A Robust and Universal Metaproteomics Workflow for Research Studies and Routine Diagnostics Within 24 h Using Phenol Extraction, FASP Digest, and the MetaProteomeAnalyzer

Abstract

The investigation of microbial proteins by mass spectrometry (metaproteomics) is a key technology for simultaneously assessing the taxonomic composition and the functionality of microbial communities in medical, environmental, and biotechnological applications. We present an improved metaproteomics workflow using an updated sample preparation and a new version of the MetaProteomeAnalyzer software for data analysis. High resolution by multidimensional separation (GeLC, MudPIT) was sacrificed to aim at fast analysis of a broad range of different samples in less than 24 h. The improved workflow generated at least two times as many protein identifications than our previous workflow, and a drastic increase of taxonomic and functional annotations. Improvements of all aspects of the workflow, particularly the speed, are first steps toward potential routine clinical diagnostics (i.e., fecal samples) and analysis of technical and environmental samples. The MetaProteomeAnalyzer is provided to the scientific community as a central remote server solution at www.mpa.ovgu.de.

Keywords: bioinformatics; environmental proteomics; gut microbiome; mass spectrometry; microbial communities; sample preparation; software.

FIGURE 1

Comparison of new (A1–A6) and old workflow (B1–B8) for metaproteomics sample preparation and analysis. In addition, methods for quality control are shown (X1–X3). The time shown represents the shortest possible time in which a single sample can be processed. Under reasonable circumstances five samples can be done in less than 24 h (or 15 samples within 48 h) using the new workflow limited by the number of available mass spectrometer. Similarly, at least 3 days are required for multiple samples using the old workflow.

FIGURE 2

MetaProteomeAnalyzer. Workflow of the MetaProteomeAnalyzer software including improvements and additions to the first MetaProteomeAnalyzer version (Muth et al., 2015a). Improvements were highlighted in red.

FIGURE 3

Visualizations of MetaProteomeAnalyzer using data from NewWF_BGP_3_B. (A) The taxonomy results view of the protein tables hierarchically orders proteins by taxonomy allowing for easy selection and filtering of specific taxonomies. (B) Pie Chart with spectral counts of the biological process ontology of the Phylum Euryarchaeota selected through the taxonomy view. (C) Interactive chord diagram visualizing the relationship between taxonomy (rank = family) and functional ontology (UniProt keywords for Biological Process) (Zoun et al., 2017). Biological processes for Methanosarcinaceae, as an example, are highlighted. (D) KEGG pathway map for central carbon metabolisms (KEGG map 01200) highlighting enzymes identified with the MPA.

FIGURE 4

Comparison of protein extraction of human gut samples of new and old workflow. For protein separation a 12% SDS-PAGE with 1 mm gel thickness was carried out and stained with colloidal coomassie. Proteins were extract by the old workflow (A) and new workflow (B). Peptide electrophoresis (C) was carried out after FASP digest according to Schägger (2006) using a 10 and a 16% acrylamide gel. (STD) molecular weight standard; (Hgut 1–3) 100 μg of human fecal sample 1–3 resp. 90 μg for peptide electrophoresis Quality and purity of protein extracts was examined by SDS-PAGE (Supplementary Presentation S1).

FIGURE 5

Increase of (A) measured spectra and (B) identified spectra using the new workflow of sample preparation compared against the old workflow. The data was analyzed with MPA v2. The four types of samples from BGP, human gut, soil, and compost, and WWTP show significant differences regarding spectral counts for old and new workflow (p-values of t-test are shown in the figure). Similar results were obtained for identified peptides, percentage of identified spectra or identified metaproteins (Supplementary Table S8). P-values: ^∗p = 0.05, ^∗∗p = 0.01, ^∗∗∗p = 0.001, ^∗∗∗∗p = 0.0001.

FIGURE 6

Amount of shared metaproteins between the old and new workflow. The upper triangular matrix shows the amount of shared metaproteins of the different BGP samples using the new workflow. The lower triangular matrix shows the amount of shared metaproteins of the different BGP samples using the old workflow. The diagonal shows the amount of shared metaproteins of the same sample analyzed by the old and the new workflow. For the calculation of the amount of shared metaproteins, the number of shared metaproteins was divided by the smaller number of metaproteins from both samples. For this analysis only metaproteins were considered which had in at least one sample a spectral count of 4. MP, metaprotein.

FIGURE 7

Amount of shared metaproteins between the old and new workflow. KEGG map for the carbon metabolism showing enzymes in the sample BGP_1 (three technical replicates combined, analyzed with MPAv2.12). The map is colored to highlight differences between functional annotation, where blue are KO numbers exclusively found in the analysis with old workflow, red are KO numbers exclusively found in the analysis with the new workflow and green are KO numbers found with both. The maps are also hosted on: http://www.mpa.ovgu.de/review/kegg_carbonmetabolism_BGP_1.png.

FIGURE 8

(A) Reproducibility using replicated samples. The spectral counts of the metaproteins from the sample NewWF_BGP_1_A were plotted against the spectral counts of the metaproteins from the sample NewWF_BGP_1_B. The points in the blue or the orange area are at least doubled in the corresponding sample. (B) Differences between samples. The spectral counts of the metaproteins from the sample NewWF_BGP_1_A were plotted against the spectral counts of the metaproteins from the sample NewWF_BGP_2_A. The points in the blue or the orange area are at least decreased (blue) or increased (orange) twofold.

FIGURE 9

Improved protein annotation via BLAST using MPAv2 in comparison to MPAv1. (A) Increase of annotated spectra. (B) Identified KO-numbers. Significance values calculated by Student’s t-test for differences between the old and the new workflow are shown above the plots. The comparison was carried out with data obtained with the new laboratory workflow. The samples BGP, human gut, soil, and compost, and WWTP as well as their averages (black line) are shown separately. For further detail see Supplementary Table S15. P-values: ^∗p = 0.05, ^∗∗p = 0.01, ^∗∗∗p = 0.001, ^∗∗∗∗p = 0.0001.

FIGURE 10

Impact of peptide database lookup on reported proteins (A) and metaproteins (B) for MPAv1 and MPAv2. The comparison was carried out with data obtained with the new laboratory workflow. The bars represent the accumulated number of proteins/metaproteins for each sample group.

FIGURE 11

Grouping of samples using PCoA. Principle coordinate analysis of all samples extracted with the previous (square) and the new (dots) workflow using the Past 3 tool and the Bray–Curtis distance as parameter. For analysis, all metaproteins that represented at least one percent of the identified spectra in at least one sample were considered. The samples comprise the three BGP samples 1–3 (aqua, cornflower blue, teal), the three human fecal samples 1–3 (light pink, purple, red), the WWTP samples (navy), the soil sample (brown) and the compost sample (dark green).

FIGURE 12

Separation of samples in cluster tree. Cluster analysis of all samples extracted with the previous and the new workflow using Matlab and the “cityblock” distance and the “average” linkage as parameter was carried out. For analysis, all metaproteins that represented at least one percent of the identified spectra in at least one sample were considered. The samples comprise the three BGP samples, the three human fecal samples 1–3, the WWTP samples, the soil sample, and the compost sample.

FIGURE 13

Chord-diagram visualizing the taxonomy-function-relationships for samples BGP 1A–C. Data was exported from MPA. All taxonomies except bacterial and archaeal orders were removed in the diagram (chord diagram for a Hgut sample is found in Supplementary Table S10).