(A and B) Expression of full-length YadAwt and mutant YadA in wild-type (wt) and mutant E. coli (Ec). Whole-cell lysates were made of bacteria grown for 2 h with 200 μg/liter AHTC. Samples were either heated (A) or not heated (B) prior to SDS-PAGE, and YadA was detected by Western blotting. YadA trimer bands are observed at 200 kDa and monomer bands at 50 kDa. (C) Coomassie blue-stained SDS-PAGE of the YadA membrane anchor domain (amino acids 335 to 422). Proteins were solubilized and purified from E. coli outer membrane fractions and heated for the times indicated prior to SDS-PAGE. Gels were stained by colloidal Coomassie blue, because the membrane anchor domain is not detected by the YadA antiserum.

Immunostaining and flow cytometry analysis of YadA surface exposure. (A) Mean fluorescence of bacteria expressing YadAwt before (bold black line, no filling) and 10 min after (thin line, gray filling) induction. (B) Bacteria expressing YadAwt (thin line, gray filling) compared to the corresponding histograms of mutated YadA versions (bold black line, no filling). The mean fluorescence of bacteria was set to 100% for YadAwt.

Adhesion to collagen type I and HeLa cells mediated by wild-type and mutant YadA expressed in E. coli BL21(DE3)Omp8. Bacteria adherent to collagen type I (A) or HeLa cells (B) were counted, and the adherence of the wild type was set to 100%. The values are means ± standard deviations and are representative of three independent experiments. Bacteria with empty vector were used as controls.

Serum resistance of E. coli (Ec) mediated by wild-type and mutant YadA expressed in E. coli BL21(DE3)Omp8. Bacteria with empty vector were used as controls. Equal amounts of bacteria were incubated with NHS or HIS and plated on LB agar. The bacterial survival was calculated as a percentage, taking the bacterial counts obtained with bacteria incubated in HIS as 100%. The data are means ± standard deviations for at least three independent experiments.

(A) Sequence alignment of the membrane anchor of TAAs. Hia (GI:21536216) and Hsf (GI:15602579) are from Haemophilus influenzae, DsrA (GI:7188575) is from Haemophilus ducreyi, VompA to -D (GI:51949816, GI:51949817, GI:51949813, and GI:51949812, respectively) are from Bartonella quintana, BadA (GI:119890727) is from Bartonella henselae, NadA (GI:83616350) is from Neisseria meningitidis, STEC (GI:73853322) is from Escherichia coli, UspA1 and UspA2 (GI:5453178 and GI:26284393, respectively) are from Moraxella catarrhalis, XadA (GI:7542317) is from Xanthomonas campestris, SadA (GI:10945148) is from Salmonella enterica serovar Typhimurium, YadA (GI:32470319) is from Yersinia enterocolitica O:8, TaaHs (GI:53728655) is from Haemophilus somnus, TaaAs (GI:50084797) is from an Acinetobacter sp., TaaOa (GI:83944938) is from Oceanicaulis alexandrii, and TaaEs (GI:85709253) is from an Erythrobacter sp. The secondary structure assignment is taken from the membrane anchor crystal structure of Hia (27) and is in good agreement with earlier bioinformatic predictions (16). The coiled-coil region (in green) has core positions marked in bold. Residues of the β-strands facing the membrane are highlighted in red. The conserved glycine residue in β-sheet 2 is highlighted in blue and its mutations in violet. (B) Schematic drawing of slices through the models of Hia and YadA membrane anchors, approximately at the level of glycine residues (highlighted in blue). For comparison of the histidine side chain size with the internal cavity around glycine residues, one of the glycines in each protein was mutated to histidine (in violet). Loops connecting β-sheet with coiled coil are highlighted in cyan. Differences in slices are the consequence of different residues building each protein and the local optimization procedures of the modeling software.

Immunofluorescence microscopy of E. coli expressing wild-type and mutant YadA. The bacteria were grown at 27°C for 15 min with 200 μg/liter AHTC, and YadA on the bacterial surface was detected with rabbit anti-YadA antibodies and Cy-2 secondary antibodies. The upper panel shows YadA surface display in wild-type E. coli (Ec) and the lower panel in E. coli ΔdegP. Note the differences in fluorescence signal intensities between YadAwt and mutant YadA when expressed in E. coli or E. coli ΔdegP. Bacteria containing the empty vector were used as controls.

Autoaggregation of wild-type and mutant YadA. After 2 h of induction, cultures were allowed to settle at room temperature for 2 h. Clearance of the culture is a qualitative measure of autoaggregation. Ec, E. coli.

YadA-induced IL-8 secretion of HeLa cells mediated by wild-type and mutant YadA expressed in E. coli BL21(DE3)Omp8 upon infection. Bacteria with empty vector were used as controls. IL-8 levels were determined by ELISA from culture supernatants. The values are means and standard deviations from four independent experiments.

Autotransport mechanism: after translation (1) the signal peptide is cleaved (2) and the polypeptide is exported via the Sec machinery (3). Insertion via C-terminal signals (4) is followed by the formation of the barrel domain (5). This is the prerequisite for autotransport (6) and folding of the extracellular adhesin domains (7). If autotransport is slowed or blocked, the protein is degraded by DegP (8). OM, outer membrane; IM inner membrane.