Dinitrotoluene-Degrading Bacteria

12 Pages·2017·1.46 MB·English

Downloaded from orbit.dtu.dk on: Jan 13, 2023

Endogenous Stress Caused by Faulty Oxidation Reactions Fosters Evolution of 2,4-
Dinitrotoluene-Degrading Bacteria
Perez-Pantoja, Danilo; Nikel, Pablo Ivan; Chavarría, Max; de Lorenzo, Victor
Published in:
P L o S Genetics
Link to article, DOI:
10.1371/journal.pgen.1003764
Publication date:
2013
Document Version
Publisher's PDF, also known as Version of record
Link back to DTU Orbit
Citation (APA):
Perez-Pantoja, D., Nikel, P. I., Chavarría, M., & de Lorenzo, V. (2013). Endogenous Stress Caused by Faulty
Oxidation Reactions Fosters Evolution of 2,4-Dinitrotoluene-Degrading Bacteria. P L o S Genetics, 9(8),
[e1003764]. https://doi.org/10.1371/journal.pgen.1003764

General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright
owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

 Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
 You may not further distribute the material or use it for any profit-making activity or commercial gain
 You may freely distribute the URL identifying the publication in the public portal

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Endogenous Stress Caused by Faulty Oxidation
Reactions Fosters Evolution of 2,4-Dinitrotoluene-
Degrading Bacteria
Danilo Pe´rez-Pantoja., Pablo I. Nikel., Max Chavarr´ıa¤, Vı´ctor de Lorenzo*
SystemsandSyntheticBiologyProgram,CentroNacionaldeBiotecnolog´ıa,CSIC,CampusdeCantoblanco,Madrid,Spain
Abstract
Environmental strain Burkholderia sp. DNT mineralizes the xenobiotic compound 2,4-dinitrotoluene (DNT) owing to the
catabolicdntgenesbornebyplasmidDNT,buttheprocessfailstopromotesignificantgrowth.Toinvestigatethislackof
physiological return of such an otherwise complete metabolic route, cells were exposed to DNT under various growth
conditions and the endogenous formation of reactive oxygen species (ROS) monitored in single bacteria. These tests
revealedthebuildupofastrongoxidativestressinthepopulationexposedtoDNT.ByeithercuringtheDNTplasmidorby
overproducingthesecondactivityofthebiodegradationroute(DntB)wecouldtracealargeshareofROSproductiontothe
first reaction of the route, which is executed by the multicomponent dioxygenase encoded by the dntA gene cluster.
Naphthalene, the ancestral substrate of the dioxygenase from which DntA has evolved, also caused significant ROS
formation. That both the old and the new substrate brought about a considerable cellular stress was indicative of a still-
evolvingDntAenzymewhichisneitheroptimalanylongerfornaphthalenenorentirelyadvantageousyetforgrowthofthe
hoststrainonDNT.WecouldassociateendogenousproductionofROSwithlikelyerror-pronerepairmechanismsofDNA
damage, and the ensuing stress-induced mutagenesis in cells exposed to DNT. It is thus plausible that the evolutionary
roadmapforbiodegradationofxenobioticcompoundslikeDNTwaslargelyelicitedbymutagenicoxidativestresscaused
byfaulty reactions ofprecursor enzymeswith novelbutstructurally related substrates-to-be.
Citation: Pe´rez-Pantoja D, Nikel PI, Chavarr´ıa M, de Lorenzo V (2013) Endogenous Stress Caused by Faulty Oxidation Reactions Fosters Evolution of 2,4-
Dinitrotoluene-DegradingBacteria.PLoSGenet9(8):e1003764.doi:10.1371/journal.pgen.1003764
Editor:IvanMatic,Universite´ParisDescartes,France
ReceivedMarch28,2013;AcceptedJuly16,2013;PublishedAugust29,2013
Copyright:(cid:2)2013Perez-Pantojaetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermits
unrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited.
Funding:ThisstudywassupportedbytheBIOandFEDERCONSOLIDER-INGENIOprogrammeoftheSpanishMinistryofScienceandInnovation,theMICROME,
ST-FLOWandARISYSContractsoftheEU,theERANETProgramandthePROMTProjectoftheCAM.DPPistheholderofaMarieCurieActionsProgramgrantof
theECforvisitingScholars(PIIF-GA-2009-253825).PINisaresearcherfromConsejoNacionaldeInvestigacionesCient´ıficasyTe´cnicas(Argentina)andholdsa
MarieCurieActionsProgramgrantoftheECforvisitingScholars(ALLEGRO,UE-FP7-PEOPLE-2011-IIF-300508).Thefundershadnoroleinstudydesign,data
collectionandanalysis,decisiontopublish,orpreparationofthemanuscript.
CompetingInterests:Theauthorshavedeclaredthatnocompetinginterestsexist.
*E-mail:[email protected]
¤Currentaddress:EscueladeQu´ımica,UniversidaddeCostaRica,SanJose´,CostaRica.
.Theseauthorscontributedequallytothiswork.
Introduction as a surprise that many xenobiotic compounds containing strong
chemicalbondsand/orsubstituentscanbedegraded byenviron-
A diversity of xenobiotic compounds has made it to the mental bacteria after being in the biosphere for only a relatively
environment in large amounts since the onset of synthetic shortperiodoftime.Oneconspicuouscaseinthisrespectisthatof
chemistry due to urban and industrial activities [1]. Although nitroaromatic compounds. Although many chemical structures
manyofsuchcompoundsbearbondsthatarerareornon-existing with C-NO bonds are found in natural products [6,7],
2
inthenaturalrealm,itisnotinfrequenttoisolatebacteriaableto nitroaromaticscouldneverbecomeasignificantselectivepressure
usethemasC,Nand/orenergysources,especiallyinplaceswith before their massive production and release in connection to the
a history of chemical pollution [1–3]. This bear witness of the modern industry of explosives [1,8]. Yet, a number of bacterial
abilityofexistingmetabolicnetworkstoconquestnewanddiverse strains have been isolated that can mineralize many of such
chemical landscapes. The process by which a given bacterium unusual chemical structures [1,2,9]. However, the mere genetic
comesacrossageneticandenzymaticsolutiontothechallengeof drift of pre-existing enzymes and operons that catabolize similar
degrading a new chemical is not trivial [4,5]. Not only enzymes compounds towards new substrates can hardly explain the high
andcompletepathwayshavetoevolvenewsubstratespecificities, rateatwhichnewstrainsandcatabolicoperonsappear[10].The
butalsotheyshouldbeexpressedattherighttimeandlocation.In question thus arises on whether evolution of given catabolic
addition, theresulting metaboliccurrency should be wiredtothe properties benefit from some inherent accelerator of the process
central biochemical network for eventual biomass buildup. This somehow encodedintheenzymes andgenes involved.
surely requires a mutual adaptation between the host and the Environmental bacteria capable of biodegradation of 2,4-
pathwayitselfthatinvolvesmultiplechangesinboththecatabolic dinitrotoluene (DNT) afford an exceptional opportunity to
genesandtherestofthecelltranscriptome.Onthisbasis,itcomes examine the factors at play in the emergence of new metabolic
PLOSGenetics | www.plosgenetics.org 1 August2013 | Volume 9 | Issue 8 | e1003764
EvolutionofDNTDegraders
Author Summary
Manybacteriahaveacquiredthecapacityofmetabolizing
chemical compounds that have never been in the
Biosphere before the onset of contemporary synthetic
chemistry. However, the factors that shape the new
metabolic properties of such microorganisms remain
obscure.Weexaminedtheperformanceofastill-evolving
metabolic pathway for biodegradation of 2,4-dinitroto-
luene(DNT,anarchetypalxenobioticcompound)borneby
a Burkholderia strain isolated from soil in an ammunition
plant. The biodegradation pathway likely arose from a
precursor set of genes for catabolism of naphthalene
Figure1.DNTmineralizationpathwayinBurkholderiasp.DNT
(although Burkholderia does not degrade this compound
and organization of thedntgene cluster. (A) The DNT catabolic
any longer), and is now advancing towards the new route starts with the DntA DNT dioxygenase that hydroxylates the
substrate, DNT. We found that the action of the first aromatic ring in positions 4 and 5 to yield 4M5NC, releasing, at the
enzyme of the biodegradation pathway, a Rieske-type same time, the first nitro substituent. The substituted catechol is
dioxygenase, on the still-suboptimal substrate (DNT) subsequently mono-oxygenated by the DntB hydroxylase, that
generates a high level of endogenous reactive oxygen eliminates the remaining nitro group in the structure, thereby
species. This, in turn, damages DNA and increases producing 2H5MQ. The rest of the pathway (executed by DntCDGE)
includes a ring cleavage reaction and channeling of the products
mutagenesis,ultimatelyresultinginthecreationofnovelty
towardsthecentralmetabolism,inwhichtheyarefinallymetabolized.
thatmayfosterevolutionofxenobiotic-degradingvariants
(B)Organizationofthedntgenecluster,includingdntR,theregulatory
ofthestrainhostingthebiodegradationpathway.Thevery
gene, and dntAabcd, encoding the multi-component DntA DNT
metabolicproblemthussomehowseemstostimulatethe dioxygenase. Amino acid identity between the orthologous dnt and
exploration of the solution space. Our data is fully nag genes is included to illustrate the relatedness of the DntA DNT
consistent with the notion that stress caused by faulty dioxygenase from Burkholderia sp. DNT and the naphthalene dioxy-
dioxygenation of DNT accelerates the rate of bacterial genasefromRalstoniasp.U2.
evolution. doi:10.1371/journal.pgen.1003764.g001
Results
abilities. One of such isolates, Burkholderia sp. strain DNT, was
Activity of the DNT pathway on its former and current
isolatedfromcontaminatedsurfacewaterofanammunitionwaste
substrate is detrimental for Burkholderia sp. DNT
plantonthebasisofutilizingDNTasCandNsources[11].This
Theexperimentalsystemofchoiceforexaminingthetransition
capacity is due to the action of a route for catabolism of the
oftheDNTpathwayfromdegradingnaphthalenetometabolizing
nitroaromaticcompound(Fig.1A).Thephylogenyofthedntgenes
DNTinvolvesonlythreeplayers:strainBurkholderiasp.DNTand
and the biochemical properties of the encoded products indicate
that this cluster for DNT biodegradation has originated from a the two chemicals at stake. It should be noted that, according to
precursor pathway for catabolism of the naturally-occurring theveryhighsimilaritybetweentheleadingringoxygenasesofthe
hydrocarbon naphthalene (Fig. 1B; [12,13]). Still, the same data nagrouteinRalstoniasp.U2fornaphthalenedegradationandthe
indicates that the pathway is even now evolving since the kinetic dntpathwayofBurkholderiasp.DNTforDNTcatabolism(Fig.1B),
couplingofeachofthebiochemicalstepsisnotwellbalanced[11], the extant DntA enzyme retains the ability to act on the first
the regulation of the system keeps the same effector profile than substrate to produce 1,2-dihydroxy-1,2-dihydronaphthalene
the precursor operon [14], there are transposon remnants and [13,17].Owingtoitscurrentcatabolicrole,thismulti-component
vestigial genes in the genetic cluster [12] and the strain hardly enzymedioxygenatesDNTinpositions4and5toyield4-methyl-
growsonthesubstrateofinterest[15].Theseareallindicatorsthat 5-nitrocatechol (4M5NC, Fig. 1A). In order to have a gross
thisstrainhasstarteditsitinerarytowardsDNTcatabolismbuthas estimateofthephysiologicalstateofthecellswhenexposedtothe
notyetfoundanoptimalsolutiontothemulti-tieredproblemofits former and current DntA substrates, we simply examined the
viable degradation. survival of Burkholderia sp. DNT pre-grown on M9 minimal
In this work we have examined possible physiological and medium containing succinate and then exposed respectively to
biochemical drivers that frame the evolution of Burkholderia sp. either chemical at a final concentration of 0.5mM. In order to
DNT towards DNT catabolism and, by extension, many other differentiatebetweentheintrinsiceffectofthesecompoundsfrom
environmental bacteria towards new chemical compounds. As theirinterplaywiththednt-encodedenzymes,wealsoemployeda
shownbelow,reactiveoxygenspecies(ROS)broughtaboutbythe variant strain cured of the dnt genes (Burkholderia sp. DNT Ddnt),
faulty performance of the first enzyme of the pathway (DntA that has no chance of transforming the substrates at stake. While
dioxygenase)onthexenobioticsubstrateseemtobeakeyagentof both naphthalene and DNT led to a very significant loss of
the process. This is because ROS is translated into DNA viability of Burkholderia sp. DNT (Fig. 2A), which went down to
mutagenesis and diversification of the host strain that foster the ,30%inthepresenceofDNT,additionofthesamechemicalsto
emergence of novel adaptive phenotypes. Moreover, the results the strain lacking the dnt genes had a considerably lesser toxicity
provide an evolutionary logic to the abundance of enzymes (Fig. 2B). It should be noted that neither naphthalene nor DNT
containing Rieske-type Fe centers (as DntA) in pathways for areinducersofthedntpathway[14],whichisexpressedthroughits
biodegradation of xenobiotic compounds recently released to the native promoter at a basal level under these experimental
environment [16]. This work showcases an evolutionary scenario conditions. The detrimental effect of the former (naphthalene)
in which the physiological stress caused by a metabolic problem andnew(DNT)substratecanthusbebothtracedtothepresence
triggersalsothegeneticdiversificationnecessaryforexploringthe of the dnt genes and plausibly to the activity of their encoded
solution spacetothesame problem. productsoneachofthechemicals.Inordertoqualifytheobserved
PLOSGenetics | www.plosgenetics.org 2 August2013 | Volume 9 | Issue 8 | e1003764
EvolutionofDNTDegraders
Figure2.EffectofDNTandnaphthaleneadditiononBurkholderiasp.DNTviability.CulturesweremadeinM9minimalmediumcontaining
0.3%(w/v)sodiumsuccinateastheCsource.WhencellsattainedanOD ofca.0.5,theculturewassplit.Onesuspensionservedasthecontrol
600
(denotedasSuccintheplots),anditwasaddedwithDMSO(carrierforDNTandnaphthalene).Theremainingcellsuspensionwasaddedwitheither
DNTornaphthalene(denotedasNaphintheplots)atafinalconcentrationof0.5mM(correspondingtodosagesof1.48and0.23mmol/mlforDNT
andnaphthalene,respectively).Survivalwasassessedin(A)thewild-typestrainor(B)theDdntstrain,andthenumberofcolonyformingunitsineach
casewasevaluatedatdifferenttimepointsbyplatingappropriatealiquotsofthecellssuspensionontoLBplates.ErrorbarsrepresentSD(n=4).
doi:10.1371/journal.pgen.1003764.g002
toxicity of the DntA substrates we measured the activity of the suggested that activity of the dnt-encoded products on the two
enzyme glucose-6-phosphate dehydrogenase (G6PDH) in cells substrates tested caused physiological stress that was more
exposed or not to DNT. The activity of this enzyme is a pronounced in thecaseof DNT.
physiological stress marker in most bacteria [18,19] and its
increased activity reports a general response to metabolic Exposure of Burkholderia sp. DNT to DNT results in
hardship. When Burkholderia sp. DNT cells were exposed to massive production of ROS in individual cells
DNT, the level of G6PDH activity more than doubled in the WhatcouldbethereasonfortheremarkabletoxicityofDNTin
presenceofDNT(Fig.3).Incontrast,theequivalentstrainlacking cells expressing dnt genes? Since an increased activity of G6PDH
thedntgeneshadaG6PDHactivityminimallyaffectedbyaddition would lead to enhanced NADPH turnover rates required to
of the nitroaromatic compound. Taken together, these data counteract oxidative stress [18–20], one plausible scenario was
Figure3.Quantificationofglucose-6-phosphatedehydrogenase(G6PDH)activityinBurkholderiasp.DNTandderivativesexposed
toDNT.Culturesofthewild-typestrainandtheDdntstrainwereexposedtoDNTatafinalconcentrationof0.5mMfor3handtheG6PDHactivity
wasassessedincell-freeextractsbyfollowingtherateofNADP+reductionaspreviouslydescribedbyNikeletal.[19].Ineachcase,theasteriskmarks
denotesignificantdifferencesintheG6PDHactivitylevelateitherP,0.05(*)orP,0.01(**)whencomparedtocontrolconditions(Succ,inwhich
cells were added with DMSO, the DNT solvent carrier), as evaluated with ANOVA. The statistical comparison between the G6PDH activity levels
betweenthetwostrainsatstakeindicatedasignificantdifferencewhencellswereexposedtoDNT.ErrorbarsrepresentSD(n=3).
doi:10.1371/journal.pgen.1003764.g003
PLOSGenetics | www.plosgenetics.org 3 August2013 | Volume 9 | Issue 8 | e1003764
EvolutionofDNTDegraders
thattheencounterofDNTwiththednt-encodedoxygenasescould DntA originates in a precursor naphthalene dioxygenase [12,13],
releaseROSasasideproductoftheoxidativestepsthatshapethe wehypothesized that some of theROS could be originatedfrom
pathway (Fig. 1A). This is a frequent situation in evolving an uncoupled, non-productive oxygen-delivery reaction between
biodegradative routes of aerobic bacteria [21–24] and it often the evolved enzyme and the suboptimal substrate DNT, as
becomes a veritable bottleneck in the catabolism of some reported for similar Rieske-type dioxygenases [29,30]. This
compounds [19,25,26]. On this background, we set out to possibility placed the focus of the ensuing experiments on the
measure ROS generation in Burkholderia sp. DNT cells treated first step oftheDNT biodegradation route.
with DNT along with their Ddnt counterparts. To quantify
oxidative stress we resorted to staining cells with the ROS- The bulk of ROS elicited by DNT in Burkholderia sp. DNT
activated green fluorescent dye 29,79-dichlorodihydrofluorescein originates in the initial dioxygenation step
diacetate (H DCF-DA) followed by quantitative flow cytometry
2 Since the most critical move in any biodegradative route of
[19]. The minimum and maximum response levels of reference
aromaticsistheinitialstepthat activatestheringforovercoming
were fixed using succinate-grown cells added with either the
the resonance energy that stabilizes their structure [31], we
H DCF-DA solvent carrier (dimethyl sulfoxide, DMSO) or the
2 concentrated on the leading reaction that converts DNT to
samewith1.5 mMH O .Fig.4accreditstheperformanceofthe
2 2 4M5NC (Fig. 1A). This compound, as is the case for other
testandshowsrawdataobtainedwhencellswereexposedtothe
catechols, is highly toxic [32,33]. Furthermore, it is known [11]
toxicantsatstake.PerusaloftheresultsindicatedthatDNTcaused
that such yellow-colored intermediate transiently accumulates in
an extraordinary and somewhat unexpected level of intracellular
the medium prior to be channeled towards the next step of the
ROS, the intensity of which nearly equaled that produced by
degradation route (Fig. 6A). The toxic effect of DNT on
straight addition of H O to the medium. In comparison, cells
2 2 Burkholderiasp.DNTcouldthusoriginateineitherfaultyreactions
lackingthecomplementofdntgenesshowedamuchlowerlevelof
of DntA on its substrate, in accumulation of 4M5NC or in both.
ROS in the presence of DNT, indicating that both the substrate
Our strategy to distinguish these possibilities was to generate a
andthegeneswererequiredforsuchasurprisingsynergytocause
strain that overproduced DntB (Fig. 1A). This enzyme is a
oxidative stress. Addition of the ancestral substrate naphthalene
monooxygenasethateliminatestheremainingnitrosubstituentof
alsocausedalower,butstillsignificant,levelofROSinBurkholderia
4M5NC to produce 2-hydroxy-5-methylquinone (2H5MQ;
sp.DNT(Fig.5AandB).Eventhoughpermeabilityofthedyecan
[34,35]). Higher levels of DntB are thus predicted to drain
be affected by the aromatic compounds tested [27,28], pairwise
4M5NC faster towards 2H5MQ -thereby allowing us toseparate
comparisons of the different strains and substrates revealed the
the intrinsic physiological effect of DntA action from that of the
extension of the stress produced by each compound. Since the
toxic catechol that results from the reaction. To this end, we
generatedstrainBurkholderiasp.DNTdntBq,inwhichextracopies
of the gene encoding the second oxygenase of the pathway were
expressed from a constitutive promoter in a broad-host-range
vector.Fig.6Bverifiessuchaprediction,asculturesofBurkholderia
sp. DNT dntBq released a much lower amount (,20%) of
4M5NC to themedium. In thesame setup, strain Burkholderia sp.
DNT Ddnt did not accumulate any intermediate, while naphtha-
lene would be biotransformed to 1,2-dihydroxy-1,2-dihydro-
naphthalene by any of the dntA+ strains. Once we had the three
isogenicvariantsinhand,wesetouttomeasureintracellularROS
production in wild-type, Ddnt and dntBq cells exposed to either
DNTornaphthalene(Fig.5A,BandC,respectively).Theseresults
provideananswertothequestionsraisedabove.First,thatDNT
causes high ROS levels in both the wild-type and dntBq strains
(Fig.5AandC)clearlyindicatesthatthebulkofoxidativestresscan
be traced to the very reaction of DntA with DNT, not to the
product of the catalysis. Second, the effects of naphthalene were
comparatively lower than those elicited by DNT, but they were
stillsignificantinrespecttotheuntreatedcontrolconditions.ROS
generated from the encounter of this substrate with the dnt-
encodedenzymesmuststemfromuncoupled,faultyreactionswith
theleadingDntAoxygenase,asthereisnootherenzymethatcan
recognize this aromatic compound as a potential but ultimately
non-productive substrate. Since both the ancestral and the new
substrateofthefirstdioxygenasereleaseROSwhenfacingDntA,
the cognate reaction has probably left behind a former biochem-
Figure 4. Flow cytometry-assisted determination of ROS ical optimum (i.e., when acting on naphthalene) but has not yet
content in Burkholderiasp. DNT exposed to the compounds
reached a newonewithDNT.
understudy.Histogramsshowtherawdatafromunstainedcells(Nil),
untreated cells added with DMSO (the DNT solvent carrier, labeled
Succ),andcellsexposedtoDNT(finalconcentration0.5mM)orH O DNT and naphthalene bring forth diversification of
2 2
(finalconcentration1.5mM)for3h.Inallcases,cellsweretreatedwith Burkholderia sp. DNT
the ROS-sensitive probe H DCF-DA, and the resulting dichlorofluor-
2 In a further effort to identify the specific types of ROS that
escein (DCF) fluorescence levels were recorded in at least 25,000
individualcellsaspreviouslydescribed[19].A.U.,arbitraryunits. resultfromtheabovementionedprocesses,wetestedcellstreated
doi:10.1371/journal.pgen.1003764.g004 with the various substrates with nitro blue tetrazolium (NBT), a
PLOSGenetics | www.plosgenetics.org 4 August2013 | Volume 9 | Issue 8 | e1003764
EvolutionofDNTDegraders
Figure5.QuantitativeanalysisofROSproductioninBurkholderiasp.DNTuponexposuretoDNTornaphthalene.Culturesof(A)the
wild-typestrain,(B)theDdntstrainand(C)thedntBqstrainwereexposedtoeitherDNTornaphthalene(Naph)atafinalconcentrationof0.5mM
andcomparedtocontrolconditions(Succ;inwhichcellswereaddedwithDMSO,theDNTandnaphthalenesolventcarrier).Thegeometricmeanof
thedichlorofluorescein(DCF)fluorescencelevelswasanalyzedforeachcondition(eachsymbolshowntotheleftrepresentsatechnicalreplicate,
along with their distribution). Box plots represent the median value and the 1st and 3rd quartiles of the geometric mean values, and different
lowercaselettersidentifysignificantdifferencesatP,0.05testedwiththeMann-WhitneyUtest(n=6).
doi:10.1371/journal.pgen.1003764.g005
reagent that is specific for superoxide production. Consistently adaptive challenge(metabolic or otherwise) than those which had
with the flow cytometry data above, the highest indications of notbeendiversifiedbecauseofthephenomenadescribedhere.
superoxide presence were found in wild-type and dntBq cells
exposed to DNT, while a lower level was detected in the Ddnt Discussion
strain and in bacteria exposed to naphthalene (Fig. 7). Since
superoxide originated in defective redox reactions promotes Bacteria that inhabit environmental niches with a history of
hydroxyl-radical formation and consequent DNA damage pollution by xenobiotic compounds offer a phenomenal experi-
[36,37] we wondered whether ROS stemming from the faulty mental system for examining the expansion of the existing
oxygenations discussed above could eventually translate into an metabolic networks into new chemical spaces [10]. At least three
insult to the genome of Burkholderia sp. DNT. One direct way of bottlenecks must be overcome for the successful emergence of a
quantifyingsuchdamageisthemeasurementofthe8-hydroxy-29- novel catabolic pathway able to deal with a new-to-nature
deoxyguanosine (8-oxoG) content of genomic DNA as a coarse chemicalstructure.First,suitableenzymesmustdeveloptheright
descriptor of HON attack to purines [37,38]. On this background substrate specificity towards the new compound [40]. In most
we resorted to an immunoassay for quantifying 8-oxoG levels in cases,suchenzymesstemfromprecursorsthatactonstructurally
cells exposed to DNT, using H O as a positive control of related chemicalsbutshift specificitiesthroughintermediatesteps
2 2
oxidative stress. As shown in Fig. 8A, DNT indeed caused a 1.4- wheretheoldandthenewactivitiescoexistinthesameproteinat
foldincreaseintheshareofdamagedpurinesintheBurkholderiasp. no fitness cost [40,41]. Second, the genes encoding the new
DNT genome(as opposed toa 2-foldincrease elicited by H O ). enzymatic activities should be expressed only when required i.e.
2 2
As such a chemical damage to DNA bases triggers the SOS when the substrate-to-be is present in sufficient concentrations to
responseandeventuallyincreasesthemutationrateofthebacteria grant a suitable return. More evolved catabolic systems are
that undergo the insult [36], we next wondered whether the typically subject to a tight regulation by inducer substrates [42],
ultimateconsequenceoftheuncouplingofthering-hydroxylating while more recent counterparts are often expressed at low
reaction performed by DntA with their current and ancestral constitutive levels. Not infrequently, still evolving biodegradative
substrateswastoincreasegeneticdiversitybyenhancingmutation operons carry along regulatory systems that respond to former
rates.SincevirtuallynothingisknownabouttheSOSresponsein substratesandnotthenewones[43].Finally,themetabolicaction
Burkholderiasp.DNT,wedirectlymeasuredsuchmutationratesasa oftheevolvedpathwaysmustbeconnectedtosomegrowthbenefit
descriptorofemerginggeneticnovelty.Forthis,weemployedthe thatultimatelyfostersitsselectiveadvantage[43].Thisisbrought
standard test of appearance of rifampicin-resistant (RifR) clones about directly by entering metabolic currency from the biodeg-
under the various conditions assayed. As shown in Fig. 8B, both radation process into the central metabolism, or indirectly, by
naphthalene and DNT triggered a considerable increase in the making cells more resistant to otherwise detrimental endogenous
appearanceofRifRcolonieswhichwasnotnoticeableinthestrain or exogenous stress. Even horizontal gene transfer granted, the
deletedofthedntgenesThemutageniceffectofnaphthaleneinthis numberofmutationsandgeneticeventsthatmightberequiredfor
contextwasslightlymorepronouncedthanwhatcouldbeexpected the emergence of a fully competent biodegradative strain able to
fromthesheerdataonROSproductionshownabove.Wespeculate deal with a novel xenobiotic compound can be considerable, far
that ROS could react chemically with this bicyclic aromatic morethanthosenecessarye.g.fornewantibioticresistances[44].
compoundandgenerateadditionalDNA-intercalatingagents(e.g. It thus comes as a surprise that a good number of xenobiotic
naphtoquinones and naphtodiols [39]) with a separate mutagenic compounds are indeed degraded by environmental strains not
action on DNA. In any case, it is worth noticing that maximum much after their production by the chemical industry [8,10]. In
novelty (as measured with this procedure) is accompanied by an view of all this, it could well be that evolution of biodegradative
acutelethality(Fig.2),sothatthesurvivorstoDNTexposureare routes for xenobiotics is accelerated by factors inherent to the
more capable to explore the possible solution space to the next nature of theenzymatic reactions involvedin theprocess.
PLOSGenetics | www.plosgenetics.org 5 August2013 | Volume 9 | Issue 8 | e1003764
EvolutionofDNTDegraders
Figure6.Accumulationandremovalof4-methyl-5-nitrocathecol(4M5NC)inBurkholderiasp.DNTandderivativesexposedtoDNT.
(A)Culturesofthewild-typestrainwereexposedtoDNTatafinalconcentrationof0.5mMfor3h,andflaskswerephotographedtoevidencethe
productionof4M5NCasthemediumacquiresayellowishcolor.Chemicalanalysesrevealedthattheyellowmetabolite,whichhasanabsorption
peakat420nm,correspondsto4M5NC[14].(B)Time-courseevolutionof4M5NC(thestructureofwhichisshownintheinset)inculturesofthewild-
type strain, the Ddnt strain and the dntBq strain amended with DNT at a final concentration of 0.5mM. Determinations were carried out as
previouslydescribedbydelasHerasetal.[14].ErrorbarsrepresentSD(n=4).
doi:10.1371/journal.pgen.1003764.g006
The metabolic pathway of Burkholderia sp. DNT that acts on metabolicallyproductiveandalsocausesconsiderableROS.These
DNT has all what is biochemically needed for its complete datanotonlypinpointstheDntAcomplextobeinthemidstofa
biodegradation but it is clearly not optimal yet for coupling the transition between the old and the new specificities, but also
processtoefficientgrowth.Furthermore,weshowabovethatthe suggest key events in evolution of new xenobiotic-degrading
DNT dioxygenation reaction that constitutes the first step of the strains. Since ROS mediate DNA damage, which in turn brings
pathwayexacerbatesROSproduction(Fig.4andFig.5).Itisvery about a mutagenic SOS response [36,37,45], it seems that the
likely that such reactive species result from the failure of the Fe- generation of diversity associated to a faulty dioxygenation
containingcenteroftheasubunitofDntAtodeliveractiveoxygen reactioncouldcontributetotheexplorationofthemetabolicnovelty
atoms to positions 4 and 5 of the DNT aromatic ring (Fig. 1A). spaceinafasterfashionthatcouldbeexpectedbyabackground,
Since the DntA enzyme complex originates in a precursor spontaneous mutagenesis. This evolutionary scenario is reminis-
naphthalenedioxygenase(Fig.1B),itislikelythatsuchuncoupling cent - but not exactly identical, to that proposed for the rapid
of the reaction stems from a still poor geometry of the substrate- evolution of antibiotic-resistant bacterial strains, which isthought
enzyme recognition. On the other hand, the enzyme still to be fostered as well by the boost of ROS and the ensuing
recognizes naphthalene as a substrate, but the reaction is not mutagenesis and novelty generation that precedes death of
PLOSGenetics | www.plosgenetics.org 6 August2013 | Volume 9 | Issue 8 | e1003764
EvolutionofDNTDegraders
Figure7.SuperoxideproductioninBurkholderiasp.DNTuponexposuretoDNTandnaphthalene.Culturesofthewild-typestrain,the
DdntstrainandthedntBqstrainwereexposedtoeitherDNTornaphthalene(Naph)atafinalconcentrationof0.5mMfor3h.Superoxidelevels
weredeterminedinvitrobytreatmentofthecellswithNBT,andtheresultswerenormalizedtocontrolconditions(inwhichcellswereaddedwith
DMSO,theDNTandnaphthalenesolventcarrier).DifferentlowercaselettersidentifysignificantdifferenceswithintreatmentsatP,0.05(ANOVA).
ErrorbarsrepresentSD(n=4).
doi:10.1371/journal.pgen.1003764.g007
bacteria treated with antimicrobials [46–49]. In the case of the hasbeenclaimedtobedetrimentalforinsituDNTdegradationby
DNT-degradingstrainpresentedabove,wecouldinfacttracethe environmental bacteria [51]. Additional mutagenic effects of
entire process to the master dioxygenation reaction that initiates biodegradation intermediates downstream of 4M5NC cannot be
the metabolic pathway. Because of its inherent enzymatic ruledout.However,sincetheonlymetabolitethatissignificantly
mechanism (e.g. Rieske-type active center) this reaction is prone accumulated during DNT turnover is precisely 4M5NC (the
to uncouple dioxygenations of suboptimal substrates and release substrateofDntB,[11]),itisunlikelythatthetoxicitygeneratedby
ROS[29,30].InthecaseofstrainDNT,theprocessmightalsobe suchintermediates could beultimately meaningful.
assisted by reactive nitrogen species connected to the nitrite [50] How general could be such a substrate-driven generation of
released as a by-product of the DntA and DntB-catalyzed diversity for evolution of other xenobiotic-biodegradation routes?
reactions (Fig. 1). This issue deserves further studies, as nitrite A good number of results are found in the literature that can be
interpreted under this light. For instance, the abundance of
Rieske-containing enzymes in biodegradative operons [16] could
reflect a key role in generation of novelty both in the catabolic
genes themselves and in the host. This is an important detail, as
changesthatoriginateabetterbiodegradativestraininvolvegenes
otherthatthosedirectlyinvolvedinthecorrespondingpathway.It
isremarkablealsothatthetranscriptomesandproteomesofstrains
degrading a suite of recalcitrant aromatics (e.g. polychlorobiphe-
nyls and polyaromatic hydrocarbons) are overrepresented with
functionsthatcounteractoxidativestress[21–23].Finally,itisnot
unusual to find oxidative pathways for xenobiotics in bacteria
whicharetypicalmacrophage-resistantpathogens(e.g.Burkholderia
andMycobacterium)andthusadaptedtoendureastronglyoxidative
Figure 8. DNA damage and RifR mutants frequency in milieu [52,53]. Such routes might only develop in hosts able to
Burkholderiasp. DNT facing DNT. (A) Relative levels of 8-oxoG in enduretheROS-mediatedlethalityassociatedtotheevolutionary
the wild-type strain treated with DNT(final concentration 0.5mM) or roadmap from the enzymatic recognition of one substrate to
H O (finalconcentration1.5mM)for6handcomparedtothecontrol
2 2 another.
condition (succinate; in which cells were added with DMSO, the DNT
In sum, the data above strongly argue that evolution of DNT
solventcarrier).AsterisksidentifysignificantdifferencesattheP,0.05
(*) and P,0.01 (**) levels (ANOVA). Error bars are SD (n=4). (B) biodegradation pathways and possibly of many other catabolic
FrequencyofspontaneousRifRmutantsinthewild-typestrainandthe systems, ultimately benefit from the mutagenic stress caused by
Ddnt strain exposed to either DNT or naphthalene at a final faultyreactions ofpre-existing enzymesonsuboptimal substrates.
concentration of 0.5mM for 24h. The dimensionless frequencies of
Thetermanti-fragility(asopposedtorobustness,[54,55])hasbeen
mutation were calculated by dividing the average number of RifR
recently coined for describing such systems that reach a higher
coloniesbythetotalnumberofviablecellsinthesameculture.Error
barsrepresentSD(n=3). peak of efficacy after having nearly collapsed with stress and
doi:10.1371/journal.pgen.1003764.g008 shocks(Fig.9).Theresultsdiscussedinthisworkseemnotonlyto
PLOSGenetics | www.plosgenetics.org 7 August2013 | Volume 9 | Issue 8 | e1003764
EvolutionofDNTDegraders
Figure9. Antifragility governsevolutionof new pathwaysforxenobioticcompounds. Thecartoonsketchesthe differences between
fragile(collapseafterashock),robust(recoverytothesamestateofaffairsafterashock)andanti-fragile(improvementafterashock)systems.Inthe
caseexaminedinthisstudy,wearguethattheROSproducedbyfaultydioxygenationreactionsproducesaconsiderablechemicalinsultwhichon
onehandplacesthepopulationatthevergeofcollapse(.70%mortality,seeFig.2)butsimultaneouslydiversifiesgeneticallythesamepopulation
andthusfosterstheexplorationofthesolutionspace.ThetermantifragilitywasoriginallycoinedforEconomics[55]buthasbeenrecentlyappliedto
biologicalsystemsaswell[54].Theconceptisreminiscent(butnotentirelyequivalent)tohormesis[67]inthatthepost-challengestrengthgoes
beyondasuperiortolerancetotheoriginalinsult,andthereforeresultsinageneraladaptivephenotype.
doi:10.1371/journal.pgen.1003764.g009
place the appearance of new xenobiotic-biodegradation strains overexpressingdntB,encodingthesecondenzymeofthepathway
withinsuchanevolutionaryscenario,butalsosuggestexperimen- that removes 4M5NC, was constructed as follows. Oligonucleo-
tal approaches toaccelerate theiremergence intheLaboratory. tides dntB-attB1-F (59-GGG GAC AAG TTT GTA CAA AAA AGC
AGG CTG CTC TGT CGA TGA TTT GAG GA-39) and dntB-
Materials and Methods attB2-R (59-GGG GAC CAC TTT GTA CAA GAA AGC TGG GTG
TTGCGCACCTGTCATCG-39),includingtheattBsecuences
Bacterialstrains,cultureconditionsandgeneralmethods required for Gateway-based cloning (shown in italics in the
Burkholderia sp. DNT has been described before as a DNT corresponding sequences), were used to amplify dntB from
degrading specimen [11]. The sequence, organization and Burkholderia sp. DNT. The amplicon was cloned in the
function of each of the dnt genes, involved in mineralization of pDONRTM/Zeo vector (Life Technologies Corp.) according to
thisaromaticcompound,havebeenreportedelsewhere[14].The themanufacturer’sinstructions,andmovedtotheGatewayvector
action of Burkholderia sp. DNT on the nitroaromatic substrate is pBAV226[58](kindlyprovidedbyJeanT.Greenberg,University
very slow i.e. depletion of 1 mM DNT takes 3–5 days [15]. A of Chicago, USA). This plasmid includes the nptII promoter to
spontaneous derivative of Burkholderia sp. DNT lackingtheDNT- constitutivelydriveexpressionofdntB.Theresultingconstructwas
degradative phenotype was isolated by three consecutive cultiva- transformed into Burkholderia sp. DNT, giving rise to strain DNT
tionroundsonLBmedium,andtheabsenceofkeydntgenes(dntA, dntBq.
dntBanddntD)inthisderivativewasconfirmedbyPCR(datanot All strains were grown in 250-ml Erlenmeyer flasks containing
shown). Accumulation of the yellow intermediate 4M5NC was 50mlofM9minimalmedium[59]addedwith2.5ml/lofatrace
spectrophotometrically quantified by absorbance readings at elementssolution[60]and0.3%(w/v)sodiumsuccinateasthesole
420nm [14]. The high segregational instability of the DNT C source. Overnight-grown Burkholderia cells were used as the
plasmid,carryingthedntgenes,hasbeenpreviouslyreportedwhen inoculum by diluting them to a starting OD of ca. 0.1. Cells
600
cells are grown in rich media [56,57]. Maintenance of the dnt+ were grown in the same culture medium until they reached an
phenotype (and thus of the DNT plasmid) was verified prior to OD ofca.0.5,atwhichpointthecellsuspensionwassplitinto
600
each of the experiments by examining accumulation of 4M5NC two 125-ml Erlenmeyer flasks, one culture served as a control
(thefirstbiodegradationintermediate;Fig.1)in200clonalcultures experimentandwasaddedwithDMSO(thevehicleforbothDNT
added with 1 mM DNT. This control indicated that growth in and naphthalene), and the other one was amended with either
minimalmediumwithsuccinateasthesoleCsourcecaused,1% DNTornaphthalene(fromconcentratedsolutionsinDMSO)toa
loss of the catabolic ability of processing DNT. A strain final concentration of 0.5 mM. These conditions were used
PLOSGenetics | www.plosgenetics.org 8 August2013 | Volume 9 | Issue 8 | e1003764
EvolutionofDNTDegraders
throughout the work. The solubility in H O of DNT at 20uC is among different strains and growth conditions as described
2
1.7 mM, and that of naphthalene at 25uC is 0.2mM; which elsewhere [64].
results in dosages of 1.48 and 0.23mmol/ml for DNT and
naphthalene, respectively, assuming similar solubility values in In vitro determination of enzymatic activity
water and in M9 minimal medium. Flasks were incubated at G6PDHactivitywasanalyzedbyfollowingtherateofNADP+
170rpmand30uC,andsamplescollectedatthetimesindicatedin reduction at 340nm at 30uC in a reaction mixture (1ml final
each case. Abiotic controls were run in parallel to estimate the volume) containing 50mM phosphate buffer (pH=7.5), 10mM
possibledisappearanceofthearomaticcompoundsintheabsence MgSO , 0.75mM NADP+, 2 mM glucose-6-phosphate and 15–
4
of cells. Non-inoculated flasks containing naphthalene (known to 100ml of the cell-free extract, obtained as previously described
be a volatile compound) were incubated during 3 h as explained [19]. An extinction coefficient (e ) of 6.22mM21cm21 was
NADH
above,andtheremainingsubstratecontentwasdeterminedbygas used to calculate the specific enzymatic activity, and the protein
chromatography.Thedeterminationsshowedthatca.88%ofthe concentration in cell-free extracts was measured by the Bradford
initialnaphthaleneremainedattheendoftheexperimentinthese method.Oneunitofenzymaticactivitywasdefinedasthequantity
abiotic controls. of enzyme that catalyzed the formation of 1 mmol product in
1 minat 30uC.
In vivo determination of ROS by flow cytometry
Fluorescence-activated cell sorter cytometry analysis was Determination of mutation frequencies
performed using the ROS-sensitive green fluorescent dye Onehundredmicrolitersofthe1025,1026and1027dilutions
H2DCF-DA (Sigma-Aldrich Co.), essentially as described previ- of treated 24-h cultures were plated by quadruplicate onto LB
ously [19,61,62]. Cells were grown and treated with either DNT agar, and 100ml of undiluted cultures were spread (by quadru-
or naphthalene as explained above. In the cases indicated, plicatealso)ontoLB-rifampicinplates(250 mg/ml).Colonycounts
1.5 mM H2O2 was separately added as a positive control of were performed after 48h of incubation. Mutation frequency
oxidative stress conditions. After being incubated for 3 h, cells valuesarereportedasthedimensionlessratiobetweenthenumber
from 1.5–5.0ml culture were pelleted by centrifugation at of RifR coloniesand thetotal viablecount [65].
5,0006g during 5 min, washed once with phosphate-buffered
saline (PBS, pH=7.5), and resuspended in PBS to adjust the DNA extraction and determination of 8-oxoG
OD600 to ca. 0.1. This cell suspension was added with 40 mM A 25-ml aliquot of the cultures treated with either DNT or
H2DCF-DA(addedfromafreshly-prepared4 mMstocksolution 1.5 mM H2O2 for 6 h was spun at 4uC at 4,5006g for 15min.
in DMSO), and then incubated in the dark for 15min at room Cells were washed twice in cold 50mM phosphate buffer
temperature. Flow cytometry analysis of fluorescence levels was (pH=7.5) and genomic DNA was extracted using the Ultra-
performedinaGalliosTMflowcytometer(BeckmanCoulterInc.) CleanTMmicrobialDNAisolationkit(MoBioLabsInc.).Genomic
equipped with an argon ion laser of 15mW at 488nm as the DNA was reconstituted in H O, quantified by horizontal gel
2
excitation source. The H2DCF-DA fluorescence emission at electrophoresis and using a NanoVue micro-volume UV/Vis
525nm was detected using a 530/30-nm band pass filter array. spectrophotometer (GE Healthcare Bio-Sciences AB), and imme-
Size-related forward scatter signals gathered by the cytometer diatelyhydrolyzedasfollows.Asuitablevolumeofthesuspension,
were used by the CyflogicTM 1.2.1 software (CyFlo Ltd.) to gate containing 1 mg DNA, was boiled at 97uC during 5min, and
fluorescence datafromonlybacteria inthestream, thus avoiding immediately placed in an ice bath. Digestion reactions (1 ml)
the mixing of data from bacteria with data from smaller, non- contained 1 mg denatured genomic DNA, 0.75mM ZnCl ,
2
livingparticlesinthesuspension.Dataforatleast25,000cellsper 10mM CH COOK (pH=5.5), 2.5U/ml nuclease P1 from
experimental condition were collected, and the CyflogicTM 1.2.1 Penicillium citr3inum, and 1.75 U/ml phosphatase acid from Ipomoea
softwarewasusedtocalculatethegeometricmeanoffluorescence batatas (enzymes purchased from Sigma-Aldrich Co.). Digestions
perbacterialcellineachsampleaswelltogeneratehistogramplots werecarriedoutduring18hat37uC,afterwhichthereactionwas
withameasureof-fluorescenceintensityshownonthex-axisand loaded into a 2-ml Centricon YM-10TM device (Millipore Corp.)
the number of bacteria (i.e., events) counted at the specific and clarified by centrifugal ultrafiltration. Hydrolysis of genomic
fluorescenceintensityusedforH2DCF-DAdetectionshownonthe DNA was assessed by horizontal gel electrophoresis and spectro-
y-axis. photometry. Samples were processed within 6h after the
hydrolysisprocedure.Thecontentof8-oxoG(inequilibriumwith
Quantifying intracellular generation of hydrogen the keto form 8-oxo-29-deoxyguanosine) was assessed by compet-
superoxide itiveELISAusingacommercialkit(CaymanChemicalCo.).The
Cell suspensions were treated with the compounds indicated assayisbasedonthecompetitionoffree8-oxoGandan8-oxoG-
before, and 1.5-ml culture aliquots were exposed to 350mg/ml acetylcholinesterase conjugate for a limited amount of 8-oxoG
NBTduring30minat30uCinthedark.HClwasthenaddedat monoclonal antibody. The free acetylcholinesterase activity was
7.5 mMandtheresultingmixturewascentrifugedat2,5006gfor evaluated with the Ellman’s reagent [66]. The detection limit for
15min at room temperature. Pellets were treated twice with 8-oxoGis30pg/ml,andcalibrationcurveswereruninparallelin
0.35 ml DMSO to extract reduced NBT, and supernatants were each setof experiments using an authentic8-oxoGstandard.
pooled and mixed with 1 ml of 50mM phosphate buffer
(pH=7.5).TheNBTcolorchangewasmonitoredbyabsorbance Statistical analysis
at 575nm and superoxide production was normalized to the All reported experiments were independently repeated at least
proteinconcentration,measuredbytheBradfordmethod[63].As twice(asspecifiedinthefigurelegends),andthemeanvalueofthe
the cultures had a similar OD value at the time of harvesting corresponding parameter 6 SD is presented. Statistical signifi-
600
and processing, the amount of total protein used in these assays cancewasassessedusinganalysisofvariance(ANOVA).Forflow
was considered to be roughly the same for each sample. Results cytometryexperiments,themedianvalueisreportedinboxplots
are expressed as the fold-change in the ratio A /mg protein alongwiththe1stand3rdquartiles,andthestatisticalsignificance
575
PLOSGenetics | www.plosgenetics.org 9 August2013 | Volume 9 | Issue 8 | e1003764