in conditions that capture essential features of the CF airway environment. Overall, our findings highlight the importance of spatial organization in community-based behaviors and the need for a more thorough understanding of the interplay between polymicrobial communities in the context of infection.

We previously reported that P. aeruginosa responds to S. aureus from a distance by using TFP to chemotax toward and surround S. aureus colonies (42), but how this behavior affects S. aureus physiology remained unclear. To test the consequences of P. aeruginosa chemotaxis on S. aureus, we first visualized P. aeruginosa interactions with S. aureus colonies in three dimensions by performing live resonant scanning confocal microscopy of S. aureus in mono-or coculture with P. aeruginosa WT or a TFPdeficient mutant (∆pilA). Here, S. aureus and P. aeruginosa constitutively expressed sgfp (pseudocolored orange) and mCherry (pseudocolored cyan), respectively. Bacteria were inoculated between a cover slip and an agarose pad for visualization in the same visual field over time. Imaging was initiated with S. aureus and P. aeruginosa as single cells, positioned approximately 30 to 50 µm apart to provide sufficient time and distance for P. aeruginosa to respond to the presence of S. aureus. As previously demonstrated (42), at approximately 5 hours, we observed that WT P. aeruginosa responds to S. aureus by breaking into single cells and moving toward it with TFP motility, which eventually leads to P. aeruginosa surrounding, invading, and disrupting S. aureus cells from the colony (Fig. 1). This invasion is depend ent on TFP motility, as P. aeruginosa ∆pilA exhibited significantly decreased invasion compared with WT (Fig. 1B). While the ∆pilA mutant is amotile, it eventually grows against the S. aureus colony at later time points (Fig. 1A and C). These data suggest that TFP motility is not only necessary for P. aeruginosa chemotaxis toward S. aureus but also enables effective invasion of P. aeruginosa into S. aureus colonies.

To investigate how P. aeruginosa invasion affects S. aureus colony physiology, we imaged S. aureus in mono-or coculture with WT or ∆pilA P. aeruginosa following ~24 hours of incubation. At later time points, visualizing P. aeruginosa becomes challenging due to reduced fluorescence from decreased mCherry production and photobleaching. Nevertheless, phase contrast microscopy confirmed that P. aeruginosa cells surround S. aureus colonies after ~24 hours (Fig. S1).

We found that in coculture with WT P. aeruginosa, S. aureus forms smaller colonies than in monoculture by measuring the area at the base of the S. aureus colony (Fig. 2A and B). Moreover, P. aeruginosa TFP motility-mediated invasion resulted in S. aureus colony edges exhibiting reduced fluorescence, likely caused by dispersed, lysed cells, or a combination thereof (Fig. 2A). In the presence of ∆pilA, the area of S. aureus colonies was comparable to that in the presence of WT P. aeruginosa (Fig. 2B). However, despite similar growth area, S. aureus colonies exhibited less dispersal at the colony edges in coculture with ∆pilA, possibly due to loss of P. aeruginosa invasion (Fig. 2A; top and middle rows). Additionally, S. aureus colonies appeared thicker and denser than in coculture with WT P. aeruginosa, likely due to reduced cell dispersal.

To further investigate this, we visualized and quantified S. aureus colony architecture in more detail. Since thickness and density were more distinct on the colony edges, images were acquired with higher magnification and spatial resolution using galvano metric point-scanning confocal microscopy at the end time point (Fig. 2A; middle row). We then measured the height at the edge of S. aureus colonies at 15 µm from the edge using the Z-plane (Fig. 2A, bottom row, and Fig. 2C). As expected, the height at the colony edge was significantly higher in coculture with ∆pilA than in mono-or coculture with WT P. aeruginosa (Fig. 2C). To quantitatively analyze colony density, we measured both cell packing and colony surface roughness using the microscopy image analysis software BiofilmQ (45). These parameters quantify density by measuring the amount of surface or volume within a specified area. In BiofilmQ, S. aureus colony edges were separated from the background by segmentation onto a 3D grid, with each cubic grid unit measuring 0.72 µm per side. Neighborhood surface roughness and cell packing were then calculated by determining the biovolume fraction and surface height variance of S. aureus for each grid cube within 4 and 6 µm, respectively. Representative heatmaps in Fig. 2D and E provide a two-dimensional visualization of the quantified data in Fig. 2F through H, using color coding to represent local surface roughness and cell packing. S. aureus colonies in monoculture show low surface roughness and uniform cell packing FIG 2 P. aeruginosa type IV pili motility-mediated invasion influences the architecture of S. aureus colonies independently of P. aeruginosa-secreted antimicrobials. Analysis of S. aureus colony edge disruption and thickness. (A) Representative resonant scanning confocal micrographs of the whole colony (top row) or Galvano scanner colony edge micrographs of WT S. aureus (orange) in monoculture or coculture with P. aeruginosa (not shown; WT or ΔpilA) at t ~ 24 hours shown from the top (middle row) of the colony or the side (bottom row). The micrographs in A (bottom row) show the colonies on the Z-plane and demonstrate how the height was quantified. Quantification of S. aureus whole colony area at t ~ 24 hours (B) or height at the edge of S. aureus (Sa) colony (µm) at t ~ 24 hours (C) in monoculture or coculture with P. aeruginosa (WT, ΔpilA, ΔlasA, ΔAM B [bacteriostatic antimicrobials; HQNO, pyoverdine, and pyochelin], ΔAM C [complete antimicrobials; HQNO, pyoverdine, pyochelin, and LasA], or ΔpilA ΔAM C ). (D-G) Representative BiofilmQ heatmaps (D and E) and quantification (F and G) of local surface roughness and cell packing analysis at S. aureus colony edge in mono-or coculture with P. aeruginosa (WT or ΔpilA). Each data point represents the average of two technical replicates within one biological replicate. Statistical significance was determined by a Mann-Whitney U-test with an ad hoc Bonferroni correction for multiple comparisons. (H) Cell packing distribution within S. aureus colony edge in the abovementioned conditions. A Kolmogorov-Smirnov cumulative distribution (Continued on next page)

(Fig. 2D and E [first column] and Fig. 2F and G). Conversely, when WT P. aeruginosa is present, S. aureus edges exhibit significantly increased surface roughness and slightly decreased cell packing (Fig. 2D and E [middle column] and Fig. 2F and G), which suggests reduced colony density is caused by WT P. aeruginosa. When cocultured with ∆pilA (i.e., lacking invasion and colony disruption), S. aureus colonies portrayed significantly reduced colony surface roughness and increased cell packing compared with WT P. aeruginosa coculture (Fig. 2D and E [last column] and Fig. 2F and G). While the colony cell roughness was not different between S. aureus coculture with ∆pilA and monoculture, the mean colony cell packing was significantly increased (Fig. 2F and G). Additionally, we analyzed the cell packing distribution within S. aureus colony edges and found all three conditions to be statistically different (Fig. 2H). The majority of the monoculture colony edge population was distributed between 0.5 and 1.0, with almost no low-density areas. In S. aureus coculture with ∆pilA, the large peak at 1.0 indicates that the majority of cells within these edges are highly packed. On the other hand, in the presence of WT P. aeruginosa, the cell packing is more evenly distributed with an increased proportion of the population at low-density values compared with monoculture or coculture with ∆pilA. However, there is also an increased proportion of the population at high-density values portrayed as a small peak at 1.0. This peak may be attributed to the colony edge height being slightly higher than monoculture S. aureus as reported in Fig. 2C, leading to higher cell packing as this calculation considers the three-dimensional space.

Thus, while the base area of the colony is similar in the presence of WT or ∆pilA, the colonies are more densely packed when P. aeruginosa lacks TFP. Altogether, these observations reveal the crucial role of P. aeruginosa TFP motility in altering S. aureus architecture. Without TFP motility, P. aeruginosa does not invade or disrupt S. aureus colonies; instead, it grows alongside them, resulting in increased compaction and altered S. aureus colony structure.

Next, we wondered how invasion changes S. aureus colony architecture and enhances competition. One hypothesis is that invasion increases the local concentration of P. aeruginosa antimicrobials within S. aureus colonies. Additionally, these anti-staphylococ cal factors could aid P. aeruginosa invasion. If the former is correct, S. aureus colonies grown in the presence of the P. aeruginosa ∆pilA would be protected from P. aeruginosa antimicrobials. P. aeruginosa secretes many factors known to inhibit or lyse S. aureus, including HQNO, a respiratory toxin that inhibits the S. aureus electron transport chain (30,36), the siderophores pyoverdine and pyochelin, which aid in iron scavenging (29,32), and an anti-staphylococcal protease, staphylolysin or LasA, which lyses S. aureus by cleaving the peptidoglycan pentaglycine cross-links (31).

Since ∆pilA had reduced invasion and disruption of S. aureus exterior structure compared with WT, we first determined if this difference is due to variations in levels of secreted antimicrobials between the P. aeruginosa strains and tested whether ∆pilA produces similar levels of exoproducts as WT. The cell-free supernatant from ∆pilA or WT P. aeruginosa was added to S. aureus to examine its growth and lysis over time. No differences were observed in either S. aureus lysis or growth rate when exposed to supernatant from WT or ∆pilA P. aeruginosa, confirming that each produces similar levels of antimicrobials (Fig. S2). Here, supernatant from P. aeruginosa ∆lasA served as a control to confirm that staphylolysin is the main driver of S. aureus lysis.

To test the hypothesis that P. aeruginosa invasion enhances competition by increas ing diffusion and local antimicrobial concentration within S. aureus colonies, we next examined S. aureus colony growth dynamics in the presence of P. aeruginosa strains lacking genes encoding antimicrobials. These include a staphylolysin mutant (∆lasA), a strain without both HQNO and siderophores (∆pqsL ∆pvdA ∆pchE), referred to as "∆AM B " (antimicrobials Bacteriostatic ), and a mutant with all four antimicrobials deleted (∆pqsL ∆pvdA ∆pchE ∆lasA) which we call "∆AM C " (AM Complete ) in Fig. 2B, C and I. The interac tions between S. aureus and these antimicrobialdeficient strains were assessed by live imaging as described for Fig. 1, and P. aeruginosa invasion and S. aureus colony height (as a proxy for biomass) were quantified. No detectable differences were observed between coculture with WT P. aeruginosa and the antimicrobial mutants for either the S. aureus colony edge height (Fig. 2C) or the number of invading cells (Fig. 2I), which suggests that these antimicrobials do not play a role in P. aeruginosa invasion or increased S. aureus colony height observed in coculture with ∆pilA. Yet, it is known that these antimicrobials can impact S. aureus growth in vitro (11,13,15,16,35,37,38,46). Therefore, we measured S. aureus colony base area to examine antimicrobial influence on growth under these conditions. The S. aureus colony area did not significantly increase when cocultured with ∆AM B or ∆lasA, compared with the WT (Fig. 2B). However, colony area did increase upon deletion of all four antimicrobials (∆AM C ), which suggests that while these antimicrobials do not influence S. aureus colony architecture, their combinatorial effect alters S. aureus growth and colony area.

To determine if S. aureus colony edge height and P. aeruginosa invasion are driven by motility alone or a combined effect of motility and antimicrobials, we deleted pilA in the ∆AM C mutant, generating ∆pqsL ∆pvdA ∆pchE ∆lasA ∆pilA (∆pilA ∆AM C ). S. aureus colony height and invasion of P. aeruginosa ∆pilA ∆AM C phenocopied ∆pilA (Fig. 2C and I), suggesting that TFP motility plays a prominent role in driving these phenotypes. Furthermore, when comparing the effect of ∆pilA on the WT background to the ∆AM C background, significant antimicrobial influence on the colony area is only observed when TFP are functional, supporting that motility may enhance antimicrobial action against S. aureus under these conditions (Fig. 2B).

Overall, these data suggest that thicker and denser S. aureus colony architecture is exclusively mediated by the absence of P. aeruginosa TFP-mediated colony invasion and that the main P. aeruginosa anti-staphylococcal factors do not substantially influence this observation. Furthermore, P. aeruginosa TFP motility may enhance antimicrobial access into the colony to fully affect S. aureus growth, revealing the important role P. aeruginosa motility plays in antagonistic interactions against S. aureus.

Although P. aeruginosa antimicrobials did not influence S. aureus structure, we next explored how colony morphology differences affect S. aureus physiological response to HQNO by utilizing a fluorescent reporter system. HQNO poisons the S. aureus respiratory chain, forcing a shift to fermentative metabolism (11); therefore, S. aureus fermentation can be used as a proxy for HQNO activity. A fluorescent transcriptional fusion to the promoter of the lactate dehydrogenase gene (P ldh1-sgfp ) was used to measure fermen tation (47). If HQNO penetrates densely packed colonies, we expect to see increased fluorescence compared with coculture with P. aeruginosa lacking HQNO production. To test this prediction, we live imaged S. aureus in mono-or coculture with P. aeruginosa and quantified the mean fluorescence intensity (MFI) per S. aureus colony over 18 hours (Fig. 3). S. aureus P ldh1-sgfp fluorescence began to increase at approximately 12 hours in coculture with WT P. aeruginosa but did not increase in ∆pqsL coculture, confirming prior reports that HQNO increases ldh expression (11). To test if fermentation increases in the absence of invasion, P ldh1-sgfp expression was quantified in coculture with P. aeruginosa ∆pilA. Notably, we observed a sharp increase in fermentation of densely packed colonies produced by coculture with ∆pilA (shown in Fig. 3A; quantified in Fig. 3B and C). One interpretation of these data is that HQNO concentrates within densely packed colonies, inducing a more dramatic change in S. aureus physiology. Additionally, since ∆pilA cells grow around and against S. aureus, it is possible that the striking increase in S. aureus fermentation is due to more HQNO-producing cells surrounding S. aureus colonies, although precisely quantifying the cell number under these conditions is technically challenging. Alternatively, the increased P ldh1-sgfp signal may result from increased cell density, independent of HQNO, potentially due to oxygen restriction within the colony. To differentiate these possibilities, fermentation was measured in the presence of a ∆pqsL ∆pilA mutant. As seen with motile ∆pqsL, the double mutant does not induce S. aureus fermentation over time (Fig. 3B), suggesting HQNO mediates this increased fermenta tion. Importantly, the phenotypes of both ∆pilA and ∆pqsL mutants were genetically complemented by expressing their respective genes in cis (∆pilA) or trans (∆pqsL) under control of an inducible promoter (Fig. S3). These findings show that HQNO likely diffuses into S. aureus colonies independently of P. aeruginosa invasion and plays a crucial role in mediating interspecies interactions by pushing S. aureus toward fermentation.

So far, we see a role for TFP motility in competition against S. aureus under conditions that constrain cells to the surface. While useful for high spatial and temporal resolution, this approach does not accurately reflect other attributes of the CF airway infection environment. Thus, we sought to determine whether TFP motility drives interactions when cocultured under conditions that mimic CF lung secretions by using artificial sputum media (ASM) (5), a modified version of SCFM2 (48). ASM captures some of the essential features of the CF environment, like constraints on movement and diffusion, that shaken liquid culture methods do not, and similar recipes have been shown to recapitulate approximately 86% of P. aeruginosa gene expression in human-expectorated CF sputum, outperforming both laboratory media and the acute mouse pneumonia model of infection (49,50).

S. aureus and P. aeruginosa (WT or ∆pilA) at a 1:1 ratio were grown statically for 22 hours and imaged with resonant scanning confocal microscopy to visualize their spatial organization. The end time point was plated for colony-forming units to assess bacterial viability. In the presence of WT motile P. aeruginosa, S. aureus was suppressed relative to its monoculture condition: very few S. aureus cells could be observed or counted (Fig. 4A and B), compared with approximately 10 8 CFUs/well recovered in S. aureus monoculture. However, when S. aureus was grown with P. aeruginosa ∆pilA, a 100-10,000-fold increase in S. aureus cells was recovered in comparison to coculture with WT P. aeruginosa (Fig. 4B). Overall, this suggests that TFP motility is necessary for effective competition with S. aureus in ASM. TFP are also necessary for P. aeruginosa biofilm formation and attachment to surfaces (51)(52)(53). However, we observed that P. aeruginosa biofilm formation and spatial organization were similar in appearance between WT and ∆pilA in coculture with S. aureus in ASM (Fig. 4A). While the CFUs/well recovered for ∆pilA were significantly lower than WT P. aeruginosa in mono-or coculture with S. aureus (Fig. 4C), the difference is modest (~95% of WT) and thus not expected to account for the increase in S. aureus survival. Next, we tested if the mere presence of TFP has a role in competition or if TFP motility is required. To differentiate between these two outcomes, a hyperpiliated, non-twitching P. aeruginosa mutant (∆pilT) was cocultured with S. aureus. This mutant lacks the main retraction ATPase of the TFP machinery, PilT, and is well documented to ineffectively retract extended pili (54). S. aureus survival in the presence of ∆pilT phenocopied ∆pilA (Fig. 4A and B), suggesting that functional TFP are necessary for competitive interactions with S. aureus in ASM. Collectively, these data demonstrate that under CF-relevant conditions, P. aeruginosa TFP motility aids in interspecies competi tion. S. aureus biofilm detaches from the surface and is blanketed by P. aeruginosa cells. Disruption was dependent on P. aeruginosa TFP motility, as WT P. aeruginosa disrupted S. aureus pre-formed biofilms significantly more than ∆pilA or ∆pilT (Fig. S4). Notably, S. aureus and P. aeruginosa ∆pilA remained segregated into monoculture aggregates, while cells were well mixed during coculture with motile P. aeruginosa (Fig. 5A, inset). These observations were consistent with results under agarose pads. We also observed that a higher number of WT P. aeruginosa colonized the surface of the coverslip compared with the ∆pilA or ∆pilT mutants in the presence of S. aureus (Fig. 5A; Fig. S4). These data suggest that TFP motility is necessary for P. aeruginosa cells to invade from the top of S. aureus biofilms and traverse through to access the coverslip, potentially disrupting and detaching the biofilms in the process. This results in a significant reduction in S. aureus viability (10-15-fold) when comparing S. aureus coculture with WT versus ∆pilA or ∆pilT (Fig. 5B). Importantly, no viability differences were observed between P. aeruginosa WT and ∆pilT (Fig. 5C). While ∆pilA shows a significant decrease in viability compared with WT, it is unlikely to have a biological influence on S. aureus growth (Fig. 5C). Altogether, these observations suggest that TFP motility enhances P. aeruginosa competitive fitness, allowing it to disrupt and potentially render S. aureus cells more vulnerable to P. aeruginosa antimicrobials.

Growing data support the hypothesis that spatial organization is crucial in shaping microbial communities and influencing community-based behaviors (4,5,39,(55)(56)(57)(58)(59)(60)(61). In this study, we found that P. aeruginosa motility plays a vital role in shaping the biogeography in S. aureus cocultures. By influencing spatial aggregation, P. aeruginosa TFP motility ultimately dictates S. aureus physiology and survival.

While P. aeruginosa antimicrobials have been well documented to influence S. aureus growth and survival (11)(12)(13)(14)(15)(16), P. aeruginosa motility in interspecies competition has only begun to be explored. We recently reported that P. aeruginosa senses S. aureus-secre ted PSM peptides from a distance by the PilJ chemoreceptor (43). Consequently, it employs TFP motility to chemotax toward S. aureus colonies or PSMs alone (42,44). In addition to chemotaxis, S. aureus PSMs also trigger a "competition sensing" response whereby P. aeruginosa upregulates type VI secretion system and pyoverdine biosynthesis pathways (44). Similarly, P. aeruginosa has been reported to utilize TFP-mediated motility to perform "suicidal chemotaxis" toward antibiotics (62). The upregulation of these common interbacterial competition pathways supports a model where P. aeruginosa senses potential danger in the environment and responds with directional twitching, while simultaneously activating defense mechanisms to combat the "enemy". Addition ally, it has been reported that P. aeruginosa upregulates antimicrobial production upon sensing N-acetylglucosamine alone or shed from Gram-positive bacteria (63).

Our single-cell level temporal analysis also revealed that P. aeruginosa TFP motility is necessary for invading and disrupting S. aureus colonies (Fig. 1). Interestingly, loss of invasion leads P. aeruginosa to grow adjacent to S. aureus colonies, potentially acting as a "wall" to prevent expansion of the S. aureus colonies, which become thicker and denser (Fig. 2). While P. aeruginosa anti-staphylococcal factors did not mediate invasion into S. aureus colonies, they did influence growth as S. aureus formed larger colonies in the absence of P. aeruginosa antimicrobials HQNO, pyoverdine, pyochelin, and staphylolysin (Fig. 2B), as expected based on prior reports (11,13,16). However, most studies have been performed with P. aeruginosa cell-free supernatant and not with live P. aeruginosa present. Imaging P. aeruginosa and S. aureus in coculture at the single-cell level has allowed us to visualize the importance of P. aeruginosa motility in their interactions and, therefore, start to build a model whereby TFP motility aids in competition by disrupting S. aureus single cells away from the colony, leaving them exposed and more vulnerable to P. aeruginosa-secreted factors (Fig. 6). Therefore, when P. aeruginosa cannot move, we hypothesize that S. aureus cells remain protected within the colony and resist infiltration of P. aeruginosa antimicrobials. Altogether, these findings provide additional support of how TFP motility can either enhance competition or foster coexistence with S. aureus.

Different S. aureus colony morphology is a consequence of limited invasion by P. aeruginosa ∆pilA, compared with WT, and shows a striking change in physiology by increasing fermentation (P ldh1-sgfp ) (Fig. 3). While we initially hypothesized that S. aureus cells remained protected from P. aeruginosa antimicrobials in the absence of invasion, these data suggest that HQNO can diffuse into S. aureus colonies and alter growth and physiology without P. aeruginosa TFP-mediated invasion. Nevertheless, deletion of antimicrobial production in the ∆pilA mutant background does not significantly improve S. aureus survival, as it does in the WT background. This observation supports the role of TFP-mediated invasion and disruption in allowing these antimicrobials to access S. aureus cells within the colonies for greater impact on its growth (Fig. 2B). Therefore, we hypothesize that without invasion and disruption, antimicrobials with higher molecular weight, such as staphylolysin (20 kDa), are precluded from freely diffusing into S. aureus colonies, while smaller compounds like HQNO (0.259 kDa) can diffuse and concentrate within S. aureus, eliciting physiological changes that could pose greater challenges for the effective treatment of infections. Importantly, P. aeruginosa TFP motility's role in mediating interspecies competition and spatial aggregation was highlighted under conditions that mimic the nutritional and viscoelastic properties of CF airways. Of note, P. aeruginosa was capable of detaching preformed S. aureus biofilms and significantly reducing S. aureus viability in a motilitydependent manner when grown in ASM. These results emphasize the importance of motility in this CF-like polymicrobial environment.

The mucoid P. aeruginosa phenotype, a common adaptation that P. aeruginosa exhibits during CF infections, is associated with decreased competition against S. aureus due to the reduced production of anti-staphylococcal factors (46). Additional studies have demonstrated that another common adaptation of P. aeruginosa linked to chronic CF infections is reduced motility (64). Interestingly, P. aeruginosa mucoid and reduced twitching phenotypes have been identified as the best phenotypic predictors of future pulmonary exacerbations in children with CF (64). Our studies revealed that impairing twitching motility hinders P. aeruginosa competitiveness and promotes coexistence with S. aureus under CF-relevant conditions. This may be a contributing explanation for why P. aeruginosa and S. aureus are still found in high numbers in people with CF (18).

Extensive discussion surrounds whether P. aeruginosa and S. aureus encounter each other in CF lungs and if they compete or coexist (4,39,46,55). CF airways are indeed a complex environment with multiple, distinct niches within; therefore, we predict that P. aeruginosa and S. aureus can coexist in some areas, while P. aeruginosa might outcompete S. aureus in others. They may be well mixed in some spaces and spatially segregated in others but still influence each other through the diffusion of secreted factors. As evidenced by both this study and others, it is clear that interspecies compe tition or coexistence can greatly depend on bacterial and host genotype and pheno type. Our in vitro data show how spatial organization can determine the outcome of microbe-microbe interactions and inform the potential interactions of these bacteria during infection. However, the CF airways are complex and involve other microorganisms and host factors that must be considered. Therefore, future experiments should explore these interactions in vivo and ex vivo to map the community biogeography and further elucidate interspecies dynamics.

Overall, our study reveals how P. aeruginosa TFP motility aids in the disruption of S. aureus colonies and biofilms, which potentiates the effect of P. aeruginosa-secreted anti-staphylococcal factors on S. aureus cells. Ultimately, P. aeruginosa motility plays a crucial and previously unexplored role in determining S. aureus outcome.

For additional details on all the methods, see the supplemental materials and methods in Text S1.

See Text S1 for details on bacterial growth conditions. A list of strains used in this study can be found in Table S1.

Markerless deletion mutants of genes in PA14 were constructed through homologous recombination as previously described (65).

P. aeruginosa and S. aureus were cocultured under agarose pads as previously described (42) and live imaged with resonant scanning confocal microscopy.