32. M. R. Grant et al., Proc. Natl. Acad. Sci. U.S.A. 95,15843 (1998).

33. Unpublished data of authors (39), comparing about 2kb of flanking sequence in 23 A. thaliana accessions.

34. To calculate changes in codons with more than oneposition substituted, current algorithms for calculating Ks and Ka assume equal rates of synonymous andreplacement changes. This assumption leads to theoverestimation of synonymous changes and under-

estimation of replacement changes when the re-placement rate actually exceeds the synonymousrate ( J. Comeron, M. Kreitman, unpublished results).

35. W.-H. Li, Molecular Evolution (Sinauer Associates,Sunderland, MA, 1997).

36. R. J. Abbott, M. F. Gomes, Heredity 62, 411 (1989).37. R. M. May, R. M. Anderson, Proc. R. Soc. London B

219, 281 (1983).38. Supplemental material is available on Science Online

at www.sciencemag.org/cgi/content/full/292/5525/2281/DC1.

39. J. Bergelson, M. Kreitman, E. A. Stahl, D. Tian, unpub-lished data.

40. R. Mauricio, unpublished data.41. We gratefully acknowledge support from the NIH

to J.B. and M.K., a Packard Fellowship to J.B., and aSloan/Department of Energy fellowship to E.A.S. K.Thornton provided computer programs.

R E V I E W

Common and Contrasting Themes ofPlant and Animal Diseases

Brian J. Staskawicz,1* Mary Beth Mudgett,1 Jeffrey L. Dangl,2 Jorge E. Galan3

Recent studies in bacterial pathogenesis reveal common and contrastingmechanisms of pathogen virulence and host resistance in plant and animaldiseases. This review presents recent developments in the study of plantand animal pathogenesis, with respect to bacterial colonization and thedelivery of effector proteins to the host. Furthermore, host defenseresponses in both plants and animals are discussed in relation to mecha-nisms of pathogen recognition and defense signaling. Future studies willgreatly add to our understanding of the molecular events defining host-pathogen interactions.

The ability of pathogenic microorganisms toharm both animal and plant hosts has beendocumented since the initial demonstration inthe 1870s that microbes were causal agents ofdisease. Since the initial discoveries by Koch(1) that Bacillus anthracis caused anthrax andby Burrill (2) that Erwinia amylovora causedfire blight in pears, our knowledge base hasexpanded enormously. Today, the genomesof the most important animal and plant patho-gens have been or will be sequenced, and themolecular basis of pathogenicity is beginningto be deciphered. Furthermore, several modelplant and animal host genomes are fully se-quenced. Basic discoveries made in the post-genomic era will fuel our quest for develop-ing new strategies for disease control.

In the past 5 years, pivotal observationshave revealed that bacterial pathogens sharecommon strategies to infect and colonizeplant and animal hosts. One is the ability todeliver effector proteins into their respectivehost cells to mimic, suppress, or modulatehost defense signaling pathways and to en-hance pathogen fitness. On the host side,plants and animals have evolved sophisticat-ed surveillance mechanisms to recognize var-ious bacterial pathogens. Interestingly, plants

recognize distinct effectors from pathogenicbacteria, whereas animals recognize con-served “molecular patterns,” such as thosederived from lipopolysaccharide (LPS) orpeptidoglycan. The discovery that surveil-lance proteins in diverse hosts share commonprotein signatures that perform similar func-tions invites speculation as to how these re-sistance mechanisms evolved.

This rapidly expanding field is obviouslya large topic, and all aspects cannot be con-sidered here. We will mainly focus on themesof plant pathogenesis. However, when appro-priate, we will discuss unique and sharedstrategies used by microbial pathogens to in-fect animal hosts. Moreover, we will compareour emerging knowledge of pathogen surveil-lance mechanisms used by plant and animalhosts.

Initial Interactions of BacterialPathogens with Host Plant CellsThe initial interactions of bacteria with theirplant hosts are critical in determining the finaloutcome of infection. Curiously, the majority ofinfections do not lead to overt disease. Micro-organisms are usually repelled by plant defensemechanisms. However, in some cases, interac-tions of microbial pathogens with their hostlead to the overt harm to the presumptive hostand to pathology. Factors intrinsic to both thepathogen and the plant determine the final out-come of the encounter.

The epiphytic (saprophytic) life stage ofphytopathogenic bacteria often precedes en-try into the host plant and the onset of patho-genicity. For example, phytopathogenic bac-

teria in the genera Pseudomonas and Xan-thomonas can colonize leaf surfaces of plantsand reach dense bacterial populations (107

colony-forming units per square centimeter)without causing disease. Under the appropriateenvironmental conditions, bacteria enter leafmesophyll tissue through natural stomatalopenings, hydothodes, or wounds, thus makingtheir first contact with internal host cells. Phy-topathogenic bacteria multiply in the intercellu-lar spaces (apoplast) of plant cells and remainextracellular. This is in contrast to many animalbacterial pathogens that gain entry into theirhost cells and then multiply intracellularly.

Mechanism for Plant Cell Infection:Conservation of Type III SecretionSystemTo grow in the apoplast, phytopathogenic bac-teria sense their environment and induce genesrequired for host infection. A primary locusinduced in Gram-negative phytopathogenicbacteria during this phase is the Hrp locus (3).The Hrp locus is composed of a cluster of genesthat encodes the bacterial type III machinerythat is involved in the secretion and transloca-tion of effector proteins to the plant cell. Mu-tations in Hrp genes affect both the induction oflocalized plant disease resistance (the hypersen-sitive response) and bacterial pathogenicity.This mutant phenotype and the subsequentdemonstration that Hrp structural proteins andtype III effectors are transcriptionally coregu-lated (4) provided important evidence that ef-fectors not only caused disease but were thecomponents of the pathogen recognized by thehost. Although the general physiology of lowsugar and low pH in the apoplast is known toinduce the assembly of the Hrp type III appa-ratus in phytopathogenic bacteria, a specificplant-derived signal has yet to be identified.

The demonstration that phytopathogenicbacteria use genes (i.e., Hrp) that are remark-ably similar to genes encoding the type IIIsecretion system in animal pathogenic bacte-ria provided an immediate conceptual frame-work to explain the molecular mechanisms

1Department of Plant and Microbial Biology, Univer-sity of California at Berkeley, Berkeley, CA 94720,USA. 2Department of Biology and Curriculum in Ge-netics and Molecular Biology, University of NorthCarolina, Chapel Hill, NC 27599, USA. 3Boyer Centerfor Molecular Medicine, Yale University School ofMedicine, New Haven, CT 06536, USA.

*To whom correspondence should addressed. E-mail:stask@nature.berkeley.edu

www.sciencemag.org SCIENCE VOL 292 22 JUNE 2001 2285

P L A N T P A T H O L O G Y

for pathogen infection and recognition byplants (5). The overall theme is that a generalmechanism for bacterial pathogenesis inplants and animals involves the direct deliv-ery of different classes of proteins to the host.These proteins, now collectively referred toas type III effectors, suppress, stimulate, in-terfere with, or modulate host responses toinvading pathogens. We now appreciate whyplant surveillance systems have evolved bothsurface-exposed and cytoplasmic resistanceproteins to recognize phytopathogenic bacte-ria type III effectors targeted to the surfaceand the interior of the host cell. Features oftype III systems in plant and animal patho-gens have been substantially reviewed (6–8);therefore, they will not be discussed here indetail.

Bacterial-Host Surface InteractionsA common theme emerging from the study ofdiverse bacterial pathogens is that they inter-act with host cell surfaces through append-agelike structures known as pili or fimbriae,which serve as scaffolds to display or presentligands that are recognized by cognate recep-tors on host cells (9). Given the distinct fea-tures of plant and animal cells, we predict thatthe ligands displayed by the surface struc-tures of pathogenic bacteria toward their re-spective hosts are substantially different.However, the scaffolds and/or the mecha-nisms of assembly of bacterial appendagesare likely to share common features. For ex-ample, the plant pathogens produce an ap-pendage-like structure that is essential for thedelivery and regulation of effector proteins tothe plant cell through its type III system. Thispilus structure is composed of a single pro-tein: HrpA in Pseudomonas syringae (10)and HrpY in Ralstonia solanacearum (11). Amorphologically similar structure is associat-ed with type III systems in the animal patho-gens Salmonella typhimurium (12) and enter-opathogenic Escherichia coli (13). The E.coli appendage structure is also composed ofa single protein, EspA, which exhibits nosequence similarity with either HrpA orHrpY (13). These animal and plant pathogensthus secrete specific proteins through theirtype III apparatus to construct a structure that,although composed of distinct subunits, isnevertheless architecturally similar.

The plant cell wall is a unique surfacestructure that differentiates plant from animalcells. The plant cell wall is a thick barrier(;200 nm as compared to 8 to 10 nm foreukaryotic plasma membranes) throughwhich bacteria must either penetrate or local-ly diffuse effectors to incite disease. Thestructural differences between the otherwiserelated surface appendages of plant and ani-mal pathogenic bacteria are therefore in keep-ing with the marked differences between thecellular organization of plant and animal

cells. It is thus remarkable that Pseudomonasaeruginosa strain PA14 is not only a humanopportunistic pathogen but is also capable ofcausing disease in the plant Arabidopsis (14).This “interkingdom” pathogen has armed it-self with the ability to effectively colonize thesurface of both plant and animal cells and toavoid diverse host surveillance mechanisms.Future studies with this pathogen, using sev-eral host model systems, should reveal com-mon and unique features that allow it to be apathogen on such a wide range of eukaryotichosts.

A vast array of structures has been char-acterized on the surface of animal pathogenicbacteria (15). These structures facilitate thecolonization of different mucosal surfacesand, in many instances, determine the specif-ic interaction of the bacterium with receptorson host cells. These recognition events oftenlead to the activation of signaling pathways inboth the host and the pathogen that have aprofound impact on the outcome of infection.For example, pathogenic E. coli deliver areceptor to the interior of host cells, and thisreceptor is then used to perceive a secondpathogen-encoded molecule to initiate infec-tion using host signaling machinery (16). Inaddition, some E. coli pathogens use FimHadhesin of type I fimbriae to interact withmannosylated glycoprotein receptors on thesurface of mammalian cells, which leads topro-inflammatory cytokine production inbladder infections (17). It remains to be de-termined whether plant pathogenic bacteriause similar surface structures for contact withplant cells.

Delivery of Type III Effectors intoPlant CellsAlthough there is extensive circumstantialevidence that phytopathogenic bacteria di-rectly deliver effectors to plant cells, there isno direct evidence demonstrating the pres-ence of delivered effectors inside plant cells.The strongest evidence in favor of directdelivery comes from experiments that dem-onstrated that the expression of various aviru-lence genes (from here on referred to as typeIII effectors) in plant cells is sufficient toinduce a localized defense response in resis-tant plant cultivars. Additional evidencecomes from the observation that the P. syrin-gae AvrRpt2 effector protein is specificallycleaved at the NH2 terminus during infection.The processing of AvrRpt2 was shown tooccur in the host and to be dependent on afunctional type III pathway, which suggeststhat proteolysis of this effector probably oc-curs after translocation into the host cell cy-toplasm (18).

In contrast to phytopathogenic bacteria,there is ample direct evidence for delivery ofeffectors by animal pathogenic bacteria. TypeIII secretion–mediated protein translocation

of effector proteins in host cells was firstdemonstrated in Yersinia and subsequentlyobserved in additional animal pathogens (6).The mechanisms by which these proteinsreach their final destination, however, remainpoorly understood. With few exceptions, typeIII effectors from both plant and animal bac-terial pathogens are structurally diverse anddo not contain an obvious conserved signalpeptide for export. Work carried out mostlywith effectors from Yersinia spp. has definedtwo regions of type III effectors that areinvolved in their secretion and/or transloca-tion into the host cell (6). One of these re-gions is located in the first ;15 amino acidsand the other is contained within the first 100amino acids. The latter serves as a bindingsite for customized chaperones that are re-quired for secretion and translocation of typeIII effectors. In addition, the first ;15 codonsof the mRNA that encodes a subset of type IIIeffectors may also be involved in their secre-tion and/or translation regulation (19). How-ever, this model is still debated. Additionalstudies are required to verify and demonstratethe universality of the proposed mechanismsof secretion in animal and plant pathogenicbacteria. It is thus conceivable that there willbe unique mechanisms by which proteins areengaged by the secretion machinery. In fact,in some type III systems (such as S. typhi-murium), the timing of host responses stimu-lated by specific effectors suggests an orderlydelivery of each protein into the host cell(20). Perhaps the existence of different secre-tion signals operating in type III systems mayprovide the molecular basis for such a hier-archy of protein secretion.

Effector proteins can be secreted in vitro,in a type III–dependent manner, from phyto-pathogenic bacteria in the genera Erwinia,Pseudomonas, Ralstonia, and Xanthomonas(8). These experiments revealed that secre-tion signals in these bacteria are also local-ized to the NH2 terminal portion of the effec-tor proteins (21). The secretion signal toler-ates alterations in the reading frame, whichsuggests an involvement of the mRNA in thesecretion process (19). The observation thatthe type III secretion systems of Xanthomo-nas campestris and E. chrysanthemi secreteeffector proteins from both plant and animalpathogens suggests some degree of ancestralconservation between type III systems (19,22). However, substantial differences areapparent. For example, although commonin type III secretion systems of animalpathogens (23), the presence of customizedchaperones for the secreted effectors intype III systems of plant pathogenic bacte-ria is formally lacking. Thus far, there isonly one report of a type III secretion–associated gene that encodes a polypeptidewith features commonly found in chaper-ones (24 ).

22 JUNE 2001 VOL 292 SCIENCE www.sciencemag.org2286

P L A N T P A T H O L O G Y

Another apparent difference betweenplant and animal pathogenic bacteria residesin a critical component of the supramolecularstructure associated with type III systems,termed the needle complex. This structure,best characterized in the animal pathogensSalmonella and Shigella, is composed of aprotein complex forming a ring base embed-ded in the bacterial envelope and a needlelikeprojection (25). Although homologs of thecomponents of the base structure are presentin type III systems from plant pathogenicbacteria, there are no apparent homologs ofthe protein subunit that makes up the needleportion (25). Because this component is high-ly conserved among animal pathogens, itsabsence from plant pathogenic bacteria sug-gests a degree of specialization in this part ofthe secretion apparatus. The chaperones andthe needle substructure are thought to becritical for functions related to the transloca-tion of effector proteins into host cells. There-fore, the theme emerging is that the type IIIsecretion components that are highly con-served between plant and animal pathogenicbacteria are those involved in secretion fromthe bacteria.

Function of Type III Effector Proteinsfrom Phytopathogenic BacteriaThe role of type III effectors in plant cellsremains elusive. Most phytopathogenicbacterial effectors show no homology to

known proteins in existing databases. Theirrole in virulence has been inferred frommutagenesis studies that have shown that,in the absence of these avirulence type IIIeffector genes, there is a loss of virulenceon host plants that do not contain cognatedisease resistance genes (26 ). The onlytype III effector that shows significant ho-mology to known proteins in a BLASTsearch is AvrBs2 from X. campestris. Thisprotein shares homology with phosphodies-terases, which suggests that AvrBs2 mayfunction in plants as an enzyme. AvrBs2 isfound in most pathovars of X. campestrisand is required for full pathogen fitness.Molecular studies show that X. campestrisis evolving under selection pressure at theavrBs2 locus to evade host recognition andto maintain larger bacterial populations inits host (27 ).

The recent application of structure predic-tion programs to predict the secondary struc-ture of the YopJ family of type III effectorshas provided new insight into the possiblefunction of these proteins (28). The YopJhomologs can be found in plant (Erwinia,Pseudomonas, and Xanthomonas) and animalpathogenic bacteria (Yersinia and Salmonel-la) and in a plant symbiont (Rhizobium) (Ta-ble 1). The predicted secondary structure ofsome YopJ homologs is similar to the knownstructure of adenovirus protease (AVP), acysteine protease. All of these effectors pos-

sess the critical residues (His, Glu, and Cys)forming the catalytic triad in the AVP cys-teine protease (Table 2). Mutation of thepredicted catalytic core in YopJ and AvrBsTinactivated both proteins, which suggests thatthese residues are critical for function. Themammalian substrates for YopJ are highlyconserved ubiquitin-like molecules, namelySUMO-1 protein conjugates (28). The plantsubstrates for AvrBsT are not yet known.This work suggests that YopJ-like effectorsmay modulate host signaling by disruptingsignal transduction events regulated bySUMO-1 post-translational modification.

Defining the molecular mechanism of thesuppression of plant host defenses is a sub-stantial challenge in the area of plant-microbeinteractions. However, new evidence sug-gests that plant pathogen type III effectorsmay work in plants to suppress host defensesignal transduction (29). Stable transgenicArabidopsis plants lacking the correspondingresistance gene and expressing the P. syrin-gae AvrRpt2 effector were more susceptibleto P. syringae infection.

One theme emerging from the elucidationof the functions of several type III effectorproteins is that of modulation of cellularfunctions by bacterial products that mimichost proteins. At the molecular level, hostmimicry may take two forms. One form ischaracterized by the use of direct homologsof host proteins subverted for the pathogen’sneeds. Examples of this are the tyrosine phos-phatases YopH (30) and SptP (31) from Yer-sinia and Salmonella and the YpkA serine-threonine kinase from Yersinia (32). The sec-ond form of mimicry is characterized by pro-karyotic protein surfaces that mimiceukaryotic protein surfaces. This form ofmimicry is less obvious, because it may in-volve proteins with no apparent sequencesimilarity to any known protein. We antici-pate that this form of mimicry may be morewidespread among type III effectors and thatit may only be deciphered by the close exam-ination afforded by crystallographic studies,as shown in the case of the guanosine triphos-phatase–activating protein (GAP) domain ofthe Salmonella effector SptP (33). The crystalstructure of SptP revealed that, although it isnot obvious at the primary sequence level,

Table 1. The YopJ family of type III protein effectors is conserved in plant and animal microbialpathogens.

Effector Target Effect

Animal pathogenS. typhimurium AvrA UnknownYersinia spp. YopJ/YopP MAPK kinases,

IKKbInhibition of TNFa, inhibition of

NFkB activation, inhibition ofSUMO-1 conjugation,proapoptotic

Plant pathogenErwinia amylovora ORFB UnknownP. syringae ORF5 Unknown

AvrPpiG1 Unknown Hypersensitive responseX. campestris AvrBsT Unknown Hypersensitive response

AvrRxv Unknown Hypersensitive responseAvrXv4 Unknown Hypersensitive response

Plant symbiontRhizobium NGR234 Y4LO Unknown

Table 2. The YopJ-like effectors are proposed to be cysteine proteases. The predicted catalytic residues (His, Glu, and Cys) conserved in all protein membersis shown in bold. Protein accession numbers (GenBank) are in parentheses.

(P31498) YopJ 109 HFSVIDYK–HINGKTSLILF– EPANF–––NSMGPAMLA––IRTKTAIERYQLPDCHFSMVEMDIQRSSSECGIFS(AAB83970) AvrA 123 HISVVDFR–VMDGKTSVILF–EPAAC–––SAFGPA–LA––LRTKAALEREQLPDCYFAMVELDIQRSSSECGIFS(AAD39255) AvrBsT 154 HHAAIDVR–FKDGKRTMLVI–EPALAYGMKDGEIKVMAGYETLGKNVQNCLGENGDMAVIQLGAQKSLFDCVIFS(AAA27595) AvrRxv 180 HRVAFDVRNHESGHTTIIAL–EPASAYN–––––PDHMPGFVKMRENLTSQFGRKISFAVIEAEALKSIGGCVIFS(AAG39033) AvrXv4 155 HHFAVDVKHHENGASTLIVL–ESASAGN–––––EIALPGYTKLASMLRSKFGGSARMVVIEAEAQKSLNDCVIFA(CAC16700) ORF5 243 HHIALDIQLRYGHRPSIVGF–ESAPGNIIDAAEREILSAL––––––––––––GNVKIKMVGNFLQYSKTDCTMFA(AAF71492) AvrPpiGl 160 HHVAVDVRNHSNGQKTLIVL–EPITAYKDDVYPPAYLPGYPQLREEVNTRLRGNAKMSVIETDAQRSWHDCVIFS(AAF71492) ORFB 235 HHRVALDIQFRPGHRPSVVGYESAPGNLAEHLKYGLEHGL––––––––––R––GAKVQVVANTIQNSVRGCSMFA(AAB92459) Y4LO 123 HHVAADVRTRAGAAPTIIVM–EGANFY–––––––TFVASYFKLRGDSFRQLGTQAKWAFIEVGAQKSAADCVMFG

www.sciencemag.org SCIENCE VOL 292 22 JUNE 2001 2287

P L A N T P A T H O L O G Y

this protein shares fundamental structural do-mains that are characteristic of eukaryoticGAPs. Therefore, despite the differences inthe general architecture, the actual surfacepresented by SptP to its cellular targets hasmuch in common with that of its eukaryoticcounterparts.

Overall, type III effectors do not exhibitobvious sequence similarities to other pro-teins, despite the fact that, at least function-ally, some behave like eukaryotic proteincounterparts. In these cases, convergent evo-lution may have sculpted protein surfaces thatmimic their eukaryotic “homologs.” Thismay be the case with the LRR (leucine-richrepeat) effector proteins identified in plant(Ralstonia) (6) and animal (Yersinia, Salmo-nella, and Shigella) bacterial pathogens (34).One role of these LRR effector proteinsmight be to interfere with or modulate LRRprotein receptors in the host that are in-volved in recognizing signals from invadingpathogens.

From the bacterial perspective, there arestill many unanswered questions in thefields of plant and animal pathogenesis.Several key questions remain. For instance,how many different bacterial effectors aredelivered by a given pathogenic strain to itsrespective hosts? Is this suite of effectorsdifferent among isolates that are pathogenicon the same host species, suggesting func-tional redundancy at the population level?More important, what are the molecular

functions of these proteins? How are theseeffectors detected by surveillance mecha-nisms in resistant hosts?

Host Surveillance Mechanisms ofPlants and AnimalsThe host environments of plant and animal cellsare vastly different and present unique chal-lenges to invading pathogens. Because plantslack a circulatory system, each cell must becapable of responding to an invading pathogen.The genetic basis of plant resistance is oftencontrolled by single resistance genes evolved torecognize organisms expressing specific aviru-lence genes. As detailed above, these generallyencode type III effector proteins. This geneticinteraction is the basis of the gene-for-genehypothesis originally proposed by Flor in the1940s (35). A flurry of activity in the past 8years has resulted in the cloning and identifica-tion of plant disease resistance genes from bothmodel and agronomically important plants(36). These studies reveal that several classes ofproteins are involved in plant disease resistance(Fig. 1). These include Cf2, Cf4, Cf5, Cf9, Ve1,and Ve2 (37), transmembrane proteins contain-ing extracellular LRRs; Xa21, a transmembraneprotein containing extracellular LRRs and acytoplasmic serine-threonine kinase; and Pto, acytoplasmic serine-threonine kinase. However,the NB/LRR (nucleotide binding site/LRR)class of proteins is the most prevalent (38). Thissame class of proteins is capable of recognizingviruses, bacteria, fungi, nematodes, and insects.

The NB/LRR class of resistance proteins can bedivided into two subclasses based on conservedNH2-terminal motifs. One class contains acoiled coil (CC) domain that contains a putativeleucine zipper domain (such as RPS2 andRPM1), whereas the other class contains signif-icant homology with the Toll/interleukin recep-tor (TIR) domain (such as N, L6, and RPP5,and not to be confused with the Tir proteinencoded by enteropathogenic E. coli). The pre-cise molecular events mediated by NB/LRRproteins, leading to pathogen recognition inplants, have not yet been established. However,NB/LRR proteins appear to be cytosolic recep-tors that are sometimes associated with theplasma membrane (39), where they may becapable of directly or indirectly perceivingpathogen effectors as they enter the plant cell.Two hybrid protein studies suggest that theremay be a direct interaction between the patho-gen effector and the host resistance protein (18),whereas other studies in plants suggest the pos-sibility of indirect interactions mediated by aprotein signaling complex with other host pro-teins (40). The molecular characterization ofmany NB/LRR disease resistance proteins fromseveral plant families immediately allowed acomparison to be drawn with proteins fromother organisms that contain similar motifs(41). The identification of the TIR domains inthe N protein of tobacco and the L protein offlax suggests that plants, insects, and mammalsmight use proteins with similar domains toresist infection.

Fig. 1. Schematic dia-gram illustrating sev-eral proteins presentin plants, Drosophila,and mammals thatinitiate signal trans-duction cascades afterpathogen infection,which ultimately leadto disease resistance.

22 JUNE 2001 VOL 292 SCIENCE www.sciencemag.org2288

P L A N T P A T H O L O G Y

The demonstration that Toll receptors inboth Drosophila and mammals play a role ininnate immunity has reinforced this concept (42,43). The Toll family encodes transmembraneproteins containing extracellular LRRs and anintracellular motif homologous to the interleu-kin-1 receptor, suggesting a role in triggeringsignal transduction. The Toll pathway originallydescribed in Drosophila controls the establish-ment of the dorsal-ventral axis in the developingembryos. However, further studies establishedthat the Drosophila Toll pathways also inducethe expression of antimicrobial peptides (42).Genetically defined components of the dorsal-ventral polarity system also function in theDrosophilia innate immune response, but cer-tain steps in the signaling pathways are uniqueto these two very different processes. Moreover,there is some specificity in the transcriptionaloutput of different Toll family members’ func-tion in the innate immune response. For exam-ple, Toll controls the induction of the antifungalgene drosomycin, whereas the Toll family mem-ber 18-wheeler controls resistance to bacterialpathogens by the induction of attacin. The Dro-sophila genome project predicts that there are atleast nine genes encoding Toll-like receptors. Arole in innate immunity for these additionalToll-like receptors remains to be determined.

It is now well established that mammalianToll-like receptors (TLRs) play a key role ininnate immunity (43, 44). For example, it hasbeen shown that the LRR domains of TLRs areinvolved in the recognition of conserved patho-gen components. TLR2 is involved in the rec-ognition of bacterial lipoprotein and lipotheicoicacid, TLR4 in the recognition of LPS, TLR9 inthe recognition of unmethylated DNA, andTLR5 in the recognition of flagellin. FLS2 fromArabidopsis uses extracellular LRRs to specifi-cally recognize a well-conserved peptide frombacterial flagellin (45), and this recognition trig-gers typical cellular responses. The recognitionof these pathogen-associated molecular patternsby the extracellular LRR motif is also involvedin induced expression of cytokines that stimu-late the adaptive immune response. The se-quences of human and mouse genomes suggestthat there are many more TLRs. The exact rolethese putative receptors play in innate immunityawaits further experimentation.

The extracellular TLRs are able to recog-nize pathogen ligands presented to the hostfrom the outside. However, very recent studiessuggest that mammalian cells also use intracel-lular protein receptors NOD1 and NOD2 thatshare homology with the NB/LRR superfamilyof plant disease resistance proteins (46). Thesestudies have demonstrated that NOD1 andNOD2 function as intracellular receptors forbacterial LPS and play a key role in innateimmunity. The exciting discovery that the geneencoding NOD2 appears to be associated withCrohn’s disease in humans has widespread im-plications for human health (47). The NOD

family proteins are structurally homologous tothe apoptosis regulators APAF1/CED4 and toNB/LRR disease resistance proteins. They con-tain an NH2-terminal caspase recruitment do-main (CARD), a central NB domain, andCOOH-terminal LRRs. An examination of thehuman genome sequence suggests that theremay be over 30 NOD gene homologs, express-ing NB/LRR domains but with potentially dif-ferent NH2-terminal domains. Future researchwill undoubtedly be exciting as researchers be-gin to genetically uncover the role of intracel-lular NOD receptors in induction of the mam-malian innate immune response.

Conclusions and Future PerspectivesThe elucidation of the molecular events in-volved in pathogen recognition and the trigger-ing of signal transduction events that lead topathogen inhibition is a common goal of re-searchers in both the plant and animal fields.The demonstration that pathogenic bacteria thatcolonize such diverse host substrates containconserved type III secretion machines is trulyremarkable. However, distinct differences existin both the proteins that make up the secretionmachinery and in the types of effector proteinsthat are delivered to the respective host cells.The structural differences that distinguish plantsand animals will most likely specify the evolu-tion and selection of particular effectors thatinterfere with or modulate host function. Theconservation of domains in host surveillanceproteins that perform similar functions raisesquestions about the mechanism of their evolu-tion. However, it remains to be shown that thebiochemical mechanisms responsible for patho-gen inhibition are also conserved.

A major goal in studying disease is to designnew and effective methods for preventing dis-ease. The uncovering of the precise molecularevents controlling the delivery of type III effec-tor proteins should eventually allow the designof compounds that specifically interfere withthese processes without deleterious side effectsin the host. One can also envision the design ofnovel plant resistance genes that target con-served pathogen ligands that are essential forpathogenicity. This will only become a realityonce we understand the molecular basis of plantand animal diseases in sufficient detail to beable to engineer resistance proteins in a predict-able fashion. The construction of transgenicplants expressing these engineered proteinsmay provide effective and durable disease re-sistance. Future research holds promise as newdiscoveries will continue to uncover commonand contrasting mechanisms of pathogen viru-lence and host resistance in plant and animaldisease.

References and Notes1. R. Koch, Beitr. Biol. Pflanz. 2, 277 (1876).2. T. J. Burrill, Trans. Ill. State Hortic. Soc. 11, 79 (1878).3. J. R. Alfano, A. Collmer, J. Bacteriol. 179, 5655 (1997).

4. T. V. Huynh, D. Dahlbeck, B. J. Staskawicz, Science245, 1374 (1989).

5. S. Fenselau, I. Balbo, U. Bonas, Mol. Plant-MicrobeInteract. 5, 390 (1992); F. Van Gijsegem, S. Genin, C.Boucher, Trends Microbiol. 1, 175 (1993).

6. G. R. Cornelis, F. Van Gijsegem, Annu. Rev. Microbiol.54, 735 (2000).

7. J. E. Galan, A. Collmer, Science 284, 1322 (1999).8. S. Y. He, Annu. Rev. Phytopathol. 36, 363 (1998).9. G. E. Soto, S. J. Hultgren, J. Bacteriol. 181, 1059

(1999).10. W. Wei et al., Proc. Natl. Acad. Sci. U.S.A. 97, 2247

(2000).11. F. Van Gijsegem, J. Vasse, J.-C. Camus, M. Marenda, C.

Boucher, Mol. Microbiol. 36, 249 (2000).12. C. Ginocchio, S. B. Olmsted, C. L. Wells, J. E. Galan,

Cell 76, 717 (1994).13. S. Knutton et al., EMBO J. 17, 2166 (1998).14. J. M. Plotnikova, L. G. Rahme, F. M. Ausubel, Plant

Physiol. 124, 1766 (2000).15. F. G. Sauer, M. A. Mulvey, J. D. Schilling, J. J. Martinez,

S. J. Hultgren, Curr. Opin. Microbiol. 3, 65 (2000).16. B. Kenny et al., Cell 91, 511 (1997).17. M. Hedlund et al., J. Infect. Dis. 183, S47 (2001).18. S. Kjemtrup, Z. Nimchuk, J. L. Dangl, Curr. Opin.

Microbiol. 3, 73 (2000).19. D. M. Anderson, D. E. Fouts, A. Collmer, O. Schnee-

wind, Proc. Natl. Acad. Sci. U.S.A. 96, 12839 (1999).20. Y. Fu, J. E. Galan, Nature 401, 293 (1999).21. M. B. Mudgett et al., Proc. Natl. Acad. Sci. U.S.A. 97,

13324 (2000); D. S. Guttman, J. T. Greenberg, Mol.Plant-Microbe Interact. 14, 145 (2001).

22. O. Rossier, K. Wengelnik, K. Hahn, U. Bonas, Proc.Natl. Acad. Sci. U.S.A. 96, 9368 (1999).

23. P. Wattiau, S. Woestyn, G. R. Cornelis, Mol. Microbiol.20, 255 (1996).

24. S. Gaudriault, L. Malandrin, J. P. Paulin, M. A. Barny,Mol. Microbiol. 26, 1057 (1997).

25. T. Kubori et al., Science 280, 602 (1998).26. F. F. White, B. Yang, L. B. Johnson, Curr. Opin. Plant

Biol. 3, 291 (2000).27. W. Gassmann et al., J. Bacteriol. 182, 7053 (2000).28. K. Orth et al., Science 290, 1594 (2000).29. Z. Y. Chen, A. P. Kloek, J. Boch, F. Katagiri, B. N.

Kunkel, Mol. Plant-Microbe Interact. 13, 1312 (2000).30. J. A. Stuckey et al., Nature 370, 571 (1994).31. K. Kaniga, J. Uralil, J. B. Bliska, J. E. Galan, Mol.

Microbiol. 21, 633 (1996).32. E. E. Galyov, S. Hakansson, A. Forsberg, H. Wolf-Watz,

Nature 361, 730 (1993).33. C. E. Stebbins, J. E. Galan, Mol. Cell 6, 1449 (2000).34. E. A. Miao et al., Mol. Microbiol. 34, 850 (1999).35. H. H. Flor, Phytopathology 45, 680 (1955).36. J. Ellis, P. Dodds, T. Pryor, Curr. Opin. Plant Biol. 3,

278 (2000).37. L. M. Kawchuk et al., Proc. Natl. Acad. Sci. U.S.A. 98,

5945 (2001).38. B. C. Meyers et al., Plant J. 20, 317 (1999).39. D. C. Boyes, J. Nam, J. L. Dangl, Proc. Natl. Acad. Sci.

U.S.A. 95, 15849 (1998).40. R. T. Leister, F. Katagiri, Plant J. 22, 345 (2000).41. B. J. Staskawicz, F. M. Ausubel, B. J. Baker, J. G. Ellis,

J. D. Jones, Science 268, 661 (1995).42. K. V. Anderson, Curr. Opin. Immunol. . 12, 13 (2000).43. R. Medzhitov, C. Janeway, Trends Microbiol. 8, 452

(2000).44. A. Aderem, R. J. Ulevitch, Nature 406, 782 (2000).45. L. Gomez-Gomez, T. Boller, Mol. Cell 5, 1003 (2000).46. N. Inohara et al., J. Biol. Chem. 274, 14560 (1999); Y.

Ogura et al., J. Biol. Chem. 276, 4812 (2001).47. Y. Ogura et al., Nature 411, 603 (2001); J.-P. Hugot et

al., Nature 411, 599 (2001).48. We gratefully acknowledge K. Anderson and M.

Axtell for critical reading of the manuscript and E.Torok Clark for preparation of the figure. Work inthe Staskawicz lab is supported by the U.S. Depart-ment of Energy (DOE), NSF, and Syngenta. Work inthe Dangl lab is supported by grants from NSF,NIH, DOE, the U.S. Department of Agriculture, andSyngenta. Work in the Galan Lab is supported byNIH.

26 April 2001; accepted 18 May 2001

www.sciencemag.org SCIENCE VOL 292 22 JUNE 2001 2289

P L A N T P A T H O L O G Y