Fig 1.
Phylogenetic tree based on core genome analysis of the strains included in the “Curated Database”.
The tree includes all strains of S. pneumoniae, S. pseudopneumoniae, S. mitis, including the experimental strains used in the study, as well as the type strain of the other species of the Streptococcus Mitis Group included in the database and S. pyogenes. The tree is based on 168,439 homologous amino acid positions and was constructed, using PhyML software and the aLRT algorithm.
Table 1.
Proteotyping results of the twelve representative strains included in the study.
For each species, the Type strain, as well as two additional well-characterized reference strains, were included. The numbers of identified proteins, peptides and species-unique peptides after analyses with TCUP are shown (averages of triplicate analyses). The accuracies (%), i.e. proportion of correctly assigned peptides of the total number of species-unique peptides, are also shown. For a confirmed identification, a minimum threshold of five peptide matches per species was used.
Table 2.
List of proteins identified from the species-unique peptides.
Peptides detected and identified in triplicate analysis of one of the twelve experimental strains, S. pneumoniae CCUG 28588T,were linked to their respective proteins. Only proteins having two or more peptide matches are shown here. Full lists of proteins and peptides for the representative strains of S. pneumoniae, S. pseudopneumoniae and S. mitis included in the study can be found in S4–S12 Tables.
Fig 2.
Proteotyping results of mixed samples.
Cells of S. pneumoniae and S. pseudopneumoniae or S. mitis were mixed in ratios of 1:1. Following sample preparation, digestion and LC-MS/MS analyses, the results were evaluated, using pre-computed correction factors, reflecting the expected proportion of unique peptides of each of the species. In both mixes, S. pneumoniae:S. pseudopneumoniae and S. pneumoniae:S. mitis, the results reflected a composition of approximately 50% of each species (standard error bars on averages from triplicate analyses).
Fig 3.
Ranking according to matching efficiency (%) of the identified peptides against complete genome sequences from RefSeq database.
Proteotyping results, following MS-proteomic analyses of strains of S. aureus (A), P. aeruginosa (B) and S. pneumoniae (C). For S. aureus and P. aeruginosa, the top ranked peptide matches are all with the correct species in the RefSeq database. The next-best ranked matches are for other species of Pseudomonas and Staphylococcus, marked with arrows, reflecting a distinct drop from almost 100% down to 30% in matching efficiencies of identified peptides. From the analysis of S. pneumoniae (C), the top ranked matches all belong to the correct species (S. pneumoniae), although, due to the phylogenetic relationships and taxonomy of this species in the Streptococcus genus, the matching efficiencies to other species, especially for S. pseudopneumoniae and S. mitis, is relatively higher (as much as 70–80%), thus making discovery of species-unique peptides more difficult.
Table 3.
Lists of genomes of species of the Mitis Group of the genus Streptococcus, used for matching the proteomic data at two different time points.
Two databases were used, created in February 2015 (“Initial Database”) and August 2016 (“Curated Database”). The (T) denotes the presence of the Type strain genome of a given species in the database.
Table 4.
Proteotyping results of the twelve representative strains included in the study (averages of triplicate analyses).
The number of species-unique peptide matches and accuracies (%), using the two databases are shown in the columns headed “Initial Database” and “Curated Database”. A minimum threshold of five peptide matches per species was used in the analysis. The improvement in accuracies for S. mitis and S. pseudopneumoniae is highlighted in bold.