Real-Time Search-Assisted Multiplexed Quantitative Proteomics Reveals System-Wide Translational Regulation of Non-Canonical Short Open Reading Frames

Abstract

Abnormal expression of histone deacetylases (HDACs) is reported to be associated with angiogenesis, metastasis and chemotherapy resistance regarding cancer in a wide range of previous studies. Suberoylanilide hydroxamic acid (SAHA) is well known to function as a pan-inhibitor for HDACs and recognized as one of the therapeutic drug candidates to epigenetically coordinate cancer cell fate regulation on a genomic scale. Here, we established a Real-Time Search (RTS)-assisted mass spectrometric platform for system-wide quantification of translated products encoded by non-canonical short open reading frames (ORFs) as well as already annotated protein coding sequences (CDSs) on the human transciptome and applied this methodology to quantitative proteomic analyses of suberoylanilide hydroxamic acid (SAHA)-treated human HeLa cells to evaluate proteome-wide regulation in response to drug perturbation. Very intriguingly, our RTS-based in-depth proteomic analysis enabled us to identify approximately 5000 novel peptides from the ribosome profiling-based short ORFs encoded in the diversified regions on presumed 'non-coding' nucleotide sequences of mRNAs as well as lncRNAs and nonsense mediated decay (NMD) transcripts. Furthermore, TMT-based multiplex large-scale quantification of the whole proteome changes upon differential SAHA treatment unveiled dose-dependent selective translational regulation of a limited fraction of the non-canonical short ORFs in addition to key cell cycle/proliferation-related molecules such as UBE2C, CENPF and PRC1. Our study provided the first system-wide landscape of drug-perturbed translational modulation on both canonical and non-canonical proteome dynamics in human cancer cells.

Keywords: cancer; histone deacetylase; proteomics; real-time search; short open reading frames.

Figure 1

Schematic representation of RTS-dependent in-depth quantitative proteomic analysis of SAHA-treated human cancer cells. The human HeLa cell lysates treated with different concentrations of SAHA (0, 1, 3, 5 and 10 µM) for 24 h were tryptic digested and labeled with discrete TMT tags (TMT126, TMT127, TMT128, TMT129 and TMT130), respectively. The TMT-labeled peptide mixture was subjected to mass spectrometry analysis in an RTS-dependent data acquisition on Orbitrap Eclipse Tribrid mass spectrometer using UniProt human reference protein sequences combined with human short ORF data from sORFs.org [22]. Protein identification and quantification was conducted by searching against the above integrated protein sequence database.

Figure 2

Identification and quantification of UniProt-defined human protein sequences by RTS-assisted shotgun MS analysis. (A) Numerical distribution of UniProt-defined human proteins identified in our RTS-assisted proteomic analysis. (B) Volcano plots for TMT-based quantitative proteomic changes in response to differential SAHA treatment. The red dots indicate the corresponding data on each quantified protein. The y-axis represents log10-transformed p-value adjusted by Benjamini–Hochberg method, whereas the x-axis indicates the log2-transformed fold change of each protein amount in response to SAHA treatment. (C) Western blot analysis of representative cell cycle/proliferation-related proteins upregulated upon SAHA treatment. The acetylation status of histone H3 was also evaluated for validating a dose-dependent effect of SAHA addition on human HeLa cells in each sample set.

Figure 2

Identification and quantification of UniProt-defined human protein sequences by RTS-assisted shotgun MS analysis. (A) Numerical distribution of UniProt-defined human proteins identified in our RTS-assisted proteomic analysis. (B) Volcano plots for TMT-based quantitative proteomic changes in response to differential SAHA treatment. The red dots indicate the corresponding data on each quantified protein. The y-axis represents log10-transformed p-value adjusted by Benjamini–Hochberg method, whereas the x-axis indicates the log2-transformed fold change of each protein amount in response to SAHA treatment. (C) Western blot analysis of representative cell cycle/proliferation-related proteins upregulated upon SAHA treatment. The acetylation status of histone H3 was also evaluated for validating a dose-dependent effect of SAHA addition on human HeLa cells in each sample set.

Figure 3

Identification and quantification of non-canonical short ORF-encoded peptides by RTS-assisted shotgun MS analysis. (A) Numerical distribution of non-canonical short ORF-encoded peptides identified from the sORFs.org database. The sORF locations classified on each corresponding Ensembl transcript sequence are indicated on the x-axis. (B) Volcano plots for TMT-based quantitative proteomic changes of non-canonical sORFs in response to differential SAHA treatment. The red dots indicate the corresponding data on each quantified sORF. The y-axis represents log10-transformed p-value adjusted by Benjamini–Hochberg method, whereas the x-axis indicates the log2-transformed fold change of each protein amount in response to SAHA treatment. (C) SAHA-dependent differential regulation of the protein products generated from Fructosamine-3-kinase-related protein (FN3KRP) gene locus. Drastic decreased regulation was observed regarding the non-canonical ORF-derived peptide encoded by ENST00000571594, whereas translation of UniProt-defined protein (Q9HA64), which is encoded by the representative transcript (ENST00000269373) on the same gene locus, was not affected by SAHA treatment.

Figure 3

Identification and quantification of non-canonical short ORF-encoded peptides by RTS-assisted shotgun MS analysis. (A) Numerical distribution of non-canonical short ORF-encoded peptides identified from the sORFs.org database. The sORF locations classified on each corresponding Ensembl transcript sequence are indicated on the x-axis. (B) Volcano plots for TMT-based quantitative proteomic changes of non-canonical sORFs in response to differential SAHA treatment. The red dots indicate the corresponding data on each quantified sORF. The y-axis represents log10-transformed p-value adjusted by Benjamini–Hochberg method, whereas the x-axis indicates the log2-transformed fold change of each protein amount in response to SAHA treatment. (C) SAHA-dependent differential regulation of the protein products generated from Fructosamine-3-kinase-related protein (FN3KRP) gene locus. Drastic decreased regulation was observed regarding the non-canonical ORF-derived peptide encoded by ENST00000571594, whereas translation of UniProt-defined protein (Q9HA64), which is encoded by the representative transcript (ENST00000269373) on the same gene locus, was not affected by SAHA treatment.