High force catch bond mechanism of bacterial adhesion in the human gut

Bacterial colonization of the human intestine requires firm adhesion of bacteria to insoluble substrates under hydrodynamic flow. Here we report the molecular mechanism behind an ultrastable protein complex responsible for resisting shear forces and adhering bacteria to cellulose fibers in the human gut. Using single-molecule force spectroscopy (SMFS), single-molecule FRET (smFRET), and molecular dynamics (MD) simulations, we resolve two binding modes and three unbinding reaction pathways of a mechanically ultrastable R. champanellensis (Rc) Dockerin:Cohesin (Doc:Coh) complex. The complex assembles in two discrete binding modes with significantly different mechanical properties, with one breaking at ~500 pN and the other at ~200 pN at loading rates from 1-100 nN s−1. A neighboring X-module domain allosterically regulates the binding interaction and inhibits one of the low-force pathways at high loading rates, giving rise to a catch bonding mechanism that manifests under force ramp protocols. Multi-state Monte Carlo simulations show strong agreement with experimental results, validating the proposed kinetic scheme. These results explain mechanistically how gut microbes regulate cell adhesion strength at high shear stress through intricate molecular mechanisms including dual-binding modes, mechanical allostery and catch bonds.

. Amino acid sequence and secondary structure elements of dockerin. The residues involved in calcium binding are shown in yellow, the residues mutated to knock out the binding mode A are shown in red and the residues mutated to knock out the binding mode B are shown in grey.

Supplementary Figure 2. Model of binding interface between dockerin and cohesin in both binding modes.
In both binding modes, the dockerin and cohesin residues at the binding interface have complementary physical properties, forming a hydrophobic core surrounded by hydrophilic residues. Magenta: negatively charged residues; blue: positive charged residues; white: non-polar residues and green: polar residues.

Supplementary Figure 3. Example force-extension curves obtained in AFM-SMFS measurements.
Examples of pathways 1 (left column, red), 2 (middle column, blue) and 3 (right column, grey) force-extension curves obtained in AFM-SMFS measurements of WT XMod-Doc:Coh complex. Some force curves showed unassigned unfolding events between ddFLN4 unfolding and complex rupture or XMod unfolding. These unfolding events broadened the contour length histogram and can be attributed to partial unfolding of Coh or Doc domains.  Table 1. showing unfolding of 2x ddFLN4 (in orange) and I27 (in blue) in all three pathways. c: Rupture force histogram of force curves filtered with both ddFLN4 and I27 fingerprint domains shows that complexes capable of unfolding I27 rarely (3%) dissociated along pathway 3. d: Rupture force histogram of force curves filtered with only ddFLN4 showed that pathway 3 was prevalent in the dataset to the same degree as for WT (~18%), but these curves lacked I27 unfolding events.

Supplementary Figure 12. AFM measurement of AF647 labeled BM A -KO.
Given the significant molecular weight of the FRET acceptor dye DBCO-AF647 (~1100 grams/mol) and its proximity to the binding interface in binding mode A, we sought to further understand the influence of dye labeling on binding. We used AFM-SMFS to measure the rupture forces between unlabeled wild-type Coh and BM A -KO labeled with AF647 at 400 nm/s pulling speed. Binding mode A (P1 + P2) consists of 61% of the force curves, which is lower than the unlabeled BM A -KO mutant (69%), indicating a decrease in the on-rate of the complex in binding mode A. In addition, the most probable rupture force of P1 was found to be 434 pN for the fluorophore labeled BM A -KO complex, which is significantly lower than the P1 rupture force measured using unlabeled BM A -KO ( Supplementary Fig.10a, 508 pN). The decrease of rupture force in binding mode A for the dye-labeled construct indicated an increase in the intrinsic off-rate of binding mode A as compared with the unlabeled construct. Therefore, we concluded that the AF647 fluorophore at the Cterminus of XMod-Doc slightly destabilized the complex in binding mode A and decreased the binding affinity in this binding mode. However, fluorophore labeling did not have a significant influence on the rupture force of P3, which corresponds to binding mode B. As a consequence, the binding mode A population was less prevalent than binding mode B in the smFRET measurement. The observed ratios of the two binding modes probed by smFRET reflect the equilibrium scenario, whereas binding mode ratios probed by AFM are governed by differences in on-rates only.   The rupture force-loading rate plots in Supplementary Fig. 5 were fitted using the following equation 2 :

Supplementary
where k0 is the intrinsic off rate in the absence of force, Δx ‡ is the distance to the energy barrier, ΔG ‡ is the height of the energy barrier in the absence of force, β -1 =kBT, and υ = 0.5, assuming the shape of the free-energy surface is cusp.

Supplementary Note 3. Quantifying dual-binding mode behavior using fingerprint domain biasing effect
As shown in Supplementary Fig. 11a, a titin I27 domain, which under our conditions has an unfolding force around ~200 pN 3,4 , was inserted between ddFLN4 and WT XMod-Doc as an additional fingerprint domain. The interaction between XMod-Doc and Coh was then probed using AFM-SMFS in the presence of two ddFLN4 and one I27 fingerprint domain. Unfolding of I27 was identified by its unfolding force of ~200 pN and the contour length increment of ~28 nm, as shown in Supplementary Fig. 11b. Force-extension curves were screened based on the contour length increments given by two ddFLN4 domains and one I27 domain and then sorted into the aforementioned three pathways based on the rupture force of the complex and the folding state of XMod. The rupture force of each pathway was plotted in a rupture force histogram ( Supplementary  Fig. 11c). In binding mode A, the complex was able to resist an external force up to ~500 pN prior to rupture, which was larger than the force required to unfold I27. Therefore, pathways 1 and 2 were not biased by the additional I27 domain and were still observable in the dataset. However, binding mode B has relatively low mechanical stability and ruptures prior to I27 unfolding. Therefore, the frequency of pathway 3 was significantly decreased to only 3% when using I27 as an additional fingerprint domain for curve selection. However, when screening the force curves only based on the two ddFLN4 domains regardless of whether the curves contained I27 unfolding or not, the frequency of different pathways in the screened curves ( Supplementary Fig. 11d) was the same as the construct lacking I27 (Supplementary Fig. 6b). This observation further demonstrated that pathways 1 and 2 (high force curves) belong to a different discrete binding mode than pathway 3 (low force curves). The two binding modes have different mechanical stabilities and cannot be converted to the other one on the timescale of the AFM-SMFS curve (~1 second).