Reduced changes in protein compared to mRNA levels across non-proliferating tissues

Perl et al. BMC Genomics (2017) 18:305 DOI 10.1186/s12864-017-3683-9 RESEARCH ARTICLE Open Access Reduced changes in protein compared to mRNA levels across non-proliferating tissues Kobi Perl1,2, Kathy Ushakov1, Yair Pozniak1, Ofer Yizhar-Barnea1, Yoni Bhonker1, Shaked Shivatzki1, Tamar Geiger1, Karen B. Avraham1*† and Ron Shamir2*† Abstract Background: The quantitative relations between RNA and protein are fundamental to biology and are still not fully understood. Across taxa, it was demonstrated that the protein-to-mRNA ratio in steady state varies in a direction that lessens the change in protein levels as a result of changes in the transcript abundance. Evidence for this behavior in tissues is sparse. We tested this phenomenon in new data that we produced for the mouse auditory system, and in previously published tissue datasets. A joint analysis of the transcriptome and proteome was performed across four datasets: inner-ear mouse tissues, mouse organ tissues, lymphoblastoid primate samples and human cancer cell lines. Results: We show that the protein levels are more conserved than the mRNA levels in all datasets, and that changes in transcription are associated with translational changes that exert opposite effects on the final protein level, in all tissues except cancer. Finally, we observe that some functions are enriched in the inner ear on the mRNA level but not in protein. Conclusions: We suggest that partial buffering between transcription and translation ensures that proteins can be made rapidly in response to a stimulus. Accounting for the buffering can improve the prediction of protein levels from mRNA levels. Keywords: Inner ear, Cochlea, Mass spectrometry, RNA-seq, Translation Background The correlation between expression levels of protein and mRNA in mammals is relatively low, with a Pearson correlation coefficient of ~0.40 [1, 2]. Suggested explanations for this low correlation include post-transcriptional regulation and measurement noise [1]. This low correlation makes it difficult to integrate protein and mRNA data. Tools for this integration are sparse and not yet adopted by the bioinformatics community (reviewed in [3]). Initial findings from such tools suggest that the transcriptional and the translational regulation evolved * Correspondence: ; † Equal contributors 1 Department of Human Molecular Genetics and Biochemistry, Sackler Faculty of Medicine and Sagol School of Neuroscience, Tel Aviv University, Tel Aviv 6997801, Israel 2 Blavatnik School of Computer Science, Tel Aviv University, Tel Aviv 6997801, Israel independently, except in the rare occasions where strong selection in favor of correlation was present [4]. However, such claims are based on data from perturbed systems, where the observed discordance between the transcriptome and the proteome is strongly affected by the lack of temporal synchronization between the transcriptional and translational regulation levels [5]. In this study we focus on the connection of mRNA and protein levels in non-proliferating tissues, through the example of the mammalian inner ear. By performing joint analysis of RNA-seq and protein mass spectrometry (MS) data from the mouse cochlea and vestibule, we aimed to shed light on the regulation of these two expression levels, identify genes that are mainly regulated in one system, and infer their general features. The two tissues are quite similar in structure, but have distinct roles in hearing and balance. This allows us to © The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Perl et al. BMC Genomics (2017) 18:305 ask questions about the contribution of each of these two systems of regulation with respect to different cellular roles. We will refer to a gene’s protein level divided by its transcript level as the gene’s protein-transcript ratio or PTR, also called the gene’s translation efficiency [6]. We note that this measure is affected by both translation and protein degradation rates, and under steady-state conditions it should be equal to the ratio of the rates [7]. It was observed that across taxa, protein levels are more conserved than mRNA levels [8], although some exceptions exist [9]. Also, it was noticed that differences in protein levels between primates are less common than differences in mRNA levels [10]. While PTR was claimed to be highly conserved between tissues for each given protein [11], it was demonstrated that it somewhat varies between tissues in a direction that buffers or compensates for the change in protein levels from changes in the transcript abundance [7], similar to what was shown across taxa. However, these observations originated from a small number of tissues, and were based mainly on regression coefficients that are affected by regression dilution bias [12]. In the first part of this study we will ask whether this phenomenon is evident in our mammalian inner ear data, and in previously obtained transcriptomic and proteomic data from different tissues. We will then use our discoveries to improve the prediction of protein levels from mRNA levels. Many experiments only measure transcript abundance in a tissue and use it as a proxy for protein levels. Previous articles that predicted protein levels from mRNA [6, 13] did not use PTR measured in other tissues, and relied mainly on sequence related features; they reached a correlation of 0.75 between the predicted and the observed levels. It has been suggested to use the average PTRs measured in other tissues in order to predict the protein levels for the tissue in question [8]. This assumes the PTR of a gene is constant across tissues. We suggest, instead, a model that assigns a higher PTR in a tissue where the mRNA level is lower. In the second part of this study we use functional analysis to compare differential expression across tissues in mRNA and protein. We give examples where inner-ear tissues maintain different levels of mRNA and similar levels of protein at rest, and hypothesize that this is done in preparation for a stimulus. Results Previous examinations of mRNA-protein relationships were mainly performed in yeast and in cancer cell lines. Aiming to examine these associations in non-transformed cells and differentiated tissue samples, we analyzed four different paired datasets of mRNA and protein. For the first dataset we generated transcriptomic and proteomics Page 2 of 14 data from the cochlea and vestibule (...truncated)