NEW YORK (GenomeWeb) – Northwestern University researchers have used top-down proteomics for unbiased discovery and characterization of intact, endogenous protein complexes.
Detailed in a paper published this week in Nature Chemical Biology, the work demonstrates the ability of mass spectrometry to profile protein complexes in their native states in great detail and with relatively high throughput, bridging fields of research including structural biology, top-down proteomics, and the study of protein complexes and interactions.
Such analyses have the potential to offer highly precise structural information, allowing for more specific and subtle characterization of the proteomic landscape, suggested Neil Kelleher, professor of chemistry at Northwestern and author on the study.
Using ion exchange chromatography or GELFrEE fractionation combined with mass spec analysis on a Thermo Fisher Scientific Q Exactive HF instrument, the researchers were able to characterize 125 intact protein complexes and 217 different proteoforms in mouse heart and human cancer cell lines, observing protein-protein interactions, binding of metal cofactors, variant isoforms, and various post-translational modifications.
They also made a number of new observations, including of unexpected processed protein forms and of post-translational modifications present in different abundances than previous research would suggest.
While proteomics has traditionally considered proteins as discrete entities within a sample, proteins typically exist and function within cells as part of complexes, interacting with other proteins and molecules. Researchers have used a variety of approaches to better characterize protein interactions, ranging from yeast two-hybrid experiments and immunoprecipitation mass spec to approaches like cross-linking mass spec and cryo-electron microscopy.
The approach presented by Kelleher and his co-authors offers another, potentially quite powerful, tool for studying intracellular proteins.
The mass spec workflow used in the paper was many years in the making, Kelleher said, noting that previous work by researchers including University of Oxford Professor Carol Robinson and the University of California, Los Angeles's Joseph Loo had established methods for upfront separation of native protein complexes and fragmenting and ejecting individual subunits from the larger complex for mass spec analysis.
Less well established were methods for top-down analysis of these ejected subunits, Kelleher said, noting that he and his lab collaborated with researchers at Thermo Fisher Scientific to work out this portion of the analysis.
While proteomic experiments commonly use two rounds of fragmentation, collecting data on molecules at the MS1 and MS2 levels, some researchers have begun using three rounds of fragmentation in certain kinds of analyses, collecting data at the MS3 level, as well.
The Northwestern team adopted an MS3 approach, which Kelleher said was one of the keys to enabling their analysis. The researchers used MS1 to measure the mass spectrum of the intact protein complex they were analyzing, looking at one or more of its different charge states. They followed this with an MS2 stage in which they used the mass spec's quadrupole to isolate one specific charge state and then ejected individual components of the complex into the HCD cell to be fragmented. They then performed MS3 in which the individual components were isolated via the quadrupole and identified and characterized.
Then, Kelleher said, the researchers took the data collected on the individual subunits, "and we pieced it all back together."
"So, you look at it and you can say, well, this is glyceraldehyde phosphate dehydrogenase, but the mass is wrong," he said. "So then you follow that, and you start to figure out, oh well, there's a metal bound [to the complex], or there's a cofactor. And then you've got to get the right stoichiometry."
It sounds like quite a puzzle to unravel, but, Kelleher said, the quality of data provided by mass spec allows for extremely confident identifications.
"Time after time, here's [a complex] where 30 years of structural biology has been done on it, and yep, it binds four magnesium, and that's exactly what we see in the native MS," he said. "The quality of the information, the fidelity, the connectedness of the information, provided by native proteomics, is so strong to the underlying biology."
Kelleher said that the identifications in the paper were all manually validated but that he believed that, while it would take time and investment, the process could be automated.
"The world can automate what it sees as high value," he said. "The engineering ethos kicks in, and you can reduce it to practice."
He added that interest among biopharma companies in native protein mass spec had helped drive the field forward and would likely continue to do so.
The most difficult portion of the workflow to automate would be the three-stage mass spec method due to the different collision energies required and charge states involved in different complexes, Kelleher said. He added, though, that Harvard University researcher Steven Gygi had managed to automate MS3 analysis as part of his lab's MS3-based isobaric labeling workflow, suggesting that such an approach was plausible.
The study was not focused on any specific biological questions, Kelleher said, but there were several interesting findings. One highlight was the method's ability to detect not only the tetrameric human mitochondrial superoxide dismutase (SOD2) complex but also its substrate (superoxide) and products (oxygen and hydrogen peroxide). The ability to capture these sorts of enzyme-substrate interactions along with stoichiometry and other information could improve understanding of "enzymatic mechanisms in a large variety of protein systems," the authors wrote.
"That's very subtle information," Kelleher said. "Having substrates of the enzyme still bound to it — you get the stoichiometry, the occupancy, how much is bound to the tetramer."
Also notable, Kelleher said, was the finding that cysteine modifications appear much more common than is apparent from previous studies. On the other hand, post-translational modifications like acetylation were much rarer than the literature would suggest.
The potential implications of these findings need further exploration, Kelleher said, but they do suggest that the proteome might be somewhat less complicated than feared.
"At least we have some data, to kind of say, 'Well, OK, it's not like these proteins all have hundreds of abundant proteoforms and it's super complicated,'" he said. "It's not that bad, and so we can actually do the whole proteome project, which is my dream. We can catalog them all, because there's not an infinite number of them."