A path from predation to mutualism

A path from predation to mutualism
of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Micro Commentary A path from predation to mutualism mmi_7341 1346..1350 Antoine Danchin* AMAbiotics SAS, Genavenir 8, 5 rue Henri Desbruères,91030 Evry Cedex, France. Summary Luminescent bacteria and nematodes associate in astrategy where the bacteria act as virulent pathogensof insects, used as their food supply, while the nema-todes graze on them. Upon reaching high density, thebacteria produce light and metabolites that turn thenematodes into hosts permitting them to be carriedover to further nematode preys. In this issue of Molecular Microbiology  , Lango and Clarke show thatthe corresponding shift in lifestyle is triggered by ametabolic switch closely linked to the tricarboxylicacid cycle, but apparently not by the well-knownacetate switch that monitors entry of bacteria into thestationary phase of growth. It took millenia for humanity to understand, with NicolasCopernic, that Man was not the centre of the world. Yet,because they have been useful in preventing the devel-opment of our pathogens or possibly killing them, we haveperceived many of the secondary metabolites synthe-sized by microbes as  antibiotics  . But should we considerthat evolution is simply the unfolding of inevitable fights,and that ‘weapons’ – this is an often-used term – arecommon tools in communities of living organisms? Whileanalysing the function of antibiotics, Julian Daviesstressed that it was not at all clear that they had evolvedto be ‘weapons of mass destruction’ (Benveniste andDavies, 1973; Davies, 1990). How do organisms thatproduce antibiotics protect themselves against theiraction? Could they not be, rather, signals that permitcommunication within and between species (Davies,2009)? In the past decades, game theory has developedprogressively into more and more intricate patterns, andcompetition – with its corollary, the right of the strongest –are now viewed as highly crude and inefficient ways ofinteracting. Group decisions, cooperation, reward andpunishment, and all kinds of subtle variations, sometimesquite unexpected, are the norm rather than the exception,and bacteria are now seen as displaying complex behav-iours in this context (Massey  et al  ., 2004; Davidson andSurette, 2008). This domain became central for biologywhen it was recognized that almost no organism liveswithout interactions with a variety of other organisms. Theconcept of quorum sensing was meant to provide a uni-fying view to this domain (Fuqua  et al  ., 1994).In this context it was of considerable interest to iden-tify situations where a small number of species exhibit aclear pattern of organized interactions. The remarkablecollaboration/conflict between three species – a nema-tode, an insect and a bacterium ( Photorhabdus lumine- scens  ) – was at the root of the effort undertaken with thelate Frank Kunst to sequence the bacterial genome(Duchaud  et al  ., 2003).  P. luminescens   is a member ofthe Gammaproteobacteria that produces a variety ofsecondary metabolites (Eleftherianos  et al  ., 2010). Thevirulence of  P. luminescens   is so high – comparable withthat of another pathogen of the same family,  Yersinia pestis  , the agent of Back Death – that just a few bacteriaare enough to kill a large caterpillar (Daborn  et al  .,2001). A pathogen of insects, but also both a foodsupply for and a commensal of nematodes,  Photorhab- dus   became a paradigm to study both pathogenicity andmutualism. Yet this organism produces no recognizablequorum-sensing molecules.How, at the molecular level, does  Photorhabdus   shiftbetween virulence and mutualism? Among attempts toexplore the question, the work of David Clarke andco-workers stands out. In recent years, experiments byClarke and others have unravelled the complex synthe-sis of the multipotent ‘antibiotic’ 3-5-dihydroxy-4-isopropylstilbene (ST), an inhibitor of phenoloxidase, theinsect’s primary defence against microbial pathogens(Eleftherianos  et al  ., 2007). ST also functions as a signalto control nematode development and as an antibioticthat might be involved in protecting the insect cadaverfrom attack by microbial saprophytes (Joyce  et al  .,2008). In a concrete illustration of several aspects ofgame theory, Lango and Clarke have shown that biosyn-thesis of ST is at the centre of a regulatory network that Accepted 3 August, 2010. *For correspondence. E-mail antoine.danchin@normalesup.org; Tel. ( + 33) 1 6091 7882; Fax ( + 33) 1 60917899. Molecular Microbiology  (2010)  77 (6), 1346–1350    doi:10.1111/j.1365-2958.2010.07341.xFirst published online 24 August 2010  © 2010 Blackwell Publishing Ltd  controls the fate of the bacteria in a fascinating interplaywith its various hosts. This is probably not the wholestory: stilbenes are common plant metabolites, but  Pho- torhabdus   is the only known producer of stilbenesoutside the plant kingdom, via an entirely different syn-thetic route. Yet another play in the game of life might befound there in the future, as insects are usually depen-dent on plants, with specific cues associating them to alimited number of food supplies, and mimicry is centralin the insect world.Let us set the stage.  Photorhabdus luminescens   inter-acts with nematodes of the genus  Heterorhabditis  ,where the bacteria colonize the gut of the infective juve-nile (IJ) stage of the nematode (Waterfield  et al  ., 2009).The IJ is a specialized non-feeding stage that is activefor dispersal in the soil. Upon finding a suitable insecthost (usually larvae such as caterpillars), the IJs enterthe insect via wounds or natural orifices and migrate tothe haemolymph where the bacterial symbionts areregurgitated. There, the bacteria begin to grow exponen-tially in their host, secreting toxins and a variety of deg-radative enzymes and finally killing the insect within 3days. In parallel, the nematodes feed on the bacteria,which now become prey, and after several rounds ofreproduction a new generation of IJs develops andemerges from the insect cadaver, carrying live  Photo- rhabdus   in their midgut – no longer prey but ratherweapons to kill a future insect larva host.Many features have to be accounted for in the  Photo- rhabdus   life cycle. The bacteria must multiply in the larvaand kill it, while also being prey of the nematodes. Atsome point they must kill the larva, while sparing thenematodes. Then they need to interact with the nema-todes in such a way that they no longer permit the nema-todes to graze on them, but rather the bacteria colonizethe nematode midgut. One postulated role for ST was totrigger the change in nematode behaviour towards thebacteria. To further explore this hypothesis Lango andClarke monitored production of ST, an anthraquinonepigment, and bioluminescence. These phenotypesappear during the transition between exponential and sta-tionary growth phases. Looking for mutants that failed toinitiate these processes, Lango and Clarke identified amutation in the  mdh   gene encoding malate dehydroge-nase, a central enzyme of the TCAcycle. This led them todiscover the existence of a metabolic switch that controlsthe transition between the insect pathogenic and nema-tode symbiont stages of  Photorhabdus  . This work hasconsiderable importance in the context of a variety ofstudies that explore phase transitions, alternative lifecycles and the transition between exponential growth tostationary phase (e.g. Frimmersdorf  et al  ., 2010). Further-more, because no standard quorum sensing system hasbeen found in  Photorhabdus  , this likely brings us closer tothe ultimate elusive metabolic target triggered by quorumsensing.Let us summarize the situation of the bacteria from theonset of infection.The bacteria begin growing by using therich haemocoel components (mainly amino acids) to mul-tiply exponentially. Upon reaching a high cell density theyenter the stationary phase of growth, as do other Entero-bacteria such as  Escherichia coli  . In the  mdh   mutant,Lango and Clarke observed abrupt cessation of growth inparallel with lack of synthesis of secondary metabolites,narrowing the search for regulatory cues to the TCAcycleor pathways close to it (Fig. 1). Crawford  et al  . (2010)have recently shown that L-proline, a major osmopro-tectant of insect haemolymph, upregulates secondarymetabolite production in  P. luminescens  . This regulationappears to be mediated by the LysR-type transcriptionalregulator HexA, previously shown to repress secondarymetabolism in the related species  Photorhabdus tem- perata   (Joyce and Clarke, 2003; Kontnik  et al  ., 2010). Intheir parallel work, Lango and Clarke established thatamino acids (particularly serine) are consumed first, yield-ing pyruvate, presumably in a manner similar to thatdescribed in  E. coli   grown in rich media (Wolfe, 2005).It has long been known that aerobic growth of bacteriasuch as  E. coli   on a carbon-rich food supply is biphasic(Wolfe, 2005). In the first growth phase, the bacteria growexponentially while accumulating acetate in the growthmedium to compensate for an overflow of carbon. In thesecond phase, the bacteria abruptly shift to acetate con-sumption while entering into the stationary phase. This‘acetate switch’ is the only known metabolic process thatparallels entry into the stationary phase of growth. It isfairly universal, and matches the function of the enzymesin the TCA ‘cycle’. During exponential growth, the TCApathway does not in fact cycle, but rather producesmetabolites via oxaloacetate and 2-ketoglutarate. Subse-quently, in the respiratory stage, the pathway cyclesaccording to the standard textbook depiction of the TCAcycle to produce protons permitting ATP synthesis andaccumulation of energy stores for the ‘difficult times’ahead, often in the form of polyphosphates (Danchin,2009). The acetate switch (overlaid in grey in Fig. 1) hasdrawn much attention because of the postulated role ofacetylphosphate (acetyl ~ P) as a regulator of cellular pro-cesses as diverse as nitrogen assimilation, protein deg-radation, osmoregulation, flagellar biogenesis, pilusassembly, capsule biosynthesis, biofilm development andpathogenicity (Wolfe, 2005; Mizrahi  et al  ., 2006; Pruss et al  ., 2010). Yet, there is no final demonstration of theeffect of acetyl ~ P as the principal mediator of theseprocesses.Lango and Clarke report experiments designed to testthe role of the acetate switch by disruption of genes con-trolling the switch (dashed lines in Fig. 1). They show that A new role for acetyl-phosphate?   1347  © 2010 Blackwell Publishing Ltd,  Molecular Microbiology  ,  77 , 1346–1350  while the ‘pathogenicity to mutualism’ metabolic switch issuppressed by disruption of  mdh   or the immediatelyupstream  fumC   gene, the switch still occurs when theenzymes involved in the acetate switch are disrupted.Thesituation in  P. luminescens   is illustrated in Fig. 1. Duringgrowth on serine (blue lines) the upper half of the TCAcycle is necessary to generate most of the biomass. Bycontrast, proline is degraded via glutamate and the lowerhalf of the cycle (green lines). During this phase acetyl-CoA is produced continuously, and it is not unexpectedthat there would be a surplus of acetate excreted in themedium, as was indeed observed (Wolfe, 2005; Langoand Clarke, 2010). This arrangement has the conse-quence that when  mdh   is disrupted the situation mustchange dramatically as soon as pyruvate-generatingmolecules have been consumed. For example, exhaus-tion of pyruvate-generating amino acids does not permituse of proline because of a lack of oxaloacetatesynthesis. Lango and Clarke show that under such con-ditions exponential growth of bacteria is not affected forsome time, although the growth yield is lower. Acetate isproduced, but cannot be used. Yet they remark that sec-ondary metabolite production was more significantlyaffected in  fumC   mutants than in  mdh   mutants. This wasunexpected and suggests that malate itself or an associ-ated metabolite (acetyl-CoA or glyoxylate or metabolitesconnected to them) plays a role in the process.As a casein point, the  ackA-pta   mutants were attenuated in viru-lence while all of the mutants analysed in the study werefully virulent. Taken together, these observations suggest PEP Pyruvate     p     p    c CO 2  ppsA  pykA  pykF citrate oxaloacetate isocitrate malatefumarate2-ketoglutarate  NADPH NADH CoA succinate ATP + CoA  succinyl-glyoxylate acetyl-CoA a c e B   a c e  A  acnA acnB icd sucA sucB lpd    sucC sucD fumC        p    c     k  mdh glycine UQUQH 2 NAD+NADH nuoA-N ndh wrbA   TT  glutamateaspartate sdhA sdhB sdhC    sdhD   a  c  e  E   a  c  e  F     l  p  d   A DP ATPATPAMP + PiCO 2 CoA acetateacetate yjcG yjcH T 1-pyrroline-5-carboxylateproline  putA  putA NAD+NADHFADHFAD 2 CO 2 proline CO 2  maeA (sfcA)maeB serineserine NADP+NAD+NADHNADPHCO 2 acetyl~P  a c k A  ATPA DPPi CoA   s  d  a  A  PiA DPATPA DP + PiNAD+NADP+ gluconeogenesis ATP CoA  p t a  m a e B  A MPPPi  ac s g l t  A     e a - oA  fumC  mdh c P   AMPPPi  ac s   cet  a c k A  m a e t a  Fig. 1.  The TCA cycle and the acetate switch in  P. luminescens  . Growth of bacteria is illustrated by acquisition of two amino acids, prolineand serine. Catabolism of proline is in green and of serine in blue. Genes inactivated in the study are in dashed lines, except for the gene ofmalate dehydrogenase, which is shown in red. In purple is shown the contribution of the malate enzymes. In orange the proposed contributionof the distal part of MaeB would produce acetyl-phosphate (in red). 1348  A. Danchin    © 2010 Blackwell Publishing Ltd,  Molecular Microbiology  ,  77 , 1346–1350  that when there is an entry point in the metabolism atmalate, but not fumarate, the effect is somewhat allevi-ated, and this may be related to a metabolite of theAckA-Pta pathway.These observations might provide a clue about a TCAcycle-related step that could be involved in a relevantparameter of the metabolic switch. Malate dehydroge-nase is essential for ubiquinone-dependent electrontransfer, but this function is also provided by succinatedehydrogenase. Its specific role, then, is to replenish thepool of oxaloacetate, which is essential for biomass con-struction, including gluconeogenesis. This should also bethe role of fumarate hydratase. We need therefore to lookfor some kind of branching out of the TCA cycle, withmalate as the entry point. In Enterobacteria,  Photorhab- dus   included, two decarboxylating malate dehydrogena-ses (malic enzymes) exist, MaeA and MaeB. While theyare not efficient in terms of biomass construction(because they produce carbon dioxide) they do replenishthe pool of pyruvate and produce NADH (resp. NADPH).It is expected that, in the absence of Mdh, this wouldpermit proline to produce all of the important cell metabo-lites during exponential growth (the loss of carbon dioxide,however, accounting for lower biomass yield). Wepropose this pathway as a tentative explanation of Langoand Clarke’s observations.Can we further refine this hypothesis to identify theTCA cycle-dependent metabolism controlling the meta-bolic shift diplayed by  P. luminescens  ? Malate enzymeMaeA (SfcA) does not display obvious features thatcould help us to find a clue. By contrast malate enzymeMaeB has a surprising domain organization that is con-served in many bacteria. Its amino terminus is similar toMaeA, but it is followed by a carboxy-terminal polypep-tide that is highly similar to phosphotransacetylase(PTA), one of the crucial enzymes of the acetate switch.Might MaeB generate some acetyl ~ P? To my knowledgethere is only one biochemical study that has investigatedthis hypothesis and this study reports that acetyl ~ Preciprocally regulates MaeA and MaeB (Bologna  et al  .,2007). This work showed that neither the whole MaeBprotein nor the truncated PTA domain was able to catal-yse the PTA reaction (interconversion of acetyl-CoA andacetyl ~ P) and it was proposed that this was a regulatorydomain permitting allosteric regulation of the Maedomain. However, these experiments were performedunder conditions where an essential regulatory factormight be missing. Therefore, although Lango and Clarkehave shown that an  ackA-pta   mutant is unaffected in thetransition from pathogen to mutualist it will be interestingto see if acetyl ~ P could still be the missing mediatorrequired for  P. luminescens   to support nematode growthand development, switching its metabolism from primarybiosyntheses to secondary metabolism. This possibilityis even more intriguing in light of the fact that in otherorganisms acetyl ~ P production is controlled by quorumsensing (Studer  et al  ., 2008). It is also noteworthy thatthe malic enzymes replenish the pool of NAD(P)H,thereby supplying energy for growth and biosynthesis.Clearly, there is still much to be learned from thissystem. Acknowledgements This work was supported by MICROME, a CollaborativeProject funded by the European Commission within its FP7Programme, contract number 222886-2. References Benveniste, R., and Davies, J. (1973) Aminoglycosideantibiotic-inactivating enzymes in actinomycetes similar tothose present in clinical isolates of antibiotic-resistantbacteria.  Proc Natl Acad Sci USA  70:  2276–2280.Bologna, F.P., Andreo, C.S., and Drincovich, M.F. (2007) Escherichia coli   malic enzymes: two isoforms with substan-tial differences in kinetic properties, metabolic regulation,and structure.  J Bacteriol   189:  5937–5946.Crawford, J.M., Kontnik, R., and Clardy, J. (2010) Regulatingalternative lifestyles in entomopathogenic bacteria.  Curr Biol   20:  69–74.Daborn, P.J., Waterfield, N., Blight, M.A., and ffrench-Constant, R.H. (2001) Measuring virulence factor expres-sion by the pathogenic bacterium  Photorhabdus luminescens   in culture and during insect infection. J Bacteriol   183:  5834–5839.Danchin, A. (2009) Natural selection and immortality.  Bioger- ontology   10:  503–516.Davidson, C.J., and Surette, M.G. (2008) Individuality inbacteria.  Annu Rev Genet   42:  253–268.Davies, J. (1990) What are antibiotics? Archaic functions formodern activities.  Mol Microbiol   4:  1227–1232.Davies, J. (2009) Everything depends on everything else. Clin Microbiol Infect   15  (Suppl. 1) :  1–4.Duchaud, E., Rusniok, C., Frangeul, L., Buchrieser, C.,Givaudan, A., Taourit, S.,  et al  . (2003) The genomesequence of the entomopathogenic bacterium  Photorhab- dus luminescens  .  Nat Biotechnol   21:  1307–1313.Eleftherianos, I., Boundy, S., Joyce, S.A., Aslam, S., Mar-shall, J.W., Cox, R.J.,  et al  . (2007) An antibiotic producedby an insect-pathogenic bacterium suppresses hostdefenses through phenoloxidase inhibition.  Proc Natl Acad Sci USA  104:  2419–2424.Eleftherianos, I., Joyce, S., ffrench-Constant, R.H., Clarke,D.J., and Reynolds, S.E. (2010) Probing the tri-trophicinteractionbetweeninsects,nematodesand Photorhabdus  . Parasitology   doi:10.1017/S0031182010000508.Frimmersdorf, E., Horatzek, S., Pelnikevich, A., Wiehlmann,L., and Schomburg, D. (2010) How  Pseudomonas aerugi- nosa   adapts to various environments: a metabolomicapproach.  Environ Microbiol   12:  1734–1747.Fuqua, W.C., Winans, S.C., and Greenberg, E.P. (1994)Quorum sensing in bacteria: the LuxR-LuxI family of cell A new role for acetyl-phosphate?   1349  © 2010 Blackwell Publishing Ltd,  Molecular Microbiology  ,  77 , 1346–1350  density-responsive transcriptional regulators.  J Bacteriol  176:  269–275.Joyce, S.A., and Clarke, D.J. (2003) A hexA  homologue from Photorhabdus   regulates pathogenicity, symbiosis and phe-notypic variation.  Mol Microbiol   47:  1445–1457.Joyce, S.A., Brachmann, A.O., Glazer, I., Lango, L., Schwar,G., Clarke, D.J., and Bode, H.B. (2008) Bacterial biosyn-thesis of a multipotent stilbene.  Angew Chem Int Ed Engl  47:  1942–1945.Kontnik, R., Crawford, J.M., and Clardy, J. (2010) Exploitinga global regulator for small molecule discovery in  Photo- rhabdus luminescens  .  ACS Chem Biol   5:  659–665.Lango, L., and Clarke, D.J. (2010) A metabolic switch isinvolved in lifestyle decisions in  Photorhabdus luminescens  .  Mol Microbiol   77:  1394–1405.Massey, R.C., Buckling, A., and ffrench-Constant, R. (2004)Interference competition and parasite virulence.  Proc Biol Sci   271:  785–788.Mizrahi, I., Biran, D., and Ron, E.Z. (2006) Requirementfor the acetyl phosphate pathway in  Escherichia coli  ATP-dependent proteolysis.  Mol Microbiol   62:  201–211.Pruss, B.M., Verma, K., Samanta, P., Sule, P., Kumar, S., Wu,J.,  et al  . (2010) Environmental and genetic factors thatcontribute to  Escherichia coli   K-12 biofilm formation.  Arch Microbiol   doi:10.1007/s00203-010-0599-z.Studer, S.V., Mandel, M.J., and Ruby, E.G. (2008) AinSquorum sensing regulates the  Vibrio fischeri   acetateswitch.  J Bacteriol   190:  5915–5923.Waterfield, N.R., Ciche, T., and Clarke, D. (2009)  Photorhab- dus   and a host of hosts.  Annu Rev Microbiol   63:  557–574.Wolfe, A.J. (2005) The acetate switch.  Microbiol Mol Biol Rev  69:  12–50. 1350  A. Danchin    © 2010 Blackwell Publishing Ltd,  Molecular Microbiology  ,  77 , 1346–1350
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks