This story is part of an ongoing series on emerging proteomic technologies. See here for the complete list of stories in the series.
NEW YORK (GenomeWeb) – Mass spec-based analysis has in recent years seen increasing uptake as a tool for research into the structures of proteins and protein complexes.
But while the approach is proving a worthy companion to techniques like cryo-electron microscopy, its true potential for structural analysis is only beginning to be realized, suggested some leading proteomics researchers.
Asked what proteomic technologies he sees as likely to emerge as particularly significant in coming years, University of Victoria researcher Christoph Borchers cited the use of mass spectrometry combined with methods like peptide crosslinking and hydrogen/deuterium exchange (HDX) to complement more traditional structural research methods like cryo-EM.
"The development of structural proteomics techniques has made significant progress [in terms of] reagents, methods, and software in the last years and are now ready for prime-time, meaning [they] are useful to answer biological questions and solve biomedical problems," he said.
Borchers noted as an example a study in Nature last year led by Max Planck researcher Patrick Cramer (on which Borchers was a co-author) that used a combination of cryo-EM and mass spec with crosslinking to elucidate the structure of the RNA polymerase II-Mediator core initiation complex in yeast.
Other proteomics researchers applying mass spec and crosslinking to structural analysis include the Swiss Federal Institute of Technology Zurich's Ruedi Aebersold, who in 2014 published a paper in Nature combining the approach with cryo-EM to determine the structure of the 39S large subunit of the mammalian mitochondrial ribosome.
Traditionally, Aebersold noted, crystallography has been the go-to approach for structural research and, indeed, a good crystal structure will have better resolution than one determined using EM-MS methods. However, a major limitation of crystallography is the large amount of sample required for the technique, which makes EM-MS an attractive approach in cases where it is difficult to generate large quantities of the target molecules.
Cross-linking chemistry uses molecules to crosslink proteins at different points, and mass spec can then be used to generate structural information based on the position of these linkages. HDX is used similarly. In that approach, the amides in protein backbones are labeled with deuterium. Amides from the regions involved in binding will, in theory, be less accessible to labeling than those from other regions, and this can be detected using mass spec to provide structural information.
Combining this sort information with the structural information provided by cryo-EM, researchers are able to generate a picture of both the overall shape of a protein or protein complex as well as an understanding of the arrangement and orientation of individual proteins within the structure.
But, even as mass spec has seen uptake in structural research as a complement to cyro-EM, Aebersold said this is not where he believes the method will ultimately prove most useful.
"This is the current state [of the technology], and it is interesting and very successful now, but I don't think this is the endpoint or the direction that is most interesting," he told GenomeWeb.
In fact, he said, improvements in cryo-EM could make mass spec data less necessary for such experiments. "Cryo-EM itself is moving quite quickly with new detectors that provide higher resolution and more information, so it's not clear whether in two or three years the crosslinking data will still be useful."
Rather, Aebersold said, he believes that mass spec and crosslinking will likely have the most impact as a tool for studying the structures of protein complexes as they exist in actual cells.
"Most of the proteins that are being crystallized eventually have to be purified in milligram amounts, which requires a very big effort to generate," he said, noting that "frequently, these complexes are reconstituted in vitro from recombinant proteins."
And while this can provide structural information about these complexes, it tells researchers little about how they exist in actual cells — for instance, how these structures change in response to various stimulants or how they differ in diseased cells versus healthy.
"So what we are particularly interested in doing is going into the cell and doing affinity purification [of a protein], and then asking what is the composition and structure of the complex around a particular protein under different cell conditions," Aebersold said. "Because these complexes are highly dynamic in their composition and are affected by mutations and the binding of factors and so on. So we see it as a means to characterize complexes in more detail than is currently possible."
George Mason University research Emanuel Petricoin similarly noted the potential of such mass spec-based structural work.
"That is somewhere our laboratory is starting to focus, and I sense that the field is starting to move away from just categorization of proteins as going up or down [in expression] to really putting the proteome together as an active information archive that is communicating," he said. "We always say that genes are the information archive but that proteins do the work. That is kind of a soapbox statement in the proteomics world. But proteins doing the work of the cell involves an active physical communication between them."
"So that to my mind is an area that is poised to explode commercially and clinically," Petricoin added.
In particular, he said, it could prove valuable for drug development as it is currently thought that targeting protein-protein interactions could prove an effective therapeutic approach, especially in the case of proteins traditionally considered "undruggable" targets.
Cancer drugs like kinase inhibitors that target enzymatic activity have proven susceptible to resistance due to the ability of tumors to develop mutations that allow, for instance, the targeted ATP-binding domains to get around the competitive inhibition used by such drugs, Petricoin said.
"But if we can identify inhibitors that can disrupt protein-protein interactions, those will be much more difficult [for the tumor] to develop resistance to, because evolutionarily it is harder to wire around," he said.
Workflows for applying mass spec to such structural analyses are still in their infancy, Aebersold noted. But, he said, workflows that provide "credible and reproducible" results do exist.
He cited as one of the first efforts to apply such a workflow a study published in Science in 2012 in which he and co-authors including Gene Center Munich researcher Franz Herzog used cross linking and mass spec to analyze the structure of protein phosphatase 2A complexes affinity purified from human cells.
More recently papers from the labs of researchers including Brian Chait at The Rockefeller University and Carol Robinson at University of Oxford have further explored and optimized such approaches, Aebersold said.
Last month his lab published a paper in Nature Methods presenting a software package for characterizing conformational changes in protein complexes via cross-linking mass spec data, which, he said, would enable researchers to look at quantitative differences before samples in an automated way.
"Before this was very difficult to do because it had to be done by basically looking at each spectra manually," he said.
Petricoin and his GMU colleague Lance Liotta have developed protein dye reagents that can be used as alternatives to crosslinking or HDX to gather structural information by mass spec.
The dyes bind to protein surfaces with high affinity and low off-rates and block trypsin cleavage sites. Researchers apply complexes of interacting proteins with these dyes, which coat all portions of the proteins except those that are blocked because they are in contact with each other. They then denature the proteins and subject them to trypsin digestion. Because all portions of the proteins save for their interaction points have been coated in the dye, only those portions containing the interaction points are still accessible to trypsin, and so, only fragments from those regions of the protein will be generated in digestion.
These fragments can then be analyzed by mass spec to identify the portions of the target proteins involved in their interaction.
In a 2014 Nature Communications paper, the GMU researchers found that the approach generated better data than conventional crosslinking or HDX when used for an analysis of the three-way interaction of the Interleukin-1β ligand, the receptor IL1RI, and the protein IL1RAcP.
In general, Aebersold said, all steps of such workflows, from enrichment of crosslinked peptides, to mass spec methods, to the backend informatics have room for improvement. "At every step there are incremental advances being realized that kind of add up over the multistep process," he said.
"In the long run we want to find optimal methods where we can ask, for instance, with a particular protein complex that is involved in a disease and for which there is a lot of genomic information: how do specific mutations affect the subunits of this protein complex and what are the implications for the topology and structure and composition when one or several of these subunits are mutated?" Aebersold said. "This is where we are going, but it is still somewhat tedious."