Intra-species signaling between Pseudomonas aeruginosa genotypes increases production of quorum sensing controlled virulence factors

The opportunistic pathogen Pseudomonas aeruginosa damages hosts through the production of diverse secreted products, many of which are regulated by quorum sensing. The lasR gene, which encodes a central quorum-sensing regulator, is frequently mutated, and loss of LasR function impairs the activity of downstream regulators RhlR and PqsR. We found that in diverse models, the presence of P. aeruginosa wild type causes LasR loss-of-function strains to hyperproduce RhlR/I-regulated antagonistic factors, and autoinducer production by the wild type is not required for this effect. We uncovered a reciprocal interaction between isogenic wild type and lasR mutant pairs wherein the iron-scavenging siderophore pyochelin, specifically produced by the lasR mutant, induces citrate release and cross-feeding from wild type. Citrate stimulates RhlR signaling and RhlI levels in LasR-but not in LasR+ strains, and the interactions occur in diverse media. Co-culture interactions between strains that differ by the function of a single transcription factor may explain worse outcomes associated with mixtures of LasR+ and LasR loss-of-function strains. More broadly, this report illustrates how interactions within a genotypically diverse population, similar to those that frequently develop in natural settings, can promote net virulence factor production.


Introduction 19
Genetic diversity frequently arises and persists within clonally-derived bacterial 20 and fungal populations in chronic infections and healthy microbiomes, and recent data 21 highlight that this heterogeneity can pose challenges to clearance and treatment (1-3). 22 Genotypic and phenotypic complexity has been particularly well-documented in the 23 chronic lung infections associated with the genetic disease cystic fibrosis (CF), and 24 these studies have convincingly demonstrated that within a species, a common set of 25 genes is under selection across strains and hosts (4-10). 26 P. aeruginosa loss-of-function mutations in lasR (LasR-) are very commonly 27 found in CF isolates, strains from acute infections, and from environmental sources (11-28 15). Although infection models often show that lasR loss-of-function mutants have 29 reduced virulence compared to strains with functional LasR (LasR+) in several animal 30 models (16,17), the presence of LasR-strains is correlated with worse disease 31 outcomes in acute and chronic infections (11,12). There are several possible 32 explanations for this apparent contradiction. Loss of lasR function confers some fitness 33 advantages including altered catabolic profiles (18) and enhanced growth in low oxygen 34 The difference in the effects of pyochelin on citrate levels in culture supernatants for WT 271 and ∆lasR was significant using data from four independent experiments (p<0.05). This 272 suggested WT-produced citrate was possibly involved in WT / ∆lasR co-culture 273 interactions, and that its increased release was enhanced by ∆lasR-produced pyochelin. To determine if citrate was sufficient to stimulate RhlR activity in ∆lasR, we 278 analyzed its effects on both rhlI promoter fusion activity and RhlI protein levels. We 279 found that citrate increased rhlI promoter activity in ∆lasR and that its effects were 280 dependent on the presence of RhlR (Fig. 5A). In contrast, the inclusion of citrate in the 281 medium caused a small but significant reduction in WT PrhlI activity compared to LB 282 control (Fig. 5A). 283 To determine if RhlI protein levels were influenced by citrate, we utilized an 284 arabinose-inducible rhlI-HA construct to assess RhlI protein levels and stability of  HA in the absence and presence of citrate. RhlI-HA was functional as swarming defects 286 of ∆rhlI were complemented upon expression of RhlI-HA but not by the empty vector 287 (Fig. 5B, inset). RhlI-HA protein levels were 3-fold higher in ∆lasR upon citrate 288 supplementation relative to controls (Fig. 5B). Consistent with the absence of an 289 increase in rhlI promoter activity in WT strains (Fig. 5A), RhlI-HA protein levels were not 290 higher with citrate in the ∆lasR complemented strain (∆lasR + lasR) (Fig. 5B). The 291 differential responses to citrate were also observed in LasR-and LasR+ pairs of clinical 292 isolates (CIs). LasR-CIs from acute (strain 388D) and chronic (strains DH2415) 293 infections had RhlI-HA levels 1.4-and 1.7-fold higher, respectively, in the presence of 294 citrate (Fig. 5C), whereas alterations in RhlI-HA protein levels in LasR+ CIs from acute 295 (550A) or chronic (DH2417) infections was not observed (Fig. 5D). Through this work, 296 we successfully identified citrate as a molecule in co-culture that specifically promoted 297 RhlI protein levels in LasR-strains, but not LasR+ strains, by a mechanism other than 298 transcriptional control. In order to identify transporters that could be involved in the 299 ∆lasR response to citrate, we deleted two organic acid transporters: dctA (65) and 300 PA14_51300 (66) in the ∆lasR background. The dctA gene was deleted in both a ∆lasR 301 and ∆lasR∆rhlR mutant. We found that the ∆lasR∆dctA strain still showed induction of 302 pyocyanin when co-cultured with the WT and that induction was dependent on RhlR 303 (Fig S7A). Similar results were obtained with the ∆lasR∆PA14_51300 (Fig S7B) 304 suggesting that these transporters were not required for the interaction perhaps due to 305 redundant functions of other proteins or the involvement of other import mechanisms. 306 The temporal pattern suggested RhlI protein induction preceded signal 307 amplification via the positive feedback loop of the quorum sensing transcriptional 308 network. This would be consistent with a primary effect on post transcriptional 309 modulation of RhlI-mediated RhlR activity. To begin to unravel the mechanisms by 310 which citrate promoted RhlR/I-dependent signaling and RhlI stability in ∆lasR, we 311 analyzed the role of two proteases previously found to target and degrade RhlI (i.e. Lon 312 and ClpXP) (67). Given knockouts of Lon protease have a less substantial rise in RhlR/I 313 expression in ∆lasR knockouts compared to WT (68), we focused on the role of ClpXP 314 in ∆lasR. We found that citrate induction of RhlI-HA protein levels in ∆lasR relative to 315 the LB control was dependent on the production of ClpX protease (Fig. 5E). More specifically, in the absence of ClpX, a protease shown to degrade RhlI (i.e. 317 ∆lasRclpX::TnM), RhlI-HA levels did not increase on LB + citrate relative to LB control, 318 unlike ∆lasR comparator (Fig. 5E). In LB conditions, RhlI-HA levels were 3.20 ± 2.1 fold 319 higher in ∆lasRclpX::TnM compared to ∆lasR, which mirrors the 3-fold induction 320 observed for ∆lasR on LB + citrate. No significant difference in RhlI-HA were observed 321 for ∆lasRclpX::TnM relative to ∆lasR in citrate supplemented conditions (fold change: 322 1.01 ± 0.53). In other words, as previously noted for WT strains, ClpX appeared to 323 degrade RhlI in ∆lasR, and played a role in ∆lasR response to citrate. The distinct 324 responses and mechanisms identified between LasR+ and LasR-strains under iron 325 limitation and exposure to the low-iron associated molecules, citrate and pyochelin, 326 enabled increases in antagonistic factor production beyond monoculture levels as an 327 emergent property of P. aeruginosa intraspecies interactions. 328 In support of the model that induction in RhlR signaling in response to citrate was 329 due to increased RhlI protein, we found that the induction of rhlI promoter activity (Fig.  330 5F) followed the increase in RhlI-HA levels ( Fig. 5G) in response to citrate. The 331 stimulation of rhlI promoter activity was greatest for citrate, but modest stimulation was 332 observed for other organic acids including TCA cycle intermediates (succinate, fumarate 333 and malate) and the common fermentation product acetate. In each case, the 334 stimulation of rhlI-promoter activity was accompanied by higher RhlI-HA protein levels in 335 ∆lasR, but not ∆lasR + lasR, aside from succinate ( Fig. 5F,G). Together, these data 336 may imply that organic acids, such as citrate, can serve as mediators of co-culture 337 interactions that can activate RhlR activity in LasR-strains. 338

Discussion 340
In this study, we described an emergent outcome of co-culturing LasR-and 341 LasR+ strains of P. aeruginosa in which their interactions promoted the increased 342 production of toxic exoproducts including pyocyanin and rhamnolipids (see Fig. 6 for 343 model). We determined that, in co-culture, the iron-binding siderophore pyochelin was 344 largely contributed by ∆lasR, and that exogenous pyochelin induced secretion of citrate 345 significantly more strongly in the WT than in ∆lasR. Citrate increased RhlI protein levels 346 and activated RhlR activity only in ∆lasR, but not WT cells (Fig. 6). Western blot 347 analysis of RhlI-HA expressed from a regulated promoter led us to propose that the 348 increase in RhlR signaling is due to decreased degradation of RhlI by ClpXP, a known 349 negative regulator of RhlI (68, 69). The differences in siderophore production, citrate 350 release, and RhlR/I activation between P. aeruginosa LasR+ and LasR-strains in co-351 culture environments reflect the pronounced differences between strains that drive the 352 reactivation of quorum sensing and enhanced production of secreted factors. Previous 353 studies have shown that LasR-strains increase their production of phenazines in the 354 presence of other species such as Candida albicans (22) and Staphylococcus aureus 355 (see Fig. 3B in (70)) and future work will determine if pyochelin and citrate also 356 participate in these interspecies interactions. Other microbial interactions have been 357 shown to be influenced by iron availability (71-74). Furthermore, the activation of RhlR 358 activity in ∆lasR strains that can occur in late stationary phase cultures (23, 75) may 359 relate to changes in iron or TCA cycle intermediates. While we found that WT 360 production of the diffusible autoinducers 3OC12HSL, C4HSL and PQS were not 361 required for co-culture stimulation, they clearly contributed to the enhanced activation of demonstrated previously (76). 364 The stimulatory relationship between LasR+ and LasR-strains was remarkably 365 stable as it was observed when strains were mixed within single spot colonies (Fig. 1A) 366 and when strains separated by either filters (Fig. 3) or mm distances on an agar plate 367 (Fig. 2B). The LasR-/ LasR+ interactions occurred across distinct media (Fig. S1A), 368 among genetically diverse LasR+ and LasR-clinical isolates (Fig. S1B) and over a wide 369 range of relative proportions of each type (Fig. 1C). The consequences of this 370 intraspecies interaction between genotypes may explain the worse outcomes exhibited 371 by patients in which LasR-strains are detected (12), but future studies with data that 372 include genotypes, mono-culture and co-culture phenotypes, and longitudinal outcome 373 data will be required. RhlR plays other important roles in host interactions (77) which 374 may benefit P. aeruginosa LasR-strains. The observation that rhlR mutants are rare in 375 natural P. aeruginosa isolates and that LasR-strains with active RhlR are virulent (21, 376 78) underscores the relevance of this mechanism and highlights the importance of 377 understanding how microbial interactions activate RhlR. 378 As the study of inter-and intra-species interactions progresses, it is becoming 379 increasingly clear that the environment can dictate the outcome of microbial interactions 380 (79). In fact, even the importance of QS regulation for fitness depends on nutrient 381 sources and conditions (80, 81). As ∆lasR-produced pyochelin was a key component of 382 the interaction, and pyochelin production is repressed under conditions of excess iron 383 availability, it was not surprising that the addition of iron to LB medium suppressed the 384 interaction without significantly altering the final colony CFUs or strain ratios relative to 385 LB control (Fig. S4). Siderophore-mediated iron uptake is often required in vivo (39,82,386 83) due to iron sequestration by host proteins (84-87), thus the in vivo settings could 387 support these interactions. Interestingly pyoverdine, the higher affinity siderophore, was 388 not required for the co-culture response mirroring findings that genes for the 389 biosynthesis of pyoverdine, but not pyochelin, are commonly disrupted in chronic CF 390 clinical isolates (44-46). In the absence of pyoverdine (i.e. ∆lasR∆pvdA), we observed 391 more pyocyanin in co-culture with WT than ∆lasR (Fig. 3D), and we speculate that this 392 is due to increased pyochelin production by ∆pvdA but future studies will be required to 393 test this model. If this is the case, it would be interesting to analyze the outcomes of 394 interactions over gradients of iron and other nutrients. It was interesting to find that in 395 WT / ∆lasR co-culture, heme-related proteins, hasAP, hasS, and hasD, were among the 396 top eight most upregulated genes by ∆lasR because the presence of lasR mutants and 397 heme utilization are both reported biomarkers of disease progression in CF patients (12, 398 88). Co-culture induced lasR mutant phenotypes may link these two correlative 399 observations. 400 Citrate, a TCA intermediate, is released under iron limitation as a result of 401 "overflow metabolism" (48-50, 52) and is also used by P. aeruginosa and other 402 microbes for iron acquisition due to its iron chelating properties (89). The higher 403 siderophore production by ∆lasR and stimulation of ∆lasR siderophore production in co-404 culture likely reflects different metabolic strategies between the two strains. Ongoing 405 work will investigate the mechanisms that drive differences in metabolism and iron 406 requirements in order to determine how these differences shape microbial and host 407 interactions. It is likely that Crc-mediated catabolite repression is involved in the response to citrate and the control of RhlI levels (67, 69). The existence of a mechanism 409 for the induction of RhlR-mediated QS in response to citrate and other TCA cycle 410 intermediates that are secreted when iron is limiting dovetails with reports of increased 411 expression of the P. aeruginosa quorum sensing regulon in low iron in LasR+ cells (90-412 92). This coordinate regulation may aid in iron acquisition as quorum sensing-controlled 413 phenazines, such as pyocyanin, reduce poorly soluble Fe3+ to Fe2+ and facilitate its 414 uptake via the Feo system (93). Additionally, rhamnolipids have been employed for iron 415 remediation (94, 95) which suggests their surfactant activity may increase P. aeruginosa 416 substrate iron uptake in part through hydroxyalkylquinolone-dependent mechanisms 417 (96). 418 As the presence of heterogeneous genotypes within single species populations 419 becomes increasingly appreciated, it is important to understand how commonly 420 encountered genotypes interact to influence the apparent behavior of the population. 421 Here, we show that inter-genotype interactions lead to increased RhlR signaling in lasR 422 strains; other work shows co-cultures can also influence the survival of other genotypes 423 (97). It is likely that a wide array of such interactions have yet to be uncovered. 424

Methods 427
Strains and Growth Conditions. Bacterial strains used in this study are listed in Table  428 S2. Bacteria were maintained on LB (lysogeny broth) with 1.5% agar. Yeast strains for 429 cloning were maintained on YPD (yeast peptone dextrose) with 2% agar. Where stated, 430 20 mM of indicated metabolite was added to the medium (liquid or molten agar). 431 Planktonic cultures were grown on roller drums at 37°C for P. aeruginosa. 432 433

Plasmid Construction 434
Plasmid constructs for making in-frame deletions, RhlI-HA expression, and for pqsA 435 promoter fusions were constructed using a Saccharomyces cerevisiae recombination Each condition had at least eight replicates each for two independent experiments. 459 460

Competition Assays 461
Competition assays were performed to determine the relative fitness of P. aeruginosa 462 mutants. Strains to be competed were grown overnight and adjusted to OD600 = 1. 463 Competing strains were combined with PA14att::lacZ strain in a 1:1 ratio, unless 464 otherwise stated. Following 15 s vortex, 5 µL of the combined suspension was spotted 465 on LB agar. After 16 h, colony biofilms (and agar) were cored, placed in 1.5 mL tubes 466 with 500 µL LB, and agitated vigorously for 5 min using a Genie Disruptor (Zymo). This 467 suspension was diluted, spread on LB plates supplemented with 150 µg / mL 5-bromo- imaged on glass sheet to reduce glare using Canon EOS Rebel T6i digital camera. To 480 process images, they were first cropped to remove background area surrounding each 481 plate, converted to 8-bit for particle analysis in ImageJ, and the threshold was 482 determined to count WT and ∆lasR CFU's, separately. Colony parameters were 483 collected for each individual CFU, including x, y coordinates and area for WT and ∆lasR 484 CFU lists. A simple distance calculation was made between every WT and ∆lasR colony 485 using the x,y coordinates and the minimum distance to a WT colony was plotted for 486 each ∆lasR CFU against its area value, representative of the approximate lacZ intensity. 487 488

Swarming Motility Assays 489
Swarm assays were performed as previously described in (99), with a few 490 modifications. Briefly, M8 medium with 0.5 % agar was poured into 60 x 15 mm plates 491 and allowed to dry at room temperature for 4 h prior to inoculation. LB grown cultures 492 (16 h at 37 °C) were diluted to OD600 = 1 in fresh LB, and co-cultures were mixed such 493 that ∆rhlA was at 0.7 proportion of the final cell suspension. Each plate was inoculated 494 with 5 µL of the final cell suspensions and incubated upright for 24 h at 37 °C in an 495 incubated chamber followed by 12 -16 h at room temperature. Each strain was 496 inoculated in four replicates and assessed on at least three separate days. 497

RNA Collection 499
A 200 µL aliquot of optical density normalized (OD600 = 1) cultures of PA14 or PA14 500 ∆lasR from three independent overnights were spread onto LB plates with glass beads 501 and briefly allowed to dry. Two isopore 0.2 µm PC membrane filters were stacked on 502 the lawn (37 mm diameter filter directly on lawn then 25 mm diameter filter on top), and 503 three 15 µL spots of normalized (OD600 = 1) ∆lasR cultures were spotted on the top-504 most filter. After 16 h incubation at 37 C, the top filter was collected and ∆lasR cells 505 were resuspended in 1 mL LB by 5 min of vigorous shaking on the genie disruptor. Cells 506 were pelleted for 10 min at 13,000 RPM and snap-frozen in ethanol and dry ice for RNA 507 extraction. RNA was extracted according to manufacturer's protocol with the QIAGEN 508 RNeasy Mini kit and DNase treated twice with Invitrogen Turbo DNA-Free kit. DNase-509 treated samples were prepared for sequencing with ribodepletion and library 510 preparation in accordance with Illumina protocols. Samples were barcoded and 511 multiplexed in a NextSeq run by the Dartmouth Sequencing Core. 512 513

RNA-Seq Processing 514
Reads were processed using CLC Genomics Workbench wherein reads were trimmed 515 and filtered for quality using default parameters. Reads were aligned to the P. aeruginosa UCPBB_PA14 genome from www.pseudomonas.com. Results were 517 exported from CLC including total counts, CPM and TPM. EdgeR was used to process 518 differential gene expression (100). Generalized linear models with mixed effect data 519 design matrices were used to calculate fold-change, p-value and FDR. Volcano plots 520 and heatmaps using EdgeR output (log fold-change and -log(p-value)) were made in R 521 (ggplot2 and pheatmap respectively) (101-103). GO term pathway enrichment analysis 522 was carried out using PantherDB (104). 523 524

Accession Number 525
Data for our RNA-Seq analysis of P. aeruginosa ∆lasR grown on ∆lasR or WT in co-526 culture has been uploaded to the GEO repository (https://www.ncbi.nlm.nih.gov/geo/) 527 with the accession number GSE149385. 528 529

Pyochelin extraction, quantification, and validation 530
Pyochelin was extracted based on the methods of Cox et al. (37). Briefly 50 mL cultures 531 of PA14 ∆pvdA and pyochelin biosynthesis deficient strain PA14 ∆pvdA∆pchE (negative 532 control) were grown in Chelex-treated (i.e. media treated for > 2 h with 0.5 g Chelex resin 533 per 10 mL media concentrate, followed by centrifugation and filtration with 0.22 µm filter 534 unit) LB for 16 h. Cultures were pelleted for 15 min at > 5000 RPM, and the supernatant 535 was passed through a 0.22 µm filter unit. The cell-free supernatant was acidified to pH 2 536 with 10 N HCl. For extraction, 5 mL ethyl acetate was added per 50 mL acidified solution 537 in a separatory funnel. The top ethyl acetate layer was concentrated using a speedvac 538 and quantified in 50 / 50 methanol:dH2O in a 1 mm quartz cuvette at 313 nm. Absorbance was checked from 200 -600 nm for expected peak profile. For some extractions, a 5 µL 540 aliquot of ∆pvdA extract in 50:50 methanol:dH2O was viewed under ultraviolet light for 541 expected fluorescence relative to ∆pvdA∆pchE extract that dissipated upon 10 µM FeSO4 542 supplementation. The concentration was determined using the molar extinction 543 coefficient at 313 nm in 50 / 50 methanol:dH2O in a 1 mm quartz cuvette (37). Upon 544 quantification, concentrated extract was lyophilized using rotovap to yellow resin, and 545 used within 2 days of initial extraction by resuspension in LB for supplementation. As 546 validation of biological activity, 10 µL of extract was spotted along with EDTA chelator as 547 control on CAS agar prepared as described in. 548 549

Citrate Quantification 550
Citrate was quantified from cell-free supernatant of 5 mL LB-grown cultures inoculated 551 from single colonies. Cultures were grown in quadruplicate and incubated on a roller 552 drum for 16 h. OD 600 nm was recorded, and cultures were pelleted for 15 min at > 5,000 553 RPM. The supernatant was passed through a 0.2 µm syringe filter unit. Citrate in the 554 filtered supernatant was quantified according to manufacturer's "manual assay" protocol 555 (Megazyme) in ½ reactions. The extinction coefficient at 340 nm in a quartz cuvette was 556 used to quantify concentration of citrate relative to OD 600 at the end of the 5 min 557 enzymatic reaction. Citrate standard and blank media conditions were included in every 558 assay. The average for 4 replicates for each experiment was reported across 3 -4 559 independent days. 560 561

Beta galactosidase assays 562
Cells with a promoter fusion to lacZ -GFP integrated at the att locus were grown in 5 mL 563 cultures of LB at 37°C for 16 h. The cultures were diluted to a starting OD 600 of 1 and 5 564 µL were spotted onto LB agar plates ± 20 mM pH 7 citrate (or other specified 565 metabolite) in triplicate. After 24 h (or other indicated time) at 37 °C, colony biofilms 566 were cored, resuspended in 500 µL LB by vigorous shaking on the Genie Disrupter for 5 567 min as previously described, and β-Gal activity was measured as described by Miller 568 (105). The average for each experiment was reported across 3-4 independent days. 569 570

Western Blot 571
Strains were grown in LB broth under selection (60 μg / mL gentamycin or 60 μg / mL 572 carbenicillin as appropriate) for 16 h at 37 C on a roller drum, and 5 μL of culture was 573 spotted onto LB plates under selection with 0.2% L-arabinose (v/v) and +/-20 mM 574 indicated carbon source. Inoculated plates were incubated at 37 C for 16 h. Colony 575 biofilms were resuspended in 325 μL of Laemmli buffer without reducing agent and 576 heated at 100 C for 15 min. Protein was quantified on 1:10 dilution of protein sample 577 according to standard procedure via Thermo scientific BCA Protein Assay Kit. Reducing 578 agent was added and samples were run on a 4 -15% SDS gradient gel (Bio-Rad) at 60 579 V for 40 min followed by 110 V for 45 min. After SDS page electrophoresis, protein was 580 transferred onto LF-PVDF (Bio-Rad) using the mixed molecular weight option on a turbo 581 blot apparatus (Bio-Rad). After transfer the membrane was dried, rehydrated, and then 582 a total protein stain was run according to manufacturer's procedure (Li-Cor). Following 583 protein quantification, the membrane was incubated in TBS blocking buffer (Li-Cor) for 1 584 h, and then purified anti-HA mouse monoclonal antibody (Biolegend) in TBS blocking 585 buffer (1:2,500 dilution) for 1 hr. Following primary antibody detection, the membrane 586 was washed 4 times in TBST 0.1%. Secondary detection was done by incubation with 587 goat anti-mouse in TBS blocking buffer (1:15000 dilution) for 1 hour in the dark. 588 Following detection, the membrane was washed 3 times in TBST 0.1% and once in 589 TBS. The membrane was then dried and imaged on the Li-Cor Odyssey CLx imager 590 relative to REVERT total protein stain. NanoString were carried out at Dartmouth Medical School in the Genomics Shared 600 Resource, which was established by equipment grants from the NIH and NSF and is 601 supported in part by a Cancer Center Core Grant (P30CA023108) from the National 602 Cancer Institute. We also thank Georgia Doing for preliminary RNAseq analysis, Pat 603 Occhipinti for swapping the antibiotic marker on the rhlI-HA expression vector, Carla 604  -(top) and co-cultures (bottom). B. Pyocyanin levels quantified for cultures described in A; ****, p<0.001. C. Pyocyanin production for wild type cocultures with ∆lasR or ∆lasR complemented with the lasR gene at the native locus (∆lasR + lasR) across several initial ( i ) proportions on LB for 20 h. D. Final proportions quantified after 16 h growth for WT and ∆lasR co-cultured with a WT tagged with lacZ. Experimental setup as described in C.

Figure 2. P. aeruginosa WT induces RhlR/I dependent activity in ∆lasR even in the absence of WT AHLs. A.
Pyocyanin production by monocultures and WT co-cultures of ∆lasR and ∆lasR derivatives that are deficient PQS or RhlR/I quorum sensing on LB after 24 h growth. B. Promoter activity, quantified by relative pixel intensity of single cellderived colony forming units (CFU) in coculture with untagged WT CFU for ∆lasR P pqsA -lacZ (grey) and ∆lasR P rhlI -lacZ (black). Inset shows representative CFUs for RhlR-dependent ∆lasR P rhlI -lacZ activity when in monoculture and co-culture with WT (red circles). C. Mono-culture and ∆lasR co-culture images for ∆pqsA, ∆rhlI, and ∆lasI∆rhlI on LB after 48 h, rather than 24 h as in panel A, due to slower color development.  Figure 3. Biosynthesis of the co-culture-induced iron scavenging siderophore pyochelin is required in ∆lasR for pyocyanin over-production when cultured with wild type (WT). A. Scheme for the collection of RNA from ∆lasR colony biofilms grown above a lawn of WT or ∆lasR. B. Volcano plot of ∆lasR expression data with each point representing the log 2 (∆lasR grown on WT / ∆lasR grown on ∆lasR) expression and -log 10 (P Value) of a single gene. Genes involved in pyoverdine (blue) and pyochelin (green) iron acquisition systems are indicated. ccmC and ccmF (indicated with arrows) of the pyoverdine GO term are involved in c-type cytochrome biosynthesis, and strains with knockouts of these genes are reported to produce more pyochelin. C. Mono-and co-cultures with ∆lasR strains deficient in pyoverdine (∆pvdA) and/or pyochelin (∆pchE). Colonies are visualized under ultraviolet light (UV) in order to see fluorescent siderophores. D. Pyocyanin production visualized for the colonies shown in panel C. E. Pyocyanin production by siderophore deficient strains grown in mono-and co-culture on LB with (+PCH) or without (LB) pyochelin-containing extract. Colonies were grown in a 12 well plate and imaged after 48 h. F. Wildtype and ∆lasR mixed colony biofilms grown on LB (-) or LB supplemented with either 10 or 100 µM FeSO4 visualized under ambient (top) and UV (bottom) light.    Citrate released by wild type in co-culture stimulates RhlR/I dependent activity (3.) by stabilizing RhlI protein in ClpXP protease dependent mechanism (4.) to promote the production of antagonistic factors like pyocyanin toxin and rhamnolipid surfactant.