be ,os-

a c -

ant in

ups

of ag- ace rti- o m

)le, en, re- ein ain ein

led, oat s in ? e l -

, in iral rth )el- ggs. .ace iger

Ini- In- /is-

nee? Molecular Genetics of Plant Disease Resista Brian J. Staskawicz,* Frederick M. Ausubel, Barbara J. Baker,

Jeffrey G. Ellis, Jonathan D. G. Jones

Plant breeders have used d isease resistance genes (R genes) to contro l p lant d isease since the turn of the century. Molecular cloning of R genes that enable plants to resist a diverse range of pa thogens has revealed that the prote ins encoded by these genes have several features in c o m m o n . These f indings suggest that p lants may have evo lved com- mon signal t ransduct ion mechan isms for the express ion of res istance to a w ide range of unrelated pathogens. Character izat ion of the molecu lar s ignals involved in pa thogen recogni t ion and of the mo lecu la r events that specify the express ion of res is tance may lead to novel strategies for p lant d isease control .

Plants, like animals, are continually ex- posed to pathogen attack. Because plants lack a circulatory system and antibodies, they have evolved a defense mechanism that is distinct from the vertebrate immune system (l). In contrast to animal cells, each plant cell is capable of defending itself by means of a combination of constitutive and induced defenses (2). Knowledge about the genetic and biochemical basis of plant dis- ease resistance has accumulated since the turn of the century, when plant breeders first recognized that resistance was often con- trolled by Mendelian genes (3). The dem- onstration that plants have geographical centers of origin (4) and have co-evolved with their pathogens was a pivotal discovery for plant breeders and has led to the use of interspecific hybrids between crops and their wild relatives as sources of resistant germ plasm (5). Until 1992, however, no plant R gene had been cloned and charac- terized at the molecular level (6). Since then, R genes from several plant species

have been cloned; this constitutes a major advance for molecular plant biology and may lead to the development of novel meth- ods for disease control.

Plant Responses to Pathogen Attack

The range of phytopathogenic organisms that attack plants is diverse and includes viruses, mycoplasma, bacteria, fungi, nema- todes, protozoa, and parasites (7). Each has a unique mode of pathogenicity. Despite the vast array of potential phytopathogens, resis- tance (lack of susceptibility) is the rule and susceptibility is the exception. Why one pathogen can cause disease in one plant but not in other plants--a phenomenon often termed nonhost resistance--remains an im- portant unsolved problem in plant pathology.

Resistance to a pathogen is manifested in a variety of ways and is often correlated with a hypersensitive response (HR), local- ized induced cell death in the host plant at

Plant host cell

Avirulent pathogen f ]

I'a.v,.u,once genq..*.n--I °.o. I

j L J

Virulent pathogen .Plant host cell

Plant host cell V ruenpa,hog.n [ ] f..,ho,enes. ~genes j DISEASE

Virulent pathogen I ~virulence g e n e ~ I Pathogenesisl ~genes

Plant host cell

DISEASE ] Fig. 1. Gene-for-gene interactions specify plant disease resistance. Resistance is only expressed when a plant that contains a specific R gene recognizes a pathogen that has the corresponding avirulence gene (upper left panel). All other combinations lead to lack of recognition by the host, and the result is disease. Green represents hypersensitive response; yellow represents susceptibility to disease.

SCIENCE • VOL. 268 • 5 MAY 1995

the site of infection (8). Although the mo- lecular mechanism is obscure, HR is thought to be responsible for the limitation of pathogen growth. Resistance does not always invo;ve visible HR, which may.re- flect either HR limited to individual plant cells or other uncharacterized defense mechanisms. Alternatively, the pathogen could lack a specific pathogenicity function required to cause disease in the host, or the host could lack a specific "susceptibility" factor. Although this review concerns the molecular basis of HR-mediated resistance, the elucidation of the mechanisms involved in nonhost resistance without HR may be an important component of future attempts to control plant disease.

The genetic basis of HR-mediated dis- ease resistance was first clarified by Flor, who demonstrated that the resistance of flax to the fungal pathogen Mehmpsora lini was a consequence of the interaction of paired cognate genes in the host and the pathogen (9). His work provided the theoretical basis for the gene-for-gene hypothesis of plant- pathogen interactions and for the molecular cloning of pathogen avirulence (avr) genes and their corresponding plant R genes. An avr gene gives the pathogen an avirulent phenotype on a host plant that carries the corresponding R gene (Fig. 1 ) (10). In gene- for-gene interactions, the induction of the plant defense response that leads to HR is initiated by the plant's recognition of spe- cific signal molecules (elicitors) produced by the pathogen; these elicitors are encoded directly or indirectly by avirulence genes, and R genes are thought to encode receptors for these elicitors. Elicitor recognition acti- vates a cascade of host genes that leads to HR and inhibition of pathogen growth (I 1 ).

Gene-for-gene systems involving HR have been described for pathosystems in- volving intracellular obligate pathogens (vi-

B. J. Staskawicz is in the Department of Plant Biology, University of California, Berkeley, CA 94720, USA. F. M Ausubel is in the Department of Genetics, Harvard Medical School, Boston, MA 02115, USA, and Department of Mo- lecular Biology, Massachusetts General Hospital, Boston, MA 02114, USA. B. J. Baker is at the Plant Gene Expres- sion Center, Agncultural Research Service, U.S. Depart- ment of Agriculture, Albany, CA 94710, USA, and Depart- ment of Plant Biology, University of California, Berkeley, CA 94720, USA. J. G. Ellis is at the CSIRO Division of Plant Industry and Cooperative Research Centre for Plant Sci- ence, Post Office Box 1600, Canberra, A.C.T., Australia. J. D. G. Jones is at the Sainsbury Laboratory, John Innes Centre, Norwich Research Park, Colney Lane, Norwich NR4 7UH, UK. *To whom correspondence should be addressed. E-mail: stask@garnet.berkeley.edu

661

ruses and mycoplasmas) as well as for in- tercellular facultative and obligate patho- gens (bacteria, fungi, and nematodes). These findings suggest that common or similar recognition and signal transduc- tion mechanisms are involved in different gene-for-gene signaling pathways (2, 1 I). Physiological features of HR common to a plant's response to different pathogens in- clude a rapid oxidative burst, ion fluxes characterized by K+-H + exchange, cellu- lar decompartmentalization, cross-linking and strengthening of plant cell wall, pro- duction of antimicrobial compounds (phy- toalexins), and induction of pathogenesis- related (PR) proteins such as chitinases

and glucanases (2). These events charac- terize a plant's defense response irrespec- tive of the pathogen, although any partic- ular defense response can vary with re- spect to timing, celt autonomy, or inten- sity. The mechanism by which these events limit the growth of specific patho- gens remains unknown.

HR (12) and other necrotic reactions are hypothesized to trigger a subsequent re- sponse, referred to as systemic acquired resis- tance (SAR), that acts nonspecifically throughout the plant: SAR reduces the se- verity of disease caused by all classes of pathogens, including normally virulent pathogens (13). Experimental evidence sug-

Fig, 2. Disease-resistant and -susceptible pheno- types of TMV, M. Iini, and P. syringae inoculated on their respective hosts. TMV was inoculated on a resistant NN tobacco plant (A) and on a sus- ceptible nn tobacco plant (B). Melampsora lini was inoculated on a resistant L 6 L 6 flax plant (C) and on a susceptible / 6 / 6 flax plant (D). Pseudomonas syringae was inoculated on a resistant RPS2 RPS2 Arabidopsis plant (E) and on a susceptible rps2 rps2 Arabidopsis plant (F).

662 SCIENCE • VOL. 268 • 5 M A Y 1995

gests that HR induces an unidentified diffi_ls- ible signal; salicylic acid is known to be involved in both HR and SAR, but may not participate in the systemic signaling pathway that induces SAR (14-16). SAR may be involved in general resistance in field situa.. tions where plants undergo HR. Manipula- tion of SAR by chemical inducers or by genetic engineering may aid disease control.

Plants that are susceptible to a given pathogen still attempt to defend them- selves. Indeed, many of the defense respons- es observed in resistant plants, with the exception of HR, are also observed in sus- ceptible plants, although usually later after the infection. For example, even disease lesions in susceptible plants frequently have a defined and delimited shape, suggesting that the host is limiting the growth of the pathogen. By contrast, Arabidopsis tMliana mutants with decreased ability to synthesize phytoalexins (17) or to induce SAR (18), and Arabidopsis and tobacco plants engi- neered to degrade salicylic acid (15), devel- op larger lesions, are susceptible to very small amounts of pathogen infiltration, and allow the pathogen to grow to high titers.

Role of Pathogen Avirulence Genes in Triggering the Plant Defense Response

Genetic, biochemical, and physiological techniques have been used to study respons- es to pathogen attack in heterozygous pop- ulations of crop plants such as maize, soy- bean, bean, parsley, tomato, potato, and bar- ley. Because plant defense responses are sim- ilar irrespective of the pathogen, it has been difficult to provide compelling evidence for the significance of particular responses in conferring specific resistance. Thus, tomato or the crucifer A. thaliana have been used as model hosts to study plant responses to pathogen attack (19). The receptor-ligand model postulates that pathogen avr genes specify elicitor molecules that induce disease resistance in host plants that contain a cog- nate R gene. This has been confirmed in studies of the interaction between tomato and its leaf mold pathogen Cladosporium fulvum (20, 21). Small peptides extracted from tomato leaves infected with a pathogen race carrying an avr gene had the features expected of an avirulence gene-encoded elicitor; the purified peptides elicited an HR-like response on tomato cultivars that were resistant to the C. fulvum race used to obtain the peptide (20). In the best-docu- mented cases, t w o C. fulvum avirulence genes, avr9 and avr4, were shown to encode precursors of elicitor peptides that specifical- ly elicited HR in tomato plants that har- bored the corresponding R genes Cf-9 and Cf-4, respectively (22).

The only bacterial avirulence gene in

'us-

be ~Ot ray be

ua-

l la- by

ro]. een ~m- ,ns-

the

;us-

fter '.ase

a v e

ing the arm size '8), ~gi- Jel- 'ery and :rs.

ical )ns-

,op- ~oy- Dar- ',im- e e n

for in

lato d as

tO

and :nes ease zog- 1 in

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~IF&I I I | [ i l

which the avr-generated signal has been controls resistance to race 1 isolates of avrRpt2, was identified by isolation of Ara- definitively identified is the Pseudomonas Cochliobolus carbonu~, was identified by bidopsis mutants that did not exhibit HR in.iS syringae pv. glycinea avrD locus. This locus transposon tagging u ith the maize (Mu) response to P. syringae strains carrying avr- ~' encodes enzymes involved in the synthesis of exported syringolides that elicit HR in soybean cultivars carrying the R gene Rpg4 (23). In the case of tobacco mosaic virus (TMV), the viral-encoded coat protein ap- pears to function as a specific intracellular elicitor' that activates HR in Nicotiana syl- vestris cultivars that carry the R gene N ' (24). Uncovering the molecular basis of how these genes and other cloned aviru- lence genes digit a plant defense response will ultimately be necessary for a complete molecular understanding of host-pathogen specificity.

Why do pathogens contain avirulence genes? The avr genes may encode pathoge- nicity factors that confer a selective advan- tage for the pathogen, as has been shown for several bacterial avirulence genes that confer enhanced virulence on susceptible hosts, that is, on hosts that do not carry a cognate R gene (I 0). In the case of fungal pathogens, no role in pathogenicity for avr-encoded elicitor peptides has yet been established.

Cloning and Characterization of Plant R Genes

The cloning of several R genes since 1992 reflects, in part, the simultaneous devel- opment of the infrastructures required for insertional mutagenesis and positional cloning in several plant species. For many years, transposons have been exploited as insertional mutagens for efficient identifi- cation and isolation of genes (a process termed transposon tagging) in a wide range of organisms, including plants. The Tam elements of snapdragon and the Ac/ Ds, Spm, and Mu elements of maize have been used for the isolation of a variety of genes. Fortunately, members of the maize Ac and Spm transposon families function when transferred into heterologous plant species; this attribute permits the engi- neering of efficient gene tagging systems in a variety of plant species (25). Trans- poson-based gene tagging systems have been used to clone R genes from maize, tobacco, tomato, and flax (26-29).

Map-based positional cloning of toma- to and Arabidopsis genes has become fea- sible with the development of high-densi- ty physical-genetic maps for these two spe- cies (30). The small genome size of Ara- bid@sis (~150 Mb) and the relatively small genome size of tomato ( - 9 5 0 Mb), and the relatively small number of repeat- ed sequences in these species, have facili- tated the successful positional cloning of two R genes in these species (31, 32).

The first plant R gene to be cloned was the maize Hml gene (26). This gene, which

transposon. Hml encodes a reduced nico- tinamide adenine dinucleotide phosphate ( N A D P H ) - d e p e n d e n t HC-toxin reduc- tase, HTRC. HTRC inactivates HC-toxin, a pathogenicity factor produced by the fungus C. carbonum Nelson race 1 that permits the fungus to infect certain genotypes of maize (26, 33). The genetics of the interaction between maize and C. carbonum differ from those of gene-for-gene systems because tox- in-deficient C. carbonum strains lose their ability to cause disease in maize cultivars that do not carry Hml.

The first plant R gene to be cloned that conforms to a classic gene-for-gene relation was the tomato PTO gene (32). The PTO locus confers resistance to strains of P. sy- ringae pv. tomato (Pst) carrying the aviru- lence gene avrPto (34). A yeast artificial chromosome (YAC) clone that spanned the PTO region was identified with the use of a map-based cloning strategy and a re- striction fragment length polymorphism (RFLP) marker tightly linked to PTO. This YAC clone was then used to isolate com- plementary DNAs (cDNAs) corresponding to the PTO region, and subsequent genetic complementation tests identified a cDNA clone corresponding to PTO. The transla- tion product of PTO predicts that it en- codes a serine-threonine protein kinase; hence, this product may play a role in signal transduction.

Interestingly, PTO appears to be part of a complex locus that consists of a cluster of five to seven genes, all homologous to PTO. One of these PTO homologs, FEN, confers sensitivity to the organophosphorous insec- ticide fenthion. The evidence that FEN is a separate gene comes from mutational anal- yses (35) and from the demonstration that a cDNA clone with - 8 0 % homology to PTO confers sensitivity to the insecticide (36). A third gene, involved in both PTO resistance and fenthion sensitivity, was identified by the isolation of tomato mutants that were simultaneously altered in both bacterial re- sistance and fenthion sensitivity. These plants carry mutations in a new locus, des- ignated PRF (to indicate Pseudomonas resis- tance and fenthion sensitivity), which is tightly linked to PTO (35). Apparently, PRF is part of the signal transduction path- way that includes PTO and FEN. Because many tools are available to dissect the PTO signal transduction pathway, this system is a promising area for future research.

Four additional plant R genes that con- form to classical gene-for-gene relations have been cloned. The Arabidopsis RPS2 gene, which confers resistance to the bac- terial pathogen P. syringae pvs. tomato and maculicola expressing the avirulence gene

Rpt2. The RPS2 gene was then cloned by means of a map-based strategy similar in concept to the method used to identify the tomato PTO gene (3 I).

The tobacco N gene, which confers re- sistance to TMV, was isolated by transpo- son tagging with the autonomous maize transposon Ac (28). At elevated tempera- tures, N does not mediate HR after TMV infection, but if the temperature is lowered after TMV infection of seedlings carrying the N gene, the seedlings become necrotic and die. Some survivors contained Ac- tagged mutations at the N locus.

The tomato Cf-9 gene, which confers resistance to the fungal pathogen C. fulvum expressing the avirulence gene avr9, was tagged by a maize Ds transposable element (27). A tomato line lacking Cf-9 was engi- neered that expressed the C. fulvum avr9 gene under the control of a plant gene promoter (37). When this line was crossed with a line containing both Cf-9 and a Ds element, most of the progeny died because the interaction of the avr9 gene product with the Cf-9 gene product resulted in the elicitation of systemic HR. However, mu- tants carrying a Ds-inactivated tagged Cf-9 gene survived.

The flax L 6 gene, which confers resis- tance to the fungal pathogen M. lini, was also identified by transposon tagging with the maize transposon Ac. However, because no selection for L 6 mutations was available, mutants were identified by visual inspection of thousands of flax plants containing puta- tive transpositions of Ac into the L 6 gene (38). The phenotypes of TMV, M. lini, and P. syringae inoculated on both resistant and susceptible hosts are shown in Fig. 2.

Although the RPS2, N, Cf-9, and L 6 genes confer resistance to bacterial, viral, and fungal pathogens, DNA sequence anal- ysis revealed that all four genes encode proteins that contain leucine-rich repeats (LRRs). LRR motifs are found in many plant and animal proteins and are usually involved in protein-protein interactions (39). Moreover, all four of these genes are fundamentally different from both the maize Hml and the tomato PTO R genes. A comparison of the sequences of the RPS2, N, CF-9, and L 6 proteins reveals that RPS2, N, and L 6 share significant homology, whereas CF-9 appears to belong to a sepa- rate class (Fig. 3). RPS2, N, and L 6 all contain a conserved nucleotide binding site (NBS) in addition to the LRRs. Further inspection of these three proteins reveals that N and L 6 are more closely related to each other than to RPS2. The NHz-termi- nus of RPS2 contains a leucine zipper, which might be involved in protein dimer-

SCIENCE • VOL. 268 ° 5 MAY 1995 6 6 3

ization, whereas the NHz-terminus of N shares homology with the Toll protein of Drosophila and the mammalian interleukin-1 receptor (IL-1R). Preliminary mutagenesis

studies and sequence analysis indicate that the RPS2 and N proteins, which lack a leader peptide, are most likely cytoplasmic and probably recognize intracellular ligands,

RPS2 LN6 [MS Y t. R E V A T A V A I. L L P F I . L L N K F WR P N S K D S $ V N D O D D S T S E V D AFS Dig ~" N PISlG S F M D_E_LL._S__~_.LL.2__~.L0-1&.CL-~

" r 0 9 r~ 00 110

L6 PSVEYEIVIFII 81F [ ] GP~D ~EQ~F~O]F L ~ " ~ R FaY KII HI F R ~ D E ~ L KGKEIGPINL IL " A I L R DG[ RPS2 Y L ..~._.,..~,>'~MINIM A. .~.8. ._.G.z._. . . ._~. .J [~] ILL~. I [ i .~.~D.. .J~J~'d . D . . . [ ~ L ~ J L ~ . t . . . . J ~ . P . L ~ } . J L .

. . . . . I.~_9.___ 130 14(t 150 . . . . 1 6 0 N FKAll E E~QF AI V r I F l E NIY A'T}8 RIWC | INE'~-]~V K I IMItlC K TIRI F K O T V I Pill FIY DVDP S H V I ]~W I ~-~,"-~N L6 }RAIl O O [ ] K I Y V P I II ~SGJ~ t A D I S K ] W C t I M E m A E I I V ] R I R Q E I E ) D P R R I I L P I F V M V D P S D V 14 F~PS2 ' $ C R t A V v ' f T K..L_6JL.~.J~IV_ F R I ! ~ M ~ _ _ . _ _ _ ~ a l L , .

- , t L _ . 19I, ' C' 210 C ~"E~E IN E I I~I'W'E-~-~ V E ~ I O R W R i IA BI N E IA A N BK IG S C O N R D K IT D A IO C I{ I~cH~ Iv D

K K F R K F D G a T ! Q N W K D V G D K G W H I G K N D SA

930 24C, 25(. 260 770 _...280 N ~ s ~ L O ~- - , v G , ~ .TTT~T I~ -~ - ITT~TgT-~ I IU ' f - g 'TTN 'T6 -g "gF~- Wl~I~L~,..Iv I i K L6 ID~WSlHII S K ~ N L } L E T O E L V G NID DIH[I IT AIV[L E K~S L O SIE}N V T[MV~L Y ~ M ~ i I ' l A

RB0 3 o 3~0 ~..L-.~-_._--. . . . . -_.. .3.&;L--- . . . . N R A I IF~D}T L t. G R MD S}SIY Q F G A C F L K [ ~ K N K - " ~ N AlL L S [: [ I ~R E K A N Y L6 KA VlYINI K I ISJS)CF~C~CC~F I I D N I I R k l T Q E K D I [ , I V V V } [ I ~ ] K K I L V S E l L ] R I D S G S V G F RPS2 .~..~.~NN.IN_L_.~. L_.L._~_..g..JG J_H_.[~. ~_.~]_VJ.LJ_JW V I~ M S IR [ ~ F I~l ~_.~_! J ~ ] I ~ _ , . ~ _ L ~ . _P~ ~_~z Jk ,_LW_..___~._. _._ _- ~ _

p__ ~ 0 ,SO 37(~ 36¢ N L6 IN NID S G G R K TII IK E RlVlS R FIK i ~V VlVlDIE KIF KIF IE DIM t GIS P K~F I ISI° I S R [ I I I I ~ 1~116 IM

N D ~ 41 4' G 440 N HLI EK ( I Y E P D H E H A F G N SL E

G 0 ~ - - - - 7, 4E...__.__.. - ,4 % 5 - N L6 ITIKIV~ ~ ISLt. FIKQEI AV~EDIIIL~QII. IRIRIT L N }t DIEIVY}DR~I IA~N! PE RPS2 I & B I L J L L J ~ I C J L , t A_H_JB_L.T_.L_~E..~LJ:LL&t&L~IY_IL]Z.IBLe_,/L£_.~_~_~.IM_.2~.LJ.Y_E_JS_LIBI~I L I . N . B I ~ . J . L~_ .

510 .._55~0 .5~C, ,40 550 560

L6 A OI F I GO N K E E P Y Y M W Y P A S N I F L O O D D E F K M

579 580 - .$90 OO 610. -. LN6 ~D~D ~I~DD~Ii I ~ H O [ ' I Q D M KYI VNIF Q D K DIPIG~E~S- 'R '~"~ 'LI~IKI~IV E E V M S N N T I ~ I T I M A M E A I WV S S " ~ ' T ~

N D Q L R D M R E I V R R E M V L F W K ~ R S R I I ~ $ 1 ~ I ~ I G I D L L L NK KI,,~ISIS K V K~Ai IS I IP] ~WGJV~K1Y RPS2 A C E T Q V K H N V V ,,~...._.,.A L~.]~j.~J~lq ~.I_~..J_~,JJL.~_LV_~J.~.JL~JLM.__G~JU.L'~#_.

_ _ 6~r0 ~30 640 6 ~ .660

RPS2 N W R Q A L V I S~L_JL P N R L_J,.~.T_BIP_.j,L,_J~II C P K L I T L M t Q N S S . . . . . £._.~..E.]

~ ~ l T W S K R L T R ~ P O F ' T G M ~6 S L R N L W ~ E T K S L RI D L D D W G G W R H M M K M A E ~ K P g V V L A S N Y S L Y G R R V R L S D C W R

RP~q2 P - . LL.~Sj

N N L E Y V N L Y QCIS N LIE E V H HIS t [GC O SIK V~I G L ~ Y I L I N D C K S L K R F P C V N V E S t E Y L G L6 K S t E V ~ A I ~ E M D E V I D I I G E L K J K L } K T ~ I V l L J K F C P I Q K { S G G T F G M L K G L R E L C L E RPS2 - - ~.-_S_LJZ-_LL..P~-LJ_& J_-L&Z- L._V.JLLJ~__ELBI~JLivlJ . . . . . . . . . . . .

790 B00 _ _ _ _ _ _ .._..B_.~.. B10 ~ 30 B40 N L R SICD)SILIE K [ P E { Y R M K P E $ Oil ~HIMQIGl'~q F~ ~ ~J~ ~ ~ F ~YKITINV'T~'~"~-T~TrWWTN"~" L6 F N W G I I N i L I R ElY V(A hi ] [.~O~ IS S L K V}t. IKI'1TIGIAIK E V E I N[E F1P L GL K E L S I ~ S R I P N ~ Q t ~ . S S R I P N L Q t L RPS2 $ G I / K I S VIL PIQE[L IL]N LIR K I~L...LD.JLL.3a_LR_LIJ.__ 1 I.. IQL_I_.LR&.D_.&JL.Ed_W!J.&

e,5o 86o 87o ~ 899,9

r0 h' I ? D EV K V D O K ~) O K L~ L' V S'L S V ~ G C S F D MP ~ A S P S E K'L'EO E S S V WWK V '$ K P ~ K I S " Q I L 'E K T R I ~'L 'P E E I' 'G O " D ' N " R I V 'F O A S D TN V N V V O D A L ' IL '~ e P S RPS2 EV N L Y Y S Y A G W E L Q S F G E D E A E E L IG.J.E~A DIL IE y ~.~JN..J~T h IL G I T V L S L E T IlJ_..._L.LL_

'I 9iCL N K t i i" C"~"~kG F K b G V' H'F E~2~ I P V A pE 9*0 949 950 GL HSI~ IE Y B I N L S Y C N L $ D G G L P E E I GS S I I R S S G G H L P R Y L L P T S L T Y L K I I I Y Q C ~ E ~ I T W L GI E N ~ I E N ~ I 3 3 S L E V N D I F Q T L G G O L D G

E~S2 ~._E. G A L HKJ.~:LJ._~J.~:LLHV E E C N E L L Y F N L I ~ i S L T N H £ RN R e~ K S C H O L E Y L V T P A O F E

L O G R S E l L I R K V N G L A R I K G L K D L L C S S T C K L R K F Y I T E C P D L ~ L O L K ° 'T

RP$2 ~.._D P $ E v L T L H.~_~ H N L T RV W G N S V S D

~ C l NI ; -NF Y L V T H R K K L N R V K L D D A H N O T M Y N L F A Y T M F N= Q L P E L P P E L L HV ° IC I IH M~A L K F[I IN Q T V V V S M A ~ L T I R D C I P I R L } E V G P I M } I R S L P F P M L K K L D L A V A N I T K E E D L D A I G S L

RPS2 H,CN K I. K N v...._~.lW V Q K L P ~1

1970 1080 1090_ O P Y P !~o0 11D. ~2o

L6 E E L ~ / ~ L H D O T L O K L T T L V V K J E L K R1 V S S S K V P S LI / ~ L Q D L Y L S S G I E E L L E R ~ ; E G L E RPS2 , ~ ~ V ! ~ ~IF D C R ~ ! EIE L I ~�IE .~.._E

H ZZ_.___.___,._ 113.0 1140 1.S 0 600 N " A V C Y S_!_JSIL I O T "T A H L , - ~ C D b K M S ~ I M T Q K . A L ~ P L6 E G C T S L L P E N W ~ ' I K P D K F L G F G G C P DL "T E L V Q t V V A V ~ S L R G L T I ~ I D C P R I l E V G MI G R L P L E L K E L D I RPS2 $ P S V E D J_E._T.J.L.~. .~,_T_JL~ T I~ID L P E I I IN S I L P __

1 20 1230 N C D T E $ $ N Y $ ~ P ~ P ~ D S L E L T L T L S M V N I T K E D E L E V L G S L E L6

1~4~.L_ .50 L I~IK [~1G P E ~ 60 ~'~"2Z~------~'-"----~ L ~2_80 N G E E K G P E V N , ~ L 70 S N E P T E E L6 Q K L T T L ' V E V P S L R E ' E ~ E E L A ~ L M [ ~ H R R I t L I L I ~ G C T S L G S L K NL N V L b RPS2 . . . . . . ~ R R T I Q I M ~ N I I I ~ ._._F . ~ . J ~ . . ~ / [ K ~ _ I K DJDJLN_J.LE_.

S °GI R R T Q Y N N R T S F Y E L NG LN6 I GO ~L P P R A R I H y K S L S L S A L K T T T W P O Q P RPS2 Y k E FL F v

Fig. 3. Amino acid alignment of the proteins expressed by the R genes N, L ~, and RPS2. The sequences for N (28), L 6 (29), and RPS2 (31) were aligned with the use of the programs AMAS (62) and Clustal W (63). Identical matches are shown in blue, and similarities are shown in green.

664 SCIENCE • VOL. Z68 • 5 MAY 1995

whereas the L 6 protein may attach to the cell membrane by means of a signal anchor.

The CF-9 protein appears to consist pri- marily of extracytoplasmic LRRs, with a COOH-terminal membrane anchor. This structure suggests that the CF-9 protein is a receptor for the extracellular ligand provid- ed by the Avr9 elicitor peptide. Whether a direct interaction occurs between CF-9 and the Avr9 peptide is unknown. Membrane protein preparations from leaves of plants that express Cf-O and Cf-9 bound the Avr9 peptide with almost equal affinity, whereas the intact leaves of the Cf-O-expressing plants did not respond to the Avr9 peptide. Thus, although CF-9 protein binds the Avr9 peptide, other plant proteins (perhaps expressed by other members of the Cf-9 multigene family) also bind the Avr9 pep- tide (40).

Additional R genes will likely be isolated by positional cloning and transposon tag- ging. Transposon tagging may be a more general method for R gene isolation from a wide variety of plant species. Because four of the six R genes cloned to date have been found to encode products with strikingly similar sequences and structural features, it is likely that other R genes will be isolated on the basis of homology to the known R genes. Indeed, homologs of RPS2, N, L 6, and Cf-9 exist in a variety of species. The L 6 gene hybridizes to RFLPs linked to the un- linked rust resistance genes at the flax M resistance locus (38).

S i g n a l T r a n s d u c t i o n E v e n t s a n d E x p r e s s i o n o f D i s e a s e R e s i s t a n c e

The mechanisms underlying gene-for-gene resistance probably involve specific recog- nition of a pathogen-generated ligand (pro- duced by an avr gene) by a plant receptor encoded by an R gene (Fig. 4). The events that occur after recognition are a matter of speculation, but the domains in R gene proteins provide clues. For example, if CF-9 is a transmembrane receptor and its LRR region binds the Avr9 peptide directly, the cytoplasmic domain of CF-9 might directly activate a kinase such as that encoded by the tomato R gene PTO. This event would be analogous to the mechanism by which CD4,_a membrane-anchored receptor on T cells, activates the tyrosine protein kinase p56 Lck (41). Alternatively, CF-9 might in- teract with other proteins, including trans- membrane protein kinases that also carry extracellular LRRs (42) or secreted LRR- carrying proteins such as polygalacturonase- inhibiting proteins (PGIPs) (43). A genetic approach holds promise for the identifica- tion of genes required for Cf-9---dependent resistance (44).

The plant cellular defense responses ac- tivated by the N protein may be analogous

~ e

.'ne ag- ro-

t o r

n t s

• of me F-9 RR the :tly by

uld ich ~T ase in- .ns-

~rry (R- ise- ~tic ,ca- ent

ac- OUS i

~ 7 ~ : ' ~ ~ - ~ 7 r 7 7 ~ - r - : : : - : ; _,__ : _ ....:...a_~27Z:2ZT..LT._.::: _ __ _:._..,._._. ...... LZ

to "natural" or innate immunity in verte- brates and insects. In mammals, perception of signals produced by pathogens results in Iranslocation of the Rel-related transcription tactor NF-KB (45) from the cytoplasm to the nucleus, where it activates transcription of defense-related genes (46); in Drosophila the same process occurs with the Rel-related transcription factor Dif (47, 48). In the mammalian immune system, the cytoplasmic domain of IL-1R is involved in the transduc- rion of the signal required for the transloca- rion of NF-KB; this domain has sequence and functional similarity to the cytoplasmic do- main of the Drosophila Toll protein (49). In Drosophila development, the perception of an extracellular signal by Toll results in the translocation of Dorsal, a homolog of NF-KB (50). The presence of a domain in the NH 2- terminus of the N protein that is similar to the cytoplasmic domains of Toll and IL-1R suggests that this domain may trigger an intracellular signal transduction cascade in plants, analogous to the Toll and IL-1R pathways in animals. The N protein, and possibly other R gene-encoded proteins, may serve as receptors that activate a Rel- related transcription factor that induces the expression of genes responsible for HR. Un- like that of the N protein, the NH2-terminal domain of the Arabidopsis RPS2 protein is not similar to those of Toll or IL-1R. How- ever, RPS2 does contain a leucine zipper motif at the NH2-terminus that may be in- volved in the formation of a heterodimer with a Toll-like protein.

The chain of events between pathogen infection in a plant and the onset of HR is not well defined. However, it has been widely observed that HR is preceded by a rapid outburst of the reactive oxygen inter- mediates (ROIs) O2-, H202, and OH" (5I). A plasma membrane muhisubunit NADPH oxidase complex, similar to the one found in mammalian phagocytes, might be in- volved in the release of ROIs in plants. If a rapid oxidative burst is crucial to HR, the activation of a protein kinase could lead to the activation of an NADPH oxidase rather than to transcriptional activation. In the mammalian innate immune response, ROIs have been shown to induce the expression of acute phase response genes by activating the transcription factors NF-KB (52) and AP-1 (53). In plants, R gene-mediated in- duction of intracellular ROIs suggests that a redox-regu|ated transcription factor may also be involved in the activation of HR.

Evolution of Plant Disease Resistance

The following scenario for the evolution of plant disease resistance has been proposed (54): The evolutionary ground state is con- sidered to be a compatible interaction in

which a pathogen has evolved to be viru- lent on a particular host plant. Selection favors the evolution and spread of host individuals that specifically recognize the pathogen and resist infection. For example, a receptor that evolved to activate defense responses to pathogens in general may be modified so that it specifically recognizes a particular pathogen product (an avirulence gene product). The pathogen responds by losing the avirulence gene by mutation. This phenomenon is absolutely essential for the survival of obligate parasites. The host is now susceptible, and again selection is brought to bear on new host R gene speci- ficities. Consequently, the evolution of gene-for-gene interactions can be seen as a continuing step-by-step or move-counter- move process, whose consequence in plant populations is a diversity of R genes in different individuals of a host species and a corresponding diversity of avirulence genes in different pathogen races.

The existing diversity of R genes is the product of an evolutionary process that ap- pears to have proceeded along two major branches. On one branch, exemplified by the M rust resistance locus in flax, tandem arrays of related R genes with different spec- ificities are found in the plant genome (38). The other evolutionary branch is exempli- fied by the flax L rust resistance locus; the specificities at this locus behave genetically as alleles of a single gene, and different specificities existing in heterozygotes can- not be recombined (38). The cloning of the L 6 allele of this locus supports the classical genetic interpretation of a simple L locus but has also provided some surprises (29). The genes at the genetically complex M

Fig. 4. Receptor-ligand model for the recognition and expression of plant disease resistance. In this hypothetical model, the R gene is thought to encode for either an ex- tracellular receptor, such as the protein product of the Cf-9 gene of tomato, or an intracellular recep- tor, such as the product of the N gene of tobac- co. The ligand in this model may represent the direct or indirect product of the pathogen's aviru- lence gene. The specific recognition event trig- gers a signal transduc- tion cascade that may involve protein kinases and may lead to the ex- pression of plant disease resistance.

r > f - / " ~- , F , f : f

Intracellular receptor?

Ligand

nce m the un- ~ locus are related in seque linked L gene, with 70 to ~ , ~ - - t ,~h~ , , , ~'90%rnucle°tide'~ identity. The M locus appears to have evolved by local duplication and divergence from an L-like R gene progenitor, whereas the L locus appears to have evolved as a multiple allelic series, with only a single L specificity capable of existing in a homozy- gore. The contrasting evolution of two such closely related genes may be the result of a rare duplication event that occurred only at the M locus and then provided the oppor- tunity for rapid amplification by unequal crossing over. Similarly, molecular analyses of the TMV resistance locus N in tobacco and of the Cf-9 locus in tomato have re- vealed a clustered gene family (27, 28).

Molecular Basis and Evolution of R Gene Specificity

R genes specifically distinguish isolates of a single pathogen species. The multiple resis- tance specificities encoded by the 13 alleles of the cloned L gene of flax provide an opportunity to study the molecular basis of this specificity. Three alleles, L 2, L 6, and L 1°, have been cloned and partially charac- terized. Although t 6 and L 1° are similar, L 2 has additional numbers of an LRR motif that occurs in the COOH-terminal region of the gene product. This region may deter- mine ligand specificity.

Pathogen propagules increase to vast numbers in comparison with their hosts, and consequently there is a greater oppor- tunity for virulent pathogen races to arise than for corresponding new host resistance specificities. Coupled with the differences in population size, pathogen evolution from

Extracellular receptor?

I Kinase

't Resistance

and Defense

SCIENCE • VOL. 268 ° 5 MAY 1995 6 6 5

avirulence to virulence usually results from loss-of-function mutation. The correspond- ing gain of function (that is, resistance in the host) is unlikely to occur by simple mutation. Do plants have a mechanism to generate new R gene specificities and keep pace with the evolutionary progress of pathogens? Pryor, Hulbert, Bennetzen, and their colleagues have observed high-fre- quency loss of rust resistance in corn asso- ciated with unequal crossing over at the Rpl locus, and they have proposed that this shuffling of preexisting coding information may provide a means for plants to generate new specificities (55). The LRR domains, which are important in receptor selectivity, may be involved in defining the recogni- tional specificity of the pathogen (56).

Bioengineering for Novel and Stable Plant Disease Resistance

The isolation of plant R genes provides opportunities for producing crop plant va- rieties with increased disease resistance. Po- tential approaches can be subdivided into those that augment classical breeding tech- niques and those that involve direct engi- neering of crop plants.

The identification of a variety of R genes on the basis of amino acid sequence conser- vation will enable plant breeders to monitor R gene segregation using appropriate DNA probes instead of testing progeny for disease resistance and susceptibility. The same ap- proach may greatly facilitate the identifica- tion and introgression of new resistances from wild species that either interbreed poorly with crop species or do not cross at all. For example, additional nematode resis- tances exist in wild accessions of Lycopersi- con peruvianum, but the amount of work necessary for the recurrent introgression of many such new resistances is daunting. Candidate nematode R genes could be iden- tified by homology, isolated by molecular cloning, and transformed into crop varieties to evaluate their effectiveness.

It is frequently the case that after pro- tracted breeding efforts of 10 or more years' duration, a new resistant plant variety is produced only to be overcome by a new pathogen race within a few years of deploy- ment. This is immensely wasteful of time and effort. Population genetic theory pre- dicts that the breakdown of resistance will happen more slowly in varietal mixtures carrying an array of different R genes (57). However, such mixtures have not been adopted in practice because they exhibit variation in multiple characteristics, includ- ing time to harvest. With several different cloned R genes responsive to the same pathogen, plant varieties could be produced consisting of mixtures of lines that differ only in the R gene allele they carry. Such an

environmentally and ecologically sound ap- proach to disease control would find accep- tance among consumers and growers.

One exception to the lack of durability of R gene-mediated resistance is the BS2 gene of pepper, which confers resistance to strains of Xanthomonas campestris pv. vesicatoria that contain the avrBs2 avirulence gene and is extremely effective in controlling the bacte- rial spot disease of pepper. The success of this gene seems to be related to the fact that all strains of the pathogen examined to date contain an active copy of the avrBs2 gene, and if the gene is lost, the pathogen suffers a severe fitness penalty (58).

The approaches described above to facil- itate the introduction of disease-resistant varieties all take advantage of cloned R genes. However, R genes appear to function at or near the beginning of a complex signal transduction cascade that leads to HR and ultimately to SAR. It would be desirable to directly manipulate HR and SAR by engi- neering the signal transduction pathways that lead to their activation. Genetic dis- section of HR and SAR and their regula- tion is beginning in Arabidopsis.

The engineering of HR and SAR cannot take place unless certain problems are ad- dressed, such as the lethality of a constitu- tively activated defense response (37). To circumvent this problem, alleles of R genes, or of genes that encode the products with which the products of R genes interact, could be found that would partially activate the defense response. The result would be a phenotype analogous to SAR that confers some degree of resistance but does not kill the plant. Mutations of this sort are likely to be selected against in natural populations because they would likely partially cripple the host in the absence of severe pathogen attack. In agricultural settings, however, they could be advantageous, even though they might be associated with yield penal- ties. Dominant mutations at R gene loci and recessive mutations at some other loci might be expected to result in partial con- stitutive expression of the defense response. Some necrotic or disease lesion mimic mu- tations may arise in this manner (59). ln- deed, the phenotype of the recessive barley mutant mlo, which has been widely used in barley breeding, can be phenocopied by application of 2,6-dichloroisonicotinic acid, a chemical that elicits a SAR-like response (60). Thus, in some circumstances, useful mutations can be identified in which the defense response has been "primed."

It has been suggested that if a suitable pathogen-inducible promoter, such as the prpl-1 promoter of potato (61), could be found, it would be possible to induce the expression of a race-specific elicitor such as the Avr9 peptide only in cells that are being challenged by a compatible pathogen

(20). If this engineered plant also contained a functional Cf-9 gene, then a previously compatible pathogen would now elicit HR. Because potato and tomato are so closely related, Cf-9 seems likely to function in potato, and this system offers real potential for increasing resistance to potato late blight caused by Phytophthora infestans.

In the course of transposon tagging of the tomato Cf-9 gene, alleles have been generated in which the Ds transposon so- matically excises from Cf-9 and thus re- stores function. In the presence of Avr9, this excision results in the formation of localized necrotic sectors in which both Avr9 and CF-9 are active (27). Preliminary experiments indicate that plants with this phenotype show some characteristics of SAR, including enhanced resistance to pathogens that would otherwise be compat- ible. This phenomenon has been designated genetic acquired resistance (GAR) to indi- cate that it is a genetically imposed SAR. Fine tuning of the system will undoubtedly be required to achieve the optimum balance between activation of the defense response and crop yield.

Summary

This is an extremely exciting time for the field of plant pathology. The cloning and characterization of several plant R genes constitutes a major breakthrough in the elucidation of the molecular basis of disease resistance to a wide range of phytopatho- gens. As a result, we are finally in a position to determine the molecular basis of plant- pathogen specificity and expression of dis- ease resistance. Future research challenges include the determination of the mecha- nisms by which R gene products recognize pathogen elicitors and the plant defense response blocks pathogen growth. The basic knowledge obtained from this research will undoubtedly help to produce novel forms of durable disease resistance and will lead to a decline in the use of environmentally dam- aging pesticides.

REFERENCES AND NOTES

1. V. Walbot, Trends Genet. 1,165 (1985). 2. C.J. Lamb, M. A. Lawton, M. Dron, R. A. Dixon, Cell

56, 2:15 (1989). 3. R. H. Biffen, J. Agric. Sci. 1,4 (1905). 4. J. R. Harlan, Science 174,468 (1971), 5. E. E. Leppik, Annu. Rev. Phytopathol. 8, 323 (1970). 6. C. J. Lamb, Ce1176, 419 (1994). 7. G. Agrios, Plant Pathology (Academic Press, San

Diego, CA, 1988). 8. N. T, Keen, A. Bent, B. J. Staskawicz, in Biotechnol-

ogy in Plant Disease Control, I. Chet, Ed. (Wiley, New York, 1993), pp. 65-88.

9. H. H. Flor, Annu. Rev. Phytopathol. 9, 275 (1971). 10. J. L. Dangl, in Bacterial Pathogenesis of Plants and

Animals: Molecular and Cellular Mechanisms, J. L. Dangl, Ed. (Springer-Verlag, Heidelberg, Germany, 1994), vol. 192, pp. 99-118.

11. N.T. Keen, Annu. Rev. Genet. 24., 447 (1990); D. W.

666 SCIENCE ° VOL. 268 ° 5 MAY 1995

ed sly . t .

:ly in ial {te

of :i : en

3o-

re -

r9, of

)th :, ary his of tO

at- ,:ed di- ,R. ! :lly Ice i ~se

:he md ties

the ase i ~o-

ion nt- tis- ges ha- dze [ise

isic rill s of o a

Gabriel and B. Rolfe, Annu. Rev. Phytopathol. 28, Biol. Chem. 267, 2605 (1992); S. Leung and N.G. G.S. Johal, S. H. Hulbert, S. P. 365 (1990). Miyamoto, Nucleic Acids Res. 17, 1177 (1989). press.

12. Z. Klement, in Phytopathogenic Procaryotes, M.S. 51. M. C. Mehdy, Plant Physiol. 105, 467 (1994); A. 60. M. Wolter, K. Hollricher, F. Salaomlli~ Pro. Schulze-'l ~' Mount and G. H. Lacy, Eds. (Academic Press, New Levine, R. Tenhaken, R. Dixon, C. J. Lamb, Cell 79, Lefert, Mol. Gen. Genet. 239, 122 (1993). York, 1982), vol. 2, pp. 149-177.

13. J. Ryals, S. Uknes, E. Ward, Plant PhysioL 104, 1109 (1994).

14. T. Gaffney et aL, Science 261,754 (1993). 15. T. P. Delaney et aL, ibid. 286, 1247 (1994). 16. B. Vemooij et al., Plant Cell 6,959 (1994); Z. Chen, J.

W. Ricigliano, D. F. Klessig, Proc. Natl. Acad. Sci. U.S,A. 90, 9533 (1993).

17. J. Glazebrook and F. Ausubel, Proc. Natl. Acad. Sci. U.S.A. 91,8955 (1994).

18. H. Cao, S. A. Bowling, A. S. Gordon, X. Dong, Plant Cell 6, 1583 (1994).

19. J. L. DangI, Adv. PlantPathol. 10, 127 (1993). 20. P. J. G. M. de~lit, Annu. Rev. Phytopathol. 30, 391

(1992). 21 . . . . Adv. Plant PathoL, in press. 22. M. H. A. J. Joosten, T. J. Cozijnsen, P. J. G. M. de

Wit, Nature 387, 384 (1994); G. F. J. M. van der Ackerveken, J. A. L. van Kan, P. J. G. M. de Wit, PlantJ. 2, 359 (1992).

23. S. L. Midland et al., J. Org. Chem. 58, 2940 (1993); N. T. Keen and R. I. Buzzell, Theor. Appl. Genet. 81, 133 (1991).

24. J. N. Culver and W. O. Dawson, MoL Plant-Microbe Interact. 4,458 (1991).

25. B. Baker, J. Schell, H. Lorz, N. Fedorofl, Proc. Natl. Acad. ScL U.S.A. 83, 4844 (1986); J. D. G. Jones, F. M. Carland, P. Malign, H. K. Dooner, Science 244, 204 (1989).

26. G. S. Johal and S. P. Briggs, Science 258, 985 (1992).

27. D. A. Jones, C. M. Thomas, K. E. Hammond-Ko- sack, P. J. Balint-Kurti, J. D. G. Jones, ibid. 268, 789 (1994).

28. S. Whitham et aL, Cell 78, 1101 (1994). 29. G. J. Lawrence, E. J. Finnegan, M. A. Ayliffe, J. G.

Ellis, Plant Cell, in press. 30. E. M. Meyerowitz, Cell 56, 263 (1989). 31. A. F. Bent etaL, Science 265, 1856 (1994); M. Min-

drinos, F. Katagifi, G.-L. Yu, F. M. Ausubel, Cell 78, 1089 (1994).

32. G. B. Martin et al., Science 282, 1432 (1993). 33. R. B. Meeley, G. S. Johal, S. P. Briggs, J. D. Walton,

Plant Cell4, 71 (1992). 34. P. C. Ronald, J. M. Satmeron, F. M. Carland, B. J.

Staskawicz, J. BacterioL 174, 1604 (1992). 35. J. M. Salmeron, S. J. Barker, F. M. Carland, A. Y.

Mehta, B. J. Staskawicz, Plant Cell 6, 511 (1994). 36. G. Martin et al., ibid., p. 1543; C. Rommens, J. Sal-

meron, D. Baulcombe, B. Staskawicz, ibid., in press. 37. K. E. Hammond-Kosack, K. Harrison, J. D. G. Jones,

Proc. Natl. Acad. Sci. U.S.A. 91, 10445 (1994). 38. J. G. Ellis, G. J. Lawrence, E. J. Finnegan, P. A.

Anderson, ibid., in press. 39. B. Kobe and J. Deisenhofer, Trends Biochem. Sci.

19, 416 (1994). 40. G. HoneeetaL, inAdvancesinMolecularGeneticsof

Plant-Microbe Interactions, M. Daniels, A. Downie, A. Osbourn, Eds. (Kluwer Academic, Dordrecht, Netherlands, 1994), vol. 3, pp. 199-206.

41. B. Tourvieille, S. D. Gorman, E. H. Field, T. Hunka- piller, J. R. Parnes, Science 234, 610 (1986); S. L. Weinstein, C. H. June, A. L. DeFranco, J. Immunol. 151,3829 (1993).

42. C. Chang, E. Schaller, S. E. Patterson, S. F. Kwok, E. Meyerowitz, Plant Cell 4, 1263 (1992); J. C. Walker, Plant& 3, 451 (1993).

43. G. De Lorenzo et aL, Biochem. "Soc. Trans. 22, 394 (1994).

44. K. Hammond-Kosack, D. A. Jones, J. Jones, Plant Cell 6, 361 (1994).

45. S. Ghosh and D. Baltimore, Nature 344, 678 (1990). 46. P. A. Baeuede, Biochim. Biophys. Acta 1072, 63

(1991). 47. Y. T. IpetaL, Ce1175, 753 (1993). 48. D. Hultmark, Trends Genet. 9, 178 (1993). 49. C. Hashimoto, K. Hudson, K. V. Anderson, Cell 52,

269 (1988). 50. D. S. Schneider, K. L. Hudson, T. Y. Lin, K. V. Ander-

son, Genes Dev. 5, 797 (1991); D. St. Johnston and C. Nusslein-Volhard, Cell 68, 201 (1992); A. Heguy, C. T. Baldafi, G. Macchia, J. L. Telforcl, M. Mefli, J.

583 (1994). 52. R. Schreck and P. A. Baeuefle, Trends Cell BioL 1,

39 (1991). 53. M. Meyer, R. Schreck, P. A. Baeuerle, EMBOJ. 12,

2005 (1993); S. Xanthoudakis and T. Curran, ibid. 11,653 (1992).

54. A. Pryor, Trends Genet. 3, 157 (1987). 55. J. L. Bennetzen and S. H. Hulbert, Plant MoL Biol.

20, 575 (1992); T. P. Pryor and J. Ellis, Adv. Plant PathoL 10, 281 (1993).

56. T. Braun, P. R. Schofield, R. Sprengel, EMBOJ. 10, 1885 (1991).

57. M. Wolfe, Annu. Rev. PhytopathoL 23, 251 (1985). 58. B. Keamey and B. J. Staskawicz, Nature 346, 385

(1990). 59. J.T. Greenberg, A. Guo, D. F. Klessig, F. M. Ausubel,

Cell 77,651 (1994); R. A. Dietrich et aL, ibid., p. 565;

61. N. Martini, M. Egen, I. ROntz, G. Strittmatter, ibid. 236, 179 (1993).

62. C. D. Livingstone and G. J. Barton, Comput. AppL Biosci. 9, 745 (1993).

63. J. D. Thompson, D. G. Higgins, T. J. Gibson, Nucleic Acids Res. 22, 4673 (1994).

64. We thank S. P. Dinesh-Kumar, S. Whitham, M. Whalen, J. Glazebrook, F. Katagiri, and K. Ham- mond-Kosack for their helpful comments and critical reading of the manuscript, and B. Osborne for help in the preparation of the figure with the amino acid alignments. Supported by grants from NSF and the U.S. Department of Energy (B.J.S.), NIH (F.M.A.), USDA (B.J.B.), CSlRO (J.G.E.), and the Gatsby Charitable Foundation, UK Biotechnology and Bio- logical Sciences Research Council (BBSRC), and EEC Bridge Program (J.D.G.J.).

The Ethylene Signal Transduction Pathway in Plants

Joseph R. Ecker

E t h y l e n e (02H4) , t he c h e m i c a l l y s i m p l e s t p l an t h o r m o n e , is a m o n g t he b e s t - c h a r a c t e r i z e d p lan t g r o w t h r e g u l a t o r s . It p a r t i c i p a t e s in a va r i e t y o f s t r e s s r e s p o n s e s a n d d e v e l o p m e n t a l p r o c e s s e s . G e n e t i c s t u d i e s in A r a b i d o p s i s h a v e d e f i n e d a n u m b e r o f g e n e s in t he ethylene s igna l t r a n s d u c t i o n p a t h w a y . I so la t i on o f t w o o f these genes has r e v e a l e d t h a t p l a n t s s e n s e th i s g a s t h r o u g h a c o m b i n a t i o n o f p r o t e i n s t h a t r e s e m b l e b o t h p r o k a r y o t i c a n d e u k a r y o t i c s i g n a l i n g p r o t e i n s . E t h y l e n e s i gna l i ng c o m p o n e n t s a re l i ke ly c o n s e r v e d f o r r e s p o n s e s a s d i v e r s e as cel l e l o n g a t i o n , cel l f a t e p a t t e r n i n g in the root e p i d e r m i s , a n d f ru i t r i pen ing . G e n e t i c m a n i p u l a t i o n o f t h e s e g e n e s wi l l p r o v i d e a g r i c u l t u r e w i th n e w tools to p r e v e n t o r m o d i f y e t h y l e n e r e s p o n s e s in a v a r i e t y o f p lan ts .

The simple gas ethylene is an endogenous regulator of developmental adaptations in higher plants (1). Exposure to ethylene can produce a myriad of effects on plant growth, development, and physiology, most notably the ripening of fruits, inhibition of stem and root elongation, promotion of seed germi- nation and flowering, senescence of leaves and flowers, and sex determination. How this simple olefin evokes such a diverse array of physiological processes has been a central question in ethylene research.

The biosynthesis of ethylene is stimulat- ed prior to several developmentally pro- grammed senescence processes and in re- sponse to environmental insults such as me- chanical trauma and pathogen infection (2, 3). As a result of biochemical analysis, the route of ethylene synthesis (the Yang Cy- cle) is now largely understood (4, 5). The rate-limiting step is the conversion of S- adenosyl-L-methionine (SAM) to 1-amino- cyclopropane-l-carboxylic acid (ACC), which is catalyzed by ACC synthase. The enzyme ACC oxidase converts ACC to

The author is in the Plant Science Institute, Department of Biology, University of Pennsylvania, Philadelphia, PA 19104-6018, USA.

ethylene, carbon dioxide, and cyanide. ACC oxidase is constitutively present in most tissues, but its synthesis is increased during fruit ripening in tomato. The genes that encode ACC synthase and ACC oxi- dase have been cloned and characterized from many plant species (5, 6). ACC syn- thase is encoded by multigene families in all species examined, and individual gene fam- ily members are transcriptionally activated by a variety of inducers. Environmental stresses (physical, chemical, and biological) and hormonal signals, such as auxin, cyto- kininin, and even ethylene itself, stimulate synthesis of the ACC synthase enzyme, thereby providing a means for autoregula- tion of its production. Although tremen- dous progress has been made since 1989, questions still remain regarding the com- plex regulation of ethylene biosynthetic genes. However, it is clear that genetic manipulation of the ACC synthase and ACC oxidase genes by expression of anti- sense RNA (7) will provide a simple means to control the ripening of fruits in a variety of plants (4, 8).

By contrast, biochemical approaches to- ward dissection of the mechanisms by

SCIENCE • VOL. 268 ° 5 MAY 1995 667