Mechanical forces and ligand binding modulate Pseudomonas aeruginosa PilY1 mechanosensitive protein

Francisco J Cao-Garcia; Jane E Walker; Stephanie Board; Alvaro Alonso-Caballero

doi:10.26508/lsa.202403111

Research Article

Published 7 March 2025. DOI: 10.26508/lsa.202403111

Abstract

Surface sensing initiates bacterial colonization of substrates. The protein PilY1 plays key roles during this process—surface detection, host adhesion, and motility—while experiencing mechanical perturbations of varying magnitudes. In Pseudomonas aeruginosa, the adhesion and motility functions of PilY1 are associated with integrin and calcium ligand-binding sites; however, how mechanical forces influence PilY1’s dynamics and its interactions with these ligands remain unknown. Here, using single-molecule magnetic tweezers, we reveal that PilY1 is a mechanosensor protein that exhibits different behaviors depending on the force load. At high forces (>20 pN), PilY1 unfolds through a hierarchical sequence of intermediates, whose mechanical stability increases with calcium binding. This enhanced stability may help counteract type IV pilus retraction forces during motility. At low forces (<7 pN), we identify the dynamics of the integrin-binding domain, which is reminiscent of the behavior of mechanosensor proteins. Integrin binding induces a force-dependent conformational change in this domain, shortening its unfolded extension. Our findings suggest that PilY1 roles are force- and ligand-modulated, which could entail a mechanical-based compartmentalization of its functions.

Introduction

Mechanical forces play key roles in cell fate, triggering crucial events such as differentiation, proliferation, or motility (Eyckmans et al, 2011). Mechanical sensing and the responses elicited by force are well-known phenomena in eukaryotic cells; however, our understanding of how prokaryotes sense and respond to mechanical cues is still scarce (Gordon & Wang, 2019).

Although explored to a lower extent, it is recognized that mechanical stimuli promote critical life changes in prokaryotes (Persat et al, 2015; Chawla et al, 2020; Dufrêne & Persat, 2020). Concretely, the mechanical cues induced by surface proximity stimulate the switch from a planktonic to a sessile lifestyle (Persat, 2017). In pathogenic bacteria like Pseudomonas aeruginosa, surface detection is followed by colonization, biofilm formation, and virulence development toward a host (Mordue et al, 2021). It has been proposed that the protein PilY1 may act as a mechanosensor mediating surface detection (Rodesney et al, 2017). This protein is located at the tip of the type IV pili (Treuner-Lange et al, 2020) (T4P), the motor-driven appendages that power bacterial translocation on substrates (O’Toole & Kolter, 1998; Persat et al, 2015; Craig et al, 2019). In addition to surface sensing, PilY1 is involved in T4P biogenesis and participates in several of its functions (Nguyen et al, 2015), such as in twitching motility (Orans et al, 2010), adhesion (Hoppe et al, 2017), biofilm formation, and virulence (Siryaporn et al, 2014; Marko et al, 2018). Therefore, PilY1 stands out as a multipurpose protein with a pivotal role in numerous processes that govern the surface-associated life of bacteria (Kuchma et al, 2010).

The different roles of PilY1 have been pinned down to specific features found in its sequence, which has enabled the compartmentalization of the protein into two functional sections (Fig 1A). On the N terminus, the von Willebrand A (vWA) domain has been assigned the mechanosensing role of PilY1. This function was inferred from the phenotype manifested after vWA domain deletion, which placed planktonic bacteria in a constitutive virulence state, a behavior typically observed after surface engagement (Siryaporn et al, 2014). The absence of the vWA domain mimicked the signaling triggered after PilY1 interaction with a surface, suggesting a role in mechanotransduction. More recently, it was found that the vWA domain could also have an adhesive function linked to conserved cysteine residues of its sequence. Mutation of these residues reduced the adhesion strength of bacteria and dwindled the responses characteristic of surface-committed cells (Webster et al, 2021). Although genetic, surface mechanics (Siryaporn et al, 2014; Wang et al, 2023), and biophysical (Webster et al, 2021) studies support the vWA domain’s involvement in mechanosensing, less is known about how force could regulate the functions of the C-terminal section of PilY1.

Figure 1. PilY1^501−1161 mechanical hierarchy and Ca²⁺ dual effect.

(A) Scheme of the PilY1 protein sequence and structural features (SP: signal peptide; vWA domain: von Willebrand A domain; IBD: integrin-binding domain; RGD: integrin-binding site; CBS1: Ca²⁺-binding site 1; blades I to VII of the β-propeller; CBS2: Ca²⁺-binding site 2). The sequence employed spans from residues 501–1,161. (B) Crystal structure of PilY1^501−1161 from top (left) and side (right) views (prediction from AlphaFold [Jumper et al, 2021; Varadi et al, 2022] and color assignation based on reference [Orans et al, 2010]). (C) Magnetic tweezer force-ramp trajectory of PilY1^501−1161 (inset i; see the main text and methods for a full description of the chimeric protein construct and tethering) under a 1-pN·s⁻¹ loading rate. PilY1^501−1161 unfolds through multiple intermediates, which are classified based on their order of appearance: I1 (1 event), I2 (1 or 2 consecutive events, inset ii), variable intermediates (two to seven events of different sizes, inset iii), and last intermediate (1 event). (D) Unfolding force distributions (top, mean ± SD) and force dependency of unfolding extensions (bottom) per intermediate group and condition (Ca²⁺ or EGTA). The extension (ΔL) versus unfolding force dependency is described with the freely jointed chain model (FJC) for polymer elasticity (lines in the bottom panels are fits of this model). The Kuhn length values, l_K, in EGTA for I1, I2, and the last intermediate are 1.5 ± 0.1 nm, 1.0 ± 0.1 nm, and 0.7 ± 0.2 nm, and in Ca2+ are 1.7 ± 0.1 nm, 1.1 ± 0.1 nm, and 0.9 ± 0.2 nm.

Unlike the vWA domain, the sequence of the C-terminal part of PilY1 is well conserved across species, and its structure has been experimentally determined (Orans et al, 2010). In P. aeruginosa strains, the sequence spanning residues ∼640–1,150 adopts a 7-bladed β-propeller configuration. This type of fold comprises several structural repeats, termed blades, arranged around a central axis (Orans et al, 2010; Mylemans et al, 2021) (Fig 1A and B). Blade IV contains an EF-hand Ca²⁺-binding motif that modulates the extension and retraction cycles of the T4P. In the metal-bound state, PilY1 inhibits the retraction activity of the T4P machinery, which can exert pulling forces ranging 30–100 pN (Maier et al, 2002; Ribbe et al, 2017). After Ca²⁺ release, pilus retraction is restored, indicating that the PilY1 C-terminal domain acts as a Ca²⁺-dependent switch that modulates T4P dynamics in biogenesis and twitching motility (Orans et al, 2010). Further scrutiny of the PilY1 sequence identified an adhesin function between the vWA domain and β-propeller. PilY1 proteins from P. aeruginosa strains harbor an RGD motif that binds to the extracellular domains of integrin. A second Ca²⁺-binding site was found in the vicinity of the RGD motif, and it was demonstrated that integrin binding required both the RGD motif and Ca²⁺ binding on this site (Johnson et al, 2011). Integrins are transmembrane proteins present in the respiratory epithelium, a common niche of P. aeruginosa pathogenic strains (Heiniger et al, 2010; Jurado-Martín et al, 2021).

PilY1 plays a key role in surface-related processes and is exposed to mechanical perturbations. Understanding the force response and ligand-binding modulation of PilY1 could shed light on the molecular mechanisms underlying surface colonization. Herein, we have employed a single-molecule approach to address the nanomechanics of the PilY1 C-terminal section, which holds the ligand-binding sites responsible for T4P dynamics and host cell adhesion. Our results indicate that PilY1 is a mechanosensor protein that exhibits different behaviors depending on the force load. In the high-force regime, PilY1 unfolds through multiple intermediates in a complex yet hierarchical sequence affected by Ca²⁺. When this metal is bound to its furthermost C-terminal site, which regulates T4P dynamics, PilY1 stability increases and becomes less sensitive to mechanical perturbations. In the low-force regime, the integrin-binding domain undergoes folding and unfolding transitions in a 2 pN range, showing a steep force dependency characteristic of mechanosensitive proteins. Experiments with integrin enabled us to detect single binding events, which partially blocked the unfolding of its cognate domain and induced a shortening of the protein. We conclude that the PilY1 C-terminal section exhibits a dual force dependency where its functions as an adhesin and T4P regulator could be invoked depending on the mechanical load experienced by bacteria during substrate colonization. Our work establishes a mechanical-based connection between the structure and functions of PilY1 and suggests that this bacterial protein has a force-sensing activity that extends beyond its N-terminal domain. These findings underpin the intricate and functional wealth of PilY1.

Results

Hierarchical unfolding and Ca²⁺-dependent mechanical stability

To explore the nanomechanics of PilY1, we employed single-molecule magnetic tweezers (Gosse & Croquette, 2002). This technique allows the application of well-calibrated forces to single proteins and monitoring their conformational changes with high spatial, temporal, and force resolutions (Popa et al, 2016; Zhao et al, 2017). To conduct single protein measurements, we designed a chimeric polyprotein with features that enable its end-to-end specific tethering between a functionalized glass cover slide and a superparamagnetic bead, the force probe. Our protein of interest is the C-terminal section of PilY1 encompassing residues 501–1,161 from the P. aeruginosa PAO1 strain (PilY1^501−1161) (Fig 1A and B). In the chimeric construct (Fig 1C, inset i), the N terminus presents a HaloTag protein for covalent anchoring on the glass surface (Taniguchi & Kawakami, 2010), and the PilY1^501−1161 sequence is flanked on both sides by two copies of the human titin I32 domain, which provide spacing between the surface and the bead. HaloTag and I32 domains exhibit high mechanical stability and act as rigid molecular handles (Li et al, 2002; Popa et al, 2013). At the forces tested in this study, their probability of unfolding is low, and they do not exhibit conformational changes during the experiments (see the Materials and Methods section). The C terminus of the polyprotein contains a biotinylated AviTag sequence that allows the tethering to a streptavidin-coated superparamagnetic bead.

Our first goal was to investigate the mechanical stability and fingerprint of PilY1^501−1161 under a steadily increasing force load. We also aimed to assess whether Ca²⁺ impacts its stability, as reported in protein mechanics (Echelman et al, 2016; Scholl et al, 2016; Milles et al, 2018). For this reason, we conducted experiments with Ca²⁺ or the calcium chelator EGTA. Fig 1C shows an unfolding trajectory of this construct, which was exposed to a loading rate of 1 pN∙s⁻¹ from its folded state up to its complete unfolding. As the load increases, the protein is stretched, and discrete jumps in the extension can be observed corresponding to the unfolding of PilY1^501−1161 through multiple intermediates. These unfolding events are heterogeneous in extension and number; nevertheless, we noticed a conserved sequence. We assigned a number to each intermediate based on their order of appearance. In this trajectory, the least stable intermediate, I1, unfolds in a single 14-nm event at 13 pN. At 15 pN, we detect the unfolding of I2, which usually occurs as a single event, but, as shown in Fig 1C, inset ii, it can proceed through two sub-intermediates. After I1 and I2, a population of unfolding events of heterogeneous quantity and size appears (Fig 1C, inset iii). The number of intermediates oscillates inter- and intramolecularly (between unfolding pulses) and ranges from two large events to up to seven smaller extensions. Because of the difficulty of identifying reproducible conformations, we pooled them into a single group termed variable intermediates. The final ∼40 nm event in this trajectory, occurring at 22 pN, was named the last intermediate. Therefore, the unfolding pattern of PilY1^501−1161 displays a clear fingerprint of four classes of events occurring hierarchically, with a conserved sequence of I1 (1 event), I2 (1–2 events), variable intermediates (2–7 events), and last intermediate (1 event).

In Fig 1D, we show the unfolding force distribution (top) and the force-dependent step size (bottom) of each of the classes of intermediates detected in the presence of Ca²⁺ or EGTA. By collecting data from multiple trajectories and molecules, we obtain the scatter plots shown in Fig 1D, which depict single unfolding events of each class occurring over a distribution of forces. The step size (∆L) force dependency of each class can be described by polymer models such as the freely jointed chain (FJC) (Flory, 1956). From the fits of the data to this model, we can obtain the extension of the unfolded intermediates (∆L_c). In the case of I1, I2, and the last intermediate classes, the data points are well described by a single fit in both Ca²⁺ and EGTA, indicating that these events can be attributed to unique structural features in PilY1^501−1161. In ∼30% of the trajectories, the I2 intermediate proceeds through two consecutive sub-intermediates (Fig 1C, inset ii, and Fig S1A), which can be described with two separate fits whose combined ∆L_c matches the ∆L_c of I2 when detected as a single event (Fig S1B). In the variable intermediate class, the unfolding events cannot be described with an FJC fit because of its heterogeneity. Nevertheless, insights can be obtained from the unfolding force means of each group and the classification standard employed (Fig 1D, top graphs). Ca²⁺ decreases the average unfolding force of both I1 and I2 (11.8 ± 0.5 pN to 9.4 ± 0.3 pN, for I1; 17.9 ± 0.7 pN to 15.7 ± 0.6 pN, for I2; mean ± SEM; see the Materials and Methods section and Table S1), also observed when I2 occurs in two events (Fig S1B). In contrast, Ca²⁺ increases the collective stability of the variable intermediates (27.0 ± 0.2 pN to 29.4 ± 0.2 pN) and the last intermediate (27.8 ± 0.4 pN to 30.8 ± 0.4 pN). Ca²⁺ has opposite effects on the PilY1^501−1161 structure, lowering the mechanical stability of I1 and I2 while increasing that of the variable intermediates and last intermediate. Furthermore, and independent of the buffer condition, there is a gap in the average unfolding forces between I1 and I2 (<25 pN), and the variable intermediates and the last intermediate (>25 pN), which suggests the existence of two mechanostability ranges across the PilY1^501−1161 structure.

Figure S1. Mechanical characterization of PilY1^501−1161 in force ramp (1 pN·s⁻¹) in the presence of Ca²⁺ or EGTA.

(A) Force-ramp trajectory of PilY1^501−1161 highlighting the unfolding of I2 through two sub-intermediates and the variable intermediates, which are numbered and colored based on their order of appearance in the unfolding sequence. (B) On the left, force-dependent extensions of I2 and its two sub-intermediates and their calculated ∆L_c and l_K from fits of the data with the FJC model in EGTA and Ca²⁺ (top and bottom, respectively). On the right, unfolding force distributions (mean ± SD) of I2 and its sub-intermediates in EGTA and Ca²⁺ (top and bottom, respectively). (C) On the left, force-dependent extensions of the variable intermediates (colored as in (A)) in EGTA and Ca²⁺ (top and bottom, respectively). On the right, unfolding force distributions (mean ± SD) of the variable intermediates in EGTA and Ca²⁺ (top and bottom, respectively). I4 and subsequent intermediates show an average unfolding force shifted to higher values with Ca²⁺.

Table S1. Comparison of unfolding force means of PilY1 intermediates in EGTA and Ca²⁺. The significance of the difference is computed in terms of the difference (of the means) and its uncertainty (which is computed from the SEMs). We state that the difference is significant when the probability of being zero is smaller than 5% (*P < 0.05) (**P < 0.01, or X if P > 0.05 and not significant). See the Materials and Methods section.

PilY1^501−1161 contains two Ca²⁺-binding sites, but from these observations, we cannot locate which group of intermediates could have these motifs. We sought to explore the variable intermediate group further, attending exclusively to their unfolding order sequence for their classification of mechanical stabilities. Fig S1C (color coding of the intermediates is unrelated to the colors chosen for Fig 1A and B) and Table S1 show the comparison of the unfolding force means and summarize the results (see the Materials and Methods section and Supplementary Information). In both EGTA and Ca²⁺, the first variable intermediate, I3, shows no differences in its average unfolding force (26.5 ± 0.4 pN to 25.8 ± 0.4 pN). In contrast, Ca²⁺ increases the unfolding forces of the following intermediates, from the second variable intermediate, I4, to the last intermediate. Comparisons inside each buffer condition reveal that in EGTA (Fig S1C, Tables S1 and S2), there is no significant change in the average unfolding forces among the variable intermediates, from I3 to I7 (26.5 ± 0.4 pN to 27.5 ± 1.0 pN), suggesting that these structures have similar stability. With Ca²⁺ (Fig S1C, Tables S1 and S3), there is a significant ∼5 pN increase in the average mechanical stability between I3 and I4 (25.8 ± 0.4 pN to 30.4 ± 0.4 pN). After I4 and until the last intermediate, the means show no significant differences, indicating similar stability. These results suggest that the variable intermediate I4 is a plausible candidate to harbor one of the Ca²⁺-binding sites and that the mechanical stabilization shift observed in subsequent intermediates could be due to Ca²⁺ binding on this intermediate. Ca²⁺ specifically induces this effect on the mechanical stability of the protein, though other divalent cations can also alter PilY1’s nanomechanics (Fig S2A–C).

Table S2. Comparison of unfolding force means (values in Table S1) between intermediates in EGTA. The significance of the difference is computed in terms of the difference (of the means) and its uncertainty (which is computed from the SEMs). We state that the difference is significant when the probability of being zero is smaller than 5% (*P < 0.05) (**P < 0.01, or X if P > 0.05 and not significant). Cells in green highlight that between I3 and subsequent variable intermediates, there is no significant difference in their unfolding force means (see Table S1, Fig S1C, and the Materials and Methods section).

Table S3. Comparison of unfolding force means (values in Table S1) between intermediates in Ca²⁺. The significance of the difference is computed in terms of the difference (of the means) and its uncertainty (which is computed from the SEMs). We state that the difference is significant when the probability of being zero is smaller than 5% (*P < 0.05) (**P < 0.01, or X if P > 0.05 and not significant). Cells in orange highlight that between I3 and subsequent intermediates, there is a significant difference in their unfolding force means. After, from I4 to the last intermediate, the means show no difference (see Table S1, Fig S1C, and the Materials and Methods section).

Figure S2. Cation-binding specificity in PilY1^501−1161.

(A) Force-ramp trajectories of a single PilY1^501−1161 molecule exposed to different buffer conditions, where the protein is repeatedly unfolded (1 pN/s) followed by force quenches to allow folding. Over time, the last intermediate’s unfolding force (mean ± SEM) is monitored in the presence of MgCl₂, CaCl₂, EGTA, MnCl₂, and CaCl₂ again. Three consecutive force-ramp trajectories are represented for each buffer condition, with breaks in time continuity indicated by symbols placed between the extension and force axes. The unfolding force under each condition indicates that calcium is the cation that significantly enhances the protein’s mechanical stability. After each buffer exchange, the offset extension changes because of the drift in the measurement (the dashed line is positioned at 0 nm as a reference). (B) Violin plot of unfolding force data for each buffer condition. (C) Computed difference in the means of the unfolding force and their uncertainties between buffer conditions indicate that the only conditions showing significant differences (*P < 0.05) are both treatments with calcium compared with EGTA, both before and after the second addition of calcium.

Ca²⁺ binding alters the unfolding pathway

The force-ramp experiments aided us in establishing a benchmark for further testing the nanomechanics of PilY1^501−1161. We defined a clear mechanical signature of the protein, which consists of a hierarchical unfolding sequence of intermediates whose stabilities are affected by Ca²⁺. We next explored PilY1^501−1161 unfolding dynamics under constant force. In these measurements, we can obtain a clearer picture of the force-dependent extensions of each of the intermediates, narrowing the variability observed in the force-ramp experiments. Because forces above 20 pN trigger the complete stretching of PilY1^501−1161, we explored its unfolding dynamics from 20 to 40 pN to capture all the intermediates.

Fig 2A shows a trajectory of PilY1^501−1161 subjected to unfolding and refolding cycles at constant forces. Changing the force from 4 to 20 pN drives the protein from its folded state to its complete unfolding through multiple intermediates. After, the force is quenched to 4 pN to allow protein folding, which reaches the same folded extension. A subsequent 20-pN pulse reproduces the same pattern of unfolding intermediates. We identified the intermediate groups previously categorized (Figs S3A and S4A). In EGTA or Ca²⁺, the average number of variable intermediates is ∼4. The number of total intermediates oscillates around seven to eight, with a slightly higher count when Ca²⁺ is present (Fig 2B). The proportion of I2 unfolding events occurring in two steps doubled in the presence of Ca²⁺ (Fig 2A, inset iii), explaining this increase. Fitting the FJC model to the average extensions of I1, I2, and the last intermediate yields similar results to those obtained in the force ramp (Figs S3D and S4D). However, the variable intermediates’ heterogeneity persists, as seen in Fig 2A, insets i and ii. For instance, I3 appears as a ∼18-nm extension in the first unfolding pulse, whereas the second shows a ∼6-nm step. This change in pattern affects all the variable intermediates except for I6, which shows a similar extension. This variability was detected intra- and intermolecularly and cannot be explained by the reshuffling of the unfolding order of the intermediates between pulses, as their sizes differ. To rule out that the variable intermediates were not properly folding between pulses, we measured in all trajectories the extension of all the unfolding events in PilY1^501−1161 and the extension of the unfolding events of the variable intermediates. Fig 2C shows FJC fits to the combined extension of all the unfolding events (**), which yields an ∆L_c ∼190 nm with both EGTA and Ca²⁺, and the variable intermediates (*), which collectively produce an ∆L_c ∼70 nm in both conditions. These results indicate that despite the different number of events and extensions, the amount of polypeptide sequence released after unfolding is the same in all trajectories; therefore, these variable intermediates achieve a compact folded state.

Figure 2. Ca²⁺-induced conformational heterogeneity and stabilization of PilY1^501−1161.

(A) Magnetic tweezer trajectory of PilY1^501−1161 exposed to unfolding and refolding cycles at different constant forces. From 4 pN, the force is jumped to 20 pN, and the sequential unfolding of PilY1^501−1161 through multiple intermediates is detected (I1→ I2→ variable intermediates → last intermediate). The force is then quenched to 4 pN to allow protein folding, and later, the 20-pN unfolding pulse is repeated, obtaining the same populations of intermediates. Insets i and ii highlight the conformational heterogeneity of the variable intermediates between consecutive unfolding pulses. In Ca²⁺, the proportion of I2 domains unfolding in two consecutive sub-intermediates doubled compared with EGTA (inset iii). Vertical arrows show the total length of the unfolding intermediates (**) and the variable intermediates (*). (B) Distribution of the total number of unfolding intermediates in PilY1^501−1161 (top, **) and variable intermediates (bottom, *) observed across trajectories in Ca²⁺ or EGTA. (C) Force-dependent extension change of the combined extension of all PilY1^501−1161 intermediates (**) and variable intermediates (*). Lines are fits of the FJC model to the average extension (mean ± SD), and the resulting values of contour length increment (∆L_c) are shown above the graph for each condition. (D) Cluster analysis results of the extension of the variable intermediates per order of unfolding and force in EGTA (left) or Ca²⁺ (right). On the top is plotted the average extension of the events classified inside each cluster identified for each intermediate at 40 pN. The size of the data points is proportional to the number of events in that cluster (most represented ones are colored). Below are shown the force-dependent extension changes (mean ± SD) of the most populated conformations for each intermediate in EGTA (left) and Ca²⁺ (right). Solid and dashed lines are fits of the FJC model to the most populated and second most populated conformations, respectively (see Figs S3C and S4C). The force axis is plotted on a logarithmic scale. (E) Unfolding kinetics under force of the variable intermediates in EGTA and Ca²⁺. Each intermediate’s unfolding kinetics (mean ± SEM) is fitted using Bell’s model for bond lifetimes. Above each plot is shown the value of the parameters determined: unfolding force at zero force (k_U⁰) and distance to the transition state (∆x_U).

Figure S3. Mechanical characterization of PilY1^501−1161 in constant force with EGTA and identification of the most prevalent conformations of the variable intermediates.

(A) Unfolding under constant force of PilY1^501−1161 in the presence of EGTA and classification of the intermediates by unfolding order and color. The category plots on the right show the stacked extensions of each intermediate per trajectory across the five unfolding forces tested (20, 25, 30, 35, and 40 pN). (B) Cluster analysis of the extension of the variable intermediates per order of unfolding and force in EGTA. Each plot shows the average extension of the events classified inside each cluster and for each intermediate across the five tested forces. The size of the data points is proportional to the number of events in that cluster (most represented ones are colored). (C) Force-dependent extensions (mean ± SD) of the most populated conformations for each intermediate (I3 to I7). The size of the data points is proportional to the number of events in that cluster. When the extension of the clusters identified in (B) is represented for each intermediate across the five unfolding forces tested, it is possible to identify the different conformations of each intermediate and model the data with individual fits of the FJC model (∆L_c data for each population shown above). (D) Force-dependent extensions (mean ± SD) of the I1, I2 (as a single and as two events), and last intermediates and their fitting with the FJC model.

Figure S4. Mechanical characterization of PilY1^501−1161 in constant force with Ca²⁺ and identification of the most prevalent conformations of the variable intermediates.

(A) Unfolding under constant force of PilY1^501−1161 in the presence of Ca²⁺ and classification of the intermediates by unfolding order and color. The category plots show the stacked extensions of each intermediate per trajectory across the five unfolding forces tested (20, 25, 30, 35, and 40 pN). (B) Cluster analysis of the extension of the variable intermediates per order of unfolding and force in Ca²⁺. Each plot shows the average extension of the events classified inside each cluster and for each intermediate across the five tested forces. The size of the data points is proportional to the number of events in that cluster (most represented ones are colored). (C) Force-dependent extensions (mean ± SD) of the most populated conformations for each intermediate (I3 to I7). The size of the data points is proportional to the number of events in that cluster. When the extension of the clusters identified in (B) is represented for each intermediate across the five unfolding forces tested, it is possible to identify the different conformations of each intermediate and model the data with individual fits of the FJC model (∆L_c data for each population shown above). (D) Force-dependent extensions (mean ± SD) of the I1, I2 (as a single and as two events), and last intermediates and their fitting with the FJC model.

Although the variable intermediates’ heterogeneity was prominent, a sequence of events repeated more frequently. To explore the existence of a canonical unfolding pathway, we resorted to clustering analysis, which enabled us to classify the variable intermediates based on their extension. This helped us to sort conformations, identify the most common ones, and resolve structural identities (Schönfelder et al, 2016) (top graphs in Figs 2D, S3B, and S4B). Fig 2D bottom plots show FJC fits to the average extension of the most common intermediates in EGTA and Ca²⁺ (Figs S3C and S4C). In EGTA, most events can be assigned to unique conformations and are well described by individual fits. The I6 intermediate deviates from this trend with two alternative conformations, although one predominates (I6A). In contrast, the behavior with Ca²⁺ differs. Beyond the I3 intermediate, the upcoming intermediates show alternative conformations that occur in a similar frequency. For instance, I4 has a predominant configuration (I4A ∆L_c ∼12 nm) that differs from the I4 population seen in EGTA (∆L_c ∼21 nm). In I5, the most represented conformation also differs (I5B ∆L_c ∼23 nm versus I5A ∆L_c ∼10 nm). For I6, the behavior differs as well in both conditions. These results suggest that Ca²⁺ alters the unfolding pathway of PilY1^501−1161, changing the conformations of the intermediates after the variable intermediate I3.

We questioned next whether Ca²⁺ alters PilY1^501−1161 unfolding kinetics. In Fig 2E, we show the force-dependent unfolding kinetics for each intermediate. We used Bell’s approximation for bond lifetimes under force to model their unfolding kinetics (Bell, 1978). From fits of the data to this model, we can obtain an extrapolation to zero force of the unfolding rates of each intermediate (k_U⁰), and the distance to the transition state (∆x_U). The comparison between conditions indicates that Ca²⁺ stabilizes intermediates I4 and I5, slowing their unfolding rates ∼2.7 and ∼2.9 times, respectively. Ca²⁺ has the opposite effect on I3, accelerating its unfolding ∼2.5 times, whereas I6 remains unchanged in both conditions. There is a change in the ∆x_U of I3 and I5, which decreases in comparison with the EGTA condition. This indicates that Ca²⁺ stabilizes PilY1^501−1161, which confirms the observations made in force-ramp experiments.

PilY1^501−1161 harbors a mechanosensing-like structure

The previous sections described PilY1^501−1161 nanomechanics at force regimes that promote its complete unfolding. The sequential unfolding of PilY1^501−1161 suggests that I1 and I2 are less stable than the other intermediates, which divides the protein into two ranges of mechanostability (Fig S5). Therefore, we conducted constant force measurements at loads below 10 pN, to test whether PilY1^501−1161 exhibits dynamics in this force regime.

Figure S5. PilY1^501−1161 shows two ranges of mechanostability.

At intermediate forces (7–20 pN range), the unfolding of I1 and I2 occurs in shorter times, as expected, with an increasing applied force. Subsequent intermediates (I3 to last intermediate) exhibit higher mechanical stability than I1 and I2, and the timescales to observe their unfolding below 20 pN require long waiting times. The trajectory depicts how, at an intermediate force (12 pN), the unfolding of I1 and I2 occurs rapidly, whereas the following intermediate (I3) takes around 300 s. This behavior highlights the presence of two ranges of mechanostability affecting I1 and I2 on the one side and the variable and last intermediates on the other.

Fig 3A shows a trajectory of PilY1^501−1161 exposed to high and low mechanical loads. After unfolding at high force, the load is quenched to 5.5 pN, which promotes the collapse of the extended polypeptide and enables folding. At this force, PilY1^501−1161 dwells over time across three levels of extension, which correspond to structure folding and unfolding. The lowest extension level corresponds to the folding of I1, hence the complete folding of PilY1^501−1161 (Fig 3A, inset i). The second and third levels correspond to I1 unfolded extension (and I2 folded extension) and I2 unfolded level, respectively. In Fig 3A, inset ii, we show long dynamics of PilY1^501−1161 at forces ranging from 4.5 to 6.0 pN, from where it can be seen how the increase in the mechanical load tilts the equilibrium toward the extended (unfolded state) levels of I1 and I2. Measuring at this force range allowed us to determine the force-dependent step size of I1 and I2 below 20 pN (Figs S3D and S4D), which enabled us to assign their identity to these dynamics. From these long equilibrium dynamics (Fig 3A, inset ii), we determined the folding probability as a function of the force for I1 and I2 with EGTA or Ca²⁺ (Fig 3B). As the mechanical load rises, the folded fraction of both intermediates decreases, adopting a sigmoidal shape. The force value at which the intermediates spend the same amount of time in the folded and unfolded state is defined as P₅₀. Ca²⁺ shifts P₅₀ to lower values for I1 (4.7 to 4.4 pN) but increases it for I2 (5.7 to 5.9 pN). Moreover, Ca²⁺ changes the force dependency of I2 folding probability, which is broadened and spans 3 pN (from fully folded at 4 pN to unfolded at 7 pN), 0.5 pN more than with EGTA (4.0–6.5 pN).

Figure 3. Conformational dynamics of PilY1^501−1161 at low force reveal a mechanosensing-like behavior.

(A) Magnetic tweezer trajectory of PilY1^501−1161 exposed to an unfolding pulse at 30 pN followed by a long 5.5-pN pulse. As shown before, the high-force pulse induces the unfolding of PilY1^501−1161 through multiple intermediates. At low force, most of the protein refolds; however, part of the protein dwells across three levels of extensions (inset i) corresponding to the folded extension of I1 (lowest), the unfolded I1/folded I2 extension (middle), and unfolded extension of I2 (highest). The unfolding hierarchy previously observed at high forces is maintained in this force regime for I1 and I2, which also applies to their folding. Inset ii shows ∼2,000-s-long trajectories of the dynamics of I1/I2 under different forces. The increase in the mechanical load promotes the occupation of the extended levels. (B) Folding probability of I1 and I2 between 4 and 6.5 pN in EGTA and Ca²⁺. The folded fraction of both intermediates sharply decreases with force, following a sigmoidal trend (line fits). The coexistence force (P₅₀) of I1 shifts to lower values, whereas that of I2 increases in the presence of Ca²⁺. Ca²⁺ also broadens the force range at which I2 exhibits folding and unfolding dynamics (∼2.5 pN in EGTA and ∼3 pN in Ca²⁺). (C) Folding and unfolding kinetics under force of I1 and I2 in EGTA and Ca²⁺. The logarithm of the rate (mean ± SEM) is plotted against the force in logarithmic scaling. I1 and I2 unfolding and folding kinetics are described with Bell’s model (fit lines) in both conditions (I1, empty circles and dashed lines for folding: EGTA k_F⁰ = [7.1 ± 0.7] ×10 s⁻¹, ∆x_F = − 4.83 ± 0.09 nm; Ca²⁺ k_F⁰ = [13.4 ± 2.5] ×10 s⁻¹, ∆x_F = − 5.62 ± 0.16 nm. Solid circles and lines for unfolding: EGTA k_U⁰ = [13.7 ± 0.4] ×10⁻² s⁻¹, ∆x_U = 0.33 ± 0.01 nm; Ca²⁺ k_U⁰ = [14.4 ± 0.4] ×10⁻² s⁻¹, ∆x_U = 0.34 ± 0.01 nm; I2 empty circles and dashed line for folding: EGTA k_F⁰ = [2.4 ± 0.6] ×10⁵ s⁻¹, ∆x_F = − 10.82 ± 0.18 nm; Ca²⁺ k_F⁰ = [2.3 ± 1.4] ×10³ s⁻¹, ∆x_F = − 6.85 ± 0.44 nm. Solid circles and lines for unfolding: EGTA k_U⁰ = [5.8 ± 0.2] ×10⁻² s⁻¹, ∆x_U = 0.23 ± 0.01 nm, Ca²⁺ k_U⁰ = [4.2 ± 0.2] ×10⁻² s⁻¹, ∆x_U = 0.24 ± 0.01 nm).

We also tested whether the kinetics of I1 and I2 were affected by Ca²⁺. Fig 3C shows both intermediates’ folding and unfolding kinetics in EGTA or Ca²⁺. Using Bell´s model, we observe that the folding and unfolding kinetics of both I1 and I2 reproduce the observations from the folding probability. From the extrapolations to zero force, in Ca²⁺ I1 unfolds and folds faster (∼1 time and ∼1.9 times) than in EGTA. I2 folding rate at zero force is slowed down in Ca²⁺ by ∼100 times, which comes from a steep force-dependent change (EGTA ∆x_F = − 10.82 ± 0.18 nm versus Ca²⁺ ∆x_F = − 6.85 ± 0.44 nm), whereas its unfolding rate is slowed down ∼1.4 times in comparison with the EGTA condition.

PilY1^501−1161 dynamics at low force reveal that the hierarchy is conserved for I1 and I2, and both exhibit a Ca²⁺-tuned mechanical sensitivity in a narrow range of forces, which showcases the prototypical behavior of mechanosensors. These results indicate that at least one of them contains one Ca²⁺-binding site, although both are affected by metal binding to this motif.

The integrin-binding domain is the mechanosensitive structure of PilY1^501−1161

After demonstrating the mechanosensing-like behavior of PilY1^501−1161, we wanted to link the observed intermediates with structural modules of this protein. The PilY1^501−1161 sequence contains the β-propeller fold and a stretch of 141 residues in the N terminus that contains the RGD and Ca²⁺-binding motifs responsible for integrin binding. Unlike the β-propeller, the structure of this section from PilY1 has not been resolved (Fig 1B shows a prediction from AlphaFold [Jumper et al, 2021; Varadi et al, 2022]).

To elucidate which of the intermediates of PilY1^501−1161 corresponds to this sequence containing the RGD and first Ca²⁺-binding motifs, we synthesized a second construct that only contains the β-propeller, named PilY1^642−1161 (Fig 4A). Fig 4B shows a trajectory of PilY1^642−1161 exposed to unfolding and refolding pulses. The unfolding trajectories show no sign of I1 and I2, but only the variable intermediate and last intermediate groups (Figs 4B, S6A–D, and S7A–D). The average number of total intermediates drops to ∼5, whereas the number of variable intermediates remains unchanged (Fig 4C), which indicates the loss of two domains. The complete unfolded extension of all intermediates (Fig 4D) is ∼60 nm shorter than PilY1^501−1161 (Fig 2C), but the variable intermediates’ complete extension is similar. After cluster analysis (Figs S6B and S7B), the variable intermediates reproduce the conformational structures observed in PilY1^501−1161 (Figs 4E and F, S6C, and S7C), and the effect that Ca²⁺ has in both the unfolding pathway and kinetics (Fig 4G, and Table S4). Therefore, the second Ca²⁺-binding site, located in blade IV of the β-propeller, is responsible for mechanically stabilizing the variable intermediates. Together with the results obtained in Fig 2, the intermediate I4 seems the most plausible candidate for being blade IV in the β-propeller.

Figure 4. Nanomechanics of PilY1^642−1161 indicates that the integrin-binding domain is responsible for the mechanosensor-like behavior observed at low force.

(A) Scheme of PilY1 protein and crystal structure (PDB:3HX6 [Orans et al, 2010]) of the sequence employed spanning residues 642–1,161. (B) Magnetic tweezer trajectory of PilY1^642−1161 exposed to unfolding and folding cycles. Removing the sequence 501–641 yields a mechanical fingerprint identical to PilY1^501−1161 but lacking the intermediates I1 and I2, where we only detect the variable intermediate and last intermediate groups. (C) Distribution of the total number of unfolding intermediates in PilY1^642−1161 (top, **) and variable intermediates (bottom, *) observed across trajectories in Ca²⁺ or EGTA. (D) Force-dependent change in the combined extension of all PilY1^642−1161 intermediates (**) and variable intermediates (*). Lines are fits of the FJC model to the average extension (mean ± SD), and the resulting values of contour length increment (∆L_c) are shown above the graph for each condition. (E, F) Force-dependent extensions (mean ± SD) of the most populated conformations determined with cluster analysis for each intermediate in EGTA (E) and Ca²⁺ (F). Solid and dashed lines are fits of the FJC model to the most populated and second most populated conformations, respectively (see Figs S6C and S7C). The force axis is plotted on a logarithmic scale. (G) Unfolding kinetics under force of the variable intermediates in EGTA and Ca²⁺. Each intermediate’s unfolding kinetics (mean ± SEM) is fitted using Bell’s model for bond lifetimes (see Table S4). (H) Based on its sequence length and predicted structure, the integrin-binding domain could be the I2 intermediate observed in PilY1^501−1161, which yields an ∆L_c ∼50 nm (plot on the left). In contrast, I1 would not be a good candidate to hold the integrin-binding site because of its smaller extension.

Figure S6. Mechanical characterization of PilY1^642−1161 in constant force with EGTA and identification of the most prevalent conformations of the variable intermediates.

(A) Unfolding under constant force of PilY1^642−1161 in the presence of EGTA and classification of the intermediates by unfolding order and color. The category plots on the right show the stacked extensions of each intermediate per trajectory across the five unfolding forces tested (20, 25, 30, 35, and 40 pN). This protein lacks the signature of the I1 and I2 intermediates observed in PilY1^501−1161. (B) Cluster analysis of the extension of the variable intermediates per order of unfolding and force in EGTA. Each plot shows the average extension of the events classified inside each cluster and for each intermediate across the five tested forces. The size of the data points is proportional to the number of events in that cluster (most represented ones are colored). (C) Force-dependent extensions (mean ± SD) of the most populated conformations for each intermediate (I3 to I7). The size of the data points is proportional to the number of events in that cluster. When the extension of the clusters identified in (B) is represented for each intermediate across the five unfolding forces tested, it is possible to identify the different conformations of each intermediate and model the data with individual fits of the FJC model (∆L_c data for each population shown above). (D) Force-dependent extension (mean ± SD) of the last intermediate and its fitting with the FJC model.

Figure S7. Mechanical characterization of PilY1^642−1161 in constant force with Ca²⁺ and identification of the most prevalent conformations of the variable intermediates.

(A) Unfolding under constant force of PilY1^642−1161 in the presence of Ca²⁺ and classification of the intermediates by unfolding order and color. The category plots show the stacked extensions of each intermediate per trajectory across the five unfolding forces tested (20, 25, 30, 35, and 40 pN). This protein lacks the signature of the I1 and I2 intermediates observed in PilY1^501−1161. (B) Cluster analysis of the extension of the variable intermediates per order of unfolding and force in Ca²⁺. Each plot shows the average extension of the events classified inside each cluster and for each intermediate across the five tested forces. The size of the data points is proportional to the number of events in that cluster (most represented ones are colored). (C) Force-dependent extensions (mean ± SD) of the most populated conformations for each intermediate (I3 to I7). The size of the data points is proportional to the number of events in that cluster. When the extension of the clusters identified in (B) is represented for each intermediate across the five unfolding forces tested, it is possible to identify the different conformations of each intermediate and model the data with individual fits of the FJC model (∆L_c data for each population shown above). (D) Force-dependent extension (mean ± SD) of the last intermediate and its fitting with the FJC model.

Table S4. Summary of the data from the Bell model fits (value ± SD) in main text Fig 4G.

These findings demonstrate that the sequence spanning residues 501–641 is involved in the dynamics observed at 4.0–6.5 pN. However, the removal of this 140-long residue sequence has a larger-than-expected impact on the total length of the protein. The predicted structure (Fig 4H) indicates that the N and C termini of this integrin-binding domain are close to each other (L₀ ∼1 nm). Assuming a length of 0.36 nm·residue⁻¹, the 140-long polypeptide would yield an ∆L_c ∼ 49 nm, in agreement with the values obtained for I2 (Figs 4H, S3D, and S4D). Although these observations support the connection between the 501–641 sequence and I2, the missing ∼60-nm length of PilY1^641−1161 accounts for the loss of I2 and I1, which was never detected. Measurements of the complete extension of the PilY1^501−1161 and PilY1^642−1161 constructs from 0.5 pN to its full unfolding at 40 pN reveal that both proteins achieve their full predicted extensions based on their sequence length (Fig S8A–C). Hence, the absence of the I1 signature in PilY1^642−1161 could be due to its inability to fold properly if the integrin-binding domain is not present.

Figure S8. Full extension of PilY1^501−1161 and PilY1^642−1161.

(A) Unfolding extension of PilY1^501−1161 from 0.5 pN to 10, 15, 20, 25, 30, 35, and 40 pN. (B) Unfolding extension of PilY1^642−1161 from 0.5 pN to 10, 15, 20, 25, 30, 35, and 40 pN. (C) Force-dependent extension of both proteins from 0.5 pN to different forces, and fitting to the FJC model. Complete extension (L_C) is calculated (0.36 nm·res⁻¹) using the sequence length and the length of folded domains from the tether after stretching (L_t). Both experimentally (left) and theoretically (right), the difference in length between both constructs is ∼50 nm, which matches the expected size of the I2 intermediate, and which is missing in the construct PilY1^642−1161.

Integrin binding blocks the unfolding of I2

The evidence from the previous section suggests that the intermediate I2 could be the putative structure that contains the RGD and Ca²⁺-binding site responsible for integrin binding. To test this hypothesis, we conducted binding experiments in the presence of the extracellular domains of human integrin αVβ5 (Fig 5A), which have shown the integrin with the highest affinity to PilY1 (Johnson et al, 2011). To conduct these experiments, we used a buffer supplemented with calcium, magnesium, and manganese (binding buffer), because these metals are important for the structural integrity of integrin (Smith et al, 1990; Johnson et al, 2011). Before proceeding with the binding experiments, we characterized PilY1^501−1161 nanomechanics in this new buffer. Figs S9A–C and S10A–G summarize these results in force-ramp and constant force measurements, respectively. The protein exhibits a mixed mechanical behavior, where the unfolding force means of I1 and I2 are like the ones observed with EGTA, but the variable and last intermediates reproduce the Ca²⁺ condition results (Fig S9A–C). This can be explained due to the presence of these three divalent cations and their influence on PilY1 nanomechanics (Fig S2A–C), where Mn²⁺ and/or Mg²⁺ could be interfering with Ca²⁺ binding in the integrin-binding domain. In contrast, the unfolding force means of the intermediates modulated by the second Ca²⁺-binding site are the same as those under the Ca²⁺ buffer condition, suggesting that both sites could have different affinities for various divalent cations.

Figure 5. Integrin binds to I2 in PilY1^501−1161 under force and prevents its full unfolding.

(A) Diagram of the magnetic tweezer-binding experiment with PilY1^501−1161 and the extracellular domains of the human integrin αVβ5 heterodimer. (B) Magnetic tweezer trajectory of PilY1^501−1161 held at 6 pN in the presence of integrin. I1 and I2 domains fold and unfold in equilibrium, and integrin binding is detected as a shortening of the unfolded extension of I2 (inset). (C) ∼4-h-long trajectory at 5 pN where integrin binds (insets i and iv) and unbinds (inset iii) to PilY1^501−1161 I2 domain, yielding a ∼4-nm change in the extension. Inset ii shows the extension distribution of the protein before binding (red) and after binding (green). The bar shows the number of points. (D) Comparison of the force-dependent extension changes of the integrin-bound I2 (mean ± SD) with force-ramp data from non-bound I2 and its two sub-intermediates I2A and I2B. Fitting the data with the FJC model reveals that integrin-bound I2 force-dependent extension falls within the fit for I2B sub-intermediate, which indicates that integrin binding prevents the unfolding of I2A.

Figure S9. Mechanical characterization of PilY1^501−1161 with force ramp (1 pN·s⁻¹) in the presence of binding buffer (Ca²⁺, Mg²⁺, and Mn²⁺).

(A) Force-dependent extension of (from top to bottom) I1, I2 (as a single or as two events), variable intermediates, and last intermediate. The lines are fits of the data with the FJC model. (B) Unfolding force distribution of the four populations of intermediates (mean ± SD). Variable intermediates are shown separately. (C) Comparison of the unfolding force distributions of the four classes of intermediates between the buffer conditions tested. Variable intermediates are shown collectively under a single distribution.

Figure S10. Mechanical characterization of PilY1^501−1161 in constant force with binding buffer (Ca²⁺, Mg²⁺, and Mn²⁺).

(A) Unfolding under constant force of PilY1^501−1161 in binding buffer and classification of the intermediates by unfolding order and color. The category plots show the stacked extensions of each intermediate per trajectory across the five unfolding forces tested (20, 25, 30, 35, and 40 pN). (B) Distribution of the total number of unfolding intermediates in PilY1^501−1161 (**) and variable intermediates (*) observed across trajectories. (C) Percentage of I2 domains unfolding in two consecutive sub-intermediates compared with EGTA and Ca²⁺ conditions. (D) Force-dependent change in the combined extension of all PilY1^501−1161 intermediates (**) and variable intermediates (*). Lines are fits of the FJC model to the average extension (mean ± SD), and the resulting values of contour length increment (∆L_c) are shown above inside the graph. (E) Force-dependent extension change (mean ± SD) of the I1, I2 (as a single and as two events), and last intermediates and their fitting with the FJC model. (F) Folding probability of I1 and I2 between 4.5 and 7.0 pN. The folded fraction of both intermediates sharply decreases with force, following a sigmoidal trend (line fits). The coexistence force (P₅₀) of I1 is 5.1 pN and of I2 is 6.3 pN. (G) Force dependency of folding and unfolding kinetics of I1 and I2. I1 and I2 unfolding and folding kinetics with Bell’s model (I1, empty circles and dashed line for folding: k_F⁰ = [8.0 ± 2.1] ×10² s⁻¹, ∆x_F = − 6.56 ± 0.20 nm. Solid circles and lines for unfolding: k_U⁰ = [11.2 ± 0.04] ×10⁻² s⁻¹, ∆x_U = 0.37 ± 0.01 nm; I2 k_F⁰ = [7.2 ± 3.6] ×10⁵ s⁻¹, ∆x_F = − 10.41 ± 0.34 nm. Solid circles and lines for unfolding: k_U⁰ = [4.0 ± 0.2] ×10⁻² s⁻¹, ∆x_U = 0.27 ± 0.01 nm).

After setting the basis of PilY1^501−1161 nanomechanics in binding buffer, we conducted integrin-binding experiments under force. Fig 5B details the fingerprint for integrin-binding events. At 6 pN, I1 and I2 domains fold and unfold in equilibrium, visiting the three different levels of extension. As shown in the inset, the protein stops visiting the unfolded extension of I2 at the moment indicated with a green arrow, but the dynamics continue unchanged below this level. This shortening of the I2 unfolded extension only occurs and is specifically induced in the presence of integrin (Fig S11A and B). In Fig 5C, we show a trajectory of PilY1^501−1161 held at 5 pN for an extended period of time in the presence of integrin. Insets i and iv indicate binding events and inset iii one unbinding event. In the bound state, the I2 domain cannot reach its unfolded extension, but a shorter one. After unbinding, I2 recovers its corresponding extension at the measured force (inset ii shows the protein extension distribution before and after binding). We measured the average extension of this new conformation of I2 across the forces where integrin binding was detected (Fig 5D). When compared to the force-dependent extension of naive I2, we can see that the integrin-bound I2 average extension falls within the FJC fit of the sub-intermediate I2B. This would indicate that integrin binding partially blocks the unfolding of I2 and that the sub-intermediate I2A cannot extend under this condition. Integrin recruitment occurs only in the presence of the RGD motif (residues 617–619) present in the PilY1^501−1161 construct (Johnson et al, 2011), which bestows the specificity of the PilY1–integrin interaction (Fig S12). These observations indicate that I2 is the conformational structure containing the integrin-binding domain, which spans residues 501–641.

Figure S11. Integrin-binding specificity to PilY1^501−1161.

(A) Control experiment in the presence of 75 μM of BSA. In the trajectories (left) at three reference constant forces (5.5, 6.0, and 6.5 pN), PilY1^501−1161 I1 and I2 domains undergo conformational changes between their folded and unfolded states. The extension changes remain constant over time (histograms on the right side), demonstrating that BSA does not induce any conformational changes, unlike in the presence of integrin. (B) Force-dependent extension changes (mean ± SD) of the I2 intermediate in the absence (circles) and presence (diamonds) of BSA overlap, indicating that BSA does not alter I2 conformation. The line represents the fit to the FJC model of I2 force-dependent extension changes in the absence of BSA.

Figure S12. Integrin binding to PilY1 is RGD-dependent.

PilY1^642−1161 control experiment in the presence of 800 nM of integrin. PilY1^642−1161 does not exhibit conformational changes at the forces where we observed integrin binding in the PilY1^501−1161 construct (below 7 pN). Therefore, we evaluate the possibility of integrin binding to PilY1^642−1161 by measuring possible changes in its final unfolded extension. In the presence of integrin, we repeat a protocol of unfolding at 40 pN, then quenching to 5.0 pN (for different times: 50, 100, 200, 500, 1,000, and 2,000 s) to allow folding (and the putative binding of integrin), and then testing the unfolding trajectory again at 40 pN. We compare the final extension after unfolding at 40 pN during the experiment (the dashed line is placed at 0 nm as a reference of the drift during the experiment). No changes are detected in the extension, which indicates that integrin binding is specific and requires the presence of the RGD motif (residues 617–619) to bind. Insets i and ii show the unfolding trajectory before and after integrin addition, and inset iii shows the mean ± SD extension of PilY1^642−1161 at 40 pN in the absence and presence of integrin.

Discussion

Bacteria colonize and adhere to substrates upon recognition of the mechanical stimuli elicited by surface proximity (Rodesney et al, 2017; Wang et al, 2023). These cues trigger signaling pathways in free-swimming cells that promote their attachment to target substrates and the formation of biofilms. This change of lifestyle reprograms gene expression and, in pathogenic bacteria, kick-starts the deployment of virulence factors, having important implications for infection progression (Heiniger et al, 2010; Luo et al, 2015).

The ability of bacteria to detect mechanical forces has been linked to their motor-driven appendages, the flagella, and the T4P (O’Toole & Kolter, 1998; Lele et al, 2013). The T4P mediate twitching motility and are essential for the surface-attached life of bacteria. PilY1 plays a crucial part in T4P biogenesis and functions, while also serving as a mechanosensor and adhesive protein. Because of its location at the tip of the T4P, PilY1 carries out its functions while exposed to mechanical stress. Here, we have explored how ligand binding and mechanical forces impact PilY1 dynamics.

The sequence comprising the residues 501–1,161 of PilY1 contains ligand-binding motifs that control T4P dynamics and host adhesion. Most of this sequence adopts a 7-bladed β-propeller fold, where blade IV contains the Ca²⁺-binding site that regulates T4P extension and retraction cycles. Forces above 20 pN lead to the sequential hierarchical unfolding of the domains of this heteropolyprotein. Under force, a polyprotein comprised of independent domains of varying levels of stability will usually undergo the unfolding of the least stable domains first before the more stable ones (Li et al, 2000). In the PilY1 constructs studied (Figs 1, 2, 3, 4, 5, S1, S2, S3, S4, S5, S6, S7, S8, S9, S10, S11, and S12), we observe a tightly conserved unzipping pattern where the unfolding of each domain depends on the unfolding of the previous one in the sequence, which reminds a Matryoshka doll–like configuration. This suggests the existence of long-range contacts and allosteric interactions over the entire structure of the protein (Batey & Clarke, 2006; Perez-Riba et al, 2018; Santorelli et al, 2023). The apparently conserved unfolding pathway is challenged by the heterogeneous conformations exhibited by the variable intermediate group. Ca²⁺ binding to its cognate site in blade IV increases the mechanical stability of the variable intermediate I4 and delays the unfolding of the subsequent domains (Figs 2 and 4, S1, and S9). Furthermore, this metal increases the frequency of conformations sparsely visited in EGTA (Figs 2 and 4, S3, S4, S6, and S7), indicating that metal binding impacts the mechanical properties and the unfolding pathway of the protein (Cao et al, 2011; Scholl et al, 2016; Löf et al, 2019; Yuan et al, 2019; Nie et al, 2022). The fact that the total unfolding extension of the variable intermediates is always the same confirms that this region of PilY1 folds between unfolding cycles (Figs 2 and 4). A source of conformational variability could be that after quenching the force and allowing folding, the variable intermediates experience misfolding events where their contacts are not achieved intradomain, but instead, non-native interactions are established with neighboring modules (Oberhauser et al, 1999). If that was the case, the unfolding rates of the variable intermediates could not be explained as single processes (Figs 2 and 4), but as a multitude of kinetic processes, each departing from a different folded state. Hence, the hierarchical pattern observed for I1, I2, and the last intermediate groups also applies inside the variable intermediate sequence.

Our findings indicate that the PilY1 β-propeller has a partially conserved unfolding pathway that Ca²⁺ alters when bound to blade IV. A rough estimation of the expected unfolded extension of each blade of the β-propeller and its candidate intermediates is shown in Fig S13. PilY1 is not a canonical β-propeller fold (Mylemans et al, 2021), and blades V-VII have a different number of strands than blades I-IV, which are four-stranded antiparallel β-sheets. However, the estimated contour length of all blades is similar, and assigning their identities is not straightforward. Furthermore, blades V and VI share one strand (Orans et al, 2010), which makes it even more challenging to devise what kind of unfolding intermediate or intermediates could generate. Ultimately, the core population of the variable intermediates could be narrowed down to four different structures of ∆L_c ∼10–20 nm (I3 to I6), and the alternative conformations observed can be explained as the rupture of substructures spanning different blades.

Figure S13. List of PilY1^501−1161 domains and a rough estimation of their size and candidate intermediates.

On the left, a cartoon representation of each domain. The domains are connected by α-helices and unstructured loops, and the criteria for defining the ends of each domain (N-C distance in the folded structure, L₀) have been set by splitting by half the length of these interdomain regions. Contour length values (L_C) are determined by multiplying the number of residues of each module by 0.36 nm·res⁻¹. On the right, ∆L_C values for each domain of PilY1^501−1161 in Ca²⁺ (Fig S4). Because of their overlapping sizes, most of the β-propeller blades have several candidates. Blades V-VI have been represented as a single unit because they share a β-strand that splits both domains. Hence, their structural integrity could be interdependent. Under force, the interdomain regions could be stretched before unfolding; therefore, the estimated values represented here could be overestimated.

From a biological standpoint, the stabilizing effect of Ca²⁺ in the β-propeller could be relevant in T4P dynamics regulation. P. aeruginosa T4P exerts pulling forces ∼30 pN that overlap with the force ranges we have tested and where Ca²⁺ has shown its stabilizing effect (Ribbe et al, 2017). How the Ca²⁺-bound state of PilY1 antagonizes T4P retraction is not yet clear, but one study hypothesized that PilY1, in association with other pilus-tip proteins involved in T4P biogenesis, might block the full retraction of the T4P beyond the outer membrane PilQ secretin pore, acting as a cork. In this scenario, Ca²⁺-induced mechanical stabilization of PilY1 would maintain the bulky structure of the β-propeller and prevent its unfolding, blocking further pilus retraction. This would ensure the structural integrity of the pilus-tip complex for priming pilus polymerization in a subsequent extension cycle (Treuner-Lange et al, 2020; Guo et al, 2024). It is also feasible that the increased mechanical stabilization of PilY1 by Ca²⁺ could be linked to its adhesive role, contributing to the structural integrity of this protein during surface attachment, as has been proposed for other bacterial pilin proteins (Echelman et al, 2016; Oude Vrielink et al, 2017; Milles et al, 2018). Across species, it has been seen that PilY1 orthologs exhibit different Ca²⁺ affinities, and its effect in T4P dynamics varies from P. aeruginosa. This would entail that the Ca²⁺-mediated binding and T4P regulation could be finely tuned to the chemical environment of bacteria (Morand et al, 2004; Cruz et al, 2014; Parker et al, 2015). Similar to chemical affinity, the nanomechanics of PilY1 is probably adapted to the environment and are specific to the host. This specificity would create a unique interplay between mechanical forces and ligand binding in regulating PilY1 functions, depending on the substrate being colonized (Cruz et al, 2012, 2014; Parker et al, 2015; Herfurth et al, 2022; Hernández-Sánchez et al, 2024).

PilY1 mechanical behavior at high force can match its expected function as a T4P retraction inhibitor, but its dynamics in the low-force regime resemble a force-sensing activity linked to its adhesin function. The I2 intermediate seems the most plausible candidate to harbor the motifs responsible for integrin binding, not only by its dimensions but also for the signature conformation it adopts upon binding (Figs 3, 4, and 5). Shortening of the unfolded extension of proteins after ligand binding has been reported in the proteins talin and vinculin, keystones of eukaryotic mechanotransduction (Yao et al, 2014; Tapia-Rojo et al, 2020). Integrin binding to PilY1 prevents the full extension of the I2 domain, but, unlike the talin–vinculin interaction, the remaining structure of I2 can still undergo folding and unfolding transitions. The conformation of this halfway structure matches the force-dependent extension of the sub-intermediate I2B; hence, integrin binding prevents the unfolding of I2A. These sub-intermediates unfold almost simultaneously in a sequence I2A-I2B and are noticeable when pulling at high forces. Their frequency increases with Ca²⁺ (Figs 2 and S10), which suggests that metal binding in I2 could promote a structural organization of the binding pocket that facilitates integrin recruitment (Johnson et al, 2011). Furthermore, Ca²⁺ binding in I2 seems to activate an allosteric mechanism that lowers the stability of I1, which might have implications for integrin-binding modulation under tensile stress. This structural connection between I1 and I2 seems plausible because the elimination of I2 abrogates the mechanical signature of I1 (Fig 4), pointing out that both domains might be interdependent for folding. This force and posttranslational modulation of the nanomechanics of the integrin-binding domain and I1 could entail an evolutionary tuning to the mechanical environment of the respiratory epithelium, as it has been proposed for other bacterial adhesive structures (Alonso-Caballero et al, 2018, 2021). Integrin binding has been shown to occur in the absence of force (Johnson et al, 2011); nevertheless, the allosteric interactions that seem to govern PilY1 mechanics could play a significant role in vivo, promoting or hindering host adhesion depending on the mechanical environment experienced by the bacterium.

Overall, our approach to PilY1 dynamics provides mechanistic details of how force could tune some of its functions at the nanoscale. We propose that PilY1 functions are mechanically compartmentalized; their recruitment depends on the stability of the structures that harbor them, ligand binding, and the mechanical load. Future work should address the nanomechanics of the vWA domain, given its central role in bacterial mechanotransduction and its spatial location in PilY1. Structural models indicate that the vWA domain may be blocking the integrin-binding site, meaning this site might only become accessible for ligand binding after the domain undergoes force-induced conformational changes (Guo et al, 2024). Our findings with the C-terminal section of PilY1 suggest a complex network of interactions among structural modules, and the multiple functions of PilY1 under force may depend on its overall structure.

Materials and Methods

Cloning and expression of chimeric constructs

All reagents employed were purchased from Sigma-Aldrich, unless otherwise indicated. The sequence of the proteins PilY1^501−1161 and PilY1^642−1161 from the P. aeruginosa PAO1 strain was optimized for expression in Escherichia coli. Both sequences were ordered (GeneArt; Thermo Fisher Scientific) with a 5′-end Kpn2I and a 3′-end NheI restriction sites, which enabled their cloning into modified pFN18A (HaloTag) T7 Flexi Vector (Promega). The expression cassette contained, from 5′- to 3′-end, an N-terminal HaloTag protein, four copies of the human titin I32 domain, HisTag sequence, and AviTag sequence. The sequence between the second and third I32 domain gene copies contained the Kpn2I and NheI restriction sites, which allowed the insertion of the proteins of interest in sense. Cloning and DNA amplification procedures were carried out in E. coli TOP10 strain (Thermo Fisher Scientific), and sequence fidelity was confirmed (GeneWiz).

Protein expression (PilY1^501−1161 construct) was conducted in E. coli BL21 Star strain (Thermo Fisher Scientific). Cell culture was grown to OD₆₀₀∼0.6 in LB broth containing 100 μg·ml⁻¹ carbenicillin at 37°C and 250-rpm shaking, and then, 1 mM IPTG (Invitrogen) was added to induce protein expression at 25ºC for 3 h and at constant shaking. After, cells were pelleted (4°C, 4,000g, 20 min) and resuspended in wash buffer (50 mM NaPi at pH 7.0, 300 mM NaCl, 20 mM imidazole, 10% vol/vol glycerol) supplemented with EDTA-free protease inhibitor (Roche). Cells were then incubated for 30 min on ice with 1 mg·ml⁻¹ lysozyme, 5 μg·ml⁻¹ DNase I, 5 μg·ml⁻¹ RNase A, and 10 mM MgCl₂. Then, the cells were mechanically lysed in a French press (G. Heinemann) and cell debris was pelleted (4°C, 40,000g, 1 h). The supernatant was incubated with HisPur Cobalt resin (Thermo Fisher Scientific) at 4°C for 1 h on an orbital shaker. The His-tagged proteins were eluted from the resin with wash buffer containing 250 mM imidazole and subsequently biotinylated in vitro at room temperature for 1 h. After, the protein solution was further purified by size-exclusion chromatography in a Superdex 200 Increase 10/300 GL column (Cytiva). The protein was eluted from the column in storage buffer (10 mM Hepes, pH 7.2, NaCl 150 mM, 1 mM EDTA, 10% vol/vol glycerol), aliquoted, snap-frozen in liquid N₂, and stored at −80°C until use. The PilY1^642−1161 construct was expressed in E. coli T7 Express strain (New England Biolabs) and cotransformed with the pBirAcm plasmid (Avidity) to biotinylate the protein in vivo. Cell culture was grown to OD₆₀₀∼0.6 in LB broth containing 100 μg·ml⁻¹ carbenicillin and 10 μg·ml⁻¹ chloramphenicol (pBirAcm). Protein expression and biotinylation were induced with 1 mM IPTG and 50 μM D-biotin (Invitrogen) for 3 h at 25°C and constant shaking. Downstream purification steps were carried out as described above. Protein purification was assessed using SDS–PAGE (Fig S14).

Figure S14. Protein integrity and purity for PilY1^501−1161 and PilY1^642−1161.

FPLC size-exclusion chromatography (left) and SDS–PAGE (right) gels of the PilY1^501−1161 (top) and PilY1^642−1161 constructs (bottom). The circles (chromatograph and SDS–PAGE lanes) indicate the fractions containing the highest-purity samples. The left lanes contain the mPAGE Color Protein Standard molecular weight marker (Millipore).

Single-molecule magnetic tweezer measurements

Single-molecule experiments were conducted on custom-built magnetic tweezer microscopes, using either a configuration consisting of a pair of permanent magnets, as previously described (Popa et al, 2016), or a magnetic tape head (Tapia-Rojo et al, 2024). In the magnetic tape head configuration, the force is applied to the magnetic bead–bound proteins with a magnetic tape head (Brush Industries), which is fixed on a custom-designed CNC aluminum piece (named card reader). In the base of the card reader lies the experimental fluid chamber where the protein under study is immobilized on the bottom glass (see below for details). The 150-μm thickness of the bottom glass places the sample-located surface at a 300-μm distance from the 25-μm gap of the magnetic tape head, which establishes a known current–force relationship (Tapia-Rojo et al, 2019). The 25-μm gap of the magnetic tape head is aligned with the optical path of the microscope. The base of the card-reader piece contains an opening to accommodate a 100X oil-immersion Plan Apochromat objective (Zeiss) that enables the visualization of the reference and magnetic beads located on the bottom glass of the fluid chamber. The objective position was controlled with a P-725 nanofocusing piezo actuator (Physik Instrumente), and the image was acquired with a CMOS camera (Ximea). Image processing was conducted with a custom-written program (C++/Qt) that allowed a ∼1.5-kHz image sampling frequency. This software communicated with a NI USB-6289 multifunction DAQ card (National Instruments), which enabled data acquisition and controlling the piezo position and the current supplied (hence, the force applied to the protein) to the magnetic tape head. The magnetic tape head electrical supply came from two 12-V lead–acid batteries, and the current was kept under feedback with a custom-built PID controller. Applying currents in the 0-1,000 mA range allowed the application of forces spanning from 0 to 42 pN using Dynabeads M-270 streptavidin–coated superparamagnetic beads (Invitrogen).

Fluid chamber preparation involved the cleaning and functionalization of a 40 × 24 mm bottom glass (150-μm thickness) and a 22 × 22 mm top glass (Ted Pella), as previously described (Popa et al, 2016). After cleaning, the top glass was treated with Repel-Silane for 30 min and the bottom glass was silanized with a 1% vol/vol solution of 3-(aminopropyl)trimethoxysilane in ethanol for 10 min. After functionalization, both types of glasses were rinsed in 100% ethanol, dried with N₂, and curated in an oven at 100°C for at least 1 h.

The fluid chambers were assembled by melting a bowtie-shaped parafilm spacer placed between the top glass and the bottom glass at 80°C for 5 min. A laser cutter was used to generate the parafilm spacer shape. After sandwiching between glasses, the parafilm shape created two wells flanking the top glass that allowed buffer exchange during experiments. Further fluid chamber functionalization steps carried out at room temperature involved the sequential treatment with 1% vol/vol solution of glutaraldehyde in 0.1 M phosphate buffer (PBS) for 1 h, covalent immobilization of a 0.05% wt/vol solution of amino-functionalized polystyrene reference beads (2.5–2.9 μm in diameter; SpheroTec) in PBS for 20 min, extensive rinse with PBS, and covalent anchoring of 20 μg·ml⁻¹ HaloTag amine ligand (Promega) in PBS overnight. After HaloTag ligand functionalization, the fluid chambers were passivated for at least 3 h with blocking buffer containing 1% wt/vol sulfhydryl-blocked BSA (Lee Biosolutions) in 20 mM Tris–HCl, pH 7.4, 150 mM NaCl, and 0.01% wt/vol NaN₃.

To conduct force experiments, the fluid chambers were incubated with a ∼3 nM solution of the chimeric constructs (PilY1^501−1161 and PilY1^642−1161) at room temperature for 30 min. The unbound protein was removed by flowing buffer from one well of the fluid chamber and removing from the other. The basic composition of the buffer used in the experiments contained 10 mM Hepes, pH 7.2, 150 mM NaCl, and 10 mM ascorbic acid. Depending on the condition, this buffer contained either 2 mM EGTA, 2 mM CaCl₂, or binding buffer (2 mM CaCl₂, 1 mM MgCl₂, and 1 mM MnCl₂). The fluid chamber was then fixed with double-sided tape (Tesa) to a custom-made metal stamped fork. The unit formed by the fork-fluid chamber assembly was placed in the base of the card-reader piece, leaving the fluid chamber below the magnetic tape head. The fork was then attached to an XY linear stage (Newport), which allowed to move the fluid chamber and scan different areas during the experiments.

The experiments started by flowing a 1:30 dilution of streptavidin-coated superparamagnetic beads, previously passivated with blocking buffer at 4°C, for >1 h, and constant rotation at 5 rpm. The beads and the surface-immobilized proteins were allowed to react for 2 min in the absence of force. Then, a constant force of 2–4 pN was applied to remove unspecific-bound beads, and the fluid chamber was further flowed with experimental buffer. From here, the different force protocols described in the work were applied. At the forces tested, the unfolding probability of the HaloTag or I32 domains remains low. In the rare event of unfolding, their identity can be easily ascertained and does not alter PilY1 behavior (Fig S15). For the integrin-binding experiments, PilY1^501−1161 proteins were held at forces that allowed folding and unfolding transitions of the I1 and I2 intermediates, and then, a binding buffer solution containing 200 nM of recombinant αVβ5 human integrin extracellular domains (R&D Systems) was slowly added to one of the wells of the fluid chamber, while simultaneously removing the previous buffer volume from the other well until complete buffer substitution.

Figure S15. Effect of tag unfolding on PilY1^501−1161 dynamics.

HaloTag and I32 exhibit high mechanical stability. At the forces explored in this study, the unfolding probability of any of these domains is low; therefore, they remain folded and act as rigid handles during the experiments. In the rare event of unfolding, the tag can be easily identified based on the extension and its known sequence and structure. In the trajectory shown, taken at 5.5 pN, the unfolding of one I32 domain (yielding an extension increase of 13 nm, indicated with an arrow) shifts the protein’s offset extension and increases the signal’s noise (more fluctuation after the release of additional unstructured polypeptide). Still, I1 and I2 dynamics continue unaltered. The gray dashed lines indicate the unfolded level of I1 and the folded level of I2 as a reference.

Analysis

All the analysis was carried out with Igor Pro 8.0 software (WaveMetrics). Magnetic tweezer recordings were captured at ∼1.0–1.5 kHz frame rate, and trajectories were smoothed with a 151 box-sized fourth-order Savitzky–Golay filter. To assess the significance of the difference between the unfolding forces between buffer conditions (EGTA or Ca²⁺) and between intermediates (Tables S1, S2, and S3), the difference of the means and its uncertainty are computed (from the SEM of the values at each condition). The difference is stated significant when the probability of being zero is smaller than 5% (∣t∣>t_critical, being t_critical=2 for P < 0.05) (Trosset, 2009).t=μ1−μ2SEM12+SEM12.

Step size determination of the unfolding intermediates was determined by measuring the extension before and after the transition, and for the equilibrium dynamics at low force, a multi-Gaussian fit was employed to find the center of the extension levels and measure their distance. The force-dependent extension of the intermediates was fitted to the FJC model for polymer elasticity (Flory, 1956):∆L(F)=∆LC[coth(FlKkBT)−kBTFlK],where ∆L_C is the increment of contour length in nm, l_K is the Kuhn length in nm, and k_BT is the thermal energy as 4.11 pN·nm. Behavior at high forces is dominated by the contour length, whereas the Kuhn length is determined by low-force behavior. Hence, the Kuhn length is only reported for those measurements and intermediates, which exhibit folding and unfolding events in the low-force regime (I1 and I2).

The force-dependent folding probability of the intermediates I1 and I2 was fitted with a sigmoidal function:FP(F)=(exp(F−P50r)+1)−1,where P₅₀ is the coexistence force (in pN) between the folded and unfolded state, and r is the rate in pN⁻¹.

Dwell time analysis in the unfolding trajectories was measured from the moment the desired force is achieved (∼40 µs) until each of the intermediates unfolds. Because their unfolding is sequential and requires that the previous structure unfolds first, their unfolding times were corrected based on this. The same procedure was used for the dynamics registered at low force. Because of their long nature, a custom-written procedure was written to automatically detect the extension levels and assign the folding and unfolding dwell times of I1 and I2 intermediates. Folding and unfolding rates were calculated as the inverse of the average dwell times, and their force dependency was fitted with the Bell model (Bell, 1978):kU/F(F)=kU/F 0e(±F∆xU/FkBT),where kU/F 0 is the unfolding/folding rate extrapolation to zero force, and ∆x_U/F is the distance to the transition state for unfolding and folding.

Cluster analysis of the variable intermediates was done with R Statistical Software (v 4.3.0; R Core Team, 2023) using the libraries “factoextra” (Kassambara & Mundt, 2020) and “cluster” (Maechler et al, 2024). The k-means algorithm and the Euclidean distance were used to establish the differences in the extension of the conformations of the variable intermediates. We tested several k-means solutions varying the number of clusters from 1 to 20 for each force, protein construct, and buffer condition. We conducted the average silhouette analysis method to find the optimal number of clusters per condition, and this value was used to cluster the data.

Data Availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgements

The authors thank Pablo Mateos-Gil, Marta Martín, and Rafael Rivilla for helpful discussions and comments. This project has received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Sklodowska-Curie grant agreement No. 101028879. This work was supported by and performed at the Francis Crick Institute, which receives its core funding from Cancer Research UK (CC0102), the UK Medical Research Council (CC0102), and the Wellcome Trust (CC0102). FJ Cao-Garcia was supported by the Spanish Ministry of Economy and Competitiveness and the European Regional Development Fund (RTI2018-095802-B-I00), the Spanish Ministry of Science, Innovation and Universities (PID2023-148319NB-I00), and the Universidad Complutense de Madrid (UCM) (PR12/24-31558). JE Walker and S Board were supported by the Leverhulme Trust (RL 2016-015). A Alonso-Caballero and this work were supported by a European Commission Marie Sklodowska-Curie fellowship (NIOBMT-101028879), the project PID2023-146082NA-I00 funded by MICIU/AEI/10.13039/501100011033 and by ERDF/EU, and the grant RYC2021-031965-I (MICIU/AEI/10.13039/501100011033 and European Union NextGenerationEU/PRTR).

Author Contributions

FJ Cao-Garcia: data curation and formal analysis.
JE Walker: methodology.
S Board: methodology.
A Alonso-Caballero: conceptualization, data curation, formal analysis, validation, investigation, visualization, and methodology.

Conflict of Interest Statement

The authors declare that they have no conflict of interest.

Received October 24, 2024.
Revision received February 28, 2025.
Accepted February 28, 2025.

https://creativecommons.org/licenses/by/4.0/

This article is available under a Creative Commons License (Attribution 4.0 International, as described at https://creativecommons.org/licenses/by/4.0/).

References

↵
1. Alonso-Caballero A,
2. Schönfelder J,
3. Poly S,
4. Corsetti F,
5. De Sancho D,
6. Artacho E,
7. Perez-Jimenez R
(2018) Mechanical architecture and folding of E. coli type 1 pilus domains. Nat Commun 9: 2758. doi:10.1038/s41467-018-05107-6
OpenUrl CrossRef PubMed
↵
1. Alonso-Caballero A,
2. Echelman DJ,
3. Tapia-Rojo R,
4. Haldar S,
5. Eckels EC,
6. Fernandez JM
(2021) Protein folding modulates the chemical reactivity of a Gram-positive adhesin. Nat Chem 13: 172–181. doi:10.1038/s41557-020-00586-x
OpenUrl CrossRef PubMed
↵
1. Batey S,
2. Clarke J
(2006) Apparent cooperativity in the folding of multidomain proteins depends on the relative rates of folding of the constituent domains. Proc Natl Acad Sci U S A 103: 18113–18118. doi:10.1073/pnas.0604580103
OpenUrl Abstract/FREE Full Text
↵
1. Bell GI
(1978) Models for the specific adhesion of cells to cells. Science 200: 618–627. doi:10.1126/science.347575
OpenUrl Abstract/FREE Full Text
↵
1. Cao Y,
2. Li YD,
3. Li H
(2011) Enhancing the mechanical stability of proteins through a cocktail approach. Biophys J 100: 1794–1799. doi:10.1016/j.bpj.2011.02.030
OpenUrl CrossRef PubMed
↵
1. Chawla R,
2. Gupta R,
3. Lele TP,
4. Lele PP
(2020) A skeptic’s guide to bacterial mechanosensing. J Mol Biol 432: 523–533. doi:10.1016/j.jmb.2019.09.004
OpenUrl CrossRef PubMed
↵
1. Craig L,
2. Forest KT,
3. Maier B
(2019) Type IV pili: Dynamics, biophysics and functional consequences. Nat Rev Microbiol 17: 429–440. doi:10.1038/s41579-019-0195-4
OpenUrl CrossRef PubMed
↵
1. Cruz LF,
2. Cobine PA,
3. De La Fuente L
(2012) Calcium increases Xylella fastidiosa surface attachment, biofilm formation, and twitching motility. Appl Environ Microbiol 78: 1321–1331. doi:10.1128/AEM.06501-11
OpenUrl Abstract/FREE Full Text
↵
1. Cruz LF,
2. Parker JK,
3. Cobine PA,
4. De La Fuente L
(2014) Calcium-enhanced twitching motility in xylella fastidiosa is linked to a single PilY1 homolog. Appl Environ Microbiol 80: 7176–7185. doi:10.1128/AEM.02153-14
OpenUrl Abstract/FREE Full Text
↵
1. Dufrêne YF,
2. Persat A
(2020) Mechanomicrobiology: How bacteria sense and respond to forces. Nat Rev Microbiol 18: 227–240. doi:10.1038/s41579-019-0314-2
OpenUrl CrossRef PubMed
↵
1. Echelman DJ,
2. Alegre-Cebollada J,
3. Badilla CL,
4. Chang C,
5. Ton-That H,
6. Fernández JM
(2016) CnaA domains in bacterial pili are efficient dissipaters of large mechanical shocks. Proc Natl Acad Sci U S A 113: 2490–2495. doi:10.1073/pnas.1522946113
OpenUrl Abstract/FREE Full Text
↵
1. Eyckmans J,
2. Boudou T,
3. Yu X,
4. Chen CS
(2011) A hitchhiker’s guide to mechanobiology. Dev Cell 21: 35–47. doi:10.1016/j.devcel.2011.06.015
OpenUrl CrossRef PubMed
↵
1. Flory PJ
(1956) Theory of elastic mechanisms in fibrous proteins. J Am Chem Soc 78: 5222–5235. doi:10.1021/ja01601a025
OpenUrl CrossRef
↵
1. Gordon VD,
2. Wang L
(2019) Bacterial mechanosensing: The force will be with you, always. J Cell Sci 132: jcs227694. doi:10.1242/jcs.227694
OpenUrl Abstract/FREE Full Text
↵
1. Gosse C,
2. Croquette V
(2002) Magnetic tweezers: Micromanipulation and force measurement at the molecular level. Biophys J 82: 3314–3329. doi:10.1016/S0006-3495(02)75672-5
OpenUrl CrossRef PubMed
↵
1. Guo S,
2. Chang Y,
3. Brun YV,
4. Howell PL,
5. Burrows LL,
6. Liu J
(2024) PilY1 regulates the dynamic architecture of the type IV pilus machine in Pseudomonas aeruginosa. Nat Commun 15: 9382. doi:10.1038/s41467-024-53638-y
OpenUrl CrossRef PubMed
↵
1. Heiniger RW,
2. Winther-Larsen HC,
3. Pickles RJ,
4. Koomey M,
5. Wolfgang MC
(2010) Infection of human mucosal tissue by Pseudomonas aeruginosa requires sequential and mutually dependent virulence factors and a novel pilus-associated adhesin. Cell Microbiol 12: 1158–1173. doi:10.1111/j.1462-5822.2010.01461.x
OpenUrl CrossRef PubMed
↵
1. Herfurth M,
2. Treuner-Lange A,
3. Glatter T,
4. Wittmaack N,
5. Hoiczyk E,
6. Pierik AJ,
7. Søgaard-Andersen L
(2022) A noncanonical cytochrome c stimulates calcium binding by PilY1 for type IVa pili formation. Proc Natl Acad Sci U S A 119: e2115061119. doi:10.1073/pnas.2115061119
OpenUrl Abstract/FREE Full Text
↵
1. Hernández-Sánchez A,
2. Páez-Pérez ED,
3. Alfaro-Saldaña E,
4. García-Meza JV
(2024) Deciphering the enigmatic PilY1 of Acidithiobacillus thiooxidans: An in silico analysis. Biochem Biophys Rep 39: 101797. doi:10.1016/j.bbrep.2024.101797
OpenUrl CrossRef
↵
1. Hoppe J,
2. Ünal CM,
3. Thiem S,
4. Grimpe L,
5. Goldmann T,
6. Gaßler N,
7. Richter M,
8. Shevchuk O,
9. Steinert M
(2017) PilY1 promotes Legionella pneumophila infection of human lung tissue explants and contributes to bacterial adhesion, host cell invasion, and twitching motility. Front Cell Infect Microbiol 7: 63. doi:10.3389/fcimb.2017.00063
OpenUrl CrossRef PubMed
↵
1. Johnson MDL,
2. Garrett CK,
3. Bond JE,
4. Coggan KA,
5. Wolfgang MC,
6. Redinbo MR
(2011) Pseudomonas aeruginosa PilY1 binds integrin in an RGD- and calcium-dependent manner. PLoS One 6: e29629. doi:10.1371/journal.pone.0029629
OpenUrl CrossRef PubMed
↵
1. Jumper J,
2. Evans R,
3. Pritzel A,
4. Green T,
5. Figurnov M,
6. Ronneberger O,
7. Tunyasuvunakool K,
8. Bates R,
9. Žídek A,
10. Potapenko A, et al.
(2021) Highly accurate protein structure prediction with AlphaFold. Nature 596: 583–589. doi:10.1038/s41586-021-03819-2
OpenUrl CrossRef PubMed
↵
1. Jurado-Martín I,
2. Sainz-Mejías M,
3. McClean S
(2021) Pseudomonas aeruginosa: An audacious pathogen with an adaptable arsenal of virulence factors. Int J Mol Sci 22: 3128–3137. doi:10.3390/ijms22063128
OpenUrl CrossRef PubMed
↵
1. Kassambara A,
2. Mundt F
(2020) factoextra: Extract and visualize the results of multivariate data analyses. R package version 1.0.7.999. Available at: https://github.com/kassambara/factoextra.
↵
1. Kuchma SL,
2. Ballok AE,
3. Merritt JH,
4. Hammond JH,
5. Lu W,
6. Rabinowitz JD,
7. O’Toole GA
(2010) Cyclic-di-GMP-mediated repression of swarming motility by Pseudomonas aeruginosa: The pilY1 gene and its impact on surface-associated behaviors. J Bacteriol 192: 2950–2964. doi:10.1128/JB.01642-09
OpenUrl Abstract/FREE Full Text
↵
1. Lele PP,
2. Hosu BG,
3. Berg HC
(2013) Dynamics of mechanosensing in the bacterial flagellar motor. Proc Natl Acad Sci U S A 110: 11839–11844. doi:10.1073/pnas.1305885110
OpenUrl Abstract/FREE Full Text
↵
1. Li H,
2. Oberhauser AF,
3. Fowler SB,
4. Clarke J,
5. Fernandez JM
(2000) Atomic force microscopy reveals the mechanical design of a modular protein. Proc Natl Acad Sci U S A 97: 6527–6531. doi:10.1073/pnas.120048697
OpenUrl Abstract/FREE Full Text
↵
1. Li H,
2. Linke WA,
3. Oberhauser AF,
4. Carrion-Vazquez M,
5. Kerkvliet JG,
6. Lu H,
7. Marszalek PE,
8. Fernandez JM
(2002) Reverse engineering of the giant muscle protein titin. Nature 418: 998–1002. doi:10.1038/nature00938
OpenUrl CrossRef PubMed
↵
1. Löf A,
2. Walker PU,
3. Sedlak SM,
4. Gruber S,
5. Obser T,
6. Brehm MA,
7. Benoit M,
8. Lipfert J
(2019) Multiplexed protein force spectroscopy reveals equilibrium protein folding dynamics and the low-force response of von Willebrand factor. Proc Natl Acad Sci U S A 116: 18798–18807. doi:10.1073/pnas.1901794116
OpenUrl Abstract/FREE Full Text
↵
1. Luo Y,
2. Zhao K,
3. Baker AE,
4. Kuchma SL,
5. Coggan KA,
6. Wolfgang MC,
7. Wong GCL,
8. O’Toole GA
(2015) A hierarchical cascade of second messengers regulates Pseudomonas aeruginosa surface behaviors. mBio 6: e02456-14. doi:10.1128/mBio.02456-14
OpenUrl Abstract/FREE Full Text
↵
1. Maechler M,
2. Rousseeuw P,
3. Struyf A,
4. Hubert M,
5. Hornik K
(2024) cluster: Cluster analysis basics and extensions. R package version 2.1.8. Available at: https://CRAN.R-project.org/package=cluster.
↵
1. Maier B,
2. Potter L,
3. So M,
4. Long CD,
5. Seifert HS,
6. Sheetz MP
(2002) Single pilus motor forces exceed 100 pN. Proc Natl Acad Sci U S A 99: 16012–16017. doi:10.1073/pnas.242523299
OpenUrl Abstract/FREE Full Text
↵
1. Marko VA,
2. Kilmury SLN,
3. MacNeil LT,
4. Burrows LL
(2018) Pseudomonas aeruginosa type IV minor pilins and PilY1 regulate virulence by modulating FimS-AlgR activity. PLoS Pathog 14: e1007074. doi:10.1371/journal.ppat.1007074
OpenUrl CrossRef PubMed
↵
1. Milles LF,
2. Unterauer EM,
3. Nicolaus T,
4. Gaub HE
(2018) Calcium stabilizes the strongest protein fold. Nat Commun 9: 4764. doi:10.1038/s41467-018-07145-6
OpenUrl CrossRef
↵
1. Morand PC,
2. Bille E,
3. Morelle S,
4. Eugène E,
5. Beretti JL,
6. Wolfgang M,
7. Meyer TF,
8. Koomey M,
9. Nassif X
(2004) Type IV pilus retraction in pathogenic Neisseria is regulated by the PilC proteins. EMBO J 23: 2009–2017. doi:10.1038/sj.emboj.7600200
OpenUrl Abstract/FREE Full Text
↵
1. Mordue J,
2. O’Boyle N,
3. Gadegaard N,
4. Roe AJ
(2021) The force awakens: The dark side of mechanosensing in bacterial pathogens. Cell Signal 78: 109867. doi:10.1016/j.cellsig.2020.109867
OpenUrl CrossRef
↵
1. Mylemans B,
2. Voet AR,
3. Tame JR
(2021) The taming of the screw: The natural and artificial development of β-propeller proteins. Curr Opin Struct Biol 68: 48–54. doi:10.1016/j.sbi.2020.11.009
OpenUrl CrossRef
↵
1. Nguyen Y,
2. Harvey H,
3. Sugiman-Marangos S,
4. Bell SD,
5. Buensuceso RNC,
6. Junop MS,
7. Burrows LL
(2015) Structural and functional studies of the Pseudomonas aeruginosa minor pilin, PilE. J Biol Chem 290: 26856–26865. doi:10.1074/jbc.M115.683334
OpenUrl Abstract/FREE Full Text
↵
1. Nie J-Y,
2. Song G-B,
3. Deng Y-B,
4. Zheng P
(2022) Single-molecule force spectroscopy reveals stability of mitoNEET and its [2Fe2Se] cluster in weakly acidic and basic solutions. ChemistryOpen 11: e202200056. doi:10.1002/open.202200056
OpenUrl CrossRef
↵
1. Oberhauser AF,
2. Marszalek PE,
3. Carrion-Vazquez M,
4. Fernandez JM
(1999) Single protein misfolding events captured by atomic force microscopy. Nat Struct Biol 6: 1025–1028. doi:10.1038/14907
OpenUrl CrossRef PubMed
↵
1. Orans J,
2. Johnson MDL,
3. Coggan KA,
4. Sperlazza JR,
5. Heiniger RW,
6. Wolfgang MC,
7. Redinbo MR
(2010) Crystal structure analysis reveals Pseudomonas PilY1 as an essential calcium-dependent regulator of bacterial surface motility. Proc Natl Acad Sci U S A 107: 1065–1070. doi:10.1073/pnas.0911616107
OpenUrl Abstract/FREE Full Text
↵
1. Oude Vrielink AS,
2. Vance TDR,
3. de Jong AM,
4. Davies PL,
5. Voets IK
(2017) Unusually high mechanical stability of bacterial adhesin extender domains having calcium clamps. PLoS One 12: e0174682. doi:10.1371/journal.pone.0174682
OpenUrl CrossRef PubMed
↵
1. O’Toole GA,
2. Kolter R
(1998) Flagellar and twitching motility are necessary for Pseudomonas aeruginosa biofilm development. Mol Microbiol 30: 295–304. doi:10.1046/j.1365-2958.1998.01062.x
OpenUrl CrossRef PubMed
↵
1. Parker JK,
2. Cruz LF,
3. Evans MR,
4. De La Fuente L
(2015) Presence of calcium-binding motifs in PilY1 homologs correlates with Ca-mediated twitching motility and evolutionary history across diverse bacteria. FEMS Microbiol Lett 362: 1–9. doi:10.1093/femsle/fnu063
OpenUrl CrossRef PubMed
↵
1. Perez-Riba A,
2. Synakewicz M,
3. Itzhaki LS
(2018) Folding cooperativity and allosteric function in the tandem-repeat protein class. Philos Trans R Soc Lond B Biol Sci 373: 20170188. doi:10.1098/rstb.2017.0188
OpenUrl CrossRef PubMed
↵
1. Persat A
(2017) Bacterial mechanotransduction. Curr Opin Microbiol 36: 1–6. doi:10.1016/j.mib.2016.12.002
OpenUrl CrossRef PubMed
↵
1. Persat A,
2. Nadell CD,
3. Kim MK,
4. Ingremeau F,
5. Siryaporn A,
6. Drescher K,
7. Wingreen NS,
8. Bassler BL,
9. Gitai Z,
10. Stone HA
(2015) The mechanical world of bacteria. Cell 161: 988–997. doi:10.1016/j.cell.2015.05.005
OpenUrl CrossRef PubMed
↵
1. Popa I,
2. Berkovich R,
3. Alegre-Cebollada J,
4. Badilla CL,
5. Rivas-Pardo JA,
6. Taniguchi Y,
7. Kawakami M,
8. Fernandez JM
(2013) Nanomechanics of HaloTag tethers. J Am Chem Soc 135: 12762–12771. doi:10.1021/ja4056382
OpenUrl CrossRef PubMed
↵
1. Popa I,
2. Rivas-Pardo JA,
3. Eckels EC,
4. Echelman DJ,
5. Badilla CL,
6. Valle-Orero J,
7. Fernández JM
(2016) A HaloTag anchored ruler for week-long studies of protein dynamics. J Am Chem Soc 138: 10546–10553. doi:10.1021/jacs.6b05429
OpenUrl CrossRef PubMed
↵
1. Ribbe J,
2. Baker AE,
3. Euler S,
4. O’Toole GA,
5. Maier B
(2017) Role of cyclic di-GMP and exopolysaccharide in type IV pilus dynamics. J Bacteriol 199: e00859-16. doi:10.1128/JB.00859-16
OpenUrl Abstract/FREE Full Text
↵
1. Rodesney CA,
2. Roman B,
3. Dhamani N,
4. Cooley BJ,
5. Katira P,
6. Touhami A,
7. Gordon VD
(2017) Mechanosensing of shear by Pseudomonas aeruginosa leads to increased levels of the cyclic-di-GMP signal initiating biofilm development. Proc Natl Acad Sci U S A 114: 5906–5911. doi:10.1073/pnas.1703255114
OpenUrl Abstract/FREE Full Text
↵
1. Santorelli D,
2. Marcocci L,
3. Pennacchietti V,
4. Nardella C,
5. Diop A,
6. Pietrangeli P,
7. Pagano L,
8. Toto A,
9. Malagrinò F,
10. Gianni S
(2023) Understanding the molecular basis of folding cooperativity through a comparative analysis of a multidomain protein and its isolated domains. J Biol Chem 299: 102983. doi:10.1016/j.jbc.2023.102983
OpenUrl CrossRef
↵
1. Scholl ZN,
2. Li Q,
3. Yang W,
4. Marszalek PE
(2016) Single-molecule force spectroscopy reveals the calcium dependence of the alternative conformations in the native state of a βγ-crystallin protein. J Biol Chem 291: 18263–18275. doi:10.1074/jbc.M116.729525
OpenUrl Abstract/FREE Full Text
↵
1. Schönfelder J,
2. Perez-Jimenez R,
3. Muñoz V
(2016) A simple two-state protein unfolds mechanically via multiple heterogeneous pathways at single-molecule resolution. Nat Commun 7: 11777. doi:10.1038/ncomms11777
OpenUrl CrossRef PubMed
↵
1. Siryaporn A,
2. Kuchma SL,
3. O’Toole GA,
4. Gitai Z
(2014) Surface attachment induces Pseudomonas aeruginosa virulence. Proc Natl Acad Sci U S A 111: 16860–16865. doi:10.1073/pnas.1415712111
OpenUrl Abstract/FREE Full Text
↵
1. Smith JW,
2. Vestal DJ,
3. Irwin SV,
4. Burke TA,
5. Cheresh DA
(1990) Purification and functional characterization of integrin alpha v beta 5. An adhesion receptor for vitronectin. J Biol Chem 265: 11008–11013. doi:10.1016/S0021-9258(19)38549-7
OpenUrl Abstract/FREE Full Text
↵
1. Taniguchi Y,
2. Kawakami M
(2010) Application of Halotag protein to covalent immobilization of recombinant proteins for single molecule force spectroscopy. Langmuir 26: 10433–10436. doi:10.1021/la101658a
OpenUrl CrossRef PubMed
↵
1. Tapia-Rojo R,
2. Eckels EC,
3. Fernández JM
(2019) Ephemeral states in protein folding under force captured with a magnetic tweezers design. Proc Natl Acad Sci U S A 116: 7873–7878. doi:10.1073/pnas.1821284116
OpenUrl Abstract/FREE Full Text
↵
1. Tapia-Rojo R,
2. Alonso-Caballero A,
3. Fernandez JM
(2020) Direct observation of a coil-to-helix contraction triggered by vinculin binding to talin. Sci Adv 6: eaaz4707. doi:10.1126/sciadv.aaz4707
OpenUrl FREE Full Text
↵
1. Tapia-Rojo R,
2. Alonso-Caballero A,
3. Badilla CL,
4. Fernandez JM
(2024) Identical sequences, different behaviors: Protein diversity captured at the single-molecule level. Biophys J 123: 814–823. doi:10.1016/j.bpj.2024.02.020
OpenUrl CrossRef
↵
1. R Core Team
(2023) R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. Available at: https://www.R-project.org/.
↵
1. Treuner-Lange A,
2. Chang YW,
3. Glatter T,
4. Herfurth M,
5. Lindow S,
6. Chreifi G,
7. Jensen GJ,
8. Søgaard-Andersen L
(2020) PilY1 and minor pilins form a complex priming the type IVa pilus in Myxococcus xanthus. Nat Commun 11: 5054–5114. doi:10.1038/s41467-020-18803-z
OpenUrl CrossRef PubMed
↵
1. Trosset MW
(2009) An Introduction to Statistical Inference and its Applications with R, 1st edn. Boca Raton, Florida: Chapman and Hall/CRC Press. doi:10.1201/9781584889489
OpenUrl CrossRef
↵
1. Varadi M,
2. Anyango S,
3. Deshpande M,
4. Nair S,
5. Natassia C,
6. Yordanova G,
7. Yuan D,
8. Stroe O,
9. Wood G,
10. Laydon A, et al.
(2022) AlphaFold Protein Structure Database: Massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Res 50: D439–D444. doi:10.1093/nar/gkab1061
OpenUrl CrossRef PubMed
↵
1. Wang L,
2. Wong Y,
3. Correira JM,
4. Wancura M,
5. Geiger CJ,
6. Webster SS,
7. Touhami A,
8. Butler BJ,
9. O’Toole GA,
10. Langford RM, et al.
(2023) The accumulation and growth of Pseudomonas aeruginosa on surfaces is modulated by surface mechanics via cyclic-di-GMP signaling. NPJ Biofilms Microbiomes 9: 78. doi:10.1038/s41522-023-00436-x
OpenUrl CrossRef
↵
1. Webster SS,
2. Mathelié-Guinlet M,
3. Verissimo AF,
4. Schultz D,
5. Viljoen A,
6. Lee CK,
7. Schmidt WC,
8. Wong GCL,
9. Dufrêne YF,
10. O’Toole GA
(2021) Force-induced changes of PilY1 drive surface sensing by Pseudomonas aeruginosa. mBio 13: e0375421. doi:10.1128/mbio.03754-21
OpenUrl CrossRef PubMed
↵
1. Yao M,
2. Goult BT,
3. Chen H,
4. Cong P,
5. Sheetz MP,
6. Yan J
(2014) Mechanical activation of vinculin binding to talin locks talin in an unfolded conformation. Sci Rep 4: 4610. doi:10.1038/srep04610
OpenUrl CrossRef PubMed
↵
1. Yuan G,
2. Liu H,
3. Ma Q,
4. Li X,
5. Nie J,
6. Zuo J,
7. Zheng P
(2019) Single-molecule force spectroscopy reveals that iron-ligand bonds modulate proteins in different modes. J Phys Chem Lett 10: 5428–5433. doi:10.1021/acs.jpclett.9b01573
OpenUrl CrossRef
↵
1. Zhao X,
2. Zeng X,
3. Lu C,
4. Yan J
(2017) Studying the mechanical responses of proteins using magnetic tweezers. Nanotechnology 28: 414002. doi:10.1088/1361-6528/aa837e
OpenUrl CrossRef PubMed

Download PDF

Email Article

Citation Tools

Tweet Widget

Subjects

Cited By...

No citing articles found.

Google Scholar

More in this TOC Section

Show more Research Article

[1] ↵
Alonso-Caballero A,
Schönfelder J,
Poly S,
Corsetti F,
De Sancho D,
Artacho E,
Perez-Jimenez R
(2018) Mechanical architecture and folding of E. coli type 1 pilus domains. Nat Commun 9: 2758. doi:10.1038/s41467-018-05107-6
OpenUrl CrossRef PubMed

[2] Alonso-Caballero A,

[3] Schönfelder J,

[4] Poly S,

[5] Corsetti F,

[6] De Sancho D,

[7] Artacho E,

[8] Perez-Jimenez R

[9] ↵
Alonso-Caballero A,
Echelman DJ,
Tapia-Rojo R,
Haldar S,
Eckels EC,
Fernandez JM
(2021) Protein folding modulates the chemical reactivity of a Gram-positive adhesin. Nat Chem 13: 172–181. doi:10.1038/s41557-020-00586-x
OpenUrl CrossRef PubMed

[10] Alonso-Caballero A,

[11] Echelman DJ,

[12] Tapia-Rojo R,

[13] Haldar S,

[14] Eckels EC,

[15] Fernandez JM

[16] ↵
Batey S,
Clarke J
(2006) Apparent cooperativity in the folding of multidomain proteins depends on the relative rates of folding of the constituent domains. Proc Natl Acad Sci U S A 103: 18113–18118. doi:10.1073/pnas.0604580103
OpenUrl Abstract/FREE Full Text

[17] Batey S,

[18] Clarke J

[19] ↵
Bell GI
(1978) Models for the specific adhesion of cells to cells. Science 200: 618–627. doi:10.1126/science.347575
OpenUrl Abstract/FREE Full Text

[20] Bell GI

[21] ↵
Cao Y,
Li YD,
Li H
(2011) Enhancing the mechanical stability of proteins through a cocktail approach. Biophys J 100: 1794–1799. doi:10.1016/j.bpj.2011.02.030
OpenUrl CrossRef PubMed

[22] Cao Y,

[23] Li YD,

[24] Li H

[25] ↵
Chawla R,
Gupta R,
Lele TP,
Lele PP
(2020) A skeptic’s guide to bacterial mechanosensing. J Mol Biol 432: 523–533. doi:10.1016/j.jmb.2019.09.004
OpenUrl CrossRef PubMed

[26] Chawla R,

[27] Gupta R,

[28] Lele TP,

[29] Lele PP

[30] ↵
Craig L,
Forest KT,
Maier B
(2019) Type IV pili: Dynamics, biophysics and functional consequences. Nat Rev Microbiol 17: 429–440. doi:10.1038/s41579-019-0195-4
OpenUrl CrossRef PubMed

[31] Craig L,

[32] Forest KT,

[33] Maier B

[34] ↵
Cruz LF,
Cobine PA,
De La Fuente L
(2012) Calcium increases Xylella fastidiosa surface attachment, biofilm formation, and twitching motility. Appl Environ Microbiol 78: 1321–1331. doi:10.1128/AEM.06501-11
OpenUrl Abstract/FREE Full Text

[35] Cruz LF,

[36] Cobine PA,

[37] De La Fuente L

[38] ↵
Cruz LF,
Parker JK,
Cobine PA,
De La Fuente L
(2014) Calcium-enhanced twitching motility in xylella fastidiosa is linked to a single PilY1 homolog. Appl Environ Microbiol 80: 7176–7185. doi:10.1128/AEM.02153-14
OpenUrl Abstract/FREE Full Text

[39] Cruz LF,

[40] Parker JK,

[41] Cobine PA,

[42] De La Fuente L

[43] ↵
Dufrêne YF,
Persat A
(2020) Mechanomicrobiology: How bacteria sense and respond to forces. Nat Rev Microbiol 18: 227–240. doi:10.1038/s41579-019-0314-2
OpenUrl CrossRef PubMed

[44] Dufrêne YF,

[45] Persat A

[46] ↵
Echelman DJ,
Alegre-Cebollada J,
Badilla CL,
Chang C,
Ton-That H,
Fernández JM
(2016) CnaA domains in bacterial pili are efficient dissipaters of large mechanical shocks. Proc Natl Acad Sci U S A 113: 2490–2495. doi:10.1073/pnas.1522946113
OpenUrl Abstract/FREE Full Text

[47] Echelman DJ,

[48] Alegre-Cebollada J,

[49] Badilla CL,

[50] Chang C,

[51] Ton-That H,

[52] Fernández JM

[53] ↵
Eyckmans J,
Boudou T,
Yu X,
Chen CS
(2011) A hitchhiker’s guide to mechanobiology. Dev Cell 21: 35–47. doi:10.1016/j.devcel.2011.06.015
OpenUrl CrossRef PubMed

[54] Eyckmans J,

[55] Boudou T,

[56] Yu X,

[57] Chen CS

[58] ↵
Flory PJ
(1956) Theory of elastic mechanisms in fibrous proteins. J Am Chem Soc 78: 5222–5235. doi:10.1021/ja01601a025
OpenUrl CrossRef

[59] Flory PJ

[60] ↵
Gordon VD,
Wang L
(2019) Bacterial mechanosensing: The force will be with you, always. J Cell Sci 132: jcs227694. doi:10.1242/jcs.227694
OpenUrl Abstract/FREE Full Text

[61] Gordon VD,

[62] Wang L

[63] ↵
Gosse C,
Croquette V
(2002) Magnetic tweezers: Micromanipulation and force measurement at the molecular level. Biophys J 82: 3314–3329. doi:10.1016/S0006-3495(02)75672-5
OpenUrl CrossRef PubMed

[64] Gosse C,

[65] Croquette V

[66] ↵
Guo S,
Chang Y,
Brun YV,
Howell PL,
Burrows LL,
Liu J
(2024) PilY1 regulates the dynamic architecture of the type IV pilus machine in Pseudomonas aeruginosa. Nat Commun 15: 9382. doi:10.1038/s41467-024-53638-y
OpenUrl CrossRef PubMed

[67] Guo S,

[68] Chang Y,

[69] Brun YV,

[70] Howell PL,

[71] Burrows LL,

[72] Liu J

[73] ↵
Heiniger RW,
Winther-Larsen HC,
Pickles RJ,
Koomey M,
Wolfgang MC
(2010) Infection of human mucosal tissue by Pseudomonas aeruginosa requires sequential and mutually dependent virulence factors and a novel pilus-associated adhesin. Cell Microbiol 12: 1158–1173. doi:10.1111/j.1462-5822.2010.01461.x
OpenUrl CrossRef PubMed

[74] Heiniger RW,

[75] Winther-Larsen HC,

[76] Pickles RJ,

[77] Koomey M,

[78] Wolfgang MC

[79] ↵
Herfurth M,
Treuner-Lange A,
Glatter T,
Wittmaack N,
Hoiczyk E,
Pierik AJ,
Søgaard-Andersen L
(2022) A noncanonical cytochrome c stimulates calcium binding by PilY1 for type IVa pili formation. Proc Natl Acad Sci U S A 119: e2115061119. doi:10.1073/pnas.2115061119
OpenUrl Abstract/FREE Full Text

[80] Herfurth M,

[81] Treuner-Lange A,

[82] Glatter T,

[83] Wittmaack N,

[84] Hoiczyk E,

[85] Pierik AJ,

[86] Søgaard-Andersen L

[87] ↵
Hernández-Sánchez A,
Páez-Pérez ED,
Alfaro-Saldaña E,
García-Meza JV
(2024) Deciphering the enigmatic PilY1 of Acidithiobacillus thiooxidans: An in silico analysis. Biochem Biophys Rep 39: 101797. doi:10.1016/j.bbrep.2024.101797
OpenUrl CrossRef

[88] Hernández-Sánchez A,

[89] Páez-Pérez ED,

[90] Alfaro-Saldaña E,

[91] García-Meza JV

[92] ↵
Hoppe J,
Ünal CM,
Thiem S,
Grimpe L,
Goldmann T,
Gaßler N,
Richter M,
Shevchuk O,
Steinert M
(2017) PilY1 promotes Legionella pneumophila infection of human lung tissue explants and contributes to bacterial adhesion, host cell invasion, and twitching motility. Front Cell Infect Microbiol 7: 63. doi:10.3389/fcimb.2017.00063
OpenUrl CrossRef PubMed

[93] Hoppe J,

[94] Ünal CM,

[95] Thiem S,

[96] Grimpe L,

[97] Goldmann T,

[98] Gaßler N,

[99] Richter M,

[100] Shevchuk O,

[101] Steinert M

[102] ↵
Johnson MDL,
Garrett CK,
Bond JE,
Coggan KA,
Wolfgang MC,
Redinbo MR
(2011) Pseudomonas aeruginosa PilY1 binds integrin in an RGD- and calcium-dependent manner. PLoS One 6: e29629. doi:10.1371/journal.pone.0029629
OpenUrl CrossRef PubMed

[103] Johnson MDL,

[104] Garrett CK,

[105] Bond JE,

[106] Coggan KA,

[107] Wolfgang MC,

[108] Redinbo MR

[109] ↵
Jumper J,
Evans R,
Pritzel A,
Green T,
Figurnov M,
Ronneberger O,
Tunyasuvunakool K,
Bates R,
Žídek A,
Potapenko A, et al.
(2021) Highly accurate protein structure prediction with AlphaFold. Nature 596: 583–589. doi:10.1038/s41586-021-03819-2
OpenUrl CrossRef PubMed

[110] Jumper J,

[111] Evans R,

[112] Pritzel A,

[113] Green T,

[114] Figurnov M,

[115] Ronneberger O,

[116] Tunyasuvunakool K,

[117] Bates R,

[118] Žídek A,

[119] Potapenko A, et al.

[120] ↵
Jurado-Martín I,
Sainz-Mejías M,
McClean S
(2021) Pseudomonas aeruginosa: An audacious pathogen with an adaptable arsenal of virulence factors. Int J Mol Sci 22: 3128–3137. doi:10.3390/ijms22063128
OpenUrl CrossRef PubMed

[121] Jurado-Martín I,

[122] Sainz-Mejías M,

[123] McClean S

[124] ↵
Kassambara A,
Mundt F
(2020) factoextra: Extract and visualize the results of multivariate data analyses. R package version 1.0.7.999. Available at: https://github.com/kassambara/factoextra.

[125] Kassambara A,

[126] Mundt F

[127] ↵
Kuchma SL,
Ballok AE,
Merritt JH,
Hammond JH,
Lu W,
Rabinowitz JD,
O’Toole GA
(2010) Cyclic-di-GMP-mediated repression of swarming motility by Pseudomonas aeruginosa: The pilY1 gene and its impact on surface-associated behaviors. J Bacteriol 192: 2950–2964. doi:10.1128/JB.01642-09
OpenUrl Abstract/FREE Full Text

[128] Kuchma SL,

[129] Ballok AE,

[130] Merritt JH,

[131] Hammond JH,

[132] Lu W,

[133] Rabinowitz JD,

[134] O’Toole GA

[135] ↵
Lele PP,
Hosu BG,
Berg HC
(2013) Dynamics of mechanosensing in the bacterial flagellar motor. Proc Natl Acad Sci U S A 110: 11839–11844. doi:10.1073/pnas.1305885110
OpenUrl Abstract/FREE Full Text

[136] Lele PP,

[137] Hosu BG,

[138] Berg HC

[139] ↵
Li H,
Oberhauser AF,
Fowler SB,
Clarke J,
Fernandez JM
(2000) Atomic force microscopy reveals the mechanical design of a modular protein. Proc Natl Acad Sci U S A 97: 6527–6531. doi:10.1073/pnas.120048697
OpenUrl Abstract/FREE Full Text

[140] Li H,

[141] Oberhauser AF,

[142] Fowler SB,

[143] Clarke J,

[144] Fernandez JM

[145] ↵
Li H,
Linke WA,
Oberhauser AF,
Carrion-Vazquez M,
Kerkvliet JG,
Lu H,
Marszalek PE,
Fernandez JM
(2002) Reverse engineering of the giant muscle protein titin. Nature 418: 998–1002. doi:10.1038/nature00938
OpenUrl CrossRef PubMed

[146] Li H,

[147] Linke WA,

[148] Oberhauser AF,

[149] Carrion-Vazquez M,

[150] Kerkvliet JG,

[151] Lu H,

[152] Marszalek PE,

[153] Fernandez JM

[154] ↵
Löf A,
Walker PU,
Sedlak SM,
Gruber S,
Obser T,
Brehm MA,
Benoit M,
Lipfert J
(2019) Multiplexed protein force spectroscopy reveals equilibrium protein folding dynamics and the low-force response of von Willebrand factor. Proc Natl Acad Sci U S A 116: 18798–18807. doi:10.1073/pnas.1901794116
OpenUrl Abstract/FREE Full Text

[155] Löf A,

[156] Walker PU,

[157] Sedlak SM,

[158] Gruber S,

[159] Obser T,

[160] Brehm MA,

[161] Benoit M,

[162] Lipfert J

[163] ↵
Luo Y,
Zhao K,
Baker AE,
Kuchma SL,
Coggan KA,
Wolfgang MC,
Wong GCL,
O’Toole GA
(2015) A hierarchical cascade of second messengers regulates Pseudomonas aeruginosa surface behaviors. mBio 6: e02456-14. doi:10.1128/mBio.02456-14
OpenUrl Abstract/FREE Full Text

[164] Luo Y,

[165] Zhao K,

[166] Baker AE,

[167] Kuchma SL,

[168] Coggan KA,

[169] Wolfgang MC,

[170] Wong GCL,

[171] O’Toole GA

[172] ↵
Maechler M,
Rousseeuw P,
Struyf A,
Hubert M,
Hornik K
(2024) cluster: Cluster analysis basics and extensions. R package version 2.1.8. Available at: https://CRAN.R-project.org/package=cluster.

[173] Maechler M,

[174] Rousseeuw P,

[175] Struyf A,

[176] Hubert M,

[177] Hornik K

[178] ↵
Maier B,
Potter L,
So M,
Long CD,
Seifert HS,
Sheetz MP
(2002) Single pilus motor forces exceed 100 pN. Proc Natl Acad Sci U S A 99: 16012–16017. doi:10.1073/pnas.242523299
OpenUrl Abstract/FREE Full Text

[179] Maier B,

[180] Potter L,

[181] So M,

[182] Long CD,

[183] Seifert HS,

[184] Sheetz MP

[185] ↵
Marko VA,
Kilmury SLN,
MacNeil LT,
Burrows LL
(2018) Pseudomonas aeruginosa type IV minor pilins and PilY1 regulate virulence by modulating FimS-AlgR activity. PLoS Pathog 14: e1007074. doi:10.1371/journal.ppat.1007074
OpenUrl CrossRef PubMed

[186] Marko VA,

[187] Kilmury SLN,

[188] MacNeil LT,

[189] Burrows LL

[190] ↵
Milles LF,
Unterauer EM,
Nicolaus T,
Gaub HE
(2018) Calcium stabilizes the strongest protein fold. Nat Commun 9: 4764. doi:10.1038/s41467-018-07145-6
OpenUrl CrossRef

[191] Milles LF,

[192] Unterauer EM,

[193] Nicolaus T,

[194] Gaub HE

[195] ↵
Morand PC,
Bille E,
Morelle S,
Eugène E,
Beretti JL,
Wolfgang M,
Meyer TF,
Koomey M,
Nassif X
(2004) Type IV pilus retraction in pathogenic Neisseria is regulated by the PilC proteins. EMBO J 23: 2009–2017. doi:10.1038/sj.emboj.7600200
OpenUrl Abstract/FREE Full Text

[196] Morand PC,

[197] Bille E,

[198] Morelle S,

[199] Eugène E,

[200] Beretti JL,

[201] Wolfgang M,

[202] Meyer TF,

[203] Koomey M,

[204] Nassif X

[205] ↵
Mordue J,
O’Boyle N,
Gadegaard N,
Roe AJ
(2021) The force awakens: The dark side of mechanosensing in bacterial pathogens. Cell Signal 78: 109867. doi:10.1016/j.cellsig.2020.109867
OpenUrl CrossRef

[206] Mordue J,

[207] O’Boyle N,

[208] Gadegaard N,

[209] Roe AJ

[210] ↵
Mylemans B,
Voet AR,
Tame JR
(2021) The taming of the screw: The natural and artificial development of β-propeller proteins. Curr Opin Struct Biol 68: 48–54. doi:10.1016/j.sbi.2020.11.009
OpenUrl CrossRef

[211] Mylemans B,

[212] Voet AR,

[213] Tame JR

[214] ↵
Nguyen Y,
Harvey H,
Sugiman-Marangos S,
Bell SD,
Buensuceso RNC,
Junop MS,
Burrows LL
(2015) Structural and functional studies of the Pseudomonas aeruginosa minor pilin, PilE. J Biol Chem 290: 26856–26865. doi:10.1074/jbc.M115.683334
OpenUrl Abstract/FREE Full Text

[215] Nguyen Y,

[216] Harvey H,

[217] Sugiman-Marangos S,

[218] Bell SD,

[219] Buensuceso RNC,

[220] Junop MS,

[221] Burrows LL

[222] ↵
Nie J-Y,
Song G-B,
Deng Y-B,
Zheng P
(2022) Single-molecule force spectroscopy reveals stability of mitoNEET and its [2Fe2Se] cluster in weakly acidic and basic solutions. ChemistryOpen 11: e202200056. doi:10.1002/open.202200056
OpenUrl CrossRef

[223] Nie J-Y,

[224] Song G-B,

[225] Deng Y-B,

[226] Zheng P

[227] ↵
Oberhauser AF,
Marszalek PE,
Carrion-Vazquez M,
Fernandez JM
(1999) Single protein misfolding events captured by atomic force microscopy. Nat Struct Biol 6: 1025–1028. doi:10.1038/14907
OpenUrl CrossRef PubMed

[228] Oberhauser AF,

[229] Marszalek PE,

[230] Carrion-Vazquez M,

[231] Fernandez JM

[232] ↵
Orans J,
Johnson MDL,
Coggan KA,
Sperlazza JR,
Heiniger RW,
Wolfgang MC,
Redinbo MR
(2010) Crystal structure analysis reveals Pseudomonas PilY1 as an essential calcium-dependent regulator of bacterial surface motility. Proc Natl Acad Sci U S A 107: 1065–1070. doi:10.1073/pnas.0911616107
OpenUrl Abstract/FREE Full Text

[233] Orans J,

[234] Johnson MDL,

[235] Coggan KA,

[236] Sperlazza JR,

[237] Heiniger RW,

[238] Wolfgang MC,

[239] Redinbo MR

[240] ↵
Oude Vrielink AS,
Vance TDR,
de Jong AM,
Davies PL,
Voets IK
(2017) Unusually high mechanical stability of bacterial adhesin extender domains having calcium clamps. PLoS One 12: e0174682. doi:10.1371/journal.pone.0174682
OpenUrl CrossRef PubMed

[241] Oude Vrielink AS,

[242] Vance TDR,

[243] de Jong AM,

[244] Davies PL,

[245] Voets IK

[246] ↵
O’Toole GA,
Kolter R
(1998) Flagellar and twitching motility are necessary for Pseudomonas aeruginosa biofilm development. Mol Microbiol 30: 295–304. doi:10.1046/j.1365-2958.1998.01062.x
OpenUrl CrossRef PubMed

[247] O’Toole GA,

[248] Kolter R

[249] ↵
Parker JK,
Cruz LF,
Evans MR,
De La Fuente L
(2015) Presence of calcium-binding motifs in PilY1 homologs correlates with Ca-mediated twitching motility and evolutionary history across diverse bacteria. FEMS Microbiol Lett 362: 1–9. doi:10.1093/femsle/fnu063
OpenUrl CrossRef PubMed

[250] Parker JK,

[251] Cruz LF,

[252] Evans MR,

[253] De La Fuente L

[254] ↵
Perez-Riba A,
Synakewicz M,
Itzhaki LS
(2018) Folding cooperativity and allosteric function in the tandem-repeat protein class. Philos Trans R Soc Lond B Biol Sci 373: 20170188. doi:10.1098/rstb.2017.0188
OpenUrl CrossRef PubMed

[255] Perez-Riba A,

[256] Synakewicz M,

[257] Itzhaki LS

[258] ↵
Persat A
(2017) Bacterial mechanotransduction. Curr Opin Microbiol 36: 1–6. doi:10.1016/j.mib.2016.12.002
OpenUrl CrossRef PubMed

[259] Persat A

[260] ↵
Persat A,
Nadell CD,
Kim MK,
Ingremeau F,
Siryaporn A,
Drescher K,
Wingreen NS,
Bassler BL,
Gitai Z,
Stone HA
(2015) The mechanical world of bacteria. Cell 161: 988–997. doi:10.1016/j.cell.2015.05.005
OpenUrl CrossRef PubMed

[261] Persat A,

[262] Nadell CD,

[263] Kim MK,

[264] Ingremeau F,

[265] Siryaporn A,

[266] Drescher K,

[267] Wingreen NS,

[268] Bassler BL,

[269] Gitai Z,

[270] Stone HA

[271] ↵
Popa I,
Berkovich R,
Alegre-Cebollada J,
Badilla CL,
Rivas-Pardo JA,
Taniguchi Y,
Kawakami M,
Fernandez JM
(2013) Nanomechanics of HaloTag tethers. J Am Chem Soc 135: 12762–12771. doi:10.1021/ja4056382
OpenUrl CrossRef PubMed

[272] Popa I,

[273] Berkovich R,

[274] Alegre-Cebollada J,

[275] Badilla CL,

[276] Rivas-Pardo JA,

[277] Taniguchi Y,

[278] Kawakami M,

[279] Fernandez JM

[280] ↵
Popa I,
Rivas-Pardo JA,
Eckels EC,
Echelman DJ,
Badilla CL,
Valle-Orero J,
Fernández JM
(2016) A HaloTag anchored ruler for week-long studies of protein dynamics. J Am Chem Soc 138: 10546–10553. doi:10.1021/jacs.6b05429
OpenUrl CrossRef PubMed

[281] Popa I,

[282] Rivas-Pardo JA,

[283] Eckels EC,

[284] Echelman DJ,

[285] Badilla CL,

[286] Valle-Orero J,

[287] Fernández JM

[288] ↵
Ribbe J,
Baker AE,
Euler S,
O’Toole GA,
Maier B
(2017) Role of cyclic di-GMP and exopolysaccharide in type IV pilus dynamics. J Bacteriol 199: e00859-16. doi:10.1128/JB.00859-16
OpenUrl Abstract/FREE Full Text

[289] Ribbe J,

[290] Baker AE,

[291] Euler S,

[292] O’Toole GA,

[293] Maier B

[294] ↵
Rodesney CA,
Roman B,
Dhamani N,
Cooley BJ,
Katira P,
Touhami A,
Gordon VD
(2017) Mechanosensing of shear by Pseudomonas aeruginosa leads to increased levels of the cyclic-di-GMP signal initiating biofilm development. Proc Natl Acad Sci U S A 114: 5906–5911. doi:10.1073/pnas.1703255114
OpenUrl Abstract/FREE Full Text

[295] Rodesney CA,

[296] Roman B,

[297] Dhamani N,

[298] Cooley BJ,

[299] Katira P,

[300] Touhami A,

[301] Gordon VD

[302] ↵
Santorelli D,
Marcocci L,
Pennacchietti V,
Nardella C,
Diop A,
Pietrangeli P,
Pagano L,
Toto A,
Malagrinò F,
Gianni S
(2023) Understanding the molecular basis of folding cooperativity through a comparative analysis of a multidomain protein and its isolated domains. J Biol Chem 299: 102983. doi:10.1016/j.jbc.2023.102983
OpenUrl CrossRef

[303] Santorelli D,

[304] Marcocci L,

[305] Pennacchietti V,

[306] Nardella C,

[307] Diop A,

[308] Pietrangeli P,

[309] Pagano L,

[310] Toto A,

[311] Malagrinò F,

[312] Gianni S

[313] ↵
Scholl ZN,
Li Q,
Yang W,
Marszalek PE
(2016) Single-molecule force spectroscopy reveals the calcium dependence of the alternative conformations in the native state of a βγ-crystallin protein. J Biol Chem 291: 18263–18275. doi:10.1074/jbc.M116.729525
OpenUrl Abstract/FREE Full Text

[314] Scholl ZN,

[315] Li Q,

[316] Yang W,

[317] Marszalek PE

[318] ↵
Schönfelder J,
Perez-Jimenez R,
Muñoz V
(2016) A simple two-state protein unfolds mechanically via multiple heterogeneous pathways at single-molecule resolution. Nat Commun 7: 11777. doi:10.1038/ncomms11777
OpenUrl CrossRef PubMed

[319] Schönfelder J,

[320] Perez-Jimenez R,

[321] Muñoz V

[322] ↵
Siryaporn A,
Kuchma SL,
O’Toole GA,
Gitai Z
(2014) Surface attachment induces Pseudomonas aeruginosa virulence. Proc Natl Acad Sci U S A 111: 16860–16865. doi:10.1073/pnas.1415712111
OpenUrl Abstract/FREE Full Text

[323] Siryaporn A,

[324] Kuchma SL,

[325] O’Toole GA,

[326] Gitai Z

[327] ↵
Smith JW,
Vestal DJ,
Irwin SV,
Burke TA,
Cheresh DA
(1990) Purification and functional characterization of integrin alpha v beta 5. An adhesion receptor for vitronectin. J Biol Chem 265: 11008–11013. doi:10.1016/S0021-9258(19)38549-7
OpenUrl Abstract/FREE Full Text

[328] Smith JW,

[329] Vestal DJ,

[330] Irwin SV,

[331] Burke TA,

[332] Cheresh DA

[333] ↵
Taniguchi Y,
Kawakami M
(2010) Application of Halotag protein to covalent immobilization of recombinant proteins for single molecule force spectroscopy. Langmuir 26: 10433–10436. doi:10.1021/la101658a
OpenUrl CrossRef PubMed

[334] Taniguchi Y,

[335] Kawakami M

[336] ↵
Tapia-Rojo R,
Eckels EC,
Fernández JM
(2019) Ephemeral states in protein folding under force captured with a magnetic tweezers design. Proc Natl Acad Sci U S A 116: 7873–7878. doi:10.1073/pnas.1821284116
OpenUrl Abstract/FREE Full Text

[337] Tapia-Rojo R,

[338] Eckels EC,

[339] Fernández JM

[340] ↵
Tapia-Rojo R,
Alonso-Caballero A,
Fernandez JM
(2020) Direct observation of a coil-to-helix contraction triggered by vinculin binding to talin. Sci Adv 6: eaaz4707. doi:10.1126/sciadv.aaz4707
OpenUrl FREE Full Text

[341] Tapia-Rojo R,

[342] Alonso-Caballero A,

[343] Fernandez JM

[344] ↵
Tapia-Rojo R,
Alonso-Caballero A,
Badilla CL,
Fernandez JM
(2024) Identical sequences, different behaviors: Protein diversity captured at the single-molecule level. Biophys J 123: 814–823. doi:10.1016/j.bpj.2024.02.020
OpenUrl CrossRef

[345] Tapia-Rojo R,

[346] Alonso-Caballero A,

[347] Badilla CL,

[348] Fernandez JM

[349] ↵
R Core Team
(2023) R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. Available at: https://www.R-project.org/.

[350] R Core Team

[351] ↵
Treuner-Lange A,
Chang YW,
Glatter T,
Herfurth M,
Lindow S,
Chreifi G,
Jensen GJ,
Søgaard-Andersen L
(2020) PilY1 and minor pilins form a complex priming the type IVa pilus in Myxococcus xanthus. Nat Commun 11: 5054–5114. doi:10.1038/s41467-020-18803-z
OpenUrl CrossRef PubMed

[352] Treuner-Lange A,

[353] Chang YW,

[354] Glatter T,

[355] Herfurth M,

[356] Lindow S,

[357] Chreifi G,

[358] Jensen GJ,

[359] Søgaard-Andersen L

[360] ↵
Trosset MW
(2009) An Introduction to Statistical Inference and its Applications with R, 1st edn. Boca Raton, Florida: Chapman and Hall/CRC Press. doi:10.1201/9781584889489
OpenUrl CrossRef

[361] Trosset MW

[362] ↵
Varadi M,
Anyango S,
Deshpande M,
Nair S,
Natassia C,
Yordanova G,
Yuan D,
Stroe O,
Wood G,
Laydon A, et al.
(2022) AlphaFold Protein Structure Database: Massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Res 50: D439–D444. doi:10.1093/nar/gkab1061
OpenUrl CrossRef PubMed

[363] Varadi M,

[364] Anyango S,

[365] Deshpande M,

[366] Nair S,

[367] Natassia C,

[368] Yordanova G,

[369] Yuan D,

[370] Stroe O,

[371] Wood G,

[372] Laydon A, et al.

[373] ↵
Wang L,
Wong Y,
Correira JM,
Wancura M,
Geiger CJ,
Webster SS,
Touhami A,
Butler BJ,
O’Toole GA,
Langford RM, et al.
(2023) The accumulation and growth of Pseudomonas aeruginosa on surfaces is modulated by surface mechanics via cyclic-di-GMP signaling. NPJ Biofilms Microbiomes 9: 78. doi:10.1038/s41522-023-00436-x
OpenUrl CrossRef

[374] Wang L,

[375] Wong Y,

[376] Correira JM,

[377] Wancura M,

[378] Geiger CJ,

[379] Webster SS,

[380] Touhami A,

[381] Butler BJ,

[382] O’Toole GA,

[383] Langford RM, et al.

[384] ↵
Webster SS,
Mathelié-Guinlet M,
Verissimo AF,
Schultz D,
Viljoen A,
Lee CK,
Schmidt WC,
Wong GCL,
Dufrêne YF,
O’Toole GA
(2021) Force-induced changes of PilY1 drive surface sensing by Pseudomonas aeruginosa. mBio 13: e0375421. doi:10.1128/mbio.03754-21
OpenUrl CrossRef PubMed

[385] Webster SS,

[386] Mathelié-Guinlet M,

[387] Verissimo AF,

[388] Schultz D,

[389] Viljoen A,

[390] Lee CK,

[391] Schmidt WC,

[392] Wong GCL,

[393] Dufrêne YF,

[394] O’Toole GA

[395] ↵
Yao M,
Goult BT,
Chen H,
Cong P,
Sheetz MP,
Yan J
(2014) Mechanical activation of vinculin binding to talin locks talin in an unfolded conformation. Sci Rep 4: 4610. doi:10.1038/srep04610
OpenUrl CrossRef PubMed

[396] Yao M,

[397] Goult BT,

[398] Chen H,

[399] Cong P,

[400] Sheetz MP,

[401] Yan J

[402] ↵
Yuan G,
Liu H,
Ma Q,
Li X,
Nie J,
Zuo J,
Zheng P
(2019) Single-molecule force spectroscopy reveals that iron-ligand bonds modulate proteins in different modes. J Phys Chem Lett 10: 5428–5433. doi:10.1021/acs.jpclett.9b01573
OpenUrl CrossRef

[403] Yuan G,

[404] Liu H,

[405] Ma Q,

[406] Li X,

[407] Nie J,

[408] Zuo J,

[409] Zheng P

[410] ↵
Zhao X,
Zeng X,
Lu C,
Yan J
(2017) Studying the mechanical responses of proteins using magnetic tweezers. Nanotechnology 28: 414002. doi:10.1088/1361-6528/aa837e
OpenUrl CrossRef PubMed

[411] Zhao X,

[412] Zeng X,

[413] Lu C,

[414] Yan J

Main menu

User menu

Search

Mechanical forces and ligand binding modulate Pseudomonas aeruginosa PilY1 mechanosensitive protein

Abstract

Introduction

Results

Hierarchical unfolding and Ca²⁺-dependent mechanical stability

Ca²⁺ binding alters the unfolding pathway

PilY1^501−1161 harbors a mechanosensing-like structure

The integrin-binding domain is the mechanosensitive structure of PilY1^501−1161

Integrin binding blocks the unfolding of I2

Discussion

Materials and Methods

Cloning and expression of chimeric constructs

Single-molecule magnetic tweezer measurements

Analysis

Data Availability

Acknowledgements

Author Contributions

Conflict of Interest Statement

References

Citation Manager Formats

In this Issue

Subjects

Related Articles

Cited By...

More in this TOC Section

Similar Articles

Content

For Authors

Other Services

More Information

Main menu

User menu

Search

Mechanical forces and ligand binding modulate Pseudomonas aeruginosa PilY1 mechanosensitive protein

Abstract

Introduction

Results

Hierarchical unfolding and Ca2+-dependent mechanical stability

Ca2+ binding alters the unfolding pathway

PilY1501−1161 harbors a mechanosensing-like structure

The integrin-binding domain is the mechanosensitive structure of PilY1501−1161

Integrin binding blocks the unfolding of I2

Discussion

Materials and Methods

Cloning and expression of chimeric constructs

Single-molecule magnetic tweezer measurements

Analysis

Data Availability

Acknowledgements

Author Contributions

Conflict of Interest Statement

References

Citation Manager Formats

In this Issue

Jump to section

Subjects

Related Articles

Cited By...

More in this TOC Section

Similar Articles

Content

For Authors

Other Services

More Information

Hierarchical unfolding and Ca²⁺-dependent mechanical stability

Ca²⁺ binding alters the unfolding pathway

PilY1^501−1161 harbors a mechanosensing-like structure

The integrin-binding domain is the mechanosensitive structure of PilY1^501−1161