H-NS is a nucleoid structuring protein and global repressor of virulence and horizontally-acquired genes in bacteria. H-NS can interact with itself or with homologous proteins, but protein family diversity and regulatory network overlap remain poorly defined. Here we present a comprehensive phylogenetic analysis that revealed deep-branching clades, dispelling the presumption that H-NS is the progenitor of varied molecular backups. With few exceptions, clades are either entirely chromosomal or entirely plasmid-encoded proteins. On chromosomes, StpA and newly discovered HlpP are core genes in specific genera, whereas Hfp and newly discovered HlpC are sporadically distributed. Six clades of H-NS plasmid proteins (Hpp) exhibit ancient and dedicated associations with plasmids, including three clades with fidelity for plasmid incompatibility groups H, F, or X. A proliferation of H-NS homologs in Erwiniaceae includes the first observation of potentially co-dependent H-NS forms. Conversely, the observed diversification of oligomerization domains may facilitate stable co-existence of divergent homologs in a genome. Transcriptomic and proteomic analysis of regulatory crosstalk in Salmonella revealed networked and hierarchical control of H-NS homologs. We also discovered that H-NS is both a repressor and activator of Salmonella Pathogenicity Island 1 gene expression, and both modes are restored by Sfh (HppH) in the absence of H-NS.

Cells from exponentially growing wildtype and mutant S. Enteritidis EN1660 cultures (100 ml) were harvested at OD_600nm 0.12 – 0.15, then lysed with bead beating. Proteins were extracted using acetone precipitation overnight at -20 °C. After resolubilization in 50 mM ammonium bicarbonate buffer, 50 µg of protein sample was digested with trypsin (Sigma-Aldrich). After drying, the resulting peptides were spiked with 40 fmol/µL rabbit phosphorylase B in running buffer (0.1% formic acid, 3% acetonitrile) before analysis with a Synapt G2 HDMS (Waters) coupled to a Nanoacquity (Waters) nano-LC with an Acquity UPLC T3HSS column (75 mm x 200 mm). The separation was conducted with a gradient from 3% acetonitrile/0.1% formic acid to 45% acetonitrile/0.1% formic acid at a flow rate of 0.3 µL/min. A total of 1 µg digest was injected in each run and eluting peptides were analysed in positive data-independent-acquisition mode (MSE) with 1s scan time. In low-energy MS mode data were collected at a collision energy of 4eV. High-energy collision energy was ramped between 18 and 42 V. Leucine enkephaline was measured as lock mass every 30s to maintain mass accuracy throughout the run.  The resulting spectra were analysed with the ProteinLynx Global Server (PLGS) v. 3.02 and searched against the UniProt Salmonella enterica reference proteome (UP000001014, accessed January 2016). Protein abundance was measured together with the identification using phosphorylase B as calibrant. The false discovery rate of protein identifications was set to 4% per run. i.e. proteins identified in all three biological replicates had an expected false discovery rate of <0.001%.

File names correspond to Salmonella Enteritidis mutants as outlined in the manuscript. Data are in Waters native file format and have been compressed with 7zip (https://www.7-zip.org/). Since the data was acquired in data-independent mode (MSE) processing using ProteinLynx Global Server is recommended. Downstream analyses are open to different workflows including Ion-accounting (Waters) Mascot, Skyline etc.