The first pilot project of the consortium for top-down proteomics: a status report

Abstract

Pilot Project #1--the identification and characterization of human histone H4 proteoforms by top-down MS--is the first project launched by the Consortium for Top-Down Proteomics (CTDP) to refine and validate top-down MS. Within the initial results from seven participating laboratories, all reported the probability-based identification of human histone H4 (UniProt accession P62805) with expectation values ranging from 10(-13) to 10(-105). Regarding characterization, a total of 74 proteoforms were reported, with 21 done so unambiguously; one new PTM, K79ac, was identified. Inter-laboratory comparison reveals aspects of the results that are consistent, such as the localization of individual PTMs and binary combinations, while other aspects are more variable, such as the accurate characterization of low-abundance proteoforms harboring >2 PTMs. An open-access tool and discussion of proteoform scoring are included, along with a description of general challenges that lie ahead including improved proteoform separations prior to mass spectrometric analysis, better instrumentation performance, and software development.

Keywords: Human histone H4; PTM analysis; Technology; Top-down proteomics.

Figure 1

Heat map of the individual PTMs marginalized from the collective top-down dataset organized by investigator. The scale of the color scheme is linear to the number of observations of proteoforms that contain each PTM, offset by 1. (0 = no observations, 1 = one observation, maximum = 29 observations) Only unambiguously localized PTMs are considered.

Figure 2

Heat map of the binary combinations of PTMs marginalized from the collective top-down dataset organized by investigator. This marginalization is achieved by summing the number of observations of proteoforms that contain both PTMs. Note that the presence of a binary combination of PTMs in the figure does not mean that there is a proteoform that contains only these two PTMs to the exclusion of other PTMs. The scale of the color scheme is linear to the number of observations of proteoforms that contain each binary combination of PTMs, offset by 1. (0 = no observations, 1 = one observation, maximum = 13 observations) Only unambiguously localized PTMs are considered.

Figure 3

(A) Scheme of top-down MS experiment. (i) The broadband mass spectrum of the intact protein. (ii) Segment of the 13+ charge state of the intact protein with methyl equivalents (m.e.) labeled. The proteoform with six m.e. is selected as the precursor for the MS² experiment. (iii) Tandem mass spectrum for precursor chosen in (ii) to localize PTMs and identify proteoforms. (B) Ambiguity of Top-Down MS. (i) With only a single proteoform existing and enough (100%) sequence coverage, there is no ambiguity for the proteoform H4K5acK8ac. (ii) With a single proteoform existing but not enough sequence coverage, there is ambiguity on the localization of the second acetylation (K8 or K12), resulting in two ambiguous proteoforms, H4K5acK8ac or H4K5acK12ac. (iii) With mixed spectra for multiple proteoforms and enough (100%) sequence coverage, there remains uncertainty on the existence of an acetylation on K12, therefore two ambiguous proteoforms (H4K5acK12ac and H4K8acK12ac) arise.

Figure 4

Depiction of three general types of scores required to capture and convey the full complexity of data generated by MS/MS of whole, multiply-modified proteins. The dashed arrows represent that these tiers of scoring are only semi-independent and each level of characterization does indeed affect and inform the other. In some cases, failure to consider all levels of information properly can result in incorrect confidence scores and misidentification.

Figure 5

Screen grab from a web-based tool for generating a Proteoform Characterization Score (PCS).