The accumulation and growth of Pseudomonas aeruginosa on ...

Abstract

Attachment of bacteria onto a surface, consequent signaling, and accumulation and growth of the surface-bound bacterial population are key initial steps in the formation of pathogenic biofilms. While recent reports have hinted that surface mechanics may affect the accumulation of bacteria on that surface, the processes that underlie bacterial perception of surface mechanics and modulation of accumulation in response to surface mechanics remain largely unknown. We use thin and thick hydrogels coated on glass to create composite materials with different mechanics (higher elasticity for thin composites; lower elasticity for thick composites) but with the same surface adhesivity and chemistry. The mechanical cue stemming from surface mechanics is elucidated using experiments with the opportunistic human pathogen Pseudomonas aeruginosa combined with finite-element modeling. Adhesion to thin composites results in greater changes in mechanical stress and strain in the bacterial envelope than does adhesion to thick composites with identical surface chemistry. Using quantitative microscopy, we find that adhesion to thin composites also results in higher cyclic-di-GMP levels, which in turn result in lower motility and less detachment, and thus greater accumulation of bacteria on the surface than does adhesion to thick composites. Mechanics-dependent c-di-GMP production is mediated by the cell-surface-exposed protein PilY1. The biofilm lag phase, which is longer for bacterial populations on thin composites than on thick composites, is also mediated by PilY1. This study shows clear evidence that bacteria actively regulate differential accumulation on surfaces of different stiffnesses via perceiving varied mechanical stress and strain upon surface engagement.

Introduction

Mechanosensing, including but not limited to responding to surface mechanics, is well-established to be an important cellular function in eukaryotes^1,2. Much less is known about mechanosensing by prokaryotes^3,4,5. Some recent studies have shown that during early biofilm formation, bacteria can sense and respond to mechanical cues, such as those arising from contacting a surface^{6,7,8,9,10,11,12} and varying fluid flow over surface-bound bacteria^13,14. For the biofilm-forming pathogen Pseudomonas aeruginosa, previous research has shown two categories of sensing pathways. One is through cell envelope-associated proteins, the membrane-bound protein WspA, which might sense cell envelope stress upon surface attachment¹², and the cell-surface-exposed protein PilY1, which has been proposed as a possible mechanosensor of surface adhesion^10,13 and fluid shear¹³. PilY1 is localized at the outer membrane^9,10 and found at the tip of type-IV pili (TFP)¹⁵. The second is through the extension and retraction of TFP, which power the twitching motility of P. aeruginosa on surfaces and are contribute to bacterial mechanosensing of surfaces^8,11 and fluid shear¹³.

In vivo, bacteria can experience a wide range of surface mechanics, from ultrasoft (dermal fillers have elastic moduli 0.02–3 kPa and living tissues 0.2–30 kPa) to hard (orthopedic implants have elastic moduli 5–300 GPa)^16,17. In such diverse settings, biofilm formation commonly causes chronic infection, resulting in prolonged illness and high medical costs^18,19,20. Recent research has indicated connections between bacterial behavior and the physical properties of the substrates to which they are attached, as follows: Other researchers have shown that the extension and retraction of TFP of P. aeruginosa actively regulates virulence-related genes in a stiffness-dependent manner indicating by a PaQa reporter²¹. A second research group has shown that increasing gel substrate stiffness above ~30 kPa correlates with increases in the speed of TFP-driven twitching in P. aeruginosa²². However, using the same PaQa reporter, intracellular signaling was not found to correlate with changes in substrate gel compositions (i.e. changed stiffness)²². These findings imply that surface mechanics seem to affect bacteria behaviors in a complicated way, and we therefore need a better understanding of how bacteria perceive and respond to the mechanics of surfaces where they attach.

The initial accumulation of bacteria on surfaces generally lead to biofilm formation. Passive adhesion of bacteria (and other colloids) will be strongly impacted by the surface energy and how this is reduced by adhesion, to reduce the system's free energy. The hydrophobicity and hydrophilicity of surfaces can be major contributors to surface energy and thereby impact bacterial adhesion^{23,24,25,26,27}. The trend of adhesion on hydrophobic and hydrophilic surfaces is different²⁸. Electrostatics of the surface and the liquid medium, as well as van der Waals forces between bacteria and the surface, can also impact bacterial adhesion^{26,27,29,30,31}. Living bacteria are not, of course, passive colloids, and may have the biological ability to respond actively to surface properties other than surface energy.

Some research has shown that the accumulation of bacteria varied on surfaces with different mechanics^32,33,34,35. These earlier studies changed surface elasticity by varying characteristics such as cross-linking density or polymer concentration. Sometimes, an inappropriate fabrication may introduce unintended changes to other surface properties, resulting in unintended changes in surface energy, porosity, or the density of adhesion sites, and thereby impact adhesivity (see Supplementary discussion), which could act as a confounding factor to obscure the impact of surface elasticity. Perhaps as a result, the literature on the effect of surface mechanics on bacterial accumulation on surfaces does not show consistent trends^32,33,35. Furthermore, many questions remain about the processes underlying how bacteria modulate their accumulation in response to surface mechanics. These questions are of critical importance because they prevent the design of strategies for controlling early biofilm development by manipulating surface mechanics.

To address these questions, in the present study, we used thin and thick hydrogels coated on glass coverslips to create composite materials with the same gel type, adhesivity and surface chemistry, but with different effective stiffnesses, and monitored P. aeruginosa through the early stages of biofilm formation on these surfaces. First, surfaces were exposed to a suspension of bacteria for one hour before quantitative microscopy was used to measure bacterial accumulation. For the immediately-following stages of biofilm formation, characterized by bacteria reproducing on a surface rather than accumulating on the surface from a suspended (planktonic) population, we measured the duration and growth rate of the biofilm lag phase and exponential growth phase, respectively.

For both the accumulation and reproduction stages of biofilm development, we show that bacteria actively recognize and respond to surface mechanics. When bacteria initially attach to a surface, both finite element modeling and experimental measurements of the activity of mechanosensitive ion channels show that attachment to thin composites causes greater changes in the mechanical stress and strain state of the bacterial cell envelopes than does attachment to thick composite. We also find that attachment to thin composites results in higher levels of intracellular c-di-GMP, which leads to greater reduction in motility, a reduced likelihood of detachment, and, as a result, greater accumulation on thin composites. Once the initial accumulation stage has passed, higher levels of cyclic-di-GMP are associated with a longer biofilm lag phase on thin composites. Modulation of c-di-GMP levels in response to surface mechanics is mediated by the cell-surface-exposed protein PilY1, a proposed mechanosensor. In short, this work uses a combination of several imaging modalities, quantitative image analysis, and physical modelling to advance our understanding of the mechanism(s) of mechanical signaling for an important human pathogen.

Results

Adhesion to thin composites leads to greater changes in mechanical stress and strain in the bacterial cell envelope than does adhesion to thick composites

To eliminate effects arising from physicochemical properties of surfaces other than elasticity, such as adhesivity and surface chemistry, we used the same gel composition to fabricate thin and thick hydrogels atop glass coverslips (Fig. 1a and Supplementary Fig. 1a). Thickness measurements, done in triplicate for each hydrogel and thickness combination, found that thin gels were ~5 μm in height and thick gels were ~150 μm in height. We chose this geometry-based approach to modifying substrate mechanics to avoid inadvertently altering material adhesivity/chemistry along with mechanics, which has been observed before (Supplementary discussion) - for instance, poly(dimethylsiloxane) (PDMS) can have different surface adhesivities associated with different stiffnesses, shown by polymer beads found to accumulate more on soft PDMS than on stiff PDMS³⁶. Using Fourier-Transform Infrared Spectroscopy, we confirmed that the surface chemistry of thick and thin composites was the same (Fig. 1b and Supplementary Fig. 1b). To confirm that the composite materials had the same surface adhesivity regardless of gel thickness, we incubated both thin and thick hydrogel composites with a suspension of fluorescent polystyrene polymer beads for one hour, and imaged the number of beads attached using confocal microscopy. We verified that the numbers of polystyrene beads that attached did not significantly differ with hydrogel thickness (Fig. 1c and Supplementary Fig. 1c). Thus, we conclude that hydrogel thickness does not impact passive physicochemical adhesion to surfaces.

**Fig. 1: Fabrication of thin gel and thick gel-coverslip composites with different surface mechanics.**

However, the thickness of the hydrogel coated onto rigid glass coverslips does impact the mechanics of the resulting composite material. Linear elasticity theory was used to derive a closed-form expression for the effective elastic modulus (E_effective) of hydrogel-coverslip composites (Supplementary Discussion and Supplementary Equation 7). For a hydrogel thickness (t_gel) comparable to the 1 micron size of a bacterium, the E_effective increases sharply with decreasing t_gel (Fig. 1d). According to this model, the composites with thin (~ 5 μm) hydrogels are approximately 16 times stiffer than those with thick (~ 150 μm) – effective composite moduli are calculated at ~1 MPa for thin composites and ~50 kPa for thick composites (Fig. 1d and Supplementary Table 1 footnote). We also used a nanoindenter to experimentally impose loads on both types of composites that achieved indentation depth that was 1 μm (comparable to bacterial size) or greater (Fig. 1e). The indentation for a given force and tip geometry was consistently less for the thin gel than for the thick gel, with the maximum indentation depth to maximum load ratio being ~125 nm/μN for thick composites and ~75 nm/μN for thin composites. This experimentally validates that the composite made with thin gel is stiffer than the composite made with thick gel (Supplementary discussion and Fig. 1e).

Cryo-electron microscopy showed no discernible difference between the surface topographies of thick and thin gels (Supplementary Fig. 1d, e). However, atomic force microscopy (AFM) shows slight differences in the topographies, with measured average roughness of 1.25 nm for thin agarose gels and 2.62 nm for thick agarose gels (Fig. 1f, g). This ~1 nm difference in roughness is 3 orders of magnitude smaller than 1 μm bacteria, and this size discrepancy makes it unlikely that differences in surface topography could be sensed by bacteria. Moreover, others have found that the virulence response of P. aeruginosa upon adhesion is dependent on the stiffness of the surface but not related to the sizes of surface pores ranging from less than 10 to more than 1000 nms²¹. AFM measures an average roughness of 6.2 nm for thin alginate gels and 9.5 nm for thick alginate gels (Supplementary Fig. 1f, g). This ~3 nm difference in roughness is similarly far too small for any impact on bacteria to be expected.

Next, we sought to evaluate the degree to which the mechanics of composite substrates impacts bacteria. Upon surface attachment, bacteria are subjected to a mechanical tension in their inner membranes arising from nanoscopic cell envelope deformation due to the adhesion force exerted by the surface^37,38,39. If topography had any effect on bacteria, we would expect that the rougher thick composite would exert a greater net adhesion force, due to the increase in surface area consequent to greater roughness. To test whether we could observe a difference in the membrane tension between bacteria adhering to the thin and to the thick composites, we compared the membrane tension in bacteria attached to both composites by measuring the activity of mechanosensitive ion channels. Other researchers have already established much about the mechanical activation of these ion channels – here, our focus is not to understand the ion channels themselves but rather to use their response as a readout to assess whether bacteria experience different degrees of mechanical change when they attach to thin composites than when they attach to thick composites.

These channels are located on the inner, cytoplasmic membrane⁴⁰ and act as transducers of membrane tension - closed when the membrane is at low tension and open when the membrane is at high tension, allowing ions to pass through^40,41. The two major mechanosensitive ion channels are large-conductance- and small-conductance- (MscL- and MscS-, respectively) type channels. When open under increased membrane tension, these channels provide non-selective pores of large and small diameter, respectively, through which sodium ions, Na⁺, can pass in very similar ways⁴¹. We pre-loaded bacteria with a fluorescent indicator for Na⁺ and then allowed them to sit for one hour attached to thin and thick agarose gels, in the presence of excess external Na⁺, before measuring the indicator brightness distribution as a proxy for internal Na⁺ levels.

The brightness distribution for bacteria on the thin composites had a peak at 100–200 arbitrary units (a.u.), whereas the brightness distribution for bacteria on thick composites had a peak, representing more than 60% of cells, at 0 to 100 a.u. (Fig. 2a). Both the median (Fig. 2a inset) and the mean fluorescence intensity of bacteria on thin composites were significantly greater than that of cells on thick ones – cells on thin gels had a mean fluorescence intensity of 2840.70 a.u. [2217.75 a.u., 3463.65 a.u.] (95% confidence interval) and cells on thick gels had a mean fluorescence intensity of 677.97 a.u. [478.28 a.u., 877.67 a.u.] (95% confidence interval). These results show that bacteria on thin composites are more permeable to Na⁺ than are bacteria on thick composites. Since mechanosensitive ion channels increase permeability upon increased membrane tension, we interpret this finding as indicating that bacteria have higher membrane tension when attached to thin gel surfaces than to the thick. This is the opposite of what we should expect if the measured differences in surface topography were impacting bacteria. However, it is entirely congruent with what we should expect if differences in composite mechanics were impacting bacteria, as follows:

**Fig. 2: Adhesion to a thin gel surface leads to greater changes in mechanical stress/strain in the bacterial envelope and increased permeation of the bacterial cell membrane by sodium.**

Adhesive forces will tend to increase the area of the bacterium in contact with the surface, by deforming the bacterium and the surface. The energy costs for deforming the bacterium and the surface will depend on the elasticity of each. Mechanical equilibrium will be found by minimizing the sum of elastic energy costs (from cell and substrate deformation) and the adhesive energy benefit (from contacting area). Therefore, for constant adhesive area and bacterial elasticity, we expect that the deformation of the bacterial cell envelope will depend on the elasticity of the substrate (i.e., surface mechanics). To validate the trends shown by our experiment results and to elucidate further the relationship between surface mechanics and mechanical stresses in adhering bacteria, we developed finite element models (Fig. 2b and Supplementary Fig. 3) to simulate bacterial attachment to gel-coverslip composites. At the molecular level, bacterial surface properties and how they impact attachment to substrates are complex and not well-known⁴². Therefore, we approximated the adhesion process by displacing bacteria into contact with surfaces (Supplementary Discussion, Supplementary Fig. 3c, d). Using our models, we compared the bacterial envelope mechanics for bacteria interacting with thin and thick gel surfaces for a range of contact-increasing displacements. For any given displacement, the total contact area is greater for bacteria on the thick gel than on the thin one (Fig. 2c), reflecting the fact that the energy cost for deforming a soft material is lower than the cost for deforming a stiff one by the same amount. The initial, free-floating cells were subjected only to a turgor pressure (biologically, this arises from the osmolarity difference between the cytoplasm and the exterior), so that bacteria were in a pre-stressed state. Contact with a surface leads to a decrease in membrane stresses on the outer membrane, an increase in circumferential strain on the inner membrane, and the development of contact pressure (Fig. 2d and Supplementary Fig. 4a–i). These changes are all more pronounced when bacteria attached to thin gel surfaces.

Thus, our modeling results, showing greater strain in the bacterial inner membrane when attached to thin composites than to thick, are consistent with experimental measurements of the activity of mechanosensitive ion channels being greater on thin composites than on thick. This confirms that adhesion to thin gel and glass composites causes greater changes in the mechanical state of the bacterial envelope than does adhesion to thick gel and glass composites, indicating that this effect arises from different composite mechanics. Therefore, the most reasonable expectation is that any difference in bacterial response to the two types of composites should be linked to the difference in effective elasticity of the thick and thin composites.

Early steps in biofilm initiation

The formation of a biofilm on a surface begins when bacteria initially encounter and attach to a surface. Bacteria can either then remain on the surface or detach back into the free-swimming, planktonic phase. The total accumulation of bacteria on a surface will depend on both the rate of attachment and the rate of detachment. Subsequent to this initial accumulation (and assuming that the reservoir of planktonic bacteria is removed), the population of bacteria on the surface will experience a phase of little change. This is known as the biofilm lag phase, and results from a combination of bacteria replication and detachment from surfaces, such that the population of surface-bound bacteria does not increase^43,44. The lag phase ends with the onset of exponential growth of the population of surface-bound bacteria. Having established in the preceding subsection that adhesion to substrates with different composite elasticities results in different mechanical changes in the bacterial cell envelope, the remainder of our study here focuses primarily on the accumulation of bacteria on surfaces with different composite elasticities, and secondarily on the lag phase and exponential growth of bacteria on surfaces with different composite elasticities.

The pore sizes of the agarose and alginate hydrogels used are far too small to allow the bacterial body to enter the hydrogel; this is confirmed by microscopy observation that all sessile bacteria are attached to the top hydrogel surface, at the gel-fluid interface. This study focuses solely on bacteria attached to the hydrogel surface, and not on any free-swimming, planktonic bacteria in the fluid phase above. This is appropriate to our goal of better characterizing the role of bacterial mechanosensing in early biofilm development. Since the hydrogel thickness does not notably impinge on surface adhesivity or other surface properties (Fig. 1b, c, f, g and Supplementary Fig. 1b–g), the mechanosensing investigated here is that arising from substrate elasticity. Different adhesion forces⁴⁵ and differences in the stiffness of bacterial cell walls³⁷ could also give rise to different mechanical changes in bacterial envelopes upon adhesion, but these two parameters are not varied in the present study.

PilY1 allows P. aeruginosa to differentially accumulate on thick and thin composites

As mentioned above, two categories of sensing pathways may be involved in the mechanosensing by P. aeruginosa. Therefore, to assess the impact of surface mechanics on the accumulation of bacteria on surfaces, we incubated the bacterial suspension of wild-type cells (WT), mutants without TFP (∆pilA), mutants without the PilT retraction motor (∆pilT) and mutants without PilY1 (∆pilY1) for one hour with hydrogel-coverslip composites and measured the bacterial accumulation on these surfaces by visualizing the number of bacteria using phase contrast microscopy. Consistent with some previous reports^32,33,34 but not with others³⁵, WT accumulated significantly more on thin composites than on thick composites (Fig. 3a, b), as did the ∆pilA and ∆pilT mutants (Fig. 3a, b). These findings were true across two types of hydrogels, agarose and alginate, supporting the idea that the difference in accumulation is likely to arise from the ~16x contrast in composite elasticity rather than details of surface chemistry.

**Fig. 3: More bacteria accumulate on thin gel composites during one hour's incubation for initial attachment.**

Accumulation on the thin composites was greater by a factor of ~3.3 for all three strains (Fig. 3c). Thus, while functional TFP can increase the "baseline" accumulation, they have no measurable impact on the greater likelihood of accumulating on thin composites. Similar effects were found for the more starkly-contrasting case of glass versus agarose gel surfaces (Supplementary Discussion and Supplementary Fig. 2a).

In contrast, ∆pilY1 accumulated equally on thin and thick composites (Fig. 3a, b); this also is true for both agarose and alginate gels. This indicates that P. aeruginosa requires the cell-surface-exposed protein PilY1 for distinguishing between, and responding to, different surface mechanics. Again, similar effects were found for the more starkly-contrasting case of glass versus agarose gel surfaces, implying a much more muted response to stiffness difference by ∆pilY1 (Supplementary Discussion and Supplementary Fig. 2a). Adhesion-induced changes can only happen following, not preceding, bacterial contact with surfaces. Since gel thickness does not impact physiochemical surface adhesivity (Fig. 1c and Supplementary Fig. 1c), we expect bacteria to have equal likelihood of encountering and initially sticking to thin and thick composites. Therefore, this finding shows that greater accumulation on thin composites must arise as the result of something that happens after initial surface engagement – i.e., there is an active bacterial response to surface mechanics.

Thus, we hypothesize that WT initially adhered to thick composites will be more likely to detach than cells initially adhered to thin composites and that this difference should require PilY1. We test this hypothesis below.

PilY1 mediates flagellar spinning and detachment rate in response to surface mechanics

Shortly after encountering a surface, many P. aeruginosa cells are reversibly tethered by their flagella, which drive spinning about the surface-attached portion of the flagellum (to optical microscopy, the axis of rotation usually appears to go through one end of the cell). Spinning facilitates detachment from surfaces^46,47. A deficiency in spinning is also associated with decreased probability of detachment⁴⁶. Bacteria can also use TFP to move laterally on surfaces, but, during the first hour after bacteria were introduced to hydrogels (i.e., what we have termed the accumulation stage), we found that the vast majority of surface motility was in the form of spinning (Supplementary Fig. 5a, b). Therefore, we tracked the center-of-mass speed of surface-adhered bacteria (Fig. 4a–d) as a measure of spinning motility. We expect that a population with faster-spinning bacteria will have a higher rate of detachment⁴⁸.

**Fig. 4: Adhered bacteria spin during the first hour of accumulation.**

For WT, the distribution of spinning speeds on thick composites was much broader than thin composites (Fig. 4e). Both the median (Fig. 4e inset) and the mean speeds on thick composites were significantly higher than on thin composites - mean speed of 20.06 μm/min [18.43 μm/min, 21.68 μm/min] (95% confidence interval) on thick composites and mean speed of 11.46 μm/min [10.95 μm/min, 11.97 μm/min] (95% confidence interval) on thin composites. In summary, WT are more likely to spin rapidly on thick composites than on thin composites. Upon tracking cells, we indeed found that WT were significantly more likely to detach from thick gels (30 detachment events among 673 tracked cells) than from thin gels (10 detachment events among 1609 tracked cells) (P < 0.001, χ² test) (Supplementary Fig. 5c). This is an active bacterial response to surface mechanics.

For the ∆pilY1 mutant, the peak spinning speed was unchanged from that of WT (Fig. 4e, f), suggesting that loss of PilY1 does not intrinsically disrupt spinning motility. However, for the ∆pilY1 mutant, neither the distributions of spinning speeds nor the median spinning speeds were significantly different on thin and thick composites (Fig. 4f). The mean speed was 15.08 μm/min [14.16 μm/min, 16.01 μm/min] (95% confidence interval) on thin composites and 14.86 μm/min [13.92 μm/min, 15.81 μm/min] (95% confidence interval) on thick composites. Furthermore, the ∆pilY1 mutant was equally likely to detach from thin and thick composites (P = 0.78, χ² test) (Supplementary Fig. 5c). These results are strikingly unlike those for WT and imply that P. aeruginosa lacking PilY1 do not adjust their spinning motility, and therefore their likelihood of detachment, in response to surface mechanics. PilY1 is linked to regulating flagellar activity either up or down - increasing spinning speed on thick composites and decreasing spinning speed on thin composites (Supplementary Fig. 5d,e). Notably, we find a linear correlation between spinning speed and the probability of detachment (Supplementary Fig. 5d,e)

These findings raise the question of what provides the causative linkage between PilY1 and changes in flagellar activity. A key regulator of flagellar motility in P. aeruginosa and many other microbes is the intracellular second messenger cyclic diguanylate (c-di-GMP)^49,50. We have shown that, compared with thick composites, thin composites could lead to higher membrane tension in the adhering bacterium. In addition, others have shown that the unfolding and misfolding of inner-membrane and periplasmic proteins associated with surface attachment results in an elevated level of c-di-GMP for P. aeruginosa¹².

Therefore, we hypothesized that different surface mechanics arising from thin and thick composites would cause different levels of c-di-GMP production upon surface attachment, and that PilY1 is involved in the c-di-GMP response. We test this hypothesis below.

Composite mechanics impact c-di-GMP signaling in a PilY1-dependent manner during bacterial accumulation

To see whether PilY1 modulates c-di-GMP dynamics in response to composite mechanics, we used a validated reporter plasmid, P_cdrA::gfp, that produces green fluorescent protein (GFP) in response to increases in c-di-GMP⁵¹. We previously used this plasmid study c-di-GMP signaling in bacterial mechanosensing of shear¹³. We measure average per-cell GFP intensity, accounting for attenuation by the substrates, as we did in earlier studies^13,52.

For bacteria containing PilY1, we found a sharp rise in c-di-GMP levels during the initial hour of accumulation (−1 to 0 h in Fig. 5a), which is consistent with previous findings that c-di-GMP levels in P. aeruginosa increase upon surface attachment^9,13,53. At the end of the "accumulation" hour (i.e., the beginning of the incubation time), WT on thin composites had significantly higher c-di-GMP levels than did WT on thick composites. This meshes with our finding that WT on thin composites had lower spinning motility than those on thick composites (Fig. 4e), as high levels of c-di-GMP inhibit bacterial motility⁴⁹. These data also suggest that the causative linkage between PilY1 and changes in flagellar activity (which, in turn, modulate the likelihood of detaching from the surface), is likely via PilY1-controlled c-di-GMP signaling.

Fig. 5: On agarose surfaces with different mechanics, PilY1 acts to mediate the duration of the lag phase in biofilm growth and the levels of the intracellular signal c-di-GMP, and PilT is required to mediate the growth rate of the exponential phase of biofilm growth.

On both thin and thick composites, the ∆pilY1 mutant had much lower c-di-GMP levels than did WT (Fig. 5a, b); this finding is consistent with the role of PilY1 in regulating c-di-GMP production⁹. For WT, the mean level of c-di-GMP at the end of the "accumulation" hour was ~2.9 times higher on thin composites than on thick composites, but it was only ~1.4 times higher for the ∆pilY1 mutant (Fig. 5a, b). This observation is consistent with a loss of the ability to discriminate surface mechanics. This finding is also consistent with the causative connection that PilY1 is required for bacterial mechanoresponse linking surface mechanics to c-di-GMP signaling levels during the initial "accumulation" phase.

This finding also raises the question of how PilY1, and consequent changes in c-di-GMP signaling, impact the growth of the bacterial population on the surface.

PilY1 impacts biofilm growth in the lag phase, in response to composite mechanics, by modulating c-di-GMP signaling

When planktonic bacteria are introduced into new liquid medium, they experience a temporary period of non-replication, termed the "lag phase"⁵⁴. After attachment to a glass surface, P. aeruginosa also undergoes a lag phase before exhibiting exponential growth^43,44. However, unlike the planktonic lag phase, the lag phase of biofilm growth involves a combination of bacteria replication and detachment from surfaces, such that the population of surface-bound bacteria does not increase^43,44.

After allowing bacteria to accumulate on the surfaces of composites for one hour, we replaced the bacterial suspension with fresh, sterile culture medium, so that no more bacteria can attach to the surface from the liquid phase. We designate this timepoint the beginning of the incubation time (0 h in Fig. 5). The duration of the lag phase, from the beginning of the incubation time to the onset of exponential growth, is given by the lag time, τ_lag, indicated by hatched color bars in Fig. 5. WT populations had a τ_lag of 4 h on thin composites and 1 h on thick composites, but ∆pilY1 populations had the same τ_lag of 1 h on both composite types (Fig. 5d, e). Similar results were found for bacterial growth on bulk gels (soft) and glass slides (stiff) (Supplementary Fig. 2b, c). These results show that composite mechanics can markedly impact the growth of the sessile bacterial population, and that PilY1 is key for this process as well as for the accumulation preceding incubation. When PilY1 was complemented back on an arabinose-inducible plasmid, the ∆pilY1 mutant populations again had different τ_lag on thin and thick composites (Supplementary Discussion and Supplementary Fig. 6c), confirming PilY1's role in surface mechanics sensing. Interestingly, although the ea...

Pseudomonas Aeruginosa Treatment