Top-down proteomics: where we are, where we are going?

Increasing Momentum

After some years spent increasing metrics of proteome coverage, the field of proteomics has increasingly focused on the quality of information generated during interrogation of living systems. Another aspect trending presently is to integrate proteomics with data from other “-omics” in order to gain deeper insights into cellular and disease biology. From this perspective, it is apparent that the analysis of intact proteoforms, or top-down proteomics,[1] presents additional advantages. In providing precise compositional information, TDP can add molecular details lost when proteoforms are dissected into proteolytic peptides used in bottom-up proteomics. Although BUP can identify and localize post-translational modifications (PTMs) on proteins, the well-known ‘protein inference problem’ greatly complicates the elucidation of their global patterns or cross-talk, aspects which can be captured using TDP. This is exemplified by the so-called “histone code”, where PTMs comprise combinatorial and highly dynamic patterns resulting from concerted interactions between prior PTMs and histone-modifying enzymes. These patterns govern the reading of histone marks and myriad biomolecular activities, and can be comprehensively described at the proteoform level to help assign the functions of PTM patterns.

It is true that top-down proteomics is technologically challenging, yet perceptions about this are often historical and not updated quickly in the minds of experts or those far afield. Rapid advances in instrumentation by most manufacturers over the decade have rendered targeted and high-throughput top-down feasible—even on the benchtop mass spectrometers of today. Moreover, a collection of top-down proteomics practitioners have founded the Consortium for Top Down Proteomics (CTDP, http://topdownproteomics.org/), many of which have contributed articles to this special issue. CTDP has the aims of accelerating impactful and collaborative research,[2] increasing the visibility of top-down proteomics within the community, and promulgating knowledge and best practices for newcomers to the field. Notably, the first CTDP manuscript rallied interest and focus around the “proteoform” and now has over 220 citations.[3] As further recognition of the accelerating progress in top-down proteomics, the UniProt Knowledgebase has begun cross-referencing their accession numbers with the permanent PFR identifiers in the Proteoform Atlas hosted by the CTDP (http://repository.topdownproteomics.org/).

Denaturing and Native Top-Down MS

Typically, top-down proteomics has been performed under denaturing conditions, following workflows originally developed for bottom-up proteomics. Namely, proteins are isolated from cells in detergent-containing buffers, and subjected to pre-fractionation to reduce sample complexity. In recent years, combined with concurrent advances in dedicated high-throughput data analysis platforms, top-down proteomics of proteins <30 kDa has been sufficiently optimized to enable a degree of qualitative and quantitative proteome coverage more typically achieved by bottom-up. Today, high-confidence identification and characterization of several thousand proteoforms is feasible.[4–6] However, the number of proteoforms identified by denaturing top-down proteomics drops off at molecular weights (MW) exceeding ~50–70 kDa due to technical issues associated with both the limitations of mass spectrometry and the need for improved separations. Solutions to extend the MW range of proteoforms identified during denaturing top-down experiments based on Fourier-transform mass spectrometry include the use of new generation instruments such as the 21T FT-ICR mass spectrometers,[7,8] or novel data acquisition strategies aimed at increasing the spectral signal-to-noise of large proteoforms by recording short time-domain transients.

Another option for high mass proteins derives from native top-down mass spectrometry, which preserves non-covalent interactions, labile PTMs, cofactors and physiological stoichiometry.[9] Due to these unique features, native top-down mass spectrometry has traditionally been employed for the targeted analysis of highly purified protein complexes, remarkably allowing the analysis of macromolecular assemblies with masses up to the megaDalton range, like in the case of the intact ribosome[10,11] or viral capside assemblies.[12] Furthermore, native top-down technology can provide unique insights into the perturbations in proteoform-proteoform interactions which regulate specific cellular processes.[13] Alternatively to its complementary role in structural biology,[14,15] the high-MW capabilities of native top-down might someday be employed on-line with the appropriate protein separation techniques such as ion exchange chromatography, potentially increasing the analysis throughput and steering native top-down toward “discovery mode” proteomics.[16] Along with the instrumentation development, this possible direction for native mass spectrometry will also encourage the creation of new bioinformatics, particularly when the analysis will focus on multiproteoform complexes, requiring database search algorithms that consider both the intact assembly and the single subunits.[17] With many analytical benefits and increased accessibility to biologically relevant protein structures and interactions, native top-down mass spectrometry has the potential to not only access the higher-MW regions of the human proteome but also to provide enhanced information and molecular detail.

When will a tipping point be reached?

As it stands today, bottom-up and top-down proteomics stand as highly complementary in terms of value proposition to both researchers and stakeholders. While bottom-up remains unmatched in depth of proteome coverage and quantitation of peptides, top-down proteomics characterizes proteoforms and their variants directly. Native top-down can hone in on endogenous protein complexes.[18] As techniques for denaturing and native top-down continue to evolve, one could extrapolate toward a tipping point where the analysis of intact proteins becomes far more widespread. The “Why?” for this change is becoming more clear — great progress in proteoform-resolved measurements has the value of high molecular specificity when it comes to protein molecules, often with strong mechanistic connections to pathology. As an example, a recent report of phosphorylated alpha-synuclein proteoforms in Parkinson’s Disease[19] demonstrates the new kind of information being provided by proteoform-resolved measurements. The “When?” is the crux of the issue: when will the value of these measurements lead to wider adoption and deployment of the technology in academic and clinical research? There is a common assertion that top-down is several years behind bottom-up in terms of acceptance and implementation in the field. If true, then perhaps increasing numbers of practitioners will adopt the direct interrogation of intact proteins and their protein complexes. Advancements in top-down proteomics like those enclosed in this special issue will be the driving agents of that change, and the implications for the study of protein molecules and the translation of this knowledge to understanding disease hold a good deal of unrealized potential.