The Frontier of Post-Translational Modification Analysis

The human proteomic repertoire is large but finite. Yet, via post-translational modifications (PTM), biology has found a way to produce a nearly infinite number of distinguishable—and often functionally different—proteins.

The field of post-translational modification analysis is also known as ‘epiproteomics’, using the prefix ‘epi’ borrowed from ‘epigenomics’. But compared to the landscape of epigenetic modifications, the landscape of PTMs is vast. Post-translational modifications include everything from adding a modifying group (like acetyl, phosphoryl, glycosyl, or methyl), to proteolytic cleavage, to structural changes (like folding or creation of disulfide bridges), to amino acid modification. There are hundreds of modifications and moieties that can be added to proteins.

From the perspective of the cell, this is very useful for reacting and adapting to changing conditions. If a cell encounters some stressor or needs to mount a response, synthesizing a whole batch of new proteins could take hours. “What can happen much, much faster is for enzymes to come in and post translationally modify a protein and give it a different structure and therefore a different function. Then the cell can get an immediate response,” says Chris Adams, Ph.D., Global Business Development Director for Bioinformatics at Bruker.

But for those trying to decode PTMs and their functions this is a big challenge. “If you think of the combinatorial possibility from any combination of 20 amino acids and just one PTM it is huge. Then the fact that there's hundreds of known PTMs, and how many other unknown PTMs. Trying to understand this whole universe gets incredibly complicated,” says Dr. Adams.

Rapidly changing field

The field of analyzing post-translational modifications is not new. The research community has been aware of and interested in PTMs for decades. But as Dr. Adams points out, “the ability to detect them and quantify them has changed rapidly in the past 10 years. To the point where we are continuously discovering new post translational modifications that we never knew were there.”

Much of the analysis done for PTMs is via mass spectrometry (MS). While affinity-based methods can be used, searching for PTMs inherently makes using affinity challenging. Dr. Adams explains it like this, “If there's a post translational modification on that protein, the whole structure looks very different. And those antibodies or those other ways to capture the protein molecule won't work. And that's why mass spectrometry has and will continue to be the major way to detect post translationally modified peptides and proteins.”

This being said, trying to look for modified peptides in the MS of whole cell lysates is challenging as often they will get lost in the noise of unmodified proteins and peptides. So many researchers who are looking at PTMs are using enrichment methods prior to putting their sample on the mass spectrometer.

Improvements in enrichment methods and sensitivity of mass spectrometry have really moved the needle on what is possible in PTM detection. “Only three or four years ago you would be lucky to run twelve enrichments in a day, and you needed to have tons of starting material. But this is all changed now. The instrumentation, particularly the timsTOF is so much more sensitive, we can start from much smaller scales and we can literally analyze hundreds of samples in a day and get similar output to what I was getting with the old enrichment techniques,” notes Dr. Adams

Challenges in epiproteomics

PTMs are already quite challenging to interrogate, but there is an added layer of complexity when using MS caused by positional isomers. A positional isomer is a peptide segment with a PTM moiety that could be attached to several different amino acids. Take the example of looking for phosphorylation. The phospho group could be on a serine, threonine, or tyrosine. The chances that a single peptide sequence contains multiple serines, threonines, or tyrosines is very high. So you could have two peptides that are phosphorylated in two different places but their molecular weight will be the same. Therefore, on a mass spec they will have the same profile.

“So, what we need in that case, is to know the gas phase structure. And we call that the collisional cross section. On our timsTOF family of mass spectrometers, we actually measure the collisional cross section for each and every analyte,” Dr. Adams says. In order to get even more information Bruker has also used large data sets of hundreds of thousands of phospho-peptides and used machine learning to be able to predict the collisional cross section and compare that with the experimental value coming off the timsTOF to make predictions about which positional isomer it might be.

Clinical applications

Investigating the epiproteome for clinical purposes looks very promising across a variety of different pathologies. Early studies have been investigating the blood plasma proteome and epiproteome. And the National Cancer Institute’s Clinical Proteomic Tumor Analysis Consortium (CPTAC) is already coupling global proteomics of tumor and healthy tissue with phosphorylation profiles as well.

Cancer researchers are particularly interested in histone modifications as PTMs on histones are well known to be altered in cancer cells both for individual genes and globally. Loss of particular histone acetylation and methylation marks have been proven to be prognostic of patient outcomes in certain types of cancer. These changes seem to take place very early in the progression of cancer and monitoring for changes of these histone modifications may be a promising method for early detection and diagnosis.

In keeping with the rest of the field, up until fairly recently the most common histone PTM analysis methods relied on antibody-based approaches. And while several histone enrichment methods have been developed in the past few years that allow researchers to interrogate histone PTMs with MS instead of antibodies, unfortunately, they require a significant amount of input material. This is a big challenge for clinical applications where only small samples may be available. Recent work has been released optimizing and improving these methods for low-abundance clinical samples, which will open the door for these methods to be used for real world clinical purposes.