Introduction:
Explaining emergent properties of complex organisms represents a grand challenge of biology (Schwenk, Padilla, Bakken, & Full, 2009). Emergent properties are often based on non-linear interactions of their molecular constituents (Stillman et al., 2011). Despite the known non-linearity of the genome to phenome continuum many biological studies rely heavily on correlations of complex organismal phenotypes (traits) with genomic variation (e.g. SNPs and other types of sequence variation) via QTL or GWAS analyses (Kratochwil & Meyer, 2015; Wray et al., 2013) and mRNA abundance changes via transcriptomics while proteome dynamics is less commonly investigated (Evans, 2015).
Unlike the genome, transcriptomes and proteomes are spatially highly heterogeneous and dynamic, i.e. they are highly responsive to environmental stimuli. Currently, there is a large gap in the literature regarding transcriptome versus proteome data and systematic comparisons aimed at discerning the rules of non-linearity between these two levels of biological organization that are not common (Buccitelli & Selbach, 2020). There is a great need for developing robust and comprehensive quantitative proteomics approaches to facilitate such comparisons and close the gap between genotypes and ecologically relevant phenotypes, e.g. environmental stress tolerance. Understanding the connection between mRNA and protein levels is fundamental to predicting how the underlying genetic code impacts changes in phenotype due to environmental changes. Non-linearity between transcriptome and proteome levels of regulation is well documented (Franks, Airoldi, & Slavov, 2017). It can be based on differential mRNA processing and degradation (Bentley, 2014), transcript-specific regulation of protein translation through all stages including initiation, elongation, and localization (Tahmasebi, Khoutorsky, Mathews, & Sonenberg, 2018), and/ or regulation of protein degradation (Pohl & Dikic, 2019).
The proteome represents the core that is central to the genome to phenome continuum. On the one hand, proteins are linked directly to specific genes via proteotypic peptides, which allows unambiguous association of each protein with a specific genomic locus (Keerthikumar & Mathivanan, 2017). On the other, proteins represent the critical molecular building blocks that define structure and carry out most biochemical processes and functions of cells, tissues and organisms (Ebhardt, Root, Sander, & Aebersold, 2015). Proteins represent the molecular constituents giving rise to phenotypic variability that is acted upon by natural selection (Clarke, 1971; Frömmel & Holzhütter, 1985; Mularoni, Ledda, Toll-Riera, & Albà, 2010). Moreover, most targets of pharmaceutical drugs are proteins, which illustrates that proteins control critical organismal phenotypes (Batchelor, Loewer, Mock, & Lahav, 2011; Ebhardt et al., 2015).
Recent developments in biological mass spectrometry have enabled robust gel- and label-free quantitative proteomics workflows that are well suited for organismal biology and molecular ecology (Huang et al., 2015; Vowinckel et al., 2013). In particular, the invention of data-independent acquisition (DIA) liquid chromatography mass spectrometry (DIA-LCMS2) holds great promise for molecular ecology studies (Crowgey, Matlock, Venkatraman, Fert-Bober, & Van Eyk, 2017; Schubert et al., 2015). The DIA approach is also referred to as Sequentially Windowed Acquisition of all theoretically possible MSMS spectra (SWATH)-MS (Arnhard, Gottschall, Pitterl, & Oberacher, 2015; Huang et al., 2015).
DIA-LCMS2 represents a merger of pre-acquisition targeted mass spectrometry approaches, i.e. selected reaction monitoring (SRM) or multiple reaction monitoring (MRM), and non-targeted data acquisition that is independent of precursor (MS1) spectra acquisition (Koopmans, Ho, Smit, & Li, 2018). In DIA-LCMS2 the targeting of specific transitions, precursors, peptides, and proteins is performed post-acquisition by interrogating all theoretically possible fragment ion (MS2 spectra) present in a sample against a previously validated DIA assay library. Here, we have constructed a DIA assay library from raw MS2 spectral libraries of Nile tilapia (Oreochromis niloticus ) kidney to facilitate quantitative studies of proteome dynamics in response to environmental stress and other ecological contexts. We demonstrate the utility of this DIA assay library by identifying proteins, biological functions, and processes that are associated with salinity acclimation in O. niloticus kidney. Furthermore, we demonstrate that quantitative proteomics provides knowledge that cannot be gained from transcriptomics data.
Methods :