Progress and Challenges in Ocean Metaproteomics and Proposed Best Practices for Data Sharing

Abstract

Ocean metaproteomics is an emerging field enabling discoveries about marine microbial communities and their impact on global biogeochemical processes. Recent ocean metaproteomic studies have provided insight into microbial nutrient transport, colimitation of carbon fixation, the metabolism of microbial biofilms, and dynamics of carbon flux in marine ecosystems. Future methodological developments could provide new capabilities such as characterizing long-term ecosystem changes, biogeochemical reaction rates, and in situ stoichiometries. Yet challenges remain for ocean metaproteomics due to the great biological diversity that produces highly complex mass spectra, as well as the difficulty in obtaining and working with environmental samples. This review summarizes the progress and challenges facing ocean metaproteomic scientists and proposes best practices for data sharing of ocean metaproteomic data sets, including the data types and metadata needed to enable intercomparisons of protein distributions and annotations that could foster global ocean metaproteomic capabilities.

Keywords: Metaproteomics; best practices; biogeochemistry; data sharing; ocean.

Figure 1.

Analysis of proteins within natural environments presents unique challenges that can be improved upon to allow this new type data to inform ecosystem function and change. These challenges span sample collection and extraction, mass spectrometry analysis, informatic approaches, and data management and dissemination.

Figure 2.

(a) Collection of ocean metaproteomic samples by in situ underwater McLane pump sampler as deployed in Terra Nova Bay of the Ross Sea in Antarctica aboard the icebreaker R/V Palmer to capture the microbial and algal communities as well as larger sinking particles by filtration of several hundreds of liters. (b) Example vertical distributions of three microbial proteins in the Equatorial Pacific Ocean using targeted metaproteomics that are biomarkers of nitrogen (N), phosphorus (P), iron (Fe) nutrient stress, and nickel (Ni) biogeochemical cycling (data from Saito et al., 2014, https://www.bco-dmo.org/dataset/646115). Proteins shown include the nitrogen PII regulator protein from Prochlorococcus (sequence VNSVIDAIAEAAK), the sulfolipid biosynthesis protein from Prochlorococcus (NEAVENDLIVDNK), UDP sulfolipid biosynthesis protein from multiple taxa (FDYDGDYGTVLNR), the IdiA iron transporter from Prochlorococcus (SPYNQSLVANQIVNK), and the nickel superoxide dismutase enzyme from Prochlorococcus and Synechococcus (VAAEAVLSMTK). Taxonomic assignments determined using METATRYP.

Figure 3.

An example environmental metaproteomic workflow where environmental samples are collected and extracted (gray), discovery proteomics are conducted (green), and peptide targets from selected proteins of interest can be assayed using isotopically labeled peptide standards whose taxonomic assignment can be queried against databases of genomes and metagenomes (yellow). The results can provide relative and absolute abundance measurements of the protein from the microbial and algal community, including functional and taxonomic information (blue).

Figure 4.

(a) Three-dimensional representation (axes of retention time (min), m/z, and intensity) of complex spectra associated with an environmental ocean sample from the Equatorial Pacific from the METZYME expedition (200 m depth, produced in MzMine2). Comparison of a 3 m/z ms¹ mass window (575–578 m/z, 140–141 min) from (b) human proteome spectra (HeLa cell line) and (c) ocean metaproteome (120 m depth) provides an example of the high complexity of environmental samples due to the biological diversity.

Figure 5.

Peak analysis of human cell line and ocean metaproteome samples by identical chromatographic and mass spectrometry conditions. (a–b) Number of peaks identified in replicates (top rows) and the total ion current (TIC, bottom rows) of the sample in Hela (Panel A, replicates Hela-A, Hela-B, and Hela-C) and an ocean metaproteome sample (Panel B, depth 40, 60, 120, 150, and 250 m, Metzyme Expedition Station 3). Samples were run during the same week on the same nanospray column (see methods) with similar amounts of protein injected (0.5 μg for Hela per injection, 1 μg for ocean metaproteome). (c) Total number of peaks by sample type showed a higher number peaks in ocean metaproteome samples consistent with Figure 4, while (d) TIC by sample showed much lower summed peak intensity within the metaproteome samples.