Molecular aspects of avirulence and
pathogenicity of thetomato pathogen
Cladosporiumfulvum
0000 0547 621<
Promotor:
dr. ir. PJ.G.M. deWit
Hoogleraar in deFytopathologie,
in hetbijzonder defysiologische aspecten
Guido F.J.M, van den Ackerveken
Molecular aspects of avirulence and
pathogenicity of thetomato pathogen
Cladosporium fitlvum
Proefschrift
ter verkrijging van degraad van doctor
in delandbouw- en milieuwetenschappen
op gezag van derector magnificus,
dr. C.M. Karssen,
in het openbaar teverdedigen
opvrijdag 12november 1993
des namiddags tevier uur in deAula
van deLandbouwuniversiteit te Wageningen.
v
kof^e
Aanmijn ouders
VoorMirjam
BSHPBOUWUN1VERSÛSMI
coverdesign: Frederik von Planta
CIP-GEGEVENSKONINKLIJKE BIBLIOTHEEK, DENHAAG
Ackerveken, Guido FJ.M. van den
Molecular aspects of avirulence andpathogenicity of the
tomatopathogen Cladosporium fulvum / Guido FJ.M. van
den Ackerveken. - [S.l. :s.n.]
Thesis Wageningen. - Withref. - With summary in Dutch.
ISBN 90-5485-148-1
Subject headings: avirulence genes ; Cladosporium fulvum
/ pathogenicity genes ; Cladosporium fulvum.
AjO0??ö
Stellingen
1.
Fysio-specifieke resistentie in tomaat tegen Cladosporiwnfulvum wordt geïnduceerd
door één specifiek peptide van het pathogeen.
Dit proefschrift
Zolang Vanderplank specifieke herkenning van pathogenen in resistente planten niet
erkent, kan het hem niet kwalijk worden genomen dathij de "avirulentie versie" van
de gen-om-gen hypothese absurd vindt.
Vanderplank (1991) Plant Pathology 40, 1-3
3.
Hetisvoorbarig deverschillen inafweerreakties vanplanten tegen pathogenen toete
schrijven aan verschillen in resistentiegenen wanneer de vorming van de betrokken
elicitoren niet nader onderzocht is.
Hammond-Kosack etal. (1993)
In: Advances inMolecular Genetics of Plant-Microbe Interactions, Vol 2 :457-461
4.
De stelling datgen-om-gen interacties tussen plant en pathogeen enkel zijn ontstaan
door de veredeling van cultuurgewassen middels stapsgewijze introductie van
resistentiegenen is ongegrond, daar in natuurlijke plant-pathogeen populaties wel
degelijk sprake is van gen-om-gen interacties.
Barrett (1985) In: Ecology and Genetics of Host-Parasite Interactions: 215-225
Crute etal. (1993) In: Advances inMolecular Genetics of Plant-Microbe Interactions, Vol 2:437-444
Het aantal plant-pathogeen interacties waarin fysio-specifieke elicitoren zijn
aangetoond iswaarschijnlijk zogeringomdatdezeelicitorennietabundantof (water-)
oplosbaar zijn.
Invertase is voor Cladosporiwnfulvum niet alleen essentieel voor de benutting van
sucrose, maar is mogelijk ook een pathogeniteitsfactor die floeembelading remt.
Dickinson etal. (1991) Plant Physiology 95,420-425
De specificiteit van verwante bacteriële avirulentiegenen uit de avrBs3 familie voor
zowel dicotyle als monocotyle waardplanten maakt het aannemelijk dat de
corresponderende resistentiegenen in dezeplanten ook nauw verwant zijn.
;etal. (1993) Molecular andGeneral Genetics 238, 261-269
Het effect van de C02-toename in de atmosfeer op de opwarming van de aarde is
gering in vergelijking tot de verhitting van de discussies over dit onderwerp.
Het streven naar de instandhouding van heidegebieden in Nederland is eerder een
zaak van cultuurbehoud dan van natuurbehoud.
10.
Een wetenschappelijk onderzoeker is meer gebaat bij een opgeruimde geest dan bij
een opgeruimd laboratorium.
11.
Het onderscheid tussen "toplay squash" en "to squash" isvoor sommige spelersvan
deze racketsport niet geheel duidelijk.
Stellingen behorend bij het proefschrift "Molecular aspects of avirulence and
pathogenicity of the tomatopathogen Cladosporium fiilvum"
Wageningen, 12 november 1993
GuidovandenAckerveken
Voorwoord
Indeafgelopen vierjaar zijn deadviezen, hulpen steunvan vele mensen essentieel geweest
vooreengoedeafronding vanhetpromotieonderzoek. IndeeersteplaatswilikPierredeWit
bedanken. Pierre, ik benje zeer erkentelijk voor de mij geboden mogelijkheid, om zonder
universitaire opleiding, als onderzoeker in opleiding het door BION gefinancierde
promotieonderzoek uit te voeren. De solide wetenschappelijke basis van het project heeft
uiteindelijk geleid tot goede resultaten. Pierre, ik dank je voor je goede begeleiding en
adviezen envoordevrijheid dieje megaf omopzelfstandige wijzeonderzoek teverrichten.
Dank aan de collega's van de Cladosporiwnfiilvumgroep met name; Jan voor het
overbrengen van zijn moleculair biologische kennis ("de wandelende Maniatis") en goede
adviezen; Matthieu voor deprettige samenwerking en hulpbij het "eiwitwerk"; Jos voorde
hulpbij demicroscopische opnamen; Henk, PaulenTon voor degoedehulpbij sequencen,
constructen maken, eiwit zuiveren ennorthern blotten. ManythankstoRuth, Nadia, Roland
and Guy for thenicecooperation and for critically reading manuscripts. De Studenten José,
Willem, Gerten Marko wil ikbedanken voor hunbijdrage aan hetonderzoek. Collega AIO
Corné wil ik bedanken voor de goede samenwerking en interessante discussies tijdens het
carpoolen. Zonder iemand te willen vergeten, dank ik alle medewerkers van de vakgroep
Fytopathologievoor deontspannen sfeer en prettige samenwerking.
Deonmisbareontspanning naasthetwerkwerd vaakgevonden opdesquashbaan dankzij
de tegenstanders en collega's Matthieu, Bill, Wynne en Paul. Daarnaast wil ik devrienden
vandeeetgroep, Ad, Hans, Matthieu, Nelleke, OlgaenTheobedanken voordegezelligheid
en ontspanning.
Dank aan stichting BIONen het LEB-fonds voor hun bijdrage in dedrukkosten van het
proefschrift.
Mijn ouders wil ikbedanken voor hun belangstelling gedurende het promotieonderzoek
en voor de geboden mogelijkheid te gaan studeren in Wageningen. Tenslotte wil ik vooral
Mirjam bedanken. Lieve Mirjam, dankje voorje niet aflatende steun en medeleven in de
afgelopen jaren.
UIM-GIX?
Theresearch described inthisthesis wascarried outatthedepartment ofPhytopathologyof
the Agricultural University Wageningen. The investigations were supported by the
Foundation for Biological Research (BION), which is subsidized by the Netherlands
Organization for Scientific Research (NWO).
Contents
List of abbreviations
1
Outline ofthisthesis
3
Chapter 1
The Cladosporiumficlvum- tomatointeraction, a model system
for fungus - plant specificity
7
Chapter2
Cloning and characterization of thecDNA encoding theAVR9
race-specific elicitor of Cladosporiumfulvum
27
Chapter 3
Molecularanalysisoftheavirulencegeneavr9 ofCladosporium
fulvum fully supports the gene-for-gene hypothesis
43
Chapter4
Disruption of theavirulence geneavi9 in Cladosporium fiilvum
causes virulence on tomato genotypes with the complementary
resistance geneCf9
57
Chapter5
The AVR9 race-specific elicitor of Cladosporium fulvum is
processed by endogenous and plant proteases
71
Chapter 6
Nitrogen limitation induces expression of the avirulence gene
avr9in Cladosporiumfulvum; a reflection ofgrowth conditions
inplantai
85
Chapter 7
Isolation and characterization of two putative pathogenicity
genes of Cladosporiumfulvum
Chapter8
General Discussion
101
115
Summary
125
Samenvatting
127
Account
131
Curriculum vitae
135
List of abbreviations
aa
act
aw
AVR
bp
BSA
Çf
ecp
ECP
EDTA
FOA
gpd
GUS
hph
HR
HRLC
HygR
IF
kb
kD
Mes
MM
Mr
nt
OMPdecase
OPRTase
ORF
PAGE
PEG
PBS
PR
SDS
SSC
TFA
Tris
Tricine
uidA
UV
amino acid(s)
geneencoding actin
avirulence gene
avrgeneproduct
basepair(s)
bovine serum albumin
genefor resistance to Cladosporiumfulvwn
geneencoding extracellular protein
ecpgeneproduct, extracellular protein
ethylene-diaminetetraacetic acid
5-fluorooroticacid
geneencoding glyceraldehyde-3-phosphate dehydrogenase
ß-glucuronidase
Hygromycin Bresistance gene
hypersensitive response
high resolution liquid chromatography
hygromycin Bresistant
intercellularfluid
kilo base (1000bp)
kilodalton
2-(iV-morpholino)-ethanesulphonicacid
tomato cultivar Moneymaker
relative molecular weight
nucleotide(s)
orotidine-5-phosphate decarboxylase
orotatepyrophosphoribosyl transferase
open reading frame
Polyacrylamide gel electrophoresis
polyethylene glycol
phosphate buffered saline
pathogenesis related
sodium dodecyl sulphate
standard salinecitrate
trifluoroacetic acid
tris(hydroxymethyl)aminomethane
iV-tris(hydroxymethyl)-methylglycine
geneencoding ß-glucuronidase
ultra violetlight
Outline of this thesis
New strategies for crop protection are required for efficient and environmentally friendly
agriculture. Durable resistance of plants to pathogens is the most effective means of crop
protection, but can often not be achieved by traditional plant breeding. Molecular breeding
for resistance isanewandalternative strategy for future cropprotection. Thisnew strategy
can only be developed when morebasic knowledge on plant-microbe interactions becomes
available. As P.R. Day once formulated: 'Wise and intelligent methods of crop protection
can only be based on an appreciation of the principles of host-parasite interaction and the
consequences ofinterfering withit' (Day, 1974).Fundamental research on molecular plantmicrobeinteractions canreveal themechanisms involved inpathogenicity and avirulenceof
pathogens. The activation of plant defence upon pathogen attack is mediated by an early
recognition of thepathogen and aquick and adequate response by the resistant plant. Basic
knowledgeonpathogen recognition andinductionofplantdefence caneventuallyleadtothe
isolation and characterization of plant resistance genes. Molecular understanding of the
structure and function of disease resistance genes could provide the basis for further
development of molecular resistance breeding.
At the beginning of the PhD project described in this thesis research on the
Cladosporiumfulvum - tomato interaction was focused on physiological and biochemical
aspects (Joosten, 1991). Plantresponse molecules such asphytoalexins (DeWit and Flach,
1979) and pathogenesis related proteins (Joosten and De Wit, 1989; De Wit et al., 1989)
were isolated and characterized. Compatible specific proteins were purified from the
intercellular fluid of C. fulvum -infected tomato leaves (De Wit et al., 1989). These
extracellularproteinsweresuggested tobeoffungal origin (JoostenandDeWit, 1988).The
abundance of these proteins in C../w/vM/n-infected tomato leaves and their absence in C.
fulvum grown invitrosuggested apossible roleinpathogenesis.
The discovery of race-specific peptide elicitors of C. fulvum (De Wit and Spikman,
1982)and the subsequent purification and characterization of theAVR9elicitor (ScholtensToma and De Wit, 1988) provided a solid basis for molecular studies on avirulence.
Avirulence of races of C.fulvum on tomato genotypes carrying resistance gene Cf9was
always correlated with the presence of the necrosis-inducing peptide elicitor in susceptible
plants infected with these same races. The hypothesis was put forward that the elicitor
originated from an avirulence gene present in races of C.fulvum avirulent on tomato
genotypeCf9.
In this thesis research on the molecular aspects of avirulence and pathogenicity of the
tomato pathogen Cladosporium fulvum is described. The interaction between the fungal
pathogen C.fulvum and tomato is an excellent model system to study fungus - plant
specificity. This willbedemonstrated in Chapter 1, which gives anoverview onthecurrent
status of the C.fulvum -tomato research.
Most of this thesis will deal with the avirulence gene avr9 of C. fulvum. Chapter 2
describes the cloning of the cDNA encoding the AVR9 race-specific elicitor. Races of C.
fiilvum virulent on tomato genotypes carrying the complementary resistance gene Cf9lack
the avirulence gene avr9. In Chapters 3 and 4 conclusive evidence is given that avr9is a
genuine avirulence gene. Chapter 3 describes the transformation of races lacking the avr9
genewiththecloned avr9 gene, resulting inachangeof cultivar-specificity from virulent to
avirulent on tomato genotype Cf9.In addition, disruption of avr9in avirulent races of C.
fiâvum, as described in Chapter 4, resulted in disruption mutants which have now become
virulent on tomatogenotypeCf9.
The AVR9 race-specific peptide elicitor could be studied in more detail by high
expression of the avr9 gene in transformants of C. fiilvum, as described in Chapter 5.
Evidenceispresented for theinvolvement ofboth thehostplant tomatoandthepathogen C.
fidvumintheprocessing oftheprimary avirulencegeneproduct toitsmatureAVR9elicitor.
Theregulation of expression of theavi9geneisdescribed in Chapter 6. Expression ofavr9
isinduced inplantaand under conditions of nitrogen limitation invitrowhich might reflect
the nutritional conditions which the fungus encounters in the tomatoleaf.
InChapter 7, thecloning andcharacterization oftwoputativepathogenicity genesofC.
fiilvumisdescribed. These ecp(encoding extracellularproteins) genes arehighly expressed
inplanta.Possiblefunctions of theecpgenesarediscussed and future research isdescribed
to elucidate their putative function in pathogenicity.
The results described in Chapters 1 to 7 are discussed in Chapter 8. The current
physiological, biochemical and molecular knowledge on the C.fiilvum- tomatointeraction
is presented in two models which describe basic compatibility and race-specific
incompatibility, respectively.
Day, P.R. (1974) Genetics ofHostParasite Interaction,Freeman and company, SanFrancisco.
De Wit, P.J.G.M. andflach, W. (1979)Differential accumulation of phytoalexinsintomatoleaves, butnot
infruits after inoculationwithvirulentandavirulentraces of Cladosporiumfulvum.Physiol.PlantPathol.
15, 257-267.
DeWit, P.J.G.M. andSpikman, G. (1982)Evidence for theoccurence ofrace andcultivar-specific elicitors
ofnecrosis inintercellular fluids of compatible interactions of Cladosporiumfulvumand tomato.Physiol.
PlantPathol.21, 1-11.
De Wit, P.J.G.M., Van den Ackerveken, G.F.J.M., Joosten, M.H.A.J. and Van Kan, J.A.L. (1989)
Apoplastic proteins involved in communication between tomato and the fungal pathogenCladosporium
fulvum. laSignalMoleculesinPlantsandPlant-MicrobeInteractions. (Lugtenberg,B.J.J., ed.), SpringerVerlag, Berlin, pp. 273-280.
Joosten, M.H.A.J. andDe Wit, P.J.G.M. (1988)Isolation, purification and preliminary characterization of
aproteinspecific forcompatibleCladosporiumfulvum(syn.Fulviafulva)-tomatointeractions.Physiol. Mol.
PlantPathol. 33,241-253.
Joosten, M.H.A.J. and De Wit, P.J.G.M. (1989) Identification of several pathogenesis-related proteins in
tomatoleaves inoculatedwith Cladosporiumfulvum(syn.Fulviafulva) as 1,3-ß-glucanasesandchitinases.
PlantPhysiol.89, 945-951.
JoostenM.H.A.J. (1991)The Cladosporiumfulvum- tomatointeraction:physiologicaland molecular aspects
of pathogenesis. PhD thesis, Wageningen.
Scholtens-Toma, I.M.J, and De Wit, P.J.G.M. (1988) Purification and primary structure of a necrosisinducingpeptidefromtheapoplasticfluidsoftomatoinfectedwithCladosporiumfulvum(syn. Fulviafulva).
Physiol.Mol. PlantPathol. 33,59-67.
Chapter 1
The Cladosporiumfulvum - tomato interaction,
a model system for fungus - plant specificity
GuidoF.J.M. van denAckerveken andPierre J.G.M. deWit
adaptedfromPathogenesis andHostSpecificity inPlantDiseases
(Kohmoto,K., Singh,R.P., Singh,U.S. andZeigler,R., eds.), PergamonPress, Oxford, Inpress
Table of contents
I.
Introduction
8
II.
Histopathologicalandultrastructural aspects of pathogenesis
8
HI.
The genetics of the Cladosporiumfulvum - tomato interaction
11
IV.
Biochemical andmolecular aspects of the interaction
A
Cladosporiumfulvum andtomato studied separately
15
B
Basiccompatibility
16
C
Race-specific incompatibility
D
Recognition anddefence responses
18
19
V.
Concluding remarks
20
VI.
References
21
I.
Introduction
The fungal pathogen CladosporiumfiilvumCooke (syn. Fulviafulva)is thecausal agentof
leaf mould of tomato (Lycopersicon esculentwn Mill.). C. fiilvum was first described by
Cookein 1883 1, from diseased material collected in South Carolina (USA). The origin of
C.fiilvumis mostprobably SouthAmerica, thecentre of origin of tomato. The fungus isa
biotrophic pathogen, which can however easily be grown in axenic culture. In infected
tomatoleaves, C.fiilvumgrows between themesophyll cellsand doesnot form specialized
feeding structures such as haustoria.
Since the 1930s many genes for resistance to C.fiilvumfrom related wild species of
tomato have been transferred to cultivated tomatoin breeding programmes 2. Often within
tenyears after theintroduction of theseresistant cultivars, newracesof C.fiilvumappeared
which could overcome the monogenic resistance 3A. At least eleven different genes for
resistance have been described of which six are available in near-isogenic lines of tomato
cultivar Moneymaker 5. Theobserved differential interactions (compatible or incompatible)
between different races of C.fiilvumand near-isogenic lines of tomato suggested that the
C.fiilvum- tomatointeraction fits thegene-for-gene hypothesis 67 .
In the last two decades, the physiology, biochemistry and molecular biology of the
C. fiilvum - tomato interaction have been studied by several groups 8"10. Fundamental
questions on pathogenicity, specificity and plant defence have been actively pursued. In
recent years wehave gained moreinsight intobasic compatibility and activeplant defence,
and have made significant progress in the molecular understanding of the gene-for-gene
interaction and theroleof fungal avirulence genes inpathogen recognition.
Inthischapter wedemonstratethattheinteraction between C.fiilvumanditshostplant
tomatoisanexcellentmodelsystemtostudythegene-for-geneinteraction.Pathogenicityand
avirulence factors and their respective involvement in induction of susceptibility and
resistance will be discussed in detail. We will only briefly discuss the defence responses
triggered by C.fiilvumas these responses are similar in other host-pathogen interactions
whichare studied worldwidebymanygroups (for reviews1113). Thehostdefence responses
are in many cases not specific for one particular host-pathogen interaction, whereas their
induction often ispathogen-specific in mostplants. Theinduction ofactiveplantdefence, in
particular the early responses and thehypersensitive response, by race-specific elicitors of
C.fiilvumwillbediscussed in moredetail. Thelatestfindingswillbecombined in amodel
explaining theinteraction at the molecular and biochemical level.
n.
Histopathological andultrastructural aspects of pathogenesis
Cladosporiumfiilvumcauses pale-yellow spots on theupper sideof infected tomatoleaves,
whicharetheresult ofabundant fungal growthinsidetheleaf.After anincubationperiodof
abouttwoweeks conidiophores emerge through stomataand sporulation becomes visibleas
brown-grey velvet areasatthelower sideoftheleaf. Whenthe infection progresses, leaves
become necrotic andeventually fall off.
Thefirstdetailed microscopic studieswereperformed byBondin193814. Germination,
penetration, developmentofintercellular myceliumand sporulation wereinitially studiedby
lightmicroscopy.Later, moredetailedstudieswereperformed byLazarovitsandHiggins1516
and DeWit17 using transmission and scanning electron microscopy. The infection process
was studied in time after inoculation of thelower side of leaves of near-isogenic linesof
susceptibleandresistantcultivarsoftomatowithconidiosporesofvirulentandavirulentraces
ofC.fiilvum.Conidiosporesgerminateontheleaf surface athighrelativehumidityandform
slender germtubes,theso-called runner hyphae (2-3/xmindiameter). Theserunner hyphae
are not growing in the direction of the stomata, but rather continue to grow until they
accidentally encounter open stomata, which are subsequently penetrated by the fungus at
three tofivedays after inoculation (Fig. 1C).Once inside theleaf, thediameter ofhyphae
enlarges twoto three fold (4-5urn). Thefungus colonizes theintercellular spaces inthe
leaves of a susceptible host without visible induction of defence responses (Fig. 1A).The
hyphae arein close contact with mesophyll cells butdonot form any specialized feeding
structures suchashaustoria (Fig. ID). Extracellularproteins (ECPs, discussed insectionIV
B)producedbyC.fiilvumareassociated withthematrixbetweenfungal hyphaeandhostcell
walls 18.Whether theseproteinsareessentialforthe adherence ofC.fiilvumtothe hostcell
wall isnotknown. Approximately 10days after penetration, anetwork of hyphae appears
in stomatal cavities andconidiophores emerge through thestomata andprofusely produce
conidiospores.
Thefungus cannot reproduce onresistant cultivars of tomato. After penetration ofthe
leaf through thestomata, fungal growth is arrested atanearly stage (Fig. IB). Theextent
of fungal growth isdependent ontheavirulence gene-resistance genecombination, which
probablydeterminestheprecisetimingofinductionofplantdefence19. Thedefence response
ischaracterized bythehypersensitiveresponse(HR,observedascellbrowningandnecrosis),
decompartimentalization andcallosedepositioninthosehostcellswhich areinclosecontact
with thehyphae of avirulent races of thepathogen 15~17. Mesophyll cells in or around the
infection site often show abundant excretion of extracellular material which occurs more
frequently inresistant than insusceptible plants1<s.
Avirulent races of C.fiilvuminduce drastic changes in theleaves of resistant tomato
cultivars, whereasvirulentracesdonotinduceanyvisiblechangesinthehostunder optimal
conditions.Incontrast, intercellular injection ofconidia ofboth virulent and avirulent races
ofC.fiilvumintoleaf mesophyllinducelocalizedresponses suchascellwallappositionsand
browning20.Anon-specific glycoproteinelicitor (discussedinsectionIVA)producedbythe
injected conidia ofvirulent aswell asavirulent races isprobably responsible for theinitial
inductionoftheseresponses21~25. Germtubesofvirulentraces, however, appeartogrowout
of the responding leaf mesophyll zone and hyphae continue to colonize the tomato leaf
withoutfurther inductionofdefenceresponses, whereasthegrowthofavirulentracesisvery
quickly arrested.
B
Figure 1. A, B) Schemetic representation of a transverse section of a tomato leaf 6-8 days after inoculation
with Cladosporiumfulvum.The fungal conidiosporegerminates on thelowerleaf surface, forms a thin runner
hyphaandenterstheleafthroughastoma.Thepenetratinghyphadevelopsintoathickerintercellularmycelium.
A) Susceptiblecultivar inoculatedwithavirulent race ofC.fitlvum.The fungus is growing abundantly inthe
intercellular spacesaround themesophyllcells;novisibledefence responsesare activated. B)resistant cultivar
inoculatedwithan avirulentrace ofC.fitlvum.After penetration fungal growthisarrested andmesophyllcells
incontactwiththefungus developahypersensitiveresponse (HR,darkcells). C)Scanningelectronmicrograph
showing penetration of C.fitlvumthrougha stoma at thelower side of the tomato leaf. D) Scanning electron
micrograph showing fungal hyphae growing around mesophyll cells (asinA).
10
Chitinases and 1,3-ß-glucanases of tomato, associated with plant defence 26'27, were
studiedbyimmunocytologyusingpolyclonalantibodies28.Subcellularlocalizationofproteins
in the intercellular spaces of C. fulvum-mîectsâtomato leaves is hindered by poor fixation
of these often highly solubleproteins. No significant differential accumulation of chitinases
and 1,3-ß-glucanases was observed between incompatible and compatible interactions. In
addition,noassociationofchitinasesand 1,3-ß-glucanases withfungal hyphaewasobserved,
suggesting that these hydrolytic enzymes play a minor role in active defence of tomato
against C.fulvum.
Future histopathological research on the C. fiilvwn - tomato interaction will involve
molecular biological techniques. In situ mRNA hybridization using labeled RNA probes
derived from plantdefence genes (encodinge.g. chitinases or 1,3-ß-glucanases) willbeused
to study the temporal and spatial expression of these genes. An alternative approach is
directedtousethereportergeneencodingtheEscherichia coliß-glucuronidase(uidA),which
was demonstrated to be fully active in C. fulvum 29. Fusion of promoters of interest to the
uidAgenewillallowtostudy thetimingandlocationofexpression offungal genesinvolved
inpathogenesis, and to search for compounds which induce those genes invitro.
m.
Thegenetics of the Cladosporiumfulvum - tomato interaction
There hasbeen acontinuous selection andbreeding of tomatofor morethan 200years 2.In
thebeginningofthiscentury resistance toC.fulvumwasfirst described andcertain resistant
varietieswereselected. Inthe 1930sLangford begantobreedfor resistanceand soonnoticed
that certain resistant varieties were still susceptible to someisolates of C. fulvum30. In the
following decades manygenes for resistance toC.fulvum (Cfgenes) havebeen crossed into
Lycopersicon esculentum from related wild species of Lycopersicon including L.
pimpinellifolium, L. hirsutum, L. hirsutum var. glabaratum andL.peruvianum 2,n. TheCf
resistance genesaresingledominantgenesandhavebeenmappedondifferent chromosomes
of tomato 32,33. Resistance genes Cf2 and Cf5 are allelic or closely linked and map on
chromosome 6 of tomato 33. Also, resistance genes against other pathogens, such as
Meloidogyne incognita (Mi) M and Oïdium lycopersicum 35, mapataround thesameposition
on chromosome 6. The resistance genes Cf4and Cf9are also allelic or closely linked and
map on chromosome 1 3236. Near-isogenic lines of tomato cultivar MoneyMaker are
available, each carrying one of these four Cfresistance genes, which are very useful for
studying the mechanisms of resistance genes separately.
Disease resistance genes have not been isolated so far. Several groups are currently
employing different strategies to clone Cfresistance genes37. Approaches used are (i) map
basedcloning,(ii)transposontagging,(iii)functionalcloningand(iv)cloninggenesencoding
receptors for race-specific elicitors (section IVD).
As described above, genes for resistance introduced in tomato were sooner or later
11
overcome by new races of C. fulvum 3A. The observed differential interactions (compatible
or incompatible) between different races of C.fulvum and near-isogenic lines of tomato
suggested that the gene-for-gene hypothesis was applicable to the C. fulvum - tomato
interaction67.Thegene-for-genehypothesisstatesthatgenesforresistanceintheplantmatch
to genes for avirulence in the pathogen. Flor, in 1942 38, developed the gene-for-gene
hypothesis based on genetic studies of the Melampsora Uni - flax interaction. Avirulence
genesinM. Uniandthecorresponding resistancegenesinflaxwhereshowntobesingleloci,
which segregated in a mendelian fashion 39.
Thegene-for-gene hypothesis in theC.fulvum - tomatointeraction is supported by the
findingofrace-specific elicitors(discussedinsectionIVC)whichfitintothespecificelicitor
- specific receptor model40,41. In this model, thegene products of fungal avirulence genes,
theso-called race-specific elicitors, arerecognized by specific receptors intheresistanthost
leading toincompatibility (Fig.2).Thereceptors recognizing thespecific elicitorsproduced
by the pathogen might be the direct products of resistance genes. Absence or mutation of
either the avirulence gene in thepathogen or the corresponding resistance gene in thehost
willabolish recognition leading tocompatibility. In this model therace-specificrecognition
is superimposed but fully dependent onbasic compatibility.
genotypeoftomatohost
R1- r2r2
genotypeof
C. fulvum
productsof
avirulencegenes
(race-specifcelicitors)
r1r1R2-
productsof Cfresistancegenes
putativespecificreceptors
A1a2
a1A2
=C
Figure 2. Schematic representation of the specific elicitor - specific receptor model explaining the gene-forgene interaction between Cladopsoriumfulvum and tomato. Avirulence genes (Al, A2) of thehaploid fungal
pathogen C.fidvumencoderace-specific peptideelicitorswhicharerecognized by theproductsofdominantCf
resistance genes (RI, R2) in tomato, theputative specific receptors. Recognition induces activeplant defence
(HR)resultinginincompatibility(I). Absenceoftherace-specific elicitororthecorresponding specificreceptor
will not result in recognition and C.fulvum continues to grow inside the tomato leaf without inducing any
defence responses (compatibility, C).
12
Thepresenceof singleavirulencegenesin C.fiilvumcannotbedemonstrated genetically,as
thesexual stage of thishaploid fungus isnot known or non-existing. Day, in 1957,tried to
prove the dominant character of an avirulence gene in C. fiilvum 42. Young sporulating
cultures of a mutant strain containing ared pigment marker and avirulent ontomatovariety
Vetomold (carryingresistance gene CfZ) wereirradiated withUVin order toobtain virulent
mutants. One mutant was obtained which had become virulent on the previously resistant
tomatovarietyVetomold.Thismutantstillcontainedtheredpigmentmarker, suggestingthat
it was a genuine mutant and not a contaminant. Mutation of C.fiilvumfrom avirulence to
virulenceasdescribedbyDay42, wasthefirst indicationfor theexistenceofavirulencegenes
in C. fiilvum.
Mutation to virulence as described above42, is the most plausible explanation for the
development of new races of C.fiilvum.Alternatively, races with new specificities might
arise as a result of parasexual recombination by hyphal anastomosis of two different races
of C.fiilvum43. However, it should be emphasized that only one report on occurrence of
hyphalanastomosis in C.fiilvumexists in theliterature43which could notberepeated with
wellcharacterized mutantsl0M.Therefore, theoccurrenceofparasexualrecombinationunder
natural conditions isvery uncertain.
Another mechanism for the development of new races of C. fiilvum might be
transposition and integration of a transposable element in avirulence genes 45. A
retrotransposon-likeelement called QT-1 ispresent in 30-100copiesat avariety of sitesin
the genome of C.fiilvum45. Evidence that QT-1 is a genuine LTR(long terminal repeat)retrotransposon wasobtainedbydemonstratingexpressionofQT-1andpackagingintoviruslikeparticles 4<s.However, no direct evidence for transposition of QT-1 was obtained, and
only circumstantial evidence suggests that QT-1 has transposed recently 46. The use of
transposons to inactivate genes by disruption can be of great help for the detection and
subsequent isolation of genes involved in pathogenicity and cultivar-specificity.
As mentioned above, no straightforward classic genetic studies can beperformed with
C. fiilvum, since it is an imperfect fungus. An artificial method for genetic analysis of
C.fiilvumwasperformed byTalbot etal.44.Byprotoplast fusion of twoparental mutantsit
was shown that a diploid strain could be obtained containing a mixed set of chromosomes
from both parents. Since the diploid is not stable, haploidization occurred spontaneously.
Recombinants were obtained from fusions between mutants in the nitrate assimilatory
pathway.AgeneticlinkagemapofC.fiilvumiscurrently beingconstructed, usingthisforced
parasexual recombination 47,48. Aninter-racial fusion between a mutant race4 and a mutant
race5,containingavirulencegeneavr5 andavrA respectively, hasbeenstudiedindetail4748.
Over 50 phenotypic, RFLP and RAPD markers have been used to analyse the haploid
progeny.Alinkagemapcontainingsevenlinkagegroupshasbeenestablished. Unfortunately,
no marker mapped toavrA and avr5 48.
Analternativestrategytoisolateavirulenceandpathogenicitygenesmightbemap-based
cloningbychromosomewalking.However, thistechniquehasnotyetbeensuccessful for C.
13
fiilvumfor two major reasons. First, there is only a smallnumber of polymorphic markers
onthelinkagemapofC.fiilvum,duetothelowgeneticvariationamongracesofC. fiilvum.
Secondly, the high content of repetitive DNA present on all chromosomes of C.fiilvum
hampers chromosome walking 49a. Hybridization of a genomic library of C. fiilvum with
labeled total genomicDNArevealed that 60-70 % oftheclonescontained repetitiveDNA.
The high degree of repeated sequences could also be a major reason for the problems
encountered in employing genomic subtraction, which was initiated to get an overall
impression of theoccurrence of deletions in C.fiilvum50.
Eleven chromosomes of C.fiilvumcanbe separated bypulsed-field gel electrophoresis
51
. The chromosomes range in size from about 2 to 7 Mbp (Fig. 3) and show few
polymorphisms in the different races and isolates of C. fiilvum. Some races contain an
additional minichromosome (nr. 12)of approximately 800kb. AUC.fiilvumgenes cloned
sofar havebeenmappedondifferent chromosomes andcanbeusedas chromosome-specific
markers.Additionalchromosome-specific markershavebeenisolatedfromagenomiclibrary
of C. fiilvum 49a, which can be used to identify individual chromosomes and to analyse
chromosome length polymorphisms.
The well documented genetics of tomato and to someextent of C.fiilvum,enables the
study of resistance and avirulence. The availability of near-isogenic lines of tomato with
different genesfor resistance andofracesofC.fiilvumwithdifferent specificities makesthe
C.fiilvum-tomatointeraction anexcellent model system to studygene-for-gene specificity.
C. fulvumrace
Figure 3. Separation of intact chromosomal
DNA molecules of 4 different races of
Cladosporium fiilvum by pulsed-field gel
electrophoresis . The size of the eleven
chromosomes of C.fulvum are estimated to
be 6.6, 6.3, 5.9, 5.7, 5.4, 4.6, 3.9,
3.4, 2.8, 2.5 and 1.9 Mb, respectively
for chromosomes 1 to 11 *. An additional
'mini'-chromosome of 0.8 Mb is present in
some races of C.fulvum. Chromosome
numbering, based on the pattern obtained
with race 4, is indicated on the left, the size
of the Schizosaccharomycespombe (Sp)
chromosomes is indicated on theright.
14
C
H
R
O
M
O
S
O
M
E
N
U
M
B
E
R
IV.
A.
Biochemical andmolecular aspects of the interaction
Cladosporiumfiilvumandtomato studied separately
The early studies on the C. fiilvum - tomato interaction were focused on the individual
partners. Resistant and susceptible plants were analysed chemically in search for factors
which could influence thegrowth and development of C.fiilvum52,53.The sugar andamino
acid content ofwholeleaves of different tomatovarieties werecompared, butdifferences in
chemicalcomposition could notbecorrelated toresistance or susceptibility52'53. The supply
ofnutrients and lightconditions during thegrowth of tomatoplants, however, influence the
severity of disease symptoms M.
TheearlybiochemicalworkonC.fiilvumwasmainlyfocused onfactors released bythe
fungus intheculturefluidofinvitro grownmycelium. Glycoproteinsproducedby C. fiilvum
invitro inducednecrosiswheninjected intomatoleavesbutwere,however, not race-specific
2125
. These non-specific elicitors induced physiological responses in tomato, such as
phytoalexin accumulation 24, electrolyte leakage 55 and increased lipoxygenase and
lipoperoxidase activity 55. Interestingly, most of these responses were only observed in
tomato and not in any other plant species tested 22. The activity of the non-specific
glycoprotein elicitors could be suppressed by incubating them with intercellular fluid
collected from C.fiilvum-infectedtomatoleaves5 6 . Ahost-derived factor was demonstrated
toinactivatethenon-specific glycoproteinelicitorbydegradation. Also, anon-proteinaceous
low molecular weight suppressor of thenon-specific glycoprotein elicitor was characterized
recently51.Sincetheproductsofdigestionofpolypectatehadsimilareffects itwassuggested
that this suppresssor originates from action of pectolytic enzymes on host cell walls. The
presence of this suppressor in theapoplast of healthy tomatoleaves, together with thehostproduced elicitor degrading activity, suggest that under natural conditions the non-specific
glycoprotein elicitor doesnot play a major rolein the C. Julvum - tomato interaction.
Transformation of C. fiilvum has been established using protoplasts 58,59. Dominant
selection markers, such as the hygromycin resistance marker (hph) M and the phleomycin
resistance marker (phleoR)61havebeen successfully used. High transformation frequencies
were obtained by using a complementing gene as a selection marker for transformation of
auxotrophic mutants of C. Julvum. Uridine auxotrophic mutants of C.fiilvum,resistant to
fluoro-orotic-acid, were complemented to prototrophy by transformation with heterologous
pyr genesofAspergillus nidulans,A. nigerand Clavicepspurpurea49,62.Theisolation ofthe
pyr gene of C. Julvumwill facilitate the development of an efficient gene targeting system
63
, which can be used in future research to study the promoters of avirulence and
pathogenicity genes of C.Julvum.
15
B. Basic compatibility
Theabilityofapathogen tosuccessfully penetrate, colonizeandreproduce onacertain host
plant is called basic compatibility 64"66. Only a few fungi are pathogens and are able to
establish andmaintain basic compatibility onacertain host species. Inaddition, biotrophic
plant pathogens are fully dependent on the living host and are therefore often highly
specialized. C.fiilvumisagoodexampleofaveryspecialized biotrophicpathogenwhichcan
onlycolonizetomato.Untilnow,toolsemployedbyC.fiilvumtosuccessfully infecttomato,
the so-called pathogenicity factors, are not known. C. fiilvum grows exclusively in the
intercellular spaces of the tomato leaf without causing any visible damage to host cells.
Factors involved in pathogenicity are therefore likely to beproduced in the intercellular
spaces of thetomato leaf, theinterface between pathogen andhost. Intercellular fluid (IF)
containing these factors can be isolated by infiltrating leaves of compatible tomato C.fiilvuminteractions with water orbuffer invacuo,followed bya centrifugation step 67.
Obviously, IFcontains besides fungal compounds alsoplant compounds.
From this IF, several low molecular weight proteins (<20kD) have been purified,
which are most probably of fungal origin since they are not present in the incompatible
C.fiilvum-tomatointeractions68. Theseextracellularproteins(ECPs)arealsonotproduced
by C.fiilvumgrown invitro.Twooftheseproteins, ECP1 (formerly called PI)69andECP2
18
werepurified andthecorresponding genes,ecpl andecp2,weresubsequentlyisolated 49,7°.
AnalysisofDNAsequencesandderived aminoacid sequencesofthe twoecpgeneshavenot
given anycluetotheir possible enzymatic or structural functions. Inallraces of C.fiilvum
the ecpgenes arehighly expressed inplanta as compared to invitro (Fig.4). This high
expression together with theabundance of theECPs intheextracellular space of infected
tomato leaves suggest arole inpathogenicity. Theextracellular localization could indicate
aroleinthematrix whichispresent attheinterface between the fungal hyphaeandthehost
cellwall18. Possibly, theECPsareactivelyinterfering withthe metabolism and/or transport
ofhostnutrientswithinthetomatoleaf,orsuppressing non-specificplantdefenceresponses.
Disruption oftheecp genesbygenereplacement with mutated ecpgeneswillreveal whether
these genes areessential forpathogenicity of C.fiilvum.
Several genes of C.fiilvumhave been isolated (av/4 71, avr949,72, ecpl andecp249,7°)
which arehighly expressed inplanta.Figure4showstheexpression ofthese four genesin
C.fiilvumgrowninvitro,anincompatibleand acompatible C.fiilvum-tomatointeraction.
Theecp andavrgenesarehighlyexpressed duringgrowthinplantabuthardlyornotduring
growth invitro.Ofthefour genes, only avirulence geneavr9showsinduced expressionin
vitrounderconditionsofnitrogen starvation49.Themechanismsofinductionorderepression
are notknown. Thepromoters of thefour genes donotreveal common structural motifs,
which couldbeinvolvedinthe regulation inplanta, although smallhomologies arepresent
between ecpl andecpl 49,7°. Future research using promoter-reporter gene fusions willbe
used todefine the promoter regions which areinvolved intheregulation of transcription.
16
Cf4/5
CfS/5
P F 4 6 8 101214 4 6 8 101214
aw4
0.7
avr9
**
w ^ ^ ^ ^ ^ ^ n &6
ecp\
as
ecp2
0.7
act
im
•
.
. ^
M
Figure 4. Northern blot analysis of poly(A)+RNAisolated from uninoculated tomatogenotype Cß (laneP),
from Cladosporiumfulvumgrowninvitro(laneF), fromanincompatibleinteractiontomatoCf4lC.fiilvumrace
5 (Cf4/5) andacompatibleinteraction tomato Cß/ C.fiilvumrace 5 (Cf5/5) atdifferent timesafter inoculation
(4, 6, 8, 10, 12and 14days, respectively). Theblotwashybridizedwith [<x-32P]dATP-labeledDNAprobesof
theC.fiilvumgenesavr4,avr9,ecp\, ecpl andact(actingene49"). Thelatterwasusedasaconstitutivecontrol
for fungal mRNA10.
SinceC.fiilvumobtainsitsnutrientsfrom theapoplast, Joostenetal.13 determined themonoand di-saccharides present in theintercellular fluid (IF)of C.fulvum-mîecXeà tomatoleaves
compared to those present in IF of healthy leaves. IF of tomato leaves inoculated with an
avirulent race of C.Julvum had a carbohydrate composition similar to IFof healthy tomato
leaves and remained unaffected within 14 days post inoculation. The carbohydrate
compositionofIFisolatedfrom tomatoleavesheavilycolonizedbyC.fiilvumwascompletely
different. In a time course experiment, a twofold increase in sucrose during thefirstseven
dayswasfollowedbyadecrease.Theleveloffructose andglucoseincreasedtenfoldbetween
17
fiveand nine days after inoculation, followed by a decrease. Mannitol, which is absent in
healthy tomato leaves, accumulates strongly until nine days after inoculation, to reach a
maximum at 14 days after inoculation. These results suggest that early in the interaction
invertases of C.fiilvumhydrolyze sucrose into fructose and glucose. The fungal enzyme
mannitol dehydrogenase is subsequently involved in the formation of mannitol, which
probably can only be utilized by thefungus but not by thehost plant.
C. Race-specific incompatibility
Amajorbreakthroughintheresearch ontheC.fidvum-tomatointeractionwasthediscovery
of race-specific elicitors, the inducers of active plant defence, in IF of C. fiilvum-infecteà
tomato leaves 67. Injection of IF in the leaves of resistant tomato plants resulted in the
inductionofHR,whichisvisibleaschlorosisornecrosis.Proteinaceouscompoundsinducing
a HR on resistant tomato cultivars were clearly demonstrated to be correlated with the
presence of avirulence genes in certain races of C. fiilvum 74'75. One such race-specific
peptideelicitor, theproduct oftheavirulencegeneavr9,waspurified from IFofC. fiilvuminfected tomatoleaves 76. Thepurified peptide elicitor specifically inducesHR (observed as
extensivenecrosis)ontomatoplantscarrying thecomplementary resistance geneCß. Races
virulent on tomatogenotype Cß do not produce the elicitor 77. Recently, the race-specific
elicitor inducing HR on tomatogenotype Cf4,theproduct of avrA has been purified 71.
The race-specific elicitor inducing HR on tomato genotypes containing the resistance
gene Cß was purified and the amino acid sequence was determined 78. The corresponding
cDNA was isolated using oligonucleotide probes deduced from theamino acid sequenceof
the peptide *9,12. The gene encoding the elicitor, named avr9, was only detected in fungal
raceswhichareavirulent ontomatogenotypescarrying the Cß gene. Racesvirulent on Cß
genotypes of tomato have evaded recognition by thehost as they lack the entire avi9 gene
49,72
and therefore do not produce the peptide elicitor. The avi9 gene was subsequently
isolated from agenomiclibrary ofC.fiilvumand sequenced49,79. Todemonstrate thecausal
relationship between a functional avr9gene encoding theelicitor and avirulence on tomato
genotype Cß, thecloned avr9genewas transferred torace2.4.5.9.11of C.fiilvum,which
lacks theavi9 geneand can overcome theresistance genes Cf2, Cf4, Cß, Cf9and Cfll of
tomato.Thecultivar-specificity of thetransformants containing theavr9genechanged from
virulent to avirulent on tomato genotype Cß- This was the first report on the cloning of a
fungal avirulence gene, which demonstrated that a singlegeneisresponsible for changes in
race-cultivar specificity 49,79.
Disruption ofavi9intworacesofC.fiilvumwasachievedbyreplacing theavr9 coding
region by the Aspergillusnidulans pyrG gene via a gene-replacement strategy in a pyr"
mutant 4962. The cultivar-specificity of these disruptants was consequently changed from
avirulent to virulent on tomato genotype Cß, confirming the conclusions drawn from the
transformation experiments described above49,79.
18
Theelicitor encoded byavr9 canbeisolated from IFof C.fulvum-infected tomatoleaves7S,
but not from C.fiilvumgrown invitro.To circumvent the laborious isolation procedure of
theAVR9elicitorfrom infected plantsandtoobtainlargequantitiesof thiselicitor, wehave
constructed a fusion between a strong constitutivepromoter (gpd from Aspergillus nidulans
80
)andtheavr9codingregion49,79. C.fiilvumtransformed with thisûw9-constructproduces
large amounts of active AVR9 elicitor invitro49,81. Milligrams of the AVR9peptide have
been purified from culture filtrates of these transformants, which can be used for AVR9
structure analysis (NMR) and elicitor-receptor binding studies. Plant breeders on the other
hand usethisAVR9elicitorinbreedingprogrammes to screen theirentries for thepresence
of the Cf9resistance gene.
Recently, theavirulence geneavrA of C.fiilvumhas beencloned bya similar approach
71
. Preliminary results indicate thatrecognition by tomato genotypes carrying theresistance
gene Cf4is overcome by single basepair mutations in the coding sequence of the gene. In
allvirulentracesanalysed, thesemutatedgenescontainonepointmutation,butthemutations
canbeatfour different locationsintheprotein. Allmutations, however, involveonechange
from cysteine to tyrosine. In one case the change from avirulence to virulence was caused
by a frame shift mutation.
D. Recognition anddefence responses
Biochemical research on the mechanism(s) of resistance to C.fiilvumwas focused on the
differences betweenincompatible (avirulentrace, resistant cultivar) andcompatible (virulent
race, susceptiblecultivar) C.fiilvum- tomatointeractions. The differential accumulation of
phytoalexins 82,83andpathogenesis related (PR)proteins, such as P14-isomers M and 1,3-ßglucanases and chitinases 26,85, was clearly demonstrated. The general idea is that in the
incompatibleinteractionanavirulentraceisrecognized atanearly stagebytheresistantplant
leading to theinduction of active defence responses.
Theprimaryrecognitionofrace-specific elicitorsandthesubsequent signaltransduction
is studied by several groups at the moment 19'86. Tomato suspension cells treated with IF
containingrace-specific elicitorsof C.fulvum were showntohaveretained thespecificity of
the intact plants 86. In contrast to the HR response of intact plants, the suspension cells
retained their viability after treatment with IF. Vera-Estrella et al. 86, demonstrated that IF
containing the AVR5 elicitor induced several physiological defence related responses,
specifically in cell suspension cultures of tomatocontaining the Cß resistance genebut not
in cultures containing the Cf4 resistance gene. Active oxygen species are rapidly generated
after treatment of suspension cells with IF, and are associated with lipid peroxidation,
cytochrome creduction, luminol-dependent chemiluminescence, oxygen uptake and release
of extracellular peroxidase and phenolic compounds. Extracellular production of active
oxygen species seems to be an important factor in the initial reaction of tomato cells to
specific elicitors of C. fulvum. The tomato cell suspension cultures responded in a similar
19
way to non-specific glycoprotein elicitors (section IV A) except for the luminol-dependent
chemiluminescence and increase, rather than decrease, in oxygen uptake87.
Using an invivo cotyledon assay, Hammond-Kosack etal. 19, confirmed the formation
ofactiveoxygenspecies (refered toasoxidativeburst)afterinjection ofIF,containingAVR9
elicitor, in cotyledons of tomato seedlings carrying the Cf9resistance gene. This oxidative
burst was detected within three hours after injection of elicitor. Theproduction of ethylene
was observed at nine to ten hours after injection, followed by increasing levels of salicylic
acid after 12hours. By 20 to 24 hours most mesophyll cells had lost membrane integrity,
leading to electrolyte leakage and alkalinization. Macroscopically visible necrosis could be
observed circa 30hours after injection with IF.
The early responses described above are induced rapidly after treatment with racespecific elicitors. The mechanism of primary recognition of elicitors by the resistant plant,
however, is still not known. Our laboratory is currently employing radioactively labeled
AVR9race-specific elicitorasaligand to studythebinding topotentialreceptors intomato.
The knowledge on avirulence genes and race-specific elicitors of the pathogen has
resulted inastrategytodevelopdisease-resistant cropplants.Intheso-called two-component
sensor system 37, theavirulence genein combination with thecorresponding resistance gene
is used to convert non-specific pathogen recognition into rapid and effective plant defence.
The gene encoding the AVR9 race-specific elicitor fused to a pathogen-inducible promoter
willbeexpressed immediately after pathogen attackin transgenicplants. TheAVR9peptide
will then bequickly recognized, leading totheinducion of activeplant defence (HR)which
will restrict thepathogen to the infection site. So far, transgenic tobacco plants containing
avi9 producing biologically active AVR9 elicitor are available 88. Hopefully, in the near
future transgenic tomato plants will be developed containing this two-component sensor
system which willberesistant to various pathogens.
V.
Concluding remarks
The datapresented in this chapter clearly show that the C.fulvum - tomatointeraction is a
very useful model system to study fungus - plant specificity. Further unravelment of this
model system mayhelpustoanswer basicquestionsonmechanisms of specificity andbasic
compatibility, not only for the C. fulvum - tomato interaction, but also for other fungus plant interactions.
20
Acknowledgements The authors are to grateful Dr. V.J. Higgins, Dr. J.D.G. Jones, Dr.
P. Lindhout and Dr. R.P. Oliver for providing unpublished data, and
to Dr. J.A.L. Van Kan and Dr. G. Honée for critically reading the
manuscript.
VI.
References
1 Cooke, M.C., New American Fungi, Grevillea, XII, 32, 1883.
2 Stevens,M.A. andRick, CM., GeneticsandBreeding, inTheTomato Crop,Atherton,J.G. andRudich,
J., Eds., Chapman and Hall, London, 1988, 35.
3 Hubbeling, N., Breakdown of resistance of the Cf5gene in tomatoby another new race of Fulviafiilva,
Med. Fac.Landbouww.Rijksuniv.Gent,43, 891,1978.
4 Laterrot, H., Race 2.5.9, anewrace of Cladosporiumfulvum (Fulviafiilva) and sources of resistance in
tomato, Neth.J. PlantPathol., 92, 305, 1986.
5 Boukema,I.W., Racesof Cladosporiumfulvum Cke.(Fulviafiilva)andgenesfor resistance inthetomato
(Lycopersicon Mill.),inProceedings oftheMeeting oftheEucarpiaTomato Working Group,Philouze, J.
Ed., Avignon, 1981,287.
6 Day,P.R., RacenamesofCladosporiumfulvum, ReportsoftheTomatoGenetics Cooperative, 6, 13,1956.
7 Day, P.R., Genetics ofHost-Parasite Interaction,Freeman and company, SanFrancisco, 1974,97.
8 DeWit,P.J.G.M., VandenAckerveken, G.F.J.M., Vossen,J.P.M.J.,Joosten,M.H.A.J., Cozijnsen,
T.J., Honée, G., Wubben, J.P.,Danhash,N., VanKan,J.A.L., Marmeisse, R. andVandenBroek,
H.W.J., Molecular cloning and functions of avirulence and pathogenicity genes of the tomato pathogen
Cladosporiumfulvum, inAdvancesinMolecularGenetics of Plant-Microbe Interactions,Vol.2, Nester,
E.W. and Verma. D.P.S., Eds., Kluwer Academic, Dordrecht, 1993,289.
9 Oliver, R.P., A model system for the study of plant-fungal interactions: tomato leaf mold caused by
Cladosporiumfulvum, mMolecularSignalsinPlant-Microbe Communications, Verma,D.P.S.,Ed., CRC
Press, Boca Raton, Florida, 1992,97.
10 Higgins, V.J., Miao, V. andHolland, J., Theuse of benomyl and cycloheximideresistance markers in
studiesofracedevelopmentbytheleafmoldpathogen Cladosporiumfulvum, Can.J. PlantPathol., 9 , 14,
1987.
11 Bowles,D.J., Defense-related proteins inhigher plants,Annu. Rev. Biochem., 59, 873, 1990.
12 Lamb, C.J., Lawton, M.A., Dron, M. and Dixon, R.A., Signals and transduction mechanisms for
activation of plantdefenses against microbial attack, Cell, 56, 215, 1989.
13 Scheel, D. andParker, J.E.,Elicitorrecognitionandsignaltransductioninplantdefense geneactivation,
Z. Naturforsch., 45c, 569, 1990.
14 Bond, T.E.T., Infection experiments with Cladosporiumfulvum Cooke and related species, Ann. Appl.
Biol., 25, 277, 1938.
15 Lazarovits, G. andHiggins, V.J., Histologicalcomparison of Cladosporiumfulvum race 1onimmune,
resistant, and susceptible tomato varieties, Can.J. Bot., 54, 224, 1976.
16 Lazarovits, G.andHiggins, V.J.,Ultrastructureofsusceptible,resistant,andimmunereactionsoftomato
toraces of Cladosporiumfulvum, Can.J. Bot., 54, 235, 1976.
17 De Wit, P.J.G.M., A light and scanning-electron microscopic study of infection of tomato plants by
virulent and avirulent races of Cladosporiumfulvum, Neth. J. PlantPathol, 83, 109, 1977.
18 Wubben, J.P. andDe Wit, P.J.G.M., unpublishedresults, 1993.
19 Hammond-Kosack,K., Oliver,R., Harrison,K., Ashfield, T. andJones,J., Studiesonthemechanism
by which tomato Cf (Cladosporium fulvum) resistance genes activate plant defence, in Advances in
MolecularGenetics ofPlant-MicrobeInteractions, Vol.2,Nester, E.W. andVerma. D.P.S.,Eds.,Kluwer
Academic, Dordrecht, 1993,457.
20 Higgins, V.J., Response of tomato to leaf injection with conidia of virulent and avirulent races of
Cladosporiumfulvum, Physiol. PlantPathol., 20, 145, 1982.
21 Dow,J.M.andCallow,J.A.,PartialcharacterizationofglycopeptidesfromculturefiltratesofFulviafiilva
(Cooke)Cifferi (syn. Cladosporiumfulvum), thetomatoleafmouldpathogen,J. Gen.Microbiol, 113, 57,
1979.
22 Lazarovits, G. andHiggins,V.J., Biologicalactivityandspecificity ofatoxinproducedby Cladosporium
21
fulvum, Phytopathol., 69, 1056, 1979.
23 Lazarovits, G., Bhullar, B.S., Sugiyama, H.J. and Higgins, V.J., Purification and partial
characterization of a glycoprotein toxinproduced by Cladosporiumfiilvum, Phytopathol., 69, 1062, 1979.
24 De Wit, P.J.G.M. and Rooseboom, P.H.M., Isolation, partial characterization and specificity of
glycoprotein elicitors from culture filtrates, mycelium and cell walls of Cladosporiumfiilvum (syn. Fulvia
Julva), Physiol. Plant Pathol., 16, 391, 1980.
25 De Wit, P.J.G.M. and Kodde, E . , Further characterization and cultivar-specificity of glycoprotein
elicitors from culture filtrates and cell walls of Cladosporiumfiilvum (syn. Fulvia fiilva), Physiol. Plant
Pathol., 18,297, 1981.
26 Joosten, M.H.A.J. and De Wit, P.J.G.M., Identification of several pathogenesis-related proteins in
tomato leaves inoculated with Cladosporiumfiilvum (syn.Fulviafulva) as 1,3-ß-glucanasesand chitinases,
Plant Physiol, 89, 945, 1989.
27 Bol, J.F., Linthorst, H.J.M, and Cornelissen, B.J.C., Plant pathogenesis-related proteins induced by
virus infection, Annu. Rev. Phytopathol., 28, 113, 1990.
28 Wubben, J.P., Joosten, M.H.A.J., Van Kan, J.A.L. and De Wit, P.J.G.M., Subcellular localization
ofplantchitinases and 1,3-ß-glucanasesin Cladosporiumfiilvum (syn. F«/via_/«/va)-infected tomato leaves,
Physiol. Mol. Plant Pathol., 4 1 , 23, 1992.
29 Roberts, I.N., Oliver, R.P., Punt, P.J. and Van den Hondel, C.A.M.J.J., Expression of the
Escherichia coli ß-glucuronidase gene in industrial and phytopathogenic filamentous fungi, Curr. Genet.,
15, 177, 1989.
30 Langford, A.N., The parasitism of Cladosporiumfiilvum and the genetics of resistance to it, Can. J. Res,
15, 35, 1937.
31 Kerr, E.A. and Bailey, D.L., Resistance to Cladosporiumfiilvum Cke obtained from wild species of
tomato, Can J. Bot., 42, 1541, 1964.
32 Van der Beek, J.G., Verkerk, R., Zabel, P. and Lindhout, P., Mapping strategy for resistance genes
in tomato based on RFLPs between cultivars: Cf9 (resistance to Cladosporiumfiilvum) on chromosome 1,
Theor. App. Genet., 84, 106, 1992.
33 Dickinson, M., Jones, D.A. and Jones, J.D.G., Molecular evidence for close linkage between the Cf2ICf-5 and Mi loci in tomato, submitted, 1993.
34 Ho, J-Y, Weide, R., Ma, H.M., Van Wordragen, M.F., Lambert, K.N., Koornneef, M., Zabel, P.
and Williamson, V.M., The root-knot nematode resistance gene (Mi) in tomato: construction of a
molecular linkage map and identification of dominant cDNA markers in resistant genotypes, Plant J., 2,
971, 1992.
35 Lindhout, P., Personal communication, 1993.
36 Jones, J.D.G., Personal communication, 1992.
37 De Wit, P . J . G . M . , Molecular characterization of gene-for-gene systems in plant-fungus interactions and
the application of avirulence genes in control of plant pathogens, Annu. Rev. Phytopathol., 30, 391, 1992.
38 Flor, H.H., Inheritance of pathogenicity in Melampsora Uni, Phytopathol., 32, 653, 1942.
39 Flor, H.H., Current status of the gene-for-gene concept, Annu. Rev. Phytopathol, 9, 275, 1971.
40 Keen, N.T., Specific recognition in gene-for-gene host-parasite systems, Adv. Plant Pathol., 1, 35, 1982.
41 Keen, N.T., Gene-for-gene complementarity in plant-pathogen interactions, Annu. Rev. Genet., 24, 447,
1990.
42 Day, P.R., Mutation to virulence in Cladosporiumfiilvum, Nature, 179, 1141, 1957.
43 Barr, R. and Tomes, M.L., Variation in the tomato leaf mold organism, Cladosporiumfiilvum, Am. J.
Bot., 48, 512, 1953.
44 Talbot,N.J., Coddington, A., Roberts, I.N. and Oliver,R.P., Diploidconstruction byprotoplast fusion
in Fulvia fulva (syn. Cladosporiumfiilvum): genetic analysis of an imperfect fungal plant pathogen, Curr.
Genet., 14,567, 1988.
45 McHale, M.T., Roberts, I.N., Talbot, N.J. and Oliver, R.P., Expression of reverse transcriptase genes
in Fulvia fulva, Mol Plant-Microbe Interact., 2, 165, 1989.
46 McHale, M.T., Roberts, I.N., Noble, S.M., Beaumont, C , Whitehead, M.P., Seth, D. and Oliver
R.P., CfT-I: an LTR-retrotransposon in Cladosporiumfiilvum, a fungal pathogen of tomato, Mol. Gen.
Genet., 233, 337, 1992.
47 Arnau, J. and Oliver, R.P., Inheritance and alteration of transforming DNA duringan induced parasexual
cycle in the imperfect fungus Cladosporiumfiilvum, Curr. Genet., In Press.
48 Arnau, J. and Oliver, R.P., personal communication, 1992.
22
49 Van den Ackerveken, G.F.J.M., Molecular aspects of avirulence and pathogenicity of the tomato
pathogen Cladosporium fulvum, thesis, 1993.
49a Van den Ackerveken, G.F.J.M., unpublished results, 1992.
50 De Wit, P . J . G . M . , unpublished results, 1991.
51 Talbot, N.J., Oliver, R.P. and Coddington, A., Pulsed field gel electrophoresis reveals chromosome
lengthdifferences betweenstrainsof Cladosporium fulvum (syn. Fulviafulva), Mol. Gen. Genet., 229,267,
1991.
52 Bailey, D.L. and Lowther, R.L., Studies on the nature of resistance in tomato to Cladosporium fulvum
Cooke, Can. J. Botany, 40, 1095, 1962.
53 Lowther, R.L., Further studies on the physiology of Cladosporium fulvum Cke. and resistance in tomato,
Can. J. Botany, 42, 1365, 1964.
54 Ghabrial, S.A. and Krone, T.P., Effect of mineral nutrition on host reaction to Cladosporium fulvum,
Phytopathol., 56, 493, 1966.
55 Peever, T.L. and Higgins, V.J., Electrolyte leakage, lipoxygenase, and lipid peroxidation induced in
tomato leaf tissueby specific and nonspecific elicitors from Cladosporium fulvum, Plant Physiol., 90, 867,
1989.
56 Peever, T.L. and Higgins, V.J., Suppression of the activity of non-specific elicitor from Cladosporium
fulvum by intercellular fluids from tomato leaves, Physiol. Mol. Plant Pathol., 34, 471, 1989.
57 Huogen, L. and Higgins, V.J., Characterization of an apoplasticnon-proteinaceous low molecular weight
suppressor of glycoprotein non-specific elicitorsof Cladosporium fulvum, manuscript inpreparation, 1993.
58 Harling, R., Kenyon, L., Lewis, B.G., Oliver, R.P., Turner, J.G. and Coddington, A., Conditions
for efficient isolation and regeneration of protoplasts hora.Fulviafulva, J. Phytopathol., 122, 143, 1988.
59 Oliver, R.P., Roberts, I.N., Harling, R., Kenyon, L., Punt, P.J., Dingemans, M.A. and Van den
Hondel, C.A.M.J.J., Transformation of Fulviafulva, a fungal plant pathogen of tomato, to hygromycin
B resistance, Curr. Genet., 12, 231, 1987.
60 Punt, P.J., Oliver, R.P., Dingemans, M.A., Pouwels, P.H. and Van den Hondel, C.A.M.J.J.,
Transformation ofAspergillus based on thehygromycin B resistance marker from Escherichia coli. Gene,
56, 117, 1987.
61 Mattern, I.E., Punt, P.J., Van den Hondel, C.A.M.J.J., A vector of Aspergillus transformation
conferring phleomycin resistance, Fungal Genet. Newsl., 35, 25, 1988.
62 Marmeisse, R., Van den Ackerveken, G.F.J.M., Goosen, T., De Wit, P.J.G.M. and Van den Broek,
H.W.J., Disruption of the avirulence gene avr9 intwo races of the tomato pathogen Cladosporium fulvum
causes virulence on tomato genotypes with the complementary resistance gene Cf9, Mol. Plant-Microbe
Interact., In Press, 1993.
63 Van Gorcom, R.F.M., Punt, P.J., Pouwels, P.H. and Van den Hondel, C.A.M.J.J., A system for the
analysis of expression signals inAspergillus, Gene, 48, 211, 1986.
64 Heath, M.C., The absence of active defense mechanisms in incompatible host-pathogen interactions, in
Active Defense Mechanisms in Plants, Wood, R.K.S., Ed., Plenum Press, New York and London, 1982,
143.
65 Heath, M.C., The role of gene-for-gene interactions in the determination of host species specificity,
Phytopathol., 81, 127, 1991.
66 Deverall, B.J., Molecular bases of fungal parasitism, Australasian Plant Pathol., 18, 79, 1989.
67 De Wit, P.J.G.M. and Spikman, G., Evidence for the occurrence of race and cultivar-specific elicitors
of necrosis in intercellular fluids of compatible interactions of Cladosporium fulvum and tomato, Physiol.
Plant Pathol, 2 1 , 1, 1982.
68 De Wit, P.J.G.M., Van den Ackerveken, G.F.J.M., Joosten, M.H.A.J. and Van Kan, J.A.L.,
Apoplastic proteins involved in communication between tomato and the fungal pathogen Cladosporium
fulvum, in Signal Molecules in Plants and Plant-Microbe Interactions, NATO ASI series, Vol. H36,
Lugtenberg, B.J.J., Ed., Springer-Verlag, Berlin Heidelberg, 1989, 273.
69 Joosten, M.H.A.J. and De Wit, P.J.G.M., Isolation, purification and preliminary characterization of a
protein specific for compatible Cladosporium fulvum (syn. Fu/via^/w/v^-tomato interactions, Physiol. Mol.
Plant Pathol, 33, 241, 1988.
70 Van den Ackerveken, G.F.J.M., Van Kan, J.A.L., Joosten, M.H.A.J., Muisers, J.M., Verbakei,
H.M. and De Wit, P.J.G.M., Characterization of two putative pathogenicity genes of the fungal tomato
pathogen Cladosporium fulvum. Mol Plant-Microbe Interact., 6, 210, 1993.
71 Joosten, M.H.A.J. and De Wit, P.J.G.M., unpublished results, 1993.
23
72 VanKan,J.A.L., VandenAckerveken, G.F.J.M. andDeWit.P.J.G.M., Cloningandcharacterization
of cDNAofavirulence geneavr9of thefungal pathogen Cladosporiumfidvum,causalagent oftomatoleaf
mold, Mol Plant-Microbe Interact., 4, 52, 1991.
73 Joosten,M.H.A.J., Hendrickx, L.J.M. andDeWit,P.J.G.M., Carbohydrate compositionofapoplastic
fluidsisolatedfromtomatoleavesinoculatedwithvirulentoravirulentraces of Cladosporiumfulvum(syn.
Fulviafulva), Neth.J. PlantPathol., 96, 103, 1990.
74 DeWit, P.J.G.M., Hofman, J.E. and Aarts, J.M.M.J.G., Originof specific elicitors of chlorosisand
necrosis occuring in intercellular fluids of compatible interactions of Cladosporiumfiilvum(syn.Fulvia
fulva) and tomato,Physiol.PlantPathol., 24, 17, 1984.
75 Higgins, V.J. andDeWit, P.J.G.M., Useof race-andcultivar-specific elicitorsfrom intercellular fluids
for characterizing races of Cladosporiumfiilvumand resistant tomato cultivars, Phytopathol., 75, 695,
1985.
76 De Wit, P.J.G.M., Hofman, A.E., Velthuis, G.C.M, and Kuc, J.A., Isolationand characterization of
anelicitor ofnecrosis isolated from intercellular fluidsof compatibleinteractionsof Cladosporium fiilvum
(syn.Fulviafulva) and tomato, PlantPhysiol, 77, 642, 1985.
77 Scholtens-Toma,I.M.J., DeWit, G.J.M. andDeWit, P.J.G.M., Characterization ofelicitoractivities
of apoplastic fluids isolated from tomato lines infected withnew races of Cladosporiumfiilvum,Neth. J.
PlantPathol, 95 Supplement 1, 161,1989.
78 Scholtens-Toma, I.M.J.andDeWit,P.J.G.M., Purificationandprimarystructureofanecrosis-inducing
peptidefromtheapoplasticfluidsoftomatoinfectedwithCladosporiumfulvum(syn.Fulviafulva), Physiol.
Mol.PlantPathol, 33,59, 1988.
79 Van den Ackerveken, G.F.J.M., VanKan, J.A.L. andDe Wit, P.J.G.M., Molecular analysis of the
avirulencegeneavr9of thefungal tomatopathogen Cladosporiumfulvumfully supports the gene-for-gene
hypothesis, PlantJ., 2, 359, 1992.
80 Punt, P.J., Dingemans, M.A., Jacobs-Meijsing, B.J.M., Pouwels, P.H. and Van den Hondel,
C.A.M.J.J., Isolation and characterization of the glyceraldehyde-3-phosphate dehydrogenase gene of
Aspergillusnidulans,Gene,69, 49, 1988.
81 VandenAckerveken, G.F.J.M., Vossen,J.P.J.M. andDeWit, P.J.G.M., TheAVR9peptideelicitor
of Cladosporiumfiilvumisprocessed by fungal and plantproteases, PlantPhysiol, In Press, 1993.
82 DeWit, P.J.G.M. andFlach, W., Differential accumulation of phytoalexinsintomatoleaves butnot in
fruits after inoculationwithvirulent and avirulent races of Cladosporiumfulvum, Physiol. Plant Pathol.,
15, 257, 1979.
83 DeWit, P.J.G.M. andKodde, E., Induction ofpolyacetylenicphytoalexinsinLycopersiconesculentum
after inoculationwith Cladosporiumfiilvum(syn. Fulviafulva), Physiol.PlantPathol., 18, 143,1981.
84 DeWit,P.J.G.M. andVanderMeer,F.E.,Accumulationofthepathogenesis-relatedtomatoleafprotein
P14as an early indicator of incompatibility in the interaction between Cladosporiumfiilvum(syn.Fulvia
fiilva) and tomato, Physiol.Mol PlantPathol, 28, 203, 1986.
85 VanKan, J.A.L., Joosten, M.H.A.J., Wagemakers, C.A.M., Van denBerg-Velthuis, G.C.M. and
De Wit, P.J.G.M., Differential accumulation of mRNAs encoding extracellular and intracellular PR
proteins in tomato inducedby virulent and avirulent races of Cladosporiumfulvum, PlantMol.Biol., 20,
513, 1992.
86 Vera-Estrella,R., Blumwald,E. andHiggins, V.J., Effect of specific elicitorsof Cladosporiumfulvum
on tomato suspension cells, PlantPhysiol., 99, 1208, 1992.
87 Vera-Estrella, R., Blumwald,E. andHiggins,V.J.,Non-specific glycopeptideelicitorsofCladosporium
fiilvum: Evidence for involvement of active oxygen species in elicitor-induced effects on tomato cell
suspensions. Physiol.Mol PlantPathol, In Press, 1993.
88 Honée, G., Personal communication, 1992.
24
Chapter 2
Cloning and characterization of the cDNA encoding the
AVR9 race-specific elicitor of Cladosporium fiilvum
Jan A.L. van Kan, Guido F.J.M. van den Ackerveken andPierreJ.G.M, deWit
adaptedfromMolecular Plant-Microbe Interactions 4 (1991), 52-59
Summary
Arace-specific peptideelicitor from Cladosporiumfiilvuminducesahypersensitive response
on tomato genotype Cf9 . We have hypothesized that the avirulence of fungal races on
genotype Cf9isdue to theproduction of this elicitor by an avirulence gene, avr9.In order
toobtain cDNA clonesof theavr9 gene, oligonucleotideprobes weredesigned based onthe
amino acid sequence, determined previously. In northern blot analysis, oneoligonucleotide
detected a mRNA of 500 nt. in tomato - C. fiilvum interactions involving fungal races
producing theelicitor. Aprimer extension experimentindicated thattheprobehybridized to
a region near position 270 of the mRNA. The probe was used to screen a cDNA library
made from poly(A)+RNA from an appropriate compatible tomato - C.fiilvuminteraction.
Oneclonewasobtained corresponding to the mRNA detected by theoligonucleotideprobe.
Sequence analysis revealed that this clone encoded the avr9 elicitor. By isolating longer
clones and by RNA sequencing, the primary structure of the mRNA was determined. The
mRNA contains an open reading frame of 63 amino acids, including the sequence of the
elicitor at the C-terminus. A time-course experiment showed that the avr9 mRNA
accumulates in a compatible tomato - C.fiilvuminteraction in correlation with theincrease
offungal biomass.Theavr9 geneisasinglecopygenewhichisabsentinfungal raceswhich
arevirulentontomatogenotype Cf9. Possiblefunctions oftheavirulencegenearediscussed.
27
Introduction
Specialization of plant pathogens has resulted in the evolution of various formae speciales
(fungi) and pathovars (bacteria) which are able to colonize only one or a small number of
hostplants. Within one forma specialis or pathovar, races can be found which are virulent
only on certain cultivars of the host, but are avirulent on other cultivars (Brasier, 1987).
Avirulence is caused by recognition of the pathogen by the host resulting in active host
defense (hypersensitiveresponse, HR).The molecularbasis ofpathogen recognition by the
hostplant is stilllargely unknown. Nevertheless, thepathogen must evade or suppress this
recognition in order to colonize theplant successfully.
Thegeneticsof manyrace-cultivarspecificplant-pathogeninteractionscanbedescribed
by a gene-for-gene model, where a pathogen-encoded avirulence geneproduct interacting
with the corresponding plant-encoded resistance gene product triggers an HR, eventually
leading to incompatibility (Crute, 1985; De Wit, 1987). Races of a pathogen lacking a
functional avirulencegenedonotactivatethehostdefense, leadingtosuccessful colonization
(compatibleinteraction).Race-specific avirulencegeneshavebeencloned from variousplant
pathogenicbacteria byashotgunapproach, involving screening for acquisition of avirulence
by virulent races transformed with genomicclones of avirulent races (Hitchin etal., 1989;
Shintakuetal., 1989;Staskawiczetal., 1984, 1987;Vivianetal., 1989).Transformants of
thevirulent races carrying a specific avirulence gene can be detected on cultivars carrying
theappropriateresistancegene.Forplantpathogenicfungi, suchacomplementation strategy
isdifficult to carry out duetolarger genomesizes, low efficiency of transformation and the
lack of suitable cloning systems. The only fungus for which an efficient transformation
system with autonomously replicating vectors has been established is Ustilago maydis
(Tsukuda et al., 1988).For many other plant pathogenic fungi, integrating cloning vectors
are available but autonomously replicating vectors still have to be developed. Therefore a
different approach, not involving random screening of transformants for acquisition of an
altered phenotype, is required in order to study genes involved in pathogenicity and
avirulence.
Ourworkonthetomatopathogen Cladosporiumfulvum (syn.Fulviafiilva) hasresulted
intheidentification offungalrace-specific elicitorswhichinducenecrosisontomatocultivars
with thecorresponding resistance genes (DeWitand Spikman, 1982;DeWit etal., 1985).
One such race-specific elicitor, the putative product of avirulence gene avi9, has been
purified to homogeneity. The purified protein induced rapid and extensive necrosis when
injected into leaves of tomato genotypes carrying the Cf9 resistance gene, but not in
genotypes containing other eigenes (Scholtens-Tomaand De Wit, 1988). The aminoacid
sequenceof thepurified elicitor has been determined (Scholtens-Toma and DeWit, 1988).
Theelicitorwasproduced inallcompatibletomato- C.fulvuminteractions involving fungal
raceswhichareavirulentontomatogenotypeCf9, butnotinanyinteractioninvolving fungal
races which are virulent on tomato genotype Çf9 (Scholtens-Toma et al., 1989). In this
28
paper, we report the cloning and characterization of the cDNA from the avirulence gene
avr9,which is thefirst fungal avirulence gene tobe cloned.
Results
Designofoligonucleotideprobesandscreening onnorthern blots
To detect the mRNA encoding the race-specific elicitor, four oligonucleotide probes were
synthesized, derived from theaminoacid sequence showninFigure 1.Theoligonucleotides
contained either mixtures of nucleotides (as in probe B) or inosines (as in probe D) at
ambiguous positions, or a combination of both (as in probes A and C). All four
oligonucleotides were 5'-end-labeled and hybridized to identical northern blots containing
equalamountsofpoly(A)+RNA from uninoculated tomatoplants, from C.fiilvumgrownin
vitro, and from three different compatible tomato - C.fiilvuminteractions. Figure 2 shows
that probe B hybridized specifically with a mRNA of approximately 500 nucleotides (nt.)
present intwocompatibleinteractions: tomatogenotype Cf41 C.fiilvumrace4 (lane3)and
tomato genotype Cß I C. fiilvum race 5 (lane 4). This mRNA was not detected in
uninoculated tomatogenotype Cß (lane 1)nor in C.fiilvumgrown invitro(lane2). Also,
no hybridization was observed in the interaction tomato genotype Cß I C. fiilvum race
2.4.5.9.11 (lane 5), as would be expected for an interaction involving a race which is
virulentontomatogenotype Cf9. Thus,thepreliminaryconclusionwasthatprobeBdetected
themRNA for thenecrosis-inducing peptide. ProbesA, Cand Ddid notdetectany specific
mRNAsonblots identical to theone shown in Fig. 2 (results not shown).
OligonucleotideprobeBwasusedinaprimerextensionexperiment.Theoligonucleotide
was 5'-end-labeled and hybridized to equal amounts of poly(A)+RNA isolated from
compatible interactions of tomato genotype Cß with either C. fiilvum race 5 or race
2.4.5.9.11(represented inFig.2, lanes4and5, respectively).Theprimerwasextendedwith
reverse transcriptase and theextension products were analyzed on a denaturing PAGE gel.
Figure3showsthata specific extensionproduct wassynthesized onpoly(A)+RNAfrom the
interaction tomatogenotype Cß IC.fiilvumrace5 (lane 1),butnotonpoly(A)+RNA from
theinteraction tomatogenotype Cß I C.fiilvumrace 2.4.5.9.11 (lane2). Thelength of the
extension product is approximately 270 nt., indicating that the avr9 mRNA contains
approximately 200nt. upstream of the sequence encoding thenecrosis-inducing peptide.
Preparation andinitialscreening ofcDNAlibrary
Poly(A)+RNAfrom theinteraction tomatogenotype Cß I C.fiilvumrace 5 (represented in
Fig. 2, lane 4) was used to prepare a cDNA library in Xgtll. A library was obtained
containing 100,000independent recombinants. Screening of filters containing 5,000phages
with 5'-end-labeled oligonucleotide probe B resulted in the isolation of two possible
candidates, onehybridizing weakly (phageA9-1), theother hybridizing significantly better
29
o
o
o
CM
CM
CM
CM
>U<0
>
o
e
O
o
u
i-i
M
o
I-I
o
<
I-l
m
o
u<u<o
3
•J
00
o
w
<^
HU
H
H
CM
U
H
O
H
o
M
o<C
V)
>^
o
HO
<!
^
i-H
O
n
:*,
u
e
i-i
o
s>->
I-I
ü
3
ID
iJ
<o
H
o
<o
o
o
PO
O
<
P
DU<;U<:O
o
u
<
H
<o
l-l
H
O
H
H
H
O
H
O
H
O
<o
o
O
<
<
o
<
P O < 0
o
o
u
<o
o
3°
HO
H
O
pu<o
o
o
po <o<o
o
o
o
o
PO
a
<
p
o<
PU
O
p
<u
l-l
3
»1
a
ta
<
H
H
l-l
PH
<!
<
HO
H
H
43
P,
w
o
<JO
u
<
<o
H
u
o
3
Ü
o
43
P-i
Pu <o
u
o
t>0
!-i
«:
u
43
H
U
l-i
(U
CO
c
ra
<
Cl
PO<O«:O
o
o
o
<
<
PO<CO
o
<!
PO
o
p
po<upo
o
o
p
<
P O < O P O
o
p
r
o
<
PO
o
u
p
H
1"
E£
12 34 5
i t
sn
«so
3 9«
3SO
'If
J
i'. •
Figure 2. Northern blot hybridization with
oligonucleotideprobeB.ProbeBwasS'-end-labeled
with fy-^PJATP and T4 polynucleotidekinase and
hybridized to a northern blot containing equal
amountsofpoly(A)+RNAisolatedfromuninoculated
tomatogenotypeCf5(lane 1), Cladosporiumfidvum
grown in vitro (lane 2), or from three compatible
tomato - C. fiilvum interactions: Cf4 with race 4
(lane 3), Cf5withrace 5 (lane4) and Cf5withrace
2.4.5.9.11 (lane5).
n
w
sg
Figure 3. Primer extension with oligonucleotide
probe B. Probe B was 5'-end labeled with [7-32P]
ATP andT4 polynucleotidekinase, hybridized to 5
Hg of poly(A)+RNA and extended with avian
myeloblastosis virus reverse transcriptase. The
extension products were analyzed on a denaturing
6% Polyacrylamide gel. Lane 1 contains the
extension products synthesized on poly(A)+RNA
from the interaction between tomato genotype Cj5
andCladosporiumfidvumrace5,andlane2contains
theextensionproductssynthesizedonpoly(A)+RNA
from the interaction between tomato genotype Cf5
and C. fiilvum race 2.4.5.9.11. Lane 3 contains
marker fragments, thelengthsofwhichare givenin
nucleotides in the margin.
(phage A9-2). Both phages were purified and their DNA was isolated. The phage DNAs
werelabeledbyrandomprimedlabelingandhybridized withblotsidenticaltotheblotshown
in Fig. 2. Phage A9-1hybridized with a mRNA of low abundance of 1.5 kbpresent in all
three tomato- C.fiilvuminteractions (datanot shown).Thisphagedid notcontain acDNA
corresponding to the mRNA detected by northern blotting (Fig. 2), and was not analyzed
further.
31
Labeled DNAfromphageA9-2hybridized with a mRNAof approximately 600nt. present
only in the compatible interactions tomato genotype Cf4I C. fiilvum race 4 and tomato
genotype Cß I C.fiilvumrace5, i.e. apattern identical to thehybridization observed with
oligonucleotide probe B (result not shown). Thus, phage A9-2 contained a copy of the
mRNA from the avi9 gene, encoding thenecrosis-inducing peptide. Restriction analysis of
theDNAfrom phageA9-2indicated thatthecDNAwasflanked byonly oneEcoBIsite,the
otherEcoEIsitehadbeenlostduringcDNAcloning.Theinsert wasestimated tobe400bp.
Sequence analysis oftheavr9 cDNA
The cDNA insert present in phage A9-2 was subcloned and the sequence was determined.
Theinsertis405bp.longandcorresponds tothe3' endofthemRNA, includingapoly(A)tail of 20 nt. The insert encodes the entire sequence of the necrosis-inducing peptide
contained within a longer open reading frame of 63 amino acids. From theposition of the
oligonucleotideprobe Bin theDNA sequenceand thesize of theprimer extension product,
itwasestimated thattheinsertofcloneA9-2lacksabout 110nt.ofthe5'-endofthemRNA.
In order to obtain full-length cDNA clones, the cDNA library was screened again with a
labeled RNA probe containing 70 nt. of the 5'-end of the insert of clone A9-2. Three
different phages (A9-3, A9-5and A9-8)wereobtained and theirinserts were subcloned and
sequenced. CloneA9-3did notcontain anEcdSI siteatthe3'-end of thecDNA insert and
1 XXXXXXXCGAAUUUCGAAAAAUCCUCAAGCCUACUAAGGUCUUUGUAGCCAUAGUCAirUU
61 UAAUAAGUCÙAUCGACUUACGUAUCUAUCÀUGAAGCUUUCCCUCCUUAGCGUAGAGCUUG
M K L S L L S V E L A
121 CUCUCCUAAÙUGCUACUACÙCUCCCACUUÙGCUGGGCAGCUGCCCUCCCÙGUAGGAUUGG
L L I A T T L P L C W A A A L P V G L G
181 GAGUCGGGCÙAGACUACUGÙAACUCAAGUÙGUACUAGGGCCUUCGACUGÙUUGGGUCAAÙ
V G L D Y C N S S C T R A F D C L G Q C
241 GUGGCAGAUGCGACUUUCAUAAGCUACAAÛGUGUCCACUÀGAGGACUAGÀGAGGAAGUGG
G R C D F H K L Q C V H
301 AGAGAAGAGGAGGGGAGAGGUACGAUAACÙAGCGAGUAAÀUCGUACAGGÙAGAAAGGGAÙ
361 AGUAAGCAGGCAGAUAGACGGACGACGUUGCGACCUUAUCCAACAUAAGÙCCUAGUCGUÂ
421 ACAUUCGUUCAUAUUGAAGGCUUUUCCUCÀAUAGUUUCUCAAAUGUGCUGCGAGGCGCAG
481 AGCCAAG 487
Figure4. Nucleotidesequenceof theavr9mRNAandtheaminoacidsequenceof theavr9translationproduct
(GenBank/EMBLaccessionnumberMS5289).Intheaminoacidsequence,thematurenecrosis-inducingpeptide
is indicated in bold characters.
32
was therefore only sequenced from the 5'-end. The sequence of all three clones was
completely identical to the sequence of clone A9-2 in the overlapping regions. All three
clones containing poly(A)-tails have different sites of polyadenylation. From the DNA
sequenceanalysisandtheprimer extension experiment showninFig. 3, itwasdeduced that
clone A9-3, which contained the most prolonged 5'-sequence of the mRNA, still lacked
approximately 35nucleotides.Therefore, anewprimer wasdesigned, hybridizingatposition
75-100. This primer was used in a primer extension experiment on poly(A)+RNA in the
presence of dideoxynucleotides. The RNA sequencing allowed the addition of another 24
nucleotides upstream of the insert of A9-3 (results not shown). The first 7 nt. of theavi9
mRNA could notberead. Three minor extension products were observed whichwere 5-20
nt. longer than the major extension product. The different end-products of the primer
extension were not a result of degradation of mRNA since an extension experiment with a
primerforadifferent fungal mRNAyieldedonediscreteextensionproductofthecorrectsize
(results not shown). The nucleotide sequence of the avr9 mRNA as obtained from
sequencing cDNA clones and mRNA is shown in Figure 4.
Timing ofexpression oftheavr9mRNA
Todeterminethetimingofexpressionoftheavr9 gene,anorthernblot wasmadecontaining
equal amounts of poly(A)+RNAs from tomato leaves harvested at different times after
inoculation withC.ficlvum.Onetimecoursewas madeof thecompatibleinteraction tomato
genotype Q5 I C. fulvum race 5, and another one of the incompatible interaction tomato
genotype Cf41C.fulvumrace5. Poly(A)+RNAsfrom uninoculatedtomato(tomatogenotype
Cf5)and from C. fulvum grown in vitro (race 5) were included as controls. The blot was
hybridized withthelabeled insertofcloneA9-2.Figure5showsthattheavr9mRNAcould
bedetected in thecompatible interaction at day6post inoculation, strongly increasing from
day 8post inoculation and onward. No hybridization could beobserved in the incompatible
interaction. AlowamountofmRNAcouldbedetectedinthefungus growninvitro(laneF).
Cf4/5
Cf5/5
P F 4 6 8 101214 4 6 8 101214
Figure5. Timecourseofaccumulationoftheavr9mRNA.TheinsertofcloneA9-2waslabeledandhybridized
toanorthern blotcontainingequal amounts ofpoly(A)+RNA isolated from uninoculated tomatogenotype Cf5
(laneP), Cladosporiumfiilvumgrown in vitro(laneF), an incompatibleinteraction between tomato genotype
Cf4andC.fulvum race 5 (lanes Cf4/5) or from acompatibleinteractionbetween tomatogenotype Cf5and C.
fulvum race 5 (lanes Cf5/5) at different times (4, 6, 8, 10, 12and 14days, respectively) after inoculation as
indicated.
33
However, whencomparing mRNAlevelsof thefungus grown invitrotothefungus grown
mplanta, it should be considered that the RNA samples from the compatible interaction
contain only low proportions of fungal mRNAs, especially in early stages of infection.
Relative tothis, only very minuteamountsof fungal RNA are obtained from incompatible
interactions,sincefungalgrowthisinhibitedcompletely shortlyafterpenetrationoftheplant.
Southern blotanalysis oftheavr9gene
Southern blotanalysisofDNAisolated from sevenracesofC.fulvum (Fig.6)indicatedthat
thecDNA clonehybridized to singlebandsinvarious restriction enzymedigests. The avr9
geneispresent ina singlecopyinraces2, 4, 5, 2.4 and2.4.5 (Fig. 6, lanes 1, 2, 3, 5and
6), respectively. The gene could not be detected in races 2.5.9 and 2.4.9.11 which are
virulentongenotype Cf9(Fig.6, lanes4and7).Analysisoffour additionalraces confirmed
the presence of single hybridizing fragments in races 2.5, 2.4.5.11 and 2.4.11 and the
absenceof hybridization in race2.4.5.9.11 (resultsnot shown).Theuseof alonger cDNA
probe (clone A9-5) or lower stringency conditions during hybridization gave hybridization
patterns identical to the blot shown in Fig. 6 (results not shown). Apparently, the coding
sequenceof theavr9geneispresent inraces thatareavirulent on tomatogenotype Cf9and
absentin races that are virulent on tomatogenotypeC/9.
ECORI
HMDIH
2 3 4 5 0 7
. • * •
n'
1 2 3 4 5 6 7
«n
xHOi
<3a«oe71 t i 4 i i r
*>••'.*a
'•<"?!
•
•
W
S T n t !n
^n £ '
f
D N A fi m d i f l f e M t
°
f
•-
>• '. • •
I- • '
*"** ° C^tPoriumjulvum.
Racesusedwere
race 24.9.11 (lane 7). Five myograms of fungal DNA was digested with EcdSl, Hindm, ft/I or Xhol as
mdicate^separated ona0 7%agarosegelandblottedontoaHybondN* membrane. T h T ï ï o t Z j £ £ S
witharandom-pnmedlabeledfragment of420nucleotidescontainingtheentireavr9cDNAinsert (fromA9-2).
34
Discussion
Usinganoligonucleotideprobe,wehavedetectedthemRNAandcloned thecDNAencoding
thenecrosis-inducing peptide, which wehypothesizeis the direct product of the avirulence
geneavr9 of Cladosporiumfiilvum. Webelievethatthisgeneistheavirulencegenesinceits
geneproduct specifically induces HR on tomatogenotype Q9. Nevertheless, it remains to
beproven that transformation of C.fiilvumraces, lacking ÖV/9, with theavr9gene confers
avirulence on tomato genotype Cf9 in a gene-for-gene relation. The isolation and
characterization of cDNA clonesrevealed that thenecrosis-inducing peptide isproduced as
a precursor protein of 63 amino acids. Surprisingly, the DNA sequence indicated an
additional histidine codon at the C-terminus of the sequence of the mature elicitor. The
elicitor was reported previously to be 27 amino acids long (Scholtens-Toma and De Wit,
1988), but reexamination of theprotein sequence data confirmed thepresence of an extra
histidine residue at position 28. The signal of histidine had been overlooked during initial
protein sequence analysis. The molecular mass of the matureavr9elicitor is 3192Daltons.
Thepresence of the extra histidine probably explains why a synthetic peptide of 27 amino
acids (lacking the C-terminal histidine) was not biologically active in bio-assays on tomato
genotype Cf9(P.J.G.M.deWitandF.Th. Brederode, unpublished results).Thepeptidewas
poorly soluble in water. Several attempts have been made to activate the 27 amino acid
syntheticpeptidebyreductionandslowoxidation,butweneverobtainedapreparation which
was biologically active. Usually, theremoval of ureum and/or acetic acid from the solvent
by dialysis made most of the peptide precipitate. The supernatant contained no necrosisinducing activity (P.J.G.M. de Wit and F.Th. Brederode). The biological activity of the
chemically synthesized peptide of 28aminoacids will betested in future experiments. The
N-terminalaminoacids which are absentin thematureelicitor havesomecharacteristics of
a signalpeptide(lengthandhydrophobiccharacter). Thecleavage site(betweenanaspartate
and a tyrosine residue), however, does not correspond to the (-3,-1) rules for cleavage of
signalpeptides (VonHeijne, 1986).Probably, theputative signalpeptideisremoved during
excretionfrom thefungalcellintotheapoplast,butitremainstobedeterminedwhetherplant
factors are needed for additional maturation toa biologically active elicitor.
The avi9 mRNA could be detected at day 6post inoculation, that is approximately 3
days after penetration of the stomata by fungal hyphae. The mRNA increases very rapidly
between day 6and day 8postinoculation and onwards. This increase mightreflect the fast
increase in fungal biomass during this period, as deduced from two parameters for fungal
biomass inplanta, mannitol concentration and mannitol dehydrogenase (MTLDH) activity
in apoplastic fluids of infected leaves. These twoparameters show a fast increase between
day 6 and day 8post inoculation (Joosten et al., 1990). Before day 6post inoculation, no
significant fungal biomasscouldbemeasuredincompatibleinteractionsandconsequentlyno
avi9 mRNA could bedetected. Alsoin theincompatibleinteraction, no fungal biomassand
noavr9mRNAcouldbedetected (Fig.5). Nevertheless, it isvery likelythattheavr9gene
35
isexpressed at very early stages of theinfection cycle, sinceitsgeneproduct is believed to
trigger HR on tomato genotype Ç/9at about day 4 post inoculation (within 1 day after
penetration of the stomata). Final evidence about the expression of avr9 mRNA in early
stages of infection willbeobtained from insituhybridization experiments.
It is difficult to speculate on a biological function of the necrosis-inducing peptide for
the fungus. From the fungal point of view the possession of an avr gene would be
disadvantageous,unlessthegenehassomebeneficial function. Therefore weproposethatthis
peptidehas another, however dispensable, function in theinfection cycleof C.fiilvum.The
function of the peptide is either not essential or can be compensated by other proteins.
Whatever maybeitsfunction, themodeofactionoftheC.fiilvumavt9geneproduct differs
in several aspects from bacterial avirulence gene products. First, the mature avi9 protein
itself directly induces HR on tomato genotype Cf9 (Scholtens-Toma and De Wit, 1988).
Purified proteins encoded by bacterial avirulence genes do not induceHR directly (Tamaki
etal., 1988;Ronald and Staskawicz, 1988)butrather seemtoactviatheproduction oflow
molecular weight elicitor-active molecules (Keen et al, 1990). Secondly, the avr9 gene
product is aprotein which is excreted by thepathogen into theapoplast of infected leaves.
In the case of bacterial avirulence gene products, all evidence points to an intracellular
location oftheseproteins (Napoliand Staskawicz, 1987)thusreinforcing thehypothesisthat
they produce elicitor-active compounds (Keen etal., 1990).
Fungal races which do notproduce theavr9elicitor, and are consequently virulent on
tomatogenotype Q9, entirely lackatleastthecodingregion oftheavr9 gene. Hybridization
with flanking sequencesofthegene(isolated from agenomiclibrary) shouldrevealwhether
theseraceslackmorethanjustthecoding sequence. Accordingly, theseracesdonotcontain
a non-functional allele which could overcome the avirulent phenotype. Therefore the
assignment of virulence gene 9 could beregarded as misleading. The absence of the avr9
gene in C.fiilvum,resulting in virulence on a formerly resistant host, is analogous to the
absence of three avirulence genes (avrA,avrBand avrQ in virulent races ofPseudomonas
syringae pv. glycinea(Staskawicz et al., 1984; Staskawicz et al., 1987). In contrast, nonfunctional alleles of avirulence genes were demonstrated in the plant pathogenic bacteria
Xanthomonas campestris pv. vesicatoria (Kearney etal., 1988)and Pseudomonas syringae
pv. glycinea(Kobayashi et al. 1989; Kobayashi et al., 1990b). In spontaneous mutants of
Xanthomonascampestrispv.vesicatoriawhichhadlostafunctional avrBslgene, mostifnot
all mutants appeared to contain a transposable element, either in the coding region or in
regulatory sequences (Kearney et al., 1988). In the case of the avrD gene isolated from
Pseudomonassyringae pv. tomato, other pathovars of P. syringae (e.g. pv. glycinea)
appeared to have a homologous allele (Kobayashi et al., 1989; Kobayashi et al., 1990b)
which seems a functional gene (with 86% amino acid homology to avrD) lacking the
necrosis-inducing phenotype (Kobayashi etal., 1990b).
It is tempting to speculate about the origin of the avr9 gene. It has been shown that
some avirulence genes of plant pathogenic bacteria reside on endogenous plasmids
36
(Kobayashi et al., 1990a; Swanson et al, 1988). Several plasmids have been detected in
plant pathogenic fungi (Leong and Holden, 1989), and recent data have indicated the
presence of pathogenicity geneson so-called B-chromosomes (Van Etten etal., 1989).The
possible presence of the avr9 gene on an episomal factor or a B-chromosome should be
considered. Alternatively, itispossiblethattheavr9 geneislocatedinanunstableregionof
thegenome which can bedeleted without dramatically affecting theviability of the fungus.
Experiments todistinguish between agenomicorepisomiclocalization of theavr9geneare
currently carried out. It remains to be determined whether virulence of other, if not all,
fungal racesonvariousother Çftomatogenotypesiscausedbydeletionofthecorresponding
avirulence genes. We are currently purifying race-specific elicitors other than AVR9, in
order to clone their encoding genes and determine whether loss of avirulence genes is a
general featureinvirulentphenotypes.Itwillbeinteresting todeterminewhether avirulence
gene products might have a beneficial effect in some stages of the fungal life cycle.
Competition experiments between a race lacking avr9 and a transformant of this race
containing thegenecould give aclueabout therelevance of theavr9genefor C.fiüvwn.
It will be of great interest to isolate and characterize the entire avr9 gene with its
regulatory elements, inorder tostudytheregulation ofthisgeneinvivo inresponse toplant
signals.Preliminary resultsindicatethat undercertain defined growth conditionsexpression
of the avr9 gene can be induced in vitro (Van den Ackerveken et al., this thesis). The
structure and regulation of expression of the avr9genewillbeanobject of further studies.
Experimental procedures
Table 1.
Interaction of Cladosporiwnfulvum races
with tomato genotype CJ9.
Race
Avirulence
genotypes
2
4
5
2.4
2.5
AiAjA^A,,
A2A5A9A11
A2A4A9A11
A5A9A11
A4A9A11
2.4.5
2.5.9
2.4.11
2.4.5.11
2.4.9.11
2.4.5.9.11
A»AU
Interaction with
tomato genotype CJ9
A4Aa
Ta
Cb
AsAg
A,
A5
-
c
c
'I = incompatible interaction, fungal race avirulent;
b
C = compatible interaction, fungal race virulent.
37
Plants,fiingiand inoculations
Near-isogenic tomato cultivars were grown as described by De Wit and Flach (1979).
Different races of Cladosporiumfidvum(Cooke) (syn. Fulvia fulva (Cooke) Cif) were
subculturedandinoculatedontotomatoplantsasdescribedbyDeWit(1977).Theracesused
in this study and the outcome of their interaction with tomato genotype Q9 are shown in
Table 1.
RNAisolation andnorthern blotting
RNA wasisolated by homogenizing tissuein guanidine-HClbuffer (8Mguanidine-HCl/ 20
mM Mes/ 20 mM EDTA/ 50 mM ß-mercaptoethanol, pH 7), extracting with
phenol/chloroform (1:1) and chloroform, and precipitating overnight with 2M LiCl.
Poly(A)+RNA was obtained by affinity chromatography on oligo(dT)-cellulose,
electrophoresed on denaturing formaldehyde-agarose gels and blotted on Hybond N or
Hybond N + membranes as described by Maniatis et al. (1982).
Hybridization witholigonucleotideprobesandDNArestrictionfragments
Oligonucleotideswere5'-end-labeledwith|/y-32P]ATPandT4polynucleotidekinaseandused
directly for hybridization. Hybridization with RNA blots was performed in 6xSSC/0.5%
SDS/0.1% NaPPj/100 /tgml1 calf thymus DNA/25 jtgml1 yeast tRNA at 35°C. The blots
were washed in lxSSC/0.5% SDS at 35°C. Hybridization of filters containing cDNA
recombinantphageswasperformed in6xSSC/0.5% SDS/0.1% NaPP;/100/igml1calfthymus
DNAat32°C. Thefilters werewashed under thesameconditions.Restriction fragments of
cDNAclones wereisolated from agarosegelsand labeled with [a-32P]dATPusing aPrimeA-Gene labeling kit (Promega). Hybridization was carried out at 42°C in 5xSSC/0.5%
SDS/5xDenhardts/100 ng ml 1 calf thymus DNA. Filters were washed at 60°C in
0.2xSSC/0.2% SDS (high stringency) or at 50°C in lxSSC/0.5% SDS (low stringency).
Primer extension andRNA sequencing
Oligonucleotides were 5'-end-labeled with [7-32P]ATP and T4 polynucleotide kinase. Five
ng oflabeled primer werehybridized to5figofpoly(A)+RNAin 2xReverse Transcriptase
buffer (Promega)byheatingto65°Candcoolingdowntoroomtemperature.Theprimerwas
extendedwithAMVReverseTranscriptase (Promega)eitherintheabsenceorinthepresence
of dideoxynucleotides for RNA sequencing. The concentrations of ddNTP's in individual
sequencing reactions were 0.2 mMexcept for ddTTP (0.4 mM).
cDNA synthesis andlibrary construction
cDNA was synthesized on 5 ng of poly(A)+RNA isolated from the interaction tomato
genotype Cß I C.fiilvumrace 5using theProtoclone cDNA kit (Promega). EcoBI sites in
thecDNA were modified with EcöSl methylase (Promega) and EcoRI-linkers were added
withT4DNAligase. After EcoRI-digestion,excesslinkerswereremoved witha Qiagentip
(Diagen) and cDNA was ligated to EcoRI-digested lambda gtll arms (Promega). After
38
ligation phage DNA waspackaged in phageparticles using aPackagene kit (Promega) and
plated onE. coliY1090 (r). Theprimary library (100.000independent recombinants) was
amplified and stored at 4GC.
Cloningprocedures andDNA sequencing
All DNA manipulations werecarried outessentially asdescribed by Maniatis etal. (1982).
DNA sequencing wasperformed with thechain termination method of Sanger etal. (1977),
using [a-35S]dATP label, on double-stranded plasmid DNA with T7 and SP6 promoter
primers specific for pGEMplasmids (Promega).
Southern blotanalysis offungalDNA
TotalDNAof Cladosporiumfulvumwasisolatedbygrindingfreeze-dried myceliuminliquid
N2, homogenizing in extraction buffer (0.5MNaCl/ 10mMTris-Cl/ 10mMEDTA/ 1%
SDS,pH7.5),extractingwithphenol/chloroform (1:1)andchloroform andprecipitating the
aqueous phase with isopropanol. DNA was treated with RNase and digested with either
EcöSl, HinäBl, Pstl orXhol. From each race, approximately 5 /xgof digested DNA was
electrophoresed on a 0.7% agarose gel. DNA was depurinated, denatured and blotted in
lOxSSContoHybond N + membranes (Amersham) using a vacuum blotter (Millipore).
Acknowledgements
Wearegrateful toDr. B.J.C. Cornelissen from MogenInt. N.V. (Leiden) forproviding the
oligonucleotides used in this study. This project was supported by Monsanto (St. Louis,
USA) and by the Foundation for Biological Research (BION), which is subsidized by the
Netherlands Organisation for Scientific Research (NWO).
References
Brasier, CM. (1987) The dynamics of fungal speciation. In EvolutionaryBiology of the Fungi. (Rayner,
A.D.M. Brasier, C M . and More, D., eds.). University Press, Cambridge (UK),pp. 231-260.
Crute, LR. (1985) The genetic bases of relationships between microbial parasites and their hosts. In
Mechanisms of Resistance to Plant Diseases. (Fraser, R.S.S., ed.). Martinus Nijhoff/Dr. W. Junk
Publishers, Dordrecht/Boston/ Lancaster, pp. 80-142.
De Wit, P.J.G.M. (1977) A light and scanning-electronmicroscopic study of infection of tomato plants by
virulent and avirulent races of Cladosporiumfulvum. Neth.J. PlantPathol. 83, 109-122.
DeWit, P.J.G.M. (1987) Specificity of activeresistance mechanisms inplant-fungus interactions. InFungal
Infectionof Plants. (Pegg, G.F. and Ayres, P.G., eds.), University Press, Cambridge (UK),pp. 1-24.
De Wit, P.J.G.M. andFlach, W. (1979)Differential accumulation ofphytoalexinsintomatoleaves, butnot
infruits after inoculationwithvirulentandavirulentraces of Cladosporiumfulvum.Physiol.PlantPathol.
15, 257-267.
DeWit,P.J.G.M. andSpikman,G. (1982)Evidencefortheoccurrenceofrace-andcultivar-specificelicitors
ofnecrosis inintercellular fluids ofcompatible interactions of Cladosporiumfulvumandtomato.Physiol.
PlantPathol. 21, 1-11.
39
De Wit, P.J.G.M., Hofman, J.E., Velthuis, G.C.M, and Kuc, J.A. (1985) Isolationand characterization
ofanelicitorofnecrosisisolatedfromintercellularfluidsofcompatibleinteractionsofCladosporiumfiilvum
(syn.Fulviafiilva) and tomato. PlantPhysiol.77, 642-647.
Hitchin, F.E., Jenner, CE., Harper, S., Mansfield, J.W., Barber, C E . and Daniels, M.J. (1989)
Determinant of cultivar specific avirulence cloned from Pseudomonas syringaepv.phaseolicola race 3.
Physiol.Mol. PlantPathol. 34, 309-322.
Joosten, M.H.À.J., Hendrickx, L.J.M. and De Wit, P.J.G.M. (1990) Carbohydrate composition of
apoplastic fluids isolated from tomato leaves inoculated withvirulent or avirulent races of Cladosporium
fitlvum (syn. Fulviafiilva). Neth.J. PlantPathol.96, 103-112.
Kearney, B.,Ronald, P.C., Dahlbeck, D. and Staskawicz, B.J. (1988)Molecular basis for evasionofhost
plant defence inbacterial spot disease ofpepper. Nature332,541-543.
Keen, N.T., Tarnaki, S., Kobayashi, D., Gerhold, D., Stayton, M., Shen, H., Gold, S., Lorang, J.,
Thordal-Christensen, H., Dahlbeck, D. andStaskawicz,B. (1990)Bacteria expressingavirulencegene
Dproduceaspecific elicitorofthesoybeanhypersensitiveresponse.Mol.Plant-MicrobeInteract.3, 112121.
Kobayashi, D.Y., Tamaki, S.J. andKeen, N.T. (1989) Cloned avirulence genes from the tomatopathogen
Pseudomonas syringae pv. tomatoconfer cultivar specificity on soybean. Proc.Natl.Acad.Sei. USA 86,
157-161.
Kobayashi, D.Y., Tamaki, S.J. and Keen, N.T. (1990a) Molecular characterization of avirulence gene D
from Pseudomonas syringaepv. tomato.Mol. Plant-Microbe Interact.3, 94-102.
Kobayashi, D.Y., Tamaki, S.J., Trollinger, D.J., Gold, S. and Keen, N.T. (1990b) A gene from
Pseudomonassyringae pv.gtycinea withhomology toavirulencegeneD from P.S.pv. tomatobutdevoid
of the avirulence phenotype.Mol. Plant-Microbe Interact.3, 103-111.
Leong,S.A. andHolden,D.W. (1989)Moleculargeneticapproachestothestudyoffungalpathogenesis.Ann.
Rev. Phytopathol.27, 463-481.
Maniatis,T., Fritsch,E.F. andSambrook,J. (1982)MolecularCloning, aLaboratoryManual.ColdSpring
Harbor Laboratory Press, New York.
Napoli, C. andStaskawicz, B. (1987)Molecular characterization andnucleic acid sequence of an avirulence
gene from race 6 of Pseudomonas syringaepv. glycinea.J. Bacterid. 169,572-578.
Ronald, P.C. and Staskawicz, B.J. (1988) The avirulence gene avrBsl from Xanthomonascampestris pv.
vesicatoria encodes a 50kDprotein. Mol. Plant-Microbe Interact.1, 191-198.
Sanger, F., Nicklen, S. andCoulson, A.R. (1977)DNAsequencingwithchainterminationinhibitors.Proc.
Natl.Acad.Sei. USA 74, 5463-5467.
Scholtens-Toma, I.M.J. and De Wit, P.J.G.M. (1988) Purification and primary structure of a necrosisinducingpeptidefromtheapoplasticfluidsoftomatoinfectedwithCladosporiumfiilvum(syn.Fulviafiilva).
Physiol.Mol. PlantPathol.33,59-67.
Scholtens-Toma,I.M.J.,DeWit, G.J.M. andDeWit,P.J.G.M. (1989)Characterization ofapoplasticfluids
isolated from tomatolinesinoculated withnew races of Cladosporiumfiilvum. Neth. J. Plant Pathol. 95
(Suppl.1), 161-168.
Shintaku, M.H., Kluepfel, D.A., Yacoub, A. andPatil, S.S. (1989)Cloningandpartialcharacterization of
anavirulencegenefromrace 1 ofPseudomonassyringaepv.phaseolicola. Physiol.Mol.PlantPathol. 35,
313-322.
Staskawicz,B.J., Dahlbeck, D.andKeen,N.T. (1984)ClonedavirulencegeneofPseudomonassyringaepv.
glycineadetermines race-specific incompatibilityon Glycine max (L.) Merr. Proc. Natl.Acad. Sei.USA
81, 6024-6028.
Staskawicz, B.J., Dahlbeck, D., Keen, N.T. and Napoli, C. (1987) Molecular characterization of cloned
avirulence genes from race 0 and race 1ofPseudomonas syringaepv. glycinea.J. Bacterid. 169,57895794.
Swanson,J., Kearney, B.,Dahlbeck,D. andStaskawicz,B. (1988)ClonedavirulencegeneofXanthomonas
campestris pv. vesicatoria complements spontaneous race-change mutants. Mol. Plant-Microbe Interact.
1, 5-9.
Tamaki,S., Dahlbeck, D., Staskawicz, B.J. andKeen, N.T. (1988)Characterization andexpressionof two
avirulence genes from race 0 and race 1ofPseudomonas syringaepv.glycinea.J. Bacteriol. 170,48464854.
Tsukuda, T., Carleton, S., Fotheringham, S. and Holloman, W.K. (1988) Isolation and characterization
of an autonomously replicating sequence from Ustilago maydis. Mol. Cell.Biol. 8, 3703-3709.
40
VanEtten,H., Matthews, D., Matthews, P., Miao, V., Maloney, A. andStraney, D. (1989)A family of
genes for phytoalexin detoxification in theplantpathogenNectriahaematococca.In SignalMoleculesin
PlantsandPlant-MicrobeInteractions. (Lugtenberg,B.J.J., ed.),SpringerVerlag,Berlin/Heidelberg,pp.
219-228.
Vivian, A., Atherton, G.T., Bevan,J.R., Crute, LR., Mur, L.A.J, andTaylor,J.D. (1989)Isolationand
characterization of cloned DNA conferring specific avirulence in Pseudomonas syringaepv. pisi to pea
(Pisumsativum)cultivars,whichpossesstheresistanceallele,R2.Physiol. Mol.PlantPathol.34,335-344.
VonHeyne, G. (1986) Anew method for predicting signal sequence cleavage sites. NucleicAcids Res. 14,
4683-4690.
41
Chapter 3
Molecular analysis of the avirulence gene avr9
of Cladosporiumfulvum fully supports
the gene-for-gene hypothesis
GuidoF.J.M. van den Ackerveken, Jan A.L. van Kan and Pierre J.G.M. de Wit
adaptedfromThe PlantJournal 2 (1992), 359-366
Summary
The interaction between the fungal pathogen Cladosporiumfulvum and tomato is supposed
tohaveagene-for-genebasis.RacesofC.fulvumwhichhave'overcome' theresistancegene
Cf9of tomato,lack theavirulencegeneavr9whichencodes arace-specific peptide elicitor.
Races avirulent on tomato genotypes carrying the resistance gene Cf9produce the racespecificpeptideelicitor, whichinducesthehypersensitiveresponse(HR)onthosegenotypes.
The causal relationship between the presence of a functional avr9gene and avirulence on
tomato genotype Cf9 was demonstrated by cloning of the avi9 gene and subsequent
transformation of C.fulvum. Aracevirulent ontomatogenotype Cf9wasshown tobecome
avirulentbytransformation withtheclonedavr9 gene.Theseresultsclearly demonstratethat
theavr9 geneisresponsiblefor cultivar-specificity ontomatogenotype Cf9andfully support
thegene-for-gene hypothesis.Theavr9 geneisthefirstfungal avirulence genetobecloned.
43
Introduction
Thegene-for-gene hypothesis, whichwasproposed morethanfiftyyearsago, statesthat for
eachgeneconditioningrace-specifcresistanceinthehostplant,thereisacorrespondinggene
conditioning avirulence in a race of the pathogen (Flor, 1942). Gene-for-gene
complementarity occursmostfrequently inplant-pathogeninteractionsinvolvingobligateand
biotrophicpathogens whicharehighly specialized andhaveanarrow hostrange (Ellingboe,
1976;Heath, 1981;Keen, 1982).Resistanceinhostplantsagainstthesepathogensisusually
inherited as a monogenic trait and isbased on early recognition of thepathogen leading to
a quicklyinducedhypersensitiveresponse (HR)whichrestricts thepathogen tothe infection
site by local necrosis of a few host cells (DeWit, 1986; Keen, 1982; Keen, 1990; Pryor,
1987). The prevailing model in which recognition in gene-for-gene systems has been
described is theelicitor-receptor model which states that a specific receptor in theresistant
host plant interacts with a molecule of the pathogen, the so-called race-specific elicitor,
leading totheinductionofHRaccompanied byacascadeofotherdefence responses (Keen,
1982; Gabriel and Rolfe, 1990).
Inorder toovercomethisrace-specific resistance thepathogen mustevade recognition
of its race-specific elicitor by the plant. Within plant pathogenic bacteria the evasion of
recognition canbeachieved in threedifferent ways: (i) theavirulence genemaybemutated
in such a way that itsproduct isno longer responsible for theavirulentphenotype as is the
case with avrD of Pseudomonas syringaepv. glycinea(Kobayashi et al., 1990), (ii) the
avirulence gene may be disrupted by a transposon as is the case with the avirulence gene
avrBslofXanthomonascampestrispv.vesicatoria (Kearneyetal., 1988)and(iii)lossofthe
avirulencegeneasdescribed foravrAofP.syringaepv.glycinea, whichisthemost rigorous
way of evasion of recognition by thehost (Staskawicz etal., 1984).
The outcome of different race-cultivar combinations is usually scored phenotypically,
by development of disease symptoms (compatible interaction) or induction of HR
(incompatible interaction), and is thought to be determined by specific host-pathogen
recognition.Themechanismofrecognition,however,isnotknownyet.Receptorsinthehost
plasmamembranemightinteractwithrace-specific elicitors, followed by signaltransduction
and defence geneactivation (Dixon and Lamb, 1990;Keen and Dawson, 1992;Scheeland
Parker, 1990). The nature and origin of race-specific elicitors are poorly described. Many
bacterial avirulence genes have been cloned, but their primary products do not induce HR
(Knoop etal., 1991;Ronald and Staskawicz, 1988;Staskawicz etal., 1988). Theprimary
productofavrDofP.syringaepv.glycinea, however, releasesalipid-likemetaboliteacting
as a race-specific elicitor (Keen etal., 1990;Keen and Buzzell, 1991).
Ourworkonthefungal pathogen C.fiilvum,thecausalagentoftomatoleaf mould,has
provided new insight into the mechanism of induction of HR by fungal avirulence gene
products.Severalrace-specific elicitors,theputativeproductsofavirulencegenes,havebeen
identified whichinducenecrosisontomatocultivarswiththecorresponding resistancegenes
(De Wit and Spikman, 1982). One such race-specific peptide elicitor, theputative product
44
oftheavirulencegeneavr9,waspurified from intercellularfluids(IF)of C.Julvum-mfected
tomatoleaves(DeWitetal., 1985;Scholtens-TomaandDeWit, 1988).Thepurified peptide
elicitor specifically induces HR (observed as extensive necrosis) on tomatoplants carrying
the complementary resistance gene Cf9. The cDNA encoding the elicitor was isolated by
usingoligonucleotideprobesderived from theaminoacid sequenceof thepeptide (VanKan
et al., 1991). The gene encoding the elicitor was only detected in fungal races which are
avirulent on tomato genotypes carrying the resistance gene Cf9. Races virulent on Cf9
genotypes of tomatowhichhaveevaded recognitionby thehost, donotproducethepeptide
elicitor (Scholtens-Toma etal., 1989)asthey lack theentireavirulence geneavr9encoding
this elicitor (Van Kan etal., 1991).
In this study, we investigated the causal relationship between the presence of a
functional avt9 geneandavirulenceof racesofC.fiilvumontomatogenotype Cf9.Wehave
isolatedagenomicclonecontainingtheavirulencegeneavr9,andusedittotransform arace
ofC.fiilvum,virulentontomatogenotypeCf9. Alltransformants becameavirulentontomato
genotype Cf9, which clearly demonstrates that avi9 is a true avirulence gene, obeying the
gene-for-gene hypothesis.
Results
Isolation andsequence analysis oftheavr9gene
A cDNA insert (A9-5) encoding therace-specific elicitor (Van Kan et al., 1991)was used
as a probe to screen 100,000 recombinant plaques of a genomic library of race 5 of C.
fulvum. Several positive plaques were obtained and purified after a second round of
screening. A restriction map of the DNA region surrounding the avt9 open reading frame
(ORF)wasobtainedbySouthernanalysisofgenomicDNA(Figure la).Themapshowsthat
a 3.8 kbXhol fragment contains the ORF and approximately 3 kb of upstream sequences.
Southern analysis of purified phage DNA indicated that one phage isolate contained the
complete 3.8 kbXhol fragment carrying the avr9gene. TheXhol fragment was subcloned
and from the resulting plasmid (pCFl) a more detailed restriction map was constructed
(Figure lb).
Sequence analysis revealed an open reading frame of 63 amino acids which is
interrupted by one short intron of 59 bp (Figure 2). The 5' upstream region contains a
putative TATA-box and 2 imperfect repeats of 61 bp which are 77% identical. The 3'
untranslated region contains 8direct repeats of 17bp.
Transformation ofC. fulvum
In order to prove that the cloned gene is responsible for the induction of genotypeCf9specific resistance, transformation studies were performed. The avr9gene was transferred
torace2.4.5.9.11of C.fulvum (virulentontomatogenotypeÇfSf) byco-transformation with
45
6' I l
b '
X
E
X P H X H E
y
yy
yy
II
VA\
1 1 1
P V S
B P
II
y
I I
y
I
1'
I
, „.
I
3'
!ü
' ' M ' ty/V/V/V/i i!
A 6 T V
<
->
VVQ H
X
X
> <r-
Figure 1. Restriction map of the DNA region surrounding the avr9ORF. (a) Restrictionmap derived from
Southern blot analysis of genomic DNA digested with several restriction enzymes and probed with theavr9
cDNA (hatched bar), (b) Restriction map of the subcloned 3.8 kb Xfwl fragment containing the avr9 ORF
(indicatedby thehatched bar) and 3kb of thepromoter region, (c)Thesequencing strategy isindicatedby the
arrows. A, Jöwl; B,BamHI; E, EcóBI; G, BgM;H, ÄmdTlI;P, PstI;S, SaR;T, SstI;V, EcoRV; X, Xhol
pAN7-l, containing thehygromycin Bresistance marker (hph).Six selected hygromycin B
resistant (hygR) transformants were screened for integration of the av/9 geneby Southern
blot analysis. Five out of six transformants (Cl, C3 to C6) were found to have multiple
copies of the avi9 gene integrated into their genomes, whereas one transformant (C2) had
only obtained the hph marker (data not shown). Two transformants were selected and
analysed in detail, C2 (HygR, avr9")and C3 (HygR, avr9+).
Analysisoftransformants C2and C3
Southern blot analysis of transformants C2 and C3, and two wild type races of C. fulvum
indicated that transformant C3 obtained several copies of the cloned avr9gene (Figure 3,
lane 4). The wild type avr9 gene, present in race 5, is located on a larger restriction
fragment (lane2). Wild typerace 2.4.5.9.11,the recipient race (lane 1), and transformant
C2 (lane 3) both lack the avr9 gene.
Figure2-*. Sequenceof the Cladosporiumfulvum avirulence geneavr9(EMBLaccession number X60284).
The gene encodes a precursor protein of 63 amino acids (VanKan et al., 1991)and is interrupted by a short
59 bp intron (—). The mature avr9elicitor peptide of 28 amino acids is located at the carboxyl-terminus of
the precursor protein (bold). A putative TATA-box (TATAAGT) is located 40 bp upstream of the main
transcription start (H). The 5' upstream region contains 2 imperfect repeats of 61 bp (i it that are 77%
identical. The 3' untranscribed region contains 8 direct repeats of 17 bp (CGCATCGACTGCTCGGG,
repeatedly over- and underlined).
46
-982
TCTAGATCACTAAGGATTATGTCGTAACGGAACCGGAAGTGCTGCTCATAGTCTGCACGT
-922
AGCAATGGCAAAGGCACATCTGGTATCGAGAAGCGAGGTCGCACACGTGTAAGTGCTTGT
-862
CGAGTCGGCAACGAGGCTGGTGCAACTAGATCTTCTGGTTGTTGGTTCGCTTGATGACCT
-802
CTCTTAGCCTTCGGCTTCCTGCTAGGCTTGCCTCCTGCGGCTGCTCCGTCCGCCTTCTTC
-742
GCCTCACGCAGTTGTTTGTGTGACTTTCCGTGGCCTAGAÀGAAGTTGGCTCCGCGCCTCG
-682
CAGCACATTTTACTAGTGAGGATGTAGCTTTCCGCAATAGGGCGCGGTTTGTCACCAGAC
-622
CGTCCGCAGAGAGCCGCCTTGGCGGTATGCCTGAGGCAGCATCTGGTCACAAGATAGCAA
-562
GGCAACTGGGGAGAATAGGÀCGAAACAGAÀTAAGCCGACÀCCGCGCATCGCTCGAAACAG
-502
GGTGACAAGTGGGCATGGGAGCTCCTTACACCTTGTCCGATGTGCCGAGAGGTGACCAGT
-442
CTCTCTATCAAGGCTAGGGAGCGACAGTGGTGCGAGGCTGTCTGTTGTTCTGCGGTATTG
-382 TGTAGCACCGGTCGTAATATGCCTAGTGTCAGGTGTTTGÀTTCGGGAGGAAGATACTGCC
-322 ATTCTGTTTTACGCTCGTCGATAGGCGCCGTCGCAAACGCTATCGGGTCTTGGATAGGCG
-262 GGCAAGATATCTATCGGCTGTTGTGCTATTAATATGTATCGCAAGICAGCTAGGTATTAGA
-202
AACCTAGATAGCTAGTTGACTTCATATTGGCTAGATATCTACCTAGKGCAATACAACCTT
-142 GAAAFCAGCTÄGGTATAGCAAACTTCAGTAGCTAGCTAACTTGATATTAACTAGATATCTA
- 82 CCTAGFCCAGTAGATCCGGCCGAGAGAGAGATATACAGGTATAAGTAGACAGTAGATCTCT
- 22
TCTACTCTACTGTTACTAAATCAACACTACGAATTTCGAAAAATCCTCAAGCCTACTAAG
39
GTCTTTGTAGCCATAGTCATTTTAATAAGTCTATCGACTTACGTATCTATCATGAAGCTT
MKL
99 TCCCTCCTTAGCGTAGAGCTTGCTCTCCTAGTAAATATACTACGAGCACTACTATTACTA
S L L S V E L A L L
159 TTACAATATÀTACTAACTACATTATTTAGÀTTGCTACTACTCTCCCACTTTGCTGGGCAG
I A T T L P L C W A A
219 CTGCCCTCCCTGTAGGATTGGGAGTCGGGCTAGACTACTGTAACTCAAGTTGTACTAGGG
A L P V G L G V G L D Y C N S S C T R A
279 CCTTCGACTGTTTGGGTCAATGTGGCAGATGCGACTTTCATAAGCTACAATGTGTCCACT
F D C L G Q C G R C D F H K L Q C V H
339 AGAGGACTAGAGAGGAAGTGGAGAGAAGAGGAGGGGAGAGGTACGATAACTAGCGAGTAA
399 ATCGTACAGGTAGAAAGGGATAGTAAGCAGGCAGATAGACGGACGACGTTGCGACCTTAT
459 CCAACATAAGTCCTAGTCGTAACATTCGTTCATATTGAAGGCTTTTGGTCAATAGTTTCT
519
CAAATGTGCTGCGAGGCGCÀGAGCCAAGCCCCGTAGAGTCCGACGAAGGCTCCTTGTTGÀ
579 ATCTACAGTATCTACGAGTGTCAACATCGCCGAGCGTACTACCGCAGCGÀGCGTACTACC
639
GCATCGACTGCTCGGGCGCATCGACTGCTCGGGCGCATCGACTGCTCGGGCGCATCGACT
699 GCTCGGGCGCATCGACTGCTCGGGCGCATCGACTGCTCGGGCGCATCGACTGCTCGGGCG
759 GACGCATCGACTGCTCGGGCGCAATGCAGÀTCCGCCTCTTTGTTTCCGGGGGAGACCCCC
819 TAGACCCCCCTTTGGCCCCTCGCCGGGCGGGTGCGC 854
47
Table1. Pathogenicityoftransformants C2andC3
on cultivar MM (lacking Ç/"-resistance genes) and
genotype CJ9 (carrying resistance geneCJ9).
Transformant
cultivar
MM
C2
C3
C
C
12 345
genotype
CJ9
C
I
C: Compatible interaction; race virulent; fungal
growth and sporulation. I: Incompatible; race
avirulent; no fungal growth.
-23.1
-9.4
-6.6
-4.4
Figure 3. Southern blot analysis of genomic DNA
from wild type C. fitlvum races and transformants
digested with PstI and EcoRl. Lane 1, wild type race
2.4.5.9.11,therecipientrace,virulentongenotypeCJ9;
lane 2, wild type race 5, avirulent on genotype CJ9;
lane3&4, transformants C2andC3respectively.;lane
S,plasmidpCFl containingtheaw9 gene.Race5(lane
2) contains the wild type gene on a 5 kb Pstl/EcoSI
fragment, whereas transformant C3 (lane 4) contains
theintroduced avr9geneon a 3kb PstIfragment (due
to a PstI site in the polylinker of the vector). Race
2.4.5.9.11 (lane 1) and transformant C2 (lane 3) lack
the avr9gene.
2.3
•2.0
Pathogenicity of transformants C2 and C3 was tested on tomato seedlings of cultivar
Moneymaker (MM, susceptible to all races of C.fidvum,carrying no (^-resistancegenes)
andgenotypeCf9(anear-isogeniclineofcultivarMoneymakercarryingresistancegene Q9).
Symptoms were visible 14-20 days after inoculation (Table 1). Transformants C2 and C3
wereboth stillvirulentoncultivarMM, indicatingthatpathogenicity wasnotaffected bythe
transformation procedure. Thepathogenicity oftransformant C3ongenotype Cf9,however,
was completely changed from virulent to avirulent, whereas transformant C2 remained
virulent.
To confirm the observed difference in growth of the virulent transformant C2 and
avirulent transformant C3, total DNA was isolated from the leaves infected with both
transformants and thepresence of fungal DNA was semi-quantified by using a single copy
probe of C. fidvum (Figure 4). The signals on the Southern blot confirm the phenotypic
observations. Visually compatible interactions (C2/MM, C2IQ9 and C3/MM) gavea clear
48
C2
C3
hybridization signal, indicating that a
significant proportion from the DNA of the
MMCf9
MMCf9
infected leaves isof fungal origin. Theleaves
ofC3-infected genotype Cf9did notcontain a
detectable amount of fungal DNA, indicating
theabsence of fungal growth.
The cultivar-specificity of transformant
C3has changed from virulent to avirulent on
tomatogenotype Cf9.Accordingly theelicitor
encoded by avr9, which induces HR on Cf9
genotypesbutnotonMM, shouldbeproduced
by transformant C3. Intercellular fluid (IF)
isolated from both compatible interactions
»* •'.-.''.• '..liifl.'ji«:'ï.-°;- :•-•
C3/MM and C2/MM was screened for the
presence of the necrosis-inducing peptide
elicitor. Necrosis-inducing activity was
assayed by injection of IF into tomato
Figure 4. Southern blot analysis of total DNA
genotype Q9 and cultivar MM. IF from the
isolatedfrom infected leaves,probedwiththesingle
interaction C3/MM contained high necrosiscopy C.fitlvumgene ecpl (Van den Ackerveken et
al., 1993).Thehybridizationsignal (a 1kb Hindm
inducing activity, indicating that active av/9
fragment) reflects the quantity of fungal biomassin
elicitor is indeed produced by this
theinfected leaves. Hybridizationwasonlydetected
transformant. Transformant C2 did not
in those interactions which were visually scored as
compatible (C2/MM, CIICJ9and C3/MM) but not
produce any avr9 elicitor (Figure 5). The
in those scored as incompatible (C3/Q9).
other four transformants carrying avr9 (CI,
C4, C5and C6) alsoappeared tobe avirulent
ontomatogenotypeCf9andaccordinglyproducedactiveav/9elicitor(resultsnotpresented).
Theseresultsclearlyindicatethatthecloned avirulencegeneavi9isresponsiblefor the
specific induction of resistance on genotype Q9, and that no other genes or factors are
required torestore avirulence in arace previously virulent on tomatogenotype Q9.
Discussion
Recognitionofthepathogenbythehost,thebasisofresistanceandsusceptibilityin gene-forgenerelationships, ispoorly understood. Much effort has beenput intounraveling induced
defence responses in resistant plants, but the mechanisms of induction are still largely
unknown (Bowles, 1990; Lamb et al, 1989; Scheel and Parker, 1990). The genetics of
fungal avirulence and cultivar-specific resistance are generally well studied (DeWit, 1992;
Ellingboe, 1981; Flor, 1971; Ilott et al, 1989). In most cases, however, the direct or
indirectproducts of avirulence genes andresistance genes are not characterized.
49
MM
Cf9
Figure5. Necrosis-inducingactivityofintercellular fluid(IF)isolatedfrom C2-and C3-infected Moneymaker
(MM)leaves (compatibleinteractions).Theintensityofnecrosisingenotype CJ9is correlated withtheamount
ofinjectedpeptideelicitor,encodedbytheavr9gene.Necrosis-inducingactivitywasfoundinIFof C3-infected
leaves, indicating that theelicitorproduced by the transformant isactive.
Ourapproach toelucidatethemechanism ofrecognition in theC.fiilvum-tomatosystemhas
resulted in thecloning of avi9, thefirst fungal avirulence genetobe cloned.
Wewereabletoclonethecvr9geneby usingaminoacid sequence datafrom the racespecific elicitorpeptide(Scholtens-TomaandDeWit, 1988).Adegenerated oligonucleotide
probe was employed to isolate the avr9cDNA clone (Van Kan et al., 1991), which was
subsequently used to isolate theavi9 gene from a genomic library of race 5 of C. fulvum,
a fungal raceavirulent on tomatogenotype Ç/9. Sequenceanalysis revealed that theORFis
interrupted by a 59 bp intron, which is common for fungal genes (Gurr et al., 1987). The
possible role of the repeats in the 5' upstream region and putative TATA-box in the
regulation of geneexpression willbe studied bypromoter analysis. The significance of the
repeatspresent in the 3' end of the gene, but are absent in the mature mRNA, is unclear.
To demonstrate the causal relationship between a functional avr9gene encoding the
elicitorandavirulenceontomatogenotype Cf9, theclonedavr9 genewastransferred torace
2.4.5.9.11 of C.fiilvum,which lacks theavr9geneand can overcometheresistance genes
Q2, Cf4, Q5, Cf9 and Cfll of tomato. The cultivar-specificity of the transformants
containing the avi9 gene changed from virulent to avirulent on tomato genotype Cf9.The
elicitor encoded by the avi9 gene could be detected in IF of compatible interactions,
demonstrating significant expression of thetransferred avr9genein these transformants.
50
Manybacterialavirulencegeneshavebeenclonedbysupplementingrecipientvirulentstrains
with DNA from an avirulent strain, followed by screening for avirulence on appropriate
cultivars (Keen and Staskawicz, 1988; Staskawicz et al., 1984; Staskawicz et al., 1988;
Swanson etal., 1988). Theprimary products of these bacterial avirulence genes, however,
are not directly responsible for theinduction of HR in cultivars carrying the corresponding
resistance genes. The primary gene products of for instance avrA, avrB and avrC from
Pseudomonas syringae pv. gtycinea lackaleader peptide secretion sequenceand significant
hydrophobicdomains, suggestingthattheseproteinsarenotexcreted (KeenandStaskawicz,
1988).Theprimaryproduct of avrBs3 ofX. campestris pv. vesicatoria seems tobe soluble
butislocalized intracellularly (Knoopetal., 1991).Theproduct of theavrD genefrom P.
syringae pv.tomatohasbeen showntoactviathereleaseoflowmolecularweightlipid-like
elicitors originating from bacterial substrates (Keen et al., 1990). Our studies on theracespecific avr9elicitor peptide is the thefirstexampleof aprimary avirulence geneproduct,
which directly inducesHRin cultivars carrying the matching resistance gene.
Further supportfor theelicitor-receptor modelfor theavr9-Q9systemcanbeobtained
by cloning the Q9 resistance gene. It is tempting to speculate about the existence of a
receptor (theputativeproductoftheresistancegeneCf9) for theavr9 elicitorintomato.The
onlyexampleofaputativereceptorisdemonstratedinsoybeanfor theheptaglucosideelicitor
from Phytophthora megasperma (Cosioetal., 1990;Yoshikawaetal., 1983). Thiselicitor
was obtained by acid hydrolysis of fungal cell walls and is not involved in race-cultivar
specific resistance. Incontrast, theavr9peptideisanaturally occuringrace-specific elicitor
and its specific and rapid induction of HR on tomato genotype Cf9strongly suggests the
presenceofaspecific receptor. Theputativeplasmamembranereceptor for theavr9 peptide
elicitor mightbetheprimary product of the Ç/9resistance gene.
The biological function of the avr9 peptide for C. fiilvum, while growing on a
susceptibleplant,isstillunknown.Theavi9geneishighlyexpressed whenthefungus grows
intheplant (VanKanetal., 1991)suggestingthatthepeptideisofphysiological importance
for thepathogen. Races of C.fiilvumthathaveovercome resistance gene Q9 of tomatoby
loss of the avr9 gene are still pathogenic, indicating that the avr9 gene is dispensible.
Nevertheless, the Cf9 resistance gene still provides a good protection to C. fiilvum in
greenhouse-growntomatocultivars, suggestingthatthecompetitiveabilityofthesenewraces
is slightly reduced. To quantify the role of the avr9 elicitor in fitness, the presence or
absence of a functional avr9 gene in the population should be studied in competition
experiments.
Experimental procedures
Subculture ofC. fiilvum andpathogenicity tests
CladosporiumfiilvumCooke (syn. Fulvia Fulva (Cooke)Cif) was grown onpotato dextrose
agar (PDA) or in liquid B5-medium in shakecultures (DeWit and Flach, 1979).Theraces
51
used and their interaction with different tomato genotypes are presented in Table 2.
Pathogenicity wastested by inoculation of 14-day-oldtomato seedlings at thelower sideof
the cotyledon with a conidial suspension of lxlO5 spores ml"1 (Talbot et al., 1988).
Symptoms were apparent 14-20 days after inoculation. In a compatible interaction, fungal
growth was observed as sporulation on the leaf surface, whereas in the incompatible
interaction (resistant plant) these symptoms were notvisible.
Table2.
Races of C.fulvumusedinthisstudy. Theinteractionwithtwocultivarsoftomato,
productionof avr9elicitorandresponse of thecultivarstotheelicitorareindicated.
HRinduction
Cultivars
Race
Proposed
genotype
MM
CJ9
5
A2A4a5A9AU
C
I
2.4.5.9.11
a2a4a5a9all
C
C
avi9
MM
CJ9
yes
no
yes
no
no
no
elicitor
C: Compatibleinteraction;race virulent; fungal growthandsporulation.
I: Incompatible; raceavirulent;no fungalgrowth.
Isolations ofgenomicDNAandconstruction ofagenomic library
DNA was isolated from liquid-grown cultures as previously described (Van Kan et al.,
1991).Total DNA of race 5 of C.fulvum waspartially digested with Sau3AIand sized by
sucrosegradient centrifugation (10-40 %). Fragments of 10-25kb wereligated intoBamHI
digested lambdaEMBL3arms (Promega, Madison, WI). Following ligation, lambdaphage
DNAwaspackaged withaPackagenekit (Promega)andplated onEscherichia coliMB406.
Theprimary library (800,000 independent recombinants) was amplified and stored at 4°C.
Isolation oftheavr9gene,subcloning andsequencing
All DNA manipulations wereconducted essentially as described by Maniatis et al. (1982).
Theprimary genomiclibrary wasplated onE. coliMB406. Two replica filters were made
on Hybond N (Amersham Nederland BV, Houten, The Netherlands) from two plates
containing 100,000 independent recombinants, according to the instructions of the
manufacturer. The filters were hybridized in 5xSSC, 5xDenhardt's, 0.5% SDS and 100
jig.mr1 denatured calf thymus DNA at 65°C, using as probe a 400 bp avrdcDNA insert
(A9-5, Van Kan et al., 1991), which was radioactively labeled with [a-32P]dATP by the
randomprimer method (HodgsonandFisk, 1987).Selectedlambdacloneswereanalysedby
digestion with several restriction enzymes and Southern hybridization with theavr9cDNAprobe. A 3.8 kb Xhol fragment was subcloned in Sail digested plasmid pGEM 3Zf+
(Promega), resulting in plasmid pCFl. The 3' region of theXhol fragment, containing the
ORF and promoter region, was sequenced by using the dideoxy chain termination method
(Sanger etal., 1977)on double-stranded DNA (Promega).
52
Fungalprotoplastisolation andfungaltransformation
Protoplasts were isolated from 48-hour-old mycelium as decribed by Hading et al. (1988)
withthemodification thatliquidB5-mediumwasusedinstead ofpotatodextrosebroth. The
avr9genewas transferred to C. fulvum race 2.4.5.9.11 by co-transformation with pAN7-l
containing a hygromycin Bresistance marker (Punt et al., 1987). Two ng of pAN7-l and
4 ng of pCFl were mixed with 1x10s protoplasts and treated with PEG as described
previously (Oliver et al., 1987). The protoplast mixture was diluted with liquid complete
medium and plated in topagar onto stabilized regeneration medium (Hading et al., 1988).
Twenty hours after plating, a second top-layer was applied containing 800 /tg.ml"1
hygromycinB(finalconcentrationof 100jtg.ml1).After approximately4weekssinglespore
cultures were madeand individual transformants were analysed.
DNAanalysis oftransformants andinfected tomato leaves
DNAofliquid-grownculturesoftransformants andwildtyperacesofC.fulvumwasisolated
as previously described (Van Kan et al., 1991), digested with EcoKL and ft/I,
electrophoresed in 0.8% agarose, transferred toHybond-N+ byvacuum blotting (Millipore
BV, Etten Leur, The Netherlands) and probed with a random primed [a-32P]dATP-labeled
0.4 kb cv/9 cDNA fragment.
Fourteen days after inoculation of tomato seedlings, leaves were collected and freezedried. Genomic DNA was isolated from the leaves, digested with EcóKL and fli'rtdlll,
electrophoresed in0.8% agarose, transferred toHybond-N+byvacuumblottingandprobed
witharandomprimed [a-32P]dATP-labeled 1 kbHwdmfragment containingthesinglecopy
C.fulvumgeneencodingtheproteinECP1(JoostenandDeWit, 1988;VandenAckerveken
etal., 1993).
Intercellularfluidisolation andelicitoractivity bioassay.
For testingelicitoractivity, theintercellular fluid (IF)ofleaves from thepathogenicity tests
was isolated. Leaves were infiltrated with distilled water in vacuoand IF was isolated by
centrifugation as described previously (De Wit and Spikman, 1982). Fifty /xl of IF was
injected intoleavesofcultivarMM(noC^-genes)andgenotypeCf9(carryingresistancegene
Cf9). One to two days after injection, the elicitor activity was visible as necrosis (HR) on
genotype Cf9but not on cultivar MM.
Acknowledgements
Theinvestigationsweresupportedbythefoundation for BiologicalResearch (BION),which
is subsidized by theNetherlands Organization for Scientific Research (NWO).
53
References
Bowles, D J . (1990) Defense-related proteins inhigher plants.Annu. Rev. Biochem. 59, 873-907.
Cosio, E.G., Frey, T., Verduyn, R., Boom, J. van and Ebel, J. (1990) High affinity of a synthetic
heptaglucosideand fungal glucanphytoalexinelicitorstosoybeanmembranes. FEBSletters271,223-226.
De Wit, P.J.G.M. andFlach, W. (1979)Differential accumulationofphytoalexinsintomatoleaves, butnot
infruits after inoculationwithvirulentandavirulentracesof Cladosporiumfidvum.Physiol.PlantPathol.
15, 257-267.
DeWit, P.J.G.M. andSpikman, G. (1982)Evidencefor theoccurence ofrace-andcultivar-specific elicitors
of necrosis in intercellular fluids of compatible interactions between Cladosporiumfiilvumand tomato.
Physiol.PlantPathol.21, 1-11.
De Wit, P.J.G.M., Hofman, A.E., Velthuis, G.C.M, andKuc, J.A. (1985) Isolationand characterization
ofanelicitorofnecrosisisolatedfromintercellularfluidsofcompatibleinteractionsofCladosporiumfulvum
(syn. Fulviafulva) and tomato. PlantPhysiol.77, 642-647.
De Wit, P.J.G.M. (1986) Elicitationof active resistance mechanisms. InBiology andMolecularBiologyof
Plant-PathogenInteractions, NATOASISeriesVolumeHI (Bailey,J., ed.),Springer-Verlag, Berlin,pp.
149-169.
DeWit,P.J.G.M. (1992)Molecularcharacterization ofgene-for-gene systemsinplant-fungusinteractionsand
theapplication of avirulence genes incontrol of plantpathogens.Annu. Rev. Phytopathol.30, 391-418.
Dixon,R.A. andLamb, C.J. (1990)Molecular communication in interactionsbetweenplantsand microbial
pathogens.Annu. Rev. PlantPhysiol.PlantMol.Biol. 41,339-367.
Ellingboe,A.H. (1976)Geneticsofhost-parasiteinteractions.InEncyclopediaofPlantPhysiology, NS,Vol.4,
Physiological PlantPathology,(Heitefuss R. andWilliamsP.H., eds.) Springer-Verlag, Heidelberg, pp.
761-778.
Ellingboe, A.H. (1981) Changing concepts inhost-pathogen genetics.Annu. Rev. Phytopathol 19, 125-143.
Flor, H.H. (1942) Inheritance ofpathogenicity inMelampsora Uni. Phytopathology32, 653-669.
Flor, H.H. (1971) Current statusof thegene-for-gene concept. Annu. Rev. Phytopathol.9, 275-296.
Gabriel,D.W. andRolfe, B.G. (1990)Workingmodelsofspecific recognitioninplant-microbeinteractions.
Annu. Rev. Phytopathol. 28, 365-391.
Gurr, S.J., Unkles, S.E. and Kinghorn, J.R. (1987) The structure and organization of nuclear genes of
filamentous fungi.In GeneStructureinEukaryoticMicrobes,(Kinghorn,J.R., ed.).I.R.L.Press, Oxford,
pp. 93-139.
Harling, R., Kenyon,L., Lewis,B.G., Oliver,R.P., Turner, J.G. andCoddington, A. (1988) Conditions
for efficient isolationand regeneration of protoplasts fromFulviafulva. J. Phytopathol.122, 143-146.
Heath, M.C. (1981) Ageneralized concept of host-parasite specificity. Phytopathology 71, 1121-1123.
Hodgson, C.P. and Fisk, R.Z. (1987) Hybridization probe size control: optimized 'oligolabeling'.Nucleic
AcidsRes. 15,6295.
Hott,T.W., Hulbert, S.H. andMichelmore, R.W. (1989) Genetic analysis of thegene-for-gene interaction
between lettuce (Lactuca sativa)andBremialactucae.Phytopathology 79, 888-897.
Joosten, M.H.A.J. andDeWit, P.J.G.M. (1988)Isolation,purification andpreliminary characterization of
aproteinspecificforcompatibleCladosporiumfulvum(syn.Fulviafulva)-tomatointeractions.Physiol. Mol.
PlantPathol.33,241-253.
Kearney, B., Ronald,P.C.,Dahlbeck, D. andStaskawicz, B.J. (1988)Molecularbasisfor evasionofplant
host defence inbacterial spot disease of pepper. Nature332,541-543.
Keen, N.T. (1982) Specific recognition in gene-for-gene host-parasite systems.Adv. PlantPathol. 1, 35-82.
Keen,N.T. andStaskawicz, B. (1988)Hostrangedeterminantsinplantpathogensandsymbionts.Annu.Rev.
Microbiol.42, 421-440.
Keen,N.T. (1990)Gene-for-gene complementarity inplant-pathogeninteractions.Annu.Rev. Genet. 24,447463.
Keen, N.T., Tamaki, S., Kobayashi, D., Gerhold, D., Stayton, M., Shen, H., Gold, S., Lorang, J.,
Thordal-Christensen, H., Dahlbeck, D. andStaskawicz, B. (1990)Bacteria expressingavirulencegene
Dproducea specific elicitorof thesoybeanhypersensitivereaction. Mol.Plant-MicrobeInteract.3, 112121.
Keen,N.T. andBuzzell,R.I. (1991)New diseaseresistance genesinsoybeanagainstPseudomonas syringae
pvglycinea: evidencethatoneoftheminteractswithabacterial elicitor. Theor. Appl. Genet. 81,133-138.
Keen,N.T.andDawson,W.O. (1992)Pathogenavirulencegenesandelicitorsofplantdefense. InPlantGene
54
Research, Vol.8, Genesinvolvedinplant defense, (Boller, T. and Meins, G., eds.), Springer-Verlag,
Vienna/New York. Inpress.
Knoop, V., Staskawicz, B. and Bonas, U. (1991) The expression of the avirulence gene avrBs3 from
Xanthomonas campestris pv. vesicatoria isnotunder the controlofhrpgenesand isindependantofplant
factors. J. Bacteriol.173, 7142-7150.
Kobayashi, D.Y., Tamaki, S.J., Trollinger, D.J., Gold, S. and Keen, N.T. (1990) A gene from
Pseudomonas syringaepv.glycineawithhomologytoavirulencegeneDfrom P. s. pv.tomatobutdevoid
of theavirulence phenotype.Mol. Plant-Microbe Interact. 3, 103-111
Lamb, C.J., Lawton, M.A., Dron, M. and Dixon, R.Â. (1989) Signals and transduction mechanisms for
activation of plant defences against microbial attack. Cell56, 215-224.
Maniatis,T., Fritsch,E.F. andSambrook, J. (1982)MolecularCloning, aLaboratoryManual.ColdSpring
Harbor Laboratory Press, New York.
Oliver,R.P.,Roberts,I.N.,Harling,R., Kenyon,L.,Punt,P.J.,Dingemans,M.A.andVandenHondel,
C.A.M.J.J. (1987) Transformation of Fulvia fitlva, a fungal pathogen of tomato, to hygromycin B
resistance. Curr. Genet.12,231-233.
Pryor, T. (1987)Theorigin and structure of fungal diseaseresistance genesinplants. TrendsGenet.3, 157161.
Punt, P.J., Oliver, R.P., Dingemans, M.A., Pouwels, P.H. and Van den Hondel, C.A.MJ.J. (1987)
Transformation aïAspergillusbasedonthehygromycineBresistancemarker fromEscherichia coli.Gene
56, 117-124.
Ronald, P.C. and Staskawicz, B.J. (1988) The avirulence gene avrBsl from Xanthomonascampestris pv.
vesicatoria encodes a 50-kDprotein. Mol. Plant-Microbe Interact.1, 191-198.
Sanger, F., Nicklen, S. andCoulson, A.R. (1977)DNAsequencingwithchainterminating-inhibitors.Proc.
Natl.Acad.Sei. USA 74, 5463-5467.
Scheel, D. and Parker, J.E. (1990) Elicitor recognition and signal transduction in plant defense gene
activation.Z. Naturforsch.45c, 569-575.
Scholtens-Toma, I.M.J, and De Wit, P.J.G.M. (1988) Purification and primary structure of a necrosisinducingpeptidefromtheapoplasticfluidsoftomatoinfectedwithCladosporiumfulvum(syn. Fulviafulva).
Physiol.Mol. PlantPathol. 33, 59-67.
Scholtens-Toma,I.M.J.,DeWit,G.J.M. andDeWit,P.J.G.M. (1989)Characterizationofelicitoractivities
of apoplastic fluids isolated from tomato lines infected with new races of Cladosporiumfulvum. Neth. J.
PlantPathol. 95 (Suppl.l), 161-168.
Staskawicz,B.J.,Dahlbeck,D. andKeen,N.T.(1984)ClonedavirulencegeneofPseudomonassyringaepv.
glycineadetermines race-specific incompatibilityon Glycine max (L.) Merr. Proc. Natl.Acad. Sei. USA
81, 6024-6028.
Staskawicz, B., Bonas, U., Dahlbeck, D., Huynh, T., Kearney, B., Ronald, P. andWhalen, M. (1988)
Molecular determinants of specificity in plant-bacterial interactions. In Physiologyand Biochemistryof
Plant-MicrobialInteractions, (Keen,N.T.,Kosuge,T.andWalling,L.L., eds.).AmericanSocietyofPlant
Physiologists,Rockville, MD, pp. 124-130.
Swanson,J., Kearney, B.,Dahlbeck,D.andStaskawicz, B. (1988)ClonedavirulencegeneofXanthomonas
campestris pv. vesicatoria complements spontaneous race-change mutants. Mol. Plant-Microbe Interact.
1, 5-9.
Talbot, N.J., Coddington, A., Roberts, I.N. and Oliver, R.P. (1988) Diploid construction by protoplast
fusion inFulviafitlva(syn. Cladosporiumfulvum): geneticanalysisofanimperfect fungal plantpathogen.
Curr. Genet. 14, 567-572.
Van denAckerveken, G.F.J.M., VanKan,J.A.L., Joosten, M.H.A.J., Muisers, J.M., Verbakei, H.M.
andDe Wit, P.J.G.M. (1993) Characterization of twoputativepathogenicity genes of thefungal tomato
pathogen Cladosporiumfulvum.Mol.Plant-Microbe Interact.6, 210-215.
Van Kan, J.A.L., Van den Ackerveken, G.F.J.M. and De Wit, P.J.G.M. (1991) Cloning and
characterization of cDNA of avirulence gene aw9 of the fungal pathogen Cladosporiumfulvum, causal
agent of tomatoleaf mold. Mol. Plant-Microbe Interact.4, 52-59.
Yoshikawa,M., Keen,N.T. andWang,M.-C.(1983)Areceptor onsoybeanmembranesfor afungal elicitor
ofphytoalexin accumulation. PlantPhysiol.73, 497-506.
55
Chapter 4
Disruption of theavirulence geneavt9in Cladosporium
fulvumcauses virulence ontomato genotypes withthe
complementary resistance gene Cf9
RolandMarmeisse, GuidoF.J.M. van denAckerveken, TheoGoosen,
PierreJ.G.M. deWit andHenk W.J. van denBroek
adaptedfromMolecular Plant-Microbe Interactions 6 (1993),Inpress
Summary
In order to study the function of the avirulence gene avr9 of the tomato pathogen
Cladosporium fulvum we developed procedures for gene disruption experiments in two
different racesofthefungus bothavirulentontomatogenotypescarrying theresistancegene
Cf9. For this purpose we selected uridine auxotrophic strains amongst fluoroorotic acid
resistant mutants. These mutants were transformed with a plasmid containing the avr9
genomic region in which the open reading frame was replaced by thepyrG gene from
Aspergillus nidulans. For eachof thetworaces used weselected onetransformant in which
theentireavr9 coding sequencewasdeletedasaresultofagenereplacement event. Thetwo
transformants were able to successfully infect Cf9tomato genotype unlike their wild-type
avr9+ parents which induced hypersensitive responses on this genotype. We also
demonstrated that these two transformants no longer produce thenecrosis-inducing elicitor
peptidespecifically interacting with Cf9tomatoplants.Theseresultsconfirm thatthecloned
avr9sequence (and therefore theAVR9peptide)isfully responsible for avirulence in wildtype avr9+ races of thefungus. These results alsoindicate that this geneis dispensable for
normal vegetative growth and pathogenicity at least in a monocyclicprocess. These results
arediscussedinrelationtothepossibleoriginofavr9"strainsandthelong-standingresistance
toward C.fiilvumoffered by thetomato CJ9resistance gene.
57
Introduction
The gene-for-gene type of interaction between a pathogen and its host plant postulates the
existence of anavirulence genein thegenomeof thepathogen whoseproduct interacts with
the product of the plant resistance gene to give an incompatible interaction usually
characterized by the development of a hypersensitive response (Flor, 1942). This type of
interaction can be best demonstrated by performing crosses between avirulent and virulent
strains of the pathogen or resistant and susceptible host cultivars and observing the
segregation of these characters in the progeny. For pathogenic fungi this is only feasible
when sexual or parasexual crosses can easily be performed. This is, for instance, the case
for the ascomycete Magnaporthe grisea (Valent et al., 1991) or the oomyceteBremia
lactucae (Michelmoreet al., 1984),pathogens of rice and lettuce, respectively.
For our model species Cladosporiumfulvum, a biotrophic fungal pathogen of tomato,
no sexual stage has ever been described and the existence of six avirulence genes hasbeen
postulatedbasedonthedifferential interactionsofwild-typefungal racesontomatocultivars
knowntopossessdifferent resistancegenestowardthisfungus. Thegeneticbasisofthegenefor-geneinteractionbetweentomatoandC.fulvumhastherefore onlybeenestablished clearly
for the host plant tomato for which the different Çf resistance genes have now been
positioned on the genetic map (Jones et al., 1991;Van der Beek et al., 1992). It has also
been shown that avirulent races of the fungus produce proteinaceous elicitors inducing a
necrotic reaction only wheninjected intotheleaves of thecorresponding resistant plant (De
Wit and Spikman, 1982). Because of the specificities of these elicitors, whose necrosisinducing activities are reminiscent of the hypersensitive responses observed following the
infection of aresistantplantbyan avirulent fungal race, itwas suggested that they could be
the primary products of the avirulence genes. The necrosis-inducing peptide produced by
fungal races avirulent on tomato genotype Cf9(avr9+ strains) was purified and sequenced
(Scholtens-Toma and De Wit, 1988). The sequence data from this peptide were used to
design oligonucleotides which were employed to isolate the cDNA and genomic clones
encodingtheAVR9peptide(VanKanetal., 1991;VandenAckervekenetal., 1992).These
authors showed that the genomic coding region is interrupted by one intron and that the
primary geneproduct is a 63 amino-acid-longpeptide, from which only the 28 C-terminal
parthasbeenfoundininfected tomatoleaves.Southernhybridizationexperimentsestablished
thepresence of this gene in the genomes of all the avr9+ fungal races tested but failed to
detect homologous sequences in thegenomes of the tworaces virulent on tomato genotype
Cf9 tested, suggesting that these isolates could have arisen following a deletion in this
genomic region. Karyotype analysis indicated that races 2.4.9.11 and 2.4.5.9.11, both
lacking cvr9, have undergone an estimated deletion of 500 kb from chromosome band 2,
suggestingthatthesevirulentracesmighthavelostfunctional chromosomalDNAinaddition
to theav/9 gene (Talbot etal., 1991).Transcripts of the avr9genecould only be detected
insusceptiblehostplantinfected byavr9+ races. Thegenomicsequencecouldbeintroduced
58
by transformation and expressed in avr9"strains, changing them from virulent to avirulent
on tomato genotype Cf9. These results clearly demonstrated that the AVR9 peptide is
sufficient for triggering thehypersensitiveresponse incombination with theplant resistance
gene Cf9supporting therefore thegene-for-gene hypothesis.
In this paper we present evidence for the disruption of the avr9 gene by gene
replacement in two different races of thefilamentousplant pathogenic fungus C. fulvum.
Theseexperiments werecarried out todefinitely confirm thefunction of thisgene, notonly
asanavirulencegenebutalsoasapotentialpathogenicityfactor, inavr9+ wild-typeisolates.
Theresultsobtained giveconclusiveevidencefor themechanismof avirulenceinthetomato
-C.fitlvuminteractionandraiseinterestingquestionsontheroleofavr9 inpathogenicityand
on the evolution of races lacking thisgene.
Results
Isolation andcharacterization of uridine auxotrophic mutants
In order toimprove the methodology for thegenetic manipulation of C.fulvum wetried to
isolate mutants which could be used for transformation with either homologous or
heterologous genes. For each of the two wild-type races (races 4 and 5, both containing
avirulence geneavr9),weselected 20mutantsresistant toFOA following UVmutagenesis.
Among them one from race 4and four from race 5were uridine auxotrophs. It was shown
that in yeast (Boeke et al., 1984) the mutants isolated using this method could result from
mutationsineitherthegeneencodingtheorotidine-5'-phosphatedecarboxylase (OMPdecase)
or the gene coding for the orotate pyrophosphoribosyl transferase (OPRTase). To decide
between these two possibilities, we transformed each of the C.fulvum mutants with the
clonedpyrA genefrom Aspergillus nigercoding for the OMPdecase (Goosen etal., 1987).
Three of the race 5 mutants and the race 4 one could be efficiently complemented by this
heterologous geneand Southern blotanalyses indeed confirmed thepresenceof theA. niger
geneintegrated in theirgenomes (datanot shown).From thisresult weconcluded thatthese
four mutantshaveamutationinthestructural genefor theOMPdecaseandthatthefifth one
whichwecouldnotcomplementmighthaveamutationinthegenecoding for theOPRTase.
Mutants5.4 from race5and4.2from race4wereusedasrecipientstrainsfor thedisruption
of theavr9gene.
Construction ofaplasmidfor thedisruption oftheavr9geneand transformations
Wedesigned aplasmid allowing asingle stepgenereplacement at theavr9locus. Theavr9
coding region present on thelambda clonedescribed by Van den Ackerveken et al. (1992)
was subcloned asa6kbEcóBI-Sall fragment intopBluescript (Stratagene). Aninternal750
bpHindTLI-Xholfragment containing thewholecoding sequence wasremoved and replaced
by the.4. nidulanspyrG genepresent ona 1.5 kbDNA fragment. Figure 1givesdetailsof
59
the construction of this pCF112 plasmid. The presence of thepyrG gene, coding for an
OMPdecase, allows thetransformation of C.fulvwn pyr"mutantstoprototrophy. Upstream
and downstream of thispyrG geneare 2.5 and 2.7 kb of C.fidvumDNA sequences which
could allow the replacement of the avr9gene following homologous recombination events
as illustrated in Figure 1.
pCF112 was linearised with EcdSI, which cuts once in thepolylinker of pBluescript,
andusedtotransform protoplastsfrom thepyr"mutantsofraces4and5ofC.fidvum.Fortythreeprototrophictransformants from race4and52from race5werepurified before further
analyses.
E
plasmidpCF112 I
WILDTYPE
GENOMIC
SEQUENCE
TRANSFORMANT
"™^*
S B B H S S X
I
I
I 1.L.U—•<J
pyiG
H
I
E
I
^
Figure 1. Construction of the plasmid pCFlll used to disrupt the avr9 gene and events leading to the
replacement of thegenefollowing transformation. Theavr9genomic clonewasobtained as a 6kbEcoRl-SaK
fragment cloned into the EcoRl-Xhol sites of pBluescript. A Hindlll site close to the EcoRl site was first
removed by partial digestion, treatment with klenow polymerase and religation of the linear molecule. The
plasmid was then reopened at the Hindlll site which covers the second and third codons of the avr9 open
reading frame. After making the site blunt with klenow polymerase, the plasmid was digested with Xhol to
remove thewholeof theavr9codingsequence. TheA. nidulanspyrG genewas obtained from plasmidpJRIS
(Oakley etal., 1987)asa 1.5 kbNdel(madebluntwithklenowpolymerase)-X7ioIfragment. Ligationof these
two molecules resulted in pCFl12which has the avr9sequences replaced by thepyrG gene and which hasa
singleEcoRl siteused to linearise theplasmid before transformation.
Only the restriction sites relevant to this work are indicated: B=Bg/II, E=£coRI, H=H/'ndIII, S=Saä and
X=Xhol. The underlined BgRl fragment was used as the probe in Fig. 2C and 3.
60
Characterization ofavrÇ' transformants
GenomicDNAsextractedfromthetransformants wereinitiallysubjected todotblotanalysis
using the 750 bp avr9 coding sequence, removed from the transforming plasmid, as the
hybridizationprobe.TheDNAfrom threetransformants (onefrom race4andtwofromrace
5)failed togiveanyhybridization signal(resultsnotshown)andonesuchtransformantfrom
eachrace(race4-A43andrace5-B51)wassubjected todetailedSouthernblotanalysesalong
withtheDNAsfrom thewild-typerace5andfrom twoothertransformants (race4-A11 and
race 5-B11)with ectopic integrations of theplasmid. Thehybridization to the labeledavr9
coding sequencewasrepeated onHindlEdigested genomicDNAs(Figure2,panelA).This
probe hybridized to a2.3 kbDNA fragment of thewild-typerace 5and transformants All
and Bll but did not hybridize to the DNAs from transformants A43 and B51 (note that a
polymorphism exists between races 4 and 5, seen as a slightly larger hybridizing fragment
intherace4-A11 transformant). [a-32P]dATP-labeledpBluescript sequencesonlyhybridized
to the DNAs of transformants All and Bll which originated from non-homologous
integrations of theplasmid (Figure2, panelB).When a 800bpBgM fragment upstream of
the avi9gene (Figure 1)was used toprobe Saildigested DNAs, it identified a 10kbDNA
fragment in thegenome of theuntransformed race 4 and one additional larger fragment in
transformants All and Bll (Figure 2, panel C; note that also in this case a DNA
polymorphism between thetransformants isvisible).Bycontrast, this sameprobe identified
a unique 3 kb fragment in the genomes of tranformants A43and B51 (Figure 2, panel C);
the size of this last fragment is in agreement with the insertion of thepyrG gene, which
contains two internal Satt, sites.
From these analyses we concluded that transformants A43 and B51 have both arisen
from asinglestepgenereplacement attheavr9locuswhichhasbeenexchangedby ÛiepyrG
genepresent on the transforming pCF112plasmid. No other integrations of theplasmid at
ectopic sites seemed to have occurred in those two transformants. Transformants Al1and
Bll, containing several ectopicintegrations of pCF112, wereused as controls to assess the
effect of transformation on pathogenicity.
Pathogenicity andvirulence ofthe transformants
Spores from thewild-typeraces4and 5, the corresponding pyr mutants, the transformants
withadisruptedcopyoftheavr9 geneandthecontroltransformants withectopicintegrations
ofpCF112,wereusedtoinoculatetwo-week-old tomatoseedlings.Foreachfungal strainan
averageof 20plantsfrom cultivarMoneymaker (MM, Cf9")andthenear-isogenicMMline
containingtheresistancegeneCf9wereused. Thesymptomswererecorded twoweeks after
inoculation.With theexceptionofthetwopyr strains, allother strainscouldinfect theMM
plantswhich wasjudged bytheappearanceof amat of sporulating myceliumontheleaves.
Tomato genotype Cf9could only be successfully colonized by transformants A43 andB51
which no longer contain an intact copy of the avr9gene. All other fungal strains failed to
infect thetomatogenotypeCf9.
61
m **
<
<
ta 4 * 4
«o
S «ura
*••
s
(0
k.
s
«0
h.
s
k-
W »- •r- «*,_
10
mi S S
<i m
m 4
in
m 0 «
0
u«
» S
w
k-
1
?
94* • •»
66.
*<
C
D <
IO
* CQ*
S 8 SSS
S S Ss s
1
2319466- —
4423-
44-
Enz.
probe
<S
i
231-
2320-"
*2 »- *" «S 2 «> 1 -
•
~
~
;
•
•
«r
06-
4
Hlnd Ut
avr9
HlmtHI
ptasmid
Sal I
pavr 9
Figure2. Southernblotanalysesdemonstratingthattransformants A43andBS1 resultfrom agenereplacement
at theavr9locus.Therestriction enzymesused todigestthegenomicDNAsandtheprobesused are indicated
below the autoradiographs. The hybridization patterns of transformants A43 and BS1 are compared to the
hybridization patterns given by the wild-type race S and two other transformants (All and Bll) which
originated from ectopic integrations of thepCF112plasmid. Size markers are inkb.
To confirm theseobservations, DNAs wereextracted from similar amountsof infected leaf
material, digested with Sail, and subjected to Southern blot analysis using the BgM-BgM
sequence,presentupstream oftheavi9gene,asaprobe. Hybridizing DNAfragments could
easily be detected from infected leaf material 14 days after inoculation (Figure 3). The
presence of a 3 kb fragment clearly identified transformants A43 and B51 (compare with
Figure 2C). The transformants with ectopic integrations contain several copies of theavi9
promoterandthehybridization signalistherefore moreintense. Noobviousdifference could
beseen intheintensity of thehybridization signals obtained between thedisruption mutants
andwild-typeC.fulvum, indicatingthatsimilaramountsoffungalbiomassarepresentinthe
leavesinfected byeithertheavr9+ strains (MMplants) or theavr9"transformants (MMand
genotype CJ9).
The ability of the avr"transformants to infect tomato genotype Q9 was not simply a
consequence of the transformation since transformants All and Bll which have ectopic
integrationsofthepCFl12plasmidretainedthehostspecificities ofthewild-typeraces4and
5, being able to infect MMplants only.
Theproduction of anactiveelicitorpeptidewasproven byinjecting intercellular fluids
collected from thedifferent interactionsintotheleavesofMMandtomatogenotype Cf9.As
summarised inTable 1, anecroticresponsecould onlybedetected whenintercellular fluids
62
A : MM
S
%
*
m
a
u
4o
10
w
g
u
(0
ui
u
l
o
u
a
B : Cf9
<
CD
4 4
m
o
o
m
U
S
CD
u
CD
CM
m
m
't
ai
u
ce
O
X
S 4
in
<
CD
U
a>
ca
CD
9
U
CD
<*
Ifl
<
<
in
*
4
U
(0
a>
u
CD
m
CD
u
m
s
u
CD
CD
.23.1.
m
•
.9.4.
.6.6.
.4.4.
.23.
2Ä
Figure 3. Southern blot analysis of DNAs extracted from the leaves of infected tomato plants 14 days after
inoculation. Genomic DNAs (from bothplant and fungal origins) were digested with Sail and the membrane
wasprobed withtheBglR fragment upstream of theavr9gene (asinFig. 2C). Each lanerepresents a different
fungal strain used to inoculate Moneymaker (A) and CJ9 (B) seedlings. The presence of a hybridizing DNA
fragment indicates fungal growthand therefore acompatibleinteraction, thesizeof the fragment distinguishes
between avr9+ strains (10kb fragment) and avr9"(avr9::pyrG) mutants (3kb fragment).
from avr9+/MM interactions were injected into the leaves of genotype Q9, while
transformants A43and B51did notproduce any elicitor. Necrotic reactions were observed
evenwhenintercellularfluidsofavr9+/MMinteractionswerediluted 10fold, whileundiluted
samples of avr97MM, Cf9interactions failed toinduce necrosis.
These results clearly demonstrate that the avt9 gene, present in the races 4 and 5 of
C.fiilvum,istheonlygeneticfactor responsible for theirinability toinfect tomatogenotype
Cf9. TheyalsoindicatethattheAVR9peptideistheonlycompoundfrom crudeintercellular
fluids which induces necrosis in tomatogenotypeCf9.
63
Table 1. Characteristics of the C.fltlvumstrains used in this study
Cultivar
MM Ç/9
Strain
Description
race 5
wild-type isolate
race 5-5.4
pyr"mutant induced in race 5
by UV mutagenesis
race 5-B51
race 5-5.4 avr9: :pyrG
insertion mutant
race 5-B11
race 5-5.4 withthe avr9::pyrG
construct at ectopic sites
race 4
wild-typeisolate
race 4-4.2
pyr"mutant induced in race 4
by UV mutagenesis
race 4-A43
race 4-4.2avr9::pyrG
insertion mutant
race 4-All
race 4-4.2 with the avr9: :pyrG
construct at ectopic sites
productionof
AVR9elicitor
Theoutcomeoftheinteractionofdifferent strainsofC.fiilvumwiththetomatocultivarsMoneymaker(MM,
Cf9") and Cf9+ is indicated; C: compatible interaction (race virulent), I: incompatible interaction (race
avirulent).Theproductionof theAVR9elicitorwasassayed by testing,in Cj9plants, thenecrosis activity
of extracellular fluids recovered from infected MM plants two weeks after their inoculation. *: non
pathogenic auxotrophicmutants.
Discussion
In thispaper wedemonstrate that a single mutation generated by gene disruption can alter
the cultivar-specificity of the fungal pathogen C.fiilvum.In the case of the avr9gene we
recovered threetransformants which hadresulted from agenereplacement outof the95we
analyzed, which corresponds to a frequency in the range of 1-5%. The data presented
confirm that gene disruption by gene replacement is possible in the filamentous plant
pathogenic fungus C. fiilvum, and provides new possibilities for the molecular genetic
analysisofavirulenceandpathogenicity.Althoughgenedisruptionoccursatalow frequency,
the use of this method could be extended in the future to assess the function of other C.
fiilvumgeneswhoseproductsarefound inintercellular fluids from infected plantsandwhich
are putative pathogenicity factors (Joosten and De Wit, 1988; Van den Ackerveken et ah,
64
1993). Successful gene replacement has been described for a number of other plant
pathogenic fungi including Ustilago maydis(Fotheringham and Holloman, 1989; Kronstad
etal., 1989),Cochlioboluscarbonum(Scott-Craigetal., 1990),Nectriahaematococca(Stahl
and Schäfer, 1992), Magnaporthe grisea(Sweigard et al., 1992) and Gibberellapulicaris
(Honnef al., 1992).
In this study we first selected uracil auxotrophic mutants and showed that the initial
selection for fluoroorotic acid resistance can also be applied to C. fiilvum. The resistant
mutants fell into three categories: theprototrophic ones, the auxotrophic ones which could
becomplemented following transformation byfungal OMPdecasegenesandathirdcategory
of auxotrophic mutant which could not be complemented by OMPdecase genes and which
mayhaveamutationintheOPRTasegene.Thesemutantsrepresentaninterestingalternative
totheuseofbacterial antibioticresistance genesas selection markers in this species (Oliver
et al., 1987).However, thesepyr mutations clearly interfere with pathogenicity on tomato
which was not thecase for other auxotrophic mutants selected by Talbot etal. (1988).
At the start of this study, we knew from the experiments performed by Van den
Ackerveken et al. (1992) that the transfer of the avr9 sequence to a virulent race of
C. fidvum was sufficient to make it avirulent on tomato genotype Cf9. From the results
presented in thispaper it becomesclear thatin thetwowild-typeraces westudied, the avr9
gene is indeed the only genetic factor responsible for the induction of an incompatible
interaction on tomato genotype Cf9. This definitively establishes the genetic basis of the
gene-for-gene relationshipinthecaseoftheavr9ICf9interactionandalsoconfirms thedirect
involvement of the AVR9peptide in the development of the hypersensitive response. The
avr9geneistheonly avirulencegenecloned sofar for which theprocessed protein product
hasbeen showntobe directly responsible for theinduction of thehypersensitive response.
This has not been established or does not seem to be the case for most of the bacterial
avirulence genes which have been characterized; there is no evidence for their translation
products to be excreted and small molecules, generated by the avirulence gene products,
seem to act as race-specific signals as in the case of the avrD gene from Pseudomonas
syringae pv. tomato (Keen et al., 1990). It remains to beproven whether other C. fiilvum
race-specific elicitorsdescribedbyDeWitandSpikman(1982)arethedirectavirulencegene
products aswell.
Thesimpleorganizationoftheavr9 locuswhichconsistsofa singleopenreading frame
supports the idea that avr9" field isolates could have arisen from a single mutation in a
previously avr9+ geneticbackground. Datafrom VanKanetal. (1991)indeed showed that
in twofieldisolates virulent ontomatogenotype Cf9theavi9 genewas deleted. Karyotype
analysisofdifferent racesofC.fiilvumrevealedthatwild-typeraces2.4.9.11and2.4.5.9.11,
both lacking avr9,haveadeletion of 500kb. (Talbot etal. 1991).Recent dataindicatethat
the deletion is somewhat smaller (R. Oliver, personal communication; G. Van den
Ackerveken, unpublished). A similar situation has recently been reported by Panaccioneet
al. (1992)whoshowed that the TOX2 locusof themaizepathogen Cochliobolus carbonum,
65
which is involved in the first step of the synthesis of the cultivar specific HC-toxin, is not
found in the genomes of toxin nonproducing (tox) strains. In a recent study, Miao etal.
(1991)demonstrated thatthePda6genefromNectriahaematococca (encodingacytochrome
P-450 enzyme responsible for the detoxification of the pea phytoalexin pisatin) was on a
dispensableBchromosomewhichdidnotsegregateduringmeiosis,givingmorepda"progeny
than expected.
Since its introduction in tomato breeding lines, the Cf9resistance gene has efficiently
contributed to theprotection of thiscroptoward C.fiilvum.Although four racesvirulent on
tomatogenotype Cf9havebeen described (Lindhout etal., 1989),they havenot yet caused
any serious problem. This contrasts with several of the other Çfgenes which have only
offered a very temporary protection due to a rapid spread of new virulent fungal races
(Hubbeling, 1978).This could beexplained by a very low mutation rate at the avi9locus,
limitingtheappearanceofinfection centres. Analternativeexplanationcouldbethattheavr9
gene is also acting as an indispensable pathogenicity factor in avr9+ races. This does not
seem to be the case since the two transformants we analyzed did not display any visible
phenotypicalteration compared totheirwild-typeparents whengrowinginvitro.In addition
theycouldinfect equally wellboth Cf9+ andCf9"tomatoplants.Weshould, however, stress
that the experimental conditions we used (infection of young seedlings, high spore
concentrations, high humidity and constant temperature) are highly favorable to the fungus
and may not reflect the environmental conditions usually found in the field. More
experiments are needed to assess thecompetitiveability of theavr9"strains with respect to
spore formation, spore viability, spore dispersal and all other traits which might affect the
spread of a fungal strain under natural conditions. However, caution should be taken in
epidemiological studies using this disruption mutant as it may become a successful new
virulentraceontomatocultivars containing thehitherto successful resistance gene Cf9.The
avr9 disruption mutantsaredifferent from wild-typeavr9"racesastheyonlyhaveonesingle
gene deletion. Wild-type avr9" races lack besides the avr9 gene a significant stretch of
flanking sequences. Possibly other genes are present on this deleted DNA that might be
involved infitnessor pathogenicity of C.fiilvum,which mightexplain why wild-typeavr9"
races are not a serious problem in the tomatogrowing areas.
Experimental procedures
Fungalstrains,mutagenesis andgrowth conditions
The C. fiilvum (syn. Fulvia fiilva) strains used in this study are listed in Table 1. The
pyrimidine auxotrophic mutants were selected among FOA resistant mutants as originally
described for yeast by Boeke etal. (1984). Conidial suspensions in water were plated onto
minimalmediumsupplemented with 10mMofuridineand2.5(race5)or3(race4)mg.ml"1
FOA; after UVirradiation togive30-40% survival, theplateswerekeptinthedark for two
66
weeks.FOAresistantmutantswerepurified ontofresh FOAplatesandsubsequentlyanalyzed
for uridineauxotrophy.
All strains were routinely maintained onpotato dextrose agar medium (PDA: Merck)
whichwasoverlaid withcellophanemembranestorecoverthemyceliafor DNAextractions.
CzapekDox(Oxoid)wastheminimalmediumused toselectprototrophictransformants. All
cultures were grown at 22°C.
Plantmaterial and inoculations
ThetomatocultivarMoneymaker (MM, susceptible to allraces of C.fiilvum)and thenearisogenic MM line containing the Cf9 resistance gene were used in this study. Seedling
inoculations were carried out according to Talbot et al. (1988). Intercellular (apoplastic)
fluids wereisolatedasdescribed byDeWitandSpikman(1982)andinjected intotheleaves
of 6to 8week-old tomatoplantsas described by the sameauthors.
Fungal transformation
Protoplasts of C. fulvum were prepared according to Harling et al. (1988) with the
modifications of Van den Ackerveken et al. (1992). Transformations were performed
according toOliveretal.(1987)withthefollowing modifications: to 107ormoreprotoplasts
in 100M1ofMTC (1MMgS04, 10mMTrispH7.5, 10mMCaCl2)upto20figofplasmid
DNA in 10/xlH 2 0 wereadded. After 15minincubationatroom temperature, 1mlofPTC
(20% polyethylene glycol 6000, 10 mM Tris pH 7.5, 10 mM CaCl2) was added and the
mixturewasleft atroomtemperaturefor anadditional 15min.Thesolutionwasthendiluted
in 5 ml of liquid minimal medium with 0.8 M sucrose and plated in a top agar layer onto
plates containing the same selective medium.
DNA manipulations
Standard DNA techniques used were as described by Sambrook et al. (1989). TheE. coli
strainDH5a (F"08Od7acZM15(/ßcZYA-argF)U169endAl recAlfedR17(rk-mk+)deoRthi1 supEM gyrA96 relAl) was used to propagate plasmids. C. fulvum genomic DNA was
extracted from freeze-dried mycelia using themethod described by Van Kan et al. (1991).
DNA from infected leaves was extracted as described by Van der Beek et al. (1992).
Following restriction-enzyme digests, DNA fragments were separated in 0.7% agarose gels
and transferred by Southern blotting onto Hybond N + membranes using 0.4 M NaOH
according to the manufacturer's instructions (Amersham). Membranes were probed with
random-primed [a-32P]dATP-labeledDNAfragments, hybridizations and post-hybridization
washes were carried out at 65°C in aqueous buffers.
67
Acknowledgements
ThisresearchwasfinancedbyagrantfromtheEuropeanCommunity(BRIDGEBIOT-CT900163)and supportedbytheFoundationfor BiologicalResearch (BION),whichissubsidized
by theNetherlands Organization for Scientific Research (NWO).
References
Boeke, J.D., LaCroute, F. and Fink, G.R. (1984) A positive selection for mutants lacking orotidine-5'phosphate decarboxylase activityinyeast:5-fluoro-oroticacidresistance.Mol. Gen. Genet.197,345-346.
DeWit,P.J.G.M. andSpikman,G. (1982)Evidencefortheoccurrenceofrace-andcultivar-specific elicitors
ofnecrosis inintercellular fluidsof compatibleinteractionsof Cladosporiumfulvumand tomato.Physiol.
PlantPathol. 21, 1-11.
Flor, H.H. (1942)Inheritanceofpathogenicity inMelampsora Uni. Phytopathology 32, 653-669.
Fotheringham, S. andHolloman, W.K. (1989) Cloninganddisruptionof Ustilagomaydisgenes.Mol. Cell
Biol. 9, 4052-4055.
Goosen, T., Bloemheuvel, G., Gysler, C , De Bie, D.A., Van denBroek, H.W.J. and Swart, K. (1987)
TransformationofAspergillusnigerusingthehomologousorotidine-5'-phosphatedecarboxylasegene. Curr.
Genet.11,499-503.
Harling,R., Kenyon,L., Lewis,B.G., Oliver,R.P., Turner, J.G. andCoddington, A. (1988) Conditions
for efficient isolationand regeneration of protoplasts from Fulviafidva. J. Phytopathol. 122, 143-146.
Hohn, T.M., Proctor, R.H. andDesjardins, A.E. (1992) Biosynthesis of sesquiterpenoid toxinsby fungal
pathogens. In Molecular Biology of FilamentousFungi (Stahl, U. and Tudzynski, P., eds.), VCH,
Weinheim, pp. 99-109.
Hubbeling, N. (1978)Breakdownofresistance oftheCJ5geneintomatoby anothernewrace of Fulviafulva.
Meded.Fac. Landb. wet. Gent.43, 891-894.
Jones, D.A., Dickinson, M.J. andJones, J.D.G. (1991) Geneticlocations of the Cf-2, Cf-4, Cf-5and Cf-9
genes for resistance to Cladosporiumfulvum. TomatoGenet. Coop.Rep. 41, 22.
Joosten, M.H.A.J. andDeWit, P.J.G.M. (1988) Isolation, purification andpreliminary characterization of
aproteinspecific forcompatibleCladosporiumfulvum(syn.F«ZWa/KZva)-tomatointeractions.Physiol. Mol.
PlantPathol. 33,241-253.
Keen, N.T., Tamaki, S., Kobayashi, D., Gerhold, D., Stayton, M., Shen, H., Gold, S., Lorang, J.,
Thordal-Christensen, H., Dahlbeck, D. andStaskawicz, B. (1990)Bacteria expressing avirulencegene
Dproduce a specific elicitor ofdiesoybeanhypersensitive reaction.Mol.Plant-MicrobeInteract. 3, 112121.
Kronstad, J.W., Wang, J., Covert, S.F., Holden, D.W., McKnight, G.L. and Leong, S.A. (1989)
Isolationof metabolic genesand demonstration of genedisruptioninthephytopathogenic fungusUstilago
maydis. Gene79, 97-106.
Lindhout, P., Korta, W., Cislik, M., Vos, I. and Gerlagh, T. (1989) Further identification of races of
Cladosporiumfulvum (Fulviafulva)ontomatooriginatingfrom theNetherlands,FranceandPoland.Neth.
J. PlantPathol. 95, 143-148.
Miao, V.P., Covert, S.F. andVanEtten,H.D. (1991)Afungal geneforantibioticresistanceonadispensable
("B") chromosome. Science254, 1773-1776.
Michehnore,R.W., Norwood, J.M., Ingram,D.S., Crute, LR. andNicholson, P. (1984)The inheritance
of virulence inBremialactucaeto match resistance factors 3, 5, 6, 8, 9, 10 and 11in lettuce(Lactuca
saliva). PlantPathol.33,301-315.
Oakley, B.R., Rinehart, J.E., Mitchell, B.L., Oakley, C E . , Carmona, C , Gray, G.L. andMay, G.S.
(1987) Cloning, mapping andmolecular analysisof OaepyrG (orotidine-5'-phosphatedecarboxylase) gene
ofAspergillusnidulans. Gene61,385-399.
Oliver, R.P., Roberts, I.N., Harling, R., Kenyon, L., Punt, P.J., Dingemanse, M.A. and Van den
Hondel, C.A.M.J.J. (1987) Transformation of Fulviafulva, a pathogen of tomato, to hygromycin B
resistance. Curr. Genet. 12,231-233.
68
Panaccione,D.G., Scott-Craig,J.S.,Pocard,J-E.andWalton,J.D.(1992)Acyclicpeptidesynthetasegene
required for pathogenicity of the fungus Cochliobolus carbonum on maize. Proc. Natl.Acad. Sei. USA.
89, 6590-6594.
Sambrook,J., Fritsch,E.F.andManiatis,T. (1989)MolecularCloning, aLaboratoryManual.ColdSpring
Harbor Laboratory Press, New York.
Scholtens-Toma, I.M.J, and De Wit, P.J.G.M. (1988) Purification and primary structure of a necrosisinducingpeptidefromtheapoplasticfluidsoftomatoinfectedwithCladosporiumfulvum(syn. Fulviafidva).
Physiol.Mol. PlantPathol. 33,59-67.
Scott-Craig, J.S., Panaccione, D.G., Cervone, F. andWalton, J.D. (1990) Endopolygalacturonase isnot
required for pathogenicity of Cochliobolus carbonum on maize. Plant Cell2, 1191-1200.
Stahl, D.J. and Schäfer, W. (1992) Cutinase isnot required for fungal pathogenicity on pea. Plant Cell4,
621-629.
Sweigard, J.A., Chumley, F.G. and Valent, B. (1992) Disruption of aMagnaporthegrisea cutinasegene.
Mol. Gen. Genet.232, 183-190.
Talbot, N.J., Coddington, A., Roberts, I.N. and Oliver, R.P. (1988) Diploid construction by protoplast
fusion inFulviafidva (syn. Cladosporiumfulvum): geneticanalysisof animperfect fungal plantpathogen.
Curr. Genet. 14, 567-572.
Talbot,N.J., Oliver, R.P. andCoddington, A. (1991)Pulsed field gel electrophoresis reveals chromosome
lengthdifferences betweenstrainsofCladosporiumfulvum(syn.Fulviafidva). Mol. Gen.Genet. 229,267272.
Valent,B., Farrall,L. andChumley,F.G. (1991)Magnaporthegriseagenesforpathogenicityandvirulence
identified through a series of back-crosses. Genetics 127, 87-101.
VandenAckerveken, G.F.J.M., VanKan,J.A.L. andDeWit,P.J.G.M. (1992)Molecularanalysisofthe
avirulencegeneavr9of thefungal tomatopathogen Cladosporiumfulvumfully supportsthegenefor gene
hypothesis.PlantJ. 2, 359-366.
Van denAckerveken, G.F.J.M., VanKan,J.A.L., Joosten, M.H.A.J., Muisers, J.M., Verbakei, H.M.
andDeWit, P.J.G.M. (1993) Characterization of twoputativepathogenicity genesof thefungal tomato
pathogen Cladosporiumfulvum.Mol. Plant-Microbe Interact.6, 210-215.
VanderBeek, J.G., Verkerk, R., Zabel,P. andLindhout,P. (1992)Mapping strategy for resistance genes
intomatobasedonRFLPsbetweencultivars: Cfi>(resistance to Cladosporiumfulvum) on chromosome 1.
Theor. Appl. Genet.84, 106-112.
Van Kan, J.A.L., Van den Ackerveken, G.F.J.M. and De Wit, P.J.G.M. (1991) Cloning and
characterisationofcDNAofavirulencegeneavr9ofthefungalpathogenCladosporiumfulvum,causalagent
of tomatoleaf mold.Mol. Plant-MicrobeInteract.4, 52-59.
69
Chapter 5
The AVR9 race-specific elicitor of Cladosporiumfulvum
is processed by endogenous and plant proteases
GuidoF.J.M. van denAckerveken, Paul Vossen andPierre J.G.M. deWit
adaptedfromPlantPhysiology 103 (1993), Inpress
Summary
The avirulence gene avr9 of the fungal tomato pathogen Cladosporiumfulvum encodes a
race-specific peptideelicitor, thatinducesahypersensitiveresponseintomatoplantscarrying
thecomplementary resistance gene Cf9.Theavr9geneis highlyexpressed when C.fulvum
is growing inplanta and the elicitor accumulates in infected leaves as a 28 aa peptide. In
C.fulvumgrowninvitro,thepeptideelicitorisnotproducedindetectableamounts.Inorder
to produce significant amounts of the AVR9 elicitor in vitro, the coding and termination
sequences of theavi9 genewere fused to theconstitutive gpd-promoter (glyceraldehyde-3phosphatedehydrogenase)ofAspergillusnidulans. Transformants ofC.fulvumwereobtained
that highly expressed the avr9 gene in vitroand produced active AVR9 peptide elicitors.
Thesepeptideswerepartially sequenced from theN-terminusandappeared toconsist of32,
33and 34 aa, respectively and are theprecursors of the mature 28 aa AVR9 peptide. We
demonstrated thatplant factors process the34aapeptide intothe mature28aapeptide. We
present a model for theprocessing of AVR9 involving cleavage of a signalpeptide during
excretion and further maturation by fungal andplant proteases intothe stable28 aapeptide
elicitor.
71
Introduction
Specific recognition in the gene-for-gene interaction between the fungal pathogen
Cladosporiumfidvumandtomatoismediatedviaso-calledrace-specificelicitors,theprimary
products of avirulence genes (Van Kan et al., 1991;Van den Ackerveken et al., 1992).
Receptors in theresistant plant, which mightbethedirectproducts of resistance genes, are
thought to recognize the specific elicitors (DeWit, 1992).The subsequent defence reaction
is characterized by a hypersensitive response (HR) which restricts growth of the pathogen
to the siteof penetration.
Race-specific elicitors in the C.fiilvum- tomatointeraction werefirstidentified by De
Wit and Spikman (1982). One such race-specific elicitor, inducing HR on tomato plants
carrying theresistance gene Q9, waspurified and characterized (DeWit et al., 1985)and
the amino acid sequence was determined (Scholtens-Toma and De Wit, 1988). The
corresponding cDNA was isolated using oligonucleotide probes and designated avirulence
gene avr9 (Van Kan et al., 1991). Races of C.fiilvumvirulent on tomato genotype Cf9
completely lack the avi9 gene and therefore do not produce AVR9 peptide elicitor.
Transformation of such a virulent race with a genomic clone containing avi9 resulted in
transformants whichhadbecomeavirulent ontomatogenotype Cf9(VandenAckerveken et
al., 1992).In addition, disruption of avi9in wild typeavirulentraces of C.fiilvumresulted
invirulenceontomatogenotypeÇf9(Marmeisseetal., 1993).Altogethertheseresultsprove
that theavirulence geneavr9is theonly genetic factor responsible for theinduction of HR
andotherdefenceresponses ontomatogenotype Cf9andthatthisismediatedbyitsproduct,
the AVR9race-specific peptide elicitor.
The AVR9 elicitor is an extracellular peptide, that is produced by C. fiilvum while
growing inplanta, but not invitro.Theavr9geneencodes a 63aapeptide, but inplantaa
28 aa peptide accumulates predominantly. Expression of avr9 in vitro, as determined by
northernblotanalysis,hasonlybeendetectedunderconditionsofnitrogen starvation.Under
these conditions, however, no elicitor activity could be detected in theculture filtrate of in
vitrogrownmycelium(VandenAckervekenetal.,submitted).Highlevelexpressionofavr9
in vitrowould enable us to isolate and purify large quantities of AVR9 elicitor, which can
beused in future studies on recognition and signal transduction pathways involved in plant
defence. Here, we report on the construction and analysis of transformants of C.fiilvum
producing AVR9invitro byemployingahigh expression promoter ofAspergillus nidulans.
Ourexperimentsprovideevidencefor theinvolvement ofbothfungal andplantproteasesin
the sequentialprocessing of the 63aaprecursor AVR9 into the mature 28 aa race-specific
peptide elicitor.
72
Results
Isolation ofAVR9producing transformants ofC.fulvum
For the development of transformants of C.fulvum which constitutively produced high
amountsof theAVR9elicitor invitro,achimericgenewasconstructed (designated pCF22)
which consisted of theA. nidulansgpdpromoter fused tothecodingand termination region
of theC.fulvum avr9 gene(Fig. 1).Co-transformation of race5of C.fulvum with pAN7-l
which contained the hygromycin B resistance marker and pCF22 resulted in stable
hygromycin B resistant transformants. Five of these transformants were grown in liquid
shakecultureand theculture filtrate wasassayed for elicitoractivity ontomato.Theculture
filtrate of three of these transformants induced intense necrosis in genotype Cf9but not in
cultivar MM (lacking resistance gene Cf9). Thus, these three transformants produced the
AVR9elicitor in vitro.Southern analysis of these AVR9producing transformants indicated
that 5-20 copies of pCF22 were integrated into their genomes.
pCF1
pNOM102
uidA
Pgpd
N
N
aw9
Pav/9
TtrpC
H E
B
H
digestwith
BspH\ andHinó\\\
digestwith
* W//?dllland A/col
aw9
Pgpd
N
B . H
Pgpd
avi9
1 kb
pCF22
H
Figure 1. ConstructionofthePgpd-avr9 genefusion. Thecodingandtermination regionoftheCladosporium
fulvum avr9genewasremoved from plasmidpCFl (VandenAckerveken etal., 1992)bydigestionwithÄspHI
and Hindlll. TheuidA geneand trpCterminator were removed from plasmidpNOM102(Robertsetal., 1989)
by digestion with Hindlll and Ncol and the plasmid containing the high level expression promoter of the
glyceraldehyde-3-phosphatedehydrogenasegene(Pgpd)ofAspergillusnidulans(Puntetal., 1988)wasisolated.
ThePgpdwassubsequently fused totheavr9codingregion at thetranslation start codonpresent on theBspHl
site, resulting inplasmid pCF21. (uidA:ß-glucuronidasecoding region, TtrpC:terminator of trpCgene of A.
nidulans,Pavr9: promoter of avr9gene of C. fulvum, E: EcoRl, N: Ncol, B:BspHl and H: Hindlll).
73
Isolation, purificationandcharacterization ofAVR9from culture filtrate
TheAVR9race-specific elicitorisolated from IFof C../w/vwm-infectedtomatoleavesisa28
aapeptide (Scholtens-Tomaand De Wit, 1988) which migrates on low pH PAGE because
ofitsslightly basicnature (Fig.2, lane 1).Analysisof theculturefiltrates of in v/Yw-grown
C. fidvum transformants producing AVR9 by low pH PAGE showed that the filtrates
contained peptides with an electrophoretic mobility different from AVR9 isolated from IF
(Fig. 2, lane 2). These samepeptides wereproduced by all three transformants. Although
these peptides had a lower mobility on low pH gel, they had similar necrosis-inducing
activity on tomato genotype Cf9as the 28 aa peptide. Purification of these peptides from
culture filtrates by preparative cation exchange chromatography and reversed phase
chromatography resulted in an enriched mixture of several peptides with necrosis-inducing
activity on tomato genotype Ç/9. These peptides were subsequently purified by HRLC
employing a combination of cation exchange chromatography and reversed phase
chromatography. The purified peptides were sequenced from the N-terminus and appeared
tobeprecursors of the28aapeptideelicitor. Theobtained sequenceof 20N-terminalaaof
a mixtureof 32and 33aaAVR9peptidesoverlapped withthesequenceof the28aaelicitor
which has been completely sequenced three times. The 33and 34 aapeptides (Fig. 2, lane
3 and 4) occur most abundantly, while the 32 aa peptide is present in relatively small
quantities,butcomigrateswiththe33aapeptideonlowpHPAGE. Theminorpeptideband
in lane 2 (indicated with an asterisk) was
characterized as a mixture of 32, 33 and 34
aa. Using the Tricine-SDS PAGE system
(Schägger and Von Jagow, 1987), which is
recommended for the separation of proteins
with aMW <10kD, the 28 and 32 to 34aa
peptides comigrated and could not be
separated. The purification of the 28 aa
peptide as described by Scholtens-Toma and
Figure 2. Low pH PAGE profiles of different
De Wit (1988) involved preparative low pH
AVR9 elicitor peptides. The 28 aa peptide, as
PAGE, whichprecluded thepossibleisolation
purified from intercellular fluid of C. fidvum (wild
of otherpeptidessuchasisolated from culture
type) infected tomato leaves, is a small basic
peptide (lane 1). Several different AVR9 peptides
filtrate as described above. We wanted to
are produced in vitro by transformants of C.
know whether these AVR9 precursors were
fidvum constitutively expressing avr9 (lane 2); the
different peptides (32, 33 and 34 aa) co-migrate
alsopresentinIFof C.,/w/vM/n-infectedtomato
and have a reduced mobility as compared to the
leaves.Wetherefore fractionated andanalyzed
28 aapeptide. The two major peptidesproduced in
proteins present in IF by a procedure similar
vitro were purified and are the 33 aa (lane 3) and
34 aa (lane 4) AVR9 peptides. The 32, 33 and 34
totheoneused for thepurification ofAVR9aa peptides are also present in IF of C. fidvum
peptides from culture filtrate. Following the
(wild type) infected tomato leaves (lane 5), albeit
preparative purification steps of cation
in significantly lower quantity than the 28 aa
peptide. The peptide bands (indicated with
exchangeandreversedphasechromatography,
asterisks, lane 2 and 5) are less basic forms of the
the fractions containing elicitor activity were
33-34 and 28 aapeptides, respectively.
1 2
74
3 4 5
^^
^
•d
0
0
0
•H
C
i-l
•ri
CU
C
eu
3
C
eu
3
tr
eu
o
XI
n)
CD
XI
0
OJ
fi
3
cr
0)
p.
w
C
0
<
Z
cd
u
m
•o
a
o
s
o
.-i
«
(-1
4-1
4J
e
O
•e
H
1
3
s
1
•K
o
o,
o,
M
P.
*!
•d
eu
p.
eu
u
3
tr
eu
n
>->
<-<
r-l
cd
•H
4J
M
ed
P.
<u
cr
eu
01
01
r->
r-l
•H
ed
•H
4J
1-1
cd
>i
1-1
r-l
OJ
p.
0.
•i-l
4J
1-1
cd
•o
eu
0
0)
CT
d)
C/l
eu
•u
eu
i-H
I*
0
0
•d
C
•Ö
cd
c
cd
C
cd
c
cd
•0
eu
•0
eu
•d
eu
•d
eu
14-1
M-l
•l-l
<u
•H
IM
•H
lt-1
•H
1-4
3
p.
•ri
3
a
T)
•i-i
IM
a
p-
•d
•i-i
•H
w
3
P.
r-~
o
1-1
r-l
r-l
iH
r-l
r-l
iH
i-l
CN
vu
r-
r~
r~
r»
r^
r~
r-l
a\
m
r-l
0
e»
in
in
ON
r-i
VO
<f
co
l-~
ro
VO
CI
m
ro
£
o
£
u
2
£
g
u
3
£
S
>
-*
r-
>4
cd
Xi
N ^
eu
•d
•H
(M
!
•G
en
>
eu
O.
M
H
•d
eu
•p
0
•ri
•Ö
•o
1-1
CU
-0
eu
•0
eu
4-1
CM
m
r-
g
1
"2
0)
o
8
e
"o.
3
1
er
4)
w
$
•o
1
s^
s
fa
a
u
^
S
E
a
u
cri
tó
e>
0
«i
o-
"3
0
1
•H
3
u
o0
3
1
8
•H
o
CJ
^
«
fa
1
2
S
H
r
.s
Ü
>
0
u
0
<
S3
CM
iJ
0
2
H
Ü
en
co
Z
U
>H
^
x
ft]
Q
u
ai
0
K
fa
O
U
PS
3
u
fa
fa
u
cy
0
3
H
u
M
C/î
U
en
CO
a
n
m
u
a
0
0
<y
0
H
H
U
>H
<f
m
^
S
fa
O
as
Z
>
0
S
U
0
!H
CJ
O
0
zu
>
3
^
u
u
o0
•J
u
9
0
a
0
fa
cS
U
CO
CO
zu
>H
0
>
>
fa
J
3
CJ
'S
2
(J
se
J
H
Ö
CM
«
<
(4-1
O
M
W
KJ
2
•J
to
3
•c
ao
Cd
S
>
6
CO
CO
rJ
S
*N
33
0)
H
•H
y\
CD
cil
H
m
v-'
«
><
cj
>
<
JU
2
H
>
>
CJ
fU
0
0
fa
zc j
Ü
&<
2
Bi
fa
O
fa
0
z
au
<s
33
M
K
u
CJ
CO
CO
CJ
>H
s
o0
J
u
03
M
U
3.
u
U
5,
M
VO
O
<f
CN
ro
03
CN
pooledandanalyzedbylowpHPAGE.Inadditiontothemajor28aapeptidetheIFcontains
several otherpeptideswhichbehavesimilarly onlowpHPAGEand cationexchangeHRLC
astheAVR9-peptidesisolated from culturefiltrate (Fig.2,lane5). Sequenceanalysisofthe
N-terminus of these purified peptides confirmed the presence of 32, 33 and 34 aa AVR9
precursors in IF. The minor peptide band present on thelow pH gel,just above the 28 aa
elicitor (Fig.2, lane5, indicated withadoubleasterisk) alsoappeared tobe28aain sizeas
determined by N-terminal sequencing. An overview of the different AVR9 peptides
characterized sofar ispresented in Table I.
Proteolyticprocessing of'AVR9byplantfactors
In culture filtrates of in vitro-gmv/nC. fulvum transformants constitutively expressing the
avr9gene the 28 aa elicitor could not be detected. The smallest necrosis-inducing peptide
isolated from culture filtrate is 32 aa in size. The final processing of AVR9 into the28 aa
peptide, which is the most abundant AVR9 y\
4.3
7.0
8.7
peptide in IF of C. yw/vw/w-infected tomato
PH
MMCS
MM
Ct9
MM
Cf9
leaves,istherefore thoughttotakeplaceinthe
plant. Purified 28 aa and 34 aa peptide
elicitors were radioactively labeled with 125I,
which is incorporated into the amino acid
tyrosine and results in mono- or diiodotyrosine (Fig. 3B, lane 1 and 2,
respectively). Iodinated AVR9 remains
B
biologically active as tested in a pilot
experiment using unlabeled Nal (results not
shown). The labeled peptides were incubated
• • m
fjfr-with IF from healthy tomato plants at 37°C
for 2hours. Analysis of thereaction mixtures
bylowpHPAGErevealed thatfactorspresent
Figure 3. Proteolytic cleavage of the 34 aa
in the IF of both tomato cultivar MM and
AVR9 peptide by plant factors. Autoradiograph of
genotype Cf9were able to process the 34 aa
123
I-labeled AVR9 peptides separated by low pH
elicitor into the 28 aa peptide elicitor (Fig.
PAGE. (B) Untreated 28 aa (lane 1, major lower
band) and 34 aa (lane 2) peptides can be separated
3A). Proteolyticactivity wasoptimalatacidic
by low pH PAGE as they significally differ in
pH (2-7) and marginal at alkaline pH (>7)
their pi. (A) Labeled 34 aa peptide was incubated
with 200 ill of IF from healthy tomato cultivar
(Fig. 3A, and data not shown). The proteoMM and genotype Cj9 for 2 hours at 37°C at
lyticactivityinIFwasdestroyedbyboilingIF
different pH values. The proteolytic activity is
for 5 minutes prior to incubation (Fig. 3B,
optimal at acidic to neutral pH and marginal at
alkaline pH. (B) The processing of the 34 aa
lane 6). No reduced proteolytic activity was
peptide into the 28 aa form at pH 4 (lane 3) is not
observedwhenpepstatin,aninhibitorofacidic
inhibited by pepstatin (40 /tg.mT1, lane 4) nor by
ethanol, a solvent for pepstatin (lane 5). Boiling
aspartyl endoprotease of tomato (Rodrigo et
the IF prior to incubation with the 34 aa peptide
al., 1989), was added to thereaction mixture
resulted in loss of proteolytic activity of IF (lane
containingthe34aapeptide(Fig.3B,lane4).
6).
76
Discussion
The avirulence gene avr9 is highly expressed in C.fulvum when growing inside tomato
leaves (VanKan etah, 1991),whileC.fulvumgrown invitrodoes notproduceanyAVR9
elicitor. Highlevelexpression ofavr9invitro hasbeenaccomplished byexpression ofavr9
under control of the heterologous gpd promoter of A. nidulans. In A. nidulansthegpd
promoter constitutively expresses the gpd gene encoding glyceraldhyde-3-phosphate
dehydrogenase (GPD) (Puntetah, 1988). InSaccharomyces cerevisiae, GPD may account
for 5% oftotalcellularproteins(Krebsetal., 1953).Transformants ofC.fulvum containing
thePgpd-avr9 fusion produced andexcreted a setofAVR9race-specific elicitors in culture
filtrates of invzVro-grownmycelium; thesefiltratesinduceHRin tomatoplantscarrying the
resistance geneCf9.
The AVR9 peptides from culture filtrate were purified to homogeneity, using a
combination of cation exchange chromatography and reversed phase chromatography, and
sequenced from theN-terminus. Instead of the28aaAVR9peptide, as found in C.fulvuminfected tomatoleaves, theculturefiltrate containeda mixtureof 32, 33and34aapeptides.
The samepeptides could alsobedetected in IF of C../«Zvwm-infected tomato leaves. These
peptides were not isolated before since the purification method described previously
(Scholtens-Toma and De Wit, 1988) involved preparative low pH PAGE followed by
excision of theband containing the28aapeptideand thusprecluded theisolation of the32,
33and34aapeptides. Thereduced mobilityof thelarger forms ofAVR9onlowpHPAGE
isduetothepresenceofanadditionalasparticacidresidue(D)reducingthepi. Thepeptides
(indicated withasterisks, Fig. 2, lane2and 5) withreduced mobility onlowpHPAGEbut
with identical N-terminal sequences might be explained by deamination of asparagine or
glutamineresidues resulting in alesspositive charge (Ahern and Klibanov, 1985).
Plantfactors mediatetheprocessing ofthe34aaAVR9peptide,producedbyC.fulvum
invitro,intothe28aaform (Fig.3).Radio-labeled (125I)34aaAVR9elicitorwasincubated
with IF isolated from healthy plants and subsequently analyzed by low pH PAGE. The
processing was inactivated by boiling the IF before incubation with the labeled peptide
substrate, suggesting enzymatic cleavage by plant proteases. Alkaline (Vera and Conejero,
1988) and acidic proteinases (Rodrigo et ah, 1989) of tomato have been reported to be
involved in degradation of plant proteins. Theprocessing of the34 aapeptide is optimalat
acidicpH, suggesting theinvolvementof anacidicaspartylproteinase. However, pepstatin,
which was reported to inhibit a tomato aspartyl proteinase (Rodrigo et ah, 1989) did not
inhibitcleavageofthe34aapeptide.ThepHintheapoplastisapproximately 5.5,indicating
thatprocessing can occur in thetomatoleaf.
The results described here allow us to propose a model for the processing of AVR9
(Fig.4).Theprecursor protein encoded bytheavr9geneis63aa longasdeduced from the
cDNA sequence (Van Kan et ah, 1991). The mature elicitor peptide, as isolated from
C../wZvH/n-infected tomato leaves, consists of the carboxy-terminal 28 aa of the precursor
77
protein(Scholtens-TomaandDeWit, 1988). Themostprobablesignalsequencecleavagesite
(VonHeijne, 1986)islocatedbetween aminoacidsAla(23)andAla(24)aspredicted bythe
program SigSeqwhichisincluded intheSequenceAnalysisSoftwarePackageoftheGenetics
Computer Group (Devereux etal., 1984). The resulting extracellular peptide of 40aa has
not been identified experimentally. Extracellular proteases produced by C.fiilvumare most
probably directly processing the40aapeptide into theintermediate forms of 32-34aa. The
finalstepin theprocessing is mediated by plant factors and results in a stable mature28aa
race-specific peptide elicitor. Wedo not know whether the predicted 40 aa peptide can be
directly processed into the 28 aa peptide by the plant or whether fungal proteases play an
essential role in the formation of 28 aa race-specific peptide elicitor in vivo. The specific
necrosis-inducing activity of the 32, 33and 34 aa peptides is indistinguishable from the 28
aapeptide.Whether processing of 32, 33and 34aapeptidesintothe28aaform bytheplant
isessential for elicitor activity is not known.
PRIMARY
TRANSLATION
PRODUCT
SIGNAL PEPTIDE
23 aa
63aa
EXTRACELLULAR
PEPTIDE40 aa
INTERMEDIATE
FORMS32,33 &34 aa
FUNGAL PROTEASES
PEPTIDE
ELICITOR 28 aa
PLANT PROTEASES
Figure 4. Model for the processing of the AVR9precursor in vivo. The primary avr9translation product of
63 aa is excreted by Cladosporiumfulvum into the intercellular space of tomato leaves during infection, after
removal ofa23aa signalpeptide.The40aa precursor protein issubsequently cleaved by fungal proteases into
32, 33 and 34 aa peptides. Plant proteases mediate the final cleavage to form the stable 28 aa race-specific
peptideelicitor. Themodel doesnotexcludethepossibility that theexcreted 40aaprecursor isdirectlycleaved
by plant proteases to form the mature 28 aa peptide in one singlestep.
Our results indicate that several processes occur either before or during the recognition of
theAVR9elicitorbytomatogenotype Cf9.Oncethepathogen enterstheleaf,theexpression
of theavirulencegeneavr9 isinduced, possiblybythenutritionalenvironment intheleafor
byunknownplantfactors (VandenAckerveken, submitted).Theextracellular AVR9peptide
78
is instantaneously processed by both fungal andplantproteases. However, wedonotknow
whether the maturation is essential for elicitor activity, as processing of AVR9 peptides
occurs tooquickly to be studied separately from biological activity of thepeptides.
The involvement of both host and pathogen in the formation of elicitors as described
hereisnotunique.Inmanyotherhost-pathogeninteractionsnon-specific elicitorsare formed
whicharereleasedfromcellwalls.Forexample,ß-l,3-glucanases from soybeanwereshown
to release elicitor-active carbohydrates from cell walls of Phytophthora megasperma (Keen
and Yoshikawa, 1982). In another system, thePGIP (polygalacturonase inhibiting protein)
from Phaseolus vulgaris inhibits the activity of fungal polygalacturonases, prolonging the
life-time of oligogalacturonide elicitors (Cervone et al., 1989). Oligogalacturonides with a
degree of polymerization higher than nine act as elicitors of phytoalexin accumulation
(Nothnageletal., 1983).In theabsenceofPGIP,pectic fragments arequickly degraded by
polygalacturonases to monomeric forms which haveno elicitor activity.
The125I-labeledAVR9peptidesarecurrentlyusedasligandsinreceptorbindingstudies.
Ahypotheticalreceptorintheresistanthostplant, whichmightbetheprimaryproductofthe
complementary resistance gene Cf9, is proposed to recognize the AVR9 peptide elicitor.
Isolation of the Cf9geneis a major goal of future studies.
Experimental procedures
Subculture of C. fulvum andAVR9 bioassay
Cladosporiumfulvum Cooke (syn. Fulviafulva (Cooke) Cif) was grown onpotato dextrose
agar (PDA) or in liquid B5-medium in shake cultures (De Wit and Flach, 1979). AVR9
elicitor activity on tomato (Lycopersicon esculentum Mill.) was assayed by injection of 50
jtl samples into leaves of cultivar Moneymaker (MM, no Cfresistance genes) and a nearisogenic line of MM carrying resistance gene Cf9(genotype Cf9)(De Wit and Spikman,
1982). Oneto two daysafter injection, AVR9 elicitor activity wasvisible as necrosis (HR)
on genotype Cf9but not on cultivar MM.
Construction oftheavr9expression vector
TheAspergillus nidulans glyceraldehyde-3-phosphate dehydrogenase igpd)promoter (Punt
etal., 1988)wasemployed toconstitutivelyexpresstheavr9 geneofC.fulvum (Fig. 1).The
avr9coding and termination region (800 bp) were isolated from plasmid pCFl (Van den
Ackerveken et al., 1992) by digestion with BspHIand Hindlll. The uidAgene andtrpC
terminator were deleted from plasmid pNOM102 (Roberts et al., 1989) by digestion with
Hindlll and Ncol and replaced by the avr9 coding and termination region, resulting in
plasmid pCF22.
79
Isolation ofAVR9 producing transformants ofC. fulvum
After race 5of C.fulvum wasgrown for 48hours in liquid B5medium, themycelium was
harvested and used for theisolation ofprotoplasts (Harling etal., 1988). Co-transformation
(Oliver et al., 1987) of 107 protoplasts with 2 /tg pAN7-l containing the hygromycin B
resistance gene (hph;Punt et al., 1987) and 4 jtg pCF22 resulted in stable hygromycin B
resistanttransformants. Fiveofthesetransformants wereassayedfortheproductionofAVR9
elicitor invitroand analyzed by Southern blotanalyses.
Purification ofAVR9 elicitorfrom culture filtrate
Transformants producing AVR9 elicitor in vitrowere grown in liquid shake cultures for 7
to 10daysin50mlB5-mediumin300mlconicalflasksat22°Cand 100strokes.min1.Cellfree culture filtrate was obtained by filtration of the cultures through filter paper over a
Büchner funnel. To 500 ml of culture filtrate one volume of acetone was added and the
majority of high molecular weight proteins was precipitated overnight at -20°C. After
centrifugation, thesupernatantwascollectedandacetonewasremovedinarotary evaporator
at 50°C. Sodiumphosphatebuffer (IM, pH5.5) wasadded totheremaining supernatant to
a final concentration of 20mMandadjusted topH 5.5 withH3P04. Theconductivity of the
samplewasadjusted to < 4mScm"1 by dilution with20mMsodiumphosphatebuffer (pH
5.5). Thebuffered samplewasapplied overnight toa CM-Sephadex C-25 column (10x 2.2
cm),whichwaspre-equilibratedwith20mMsodiumphosphatebuffer, pH5.5.Thecolumn
was subsequently washed with 20 mM sodium phosphate buffer, pH 5.5, until the
absorbance at 280nmof theeffluent returned tobaselinelevel. AVR9peptideswereeluted
by washing the column with 20 mM sodium phosphate buffer, pH 8.0. The fractions
containing AVR9elicitor activity, as assayed by HRinduction on genotype Cf9of tomato,
were pooled and TFA was added to a concentration of 0.1% (v/v). The sample was
subsequently applied to a Sep-Pak C18 cartridge (Waters), a reversed phase cartridge for
rapid desalting, thathadbeenpreconditioned with 10ml90% (v/v)acetonitrile, 0.1%(v/v)
TFAfollowed by 10mlof0.1% (v/v)TFA.Followingapplicationthecartridgewaswashed
with 10ml0.1%(v/v)TFA.Theelicitorpeptideswereeluted from thecartridgewith 10ml
90% (v/v) acetonitrile, 0.1% (v/v) TFA. Acetonitrile was removed from the eluate in a
rotary evaporator at 50°C and the remaining sample was freeze-dried. High resolution
purification was achieved by reversed phase HRLC (High Resolution Liquid
Chromatography, BioRad) using a SuperPac Pep-S column (5 fim, C2/C18, 4 x 250 mm,
Pharmacia). The sample (200/d in 0.1%(v/v) TFA) wasapplied to thecolumn whichwas
pre-equilibrated with0.1% (v/v)TFA.Thecolumnwaselutedwithalineargradientof0.1%
(v/v)TFAto90% (v/v)acetonitrile,0.1 % (v/v)TFAin30minataflowrateof 1 mLmin1.
Fractionsof0.5mlwerecollected andfreeze-dried. Thedriedproteinfractions inindividual
peaks werepooled and dissolved in 20 mMMes, pH 5.5. The sample (200jtl)wasapplied
to a MA7S cation exchange column (HRLC, 7.8 x 50 mm, BioRad) which was preequilibrated with 20 mMMes, pH 5.5. Following application, thecolumn was eluted with
agradient of 20mMMes,pH 5.5 to20mMMes,pH5.5, 0.4 MNaClin 30minata flow
80
rate of 1 ml.min1. Individual peaks containing elicitor activity on genotype Cf9 were
analyzed by low pHPAGE.
Polyacrylamide gelelectrophoresis (PAGE)
LowpHPAGEwasperformed on 15% (w/v)Polyacrylamideslabgelsundernon-denaturing
conditions using pyronine Y as a front marker (Reisfeld et al., 1962). Following
electrophoresis at 200 V, the gels were stained and fixed according to Steck et al. (1980).
Omitting formaldehyde in the staining and destaining solutions resulted in loss of AVR9
peptides from thegel. PAGE under denaturing conditions wasperformed on Tricine-SDSPAGEgels (16.5%T, 6%C) as described by Schäggerand VonJagow (1987).
Peptide sequencing
Purifiedpeptidesweresequencedonagas-phase-sequenator (SON,sequencefacility, Leiden,
The Netherlands). Atleast five N-terminalaa were determined for allpeptides. Twenty aa
were determined for a mixture of 32 and 33 aa peptides. The 28 aa peptide elicitor was
completely sequenced.
Radio-Iodination ofAVR9peptides
AVR9 peptides of 28 and 34 aa were radiolabeled by iodination with Na125I (Amersham)
usingIodobeads(Pierce)asacatalyst. OneIodobead waswashedwithPBSbeforeincubation
with 19.5MBq Na125Iin 95/xlPBS for 5 minutes in an eppendorf vial. AVR9peptide was
added (5 id, 5 nmol) and incubated for an additional 15 minutes at room temperature.
Following incubation the reaction mixture (100 /xl) was transferred to a new vial. The
remainingIodobeadwaswashed with 100/xlPBS.Thewashingfluid wascombined withthe
reaction mixture(200/xl)andappliedtoaSep-PakCI8cartridgetoremovethefree 125Ifrom
labeled AVR9elicitor whichbinds. Thelabeled peptideswere eluted from thecolumn with
90% (v/v) acetonitrile, 0.1% (v/v) TFA. The eluted fractions were freeze-dried and
redissolved in milli-Q water. The specific activity of the labeled 34 aa AVR9 was 1to 2
MBq.nmol"1as determined by liquid scintilation counting.
Proteolytic activity assay
Proteolyticprocessing byplant factors ofAVR9produced byC.fulvum invitrowasstudied
by using 125I-labeled34aaelicitorpeptide.Intercellular fluid (200id)isolated from healthy
tomatoplants (DeWitand Spikman, 1982)wasincubated withlabeled AVR9elicitor (5til,
50,000 cpm) for 2 hours at 37°C. The buffers used were: (i) acidic buffer; 84 mM citric
acid, 32mMNa2HP04,pH4(Rodrigoetal., 1989),(ii)alkalinebuffer; 150mMTris-HCl,
2 mM CaCl2, 0.1 mM DTT, pH 9 (Vera and Conejero, 1988) and (iii) unbuffered; IF
adjusted topH 7 with NaOH. The effect of theprotease inhibitor pepstatin was studied by
pre-incubating theassay mixturefor 30minat4°C withinhibitor (40/xg.ml"1)before adding
labeled AVR9. Since pepstatin was dissolved in ethanol, an equal volume of ethanol was
addedtoacontrolassay, toafinalconcentrationof2% (v/v).FollowingincubationofAVR9
81
elicitorunderdifferent conditions,sampleswerefreeze-dried andanalyzedbylowpHPAGE.
Acknowledgements
TheinvestigationsweresupportedbytheFoundationforBiologicalResearch (BION),which
is subsidized by the Netherlands Organization for Scientific Research (NWO). We are
grateful to Dr. J.A.L. Van Kan and Dr. G. Honéefor critically reading the manuscript.
References
Ahern,T.J. andKlibanov,M.K. (1985)Themechanismofirreversibleenzymeinactivationat 100°C. Science
228, 1280-1284.
Cervone, F., Hahn, M.G., De Lorenzo, G., Darvill, A. and Albersheim, P. (1989) Host-pathogen
interactionsXXXIII.Aplantproteinconvertsafungal pathogenicity factor intoanelicitorofplant defence
response. PlantPhysiol.90, 542-548.
Devereux, J., Haeberli, P. and Smithies, O. (1984)Acomprehensive setof sequenceanalysisprograms for
the VAX. NucleicAcidsRes. 12,387-395.
DeWit, P.J.G.M. andFlach, W. (1979)Differential accumulation ofphytoalexinsintomatoleaves,butnot
infruits after inoculationwithvirulentandavirulentraces of CladosporiumfUlvum. Physiol.PlantPathol.
15, 257-267.
DeWit,P.J.G.M. andSpikman, G. (1982)Evidencefor theoccurence ofrace-andcultivar-specific elicitors
of necrosis in intercellular fluids of compatible interactions between Cladosporiumfidvumand tomato.
Physiol.PlantPathol.21, 1-11.
De Wit, P.J.G.M., Hofman, A.E., Velthuis, G.C.M, andKuc, J.A. (1985) Isolation and characterization
ofanelicitorofnecrosisisolatedfromintercellularfluidsofcompatibleinteractionsofCladosporiumfulvum
(syn. Fulviafulva) and tomato. PlantPhysiol.77, 642-647.
DeWit,P.J.G.M. (1992)Molecularcharacterization ofgene-for-gene systemsinplant-fungus interactionsand
the application of avirulence genes in control of plant pathogens.Annu. Rev. Phytopathol.30, 391-418.
Harling,R., Kenyon,L., Lewis, B.G., Oliver,R.P., Turner, J.G. andCoddington, A. (1988) Conditions
for efficient isolationandregenerationof protoplasts from Fulviafidva. J. Phytopathol.122, 143-146.
Keen, N.T. and Yoshikawa, M. (1982) £-l,3-glucanase from soybean releases elicitor-active carbohydrates
from fungal cell walls. PlantPhysiol.71,460-465.
Krebs,E.G., Rafter, G.W. andJunge,J.M. (1953)Yeastglyceraldehyde-3-phosphatedehydrogenase.Yeast
protein, II.J. Biol. Chem.200, 479-492.
Marmeisse, R., Van den Ackerveken, G.F.J.M., Goosen, T., De Wit, P.J.G.M. and Van den Broek,
H.W.J. (1993) Disruptionof theavirulence geneavr9in tworaces of the tomatopathogen Cladosporium
fiilvum causes virulence on tomato genotypes with the complementary resistance gene Cfi>. Mol. PlantMicrobeInteract., Inpress.
Nothnagel, E.A., McNeil, M., Albersheim, P. and Dell, A. (1983) Host-pathogen interactions XXII. A
galacturonic acid oligosaccharide from plant cellwalls elicitsphytoalexins. PlantPhysiol.71,916-926.
Oliver,R.P.,Roberts,I.N.,Harling,R., Kenyon,L., Punt,P.J.,Dingemans,M.A. andVandenHondel,
C.A.M.J.J. (1987) Transformation of Fulvia fidva, a fungal pathogen of tomato, to hygromycin B
resistance. Curr. Genet.12,231-233.
Punt, P.J., Oliver, R.P., Dingemans, M.A., Pouwels, P.H. and Van den Hondel, C.A.M.J.J. (1987)
Transformation ofAspergillus based on thehygromycin Bresistance marker from Escherichia coli.Gene
56, 117-124.
Punt,P.J.,Dingemans,M.A., Jacobs-Meijsing, B.J.M., Pouwels,P.H. andVandenHondel, C.A.M.J.J.
(1988)Isolationandcharacterization oftheglyceraldehyde-3-phosphatedehydrogenasegeneofAspergillus
nidulans. Gene69,49-57.
Reisfeld,R.A., Lewis,U.J. andWilliam,D.E. (1962)Diskelectrophoresisofbasicproteinsandpeptideson
Polyacrylamide gels. Nature195,281-283.
82
Roberts, I.N., Oliver, R.P., Punt, P.J. and Van den Hondel, C.A.M.J.J. (1989) Expression of the
Escherichiacoliß-glucuronidasegene in industrialand phytopathogenicfilamentousfungi. Curr.Genet.
15, 177-180.
Rodrigo, I, Vera, P. and Conejero, V. (1989) Degradation of tomato pathogenesis-related proteins by an
endogenous 37-kDa aspartyl endoproteinase. Eur. J. Biochem. 184, 663-669.
Schägger,H. andVonJagow, G. (1987)Tricine-sodiumdodecylsulphate-polyacrylamidegel electrophoresis
for the separation of proteins in the range from 1to 100kDa. Anal. Biochem. 166, 368-379.
Scholtens-Toma, I.M.J. and De Wit, P.J.G.M. (1988) Purification and primary structure of a necrosisinducingpeptidefromtheapoplasticfluidsoftomatoinfectedwithCladosporiumfulvum(syn. Fulviafiilva).
Physiol.Mol. PlantPathol. 33,59-67.
Steck, G., Leuthärd, P. and Biirk, R.R. (1980) Detection of basic proteins and low molecular weight
peptides inPolyacrylamide gels by formaldehydefixation.Anal. Biochem. 107, 21-24.
VandenAckerveken, G.F.J.M., VanKan,J.A.L. andDeWit, P.J.G.M. (1992)Molecularanalysisof the
avirulence geneavr9of thefungal tomatopathogen Cladosporiumfulvum fully supportsthe gene-for-gene
hypothesis.PlantJ. 2, 359-366.
Van Kan, J.A.L., Van den Ackerveken, G.F.J.M. and De Wit, P.J.G.M. (1991) Cloning and
characterization of cDNA of avirulence gene avr9of the fungal pathogen Cladosporiumfulvum, causal
agent of tomato leaf mold. Mol. Plant-Microbe Interact.4, 52-59.
Vera, P.andConejero,V. (1988)Pathogenesis-relatedproteinsoftomato:P-69asanalkalineendoproteinase.
PlantPhysiol.87, 58-63.
Von Heijne, G. (1986) Anew method for predicting signal sequence cleavage sites. NucleicAcids Res. 14,
4683-4690.
83
Chapter 6
Nitrogen limitation induces expression of the
avirulence gene avr9in Cladosporium fiilvum;
a reflection of growth conditions inplanta?
GuidoF.J.M. van denAckerveken, RuthM. Dunn, TonJ. Cozijnsen,
Paul Vossen, Henk W.J. Van denBroek andPierre J.G.M. DeWit
submittedto Molecular andGeneral Genetics
Summary
The avirulence gene avr9 of the fungal tomato pathogen Cladosporiumfiilvumencodes a
race-specific peptide elicitor that induces the hypersensitive response in tomato plants
carrying the complementary resistance gene Cf9. The avr9 gene is not expressed under
optimalgrowthconditions invitro,butishighlyexpressed whenthefungus growsinsidethe
tomatoleaf. In thispaper wepresent evidence for theinduction of avr9geneexpression in
C.fiilvumgrown invitro under conditionsof nitrogenlimitation. Only growth mediumwith
very low amounts of nitrogen (nitrate, ammonium, glutamate or glutamine) induced the
expression of avr9.Limitation of other macro nutrients or theaddition of plant factors did
not inducetheexpression of avr9.The induced expression of avr9ispossibly mediated by
a positive-acting nitrogen regulatory protein, homologous to the Neurospora crassaNIT2
protein,whichinducestheexpressionofmanygenesunderconditionsofnitrogenlimitation.
The avr9 promoter contains several putative NIT2 binding sites. The expression of avr9
during the infection process was explored cytologically using transformants of C.fiilvum
carrying an avr9 promoter - ß-glucuronidase reporter gene fusion. The possibility that
expression of avr9in C.fiilvumgrowing inplanta is caused by nitrogen limitation in the
apoplast of thetomatoleaf is discussed.
85
Introduction
The fungal pathogen Cladosporiumfulvum Cookeis thecausal agent of tomatoleaf mould.
Colonization of susceptible tomato cultivars by C.fulvum is characterized by abundant
intercellular fungal growthandtheabsenceofactiveplantdefence (DeWit, 1977).Resistant
tomato genotypes, however, are able to recognize C.fulvum at an early stage of infection,
resulting in theinduction of the hypersensitive response and other defence responses which
restrict growth of thepathogen to theprimary infection site (DeWit, 1986).Recognition is
mediatedviarace-specificelicitors,theprimaryproductsofavirulencegenesofthepathogen,
whicharerecognized bytomatogenotypescarrying thecomplementary resistancegenes(De
Wit and Spikman, 1982; Scholtens-Toma and De Wit, 1988; Van den Ackerveken et al.,
1992).
Theavirulence geneavi9of C.fulvum, thefirst fungal avirulencegenecloned, iswell
characterized. Racesvirulent on tomatogenotypescarrying theresistance gene Q9 lackthe
avr9 geneandthereby avoidrecognition (VanKanetal., 1991).Virulentraces transformed
withtheclonedavi9geneareavirulentontomatogenotype Cf9(VandenAckerveken etal.,
1992). Disruption of the avr9 gene in wild-type avirulent races of C.fulvum resulted in
mutants virulent on tomato genotype Çf9(Marmeisse et al., 1993). These results clearly
demonstrate the role of the avr9 gene in the primary interaction of the pathogen with a
resistant tomatogenotype and the subsequent induction of defence responses. The intrinsic
function of the avi9 gene for C. fulvum is still unknown.
The avr9gene is highly expressed during growth of C. fulvum inside the tomatoleaf,
while the expression of avr9in C. fulvum under normal growth conditions in vitrois very
low (Van Kan et al., 1991). These observations prompted us to study the mechanisms of
induction and/or derepression of avr9. The expression of avr9in C. fulvum grown under
different conditions in vitrowas tested by northern blot analysis. The temporal and spatial
expression ofavr9 inC.fulvumgrowing inplantawasstudiedbyusing theEscherichia coli
ß-glucuronidase reporter gene fused to the avr9promoter. Nitrogen limitation appeared to
induceavr9geneexpression, which mightbea reflection of growth conditions inplanta.
Results
Accumulation ofavr9mRNA inC.fulvum growninvitro
Theavirulence geneavr9is highly expressed when C.fulvum grows inplanta (Van Kanet
al., 1991).Thelevelofavr9 expression duringgrowthinvitro instandardB5mediumislow
orundetectable.Inordertocharacterizeinducingand/orrepressing factorsorconditions,the
effect ofplant factors anddifferent growth mediaonavr9 mRNAaccumulationin C.fulvum
grown invitrowas studied.
Plant factors were added to 48-hours-old C.fulvum mycelium grown in liquid B5
86
mediumand after 24hourstheir effect onavî9mRNAaccumulation wastestedby northern
blot analysis. Tomato intercellular fluid (De Wit and Spikman, 1982), tomato conditioned
medium and homogenized tomato suspension cells (supernatant and pellet; Wolters et al.,
1991)weretested for theinduction of avr9geneexpression. Northern blot analysis of total
RNAdidnotrevealanyavr9 mRNAaccumulation (datanotshown)indicatingthatunderthe
conditions tested, plant factors do notinducetheexpression of theavr9gene.
Theeffect of macro-nutrients wastestedby growing C.fidvuminvitrofor twodaysin
B5mediumand subsequent transfer ofthemyceliumtodifferent testmedia.After additional
incubation for one day, mycelium was collected and the expression of avr9was tested by
northern blotanalysis. Replacement of sucrose, thecarbon sourceinB5mediumby glucose
and fructose did not result in accumulation of avr9mRNA (datanot shown).Limitationof
thecarbon, sulphur andphosphate sources (reduced with a factor 10and 100relative toB5
medium) did not result in any induced expression either (data not shown). Only limitation
of nitrogen resulted in increased levelsof avt9 mRNAindicating that avr9gene expression
is induced under nitrogen limiting conditions (seenext section).
N03
NH4
glutamateglutamine highC
40 10 1 40 10 1 40 10 1 40 10 1 40 1
aw9
act
• •• *• i
# i* .* m m m:m
SN A * •-*• * » t Ü
Figure 1. avr9mRNAaccumulationinC.fidvumgrown invitroinB5mediumcontaining40, 10and1 mM
of four different sources of nitrogen (nitrate, ammonium, glutamate and glutamine) and in BS medium
containing40 and 1mM nitrateand 20% sucroseinstead of 2% (laneshigh C). Twenty /ig of total RNAwas
electrophoresed in a 1.5% denaturing formaldehyde-agarose gel, blotted onto Hybond N + membrane and
hybridized with the avr9 cDNA and act gene (encoding actin, Van den Ackerveken, unpublished) as a
constitutivecontrol.
Accumulation ofavr9 mRNAunder conditions ofnitrogen limitation
Theeffect of nitrogen concentration onavr9geneexpression was studied in moredetailby
testing different sources andconcentrations ofnitrogen. Figure 1showstheeffect ofnitrate,
87
ammonium, glutamateandglutamineonavr9 expression. Allthesenitrogen sourcesinduced
no avr9mRNA accumulation at a concentration of 40 mM. However, high levels ofavr9
mRNA wereobserved whennitrogen concentrations werereduced to 1mM, indicating that
the avr9 gene expression is strongly induced or derepressed under low nitrogen
concentrations. At a concentration of 10 mM, the avr9 mRNA level was lower with
ammonium compared to the other nitrogen sources. Theinduced expression of avr9is not
caused by a high C/N ratio as medium with a 10fold higher sucrose concentration did not
result in induced avi9expression (Fig. 1).Theconstitutively expressed C.fulvum actgene,
encoding actin (Van denAckerveken, unpublished), was used as internal controlprobeand
demonstrates that comparable amounts of mRNA were loaded on thenorthern blot.
The effect of nitrate concentration on avr9 expression was studied in more detail.
Mycelium of C.fiilvwn(growninB5mediumwith 50mMnitratefor 48h) was transferred
to fresh B5mediumwith different concentrations of nitrate (0- 100mM) and grown for an
additional 24 h. At the end of this growth period the expression of avr9 was tested by
northern blot analysis. Theremaining concentration ofnitrate in theculturefiltrate after 24
hours of growth was measured, to allow estimation of nitrogen uptake during the growth
period. Theresultsare showninFigure2. During the24hgrowthperiod, theconcentration
of nitratedecreased 7-8 mM. Accumulation of avr9mRNA wasdetectablewhen thenitrate
concentration at theend of theincubation period wasbelow 1mM.
JSWrww
*HP
— IM
9
£
'S
CC
E
0
0.025 0.25
- O -
0.5
1
2.5
5
10
25
(N03] at start of growth period (mM)
avrtmRNA
- A -
50
100
(N03J
Figure 2. avr9 mRNA accumulation (% of maximum) in C. fiilvum grown in BS medium with different
concentrations of nitrate. The avr9mRNA levelwas determined by densitometric scanning of a northern blot
autoradiograph (shownabovegraph)containingequal amountsof totalRNA hybridized with theavr9cDNA.
Thenitrateconcentration in theculture filtrate after incubationof thefungus for 24hourswas determined and
is indicated on the right-handaxis.
88
Putative nitrogen regulatory elements intheavr9promoter
Data available on nitrogen regulation in Neurospora crassa (Marzluf et al., 1992) and
Aspergillus nidulans (Caddick, 1992) enabled us to examine the avr9 promoter for
homologies with known binding sites of regulatory proteins. Six copies of the sequence
TAGATA, theconsensussequencefor therecognition siteoftheN.crassa NTT2protein(Fu
and Marzluf, 1990), are present in the promoter region -300 to -1 (Fig. 3). Six additonal
copies of the core sequence GATA are present in this region as well, nit-2 is a major
positive-acting nitrogen regulatory genewhich turns ontheexpression of manygenesunder
conditions of nitrogen limitation (Fu and Marzluf, 1990).
-300 +
AGGCGCCGTCGCAAACGCTATCGGGTCTTGGATAGGCGGGCAAGATATCTATCGGCTGTT
+
+
++
+
TCCGCGGCAGCGTTTGCGATAGCCCAGAACCTATCCGCCCGTTCTATAGATAGCCGACAA
1
2
GTGCTATTAATATGTATCGCAAGCAGCTAGGTATTAGAAACCTAGATAGCTAGTTGACTT
-240 +
+
+
+
+
+
CACGATAATTATACATAGCGTTCGTCGATCCATAATCTTTGGATCTATCGATCAACTGAA
3
CATATTGGCTAGATATCTACCTAGAGCAATACAACCTTGAAACAGCTAGGTATAGCAAAC
-180 +
+
+
+
+
+
GTATAACCGATCTATAGATGGATCTCGTTATGTTGGAACTTTGTCGATCCATATCGTTTG
4
5
TTCAGTAGCTAGCTAACTTGATATTAACTAGATATCTACCTAGGCAGTAGATCCGGCCGA
-120 +
++
+
+
+AAGTCATCGATCGATTGAACTATAATTGATCTATAGATGGATCCGTCATCTAGGCCGGCT
6
1
GAGAGAGATATACAGGTATAAGTAGACAGTAGATCTCTTCTACTCTACTGTTACTAAATC4.
-60 +
+
+
+
+
+
CTCTCTCTATATGTCCATATTCATCTGTCATCTAGAGAAGATGAGATGACAATGATTTAG
Figure 3. Putative binding sites in the avr9promoter of C.fiilvumfor a positive-acting nitrogen regulatory
proteinhomologoustotheN. crassa NIT2protein (FuandMarzluf, 1990).Thenucleotidesequence of 300bp
oftheavr9promoter, upstreamofthemaintranscriptionstartisshown(VandenAckerveken etal., 1992).The
sixboxeswith theNIT2consensus sequence (TAGATA) are numbered (1-6). Six additionalboxes containing
thecoresequence(GATA),aputativeTATA-box(TATAAGT)andthemajor transcriptionstartpoint(position
1, indicated by the arrow) are also indicated.
89
Studieswiththeavr9promoter-GUSfiisions
The construction of theavr9promoter - reporter gene GUS fusion isdepicted in Figure 4.
A perfect fusion of the avr9 promoter fragment with the uidA start codon was possible
because the avr9gene start codon (ATG) was located in the ÄspHI restriction site. Three
avr9 promoter deletions were also constructed. Plasmids pCF23, PCF24 and pCF25
contained fragments of 2, 0.8 and 0.6 kb of the original promoter fused to the uidA gene,
respectively (Fig. 4).
•partial Noo I
•klenowfill up
' religation
pCF20
uidA
SpSt
B
H
E
TfraC
N
digest with
f coRI and Noo I
d e l e t e god promoter
digest with
EooRI and BspHl '
Isolate Paw9
PCF21
SpSt
A EcoRI - Sal I
pCF23
SpSt
A EooRI - S p eI
PCF24
E St
pCF25
A ÊcoRI - SefI
Figure 4. Construction of the avr9 promoter - reporter gene GUS fusions.
pNOM102(Robertset al., 1989), a plasmid containing the gpdpromoter of Aspergillus nidulans(Pgpd),the
E. coliuidAgene, coding for ß-glucuronidase(GUS) and trpCterminator (JtrpC) was modified by removing
one Ncol site. From the resulting plasmid pCF20 thegpd promoter was deleted by digestion with EcoRIand
Ncol. The 3kb avr9promoter region (Pavr9) was isolated from pCFl (Van den Ackerveken et al., 1992)by
digestion with EcoRI and BspHl and ligated to the uidAgene, resulting in plasmid pCF21. Three promoter
deletionderivativeswereconstructed with 2, 0.8 and 0.6kbof theoriginal avr9promoter fused tothereporter
gene and designated pCF23, pCF24 and pCF25, respectively. B=Äs/>HI, E=£coRI, H=Wi'/»dIII, N=M»I,
Sa=&i/I, Sp=S/>eI and St=SrtI.
90
C.fulvum race4 and 5were transformed to hygromycin Bresistance with pAN7-l and cotransforming plasmid (pCF21, 23,24 or 25). Transformants containing the co-transformed
ûvr9-reporter gene GTJS fusion were selected by Southern blot analysis. Transformants
containinglowß-glucuronidaseactivityinmyceliumgrowninvitroandhighß-glucuronidase
activity inC.fulvum-mtecteA tomatoleaves werefurther analysed bynorthern blotanalysis.
The avr9and uidAmRNA accumulation was determined in C. fulvum grown in liquidB5
medium with 50 mM nitrate (low avr9mRNA level) and from C. fulvum grown inplanta
(10daysafter inoculation, highavr9 mRNAlevel).Intheselected transformants, expression
of uidA (GUS)isdetected whentheavr9mRNAlevelishighand notwhentheavr9mRNA
level islow, indicating thatthereporter geneuidA isregulated similar tothe wild type avr9
gene(datanot shown).TheuidA mRNAlevelismuchlower than theavr9mRNA leveland
mightbetheresult of alow uidAmRNA stability. The selected transformants, co-regulated
foravr9 anduidA,wereusedforhistochemicalanalysisofavr9 expression.Theseresultsand
additionaldata(notshown)obtainedbystudyingthesetransformants whilegrowinginplanta
indicatethat thereisnoobviousdifference between transformants carrying pCF21, -23,-24
and -25, suggesting that full promoter activity is contained within the 0.6 kb promoter
fragment present inpCF25.
Localization andtiming ofexpression ofavr9inplanta
The different transformants of C.fulvum containing avr9promoter - reporter gene GTJS
fusions showed similar expression patterns during the infection of tomato seedlings. The
spatialand temporal expression of avr9 wasassessed by staining tomatocotyledons infected
with several different selected transformants for GTJSactivity (Fig. 5). Three to four days
after inoculation the fungus has penetrated the leaf through the stomata. Six days after
inoculation thegerminating conidiospores and runner hyphaeon theleaf surface showed no
significant GTJSactivity (Fig. 5 A &B). The penetrating hyphae and intercellular hyphae
showed high GTJSactivity (Fig. 5D, E &F). Ten days after inoculation the fungus grows
abundantly in the intercellular space of the mesophyll. The GTJS activity, however, was
mainlylocalized inthemyceliumgrowing near thevascular tissuein theleaf (Fig.5G)and
mightbetheresult of high metabolic activity of the fungus near thevascular tissue.
Twelveto fourteen daysafter inoculation, a networkof hyphaeappears in stomatal cavities
and conidiophores emerge through the stomata, all showing high GTJS activity (data not
shown).
Discussion
The specificity of host-pathogen interactions at the molecular level is determined by the
interaction between resistance genes of the host and avirulence genes of thepathogen. The
tomato-Cfulvum interactionprovides anexcellent modelsystem to study thesegenes.Two
91
92
Figure 5. Histochemical localization of GUS activity in cotyledons of the susceptible cultivar Moneymaker
inoculated with transformants of Cladosporiumfitlvumcontaining the aw9 promoter - GUS fusion. A,B)
Conidiospores (C) and runner hyphae (R) on the surface of the ab-axial side of the cotyledon show very low
or no GUS activity (6 days post inoculation). C) Conidiospores (C) and runner hyphae (R) of a control
transformant containingpNOM102(highconstitutiveexpression promoter, seefigure4) showvery highGUS
activity (S=Stoma). D) The runner hypha (R) from a germinated conidiospore (C) enters the cotyledon via a
stoma (S).Thepenetratinghypha(PH)becomesthickened(H)andshowsGUSactivity(6dayspostinoculation,
composite picture taken in different focal planes). E) The intercellular hyphae (H)just below the epidermal
puzzle cells (P) show very high GUS activity (6 days after inoculation). F) The intercellular hyphae (H)
continuetocolonize theintercellular spaces between thelowermesophyll cells (M). Mosthyphaeshow ahigh
GUSactivity,however, somehyphaeshowverylowGUSactivity(H-»)althoughtheyarelocalizedinthesame
cell layer (8 days postinoculation). G) Thehyphae (H)near thevascular tissue (V) show high GUS activity,
whereas thehyphae (H-»)around the mesophyll cells (M)often show lowor no GUS activity.
avirulence genes of C.fiilvum,avr9(Van den Ackerveken etal., 1992)and ÛVA4(Joosten
et al., unpublished), have been cloned and were shown to be highly expressed in planta
compared to normal growth conditions in vitro (Van Kan et al., 1991; Joosten et al.
unpublished). Two putative pathogenicity genes of C. fiilvum, ecpl and ecp2 showed a
similar expression pattern (Van den Ackerveken etal., 1993).Unravelling the mechanism
of regulation of these genes inplanta might help to reveal the primary function of these
genes for thepathogen.
In this paper we describe the induction of avr9gene expression under conditionsof
nitrogen limitation. The C.fiilvumavrA,ecpl and ecpl genes are not induced under these
conditions (Van den Ackerveken unpublished). Deprivation of any nitrogen source is
sufficient for theinductionofavr9expression (Fig. 1).Atnitrateconcentrationslowerthan
1 mMtheexpression ofavr9 isinducedorderepressed (Fig.2).Undertheseconditionsthe
cv/9 geneproduct could never bedetected in theculture filtrate, possiblyby thedecreased
levelofprotein synthesisundertheapplied nitrogenlimitationconditions.
In Neurospora crassaandAspergillus nidulans regulatory genes involved in nitrogen
metabolism havebeen well studied genetically. Nitrogen metaboliterepression is mediated
by theglobal regulatory genes nit-2and areA in N. crassa(Marzluf et al., 1992) and A.
nidulans(Caddick, 1992), respectively. The products of these genes are DNA binding
proteinswhichturnontheexpressionofmanygenesunderconditionsofnitrogenlimitation.
TheN. crassa NIT2proteinbindstopromoter sequencesthat containat least twocopiesof
thesequenceTAGATA,orthecoresequenceGATA(FuandMarzluf, 1990).Thesamecore
sequenceisrecognized by the mammalian GF1protein, which has a high homology tothe
NIT2andAREAproteinsinthezincfinger domain(Marzlufetal., 1992).TheNIT2protein
has alsobeen shown to bind to thepromoter of the tomato nitrate reductase gene invitro
(Jarai etal., 1992)which suggestshigh conservation of theseDNA-bindingproteins.
Theavr9 promotercontainssixcopiesof theTAGATA sequenceand anadditionalset
of six copies containing thecore sequence GATA (Fig. 3). Theinduced expression of the
avr9gene under conditions of nitrogen limitation could be the result of the binding of a
93
positiveactingregulatoryproteinhomologoustoNIT2andAREA.Infuture experimentsthe
C.fulvum avi9 promoter will be tested for AREA regulation in areA deletion mutantsand
constitutiveareA mutantsof A. nidulans.
Theinducedexpression ofavr9 underconditionsof nitrogen limitation andtheputative
binding sites in the avr9promoter for a major nitrogen regulatory protein suggest that the
AVR9 protein might be involved in nitrogen metabolism. Data from Neurosporaand
Aspergillus indicate, however, that of themanygenes regulated by NIT2 and AREA, some
are not directly involved in nitrogen metabolism.
Bacterial plant pathogens possess hrp genes, required for the induction of the
hypersensitive responseinresistant plantsand for pathogenicity in susceptibleplants. These
genesare highlyexpressed whenthebacteria aregrowing in thehostplant, when compared
to levels of expression in vitro.Thehrpgenes are induced invitroin response to different
cultureconditions.Thehrp genesofXanthomonascampestris pv. vesicatoria areinducedby
sucrose and sulphur containing aminoacids (Schulteand Bonas, 1992a&b). Theexpression
ofthehrpgenesofPseudomonassyringaepv.phaseolicola wasshowntobeaffected bypH,
osmotic strength and typeof carbon source (Rahmeetal., 1992).Theavirulence gene avrB
ofPseudomonas syringae pv.glycinea couldbeinduced invitro by manipulating thecarbon
source (Huynh etal., 1989). So far, only the hrploci of Erwiniaamylovora were clearly
shown to beinduced under conditions of nitrogen deficiency (Wei etal., 1992).
Thebacterial genes described above and theavirulence geneavr9of C.fulvum are all
induced under different growth conditions invitro. Specific plant substances are apparently
not required for the expression of these genes. This is in contrast to the nodulation and
virulencegenes ofRhizobium andAgrobacterium, respectively (Peters etal., 1986;Stachel
etal., 1985), which are induced by specific secondary plant metabolites.
The most interesting question is whether the high level of avr9expression in theplant
(Fig. 5) is due to low nitrogen concentrations in the apoplast of the tomato leaf where the
fungus grows. Nitrogen and other nutrients aretaken up from thesoilby theplantandmay
be metabolized in the root before being transported to the leaves through the xylem. The
major nitrogen compounds present in the xylem fluid of tomato are glutamine and nitrate
(Van Die, 1958). The concentration of nitrogen in the xylem, albeit dependant on the
availability and source of nitrogen in the soil, is usually in the range of 0.01% to 0.20%
(Pate, 1980).However, thenitrogen contentoftheapoplastcannotbemeasured accurately.
Apoplasticnitrogen concentration isdetermined byimportviathexylem,absorption bycells
andexportviathephloem(GrignonandSentenac, 1991).InC../«/vwm-infectedtomatoleaves
thefungus utilizesthisapoplasticnitrogen, therebyreducingtheconcentrationintheapoplast
which in turn could induce avr9 gene expression. The constant flow of nutrients into the
apoplast of the leaf via the xylem would support the growth of C.fulvum although the
absolutenitrogen concentration would remain limiting.Detailed avr9 promoter analysis will
reveal whether thenitrogen responsive elements in thepromoter areindeed responsible for
the high expression of avr9in C. fulvum growing inplanta.
94
Experimental procedures
Mediaandgrowth conditions
CladosporiumfitlvumCooke (syn. Fulviafiilva(Cooke) Cif) wasgrown onpotato dextrose
agar (PDA) or in liquid B5 medium in shake cultures at 22°C unless otherwise stated (De
Wit and Flach, 1979).Ten-day-old PDA cultures of C. fulvum were used toprepare spore
suspensions for inoculation of plants and for liquid shake cultures (5.105 conidia ml"1).
Fourteen-day old tomato seedlings were inoculated with spore suspensions as described
previously (VandenAckervekenetal., 1992).ModifiedB5mediumwithoutnitrogen(B5-N)
contained,perliter, thefollowing components:20gsucrose, 250mgMgS04.7H20, 150 mg
NaH2P04.H20, 150 mg CaCl2.2H20, 134 mg K2S04, 37.2 mg Na2EDTA, 27.8 mg
FeS04.7H20, 1mgKI, 1ml micronutrients; thepH was adjusted to 6.0 with KOH before
autoclaving. One ml micronutrient solution contained the following components: 10 mg
MnS04.H20, 3 mg H3B03, 2 mg ZnS04.7H20, 0.25 mg Na2Mo04.2H20, 0.025 mg
CuS04.5H20 and 0.025 mg CoCl2.6H20. Nitrogen was added from a sterile stock solution
(1 M nitrogen: 1M KN03, 0.5 M (NH4)2S04, 1M glutamate or 0.5 M glutamine). The
induction experiments in vitro were performed with mycelium, which was grown for 48
hours in 25 ml B5 medium containing 50 mMnitrate in 100ml conical flasks at 22°Cand
100strokes min1. Themyceliumwasfiltered over miracloth orcheesecloth andrinsed once
with the test medium. The collected mycelium was subsequently resuspended in 25 mltest
medium and grown for an additional 24 hours under different growth conditions. After 24
hours, the mycelium was collected by filtration over a Büchner funnel, freeze-dried and
stored at -20°C until used for RNA analysis.
RNA isolation andnorthern blot analysis
RNA was isolated from freeze-dried mycelium using guanidine HCl extraction buffer and
LiCl precipitation as described by Van Kan et al. (1991). Following electrophoresis in
denaturing formaldehyde-agarose gels(Maniatisetal., 1982)RNAwasblotted overnightto
HybondN + membranes with lOxSSCbuffer. Northern blotswerehybridized overnightwith
[a-32P]dATPrandom primed labeled DNA probes at 65°C in modified Church buffer (0.5
MNaP04, 1 mMEDTA,7% SDSpH7.2).After 16hourshybridization, blotswerewashed
to a final stringency of 0.5xSSC, 0.1% SDS at 65°C. The mRNA accumulation was
visualizedbyautoradiography andquantified bydensitometricscanningwithaCybertechCS1 imagedocumentation andprocessing system (Cybertech, Berlin).
Determination ofnitrate
The concentration of nitrate in culturefiltrateswas determined essentially as described by
Singh (1988). Acetic acid was added to the culturefiltrateto a final concentration of 2%.
Thefiltratewas further diluted with 2% aceticacid to giveafinalabsorbance at 540nmof
0 . 1 - 1 (equivalent to approximately 0.2 - 2 mMnitrate). Fifty mg of powder mixture (3.7
g citric acid, 0.5 g MnS04.H20, 0.2 g sulphanilamide, 0.1 g N-1-naphtyethelenediamine
95
dihydrochloride and0.1gpowdered zinc) wasadded toeach 1ml sample. The samplewas
mixed thoroughly and centrifuged for 2 minutes. Theabsorbance of thecleared supernatant
was measured spectrophotometrically at 540 nm and theconcentration of nitrate calculated
by reference to a standard curve.
Construction ofaw9promoter- GUS fission
pNOM102,aplasmidcontaining thegpdpromoter ofAspergillus nidulans, theE. coli uidA
gene, coding for ß-glucuronidase(GUS,Jefferson etal., 1987)and theterminator of the.A.
nidulanstrpC gene (Roberts et al., 1989) was partially digested with Ncol. Full-length
linearized plasmid DNA was treated with Klenow large fragment DNA polymerase in the
presence of dNTPs to fill up the Ncol sticky ends. After blunt-end religation and
transformation, plasmid pCF20was selected thatcontained a singleNcolsiteatthejunction
between thegpdpromoter andtheuidA gene. A2.5kbEcóSlINcólfragment containingthe
gpdpromoter wasdeleted from pCF20andreplaced by theavr9 promoterpresent ona3kb
EcoRI/BspHIfragment that was isolated from pCFl (Van den Ackerveken et al., 1992),
resulting in plasmid pCF21 (Fig. 4). Three avr9promoter deletions (pCF23, -24 and -25)
wereconstructed bydigestion ofpCFl (with Sail,SpelorSstI, which areuniquein the 3.8
kbavr9 insert ofpCFl), isolation of thelinearized plasmid DNA, creation of anEcóBIsite
byligationofanEcöSl linker, andligationofthenewEcöRI/BspïHfragment inEcdSllNcoI
digested pCF20 (Fig. 4).
Analysisof C. fulvum transformants
C.fulvum (race 4 and 5) was grown for 48 h in liquid B5 medium, the mycelium was
collected and used for theisolation ofprotoplasts (Harling etal., 1988). Co-transformation
(Oliver et al., 1987) of 107 protoplasts with 2 jtg pAN7-l containing the hygromycin B
resistance gene (hph,Punt etal., 1987)and4jtg of eitherpCF21, -23,-24or -25,resulted
in stable hygromycin B resistant transformants. Transformants were analysed by Southern
blot analysis for thepresence and copy number of the co-transformed avr9promoter-GUS
constructs.Transformants withalowcopynumberweresubsequentlytestedbynorthernblot
analysis for the accumulation of avt9 and uidAmRNA. The expression of both genes was
tested in C. fulvum growing inplanta at ten days after inoculation and in C.fulvum grown
invitroafter transfer of 48 h old mycelium to B5 medium with 50 mMnitrate and growth
for oneadditional day.
ß-Glucuronidase assay
Histochemical localization of ÛVA9-GUS expression in C.fulvum growing inplanta was
assayedbyvacuuminfiltration ofinfected tomatocotyledonswithasolutionconsistingof50
mMsodiumphosphate (pH7.0), 1mMpotassium ferri/ferro cyanide, 0.05% TritonX-100
and0.5mg/mlX-gluc(5-bromo-4-chloro-3-indolylß-D-glucuronide, Sigma).Theinfiltrated
cotyledons were incubated at 37°C for eight hours in the dark. Finally, cotyledons were
96
cleared of chlorophyll by incubation in 70% ethanol and analysed by light microscopy.
Acknowledgements
TheauthorsliketothankW.Tijs forperforming northernblotexperiments,Mrs. C. Hanhart
(Departmentof Genetics,Wageningen) forprovidinguswithtomatocellsuspension cultures
andC.M.J. Pietersefor criticallyreadingthemanuscript. Theinvestigationswere supported
by theFoundation for BiologicalResearch (BION), which is subsidized by theNetherlands
Organization for Scientific Research (NWO).
References
Caddick, M.X. (1992) Characterization of a major Aspergillus regulatory gene, areA. In MolecularBiology
ofFilamentous Fungi, (Stahl, U. and Tudzynski, P., eds.), VCH, Weinheim, pp. 141-152.
De Wit, P.J.G.M. (1977) A light and scanning-electron microscopic study of infection of tomato plants by
virulent and avirulent races of Cladosporiumfiilvum.Nelh.J. PlantPathol. 83, 109-122.
De Wit, P.J.G.M. (1986) Elicitation of active resistance mechanisms. In Biology andMolecularBiologyof
Plant-PathogenInteractions, NATOASISeriesVolHI, (BaileyJ., ed.),SpringerVerlag,Berlin,pp. 149169.
De Wit,P.J.G.M. andFlach, W. (1979)Differential accumulation ofphytoalexinsintomatoleaves, butnot
in fruits after inoculationwithvirulentandavirulent races of Cladosporiumfulvutn. Physiol.PlantPathol.
15, 257-267.
DeWit,P.J.G.M. andSpikman, G. (1982)Evidence for theoccurrence ofrace-andcultivar-specific elicitors
of necrosis in intercellular fluids of compatible interactions between Cladosporiumfulvum and tomato.
Physiol.PlantPathol. 15, 257-267.
Fu, Y.H. and Marzluf, G.A. (1990) nit-2, the major positive-actingnitrogen regulatory geneofNeurospora
crassa,encodes a sequence-specific DNA-bindingprotein. Proc. Natl.Acad.Sei. USA 87, 5331-5335.
Grignon, G. andSentenac, H. (1991)pHandionicconditionsintheapoplast.Annu.Rev. PlantPhysiol. Plant
Mol.Biol. 42, 103-128.
Harling, R., Kenyon, L., Lewis, B.G., Oliver, R.P., Turner, J.G. and Coddington, A. (1988) Conditions
for efficient isolationand regeneration of protoplasts from Fulviafulva. J. Phytopathol.122, 143-146.
Huynh,T.V., Dahlbeck,D. andStaskawicz,B.J. (1989)Bacterialblightofsoybean:regulationofapathogen
gene determining host-cultivarspecificity. Science245, 1374-1377.
Jarai, G., Truong, H.N., Daniel-Vedele, F. and Marzluf, G.A. (1992) NTT2, the regulatory protein of
Neurospora crassa, bindsupstreamofnia,thetomatonitratereductase gene,invitro. Curr.Genet. 21,3741.
Jefferson, R.A., Kavanagh, T.A. andBevan, M.W. (1987) GUSfusions: ß-glucuronidaseas asensitiveand
versitalegene fusion marker inhigher plants. EMBOJ. 6, 3901-3907.
Maniatis, T., Fritsch,E.F. andSambrook,J. (1982)Molecular Cloning, aLaboratoryManual.ColdSpring
Harbor Laboratory Press, New York.
Marmeisse, R., Van den Ackerveken, G.F.J.M., Goosen, T., De Wit, P.J.G.M. and Van den Broek,
H.W.J. (1993)Disruption of the avirulence geneavr9 in tworaces of the tomatopathogen Cladosporium
fulvum causes virulence on tomato genotypes with the complementary resistance gene CJ9. Mol. PlantMicrobeInteract.Inpress.
Marzluf, G.A., Kanaan,M.andFu,Y.H. (1992)MolecularanalysisofDNA-binding,trans-actingregulatory
proteins ofNeurospora crassa.InMolecularBiologyofFilamentousFungi, (Stahl, U. and Tudzynski, P.,
eds.), VCH, Weinheim, pp. 153-166.
Oliver,R.P., Roberts,I.N.,Harling,R., Kenyon,L., Punt,P.J., Dingemans,M.A. andVandenHondel,
C.A.M.J.J. (1987) Transformation of Fulviafulva, a fungal pathogen of tomato, to hygromycin B
resistance. Curr. Genet. 12,231-233.
97
Pate, J.S. (1980) Transport and partitioning of nitrogenous solutes.Annu. Rev. PlantPhysiol.31, 313-340.
Peters, N.K., Frost, J.W. and Long, S.R. (1986) A plant flavone, luteolin, induces the expression of
Rhizobium melilotinodulationgenes. Science233, 977-980.
Punt, P.J., Oliver, R.P., Dingemans, M.A., Pouwels, P.H. and Van den Hondel, C.A.M.J.J. (1987)
Transformation ofAspergillus based on thehygromycin Bresistance marker from Escherichiacoli.Gene
56, 117-124.
Rahme,L.G., Mindrinos,M.N. andPanopoulos,N.J. (1992)Plantandenvironmentalsensorysignalscontrol
the expression of hrpgenes inPseudomonas syringaepv. phaseolicola. J. Bacteriol. 174, 3499-3507.
Roberts, I.N., Oliver, R.P., Punt, P.J. and Van den Hondel, C.A.M.J.J. (1989) Expression of the
Escherichia coliß-glucuronidasegeneinindustrialandphytopathogenicfilamentous fungi. Curr.Genet. 15,
177-180.
Scholtens-Toma, I.M.J, and De Wit, P.J.G.M. (1988) Purification and primary structure of a necrosisinducingpeptidefromtheapoplasticfluidsoftomatoinfected withCladosporiumfitlvum (syn. Fulviafitlva).
Physiol.Mol. PlantPathol. 33, 59-67.
Schulte,R. andBonas,U. (1992a)ExpressionoftheXanthomonascampestrispv.vesicatoriahrpgenecluster,
whichdeterminespathogenicityandhypersensitivityonpepper andtomato,isplantinducible.J.Bacteriol.
174, 815-823.
Schulte, R. and Bonas, U. (1992b) AXanthomonas pathogenicity locus is induced by sucrose and sulphur
containing amino acids. Plant Cell4, 79-86.
Singh, J.P. (1988) Arapid method for determination of nitrate in soiland plant extracts. Plant and Soil110,
137-139.
Stachel, S.E., Messens, E., Van Montagu, M. and Zambryski, D. (1985) Identification of the signal
molecules produced by wounded plant cells that activate T-DNA transfer inAgrobacterium tumefaciens.
Nature318, 624-629.
VandenAckerveken, G.F.J.M., VanKan,J.A.L. andDe Wit, P.J.G.M. (1992)Molecular analysis ofthe
avirulence geneavr9of the fungal tomatopathogen Cladosporiumfitlvumfully supports the gene-for-gene
hypothesis.PlantJ. 2, 359-366.
Van den Ackerveken, G.F.J.M., VanKan, J.A.L., Joosten, M.H.A.J., Muisers, J.M., Verbakei, H.M.
and De Wit, P.J.G.M. (1993) Characterization of twoputativepathogenicity genes of the fungal tomato
pathogen Cladosporiumfitlvum.Mol.Plant-Microbe Interact.6, 210-215.
Van Die, J. (1958) On the occurence of a-keto acids and organic nitrogen compounds in xylem exudates of
cucumber and tomatoplants. Proc. Kon. Ned. Acad.Wetensch. AmsterdamC61,572-578.
Van Kan, J.A.L., Van den Ackerveken, G.F.J.M. and De Wit, P.J.G.M. (1991) Cloning and
characterization of cDNA of avirulence gene avr9 of the fungal tomato pathogen Cladosporium fitlvum,
causal agent of tomato leaf mold. Mol.Plant-Microbe Interact.4, 52-59.
Wei, Z-M., Sneath, B.J. and Beer, S.V. (1992) Expression of Erwiniaamylovera hrp genes in response to
environmental stimuli.J. Bacteriol. VIA,1875-1882.
Wolters,A.M.A., Schoenmakers, H.C.H., VanderMeulen-Muisers, J.J.M., VanderKnaap,E.,Derks,
F.H.M., Koornneef, M.andZeker, A. (1991)LimitedDNAeliminationfrom theirradiatedpotatoparent
infusionproductsofalbinoLycopersiconesculentum andSolanumtuberosum.Theor. Appl. Genet. 83,225232.
98
Chapter 7
Isolation and characterization of two putative
pathogenicity genes of Cladosporium fiilvum
GuidoF.J.M. vandenAckerveken, JanA.L. vanKan, MatthieuH.A.J. Joosten,
José M. Muisers, HenkM. Verbakei andPierreJ.G.M deWit
adaptedfromMolecularPlant-Microbe Interactions 6 (1993), 210-215
Summary
Thefungus Cladosporiumfiilvumisabiotrophicpathogen of tomato. Onsusceptibletomato
plants, thefungus growsabundantly intheextracellular spacesbetween themesophyllcells.
Themechanism by which C.fiilvumis abletoestablish and maintainbasic compatibility on
itsoneand only host species, thetomato, isunknown. Theisolation andcharacterization of
pathogenicity factors and the corresponding genes will provide insight into the mechanism
by which C. fiilvum colonizes its host. Two putative pathogenicity genes of C. fiilvum,
encoding proteins which occur abundantly in the extracellular space of infected tomato
leaves, were isolated and characterized (ecpl and ecp2). TheDNA-sequences of theseecpgenes (encodingextracellularproteins)donot sharehomology withany sequencepresentin
theDNAdatabases.Theecp-genesarehighlyexpressed inplantabutnotinvitro,suggesting
that they play a significant role inpathogenesis.
101
Introduction
Theinteractionbetween thebiotrophic fungal pathogen Cladosporiumfiilvumand tomatois
confined to the leaf surface and the extracellular space between the mesophyll cells. No
specialized structures, involved in penetration and nutrient uptake, could be detected
microscopically(DeWit, 1977;LazarovitsandHiggins, 1976a; 1976b).Thepathogenisable
togrowin theextracellular space without destroying hostcells. Studieson the carbohydrate
composition of the apoplastic fluid of C. fulvum-infecteà leaves indicated that the fungus is
probablydependent onextracellular sucroseasitsmaincarbon source(Joostenetal., 1990).
Incompatibility between C.fiilvumand tomato is caused by the recognition of fungal
avirulence gene-products, the so-called race-specific elicitors, by the resistant host leading
totheactivation ofthehypersensitive response (HR),whichrestricts thepathogen tothesite
ofinfection (VandenAckerveken etal., 1992).IncompatibleinteractionsnoHRisinduced
and other defence responses, such as the accumulation of PR proteins, are activated much
laterascompared toincompatibleinteractions(JoostenandDeWit, 1989).Theaccumulation
ofPRproteinsandother compounds incompatibleinteractions ismostprobably induced by
general stress and possibly by cell wall fragments released from the fungus during
pathogenesis.
ThewayC.fiilvumcolonizestomato,withouttheformationofspecialized structuresand
the production of damaging hydrolytic enzymes, raises questions about the mechanism by
which the fungus establishes and maintains basic compatibility. The isolation and
characterization of putative pathogenicity factors and their corresponding genes should
provide aninsight intothestrategy adopted by C.fiilvumto successfully infect itsoneand
only host species, thetomato.
In compatible C.fiilvum- tomato interactions several low molecular weight proteins
(<20kD)accumulatein theextracellular spaceduring theinfection process (DeWit et al.,
1989).Theseproteins are thought tobeof fungal origin, as they arenot detected in healthy
plants nor in incompatible C. fiilvum - tomato interactions. These putative pathogenicity
factors might play an important role in establishing basic compatibility. Two such
extracellular proteins, ECP1 (synonymous to PI; Joosten and De Wit, 1988) and ECP2
(Wubben etal., inpreparation) havebeenpurified andpolyclonal antisera havebeenraised
against these proteins. Western blot analyses indicated that proteins ECP1 and ECP2 are
neitherpresent in C.fiilvumgrown invitro,nor in healthy tomatoplants. In thispaper we
describe the isolation and characterization of the genes ecpl and ecp2. They are shown to
be of fungal origin and are highly expressed in C.fiilvum-infectedtomato leaves, whereas
their expression invitroislow or not detectable.
102
Results
Isolation oftheecpl gene
TheECP1protein waspurified aspreviously described (Joostenand DeWit, 1988)andthe
sequence of the 36 N-terminal amino acids was determined. A degenerate oligonucleotide
probe was designed, complementary to the derived mRNA sequence encoding amino acids
27-32 (Fig. 1). By using this probe on northern blots, a clear hybridization signal was
detected specifically inRNAisolated from compatible C.,/w/vHm-tomatointeractions (results
not shown).In order toobtain a 100% matchingprobe, RNA from acompatible C.fulvumtomatointeractionwasusedastemplateforRNA-sequencing usingtheoligonucleotideprobe
as primer. The sequence of the RNA confirmed the N-terminal amino acid sequence of
ECP1, andallowed thesynthesisofaperfectly matching oligonucleotideprobe (aminoacids
11-20,Fig. 1), which was used to screen a XZAPlibrary containing cDNA obtained from
C. fulvum infected tomato leaves. Severalpositive clones were selected of which two were
analyzed in detail. Both clones contained an insert encoding aprotein which corresponded
totheaminoacid sequence obtained for ECP1. ThecDNA insert was subsequently used to
screen agenomiclibrary of C.fulvum. A7kbXholfragment containing theecpl genewas
subcloned from a purified positivephage (pCF140).
Isolation oftheecp2 gene
Antiserum raised against the ECP2 protein was used to screen a Xgtll expression library
containing cDNA obtained from C.fulvum-infeetedtomato leaves. Two positive phages,
containinginserts of 600and 650bp, werepurified and tested onanorthern blot containing
RNA from C.fulvum grown in vitro, healthy tomato and C. fulvum infected tomato,
respectively. The 600bpEcóKL insert was shown to be of fungal origin and hybridized to
mRNA whichwashighlyabundantincompatibleC.fulvum -tomatointeractions (resultsnot
shown).ThecDNA insert was subsequently used to screen agenomic library of C.fulvum.
Severalpositiveclones were obtained and a 4.3 kbEcoJU/BamEI fragment, containing the
ecpl gene, was subcloned (pCF170).
Sequence analysis ofecpl and ecp2
Detailed restriction maps of the subcloned DNA fragments containing the ecpl and ecpl
genes were obtained by Southern analysis (Fig. 2). Theecp-genes are singlecopy genesin
allraces of C.fulvum tested (datanot shown). TheDNA sequence of thecoding regionand
approximately 1 kb of the upstream region was obtained for both genes (Figs. 3 and 4).
Introns were identified by comparison of cDNA sequences with genomic sequences. Two
introns were found in ecpl, and one in the ecpl gene. The transcription start sites, as
determined by primer extension, and polyadenylation sites are indicated infigures3 and 4
for ecpl and ecpl, respectively. Apossible TATA-box and CAAT-box arepresent inboth
genes in the 5'-upstream region.
103
<
Ü
àS°
U <
m
i
u
u
9°
< a
<
S
u
1"
1
u o
CO
i < a
1U
c
<
0
M
eu
Ä
o
£
U
s
1
c
m
<
3
a
^
EH
H
EH
U
8
u
S
è
à
CD
§
3
S
1
u
u
M
<D
CO
M
E-«
M
H
o
9
O
<
(H
< u
3
Su
r
3°
fi
<
1"
M
>>
5 to
m
3 G
g
H
The amino acid sequence which was determined for the purified proteins (ECP1:Fig. 1,
ECP2: seebelow)wasconfirmed bythecDNAandgenomicsequences.Thematureprotein,
as isolated from C.,/w/vM/n-infected tomatoplants, isindicated inbold (Figs. 3and 4).Both
precursor proteins contain a signal sequence which is hydrophobic. The signal peptide
cleavage site, as predicted by the computer program SigSeq (GCG), is indicated by the
arrows (Von Heijne, 1986). Comparison of both DNA and protein sequences with the
GenBank, EMBL, PIR and SwissProt databases (Devereux et al., 1984) did not reveal
significant homology toany known sequence.
X
ecpl
TK
s'l
C
ll
I
U
I
C P
KVPP
H
M
P HX
I I IIII jjaila l lia'
->—>
X
ecp2
s'l
I
UC HC
\VÀYM\\
P
E
1 kb
1 zJ3'
<—> <—>
Figure2. Restriction mapof sequences surroundingtheecpl and ecpl open reading frames (ORF,hatched
bar). The restriction map is derived from Southern blot analysis of genomic DNA digested with several
restriction enzymes and probed with the ecpl and ecpl cDNAs, respectively. The sequencing strategy is
indicatedbythearrows. B,BamHl;C, Hindi; E, £coRI; H, Hindlll;K, Kpnl;M, Smal;P, Pstl;S,Sail;T,
Sstl;U, PvuII; V, EcoKV;X,Xhol
105
CTGCAGTGTGCCATGCGCTTATTACAGACGACTTCTCGGGGATACCTGCAGGAAGTGGAC
-605
GTTCGTGCTTTGATGAGTTGAAGCÂCTCCCCGAGÀGTCGAGACCGATTGTGACCGTCATT
-545
TGCGCTTCCACGTTCGCGGTTCGCCGCATTCGAGGCCGGGATGGGTTTGGAAAGTCGCTG
-485
TTAAÀGAAAGGCGTÀGTTGCCATTCCTGGGCGTCÀCTTCGGGGCCGCGCGCGATTTAGCT
-425
CATCGAGCCAGGTTGCCACATGCCTGCCCTCCCACGTGCAACGATCATGTTGGTÀCGAAG
-365
CCGCÂGAACACTGCÀCTTTCACGGÀCTCGTAACTCGTAACATCTÀCCAAGGTCATCTCGA
-305
ATCTCGGCCATCTCGCAAGCGCACGGGTATGTCCTCTCACGACCÀGGATAAGTTCGCAGT
-245
CCGGÀCGGCGTATTTACCAGCAAGCTTACGAGATGGCCGCGATTCGAGATGACCCTGGTA
-185
GATGTTATACÄÄTGAGTTGCGGAACTTCGCGGAGCTTCGTAATATTGTGCAGCGTACTGT
-125
ACAATGCTTGATGTÀAGCTTGGCACAACATCCAAGAAGCCTGATCGGTGCGGCCÀGTGCT
-65
TGTAGACGGCACCCTAGGTGTGCAÀCCTATAAAGCAGCCCGTCTGCGCTGCGCATTATTG
T
- 5
CTTCATCACATCCTTAGTTACCGAGACATCACCACTCCCACTTCCCTCCAAGATGCACTT
M H F
CGTCACATCTCTCÀTCGCCGGCGTCGCCATGCTCGCCACGGCGÀGCGCCACTGTCCAAGG
V T S L I A G V A M L A T A S A T V Q G
56
116
CGGCGCCCCCGTGGACGATCTCAÀGTTCGCAAAÀAAGTTCAACCAGAACTGCCÀGCAAAT
G A P V D D L K F A K K F N Q N C Q Q I
176
CTCTGGCGGTCCAÀACGGGTAAGÀCCCTCGACCCTCATCCAACCCACTCGGCCÀCAATAG
S G G P N G
-
236
ATCGCGAAGACTCCACGAAACTAÀCACGAATGATACCATCTTAGAGCGATCTGCCCAGAC
296
-
---
A I C P D
GGCGACTTGGTATGACCACTCGTÀCCCCTCAGCGATCGATGCTGTCCTCGAACTAATGCG
G D L
356
GGAÀTGCCCCGGGÀACAGTATTGGTGCAAAGACGGTAGGGCCATATTTTGCCAGACATGC
416
Y
W
C
K
D
G
R
A
I
F
C
Q
T
C
CAAÀCTGGATGTACCGCCGATGAÀAATGGCAAGGTCGGCTACTGCAACGAAGGGCCTACA
Q
T
G
C
T
A
D
E
N
G
K
V
G
Y
C
N
E
G
P
476
T
AACCCCAAGTGCCTTTAATGGAGGAAGCATCGCGGAGGCTGAATGCCACCGTTÀCCATTT
N P K C L
536
GAGÀCGATGTGGAGAATGTGCCCTTCTACGTCGÀGGGGCAGAGCGACCGTACGCCTTACG
596
TAGÀCATCCAAGCÂACGAGGAGACGTTAAGCCCTGTAATTGCTCAAATCGTCTÀTCCCTC
v
.
TAGATGCCCAGCTTCCCTTGCCTAAAGCGAAACGTGCGCATCCTGTGAGCGCAACGACTC
656
716
CAGCTCGGCCTGCÀGTGGACCGCTCGTCGATGTÀTTGGGCAGCCACAGCTTCCCGCCCAT
776
CCTCTACACCAAAGCAGACCGGTGTTATACAGCCACACTAGCCCAGTACGGCCCCATACA
836
GCGÀGATCGTCGCÂCTAAACATTÂGCGAAGAGCÀTGATCCACGGCCTACATGTCTTTCGC
896
CCTGTAGTAGACAÀACCGCAGACTTCTACATTTÀTTGAGCGCAÀCGACCGACGÀGCGCTA
956
GCCÀCACCGTAGCGACGTATAAGCTT 982
Figure 3. Sequence of the C. fiilvum gene ecpl (EMBL accession number Z14023). The gene encodes a
precursorproteinof96aminoacidsandisinterruptedbytwoshortintrons(—). Thepredictedsignalsequence
cleavage site is indicated by an arrow (4). The mature ECP1 protein of 65 amino acids is located at the
carboxyl-terminusoftheprecursorprotein(bold).AputativeTATA-box(TATAAA1 islocated37bpupstream
of the main transcriptionstart (T). The5' upstream region containsaputative CAAT-box (overlined)located
atposition-175. In themRNA, the3'untranslatedregionof 170bpis followedby apoly(A)tailstartingat the
positionindicated(v).
106
GATCGGAGGTGAACAGAGACTGTAGGTGTTTATGAGACGGGGGCAATCCTAGTGCGAGGC -630
TGTGGTTAGCCGAAAATCGTTGTCAATCGTCTGTCAAATTAGCATCGTCAAGGGTTGCAG -570
CCAAGGTTTCCTTGTCTTCÀGATGGAAACTAGCGGGACTÂTTTGGCGGTGAGTCTTCAGÀ
-510
GGGATGCAAGTGGATGCTCGTACTCCCAGGTCGTGACGCGGCACGTGGCGCTACTCAGAÀ
-450
CAACGAGTGTCACCACCTTTAGCAACATACAGCGTCTGTCAGCACGGAGÀACATCTTTAC
-390
TTAGCGTTCGCAGTGTTTGCGCCGGCCGATGCTATGATCCTGCTTTCCGCTGGGATAAGÀ -330
CGATCTGCACTAACGGGAAÀTGTAACTCCGCTGCTGACAGACGCTGTACGATTCCAAGAC
-270
GAAGGAGTTTACCGTCTAGGTGTCAAATCCTAGATCCCAGGAAATCCAAÀTCTGGACCCT
-210
TGGCTACATGAGTTACGAAGCAGGTTGCAÀCAGTGCACTGCCCAAGTCGAACGCTATGGC
-150
CTGGGTAGGGCGTTTCGTGCGATGCCCAGTCCAAGGATCGTGTCTCAAGÀTGCAGCCTTÀ
- 90
CAGTTCGATCAAATTCGCCACCAGAAATGACCTCGAGTGTTGGGAACGGTATTTATAGTC
-30
CCCTAAGTATGCGCCTCAACGCCCACCAAGTTTACCTTCACCGAACTTCGAACAACAGTA
31
CACTCTCCÀCACAACACCÂCTCTTCCTCCTGTGCCAACÀTCACACGACCATACAACCAGC
TTCACGATGCTCTTCAACGCCGCCGCCGCCGCCGTGTTCGCCCCTCTGCTAGTCATGGGC
M L F N A A A A A V F A P L L V M G
4.
AACGTTCTGCCTCGGAACGCTGGCAACTCGCCCGGCTCAAACCGCTGCGATGCCTCAACT
N V L P R N A G N S P G S N R C D A S T
91
151
TTTAACAACGGCCAAGACTTTGACATTCCACAGGCACCÀGTCAATGACTGCCGGCAGATG
F N N G Q D F D I P Q A P V N D C R Q M
271
GTTGAAAACATCAATAGAGATTCTCAATTCTCAGTGTCCCACAGCTGGGCGAGACCATTC
V E N I N R D S Q F S V S H S W A R P F
GGCGGATATGGCGATTGTÀCGTCACAGAGCCTCCCACGÀCAATTCTTGAATATACTGACA
G GYGD C
TACCATTTCCAGGTGCATTCAACGTCCGCGTCATTGCTGGCTGGCGGAÀTGGCTTGGTTG
A F N V R V I A G W R N G L V G
331
391
GCGGTGCGGATGCCGTTGÀCCTTCTCACTGATTCTGTCAAAAACTTCGGTGAGGCCAACA
G A D A V D L L T D S V K N F G E A N K
511
AGGTTTCCÀGCAAGGGCACCTACAACCAAATTGTCTCCGCGGAGGGCGAAGTTACGTGCG
V S S K G T Y N Q I V S A E G E V T C D
571
ATTCTGTGGATCGTGGTGGTCAGGTCAGÀGTTCAGTGGÀTTGTTGCCAGCTCATCGTACA
S V D R G G Q V R V Q W I V A S S S Y N
631
ACCCGTCCAACGATGACTÀGAGGCCGCGTCAGACGGGTTTGCTAGAACCAGGTTGTAGCA
P SN DD
GAGCTGAACATGGCATTTCTGGTGTCGAÂGCTTGGGCATAATCAAATATACAGAGCAACG
v
GAATGGATCTTGAAGCTAGTAGTCAACCATACGACTTGATCTTTCAACCACACACAGCTC
691
811
TGATACAGÀCATCACAGTTCGGCATCCACACTAAAGCCTTTCGATAACCGTCCATGCCTC
871
CAGCCTGATTCGCAGAATCGTCGAACCAGACCATCATCTCCTGGTCTGCTTCGGCAGCÀG
931
211
451
751
TCATGATC 939
Figure 4. Sequence of the C. fitlvum gene ecp2 (EMBL accession number Z14024). The gene encodes a
precursorproteinof 165 aminoacidsandisinterruptedbyoneshortintron(—). Thepredicted signalsequence
cleavage site is indicated by an arrow (I). The mature ECP2 protein of 142 amino acids is located at the
carboxyl-terminusoftheprecursorprotein(bold).AputativeTATA-box(TATAGT) islocated36bpupstream
of the main transcription start ( T ) . The5' upstream region containsaputative CAAT-box (overlined) located
atposition -80.In the mRNA, the 3' untranslated region of 125bp is followed by a poly(A)tail starting atthe
positionindicated (v).
107
Expression ofecpl andecp2invitroandinplanta
The cDNAs of both ecpl and ecp2 were used as a probe to hybridize to northern blots
containingpoly(A)+RNAisolatedfrom anincompatibleandacompatibleC.fiilvum-tomato
interaction at different times after inoculation (Fig.5). Nohybridization wasdetected in
healthyplantsnorinincompatibleinteractions. Theecp2transcript ispresentatalow level
in inwro-grown C.fiilvum,whereas theecpl transcript could notbedetected under these
conditions. The time course of accumulation of the ecpl and ecpl transcripts is rather
similar, andfollows theincreaseoffungal biomassintheinfected leaves.Theexpressionof
the ecp-genes in thecompatible interaction, as compared tothefungus grown invitro,is
highlyinduced, takingintoconsiderationtheminorproportionoffungalRNAintheinfected
leaf samples.TheC.fiilvumactin gene (act) wasusedasaconstitutivecontroltodetermine
the presence of fungal mRNA in the infected tomato leaves (Van den Ackerveken,
unpublished). Theacttranscript ispresent in C.fiilvumgrown invitrobutabsent inthe
incompatible interaction because of thelow fungal biomass in theleaf. Thelevel ofact
transcriptincreasesduringthecompatibleinteractionfollowingtheincreaseinfungalbiomass
but stays significantly lower than thelevel ofinvifro-grown C. fiilvum.
Cf4/5
Cf5/5
P F 4 6 8 101214 4 6 8 101214
ecpl
0.5
"**
act
•••«„
— «*
1.8
Figure 5. Timecourse ofaccumulationoftheecpl, ecpl andactmRNA. ThecDNA-inserts (eçp-genes) and
genomic subclone {act)were labeled by random priming andhybridized toa northern blot containing equal
amountsofpoly(A)+-RNAisolatedfromuninoculatedtomatogenotypeCf5(laneP), fromC.fiilvumgrownin
vitro(lane F),from anincompatible interaction Cf4/race 5 (Cf4/5) andacompatible interaction Cf5/ race 5
(Cf5/5) atdifferent times(4,6, 8, 10,12, 14days, respectively) after inoculation.Thesizesofthehybridizing
mRNAsinkilobasesis indicated.
108
Further characterization ofproteinsECP1andECP2
TheTricine-SDS-PAGE system (Schäggerand von Jagow, 1987) was applied to determine
themolecular weight of ECP1protein. Theconventional SDS-PAGEsystem was not suited
for the separation of proteins smaller than 14kD. Acrylamide gels of 16.5% T and 6% C
enabled us to clearly separateproteins in the5 - 20kD range. The molecular weight of the
ECP1 protein was estimated to be 9 kD, which is in agreement with the molecular weight
deduced from theDNA sequence.
Toconfirm that theclonedecpl geneindeed encodestheECP2protein wepurified and
sequenced theN-terminalpart of theprotein. The sequence obtained (NH2-X A GXXP
G S N R C D A S T F N N G Q - COOH) wasin agreement with theaminoacid sequence
as deduced from the DNA -sequence. The mature ECP2 protein, as isolated from the
extracellular space, is one N-terminal amino acid shorter than the processed protein, as
predicted by the SigSeqprogram.
Discussion
Apoplastic fluid of C.yw/vM/n-infected tomatoleaves contains several low molecular weight
proteins (<20 kD) which are absent in culture filtrates of in vitro-grovm C. fiilvwn and
healthy tomato plants. The accumulation of these proteins in infected tomato plants is
correlated with theincrease in fungal biomassin theleaf, and theseproteinswere therefore
thought to be of fungal origin (De Wit et al, 1989). Two of these proteins, ECP1
(synonymous to PI; Joosten and DeWit, 1988)and ECP2 (Wubben et al., in preparation)
were purified. Amino acid sequence data of the purified ECP1 enabled us to employ
oligonucleotide probes to isolate the corresponding cDNA and gene (ecpl). Polyclonal
antiserum raised against ECP2 was used to screen a cDNA-expression library of infected
tomatoleaves, resulting in theisolation of thecDNA and gene (ecpl).
The ecpl gene encodes a precursor protein of 96 amino acids. The processed protein
is65aminoacids in size (7kD),asdeduced from theDNA-sequenceandN-terminalamino
acid sequence of thepurified protein from C.fiilvum-'mfectedtomato (Figs. 1and 3). The
originalpaper onthepurification of ECP1reports an estimated molecular weight of 14kD
(Joosten and DeWit, 1988), which is much higher than thededuced molecular weight of 7
kD. By using the Tricine-SDS-PAGE system, the molecular weight was estimated to be 9
kD, which is more in agreement with the deduced size of 7 kD. Theprocessing of the 96
aminoacidsprecursor protein tothe 65aminoacids matureprotein mostprobably involves
two steps. Firstly, theprotein is secreted and the signalpeptide is cleaved off. The signal
sequencecleavagesite,aspredictedbyVonHeijne(1986)isbetweentheaminoacidresidues
Gly(23)andGly(24)(Fig.3).Thesecond stepprobablytakesplaceintheextracellular space
andinvolvestheactivityof (plantorfungal) proteasesontheN-terminalpartofthesecreted
protein, resultinginastableprotein of 65aminoacids.Thisphenomenonhasbeenobserved
for therace-specific peptideelicitorencodedbytheavirulencegeneavr9 ofC.fitlvum,which
109
issecreted intheextracellular spaceasa40aminoacidspeptidebutissequentiallydegraded
to a very stable 28amino acids peptide (Van den Ackerveken et al., 1993).
The processing of the ECP2 protein is similar to that of the ECP1 protein. The ecpl
encoded precursor protein of 165amino acids is secreted into the extracellular space after
cleavage of the signalpeptide between residues Pro (22) and Arg (23) aspredicted by Von
Heijne (1986). Of theresulting 143amino acid pre-protein aminoacid residue Arg (23) is
mostprobably removed byproteasespresentintheextracellular space,asthematureprotein
isolated from C.fulvum-infected tomatoleaves startswith aminoacid residue 24 (XAG).
The estimated size of 17 kD from SDS-PAGE (Wubben et al., in preparation) is in good
agreement with the 142amino acidsprotein as deduced from theDNA-sequence.
The structure and organization of thetwo cloned eçp-genes of C.fulvum is typical for
filamentous fungi (Figs. 3 and 4). Motifs which are commonly found in genes of higher
eukaryotes arealsopresent ingenes of filamentous fungi, butwhether they are functional is
not always clear (Gurr et al., 1987). The introns in the ecp-genes are small, which is
common for filamentous fungi. The splice junctions (GT..AG) and internal consensus
sequence (TACTAAC) are present in all three introns. The two ecp-genes both contain a
possible TATA-box and a CAAT-box which might be involved in the regulation and
positioningoftranscription.Thetranscription startoftheecp2-geneislocatedontheCAAGsequencewhichisacommontranscription initiation siteinmanyyeastgenes.Theecp-genes
bothlackatypicalpolyadenylation signal (AATAAA)whichiscommonly found ingenesof
higher eukaryotes, but not frequently in genes of filamentous fungi (Gurr et al., 1987).
The ecp-genes and thepreviously cloned avirulence gene avr9(Van Kan etal., 1991)
are highly expressed inplanta as compared to the in vitro situation. The mechanisms of
induction or derepression for these genes are not known. The promoters of these induced
genes donot reveal common structural motifs, which could beinvolved in theregulation in
planta, although small homologies are present between ecpl and ecpl (data not shown).
Future research usingpromoter-reporter gene fusions willenable us to define the promoter
regions which areinvolved in the regulation of transcription.
The role of the ecp-genes, which are functionally present in all races of C.fulvum
tested, in establishing and maintaining basic compatibility remains unclear yet. The DNA
sequences andderived aminoacid sequences oftheecp-genes havenot givenanyindication
ofpossibleenzymaticor structural functions. Thehighexpression oftheecp-genes inplanta
as compared to invitroand theabundance of the eçp-proteins in the extracellular spaceof
infected tomato leaves suggest a role in pathogenicity. The extracellular localization as
described by Wubben et al. (inpreparation) suggests a role in the matrix which is present
between the fungal hyphae and the host cell wall. Possibly, the ecp-proteins are actively
interfering with the metabolism and/or transport of host nutrients within the tomato leaf.
Gene-disruption by transformation and homologous recombination with mutated ecp-genes
iscurrentlybeingcarried outtodeterminewhetherthesegenesareessentialforpathogenicity
of C.fulvum.
110
Experimental procedures
Subculture ofC. fulvum andinoculation oftomato
Cladosporiumfulvum Cooke (syn. Fulviafitlva (Cooke)Cif) was grown on potato dextrose
agar (PDA) or in liquid B5-medium in shake cultures (De Wit and Flach, 1979). Tomato
genotypes containing different genes for resistance to C.fulvum were inoculated with a
conidial suspension ofC.fulvum containing5x10sconidiaml1. Sixtoseven-week-old plants
were sprayed twice at the lower side of the leaf. The plants were allowed to dry and
subsequently incubated at 100% humidity for 2 days to allow the spores to germinate (De
Wit, 1977).Theinfection of susceptibletomatogenotypes canbedescribed as follows. The
fungus penetrates the leaf through the stomata at 3-5 days after inoculation, followed by
abundant fungal growth in the intercellular spaces of the leaf (days 4-11). Sporulation
becomes visible at 12 days after inoculation, and becomes intense between two and three
weeks after inoculation (DeWit, 1977;Lazarovits and Higgins, 1976a).
Purification ofcompatible interaction specificproteins
Intercellular fluid wasisolated from thecompatibletomatogenotype Cf5lC.fulvum, race5
interaction at 14 days after inoculation. Protein ECP1 was purified by cation exchange
chromatography, chromatofocusing andgelfiltrationasdescribedpreviously (JoostenandDe
Wit, 1988)and theN-terminus was sequenced. Polyclonal antiserum, raised againstprotein
ECP2, purified bygelfiltrationandanion exchangeHPLC (Wubben etal., inpreparation),
was used to screen a cDNA expression library.
Screening ofcDNA andgenomic libraries
A XZAP library containing cDNA synthesized on poly(A)+ RNA isolated from infected
leaves of a compatible tomato genotype Cf5IC. fulvum, race 5 interaction at 11 days after
inoculation (Van Kan etal., 1992)wasplated with E. colistrain PKLF' to obtain 100,000
plaques.Filterswerehybridizedwitha[?-32P]ATP5'-end-labeledoligonucleotideprobe(Fig.
1, probe II).Positivephages werepurified by replating and a second round of screening.
A Xgtll expression library containing cDNA synthesized on poly(A)+ RNA isolated
from infected leaves of a compatible genotype Cf5lia.ce. 5 interaction at 14 days after
inoculation (Van Kan etal., 1991) was plated with E. coli strain Y1090 to obtain 250,000
plaques. Three hours after plating and incubation at 42°C, nitrocellulose filters (Schleicher
and Schüell), soaked in 0.25% IPTG and dried, wereplaced on the top agar and incubated
for 4 hours at 37°C. The filters were removed, blocked with 1% gelatine in TBST (Trisbuffered Saline, Tween) and subsequently incubated overnight with polyclonal antiserum
raised against purified ECP2 (diluted 1:500 in TBST, 1% E. coli lysate, 0.5% BSA and
0.02% NaN3). Theantigen-antibody complexes weredetected with theBioRad Immun-Blot
goat anti-rabbit alkalinephosphatase (GAR-AP) assay kit. Positivephages werepurified by
replating and a second round of screening.
Ill
AX EMBL3 genomiclibrary of race5of C.fiilvumwasconstructed and screened with the
cDNA clones obtained as described (Van den Ackerveken et al., 1992). The cDNA inserts
wereused asprobesandwereradioactively labeled with [a-32P]dATPbytherandomprimer
method (Hodgson and Fisk, 1987).
Clovingprocedures andDNA sequencing
All DNA manipulations were conducted essentially as described by Maniatis et al. (1982).
DNAsequencingwasperformed usingthechaintermination methodofSangeretal.(1977).
RNAisolation,northern blotting, primerextension andRNA sequencing
RNAwasisolated from freeze-dried materialasdescribed (VanKanetal., 1991).Poly(A)+
RNA was obtained by affinity chromatography on oligo(dT)-cellulose, electrophoresed on
denaturing formaldehyde-agarose gels, andblotted ontoHybondN membranes (Amersham)
asdescribed byManiatisetal. (1982).TheC.fiilvumactingene(act) wasisolated from the
genomiclibrary using thePhytophthora infestans actA gene (Unkles etah, 1991)andused
as a constitutive control for the northern blot time course. RNA sequencing and primer
extension wasperformed onpoly(A)+RNA from acompatible C.Julvum-tomalo interaction
using 5'-end-labeled oligonucleotide primers as described (Van Kan etal., 1991).
Polyacrylamide gelelectrophoresis (PAGE)
Sodium dodecyl sulphate (SDS)-PAGE on 15% (w/v) Polyacrylamide slab gels was
performed asdescribed (Joosten andDeWit, 1988).Themolecular weight ofproteinECP1
was determined onTricine-SDS-PAGEgels (16.5%T, 6%C) as described by Schäggerand
VonJagow (1987).
Acknowledgements
These investigations were supported by the foundation for Biological Research (BION),
whichissubsidized bytheNetherlands Organization for ScientificResearch (NWO).Weare
grateful to Dr. Mösinger of Sandoz for sequencing theN-terminalpart ofprotein ECP1, to
Mogenfor synthesizing different oligonucleotideprobesandtoDr. N.Danhashfor critically
reading the manuscript.
References
De Wit, P.J.G.M. (1977) A light and scanning-electron microscopic study of infection of tomato plants by
virulent and avirulent races of Cladosporiwnfiilvum.Neih. J. PlantPathol. 83, 109-122.
DeWit, P.J.G.M. andFlach, W. (1979)Differential accumulation ofphytoalexinsintomatoleaves, butnot
infruits after inoculationwithvirulent andavirulentraces of Cladosporiumfiilvum.Physiol.PlantPathol.
15, 257-267.
112
De Wit, P.J.G.M., Van den Ackerveken, G.F.J.M., Joosten, M.H.A.J. and Van Kan, J.A.L. (1989)
Apoplastic proteins involved in communication between tomato and the fungal pathogen Cladosporium
fidvum. laSignalMolecules inPlantsandPlant-MicrobeInteractions. (Lugtenberg,B.J.J., ed.), SpringerVerlag, Berlin, pp. 273-280.
Devereux, J., Haeberli, P., andSmithies, O. (1984)Acomprehensive set of sequenceanalysisprograms for
theVAX. NucleicAcidsRes. 12,387-395.
Gurr, S.J., Unkles, S.E. and Kinghorn, J.R. (1987) The structure and organization of nuclear genes of
filamentous fungi. In GeneStructureinEukaryotic Microbes. (KinghornJ.R., ed.), IRL Press, Oxford,
pp. 93-139.
Hodgson, C.P. and Fisk, R.Z. (1987) Hybridization probe size control: optimized 'oligolabeling'.Nucleic
AcidsRes. 15,6295.
Joosten, M.H.A.J. andDeWit, P.J.G.M. (1988)Isolation, purification andpreliminary characterization of
aproteinspecific forcompatibleCladosporiumfulvum(syn.Fulvia_/u/va)-tomatointeractions.Physiol. Mol.
PlantPathol. 33,241-253.
Joosten, M.H.A.J. and De Wit, P.J.G.M. (1989) Identification of several pathogenesis-related proteins in
tomatoleavesinoculatedwithCladosporiumfulvum(syn.Fulviafulvd) as 1,3-ß-glucanasesandchitinases.
PlantPhysiol.89, 945-951.
Joosten, M.H.A.J., Hendrickx, L.J.M. and De Wit, P.J.G.M. (1990) Carbohydrate composition of
apoplastic fluids isolated from tomato leaves inoculated with virulent or avirulent races of Cladosporium
fitlvum (syn. Fulviafulvd). Neth.J. PlantPathol.96, 103-112.
Lazarovits, G.andHiggins,V.J. (1976a)HistologicalcomparisonofCladosporiumfulvumrace 1 onimmune,
resistant, and susceptible tomato varieties. Can.J. Bot. 54, 224-234.
Lazarovits, G. and Higgins, V.J. (1976b) Ultrastructure of susceptible, resistant, and immune reactions of
tomato to races of Cladosporiumfulvum. Can.J. Bot. 54, 235-249.
Maniatis,T., Fritsch,E.F. andSambrook,J. (1982)MolecularCloning, aLaboratoryManual. ColdSpring
Harbor Laboratory Press, New York.
Sanger, F., Nicklen, S. and Coulson, A.R. (1977)DNAsequencing withchainterminationinhibitors. Proc.
Natl.Acad. Sri. USA74, 5463-5467.
Schägger, H andVonJagow, G. (1987)Tricine-sodiumdodecyl sulphate-polyacrylamide gel electrophoresis
for the separation of proteins in therange from 1to 100kDa. Anal. Biochem. 166, 368-379.
Unkles, S.E., Moon, R.P., Hawkins, A.R., Duncan, J.M. and Kinghorn, J.R. (1991) Actin in the
oomycetous fungus Phytophthora infestonsis theproduct of several genes. Gene100, 105-112.
VandenAckerveken, G.F.J.M., VanKan,J.A.L. andDeWit,P.J.G.M. (1992)Molecularanalysisofthe
avirulencegeneavr9ofthe fungal tomatopathogen Cladosporiumfulvumfully supportsthe gene-for-gene
hypothesis. PlantJ. 2, 359-366.
VandenAckerveken, G.F.J.M., Vossen, P. andDeWit,P.J.G.M. (1993)TheAVR9race-specific elicitor
of Cladosporiumfulvumisprocessed by endogenousandplantproteases. PlantPhysiol., Inpress.
Van Kan, J.A.L., Van den Ackerveken, G.F.J.M. and De Wit, P.J.G.M. (1991) Cloning and
characterization of cDNA of avirulence gene avr9 of the fungal pathogen Cladosporiumfulvum, causal
agent of tomato leaf mold. Mol. Plant-Microbe Interact.4, 52-59.
VanKan,J.A.L.,Joosten,M.H.A.J., Sibbel-Wagemakers, C.A.M., VandenBerg-Velthuis, G.C.M, and
DeWit,P.J.G.M. (1992)Differential accumulationofmRNAsencodingextracellularandintracellularPR
proteins in tomato inducedby virulent and avirulent races of Cladosporiumfulvum. PlantMol. Biol. 20,
513-527.
VonHeijne, G. (1986) A new method for predicting signal sequence cleavage sites. NucleicAcids Res. 14,
4683-4690.
113
Chapter 8
General Discussion
Inthelastdecadesignificantprogresshasbeenmadeinunravelinghost-pathogeninteractions
atthemolecularlevel. Avirulencegenes ofmanybacterial andafew fungalplantpathogens
havebeenclonedandcharacterized indetail.Thegene-for-genehypothesisasproposed more
than 50 years ago by Flor (1942)is still standing andis now supported by molecular data.
Aspectsofpathogenicity cannowbeanalysed byemployingmoleculargenetictechniquesto
provide conclusive evidence on the role of putative pathogenicity genes in host-pathogen
interactions.Inthischaptermajor advancesinthemolecularunderstanding ofavirulenceand
pathogenicity of bacterial and fungal plant pathogens, with emphasis on Cladosporium
fulvum, are discussed.
The biological data on the C. fulvum - tomato interaction presented in the preceding
chapters canbeintegrated in twoconceptual models.Pathogenicity factors of C.fulvum and
their possible role in establishing basic compatibility on tomato are described in the first
model. The second model describes race-specific incompatibility as a result of early
recognition of avirulent races of C. fulvum by resistant tomato cultivars, carrying the
complementary genes for resistance.
115
Basic compatibility
Theabilityof apathogen tosuccessfully infect andcolonizeaparticular hostplantiscalled
basiccompatibility(Ellingboe, 1976;Heath, 1981).Mostpathogensareabletocolonizeonly
a limited number of host plants. Successfull infection of a host requires a range of
pathogenicity factors, among which are attachment to the host plant (Hamer et al., 1988),
formation ofpenetration structures (Staplesetal., 1986),degradationofhostcuticleandcell
walls (Collmer, 1986; Dickman et al., 1989), toxin production (Panaccione et al., 1992),
phytoalexin detoxification (Van Etten et al., 1989) and suppression of defence responses
(Heath, 1982),etc.
Aspects of pathogenicity and basic compatibility have been studied in bacterial plant
pathogens by using non-pathogenic mutants. Genes involved in pathogenicity have been
cloned by complementing these mutants with a genomic library of wild-typepathogens. In
this way the so-called hrp genes have been cloned from several bacterial plant pathogens
(Willis et al., 1991). The hrpgenes are involved in pathogenicity on a susceptible host as
wellasintheinductionofthehypersensitiveresponseonresistanthosts.Recently, homology
betweenhrpgeneproducts ofbacterialplantpathogensandproteins ofmammalian bacterial
pathogens involved in excretion of pathogenicity factors has been observed (Gough et al.,
1992).
All pathogenicity genes from plant pathogenic fungi characterized so far have been
cloned by function, suchas cutinase (Dickman etal., 1989),pisatin demefhylase (Weltring
etal., 1988)and toxin synthetase (Panaccioneetal., 1992).Pathogenicity genes withasyet
unknown functions are difficult to isolate unless other approaches are addressed.
Inventarization and analysis of fungal genes expressed duringpathogenesis mightprovidea
broader view on the molecular basis ofpathogenicity. Onesuch approach is the differential
hybridization of agenomiclibrary of thefungus with cDNA synthesized onmRNA from in
vitrogrown and inplanta grown fungus. Nine inplanta induced genes of Phytophthora
infestonswere isolated in this way (Pieterse et al., 1993). In the C. fulvum - tomato
interaction, inplanta induced fungal genes have been characterized and cloned via their
proteinproductswhichspecifically accumulateininfected tomatoleavesbutnotinthe fungus
grown invitro(Chapter 7).
The knowledge on factors of C. fulvum involved in establishing and maintaining basic
compatibility on tomato is very limited. A number of putative pathogenicity factors of
C.fulvumisproposed anddepictedinthefirst modeldescribing theinfection ofasusceptible
tomatogenotypebyvirulent races of C.fulvum (Fig. 1).Theseraces arenot recognized by
resistant tomato genotypes as they either do not produce a race-specific elicitor or the
producedelicitorisnotrecognized. Besides,in susceptibletomatogenotypeselicitorscannot
berecognized duetotheabsenceof thecorresponding resistance gene(seenext section).C.
fulvum can grow abundantly in the leaves of susceptible plants while no defence responses
are induced. Possibly suppressors produced by C.fulvum abolish the induction of non-
116
specificplantdefenceresponses.Also,hostfactors couldsuppresstheactivityofnon-specific
glycoproteinelicitorsproducedbythefungus (PeeverandHiggins, 1989).Thereforeitmight
be concluded that C.fiilvumis adapted so well to its host plant tomato that none ofits
excreted compoundsinducenon-specific defence responseswhilethefungusgrowsinsidethe
leaf.
C.fiilvumhastoobtain allnutrients necessary for growth and reproduction from the
intercellularspacesbetweenthemesophyllcells.Possibly,pathogenicityfactorsinterferewith
metabolismand translocation ofnutrientsinthehostplant. Invertases, forinstance, excreted
by C.fiilvumin theapoplast convert sucrose into glucose andfructose andmight thereby
interfere with sugar transport inthehost (Joostenetal., 1990).Ithasbeen shown thatyeast
invertase targeted to the apoplast in tomato, tobacco andArabidopsis leaves significantly
reduces the source activity of photosynthesizing leaves (Von Schaewen et al., 1990;
Dickinson etal., 1991).Itwas suggested thatphloem loading from the apoplastisachieved
by sucrosecarriers andthat glucoseandfructose, the products of invertase activity, can no
longerbeloadedeffectively. Similarly, C.fiilvumcouldreducethephloemloadingintomato
by itsinvertase activity, preventing translocation ofcarbon sources from theinfected leaf.
Possibly, theconversion of monosaccharides to mannitol enhances this effect, as mannitol
cannot beutilized bythehost (Joosten etal., 1990).
EXTRACELLULAR
PROTEINS (ECPs)
PATHOGENICITY
FACTORS
Figure 1. Schematic representation of events involved in the establishment of basic compatibility between
virulent races ofC.fiilvumandsusceptibletomatocultivars(detailsdiscussed intext;adapted from
VandenAckerveken andDeWit, 1993).
117
InplantainducedgenesofC.fulvum, suchastheputativepathogenicitygenesecpl and ecp2
(chapter 7) and the avirulence genes avr9 (chapter 2) and avrA (Joosten et al., in
preparation), represent a groupof genes with no knownprimary function for thepathogen.
Their possible involvement in pathogenicity can now be addressed by gene disruption
(Chapter 4). Studies on the regulation of expression of these genes and the possible
involvementofplantfactorsintheinductiongeneexpression mightprovidemoreinformation
on the function of these genes for thepathogen. Future research will likely reveal thus far
unknownstrategies employedbyC.fulvumtosuccessfully penetrate, colonizeandreproduce
on its host plant.
Race-specific incompatibility
Recognition of pathogens by resistant host plants leads to the induction of active plant
defence and will result in incompatibility. The mechanism of pathogen recognition is still
unknown. Several models have been proposed in which the recognition is mediated via
resistance gene encoded receptors. In the last decade, avirulence genes and their direct or
indirect products, responsible for the recognition by the resistant plant, have been isolated
from different pathogens and characterized in detail (Keen, 1990).
Thefirst bacterial avirulencegenewasclonedin 1984by mobilizingDNAclones from
anavirulentracetoavirulentraceofPseudomonas syringaepv.glycinea (Staskawiczetal.,
1984). This was the first example of a single gene from thepathogen that determines the
outcomeofahost-pathogen combination. Ever sincemanyotherbacterialavirulencegenes
have been isolated and characterized. Recognition by the host can be overcome by the
pathogen dueto (i)lossof theavirulencegene(Staskawicz etal., 1984), (ii)inactivationby
transposon insertion (Kearney et al., 1988) and (iii) mutation within the avirulence gene
(Kobayashma/., 1990).
The primary products of bacterial avirulence genes do not induce a hypersensitive
responsein resistant hostplants,butareregulatory proteinsor areinvolved inthe formation
ofelicitors.ThereisoneexamplewheretheavirulencegeneavrD of Pseudomonas syringae
pv. tomato is involved in the formation of a non-proteinaceous race-specific elicitor (Keen
etal., 1990).Thedirectinteraction oftheprimary avirulencegeneproductwiththehosthas
beensuggested foravrBs3 ofXanthomonas campestrispv. vesicatoria. Thepredictedprotein
productof avrBs3 consists of 17.5nearlyidentical 34aminoacid repeat units (Bonas et al.,
1989). The number and type of repeats affect the avirulence spectrum, suggesting that the
primarygeneproductsfunction aselicitorsbydirectinteractionwithreceptorsintheresistant
hostplant (Herbers etal., 1992).
Theonlyexamplesofrace-specific elicitorswhichareprimaryavirulencegeneproducts
are the capsid protein of tobacco mosaic virus (Knorr and Dawson, 1988), the necrosisinducing peptides of Rhynchosporium secalis (Wevelsiep et al., 1992) and the peptide
118
elicitorsof Cladosporiumfulvwn (DeWitand Spikman, 1982;Scholtens-Tomaand DeWit,
1988; Joosten etal., in preparation).
Fungal avirulence genes have not been isolated by a shot-gun cloning approach as
reported for bacterial avirulence genes. Thelargegenome sizeof filamentous fungi andthe
low transformation frequency as compared to that of bacteria make this technique time
consuming.Thefirst fungal avirulencegenetobecloned, avr9,wasisolatedfrom C.fulvwn
byusing aminoacid sequencedataof therace-specific peptideelicitor (Chapter 2and 3).A
second avirulence genefrom C.fulvwn, avr4,wasclonedby a similar approach (Joosten et
al.,inpreparation).Recently, twoavirulencegenesofMagnaporthe grisea wereisolatedby
map-based cloning (Valent etal, 1993).
Thestudiesontheavirulencegeneavr9 described inthisthesisprovideapictureof the
events preceding the recognition of the AVR9 race-specific elicitor by the resistant host.
Followingpenetrationof theleaf through stomatatheexpression oftheavr9 geneisinduced.
In vitroexperiments indicate that avr9is induced under conditions of nitrogen limitation,
which might reflect growth conditions inplanta (Chapter 6). The AVR9precursor peptide
of 63 aa is excreted to the intercellular space and the putative signal peptide of 23 aa is
removed. The extracellular peptide of 40 aa is sequentially processed by fungal and plant
proteases to a stable 28 aa peptide, which can accumulate to high concentrations in the
apoplast of susceptible plants (Chapter 5). This mature peptide is an active elicitor of the
hypersensitive response inresistantplants, althoughwedonotexcludethattheintermediate
forms of 32, 33and34aaareactiveelicitorsaswell.Theprimary function of theavr9 gene
for the pathogen itself is still unknown. Wild-type races of C. fulvwn virulent on tomato
genotype Q9 havelostthecorresponding avirulencegeneavr9.Preliminary studies indicate
that disruption mutants lacking the avr9gene are stillpathogenic on tomatoindicating that
av/9 mightbedispensible(chapter4). Thecorresponding resistancegene Cf9, however, still
proves tobestablein thetomatogrowing areas, suggesting thattheraceslacking avr9 have
a reduced fitness.
Race-specific incompatibility is depicted in the second model (Fig. 2) describing the
interaction between a resistant tomato genotype and an avirulent race of C. fulvwn.
According to this model, expression of avirulence genes inplanta is induced by factors
and/or environmental stimuli from theplant. Thegeneproducts, therace-specific elicitors,
are recognized by putative receptors present in the plant plasma membrane. Upon
recognition, defence responses aretriggered andas aresult fungal growthisrestricted atan
early stage. The crucial factors in thecomplex defence response which are responsible for
theultimate restriction of fungal growth are not yet known. Most probably, a combination
of different defence responses, including the hypersensitive response, will finally result in
effective inhibition of fungal growth.
The fungal component involved in therecognition event inducing resistance, theracespecific elicitor, is now well documented and supports the gene-for-gene hypothesis.
Recognition of race-specific elicitors induces active plant defence and thereby blocks the
119
basic ability of C.Julvum to successfully colonize tomato. The underlying mechanisms
involvedinrecognitionandsubsequentsignaltransductionare,however, notyetunderstood.
Recently, evidence was obtained for the involvement of G-protein mediated signal
transduction following therecognition of arace-specific peptideelicitorof C.fiilvum(VeraEstrella et al., 1993). The results suggest that a membrane bound receptor is coupled toa
G-protein whichisinvolved intheactivation of phosphataseswhich inturn stimulateplasma
membrane H+-ATPases by dephosphorylation. Future research will tell us more about the
structure of these receptors, the putative products of resistance genes. Whether the genes
encoding receptorsareindeed Çfresistance genescanonlybeprovenbycomplementingÇflessplantswiththesegenesandscorefor resistanceofthesetransformed plantstoraceswith
thecomplementary avirulence gene.
RESISTANCE
GENE
RESTRICTIONOF
FUNGALGFiOWm
ATANEARLY •
STAGE
< PLANT
FACTORS
^
^
PATHOGENICITY
FACTORS
Figure 2. Schematic representation ofeventsinvolvedinrace-specific incompatibilitybetweenavirulentraces
of C. fiilvum and resistant tomato cultivars (details discussed in text; adapted from Van den
Ackerveken and De Wit, 1993).
References
Bonas, U., Stall, R.E. and Staskawicz, B. (1989) Genetic and structural characterization of the avirulence
gene awBs3 from Xanthomonas campestris pv. vesicatoria. Mol. Gen. Genet.218, 127-136.
Collmer, A. (1986)Themolecular biologyof pectic enzyme production and bacterial soft-rot pathogenesis.In
BiologyandMolecularBiologyofPlantPathogenInteractions, (Bailey,J.A.,ed.),Springer-Verlag, Berlin,
pp. 277-289.
120
DeWit,P.J.G.M. andSpikman, G. (1982)Evidencefor theoccurence ofrace-andcultivar-specific elicitors
of necrosis in intercellular fluids of compatible interactions between Cladosporiumfidvumand tomato.
Physiol.PlantPathol. 21, 1-11.
Dickinson, CD., Altabella, T. and Chrispeels, M.J. (1991) Slow-growth phenotype of transgenic tomato
expressing apoplastic invertase, PlantPhysiol.95,420-425.
Dickman, M.B., Podila, G.K. and Kolattukudy, P.E. (1989) Insertion of cutinase gene into a wound
pathogen enables it to infect intacthost. Nature342,446-448.
EUingboe,A.H. (1976)Geneticsofhost-parasiteinteractions.InEncyclopedia ofPlantPhysiology, NS,Vol.4,
PhysiologicalPlantPathology, (Heitefuss, R. andWilliams,P.H., eds.),Springer-Verlag,Heidelberg,pp.
761-778.
Flor, H.H. (1942)Inheritance of pathogenicity inMelampsora Uni. Phytopathol.32, 653-669.
Hamer, J.E.,Howard,R.J., Chumley,F.G. andValent, B. (1988)Amechanism for surface attachment in
spores of a plantpathogenic fungus. Science239, 288-290.
Gough, C.L., Genin, S., Zischek, C. andBoucher, C.Â. (1992)hrp genes of Pseudomonas solanacearum
arehomologoustopathogenicitydeterminantsofanimalpathogenicbacteria andareconserved amongplant
pathogenicbacteria. Mol. Plant-MicrobeInteract.5, 384-389.
Heath, M.C. (1981) A generalized concept of host-parasite specifictity. Phytopathology 71, 1121-1123.
Heath, M.C. (1982)The absence of active defence mechanisms in compatiblehost-pathogen interactions.La
ActiveDefenceMechanisms inPlants, (Woods,R.K.S., ed.),PlenumPress, New YorkandLondon, pp.
143-156.
Herbers, K., Conrads-Strauch,J. andBonas,U. (1992)Race-specificity ofplantresistance tobacterial spot
disease determined by repetitive motifs ina bacterial avirulence protein. Nature356, 172-174.
Joosten, M.H.A.J., Hendrickx, L.J.M. and De Wit, P.J.G.M. (1990) Carbohydrate composition of
apoplastic fluids isolated from tomato leaves inoculated with virulent or avirulent races of Cladosporium
fulvum (syn. Fulviafulva). Neth.J. PlantPathol. 96, 103-112.
Kearney B., Ronald, P.C.,Dahlbeck, D. andStaskawicz, B.J. (1988)Molecularbasis for evasionofplant
hostdefence inbacterial spotdisease of pepper. Nature332,541-543.
Keen, N.T., Tamaki, S., Kobayashi, D., Gerhold, D., Stayton, M., Shen, H., Gold, S., Lorang, J.,
Thordal-Christensen, H., Dahlbeck, D. andStaskawicz, B. (1990)Bacteria expressing avirulencegene
Dproduce a specific elicitor ofthesoybeanhypersensitive reaction.Mol.Plant-MicrobeInteract.3, 112121.
Keen, N.T. (1990)Gene-for-gene complementarity inplant-pathogeninteractions.Annu.Rev. Genet.24,447463.
Knorr, D.A. andDawson, W.O. (1988) A point mutation in the tobacco mosaicvirus capsid protein gene
induceshypersensitivity inNicotiana sylvestris.Proc. Natl.Acad.Sei. USA86, 170-174.
Kobayashi, D.Y., Tamaki, S.J., Trollinger, D.J., Gold, S. and Keen, N.T. (1990) A gene from
Pseudomonassyringaepv.glycinea withhomologytoavirulencegeneDfrom P. s. pv.tomatobutdevoid
of theavirulence phenotype. Mol. Plant-Microbe Interact.3, 103-111
Panaccione,D.G., Scott-Craig,J.S.,Pocard,J-E.andWalton,J.D.(1992)Acyclicpeptidesynthetasegene
required forpathogenicityofthefungus Cochlioboluscarbonum onmaize.Proc.Natl.Acad.Sei. USA 89,
6590-6594.
Peever, T.L. andHiggins, V.J. (1989)Suppressionoftheactivityofnon-specific elicitorfrom Cladosporium
fulvum by intercellular fluids from tomato leaves. Physiol.Mol. PlantPathol. 34, 471-482.
Pieterse, C.M.J., Riach, M.B.R., Bleker, T., Van den Berg-Velthuis, G.C.M. and Govers, F. (1993)
Isolation of putative pathogenicity genes of the potato late blight fungus Phytophthora infestons by
differential hybridization of a genomic library. Physiol.Mol. PlantPathol. Inpress.
Scheel, D. and Parker, J.E. (1990) Elicitor recognition and signal transduction in plant defense gene
activation. Z Naturforsch.45c, 569-575.
Scholtens-Toma, I.M.J, and De Wit, P.J.G.M. (1988) Purification and primary structure of a necrosisinducing peptide from the apoplastic fluids of tomato infected with Cladosporiumfulvum (syn. Fulvia
Fulva). Physiol.Mol.PlantPathol 33, 59-67.
Staples, R.C., Yoder, O.C., Hach, H.C., Epstein, L. and Bhairi, S. (1986) Gene expression during
infection structuredevelopmentbygermlingsoftherust fungi. InBiologyandMolecularBiologyofPlant
Pathogen Interactions,(Bailey, J.A., ed.), Springer-Verlag, Berlin, pp. 331-341.
121
Staskawicz,B.J., Dahlbeck,D.andKeen, N.T. (1984)ClonedavirulencegeneofPseudomonassyringaepv.
glycineadetermines race-specific incompatibilityon Glycine max(L.) Merr. Proc. Natl. Acad.SeiUSA,
81, 6024-6028.
Van den Ackerveken, G.F.J.M. and De Wit, P.J.G.M. (1993) The Cladosporium fiilvum - tomato
interaction: a model system for fungus - plant specificity. In Pathogenesis andHost SpecificityinPlant
Diseases, (Kohmoto, K., Singh, R.P., Singh, U.S. and Zeigler, R., eds.), Pergamon Press, Oxford, In
press.
Valent, B., Sweigard, J., Orbach, M., Farrall,M., Walter, A., Kang, S. andChumley, F. (1993)Book
of abstracts of theFungal Genetics Conference, Assilomar, CA, pp. 18.
Van Etten, H.D., Matthews, D.E. and Matthews, P.S. (1989) Phytoalexin detoxification: important for
pathogenicity andpractical implications.Annu. Rev. Phytopathol.27, 143-164.
Vera-Estrella,R., Barkla,B.J., Higgins,V.J. andBlumwald,E. (1993)Signaltransductionviaa G-protein
mediates a plantdefense response to fungal pathogens. Inpress.
VonSchaewen, A., Stitt,M., Schmidt,R., Sonnewald,U. andWillmitzer,L. (1990)Expressionofayeastderived invertase inthe cellwallof tobacco andArabidopsis plantsleads toaccumulation of carbohydrate
andinhibitionofphotosynthesisandstronglyinfluences growthandphenotypeoftransgenic tobaccoplants,
EMBOJ. 9, 3033-3044.
Weltring, K-M., Turgeon, B.G., Yoder, O.C. and Van Etten, H.D. (1988) Isolation of a phytoalexindetoxification gene from theplantpathogenic fungus Nectriahaematococca by detecting itsexpression in
Aspergillusnidulans. Gene68, 335-344.
Wevelsiep, L., Kogel, K.H. and Knogge, W. (1992) Purification and characterization of peptides from
Rhynchosporium secalisinducing necrosis in barley. Physiol.Mol. PlantPathol. 39, 471-482.
Willis,D.K.,Rich,J.J.andHrabak,E.M. (1991)hrpgenesofphytopathogenicbacteria.Mol. Plant-Microbe
Interact.4, 132-138.
122
Summary
The molecular understanding of host-pathogen interactions and particularly of specificity
forms the basis for studying plant resistance. Understanding why a certain plant species or
cultivaris susceptibleand whyother species orcultivars areresistant is ofgreat importance
in order to design new strategies for future cropprotection by molecular plant breeding.
In this thesis molecular aspects of avirulence and pathogenicity of the tomatopathogen
Cladosporiumfulvumaredescribed.TheinteractionC.fulvum -tomatoisanexcellentmodel
system to study fungus -plant specificity asthecommunication betweenpathogen andplant
isconfined totheapoplast (intercellular space).Theability toobtainintercellular fluid from
C.,/w/vw/n-infected tomato leaves enabled the isolation and characterization of plant and
fungal compounds that mightplayanimportant roleinpathogenesis and/or theinductionof
resistance. Thepurification and characterization of a race-specific peptide elicitor provides
thebasisfor mostoftheexperiments described inthisthesis.Thispeptidewasthoughttobe
produced only by races of C. fulvum avirulent on tomatogenotypes carrying theresistance
gene Cf9,on which theelicitor induced necrosis.
Molecular aspects of avirulence of C.fulvum were first studied by the isolation and
characterization ofthecDNAencoding theAVR9race-specific peptideelicitor (Chapter2).
The peptide was shown to be indeed produced by C.fulvum. Races virulent on tomato
genotype Q9 lack the avi9 gene and do not produce the peptide elicitor thereby evading
recognition by tomatogenotypes carrying thecorresponding resistance gene Q9. Toprove
thattheav/9 geneisagenuineavirulencegene, racesvirulent ontomatogenotype Cf9were
transformed with the cloned av/9 gene (Chapter 3). The cultivar-specificity of the
transformants waschangedfrom virulenttoavirulentontomatogenotype Cf9. Theavr9 gene
cantherefore beconsidered tobeagenuineavirulencegene, thefirst fungal avirulencegene
cloned. Additionalproof for theroleof theavirulencegeneavr9 in specificity wasprovided
by the disruption of avt9 in two races avirulent on tomato genotype Cf9, by gene
replacement, resulting in transformants virulent on tomatogenotype Cf9(Chapter 4).
Theav/9 geneencodesa 63aminoacidsprecursor protein. Removal ofa signalpeptide
resultsin anextracellular peptideof 40amino acids.Proteases of C.fulvum areinvolved in
further processing thisextracellular peptideby removal ofN-terminalamino acidsresulting
inpeptidesof32,33and34aminoacids.Plantfactors areresponsiblefor further processing,
resulting in the stablepeptideelicitor of 28aminoacids (Chapter 5).
The avirulence gene avr9 is highly expressed in C.fulvum while growing inside the
tomatoleaf.Theexpression ofav/9isinduced inC.fulvum grown invitro under conditions
of nitrogen limitation. The high expression of av/9 in C.fulvum growing inside thetomato
leaf mightbecaused by nitrogen limiting conditions in theapoplast (Chapter 6).
125
Pathogenicity of C.fulvum was studied at the molecular level by theisolation of twogenes
encodingextracellularproteins(ECPs).Theecpl andecp2geneswereisolatedviatheamino
acidsequenceofECP1,andpolyclonalantibodiesraisedagainstECP2,respectively (Chapter
7). Theexpression of the ecpgenes ishighly induced inplantaas compared to theinvitro
situation. The availability of the cloned ecp genes now enables us to study the role and
importanceofthesegenesduringpathogenesisbyreportergeneanalysisandgenedisruption.
TwomodelsdescribingtheC.fidvum-tom&iointeractionarepresented, dealingwithbasic
compatibility and race-specific incompatibility, respectively (Chapter 8). The improved
understandingofpathogenrecognitioncanbeexploitedinfuture research toelucidatetherole
of putative receptors in theresistant plant involved in perception of elicitors and induction
of activeplant defence.
126
Samenvatting
De moleculaire kennis van plant-pathogeen interacties en in het bijzonder van specificiteit
vormtdebasisvoorhetonderzoek naar resistentievanplanten. Dewetenschap waaromeen
bepaalde plantesoort of cultivar vatbaar is en waarom een andere soort of cultivar resistent
is, is van groot belang voor het ontwikkelen van nieuwe strategieën voor de
gewasbescherming door middelvan moleculaire plantenveredeling.
In dit proefschrift worden de moleculaire aspecten van avirulentie en pathogeniteit van
het tomatepathogeen Cladosporiumfiilvwnbeschreven. De interactie C.fidvwn- tomaat is
een uitstekend modelsysteem voor het onderzoek naar schimmel -plant specificiteit, omdat
decommunicatietussenpathogeenenplantbeperktistotdeapoplast (intercellulaireruimte).
De techniek om van C. ./w/vw/n-geïnfecteerde tomatebladeren de intercellulaire vloeistof te
isoleren, maakte het mogelijk om schimmel- en plantfactoren te isoleren en karakteriseren
die mogelijk een belangrijke rol spelen tijdens de Pathogenese en/of de inductie van
resistentie. Dezuivering enkarakterisering van eenfysio-specifieke peptide elicitorvormen
debasisvoor de meestevan de inditproefschrift beschreven experimenten. Dit eiwit werd
verondersteld alleen geproduceerd te worden door fysio's van C.fiilvumdie avirulent zijn
op tomaat-genotypen met het resistentiegen Cf9,waarop deelicitor necrose induceert.
De moleculaire aspecten van avirulentie van C. fiilvum werden onderzocht door de
opsporingenkarakterisering vanhetcDNAcoderend voordefysio-specifieke elicitorAVR9
(Hoofdstuk 2).Hetpeptidebleekinderdaad doorC.fiilvumgeproduceerd teworden. Fysio's
dievirulent zijn op tomaat-genotype Cf9missen hetavirulentiegen avr9enproduceren geen
fysio-specifieke elicitor, waardoor ze de herkenning door tomaat-genotypen met het
resistentiegen Cf9omzeilen. Om te bewijzen dat het avr9gen een echt avirulentiegen is,
werden fysio's virulent op tomaat-genotype Cf9 getransformeerd met het gekloneerdeavr9
gen (Hoofdstuk 3).Decultivar-specificiteit van detransformanten werd hierdoor veranderd
van virulent naar avirulent op tomaat-genotype Cf9. Het avr9 gen blijkt dus een echt
avirulentiegen te zijn, het eerste gekloneerde avirulentiegen van een schimmel. Aanvullend
bewijs voor de rol van het avirulentiegen avi9 in schimmel - plant specificiteit werd
verkregen door de disruptie van het avr9gen in twee fysio's avirulent op tomaat-genotype
CJ9, door middel van vervanging van het avi9 gen (gene replacement). Dit resulteerde in
transformanten dievirulent zijn op tomaat-genotype Cf9(Hoofdstuk 4).
Hetavr9 gencodeertvooreenprecursor eiwitvan63aminozuren. Verwijdering vanhet
signaalpeptide resulteert in een extracellulair peptide van 40 aminozuren. Proteases van
C.fiilvumzijn betrokken bij deprocessing van dit extracellulaire peptidedoor verwijdering
van de N-terminale aminozuren resulterend in peptiden van respectievelijk 32, 33 en 34
aminozuren. Factoren van de plant zijn verantwoordelijk voor de verdere processing,
resulterend in een stabielepeptide elicitor van 28 aminozuren (Hoofdstuk 5).
127
Het avirulentiegen avi9 komt sterk tot expressiein C.fulvwn tijdens dekolonisatievanhet
tomateblad.Deexpressievanhetavr9 genwordtgeïnduceerd ininvitrogekweektC. fidvum
mycelium onder stikstoflimiterende condities. De hoge expressie van avr9 in C. fidvum
tijdensdekolonisatievanhettomatebladwordtmogelijkveroorzaakt door stikstoflimiterende
condities in deapoplast (Hoofdstuk 6).
PathogeniteitvanC.fidvumisopmoleculairniveaubestudeerd doordeisolatievantwee
genen die coderen voor extracellulaire eiwitten (extracellular ßroteins, ECPs). De ecpl en
ecpl genen zijn achtereenvolgens geïsoleerd via de aminozuurvolgorde van ECP1, en
polyklonaleantilichamen tegenECP2 (Hoofdstuk 7).DeexpressievandezeecpgeneninC.
fidvum is sterk geïnduceerd in planta in vergelijking met de situatie in vitro. De
beschikbaarheid van de gekloneerde ecp genen maakt het nu mogelijk om de rol en het
belang van deze genen tijdens de Pathogenese te bestuderen met behulp van reportergen
analyse en gendisruptie.
Twee modellen voor de C.fidvum-tomaat interactieworden gepresenteerd, betrekking
hebbend op respectievelijk basis compatibiliteit en fysio-specifieke incompatibiliteit
(Hoofdstuk 8). Deverkregen kennis metbetrekking tot herkenning vanpathogenen door de
plant kan aangewend worden ominzicht teverkrijgen in derolvan mogelijke receptoren in
de resistente plant die betrokken zijn bij de herkenning van elicitoren en de inductie van
actieveafweer van deplant.
128
Account
Mostof theresults presented in this thesis havebeen published, or will bepublished in the
near future. The contents of the preceding chapters have been based on most of these
publications.
De Wit, P.J.G.M., Van den Ackerveken, G.F.J.M., Joosten, M.H.A.J. andVanKan,
J.A.L. (1989)Apoplasticproteinsinvolvedincommunicationbetweentomatoandthe fungal
pathogen Cladosporium fiilvum. In Signal Molecules in Plants andPlant-Microbe
Interactions, (Lugtenberg, B.J.J., ed.), Springer-Verlag, Berlin, pp. 273-280.
De Wit, P.J.G.M., Van Kan, J.A.L., Van den Ackerveken, G.FJ.M. and Joosten,
M.H.A.J. (1991) Specificity of plant-fungus interactions: molecular aspects of avirulence
genes.InAdvancesinMolecularGeneticsofPlant-MicrobeInteractions, (Hennecke,H.and
Verma, D.P.S., eds.), Kluwer Academic Publishers, Dordrecht, pp. 233-241.
VanKan,J.A.L., VandenAckerveken,G.FJ.M. andDeWit,P.J.G.M. (1991)Cloning
and characterization of cDNA of avirulence gene avr9 of the fungal tomato pathogen
Cladosporiumfulvum, causal agentof tomatoleaf mold.Mol. Plant-Microbe Interact. 4,5259.
Van den Ackerveken, G.FJ.M., Van Kan, J.A.L. and De Wit, PJ.G.M. (1992)
Molecular analysis of theavirulencegeneavr9 of thefungal tomatopathogen Cladosporium
fulvum supports the gene-for-gene hypothesis. PlantJ. 2, 359-366.
Van Kan, J.A.L., Van den Ackerveken, G.FJ.M. and De Wit, PJ.G.M. (1992)
Molecular aspectsofvirulenceandavirulenceofthetomatopathogen Cladosporiumfulvum.
In MolecularBiologyof Filamentous Fungi, (Stahl, U. and Tudzynski, P., eds.), VCH,
Weinheim, pp. 87-97.
De Wit, PJ.G.M., Van den Ackerveken, G.FJ.M., Vossen, J.P.M.J., Joosten,
M.H.AJ., Cozynsen, T.J., Honée,G., Wubben,J.P., Danhash,N., VanKan,J.A.L.,
Marmeisse, R. and Van den Broek, H.WJ. (1992) Molecular cloning and functions of
avirulenceandpathogenicitygenesofthetomatopathogen Cladosporiumfulvum. InAdvances
in MolecularGenetics of Plant-Microbe Interactions, Vol.2., (Nester, E.W. and Verma,
D.P.S., eds.), Kluwer Academic Publishers, Dordrecht, pp.289-298.
131
Van denAckerveken, G.F.J.M., VanKan,J.A.L.,Joosten,M.H.A.J., Muisers, J.M.,
Verbakel, H.M. and De Wit, P.J.G.M. (1993) Characterization of two putative
pathogenicitygenesofthefungaltomatopathogenCladosporiumfiilvum.Mol. Plant-Microbe
Interact.6, 210-215.
De Wit, P.J.G.M., Van den Ackerveken, G.FJ.M., Vossen, J.P.M.J., Joosten,
M.H.A.J., Cozynsen, T.J., Honée, G., Wubben,J.P.,Danhash,N., VanKan,J.A.L.,
Marmeisse, R. and Van den Broek, H.WJ . (1993) Avirulence genes of the tomato
pathogen Cladosporiumfiilvumand their exploitation in molecular breeding for diseases
resistant plants. In Developments in Plant Pathology; Mechanisms of PlantDefense
Responses, (Fritig,B.andLegrand,M., eds.),KluwerAcademicPublishers,Dordrecht, pp.
24-32.
Marmeisse, R., Van den Ackerveken, G.FJ.M., Goosen, T., De Wit, PJ.G.M. and
Van denBroek, H.WJ . (1993) Disruption of theavirulence geneavr9in tworacesof the
tomato pathogen Cladosporiumfiilvumcauses virulence on tomato genotypes with the
complementary resistance gene Çf9.Mol.Plant-Microbe Interact. In press.
Van den Ackerveken, G.FJ.M., Vossen, P. and De Wit, PJ.G.M. (1993) The AVR9
race-specificelicitorof Cladosporiumfiilvumisprocessedbyendogenousandplantproteases.
PlantPhysiol.Inpress.
Van den Ackerveken, G.FJ.M. and De Wit, PJ.G.M. 1993/94. The Cladosporium
fiilvum - tomato interaction: a model system for fungus - plant specificity. In Pathogenesis
andHostSpecificityInPlantDiseases, (Kohmoto,K., Singh,R.P., Singh,U.S.and Zeigler,
R., eds.), Pergamon Press, Oxford, In press.
Van den Ackerveken, G.FJ.M., Dunn, R.M., Cozù'nsen, T.J., Vossen P., Van den
BroekH.WJ . andDeWit, PJ.G.M. (1993)Nitrogen limitation inducesexpression ofthe
avirulence gene avr9in the tomato pathogen Cladosporiumfulvum; a reflection of growth
conditions inplantai submitted for publication toMol.Gen. Genet.
Yil
Curriculum vitae
Augustinus FranciscusJohannesMariavan denAckerveken werd op21april 1967teBreda
geboren.In 1984behaaldehijhetdiplomaHAVOaandescholengemeenschap "Markenhage"
teBreda.Inhetzelfdejaarbegonhij aandeopleidingtotbotanisch laboratorium medewerker
aan de toenmalige Rijks Hogere Agrarische School (de huidige Internationale Agrarische
Hogeschool "Larenstein", afdeling laboratoriumonderwijs) teWageningen. Na een stageop
de toenmalige Stichting voor Plantenveredeling bij de afdeling Grassen en een
afstudeeropdracht op de vakgroep Fytopathologie van de Landbouwuniversiteit te
Wageningenwerdhetdiplomavoorbotanischlaboratoriummedewerkerin 1988behaald.Het
afstudeerproject bij dr.PJ.G.M. deWitwerdvervolgens 6maandenvoortgezet middelseen
aanstelling als analist. Hierna heeft hij 6 maanden ervaring opgedaan in het moleculair
genetisch onderzoek aan schimmels in de groep van dr. H.W.J. van den Broek op de
vakgroep Erfelijkheidsleer van de Landbouwuniversiteit te Wageningen. In augustus 1989
werd hij aangesteld als onderzoeker in opleiding (BION, NWO) bij de vakgroep
Fytopathologie van de Landbouwuniversiteit. In de onderzoeksgroep van dr. PJ.G.M. de
Witwashijtotfebruari 1993werkzaamaanhetprojectgetiteld: "Opsporingenkloneringvan
genen die coderen voor fysio-specifieke elicitoren in de waardplant-schimmel interactie
Cladosporiumfiilvwn-tomazt".De resultaten van het gedurende deze periode uitgevoerde
onderzoek zijn beschreven in dit proefschrift. Sinds februari 1993 is hij werkzaam als
toegevoegd onderzoeker bij devakgroep Erfelijkheidsleer van de Landbouwuniversiteit.
135