Identification of cardiac myofilament protein isoforms using multiple mass spectrometry based approaches

Abstract

Purpose: The identification of protein isoforms in complex biological samples is challenging. We, therefore, used an MS approach to unambiguously identify cardiac myofilament protein isoforms based on the observation of a tryptic peptide consisting of a sequence unique to a particular isoform.

Experimental design: Three different workflows were used to isolate and fractionate rat cardiac myofilament subproteomes. All fractions were analyzed on an LTQ-Orbitrap MS, proteins were identified using various search engines (MASCOT, X!Tandem, X!Tandem Kscore, and OMSSA) with results combined via PepArML Meta-Search engine, and a postsearch analysis was performed by MASPECTRAS. All MS data have been deposited in the ProteomeXchange with identifier PXD000874 (http://proteomecentral.proteomexchange.org/dataset/PXD000874).

Results: The combination of multiple workflows and search engines resulted in a larger number of nonredundant proteins identified than with individual methods. A total of 102 myofilament annotated proteins were observed overlapping in two or three of the workflows. Literature search for myofilament presence with manual validation of the MS spectra was carried out for unambiguous identification: ten cardiac myofilament and 17 cardiac myofilament-associated proteins were identified with 39 isoforms and subisoforms.

Conclusion and clinical relevance: We have identified multiple isoforms of myofilament proteins that are present in cardiac tissue using unique tryptic peptides. Changes in distribution of these protein isoforms under pathological conditions could ultimately allow for clinical diagnostics or as therapeutic targets.

Keywords: Cardiac; Myofilament proteins; Protein isoforms; Protein subisoforms.

Figure 1. Experimental workflows

Three workflows are combined to identify the myofilament protein isoforms in cardiac samples from rat.1DLC’InSeq’: myofilament proteins were enriched using the “IN Sequence” method, fractionated using HPLC and in-solution digested. 1DE’InSeq’: enrichment using the “IN Sequence” method combined with SDS-PAGE fractionation and in-gel digestion. 1DE’triton’: enriched using the Triton X-100, fractionated using SDS-PAGE and in-gel digested. All samples were run on an LTQ-Orbitrap, proteins identified using Mascot, X!Tandem, X!Tandem Kscore, and OMSSA with results merged using PepArML Meta-Search engine, and post-search analysis performed using MASPECTRAS.

Figure 2. Identified cardiac myofilament proteins using the meta-search engine PepArML

Clustered proteins identified using PepArML Meta-Search engine are portrayed in a Venn diagram shown in Figure 2A. The 1DLC”InSeq” workflow resulted in 1198 proteins, 1DE”InSeq” in 656 proteins and 1DE”triton 459 proteins (with ≥1 peptides) (Supplemental Table 1B). 461 proteins were identified using two or three workflows (Supplemental Table 3). Figure 3B illustrates the cellular localization of these proteins using Uniprot Subcellular Location. Uniprot annotated 102 proteins as myofilament or myofilament-associated proteins (Supplemental Table 4). Some proteins have multiple location annotations and where therefore counted double. All proteins identified in this Uniprot annotated myofilament list were validated in the literature for direct association with the myofilament. This resulted in 27 “core” myofilament proteins and the identification of 39 isoforms and sub-isoforms (Table 2).

Figure 3. Schematic sarcomere overview with myofilament protein localization

The localization of the 27 “core” myofilament proteins listed in Table 2 are depicted in the sarcomere schematic. Proteins identified with one or more isoforms have been marked (*).

Figure 4. Schematic representation of the isoforms of α-actinin and tropomyosin

A. Isoforms 1, 2 and 4 of α-actinin with the known binding sites to actin, dendrin, Ca²⁺ and PIP₂ marked. B. Isoforms 1-1, 1-2 and 3 of tropomyosin with the overlapping sites and the actin binding regions marked. Unique peptide denotes the region where the unique peptide of the isoform was located.

Figure 5. Schematic representation of the isoforms of MLC and MHC

A. Isoforms 2a, 2c, 3, 4 and 6 of MLC with marked EF-hand regions. B. Isoforms 4, 6, 7, 8, 9, 13, and 14 of MHC with the ATP-binding sites, IQ-motifs and actin binding sites marked. Unique peptide denotes the region where the isoforms unique peptide was located.