Fig 1.
PCA plot of the location of aligned sequences of HU/IHF proteins on three axes.
Three main groups of proteins, HU, IHF_A and IHF_B, are indicated as well as results of further subdivision of the protein sequences (HU and IHF clades).A. Most populated HU clades: clade HU_Firmicutes (mainly originated from HU of Firmicutes species) and clades HU_ecoA and HU_ecoB (mainly from proteobacteria) are shown in magenta, bleu and cyan, respectively. Other apparent HU/IHF clades are indicated. Position HUs and IHFs proteins of E. coli are shown in red, as well as positions of HU of N. gonorrhoeae (NG), M. gallisepticum (MG), and S. melliferum (SM). E. coli HUα and Huβ are very close to each other (62 identities in 90 amino acid core sequence), and IHFα and IHFβ are far from each other (24 identities of 90). We believe that it is the reason why HU separation onto two groups, one close to HUα, and another close to HUβ, is ambiguous. B. IHF clades. Result of two major IHF group subdivisions shown in color: IHF_A magenta, cyan and green, IHF_B red, yellow, and blue. HU sequences are shown in grey.
Fig 2.
Consensus sequences of three IHF_A, three IHF_B, and several HU clades revealed in this study.
Conservative residues are highlighted. Alpha helixes and beta sheets are indicated with yellow and blue blocks, respectively.
Fig 3.
Average number of identic amino acid residues in protein sequences of HU/IHF clades.
Number of identic amino acid residues in protein sequences was calculated for each pair of HU/IHF proteins. Each cell represents an average number of identic amino acids for sequences from two clades. Only residues within 90 amino acid long core of HU/IHF were taken into account; non-equivalent insertion or deletion in protein sequence was calculated as one non-identic residue.
Fig 4.
Model of E.coli HUα dimer with hotspots for amino acid insertions and deletions.
Each HU monomer contains three alpha helixes and five beta strands. HU body (helixes 1 and 2) is responsible for dimer stabilization. HU arms are responsible for DNA binding. Hotspots for amino acid insertions and deletions in HU are shown.
Fig 5.
Model of E.coli HUα dimer with distances between C-alpha atoms that were measured.
Positions of amino acid residues corresponding to the C-termini of alpha helixes 2 are indicated and distance between corresponding C-alpha atoms of both monomers is shown with khaki dotted line. Positions of the amino acid residues at the C-termini of beta strands 2 and N-termini of beta strands 5 as well as their neighboring residues are indicated and the corresponding distances between C-alpha atoms of both monomers are shown with rose and blue dotted line, respectively.
Fig 6.
Model of Pseudomonas syringae HU (magenta) superimposed with E. coli HUα model (cyan).
Angle between alpha helix 1 and alpha helix 2 is shown in green.
Fig 7.
Model of HU Bifidobacterium longum from Actinobacteria (clade HU_acti_0), superimposed with E. coli HUα model.
B. longum HU is shown in yellow, E. coli HUα—in cyan. Angle between alpha helix 2 and alpha helix 3 is shown in magenta.
Fig 8.
Models of S. melliferum (blue), N. gonorrhoeae (green), M. gallisepticum (grey), and H. pylori HU (deep teal) superimposed with E. coli HUα model (cyan).
Two-residue insertion between alpha helix 1 and 2 in S. melliferum HU; six amino acid long extension at the N-termini of M. gallisepticum HU and four amino acid long extension at the C-termini of H. pylori HU are indicated with red circles.
Fig 9.
HU binding to dsDNA of various lengths.
Binding of labeled DNA to HU proteins was analyzed by polyacrylamide gel electrophoresis. The gel was buffered with 50 mM Tris–borate; binding mixture contains 40 mM NaCl. DNA samples were: dsDNA of sequence ‘D’ with the length varying from 21 to 48 bp (indicated at the bottom). HU origin and concentration is indicated at the top (“-“, no HU was added). Bands corresponding to HU-DNA complexes are marked with arrows, the number of HU dimers in each complex is indicated on the left of the arrow. Panels correspond to HU proteins of various bacteria, protein concentrations are indicated.
Fig 10.
HU binding to “distorted” DNA structures checked by polyacrylamide gel mobility assay.
HU protein at concentrations indicated above the gel image (“-“, no HU was added) was mixed with 5’-labelled DNA in a buffer containing 150 mM NaCl; the bound and free DNA were gel-separated. DNA structures indicated at the bottom of the gel images: n, nicked DNA; ds, dsDNA; A1, A3 and A7, DNA bulges, containing one, three or seven non-paired adenines in one of DNA strands; J–four-way junction; fork, ssDNA fork; ov, DNA overhang; iJ, incomplete junction lacking one DNA strand; inv, DNA invasion. Panels correspond to HU proteins of various bacteria.
Fig 11.
The profiles of HU affinities to various DNA-structures.
Each column represents the association constant of HU complex with one DNA structure (see Fig 10 legend for structures description) normalized on the association constants of HU-dsDNA complex of the same protein. Data from at least three independent experiments were combined.