Small open reading frame encoded proteins (SEPs) gained increasing interest during the last years due to their broad range of important functions in both, prokaryotes and eukaryotes. In bacteria, signalling, virulence or regulation of enzyme activities have been associated with SEPs. Nonetheless, the number of SEPs detected in large-scale proteome studies is often low as classical methods are biased towards the identification of larger proteins. In the accompanying manuscript, we present a workflow that allows enhanced identification of small proteins compared to traditional protocols. For this aim, the steps of small protein enrichment, proteolytic digest and database search were reviewed and adjusted to the special requirement of SEPs. Enrichment by the use of small-pore-sized solid-phase material increased the number of identified SEPs by a factor of two and the utilisation of alternative proteases to trypsin reduced spectral counts for larger proteins. The application of the optimised protocol allowed the detection of 210 already annotated proteins up to 100 amino acids length, including 16 proteins below 51 amino acids in the Gram-positive model organism Bacillus subtilis. Moreover, 12% of all identified proteins were up to 100 amino acids which is a significant larger fraction than reported in studies involving traditional proteomics workflows. Finally, an integrated proteogenomics search database was applied to identify potentially novel SEPs. This submission contains PRM data generated in order to verify the identification of novel very small proteins in Bacillus subtilis. Three SEPs, which are 21, 26 and 42 amino acids long, respectively were successfully verified.

In order to validate the identification of peptides assigned to proteins not annotated in the NCBI RefSeq database of Bacillus subtilis, 68 reference peptides for 37 proteins were synthesised by JPT Peptide Technologies (Berlin). Upon extensive manual curation, selective PRM assays were obtained for 37 reference peptides belonging to 27 proteins. Assays could not be established for the remaining peptides, either because the peptide did not elute well, generated only a low number of suitable fragment ions, or was interfered by an abundant co-eluting precursor with a similar mass and interfering transitions (as tested by SPE enriched samples with spiked synthetic peptides). After the establishment of the assays, the LC was equipped with a fresh column and about 1 µg peptide of small protein-enriched (solid phase-enrichment with surface access-limited reverse-phase material) and Lys-C digested samples from 14 cultivation conditions was subjected to these assays. The Q-Exactive was coupled to an EASY-nLC 1200 (Thermo-Fisher Scientific), which was equipped with an in-house built 20 cm reversed-phase column. The column had an integrated emitter tip and was packed with C18 particles (3 µm diameter, ReproSil-Pur 120 C18-AQ, Dr. Maisch). The instrument was operated in a time-scheduled SIM-mode, where one survey scan (300 – 1650 Th mass range; 70,000 resolution at m/z 200; 3 x 1e6 predictive automatic gain control target; max. 200 ms injection time; activated lock mass correction) was followed by up to 10 fragment-scans (higher-energy collisional dissociation (HCD) at normalised energy of 27 for pre-defined precursor ions, mass range dependent on precursor m/z; 70,000 resolution at m/z 200; 1 x 1e5 predictive automatic gain control target; max. 200 ms injection time, isolation window 1.2 m/z, isolation offset 0.2 m/z ). Peptides with a library dot product (in comparison to the synthetic peptides measured under the same conditions) of at least 0.7, a co-eluting monoisotopic precursor ion and at least 5 co-eluting fragment ions (including a minimum of 3 b-or y-ions with a summed intensity of at least 1 x 1e5) in measurements from three or more different cultivation conditions were considered as verified.

Samples obtained from 14 different growth conditions were analysed to characterise the small proteome of B. subtilis. The selected conditions included chemically defined and rich media, induction of diverse stresses, different carbon sources and conditions that initiate sporulation. Cells were collected by centrifugation (10,000 x g, 4 °C, 5 min), the remaining growth medium was removed by washing with Tris-HCl buffer (50 mM, pH 7.4) and cells were stored in aliquots at 80 °C until further processing. A method based on the interaction of proteins with a solid phase was developed to enrich small proteins present in bacterial protein extracts. For this, B. subtilis cells in Tris-HCl buffer were disrupted by ultrasonication with an MS-72 sonotrode operated at 40 W output (Bandelin, 4 x 30 sec). A mild centrifugation step at 8,000 x g for 5 minutes ensured the minimal loss of membrane-associated small proteins, while undisrupted cells and cell debris were sufficiently removed. A disposable solid-phase enrichment (SPE) column (Phenomenex, 8B-S100-AAK), packed with an 8.5 nm pore size, modified styrene-divinylbenzene resign was equilibrated by passing twice 1 ml acetonitrile and twice 1 ml water through it. Subsequently, 500 µg protein (as determined by Bradford assay using bovine serum albumin as external calibrant) was loaded, and unbound, potentially larger proteins were removed by washing twice with 1 ml water. The enriched small protein fraction was eluted with 500 µl 70% (v/v) acetonitrile and evaporated to dryness in a vacuum centrifuge. This procedure typically yielded about 20 µg small protein sample. For digestion the enriched small protein sample was suspended in 100 µl 50 mM triethylammonium bicarbonate buffer containing 0.2% RapiGest. All samples were subsequently reduced for 45 minutes with 0.5 µmol Tris(2-carboxyethyl) phosphine (TCEP) at 65 °C. The reaction was quenched by adding 1 µmol iodoacetamide; the mixture was allowed to alkylate for another 15 minutes in the dark at room temperature and Lys-C was added in a 1: 100 enzyme to protein ratio. The samples were digested in a temperature-controlled shaker for 12 h at 37 °C. After completion of the digestion, RapiGest was removed by hydrolysis with trifluoroacetic acid for 30 minutes. Prior to analysis, peptides were purified by Pierce C18 Tips (Thermo Fisher Scientific) according to the manufacturer's protocol.