternative approach was discussed by Samuel A. Myers, Ph.D., a researcher at the Broad Institute of MIT and Harvard. With his colleagues in the laboratory of Steven A. Carr, Ph.D., the Broad’s senior director of proteomics, Dr. Myers developed a means of identifying proteins associated with a particular genomic locus within the native cellular context (Figure 2).

“We developed a new technology by combining recent advances in chemical biology, genome targeting, and quantitative mass spectrometry,” he explained. “This technology provides a genomic locus proteomics (GLoPro) workflow that can identify proteins occupying a specific genomic site.”

In a proof-of-principle study, Dr. Carr’s team employed an inducible expression system featuring CRISPR/Cas9-APEX-mediated proximity labeling. To create this system, the team fused the catalytically dead RNA-guided nuclease Cas9 (dCas9) to the engineered ascorbate peroxidase APEX2. Then the team transfected its dCas9-APEX2 (Caspex) plasmid along with single guide RNAs (sgRNAs) into HEK293T cells.

“The GloPro system relies on the localization of the affinity labeling enzyme, APEX2, directed by a catalytically dead CRISPR/Cas9 system and sgRNA to biotinylate proteins in a very small radius at a specific genomic site,” explained Dr. Myers. “Cells are subsequently lysed, biotinylated proteins are captured, and enzymatic digestion is performed for protein identification via liquid chromatography–mass spectrometry. Aside from expressing the CASPEX protein and the sgRNAs, there is no need for genomic engineering or cellular disruption to obtain a snapshot of proteins associated with the genomic locus.”

Dr. Myers and colleagues will continue to develop and optimize the system, which they consider to be an orthogonal and highly complementary approach to ChIP. “We plan to make this system available to the whole scientific community so that others can optimize it for their particular use,” noted Dr. Myers. “We see many applications for the unbiased discovery of proteins that regulate gene expression and chromatin structure.”


Figure 2. Genomic locus proteomics (GLoPro) relies on CRISPR/Cas9-APEX-mediated proximity labeling to identify proteins that occupy specific genomic sites. (A) Expression of a gene of interest (white box) is regulated by the proteins that occupy its proximal promoter (blue circles) and, potentially, enhancer elements containing their own proteins (green circles), provided these elements are near. (B) CASPEX binds the predefined locus determined by the sgRNA sequence and creates a reactive cloud of biotin-phenoxy-radicals from biotin-phenol upon addition of hydrogen peroxide. (C) Only proteins proximal to CASPEX are covalently labeled with a biotin moiety (orange stars) for subsequent enrichment and identification via mass spectrometry. Samuel A. Myers, Ph.D., the Broad Institute of MIT and Harvard.

Quantitative Chemical Proteomics

Markus Schirle, Ph.D., senior investigator at Novartis Institutes for BioMedical Research, said that he is focused a comprehensive strategy for the identification of targets such as small-molecule drug candidates. This strategy, he emphasized, incorporates quantitative chemical proteomics.

Dr. Schirle and colleagues are asking two key questions: 1) What is the cellular target of a compound, and 2) what is its mechanistic link to the observed cellular effect?

The common theme of their studies is the combination of affinity chromatography using an affinity probe derived from the original bioactive compound for enrichment of protein interactors, and quantitative MS based on isobaric labeling tags for protein identification and quantitation. Dr. Schirle and colleagues also rely on several orthogonal strategies.

Dr. Schirle concluded by expressing confidence in the power of chemical proteomics: “Chemical proteomics has the ability to provide proteome-wide compound interaction profiles from any disease-relevant cell line, tissue, or organism. This ability will continue to improve with the development of more sensitive and faster next-generation mass spectrometers and higher-order multiplexing approaches.”

Envisioning Proteomic Possibilities

Ralph A. Bradshaw, Ph.D., CSO at Trefoil Therapeutics, and professor emeritus, department of physiology and biophysics, University of California, Irvine, provided his perspective on the future of proteomics. He pointed out that one of the challenges in contemporary proteomics is the need to better characterize and understand protein post-translational modifications.

“Although studies often focus primarily on phosphorylation and acetylation, many other post-translational modifications are also important in signal-transduction events,” he advised. “Further, post-translational modifications are much more abundant and widespread than was expected from early studies.

“To complicate matters, a protein often has only a low level of occupancy at any given site (only a few percent or less) of any of these modifications. Current challenges include how to quantify such targets and even identify the modifications themselves.”

Dr. Bradshaw emphasized that it’s only a matter of time until today’s challenges are solved. “If we look back, we think, ‘We have come such a huge distance!’ But if we look forward, we fear that we still have so far to go. The next-generation technologies that will better sort out all this proteomic complexity may take 10 years to develop, but they are already percolating in the minds of people today.”

He predicted that proteomic scientists eventually will be able to garner measurements down to the level of single cells and even individual molecules: “Microfluidics and nanotechnologies are rapidly emerging. The next game-changer that will provide a paradigm shift in proteomics will be the ability to routinely scale down to nanoproteomics. This will help solve protein complexity and allow us to take our first views dynamically into the cell.”

Previous articleSmart Contacts Can Monitor the Glucose in Tears
Next articleNIH Budget Proposal