Fig 1.
A core motif in the biofilm-derived peptide is identified as key for lipid binding through MD simulations.
(a) Schematic of Bap1 structure and sequence of the Bap1-derived peptide (residues 415-471, abbreviated as Bap1-57aa). Bap1-57aa corresponds to a loop in the biofilm-specific adhesion molecule Bap1 in Vibrio cholerae. Four pseudo repeats in the Bap1-57aa sequence are shown in blue, magenta, black, and red. The core hydrophobic motif (WFFG) is shown in cyan and the rest of the residues are shown in green. (b) Snapshot from a simulation of Bap1-57aa-membrane binding, started from a disordered conformation for the peptide with the Cα atoms of the N- and C-terminal residues restrained at 28 Å. The middle linker W440FFG443 inserts into the membrane (zoomed view). (c) Membrane contact frequencies of all amino acid residues from the simulations without N-C distance restraint (IDP-wor), with N-C distance restraint (IDP-wr), and in a conformation melted from the AlphaFold structure (AF-melt). (d) Mean Ztip values from the IDP-wor, IDP-wr, and AF-melt simulations, respectively. Ztip corresponds to the Z coordinate (along the membrane normal) of the sidechain tip heavy atom of each residue relative to the proximal phosphorus plane.
Fig 2.
Experimental evidence for the importance of the core motif and avidity of the pseudo repeats in lipid binding.
(a) Schematic of the fluorescence-based bead adsorption assay. Created in BioRender. Yan, J. (2026) https://BioRender.com/3t4yegs. (b) Robust lipid binding of Bap1-57aa with respect to lipid composition. Excess fluorescence signals on the bead surface relative to the solution signals are plotted against initial peptide concentration, using beads coated with various lipids. Dextran (4 kDa) conjugated to FITC was used as a negative control. DOPC = 1,2-dioleoyl-sn-glycero-3-phosphocholine. DOPS = 1,2-dioctadecenoyl-sn-glycero-3-phosphoserine. PIP2 = phosphatidylinositol 4,5-bisphosphate. Chol = cholesterol. SM = sphingomyelin. (c) Adsorption curves with Langmuir model fitting for Bap1-57aa and sequence variants on DOPC-coated beads. (d) Dissociation constants derived from fitting adsorption curves to the Langmuir model, for peptides with different sequences. (e) Schematic of the biofilm-based adsorption assay. Rh-PE labeled SUVs were added to a confluent layer of biofilm grown from V. cholerae cells constitutively expressing mNeonGreen to assess the ability of the biofilm to capture SUVs. Non-adherent SUVs were removed during the washing step. (f) Cross-sectional images of a biofilm from cells constitutively expressing SCFP3A (cyan); Bap1 was tagged with 3 × FLAG and labeled with anti-FLAG antibody conjugated to FITC (green). Rh-PE labeled SUVs (red) bind to the periphery of the cell cluster. Cyan cells were used in this particular experiment to avoid spectral overlapping with the FITC-anti-FLAG antibody. (g) Quantification of SUV capture by biofilms formed by various V. cholerae mutants using the total SUV signal intensity normalized by the biofilm surface area. a.u. stands for arbitrary unit. A genetic background lacking rbmC was used to avoid the confounding effect of the other adhesin. All data represent mean ± SD (n = 3 biological replicates). Statistical analyses were performed using two-tailed t-tests with Welch’s correction. ns, not significant; ***, p < 0.001; ****, p < 0.0001. p values from left to right: 0.1006, 0.7815, 0.3136, < 0.0001, 0.0005, and 0.0006.
Fig 3.
The core motif binds lipid membranes in a β-hairpin conformation. (a-b) Normalized tryptophan fluorescence spectra of (a) 3 μM core motif or (b) 3 μM Bap1-57aa mixed with 3:1 DOPC/DOPS SUV at molar ratios between 1:0 and 1:150 in 10 mM Tris buffer pH 7.4 and 150 mM NaCl. Insets: Maximum tryptophan fluorescence intensity of the corresponding peptide plotted over a series of peptide:lipid molar ratios. (c) CD spectra of 50 μM core motif mixed with 3:1 DOPC/DOPS SUV at peptide:lipid molar ratios between 1:0 and 1:12 in 2 mM Tris buffer pH 7.4 and 5 mM NaCl at 20 °C. The characteristic peak at 205 nm indicates a β-turn structure, which is absent in the lipid-free CD spectrum of the mutant sequence in which the WFFG residues are replaced with LGPE. (d) CD spectra of 25 μM Bap1-57aa peptide mixed with 3:1 DOPC/DOPS SUV at molar ratios between 1:0 and 1:4 in 2 mM Tris buffer pH 7.4 and 5 mM NaCl at 20 °C. a.u. stands for arbitrary unit.
Fig 4.
Simulation of the Bap1-57aa peptide interacting with membranes in a β-hairpin conformation. (a) A snapshot showing the core motif (S438YWFFGWHTK447; cyan) inserting into the lipid bilayer as a β-hairpin. Inset: schematic of the β-hairpin model for the core motif. (b-c) Frequency of secondary structures (SS) for each residue in simulations starting from a β-hairpin conformation, (b) at the lipid bilayer surface or (c) in solution. Inset in (c) shows a representative snapshot of Bap1-57aa in solution. In (b), the two anti-parallel β-sheets adjacent to each other signify a tight β-turn structure. (d) Contact frequency of the amino acids and the membrane surface and (e) Ztip distance from simulations starting from the β-hairpin conformation. Eight independent simulations were performed for 1.1 µs each; average properties are plotted.
Fig 5.
Bap1-57aa binds to cell surfaces.
Shown are 3D renderings of confocal images of Caco-2 cells stained with 300 nM of DAPI (blue) for nuclei, 3 μg/mL of FM 4-64 (red) for membranes, and 1.5 µM FITC-labeled peptides (green). The total size of each image is 220 × 220 × 28 µm.
Fig 6.
Bap1-57aa interaction with liposomes may be curvature sensitive.
(a) Schematic of the floatation assay. A mixture of liposomes (extruded through a 50 nm filter, 50 μM total lipid, containing 0.5% Cy5-DOPE) and FITC labeled Bap1-57aa peptide (40 nM) are layered at the bottom of an iodixanol gradient. The three layers of iodixanol media (volume and concentration labeled) correspond to buoyant densities of 1.006, 1.111, and 1.155 g/mL (top to bottom). Upon centrifugation, the liposome-peptide complexes distribute along the gradient according to their average density. (b) SDS-PAGE analysis of 11 fractions (60 μL each, F1-F11) recovered from the post-centrifugation gradient and a final 60 μL wash of the centrifuge tube (F12*). L: ladder. Ctrl: pre-flotation mixture as a control. See methods for details. Pseudo-colors: Cy5, red; FITC, green. (c) FITC to Cy5 fluorescence ratio (FITC/Cy5, measured from the gel image in b and normalized to the control) plotted as a function of vesicle diameter (measured by negative-stain TEM). Box-and-whisker plots display the 25th–75th percentile and the minimum-to-maximum range, with the line and ‘+’ indicating the median and mean, respectively. Number of liposomes measured: 262 (F2, brown), 212 (F3, red) and 122 (Ctrl, pink). ****: p < 0.0001 in two-sample Kolmogorov-Smirnov test. a.u.: arbitrary unit. (d) Representative negative-stain TEM images of vesicles in F2, F3 and the pre-flotation sample. Scale bars: 50 nm.
Fig 7.
Distribution and conservation of Bap1-57aa across the Vibrio genus.
(a) Distribution of Bap1-57aa across the Bap1 encoded gene tree. The gene tree is rooted with four RbmC-encoded genes as outgroups. The presence or absence of Bap1-57aa in each gene is denoted by dark blue and light gray circles at the tree tips, respectively. Ancestral states for the presence or absence of Bap1-57aa are shown at internal nodes. A color bar indicates the species origins of the Bap1 encoded genes. (b) Sequence logo of Bap1-57aa. The X-axis represents amino acid positions, while the Y-axis represents information content, indicating amino-acid conservation within the sequence. (c) Contact map and potential co-evolved residues within Bap1-57aa. The X-axis and Y-axis represent amino acid positions. Black points represent potential co-evolved residues identified using EVcouplings. Red boxes highlight residues two positions upstream and three positions downstream of W440FFGW444, which are predicted to co-evolve.
Fig 8.
Proposed model for conformational changes and sequence-specific interactions of Bap1-57aa with lipid bilayers.
(a) Conformational ensemble in solution, featuring transient parallel or antiparallel β-sheets. The chain is in a relatively collapsed state due to intramolecular interactions. (b) Potential pathway for tight membrane binding of Bap1-57aa. After the initial contact through the peripheral repeats, the central linker moves toward the membrane hydrophobic core. This positioning brings the downstream and upstream residues together to form a β-hairpin. β-hairpin formation in turn leads to deeper insertion of the central linker and tighter membrane binding for the entire chain.