A recent histone modification study led by Broad Institute researchers demonstrates the potential of high-resolution mass spectrometry for targeted proteomics.
The study, detailed in a paper published this week in Nature Genetics, used mass spectrometry to profile chromatin modifications in more than 100 cell lines from the Cancer Cell Line Encyclopedia, identifying several chromatin modification signatures linked to genetic mutations in pediatric acute lymphoblastic leukemia.
The researchers looked at 42 different combinations of modifications on peptides from the histone 3 protein across 115 cancer cell lines. Using unsupervised clustering to analyze the resulting data, they identified six clusters featuring distinct chromatin signatures, including a cluster characterized by an increase in dimethylation of H3 lysine 36 along with a decrease of unmodified H3K36.
Of the 13 lines that comprised this cluster, six were known to contain a translocation leading to the overexpression of NSD2, which is associated with the observed chromatin modification signature. And, upon examining the other seven lines, the researchers discovered in these lines a previously unknown NSD2 variant that, likewise, was linked to the cluster's characteristic histone modifications.
This variant was also highly enriched in pediatric acute lymphoblastic leukemia, suggesting, the Nature Genetics authors wrote, that the mutated protein could prove to be a therapeutic target for the disease.
Their findings also suggest that the associated chromatin signature might be useful as biomarkers for tracking disease progression or efficacy of treatment, said Jacob Jaffe, first author on the paper and assistant director of the Broad's proteomics platform.
"Say we have this signature of aberrant chromatin behavior that you are trying to correct with a therapeutic. Resequencing the [mutated NSD2] gene isn't going to tell you anything about how you are doing there," he said. Longitudinally tracking changes in the histone modifications, on the other hand, could provide a measure of a treatment's effectiveness.
Beyond its potential insights into the genetics underpinning pediatric ALL, the study also offers an example of the potential of high-res mass spec as a tool for targeted protein quantitation. While triple quadrupoles have traditionally been the instrument of choice for such experiments, interest in high-res quantitation has grown in recent years with a number of researchers exploring the potential of the technique (PM 10/12/2012).
In the case of the Nature Genetics study, the high-res approach was necessary for distinguishing between highly similar histone modification patterns, Jaffe told ProteoMonitor.
Such analyses involve "very tricky identifications [of modifications] that are near isobars or in certain cases exact isobars – just swaps of modification positions," he said. "So we used [hi-res] mass spectrometry to be absolutely sure that the modifications we were calling and quantifying were what we said they were."
Jaffe and his colleagues synthesized the peptides of interest, as well, allowing them to generate synthetic peptide reference spectra for the various modified histone peptides they looked at – "so we are rock-solid sure about what we are calling," he said.
They built selected-reaction monitoring-style assays to each of the modified histone peptides using the Skyline targeted proteomics software developed by University of Washington researcher Michael MacCoss, and performed the assays on a Thermo Fisher Scientific Q Exactive mass spec, which has emerged as one of the main instruments used for high-res targeted quantitation.
Indeed, the recent trend toward using high-res mass spec for protein quantitation began, in large part, with Thermo Fisher's 2011 launch of the Q Exactive. Several months after the machine was introduced, Bruno Domon, director of the Luxembourg Clinical Proteomics Center, presented results indicating that the instrument could compete with triple quads for quantitation (PM 9/16/2011).
The following year, University of Wisconsin-Madison research Joshua Coon presented work that similarly demonstrated the Q Exactive could compete with triple quads as a tool for targeted quantitation (PM 8/10/2012).
In a triple quad-based SRM assay, the first quadrupole isolates a target precursor ion, which is then fragmented in the second quadrupole, after which a set of pre-selected product ions are detected in the third quadrupole. In high-res instruments like Q-TOFs or the Q Exactive, on the other hand, researchers can isolate a target precursor ion using the upfront quadrupole and then monitor not just a few, but all of the resulting product ions using the time-of-flight or Orbitrap analyzer.
The high resolution of these analyzers – up to 140,000 in the case of the Q Exactive – allows researchers to quantify peptides that might otherwise be lost amidst background peaks. The larger number of product ions monitored via high-res quantitation can also improve the specificity of an analysis, since more transitions are available to confirm a peptide ID.
While hi-res machines are typically not as sensitive for quantitation as high-end triple quads, this was a necessary trade-off for the histone work, Jaffe said.
"It's probably not quite as good [in terms of sensitivity], but for this work we absolutely needed the high resolution because of the near isobaric modifications," he said. He noted that the power of the high-res approach allowed the researchers to build an extremely comprehensive histone modification dataset across the 115 cancer lines.
"What we generated using this approach is a true matrix of data where very few data points are missing," he said. "Across the 100-plus lines versus the 42 [modifications] we reported on in this paper, there are very few blank spots."
This, Jaffe said, allowed the researchers to perform a genomic-style analysis of the modification data "where we are clustering [the cell lines] and identifying [modification patterns] that are enriched."
"And, ultimately, it is that that lets us make these correlations back to the genomic data," he said. "The ability to take proteomic data and plug it into genomic clustering tools is, I think, a pretty big breakthrough for us."
In the future, the researchers aim to expand their histone modification analyses to the more than 1,000 cancer cell lines comprising the Cancer Cell Line Encyclopedia, Jaffe said. They also hope to apply the methodology to the study of other diseases as well as normal biological processes such as stem cell differentiation, he added.
"These histone marks are not just about disease. They are also about fine control of how the genome gets expressed in many biological processes," he said. "So I think we have a lot to learn in many different areas."