Lectures 19 and 20: Mycorrhizas – Reading: Text, Chapter 17. ALSO great website from CSIRO in Austalia http://www.ffp.csiro.au/research/mycorrhiza/index.html http://mycorrhizas.info/index.html Rodriguez R and Redman R. 2008. More than 400 years of evolution and some plants still canʼt make it on their own. Journal of Experimental Botany 59: 1109-1114. Parniske M. 2008. Arbuscular mycorrhiza: the mother of plant root endosymbioses. Nature Reviews Microbiology 6: 763-773. Stinson, K.A. et al. 2006. Invasive plant suppresses the growth of native tree seedlings by disrupting belowground mutualisms. PLOS Biology Objectives: Understand the importance of mycorrhizas. Know how to recognize ectomycorrhizas, and the types of endomycorrhizas. Know how to prove that a mycorrhizal symbiosis has been formed (differentiating between what you expect with ecto- vs. endomycorrhizas). Know which groups of fungi (phyla, orders, families) form ectomycorrhizas, which form endomycorrhizas. Keywords: ectomycorrhiza, mantle, Hartig net, endomycorrhiza, vesicle, arbuscule, spore, Glomeromycota. Study questions: 1) What are the soil and root-associated structures found with VAM? With orchid mycorrhizas? Compare and contrast the structures of VAM, orchid mycorrhizas, and ECMs. Which fungal phyla form each of the three types of mycorrhizas that we have discussed? 2) Are VAM common in plant roots? Explain. 3) VAM appears to be a balanced mutualism – is this true? 4) You wish to prove that Russula mississaugii is a mycorrhizal fungus. How could you convince our class? Use your naked eye, a dissecting microscope, a compound microscope, and radio-isotope study of tree seedlings to prove your point! What are the controls? 5) What are the benefits to the plant and to the ecosystem of mycorrhizas? 6) Why do plant species that depend on mycorrhizas tend to lack root hairs?

7) Garlic mustard is very common around Mississauga and on this campus. Despite the wet weather this year, mushroom biomass and diversity was not high on campus. Also, young tree seedlings may not be competing well to restore the tree canopy as older trees die. Explain a possible role for garlic mustard and speculate on additional factors behind these observations. 8) What are myco-heterotrophs and why are they “cheaters”? 9) What do we mean when we say that mycorrhizas are “dynamic” and form “nutrient networks”? 10) What is Parniskeʼs theory about VAM as the “mother of plant root endosymbioses”? There are two types of evidence – what are they?

Invasive Plant Suppresses the Growthof Native Tree Seedlings by DisruptingBelowground MutualismsKristina A. Stinson

1, Stuart A. Campbell

2, Jeff R. Powell

2, Benjamin E. Wolfe

2, Ragan M. Callaway

3, Giles C. Thelen

3,

Steven G. Hallett4

, Daniel Prati5

, John N. Klironomos2*

1 Harvard Forest, Harvard University, Petersham, Massachusetts, United States of America, 2 Department of Integrative Biology, University of Guelph, Guelph, Ontario,

Canada, 3 Division of Biological Sciences, University of Montana, Missoula, Montana, United States of America, 4 Department of Botany and Plant Pathology, Purdue

University, West Lafayette, Indiana, United States of America, 5 Department of Community Ecology, UFZ Centre for Environmental Research, Halle, Germany

The impact of exotic species on native organisms is widely acknowledged, but poorly understood. Very few studieshave empirically investigated how invading plants may alter delicate ecological interactions among resident species inthe invaded range. We present novel evidence that antifungal phytochemistry of the invasive plant, Alliaria petiolata, aEuropean invader of North American forests, suppresses native plant growth by disrupting mutualistic associationsbetween native canopy tree seedlings and belowground arbuscular mycorrhizal fungi. Our results elucidate an indirectmechanism by which invasive plants can impact native flora, and may help explain how this plant successfully invadesrelatively undisturbed forest habitat.

Citation: Stinson KA, Campbell SA, Powell JR, Wolfe BE, Callaway RM, et al. (2006) Invasive plant suppresses the growth of native tree seedlings by disrupting belowgroundmutualisms. PLoS Biol 4(5): e140. DOI: 10.1371/journal.pbio.0040140

Introduction

Widespread anthropogenic dispersal of exotic organismshas raised growing concern over their devastating ecologicalimpacts, and has prompted decades of research on theecology of invasive species [1–3]. Exotic plants may becomeaggressive invaders outside their home ranges for a numberof reasons, including release from native, specialized antag-onists [4], higher relative performance in a new site [5], directchemical (allelopathic) interference with native plant per-formance [6], and variability in the responses and resistanceof native systems to invasion [7,8]. Thus, successful invasion inmany cases appears to involve the fact that invasive speciesare not at equilibrium, and are either freed of long-standingbiotic interactions with their enemies in the home range, and/or disrupt interactions among the suite of native organismsthey encounter in a new range [9]. Nevertheless, experimentaldata on species-level impacts of exotic plants are still limited[10]. One particularly understudied area is the potential forinvasive plants to disrupt existing ecological associationswithin native communities [6,10]. Many exotic and nativeplants alike depend upon mutualisms with native insects,birds, or mammals for pollination and seed dispersal [11], andwith soil microbes for symbiotic nutrient exchange [12]. Thus,when an introduced species encounters a new suite ofresident organisms, it is likely to alter closely interlinkedecological relationships, many of which have co-evolvedwithin native systems [6,11].

One such relationship is that between plants and mycor-rhizal fungi [12]. Most vascular plants form mycorrhizalassociations with arbuscular mycorrhizal fungi (AMF) [12],and many plants are highly dependent on this association fortheir growth and survival [12], particularly woody perennialsand others found in late-successional communities [13]. Incontrast, many weedy plants, in particular non-mycotrophic

plants, can be negatively affected by AMF [14–16]. Natural-ized exotic plants have been found to be poorer hosts anddepend less on native AMF than native plants [17]. They oftencolonize areas that have been disturbed [2], and disturbancesto soil have been shown to negatively impact AMF function-ing [18]. Furthermore, it has been proposed that theproliferation of plants with low mycorrhizal dependencymay degrade AMF densities in the soil [17]. However, a fewinvasive plants proliferate in the understory of maturetemperate forests [2], where AMF density is typically high[19]. The existing mycelial network in mature forest soils mayfacilitate the establishment of exotic, mycorrhizal-dependent,recruits [20,21], but this should not be the case for non-mycorrhizal invaders. If non-mycorrhizal invasive plantsestablish and degrade AMF in mature forests, then the effectson certain resident native plants could be substantial.One of the most problematic invaders of mesic temperate

forests in North America is Alliaria petiolata (garlic mustard;Brassicaceae), a non-mycorrhizal, shade-tolerant, Eurasianbiennial herb which, like most other mustards, primarilyoccupies disturbed areas. Garlic mustard is abundant inforest edges, semishaded floodplains, and other disturbedsites in its home range [22]. However, this species has recentlybecome an aggressive and widespread invader of both

Academic Editor: Michel Loreau, McGill University, Canada

Received December 5, 2005; Accepted March 1, 2006; Published April 25, 2006

DOI: 10.1371/journal.pbio.0040140

Copyright: � 2006 Stinson et al. This is an open-access article distributed underthe terms of the Creative Commons Attribution License, which permits unrestricteduse, distribution, and reproduction in any medium, provided the original authorand source are credited.

Abbreviations: AMF, arbuscular mycorrhizal fungi; ANOVA, analysis of variance;REGW, Ryan-Einot-Gabriel-Welsch

* To whom correspondence should be addressed. E-mail: jklirono@uoguelph.ca

PLoS Biology | www.plosbiology.org May 2006 | Volume 4 | Issue 5 | e1400727

PLoS BIOLOGY

disturbed areas and closed-canopy forest understory acrossmuch of the United States and Canada [23], where itapparently suppresses native understory plants, includingthe seedlings of dominant canopy trees [22,24]. The mecha-nism underlying garlic mustard’s unusual capacity to enterand proliferate within intact North American forest com-munity has not yet been established.

As shown in recent greenhouse experiments, garlicmustard’s impact on native understory flora may involvecompetitive [25] or allelopathic effects on native plants [26],but it has also been hypothesized that this species interfereswith plant–AMF interactions in its invaded range [27].Members of the Brassicaceae, including garlic mustard,produce various combinations of glucosinolate products[28], organic plant chemicals with known anti-herbivore,anti-pathogenic and allelopathic [29] properties, that mayalso prevent this non-mycorrhizal plant family from associat-ing with AMF [30]. These phytochemicals may be releasedinto soils as root exudates, as a result of damaged root tissue,or in the form of leaf litter. High densities of garlic mustardin the field correlate with low inoculum potential of AMF,and extracts of garlic mustard leaves have been shown toreduce the germination of AMF spores and impair AMFcolonization of cultivated tomato roots in laboratory settings[27]. Although not all Brassicaceae are invasive, it is possiblethat garlic mustard’s successful invasion of understoryhabitats involves the negative effects of its phytochemistryon the native plant and AMF species it encounters outside itshome range. Others have shown that exotic plants can recruitdifferent suites of microbial organisms in their new rangesthat can be antagonistic to native plants [6]. However, to ourknowledge, no previous studies have directly tested whetherthis species or any other exotic plant disrupts native plant–AMF mutualisms within natural communities. Here, wepresent novel evidence that garlic mustard negatively impactsthe growth of AMF-dependent forest tree seedlings by itsdisruption of native mycorrhizal mutualisms. We furthershow that, because seedlings of dominant tree species inmature forest communities are more highly dependent onAMF than plants that typically dominate earlier successionalcommunities, garlic mustard invasion may disproportionatelydamage mature forests relative to other habitats.

Results/Discussion

We first tested whether native tree seedlings were less ableto form mycorrhizal associations when grown in forestunderstory soils with a history of garlic mustard invasionthan when grown in soils that had not experienced invasions(Experiment 1). We found that dominant native hardwoodtree species of northeastern temperate forests, Acer saccharum(sugar maple), Ac. rubrum (red maple), and Faxinus americana(white ash), showed significantly less AMF colonization ofroots (Figure 1A) and slower growth (Figure 1B) when grownin soil that had been invaded by garlic mustard. AMFcolonization was almost undetectable in soil that had beeninvaded by garlic mustard. These reductions were similar tothose observed when seedlings were grown in sterilized soilfrom both garlic mustard–invaded and garlic mustard–freesites (Figure 1B), strongly suggesting that the mechanism bywhich garlic mustard suppresses the growth of native tree

species is microbially-mediated, and not the result of soildifferences or direct allelopathy.We then conducted additional experiments to confirm that

garlic mustard specifically caused AMF decline in the nativesoils (Experiment 2–4). We grew seedlings of the same threenative tree species used in Experiment 1 in uninvaded forestsoils that were conditioned for 3 mo with either garlicmustard plants or with one of the three native tree species.All three tree species demonstrated significantly lower AMFcolonization in soils conditioned by Al. petiolata (0%–10%)than in soils conditioned by the native plants (20%–65%;Figure 2A). AMF colonization was similar in unconditioned(control) soils and soils conditioned with native plants. Inaddition, growth of the tree seedlings was the lowest in soilsconditioned by garlic mustard (Figure 2B), confirming thatgarlic mustard plants reduce native plant performance byinterfering with the formation of mycorrhizal associations.We investigated whether there is a phytochemical basis to

garlic mustard’s observed antifungal effects on AMF inExperiments 3–4. In an earlier study, Vaughn and Berhow[31] isolated the phytotoxic glucosinolate hydrolysis productsallyl isothiocyanate, benzyl isothiocyanate, and glucotropaeo-lin from extracts of Al. petiolata root tissues and foundevidence for their allelopathic effects on certain plants in theabsence of mycorrhizas. These phytochemicals could havedirect effects on plant growth through allelopathy as well asindirect effects via disruption of AMF. To experimentallyestablish that garlic mustard’s effect on AMF is phytochemi-cally based, we grew native tree seedlings on uninvaded soilsto which we added individual aqueous extracts of garlic

Figure 1. Experiment 1

The influence of field soils that were invaded or uninvaded by Al.petiolata (6 sterilized) on (A) mycorrhizal colonization (Fsugar maple¼ 77.7,df¼3,39, p , 0.001; Fred maple¼60.5, df¼3,39, p , 0.001; and Fwhite ash¼116.6, df¼ 3,39, p , 0.001) and (B) biomass accumulation (Fsugar maple¼57.8, df¼ 3,39, p , 0.001; Fred maple¼61.4, df¼3,39, p , 0.001; and Fwhite

ash ¼ 70.1, df ¼ 3,39, p , 0.001) of native tree seedlings. Bars representthe mean and standard error.DOI: 10.1371/journal.pbio.0040140.g001

PLoS Biology | www.plosbiology.org May 2006 | Volume 4 | Issue 5 | e1400728

Invasive Plant Disrupts Mycorrhizas

mustard or each of the native trees species (Experiment 3).We found that garlic mustard extract was just as effective asthe living plant at reducing AMF colonization (Figure 3A) andgrowth (Figure 3B) of the native plants. Moreover, exposingAMF spores to extract of garlic mustard severely andsignificantly reduced germination rates of those spores(Experiment 4; Figure 3C). Collectively, our results clearlydemonstrate that garlic mustard, probably through phyto-chemical inhibition, disrupts the formation of mycorrhizalassociations. Our results thus reveal a powerful, indirectmechanism by which an invasive species can suppress thegrowth of native flora.

Because plants vary in their dependency on AMF [32],garlic mustard’s disruption of native plant–fungal mutualismsshould not inhibit the growth of all plants equally, but rathershould correlate strongly with the mycorrhizal dependence ofspecies encountered in the invaded range. Specifically,courser root production, which impedes the nutrient uptakeof typically slow-growing, woody plants such as tree seedlings,may explain the stronger AMF dependency of certain species[19,33]. To test whether garlic mustard’s effects correlate withAMF dependency, and whether garlic mustard has strongernegative effects on forest tree seedlings than on other plants,we conducted another experiment (Experiment 5) using 16plant species for which we determined AMF-dependency bycomputing the difference in plant growth in the presence andabsence of AMF. We then tested the impact of garlic mustardon the AM fungal colonization and growth of each plantspecies as above. All 16 plants were successfully colonized byAMF, and the presence of garlic mustard heavily reducedAMF colonization in all plants (Figure 4A). However, thepresence of garlic mustard had a much stronger effect onplants that had high mycorrhizal dependency than those withless dependency (Figure 4B). The strongest effects wereobserved for woody species most typically found in forestedsites. These results indicate that the invasion of garlicmustard is more likely to negatively impact highly mycor-

rhizal-dependent tree seedlings than less-mycorrhizal-de-pendent plants. Thus, garlic mustard’s successfulcolonization of understory habitat may be attributed in partto its ability to indirectly suppress woody competitors, and itseffect on the native flora may be more detrimental in intactforests than disturbed sites. In addition, the data suggest thatinvasion by garlic mustard may have profound effects on thecomposition of mature forest communities (e.g., by repres-sing the regeneration of dominant canopy trees, and byfavoring plants with low mycorrhizal dependency such asweedy herbs).In conclusion, our results reveal a novel mechanism by

which an invasive plant can disrupt native communities: byvirtually eliminating the activity of native AMF from the soiland drastically impairing the growth of native canopy species.It is currently unclear precisely which phytochemicalsproduced by garlic mustard have the observed antifungalproperties, whether and how they interact with other soilmicrobes, and whether these anti-fungal effects extend toother functionally important forest soil fungi such asectomycorrhizal fungi and saprotrophic fungi. In addition,within the home range, it is not known if evolutionary naturalresistance of co-occurring European neighbors may bufferthe effects of garlic mustard’s antifungal properties [34–36].Further research in these directions is needed to betterunderstand the effects of this invader on natural ecosystemsand the mechanisms involved. In North America; however,the disruption of native tree seedling–AMF mutualisms mayfacilitate garlic mustard’s invasion into mature forest under-

Figure 2. Experiment 2

The effect of soils conditioned with garlic mustard Al. petiolata (gm),sugar maple (sm), red maple (rm), or white ash (wa) on (A) mycorrhizalcolonization (Fsugar maple¼31.2, df¼4,49, p , 0.001; Fred maple¼18.2, df¼4,49, p , 0.001; and Fwhite ash¼22.1, df¼4,49, p , 0.001) and (B) increasein biomass (Fsugar maple¼ 15.1, df¼ 4,49, p , 0.001; Fred maple¼ 18.1, df¼4,49, p , 0.001; and Fwhite ash¼ 13.2, df¼ 4,49, p , 0.001) of native treeseedlings. Bars represent the mean and standard error.DOI: 10.1371/journal.pbio.0040140.g002

Figure 3. Experiments 3 and 4

The effects of extract of garlic mustard (gm), sugar maple (sm), red maple(rm), white ash (wa), or a water control on (A) mycorrhizal colonization ofnative tree seedlings (Fsugar maple¼ 20.3, df¼ 4,49, p , 0.001; Fred maple¼19.8, df ¼ 4,49, p , 0.001; and Fwhite ash ¼ 25.4, df ¼ 4,49, p , 0.001[Experiment 3]), (B) increase in biomass of native tree seedlings (Fsugar

maple¼ 11.7, df¼ 4,49, p , 0.001; Fred maple¼ 14.2, df¼ 4,49, p , 0.001;and Fwhite ash¼ 27.9, df¼ 4,49, p , 0.001 [Experiment 3]), and (C) percentgermination of native AMF spores (FGlomus ¼ 17.3, df ¼ 4,49, p , 0.001;and FAcaulospora ¼ 21.8, df ¼ 4,49, p , 0.001 [Experiment 4]). Barsrepresent the mean and standard error.DOI: 10.1371/journal.pbio.0040140.g003

PLoS Biology | www.plosbiology.org May 2006 | Volume 4 | Issue 5 | e1400729

Invasive Plant Disrupts Mycorrhizas

story and have particularly negative effects on the growth,survival, and recruitment of native trees, and the compositionof forest communities.

Materials and Methods

Experiment 1. Using a 15-cm–wide corer, we collected soil fromgarlic mustard–invaded and nearby garlic mustard–free locations ateach of five forested areas dominated by Acer rubrum L. (red maple),Ac. saccharum Marsh. (sugar maple), Fraxinus americana L. (white ash),and Fagus grandifolia Ehrh. (American beech) near Waterloo, Ontario,Canada. Invaded and uninvaded sites were randomly chosen within a40-m2 plot within each forested area. Soils from the invaded anduninvaded areas were pooled separately in the lab and screened toremove coarse roots and debris. Half the soil from each pool was thensterilized by autoclaving at 120 8C to create four soil treatments: (1)soil with a history of garlic mustard, (2) sterile soil with a history of

garlic mustard, (3) soil without a history of garlic mustard, and (4)sterile soil without a history of garlic mustard. Six-inch pots werefilled with a 1:1 mixture of sterilized silica sand and one of the foursoil types. To each pot, we added a single seedling (seeds germinatedon Turface [Aimcor, Buffalo Grove, Illinois, United States], a claysubstrate) of one of the three native overstory tree species (sugarmaple, red maple, or white ash) in a complete 4 3 3 factorial designwith ten replicates of each treatment combination. The initial wetbiomass of each seedling was recorded prior to planting, and dryweights were estimated using a dry–wet regression calculated fromtwenty extra seedlings. Pots were randomly placed on a greenhousebench. Plants were watered (400 ml) once per week. Fertilizer was notadded. After 4 mo of growth, shoots and roots were harvested, driedat 60 8C for 48 h, and weighed to determine biomass. Anapproximately 1-g subsample of roots from each seedling wasextracted, stained with Chlorazol Black E [37] and analyzed forpercent colonization by AMF [38]. Biomass and percent colonizationdata were analyzed using analysis of variance (ANOVA) for two fixedeffects (soil type and species) and their interaction, followed by theRyan-Einot-Gabriel-Welsch (REGW) multiple-range test.

Experiment 2. Using field soil without a history of garlic mustardinvasion (see Experiment 1), we grew garlic mustard, sugar maple, redmaple, and white ash seedlings in separate 6-in pots (n ¼ 10) tocondition the soil to each plant species. After 3 mo of conditioning,shoots and roots were removed. Unconditioned soil served as acontrol to the four plant-conditioning treatments. We added a singleseedling of each of the three tree species to each of the five soiltreatments. Pots were randomly placed on a greenhouse bench. Plantswere watered (400 ml) once per week, without fertilizer. After 4 mo ofgrowth, plants were harvested, biomass was determined, and percentmycorrhizal colonization of roots was assessed as in Experiment 1.Data were analyzed using ANOVA for two fixed effects (species andsoil condition treatment). Means from the three species were pooled,and the effect of conditioning treatment was tested with a single-factor ANOVA followed by the REGW multiple-range test.

Experiment 3. To 6-in pots containing field soil without a historyof garlic mustard (see Experiment 1), we added a one-time, 100-mlaqueous extract [27] of whole plants of either garlic mustard, sugarmaple, red maple, or white ash. A water control was included to givefive treatments. Whole-plant extract was used to account forsecondary compounds exuded through roots and leaf litter. After 1wk of exposure to the extract, seedlings of each tree species wereplanted in each of these five treatments to give a full factorial design(extract source 3 tree species) with ten replicates of each treatmentcombination. Plants were watered (40 ml) every week, withoutfertilizer. After 4 mo of growth, plants were harvested, biomass wasdetermined, and roots were assayed for mycorrhizal colonization asin Experiment 1. Data were analyzed by two-factor ANOVA.

Experiment 4. Spores from AMF native to the forest sites wereobtained using trap cultures (as described in [39], but with a mix ofnative plants) of soil samples from the uninvaded locations. Wevisually collected and separated Glomus and Acaulospora spores fromthese cultures, and compared germination rates of each genus in fivetreatments: a water agar control and water agar amended with anaqueous extract from each of the four plants, as above. Ten randomlydrawn spores were added into each plate, which was then incubatedat 18 8C for 10 d. Ten replicate plates were prepared for each of theten treatment combinations (two AMF genera 3 five extracts). Plateswere monitored microscopically for spore germination. Percentgermination data were analyzed using ANOVA for two fixed effects(extract source and AMF genus), and because of a significantinteraction, each AMF genus was then analyzed separately usingsingle-factor ANOVA followed by the REGW multiple-range test.

Experiment 5. We investigated the effects of garlic mustard onwoody and herbaceous plants using the following 16 native plantspecies: Cichorium intybus, Trifolium repens, Plantago major, and Tarax-acum officinale (dominant herbaceous colonizers of disturbed edgesand bare ground); Solidago canadensis, Chrysanthemum leucanthemum,Daucus carota, and Asclepias syriaca (dominant herbaceous edge and gapspecies); Juniperus virginiana, Populus deltoides, Morus alba, and Prunusvirginiana (dominant woody colonizers of forest edges and gaps); andFr. americana, Ac. saccharum, Ac. rubrum, and Pr. serotina (dominant treespecies of mature forest). Seedlings of each plant were transplantedinto 8-in pots. For each species, growth was compared under thefollowing soil treatments: (1) soil without a history of garlic mustardand inoculated with AMF, (2) soil without a history of garlic mustard,without AMF, and (3) soil with a history of garlic mustard, andinoculated with AMF. Experimental soil was collected within amature-canopy maple forest from locations with and without garlicmustard. Soils from each location type were then mixed, cleaned of

Figure 4. Experiment 5

(A) Effect of mycorrhizal dependency on Al. petiolata reduction of AMFcolonization.(B) Effect of mycorrhizal dependency on Al. petiolata reduction in plantgrowth. Mycorrhizal dependency was calculated separately as thedifference between plant growth in the presence and absence of AMF.Different colors represent plants with different life-history strategies, asfollows: yellow dot, herbaceous colonizers of disturbed edges and bareground; reddish brown dot, herbaceous edge and gap species; blue dot,woody colonizers of forest edges and gaps; black dot, tree species ofmature forest. Species are labeled as follows (with mean mycorrhizalcolonization in soil not conditioned by garlic mustard 6 standard errorin parentheses): 1¼Ci. intybus (18.5 6 4.1), 2¼ Tr. repens (46.7 6 6.3), 3¼Pl. major (28.2 6 3.7), 4 ¼ Ta. officinale (37.3 6 2.5), 5 ¼ S. canadensis(48.0 6 6.2), 6¼ C. leucanthemum (34.6 6 3.1), 7¼D. carota (40.4 6 6.2),8¼ As. syriaca (52.1 6 5.8), 9¼ J. virginiana (31.2 6 4.4), 10¼ Po. deltoids(63.9 6 4.5), 11¼M. alba (38.6 6 5.9), 12¼ Pr. virginiana (28.4 6 4.2), 13¼ Fr. americana (65.9 6 5.3), 14 ¼ Ac. saccharum (46.3 6 3.7), 15 ¼ Ac.rubrum (59.5 6 5.7), 16 ¼ Pr. serotina (34.8 6 5.5).DOI: 10.1371/journal.pbio.0040140.g004

PLoS Biology | www.plosbiology.org May 2006 | Volume 4 | Issue 5 | e1400730

Invasive Plant Disrupts Mycorrhizas

all coarse roots and debris, autoclaved, and added to the pots as a 1:1mix of soil and silica sand. AMF spores were extracted from field soilcollected from sites representing the four different habitats, andpooled. The AMF-inoculation treatment consisted of adding 200randomly picked spores to each pot, 2 cm below the surface, andbeneath the newly transplanted seedlings. Plants were watered (500ml) once per week, without fertilizer. They were harvested after 4 moof growth, dried at 60 8C for 36 h, and weighed to determine biomass.AMF dependency of each plant species was determined by computingthe difference in plant growth in the presence and absence of AMF,i.e., contrast of treatments (1) and (2) [32]. The effects of garlicmustard on plant growth and percent colonization of each plant weredetermined by contrasting treatments (1) and (3). To ask whether anyrelationships existed among mycorrhizal dependency, life form, andgarlic mustard effects, we performed two regressions: percentreduction in AMF colonization by garlic mustard on AMF depend-ency and percent reduction in plant biomass by garlic mustard onAMF dependency.

Acknowledgments

We thank T. Denich, V. Grebogi, G. Herrin, P. Hudson, G. Kuenen, J.Lozi, B. Shelton, P. Stephens, J. Van Houten, and Z. Zhu for technicalassistance, and P. Antunes, G. De Deyn, and M. Hart for helpfulcomments on the text.

Author contributions. KAS, RMC, and JNK conceived and designedthe experiments. KAS and JNK performed the experiments. KAS,SAC, JRP, BEW, RMC, GCT, SGH, DP, and JNK analyzed the data. JNKcontributed reagents/materials/analysis tools. All authors wrote thepaper.

Funding.We thank the Natural Sciences and Engineering ResearchCouncil of Canada, and the Harvard University Bullard Foundationfor financial support.

Competing interests. The authors have declared that no competinginterests exist. &

References1. Rejmanek M (2000) Invasive plants: Approaches and predictions. Austral

Ecol 25: 497–506.2. Mooney HA, Hobbs RJ (2000) Invasive species in a changing world.

Washington (D. C.): Island Press. 457 p.3. Ewel JJ, O’Dowd DJ, Bergelson J, Daehler CC, D’Antonio CM, et al. (1999)

Deliberate introductions of species: Research needs. Bioscience 49: 619–630.

4. Mitchell CE, Power AG (2003) Release of invasive plants from viral andfungal pathogens. Nature 421: 625–627.

5. Thebaud C, Simberloff D (2001) Are plants really larger in their introducedranges? Am Nat 157: 231–236.

6. Callaway RM, Ridenour WM (2004) Novel weapons: Invasive success and theevolution of increased competitive ability. Front Ecol Environ 2: 436–443.

7. Hobbs RJ, Huenneke LF (1992) Disturbance, diversity, and invasion:Implications for conservations. Conserv Biol 6: 324–337.

8. Levine JM, D’Antonio CM (1999) Elton revisited: A review of evidencelinking diversity and invasibility. Oikos 87: 15–26.

9. Keane RM, Crawley MJ (2002) Exotic plant invasions and the enemy releasehypothesis. Trends Ecol Evol 17: 164–170.

10. Levine JM, Vila M, D’Antonio CM, Dukes JS, Grigulis K, et al. (2003)Mechanisms underlying the impacts of exotic plant invasions. Proc Biol Sci270: 775–781.

11. Richardson DM, Allsopp N, D’Antonio CM, Milton SJ, Rejmanek M (2000)Plant invasions—The role of mutualisms. Biol Rev Camb Philos Soc. 75: 65–93.

12. Smith SE, Read DJ (1997) Mycorrhizal symbiosis. 2nd edition. New York:Academic Press. 605 p.

13. Janos DP (1980) Mycorrhizae influence tropical succession. Biotropica 12:56–64.

14. Grime JP, Mackey JML, Hillier SH, Read DJ (1987) Floristic diversity in amodel system using experimental microcosms. Nature 328: 420–422.

15. Francis R, Read DJ (1995) Mutualism and antagonism in the mycorrhizalsymbiosis, with special reference to impacts on plant community structure.Can J Bot 73: 1301–1309.

16. Van der Heijden MGA, Klironomos JN, Ursic M, Moutoglis P, Streitwolf-Engel R, et al. (1998) Mycorrhizal fungal diversity determines plantbiodiversity, ecosystem variability and productivity. Nature 396: 69–72.

17. Vogelsang KM, Bever JD, Griswold M, Schultz PA (2004 June) The use ofmycorrhizal fungi in erosion control applications. Final Report forCaltrans. Sacramento (California): California Department of Transporta-tion Contract No. 65A0070. 150 p.

18. Haselwandter K (1997) Soil micro-organisms, mycorrhiza, and restorationecology. In: Urbanska KM, Webb NR, Edwards PJ, editors. Restorationecology and sustainable development. Cambridge: Cambridge UniversityPress. pp. 65–80.

19. Read DJ (1991) Mycorrhizas in ecosystems–Nature’s response to the ‘Law ofthe minimum.’ In: Hawksworth DL. Frontiers in mycology. Wallingford(United Kingdom): CAB International. pp. 101–130.

20. Marler MM, Zabinski CA, Callaway RM. (1999) Mycorrhizae indirectly

enhance competitive effects of an invasive forb on a native bunchgrass.Ecology 80: 1180–1186.

21. Van der Heijden MGA (2004) Arbuscular mycorrhizal fungi as supportsystems for seedling establishment in grassland. Ecol Lett 7: 293–303.

22. Nuzzo V (1999) Invasion pattern of the herb garlic mustard (Alliariapetiolata) in high quality forests. Biol Invasions 1: 169–179.

23. Nuzzo V (2000) Element stewardship abstract for Alliaria petiolata. Arlington(Virginia): The Nature Conservancy. Available: http://tncweeds.ucdavis.edu/esadocs/documnts/allipet.html. Accessed 16 March 2006.

24. Blossey B, Nuzzo V, Hinz H, Gerber E. (2001) Developing biological controlof Alliaria petiolata (M. Bieb.) Cavara and Grande (garlic mustard). Nat AreasJ 21: 357–367.

25. Meekins JF, McCarthy BC (1999) Competitive ability of Alliaria petiolata(garlic mustard, Brassicaceae), an invasive, nonindigenous forest herb. Int JPlant Sci 160: 743–752.

26. Prati D, Bossdorf O (2004) Allelopathic inhibition of germination byAlliaria petiolata (Brassicaceae). Am J Bot 91: 285–288.

27. Roberts KJ, Anderson RC (2001) Effect of garlic mustard [Alliaria petiolata(Beib. Cavara & Grande)] extracts on plants and arbuscular mycorrhizal(AM) fungi. Am Midl Nat 146: 146–152.

28. Renwick JAA (2002) The chemical world of crucivores: Lures, treats andtraps. Entomol Exp Appl 104: 35–42.

29. Siemens DS, Garner S, Mitchell-Olds T, Callaway RM (2002) Cost of defensein the context of plant competition: Brassica rapa may grow and defend.Ecology 83: 505–517.

30. Schreiner RP, Koide RT (1993) Mustards, mustard oils and mycorrhizas.New Phytol 123: 107–113.

31. Vaughn SF, Berhow MA (1999) Allelochemicals isolated from tissues of theinvasive weed garlic mustard (Alliaria petiolata). J Chem Ecol 25: 2495–2504.

32. Klironomos JN (2003) Variation in plant response to native and exoticarbuscular mycorrhizal fungi. Ecology 84: 2292–2301.

33. Newsham KK, Fitter AH, Watkinson AR (1995) Multifunctionality andbiodiversity in arbuscular mycorrhizas. Trends Ecol Evol 10: 407–411.

34. Reinhart KO, Packer A, Van der Putten WH, Clay K (2003) Plant-soil biotainteractions and spatial distribution of black cherry in its native andinvasive ranges. Ecol Lett 6: 1046–1050.

35. Callaway RM, Thelen C, Rodriguez A, Holben WE (2004) Soil biota andexotic plant invasion. Nature 427: 731–733.

36. Callaway RM, Thelen GC, Barth S, Ramsey PW, Gannon JE (2004) Soil fungialter interactions between the invader Centaurea maculosa and NorthAmerican natives. Ecology 85: 1062–1071.

37. Brundrett MC, Piche Y, Peterson RL (1984) A new method for observingthe morphology of vesicular-arbuscular mycorrhizae. Can J Bot 62: 2128–2134.

38. McGonigle TP, Miller MH, Evans DG, Fairchild GL, Swan JA (1990) A newmethod which gives an objective measure of colonization of roots byvesicular arbuscular mycorrhizal fungi. New Phytol 115: 495–501.

39. Klironomos JN, Allen MF, Rillig MC, Piotrowski J, Makvandi-Nejad S, et al.(2005) Abrupt rise in atmospheric CO2 overestimates community responsein a model plant-soil system. Nature 433: 621–624.

PLoS Biology | www.plosbiology.org May 2006 | Volume 4 | Issue 5 | e1400731

Invasive Plant Disrupts Mycorrhizas

Journal of Experimental Botany, Vol. 59, No. 5, pp. 1109–1114, 2008

doi:10.1093/jxb/erm342 Advance Access publication 10 February, 2008

SPECIAL ISSUE REVIEW PAPER

More than 400 million years of evolution and some plantsstill can’t make it on their own: plant stress tolerancevia fungal symbiosis

Rusty Rodriguez1,2,* and Regina Redman2,3

1 US Geological Survey, Seattle, WA 98115, USA2 University of Washington, Seattle, WA, USA3 Montana State University, Bozeman, MT, USA

Received 19 June 2007; Revised 25 November 2007; Accepted 30 November 2007

Abstract

All plants in natural ecosystems are thought to be

symbiotic with mycorrhizal and/or endophytic fungi.

Collectively, these fungi express different symbiotic

lifestyles ranging from parasitism to mutualism. Analy-

sis of Colletotrichum species indicates that individual

isolates can express either parasitic or mutualistic

lifestyles depending on the host genotype colonized.

The endophyte colonization pattern and lifestyle ex-

pression indicate that plants can be discerned as either

disease, non-disease, or non-hosts. Fitness benefits

conferred by fungi expressing mutualistic lifestyles

include biotic and abiotic stress tolerance, growth

enhancement, and increased reproductive success.

Analysis of plant–endophyte associations in high stress

habitats revealed that at least some fungal endophytes

confer habitat-specific stress tolerance to host plants.

Without the habitat-adapted fungal endophytes, the

plants are unable to survive in their native habitats.

Moreover, the endophytes have a broad host range

encompassing both monocots and eudicots, and confer

habitat-specific stress tolerance to both plant groups.

Key words: Colletotrichum, fungal endophytes, stress

tolerance, symbiosis, symbiotic lifestyle.

Introduction

Throughout evolutionary time plants have been con-fronted with various abiotic and biotic stresses. Lackingany form of locomotion, plants have depended on seed

dispersal, vegetative growth, and complex physiologyeither to escape or to mitigate the impacts of stress. Allplants are known to perceive and transmit signals, and

respond to stress such as drought, heat, salinity, and disease

(Bohnert et al., 1995; Bartels and Sunkar, 2005). Some

biochemical processes are common to all plant stress

responses including the production of osmolytes, altering

water movement, and scavenging reactive oxygen species

(ROS) (Leone et al., 2003; Maggio et al., 2003; Tuberosa

et al., 2003). Although there has been extensive research

in plant stress responses, it is not known why so few

species are able to colonize high stress habitats. However,

plant stress research rarely takes into consideration

a ubiquitous aspect of plant biology—fungal symbiosis.Since the first description of symbiosis (De Bary, 1879),

several symbiotic lifestyles have been defined based on

fitness benefits to or impacts on host and symbiont

(Lewis, 1985). After >100 years of research it is reason-

able to conclude that most, if not all, multicellular life on

earth is symbiotic with micro-organisms. For example, all

plants in natural ecosystems are thought to be symbiotic

with mycorrhizal and/or endophytic fungi (Petrini, 1996;

Brundrett, 2006). Recent studies indicate that fitness

benefits conferred by mutualistic fungi contribute to or

are responsible for plant adaptation to stress (Read, 1999;

Stone et al., 2000; Rodriguez et al., 2004). Collectively,

mutualistic fungi may confer tolerance to drought, metals,

disease, heat, and herbivory, and/or promote growth and

nutrient acquisition. It has become clear that at least some

plants are unable to endure habitat-imposed abiotic and

biotic stresses in the absence of fungal endophytes

(Redman et al., 2002b). Since there are several excellent

* To whom correspondence should be addressed. E-mail: rustyrod@u.washington.edu

ª The Author [2008]. Published by Oxford University Press [on behalf of the Society for Experimental Biology]. All rights reserved.For Permissions, please e-mail: journals.permissions@oxfordjournals.org

at University of T

oronto Library on Novem

ber 19, 2010jxb.oxfordjournals.org

Dow

nloaded from

reviews on mycorrhizal and endophytic fungi (Carroll,1988; Read, 1999; Stone et al., 2000; Schardl andLeuchtmann, 2005; Brundrett, 2006), the focus of thisdiscussion will be on two aspects of fungal endophytebiology: symbiotic lifestyle switching (Redman et al.,2001) and the recently observed ecological phenomenonhabitat-adapted symbiosis (HA-symbiosis; Rodriguezet al., 2008). It is hypothesized that HA-symbiosis allowsplants to establish in high stress habitats.

Fungal endophytes

Unlike mycorrhizal fungi, endophytes reside entirelywithin host tissues and emerge during host senescence.These fungi comprise a phylogentically diverse group thatare members of the dikarya (Carroll, 1988; Schardl andLeuchtmann, 2005; Van Bael et al., 2005; Girlanda et al.,2006; Arnold and Lutzoni, 2007). While most endophytesbelong to the Ascomycota clade, some belong to theBasidiomycota. Although these fungi are often groupedtogether, they can be discriminated into different func-tional groups just as has been done with mycorrhizal fungi(Brundrett, 2006). Currently, endophytes can be sub-divided into four classes based on host range, colonizationpattern, transmission, and ecological function (Rodriguezet al., in review). Nevertheless, endophytes have beenshown to confer fitness benefits to host plants includingtolerance to herbivory, heat, salt, disease, and drought,and increased below- and above-ground biomass (Baconand Hill, 1996; Clay and Holah, 1999; Sahay and Varma,1999; Redman et al., 2001, 2002b; Arnold et al., 2003;Waller et al., 2005; Marquez et al., 2007).

The symbiotic continuum, lifestyle switching,and host range

Collectively, fungi express several different symbioticlifestyles that are defined by fitness benefits to plant hostsand symbionts (Lewis, 1985). The range of symbioticlifestyle expression from mutualism to parasitism has beendescribed as the symbiotic continuum (Carroll, 1988;Johnson et al., 1997; Saikkonen et al., 1998; Schulzet al., 1999; Schardl and Leuchtmann, 2005). Within eachgroup of fungal symbionts there are isolates and/or speciesthat span the symbiotic continuum by expressing differentlifestyles. For example, the endophyte genus Epichloecomprises species that express either mutualistic orparasitic lifestyles (Schardl and Leuchtmann, 2005).Several studies that focused on the isolation of endophytesfrom asymptomatic plant tissues indicate that individualspecies express either mutualistic, commensal, or parasiticlifestyles when re-inoculated back on the original hostspecies (Schulz et al., 1999). This indicates that bothmutualists and pathogens infect plants and remain quies-

cent until plant senescence. This represents an excellentecological strategy for fungi to capitalize on plantnutrients. By already being established in tissues, endo-phytes have immediate access to plant nutrients madeavailable during plant senescence.

Studies on host genotype versus symbiotic lifestyleexpression revealed that individual isolates of some fungalspecies could span the symbiotic continuum by expressingeither mutualistic or pathogenic lifestyles in different hostplants (Redman et al., 2001). For example, Colletotrichumspecies are classified as virulent pathogens, yet severalspecies can express mutualistic lifestyles in non-diseasehosts (Table 1). Mutualistic benefits conferred by Colleto-trichum spp. include disease resistance, growth enhance-ment, and/or drought tolerance (Redman et al., 2001).Although the genetic basis of symbiotic communication isnot yet known, subtle differences in host genomes haveprofound effects on the outcome of symbiotic interactions.For example, commercially grown tomato (Solanumlycopersicum) is known to possess relatively few geneticdifferences between varieties yet is able to express highlevels of phenotypic plasticity (Miller and Tanksley, 1990;Tanksley, 2004; Brewer et al., 2007). When C. magna isintroduced into different tomato cultivars, the fungus mayexpress either mutualistic, commensal, or parasitic life-styles. While parasitic and mutualistic lifestyles are easilyobserved, commensal lifestyles are often designated whenno host fitness benefit is observed. However, dependingon the traits being assessed, the commensal designationmay be misleading. For example, C. gloeosporioides wasdesignated a pathogen of strawberry and a commensal oftomato because it conferred no disease protection(Redman et al., 2001). However, C. gloeosporioides

Table 1. Symbolic lifestyle expression of Colletotrichum speciesversus plant host

Fungalpathogen

Diseasehosta

Non-diseasehostb

Lifestyle expressed

Diseasestressc

Droughtstressd

C. magna Watermelon Tomato Mutualism MutualismC. musae Banana Pepper Mutualism MutualismC. orbiculare Cucumber Tomato Mutualism MutualismC. acutatum Strawberry Watermelon Commensalism MutualismC. gloeosporioides Strawberry Watermelon Commensalism Mutualism

a Species were isolated from disease lesions on the indicated hostplants.

b Host plants that are asymptomatically colonized by the respectiveColletotrichum spp.

c Symbiotic lifestyle expressed after asymptomatic colonization.Lifestyles were defined by the ability of each Colletotrichum sp. toconfer disease resistance against virulent Colletotrichum pathogens ofthe non-disease hosts (data from Redman et al., 2001).

d Symbiotic lifestyle expressed after asymptomatic colonization.Lifestyles were defined by the ability of each Colletotrichum sp. toconfer drought tolerance based on the length of time before wilting aftercessation of watering (data from Redman et al., 2001).

1110 Rodriguez and Redman

at University of T

oronto Library on Novem

ber 19, 2010jxb.oxfordjournals.org

Dow

nloaded from

increased plant biomass and conferred drought toleranceto tomato plants, and was therefore designated a mutualist.

A series of experiments were performed to characterizethe genetic basis of fungal symbiotic lifestyles. UVmutagenesis of a virulent isolate (CmL2.5) of C. magnaresulted in the isolation of a non-pathogenic mutant (Path-1)that was able to colonize host plants asymptomatically(Freeman and Rodriguez, 1993). Path-1 conferred severalfitness benefits to hosts, including disease and droughtresistance, and growth enhancement. Based on thesefitness benefits, it was concluded that Path-1 was express-ing a mutualistic lifestyle in host plants. Additionalstudies involving gene disruption by restriction enzyme-mediated integration (REMI) with a selectable plasmidresulted in the generation of non-pathogenic mutants thatdiffered in the ability to confer disease resistance (Fig. 1;Redman et al., 1999). The UV and REMI mutants lost theability to switch between lifestyles and were ‘locked’ intoone lifestyle (either mutualism, intermediate-mutualism, orcommensal). These results indicate that the ability toswitch between symbiotic lifestyles, at least in thisspecies, is controlled by single genetic loci.

Although the original experiments on lifestyle switchingwere performed with Colletotrichum species known to bepathogenic, similar results have been observed with otherendophytes from plants in natural habitats (RY Rodriguez,unpublished results). What does this mean with regard tohost specificity? It appears that there are non-hosts thata fungus is unable to infect and two types of hosts thatfungi can colonize: disease hosts that they parasitize andnon-disease hosts that they asymptomatically colonize.

Colonization of non-disease hosts by pathogenic Colleto-trichum species is asymptomatic and there are no observ-able differences between colonized and uncolonized plantsin the absence of stress, unless the endophyte promotesplant growth (Redman et al., 1999, 2002a). Conventionalviews suggest that pathogens either cause disease or inducehost defence systems which terminate the infection process.When Colletotrichum species express mutualistic lifestylesand confer disease resistance, host defence systems are notactivated unless the symbiotic plants are challenged witha virulent pathogen (Redman et al., 1999, 2002a). Oncechallenged, the host defence systems activate very rapidly(<24 h) to maximal levels (Redman et al., 1999).

The ability to switch lifestyles brings up someinteresting questions:

(i) Is there an evolutionary direction to symbiotic life-styles? Clavicipitaceous endophytes expressing mutu-alisms are hypothesized to have evolved directionallyfrom pathogenic ancestors (Schardl and Leuchtmann,2005). The situation with at least some other endo-phytes appears to be quite different, where theevolution of symbiotic lifestyle appears to lack specificdirectionality (Arnold and Lutzoni, 2007). Endophytesthat can switch lifestyles may represent evolutionarytransitions or simply fungi that have achieved a higherdegree of ecological flexibility to ensure optimalgrowth and reproduction in a variety of hosts.

(ii) Do plants inadvertently participate or possibly in-stigate disease processes? Individual fungal isolatescan equally colonize different plants irrespective ofthe symbiotic lifestyle they express. For C. magna toexpress mutualism in one tomato cultivar and parasit-ism in another suggests that disease may reflectmiscommunication rather than aggressive pathogenicity.

Symbiosis and stress tolerance

There are numerous reports of fungal symbionts confer-ring tolerance to stress to host plants, including herbivory,drought, heat, salt, metals, and disease (Bacon and Hill,1996; Clay and Holah, 1999; Sahay and Varma, 1999;Redman et al., 2001, 2002b; Arnold et al., 2003; Walleret al., 2005; Marquez et al., 2007; Rodriguez et al., 2007).It is interesting that the stress tolerance conferred by someendophytes involves habitat-specific fungal adaptations.For example, within the geothermal soils of YellowstoneNational Park, a small number of plant species reside. Oneplant species (Dichanthelium lanuginosum) has beenstudied and found to be colonized by one dominantendophyte (Curvularia protuberata). Curvularia protu-berata confers heat tolerance to the host plant, and neitherthe fungus nor the plant can survive separate from oneanother when exposed to heat stress >38 �C (Redman

Restriction Enzyme Mediated Integration

14,400 Transformants screened on plants

176 nonpathogenic REMI mutants

Four phenotypes elucidated based on ability to colonize and confer diseaseresistance

A100

80 - 100 Mutualist

B100

20 - 65 IM

C

1000

Commensal

REMI Mutant Class

Symbiotic Lifestyle

D00

Abortive

Fig. 1. Gene disruption (restriction enzyme-mediated integration,REMI) of fungal symbiotic lifestyle loci in Colletotrichum magna.Symbiotic lifestyles reflect the ability of REMI mutants to colonize hostplants (watermelon) asymptomatically and confer disease resistanceagainst the virulent wild type. REMI mutants were designated either asmutualists, intermediate mutualists, or commensals based on diseaseprotection, or abortive if they were unable to colonize host tissues.Although these lifestyle designations reflect quantitative differences,they probably reflect a continuum of symbiotic lifestyles representedamong the mutants. Methods and data are from Redman et al. (1999a).

Plant–fungal symbiosis 1111

at University of T

oronto Library on Novem

ber 19, 2010jxb.oxfordjournals.org

Dow

nloaded from

et al., 2002b). The ability of the endophyte to confer heattolerance requires the presence of a fungal RNA virus(Marquez et al., 2007). While the genetic/biochemical roleof the virus in symbiotically conferred heat tolerance isnot known, it is surmised that the virus is providingbiochemical functionality to the fungus and it is not thevirus that directly confers heat tolerance. A comparison ofC. protuberata isolates from geothermal and non-geothermal plants revealed that the ability to confer heattolerance was specific to isolates from geothermal plants(Rodriguez et al., 2008). Therefore, the ability to conferheat tolerance is a habitat-adapted phenomenon.

Another example of habitat-specific fungal adaptationinvolves a native dunegrass (Leymus mollis) on coastalbeaches of Puget Sound, WA (Rodriguez et al., 2008).Leymus mollis is colonized by one dominant fungalendophyte (Fusarium culmorum) that can be isolated fromabove- and below-ground tissues and seed coats. Fusa-rium culmorum confers salt tolerance to the host plantwhich cannot survive in coastal habitats without thehabitat-adapted endophyte. A comparison of F. culmorumisolates from L. mollis and a non-coastal plant revealedthat the ability to confer salt tolerance was specific toisolates from the coastal plants, indicating that the abilityto confer salt tolerance is a habitat-adapted phenomenon(Rodriguez et al., 2008).

A comparison of C. protuberata, F. culmorum, andC. magna isolates further supports habitat-specific adapta-tion of endophytes: C. protuberata confers heat but notdisease or salt tolerance; F. culmorum confers salt but notheat or disease tolerance; and C. magna confers diseasebut not heat or salt tolerance (Rodriguez et al., 2008).These symbiotically conferred stress tolerances conformto the evolutionary dynamics that must play out in thedifferent habitats, with fungi adapting to habitat-specificstresses and conferring stress tolerance to host plants. Thishabitat-specific adaptation is defined as HA-symbiosis,and it is hypothesized that this allows plants to establishand survive in high stress habitats.

Biochemical basis of endophyte-conferredstress tolerance

It is fascinating that after 400 million years of evolutionthere are plants that require symbiotic associations forstress tolerance. There has been an enormous researcheffort in plant stress physiology that is described inseveral excellent books and reviews. Although previousstudies have elucidated how plants respond to stress, theyrarely consider symbiotic contributions.

Symbiotically conferred disease tolerance appears toinvolve different mechanisms depending on the endo-phyte. For example, the ability of a non-pathogenicColletotrichum mutant (Path-1 that expresses a mutualism)to confer disease resistance is correlated to a rapid and

strong activation of biochemical processes known toconfer resistance (Redman et al., 1999). In the absence ofpathogen challenge, Path-1-colonized plants do not appearto activate host defence systems. However, when Path-1-colonized watermelon and cucumber seedlings wereexposed to a virulent pathogen, peroxidase and phenyl-alanine ammonia lyase activity and lignin depositionincreased within 24 h to levels that non-symbiotic plantsnever achieved (Table 2; Redman et al., 1999). Interest-ingly, Colletotrichum-conferred disease resistance is local-ized to tissues that the fungus has colonized, and is notsystemic. The results suggest that the endophyte may beacting as a type of biological trigger that activates hostdefence systems. The fact that Colletotrichum spp.expressing non-pathogenic lifestyles do not activate hostdefence in the absence of pathogen challenge may beviewed as either suppression of host defences or eludinghost recognition. However, the dynamics of host defenceactivation suggest that the endophytes are recognized anddo not suppress defence systems.

In barley, the root endophyte Piriformospora indicaconfers disease resistance by a different mechanism.Symbiotic plants are thought to resist necrotrophic rootpathogens due to increased activity of glutathione–ascorbate antioxidant systems (Waller et al., 2005).Unlike Colletotrichum endophytes, disease resistanceconferred by P. indica appears to be systemic. It is notclear if P. indica increases antioxidation systems in theabsence of pathogens or if other aspects of hostphysiology are involved in resistance.

The differences between Colletotrichum spp.- andP. indica-conferred disease resistance may indicate thata greater diversity of mechanisms may yet be elucidated.Regardless, these results warrant a more comprehensiveanalysis of endophyte-conferred disease resistance.

Table 2. Physiological defence activity versus symboticallyconferred disease conferred disease resistance by Colletotri-chum magna

Methods and physiological data are from Redman et al. (1999).

Host Peroxidaseactivitya

PALactivityb

Lignindepositionc

24 h 48 h 24 h 48 h 24 h 48 h

Watermelon (E–)d 2.76 3.46 2.27 2.90 – +Watermelon (E+)e 5.77 6.30 2.50 3.70 +++ ++++Cucumber (E–) 0.63 1.31 0.02 0.25 – +Cucumber (E+) 1.80 2.34 .27 0.34 +++ ++++

a Activity based on a guaiacol/H2O2 assay, and units indicate changein A470 min�1 lg�1 protein.

b Activity based on the production of cinnamic acid, and unitsindicate change in A290 min�1 lg�1 protein.

c Qualitative assessment of the absence (–) or presence (+) of ligninvisualized with acidic phloroglucinol.

d (E–)¼endophyte (C. magna) free.e (E+)¼endophyte (C. magna) colonized.

1112 Rodriguez and Redman

at University of T

oronto Library on Novem

ber 19, 2010jxb.oxfordjournals.org

Dow

nloaded from

Symbiotic plasticity and fungal taxonomy

One deficiency in species designations is a dearth offunctional ecological descriptions, symbiotic lifestylepotential, and host ranges. A good example of this issueis the fact that C. protuberata is described as a plantpathogen of several monocots (Farr et al., 1989). Yet,C. protuberata isolate Cp4666D is a mutualist in Dichan-thelium lanuginosum, conferring heat and drought toler-ance (Rodriguez et al., 2008). While Curvularia speciesare not known to have broad disease-host ranges,C. protuberata from the monocot D. lanuginosum isa mutualist (confers heat tolerance) in the eudicot tomato,and isolates from non-geothermal plants do not confer heattolerance (Marquez et al., 2007; Rodriguez et al., 2008).A similar scenario occurs with F. culmorum. Designatedas a virulent plant pathogen, F. culmorum causes diseaseon a variety of crop plants (Farr et al., 1989). However,the F. culmorum isolate FcRed1 from dunegrass is amutualist in dunegrass and tomato conferring salt toler-ance, and isolates from non-coastal plants do not confersalt tolerance (Rodriguez et al., 2008). These examplesindicate that the current concept of a fungus beingcategorized as either a pathogen, saprophyte, or mutualistis inadequate to address the fact that individual species canrepresent significant ecological plasticity.

The ability of ‘pathogenic’ Colletotrichum species toswitch symbiotic lifestyles and express mutualisms pro-vides insight into why these species are so ubiquitous. Ithas been suggested previously that pathogens may bepresent in non-disease host plants constituting potentialinocula for disease. In fact, C. acutatum asymptomaticallycolonizes pepper, eggplant, bean, and tomato plants,which can subsequently provide inoculum for diseaseoutbreaks in strawberry plants (Freeman et al., 2001). So,at least in this genus, species may move freely betweenlifestyles and hosts, thereby expanding bio-geographicdistribution. It is unlikely that this phenomenon is specificto Colletotrichum as asymptomatic colonization of hostshas been reported for other genera such as Fusarium(Bacon and Yates, 2006).

Incorporating information on lifestyle expression andecological functionalities would allow ecologists to un-derstand better the role of fungi in ecosystem processes,geneticists to understand better genome differences be-tween isolates, and mycologists to understand phenotypicand ecological plasticity.

Symbiotic communities

While this discussion has focused on fungal symbionts, itis important to point out that plants represent communitiesof fungi, bacteria, viruses, and/or algae. All of thesemicro-organisms contribute to the outcome of symbiosisand hence increase the complexity of studying plant

biology. Moreover, fungal symbionts may also harbourbacteria and viruses that can have dramatic effects onsymbiotic communication. For example, the class2 endophyte C. protuberata (Cp4666D), originally iso-lated from plants growing in geothermal soils, containsa double-stranded RNA (dsRNA) virus that is required forsymbiotically conferred heat tolerance (Marquez et al.,2007). In the absence of the virus, Cp4666D asymptom-atically colonizes plants but confers no heat tolerance.Therefore, a three-way symbiosis (a virus in a fungus ina plant) is required for thermal tolerance. This was anunexpected result and reflects our limited understanding ofsymbiotic systems and how they function. More impor-tantly, it indicates the need to study plants froma symbiotic systems perspective to elucidate the contribu-tions of all symbionts.

Summary

Both laboratory and field studies have demonstrated thatat least some plant species in natural habitats requirefungal endophytes for stress tolerance and survival. Sincecolonizing land ;400 million years ago, plants haveevolved intragenomic mechanisms to perceive and trans-mit signals, and respond to stress (Bohnert et al., 1995;Bartels and Sunkar, 2005), but most plants lack theadaptive capability to mitigate the impacts of stress(Alpert, 2000). At least some plants depend on inter-genomic epigenetic processes provided by symbiotic fungifor stress adaptation. The observations described in thismanuscript raise some fundamental questions in plantbiology. Why have plants in high stress habitats notevolved intragenomic capabilities for stress adaptation?Can plants adapt to stress without symbiotic involvement?Why are so few plants adapted to high stresshabitats? Answers to these questions will require extensiveresearch efforts over the coming decades and are necessarybefore ecosystem processes are fully understood.

Acknowledgements

This project was made possible by the permission, assistance, andguidelines of YNP and the UW Cedar Rocks Biological Preserve.This work was supported by the US Geological Survey, NSF(0414463) and US/IS BARD (3260-01C).

References

Alpert P. 2000. The discovery, scope, and puzzle of desiccationtolerance in plants. Plant Ecology 151, 5–17.

Arnold AE, Lutzoni F. 2007. Diversity and host range of foliarfungal endophytes: are tropical leaves biodiversity hotspots?Ecology 88, 541–549.

Arnold EA, Mejia LC, Kyllo D, Rojas E, Maynard Z,Robbins N, Herre EA. 2003. Fungal endophytes limit pathogen

Plant–fungal symbiosis 1113

at University of T

oronto Library on Novem

ber 19, 2010jxb.oxfordjournals.org

Dow

nloaded from

damage in a tropical tree. Proceedings of the National Academyof Sciences, USA 100, 15649–15654.

Bacon CW, Hill NS. 1996. Symptomless grass endophytes:products of coevolutionary symbioses and their role in theecological adaptations of grasses. In: Redkin SC, Carris LM, eds.Endophytic fungi in grasses and woody plants. St Paul, MN: APSPress, 155–178.

Bacon CW, Yates IE. 2006. Endophytic root colonization byfusarium species: histology, plant interactions, and toxicity. In:Schulz BJE, Boyle CJC, Sieber TN, eds. Microbial rootendophytes. Berlin: Springer-Verlag, 133–152.

Bartels D, Sunkar R. 2005. Drought and salt tolerance in plants.Critical Reviews in Plant Science 24, 23–58.

Bohnert HJ, Nelson DE, Jensen RG. 1995. Adaptations toenvironmental stresses. The Plant Cell 7, 1099–1111.

Brewer MT, Moyseenko1 JB, Monforte AJ, van der Knaap E.2007. Morphological variation in tomato: a comprehensive studyof quantitative trait loci controlling fruit shape and development.Journal of Experimental Botany 58, 1339–1349.

Brundrett MC. 2006. Understanding the roles of multifunctionalmycorrhizal and endophytic fungi. In: Schulz BJE, Boyle CJC,Sieber TN, eds. Microbial root endophytes. Berlin: Springer-Verlag, 281–293.

Carroll G. 1988. Fungal endophytes in stems and leaves: fromlatent pathogen to mutualistic symbiont. Ecology 69, 2–9.

Clay K, Holah J. 1999. Fungal endophyte symbiosis and plantdiversity in successional fields. Science 285, 1742–1745.

De Bary A. 1879. Die erschenung symbiose. In: Trubner KJ, ed.Vortrag auf der versammlung der naturforscher und artze zucassel. Strassburg: 1–30.

Farr DF, Bills GF, Chamuris GP, Rossman AY. 1989. Fungi onplants and plant products in the United States. St Paul, MN: APSPress.

Freeman S, Horowitz S, Sharon A. 2001. Pathogenic andnonpathogenic lifestyles in Colletotrichum acutatum from straw-berry and other plants. Phytopathology 91, 986–992.

Freeman S, Rodriguez RJ. 1993. Genetic conversion of a fungalplant pathogen to a nonpathogenic, endophytic mutualist. Science260, 75–78.

Girlanda M, Perotto S, Luppi AM. 2006. Molecular diversity andecological roles of mycorrrhiza-associated sterile fungal endo-phytes in mediterranean ecosysems. In: Boyle CJC, Sieber TN,eds. Microbial root endophytes. Berlin: Springer-Verlag,207–226.

Johnson NC, Graham JH, Smith FA. 1997. Functioning ofmycorrhizal associations along the mutualism–parasitism contin-uum. New Phytologist 135, 575–586.

Leone A, Perrotta C, Maresca B. 2003. Plant tolerance to heatstress: current strategies and new emergent insight. In: di ToppiLS, Pawlik-Skowronska B, eds. Abiotic stresses in plants.London: Kluwer Academic Publishers, 1–22.

Lewis DH. 1985. Symbiosis and mutualism: crisp concepts andsoggy semantics. In: Boucher DH, ed. The biology of mutualism.London: Croom Helm Ltd, 29–39.

Maggio A, Bressan RA, Ruggiero C, Xiong L, Grillo S. 2003.Salt tolerance: placing advances in molecular genetics intoa physiological and agronomic context. In: di Toppi LS, Pawlik-Skowronska B, eds. Abiotic stresses in plants. London: KluwerAcademic Publishers, 53–70.

Marquez LM, Redman RS, Rodriguez RJ, Roossinck MJ. 2007.A virus in a fungus in a plant—three way symbiosis required forthermal tolerance. Science 315, 513–515.

Miller JC, Tanksley SD. 1990. Rflp analysis of phylogeneticrelationships and geneteic variation in the genus Lycopersicon.Theoretical and Applied Genetics 80, 437–448.

Petrini O. 1996. Ecological and physiological aspects of host-specificity in endophytic fungi. In: Redlin SC, Carris LM, eds.Endophytic fungi in grasses and woody plants. St Paul, MN: APSPress, 87–100.

Read DJ. 1999. Mycorrhiza—the state of the art. In: Varma A,Hock B, eds. Mycorrhiza. Berlin: Springer-Verlag, 3–34.

Redman RS, Dunigan DD, Rodriguez RJ. 2001. Fungal symbio-sis: from mutualism to parasitism, who controls the outcome, hostor invader? New Phytologist 151, 705–716.

Redman RS, Freeman S, Clifton DR, Morrel J, Brown G,Rodriguez RJ. 1999. Biochemical analysis of plant protectionafforded by a nonpathogenic endophytic mutant of Colletotri-chum magna. Plant Physiology 119, 795–804.

Redman RS, Rossinck MR, Maher S, Andrews QC,Schneider WL, Rodriguez RJ. 2002a. Field performance ofcucurbit and tomato plants infected with a nonpathogenic mutantof Colletotrichum magna (teleomorph: Glomerella magna; Jen-kins and Winstead). Symbiosis 32, 55–70.

Redman RS, Sheehan KB, Stout RG, Rodriguez RJ,Henson JM. 2002b. Thermotolerance conferred to plant host andfungal endophyte during mutualistic symbiosis. Science298, 1581.

Rodriguez RJ, Henson J, Van Volkenburgh E, Hoy M,Wright L, Beckwith F, Kim Y, Redman RS. 2008. Stresstolerance in plants via habitat-adapted symbiosis. InternationalSociety of Microbial Ecology in press.

Rodriguez RJ, Redman RS, Henson JM. 2004. The role of fungalsymbioses in the adaptation of plants to high stress environments.Mitigation and Adaptation Strategies for Global Change 9,261–272.

Sahay NS, Varma A. 1999. Piriformospora indica: a new bi-ological hardening tool for micropropagated plants. FEMSMicrobiology Letters 181, 297–302.

Saikkonen K, Faeth SH, Helander M, Sullivan TJ. 1998. Fungalendophytes: a continuum of interactions with host plants. AnnualReview of Ecology and Systematics 29, 319–343.

Schardl C, Leuchtmann A. 2005. The epichloe endophytes ofgrasses and the symbiotic continuum. In: Dighton J, White JF,Oudemans P, eds. The fungal community: its organization androle in the ecosystem. Boca Raton, FL: Taylor & Francis,475–503.

Schulz B, Rommert AK, Dammann U, Aust HJ, Strack D. 1999.The endophyte–host interaction: a balanced antagonism? Myco-logical Research 10, 1275–1283.

Stone JK, Bacon CW, White JF. 2000. An overview ofendophytic microbes: endophytism defined. In: Bacon CW, WhiteJF, eds. Microbial endophytes. New York: Marcel Dekker, Inc,3–30.

Tanksley SD. 2004. The genetic, development, and molecular basesof fruit size and shape variation in tomato. The Plant Cell 16,S181–S189.

Tuberosa R, Grillo S, Ellis RP. 2003. Unravelling the geneticbasis of drought tolerance in crops. In: di Toppi LS, Pawlik-Skowronska B, eds. Abiotic stresses in plants. London: KluwerAcademic Publishers, 71–122.

Van Bael SA, Maynard Z, Rojas E, Mejia LC, Kyllo DA,Herre EA, Robbins N, Bischoff JF, Arnold AE. 2005.Emerging perspectives on the ecological roles of endophyticfungi in tropical plants. In: Dighton J, White JF, Oudemans P,eds. The fungal community: its organization and role in theecosystem. Boca Raton, FL: Taylor & Francis, 505–518.

Waller F, Achatz B, Baltruschat H, et al. 2005. The endophyticfungus Piriformospora indica reprograms barley to salt-stresstolerance, disease resistance, and higher yield. Proceedings of theNational Academy of Sciences, USA 102, 13386–13391.

1114 Rodriguez and Redman

at University of T

oronto Library on Novem

ber 19, 2010jxb.oxfordjournals.org

Dow

nloaded from

Plant root symbioses with fungi occur in several dif-ferent forms and are referred to as mycorrhiza (from the Greek ‘mycos’, meaning fungus and ‘rhiza’, mean-ing root). In ectomycorrhiza, which is predominant on trees in temperate forests, the fungal partner remains outside of plant cells, whereas in endomycorrhiza, including orchid, ericoid and arbuscular mycorrhiza (AM), part of the fungal hyphae is inside. AM is probably the most widespread terrestrial symbiosis1 and is formed by 70–90% of land plant species2 with fungi that belong to a monophyletic phylum, the Glomeromycota3,4 (FIG. 1). Symbiotic development results in the formation of tree-shaped subcellular structures within plant cells. These structures, which are known as arbuscules (from the Latin ‘arbusculum’, meaning bush or little tree) are thought to be the main site of nutrient exchange between the fungal and plant symbiotic partners (FIG. 2). AM intimately connects plants to the hyphal network of the fungi, which can be in excess of 100 metres of hyphae per cubic centimetre of soil5. This hyphal network is specialized for nutri-ent (predominantly phosphate) and water uptake6. In return for supplying plants with nutrients and water, AM fungi obtain carbohydrates from plants7,8. Up to 20% of the photosynthesis products of terrestrial plants (approximately 5 billion tonnes of carbon per year) are

estimated to be consumed by AM fungi9. Therefore, AM symbiosis contributes significantly to global phosphate and carbon cycling and influences primary productivity in terrestrial ecosystems1. The beneficial effects of AM are most apparent under conditions of limited nutrient availability. Although the underly-ing regulatory mechanisms are not understood, the amount of root colonization typically decreases when nutrients are in abundance. Interestingly, the colo-nization of roots with AM fungi has been observed to lead to an inhibition of bacterial leaf pathogens10. Whether such increased resilience to pathogens is a consequence of improved plant fitness or is due to specific defence responses that are induced by AM fungi is unknown.

AM fungi are unusual organisms because of their age, lifestyle and genetic make-up; they have existed for more than 400 million years morphologically unaltered and could therefore qualify as living fossils. They are considered by many to be ancient asexuals, a character-istic that defies the predictions of evolutionary theory. The hyphal network of AM fungi is usually aseptate and coenocytic, with hundreds of nuclei sharing the same cytoplasm. Likewise, individual spores contain hun-dreds of nuclei and the question of how the different polymorphic DNA-sequence variants that are present

Faculty of Biology, University of Munich, Großhaderner Straße 2-4, 82152 Planegg-Martinsried, Germany.e-mail: parniske@lmu.dedoi:10.1038/nrmicro1987

AseptateNot containing septae.

CoenocyticMultiple nuclei within the same cell.

Arbuscular mycorrhiza: the mother of plant root endosymbiosesMartin Parniske

Abstract | Arbuscular mycorrhiza (AM), a symbiosis between plants and members of an ancient phylum of fungi, the Glomeromycota, improves the supply of water and nutrients, such as phosphate and nitrogen, to the host plant. In return, up to 20% of plant-fixed carbon is transferred to the fungus. Nutrient transport occurs through symbiotic structures inside plant root cells known as arbuscules. AM development is accompanied by an exchange of signalling molecules between the symbionts. A novel class of plant hormones known as strigolactones are exuded by the plant roots. On the one hand, strigolactones stimulate fungal metabolism and branching. On the other hand, they also trigger seed germination of parasitic plants. Fungi release signalling molecules, in the form of ‘Myc factors’ that trigger symbiotic root responses. Plant genes required for AM development have been characterized. During evolution, the genetic programme for AM has been recruited for other plant root symbioses: functional adaptation of a plant receptor kinase that is essential for AM symbiosis paved the way for nitrogen-fixing bacteria to form intracellular symbioses with plant cells.

R E V I E W S

NATUre revIeWS | microbiology voLUMe 6 | ocTober 2008 | 763

f o c u S o n S y m b I o S I S

Nature Reviews | Microbiology

0.01

Glomus group Ab

Glomus group Aa

Glomus group B

Diversisporaceae(Diversispora)

Acaulosporaceae(Acaulospora)

Gigasporaceae(Gigaspora and Scutellospora)

Archaeosporaceae(Archaeospora)

Ambisporaceae(Ambispora)

Glomerales

Archaeosporales

Diversisporales

Paraglomerales

Glomeraceae 1 and 2

OutgroupsBasidiomycotaand Ascomycota

Pacisporaceae(Pacispora)

RussulaBoletus

AspergillusPenicillium

CandidaKluyveromyces

Paraglomeraceae(Paraglomus)

Geosiphonaceae(Geosiphon)

Glomus sp. W3347

0.05

Glomeromycota

Ascomycota

Basidiomycota

Endogone and Mortierella

Chytridiales, including Basidiobolus

Blastocladiales

Kickxellales andHarpellales

Mucorales

Entomophthorales

Choanoflagellida

Zea mays and Chlorella

Homo sapiens and Aphrodita

Outgroups

Stylonychia

Ulkenia andThraustochytridium

Dictyostelium

a

b

R E V I E W S

764 | ocTober 2008 | voLUMe 6 www.nature.com/reviews/micro

R E V I E W S

AnastomosisA hyphal fusion with a cytoplasmic connection.

Obligate biotrophAn organism that is unable to complete a reproductive cycle in the absence of a living host.

MycoheterotrophicObtains carbon sources from a fungal symbiont.

within a single cell are distributed between genomes or nuclei is the subject of an ongoing debate11–13. Although there is no confirmed report of a sexual stage in the life cycle of AM fungi, it is possible that genetic material is exchanged and recombined; anastomosis between hyphae14,15 allows the exchange of nuclei15 but has so far only been observed between hyphae of closely related fungal strains. It will be interesting to determine the level of relatedness that is required for these fusions to occur. Molecular evidence for recom-bination in AM fungi16,17 has been controversial13. As an important step towards the genetic manipulation of these fungi, transient transformation by particle bombardment has been achieved18.

Although spores of AM fungi can germinate in the absence of host plants, they are obligate biotrophs, and therefore depend on a living photoautotrophic partner to complete their life cycle and produce the next generation of spores. In one reported case, however, co-culture with Paenibacillus validus led to the production of secondary and infective spores in the absence of a host19.

Individual fungal strains exhibit little host specifi-city when grown with different plants under laboratory conditions2. Likewise, a single plant can be colonized by many different AM fungal species within the same root1,20. Therefore, on the one hand, the AM symbiosis is thought to show little host specificity at the level of colonization. on the other hand, the biodiversity of fungal and plant communities are positively correlated with each other21 and host preference seems to play an important role in natural ecosystems22,23. It is likely that these host preferences reflect different fungal strategies and levels of functional compatibility24,25 (FIG. 3). High specificity has been observed between mycoheterotrophic plants and their mycosymbionts26. Most AM fungi that can be detected in natural ecosystems have not been cultured1 and it is possible that many have a more restricted host range than ‘generalists’, such as Glomus mossae or Glomus intraradices, which are intensively investigated because they are easily cultured.

This review will describe AM development, an area in which substantial progress has been made over the past few years. Two novel signalling molecules have been identified that alter fungal development or plant gene expression. Seven plant genes that are involved in symbiotic reprogramming of plant cell development

have been cloned. This reprogramming involves the formation of a newly discovered prepenetration appa-ratus (PPA) by the plant cell in anticipation of fungal infection, by which the plant cell dictates the route of fungal intracellular passage. Moreover, our current knowledge of the function of the AM symbiosis is sum-marized, including the nutrient exchange and metabo-lite fluxes in AM. Finally, evolutionary aspects of AM fungi and the evolution of the plant genetic programme for symbiosis development are considered.

AM developmentThe presymbiotic phase. Multiple, successive rounds of spore germination and retraction of nuclei and cytoplasm can occur in AM fungi. This exploratory hyphal development changes dramatically in the pres-ence of plant-derived signals (FIG. 4). The stimulatory effect of plant root exudates on AM fungal hyphae has been recognized for a long time, but the molecular identity of the ‘branching factors’ has only recently been identified. In two landmark papers, strigolac-tones were found to be responsible for the induction of branching27 and alterations in fungal physiology and mitochondrial activity28. Strigolactones can also stimulate spore germination in some AM fungi. Strigolactones are short-lived in the rhizosphere owing to a labile ether bond that spontaneously hydrolyses in water. This ephemeral compound forms a steep concentration gradient, and therefore its per-ception has been suggested to be a reliable indicator of the proximity of a host root29. Interestingly, the same class of compounds was identified 50 years ago as a potent germination inducer of seeds of the parasitic plant genus Striga. The discovery that strigolactones

Nature Reviews | Microbiology

Fungus

PAS

PAM

Plant

Fungal cell wall

Cytoplasm

Plant cell wallFungal cytoplasm

Fungal plasmamembrane

Figure 1 | Arbuscular mycorrhiza (Am) fungi form an independent phylum, the glomeromycota. a | A phylogenetic tree showing the Glomeromycota in relation to other main fungal lineages: the Ascomycota and Basidiomycota and the non-monophyletic Chytridiomycota (green) and Zygomycota (blue)3. All tested members of the Glomeromycota form AM and all AM fungi are members of the Glomeromycota. b | Phylogenetic relationships between members of the Glomeromycota. Among the four orders that are currently recognized, the Archaeosporales and Paraglomerales are clearly distinct from the subgroup Glomerales and Diversisporales. The phylogeny and taxonomy of AM fungi is still under substantial debate. For example, owing to the significant divergence among the Glomeraceae, this family will probably be taxonomically separated in the future. The scale bar represents the number of substitutions per site. Panel a modified, with permission, from REF. 3 (2001) Cambridge University Press. Panel b modified, with permission, from REF. 122 (2002) Springer Netherlands.

◀

Figure 2 | The arbuscule. Schematic drawing of an arbuscule, the symbiotic structure and arbuscular mycorrhiza (AM). Each fungal branch within a plant cell is surrounded by a plant-derived periarbuscular membrane (PAM) that is continuous with the plant plasma membrane and excludes the fungus from the plant cytoplasm. The apoplastic interface between the fungal plasma membrane and the plant-derived PAM is called the periarbuscular space (PAS). Because of the cell-wall synthesizing potential of both the fungal membrane and the PAM, the PAS comprises fungal and plant cell-wall material.

R E V I E W S

NATUre revIeWS | microbiology voLUMe 6 | ocTober 2008 | 765

f o c u S o n S y m b I o S I S

Nod factorsThe bacterial symbionts of legumes (rhizobia) produce signalling molecules named Nod factors. They consist of an N-acetylglucosamine backbone that carries various strain-specific decorations including a lipid side chain.

Calcium spikingA sharp periodic increase in calcium concentration around the nucleus of symbiotically stimulated root cells.

act as signals for AM fungi has revealed that species of Striga exploit a conserved and ancient communica-tion system between symbiotic fungi and their host plants29. The recent pioneering discovery of strigolac-tones as novel endogenous plant hormones in diverse angiosperms that range from Arabidopsis thaliana to pea and rice30,31 suggests that the strigolactone percep-tion system of Striga is less unique than previously thought. by contrast, it is probably derived from a general hormonal perception system of angiosperms. Whether strigolactones first evolved as endogenous plant hormones or as signals in AM remains an open question. Strigolactone perception by the fungus induces the so-called presymbiotic stage, which is characterized by continued hyphal growth, increased physiological activity and profuse branching of hyphae. Plant mutants that are unable to produce strigolactones are now available in pea and rice30, and can be useful for refined functional analyses of these signalling molecules in AM development.

Fungal signalling molecules and plant receptors. There is currently much interest in the molecular identifica-tion of fungal signalling molecules that induce sym-biosis-specific responses in the host root (collectively called Myc factors). The existence of such molecules became apparent in experiments in which direct contact between the fungus and plant was prevented by fungus-impenetrable membranes. In these experi-ments, a symbiosis-responsive Enod11-promoter GUS (b-glucuronidase) reporter gene fusion in roots of Medicago truncatula was activated in the vicinity of fungal hyphae32. This Myc factor was found to be a diffusible molecule that induced transcriptional acti-vation of symbiosis-related genes. Whether the produc-tion of this Myc factor is stimulated by strigolactones is unclear. A weak but significant increase in lateral root initiation was observed when roots were treated with a diffusible factor from AM fungi33. However, it is unknown whether the root-inducing and Enod11-inducing molecules are the same. calcium signatures that were reminiscent of, but clearly distinct from, Nod-factor-induced calcium spiking were recently observed in root hair cells in the vicinity of, but before contact with, approaching AM hyphae34. Interestingly, calcium oscillations of lower frequency and amplitude than the Nod-factor-induced calcium spiking can be induced by oligo-n-acetylglucosamine35,36. LysM domains have been implicated in n-acetylglucosamine binding. Two different receptor-like molecules that are required for chitin perception, both of which have LysM domains in their extracellular domain, have recently been identified in rice and A. thaliana37,38. These receptor molecules are involved in the induction of resistance responses and constitute part of an ancient perception system for the detection of microbial-associated molecular pat-terns (MAMPs). The Nod factor receptors also contain LysM domains, which are likely to bind the lipochitoo-ligosaccharide Nod factors39,40. Their close structural relationship indicates that the Nod factor and chitin receptors share a common ancestry. AM fungi have chitin in their cell walls and could be recognized by the chitin-perception system. In addition to chitin and Nod factor receptors, both A. thaliana and rice contain several LysM-containing receptor kinases of unknown function. Whether, in common with rhizobia, AM fungi produce chitooligosaccharides or derivatives that function as symbiotic signals which are recognized by LysM receptor kinases remains to be determined. The search for the Myc factor requires a specific assay system, because plant cells are responsive to a range of MAMPs41 that trigger related downstream responses, including calcium responses42. Therefore, the challenge is to show that candidate Myc factors, such as an as-yet-unidentified small molecule from Gigaspora margarita, which elicits calcium responses in soybean cells, can induce symbiosis-specific responses43.

The prepenetration apparatus. The paradigm-shifting discovery that the plant cell actively prepares the intracellular environment for AM fungal hyphae44,45 changed our view of the role of the plant cell during

Nature Reviews | Microbiology

Scutellospora hyphal networks

Plant root

Glomus hyphal networks

Small infectedpatch

Large infectedpatch

Profuse branching and anastomosis close to root

Branching hyphae

Hyphae bridginglong distances;few branchesor anastomoses

Spore closeto or evenwithin root

Spore distant from root

Figure 3 | Different hyphal growth and branching strategies in arbuscular mycorrhiza (Am) fungi. AM fungi have different hyphal growth patterns, anastomoses and branching frequencies. These differences probably reflect different strategies and the occupation of different niches within the soil. Many Glomus species form highly branched and anastomosing hyphal networks. These networks are more recalcitrant to disturbances of the soil than the mycelia of species of Scutellospora or Gigaspora, which form longer hyphae and can probably explore more distant regions of the soil14,123. Most of the fungal biomass in members of the Gigasporaceae family is found in the hyphae that are located outside the plant root, whereas in members of the Glomeraceae family, most of the hyphal biomass is inside the root21.

R E V I E W S

766 | ocTober 2008 | voLUMe 6 www.nature.com/reviews/micro

R E V I E W S

AppressoriumA flattened, hyphal organ that facilitates the penetration of cells or tissues of other organisms.

MicrofilamentStrong, but flexible, linear polymer of actin subunits and component of the cytoskeleton.

infection by biotrophic fungi. The PPA is a subcellular structure that predetermines the path of fungal growth through the plant cell and is formed 4–5 hours after the formation of a fungal appressorium, also called a hyphopodium. Formation of the PPA is preceded by a migration of the plant cell nucleus towards the point of anticipated fungal entry. The nucleus then moves ahead of the developing PPA, as if to guide its growth direction through the cell. The PPA is a thick cyto-plasmic bridge across the vacuole of the host cell. It contains cytoskeletal microtubules and microfilaments, which together with dense endoplasmic reticulum cisternae form a hollow tube within the PPA that con-nects the leading nucleus with the site of appressorial contact44,46 (FIG. 5). only after this ‘transcellular tunnel’ is completed can the fungal hypha penetrate the host cell. endoplasmic reticulum membranes that decorate the tunnel are ideally positioned for the synthesis of the perifungal membrane. However, the signals that trig-ger the formation of the PPA44 are unknown. Purely mechanical stimulation of plant cells with a needle can induce the nucleus to migrate towards the site of distur-bance47. This might be the initial trigger during AM, as this response is independent of the common plant SyM genes dMi2 and dMi3. To induce formation of the PPA, however, additional chemical cues are probably needed to provide specificity. The structurally related

‘pre-infection thread’ of legumes, which forms in response to rhizobia48 in anticipation of bacterial infection, probably evolved from the PPA (FIG. 5).

Plant genes required for AM developmentAt least seven genes that are required for both the AM symbiosis and the root-nodule symbiosis with rhizobia have been identified in legumes49 (TABLE 1). These genes encode proteins that are directly or indirectly involved in a signal transduction network that is required for the development of intracellular accommodation structures for symbiotic fungi and bacteria by the host cell (FIGS 5,6). The AM phenotype of a mutant that is defective in a common symbio-sis gene is characterized by an early block of fungal infection in the outer cell layers49. Phenotypic analy-sis of M. truncatula symbiotic mutants shows that the common SyM genes dMi2 and dMi3 (TABLE 1) are required for PPA induction44 and that dMi3 is required for a subset of genes to be induced during PPA formation46. Transcriptome analysis revealed that most AM-induced genes are not activated in com-mon sym mutants46,49,50. Similarly, the transcriptional response to Nod factors is largely abolished51.

An analysis of calcium spiking in L. japonicus in response to Nod factor revealed that symrk, castor, pollux, nup85 and nup133 mutants are defective for

Outer cortex

Inner cortex

Endodermis

Epidermis

Vascular cylinder

Nature Reviews | Microbiology

Formation ofPPA

Mutual recognition‘presymbiotic phase’

Hyphopodium

Strigolactone

Striga seedling

Myc factor

Spore

Calcium spiking

Fungal penetration

Arbuscule

Nucleus

Plan

t ro

ot

Figure 4 | Steps in arbuscular mycorrhiza (Am) development. Plant roots exude strigolactones which induce spore germination and hyphal branching and increase physiological activity in fungal spores and hyphae. Strigolactones also induce seed germination in parasitic plants, such as Striga124. Fungi produce mycorrhiza (Myc) factors that are operationally defined through their ability to induce calcium oscillations in root epidermal cells34 and to activate plant symbiosis-related genes32. AM fungi form special types of appressoria called hyphopodia, which by definition develop from mature hyphae and not from germination tubes125. As a consequence of sequential chemical and mechanical stimulation, plant cells produce a prepenetration apparatus (PPA). Subsequently, a fungal hypha that extends from the hyphopodium enters the PPA, which guides the fungus through root cells towards the cortex. Here, the fungus leaves the plant cell and enters the apoplast, where it branches and grows laterally along the root axis. These hyphae induce the development of PPA-like structures in inner cortical cells45, subsequently enter these cells and branch to form arbuscules. Vesicles, which are proposed to function as storage organs of the fungus, are sometimes, but not always, formed in AM and are present in the apoplast (not shown). New spores are typically synthesized outside of the plant root at the leading tip of individual fungal hyphae. Figure modified, with permission, from REF. 45 (2008) © American Society of Plant Biologists.

R E V I E W S

NATUre revIeWS | microbiology voLUMe 6 | ocTober 2008 | 767

f o c u S o n S y m b I o S I S

calcium spiking, whereas CCaMK and CyCLoPS act downstream52,53. The data obtained for the AM-induced calcium signatures in M. truncatula mutants are consistent with those for Nod factor, in that dmi1 and dmi2 do not show this response, whereas dmi3 mutants do34. Mutants that have defective com-mon SyM genes do not form infection threads and, with the exception of cyclops mutants, do not initi-ate nodule organogenesis55,56. This suggests that the common SyM gene products are involved in the early stages of symbiotic signal transduction, which involves the generation and decoding of calcium oscillations in and around the nucleus and causes the induction of early symbiosis-related gene expression. consistent with this, some of the predicted protein products are typical signal transduction components, although the contribution of the nucleoporins is likely to be indirect (FIG. 6; TABLE 1). There are subtle differences in the AM phenotypes of common sym mutants in the epidermis, the outer corti-cal cell layers and the arbuscule-forming cells57 (TABLE 1). For example, there is a clear requirement for CCaMK and CyCLoPS in arbuscule development, whereas arbuscules can develop normally in symrk mutant roots. This is indicative of substantial plasticity, and

probably cell-type-specific functionality, in the signal-ling network that is defined by the common SYM pro-teins. Moreover, accumulating evidence suggests there is common SyM-independent signalling32,46. TABLE 1 and FIGURE 6 provide an overview over the common SYM pathway, whereas the following sections describe individual components.

SYMRK. SyMRK (also known as M. truncatula dMi2 or Medicago sativa noRK) encodes a receptor-like kinase58,59 that has an enzymatically functional kinase domain60. owing to the structure of SYMrK and the symbiotic phenotype of corresponding mutants, this molecule is typically portrayed as the entry point into the common symbiotic signalling pathway. In this model, SYMrK has the potential to directly or indirectly perceive extracel-lular signals from microbial symbionts and transduce this perception event through its intracellular kinase domain. The ligands of the SYMrK extracellular domain have not been identified, however. Interestingly, there are at least three different types of SYMrK in the angiosperm lineage which differ in length and the domain structure of their predicted extracellular regions. The shortest type (found in rice) is sufficient for restoring AM in Lotus symrk mutants, whereas the full-length extracellular extension of SYMrK only seems to be required during interactions with rhizobia61. It is therefore possible that during AM and root-nodule symbiosis different extracellular lig-ands bind to different parts of the SYMrK extracellular domain. SYMrK can be exchanged between M. trunca-tula and Lotus japonicus, and the corresponding comple-mented Lotus and Medicago symrk mutant roots regain their ability to form nodules with Mesorhizobium loti and Sinorhizobium meliloti, respectively. This indicates that SYMrK is not involved in determining rhizobial recog-nition specificity. even SYMrK from actinorhiza plants that nodulate with Gram-positive bacteria of the genus Frankia or non-nodulating eurosid species restores nodu-lation and AM in Lotus symrk mutants61,62. Therefore, SYMrK does not contribute to recognitional specificity, and probably does not directly bind to the Nod factor61.

CASTOR and POLLUX. Mutants that are defective in CASToR or PoLLUX (also known as pea SyM8 and Medicago dMi1)63 are also defective in Nod-factor-induced calcium spiking64,65. cASTor and PoLLUX (or DMI1) share similar overall domain structures and high sequence similarity65. electrophysiological measurements of channels that were reconstituted in lipid membranes and yeast-complementation experiments unambiguously showed that these proteins are potassium-permeable cation channels (M. charpentier and colleagues, per-sonal communication). Importantly, these proteins are much less permeable for calcium, which indicates that they are unlikely to be the channels that release calcium from the storage compartment. Nuclear localization of the cASTor and PoLLUX proteins is consistent with their proposed role as counter-ion channels that com-pensate for the rapid charge imbalance that is produced during calcium spiking (M. charpentier and colleagues, personal communication).

Nature Reviews | Microbiology

PPA

Rhizobia

Vacuole

HyphopodiumFungal cell wall

Fungal plasmamembrane

ER cisternae

a b

Tonoplast

Peri-infection thread membrane

PIT

Weakenedcell wall

Infection threadmatrix

PerifungalspacePerifungalmembrane

Plant cellwall

Plant plasmamembrane

Figure 5 | comparison of intracellular accommodation structures in bacterial and fungal root endosymbioses. a | The prepenetration apparatus (PPA) is a cytoplasmic bridge across the vacuole of a plant cell that forms in anticipation of fungal infection (lower cell). The plant cell nucleus migrates ahead of the growing PPA and determines its orientation within the cell. The PPA contains a hollow tube that is formed by microtubules and is lined with endoplasmic reticulum cisternae. Only after completion of the PPA does a fungal hypha enter the PPA (upper cell). b | A preinfection thread (PIT) forms ahead of the bacteria-filled infection thread. The PIT can be induced by bacterial signals alone48 and contains an array of microtubules that resemble the arrangement within the PPA54. The PIT is unique to the nodulating clade and is likely to have evolved from the PPA of arbuscular mycorrhiza (AM). A plant-derived perimicrobial membrane encloses the bacteria-filled infection thread and the fungal hypha and prevents microbial contact with the plant cytoplasm. This membrane synthesises cell wall material, which contributes to the composition of the apoplastic interface between the symbiotic organisms. Part a modified, with permission, from REF. 44 (2005) American Society of Plant Biologists and REF. 45 (2008) American Society of Plant Biologists. Part b modified, with permission, from REF. 118 (2000) Elsevier Science.

R E V I E W S

768 | ocTober 2008 | voLUMe 6 www.nature.com/reviews/micro

R E V I E W S

Nucleoporins. Two genes that encode proteins with similarity to nucleoporins 85 and 133 are required for the temperature-dependent initiation of symbiosis66,67 (FIG. 6; TABLE 1). In humans and yeast, both of these pro-teins belong to the same NUP107–160 subcomplex of the nuclear pore and are not in contact with substrates of the canonical import and export pathways68. However, the transport of proteins that are larger than 75 kDa to the inner nuclear envelope is not well understood69. because Lotus NUP85 and NUP133 act upstream of calcium spik-ing, a plant version of the vertebrate NUP107–160 complex might be involved in transporting cASTor or PoLLUX or both to the inner nuclear envelope. considering the central role of the nuclear pore in transport processes into and out of the nucleus, it is surprising that nup85 and nup133 mutants lack major pleiotropic defects. This could be explained by either a partial redundancy of plant nuclear pore components or functional diversification of the symbiotic nucleoporin homologues.

CCaMK. A calcium–calmodulin-dependent protein kinase (ccaMK) is essential for AM70,71. The calmod-ulin-binding domain and calcium-binding eF hand motifs of ccaMK allow the protein to sense calcium, which makes it a prime candidate for the response to calcium signatures that are induced by AM fungi34 or the Nod factor that induces calcium spiking. Interestingly, a deregulated version of the protein can trigger sponta-neous nodule formation in the absence of rhizobia72,73, which indicates that deregulation of ccaMK alone is sufficient to trigger the organogenesis programme. An

unresolved conundrum is the observation that AM fungi induce calcium signatures but no nodules. Perhaps dur-ing AM, ccaMK is not activated to the same extent or in the same cell types as during nodulation. Alternatively, additional layers of negative regulation might be operating to inhibit nodule organogenesis during AM.

CYCLOPS. cyclops mutants severely impair the infec-tion process of bacterial or fungal symbionts, and are also defective in arbuscule development49. During root- nodule symbiosis, cyclops mutants exhibit specific defects in infection-thread initiation, but not in nod-ule organogenesis (K. Yano and colleagues, personal communication), indicating that cYcLoPS acts in an infection-specific branch of the symbiotic signalling network. CyCLoPS encodes a protein with no overall sequence similarity to proteins with known function, but contains a functional nuclear localization signal and a carboxy-terminal coiled-coil domain. cYcLoPS interacts with ccaMK in yeast and in planta and can be phosphorylated by ccaMK in vitro.

The signalling network that enables symbiotic infection by AM fungi is starting to emerge from the analysis of these cloned plant genes. Additional insights are expected from the identification of the mutations in Petunia arbuscule-development mutants74 and in maize AM mutants74,75.

Arbuscule developmentArbuscules are the result of coordinated subcellular development of the host plant cell and the AM fungus.

Table 1 | Overview of common sym mutants and their corresponding phenotypes

gene mutants Phenotypes of Lotus mutants Predicted function of gene productLotus japonicus

(previous designation)

Medicago truncatula

Pisum sativum

Arbuscular mycorrhiza (Am) phenotype*

root-nodule symbiosis phenotype‡

calcium spiking§

SYMRK symrk59 (sym2) dmi2 (REF. 58) and Medicago sativa nork

sym19 (REF. 126)

Type II Non- nodulating||

No Leucine-rich-repeat receptor kinase60

CASTOR castor65 (sym4 and sym71)

Unknown Unknown Types II and III Non- nodulating||

No Cation channel

POLLUX pollux65 (sym23 and sym86)

dmi1 (REF. 127) sym8 (REF. 127)

Type II Non- nodulating||

No Cation channel

NUP85 nup85 (REF. 67) (sym24, sym73 and sym85)

Unknown Unknown Type II temperature sensitive

Temperature sensitive¶

No Putative nuclear pore component

NUP133 nup133 (REF. 66) (sym3 and sym45)

Unknown Unknown Type II temperature sensitive

Temperature sensitive#

No Putative nuclear pore component

CCaMK ccamk73 (sym15 and sym72)

dmi3 (REFS 70,71)

sym9 (REF. 128)

Types I, II and III Non- nodulating||**

Yes Calcium and calmodulin-dependent protein kinase

CYCLOPS cyclops (sym6, sym30 and sym82)‡‡

ipd3 (REF. 129)§§ Unknown Types II and III Small, non- infected nodules||||

Yes Unknown protein that features a nuclear localization signal and a carboxy-terminal coiled-coil domain

*Arbuscular mycorrhiza (AM) common sym mutant phenotypes I–III are characterized by impaired epidermal opening (type I), impaired intracellular passage through the outer cell layer (or layers) (type II) and/or impaired arbuscule formation (type III). ‡Based on phenotypes described in REFS 115,130–132. §Based on phenotypes described in REFS 52,53. ||Root-hair swelling and branching occurs after inoculation with Mesorhizobium loti, but neither infection threads nor nitrogen-fixing nodules are formed. ¶More arbuscules and nodules form at 18°C compared with 22°C67. #Few nodules form at 22°C, and almost no nodules form at 26°C66. **Deregulated versions of CCaMK induce spontaneous nodule formation in the absence of rhizobia72,73. ‡‡ K. Yano and colleagues, personal communication. §§Mutants not available. |||| cyclops mutants form non-fixing, small white bumps after inoculation with M. loti (K. Yano and colleagues, personal communication).

R E V I E W S

NATUre revIeWS | microbiology voLUMe 6 | ocTober 2008 | 769

f o c u S o n S y m b I o S I S

The fungal hyphae branch repeatedly to produce the tree-shaped arbuscule structure. The exact structure that is formed can vary depending on the fungal and host genotype2. The branches of the fungi are excluded from the host cytoplasm by a plant-derived periarbuscular membrane (PAM). Nutrients, and perhaps signals, are exchanged across the symbiotic interface between the fun-gus and the plant (which constitutes the PAM, the fungal plasma membrane and the periarbuscular space that exists between these two membranes (FIGS. 2,7))76. The trans-porters that mediate metabolite exchange at the interface between the plant and the fungus are of key biotechnologi-cal interest, and some candidate transporter genes have been cloned, although only the PT4 transporter has been specifically localized to the PAM77. The PAM is continu-ous with the plant plasma membrane, but has a distinct protein composition, as revealed by the observation that

the M. truncatula phosphate transporter PT4 is present in the PAM but absent from the plasma membrane77. Ultrastructural analyses have detected molecules in the periarbuscular space that are typically found in the plant primary cell wall, including b-1,4-glucans, non-esterified homogalacturonans, xyloglucans, proteins that are rich in hydroxyproline and arabinogalactan proteins78.

Arbuscules have a shorter lifetime than the host cell (perhaps as short as 8.5 days79), and consequently, a single host cell is thought to be competent for several rounds of successive fungal invasions. In a recent study, Javot and colleagues80 analysed the time-scale of arbuscule development in more detail. They found that arbuscules undergo a phase of growth until a certain maximum size is reached, after which arbuscule degradation or senescence is induced and the arbuscular hyphae become separated from the remaining cytoplasm by septation. Arbuscules subsequently collapse over time and ultimately disappear. This succession of arbuscules is costly in terms of plant and fungal resources, so why are arbuscules so short-lived? The observation that mutation of the arbuscule-specific phosphate transporter PT4 results in premature degradation of arbuscules80 suggests that the lifetime of arbuscules is influenced by their ability to deliver phos-phate and probably other nutrients. This provides the plant with a means to maintain efficient arbuscules and penalize inefficient arbuscules with early degradation. conceptually, this mechanism allows the plant not only to discriminate between efficient and inefficient fun-gal species but also to remove potientally ‘good’ fungal symbionts that are attached to a poor phosphate source. This concept allows fungal clones and species to compete for arbuscule formation, which allows succession in an established root system. The spatial distribution of nutri-ents in the soil will change over time, and well-connected hyphae replace ‘non-providers’. Thus, a limited arbuscule lifetime allows constant renewal and rewiring of the hyphal network and allows connections to be made to the most efficient providers (FIG. 3). AM is a living fos-sil, and therefore mechanisms to specifically promote beneficial symbiotic fungi, and to counter-select against inefficient ‘parasitic’ fungi, might have contributed to the long-term evolutionary stability of this symbiosis. A special role in the development of arbuscules has been ascribed to lysophosphatidylcholine (LPc). LPc is a normal product of phospholipid metabolism and was recently described as a signalling molecule that activates the expression of phosphate-transporter genes, includ-ing the potato gene PT3 (REF. 81). This type of phosphate transporter was found to be required for arbuscule main-tenance80. A phosphate-containing molecule, such as LPc, might be a cell autonomous molecular measure for how much phosphate is made locally available to the plant.

Arbuscule development is accompanied by plastid proliferation and the formation of a plastidial network in close physical contact with the arbuscule82. The plastid is involved in numerous biosynthetic activities, including the production of apocarotenoids that specifically accu-mulate in AM roots83. Given the involvement of hormones in almost all plant developmental processes, it is thought that hormones have key roles during the development of

CaM

SYMRK

Plasma membrane

Extracellular space

Cytoplasm

NFR5

Secondmessenger

Ion channelactivation

NFR1

Ca2+

K+

K+

K+

Calcium spiking

Nucleoplasm

POLLUX

Inner membraneOuter membrane

CASTOR

Nod factor

CYCLOPSNoduleorganogenesis

Nuclearenvelope

Intracellular infection

CCaMK

NUP133, NUP85

Nature Reviews | Microbiology

Nuclear pore complex

Myc factor

HypotheticalMyc factor receptor

Figure 6 | common symbiosis signalling components for arbuscular mycorrhiza (Am) and root-nodule symbiosis. Perception of AM fungal or rhizobia-derived signals triggers early signal transduction, which is mediated by at least seven shared components. The symbiosis receptor kinase SYMRK acts upstream of the Nod factor- and Myc factor-induced calcium signatures that occur in and around the nucleus34. Perinuclear calcium spiking involves the release of calcium from a storage compartment (probably the nuclear envelope) through as-yet-unidentified calcium channels. The potassium-permeable channels CASTOR and POLLUX might compensate for the resulting charge imbalance. The nucleoporins NUP85 and NUP133 are required for calcium spiking, although their mode of involvement is currently unknown. The calcium–calmodulin-dependent protein kinase (CCaMK) forms a complex with CYCLOPS, a phosphorylation substrate, within the nucleus. Together with calmodulin, this complex might decode the symbiotic calcium signatures (K. Yano and colleagues, personal communication). Upstream of the common pathway, the Nod factor receptor kinases NFR1 and NFR5 are specifically required for Nod factor perception39. It is possible that similar receptors are involved in Myc factor perception. Lotus japonicus protein nomenclature is used (see TABLE 1 for the names of common SyM gene orthologues of other species).

R E V I E W S

770 | ocTober 2008 | voLUMe 6 www.nature.com/reviews/micro

R E V I E W S

AM. This is still a developing area of research, but abscisic and jasmonic acid have emerged as potential regulators of AM84,85.

AM FunctionNutrient uptake and transport in the extraradical mycelium. Fungal hyphae explore the soil substratum, but different AM fungi seem to use different strategies to do so (FIG. 3). The fungal hyphal network is ideally positioned to efficiently take up nutrients and water from the soil, but only a few fungal transporters that are involved in this process, including those that transport phosphate86,87, ammonium88 and zinc89, have been cloned. because diffusion is too slow, nutrients are moved in a pack-aged form between the extraradical and the intraradical mycelium (FIG. 7).

Carbon metabolism. our understanding of the meta-bolic functions of the AM and AM fungi has been boosted by the development of axenic culture systems and the ability to restrict fungal and plant tissue to separate compartments, together with isotope labelling and in situ NMr (metabolism and transport in AM has been reviewed previously90,91). The plant can control the flux of sucrose directed to the root, including the fungus. Jasmonic acid has been proposed to be involved in the regulation of sink strength of AM roots92. Sucrose that is delivered to the AM root is cleaved either by symbiosis-induced sucrose synthases93 or invertases94. in vivo NMr studies indicate that AM fungi obtain hexoses from the plant and convert them into lipids and glycogen for long-distance transport8,95. A member of the novel clade of hexose transporters that was identified from

Nature Reviews | Microbiology

Sucrose

ArbusculePlant cytoplasm

Plant cellwall

Plant plasmamembrane

Plant cellwall

Phosphate transporter (plant)

Phosphate transporter (fungus)

Hexose

Hexose

Fungalmonosaccharidetransporter

NH4

Glycogen Glycogen

Acetyl coenzyme AGlycerol

Arg

Pi

Urea

TAG TAG

NO3

NH4

Gln

ERH

ChitinTrehalose

Arg

Pi

Poly-PPoly-P

Soil

Fungal hypha

Ammoniumtransporter

Fungal cell wall

Fungal cytoplasm

Amino acidtransporter

Fungal plasma membrane

Periarbuscularspace

Periarbuscularmembrane

Figure 7 | metabolic fluxes and long-distance transport in arbuscular mycorrhiza (Am). Plant-derived carbon is transported to the fungus through the two membranes at the symbiotic interface. This carbon is first released into the periarbuscular space (PAS), probably in the form of sucrose, then cleaved into hexoses and taken up by AM fungi through transport across the fungal plasma membrane. Within the fungal cytoplasm, hexoses are converted into glycogen granules and triacylglycerol (TAG) lipid droplets, which serve as suitable units for long-distance transport through the hyphal network. Nutrients that are acquired by the fungus from the soil and are delivered to the plant cell have to cross the fungal plasma membrane, be transported long distance to the intraradical hyphae (IRH), including the arbuscules, and subsequently reach the plant cytoplasm across the fungal plasma membrane and the plant periarbuscular membrane (PAM). Phosphate is imported by fungal phosphate transporters (cloned from Glomus intraradices87 and Glomus versiforme86) that are present in extraradical hyphae (ERH). Phosphate is transported towards the root and IRH in the form of polyphosphate granules, which reside in membrane-enclosed vesicles. The negative charge of these granules makes them likely transport vehicles for metal ions and arginine. Phosphate is released from polyphosphate granules within IRH. Plant transporters that are involved in phosphate transport across the PAM have been cloned and characterized80,101–103, whereas the fungal phosphate transporters that are responsible for the release of phosphate from IRH are still unknown. Nitrogen is taken up by ammonium88, nitrate or amino-acid transporters in ERH. In AM fungal hyphae, nitrogen is mainly transported as arginine106. Within the IRH, nitrogen is released from arginine as urea and either transported to the plant directly or after cleavage to ammonium. Figure modified, with permission from REF. 8 (2003) American Society of Plant Biologists, REF. 90 (2005) Blackwell Publishing and Nature REF. 106 (2005) Macmillan Publishers Ltd. All rights reserved.

R E V I E W S

NATUre revIeWS | microbiology voLUMe 6 | ocTober 2008 | 771

f o c u S o n S y m b I o S I S

the symbiotic organ of the glomeromycotan Geosiphon pyriformis might also be expressed in the fungal interface membrane of the arbuscule96. In a typical AM, no carbon transport from the fungus to the plant was detected97. However, mycoheterotrophic plants that associate with AM fungi are likely to receive carbon from the fungus26, and it is proposed that photosynthetic sporophytes of species of Huperzia deliver carbon to mycoheterotrophic gametophytes through shared fungal networks98. This would require carbon transporters that work in the efflux direction, which have not yet been reported in fungi. Alternatively the mycoheterotrophic plant must obtain carbon by efficiently digesting fungal hyphae, similar to the orchid symbiosis.

Phosphate. Improved phosphate uptake is the main bene-fit of the AM symbiosis99,100. The extensive hyphal network of AM fungi influences the physicochemical properties of the soil and directly or indirectly contributes to the release of phosphate from inorganic complexes of low solubility6. Fungal phosphate transporters that are expressed in the extraradical mycelium are probably involved in the uptake of phosphate from the substratum86,87. Polyphosphate granules are used as transport vehicles to move phosphate (and possibly arginine and trace elements) to the host root. Symbiosis-induced plant phosphate transporter genes have been identified in different plant species (reviewed in REF. 101), and accumulating evidence suggests a role for at least a subset of the corresponding proteins in symbiotic phosphate transport80,102,103. Fusions of potato PT3 or M. truncatula PT4 promoters to GUS targeted expression specifically to arbuscule-containing cells, which is con-sistent with results from laser-capture microdissection of tomato arbuscules, in which transcripts of five isoforms were detected in arbuscule-containing cells104.

Nitrogen. AM fungi can accelerate decomposition and directly acquire nitrogen from organic material105. A fungal amino-acid transporter46 and an ammonium transporter that might be involved in nitrogen uptake by extraradical hyphae have been cloned88. Long-distance transport to the plant probably proceeds mainly through arginine106,107 (FIG. 7). Nitrogen is released in a carbon-free form (probably ammonium) to the plant106, although the ammonium transporters in the symbiotic interface membranes have not yet been identified.

Evolution of plant root endosymbiosisAll members of the Glomeromycota phylum require a photosynthetic partner to complete their life cycle. Not a single member of this lineage has escaped from this dependency, which suggests that the ancestral fungus was already an obligate biotroph. This extreme specializa-tion of an entire fungal clade is unique, as members of all phylogenetically comparable lineages — the ascomycetes, basidiomycetes and the non-monophyletic zygomycetes and chytridiomycetes (in the classical sense; for a more recent classification see REF. 4) — inhabit a wide range of ecological niches and include plant and animal patho-gens and symbionts, as well as free-living saprophytes. AM is indeed an ancient symbiosis, and the excellent

fossil record of early land plants from the rhynie chert in Scotland provided ‘rock-solid’ evidence that typical AM fungal structures, such as arbuscules and spores, were already present 400 million years ago108,109. The high level of organization in these fossils and the wide distribution of AM in all branches of the phylogenetic tree of plants suggest that AM might have been present in a common ancestor and perhaps was instrumental in the initial colonization of land. Interestingly, the rhynie chert fossils contain an impressive range of other fun-gal endophytes, including potential parasites110, that provoked the formation of structural ‘defences’ by the plant, which reveals the ancient nature of symbiosis and defence programmes in land plants. Molecular-clock estimates of the age of members of the Glomeromycota differ by several hundred million years111,112, but raise the possibility that they evolved before land plants. In this context, it is interesting that the glomeromycotan fungus G. pyriformis forms a symbiosis with photosynthetic (and nitrogen-fixing) species of nostoc cyanobacteria and that similar fungus–bacteria interactions might have preceded the AM symbiosis113. consistent with an ancient fungal-uptake mechanism for bacteria, most AM fungi harbour endosymbiotic bacteria, including Gram-negative Burkholderia species114 and uncharacterized Gram-positive species113, indicating multiple independent uptake events of symbiotic bacteria.

A conserved ancient genetic programme for AMThe presence of AM in the earliest land plants raises the possibility that the underlying genetic programme is conserved among extant AM-forming plants115. Indeed, phylogenetic and functional conservation of common SyM genes, at least in the angiosperm lineage, has recently been described61,116. evidence from legumes indicates that the common symbiosis genes are required for the formation of the intracellular accommodation structure PPA44. This suggests that these genes are com-ponents of an ancient and conserved programme that evolved before the divergence of the angiosperms and was retained in most lineages because of the selective advantages conferred by AM.

AM is the ancestor of bacterial root endosymbioses. The discovery that some nodulation-defective legume mutants are also defective in AM development revealed a genetic link between bacterial and fungal symbiosis, which has led to the hypothesis that the root-nodule symbiosis evolved from AM functions. Now that common symbiosis genes have been cloned and functionally characterized from different nodulating and non-nodulating angiosperm species, we can draw a more detailed picture of the events that led to the evolution of the root-nodule symbiosis. The combined results suggest that the common sym-biosis programme evolved in the context of AM and was recruited for the bacterial root-nodule symbiosis115. The identified genes are all required for induction of the intracellular accommodation programme, a common feature of both bacterial and fungal root endosymbiosis (FIG. 5). Most common SyM genes are conserved in over-all domain composition between legumes and rice.

R E V I E W S

772 | ocTober 2008 | voLUMe 6 www.nature.com/reviews/micro

R E V I E W S

A notable exception is SYMrK, which exhibits remarkable variation in its extracellular domain composition across angiosperms. Full-length SYMrK is consistently found in all tested members of the eurosid clade, whereas all other tested angiosperms contain shorter types that lack one of the leucine-rich repeats (Lrrs), or one Lrr and a long amino-terminal extension. A central role of SYMrK in the evolutionary events that led to nodulation was revealed by the observation that only full-length SYMrK can fully complement Lotus symrk mutants and restore nodulation, whereas shorter versions are sufficient only for AM61. This finding suggests that exon acquisition in an ancestor of the eurosids was associated with functional adaptation of SYMrK and provided the basis for bacterial triggering of the common symbiosis programme by bacteria61. In a hypothetical scenario, SYMrK evolution was a prereq-uisite for an intracellular symbiosis with nitrogen-fixing bacteria that was manifested by the formation of intracel-lular infection threads. Such an early bacterial symbiosis might not have been associated with nodule organo-genesis115. A situation that mimics such early bacterial endosymbiosis is found in the hit1/har1 double mutant of L. japonicus, which develops abundant infection threads in the absence of nodule formation117.

Phylogenetic and functional analysis of the symbiosis receptor kinase gene SyMRK has revealed functional and structural polymorphism across angiosperms, which suggests that this gene had a key role in the evolution of the nitrogen-fixing root-nodule symbiosis on the basis of a pre-existing AM genetic programme.

because many of the important plant pathogenic fungi share an intracellular biotrophic lifestyle with

AM fungi, it has been suggested that both parasitic and symbiotic fungi rely on partially overlapping intracellular accommodation programmes of the plant118. However, L. japonicus mutants that are defective in common symbiosis genes support completion of the life cycle of a leaf rust fungus, including the formation of intracellular haustoria119. So far, plant genes that support biotrophic fungal pathogens are largely unknown, and therefore the molecular components of such a shared programme have been elusive.

Conclusions and outlookover the past few years, a novel and unexpected devel-opmental capacity of plant cells has been discovered that is essential for the intracellular uptake of AM fungi. Plant genetics will continue to be a major tool in the identification of genes that are required for AM develop-ment and function. It is expected that in the near future, the chemical structure of the fungal Myc factor that triggers the symbiotic responses of the root will be pub-lished, which will help us to identify the cognate plant receptors. To unlock the potential of AM for sustainable agriculture, we must identify the key molecular players. equally importantly, we must investigate the natural variation for AM function and responsiveness within biodiversity collections of important crop plants120 and between different fungal lineages. The long-term aim is to identify or design crop–fungus combinations with optimized AM performance, which would be instrumen-tal in reducing the application of fertilizer and energy input, a goal that is mandatory in a world of depleting non-renewable resources121.

1. Fitter, A. H. Darkness visible: reflections on underground ecology. J. Ecol. 93, 231–243 (2005).

2. Smith, S. E. & Read, D. J. Mycorrhizal Symbiosis (Academic, London, 2008).

3. Schübler, A., Schwarzott, D. & Walker, C. A new fungal phylum, the Glomeromycota: phylogeny and evolution. Mycol. Res. 105, 1413–1421 (2001).Described the fifth fungal phylum, the Glomeromycota, which, with the possible exception of the Geosiphon genus, exclusively comprises AM fungi.

4. Hibbett, D. S. et al. A higher-level phylogenetic classification of the Fungi. Mycol. Res. 111, 509–547 (2007).

5. Miller, R. M., Reinhardt, D. R. & Jastrow, J. D. External hyphal production of vesicular-arbuscular mycorrhizal fungi in pasture and tallgrass prairie communities. Oecologia 103, 17–23 (1995).

6. Finlay, R. D. Ecological aspects of mycorrhizal symbiosis: with special emphasis on the functional diversity of interactions involving the extraradical mycelium. J. Exp. Bot. 59, 1115–1126 (2008).

7. Solaiman, M. D. Z. & Saito, M. Use of sugars by intraradical hyphae of arbuscular mycorrhizal fungi revealed by radiorespirometry. New Phytol. 136, 533–538 (1997).

8. Bago, B. et al. Carbon export from arbuscular mycorrhizal roots involves the translocation of carbohydrate as well as lipid. Plant Physiol. 131, 1496–1507 (2003).

9. Bago, B., Pfeffer, P. E. & Shachar-Hill, Y. Carbon metabolism and transport in arbuscular mycorrhizas. Plant Physiol. 124, 949–958 (2000).

10. Liu, J. et al. Arbuscular mycorrhizal symbiosis is accompanied by local and systemic alterations in gene expression and an increase in disease resistance in the shoots. Plant J. 50, 529–544 (2007).

11. Hijri, M. & Sanders, I. R. Low gene copy number shows that arbuscular mycorrhizal fungi inherit genetically different nuclei. Nature 433, 160–163 (2005).

12. Pawlowska, T. E. & Taylor, J. W. Organization of genetic variation in individuals of arbuscular mycorrhizal fungi. Nature 427, 733–737 (2004).

13. Rosendahl, S. Communities, populations and individuals of arbuscular mycorrhizal fungi. New Phytol. 178, 253–266 (2008).

14. de la Providencia, E. G., de Souza, F. A., Fernandez, F., Delmas, N. S. & Declerck, S. Arbuscular mycorrhizal fungi reveal distinct patterns of anastomosis formation and hyphal healing mechanisms between different phylogenic groups. New Phytol. 165, 261–271 (2005).

15. Giovannetti, M., Sbrana, C., Avio, L. & Strani, P. Patterns of below-ground plant interconnections established by means of arbuscular mycorrhizal networks. New Phytol. 164, 175–181 (2004).

16. Gandolfi, A., Sanders, I. R., Rossi, V. & Menozzi, P. Evidence of recombination in putative ancient asexuals. Mol. Biol. Evol. 20, 754–761 (2003).

17. Vandenkoornhuyse, P., Leyval, C. & Bonnin, I. High genetic diversity in arbuscular mycorrhizal fungi: evidence for recombination events. Heredity 87, 243–253 (2001).

18. Helber, N. & Requena, N. Expression of the fluorescence markers DsRed and GFP fused to a nuclear localization signal in the arbuscular mycorrhizal fungus Glomus intraradices. New Phytol. 177, 537–548 (2008).

19. Hildebrandt, U., Ouziad, F., Marner, F. J. & Bothe, H. The bacterium Paenibacillus validus stimulates growth of the arbuscular mycorrhizal fungus Glomus intraradices up to the formation of fertile spores. FEMS Microbiol. Lett. 254, 258–267 (2006).

20. Santos-Gonzalez, J. C., Finlay, R. D. & Tehler, A. Seasonal dynamics of arbuscular mycorrhizal fungal communities in roots in a seminatural grassland. Appl. Environ. Microbiol. 73, 5613–5623 (2007).

21. Maherali, H. & Klironomos, J. N. Influence of phylogeny on fungal community assembly and ecosystem functioning. Science 316, 1746–1748 (2007).

22. Vandenkoornhuyse, P., Ridgway, K. P., Watson, I. J., Fitter, A. H. & Young, J. P. W. Co-existing grass species have distinctive arbuscular mycorrhizal communities. Mol. Ecol. 12, 3085–3095 (2003).

23. Rosendahl, S. & Stukenbrock, E. H. Community structure of arbuscular mycorrhizal fungi in undisturbed vegetation revealed by analyses of LSU rDNA sequences. Mol. Ecol. 13, 3179–3186 (2004).

24. Sykorova, Z., Ineichen, K., Wiemken, A. & Redecker, D. The cultivation bias: different communities of arbuscular mycorrhizal fungi detected in roots from the field, from bait plants transplanted to the field, and from a greenhouse trap experiment. Mycorrhiza 18, 1–14 (2007).

25. Munkvold, L., Kjoller, R., Vestberg, M., Rosendahl, S. & Jakobsen, I. High functional diversity within species of arbuscular mycorrhizal fungi. New Phytol. 164, 357–364 (2004).

26. Bidartondo, M. I. et al. Epiparasitic plants specialized on arbuscular mycorrhizal fungi. Nature 419, 389–392 (2002).

27. Akiyama, K., Matsuzaki, K. & Hayashi, H. Plant sesquiterpenes induce hyphal branching in arbuscular mycorrhizal fungi. Nature 435, 824–827 (2005).Chemically identified minute amounts of strigolactones from root exudates and showed that they function as root-derived branching stimulants of AM fungal hyphae.

28. Besserer, A. et al. Strigolactones stimulate arbuscular mycorrhizal fungi by activating mitochondria. PloS Biol. 4, 1239–1247 (2006).Identifies strigolactones as stimulants of mitochondrial activity and spore germination in phylogenetically diverse AM fungi.

29. Parniske, M. Cue for the branching connection. Nature 435, 750–751 (2005).

30. Umehara, M. et al. Inhibition of shoot branching by new terpenoid plant hormones. Nature (in the press).

R E V I E W S

NATUre revIeWS | microbiology voLUMe 6 | ocTober 2008 | 773

f o c u S o n S y m b I o S I S

31. Gomez-Roldan, V. et al. Strigolactone inhibition of shoot branching. Nature 10 Aug 2008 (doi:10.1038/nature07271).References 30 and 31 describe the ground-breaking identification of strigolactones, a novel class of plant hormone, through the analysis of plant mutants with exaggerated branching phenotypes. Importantly, strigolactone mutants are impaired in AM.

32. Kosuta, S. et al. A diffusible factor from arbuscular mycorrhizal fungi induces symbiosis-specific MtENOD11 expression in roots of Medicago truncatula. Plant Physiol. 131, 952–962 (2003).

33. Oláh, B., Brière, C., Bécard, G., Dénarié, J. & Gough, C. Nod factors and a diffusible factor from arbuscular mycorrhizal fungi stimulate lateral root formation in Medicago truncatula via the DMI1/DMI2 signalling pathway. Plant J. 44, 195–207 (2005).

34. Kosuta, S. et al. Differential and chaotic calcium signatures in the symbiosis signaling pathway of legumes. Proc. Natl Acad. Sci. USA 105, 9823–9828 (2008).The genetic overlap between AM and the root nodule symbiosis led to the prediction that calcium signatures are an intermediate signal in response to AM fungi. This paper provides experimental proof that such calcium signatures are induced by AM fungi.

35. Walker, S., Viprey, V. & Downie, J. Dissection of nodulation signaling using pea mutants defective for calcium spiking induced by Nod factors and chitin oligomers. Proc. Natl Acad. Sci. USA 97, 13413–13418 (2000).

36. Oldroyd, G. E., Mitra, R. M., Wais, R. J. & Long, S. R. Evidence for structurally specific negative feedback in the Nod factor signal transduction pathway. Plant J. 28, 191–199 (2001).

37. Kaku, H. et al. Plant cells recognize chitin fragments for defense signaling through a plasma membrane receptor. Proc. Natl Acad. Sci. USA 103, 11086–11091 (2006).

38. Miya, A. et al. CERK1, a LysM receptor kinase, is essential for chitin elicitor signaling in Arabidopsis. Proc. Natl Acad. Sci. USA 104, 19613–19618 (2007).

39. Radutoiu, S. et al. LysM domains mediate lipochitin–oligosaccharide recognition and Nfr genes extend the symbiotic host range. EMBO J. 26, 3923–3935 (2007).

40. Smit, P. et al. Medicago LYK3, an entry receptor in rhizobial nodulation factor signaling. Plant Physiol. 145, 183–191 (2007).

41. Boller, T. Chemoperception of microbial signals in plant cells. Annu. Rev. Plant Physiol. Plant Mol. Biol. 46, 189–214 (1995).

42. Müller, J., Staehelin, C., Xie, Z. P., Neuhaus-Url, G. & Boller, T. Nod factors and chitooligomers elicit an increase in cytosolic calcium in aequorin-expressing soybean cells. Plant Physiol. 124, 733–740 (2000).

43. Navazio, L. et al. A diffusible signal from arbuscular mycorrhizal fungi elicits a transient cytosolic calcium elevation in host plant cells. Plant Physiol. 144, 673–681 (2007).

44. Genre, A., Chabaud, M., Timmers, T., Bonfante, P. & Barker, D. G. Arbuscular mycorrhizal fungi elicit a novel intracellular apparatus in Medicago truncatula root epidermal cells before infection. Plant Cell 17, 3489–3499 (2005).This paper represented a paradigm shift in the field of plant–microorganism interactions. Plant cells were found to actively promote fungal infection by forming an intracellular prepenetration apparatus that guides the invading hyphae through the plant cell.

45. Genre, A., Chabaud, M., Faccio, A., Barker, D. G. & Bonfante, P. Prepenetration apparatus assembly precedes and predicts the colonization patterns of arbuscular mycorrhizal fungi within the root cortex of both Medicago truncatula and Daucus carota. Plant Cell 20, 1407–1420 (2008).Extended the key discovery from Reference 44 by showing that cortical cell layers also form a prepenetration apparatus.

46. Siciliano, V. et al. Transcriptome analysis of arbuscular mycorrhizal roots during development of the prepenetration apparatus. Plant Physiol. 144, 1455–1466 (2007).

47. Gus-Mayer, S., Naton, B., Hahlbrock, K. & Schmelzer, E. Local mechanical stimulation induces components of the pathogen defense response in parsley. Proc. Natl Acad. Sci. USA 95, 8398–8403 (1998).

48. van Brussel, A. et al. Induction of pre-infection thread structures in the leguminous host plant by mitogenic lipo-oligosaccharides of Rhizobium. Science 257, 70–72 (1992).

49. Kistner, C. et al. Seven Lotus japonicus genes required for transcriptional reprogramming of the root during fungal and bacterial symbiosis. Plant Cell 17, 2217–2229 (2005).

50. Liu, J. et al. Transcript profiling coupled with spatial expression analyses reveals genes involved in distinct developmental stages of an arbuscular mycorrhizal symbiosis. Plant Cell 15, 2106–2123 (2003).

51. Mitra, R. M., Shaw, S. L. & Long, S. R. Six nonnodulating plant mutants defective for Nod factor-induced transcriptional changes associated with the legume–rhizobia symbiosis. Proc. Natl Acad. Sci. USA 101, 10217–10222 (2004).

52. Miwa, H., Sun, J., Oldroyd, G. E. & Downie, J. A. Analysis of Nod-factor-induced calcium signaling in root hairs of symbiotically defective mutants of Lotus japonicus. Mol. Plant Microbe Interact. 19, 914–923 (2006).

53. Harris, J. M., Wais, R. & Long, S. R. Rhizobium-induced calcium spiking in Lotus japonicus. Mol. Plant Microbe Interact. 16, 335–341 (2003).

54. Timmers, A. C., Auriac, M. C. & Truchet, G. Refined analysis of early symbiotic steps of the Rhizobium–Medicago interaction in relationship with microtubular cytoskeleton rearrangements. Development 12, 3617–3628 (1999).

55. Szczyglowski, K. et al. Nodule organogenesis and symbiotic mutants of the model legume Lotus japonicus. Mol. Plant Microbe Interact. 11, 684–697 (1998).

56. Catoira, R. et al. Four genes of Medicago truncatula controlling components of a Nod factor transduction pathway. Plant Cell 12, 1647–1666 (2000).

57. Parniske, M. Molecular genetics of the arbuscular mycorrhizal symbiosis. Curr. Opin. Plant Biol. 7, 414–421 (2004).

58. Endre, G. et al. A receptor kinase gene regulating symbiotic nodule development. Nature 417, 962–966 (2002).

59. Stracke, S. et al. A plant receptor-like kinase required for both bacterial and fungal symbiosis. Nature 417, 959–962 (2002).

60. Yoshida, S. & Parniske, M. Regulation of plant symbiosis receptor kinase through serine and threonine phosphorylation. J. Biol. Chem. 280, 9203–9209 (2005).

61. Markmann, K., Giczey, G. & Parniske, M. Functional adaptation of a plant receptor-kinase paved the way for the evolution of intracellular root symbioses with bacteria. PLoS Biol. 6, e68 (2008).Shows that symrk is a genetic link between all root endosymbioses. Also showed that AM functionality of the common symbiosis receptor kinase SYMRK is conserved across angiosperm lineages. Importantly, the evolution of root-nodule symbiosis was associated with structural polymorphism in the extracellular domain of SYMRK, which could be involved in ligand perception.

62. Gherbi, H. et al. SymRK defines a common genetic basis for plant root endosymbioses with arbuscular mycorrhiza fungi, rhizobia, and Frankiabacteria. Proc. Natl Acad. Sci. USA 105, 4928–4932 (2008).Showed that all three root endosymbioses are genetically linked, which suggests that both types of root-nodule symbiosis (the actinorhiza–frankia and legume–rhizobium symbioses) evolved by recruiting functions from older AM.

63. Peiter, E. et al. The Medicago truncatula DMI1 protein modulates cytosolic calcium signaling. Plant Physiol. 145, 192–203 (2007).

64. Oldroyd, G. E. D. & Downie, J. A. Coordinating nodule morphogenesis with rhizobial infection in legumes. Annu. Rev. Plant Biol. 59, 519–546 (2008).

65. Imaizumi-Anraku, H. et al. Plastid proteins crucial for symbiotic fungal and bacterial entry into plant roots. Nature 433, 527–531 (2005).

66. Kanamori, N. et al. A nucleoporin is required for induction of Ca2+ spiking in legume nodule development and essential for rhizobial and fungal symbiosis. Proc. Natl Acad. Sci. USA 103, 359–364 (2006).

67. Saito, K. et al. NUCLEOPORIN85 is required for calcium spiking, fungal and bacterial symbioses, and seed production in Lotus japonicus. Plant Cell 19, 610–624 (2007).

68. Alber, F. et al. The molecular architecture of the nuclear pore complex. Nature 450, 695–701 (2007).

69. Lusk, C. P., Blobel, G. & King, M. C. Highway to the inner nuclear membrane: rules for the road. Nature Rev. Mol. Cell Biol. 8, 414–420 (2007).

70. Levy, J. et al. A putative Ca2+ and calmodulin-dependent protein kinase required for bacterial and fungal symbioses. Science 303, 1361–1364 (2004).

71. Mitra, R. M. et al. A Ca2+/calmodulin-dependent protein kinase required for symbiotic nodule development: gene identification by transcript-based cloning. Proc. Natl Acad. Sci. USA 101, 4701–4705 (2004).

72. Gleason, C. et al. Nodulation independent of rhizobia induced by a calcium-activated kinase lacking autoinhibition. Nature 441, 1149–1152 (2006).

73. Tirichine, L. et al. Deregulation of a Ca2+/calmodulin-dependent kinase leads to spontaneous nodule development. Nature 441, 1153–1156 (2006).

74. Sekhara, R., Schorderet, M., Feller, U. & Reinhardt, D. A petunia mutant affected in intracellular accommodation and morphogenesis of arbuscular mycorrhizal fungi. Plant J. 51, 739–750 (2007).

75. Paszkowski, U., Jakovleva, L. & Boller, T. Maize mutants affected at distinct stages of the arbuscular mycorrhizal symbiosis. Plant J. 47, 165–173 (2006).

76. Harrison, M. J. Signaling in the arbuscular mycorrhizal symbiosis. Annu. Rev. Microbiol. 59, 19–42 (2005).

77. Harrison, M. J., Dewbre, G. R. & Liu, J. A phosphate transporter from Medicago truncatula involved in the acquisition of phosphate released by arbuscular mycorrhizal fungi. Plant Cell 14, 2413–2429 (2002).

78. Bonfante, P. in The Mycota Vol. 9 (ed. Hock, B.) 45–61 (Springer-Verlag, Berlin, 2001).

79. Alexander, T., Toth, R., Meier, R. & Weber, H. C. Dynamics of arbuscule development and degeneration in onion, bean and tomato with reference to vesicular-arbuscular mycorrhizae in grasses. Can. J. Bot. 67, 2505–2513 (1989).

80. Javot, H., Penmetsa, R. V., Terzaghi, N., Cook, D. R. & Harrison, M. J. A Medicago truncatula phosphate transporter indispensable for the arbuscular mycorrhizal symbiosis. Proc. Natl Acad. Sci. USA 104, 1720–1725 (2007).Showed that mutation of the arbuscule-localized phosphate transporter PT4 leads to a defective AM symbiosis. A detailed analysis of the kinetics of arbuscule development revealed that the plant penalizes inefficient arbuscules by degrading them prior to their maturation.

81. Drissner, D. et al. Lyso-phosphatidylcholine is a signal in the arbuscular mycorrhizal symbiosis. Science 318, 265–268 (2007).Identified a common phospholipid metabolite that can activate the transcription of phosphate transporter genes. Whether the metabolite originates from the plant, the fungus or both is unknown.

82. Strack, D. & Fester, T. Isoprenoid metabolism and plastid reorganization in arbuscular mycorrhizal roots. New Phytol. 172, 22–34 (2006).

83. Walter, M. H., Floss, D. S., Hans, J., Fester, T. & Strack, D. Apocarotenoid biosynthesis in arbuscular mycorrhizal roots: contributions from methylerythritol phosphate pathway isogenes and tools for its manipulation. Phytochemistry 68, 130–138 (2007).

84. Herrera-Medina, M. J., Steinkellner, S., Vierheilig, H., Ocampo Bote, J. A. & García Garrido, J. M. Abscisic acid determines arbuscule development and functionality in the tomato arbuscular mycorrhiza. New Phytol. 175, 554–564 (2007).

85. Hause, B., Mrosk, C., Isayenkov, S. & Strack, D. Jasmonates in arbuscular mycorrhizal interactions. Phytochemistry 68, 101–110 (2007).

86. Harrison, M. & van Buuren, M. A phosphate transporter from the mycorrhizal fungus Glomus versiforme. Nature 378, 626–629 (1995).

87. Maldonado-Mendoza, E. G., Dewbre, G. R. & Harrison, M. J. A phosphate transporter gene from the extra-radical mycelium of an arbuscular mycorrhizal fungus Glomus intraradices is regulated in response to phosphate in the environment. Mol. Plant Microbe Interact. 14, 1140–1148 (2001).

88. Lopez-Pedrosa, A., Gonzalez-Guerrero, M., Valderas, A., Azcon-Aguilar, C. & Ferrol, N. GintAMT1 encodes a functional high-affinity ammonium transporter that is expressed in the extraradical mycelium of Glomus intraradices. Fungal Genet. Biol. 43, 102–110 (2006).

89. Gonzalez-Guerrero, M. et al. Characterization of a Glomus intraradices gene encoding a putative Zn transporter of the cation diffusion facilitator family. Fungal Genet. Biol. 42, 130–140 (2005).

R E V I E W S

774 | ocTober 2008 | voLUMe 6 www.nature.com/reviews/micro

R E V I E W S

90. Jin, H. et al. The uptake, metabolism, transport and transfer of nitrogen in an arbuscular mycorrhizal symbiosis. New Phytol. 168, 687–696 (2005).

91. Shachar-Hill, Y. Quantifying flows through metabolic networks and the prospects for fluxomic studies of mycorrhizas. New Phytol. 174, 235–240 (2007).

92. Hause, B., Maier, W., Miersch, O., Kramell, R. & Strack, D. Induction of jasmonate biosynthesis in arbuscular mycorrhizal barley roots. Plant Physiol. 130, 1213–1220 (2002).

93. Hohnjec, N., Perlick, A. M., Puhler, A. & Kuster, H. The Medicago truncatula sucrose synthase gene MtSucS1 is activated both in the infected region of root nodules and in the cortex of roots colonized by arbuscular mycorrhizal fungi. Mol. Plant Microbe Interact. 16, 903–915 (2003).

94. Schaarschmidt, S., Roitsch, T. & Hause, B. Arbuscular mycorrhiza induces gene expression of the apoplastic invertase LIN6 in tomato (Lycopersicon esculentum) roots. J. Exp. Bot. 57, 4015–4023 (2006).

95. Bago, B. et al. Translocation and utilization of fungal storage lipid in the arbuscular mycorrhizal symbiosis. Plant Physiol. 128, 108–124 (2002).

96. Schübler, A., Martin, H., Cohen, D., Fitz, M. & Wipf, D. Characterization of a carbohydrate transporter from symbiotic glomeromycotan fungi. Nature 444, 933–936 (2006).Identification of a hexose transporter from an AM fungus. On a global scale, this class of transporter could contribute significantly to the transfer of up to 5 billion tonnes of photosynthetically fixed carbon per year to the fungal symbiont.

97. Pfeffer, P. E., Douds, D. D., Bucking, H., Schwartz, D. P. & Shachar-Hill, Y. The fungus does not transfer carbon to or between roots in an arbuscular mycorrhizal symbiosis. New Phytol. 163, 617–627 (2004).

98. Winther, J. L. & Friedman, W. E. Arbuscular mycorrhizal associations in Lycopodiaceae. New Phytol. 177, 790–801 (2008).

99. Bucher, M. Functional biology of plant phosphate uptake at root and mycorrhiza interfaces. New Phytol. 173, 11–26 (2007).

100. Javot, H., Pumplin, N. & Harrison, M. J. Phosphate in the arbuscular mycorrhizal symbiosis: transport properties and regulatory roles. Plant Cell Environ. 30, 310–322 (2007).

101. Karandashov, V. & Bucher, M. Symbiotic phosphate transport in arbuscular mycorrhizas. Trends Plant Sci. 10, 22–29 (2005).

102. Poulsen, K. H. et al. Physiological and molecular evidence for Pi uptake via the symbiotic pathway in a reduced mycorrhizal colonization mutant in tomato associated with a compatible fungus. New Phytol. 168, 445–453 (2005).

103. Maeda, D. et al. Knockdown of an arbuscular mycorrhiza-inducible phosphate transporter gene of Lotus japonicus suppresses mutualistic symbiosis. Plant Cell Physiol. 47, 807–817 (2006).

104. Balestrini, R., Gomez-Ariza, J., Lanfranco, L. & Bonfante, P. Laser microdissection reveals that transcripts for five plant and one fungal phosphate transporter genes are contemporaneously present in

arbusculated cells. Mol. Plant Microbe Interact. 20, 1055–1062 (2007).

105. Hodge, A., Campbell, C. D. & Fitter, A. H. An arbuscular mycorrhizal fungus accelerates decomposition and acquires nitrogen directly from organic material. Nature 413, 297–299 (2001).

106. Govindarajulu, M. et al. Nitrogen transfer in the arbuscular mycorrhizal symbiosis. Nature 435, 819–823 (2005).Shows that arginine transports nitrogen in AM hyphae and that nitrogen is transferred to the plant without carbon.

107. Cruz, C. et al. Enzymatic evidence for the key role of arginine in nitrogen translocation by arbuscular mycorrhizal fungi. Plant Physiol. 144, 782–792 (2007).

108. Remy, W., Taylor, T. N., Hass, H. & Kerp, H. Four hundred-million-year-old vesicular arbuscular mycorrhizae. Proc. Natl Acad. Sci. USA 91, 11841–11843 (1994).

109. Dotzler, N., Krings, M., Taylor, T. N. & Agerer, R. Germination shields in Scutellospora (Glomeromycota: Diversisporales, Gigasporaceae) from the 400 million-year-old Rhynie chert. Mycol. Prog. 5, 178–184 (2006).

110. Krings, M. et al. Fungal endophytes in a 400-million-yr-old land plant: infection pathways, spatial distribution, and host responses. New Phytol. 174, 648–657 (2007).

111. Heckman, D. S. et al. Molecular evidence for the early colonization of land by fungi and plants. Science 293, 1129–1133 (2001).

112. Redecker, D., Kodner, R. & Graham, L. E. Glomalean fungi from the Ordovician. Science 289, 1920–1921 (2000).

113. Schübler, A., Krüger, M. & Walker, C. in The Mycota Vol. 14 (eds Wöstemeyer, J. & Martin, W.) (Springer-Verlag, Berlin, in the press).

114. Bianciotto, V., Lumini, E., Bonfante, P. & Vandamme, P. ‘Candidatus Glomeribacter gigasporarum’ gen. nov., sp nov., an endosymbiont of arbuscular mycorrhizal fungi. Int. J. Syst. Evol. Microbiol. 53, 121–124 (2003).

115. Kistner, C. & Parniske, M. Evolution of signal transduction in intracellular symbiosis. Trends Plant Sci. 7, 511–518 (2002).

116. Chen, C. Y., Gao, M. Q., Liu, J. Y. & Zhu, H. Y. Fungal symbiosis in rice requires an ortholog of a legume common symbiosis gene encoding a Ca2+/calmodulin-dependent protein kinase. Plant Physiol. 145, 1619–1628 (2007).

117. Murray, J. D. et al. A cytokinin perception mutant colonized by Rhizobium in the absence of nodule organogenesis. Science 315, 101–104 (2007).

118. Parniske, M. Intracellular accommodation of microbes by plants: a common developmental program for symbiosis and disease? Curr. Opin. Plant Biol. 3, 320–328 (2000).

119. Mellersh, D. & Parniske, M. Common symbiosis genes of Lotus japonicus are not required for intracellular accommodation of the rust fungus Uromyces loti. New Phytol. 170, 641–644 (2006).

120. Sawers, R. J., Gutjahr, C. & Paszkowski, U. Cereal mycorrhiza: an ancient symbiosis in modern

agriculture. Trends Plant Sci. 13, 93–97 (2008).

121. Vance, C. P. Symbiotic nitrogen fixation and phosphorus acquisition. Plant nutrition in a world of declining renewable resources. Plant Physiol. 127, 390–397 (2001).

122. Schüßler, A. Molecular phylogeny, taxonomy, and evolution of Geosiphon pyriformis and arbuscular mycorrhizal fungi. Plant Soil 244, 75–83 (2002).

123. Voets, L., de la Providencia, E. G. & Declerck, S. Glomeraceae and Gigasporaceae differ in their ability to form hyphal networks. New Phytol. 172, 185–188 (2006).

124. Cook, C. E., Whichard, L. P., Turner, B. & Wall, M. E. Germination of witchweed (Striga Lutea Lour): isolation and properties of a potent stimulant. Science 154, 1189–1190 (1966).

125. Bastmeyer, M., Deising, H. & Bechinger, C. Force exertion in fungal infection. Annu. Rev. Biophys. Biomol. Struct. 31, 321–341 (2002).

126. Schneider, A. et al. Genetic mapping and functional analysis of a nodulation-defective mutant (sym19 of pea (Pisum sativum L.). Mol. Gen. Genet. 262, 1–11 (1999).

127. Ané, J. M. et al. Medicago truncatula DMI1 required for bacterial and fungal symbioses in legumes. Science 303, 1364–1367 (2004).

128. Schneider, A. et al. Mapping of the nodulation loci sym9 and sym10 of pea (Pisum sativum L.). Theor. Appl. Genet. 104, 1312–1316 (2002).

129. Messinese, E. et al. A novel nuclear protein interacts with the symbiotic DMI3 calcium- and calmodulin-dependent protein kinase of Medicago truncatula. Mol. Plant Microbe Interact. 20, 912–921 (2007).

130. Demchenko, K., Winzer, T., Stougaard, J., Parniske, M. & Pawlowski, K. Distinct roles of Lotus japonicus SYMRK and SYM15 in root colonization and arbuscule formation. New Phytol. 163, 381–392 (2004).

131. Bonfante, P. et al. The Lotus japonicus LjSym4 gene is required for the successful symbiotic infection of root epidermal cells. Mol. Plant Microbe Interact. 13, 1109–1120 (2000).

132. Novero, M. et al. Dual requirement of the LjSym4 gene for mycorrhizal development in epidermal and cortical cells of Lotus japonicus roots. New Phytol. 154, 741–749 (2002).

AcknowledgementsI thank A. Schübler for helpful discussions and for adapting FIG. 1, and K. Haage for compiling TABLE 1. I apologize to all those colleagues whose work was not cited owing to space restrictions.

DATABASESEntrez Genome Project: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=genomeprjLotus japonicus | Medicago truncatula | Mesorhizobium loti | Sinorhizobium meliloti

FURTHER INFORMATIONmartin Parniske’s homepage: http://www.genetik.bio.lmu.de

All linkS Are AcTive in The online PDf

R E V I E W S

NATUre revIeWS | microbiology voLUMe 6 | ocTober 2008 | 775

f o c u S o n S y m b I o S I S