Mass spectrometry, the primary tool of proteome science, is hammering away at biomedical research, biomarker discovery, and drug development. These clinically oriented applications posed daunting barriers to mass spectrometry back when it was limited to compiling protein inventories—woefully incomplete inventories at that. Today, mass spectrometry isn’t just recognizing formerly overlooked proteins, such as low-abundance proteins or microproteins, it is also distinguishing between proteoforms, that is, proteins that arise from the same gene but differ, often subtly, as a result of genetic variation, alternatively spliced RNA transcripts, or post-translational modifications.
What’s really causing old barriers to crumble, however, isn’t mass spectrometry’s ability to collect information about additional proteins and potentially relevant proteoforms. It’s mass spectrometry’s analytical power. Mass spectrometry is integrating “big data” technology to derive meaning from proteome inventories and how they shift in response to changing circumstances. Proteome inventories may fluctuate as disease processes unfold, as the immune system under- or overreacts, or as therapeutics intervene.
By capturing proteome fluctuations that reflect intercellular and intracellular communications, mass spectrometry may tap into the miscommunications that initiate and sustain cancer. Such miscommunications may involve signaling proteins that acquire infelicitous phosphate groups and, consequently, trigger aberrant signaling cascades. Or they may involve proteins that mishandle delicate protein-protein interactions, short-circuiting or overloading signaling pathways.
Essentially, mass spectrometry is beginning to work on a higher plane. Not only is mass spectrometry familiarizing us with protein structures that flicker into and out of existence, it is also revealing that the proteome, too, has dynamic structural qualities. By opening our eyes to the proteome’s shifting pathway- and network-level contours, mass spectrometry helps us recognize and even anticipate disease. Such insights may help us preserve or restore proteomic contours consistent with health.
A proteogenomic take on colon cancer
To obtain a comprehensive view of colon cancer, scientists based at the Baylor College of Medicine, the Pacific Northwest National Laboratory (PNNL), and Vanderbilt University stepped back far enough to take in proteomic and genomic details about tumorous and nontumorous colon tissues. The scientists’ proteogenomic approach, which was described April 2019 in Cell, is actually a phoshoproteogenomic approach.
“Signaling proteins and pathways are often attractive therapeutic targets for cancer treatment, yet global phosphoproteomic analyses on human colorectal cancer are lacking,” the scientists noted. To help fill the void, the scientists performed whole-exome sequencing, copy-number array, RNA sequencing, microRNA sequencing, and label-free shotgun proteomic analyses on tissue samples. Then the scientists systematically identified colon cancer–associated proteins and phosphosites.
“In addition to reinforcing or complementing genomics data, proteogenomic integration also may correct inaccurate genomics data-based inferences and lead to unexpected discoveries and therapeutic opportunities,” the Cell article indicated. “One example is the proteomic identification of SOX9 as an oncogene, whereas it was predicted to be a tumor suppressor based on somatic mutation data. Another example is the phosphoproteomics data–enabled discovery of Rb phosphorylation as an oncogenic driver of colon cancer, suggesting a unique opportunity to target Rb phosphorylation in colon cancer through CDK2 inhibition.”
The article emphasized that genomic and proteomic data complement each other in ways that can provide researchers with a better understanding of what goes on inside colon cancer cells. “Analysis of the genes tells us what might possibly go wrong,” Tao Liu, PhD, a senior scientist at PNNL and a co-corresponding author of the Cell article, indicated in a press statement. “But we don’t know exactly what actually has gone wrong until we analyze the proteins.”
This point is illustrated by another finding from the study. “Somatic mutation analyses identified new significantly mutated genes among microsatellite instability-high (MSI-H) tumors,” the study indicated, “and the proteomics data revealed unexpected functional complexity that could not have been predicted from mutation data alone.”
By identifying an association between decreased CD8 T-cell infiltration and increased glycolysis in MSI-H tumors, the study’s proteomic analysis suggests glycolysis as a potential target to overcome the resistance of MSI-H tumors to immune checkpoint blockade.
All the data is accessible in LinkedOmics, a web-based tool. It could be useful to researchers investigating colon cancer, suggested Bing Zhang, PhD, professor of molecular and human genetics at Baylor and another co-corresponding author of the Cell article. Computational analysis of unbiased measurements of protein abundance and activity “can reveal proteins, signaling pathways, and protein networks that are dysregulated in tumor samples,” he tells GEN. “Such information can help scientists identify putative biomarkers and drug targets for further investigation.”
Methyl tags and drug targets
Although the human genome contains genes that encode about 20,000 proteins, many additional proteoforms exist, partly as a result of post-translational modifications, many of which involve the addition of a small chemical tag. Of the hundred or so types of tags that exist, one in particular—the methyl group—interested scientists based at the Van Andel Research Institute (VARI). The VARI team, led by associate professor Scott B. Rothbart, PhD, collaborated EpiCypher to track the activities of lysine (K) methyltransferases.
These enzymes alter the function of proteins by marking them with methyl groups. “Several inhibitors of these enzymes are clinical development for cancer therapy,” Rothbart noted in a statement. “Defining the spectrum of their activity is critical for understanding exactly how these drugs work and for selecting reliable biomarkers to track their activity in patients.”
VARI and EpiCypher developed a functional proteomics platform to enable the rapid mapping of lysine methyltransferase (KMT) substrate selectivity without a priori knowledge of a substrate or target proteome. The platform, which is built around a lysine-oriented peptide library (K-OPL), can generate a KMT substrate selectivity profile (plus or minus three amino acids from a fixed central lysine) for any KMT.
Details about the platform appeared November 2018 in Science. “By comparing KMT selectivity profiles to available lysine methylome datasets, we reveal a disconnect between preferred KMT substrates and the ability to detect these motifs using standard mass spectrometry pipelines,” wrote the article’s authors. “Our studies validate the use of this platform for guiding the study of lysine methylation signaling and suggest that substantial gaps exist in proteome-wide curation of lysine methylomes.”
“The beauty of this technology is its simplicity and throughput, which is staggering compared to current mass spectrometry-based approaches,” added Martis Cowles, PhD, EpiCypher’s chief business officer and study co-author. “We are excited to use this technology to help drug developers identify new therapeutic targets and even identify optimal target substrates for high-throughput inhibitor screening.”
Toward single-molecule resolution
Although current mass spectrometry platforms already detect proteins of widely varying abundance and highly variegated form, proteomic scientists still want more. They want to determine protein sequences the way genomic scientists determine gene sequences. At present, doing so requires investigators to analyze samples that contain millions of molecules. Otherwise, molecules of interest fall below detection limits.
Consequently, proteomics researchers are looking forward to using platforms that could provide the ultimate in sensitivity, platforms that would identify single molecules. To devise such platforms, developers are working with technologies that exploit fluorescence, tunneling currents, or nanopores.
For example, scientists based at the University of Texas at Austin have developed a method called single-molecule fluorosequencing. These scientists, led by professor of molecular sciences Edward M. Marcotte, PhD, described their work in a paper that appeared October 2018 in Nature Biotechnology. According to this article, sparse amino acid–sequence information can be obtained for individual protein molecules for thousands to millions of molecules in parallel.
“We demonstrate selective fluorescence labeling of cysteine and lysine residues in peptide samples, immobilization of labeled peptides on a glass surface, and imaging by total internal reflection microscopy to monitor decreases in each molecule’s fluorescence after consecutive rounds of Edman degradation,” the article’s authors wrote. “The obtained sparse fluorescent sequence of each molecule was then assigned to its parent protein in a reference database.”
Marcotte and colleagues anticipate that their approach could detect elusive biomarkers of disease and improve biological research in general. In cancer studies, for example, researchers could reveal how tumors evolve from a small collection of identical cells to a heterogeneous mass of genetically divergent cells. Conceivably, different cell types could possess distinct weaknesses, which might help identify drug targets.
An alternative single-molecule approach is being developed by associate professor of chemical biology Giovanni Maglia, PhD, and his colleagues at the University of Groningen. They hope to use nanopore technology to construct an inexpensive mass spectrometer that could directly determine peptide masses. Biological nanopores, Maglia and colleagues have demonstrated, can be used to measure metabolites and to identify proteins and peptides. The pores, which consist of large proteins, may be incorporated into a membrane. When a molecule passes through the membrane, an electric signal is generated.
In early work, constructing sufficiently small pores was a problem. With even the smallest pores available, Maglia complained, the passage of peptides occurred too quickly for a reading to be detected.
This problem looks less daunting now that Maglia and colleagues have completed a proof-of-concept study that describes the construction of exceptionally small biological pores. The usual construction approach is to tinker with a pore’s constituent monomers, adjusting their interactions. Maglia’s team, however, manipulated the interactions between the monomers and the lipids of the membrane surrounding the pore.
In a paper that appeared February 2019 in Nature Communications, Maglia and colleagues explained how they constructed tiny funnel-shaped pores, “the smallest biological pores ever produced,” Maglia asserted in a statement.
“We have engineered the assembly of Fragaceatoxin C (FraC) to obtain three nanopores types with 1.6-, 1.1-, and 0.84-nm inner diameters,” the article’s authors wrote. “The nanopores can accommodate peptides ranging from 22 to 4 amino acids in length.
“Ionic blockades through engineered nanopores distinguish a variety of peptides, including two peptides differing only by the substitution of alanine with glutamate. We also find that at pH 3.8 the depth of the peptide current blockades scales with the mass of the peptides irrespectively of the chemical composition of the analyte.”
At present, the nanopores lack the sensitivity that would be needed to build a nanopore-based platform capable of rivaling the resolution of conventional mass spectrometers. Better sensitivity, however, might be obtained through various refinements, such as the incorporation of artificial amino acids, the use of different ions in the solution phase, or the maintenance of temperatures low enough to reduce background noise.
A potential advantage of the approach is the use of many different pores, which enables simultaneous measurements of differently sized peptides and even peptide modifications. “A versatile and cheap mass spectrometer for peptide analysis is feasible,” Maglia concluded. “And that would mean that more laboratories would be able to afford to conduct very important proteomics studies.”
Progress in Biological Mass Spectrometry
By Biswapriya B. Misra, PhD
During the recent Pittcon 2019 event, several presentations described how mass spectrometry can be used to illuminate molecular events of interest to clinicians and drug developers. In a presentation on functional proteomics, Forest M. White, PhD, a professor of biological engineering at MIT, detailed an approach his lab uses to profile protein expression and analyze phosphotyrosine signaling. The approach, he emphasized, could uncover mechanisms behind cancers such as glioblastoma multiforme.
Most clinical trials in glioblastoma are conducted with too little knowledge of underlying molecular mechanisms. To address this shortcoming, the White lab contributes to an effort to spatially resolve drug distribution and metabolism in biopsy tissues from clinical trials. Other participating labs are affiliated with Brigham and Women’s Hospital and Dana Farber Cancer Institute—one of these labs is led by Nathalie Y.R. Agar, PhD; the other, by Sandro Santagata, MD, PhD. White’s lab is focused on functional proteomics; Agar’s lab, MALDI-mass spectrometry imaging; and Santagata’s lab, multiplexed immunofluorescence.
“We aim to [use] this combined platform to quantify drug distribution, drug efficacy, and tumor cell response, all with micron-scale spatial resolution,” said White.
Ronghu Wu, PhD, an associate professor at the Georgia Institute of Technology, carries out global and site-specific analyses of surface glycoproteins to uncover glycoprotein functions and to help identify surface glycoproteins as drug targets and biomarkers. These analyses involve metabolic labeling, click chemistry, and enzymatic reactions with mass spectrometry-based proteomics.
Wu described an investigation of surface glycoprotein dynamics. He emphasized that compared with receptors and transporters, surface glycoproteins as enzymes had longer half-lives, indicating that enzymes on the cell surface were protected by glycans.
“This method will have extensive applications in the biological and biomedical research fields,” said Wu. “For instance, it allows us to study cell surface glycoprotein dynamics in stem cell differentiation and cell immune responses during infections.”
Biswapriya B. Misra, PhD (email@example.com), is an assistant professor of molecular medicine at the Wake Forest School of Medicine.