ANRV384-PY47-18 ARI 2 July 2009 19:16

Ustilago maydis as a PathogenThomas Brefort, Gunther Doehlemann,Artemio Mendoza-Mendoza, Stefanie Reissmann,Armin Djamei, and Regine KahmannMax Planck Institute for Terrestrial Microbiology, Department of Organismic Interactions,D-35043 Marburg, Germany; email: kahmann@mpi-marburg.mpg.de

Annu. Rev. Phytopathol. 2009. 47:423–45

The Annual Review of Phytopathology is online atphyto.annualreviews.org

This article’s doi:10.1146/annurev-phyto-080508-081923

Copyright c© 2009 by Annual Reviews.All rights reserved

0066-4286/09/0908/0423$20.00

Key Words

corn smut, dimorphism, signal transduction, host response, fungaleffectors

AbstractThe Ustilago maydis–maize pathosystem has emerged as the currentmodel for plant pathogenic basidiomycetes and as one of the few modelsfor a true biotrophic interaction that persists throughout fungal devel-opment inside the host plant. This is based on the highly advanced ge-netic system for both the pathogen and its host, the ability to propagateU. maydis in axenic culture, and its unique capacity to induce prominentdisease symptoms (tumors) on all aerial parts of maize within less thana week. The corn smut pathogen, though economically not threaten-ing, will continue to serve as a model for related obligate biotrophicfungi such as the rusts, but also for closely related smut species thatinduce symptoms only in the flower organs of their hosts. In this re-view we describe the most prominent features of the U. maydis–maizepathosystem as well as genes and pathways most relevant to disease.We highlight recent developments that place this system at the fore-front of understanding the function of secreted effectors in eukaryoticpathogens and describe the expected spin-offs for closely related speciesexploiting comparative genomics approaches.

423

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

INTRODUCTION

Hemibasidiomycete smut fungi comprise morethan 1500 species, and among them are plantpathogens of considerable economic impor-tance. Most smuts have a narrow host rangeand are specialized on members of the Poaceae(grasses) that include valuable crop species suchas maize, sorghum, sugar cane, wheat, and bar-ley. A recent investigation involving five smutspecies indicates that contemporary smut taxadiverged millions of years ago, i.e., before thetime of domestication and modern agriculture,presumably through coevolution within natu-ral populations of the ancestors of our currentcrop plants (86).

In all smut fungi, pathogenic developmentis coupled to sexual development. Haploid cellsof opposite mating type have to fuse and gen-erate a dikaryotic form that is then able toinvade the host plant either through naturalopenings or directly. Most smut fungi infectplants during the seedling stage and then con-tinue growth within meristematic tissue duringplant development. Usually, this phase is sys-temic, and symptom development is restrictedto the female and/or male inflorescences ofthe host. In the flowers, massive fungal pro-liferation and spore formation occurs. As aconsequence, seeds are replaced by smut soriconsisting of masses of black teliospores (74).Teliospores represent the resting stage that cansurvive harsh environments for long periods.Under appropriate conditions teliospores ger-minate, undergo meiosis, and produce haploidprogeny. Ustilago maydis, in contrast to the othersmut fungi, produces prominent symptoms onall aerial parts of its host plant, maize. In prac-tice, maize seedlings can be infected at thethree-leaf stage, and about one week later thesymptoms, in this case tumor formation, can bescored.

U. maydis represents a fully developedgenetic system that is amenable to highlyefficient reverse genetics and cell biologi-cal approaches (11, 58, 98). The first stepstoward understanding the disease were under-taken by the characterization of the mating

type loci (1), which enabled the design ofsolopathogenic haploid strains that causedisease without the need for a mating partner.Such strains allowed unbiased approachesto analyze pathogenicity (9, 92). In 2006 anew era began with the publication of the20.5 Mb U. maydis genome sequence (59).The genome is highly compact and codes forapproximately 6900 protein encoding genes.The majority of these genes lack introns, andthe genome is largely devoid of repetitiveDNA (59). More than 99% of the sequence isrepresented in 24 large scaffolds that corre-spond to the 23 chromosomes. Access to themanually annotated genome sequence (http://www.broad.mit.edu/annotation/genome/ustilago maydis/Home.html and http://mips.gsf.de/genre/proj/ustilago/) alloweddetection of a large set of novel genes that havecrucial roles during pathogenesis. Many ofthese novel genes code for secreted proteins,so-called effectors (59). Effectors have so farmostly been studied in prokaryotic pathogens,where they have been shown to be translocatedto plant cells and interfere with a varietyof defense responses (43). Discovery of alarge repertoire of effectors in oomycete plantpathogens has led to the definition of two effec-tor classes: those that function in the apoplastand those that are taken up by plant cells via aconserved motif (RXLR) (57, 103). Effectorssecreted by fungi including U. maydis lacksuch a common motif (34). In the U. maydisgenome, effector genes were detected byserendipity through their organization inclusters that are upregulated during plantcolonization (59).

In this review, we will focus on aspects thatrelate to pathogenic development of U. maydis.This includes morphological transitions, sig-naling networks, nutrition in the host environ-ment, plant defense responses, and secreted ef-fectors. A section on comparative genomics andwhat we can expect from such approaches is in-cluded, as well as an outlook on questions thatmight be addressed in the near future. For allother aspects of U. maydis biology the reader isreferred to a number of recent comprehensive

424 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

reviews (1, 56, 64, 87, 97, 98) and to a specialissue of Fungal Genetics and Biology (46) thathighlights several distinct features of thegenome.

THE DISEASE CYCLE

The haploid yeast-like form of U. maydis can bepropagated on artificial media. However, thisform is unable to cause disease. The infectiousform is generated when two compatible haploidstrains fuse and generate a dikaryotic filament.This process is controlled by a tetrapolar mat-ing system that is specified by the biallelic alocus and the multiallelic b locus (1, 56). Strainsthat differ in a and b can fuse and form dikary-otic hyphae. On agar plates containing activatedcharcoal, such dikaryotic filaments extend intothe air and profusely cover the colonies, result-ing in a white, fuzzy appearance (2). The a lo-cus encodes a pheromone and receptor systemthat allows haploid cells of the opposite a mat-ing type to sense each other and to fuse. Thefate of the resulting dikaryon depends on the blocus, which codes for a pair of homeodomaintranscription factors, termed bE and bW. bEand bW proteins dimerize when derived fromdifferent alleles, and the heterodimeric bE/bWcomplex triggers filamentation as well as sex-ual and pathogenic development (1). The fil-amentous dikaryon of U. maydis establishes abiotrophic interaction that persists throughoutthe disease cycle (Figure 1).

The dikaryotic hyphae show tip-directedgrowth, and the cytoplasm accumulates in thetip cell compartment, whereas older parts of thehyphae become vacuolated and are sealed offby regularly spaced septa. Often, these emptysections collapse. On the plant surface, hy-phae stop polar growth in response to an asyet unidentified signal, and their tips swell toform poorly differentiated, nonmelanized ap-pressoria (Figure 1a). These develop infectionhyphae (Figure 1b) that directly penetrate intothe plant tissue, presumably aided by lytic en-zymes. The plasma membrane of the plant cellinvaginates and tightly surrounds the intracel-lular hyphae. This creates a special biotrophic

interface, an interaction zone in which fungalsecretion leads to an accumulation of depositsthat make up a vesicular matrix along the con-tact area (7; Figure 1c).

U. maydis does not form specialized feed-ing structures (haustoria) suggesting that sig-nal exchange and nutrient uptake have to oc-cur via the biotrophic interface. During earlyin planta development, the fungus grows in-tracellular, and cell-to-cell passages into deeperlayers of the infected tissue are observed. Theformation of empty sections ceases, and mi-totic divisions take place, concomitant withthe development of clamp-like structures thatallow the correct sorting of nuclei (91) andbranching. After these early stages the fungusis found preferentially in and near the vascula-ture (Figure 1d ). Five to six days after infectionmassive fungal proliferation, presumably of thediploid form, is observed in tumor tissue, andthis occurs intracellularly and extracellularly inapoplastic cavities where large fungal aggre-gates form (27; Figure 1e). In these aggregateshyphae are embedded in a mucilaginous ma-trix that could have a protective role. Prolifer-ation is followed by sporogenesis where hyphalsections fragment, round up, and differentiateinto heavily melanized diploid teliospores (3,94, 95). The spores are released when the tu-mors dry up and rupture. Under favorable con-ditions these spores germinate, the diploid nu-cleus undergoes meiosis, and budding off froma promycelium, haploid cells are again pro-duced. So far it has not been possible to recon-stitute this cycle in vitro, indicating that manyof its steps require stage-specific signals fromthe host.

Mating and Pathogenic DevelopmentRequire cAMP and MAPK Signaling

During mating the lipopeptide pheromone isperceived, and the signal is transmitted bytwo conserved signaling cascades: a cAMP-dependent protein kinase A (PKA) pathway anda mitogen-activated protein kinase (MAPK)module that have been comprehensively re-viewed recently (40, 64, 87). Components of

www.annualreviews.org • Ustilago maydis as a Pathogen 425

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

these pathways with crucial roles during matingand their epistatic relationships are depicted inFigure 2a. After pheromone-induced activa-tion both signaling pathways converge on thekey transcription factor Prf1, which in turn in-duces transcription of a large set of genes in-cluding the a and b mating-type genes (112) andis thus essential for mating. To induce expres-sion of the a genes, mfa and pra, Prf1 needs to bephosphorylated by the PKA Adr1, whereas in-duction of the b genes requires Prf1 phosphory-lation by Adr1 and the MAPK Kpp2 (55, 85). In

Plant responses U. maydis effectors

PR genes

JA signaling

Protease inhibitors

Receptor kinases

PR genes*

JA signaling

No SA signaling

Cell death suppression

PR genes*

Cell death suppression

Antioxidants

Photosynthesis

Antioxidants

Photosynthesis

Shikimate pathway

Hexoses

Auxin and GA

Redirection of carbon flow

Photosynthesis

Shikimate pathway

Auxin and GA

Cell division and enlargement

Rep1

Pep1

Pep1

Cluster 5A

Mig2_5-6

Rsp1/Hum3

Mig1_1

Mig2_1–6

Mig1_1

Mig2_1–6

Cluster 19A

Cluster 19A

Cluster 2A

b

c

d

e

12 h

24 h

48 h

4 d

8 d

a

5 μm

5 μm

5 μm

25 μm

50 μm

addition, prf1 transcription is controlled by thetwo MAPKs Kpp2 and Crk1 (41, 85) through acomplex interplay of at least four downstreamtranscription factors that bind to discrete ele-ments in the prf1 promoter: Prf1 itself, Rop1,Hap2, and an as yet unidentified factor down-stream of Crk1 acting via an upstream activationsequence (15, 41, 48, 78, 105). Except for therecently identified dual-specificity phosphataseRok1 that downregulates mating (23), all

←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−Figure 1Stage specificity of plant responses and activity ofsecreted U. maydis effectors. Prominent plantresponses to U. maydis infection (left), micrographsof typical infection structures (middle), and secretedeffectors that are required at distinct stages ofinfection between 12 hpi and 8 dpi are shown.Green arrows: Induced plant genes or processes.Red arrows: Repressed plant genes or processes.Asterisks: Expression reduced compared to the 12 htime point. U. maydis effectors that are required forvirulence are shown in red letters; effectors thatrestrict virulence are shown in green. Effectorswithout known function for pathogenicity are shownin black. For microscopy, maize plants were infectedwith solopathogenic haploid strains of U. maydis thatexpressed fluorescently tagged fusion proteins in (c).(a) 12-h postinfection, filaments of a solopathogenicU. maydis strain are forming appressoria on themaize epidermis. Blue: Fungal hypha stained withcalcofluor white. (b) 24-h postinfection, confocalprojection of an appressorium of the samesolopathogenic strain as in (a); the penetratinghypha has entered the maize epidermis cell (arrow).Blue: Fungal hypha stained with calcofluor white.(c) Confocal projection of a hypha that is growingintracellularly in a maize epidermis cell, 48-hpostinfection. Blue: Plant-cell-wallautofluorescence. Red: Cytoplasmic red fluorescentprotein (RFP) expressed by U. maydis. Green:Mig2-6:GFP fusion protein, which is secreted fromthe hypha to the biotrophic interface. (d ) Confocalprojection of U. maydis hyphae colonizing maizetissue in and around a vascular bundle, 4-dayspostinfection. Red: Plant-cell-wall structures stainedwith propidium iodide. Green: Hyphae of U. maydisstained with WGA-AF488. (e) Confocal projectionof U. maydis hyphae in maize leaf tumor tissue,8-days postinfection. Fungal hyphae form largeaggregates between the enlarged maize tumor cells.Green: U. maydis hyphae stained with WGA-AF488.Blue: Plant-cell-wall autofluorescence.

426 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

a Mating

a1b1

?

Ras2 Ras2

P

PP

P

P

PP

PP

Uac1

Ubc1

Ubc2

Kpp4/Ubc4

Fuz7/Ubc5

Kpp2/Ubc3Crk1Ubc1

? Rop1

Rok1

Hap2

cAMP

cAMPATP

Adr1

Adr1

Prf1

a mating type genes:mfa1, pra1

b mating type genes:bE1, bW1

prf1

Gpa3

a1b1

U

a3

Pra1

Gγ

Bp

p1

a2b2

b Plant invasion

?

? ?Ras1 Ras1

Sql2

P

P

P

P

P

P Kp

p2

P

bE1 bW2

P

PP

P

P

P

Uac1

Ubc1

Ubc2

Kpp4

Fuz7

Crk1Ubc1

?Hgl1

Rok1

Hap2

cAMP

cAMPATP

Cutin monomers/

Adr1

Adr1

Prf1

a mating type genes:mfa1, pra1

b mating type genes

prf1

Gpa3

? ?

U

Cutin monomers/

a3

Pra1

Gγ

Kpp6 P

Figure 2cAMP and MAPK signaling during mating and pathogenic development. Components of the cAMP ( green) and MAPK (red ) signalingpathways and their interactions are indicated ( green or red arrows, respectively). Phosphorylation is symbolized by small circles labeledP in the color of the respective upstream kinase. Transcriptional activation is symbolized by black arrows. Postulated interactions andunknown components are indicated by question marks. (a) Components required for mating in a haploid a1 b1 cell. (b) Componentsrequired for pathogenic development in the dikaryon. Components that are specifically required during pathogenesis are highlighted inbright yellow. Components or domains with reduced or no function during pathogenic development are drawn smaller in size andfainter in color or are not shown, respectively. The seven transmembrane receptor Pra for perception of the Mfa pheromone andpostulated receptors for environmental cues and plant-derived stimuli are depicted in the schematic membrane. Characterizedcomponents of the cAMP pathway are the α and β subunits of a heterotrimeric G-protein, Gpa3 and Bpp1, respectively; the smallG-protein Ras1; the adenylate cyclase Uac1; the inhibitory and catalytic subunits of PKA Ubc1 and Adr1, respectively; and the putativeregulatory protein Hgl1. Hg1 is predicted to become inactive upon phosphorylation by Adr1, and this is indicated by brackets. Gpa3,Bpp1, and Ras1 all activate the adenylate cyclase Uac1. Components of the MAPK pathway are the small G-protein Ras2; the putativeCDC25-like guanyl nucleotide exchange factor for Ras2, Sql2; and the Ste50p-like adaptor protein Ubc2. The N terminus of Ubc2 isessential for mating but dispensable for pathogenic development, whereas the C terminus is essential for pathogenic development butplays no role in mating. The core MAPK module consists of the hierarchical kinases Kpp4 (Ubc4, MAPKKK), Fuz7 (Ubc5, MAPKK),and Kpp2 (Ubc3, MAPK) and the two alternative MAPKs, Crk1 and Kpp6. Downstream the transcriptional regulators Rop1 andHap2, a postulated Crk1-induced factor, and the central regulator Prf1 (blue) bind to four distinct elements in the prf1 promoter(indicated by different shapes and colors). Phosphorylated Prf1 induces expression of the a and b mating type genes. After cell fusionthe bE/bW heterodimer is formed that is the key regulator of pathogenic development. (1, 12, 15, 23, 29, 40, 41, 44, 47, 48, 55, 63, 64,66, 68, 76, 78, 79, 83, 84, 85, 105, 112).

www.annualreviews.org • Ustilago maydis as a Pathogen 427

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

components shown in Figure 2a are requiredfor efficient cell fusion in axenic culture.

The analysis of mutants in solopathogenicstrains has revealed that most components re-quired for pheromone signaling play additionalroles during pathogenesis, indicating that thecore constituents of the cAMP and MAPKpathways transmit signals from the plant en-vironment in addition to pheromone. How-ever, some additional components are requiredonly for pathogenesis, whereas others are dis-pensable (Figure 2a,b). Interestingly, the re-sulting signaling outputs during pathogenic de-velopment are fundamentally different fromthe responses during mating. We attribute theprime differences in signaling output betweenhaploid strains (Figure 2a) and dikaryotic (orsolopathogenic) strains to the absence and pres-ence of an active bE/bW heterodimer, respec-tively (Figure 2) (see following section for de-tails). Support for this scenario is provided bythe putative adaptor protein Ubc2, where ithas been demonstrated that the N-terminaldomain is required for mating, whereas thebasidiomycete-specific C terminus is essentialfor pathogenic development but plays no rolein mating (Figure 2; 63, 76). This could in-dicate that the distinct Ubc2 domains mediatethe assembly of specific MAPK-signaling com-plexes that differ during mating and pathogenicdevelopment. On the plant surface the tran-scription factor Rop1 and the β subunit bpp1of the heterotrimeric G-protein are dispens-able, although both are needed for cell fusion inaxenic culture (15, 84). It has been speculatedthat the role of Rop1 is taken over by a yet un-known regulator supposedly involved in sensingplant surface cues and acting via the upstream-activating sequence (UAS) in the prf1 promoter(15, 48). Bpp1 might be replaced by an uncon-ventional Gβ subunit (Figure 2b). Signalingcomponents that are required for pathogenesisbut not for mating are the Kpp2-related MAPKKpp6, Sql2, a Cdc25-like guanyl nucleotide ex-change factor predicted to control activity ofthe small G-protein Ras2, and the putative reg-ulatory protein Hgl1 (12, 29, 68, 83). The alter-native MAPK Kpp6 shares partially redundant

functions with Kpp2 during mating but has ahighly specific function in appressoria, where itis required for penetration of the plant cuticle(12). The phosphorylation status of Kpp2 andKpp6 is regulated by the dual specificity phos-phatase Rok1, and as a result rok1 mutants arehypervirulent (23). The specific need for Sql2during pathogenesis could indicate an involve-ment in transmission of plant-derived signalsthat feed into the MAPK module.

The function of the PKA target Hgl1, whichis predicted to be inactivated upon phospho-rylation by Adr1, becomes apparent late dur-ing the infection cycle, where it is essential forprogression from hyphal fragments to matureteliospores, possibly because of low PKA activ-ity during this stage (29). The need for tightlycontrolled PKA activity also becomes apparentin ubc1 mutants lacking the inhibitory subunitof the PKA or gpa3QL mutants in which cAMPsignaling is constitutive. Both mutants are ableto colonize plants, but ubc1 mutants fail to in-duce tumors, whereas gpa3QL mutants inducetumors that do not contain teliospores but tendto develop shoot-like structures (44, 66). Regu-lated PKA activity thus emerges as a critical de-terminant for the production and/or functionof U. maydis–emitted signals leading to tumorinduction and affecting tumor morphology.

The bE/bW Master Regulatorfor Pathogenic Development

After fusion of compatible haploid cells, an ac-tive heterodimeric bE/bW homeodomain tran-scription factor is generated that recognizes asequence containing a TGA-N9-TGA motif(56). The formation of this heterodimer is cen-tral for pathogenic development and cannot bebypassed by other means.

A U. maydis strain that expresses an in-ducible combination of bE1/bW2 genes was in-strumental to obtain a comprehensive view ofthe genes that are regulated by the bE/bWcomplex and because it made it possible to ob-tain a time-resolved view (13). Application ofsuch b-inducible strains in RNA fingerprint-ing and Affymetrix array analyses defined a set

428 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

of 347 b-regulated genes, of which 212 wereupregulated, and 135 were repressed after in-duction of the bE/bW heterodimer (12, 13,56). Functional classification revealed that cel-lular processes such as cell wall remodeling,lipid metabolism, cell cycle control, mitosis, andDNA replication were differentially regulatedby the active bE/bW heterodimer (56). Giventhat promoters of the majority of these geneslack putative bE/bW binding sites, the bE/bWheterodimer must induce a transcriptional cas-cade. Surprisingly, only a small fraction of thebE/bW-regulated genes proved to be requiredfor pathogenic development (13).

A bE/bW-regulated protein with an impor-tant role during plant colonization is the zincfinger transcription factor Biz1 (37). biz1 mu-tant cells are unaffected in mating and fila-mentation, but appressoria formation in biz1mutants is about 10-fold reduced compared towild type, and the few hyphae that invade theplant are found only in the epidermal layer (37).biz1 is expressed throughout biotrophic devel-opment, and this has been interpreted to indi-cate that the protein could have additional func-tions during these stages (37).

Another bE/bW-regulated protein is theMAPK Kpp6 (see previous section). Using aKpp6 version that cannot be phosphorylatedby the upstream MAPK, it was shown that re-spective mutants are able to form appresso-ria, but these fail to penetrate (12). This couldsuggest that these mutants are defective in theproduction of cell-wall-degrading or cell-wall-loosening enzymes.

One of the potential direct bE/bW targetsis clp1 (91). The clp1 gene product is related toClp1 from Coprinus cinereus, where it is requiredfor clamp cell formation (53), structures thatserve to guarantee correct nuclear distributionin the mitotically dividing dikaryotic filament.In a haploid strain of U. maydis, clp1 overexpres-sion has no effect, but when clp1 is expressedin strains with an active bE/bW complex, fila-mentation is strongly attenuated. clp1 is essen-tial for host colonization; i.e., clp1 mutants formappressoria, and these penetrate, but growth isarrested prior to the first mitotic division. The

defect could be tracked to the inability of clp1mutant strains to form clamp-like structures,resulting in disturbed nuclear distribution (91).Based on the finding that bE and bW expressionis unaffected by clp1 overexpression but a num-ber of genes that are upregulated by the bE/bWheterodimer are repressed, Scherer et al. (91)hypothesized that Clp1 could affect the regula-tory activity of the bE/bW heterodimer on theprotein level. Because the bE and bW proteinsare needed throughout fungal development in-side the plant, this inhibition must be of a tran-sient nature and has been speculated to operateat the level of nuclear localization of Clp1 (91).

An additional direct target of the bE/bWheterodimer is lga2, a gene in the a2 locus. lga2is upregulated through the pheromone cascade,and additional strong induction is conferredby the bE/bW heterodimer. The Lga2 proteinlocalizes to mitochondria, and overexpressionof lga2 (as is expected to occur in the dikaryon)interferes with cell growth, induces mitochon-drial fragmentation, and lowers mitochondrialrespiratory activity (10). lga2 interferes with mi-tochondrial fusion (10), and recent studies havedemonstrated that Lga2, together with Rga2,another mitochondrial protein encoded by thea2 locus, directs uniparental mitochondrialinheritance and represses recombination of mi-tochondrial DNA during sexual developmentof U. maydis (36). The other genes in the a locus,i.e., the pheromone and receptor genes, on theother hand, are downregulated by the bE/bWheterodimer (105). This effect is likely to beindirect, as no potential binding sites for thebE/bW heterodimer could be detected in therespective promoters. Downregulation of the amating type genes provides an explanation forthe observed attenuation of mating in strainsexpressing an active bE/bW heterodimer (67).These studies illustrate that the processesregulated through the bE/bW heterodimerare diverse and concern developmental stepson the leaf surface and after penetration. It isalso becoming evident that the cascade thatis triggered through bE/bW is not a meretranscriptional cascade but involves feedbackloops operating on the posttranscriptional

www.annualreviews.org • Ustilago maydis as a Pathogen 429

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

level. A comprehensive characterization ofthis cascade was once thought to identify thecomplete pathogenicity program of U. maydis.However, it is now becoming increasingly clearthat additional levels of regulation are switchedon once the fungus has come into contactwith the plant, and these have stimulatory orinhibitory effects on the processes regulated bythe bE/bW master regulator for pathogenicity.

Morphological Transitions AreIntimately Coupled to CellCycle Control

During its life cycle, U. maydis undergoes anumber of discrete morphological transitionswhose timing and execution are decisive forpathogenic development. The most dramaticof these transitions is the switch from buddingto polarized hyphal growth that is initiated afterperception of pheromones and continues aftercell fusion through much of the disease cycle.Polarized growth of hyphae requires bothmicrotubules and F-actin. As a consequence,pharmacological disruption of actin cables ormicrotubules leads to defects in conjugationhyphae, cell fusion, and development of dikary-otic hyphae. This is supported by the findingthat Myo5, a class V myosin predicted totransport membranous vesicles along F-actinfilaments, is required for hyphal growth as wellas pathogenicity. Microtubules, on the otherhand, are specifically required for long-distancetransport in all U. maydis hyphae that extendbeyond 50–60 μm, and this is reflected by theneed of kinesin-1 and kinesin-3 motors (98).

To set up a new morphogenetic programlike the switch from budding to filamentousgrowth requires the cell cycle to pause. Thefinding that overexpression of the G1 cyclincln1 blocks sexual development but that itsabsence enables cells to express mating typegenes has led to the proposition that an activecell cycle represses mating, whereas the induc-tion of mating arrests the cell cycle; i.e., thesechoices become incompatible (17). Underpoor nutrient conditions (as are expected tooccur on the leaf surface) both mitotic cyclin

genes of U. maydis are downregulated, andthis was shown to stimulate prf expression andfacilitates entry into the mating program (17).Upon pheromone perception, the haploid cellsarrest budding growth in the G2 phase (39) andstart formation of conjugation tubes (96). Con-jugation tubes are filled with cytoplasm, fuse attheir tips, and establish the dikaryotic hyphae.These hyphae can be up to 20 times the lengthof haploid cells and develop the characteristicsof hyphae seen on the plant surface, that is, theyshow tip growth, the cytoplasm accumulates inthe tip cell compartment, and the older parts ofthese hyphae appear highly vacuolated and aresealed off by regularly spaced septae (38, 93).

After cell fusion the cell cycle arrest is main-tained, at least partially through the action ofBiz1, one of the bE/bW-induced transcriptionfactors. Biz1 is required for appressorium for-mation and proliferation inside the host plant(37). Biz1 represses the expression of the cyclinclb1, resulting in G2-arrest (37). Based on theseobservations, it was suggested that the Biz1-induced cell cycle arrest is important thoughprobably not sufficient for appressorium forma-tion (37). The fact that biz1 is highly expressedduring the entire infection process might in-dicate that U. maydis cells growing within theplant tissue continuously need to go throughcell cycle arrest stages. We consider it an attrac-tive possibility that this may turn out to be par-ticularly important during intracellular growthwhen cells traverse from one cell to the next.This process is very similar to the penetrationstep of the epidermis (25) and may need a cellcycle–arrested hyphae to reform appressoria-like structures.

However, the open question is how the cellcycle arrest is (intermittently) released duringbiotrophic development. It is evident that thehyphae penetrating the epidermis are still cellcycle arrested, as they show the typical emptysections also seen in the dikaryotic filamentsgrowing on the leaf surface (94). We considerit feasible that Clp1, the protein needed forclamp formation, assumes this role: clp1 mu-tants are unable to release the bE/bW-triggeredcell cycle arrest, and the forced expression of

430 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

clp1 in strains where filamentation is inducedby an active bE/bW heterodimer suppresses fil-amentation, that is, the cell cycle block (91).Currently, it is not clear how this presumedcycling could be regulated, but we would liketo speculate that the timing may be set by theappressorium-like structures that allow hyphaeto spread from one cell to the next. Later, whenthe hyphae grow in the apoplast, branch, formlobed structures and proliferate massively, frag-ment, and produce rounded spores, there mustbe additional regulators that control these mor-phogenetic events. Their identity is currentlyunknown, and without availability of a systemwhere these events can be triggered in axenicculture, it will be difficult to identify the genesinvolved. The tight link between cell cycle andpathogenic differentiation makes it mandatoryto study in detail the mechanisms that regulatethe cell cycle during the biotrophic phase.

Sensing the Plant Surface InvolvesChemical and Physical Cues

The plant cuticle is composed of two lay-ers, an inner layer consisting of intracuticu-lar waxes associated with a polyester matrix ofcutin and a continuous surface layer of epicu-ticular wax without cutin (32, 65). Cutin con-sists of a network of interesterified hydroxyland epoxy derivatives of C16 and C18 fatty acids(89). Plant surfaces thus provide a reservoir ofchemical and physical cues, of which epicutic-ular waxes and cutin monomers, as well as to-pography, hardness, and hydrophobicity, havebeen shown to trigger morphogenetic events inphytopathogenic fungi (101). On the leaf sur-face, compatible haploid strains of U. maydismate and then switch to filamentous growth,whereas solopathogenic haploid strains switchto filamentous growth without prior mating.

Following hyphal tip swelling, non-melanized appressoria are formed that directlypenetrate the epidermis, most likely aided bycell wall–degrading enzymes (92, 94). Recently,a marker gene that is specifically expressed inthe tip cell of those hyphae that have formedan appressorium has been identified, and after

fusion to green fluorescent protein (GFP),this allowed simple scoring of appressoriumformation (79). With the help of this marker,it was demonstrated that hydrophobicity is anessential signal for appressoria development(79). Hydroxy-fatty acids, like cutin monomersfrom maize leaves, enhance appressoriumdevelopment on hydrophobic surfaces (79).These morphogenetic events require an activebE/bW transcription factor, suggesting thatthe program that is induced by these cuesfrom the plant surface relies on the programthat is triggered by bE/bW. Hydroxy-fattyacids induce short and nonseptated filamentsresembling conjugation tubes, whereas ahydrophobic surface induces long and septatedfilaments that look like the filaments observedon the plant surface (79, 94). In haploidcells hydroxy-fatty acids induce pheromonegene transcription but not conjugation tubeformation (79). It is currently unknown howthe hydroxy-fatty acid signal is perceivedand transmitted to induce the pheromonegenes. Also, in solopathogenic haploid strainswhere the pheromone genes are repressed bythe bE/bW heterodimer, the hydroxy-fattyacid signal elevates pheromone gene expres-sion. The following increase in autocrinepheromone stimulation is responsible for theobserved morphological response (79).

The effect of hydroxy-fatty acids on U. may-dis development depends on the phosphoryla-tion of Kpp2. The genetic activation of kpp2bypasses the need of hydroxy-fatty acids for ap-pressorium formation but is not able to bypassthe need for the hydrophobic stimulus (79).The hydroxy-fatty acid signal is also dispens-able when rok1, which negatively modulatesthe phosphorylation status of Kpp2, is deleted(Figure 2b; 23). Kpp2 activity is also essentialfor hydrophobicity-induced filamentation, butso far it is unknown how hydrophobicity is per-ceived. Pth11, a nine-transmembrane proteinin Magnaporthe grisea has been suggested as aputative surface-sensing protein, and pth11 mu-tants are impaired in their response to both hy-drophobicity and cutin monomers (22). How-ever, U. maydis lacks an obvious ortholog to

www.annualreviews.org • Ustilago maydis as a Pathogen 431

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

this protein. Because in U. maydis hydroxy-fattyacids and hydrophobicity are both needed forthe efficient induction of appressoria, a certainthreshold activation of the underlying MAPKsignaling pathway might be needed that wouldrequire combination of both signals (79). Onthe leaf surface, U. maydis differentiation couldthen proceed in two stages. After cell fusion, fil-amentation could be initially induced throughthe hydrophobic surface. The filament mightthen secrete cutin degrading enzymes that lib-erate cutin monomers. This would amplify thefilamentation response and at the same timeinduce appressoria. Based on the finding thatlipids induce a filamentous phenotype that re-sembles the infectious cell type, lipids are alsoproposed to act as one of the signals that pro-motes and maintains filamentous growth ofU. maydis in the host environment (62).

Identification of cutin monomers and hy-drophobicity as in vitro cues for filamenta-tion and appressoria development in U. may-dis now opens the possibility to analyze howthese signals are perceived and how they triggerdifferentiation and function of nonmelanizedappressoria.

U. maydis Infection Induces DistinctPlant Responses and SuppressesInduction of Programmed Cell Death

As a biotrophic pathogen, U. maydis depends onsurvival of colonized host cells. Nevertheless,one does observe early symptoms that becomemacroscopically visible 12–24 h after inocula-tion and include chlorosis and small necroticspots at the sites of infection. In some casesthese symptoms are likely to result from un-successful penetration events, but in addition,the plant appears to recognize U. maydis hy-phae during intracellular development in theepidermal layer and during cell-to-cell move-ment. Usually, the colonized plant cells stayalive, whereas plant cells containing older hy-phae that lack cytoplasm undergo cell death (26,27). At later stages, U. maydis–induced tumorsare formed by enlargement and proliferationof plant cells. Large fungal aggregates form in

tumors, and this occurs without the elicitationof programmed cell death in the surroundingplant tissue (27). Induction of tumor growthis also accompanied by accumulation of antho-cyanins, resulting in a red pigmentation of in-fected tissue. The transcriptional responses ofmaize plants were initially studied by differen-tial display and subsequently at distinct stagesafter infection with a solopathogenic U. may-dis strain by using the Affymetrix microarraysystem (4, 26). This latter approach shows aninitial recognition of the fungus on the leaf sur-face that elicits strong, rather unspecific plantdefense responses (Figure 1a; 26).

The early transient upregulation of plantdefense genes suggests that the plant recognizesU. maydis via conserved pathogen-associatedmolecular patterns (PAMPs). Recently, it hasbeen shown that specific PAMP receptors aretranscriptionally upregulated after elicitation(113, 114). Similarly, during the early phaseof U. maydis infection, upregulation of twoputative membrane-bound leucine-rich repeat(LRR) receptor-like kinases is observed.Moreover, a somatic embryogenesis receptor(SER)-like kinase is induced that belongs toa group including BAK1, which has recentlybeen shown to act as a positive regulatorof infection-induced cell death signaling(18, 61). These observations suggest thatPAMP-induced signaling is activated priorto and during the penetration phase of U.maydis. With establishment of the biotrophicinteraction 24 h after infection, these initialresponses are attenuated (Figure 1b; 26). Inparallel, induction of genes coding for celldeath suppressors such as Bax-Inhibitor-1 (31)and the repression of caspases is observed(26). This is likely to reflect the suppression ofprogrammed cell death by U. maydis once thebiotrophic interface has been established.

Reactive oxygen species (ROS) that areformed in response to fungal pathogens areimportant regulators of cell death (109).Necrotrophic pathogens such as Botrytis cinereamassively induce the production of H2O2 andsuperoxide radicals, which leads to runaway celldeath in the infected tissue (45). It is obvious

432 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

that massive accumulations of ROS must beprevented in a biotrophic interaction. ForU. maydis it has been shown that the fungusactively needs to detoxify H2O2 to be fully vir-ulent, and this occurs through a system relatedto the Saccharomyces cerevisiae Yap1p regulator(80). U. maydis yap1 mutants show higher sensi-tivity to H2O2 than wild-type cells and displayreduced proliferation in the infected tissue (80).Around intracellularly growing yap1 mutant hy-phae, H2O2 accumulates, which is never seenaround infecting wild-type hyphae. BecauseU. maydis does not code for NADPH oxidasegenes, pharmacological inhibition of NADPHoxidase could be used to specifically target plantenzymes in infected tissue. Such conditionslargely restore virulence of yap1 deletion strains(80). This illustrates that virulence of U. maydisdepends on its ability to detoxify ROS.

A number of plant proteins such as glu-tathione S-transferases (GSTs) are associ-ated with scavenging of oxygen radicals afterpathogen attack (75). Just 12 h after infectionwith U. maydis, seven genes coding for maizeGSTs are transcriptionally induced. Moreover,Tau-class GSTs, which have been shown to sup-press Bax-mediated cell death induction (60),are strongly induced 24 h postinfection whenbiotrophy is established (26). The total con-tent of glutathione is also increased through-out U. maydis infection. This likely reflects therequirement for a higher antioxidative capac-ity in colonized tissue (26), which may aid inprevention of cell death.

Progression of U. maydis infection is ac-companied by distinct transcriptional changesof plant hormone signaling genes. Whereasnecrotrophic pathogens induce salicylic acid(SA)-dependent cell death responses includingexpression of defense-related genes suchas PR1 (106), biotrophic pathogens induceprimarily JA and ethylene responses duringcompatible interactions (42). These responsesare associated with an induction of tryptophanbiosynthesis, the accumulation of secondarymetabolites, and the induction of plant genesencoding defensins (14, 42, 108). In linewith this, PR1 expression was undetectable

during the early biotrophic phase of U. maydisinfection, whereas activation of typical JA-responsive defense genes such as defensins andBowman-Birk-like proteinase inhibitors wasobserved (Figure 1c,d; 26).

At later stages, when tumors are formed,both auxin synthesis and auxin-responsivegenes are induced (Figure 1d,e; 26). Thissuggests a role of plant-derived auxin in cellenlargement during tumor formation. Becauseauxin signaling was shown to be antagonisticto SA signaling in Arabidopsis thaliana (107),auxin could contribute to the suppression of SAsignaling and thereby promote fungal growthand host susceptibility to U. maydis. The ob-servation that U. maydis–induced plant tumorscontain up to 20-fold higher auxin levels thannoninfected tissue was made almost 50 yearsago (102). U. maydis can produce auxin, andin the meantime the underlying biosyntheticpathways for auxin have been identified (90).The starting point of auxin biosynthesis istryptophan, which can be deaminated toindole-3-pyruvate (IP) by the enzymes tam1and tam2. This is followed by the decarboxy-lation to indole-3-acetaldehyde, which is thenconverted to indole-acetic acid (IAA) by twoindole-3-acetaldehyde dehydrogenases Iad1and Iad2. U. maydis strains in which tam1,tam2, iad1, and iad2 are simultaneously deletedare not compromised in tumor formation,but these tumors contain significantly lowerlevels of free IAA compared to tumors inducedby wild-type strains (90). This rules outinvolvement of U. maydis–produced auxin intumor formation. However, the possibilityremains that elevated auxin biosynthesis, as wasobserved in U. maydis–infected maize plants(26), stimulates tumor development throughhost cell enlargement.

Together, these observations suggest thatU. maydis is initially recognized by the plant in-nate immune system and induces a nonspecific,PAMP-triggered response. Once U. maydis hasentered the host tissue, a biotrophic interfaceis established in which these defense responsesand, in particular, programmed cell death aresuppressed.

www.annualreviews.org • Ustilago maydis as a Pathogen 433

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

Resistance to U. maydis InfectionIs a Polygenic Trait

In the U. maydis–maize system, gene-for-geneinteractions have not been identified, althoughthey provide durable resistance in other smutpathosystems. For example, in the 14 races ofUstilago hordei, six avirulence genes have beenidentified that have corresponding resistancegenes in barley cultivars (99). In current workthese avirulence genes have been mapped, andtheir molecular characterization is under way(70). U. maydis resistance in maize, on the otherhand, is a polygenic trait, and specific loci maycontribute to resistance in either the ear or thetassel (8). In the most recent study using tworecombinant inbred populations, three quanti-tative trait loci (QTL) mapping to similar re-gions in both populations have been identifiedthat contribute to the frequency and severityof U. maydis infection over the entire plant (8).Depending on the population, several QTLscontribute to the frequency or severity of in-fection in only the tassel, the stalk, or the eartissue (8). Currently, it is not clear whetherthe tissue-specific resistance provided by theseQTLs is direct or operates through secondarytraits such as husk cover, surface properties, orflowering time that could limit access of the fun-gus to specific plant organs. Interestingly, sev-eral of the QTL regions defined in the studyby Baumgarten et al. (8) contain genes witha known role in pathogen resistance such asnucleotide-binding site (NBS)-LRR resistancegene homologs, a pathogenesis-related protein,a chitinase, a basal antifungal protein, and awound-inducible protein. Future studies willshow whether these genes contribute to the ac-tivity of a given QTL. It will also be of inter-est to elucidate whether U. maydis mutants thatlack specific effectors (see below) show a differ-ential response on the respective inbred lines,as this could indicate that certain QTL geneproducts might be targeted by the respectiveeffectors. Another interesting area of investi-gation would be to assess whether the QTLsmapped for U. maydis resistance also provideresistance to Sporisorium reilianum, the cause

of head smut in corn, a disease with the moretypical smut symptoms occurring only in theinflorescences.

Secreted Effectors CounteractPlant Defenses

To establish compatibility, biotrophicpathogens need to overcome the basal, PAMP-triggered plant defense mechanisms (54).To counteract defense responses, microbialpathogens secrete a variety of proteins, so-called effectors, that interfere with plantdefenses and thereby trigger susceptibility (54).Effector proteins show a high structural andfunctional variety and are often completelynovel proteins that lack conserved functionaldomains. To establish and maintain biotrophy,effectors have to interfere with various pro-cesses in the host tissue such as primary (andsecondary) metabolic pathways, cell-deathregulation, hormone signaling and/or cellgrowth (43). To cope with plant defenseresponses, fungal pathogens have developedseveral strategies: (a) They can modify their cellwall to evade host recognition, (b) they can se-quester breakdown products of fungal cell wallsthat trigger host defenses, (c) they can coun-teract the activity of antimicrobial enzymesproduced by the plant via apoplastic effectors,or (d ) they can interfere with plant defenseresponses via intracellular effectors that aretranslocated to plant cells (88). For two rustfungi and Colletotrichum graminicola, it has beendemonstrated that invading hyphae modifytheir chitin to chitosan, and this is discussed as amechanism that could protect invading hyphaefrom hydrolysis by host chitinases or to avoidthe generation of PAMPs that could triggerdefense responses (33). The U. maydis genomecontains six putative chitin deacetylase genes,but whether these genes have the expectedfunction has not been investigated so far, norhas it been shown whether they contribute toescape from host recognition. Recently, twoproteins with a LysM domain have been iden-tified in the U. maydis genome (20). Both could

434 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

potentially be secreted. The LysM domain is acarbohydrate-binding module that is proposedto sequester chitin oligosaccharides that couldelicit host immune responses and/or protectfungal hyphae against chitinases (20). Basedon the finding that the Cladosporium fulvumLysM effector Ecp6 is a virulence factor (20),it will be rewarding to determine whether theU. maydis LysM proteins also have a virulencefunction.

U. maydis encodes 386 putatively secretedproteins that are not predicted to have an enzy-matic function, and of these, 272 code for eitherU. maydis–specific proteins or conserved pro-teins lacking defined InterPro domains. Thesesecreted proteins have recently been groupedinto repetitive proteins with and without inter-nal Kex2 protease cleavage sites, cysteine-richproteins, proteins with a presumed regulatoryfunction, and a large group of proteins with-out recognizable features (81). Only a smallfraction of the corresponding genes has beencharacterized molecularly (Figure 1). These in-clude the repellent gene rep1 that encodes 11secreted repellent peptides generated by Kex2-dependent cleavage of a precursor protein. Inaddition, two hydrophobin genes have beencharacterized. hum2 encodes a typical class I hy-drophobin, whereas hum3 encodes a repeateddomain likely to be processed by Kex2 fusedto a hydrophobin domain (100). In contrast toother fungi, the repellents, rather than the hy-drophobins, are required for hyphal hydropho-bicity, aerial hyphae formation, and attachmentto hydrophobic surfaces. However, neither thedeletion of rep1 nor the simultaneous deletionof hum2 or hum3 affects pathogenic develop-ment (100, 110). This makes it unlikely thatthese molecules, which attach tightly to the fun-gal cell wall, have a protective function duringpathogenesis. Interestingly, a third secreted andKex2-processed repetitive protein, Rsp1, has arole during pathogenesis, but this becomes ap-parent only when hum3 is deleted simultane-ously (82). The fact that the hum2/hum3 doubledeletion strains are fully virulent strongly sug-gests that the processed peptides from Rsp1 andthe repetitive domain of Hum3 have redundant

functions. The rsp1/hum3 double mutants areunaffected in mating and penetration but arrestgrowth during early intracellular developmentwithout signs of enhanced plant defense (82).This could indicate that the respective peptideshave a signaling function. Alternatively, theymight facilitate nutrient uptake by changing theinterface between fungus and host membrane.

Another family of secreted effectors thatwere originally identified because of their enor-mous transcriptional upregulation in infectedtissue is encoded by the mig1 and mig 2 geneclusters (Figure 1). These clusters consist oftwo and six genes, respectively, that each codefor small related proteins with a characteris-tic spacing of Cys residues reminiscent of fun-gal Avr gene products from Cladosporium ful-vum (5, 6, 21, 35). However, under the testedconditions neither the mig1 nor the mig2 genesare required for pathogenic development. Thiscould suggest additional redundancy with othercysteine-rich effectors. Alternatively the miggene products might indeed confer avirulenceon some hosts which carry the cognate resis-tance gene(s). In order to identify such host va-rieties, one might screen different accessions ofteosinte, the closest wild relative to maize (24),which is the only additional host for U. maydis(19).

U. maydis was the first eukaryotic pathogenin which it was discovered that novel effectorswith relevant functions for pathogenic devel-opment are situated in gene clusters that aretranscriptionally upregulated in tumor tissue(59). Twelve such gene clusters were identified,which all code for novel secreted proteins, andin several instances these belong to small genefamilies. Four of the cluster deletion mutantsare significantly attenuated in virulence andshow defects at different stages of pathogenicdevelopment. Remarkably, deletion of one clus-ter results in increased virulence (59). The exis-tence of secreted effectors negatively affectingvirulence indicates that U. maydis does not useits full virulence potential, presumably becausethis might result in host death prior to com-pletion of the life cycle and spore formation.One cluster encoding 24 secreted effectors is

www.annualreviews.org • Ustilago maydis as a Pathogen 435

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

of particular interest with respect to tumor de-velopment, as mutants still proliferate in the in-fected tissue but fail to induce tumors (59). Cur-rently, it is not clear which or how many effec-tors from a given cluster are responsible for theobserved phenotypes or with which processesthey interfere. Nevertheless, because individ-ual cluster mutants are affected at discrete stepsof biotrophic development (Figure 1), the re-sponsible effectors must have distinct cellulartargets. To elucidate their functions, the sub-cellular localization of the respective proteinsneeds to be determined, in particular whetherthey act in the apoplast and biotrophic inter-face or after being translocated into plant cells.Furthermore, we expect that identification ofhost proteins interacting with secreted effec-tors will provide the crucial leads to effectorfunction.

In addition to these gene clusters, a novelsecreted effector has recently been shown to beessential for establishment of the biotrophic in-teraction of U. maydis with maize plants. Thisprotein, termed Pep1, is dispensable for sapro-phytic growth of U. maydis, but deletion mu-tants are completely blocked in biotrophic de-velopment (25). Confocal microscopy showedthat pep1 mutants are able to penetrate the plantcell wall, but subsequent invasion of the hostcell stops immediately after invagination of theplant plasma membrane. At the same time, themutant induces strong plant defense responsesincluding programmed cell death of attackedepidermis cells. Fluorescently tagged versionsand immunolocalization show that Pep1 pro-tein is secreted from intracellular hyphae intothe biotrophic interface, and particularly strongaccumulations are seen when hyphae invadeother cells. The molecular function of Pep1 ispresently unclear. Because a Pep1 ortholog ofthe barley-covered smut fungus U. hordei is ableto complement U. maydis pep1 deletion mutantsand because U. hordei pep1 deletion strains ar-rest at a similar stage as corresponding U. maydismutants, it has been proposed that Pep1 pro-teins are essential compatibility factors in smutfungi (25).

Extracellular Enzymes and Nutritionin the Host Environment

Genome comparisons between thenecrotrophic fungi Magnaporthe grisea andFusarium graminearum with the biotrophU. maydis revealed a relatively small number(33 versus 138 and 103) of hydrolytic secretedenzymes (59, 81). This is in line with thebiotrophic lifestyle of U. maydis, in whichextensive tissue damage that could result in theproduction of cell-wall degradation productsthat may trigger host defense responses (28)is avoided. The genes for cell-wall-degradingenzymes follow distinct expression profiles atearly and late stages of tumor development(27), but so far no systematic analysis ofmutants has been performed. For the set ofthree genes coding for pectinolytic enzymes,deletion mutants exist, but neither single nortriple mutants show defects in pathogenicdevelopment (27). Based on these studies, ithas been speculated that U. maydis uses its setof plant-cell-wall-degrading enzymes largelyfor softening the plant-cell-wall structure as aprerequisite for in planta growth rather thanfor feeding on carbohydrates derived from thedigestion of plant-cell-wall material (27). Fornone of the other secreted enzyme families,that is, the 23 proteases or the 11 lipases(81), has an attempt been made to generatemutants, as the fear has been that owing tofunctional redundancy no phenotype wouldbe observed. With the establishment of theFLP-FRT recombination system for U. maydis(Y. Khrunyk and R. Kahmann, unpublished)such experiments may be feasible in the nearfuture.

As part of the establishment of thebiotrophic lifestyle, U. maydis must enter a com-petition with his host for carbohydrates andtrace elements. So far only the aspect of iron ac-quisition by U. maydis has been studied in detail.

The essential trace element iron is takenup by two high-affinity iron uptake systemsin U. maydis. One system is permease based;the other relies on the secreted hydroxam-ate siderophores ferrichrome and ferrichrome

436 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

A (16, 30). Both siderophores are cyclic pep-tides produced by nonribosomal peptide syn-thases. The genes necessary for the synthe-sis of siderophores and for the permease-basedhigh-affinity iron uptake are organized in threeclusters that are under negative control bythe iron-responsive GATA transcription fac-tor Urbs1 (30, 69). Under low-iron condi-tions the biosynthesis genes are induced andsiderophores are secreted. The siderophore sys-tem is not required for virulence (77), butstrains with deletions in the permease-basedhigh-affinity iron uptake system show a dras-tic virulence reduction (30). Recent data sug-gest that the siderophore system is essentialfor iron storage in U. maydis spores and is up-regulated transcriptionally during spore for-mation (B. Winterberg, R. Kahmann, and J.Schirawski, personal communication).

Transcript profiling of infected and unin-fected maize leaves at different stages of de-velopment provided significant new insightsinto the influence of U. maydis infection onmetabolic regulation and carbon fluxes. Dur-ing development, maize leaves undergo a tran-sition from a sink to a source tissue. Whenleaves emerge from the stem and the blade un-rolls, it becomes exposed to light, which inducesmassive transcriptional changes in the devel-oping leaf tissue. Major aspects of these de-velopmental changes are the activation of thephotosynthetic apparatus, RNA synthesis, andsucrose as well as starch synthesis (26). On themetabolic level, a significant decrease of freehexoses can be observed during leaf develop-ment, which is indicative of photosynthetic ac-tive source tissue (50). These transitions duringnormal leaf development are massively changedin U. maydis–infected tissue. Transcriptional ac-tivation of genes involved in photosynthesis isalmost completely blocked, and the switch fromC3 to C4 metabolism that is observed during de-velopment of noninfected maize leaves does notoccur (26, 50). At the same time, genes involvedin sucrose degradation, glycolysis, and the tri-carboxylic acid cycle are strongly induced in in-fected tissue, whereas sucrose synthesis genesare downregulated. In line with this, the content

of free hexoses remains at the high level charac-teristic for immature sink leaves. These obser-vations indicate that sucrose is imported fromuninfected tissue to the developing U. maydistumors. Tumors thus represent an artificial sinkorgan supporting fungal growth by providingplant-derived hexoses (26).

For acquisition of these carbon sources,U. maydis is equipped with 19 potential hex-ose transporters. Of these, one general hexosetransporter and one novel transporter for su-crose were recently shown to be required forfull virulence (R. Wahl, K. Wippel, J. Kamperand N. Sauer, personal communication). Thisis an exciting discovery, as it links the phys-iological status of the infected tissue to thecarbon sources preferred by U. maydis duringbiotrophic growth.

Comparative Genomics: FromSequence Informationto Biological Insights?

Comparative genomics of the few sequencedbasidiomycetes, U. maydis, Laccaria bicolor, Cryp-tococcus neoformans, Phanerochaete chrysosporium,and Malazezzia globosa, reveal dramatic differ-ences in genome size. Whereas the ectomy-corrhizal symbiont L. bicolor (65 Mb; 20,614predicted genes) to date carries the largest se-quenced fungal genome, the skin-inhabitingfungus M. globosa contains only 4285 predictedgenes in an assembled genome sequence ofabout 9 Mb, which is among the smallest forsequenced free-living fungi (72, 111). U. may-dis (20.5 Mb; 6785 predicted genes) and theopportunistic human pathogen C. neoformans(20 Mb; 6574 predicted genes) have similar sizegenomes and gene numbers, whereas the whiterot fungus P. chrysosporium has an intermedi-ate genome size of 30 Mb (11,777 predictedprotein-encoding genes) (52, 59, 71, 73).

In addition to these significant differencesin size, the sequenced genomes show tremen-dous variation in structure, that is, with respectto the abundance of introns and repeti-tive transposon-derived sequences. Further-more, comparison of these genomes reveals a

www.annualreviews.org • Ustilago maydis as a Pathogen 437

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

remarkable absence of synteny. Even M. glo-bosa and U. maydis, where 82% of the predictedM. globosa proteins show significant sequenceconservation to U. maydis proteins (52% aver-age amino acid identity), show strikingly littlesynteny. In addition, the clusters of secreted ef-fector genes, which are crucial determinants forthe biotrophic lifestyle in U. maydis, do not existin the M. globosa genome (111). Gene invento-ries in the sequenced basidiomycetes stronglyreflect their specialized environmental niches.The Malassezia species that live in close associ-ation with human and animal skin lack a func-tional fatty acid synthase gene and satisfy theirneed for fatty acids from external sources byan expanded set of secreted lipases, phospholi-pases, and acid sphingomyelinases (111). Thesaprophytic, wood-degrading basidiomyceteP. chrysosporium, on the other hand, encodeslarge number of secreted oxidases, peroxidases,and hydrolytic enzymes that cooperate in lignindegradation (73). The genome sequence of theectomycorrhizal fungus L. bicolor, a biotrophicsymbiont of trees, shows striking expansions ingenes, which are predicted to play roles in sig-nal transduction (72). This might reflect an in-timate cross talk between symbiont and host aswell as the need to coordinate the complex dif-ferentiation processes during the developmentof multicellular fruiting bodies. Interestingly,the L. bicolor gene inventory revealed some sim-ilarities to the biotrophic pathogen U. maydis,that is, a restricted number of enzymes for thedegradation of plant-cell-wall polysaccharides(59, 72) and an array of small secreted effector-type proteins with unknown functions that arespecifically expressed in symbiotic tissue (72).Analogous to the role of the U. maydis geneclusters encoding secreted effectors (see pre-vious section; 59), this set of mycorrhizal pro-teins might play decisive roles in the establish-ment and subsequent beneficial colonization ofthe host. However, most (82%) of these 2931predicted secreted proteins from L. bicolor areexclusively found in this fungus (72).

It is tempting to speculate that the species-specific effector sets may manipulate leaf andflower tissue for parasitic colonization by

U. maydis and facilitate the ectomycorrhizalsymbiosis with roots by L. bicolor, respectively.These comparisons illustrate that with respectto pathogenicity, relatively little can be gainedfrom a comparison of fungal genomes thatare all classified as basidiomycetes but inhabitvery different ecological niches. A much morepromising approach that has been pioneeredin oomycete pathogens is genome compar-isons of closely related species: Many of theoomycete genomes are syntenic (104), but ef-fector gene families, despite their common evo-lutionary origin, are highly diverse and areevolving rapidly, presumably owing to coevo-lutionary pressure from the host species (103).We have recently sequenced the genome of thehead smut fungus S. reilianum, a close relativeof U. maydis, which also infects maize but elic-its distinct disease symptoms only in the flowertissue. The S. reilianum genome shows a veryhigh level of synteny to the U. maydis genome.Interestingly, many of the effector genes iden-tified in the U. maydis genome are also found inS. reilianum, but they display low sequence con-servation and often differ in copy number. Inaddition, a number of species-specific genes en-coding effectors were found ( J. Schirawski andR. Kahmann, unpublished). This serves as animportant guide for experimental approaches(i.e., for prioritizing which genes to delete),will permit study of the relevance of species-specific constriction and expansions of effectorgene families, and might help map functionallyrelevant protein domains.

CONCLUSIONS

The studies of U. maydis as a pathogen haveachieved a new level of understanding thatgoes far beyond what has started it all, thatis, the characterization of the mating-typeloci and how they control morphogenesis andpathogenic development. We are now in a sit-uation where the processes that are controlledby the central regulator for pathogenicity, thebE/bW complex, are studied on the genome-wide level and where plant inputs into theseprocesses can be analyzed. The same holds for

438 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

the fungal signaling pathways that were initiallystudied because of their role during mating, andit is now clear that they assume new roles dur-ing virulence. The connection between mor-phological transitions and the cell cycle hasbeen established, and from this insights into theelaborate control of cell cycle progression dur-ing fungal in planta development are emerg-ing. The cues that trigger U. maydis differen-tiation on the leaf surface up to the formationof appressoria have been uncovered, and thisshould pave the way for elucidating how thesesignals are perceived. Access to the genomesequence has provided unprecedented insightsinto how U. maydis, with the help of novel se-creted effector proteins, adapts to growth anddevelopment in the plant environment and es-tablishes a biotrophic relationship with its hostplant. One of the so far unique features of U.maydis is that mutants in single effector genesor effector gene clusters have rather dramatic

effects on biotrophic development. Such a lowdegree of genetic redundancy is in stark contrastto oomycetes, where it seems likely that largenumbers of effector genes with similar func-tions exist (103). Given the low level of geneticredundancy in the effector repertoire of U. may-dis and given the ease with which gene replace-ment mutants can be generated, a systematiceffector gene knockout strategy promises in-sights into all aspects and steps of biotrophicdevelopment that are modulated by effectors.The genome-wide analysis of plant responses toinfection by U. maydis has uncovered several keyprocesses that are changed, that is, defense re-sponses, cell death suppression, phytohormonesignaling, and photosynthesis. The challengeahead will be to connect these processes withspecific effectors and to elucidate which of theseprocesses are necessary for the establishmentof a compatible relationship between U. maydisand maize plants.

SUMMARY POINTS

1. U. maydis continues to serve as an instructive model for biotrophic fungal plant pathogens.

2. Studies of the U. maydis mating program, morphological transitions, cell cycle, regulatorycascades, signal transduction pathways, and nutrition during the biotrophic phase areleading to an integrated view of the fungal processes that need to be adjusted for growthand differentiation in the plant environment.

3. Plant responses to infection by U. maydis are providing leads to those processes in theplant that need to be reprogrammed for compatibility.

4. U. maydis uses a multilayered system to deal with plant responses that relies on thedetoxification of ROS as well as on secreted effectors.

5. The growing repertoire of secreted U. maydis effectors with crucial roles during differentstages of pathogenic development provides insights into the different levels of fungalcommunication with the host plant.

FUTURE ISSUES

1. The established framework for the signaling pathways that are relevant for disease shouldallow determination of how the outputs of these pathways are modulated by plant signals.

2. Based on the success of mutant complementation in U. maydis by genes from moredistantly related biotrophic pathogens such as rust fungi (51), it should be possible to de-termine whether effectors from such obligate biotrophs can complement the phenotypeof U. maydis effector mutants.

www.annualreviews.org • Ustilago maydis as a Pathogen 439

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

3. With respect to the effectors of U. maydis, one of the big challenges will be to determinewhere they function, that is, whether they reside in the apoplast or are taken up by plantcells, which processes they control in the host plant, and which of these processes arerelevant for disease progression.

4. It is presently an open question whether it will be possible to study effectors from smutfungi that parasitize on grasses in A. thaliana, where a wealth of mutants and tools wouldbe available to speed up functional analyses. Therefore, the accessibility of hosts suchas maize for reverse genetics needs major improvement, and efficient transient genesilencing systems need to be established.

5. The manually annotated genome sequence of U. maydis is expected to serve as a scaffoldfor the assembly of the genomes of closely related pathogens that can now be cost-effectively sequenced by 454-technology. By extending comparative approaches to closelyrelated smuts that parasitize on different hosts, we expect to obtain insights not only intothe repertoire of critical determinants for disease and their evolution but also into hostrange.

DISCLOSURE STATEMENT

The authors are not aware of any affiliations, memberships, funding, or financial holdings thatmight be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS

We apologize to those whose original work could not be cited in this review because of spacelimitations. We thank the groups of Jorg Kamper and Jan Schirawski for sharing unpublisheddata and Gertrud Mannhaupt for bioinformatics support. Our work has been supported throughthe DFG Collaborative Research Center 593, the DFG Research Group FOR 666, DFG grantDJ64/1-1 to A. D., a Humboldt fellowship to A. M.-M., and funds from the Max Planck Society.

LITERATURE CITED

1. Banuett F. 2007. History of the mating types in Ustilago maydis. See Ref. 49, pp. 351–752. Banuett F, Herskowitz I. 1989. Different a alleles of Ustilago maydis are necessary for maintenance of

filamentous growth but not for meiosis. Proc. Natl. Acad. Sci. USA 86:5878–823. Banuett F, Herskowitz I. 1996. Discrete developmental stages during teliospore formation in the corn

smut fungus. Ustilago maydis. Development 122:2965–764. Basse CW. 2005. Dissecting defense-related and developmental transcriptional responses of maize during

Ustilago maydis infection and subsequent tumor formation. Plant Physiol. 138:1774–845. Basse CW, Kolb S, Kahmann R. 2002. A maize-specifically expressed gene cluster in Ustilago maydis.

Mol. Microbiol. 43:75–936. Basse CW, Stumpferl S, Kahmann R. 2000. Characterization of a Ustilago maydis gene specifically induced

during the biotrophic phase: evidence for negative as well as positive regulation. Mol. Cell. Biol. 20:329–397. Bauer R, Oberwinkler F, Vanky K. 1997. Ultrastructural markers and systematics in smut fungi and

allied taxa. Can. J. Bot. 75:1273–3148. Baumgarten AM, Suresh J, May G, Phillips RL. 2007. Mapping QTLs contributing to Ustilago maydis

resistance in specific plant tissues of maize. Theor. Appl. Genet. 114:1229–38

440 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

9. Bolker M, Bohnert HU, Braun KH, Gorl J, Kahmann R. 1995. Tagging pathogenicity genes in Ustilagomaydis by restriction enzyme-mediated integration (REMI). Mol. Gen. Genet. 248:547–52

10. Bortfeld M, Auffarth K, Kahmann R, Basse CW. 2004. The Ustilago maydis a2 mating-type locus geneslga2 and rga2 compromise pathogenicity in the absence of the mitochondrial p32 family protein Mrb1.Plant Cell 16:2233–48

11. Brachmann A, Konig J, Julius C, Feldbrugge M. 2004. A reverse genetic approach for generating genereplacement mutants in Ustilago maydis. Mol. Genet. Genomics 272:216–26

12. Brachmann A, Schirawski J, Muller P, Kahmann R. 2003. An unusual MAP kinase is required for efficientpenetration of the plant surface by Ustilago maydis. EMBO J. 22:2199–210

13. Brachmann A, Weinzierl G, Kamper J, Kahmann R. 2001. Identification of genes in the bW/bE regu-latory cascade in Ustilago maydis. Mol. Microbiol. 42:1047–63

14. Brader G, Tas E, Palva ET. 2001. Jasmonate-dependent induction of indole glucosinolates in Arabidopsisby culture filtrates of the nonspecific pathogen Erwinia carotovora. Plant Physiol. 126:849–60

15. Brefort T, Muller P, Kahmann R. 2005. The high-mobility-group domain transcription factor Rop1 isa direct regulator of prf1 in Ustilago maydis. Eukaryot. Cell 4:379–91

16. Budde AD, Leong SA. 1989. Characterization of siderophores from Ustilago maydis. Mycopathologia108:125–33

17. Castillo-Lluva S, Perez-Martin J. 2005. The induction of the mating program in the phytopathogenUstilago maydis is controlled by a G1 cyclin. Plant Cell 17:3544–60

18. Chinchilla D, Zipfel C, Robatzek S, Kemmerling B, Nurnberger T, et al. 2007. A flagellin-inducedcomplex of the receptor FLS2 and BAK1 initiates plant defence. Nature 448:497–500

19. Christensen J. 1963. Corn smut caused by Ustilago maydis. Monograph No. 2, Am. Phytopathol. Soc., St.Paul

20. De Jonge R, Thomma BPHJ. 2009. Fungal LysM effectors: extinguishers of host immunity? TrendsMicrobiol. 17:151–157

21. De Wit PJ. 2007. How plants recognize pathogens and defend themselves. Cell Mol. Life Sci. 64:2726–3222. DeZwaan TM, Carroll AM, Valent B, Sweigard JA. 1999. Magnaporthe grisea Pth11p is a novel plasma

membrane protein that mediates appressorium differentiation in response to inductive substrate cues.Plant Cell 11:2013–30

23. Di Stasio M, Brefort T, Mendoza-Mendoza A, Munch K, Kahmann R. 2009. The dual specificityphosphatase Rok1 regulates mating and pathogenicity in Ustilago maydis. Mol. Microbiol. In press

24. Doebley J. 2004. The genetics of maize evolution. Annu. Rev. Genet. 38:37–5925. Doehlemann G, Van Der Linde K, Assmann D, Schwammbach D, Hof A, et al. 2009. Pep1, a secreted

effector protein of Ustilago maydis is required for successful invasion of plant cells. PLoS Pathogens 5(2):e1000290

26. Doehlemann G, Wahl R, Horst RJ, Voll LM, Usadel B, et al. 2008. Reprogramming a maize plant:transcriptional and metabolic changes induced by the fungal biotroph Ustilago maydis. Plant J. 56:181–95

27. Doehlemann G, Wahl R, Vranes M, de Vries RP, Kamper J, Kahmann R. 2008. Establishment ofcompatibility in the Ustilago maydis/maize pathosystem. J. Plant Physiol. 165:29–40

28. D’Ovidio R, Mattei B, Roberti S, Bellincampi D. 2004. Polygalacturonases, polygalacturonase-inhibitingproteins and pectic oligomers in plant-pathogen interactions. Biochim. Biophys. Acta 1696:237–44

29. Durrenberger F, Laidlaw RD, Kronstad JW. 2001. The hgl1 gene is required for dimorphism andteliospore formation in the fungal pathogen Ustilago maydis. Mol. Microbiol. 41:337–48

30. Eichhorn H, Lessing F, Winterberg B, Schirawski J, Kamper J, et al. 2006. A ferroxidation/permeationiron uptake system is required for virulence in Ustilago maydis. Plant Cell 18:3332–45

31. Eichmann R, Schultheiss H, Kogel KH, Huckelhoven R. 2004. The barley apoptosis suppressor ho-mologue BAX inhibitor-1 compromises nonhost penetration resistance of barley to the inappropriatepathogen Blumeria graminis f. sp. tritici. Mol. Plant-Microbe Interact. 17:484–90

32. Eigenbrode SD, Espelie KE. 1995. Effects of plant epicuticular lipids on insect herbivores. Annu. Rev.Entomol. 40:171–94

33. El Gueddari NE, Rauchhaus U, Moerschbacher BM, Deising HB. 2002. Developmentally regulatedconversion of surface-exposed chitin to chitosan in cell walls of plant pathogenic fungi. New Phytol.156:103–12

www.annualreviews.org • Ustilago maydis as a Pathogen 441

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

34. Ellis JG, Dodds PN, Lawrence GJ. 2007. The role of secreted proteins in diseases of plants caused byrust, powdery mildew and smut fungi. Curr. Opin. Microbiol. 10:326–31

35. Farfsing JW, Auffarth K, Basse CW. 2005. Identification of cis-active elements in Ustilago maydis mig2promoters conferring high-level activity during pathogenic growth in maize. Mol. Plant-Microbe Interact.18:75–87

36. Fedler M, Luh K-S, Stelter K, Nieto-Jacobo F, Basse CW. 2009. The a2 mating-type locus genes lga2 andrga2 direct uniparental mtDNA inheritance and constrain mitochondrial DNA recombination duringsexual development of Ustilago maydis. Genetics 181:847–60

37. Flor-Parra I, Vranes M, Kamper J, Perez-Martin J. 2006. Biz1, a zinc finger protein required for plantinvasion by Ustilago maydis, regulates the levels of a mitotic cyclin. Plant Cell 18:2369–87

38. Fuchs U, Manns I, Steinberg G. 2005. Microtubules are dispensable for the initial pathogenic develop-ment but required for long-distance hyphal growth in the corn smut fungus Ustilago maydis. Mol. Biol.Cell 16:2746–58

39. Garcia-Muse T, Steinberg G, Perez-Martin J. 2003. Pheromone-induced G2 arrest in the phy-topathogenic fungus Ustilago maydis. Eukaryot. Cell 2:494–500

40. Garcia-Pedrajas MD, Nadal M, Bolker M, Gold SE, Perlin MH. 2008. Sending mixed signals: redun-dancy vs. uniqueness of signaling components in the plant pathogen, Ustilago maydis. Fungal Genet. Biol.45(Suppl 1):S22–30

41. Garrido E, Voss U, Muller P, Castillo-Lluva S, Kahmann R, Perez-Martin J. 2004. The induction ofsexual development and virulence in the smut fungus Ustilago maydis depends on Crk1, a novel MAPKprotein. Genes Dev. 18:3117–30

42. Glazebrook J. 2005. Contrasting mechanisms of defense against biotrophic and necrotrophic pathogens.Annu. Rev. Phytopathol. 43:205–27

43. Gohre V, Robatzek S. 2008. Breaking the barriers: microbial effector molecules subvert plant immunity.Annu. Rev. Phytopathol. 46:189–215

44. Gold SE, Brogdon SM, Mayorga ME, Kronstad JW. 1997. The Ustilago maydis regulatory subunit of acAMP-dependent protein kinase is required for gall formation in maize. Plant Cell 9:1585–94

45. Govrin EM, Rachmilevitch S, Tiwari BS, Solomon M, Levine A. 2006. An elicitor from Botrytis cinereainduces the hypersensitive response in Arabidopsis thaliana and other plants and promotes the gray molddisease. Phytopathology 96:299–307

46. Gow N, ed. 2008. Fungal Genetics and Biology. Vol. 45. St. Louis, MO: Academic Press. 94 pp.47. Hartmann HA, Kahmann R, Bolker M. 1996. The pheromone response factor coordinates filamentous

growth and pathogenicity in Ustilago maydis. EMBO J. 15:1632–4148. Hartmann HA, Kruger J, Lottspeich F, Kahmann R. 1999. Environmental signals controlling sexual

development of the corn smut fungus Ustilago maydis through the transcriptional regulator Prf1. PlantCell 11:1293–306

49. Heitman J, Kronstad JW, Taylor JW, Casselton LA, eds. 2007. Sex in Fungi: Molecular Determination andEvolutionary Implications. Washington, DC: ASM Press

50. Horst RJ, Engelsdorf T, Sonnewald U, Voll LM. 2008. Infection of maize leaves with Ustilago maydisprevents establishment of C4 photosynthesis. J. Plant Physiol. 165:19–28

51. Hu G, Kamp A, Linning R, Naik S, Bakkeren G. 2007. Complementation of Ustilago maydis MAPKmutants by a wheat leaf rust, Puccinia triticina homolog: potential for functional analyses of rust genes.Mol. Plant-Microbe Interact. 20:637–47

52. Idnurm A, Bahn YS, Nielsen K, Lin X, Fraser JA, Heitman J. 2005. Deciphering the model pathogenicfungus Cryptococcus neoformans. Nat. Rev. Microbiol. 3:753–64

53. Inada K, Morimoto Y, Arima T, Murata Y, Kamada T. 2001. The clp1 gene of the mushroom Coprinuscinereus is essential for A-regulated sexual development. Genetics 157:133–40

54. Jones JD, Dangl JL. 2006. The plant immune system. Nature 444:323–2955. Kaffarnik F, Muller P, Leibundgut M, Kahmann R, Feldbrugge M. 2003. PKA and MAPK phosphory-

lation of Prf1 allows promoter discrimination in Ustilago maydis. EMBO J. 22:5817–2656. Kahmann R, Schirawski J. 2007. Mating in the smut fungi: from a to b to the downstream cascades. See

Ref. 49, pp. 377–87

442 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

57. Kamoun S. 2006. A catalogue of the effector secretome of plant pathogenic oomycetes. Annu. Rev.Phytopathol. 44:41–60

58. Kamper J. 2004. A PCR-based system for highly efficient generation of gene replacement mutants inUstilago maydis. Mol. Genet. Genomics 271:103–10

59. Kamper J, Kahmann R, Bolker M, Ma LJ, Brefort T, et al. 2006. Insights from the genome of thebiotrophic fungal plant pathogen Ustilago maydis. Nature 444:97–101

60. Kampranis SC, Damianova R, Atallah M, Toby G, Kondi G, et al. 2000. A novel plant glutathioneS-transferase/peroxidase suppresses Bax lethality in yeast. J. Biol. Chem. 275:29207–16

61. Kemmerling B, Schwedt A, Rodriguez P, Mazzotta S, Frank M, et al. 2007. The BRI1-associated kinase1, BAK1, has a brassinolide-independent role in plant cell-death control. Curr. Biol. 17:1116–22

62. Klose J, de Sa MM, Kronstad JW. 2004. Lipid-induced filamentous growth in Ustilago maydis. Mol.Microbiol. 52:823–35

63. Klosterman SJ, Martinez-Espinoza AD, Andrews DL, Seay JR, Gold SE. 2008. Ubc2, an ortholog ofthe yeast Ste50p adaptor, possesses a basidiomycete-specific carboxy terminal extension essential forpathogenicity independent of pheromone response. Mol. Plant-Microbe. Interact. 21:110–21

64. Klosterman SJ, Perlin MH, Garcia-Pedrajas M, Covert SF, Gold SE. 2007. Genetics of morphogenesisand pathogenic development of Ustilago maydis. Adv. Genet. 57:1–47

65. Kolattukudy PE. 2001. Cutin from Plants. Weinheim, Germany: Wiley-VCH, 35 pp.66. Kruger J, Loubradou G, Wanner G, Regenfelder E, Feldbrugge M, Kahmann R. 2000. Activation of the

cAMP pathway in Ustilago maydis reduces fungal proliferation and teliospore formation in plant tumors.Mol. Plant-Microbe Interact. 13:1034–40

67. Laity C, Giasson L, Campbell R, Kronstad J. 1995. Heterozygosity at the b mating-type locus attenuatesfusion in Ustilago maydis. Curr. Genet. 27:451–59

68. Lee N, Kronstad JW. 2002. ras2 controls morphogenesis, pheromone response and pathogenicity in thefungal pathogen Ustilago maydis. Eukaryot. Cell 1:954–66

69. Leong SA, Winkelmann G. 1998. Molecular biology of iron transport in fungi. Met. Ions Biol. Syst.35:147–86

70. Linning R, Lin D, Lee N, Abdennadher M, Gaudet D, et al. 2004. Marker-based cloning of the regioncontaining the UhAvr1 avirulence gene from the basidiomycete barley pathogen Ustilago hordei. Genetics166:99–111

71. Loftus BJ, Fung E, Roncaglia P, Rowley D, Amedeo P, et al. 2005. The genome of the basidiomycetousyeast and human pathogen Cryptococcus neoformans. Science 307:1321–24

72. Martin F, Aerts A, Ahren D, Brun A, Danchin EG, et al. 2008. The genome of Laccaria bicolor providesinsights into mycorrhizal symbiosis. Nature 452:88–92

73. Martinez D, Larrondo LF, Putnam N, Gelpke MD, Huang K, et al. 2004. Genome sequence of thelignocellulose degrading fungus Phanerochaete chrysosporium strain RP78. Nat. Biotechnol. 22:695–700

74. Martinez-Espinoza AD, Garcia-Pedrajas MD, Gold SE. 2002. The Ustilaginales as plant pests and modelsystems. Fungal Genet. Biol. 35:1–20

75. Mauch F, Dudler R. 1993. Differential induction of distinct glutathione-S-transferases of wheat byxenobiotics and by pathogen attack. Plant Physiol. 102:1193–201

76. Mayorga ME, Gold SE. 2001. The ubc2 gene of Ustilago maydis encodes a putative novel adaptor proteinrequired for filamentous growth, pheromone response and virulence. Mol. Microbiol. 41:1365–79

77. Mei B, Budde AD, Leong SA. 1993. sid1, a gene initiating siderophore biosynthesis in Ustilago maydis:molecular characterization, regulation by iron, and role in phytopathogenicity. Proc. Natl. Acad. Sci. USA90:903–7

78. Mendoza-Mendoza A, Eskova A, Weise C, Czajkowski R, Kahmann R. 2009. Hap2 regulates thepheromone response transcription factor prf1 in Ustilago maydis. Mol. Microbiol. 71(4):895–911

79. Mendoza-Mendoza ABP, Djamei A, Weise C, Linne U, Marahiel M, et al. 2009. Physical-chemicalplant-derived signals induce differentiation in Ustilago maydis. Mol. Microbiol. 71:895–911

80. Molina L, Kahmann R. 2007. An Ustilago maydis gene involved in H2O2 detoxification is required forvirulence. Plant Cell 19:2293–309

81. Muller O, Kahmann R, Aguilar G, Trejo-Aguilar B, Wu A, de Vries RP. 2008. The secretome of themaize pathogen Ustilago maydis. Fungal Genet. Biol. 45(Suppl 1):S63–70

www.annualreviews.org • Ustilago maydis as a Pathogen 443

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

82. Muller O, Schreier PH, Uhrig JF. 2008. Identification and characterization of secreted and pathogenesis-related proteins in Ustilago maydis. Mol. Genet. Genomics 279:27–39

83. Muller P, Katzenberger JD, Loubradou G, Kahmann R. 2003. Guanyl exchange factor Sql2 and Ras2regulate filamentous growth in Ustilago maydis. Eukaryot. Cell 2:609–17

84. Muller P, Leibbrandt A, Teunissen H, Cubasch S, Aichinger C, Kahmann R. 2004. The Gbeta-subunit-encoding gene bpp1 controls cyclic-AMP signaling in Ustilago maydis. Eukaryot. Cell 3:806–14

85. Muller P, Weinzierl G, Brachmann A, Feldbrugge M, Kahmann R. 2003. Mating and pathogenic de-velopment of the smut fungus Ustilago maydis are regulated by one mitogen-activated protein kinase.Eukaryot. Cell. 2:1187–99

86. Munkacsi AB, Stoxen S, May G. 2007. Domestication of maize, sorghum, and sugarcane did not drivethe divergence of their smut pathogens. Evolution 61:388–403

87. Nadal M, Garcıa-Pedrajas MD, Gold SE. 2008. Dimorphism in fungal plant pathogens. FEMS Microbiol.Lett. 284:127–34

88. O’Connell RJ, Panstruga R. 2006. Tete a tete inside a plant cell: establishing compatibility betweenplants and biotrophic fungi and oomycetes. New Phytol. 171:699–718

89. Purdy RE, Kolattukudy PE. 1975. Hydrolysis of plant cuticle by plant pathogens. Purification, aminoacid composition, and molecular weight of two isozymes of cutinase and a non-specific esterase fromFusarium solani f. pisi. Biochemistry 14:2824–31

90. Reineke G, Heinze B, Schirawski J, Buettner H, Kahmann R, Basse CW. 2008. Indole-3-acetic acid (IAA)biosynthesis in the smut fungus Ustilago maydis and its relevance for increased IAA levels in infected tissueand host tumour formation. Mol. Plant Pathol. 9:339–55

91. Scherer M, Heimel K, Starke V, Kamper J. 2006. The Clp1 protein is required for clamp formation andpathogenic development of Ustilago maydis. Plant Cell 18:2388–401

92. Schirawski J, Bohnert HU, Steinberg G, Snetselaar K, Adamikowa L, Kahmann R. 2005. Endoplasmicreticulum glucosidase II is required for pathogenicity of Ustilago maydis. Plant Cell 17:3532–43

93. Snetselaar KM, Bolker M, Kahmann R. 1996. Ustilago maydis mating hyphae orient their growth towardpheromone sources. Fungal Genet. Biol. 20:299–312

94. Snetselaar KM, Mims CW. 1993. Infection of maize stigmas by Ustilago maydis: light and electronmicroscopy. Phytopathology 83:843–50

95. Snetselaar KM, Mims CW. 1994. Light and electron microscopy of Ustilago maydis hyphae in maize.Mycol. Res. 98:347–55

96. Spellig T, Bolker M, Lottspeich F, Frank RW, Kahmann R. 1994. Pheromones trigger filamentousgrowth in Ustilago maydis. EMBO J. 13:1620–27

97. Steinberg G. 2007. On the move: endosomes in fungal growth and pathogenicity. Nat. Rev. Microbiol.5:309–16

98. Steinberg G, Perez-Martin J. 2008. Ustilago maydis, a new fungal model system for cell biology. TrendsCell Biol. 18:61–67

99. Tapke VF. 1945. New physiological races of Ustilago hordei. Phytopathology 35:970–76100. Teertstra WR, Deelstra HJ, Vranes M, Bohlmann R, Kahmann R, et al. 2006. Repellents have func-

tionally replaced hydrophobins in mediating attachment to a hydrophobic surface and in formation ofhydrophobic aerial hyphae in Ustilago maydis. Microbiology 152:3607–12

101. Tucker SL, Talbot NJ. 2001. Surface attachment and pre-penetration stage development by plantpathogenic fungi. Annu. Rev. Phytopathol. 39:385–417

102. Turian G, Hamilton RH. 1960. Chemical detection of 3-indolylacetic acid in Ustilago zeae tumors.Biochim. Biophys. Acta 41:148–50

103. Tyler BM. 2009. Entering and breaking: virulence effector proteins of oomycete plant pathogens. CellMicrobiol. 11:13–20

104. Tyler BM, Tripathy S, Zhang X, Dehal P, Jiang RH, et al. 2006. Phytophthora genome sequences uncoverevolutionary origins and mechanisms of pathogenesis. Science 313:1261–66

105. Urban M, Kahmann R, Bolker M. 1996. Identification of the pheromone response element in Ustilagomaydis. Mol. Gen. Genet. 251:31–37

106. van Loon LC, Rep M, Pieterse CMJ. 2006. Significance of inducible defense-related proteins in infectedplants. Annu. Rev. Phytopathol. 44:135–62

444 Brefort et al.

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

ANRV384-PY47-18 ARI 2 July 2009 19:16

107. Wang D, Pajerowska-Mukhtar K, Culler AH, Dong X. 2007. Salicylic acid inhibits pathogen growth inplants through repression of the auxin signaling pathway. Curr. Biol. 17:1784–90

108. Wasternack C. 2007. Jasmonates: an update on biosynthesis, signal transduction and action in plant stressresponse, growth and development. Ann. Bot. (Lond.) 100:681–97

109. Wojtaszek P. 1997. Oxidative burst: an early plant response to pathogen infection. Biochem. J. 322:681–92110. Wosten HA, Bohlmann R, Eckerskorn C, Lottspeich F, Bolker M, Kahmann R. 1996. A novel class of

small amphipathic peptides affect aerial hyphal growth and surface hydrophobicity in Ustilago maydis.EMBO J. 15:4274–81

111. Xu J, Saunders CW, Hu P, Grant RA, Boekhout T, et al. 2007. Dandruff-associated Malassezia genomesreveal convergent and divergent virulence traits shared with plant and human fungal pathogens. Proc.Natl. Acad. Sci. USA 104:18730–35

112. Zarnack K, Eichhorn H, Kahmann R, Feldbrugge M. 2008. Pheromone-regulated target genes responddifferentially to MAPK phosphorylation of transcription factor Prf1. Mol. Microbiol. 69:1041–53

113. Zipfel C, Kunze G, Chinchilla D, Caniard A, Jones JD, et al. 2006. Perception of the bacterial PAMPEF-Tu by the receptor EFR restricts Agrobacterium-mediated transformation. Cell 125:749–60

114. Zipfel C, Robatzek S, Navarro L, Oakeley EJ, Jones JD, et al. 2004. Bacterial disease resistance inArabidopsis through flagellin perception. Nature 428:764–67

RELATED RESOURCES

Broad Institute Ustilago maydis database: http://www.broad.mit.edu/annotation/genome/ustilago maydis/Home.html

MUMDB: MIPS Ustilago maydis database: http://mips.gsf.de/genre/proj/ustilago/Special Ustilago maydis issue: Fungal Genetics & Biology 2008. 45 (Suppl. 1):S1–S94

www.annualreviews.org • Ustilago maydis as a Pathogen 445

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

AR384-FM ARI 14 July 2009 23:48

Annual Review ofPhytopathology

Volume 47, 2009Contents

Look Before You Leap: Memoirs of a “Cell Biological” PlantPathologistMichele C. Heath � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1

Plant Disease Diagnostic Capabilities and NetworksSally A. Miller, Fen D. Beed, and Carrie Lapaire Harmon � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �15

Diversity, Pathogenicity, and Management of Verticillium SpeciesSteven J. Klosterman, Zahi K. Atallah, Gary E. Vallad, and Krishna V. Subbarao � � � � � � �39

Bacterial/Fungal Interactions: From Pathogens to MutualisticEndosymbiontsDonald Y. Kobayashi and Jo Anne Crouch � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �63

Community Ecology of Fungal Pathogens Causing Wheat Head BlightXiangming Xu and Paul Nicholson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �83

The Biology of Viroid-Host InteractionsBiao Ding � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 105

Recent Evolution of Bacterial Pathogens: The Gall-FormingPantoea agglomerans CaseIsaac Barash and Shulamit Manulis-Sasson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 133

Fatty Acid–Derived Signals in Plant DefenseAardra Kachroo and Pradeep Kachroo � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 153

Salicylic Acid, a Multifaceted Hormone to Combat DiseaseA. Corina Vlot, D’Maris Amick Dempsey, and Daniel F. Klessig � � � � � � � � � � � � � � � � � � � � � � � � 177

RNAi and Functional Genomics in Plant Parasitic NematodesM.N. Rosso, J.T. Jones, and P. Abad � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 207

Fungal Effector ProteinsIoannis Stergiopoulos and Pierre J.G.M. de Wit � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 233

Durability of Resistance in Tomato and Pepper to XanthomonadsCausing Bacterial SpotRobert E. Stall, Jeffrey B. Jones, and Gerald V. Minsavage � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 265

v

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

AR384-FM ARI 14 July 2009 23:48

Seed Pathology Progress in Academia and IndustryGary P. Munkvold � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 285

Migratory Plant Endoparasitic Nematodes: A Group Rich in Contrastsand DivergenceMaurice Moens and Roland N. Perry � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 313

The Genomes of Root-Knot NematodesDavid McK. Bird, Valerie M. Williamson, Pierre Abad, James McCarter,

Etienne G.J. Danchin, Philippe Castagnone-Sereno, and Charles H. Opperman � � � � � 333

Viruses of Plant Pathogenic FungiSaid A. Ghabrial and Nobuhiro Suzuki � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 353

Hordeivirus Replication, Movement, and PathogenesisAndrew O. Jackson, Hyoun-Sub Lim, Jennifer Bragg, Uma Ganesan,

and Mi Yeon Lee � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 385

Ustilago maydis as a PathogenThomas Brefort, Gunther Doehlemann, Artemio Mendoza-Mendoza,

Stefanie Reissmann, Armin Djamei, and Regine Kahmann � � � � � � � � � � � � � � � � � � � � � � � � � � � � 423

Errata

An online log of corrections to Annual Review of Phytopathology articles may be found athttp://phyto.annualreviews.org/

vi Contents

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

AR384-FM ARI 14 July 2009 23:48

Related Articles

From the Annual Review of Analytical Chemistry, Volume 2 (2009)

Nanoparticle PEBBLE Sensors in Live Cells and In VivoYong-Eun Koo Lee, Ron Smith, and Raoul Kopelman

Micro- and Nanocantilever Devices and Systems for Biomolecule DetectionKyo Seon Hwang, Sang-Myung Lee, Sang Kyung Kim, Jeong Hoon Lee,

and Tae Song Kim

Applications of Aptamers as SensorsEun Jeong Cho, Joo-Woon Lee, and Andrew D. Ellington

From the Annual Review of Biochemistry, Volume 78 (2009)

The Structural and Functional Diversity of Metabolite-Binding RiboswitchesAdam Roth and Ronald R. Breaker

Sphingosine 1-Phosphate Receptor SignalingHugh Rosen, Pedro J. Gonzalez-Cabrera, M. Germana Sanna, and Steven Brown

From the Annual Review of Biomedical Engineering, Volume 11 (2009)

Fluorescent Probes for Live-Cell RNA DetectionGang Bao, Won Jong Rhee, and Andrew Tsourkas

From the Annual Review of Biophysics, Volume 38 (2009)

Comparative Enzymology and Structural Biology of RNA Self-CleavageMartha J. Fedor

From the Annual Review of Cell and Developmental Biology, Volume 24 (2008)

Auxin Receptors and Plant Development: A New Signaling ParadigmKeithanne Mockaitis and Mark Estelle

Systems Approaches to Identifying Gene Regulatory Networks in PlantsTerri A. Long, Siobhan M. Brady, and Philip N. Benfey

vii

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

AR384-FM ARI 14 July 2009 23:48

From the Annual Review of Ecology, Evolution, and Systematics, Volume 39 (2008)

Sanctions, Cooperation, and the Stability of Plant-Rhizosphere MutualismsE. Toby Kiers and R. Ford Denison

The Performance of the Endangered Species ActMark W. Schwartz

Pandora’s Box Contained Bait: The Global Problem of Introduced EarthwormsPaul F. Hendrix, Mac A. Callaham, Jr., John Drake, Ching-Yu Huang,

Sam W. James, Bruce A. Snyder, and Weixin Zhang

From the Annual Review of Entomology, Volume 54 (2009)

Adaptation and Invasiveness of Western Corn Rootworm: Intensifying Researchon a Worsening PestMichael E. Gray, Thomas W. Sappington, Nicholas J. Miller, Joachim Moeser,

and Martin O. Bohn

Impacts of Plant Symbiotic Fungi on Insect Herbivores: Mutualism in aMultitrophic ContextSue E. Hartley and Alan C. Gange

Cellular and Molecular Aspects of Rhabdovirus Interactions with Insectand Plant HostsEl-Desouky Ammar, Chi-Wei Tsai, Anna E. Whitfield, Margaret G. Redinbaugh,

and Saskia A. Hogenhout

From the Annual Review of Genetics, Volume 42 (2008)

How Saccharomyces Responds to NutrientsShadia Zaman, Soyeon Im Lippman, Xin Zhao, and James R. Broach

The Organization of the Bacterial GenomeEduardo P.C. Rocha

Genomic Insights into Marine MicroalgaeMicaela S. Parker, Thomas Mock, and E. Virginia Armbrust

From the Annual Review of Genomics and Human Genetics, Volume 9 (2008)

Phylogenetic Inference Using Whole GenomesBruce Rannala and Ziheng Yang

From the Annual Review of Microbiology, Volume 62 (2008)

Evolution of Intracellular PathogensArturo Casadevall

Chlamydiae as Symbionts in EukaryotesMatthias Horn

viii Related Articles

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

AR384-FM ARI 14 July 2009 23:48

From the Annual Review of Pharmacology and Toxicology, Volume 49 (2009)

Lipid Mediators in Health and Disease: Enzymes and Receptors as TherapeuticTargets for the Regulation of Immunity and InflammationTakao Shimizu

From the Annual Review of Plant Biology, Volume 60 (2009)

DNA Transfer from Organelles to the Nucleus: The Idiosyncratic Geneticsof EndosymbiosisTatjana Kleine, Uwe G. Maier, and Dario Leister

Jasmonate Passes Muster: A Receptor and Targets for the Defense HormoneJohn Browse

A Renaissance of Elicitors Perception of Microbe-Associated Molecular Patternsand Danger Signals by Pattern-Recognition ReceptsThomas Boller and Georg Felix

Biosynthesis of Plant Isoprenoids: Perspectives for Microbial EngineeringJames Kirby and Jay D. Keasling

Roles of Plant Small RNAs in Biotic Stress ResponsesVirginia Ruiz-Ferrer and Olivier Voinnet

Related Articles ix

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.

AR384-FM ARI 14 July 2009 23:48

Annual Reviews is a nonprofit scientific publisher established to promote the advancement of thesciences. Beginning in 1932 with the Annual Review of Biochemistry, the Company has pursued as itsprincipal function the publication of high-quality, reasonably priced Annual Review volumes. The volumesare organized by Editors and Editorial Committees who invite qualified authors to contribute criticalarticles reviewing significant developments within each major discipline. The Editor-in-Chief invites thoseinterested in serving as future Editorial Committee members to communicate directly with him. AnnualReviews is administered by a Board of Directors, whose members serve without compensation.

2009 Board of Directors, Annual Reviews

Richard N. Zare, Chairman of Annual Reviews, Marguerite Blake Wilbur Professor of Chemistry,Stanford University

John I. Brauman, J.G. Jackson–C.J. Wood Professor of Chemistry, Stanford UniversityPeter F. Carpenter, Founder, Mission and Values Institute, Atherton, CaliforniaKaren S. Cook, Chair of Department of Sociology and Ray Lyman Wilbur Professor of Sociology,

Stanford UniversitySandra M. Faber, Professor of Astronomy and Astronomer at Lick Observatory, University of California

at Santa CruzSusan T. Fiske, Professor of Psychology, Princeton UniversityEugene Garfield, Publisher, The ScientistSamuel Gubins, President and Editor-in-Chief, Annual ReviewsSteven E. Hyman, Provost, Harvard UniversitySharon R. Long, Professor of Biological Sciences, Stanford UniversityJ. Boyce Nute, Palo Alto, CaliforniaMichael E. Peskin, Professor of Theoretical Physics, Stanford Linear Accelerator CenterHarriet A. Zuckerman, Vice President, The Andrew W. Mellon Foundation

Management of Annual Reviews

Samuel Gubins, President and Editor-in-ChiefRichard L. Burke, Director for ProductionPaul J. Calvi Jr., Director of Information TechnologySteven J. Castro, Chief Financial Officer and Director of Marketing & SalesJeanne M. Kunz, Human Resources Manager and Secretary to the Board

Annual Reviews of

Analytical ChemistryAnthropologyAstronomy and AstrophysicsBiochemistryBiomedical EngineeringBiophysicsCell and Developmental BiologyClinical PsychologyEarth and Planetary SciencesEcology, Evolution, and

SystematicsEconomicsEntomologyEnvironment and ResourcesFinancial EconomicsFluid Mechanics

GeneticsGenomics and Human

GeneticsImmunologyLaw and Social ScienceMarine ScienceMaterials ResearchMedicineMicrobiologyNeuroscienceNuclear and Particle

ScienceNutritionPathology: Mechanisms of

DiseasePharmacology and Toxicology

Physical ChemistryPhysiologyPhytopathologyPlant BiologyPolitical SciencePsychologyPublic HealthResource EconomicsSociology

SPECIAL PUBLICATIONSExcitement and Fascination of

Science, Vols. 1, 2, 3, and 4

Ann

u. R

ev. P

hyto

path

ol. 2

009.

47:4

23-4

45. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by W

ashi

ngto

n St

ate

Uni

vers

ity o

n 02

/23/

10. F

or p

erso

nal u

se o

nly.