NEW YORK (GenomeWeb News) – The lab of Northwestern University researcher Neil Kelleher has completed a top-down analysis of intact human histone H3 proteins.
Detailed in a paper published this month in Molecular & Cellular Proteomics, the study represents a leap forward in mass spec-based histone work, Kelleher told GenomeWeb, noting that it marks the first time researchers have managed to do in-depth characterization of intact histone H3 proteoforms.
The study was significantly enabled by the performance of Thermo Fisher Scientific's Orbitrap Fusion Lumos instrument, Kelleher said.
Kelleher has an ongoing relationship with Thermo Fisher developing methods for top-down proteomics, and the company offers a commercial version of the ProSight software for top-down proteomics that he developed.
Bottom-up, peptide-based proteomics has dominated the field since its inception more than a decade ago, but interest in top-down methods looking at undigested full-length proteins has grown as improvements in instrumentation have made analysis of intact proteins easier and researchers have become increasingly aware of the importance of protein isoforms and post-translational modifications.
In conventional bottom-up proteomics, proteins are digested into smaller peptides prior to analysis, so it can be difficult or impossible to determine, for instance, what different combinations of post-translational modifications exist on single proteins.
Top-down analyses, on the other hand, look at undigested proteins, meaning researchers can profile the intact molecule with all its modifications rather than tying that information together from collections of identified peptides.
Being able to analyze combinations of protein post-translational modifications is potentially valuable in a number of contexts, but especially when analyzing histones, where top-down analyses have the potential to unravel what is often called the "histone code" — the different histone PTM combinations thought to play a significant role in DNA transcription.
To date, however, in-depth top-down analysis of human histones has run up against the technical challenges presented by the complexity of these proteins' modifications.
"You have 200 different proteoforms, minimum, and they are separated by about 1 m/z — good luck," Kelleher said, describing the challenge facing researchers. In the past, he said, his group has used a combination of bottom-up and middle-down mass spec assays to analyze H3 modifications, but their ability to do top-down analyses has been limited.
The MCP work, suggests, though, that mass spec instrumentation has now reached a point where such in-depth profiling is now possible. Kelleher's group used Thermo Fisher's Orbitrap Fusion Lumos instrument for the project, the second generation of the company's Orbitrap Fusion, which combines a quadrupole, ion trap, and Orbitrap in one machine.
Compared to the original Fusion, the Lumos, which was released in June at the American Society for Mass Spectrometry annual meeting, has better inlet optics, an electrodynamic ion funnel, and an improved quadrupole, all of which serve to increase the intake and flow of ions through the machine, offering an increase in sensitivity of between three- and five-fold compared to the original system, according to the company.
This improvement of ion intake and flow along with the upgraded quadrupole were key to the MCP study, Kelleher said, as they allowed him and his colleagues to select particular H3 proteoforms with the required specificity while still getting enough analytes into the Orbitrap analyzer to generate good data.
"Those histone H3 proteoforms are separated by maybe 1.2 or 1.5 m/z, and so you have to pluck them out from the sea of others and not let the others leak through and still get enough through the pipe" to enable a good analysis, he said.
"With the [Lumos] quadrupole you can go down to sub-thomson isolation windows and still get high transmission. So you just target a [particular] histone H3 proteoform and collect," he said, noting that the researchers were able to get good data on proteoforms present in their samples at relative abundances as low as one percent.
"These proteoforms for intact H3 are so narrowly spaced and so complicated that you really couldn't isolate these different isobaric peaks out from one another and then get really rich fragmentation data on them once you isolated," Kelleher said. However, he noted, within two months of beginning their analysis on the Lumos, he and his team were able to "get data on probably about a couple hundred [intact] proteoforms in short order."
What to do with that data, now, is another question. As Kelleher said, teasing out the different biological implications of the presence of these hundreds of proteoforms "gets very complicated."
"We can produce this kind of data at a much faster rate on full-length, highly modified proteins — ok, show them what they've won," he said.
"You have 100 million nucleosomes in a diploid cell, and so that means you have like  copies of these histones, and they are spread all throughout the genome," he said. "What we are doing is getting a quantitative readout of a mix of all the proteoforms that exist at all of these different nucleosomes, and they are spread through all the bases of the genome. So the chasing of function is going to be interesting, but it is going to take some time."
One possibility, Kelleher said, is to combine the proteoform data with data from CHiP-Seq experiments to try to get at where in the genome particular H3 proteoforms are active.
"If we know [for instance] that H3 k27 trimethlylation occurs with H3 K36 dimethylation, then we can look on the CHiP-Seq maps for where those two marks occur [together] and we can infer that that proteoform might exist there and on related loci like it," he said.
Kelleher noted that the MCP study, which profiled histone H3 proteoforms in multiple myeloma cells and normal cells did, however, generate one somewhat straightforward result — identification of one specific proteoform that was observed only in multiple myeloma cells.
He and his colleagues now hope to apply their analysis to actual clinical samples, he said, noting that they could do it with a few as 200,000 to 300,000 patient B cells.