University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Faculty Publications in the Biological Sciences Papers in the Biological Sciences

5-1990

PYOVERDINE PRODUCTION BY AERUGINOSA EXPOSED TO METALS OR AN OXIDATIVE STRESS AGENT

Kim-Hien Thi Dao University of Nebraska-Lincoln

Katherine E. Hamer University of Nebraska-Lincoln

Christine L. Clark University of Nebraska-Lincoln

Lawrence G. Harshman University of Nebraska - Lincoln, [email protected]

Follow this and additional works at: https://digitalcommons.unl.edu/bioscifacpub

Thi Dao, Kim-Hien; Hamer, Katherine E.; Clark, Christine L.; and Harshman, Lawrence G., "PYOVERDINE PRODUCTION BY EXPOSED TO METALS OR AN OXIDATIVE STRESS AGENT" (1990). Faculty Publications in the Biological Sciences. 274. https://digitalcommons.unl.edu/bioscifacpub/274

This Article is brought to you for free and open access by the Papers in the Biological Sciences at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Faculty Publications in the Biological Sciences by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Dao, Hamer, Clark & Harshman in Ecological Applications (May 1999) 9(2). Copyright 1999, Ecological Society of America. Used bypermission.

May 1999 STRESS IN ECOLOGICAL SYSTEMS 441

Ecological Applications, 9(2), 1999, pp. 441-448 C) 1999 by the Ecological Society of America

PYOVERDINE PRODUCTION BY PSEUDOMONAS AERUGINOSA EXPOSED TO METALS OR AN OXIDATIVE STRESS AGENT

KIM-HIEN THI DAO, KATHERINE E. HAMER, CHRISTINE L. CLARK, AND LAWRENCE G. HARSHMAN' School of Biological Sciences, Universityof Nebraska-Lincoln,Lincoln, Nebraska 68588 USA

Abstract. Siderophoresare low molecular mass compounds used by many microor- ganisms to scavenge dissolved ,which is typicallyrare in environments.Pseudomonas aeruginosa (PAO1) was exposed to metals and methylviologen in low-iron medium in orderto monitorthe abundance of a specificsiderophore (pyoverdine) and monitorgrowth over time.In thisstudy it was discoveredthat cadmium can stimulatepyoverdine abundance. Cadmium may interactwith and sequester the Fur regulatoryprotein, which represses siderophoresynthesis under conditions of ironsufficiency. Mercury did not stimulatepyov- erdineproduction at concentrationsthat strongly inhibited bacterial growth, suggesting that pyoverdineproduction is not governedas partof a generalstress response. Methyl viologen is an oxygen radical generator,and it was discovered thatexposure to it decreased pyov- erdineproduction. Decreased pyoverdineproduction may be a mechanismfor reducing the iron potentiationof oxygen toxicity.It is hypothesizedthat factors stimulating production of siderophoresmay increase the susceptibilityof microorganismsto oxidativedamage. In general,our researchsuggests new opportunitiesfor predicting ecotoxicological outcomes based on understandingof molecular mechanismsand the effectof xenobiotics or stress factorson fundamentallyimportant microbial processes. Key words: cadmium,mercury; methyl viologen; oxidativestress; Pseudomonasaeruginosa; sid- erophore.

INTRODUCTION are model systemsfor the studyof physiologicalpro- cesses. One example of the use of microbes for this It is generallyaccepted thatour abilityto make pre- approach is the studyof phosphoenolpyruvatecarbox- dictionsabout theecological effectof pollutantsis very ylase regulation,which is motivatedby the important limited (Moriarty1988). Studies of the effectsof xe- role of this enzyme in the metabolismof (often lim- nobiotics or other stressorson critical physiological iting) dissolved organic matterin aquatic ecosystems processes in organisms may be especially useful for by heterotrophicbacteria (Overbeck 1990). The effect thepurpose of improvingour abilityto predictthe eco- of pollutantsand stressorson critical aspects of mi- logical impact of pollutants and stressors. The as- crobial physiologyis understudiedwith respect to the sumptionis thatphysiological functionis the logical importanceof microbes in the functionof ecological point of integrationfor ecotoxicological studies. Spe- communities.Such studies are valuable for the devel- cifically,the impact of pollutants,or environmental opmentof the fieldof ecotoxicology. stress,on physiologyis relevantto populationgrowth, Iron acquisitionis an essentialphysiological process interactionbetween species, relative fitness,indicator for organisms.Iron is almost universallyrequired for expression,and theinteraction between organisms life, yet it is poorly soluble in the form of polyhy- and the environment.Physiological measures thatare complexes at neutralpH (Neilands 1981). fundamentallyimportant and connected with specific droxylated In some regions of the open ocean iron availabilityis environmentalstates are attractivecandidates for study. population size Microbes are obvious candidates for such studies the factor that limits phytoplankton In order to meet this demand for iron, because theyplay an essential role in ecological com- (Kerr 1994). low molecularmass munities.Moreover, they are capable of modifyingthe microbes,and otherorganisms use environmentby theirmetabolic processes. This power compounds,collectively called ,to scav- is manifestin microenvironments,or on a largerscale, enge for iron (Guerinot 1994, Neilands 1995). Sider- as in geochemical change. In addition,certain bacteria ophores are importantin relationshipto the growthof aerobic and facultative anaerobic bacteria, fungal growth,plant growth,as well for theireffects on the Manuscriptreceived 20 October 1997; accepted 27 April 1998; finalversion received 8 June 1998. For reprintsof this virulenceof plant and humanmicrobial diseases (Nei- Invited Feature, see footnote1, page 429. lands 1981, Neilands and Leong 1986, Barton and 1Address corresDondenceto this author. Hemming 1993, Payne 1993).

441 Ecological Applications 442 INVITED FEATURE Vol. 9, No. 2

Pseudomonas aeruginosa is associated withaquatic low-iron medium that promotes siderophoreproduc- and otherhabitats, and is also known as an opportu- tion. Our experimentswere designed to address the nistic of humans. This species belongs to a questions of whetheradded metals or an oxidative group of ,the fluorescentpseudomo- stress agent could stimulateor suppress the existing nads, thatare widely distributedacross a range of hab- level of pyoverdineproduction in low iron medium. itats. Among the fluorescentpseudomonads, P. aeru- Growthin SSM was estimatedby measurementof tur- ginosa is the taxon best known in termsof genetics bidityat an optical densityof OD600 nm(Perkin-Elmer and physiology.P aeruginosa produces two siderop- SpectrometerLambda 3B). All solutions and media hores;pyoverdine is predominant,and has a muchhigh- were preparedusing double-distilleddeionized . er affinityfor iron than the othersiderophore, which Chemicals were fromSigma Chemical Company, St. is called pyochelin.The structureof pyoverdineis pre- Louis, Missouri unless otherwisestated. sentedin Briskotet al. (1989). Pyoverdineis a containingeight amino acids: 1 L-arginine,1 L-lysine, Metal assays 2 L-threonines,2 L-6N-hydroxyornithines,and 2 D- In orderto test the hypothesisthat pyoverdine pro- serines. The octahedral coordinationof Fe(III) is ac- ductionby P. aeruginosa is influencedby thepresence complishedby two 6N-hydroxyornithinesand an (IS)- of heavy metals,dose response assays were conducted 5-amino-2,3 -dihydro-8,9-dihydroxy-1H-pyrimido- in whichgrowth and pyoverdineproduction were mon- [1,2-a]-1-carboxylicacid providinghydrox- itored over time in samples taken fromSSM culture. ymate and catecholate donor oxygen atoms For these assays, PAO1 was inoculated froman over- (Wendenbaumet al. 1983). nightculture into 13 x 100 mmborosilicate glass tubes The presentstudy addresses the question of whether (Fisher Scientific,Pittsburgh, Pennsylvania) containing exposureto xenobioticstress factors results in a change SSM. All glassware used for cultures or assays was in pyoverdineproduction by P. aeruginosa. Two of the soaked in 0.1 mol/LHCl (MallinckrodtChemical, Par- metals employedin this study,cadmium and mercury, is, Kentucky)overnight to removecontaminating met- are significantenvironmental pollutants. In addition, als, followed by rinses withdouble-distilled water be- experimentswere conductedto determineif a nonmetal fore autoclave sterilization.Overnight cultures were oxygen radical generator,methyl viologen (paraquat), producedby inoculatinga PAO1 colony fromPAF me- could influencethe production of pyoverdine.Paraquat dium into a 50-mL long-neckculture flask with 25 mL is well establishedas a radical generator(Hassan and SSM. The growthperiod was - 16 h at 25?C withmod- Fridovich 1979), and is used routinelyto produce ox- erate shakingat 150 cycles/min. idative stress.Since iron is knownto participatein the Afterovernight growth, cells were centrifugedfor 5 generationof oxygen radicals, exposure to this com- min at 59 x 103 m/s2at room temperature.The pellet pound was motivatedby the issue of iron management was resuspendedin sterile0.85% NaCl. The centrifu- underconditions of oxidative stress. gationand resuspensionsteps were repeated. The pellet MATERIALS AND METHODS from the second centrifugationwas resuspended in 0.85% NaCl to a standardfinal OD600 nm of 0.200. and Pyoverdinemeasurement bacterial growth A 20-j,L aliquot of thisbacterial suspension was used The fluorescentchromophore group of pyoverdine as the inoculum into each 13 X 100 mm assay tube producesa characteristicpattern of excitationand emis- containing3 mL of SSM. The controltubes contained sion (Meyer and Abdallah 1978). In the presentstudy, 3 mL SSM withno added metals. The treatmenttubes a Perkin-Elmer(Oak Brook, Illinois) LS-5B lumines- contained SSM with added metal chlorides (CdCl2, cent spectrometerwas used to qualitativelymeasure HgCl2,ZnCl2). For cadmiumand zinc, the experiments the abundance of pyoverdineby measuring fluores- were conductedat finalconcentrations of 0.125, 0.25, cence (excitation405 nm and emission450 nm), which 0.50, and 1.00 mmol/L.Solutions were prepared the is functionallyequivalent to an establishedmethod of day before inoculationof the assay tubes. The exper- pyoverdineabundance determination(Meyer and Ab- iments with these metals were conducted four times. dallah 1978, Cox and Adams 1985, Visca et al. 1992). Mercuryis highly toxic, and thus the concentrations Low concentrationof iron stimulatespyoverdine pro- employed(0.003-0.20 ,umol/L)were much lower than duction,and higherconcentrations, such as 10 ,umol/L the othermetals. Replicate experimentsusing mercury of supplementaliron, allows growthof P. aeruginosa were conductedwith an extensiveseries of concentra- with no siderophoreproduction (Stintzi and Meyer tions for the purpose of bracketingthe transitionto 1994). conditionsthat reduced growthof PAO1 in SSM. The P. aeruginosa strain used in this study was Growth (OD600) and the qualitative estimate of PAO1. Colonies were obtained by streakplating from pyoverdineproduction derived from the measure of -80?C 15% glycerol storage onto Pseudomonas Agar ,excitation at 405 nm (ex405) and emis- F (PAF) medium. For aqueous culture, cells were sion at 450 nm (em450), were recordedat 8-h intervals grownin standardsuccinate medium (SSM). SSM is a duringa 64-h incubationperiod. The experimentwas May 1999 STRESS IN ECOLOGICAL SYSTEMS 443

tachedto a vacuum source and thecontents freeze dried 25 to a dry powder. The supernatantpowder was resus- a) 2015C= pended in distilled water and solvent extractedas de- 10 - -- scribed by Meyer and Abdallah (1978). n Z~~~~~inc Iron is known to quench pyoverdinefluorescence 5 (Abdallah 1991). For the fluorescencequench assay, semipurifiedpyoverdine was measuredbefore and after 0 0.2 0.4 0.6 0.8 1.0 1.2 additionof differentmetals into solution.Semipurified Metal Concentration(mmol/L) pyoverdinewas standardizedwith distilled water to FIG. 1. Mean slope of pyoverdineproduced (log excita- -1000 luminometerunits (ex405em450) beforeaddi- tion 405 nm, emission 450 nm/logOD600) as a functionof tion of the test metals. added metal. Pseudomonas aeruginosa was continuouslyex- posed to metals and sampled over time in standardsuccinate Methylviologen exposure medium. The mean slope with no added metal was 6.85. In orderto test the hypothesisthat pyoverdine pro- designed so that there were replicate tubes for each duction is influencedby the presence of methylviol- assay time point,and only a single growthand pyov- ogen (paraquat), assays were conducted in which erdine reading was taken from any one tube. Assay growth and pyoverdine production were monitored tubes were not reused for successive readings. Lumi- over time in SSM culture. Overnightcultures were nometermeasurements (ex405em450) were typically grown,washed, and standardizedas described for the obtained froma 1:1000 dilution,but rangingfrom di- metal assays. From a 100-mmol/Lstock solution of lutionsof 1:100 to 1:10 000 dependingon the level of methylviologen, dilutions were made intoSSM to final fluorescencepresent in a sample. methylviologen concentrationsof 100 Fmol/Land 500 Samples fromSSM medium,and SSM mediumwith VLmol/L.Control tubes employed SSM withoutaddition the high concentrationof cadmium,were testedto de- of methylviologen. The tubes with SSM were inocu- terminehow much of thecharacteristic pyoverdine flu- lated with 20 p1Lof standardizedovernight culture of orescence was quenched by additionof iron. The pur- PAO1, as described for the metals assay. It is known pose was to estimate the proportionof residual fluo- that P. aeruginosa does not metabolize paraquat rescent signal that could not be attributableto pyov- (Brown et al. 1995), and presumablymetabolism of the erdine presentin samples. The method was to dilute compound does not play a role in these experiments. 64-h postinoculationsamples to -1000 luminometer These assays were repeateda total of fourtimes. The unitsat ex405em450 and thenadd iron.Iron was added samplingintervals and totalduration of incubationwas to a finalconcentration of 1.0 mmol/L,which has been longerthan employed in the metalassays, because pre- shown in this studyto quench almost all of the char- liminaryexperiments indicated thatit requiredlonger acteristicfluorescent signal from semipurifiedpyov- to reach a reasonably consistentlevel of pyoverdine erdine. abundance.

Semipurificationof pyoverdineand RESULTS metal quench tests Metal assays Pyoverdinewas semipurifiedto determineif themet- als we employed in assays quenched the fluorescent The effectsof cadmiumand zinc on pyoverdinepro- signal used to estimate pyoverdineabundance. This ductionare shown in Fig. 1. For each compoundcon- was accomplishedin a mannersimilar to thatdescribed centration,the average slope of the log of fluorescence in Meyer and Abdallah (1978). Overnightgrowth of (ex405em450) divided by log OD600 (pyoverdinepro- PAO1 was obtained as described forthe metals assay. ductionadjusted by growth)is presentedfor each con- Bacterial cells were pelleted,washed, and dilutedto a centrationof a testedmetal. The average slope forthe finalOD600 of 0.200 in 0.85% NaCl. A 1.0-mL aliquot controls,SSM withoutadded metal, is 6.85. To show of the overnightculture was used as the inoculuminto theeffect of metalson thegrowth response, the average a 2-L Erlenmeyerflask containing 1.0 L of SSM. The culturein this flaskwas incubatedat 25?C with mod- TABLE 1. Mean slope of growthresponses (log OD600) of erate shaking for 72 h. Thereafter,the cells were pel- Pseudomonas aeruginosa continuouslyexposed to different leted by centrifugationat 127.5 x 103m/s2 for 10 min metals in standardsuccinate medium. at 25?C. The supernatantwas retainedand concentrated Concentration(mmol/L) by lyophilization.For lyophilization,450 mL of culture supernatantwas used to shell-coatthe inside of bottles Metal 0.125 0.25 0.50 1.0 by rotationand freezingon theinside surfaceof a 1200- Cadmium 0.018 0.017 0.013 0.005 mL lyophilizationbottle using a dry ice-ethanolbath Iron 0.027 0.027 0.023 0.014 Zinc 0.019 0.019 0.020 0.015 applied to the exteriorof the bottle. Bottles were at- 444 INVITED FEATURE Ecological Applications Vol. 9, No. 2

TABLE 2. Mean slope of the growthresponse and pyoverdineproduction by Pseudomonas aeruginosa exposed to mercury concentrationsin standardsuccinate medium for 64 h.

HgCl2 concentration(pmol/L) Measure 0 0.0031 0.0062 0.0125 0.025 0.050 0.100 0.200 Mean slope of growth(log OD600) 0.286 0.325 0.340 0.332 0.362 0.305 0.252 0.212 Mean slope of pyoverdineproduc- tion (log excitation405 nm, emis- sion 450 nm/logOD600) 0.108 0.143 0.127 0.126 0.148 0.124 0.087 0.068 slope of the log of OD600 measurementsis presented mmol/LFeCl3 the ex405em450 measure of semipure in Table 1. pyoverdinewas quenched 99.8%. At the same concen- In the presentstudy, zinc did not have a substantial trationof iron using ex405em455, the fluorescenceof effecton growthor pyoverdineproduction, but cad- semipurepyoverdine was quenched 99.7%. At a final miumwas observed to affectboth. The linearfit of the concentrationof 0.1 mmol/LFeCl3 and ex405em450, slope across a range of zinc concentrationsis not sig- thefluorescence of semipurepyoverdine was quenched nificantlydifferent from a slope of zero (P = 0.3950, 96.9%. At the same concentrationof iron using df = 1). In contrast,cadmium acts to stimulatecell- ex405em455, the fluorescenceof semipurepyoverdine density-adjustedpyoverdine production. In the case of was quenched 97.3%. At 0.1 mmolIL finalconcentra- cadmium,the linear fitof the slope of pyoverdinepro- tionof iron,mercury quenched fluorescence 7.5%, cad- duction is significant(P < 0.0001, df = 1), as is the mium 2%, and zinc <0.5%. Given the concentrations fitto a quadraticslope (P = 0.0075, df = 1). The slope of mercury,cadmium, and zinc in solution afterdilu- produced by 0.5 mmol/LCd is 140% greaterthan the tion of SSM cultures for fluorescencereadings, it is control and 1 mmol/L Cd is 300% greaterthan the unlikely that the metals significantlyquenched fluo- control.The two higherconcentrations of Cd stimulate rescence and thusimpacted the measureof pyoverdine pyoverdineproduction. This interpretationis supported abundance in the experimentaltreatments. Moreover, by the statisticalanalysis, and inspectionof the data thereis no evidence thatcadmium increases the fluo- suggests that the response is somewhat greaterthan rescent signal as an explanation for the increase in loglinear (Fig. 1). There is evidence that the higher pyoverdineabundance when P. aeruginosa is exposed concentrationsof cadmium are reducing the growth to cadmiumin SSM. response (Table 1). Cadmium stimulatesproduction of Samples (64 h postinoculation)from the medium pyoverdine,but a question arises as to whetherthis is withthe highest concentration of cadmiumand control a nonspecific effect common to growth-inhibited SSM were testedto determinehow much of the char- (stressed) cells. acteristicpyoverdine fluorescence was quenchedby ad- Table 2 presentsthe effect of mercuryconcentrations dition of iron. At a finalconcentration of 1.0 mmol/L on growthand pyoverdineabundance. The mean slope FeCl3 and fluorescencereadings at ex405em450, con- values of log OD600 growthand log fluorescenceat trolsample fluorescencewas quenched 94.4% and the ex405em450 divided by log OD600 (pyoverdinepro- fluorescencefrom the SSM with the highest concen- duction adjusted by growth) are shown for the 64-h trationof cadmiumwas quenched 95.6%. At the same growthperiod for each mercuryconcentration. In the concentrationof iron using ex405em455, controlsam- concentrationrange from0.003 to 0.05 ,umol/L,the ple was quenched 88.4% and the fluorescencefrom the effectof mercuryrelative to the controlappears to be SSM with the highestconcentration of cadmium was a slightincrease in growthand slightdecrease of the quenched 93.5%. Overall, the residual characteristic measure of -density-adjustedpyoverdine abun- fluorescencethat cannot be quenched by the addition dance. The higherconcentrations of mercurydecrease of iron is small, suggestingthat in thisregard the mea- growth,as indicatedby Table 2 and supportedby the sure of pyoverdineabundance is not subject to a bias fact thatthe 64-h OD600 was at least 10-foldlower in sufficientto change the interpretationof the results. the presence of the highestconcentration of mercury There is no indication that exposure to cadmium in- comparedto controlgrowth (data notshown). The mea- creases the proportionof this residual fluorescence. sure of pyoverdineadjusted by cell densitydecreases at higher mercuryconcentrations (Table 2). Conse- Methylviologen assay quently,it is apparentthat there is no universalasso- Fig. 2 presentsthe pyoverdineproduction response ciation between stress (reduction of growth) and an of PAO1 cells to concentrationsof increase in pyoverdineproduction. methylviologen. The mean variatesare standardizedto the controlval- Metal quench tests ues to emphasize thepattern of responseas a summary The quench effectof ironand othermetals was tested of fourindependent experiments. In each experiment, on semipurepyoverdine. At a finalconcentration of 1.0 therewas a significantreduction in theOD600-adjusted May 1999 STRESS IN ECOLOGICAL SYSTEMS 445

2.0 .

-- 1.8 * -- control 100 ,mol/L a) 1.6 -A- 500 ,mol/L S._ D 1.4 0 0~

U\\CZ

~0.8

>? 0.6

10-12*4-8729- . 2 0.4

0.2

0 12 24 48 72 96 120 Time (hours) FIG. 2. Mean fluorescence(excitation 405 nm, emission 450 nm) adjusted by cell density (OD600) as a measure of pyoverdineproduced by Pseudomonas aeruginosa continuouslyexposed to methylviologen in standardsuccinate medium. The means of the treatmentvariates are presentedrelative to the control;the controlwith no methylviologen added was set at 1.0.

fluorescence.The level of supportfor significant treat- DISCUSSION menteffects is P = 0.0004 (experiment1), P = 0.0495 We are the firstto documenta stimulatoryeffect of (experiment2), P = 0.0001 (experiment3), and P = cadmium on pyoverdineproduction, and also the first 0.0039 (experiment4). The data indicatethat exposure to documentthat exposure to methylviologen can de- to methylviologen reduces pyoverdineproduction by crease siderophoreproduction. In this discussion, we PAO1 in SSM. Two earlierexperiments (Hamer 1997) argue thatthese effectsmay be general,and we predict indicated that 100 pLmol/Lmethyl viologen reduced a negativesynergistic effect of cadmiumand paraquat fluorescence-50% relative to to controls,compared based on theeffect of cadmiumon themolecular mech- 60% in the experimentsshown in Fig. 2, and 500 anisms thatcontrol synthesis and ironpo- pLmol/Lmethyl viologen reducedfluorescence by 65%, tentiationof oxidative stress. compared to -80% in the experimentsshown in Fig. Aspects of the mechanismof microbial control of 2. From the earlier experimentsit was observed that siderophoreproduction may be highlyconserved. First plated cells taken from the last time point of 500 describedin Escherichia coli, siderophoresynthesis is ,umol/Lmethyl viologen culturesshowed lower levels negativelyregulated by iron interactionwith Fur a re- of pyoverdineproduction than control PAO1 when re- pressorprotein (Bagg and Neilands 1987). Specifically, introducedinto SSM aqueous medium (Hamer 1997). Fe(II) is the thatinteracts with Fur, which then This observation suggests that exposure of cells to binds to an operatorsuppressing expression of sider- paraquat selects for genetic variantsthat produce less ophore synthesisgenes. Fur and fur gene homologs pyoverdine.Strong selection is consistentwith the re- have been characterizedfrom P. aeruginosa (Prince et duction in cell densityobserved at early time points al. 1991, 1993) and other taxa (Guerinot 1994). The after inoculation. The apparent temporal oscillatory DNA sequences requiredfor iron regulationmay also patternof the treatmentpyoverdine measurements pre- be conserved (Neilands 1995, Rombel et al. 1995). sentedin Fig. 2 are not a significantfeature of thedata, Moreover,in E. coli, an enterobactin(siderophore) iron given the standard errors of the 48- and 96-h time complex is taken into the cell by a specific points.Overall, thereduction of pyoverdineproduction protein(Raymond 1994). Similarly,after pyoverdine in response to oxidative stressis interpretablein terms formsa complex withiron, membrane-bound receptors of an adaptive response thatreduces theiron-mediated participate in ferripyoverdineuptake (Meyer et al. potentiationof oxygen toxicity. 1990, Poole et al. 1993). In general,when iron is lim- 446 INVITED FEATURE Ecological Applications Vol. 9, No. 2 ited, gram-negativebacteria produce multiple outer gene (Hofte et al. 1994). Their studiesinvestigated ac- membraneproteins as receptorswhich take up the fer- tivationof siderophoreexpression in the presence of risiderophorecomplex, perhaps in a conservedmanner relativelyhigh iron concentration,whereas our study by interactionwith the TonB outer membraneprotein was an investigationof the possible stimulatoryeffect (Guerinot 1994). As a resultof mechanismconserva- of zinc when iron concentrationis sufficientlylow to tion, data on the effectsof xenobiotics and otheren- allow pyoverdineproduction. Noteworthy in this re- vironmentalstressors on siderophoreproduction in P. gard is the data showing that 100 VLmol/L(or less) of aeruginosa may extendto a range of microorganisms. zinc in low iron mediumdid not stimulatepyoverdine Various factorsare known to influencesiderophore production by P. aeruginosa PAO1 (Visca et al. 1992). productionin fluorescentpseudomonads. As summa- Alternatively,there may be differencesbetween the rized by Meyer and Abdallah (1978), studies have 7NSK2 strainused by Hofte et al. (1993, 1994) and shown thatthe natureof the carbon source in culture PAO1 used in our studyand the work by Visca et al. medium can influenceproduction of pyoverdine-like (1992). Finally, differentialoutcomes may be derived pigmentproduction. For P. aeruginosa, succinate was fromthe use of plate assays (Hofte et al. 1993, 1994) established as an optimal medium for productionof vs. theaqueous cultureassays used in thepresent study. this class of compounds (King et al. 1948) and this is In supportof this possibility,a recombinasemutation likely to be a result of an iron requirementfor the continuedto synthesizepyoverdine in LB brothsup- enzyme succino-dehydrogenase(Stinzi and Meyer plementedwith zinc, but not on an LB plate supple- 1994). Iron plays a pivotal role in controllingpyov- mentedwith zinc (Hofte et al. 1994). erdineproduction by P. aeruginosa (Totterand Mose- Physiological measures can be used to detect and ley 1953) and othermicrobes. integratethe biological impactof stressfactors. In spite An additional layer of control of siderophorepro- of theintuitive appeal ofphysiological indicators, Bayne duction has been documentedin fluorescentpseudo- et al. (1979) has noted thatgeneral measures such as monad taxa includingP. aeruginosa (Leong et al. 1991, growthtend to obscure the identityof specific stress Cunliffeet al. 1995, Miyazaki et al. 1995). Totterand factors.Pyoverdine production is a specificphysiolog- Moseley (1953) presentedevidence thatpenicillin, but ical response that may provide some evidence about not streptomycinor chloromycetin,increased the pro- the natureof the stressor.However, it is increasingly duction of pyoverdine-likepigment production in P. apparent that the P. aeruginosa Fur regulatory system aeruginosa. Partiallybased on the workof Mergeay et is complex, involving an array of coordinatelyregu- al. (1978), Hofte et al. (1989) suggested that modu- lated as well as various sigma factorsand other lation of siderophoreproduction was associated witha global regulators(Tardat and Touati 1993, Cunliffeet general stress response in various pseudomonads in- al. 1995, Ochsner and Vasil 1996). Consequently,it is cluding P. aeruginosa. Antibiotics(ampicillin, tetra- not clear to what degree a pyoverdineresponse could cycline, and gentamycin)and metals (zinc and nickel) provide evidence about the specificidentity of a stres- were implicated as stimulatorystressors. In contrast, sor. the effectof mercuryin our studydoes not supportthe Iron potentiatesoxygen toxicity, and virtuallyall or- hypothesisof increased pyoverdineproduction as a ganisms are expected to have evolved mechanismsfor general stressresponse. the purpose of iron management (Kaplan and In our study,cadmium was foundto stimulatepyov- O'Halloran 1996). Our studysuggests that methyl viol- erdine abundance but zinc had no effect.The earlier ogen (paraquat) may act to suppress pyoverdinepro- assertionthat cadmium inhibitedpyoverdine produc- duction. During P. aeruginosa culture, the reduction tion was based on a descriptionof an interactionwith of oxygen in cultureduring stationary phase is asso- cobalt resistance,but no data were presentedto doc- ciated witha burstof pyoverdineproduction under low umentthis assertion(Mergeay et al. 1978). Our work, ironconditions (Cox 1986) and an increasein potential documentinga dose-dependentstimulation of pyov- in the cell (Hassan and Moody 1987). Methylviologen erdine abundance, supportsthe hypothesisthat cad- is known to induce catalase and superoxidedismutase mium has a stimulatingeffect on pyoverdineproduc- activities, presumably as oxygen stress protection tion. Workon E. coli suggeststhat our resultsare cor- mechanisms(Hassett et al. 1993, Brown et al. 1995). rect. Specifically,in E. coli it has been shown that Fur mutationscan have alteredcatalase and superoxide cadmium interactswith Fur proteinthiols with high dismutase activities,and are defectivein pyoverdine avidity,but the liganded repressorhas reduced affinity synthesis(Hassett et al. 1996). The sensitivityof these forregulatory DNA sequences as determineby in vitro mutantsto hydrogenperoxide and paraquatwas greater footprintingassays (Coy and Neilands 1991). than wild-type,and the suggestionwas made thatre- We did not observe an effectof zinc on pyoverdine ductionin pyoverdinesynthesis could be a contributing productionas mightbe expectedby theresults of Hofte factor (Hassett et al. 1996). Perhaps in contrast,our et al. (1993, 1994). The activationby zinc (Hofte et al. results indicate that reduced pyoverdine production 1993) is associated with a site-specificrecombinase may be a consequence of exposure to paraquat, pre- May 1999 STRESS IN ECOLOGICAL SYSTEMS 447 sumablyto reduce the amountof iron assimilationand of microbial iron chelates. CRC Press, Ann Arbor,Mich- therebyreduce the potentiation of its toxic effects.Giv- igan, USA. Arrage, A. A., T. J. Phelps, R. E. Benoit, and D. C. White. en the complexityof iron and oxygen regulationof 1993. Survival of subsurfacemicroorganisms exposed to gene expression in pseudomonads (Tardat and Touati UV radiation and hydrogenperoxide. Applied Environ- 1993), it is plausible that under some circumstances mentalMicrobiology 59:3545-3550. pyoverdineexpression could be decoupled from co- Bagg, A., and J. B. Neilands. 1987. Ferricuptake regulation ordinate with proteinacts as a repressor,employing iron(II) as a cofactor upregulation catalase and superoxidedis- to bind the operatorof an iron transportoperon in Esch- mutase activities. erichia coli. Biochemistry26:5471-5477. The ecological consequences of maladaptive regu- Barton,L. L., and B. C. Hemming. 1993. Iron in lation of siderophoreproduction is significantbecause plants and soil microorganisms.Academic Press, New ironmanagement is a universalproblem for organisms. York, New York, USA. Bayne, B. L., D. A. Brown, K. Burns, D. R. Dixon, A. Iva- For example, constitutivesiderophore expression in E. novici, D. R. Livingstone,D. M. Lowe, M. N. Moore, A. coli resultsin increased sensitivityto near UV irradi- R. D. Stebbing,and J. Widdows. 1979. An experimental ation (Hoerteret al. 1996). In general,DNA-damaging approach to the determinantsof biological water quality. oxygenradicals challenge the viabilityof microorgan- Philosophical Transactionsof the Royal Society of London isms B 286:563-581. (Linn and Imlay 1987, Arrage et al. 1993). Briskot,G., K. Taraz, and H. Budikiewicz. 1989. Pyoverdin- Our research suggests new opportunitiesfor pre- type siderophoresfrom Pseudomonas aeruginosa. Liebigs dictingecotoxicological outcomes based on the inter- Annalen Der Chemie 4:375-384. action of xenobioticswith molecular mechanisms that Brown, S. M., M. L. Howell, M. L. Vasil, A. J. Anderson, regulate key microbial processes. Specifically,based and D. J. Hassett. 1995. Cloning and characterizationof the katb gene of Pseudomonas aeruginosa encoding,a hy- on the results of this studywe are able to propose a drogen peroxide-induciblecatalase: purificationof KatB, hypothesisthat relates pyoverdineproduction, cellular cellular localization, and demonstrationthat it is essential ironaccumulation, and oxygentoxicity. The hypothesis for optimal resistance to hydrogenperoxide. Journalof could have general implicationsgiven the conserved Bacteriology 177:6536-6544. mechanisms Cox, C. D. 1986. Relationship between oxygen and sider- underlyingcontrol of siderophoresynthe- ophore synthesisin Pseudomonas aeruginosa. CurrentMi- sis. Importantly,we can make the detailed prediction crobiology 14:19-23. that the effectof cadmium in stimulatingpyoverdine Cox, C. D., and P. Adams. 1985. Siderophore activityof productionwill predominateover theeffect of paraquat pyoverdinefor Pseudomona aeruginosa. Journalof Bac- in reducingpyoverdine production when P. aeruginosa teriology137:357-364. Coy, M., and J. B. Neilands. 1991. Structuraldynamics and is exposed simultaneouslyto both xenobiotics.In this functionaldomains of the Fur protein.Biochemistry 30: scenario, the stimulatoryeffect of cadmium would be 8201-8210. predominantbecause a diffusablefactor, cadmium, is Cunliffe,H. E., T. R. Merriman,and I. L. Lamont. 1995. expected to reduce the functionalavailability of the Cloning and characterizationof pvdS, a gene requiredfor Fur pyoverdinesynthesis in Pseudomonas aeruginosa: PvdS is protein,which serves to repress gene expression probably an alternativesigma factor.Journal of Bacteri- forpyoverdine synthesis. Consequently when both cad- ology 177:2744-2750. mium and methylviologen are present,the cell is hy- Guerinot,M. L. 1994. Microbial iron transport.Annual Re- pothesized to take up an excess of iron by the vehicle view of Microbiology 48:743-772. of pyoverdineoverproduction under conditions of ox- Hamer, K. 1997. Productionof pyoverdinby Pseudomonas aeruginosa afterexposure to stressconditions. Thesis. Uni- idative stressdue to the presence of methylviologen. versityof Nebraska-Lincoln,Lincoln, Nebraska, USA. In this manner,the cell would amplifyoxidative stress Hassan, H. M., and I. Fridovich. 1979. Paraquat and Esch- because iron potentiatesoxygen toxicity.In general, erichia coli: mechanismof productionof extracellularsu- we are arguingthat a futureavenue of ecotoxicological peroxide radical. Journal of Biological Chemistry253: 10846-10852. researchshould be to conductmechanism-driven stud- Hassan, H. M. and C. S. Moody. 1987. Regulation of man- ies of synergisticeffects of xenobiotics,or otherstres- ganese-containing superoxide dismutase in Escherichia sors, based on fundamentallyimportant features of mi- coli: anaerobic inductionby nitrate.Journal of Biological crobial physiological processes. Chemistry262:17173-17177. Hassett, D. J., P. A. Sokol, M. L. Howell, J. E Ma, H. T. ACKNOWLEDGMENTS Schweizer, U. Ochsner,and M. L. Vasil. 1996. Ferricup- take regulators(Fur) mutantsof Pseudomonas aeruginosa This research was supportedby the metallobiochemistry demonstratedefective siderophore-mediatediron uptake, programat the Universityof Nebraska-Lincoln and by NSF- altered aerobic growth,decreased superoxide dismutase EPSCOR project 92-555225. We thank ProfessorNielands and catalase activities.Journal of Bacteriology178:3996- for sharinghis experience and insight,Ming-Hoi Wu forca- 4003. pable assistance, and Tony Joernfor editorialcontributions. Hassett, D. J., W. A. Woodruff,D. J. Wozniak, M. L. Vasil, In addition,we would like to thankRonald Gormanand Kathie M. S. Cohen, and D. E. Ohman. 1993. Cloning and char- Helzer for diligenthelp in data. collecting acterizationof thePseudomonas aeruginosa sodA and sodB genes encoding manganese- and iron-cofactoredsuperox- LITERATURE CITED ide dismutase: demonstrationof increased manganese su- Abdallah, M. A. 1991. Pyoverdines and pseudobactins. peroxide dismutaseactivity in alginate-producingbacteria. Pages 139-153 in G. Winkelman,editor. CRC handbook Journalof Bacteriology 175:7658-7665. INVITED FEATURE Ecological Applications 448 Vol. 9, No. 2

Hoerter,J., A. Pierce, C. Troupe, J. Epperson, and A. Eisen- crobial iron transportcompounds. Journal of Biological stark. 1996. The role of enterobactinand intracellulariron Chemistry270:26723-26726. in cell lethalityduring near-UV irradiationin E. coli. Pho- Neilands, J. B., and S. A. Leong. 1986. Siderophores in tochemistryand Photobiology64:537-541. relationshipto plant growthand disease. Annual Review Hofte, M., S. Buysens, N. Koedam, and P. Cornelis. 1993. of Plant Physiology37:187-208. Zinc affectssiderophore-mediated high affinity iron uptake Ochsner,U. A., and M. L. Vasil. 1996. Gene repressionby systems in the rhizoshere Pseudomonas aeruginosa the ferricuptake regulator in Pseudomonas aeruginosa: cy- 7NSK2. Biological Metals 6:85-91. cle selection by iron-regulatedgenes. Proceedings of the Hofte,M., Q. Dong, S. Kourambas,V. Krishnapillai,D. Sher- National Academy of Science (USA) 93:4409-4414. rat, and M. Mergeay. 1994. The sss gene product,which Overbeck,J. 1990. Aspects of aquatic microbialcarbon me- affectspyoverdine production in Pseudomonas aeruginosa tabolism: regulationof phosphoenolpyruvatecarboxylase. 7NSK2, is a site-specificrecombinase. Molecular Micro- Pages 70-95 in J. Overbeck and R. Chrost,editors. Aquatic biology 14:1011-1020. microbial ecology. Springer-Verlag,Berlin, Germany. Hofte, M., L. Mergeay, L. Diels, and W. Verstraete. 1989. Payne, S. M. 1993. Iron acquisition in microbialpathogen- Influence of stress factors on siderophoreproduction by esis. Trends in Microbiology 1:66-69. fluorescentPseudomronads. Societe Belge de Biochimie, Poole, K., S. Neshat, K. Krebes, and D. E. Heinrichs. 1993. Gent. 4:B96. Cloning and nucleotide sequence analysis of the ferri- pyoverdinereceptor genefpvA of Pseudomonasaeruginosa. Kaplan, J., and T. V. O'Halloran. 1996. Iron metabolismin Journalof Bacteriology 175:4597-4604. Mars and Venus at it again. Science 271:1510- eukaryotes: Prince,R. W., C. D. Cox, and M. L. Vasil. 1993. Coordinate 1512. regulationof siderophoreand exotoxin A production:mo- Kerr,R. A. 1994. Iron fertilization:a tonic, but no cure for lecular cloning and sequencing of the Pseudomonas aeru- the greenhouse.Science 263:1089-1090. ginosa fur gene. Journalof Bacteriology 175:2589-2598. King, J. V., J. J. R. Campbell, and B. A. Eagles. 1948. The Prince, R. W., D. G. Storey, A. I. Vasil, and M. L. Vasil. mineralrequirements for fluorescein production. Canadian 1991. Regulationof toxAand regAby the Escherichia coli Journalof Research 26C:514-519. fur gene and identificationof a Fur homologue in Pseu- Leong, J.,W. Bitter,M. Koster,V. Venturi,and P. J.Weisbeek. domonas aeruginosa PA103 and PAO1. Molecular Micro- 1991. Molecular analysis of irontransport in plantgrowth- biology 5:2823-2831. promotingPseudomonas putida WCS358. Biological Met- Raymond,K. N. 1994. Recognitionand transportof natural als 4:36-40. and syntheticsiderophores by microbes.Pure and Applied Linn, S., and J. A. Imlay. 1987. Toxicity,mutagenesis and Chemistry66:773-78 1. stressresponses induced in E. coli by hydrogenperoxide. Rombel, I. T., B. J. McMorran, and I. L. Lamont. 1995. Journalof Cell Science, Supplement6:289-301. Identificationof a DNA sequence motifrequired for ex- Mergeay, M., J. Gerits,and C. Houra. 1978. Facteur trans- pressionof iron-regulatedgenes in pseudomonads.Molec- missible de la resistance au cobalt chez un Pseudomonas ular and General Genetics 246:519-528. du type hydrogenom.onas.C. R. Soc. Biol. 172:575-579. Stintzi,A., and J.-M. Meyer. 1994. Search forsiderophores Meyer, J. M., and M. A. Abdallah. 1978. The fluorescent in micro-organisms.Microbes for Better Living 35:169- pigmentof Pseudomnonasfluorescens: biosynthesis,puri- 176. fication,and physiochemicalproperties. Journal of General Tardat,B., and D. Touati. 1993. Iron and oxygenregulation Microbiology 107:319-328. of Escherichia coli MnSOD expression: competitionbe- tween the global regulatorsFur and ArcA for binding to Meyer,J. M., D. Hohnadel, A. Khan, and P. Cornelius. 1990. DNA. Molecular Microbiology 9:53-63. Pyoverdine-facilitatediron uptake in Pseudomonas aeru- Totter,J. R., and F T. Moseley. 1953. Influencesof the con- immunological characterizationof the ferripyov- ginosa: centrationof ironon the productionof fluoresceinby Pseu- erdine receptor.Molecular Microbiology4:1401-1405. domonas aeruginosa. Journalof Bacteriology65:45-47. Miyazaki, H., H. Kato, T. Nakazawa, and M. Tsuda. 1995. Visca, P., G. Colotti, L. Serino, D. Verzili, N. Orsi, and E. A positive regulatorygene, pvdS, for expression of pyov- Chiancone. 1992. Metal regulationof siderophoresynthe- erdinebiosynthetic genes in Pseudomonasaeruginosa. Mo- sis in Pseudomonas aeruginosa and functionaleffects of lecular and General Genetics 248:17-24. siderophore-metalcomplexes. Applied and Environmental Moriarty,F 1988. Ecotoxicology, the studyof pollutantsin Microbiology 58:2886-2893. ecosystems. Academic Press, London, UK. Wendenbaum,S., P. Demange, A. Dell, J. M. Meyer,and M. Neilands, J. B. 1981. Microbial iron compounds. Annual A. Abdallah. 1983. The structureof pyoverdinePa, the Review of Biochemistry50:715-73. siderophoreof Pseudomonas aeruginosa. TetrahedronLet- . 1995. Siderophores: structureand functionof mi- ters 24:4877-4880.