Surface hydrophilicity promotes bacterial twitching motility

ABSTRACT Twitching motility is a form of bacterial surface translocation powered by the type IV pilus (T4P). It is frequently analyzed by interstitial colony expansion between agar and the polystyrene surfaces of petri dishes. In such assays, the twitching motility of Acinetobacter nosocomialis was observed with MacConkey but not Luria-Bertani (LB) agar media. One difference between these two media is the presence of bile salts as a selective agent in MacConkey but not in LB. Here, we demonstrate that the addition of bile salts to LB allowed A. nosocomialis to display twitching. Similarly, bile salts enhanced the twitching of Acinetobacter baumannii and Pseudomonas aeruginosa in LB. These observations suggest that there is a common mechanism, whereby bile salts enhance bacterial twitching and promote interstitial colony expansion. Bile salts disrupt lipid membranes and apply envelope stress as detergents. Surprisingly, their stimulatory effect on twitching appears not to be related to a bacterial physiological response to stressors. Rather, it is due to their ability to alter the physicochemical properties of a twitching surface. We observed that while other detergents promoted twitching like bile salts, stresses applied by antibiotics, including the outer membrane-targeting polymyxin B, did not enhance twitching motility. More importantly, bacteria displayed increased twitching on hydrophilic surfaces such as those of glass and tissue culture-treated polystyrene plastics, and bile salts no longer stimulated twitching on these surfaces. Together, our results show that altering the hydrophilicity of a twitching surface significantly impacts T4P functionality. IMPORTANCE The bacterial type IV pilus (T4P) is a critical virulence factor for many medically important pathogens, some of which are prioritized by the World Health Organization for their high levels of antibiotic resistance. The T4P is known to propel bacterial twitching motility, the analysis of which provides a convenient assay for T4P functionality. Here, we show that bile salts and other detergents augment the twitching of multiple bacterial pathogens. We identified the underlying mechanism as the alteration of surface hydrophilicity by detergents. Consequently, hydrophilic surfaces like those of glass or plasma-treated polystyrene promote bacterial twitching, bypassing the requirement for detergents. The implication is that surface properties, such as those of tissues and medical implants, significantly impact the functionality of bacterial T4P as a virulence determinant. This offers valuable insights for developing countermeasures against the colonization and infection by bacterial pathogens of critical importance to human health on a global scale.


Introduction
Twitching motility is a form of non-flagellated bacterial translocation that allows bacteria to move on or between solid surfaces [1].This type of motility is powered by the bacterial type IV pilus (T4P) and the supramolecular T4P machinery (T4PM) [2][3][4][5].In the current model, it is the recurrent cycles of T4P extension and retraction that powers this bacterial surface motility [6].The T4PM assembles the long T4P protein filament that protrudes from a cell into its surroundings.When the tip of an assembled or extended T4P attaches to a solid substratum, the retraction of the T4P by the T4PM moves a bacterium directionally.This results in the translocation of bacterial cells on or between solid surfaces, resulting in bacterial twitching motility [7].
More important to human health, T4P plays a crucial role in the pathogenesis of many important bacterial pathogens [8][9][10][11][12][13].These include World Health Organization (WHO) priority pathogens [14], Pseudomonas aeruginosa and Acinetobacter baumannii which use T4P to adhere to human cells or tissues to initiate colonization and invasion [1,15,16].Acinetobacter nosocomialis, a close relative of A. baumannii, is itself an opportunistic pathogen primarily causing nosocomial or hospital-acquired infections [17].Additionally, the A. nosocomialis M2 strain has been used as a model system for Acinetobacter studies of pathogenesis [16,18,19].
Although the lack of flagellated motility led to the acineto-or non-motile designation for this genus, many Acinetobacter species, in fact, possess T4P and exhibit the T4P-dependent twitching motility [20].As such, twitching motility provides a convenient assay for bacterial T4P in these medically important pathogens.
Twitching motility is commonly analyzed by observing interstitial colony expansion between the surfaces of solidified nutrient agar and of plastic Petri dishes made of polystyrene [21].It was observed previously that A. nosocomialis exhibited twitching only with MacConkey agar, but not Luria-Burtani (LB) agar on these polystyrene petri dishes [22].Notwithstanding their commonalities, LB lacks peptone, lactose and bile salts that are present in MacConkey.
Peptone is a proteinous nutrient source and lactose is a carbon and energy source.Bile salts, primarily sodium glycocholate and sodium taurocholate, are detergents produced from cholesterol degradation in humans and other vertebrates [23].They are included in MacConkey as a selective agent against Gram + bacteria and for enteric bacteria.The amphipathic nature of bile salts allows them to interact with hydrophilic and hydrophobic molecules, thus enabling them to result in envelope stress on bacteria [24] or alter the physicochemical properties of solid surfaces they are applied to [25].
In this study, we examined the composition of MacConkey media and determined that bile salts are stimulatory of A. nosocomialis twitching motility on polystyrene Petri dishes.We also observed similar stimulatory effects of bile salts on the twitching of P. aeruginosa and A. baumannii.Our results further demonstrate that other detergents likewise can enhance bacterial twitching on polystyrene surfaces.Antibiotics, including outer membrane-targeting polymyxin B, do not increase P. aeruginosa twitching motility, suggesting that the mechanism for the stimulatory effect of bile salts is unlikely related to a bacterial response to envelope or antimicrobial stress.Instead, we discovered that it was the hydrophilicity of surfaces that promotes bacterial twitching.Plasma-treated polystyrene or glass surfaces, which are more hydrophilic, for example, promote twitching motility significantly without detergents.The results here suggest that bacterial pathogens may have evolved mechanisms to differentially interact with surfaces with different physicochemical properties to optimize host recognition, colonization and infections.

Results
Bile salts enable A. nosocomialis twitching motility in agar plates.It has been reported in the literature that Acinetobacter nosocomialis displays twitching motility in agar plate assays with MacConkey but not in LB [22].In these assays, bacterial cells are stabbed through the agar to form an interstitial colony on the Petri dish surface below agar [21].The size of an interstitial colony can be measured to quantify twitching motility in such assays.As shown in Figure 1, the A. nosocomialis M2 strain shows clear twitching with MacConkey, but not with LB agar.These results here confirmed the effect of agar media on A. nosocomialis twitching motility in the literature.
To investigate the underlying reason for this observation, the composition of these two commonly used growth media were compared.As shown in Table S1, both media use the same concentration of agar and sodium chloride, 1.2% and 0.5% respectively.The key components in MacConkey that are absent in LB include peptone, lactose, and bile salts.Neutral red, another component that differed between them, is a pH indicator dye for the detection of lactose fermentation and we considered it unlikely to be responsible for the differences in twitching.We supplemented the LB agar with peptone, lactose or bile salts at the same concentration as in MacConkey agar to determine if one of these could enable A. nosocomialis to twitch.As shown in Figure 1A, bile salts, but not the others, were found to permit A. nosocomialis twitching motility in LB agar plates.This result suggests a bile salt-specific stimulation of twitching motility in A. nosocomialis.
Twitching motility as analyzed by the agar stabbing method has been observed in Acinetobacter baumannii, a closely related Acinetobacter species and WHO priority pathogen [14].The twitching motility of this bacterium is similarly noted in MacConkey agar and not LB agar.We tested two A. baumannii strains, AYE and AB0057, in LB supplemented with bile salts in comparison with A. nosocomialis M2.As shown in Figure 1C, bile salts stimulated the twitching motility of the two A. baumannii strains similar to the observation of A. nosocomialis.
These results suggested that the stimulatory effects of bile salts on twitching motility occurred in Acinetobacter species as a general phenomenon.
Bile salt-stimulated twitching motility of P. aeruginosa.These above observations prompted us to investigate if bile salts enhance bacterial twitching by a more general mechanism in other bacteria.
Pseudomonas aeruginosa, a WHO priority pathogen [14], has been used as a model for studies of twitching motility [6,26,27].Its twitching motility has been routinely analyzed with LB instead of MacConkey agar plates.As shown in Figure 2A, bile salts supplemented at 0.5%, the same concentration present in MacConkey, significantly increased the twitching motility of the P. aeruginosa strain PAO1, the frequently used strain for the studies of twitching motility in this bacterium [28,29].
The dose response of bile salts on PAO1 twitching was examined and as shown in Fig. 2B, there is a concentration-dependent response, reaching a maximum stimulation around 0.4% bile salts.At higher concentrations, bile salts start to inhibit P. aeruginosa growth and reduce its twitching area in this assay.
We also observed that the twitching motility of PA14, another common P. aeruginosa strain in research laboratories [30][31][32], was stimulated by bile salts in LB media (Figure S2).These results demonstrate that the stimulatory effect of bile salts on twitching is applicable to both Acinetobacter species and P. aeruginosa.For the remaining part of this study, we mostly used P. aeruginosa PAO1 as the model organism to investigate the mechanisms by which bile salts stimulate bacterial twitching motility.
Detergents stimulate twitching motility.Bile salts are anionic detergents, produced from cholesterol degradation [33], that can elicit a lipid membrane stress response in bacteria [34].Thus, it is possible that bile salts function as a detergent and apply envelope stress to stimulate bacterial twitching [34].To examine this possibility, we investigated the effect of other detergents on the twitching motility of P. aeruginosa.To avoid crosstalk between growth inhibition and twitching motility, non-inhibitory concentrations of detergent were determined to guide their use in the twitching motility assays (Table S2).For this experiment, we tested sodium dodecyl sulfate (SDS), an anionic detergent, as well as Triton X-100 and Triton X-114, which are non-ionic detergents (Table S2).As shown in Fig. 3A, all the detergents examined, whether anionic or non-ionic, significantly stimulated the twitching motility of P. aeruginosa much like bile salts.These results suggest that the stimulatory effects of bile salts on twitching is related to their chemicophysical properties as detergents.
The stimulatory effects of detergent on twitching motility could be explained by a physiological response of a bacterium to envelope stress applied by these amphipathic molecules or a general stress applied by other stressors [35].To investigate this, we tested a few antibiotics with different targets at non-inhibitory concentrations as stressors distinct from detergents.These included ampicillin, gentamicin, and ciprofloxacin, which target cell wall biosynthesis, ribosome function and DNA topology, respectively.Maximum non-inhibitory concentrations (MaxNICs) of these antibiotics were determined by testing the effect of different concentrations on P. aeruginosa growth in liquid culture (Table S3).Antibiotics at their respective MaxNICs were tested for their effect on P. aeruginosa twitching.As shown in Figure 3B, none of these three antibiotics increased P. aeruginosa twitching motility.This suggested that the stimulation of twitching was unlikely the result of a general physiological response to stressors.We additional tested the effect of polymyxin B, which targets the outer membrane and would apply envelop stress similar to bile salts and the detergents.Somewhat surprisingly, this antibiotic at its MaxNIC (Table S3) showed no stimulatory effect on P. aeruginosa twitching (Figure 3B).These results suggested that the observed stimulation of twitching motility by bile salts and other detergents in Figures 1 and 2 was not a response to envelope stress.
Hydrophilic surfaces promote twitching motility in P. aeruginosa.Detergents are amphipathic molecules that can mediate both hydrophobic and hydrophilic interactions [33].Besides physiological effects on bacteria as envelope stressors, they can also change properties of surfaces on which bacterial twitching motility occurs [25].In the stabbing assay, bacteria cells twitch in the interstitial space between the agar media and the surface of the hydrophobic polystyrene Petri dish.We considered the possibility that detergents such as bile salts may interact with the hydrophobic surface of the polystyrene Petri dishes.Such interactions could coat the surface and make it more hydrophilic to facility twitching motility.Glass is a hydrophilic surface, so we compared P. aeruginosa twitching on glass versus polystyrene microscope slides placed within Petri dishes containing LB agar without bile salts or detergents.As shown in Figure 4, twitching motility on glass is significantly higher than that on polystyrene for the wildtype (WT) strain.The addition of bile salts had minimum effects on twitching on glass surfaces, in contrast to the observations with polystyrene ones (Figure 4).These results indicate that surface hydrophilicity likely enhances twitching motility and the effects of detergents and bile salts on twitching could be attributed to their ability to change a hydrophobic surface to a more hydrophilic one.
To further examine this possibility, we used a coating to change the surface hydrophilicity of glass by pretreating it with a polydimethylsiloxane solution (PDMS).When the glass surface was layered with PDMS to mimic the hydrophobicity of a polystyrene plate, P. aeruginosa twitched at similar levels as it did with polystyrene Petri dishes in LB media without bile salts (Figure 4B).PDMS coating did not impact twitching motility of P. aeruginosa on polystyrene surfaces which is already hydrophobic (Figure 4B).A P. aeruginosa pilA mutant was used as control.It showed no twitching with any of the surfaces with or without PDMS treatment.These results are consistent with the notion that hydrophilic surfaces enhance P. aeruginosa twitching and that bile salts and other detergents stimulate twitching motility on the hydrophobic polystyrene surfaces by making it more hydrophilic.
While natural polystyrene surfaces are hydrophobic, they can be treated with plasma gas to increase their hydrophilicity for tissue culture purposes [36].We therefore compared P. aeruginosa twitching motility on 6-well polystyrene plates either untreated or treated by plasma gas for tissue culture.As shown in Figure 5A, the wildtype P. aeruginosa PAO1 strain exhibited significantly increased twitching motility on plasma-treated surfaces in LB media over those untreated surfaces.In contrast, while the addition of bile salts significantly enhanced P. aeruginosa twitching on untreated plates, it had no significant effect on PAO1 twitching on the treated surfaces.These results led us to conclude that the stimulatory effect of bile salts on twitching motility is related to the changes in surface hydrophobicity instead of a physiological response of a bacterium elicited by the presence of bile salts or other detergents.
The increase in P. aeruginosa twitching motility without any media additives on hydrophilic surfaces was striking, thus we wanted to confirm if this phenotype was consistent with Acinetobacter nosocomialis, which is non-motile in LB media.Markedly, the wildtype A. nosocomialis M2 strain (M2 WT) also exhibited increased twitching motility on plasma-treated polystyrene in LB media and no twitching on hydrophobic polystyrene in LB media (Figure 5B).The difference between these two polystyrene structures is their hydrophilicity, therefore we provided significant evidence that surface hydrophilicity promotes twitching motility in P. aeruginosa and A. nosocomialis.

Discussion
Here we describe the discovery of detergent-stimulated twitching motility and identify a relationship between surface hydrophilicity and twitching motility of Acinetobacter species as well as P. aeruginosa.It was previously observed that A. nosocomialis can perform twitching motility on MacConkey agar but not Luria-Bertani (LB) agar [22].We discovered that bile salts were the component in MacConkey agar that facilitated the twitching motility of A. nosocomialis as well as A. baumannii on polystyrene surfaces.This stimulatory effect further extended to P. aeruginosa.We observed that other ionic and anionic detergents, such as Triton X-100, Triton X-114 and sodium dodecyl sulfate (SDS) also promoted P. aeruginosa twitching motility.The stimulatory effect of bile salts and other detergents on bacterial twitching is likely due to their effects on physicochemical properties of surfaces instead of a physiological response of the bacteria to detergents.This conclusion is supported by our experimental results with antibiotics that have various modes of actions, as they all failed to enhance twitching motility.Rather, we found that physicochemical properties of surfaces play a critical role in modulating the level of bacterial twitching motility: surface hydrophilicity promotes whereas hydrophobicity suppresses bacterial twitching.
The interaction between pathogens and host surfaces is essential for the establishment of a bacterial infection.Additionally, interactions with the surfaces of medical implants and devices play a major role in the establishment of hospital-acquired bacterial infections [15,[37][38][39][40].The bacterial type IV pilus (T4P) is an important virulence factor needed for the establishment of infection for many WHO priority pathogens [14], such as P. aeruginosa and A. baumannii [15].Thus, studying the T4P as a motility apparatus is essential to better understanding the first step of infection.Interested in understanding how MacConkey agar could enable the T4P-mediated twitching motility of A. nosocomialis, we discovered an important relationship between surface chemistry and twitching motility.Not only did hydrophilic surfaces, such as glass and plasma-treated polystyrene increase bacterial twitching motility, but hydrophobic surfaces, like polystyrene, completely disabled twitching motility of A. nosocomialis.This result highlights a powerful control point for T4P functionality via surface polymer construction.Hence, we propose that the hydrophilicity of surfaces, such as those of host tissues or medical implants, may play an important role in the activity of the T4P as a virulence factor responsible for the establishment and persistence of a bacterial infection.
While the implications of this discovery are a much-needed consideration in future research, we note that the mechanism of surface hydrophilicity-promoted twitching motility remains unclear.
Hydrophobic surfaces allow for better attachment by P. aeruginosa for biofilm formation, a nonmotile or sessile state [41].Thus, it is possible that the interaction between the T4P head and surfaces during twitching motility has better attachment to hydrophilic surfaces than hydrophobic surfaces.Additionally, hydrophilicity may play a role in reducing the friction between the cell surface during translocation over the surface via pilus retraction.In general, there are like-to-like observations for surface attachment, in which bacteria with hydrophobic cell surfaces adhere better to hydrophobic surface materials [42].Both P. aeruginosa and A. baumannii have high surface hydrophobicity [43,44], thus adhering better to hydrophobic twitching surfaces.When these hydrophobic cells move over hydrophilic surfaces, it is expected that the cells will adhere less which may promote increased motility.Although we are unsure of the mechanism of surface hydrophilicity-promoted twitching, it is reasonable to conclude that increased hydrophilicity of a material surface would stimulate twitching motility.
While our observations are consistent with surface hydrophilicity promoted twitching motility, it is also possible the changes in motility are due to the surface sensing capacity of the T4P.The T4P has been implicated as a surface sensor in Caulobacter crescentus [45][46][47].The T4P retraction motor and minor pilins were also found to be involved in sensing surface association necessary for quorum sensing in P. aeruginosa [48].Additionally, twitching motility has been shown to require bacteria to sense and respond to surfaces [49].Thus, it is possible that bacteria are sensing the hydrophilicity of a surface, and this results in the increased twitching motility phenotype we describe in this study.Since surface hydrophilicity is less conducive for hydrophobic cell attachment [42], it is possible that surface sensing of hydrophobicity would result in increased twitching via cell signaling.
A few lines of evidence suggest that hydrophilic surfaces increase the twitching motility of P. aeruginosa and A. nosocomialis.Mainly, two hydrophobic surfaces, glass and plasma-treated polystyrene, resulted in significantly increased twitching.On the other hand, hydrophobic surfaces decreased twitching motility.Using detergents and antibiotics we also supported the physicochemical twitching response and demonstrated that bacterial physiology was not the mechanism of detergentstimulated motility.While this result highlights the discovery of surface hydrophilicity-promoted motility, future work is necessary to elucidate the mechanisms of this phenomenon.Additionally, it would be interesting to investigate the extents of this phenomenon as it may play a role in other T4Pdependent motilities, such as social gliding motility in Myxococcus xanthus [50], or other pathogens that utilize twitching motility, like Neisseria gonorrhoeae [1,51].We suggest that the hydrophilicity of surfaces, such as those of animal tissues as well as of medical implants and devices, may significantly impact the functionality of T4P as a virulence factor and its interaction with hosts in the establishment and persistence of a bacterial infection.

Materials and Methods
Strains and culture conditions.The bacterial strains used in this study are detailed in Table 1.These include wildtype and pilA mutant Acinetobacter nosocomialis M2 and Pseudomonas aeruginosa PAO1.
LB broth supplemented with 1.2% granulated agar was also utilized for twitching assays throughout this study.P. aeruginosa strains were grown and maintained on 1.5% Luria-Bertani agar (LBA) at 37℃, while A. nosocomialis and A. baumannii strains were grown and maintained on 1.5% MacConkey agar at 37℃.
Polystyrene Petri dish twitching motility assay.Twitching assays were performed on fresh 25 mL 1.2% LBA plates prepared the day before.When additives were added to the media, antibiotics or detergents were utilized at their maximum non-inhibitory concentration.Before performing the assay, the 1.2% LBA plates were dried face up without their lids on in the biosafety cabinet for 20 minutes.
Condensation was wiped off the Petri dish lids, and after 20 minutes of drying, the plates and the lids were sterilized under UV light for 2 minutes.The inoculum was prepared by spreading a loopful of bacteria from the outside edge of the overnight culture onto an untouched 1.5% LBA plate to create a thin layer of culture.Using a toothpick, a unified amount of inoculum was picked up and inoculated the 1.2% LBA plates by stabbing through the agar, making sure to touch the bottom of the plate.The inoculated 1.2% LBA plates were incubated in stacks at 37℃ for 48 hours in a humidity chamber.The humidity chamber was prepared by placing a dampened paper towel to cover the bottom of a plastic box.
Every 24 hours the paper towel was removed and replaced with a fresh one.After 48 hours of incubation, the twitching motility was visualized by removing the agar from the Petri dish and stained with 1% crystal violet.Twitching area was determined by analyzing the stained area of the plate with ImageJ software.This protocol was modified from a previously described twitching motility assay [21].
Twitching motility on glass and polystyrene microscope slides coated with a polydimethylsiloxane solution.Glass or polystyrene microscope slides were submerged in either a filter-sterilized polydimethylsiloxane solution or 70% ethanol as a control.These slides were left on a rack to dry for a few hours at 40℃.Once dry, the microscope slides were laid inside polystyrene Petri dishes, and 25 mL of 1.2% LBA was poured on top of them.The modified plates were inoculated as described above.The inoculated 1.2% LBA plates were incubated in stacks at 37℃ for 18 hours in a humidity chamber.After 18 hours of incubation, the twitching motility was visualized without removing the agar from the Petri dish by holding the plates up to the light and tracing the visible twitching area with a permanent marker.
Twitching area was determined by analyzing the traced area of the plate with ImageJ software.

Optimized 6-well twitching motility assay on polystyrene and plasma-treated polystyrene.
Twitching motility assays were performed as described above, with modifications to the volume of media placed in each well and incubation times. 2 mL of 1.2% LBA, with or without 0.5% bile salts, was dispensed into each well of a 6-well polystyrene or plasma-treated polystyrene plate.Inoculated plates are placed in a humidity chamber and incubated for 24 hours at 37℃.Rather than staining the plate with 1% crystal violet, once the agar is removed from the wells, the twitching area was traced with a permanent marker.The twitching area was determined by analyzing the marked area with ImageJ software.
Table 1.Bacterial strains used in this study.Subsurface twitching motility assays of wildtype (WT) and a pilA mutant strain (pilA -) of P. aeruginosa (PAO1) were performed in LB agar on polystyrene and glass microscope slides placed within a polystyrene Petri dish.The twitching motility of WT and pilA -P.aeruginosa was also analyzed in LB supplemented with 0.5% bile salts (BSs) on polystyrene and glass microscope slides.To test the effect of hydrophobic surfaces, the twitching motility of WT and pilA -P.aeruginosa was additionally analyzed on polystyrene and glass microscope slides coated with a polydimethylsiloxane (PDMS) solution.The microscope slides not coated in PDMS, were coated with 70% ethanol as a control.All of these conditions were tested via three biological replicates in triplicate.After 18 hours of incubation, the agar was removed, and the twitching area was quantified using image analysis.Asterisks (*) specify that the twitching area values on the indicated surfaces are statistically different (P<0.05) from those of the polystyrene control, and twitching values that are not statistically different (P>0.05) are represented by (ns).

A B
Fig. 4 28 Fig. 5 A B

Figure 1 .
Figure 1.Bile salts alone enable the twitching motility of A. nosocomialis in LB medium.

Figure 2 .
Figure 2. Twitching motility analysis of Pseudomonas aeruginosa in the presence of bile

Figure 3 .
Figure 3. Detergents, and not antibiotics, promote twitching motility equivalent to the bile

Figure 4 .
Figure 4. Hydrophilic glass surfaces increase twitching motility of P. aeruginosa in LB