Molecular genetics of the arbuscular mycorrhizal symbiosis

Molecular genetics of the arbuscular mycorrhizal symbiosisMartin Parniske

During arbuscular mycorrhiza (AM) development, fungal

hyphae grow throughout root epidermal, exodermal and cortical

cell layers to reach the inner cortex where the symbiosis’

functional units, the arbuscles, develop. Three essential

components of a plant signalling network, a receptor-like kinase,

a predicted ion-channel and a calmodulin-dependent protein

kinase have been identified. A detailed morphological study

of symbiotic plant mutants revealed that different subsets of

plant genes support the progress of fungal infection in

successive root cell layers. Moreover, evidence of a diffusible

fungal signalling factor that triggers gene activation in the

root has recently been obtained.

AddressesThe Sainsbury Laboratory, Colney Lane, Norwich NR4 7UH, UK

e-mail: [email protected]

Current Opinion in Plant Biology 2004, 7:414–421

This review comes from a themed issue on

Biotic interactions

Edited by Maria J Harrison and Ian T Baldwin

Available online 9th June 2004

1369-5266/$ – see front matter

� 2004 Elsevier Ltd. All rights reserved.

DOI 10.1016/j.pbi.2004.05.011

AbbreviationsAM arbuscular mycorrhiza(l)

DMI1 DOES NOT MAKE INFECTIONS1

ENOD11 EARLY NODULATION11

GUS b-glucuronidase

NFR Nod-factor receptor

NORK NODULATION RECEPTOR-LIKE KINASE

RNS root nodule symbiosis

SYMRK symbiosis receptor kinase

Introduction: the fungal partnerWhen phylogeny and ecology are considered, the arbus-

cular mycorrhizal (AM) symbiosis is probably the most

important interaction between plants and microbes. More

than 80% of land plants form AM, which improve their

nutrient (especially phosphate) uptake [1]. In turn, the

plant provides both carbohydrates and lipids to the fungus

[2]. Worldwide, it has been estimated that 5 billion tons of

carbon are transferred from plants to fungi via AM each

year [3]. AM are an integral part of many ecosystems, and

are considered to be particularly advantageous to plants

growing in tropical soils, in which nutrients are cycled

rapidly and are present in low steady-state concentrations

[4]. All AM fungi belong to the Glomeromycota, an ancient

group of fungi that was present about 450 million years

ago, and were instrumental for plants to colonise land

[5–7]. AM fungi cannot complete their life cycle in the

absence of a host root. The indispensable compounds that

are provided by the host root have not yet been identified,

although fungal growth can be promoted by specific

bacteria [8]. Moreover, the genetics of the Glomeromycotaare still a mystery; no sexual stages are known and single

spores can contain hundreds of nuclei. The observation of

unusual rDNA polymorphism indicated that nuclei that

coexist in a single fungal cell are not genetically identical

[9], a view that has been challenged recently [10]. The

genome sequence of Glomus intraradices, estimated to be

only 16 Mbp in size [11], is expected to shed light on some

of these issues. Glomus expressed sequence tag (EST)

sequences from spores, extraradical mycelium and

hyphae within plant roots have provided insights in to

the gene repertoire of this fungus [12–17] and its meta-

bolic capabilities [18]. The establishment of a robust

transformation system is a prerequisite for studies of

the functional genomics of AM fungi, and transient

expression of marker genes using biolistic [19] and

Agrobacterium-based approaches (N Requena, pers.

comm.) has been achieved.

Plant genes required for symbioticdevelopmentPlant mutants are the key tools for the genetic dissection

of AM development. AM mutants have been identified in

non-legume species, most prominently in tomato [20,21].

By far the largest number of AM-mutants, however, have

been isolated from legumes by examining the ability of

mutants that have defects in the nitrogen-fixing root

nodule symbiosis (RNS) to develop AM ([22–32];

Table 1). The mutants isolated this way are defective

in both bacterial and fungal symbiosis. The impaired

genes are thus collectively referred to as the ‘common’

SYM genes [33]. These genes define a partial overlap

between the genetic programmes for fungal and bacterial

endosymbiosis [34,35]. The evolutionary implication is

that the younger RNS has recruited functions from the

ancient AM symbiosis [33,36]. The common SYM genes

are therefore in the centre of interest from both a mechan-

istic and an evolutionary viewpoint.

The bacterial symbionts of legumes, the rhizobia that

inhabit root nodules, produce lipochitin-oligosaccharide

signals that induce symbiotic responses in the host root.

Most of the more thoroughly analysed common SYMgenes play a role in the transduction of the Nod-factor

signal [29,37��]. However, two nodulation-specific Nod-

factor receptor (NFR) kinases, NFR1 and NFR5, have

Current Opinion in Plant Biology 2004, 7:414–421 www.sciencedirect.com

been positioned upstream of these common pathway

components. These kinases contain LysM motifs in their

extracellular domains, which are implicated in binding N-

acetyl-glucosamine-containing molecules. The presence

of LysM domains makes NFR1 and NFR5 prime candi-

dates to be Nod-factor receptors [38��–40��,41]. Muta-

tions in NFR1 or NFR5 affect the earliest Nod-factor

responses [39��] but not the AM symbiosis [28], suggest-

ing that the essential fungal signalling factor(s) [42��] are

dissimilar to the Nod-factor. The Nod-factor perception

event that involves NFR1 and NFR5 is likely to be

transmitted through SYMRK [37��,43��]. Whether this

occurs directly through the formation of SYMRK–NFR

heterocomplexes or indirectly via secondary signals that

are released from a NFR1/5 complex, as suggested by

Radutoiu et al. [39��], remains to be established. Within

seconds, Nod-factors induce ion fluxes across the plasma

membrane of the root hair cell, as evidenced by an

extracellular alkalinisation and membrane depolarisation.

NFR5 is absolutely required for this response, whereas

the different responses observed in single and double

mutants suggest that the contribution of SYMRK and

NFR1 might be synergistic [39��]. Localised changes

of ion concentration, similar to those induced by Nod-

factors, are likely to occur in the interaction with AM

fungi but have not yet been experimentally verified.

Only a few markers are available that can differentiate

between the phenotypes of common SYM mutants, and

that provide a framework to position the genes that are

affected in these mutants in a conceptual signalling

cascade. One such marker is the calcium-spiking response

of root hair cells of various legume species that occurs 10–

30 min after Nod-factor treatment [35]. These rhythmic

oscillations in cytoplasmic calcium concentration are

probably a signal, but to prove this experimentally is a

key challenge for the future. The common SYM gene

DOES NOT MAKE INFECTIONS1 (DMI1) from Medicagotruncatula encodes a predicted ion channel [44��] that acts

upstream of calcium spiking. The receptor-kinase struc-

ture of SYMRK (or its orthologue DMI2 in M. truncatula)

suggests that it acts mechanistically upstream of the ion

channel. In a hypothetical scenario, SYMRK is involved

in the direct or indirect perception of a fungal signal,

which is then transduced through the intracellular kinase

to activate the ion channel (Figure 1). It will be interest-

ing to establish which of the early-symbiosis-induced ion

fluxes DMI1 is involved in.

DMI3 is a calcium- and calmodulin-dependent protein

kinase (CCaMK), which contains three EF hands impli-

cated in calcium-binding and a calmodulin-binding site

[45��,46��]. These sequence features open the exciting

possibility that the CCaMK is able to interpret the

calcium-spiking signal, resulting in a phosphorylation

event [45��,46��]. It appears likely that this phosphor-

ylation event is not the only signal that is required for

the onset of symbiotic development, and that additional

components will be revealed through genetics and

biochemistry.

Signal exchange before infectionHow do fungal and plant partners find each other within

the soil? One possibility is that root-derived signals redir-

ect fungal growth. At least on Petri-dishes, however, the

negative geotropism of Gigaspora germtubes does not

appear to be influenced by the presence of roots [47].

An alternative strategy would be to ramify hyphae in the

Table 1

Orthologous relationship between some legume symbiosis genes.

L. japonicus Pisum sativum M. truncatula Predicted function Reference(s)

(a) Cloned nodulation genes that are required for early stages of RNS

Cloned gene Initial designation

NFR1 LjSYM1 (PsSYM2?) LYK3 Nod-factor receptor kinase [38��,39��]

NFR5 LjSYM5 PsSYM10 (NFP?)

NIN PsSYM35 ? Membrane protein with [76,77]

DNA-binding domain

(b) Common SYM genes required for AM and RNS? (PsSYM8?) DMI1 Ion channel [44��]

SYMRK LjSYM2 PsSYM19 DMI2 (MsNORK) Symbiosis receptor kinase [37��,43��]

? PsSYM9/30 DMI3 Calcium- and calmodulin- [45��,46��]

dependent kinase

LjSYM3 ? ? [26]

LjSYM4 ? ? [26]

LjSYM15 ? ? [26]

LjSYM23 ? ? [27]

LjSYM30 ? ? [27]? PsSYM36 ? [57]

Symbiotic mutants that are affected in bacterial and fungal symbiosis have also been isolated in Melilotus alba, Vicia faba and Phaseolus

vulgaris (reviewed in [34]) but their relationships to the listed genes are unclear.

Molecular genetics of the arbuscular mycorrhizal symbiosis Parniske 415

www.sciencedirect.com Current Opinion in Plant Biology 2004, 7:414–421

vicinity of roots to increase the chance of an encounter

with the host. A root-derived signal induces the branching

of (free-living) fungal hyphae [48]. This ‘branching factor’

has been partially purified [49] but its chemical structure

has not been published yet (Figure 2a).

Evidence for a fungal signalling molecule that induces

plant gene activation was gained from experiments by

Kosuta et al. [42��] in which fungal hyphae and host roots

were grown in close proximity but separated from each

other by membranes that were impenetrable for roots or

fungal hyphae. In this set-up, a Medicago EARLY NODU-LATION11 (ENOD11)-promoter::b-glucuronidase (GUS)

fusion, which was responsive to both arbuscular mycor-

rhiza fungi and rhizobial Nod-factor [50], was activated at

a distance from fungal hyphae [42��]. This activation was

the first experimental evidence of a long-postulated fun-

gus-derived diffusible signalling molecule. GUS-staining

was limited to roots sectors located in the vicinity of

ramified hyphae, suggesting that signal production by

individual hyphae either was too low to elicit a response

or was induced concomitantly with the hyphal-branching

response [42��]. Whether a single plant molecule induces

hyphal branching and production of the fungal factor

remains to be shown. In further work using the

ENOD11-promoter::GUS transgenic line carried out in

the same laboratory, Chabaud et al. [47] and Kosuta et al.[42��] found that gene expression patterns differed

dramatically depending on whether fungal contact with

the root was permitted or not. In the absence of fungal

contact, large root sections of several centimetres in

length responded, and GUS staining was observed in

most epidermal and cortical cell layers of these sections

[42��]. In contrast, once fungal contact with the root was

established, ENOD11 expression was confined to infected

cells [47]. This striking difference provides a conundrum

that can be resolved by postulating a negative regulatory

mechanism that, only upon fungal contact, suppresses

Figure 1

?SYMRK/NORK

DMI1

?

KinaseCurrent Opinion in Plant Biology

Symbiosis DMI3

Ca2+-spiking

Fungal signal

A model of events that are mediated by the predicted protein

products of cloned common SYM genes. Among the identified

components, the SYMRK/NORK/DMI2 receptor kinase may be the

earliest to act in the AM signalling pathway [37��,43��]. It perceives

signals emanating from the fungal microsymbiont either directly or

indirectly, and transduces the event through its intracellular kinase

domain. This, in turn, activates the predicted ion channel, DMI1 [44��].

The availability of purified bacterial signalling compounds and

experimental difficulties arising from the obligate biotrophic nature

of the fungus have contributed to a situation in which we know more

about early signalling events in root nodule symbiosis than in AM.

In particular, we do not know whether the calcium-spiking response

that is characteristic of the rhizobial symbiosis also occurs in the

mycorrhizal interaction. The DMI3 kinase potentially responds directlyto oscillations in calcium-concentration, however, implying that

Ca2þ is also a messenger in mycorrhizal signalling [45��,46��].

Figure 2

Fungal signal

Branching factor

+ + + + + + + +

– –––

(a)

(b) (c)

Current Opinion in Plant Biology

Signal exchange between the plant root and the hyphae of AM fungi

before infection. (a) Roots (left) release a branching factor that

induces alterations in the growth pattern of the fungus. In turn,

the fungus (right) releases a diffusible signal that is recognised by

the plant and that leads to symbiosis-related gene activation.

(b) The Kosuta experiment [42��]. Fungal hyphae (brown) growing

in the vicinity of the host root but separated from the host root by a

cellophane membrane release a diffusible signal (þ) that induces the

widespread expression of a ENOD11-promoter::GUS fusion in

M. truncatula roots. This gene induction also occurs in the roots of dmi1,

dmi2 and dmi3 mutants. (c) The Chabaud experiment [47]. When

fungal contact with the root is established ENOD11-promoter::GUS

expression is limited to cells that are contacted by the fungus. The

different root responses in (b) and (c) can be explained by postulatinga negative regulatory signal (�) that limits the symbiotic response to

those cells that are actually engaged. This negative regulatory circuit

also appears to be active in dmi2 mutants, as no ENOD11-

promoter::GUS staining was observed when fungal contact with

the roots of this mutant was permitted.

416 Biotic interactions


ENOD11 in those cells that are not infected by the fungus

(Figure 2c).

The use of symbiotic plant mutants revealed additional

switches in the regulatory circuits that are involved

in fungal infection. ENOD11-promoter::GUS was not

expressed at all in dmi2 mutant roots that were contacted

by the fungus [47], whereas wildtype-like GUS expres-

sion was observed when the fungus and mutant roots were

separated by a membrane [42��]. The latter result was

entirely unexpected; the ENOD11-promoter::GUS fusion

does not respond to Nod-factor treatment in dmi mutant

backgrounds [29], so a positive response to the diffusible

fungal factor indicates significant differences in the ways

that signalling from Nod-factor and from fungal factor are

integrated. In contrast to Nod-factor signalling, signalling

from the fungal factor to ENOD11-promoter::GUS induc-

tion is independent of DMI1, DMI2 and DMI3. What’s

more, the absence of gene induction upon fungal contact

with the mutant roots shows that the mechanisms that

repress gene expression upon fungal contact are intact in

the mutants. This is exciting, because it opens genetic

avenues to identify components that are involved in these

pathways.

Cell-type-specific symbiotic programmesThe most frequently described AM phenotype of common

SYM mutants in Medicago and pea is appressorium forma-

tion at the root surface and a block of all subsequent

infection steps (reviewed in [34]). More recently, a series

of mutant alleles in L. japonicus, and the relative vigour with

which the AM fungi G. intraradices or Gigaspora margaritaattack plant roots have enabled studies of the role of

common SYM genes in cell types other than the epidermis.

Epidermal openingAfter colonisation of the root surface, fungal hyphae

typically enter the root through a cleft that opens between

the anticlinal walls of two adjacent epidermal cells.

Genetic evidence indicates that the opening of this cleft

is, to a large extent, an activity of the host plant [51�].These clefts are not fully formed in plant mutants that are

affected in the LjSYM15 gene, and fungal hyphae con-

tinue to grow on the root surface of these mutants pre-

cisely along the boundaries of epidermal cells, as if in

search of such opening clefts. This epidermal opening

occurs along the middle lamella and is probably mediated

through the production and localised release of pectino-

lytic enzymes by epidermal cells. LjSYM15 is presumably

involved in the regulation of this process. Surprisingly,

neither LjSYM4 [52�] nor the receptor kinase SYMRK are

required for the epidermal-opening response, which must

therefore rely on alternative signal perception and trans-

duction components [51�].

Intracellular passage to the inner cortexFungal hyphae form appressoria to gain entry into root

epidermal or exodermal cells. Appressorium formation by

Figure 3

Epidermal opening:LjSYM15

Intracellular accommodationin epi- and exodermis andouter cortex: SYMRK, LjSYM4

Arbuscule development:LjSYM4, LjSYM15

Epidermis

Exodermis

Outer cortex

Inner cortex


Cell-layer-specific responses that are mediated by distinct subsets of common SYM genes. A schematic overview of AM development,

as compiled from cytological analysis of the interaction of L. japonicus roots with G. intraradices or G. margarita [51�,52�,55]. Fungal hyphae

(blue) grow through epidermis and exodermis of the root to form arbuscules in the inner cortex. Plant genes and cell types that are involved indifferent steps are depicted. Directly below a fungal hypha, the walls of two adjacent epidermal cells separate from each other, a process that

the LjSYM15 gene appears to contribute to. A fungal hypha enters this gap, and subsequently penetrates an epi- or exodermal cell. Successful

intracellular accommodation within this and the subsequent cell layer requires the Lotus SYMRK and LjSYM4 genes [51�,52�]. The fungus leaves

the plant cell and re-enters the apoplast in the inner cortex. There, hyphae can grow relatively rapidly in the spaces between adjacent cells. They

branch to penetrate cells of the two innermost cortical layers, forming arbuscules. The uptake of hyphae into the inner cortical cells requires the

LjSYM4 and LjSYM15 genes [51�,52�]. Once inside these cells, proper arbuscule development in pea depends on the PsSYM36 gene [56,57].



an AM fungus was induced by purified epidermal cell wall

fragments, suggesting that the physical properties of

these fragments are sufficient to elicit this response

[53]. Fungal hyphae grow through the initially infected

cell and are guided to the underlying cell layer. The

hyphae then pass through the outer layers of living cells,

where they are surrounded by a plant-derived accommo-

dation structure that comprises a perifungal membrane

that is continuous with the plasma membrane. Additional

symptoms of the induction of symbiotic programmes in

these cells are the activation of a M. truncatula serine

carboxypeptidase II promoter::green fluorescent protein

(GFP) [54] and an ENOD11-promoter::GUS fusion [47],

both of which are also induced in arbuscule-containing

cells. The activation of the host’s intracellular accommo-

dation programme in the outer cell layers is dependent on

the LjSYM4 and LjSYMRK genes [51�,52�]. When these

genes are defective, fungal entry into the first cell occurs

but the interaction aborts. In the Ljsym4 mutant at least,

this aborted interaction is associated with the death of

both the plant cell and the hyphal tip that is physically

involved in the penetration event [55].

Release into the apoplast and fungalspread in the cortexThe initial intracellular passage through the outer cell

layers constitutes a bottleneck of AM establishment, after

which the fungus is released into the apoplast between

cortical cells. This extracellular phase allows for the rapid

proliferation of fungal hyphae along the longitudinal axis

of the root, and relatively large root sectors can be

colonised as the result of a single successful initial infec-

tion event (Figure 3). No plant mutants that are affected

at this phase of symbiotic development have yet been

described.

Arbuscule development and functionAs the fungus grows through the apoplast, hyphal branches

are formed that penetrate inner cortical cells to initiate

arbuscule formation. No obvious appressoria are formed

before fungal penetration of the plant cortical cell wall.

The entry of hyphae into inner cortical cells is impaired in

LjSYM4 and LjSYM15 mutants, indicating that the plant

has an active role in this process [51�,52�]. Lotus plants

carrying weak mutant alleles of the LjSYM15 gene suffer

from delayed nodulation and arbuscule development [51�].

Once inside inner cortical cells, fungal hyphae branch

profusely and form a tree-like structure, the arbuscule.

Host genes that are required for arbuscule development

have been identified: mutants affected in the pea

PsSYM36 gene are low and late nodulating and form only

stumpy branches instead of proper arbuscules [56,57].

The periarbuscular membrane in this interaction does not

stain for an ATPase activity observed in the wildtype,

indicating that these rudimentary arbuscules are probably

non-functional [58].

Several genes have been characterised using promoter::

reporter gene fusions, such as those involving M. trunca-tula CEL1 (b-1,4 glucanase) and phosphate transporter

genes, that are specifically activated in arbuscule-contain-

ing cells ([54,59�]; Figure 4). Such promoters can be

useful for the dissection of components that specify

the transcriptome of arbuscules.

The extent of AM colonisation is strictly controlled by the

plant. The HYPERNODULATION AND ABERRANT

ROOT FORMATION1 (HAR1) receptor kinase, which

controls the number of root nodules in legumes [60–62],

also appears to be involved in the regulation of AM

because har1 mutants have higher fungal colonisation

rates than do wildtype plants [63]. ENOD40 overexpres-

sion leads to enhanced arbuscule formation [64], but the

mode of action of this gene is still not understood.

Arbuscules are ephemeral structures and their turnover

seems to be affected in the pea mutants sym33 and sym40,

which also form inefficient root nodules [65]. The genetic

determinants that control the functional efficiency of AM

Figure 4


Soil

Plant

Pi

Pi(a)

(b)

(c)

Fungus

Phosphate transport from the soil via the fungus to the plant. The

primary function of AM is phosphate uptake, and the cloning of the

phosphate transporters that are responsible for this uptake is therefore

a key goal. (a) A fungal phosphate transporter that is involved in the

uptake of inorganic phosphate from the soil has been cloned from two

AM fungi [72,73] and is regulated by phosphate concentration in the

environment [73]. More recently, three groups reported the cloning of

mycorrhiza-induced plant phosphate transporter genes from potato

[74�], M. truncatula [59�] and rice [75�]. Phylogenetic analysis

suggests that the genes from rice and M. truncatula are closely

related whereas the potato protein is more distantly related [59�],indicating that at least two classes of plant transporters are involved.

(b) Immunolocalisation of the Medicago protein on the periphery of

the arbuscule [59�] suggests that this transporter operates within the

periarbuscular membrane. How the fungus (c) transports phosphate

out of its cytoplasm into the periarbuscular space remains to be

determined.



are of great agricultural potential. The use of recombinant

inbred lines in white clover is a promising approach to

identify some the responsible genes [66].

Plant gene expressionOver the past few years, significant efforts have been

directed towards the identification of genes that are

differentially regulated in root symbiosis. These efforts

comprise EST sequencing from symbiotic tissues [16,67]

and the development of online database systems, such as

MtDB, that allow in-silico gene expression analysis [68]

and hybridisation of cDNA arrays [54,69]. Hundreds of

novel nodulin and mycorrhizin genes that are expressed

in response to rhizobia or AM fungi, respectively, have

been identified. In addition, the increasing number of

identified plant genes that are activated by both fungal

and bacterial symbionts (symbiosins) [70] has the poten-

tial to provide information about the functional overlap

between the root symbioses. In the future, reverse

genetic approaches, such as RNA interference (RNAi)

[40��] or TILLING [71�], will be crucial to determine the

relevance of these genes in symbiosis.

Conclusions and outlookWe are experiencing exciting times in the genetic analysis

of arbuscular mycorrhiza. The cloning of novel symbiosis

genes from plant mutants that have been identified

through forward genetic screens is proving to be a very

successful strategy for the identification of signalling

components. Predictions of the functions of these com-

ponents are allowing researchers to map out the pathways

leading to the intracellular accommodation of a fungal

symbiont. Despite the large amount of research directed

towards plant resistance against intracellular fungal

pathogens such as downy mildew and rust fungi, little

is known about the plant genes that are required for

compatibility in these systems. Over the past two years,

the arbuscular mycorrhiza has become probably the best-

understood biotrophic plant–fungus interaction, thanks to

the power of genetics.

AcknowledgementsWe thank Eva Wegel for preparation of figures, and Sonja Kosuta,Katharina Pawlowski, Katharina Markmann and Thilo Winzer forcomments on the manuscript. Research at The Sainsbury Laboratoryis funded by the Gatsby Charitable Foundation.

References and recommended readingPapers of particular interest, published within the annual period ofreview, have been highlighted as:

� of special interest��of outstanding interest

1. Newman EI, Reddell P: The distribution of mycorrhizasamong families of vascular plants. New Phytol 1987,106:745-751.

2. Bago B, Pfeffer PE, Abubaker J, Jun J, Allen JW, Brouillette J,Douds DD, Lammers PJ, Shachar-Hill Y: Carbon export fromarbuscular mycorrhizal roots involves the translocationof carbohydrate as well as lipid. Plant Physiol 2003,131:1496-1507.

3. Bago B, Pfeffer PE, Shachar-Hill Y: Carbon metabolism andtransport in arbuscular mycorrhizas. Plant Physiol 2000,124:949-958.

4. Smith SE, Read DJ: Mycorrhizal Symbiosis edn 2nd. London:Academic Press; 1997.

5. Remy W, Taylor TN, Hass H, Kerp H: Four hundred-million-year-old vesicular arbuscular mycorrhizae. Proc Natl Acad Sci USA1994, 91:11841-11843.

6. Redecker D, Kodner R, Graham LE: Glomalean fungi from theOrdovician. Science 2000, 289:1920-1921.

7. Schussler A, Schwarzott D, Walker C: A new fungal phylum, theGlomeromycota: phylogeny and evolution. Mycol Res 2001,105:1413-1421.

8. Hildebrandt U, Janetta K, Bothe H: Towards growth of arbuscularmycorrhizal fungi independent of a plant host. 2002,68:1919-1924.

9. Kuhn G, Hijri M, Sanders IR: Evidence for the evolution ofmultiple genomes in arbuscular mycorrhizal fungi. Nature 2001,414:745-748.

10. Pawlowska TE, Taylor JW: Organization of genetic variation inindividuals of arbuscular mycorrhizal fungi. Nature 2004,427:733-737.

11. Hijri M, Sanders IR: The arbuscular mycorrhizal fungus Glomusintraradices is haploid and has a small genome size in the lowerlimit of eukaryotes. Fungal Genet Biol 2004, 41:253-261.

12. Jun J, Abubaker J, Rehrer C, Pfeffer PE, Shachar-Hill Y,Lammers PJ: Expression in an arbuscular mycorrhizal fungusof genes putatively involved in metabolism, transport, thecytoskeleton and the cell cycle. Plant Soil 2002, 244:141-148.

13. Sawaki H, Saito M: Expressed genes in the extraradical hyphaeof an arbuscular mycorrhizal fungus, Glomus intraradices, inthe symbiotic phase. FEMS Microbiol Lett 2001, 195:109-113.

14. Stommel M, Mann P, Franken P: EST-library construction usingspore RNA of the arbuscular mycorrhizal fungus Gigasporarosea. Mycorrhiza 2001, 10:281-286.

15. Delp G, Timonen S, Rosewarne GM, Barker SJ, Smith S:Differential expression of Glomus intraradices genes inexternal mycelium and mycorrhizal roots of tomato and barley.Mycol Res 2003, 107:1083-1093.

16. Wulf A, Manthey K, Doll J, Perlick AM, Linke B, Bekel T, Meyer F,Franken P, Kuster H, Krajinski F: Transcriptional changes inresponse to arbuscular mycorrhiza development in the modelplant Medicago truncatula. Mol Plant Microbe Interact 2003,16:306-314.

17. Requena N, Mann P, Hampp R, Franken P: Early developmentallyregulated genes in the arbuscular mycorrhizal fungus Glomusmosseae: identification of GmGIN1, a novel gene withhomology to the C-terminus of metazoan hedgehog proteins.Plant Soil 2002, 244:129-139.

18. Lammers PJ, Jun J, Abubaker J, Arreola R, Gopalan A, Bago B,Hernandez-Sebastia C, Allen JW, Douds DD, Pfeffer PE et al.: Theglyoxylate cycle in an arbuscular mycorrhizal fungus. Carbonflux and gene expression. Plant Physiol 2001, 127:1287-1298.

19. Harrier LA, Millam S: Biolistic transformation of arbuscularmycorrhizal fungi — progress and perspectives.Mol Biotechnol 2001, 18:25-33.

20. David-Schwartz R, Badani H, Smadar W, Levy AA, Galili G,Kapulnik Y: Identification of a novel genetically controlled step inmycorrhizal colonization: plant resistance to infection by fungalspores but not extra-radical hyphae. Plant J 2001, 27:561-569.

21. David-Schwartz R, Gadkar V, Wininger S, Bendov R, Galili G,Levy AA, Kapulnik Y: Isolation of a premycorrhizal infection(pmi2) mutant of tomato, resistant to arbuscular mycorrhizalfungal colonization. Mol Plant Microbe Interact 2003, 16:382-388.

22. Duc G, Trouvelot A, Gianinazzi-Pearson V, Gianinazzi S: Firstreport of non-mycorrhizal plant mutants (Myc–) obtained in pea(Pisum sativum L.) and fababean (Vicia faba L.). Plant Sci 1989,60:215-222.



23. Bradbury SM, Bowley SR: Interaction between three alfalfanodulation genotypes and two Glomus species.New Phytol 1991, 119:115-120.

24. Sagan M, Morandi D, Tarenghi E, Duc G: Selection of nodulationand mycorrhizal mutants in the model plant Medicagotrunculata (Gaertn.) after c-ray mutagenesis. Plant Sci 1995,111:63-71.

25. Shirtliffe SJ, Vessey JK: A nodulation (NodR/Fix–) mutant ofPhaseolus vulgaris L. has nodule-like structures lackingperipheral vascular bundles (Pvb–) and is resistant tomycorrhizal infection (Myc–). Plant Sci 1996, 118:209-220.

26. Schauser L, Handberg K, Sandal N, Stiller J, Thykjær T, Pajuelo E,Nielsen A, Stougaard J: Symbiotic mutants deficient in noduleestablishment identified after T-DNA transformation of Lotusjaponicus. Mol Gen Genet 1998, 259:414-423.

27. Szczyglowski K, Shaw RS, Wopereis J, Copeland S, Hamburger D,Kasiborski B, Dazzo FB, de Bruijn FJ: Nodule organogenesis andsymbiotic mutants of the model legume Lotus japonicus.Mol Plant Microbe Interact 1998, 11:684-697.

28. Wegel E, Schauser L, Sandal N, Stougaard J, Parniske M:Mycorrhiza mutants of Lotus japonicus define geneticallyindependent steps during symbiotic infection. Mol PlantMicrobe Interact 1998, 11:933-936.

29. Catoira R, Galera C, De Billy F, Penmetsa RV, Journet EP, Maillet F,Rosenberg C, Cook D, Gough C, Denarie J: Four genes ofMedicago truncatula controlling components of a Nod factortransduction pathway. Plant Cell 2000, 12:1647-1666.

30. Senoo K, Solaiman MZ, Kawaguchi M, Imaizumi-Anraku H, Akao S,Tanaka A, Obata H: Isolation of two different phenotypes ofmycorrhizal mutants in the model legume plant Lotusjaponicus after EMS-treatment. Plant Cell Physiol 2000,41:726-732.

31. Kawaguchi M, Imaizumi-Anraku H, Koiwa H, Niwa S, Ikuta A,Syono K, Akao S: Root, root hair, and symbiotic mutants ofthe model legume Lotus japonicus. Mol Plant MicrobeInteract 2002, 15:17-26.

32. Lum MR, Li Y, LaRue TA, David-Schwartz R, Kapulnik Y, Hirsch AM:Investigation of four classes of non-nodulating whitesweetclover (Melilotus alba annua Desr.) mutants and theirresponses to arbuscular-mycorrhizal fungi. Integr Comp Biol2002, 42:295-303.

33. Kistner C, Parniske M: Evolution of signal transduction inintracellular symbiosis. Trends Plant Sci 2002, 7:511-518.

34. Marsh JF, Schultze M: Analysis of arbuscular mycorrhizasusing symbiosis-defective plant mutants. New Phytol 2001,150:525-532.

35. Oldroyd GED, Downie JA: Calcium, kinases and nodulationsignalling in legumes. Nat Rev Mol Cell Biol 2004.

36. LaRue TA, Weeden NF: The symbiosis genes of the host. InProceedings of the 1st European Nitrogen Fixation Conference.Edited by Kiss GB, Endre G: Officina; 1994:147-151.

37.��

Stracke S, Kistner C, Yoshida S, Mulder L, Sato S, Kaneko T,Tabata S, Sandal N, Stougaard J, Szczyglowski K et al.:A plant receptor-like kinase required for both fungal andbacterial symbiosis. Nature 2002, 417:959-962.

The authors of this paper, together with Endre et al. [43��], describe thecloning of the first common SYM gene. The Lotus SYMRK receptor kinaseis orthologous to Pea PsSYM19 and is involved in early symbioticsignalling.

38.��

Madsen EB, Madsen LH, Radutoiu S, Olbryt M, Rakwalska M,Szczyglowski K, Sato S, Kaneko T, Tabata S, Sandal N et al.:A receptor kinase gene of the LysM type is involved in legumeperception of rhizobial signals. Nature 2003, 425:637-640.

See annotation for [39��].

39.��

Radutoiu S, Madsen LH, Madsen EB, Felle HH, Umehara Y,Gronlund M, Sato S, Nakamura Y, Tabata S, Sandal N et al.:Plant recognition of symbiotic bacteria requires two LysMreceptor-like kinases. Nature 2003, 425:585-592.

The work described in this report, together with that described byLimpens et al. [40��] and Madsen et al. [38��], constitutes a scientific

breakthrough: the cloning of the Nod-factor receptor kinases! A moredetailed appreciation of this achievement can be found in [41].

40.��

Limpens E, Franken C, Smit P, Willemse J, Bisseling T, Geurts R:LysM domain receptor kinases regulating rhizobial Nodfactor-induced infection. Science 2003.

RNAi-mediated silencing of candidate Nod-factor receptors that origi-nated from a positional cloning approach showed that M. truncatulahomologues of the NFR1 gene are required for infection-thread devel-opment. See also [38��,39��,41].

41. Parniske M, Downie JA: Plant biology: locks, keys andsymbioses. Nature 2003, 425:569-570.

42.��

Kosuta S, Chabaud M, Lougnon G, Gough C, Denarie J, Barker DG,Becard G: A diffusible factor from arbuscular mycorrhizal fungiinduces symbiosis-specific MtENOD11 expression in roots ofMedicago truncatula. Plant Physiol 2003, 131:952-962.

The work described in this paper provides experimental evidence of along-postulated signalling molecule from AM fungi. AM fungi can induceplant gene activation from a distance when separated from host roots bymembranes that are impassable for hyphae.

43.��

Endre G, Kereszt A, Kevei Z, Mihacea S, Kalo P, Kiss GB:A receptor kinase gene regulating symbiotic noduledevelopment. Nature 2002, 417:962-966.

The authors carried out the map-based cloning of a symbiosis gene fromMedicago sativa, a particularly difficult assignment as this plant is tetra-ploid. The receptor kinase that they identified, NODULATION RECEP-TOR-LIKE KINASE (NORK), is orthologous to M. truncatula DMI2, peaSYM19 and Lotus SYMRK [37��].

44.��

Ane JM, Kiss GB, Riely BK, Penmetsa RV, Oldroyd GE, Ayax C,Levy J, Debelle F, Baek JM, Kalo P et al.: Medicago truncatulaDMI1 required for bacterial and fungal symbioses in legumes.Science 2004, 303:1364-1367.

The DMI1 gene encodes a predicted ion channel that is essential forsymbiosis. Interestingly, a homologous gene has been found in twobacteria, Mesorhizobium loti, a bacterium that forms root nodules withLotus japonicus, and Streptomyces, suggesting the possibility of hor-izontal gene transfer from plants to a limited number of bacteria.

45.��

Levy J, Bres C, Geurts R, Chalhoub B, Kulikova O, Duc G,Journet EP, Ane JM, Lauber E, Bisseling T et al.: A putativeCa2R and calmodulin-dependent protein kinase required forbacterial and fungal symbioses. Science 2004, 303:1361-1364.

The kinase that is cloned in the study is potentially able to interpret thecalcium-spiking response observed in root hairs after Nod-factor treat-ment, and to translate it into a phosphorylation event as a readout. Thiskinase is also required for AM, which opens up the possibility that calciumspiking plays a role in fungal-mediated signalling.

46.��

Mitra RM, Gleason CA, Edwards A, Hadfield J, Downie JA,Oldroyd GE, Long SR: A Ca2R/calmodulin-dependent proteinkinase required for symbiotic nodule development: geneidentification by transcript-based cloning. Proc Natl AcadSci USA 2004, 101:4701-4705.

The authors used a candidate-gene approach aided by transcriptionalprofiling to identify the DMI3 gene, which was independently cloned usinga map-based strategy by [45��].

47. Chabaud M, Venard C, Defaux-Petras A, Becard G, Barker DG:Targeted inoculation of Medicago truncatula in vitro rootcultures reveals MtENOD11 expression during early stages ofinfection by arbuscular mycorrhizal fungi. New Phytol 2002,156:265-273.

48. Giovannetti MC, Sbrana LA, Citernesi AS, Logi C: Differentialhyphal morphogenesis in arbuscular mycorrhizal fungi duringpre-infection stages. New Phytol 1993, 125:587-593.

49. Buee M, Rossignol M, Jauneau A, Ranjeva R, Becard G: The pre-symbiotic growth of arbuscular mycorrhizal fungi is induced bya branching factor partially purified from plant root exudates.Mol Plant Microbe Interact 2000, 13:693-698.

50. Journet EP, El-Gachtouli N, Vernoud V, de Billy F, Pichon M,Dedieu A, Arnould C, Morandi D, Barker DG, Gianinazzi-Pearson V:Medicago truncatula ENOD11: a novel RPRP-encoding earlynodulin gene expressed during mycorrhization in arbuscule-containing cells. Mol Plant Microbe Interact 2001, 14:737-748.

51.�

Demchenko K, Winzer T, Stougaard J, Parniske M, Pawlowski K:Distinct roles of Lotus japonicus SYMRK and SYM15 in rootcolonization and arbuscule formation. New Phytol 2004, in press.



A detailed cytological analysis of the symbiotic phenotype of LotusLjSYM15 mutants detected a new plant function in AM development:the separation of epidermal cell walls to allow the trespass of fungalhyphae (Figure 3).

52.�

Novero M, Faccio A, Genre A, Stougaard J, Webb KJ, Mulder L,Parniske M, Bonfante P: Dual requirement of the LjSym4 gene formycorrhizal development in epidermal and cortical cells ofLotus japonicus roots. New Phytol 2002, 154:741-749.

Weak mutant alleles of the Lotus LjSYM4 gene revealed that this gene isrequired for the symbiotic response of not only outer cell layers but also ofinner cortical cells (Figure 3).

53. Nagahashi G, Douds DD: Appressorium formation by AMfungi on isolated cell walls of carrot roots. New Phytol 1997,136:299-304.

54. Liu J, Blaylock LA, Endre G, Cho J, Town CD, VandenBosch KA,Harrison MJ: Transcript profiling coupled with spatialexpression analyses reveals genes involved in distinctdevelopmental stages of an arbuscular mycorrhizal symbiosis.Plant Cell 2003, 15:2106-2123.

55. Bonfante P, Genre A, Faccio A, Martini I, Schauser L, Stougaard J,Webb J, Parniske M: The Lotus japonicus LjSym4 gene isrequired for the successful symbiotic infection of rootepidermal cells. Mol Plant Microbe Interact 2000,13:1109-1120.

56. Gianinazzi-Pearson V: Plant cell responses to arbuscularmycorrhizal fungi: getting to the roots of the symbiosis.Plant Cell 1996, 8:1871-1883.

57. Guinel FC, Geil RD: A model for the development of the rhizobialand arbuscular mycorrhizal symbioses in legumes and its useto understand the roles of ethylene in the establishment ofthese two symbioses. Can J Bot 2002, 80:695-720.

58. Gianinazzi-Pearson V, Gollotte A, Lherminier J, Tisserant B,Franken P, Dumas-Gaudot E, van Tuinen D, Gianinazzi S: Cellularand molecular approaches in the characterization of symbioticevents in functional arbuscular mycorrhizal associations.Can J Bot 1995, 73:S526-S532.

59.�

Harrison MJ, Dewbre GR, Liu J: A phosphate transporter fromMedicago truncatula involved in the acquisition of phosphatereleased by arbuscular mycorrhizal fungi. Plant Cell 2002,14:2413-2429.

A Medicago phosphate transporter was immunolocalised to the peripheryof arbuscules, suggesting that it functions in phosphate uptake from theperiarbuscular space (Figure 4).

60. Nishimura R, Hayashi M, Wu GJ, Kouchi H, Imaizumi-Anraku H,Murakami Y, Kawasaki S, Akao S, Ohmori M, Nagasawa M et al.:HAR1 mediates systemic regulation of symbiotic organdevelopment. Nature 2002, 420:426-429.

61. Krusell L, Madsen LH, Sato S, Aubert G, Genua A, Szczyglowski K,Duc G, Kaneko T, Tabata S, de Bruijn F et al.: Shoot control of rootdevelopment and nodulation is mediated by a receptor-likekinase. Nature 2002, 420:422-426.

62. Searle IR, Men AE, Laniya TS, Buzas DM, Iturbe-Ormaetxe I,Carroll BJ, Gresshoff PM: Long-distance signaling in nodulationdirected by a CLAVATA1-like receptor kinase. Science 2003,299:109-112.

63. Solaiman MZ, Senoo K, Kawaguchi M, Imaizumi-Anraku H, Akao S,Tanaka A, Obata H: Characterization of mycorrhizas formedby Glomus sp. on root of hypernodulating mutants ofLotus japonicus. J Plant Res 2000, 113:443-448.

64. Staehelin C, Charon C, Boller T, Crespi M, Kondorosi A: Medicagotruncatula plants overexpressing the early nodulin geneenod40 exhibit accelerated mycorrhizal colonization andenhanced formation of arbuscules. Proc Natl Acad Sci USA2001, 98:15366-15371.

65. Jacobi LM, Zubkova LA, Barmicheva EM, Tsyganov VE,Borisov AY, Tikhonovich IA: Effect of mutations in the pea genesSym33 and Sym40. II. Dynamics of arbuscule developmentand turnover. Mycorrhiza 2003, 13:9-16.

66. Eason WR, Webb KJ, Michaelson-Yeates TPT, Abberton MT,Griffith GW, Culshaw CM, Hooker JE, Dhanoa MS: Effect ofgenotype of Trifolium repens on mycorrhizal symbiosis withGlomus mosseae. J Agric Sci 2001, 137:27-36.

67. Journet EP, van Tuinen D, Gouzy J, Crespeau H, Carreau V,Farmer MJ, Niebel A, Schiex T, Jaillon O, Chatagnier O et al.:Exploring root symbiotic programs in the model legumeMedicago truncatula using EST analysis. Nucleic Acids Res2002, 30:5579-5592.

68. Lamblin AF, Crow JA, Johnson JE, Silverstein KA, Kunau TM,Kilian A, Benz D, Stromvik M, Endre G, VandenBosch KA et al.:MtDB: a database for personalized data mining of the modellegume Medicago truncatula transcriptome. Nucleic Acids Res2003, 31:196-201.

69. Colebatch G, Kloska S, Trevaskis B, Freund S, Altmann T,Udvardi MK: Novel aspects of symbiotic nitrogen fixationuncovered by transcript profiling with cDNA arrays.Mol Plant Microbe Interact 2002, 15:411-420.

70. van Rhijn P, Fang Y, Galili S, Shoul O, Atzmon N, Wininger S,Eshed Y, Lum M, Li Y, To V et al.: Expression of early nodulingenes in alfalfa mycorrhizae indicates that signal transductionpathways used in forming arbuscular mycorrhizae andRhizobium-induced nodules may be conserved. Proc Natl AcadSci USA 1997, 94:5467-5472.

71.�

Perry JA, Wang TL, Welham TJ, Gardner S, Pike JM, Yoshida S,Parniske M: A TILLING reverse genetics tool and a web-accessible collection of mutants of the legume Lotusjaponicus. Plant Physiol 2003, 131:866-871.

This paper describes the construction and exploitation of the first TILLINGfacility for a legume. Lotus plants with point mutations in any gene ofinterest can be isolated from a ethyl methanesulfonate (EMS)-mutagen-ised population.

72. Harrison M, van Buuren M: A phosphate transporter from themycorrhizal fungus Glomus versiforme. Nature 1995,378:626-629.

73. Maldonado-Mendoza IE, Dewbre GR, Harrison MJ: A phosphatetransporter gene from the extra-radical mycelium of anarbuscular mycorrhizal fungus Glomus intraradices isregulated in response to phosphate in the environment.Mol Plant Microbe Interact 2001, 14:1140-1148.

74.�

Rausch C, Daram P, Brunner S, Jansa J, Laloi M, Leggewie G,Amrhein N, Bucher M: A phosphate transporter expressed inarbuscule-containing cells in potato. Nature 2001, 414:462-470.

Arbuscule-expressed phosphate transporters are likely to play a key rolein AM function. This is the first paper to describe a plant gene thatencodes such a transporter.

75.�

Paszkowski U, Kroken S, Roux C, Briggs SP: Rice phosphatetransporters include an evolutionarily divergent genespecifically activated in arbuscular mycorrhizal symbiosis.Proc Natl Acad Sci USA 2002, 99:13324-13329.

Transcriptional profiling of rice during AM development using an Affyme-trix-oligonucleotide chip led to the identification of phosphate transpor-ters that are specifically expressed in arbuscule-containing cells.

76. Schauser L, Roussis A, Stiller J, Stougaard J: A plant regulatorcontrolling development of symbiotic root nodules.Nature 1999, 402:191-195.

77. Borisov AY, Madsen LH, Tsyganov VE, Umehara Y, Voroshilova VA,Batagov AO, Sandal N, Mortensen A, Schauser L, Ellis N et al.: Thesym35 gene required for root nodule development in pea is anortholog of nin from Lotus japonicus. Plant Physiol 2003,131:1009-1017.



Documents

Molecular genetics of the arbuscular mycorrhizal symbiosis