1531RESEARCH ARTICLE

INTRODUCTIONWhen land was colonised, plant morphology significantlydiversified (Kenrick and Crane, 1997b; Langdale, 2008). Theunderlying developmental mechanisms of this diversificationduring evolution are still poorly understood. Shoot branching is oneof the key innovations that led to more complex architecture.Recently, strigolactones were shown to represent a novel class ofplant hormones controlling this process (Gomez-Roldan et al.,2008; Umehara et al., 2008). The same molecules act in therhizosphere as signalling molecules in establishing the symbioticrelationship between arbuscular mycorrhizal fungi and over 80%of extant land plants (Akiyama et al., 2005; Besserer et al., 2006)and in stimulating seed germination of root parasitic weeds(Bouwmeester et al., 2007). Arbuscular mycorrhizal symbiosisevolved for 460 million years before the appearance of roots andis coincident with the appearance of the first bryophyte-like landplants (Remy et al., 1994). In seed plants, strigolactone-deficientmutants in pea (Fig. 1A), Arabidopsis thaliana and rice show anincreased branching phenotype (Gomez-Roldan et al., 2008;Umehara et al., 2008). To gain a better understanding of the originof branching in seed plants and to explore strigolactone functionsin early land plants, we investigated whether strigolactones areproduced in the moss Physcomitrella patens and what roles theyplay in a non vascular plant.

Strigolactones are apocarotenoids (Matusova et al., 2005) andthe first step in their biosynthesis is catalysed by CAROTENOIDCLEAVAGE DIOXYGENASES (CCD). The CCDs are an ancient

family of enzymes with members identified in bacteria, plants andanimals (Auldridge et al., 2006). In Arabidopsis and pea, the CCD7and CCD8 enzymes are encoded by the branching genesMAX3/RMS5 (Booker et al., 2004; Johnson et al., 2006) andMAX4/RMS1 (Sorefan et al., 2003), respectively, and are bothinvolved in strigolactone biosynthesis (Gomez-Roldan et al., 2008;Umehara et al., 2008).

The model moss Physcomitrella patens provides unique geneticand developmental facilities to undertake evolutionary studies inplant biology (Cove et al., 2006; Lang et al., 2008; Schaefer, 2002).Its life cycle is dominated by the haploid gametophyte that displaysa succession of simple and well described developmentaltransitions that leads from the germinating spore to the colony ofdifferentiated leafy shoots carrying mature diploid sporophytes(Cove et al., 2006).

We generated a mutant in the moss homolog of the strigolactonebiosynthesis gene CCD8. Here, we demonstrate that in P. patens,PpCCD8 is involved in strigolactone biosynthesis and regulates thebranching of filament and colony extension. Furthermore, our datademonstrate that in wild-type (WT) P. patens, secretedstrigolactones are directly involved in the regulation of colonyextension in response to internal cues or population density. Thisnovel function of strigolactones in plants is reminiscent of quorum-sensing signalling processes that are well described in bacteria(Fuqua and Greenberg, 2002; Ng and Bassler, 2009), and wepropose that moss uses strigolactones as quorum-sensing moleculesto regulate colonial growth in distinct environmental conditions.

MATERIALS AND METHODSP. patens growth conditions and treatmentsThe Gransden wild-type strain was used and conditions for cultures wereas described (Ashton and Cove, 1977; Ashton et al., 1979; Trouiller et al.,2006). For phenotypic analysis, spores were grown on minimal PP-NO3

medium (Ashton et al., 1979). Sporogenesis was performed in Magentavessels in which 15-day-old wild-type and mutant colonies were grown for2 months, then irrigated with sterile water and transferred to growth

Development 138, 1531-1539 (2011) doi:10.1242/dev.058495© 2011. Published by The Company of Biologists Ltd

1Institut Jean-Pierre Bourgin, UMR1318 INRA-AgroParisTech, Centre de Versailles-Grignon, 78026 Versailles Cedex, France. 2Weed Science Center, UtsunomiyaUniversity, Utsunomiya 321-8505, Japan. 3Institute of Biology, Laboratory of Cell andMolecular Biology, University of Neuchatel, CH-2009 Neuchatel, Switzerland.

*Author for correspondence (catherine.rameau@versailles.inra.fr)

Accepted 28 January 2011

SUMMARYStrigolactones are a novel class of plant hormones controlling shoot branching in seed plants. They also signal host root proximityduring symbiotic and parasitic interactions. To gain a better understanding of the origin of strigolactone functions, wecharacterised a moss mutant strongly affected in strigolactone biosynthesis following deletion of the CAROTENOID CLEAVAGEDIOXYGENASE 8 (CCD8) gene. Here, we show that wild-type Physcomitrella patens produces and releases strigolactones into themedium where they control branching of protonemal filaments and colony extension. We further show that Ppccd8 mutantcolonies fail to sense the proximity of neighbouring colonies, which in wild-type plants causes the arrest of colony extension. Themutant phenotype is rescued when grown in the proximity of wild-type colonies, by exogenous supply of synthetic strigolactonesor by ectopic expression of seed plant CCD8. Thus, our data demonstrate for the first time that Bryophytes (P. patens) producestrigolactones that act as signalling factors controlling developmental and potentially ecophysiological processes. We proposethat in P. patens, strigolactones are reminiscent of quorum-sensing molecules used by bacteria to communicate with one another.

KEY WORDS: Physcomitrella, Strigolactones, CCD8

Strigolactones regulate protonema branching and act as aquorum sensing-like signal in the moss Physcomitrella patensHélène Proust1, Beate Hoffmann1, Xiaonan Xie2, Kaori Yoneyama2, Didier G. Schaefer3, Koichi Yoneyama2,Fabien Nogué1 and Catherine Rameau1,*

DEVELO

PMENT

1532

chambers at 15°C with 8 hours of light per day and a quantum irradianceof 15 E m–2 s–1. Maturation of spore capsules was followed visually. Thedifferentiation of caulonemata and buds was followed using a binocular(Nikon SMZ1000). For a precise estimation of caulonema differentiation,colonies of different ages were transferred in darkness on minimal mediumsupplemented with 5 g/l glucose for 11 days. Because of negativegravitropism, Petri dishes were kept upside-down in the dark for an easyobservation and counting of caulonemal filaments. In the cross-feedingexperiment, the central colony was transferred to the medium 10 days afterthe four surrounding colonies. In the density experiment, to eliminate aputative starvation effect, the experiment was carried out on large Petridishes (12 cm) and glucose (0.5%) was added in another similarexperiment (not shown). Cell measurements were performed with ImageJsoftware (http://rsb.info.nih.gov/ij/).

Phylogenetic analysisEvolutionary relationships between CCD proteins were analysed using aselection of proteins from different databases aligned with the programCLUSTALX (Thompson et al., 1997). The evolutionary history wasinferred using the Neighbour-Joining method (Saitou and Nei, 1987) andthe optimal tree was generated. Phylogenetic analyses were conducted inMEGA4 (Tamura et al., 2007).

PrimersThe primers used in this study are described in Table S1 in thesupplementary material.

Molecular cloningPpccd8-KO constructA 2855 bp PCR amplified PpCCD8 genomic fragment (scaffold 14:2616535-2619389) was cloned into vector pCRII Blunt TOPO(Invitrogen). To disrupt the PpCCD8 gene, a 664 bp 3� targeting fragment(scaffold 14: 2616135-2616798) was first recovered by enzyme digestionwith XmnI and BamHI, and inserted into BamHI- and SmaI-digestedpBNRF, which carries a loxP-neoR-loxP selectable marker (Schaefer et al.,2010). The 925 bp 5� targeting fragment (scaffold 14: 2618287-2619211),obtained by digestion with AcyI and BseJI, was then inserted between theClaI/HpaI sites of vector pBNRF carrying the 3� targeting fragment togenerate pPpccd8-KO (see Fig. S1 in the supplementary material).

OsACTpro-108-HygR

The CaMVter terminator fragment was excised from pDH51 (Pietrzak etal., 1986) using XbaI/HindIII (HindIII site was blunted) and ligated topCOR 104 (McElroy et al., 1991) digested by XbaI/SacII (SacII wasblunted) to generate pCOR104-CaMVter. A 1800 pb fragment carrying therice ACTIN-1 gene promoter (GenBank: S44221) and the CaMVter wasobtained by digestion of pCOR104-CaMVter vector with XhoI and PvuIIenzymes. This fragment was cloned into SwaI and SalI sites of BS-108vector which contains a genomic P. patens fragment corresponding to locus108 (Schaefer and Zryd, 1997) giving the OsACTpro-108 vector. The35S::hygroR cassette (Hajdukiewicz et al., 1994) was excised of pBKS-hygro vector with HindIII and EcoRI enzymes, and integrated in theHindIII site of p-Act-108 vector to generate OsACTpro-108-HygR (see Fig.S2A in the supplementary material).

OsACTpro:RMS1 constructThe pea RMS1 cDNA (GenBank: AY557341) obtained by PCR wassubcloned in pTOPO vector and inserted in NcoI/SphI sites of OsACTpro-108-H (see Fig. S2A,B in the supplementary material).

PpCCD8pro:GUS constructThis construct was made in the OsACTpro-108-H vector. A 1616 bpamplified fragment of the PpCCD8 promoter (scaffold 14: 2619186-2620801) was cloned behind the GUS:Noster (Fourgoux-Nicol et al., 1999)cassette of the pAF1 vector digested with SacII/HindIII. A fragmentcontaining PpCCD8pro:GUS:Noster was excised using SacII and NcoI,blunted and used to replace the deleted rice actin promoter CaMVtercassette of the OsACTpro-108-H vector with KpnI sites (see Fig. S2E inthe supplementary material).

Generation of Ppccd8, Ppccd8::OsACTpro:RMS1 andPpCCD8pro:GUS strainsMoss protoplasts were transformed and transgenic strains were selected asdescribed previously (Trouiller et al., 2006). Transformants were selectedon G418 (50 g/ml) (Duchefa GO 175, Haarlem, The Netherlands). Stabledisruptants of the PpCCD8 gene were confirmed by PCR using primersspecific for the PpCCD8 sequence outside of the transformation construct.A clean deletion in the PpCCD8 gene was obtained by transient expressionof the Cre recombinase as described previously (Trouiller et al., 2006). Therecombinant locus was analysed by PCR with the same primers (see Fig.S1 in the supplementary material). Several independent deletion strainswith similar phenotype were obtained and one of them, Ppccd8, was usedfor subsequent analyses.

OsACTpro:RMS1 and PpCCD8pro:GUS were linearised with SacIIbefore transformation, and transformants were selected on hygromycin (25g/ml). To confirm the presence of constructs in stable transformants, PCRamplification with specific primers designed on sequences inside and outsidethe construct was performed (see Fig. S2B-F in the supplementary material).

Quantification of strigolactones released by wild type andPpccd8� mutantApproximately 1 g of 7-day-old protonemata from the wild type orPpccd8� mutant was used to inoculate 150 ml of BCD medium in a 300ml flask on a rotary shaker. Cultures were grown under standard lightconditions for 30 days and the growth medium was replaced with freshmedia weekly. Thirty-day-old tissue was transferred to fresh media andgrown for an additional 3 days. Plant material was collected by filtrationand weighed, while the liquid culture medium was collected and extractedthree times with an equal volume of ethyl acetate. The combined ethylacetate solutions were washed with 0.2 M K2HPO4 for removal of acidiccompounds, dried over anhydrous MgSO4, filtered and concentrated invacuo. These crude extracts were stored in sealed glass vials at 4°C untiluse. Strigolactones released to liquid culture media were determined onLC-MS/MS as described previously (Yoneyama et al., 2008). Forquantitative analysis using LC-MS/MS, orobanchol-d1, orobanchyl acetate-d1 and [13C2]strigol were used as internal standards. Levels of otherstrigolactones were determined as reported previously (Yoneyama et al.,2008) and their recovery rates were estimated based on that of orobanchylacetate-d1. The experiments were conducted with three replicate samples.

GUS staining of Physcomitrella patensColonies developed from ground tissues were incubated for 12 hours at37°C in 0.8 mM 5-bromo-4-chloro-3-indolyl--D-glucuronide, 2 mMpotassium ferrocyanide, 2 mM potassium ferricyanide and 100 mMphosphate buffer (pH 7).

Gene expression analysisTotal RNA was extracted using the RNeasy plant mini kit (Qiagen. Ref74904). cDNA was synthesised from 2 g of total RNA with RevertAid HMinus M-MuLV Reverse Transcriptase (Fermentas) and dT(18) primer.Quantitative PCR was performed in a Mastercycle ep realplex (Eppendorf)with Realmaster mix SYBR ROX 2.5� (5 PRIME). PpAPT, which encodesan ADENOSINE PHOSPHORIBOSYL TRANSFERASE, was used as theconstitutive gene (Trouiller et al., 2006). Relative expression of PpCCD wasestablished with the formula 2–Ct(PpCCD)–Ct(PpAPT). Three biologicalreplicates were carried out with two technical replicates for each sample.

RESULTSCharacterisation of P. patens strigolactonebiosynthesis genesUnique CCD7-like and CCD8-like sequences were found in thegenome of P. patens that were designated PpCCD7 (GenBankAccession Number HM007803) and PpCCD8 (GenBankAccession Number HM007802). The predicted moss proteinsshare 45% and 57% amino acid identity with CCD7 andCCD8 from Arabidopsis, respectively. Phylogenetic analysesdemonstrated that they grouped with their respective AtCCD

RESEARCH ARTICLE Development 138 (8)

DEVELO

PMENT

clades (see Fig. S3 in the supplementary material). Two othergenes involved in strigolactone biosynthesis have been identifiedby genetic analyses: MAX1 in Arabidopsis, which encodes amember of the CYP450 family; CYP711A1 (Booker et al.,2005); and the rice D27 gene, which encodes an iron-containingprotein (Lin et al., 2009). No MAX1 homologue was identifiedin the P. patens genome, whereas several homologues for D27were found, one (XP001767010) encoding a protein with 47%of identity with D27 from rice.

To characterise the function of the CCD8 gene in P. patens, wegenerated a Ppccd8 mutant in which genomic sequencesencoding part of exon V, exons VI and VII were deleted. RT-PCRanalyses confirmed the absence of a full-length PpCCD8 transcriptin the mutant, showing that the Ppccd8 strain is a null allele (seeFig. S1 in the supplementary material). The Ppccd8 strain wasfertile and we used its selfed progeny to analyse single spore-derived colonies in all the experiments described below. Eachcolony should, thus, be considered as an individual sibling and notas a clonal replicate.

Ppccd8 strains display several developmentalphenotypesFollowing spore germination, P. patens colonies develop intophotosynthetic primary chloronemal filaments that furtherdifferentiate into caulonemal filaments implicated in rapid radial

extension of the colony (Fig. 1B). Most caulonemal subapical cellsfurther form secondary filaments and some buds (Fig. 1B) that willdevelop into leafy gametophores. The developmental fate of thesecaulonemal side branch initials is rather constant in P. patens, withmost of them developing into secondary chloronemata and only afew percentage into new caulonemata or buds (Cove et al., 2006).

Under our growth conditions, spore germination occurredsignificantly earlier in Ppccd8 spores than in the wild type (seeFig. S4A in the supplementary material) and was associated with ahigher frequency of multipolar germination patterns (Fig. 1C,D).Branching of primary chloronemata was also higher in the mutantduring this early developmental stage. The percentage of lateralcells was higher in the mutant (37.2%) than in the wild type(10.8%) 93 hours after spore transfer on the medium. This increasein lateral cell number is not due to early germination of thePpccd8 mutant as the percentage was still higher for the mutantat 81 hours (25.6%) (see Fig. S4B in the supplementary material).The first caulonemal cells differentiated ~7 days after sporegermination, followed by the formation of the first buds at day 15;timing of chloronemata to caulonemata transition was similar inboth wild type and Ppccd8 mutant with a higher number ofcaulonemata in the mutant very likely to be due to higher branchingof chloronemal cells (see Fig. S4C in the supplementary material).Fifteen days after spore germination, branching (i.e. production ofsecondary filaments or new buds) of caulonemata protruding from

1533RESEARCH ARTICLEStrigolactones function in moss

Fig. 1. CCD8 controls shoot branching in pea and the synthesis of a diffusible signal regulating caulonema branching and colonyextension in P. patens. (A)Wild-type (left) and rms1 (Psccd8) (right) pea plants. (B)Caulonemata (ca), secondary chloronemata (ch) and bud (bud)from a 18-day-old wild-type colony. (C,D)Development of wild-type (C) and Ppccd8� (D) colonies, 56 to 106 hours after spore transfer to the medium.(E-H)Wild-type (E,F) and Ppccd8� (G,H) colony, periphery of 25-day-old colonies (E,G), 40-day-old colonies (F,H). (I-L)Cross-feeding experiment with a30-day-old central colony surrounded by four 40-day-old colonies. Scale bars: 200m in B; 100m in C,D; 500m in E,G; 1 cm in F,H-L.

DEVELO

PMENT

1534

the colonies was similar in both genotypes (see Fig. S4F in thesupplementary material). Until day 21, diameter of mutant colonies(measured from one caulonema extremity to the other on theopposite side of the colony) are comparable with the diameter ofwild-type colonies (see Fig. S4G in the supplementary material).Moreover, size of chloronemal cells for Ppccd8 is notsignificantly different from wild type and length of caulonemalcells is slightly higher in Ppccd8 compared with wild type (seeFig. S4D,E in the supplementary material). Taken together, theseresults suggested that rate of apical cell division was not increasedin the Ppccd8 mutant and that the higher branching observed inthe mutant was not linked with a general higher cell division rate.A sudden and significant difference appeared 20 days after sporegermination: the diameter of wild-type colonies reached a plateau,while mutant colonies continued to extend (Fig. 1F,H; see Fig. S4Gin the supplementary material). At this stage, all caulonemal-typefilaments, a mixture of caulonemata and rhizoids, protruding fromthe wild-type colonies were unbranched (Fig. 1E) and had stoppedto elongate. By contrast, elongation and branching of Ppccd8caulonemata continued up to 45 days after spore germinationgiving rise to much larger colonies (Fig. 1G; see Fig. S4F,G in thesupplementary material).

In both genotypes, the first buds appeared ~8 days aftercaulonema differentiation. Size of gametophores was affected inPpccd8 compared with wild type. Mutant gametophores were inaverage half the size of wild type, had reduced number of leavesand shorter internodes after 45 days (see Fig. S4H-J in thesupplementary material). Yet leaves were normally shaped anddisplayed normal phyllotaxy, and the general architecture of theleafy shoot was unaffected. Most importantly, we did not observeany increase in gametophore branching, and all Ppccd8gametophores were formed by a single stem as in wild type (datanot shown). Gametophores bearing a capsule were less frequent inthe mutant (0.55 sporophyte/gametophore in wild type versus 0.04in Ppccd8). Nevertheless, capsules from mutant contained viablespores but in reduced numbers compared with wild type (see Fig.S4K,L in the supplementary material).

These results indicated that while cells and organs forminggametophyte and sporophyte were not altered, except in size,control of protonema branching and colony extension was stronglyaffected in Ppccd8.

A diffusible signal secreted by wild-type colonycomplements Ppccd8 phenotypeIn P. patens, hormones such as auxins and cytokinins are secretedin the medium to control the protonemal stage of gametophytedevelopment (Cove et al., 2006). To assess whether the PpCCD8gene is involved in the biosynthesis of a signal secreted into themedium, we performed a series of cross-feeding assays.

After 2 weeks of co-culture, the Ppccd8 colony surrounded bywild-type colonies displayed a wild-type phenotype with a strongdecrease in colony diameter (Fig. 1I-L; see Fig. S5 in thesupplementary material). These results suggested that wild-typemoss produces a signal that diffuses into the medium andcomplements the phenotype of Ppccd8. This signal inhibitscolony extension and filament branching. The mutant does notappear to be producing this inhibitor that accounts for itsdevelopmental phenotype.

When a wild-type colony was surrounded by Ppccd8 colonies,a slight but significant increase in the diameter of the central wild-type colony was observed (Student’s t-test, P<0.01) (see Fig. S5 inthe supplementary material). The central wild-type colony also had

an effect on the surrounding Ppccd8 colonies. The mutant coloniesgrew asymmetrically owing to reduced extension on the side closestto the wild-type colony (Fig. 1K). The difference in diameter of thewild-type colony when surrounded by either WT or mutant coloniesalso indicated that the colony stopped extending once the signalreached a certain threshold. This is attained faster when surroundingcolonies also participate in the production of the signal.

Feedback regulation of PpCCD7 and PpCCD8In several seed plants, CCD7 and CCD8 transcript levels areupregulated in strigolactone-deficient mutants suggesting thatnegative feedback controls strigolactone biosynthesis (Beveridgeand Kyozuka, 2010). In rice, application of the syntheticstrigolactone GR24 to the strigolactone deficient ccd8 (d10) mutantrestored CCD8 transcript levels to wild type (Umehara et al.,2008).

Expression of PpCCD8 in the wild type was first investigated byquantitative RT-PCR. The expression level of PpCCD8 was lowbut detectable 10 days after spore germination and increasedapproximately eightfold between days 15 and 21 to reach a plateau(see Fig. S6A in the supplementary material). These results showthat PpCCD8 is expressed at low levels in the protonemata and isupregulated along with the differentiation of new gametophoreswhere it is more abundantly expressed (see below).

Quantitative RT-PCR further revealed that PpCCD7 displayed asimilar expression profile in both wild-type and Ppccd8 colonies.Yet PpCCD7 transcript levels were five to 20 times higher inPpccd8 than in wild type (see Fig. S6A,B in the supplementarymaterial). This observation suggests that, in Physcomitrella, thetwo CCD genes are co-regulated with feedback control ofPpCCD7, as previously observed in seed plants (Beveridge andKyozuka, 2010).

To gain more insight in the regulation of PpCCD8 expression,the -glucuronidase (GUS) reporter gene driven by PpCCD8promoter sequence was introduced into locus 108 in wild-type andPpccd8 strains. Locus 108 is often used as target sequence as itdoes not contain any coding sequence (Schaefer and Zryd, 1997).GUS expression driven by the PpCCD8 promoter was notdetectable in wild type by histochemical staining (data not shown),indicating that the expression level of PpCCD8 was very low. Bycontrast, GUS staining was visible following overnight incubationin the Ppccd8 background, and was observed in everydifferentiated bud in the protonemata and at the foot of eachgametophore (Fig. 2A,B). These data indicate that PpCCD8expression is also regulated by feedback control, as observed abovefor the PpCCD7 transcript. They further indicate that the biologicalactivity of PpCCD8 is developmentally regulated and spatiallyconcentrated at the base of gametophores.

Strigolactone identification and production inwild type and Ppccd8The presence of strigolactones in Physcomitrella was tested furtherwith the analysis of liquid medium in which differentiated wild-type and Ppccd8 tissue were grown for 3 days. As shown forseveral seed plants, P. patens exuded a mixture of strigolactones(see Fig. S7A,B in the supplementary material). The majorstrigolactones produced by the wild type were 7-oxoorobanchylacetate, orobanchyl acetate, 7-hydroxyorobanchyl acetate, fabacylacetate, strigol and orobanchol (Fig. 3A). The mutant producedonly strigol and 7-oxoorobanchyl acetate (Fig. 3A) along withsmall amounts of 5-deoxystrigol, which could not be quantified ineither strains. Wild type produced and released approximately 350

RESEARCH ARTICLE Development 138 (8)

DEVELO

PMENT

pg/g fresh weight (FW) of strigolactones in 3 days, while themutant produced 40 pg/g FW, less than one-eighth that of wildtype. These data provide the first characterisation of strigolactonesin a moss. They further demonstrate that strigolactone biosynthesisis severely affected in Ppccd8, indicating the involvement ofPpCCD8 in strigolactone biosynthesis in P. patens and the potentialinvolvement of the strigolactone pool in the developmentalphenotype described above.

Ppccd8 phenotype is complemented byexpression of pea CCD8 and by exogenousstrigolactone applicationTo examine whether expression of CCD8 genes from seed plantscan complement the Ppccd8 phenotype, the pea CCD8 gene(RMS1) driven by the rice Actin-1 gene promoter was introducedat locus 108 in the Ppccd8 mutant. Several independentPpccd8::OsACTpro:RMS1 strains were obtained that exhibitedthe same wild-type developmental pattern: colony extensionstopped after 3 weeks and colonies developed leafy shoots ofnormal size. Furthermore, these strains perfectly restored normaldevelopment of Ppccd8 colonies in cross-feeding experiments(see Fig. S8A,B in the supplementary material). Consistently, semi-quantitative RT-PCR indicated that the expression of PpCCD7 wasreduced to wild-type levels in Ppccd8::OsACTpro:RMS1 strains(see Fig. S8C in the supplementary material). These results providegenetic evidence that pea RMS1 is able to restore both normaldevelopment and feedback control of PpCCD7 expression inPpccd8 mutant. These data suggest that pea and moss CCD8 havethe same biochemical activity that has been conserved during landplant evolution.

To determine whether the synthetic strigolactone GR24 couldcomplement the Ppccd8 developmental phenotype, 15-day-oldPpccd8 colonies were transferred on media containing 1 M ofGR24. After 2 weeks, these colonies showed an intermediatediameter and were surrounded by unbranched caulonema-typefilaments (Fig. 3B-G). In a dose-response experiment, thediameters of both wild-type and Ppccd8 colonies weresignificantly smaller after 23 days on medium supplemented with100 nM or 1 M GR24 (test of ANOVA, P<0.01, n12; see Fig.S9 in the supplementary material). At the same stage, we observedin both strains that gametophores were larger than in the non-supplemented controls (Fig. 3B-D). To assess whether GR24 wasable to restore feedback control on PpCCD7 expression, weanalysed its expression in 21-day-old wild-type and Ppccd8colonies during the 2 days following transfer on media containing

1 M GR24. These analyses showed that GR24 downregulatedPpCCD7 expression in Ppccd8 (Fig. 3H). To determine whetherthe wild-type diffusible signal identified in cross-feedingexperiments also regulates PpCCD7 expression, we analysedPpccd8 samples that were transferred on media where wild-typeor Ppccd8 colonies had grown previously. The same downregulation of PpCCD7 expression was observed in Ppccd8 tissuetreated with wild-type preconditioned medium (Fig. 3H).Altogether, these data show that GR24 almost completelycomplements the developmental phenotype of Ppccd8 mutant, the

1535RESEARCH ARTICLEStrigolactones function in moss

Fig. 2. PpCCD8 is expressed at the base of each gametophore.(A)PpCCD8pro:GUS in a 21-day-old Ppccd8� colony. (B)Younggametophore (enlargement of A as indicated). Scale bars: 500m in A;100m in B. Arrowheads indicate bud or gametophore.

Fig. 3. The signal that regulates colony extension in P. patens is astrigolactone. (A)Quantification of strigolactones released by wildtype (open bars) and Ppccd8� (black bars) in the growth media for 3days. Data are presented as mean±s.e.m. (n3).(B-G)Complementation of Ppccd8� by 1M GR24 application. Twenty-five-day-old colonies (B-D). Periphery of the same colonies (E-G).ch, chloronema side branches. Scale bars: 1 cm in B-D; 1 mm in E-G.(H)Expression of PpCCD7 in wild-type (white) and Ppccd8� colonies(21-day-old) 24 and 48 hours after transfer to new media containing0M (black), 1M GR24 (blue) or to media on which wild-type (lightgrey) or Ppccd8� colonies (dark grey) were grown for 21 days. Errorbars indicate s.e.m., n3. Letters indicate means that differ significantly(P<0.01, ANOVA test).

DEVELO

PMENT

1536

most striking feature being the complete arrest of secondarychloronemata formation at the edge of the colonies. They furtherdemonstrate that both GR24 and the wild-type diffusible signaldownregulate PpCCD7 expression.

Colony extension responds to colony density inwild-type moss but not in Ppccd8To analyse further the threshold effect of signal accumulation in themedium, we tested if by varying the density of colonies, thediameter of wild-type colonies changed and examined the effect onmutant colonies. We followed the evolution of wild-type andPpccd8 colonies inoculated at four different densities (Fig. 4A-D). At high densities, the Ppccd8 colonies started to merge at 28days and at 39 days, the surface of the Petri dish was completelycovered by mutant colonies (Fig. 4C,D). Consequently, after 28days we were unable to measure colony diameter for Ppccd8. Bycontrast, at 39 days, regardless of the initial colony density, wild-type colonies showed a well-delimited round surface (Fig. 4A-D).ANOVA confirmed a significant effect of density on diameter at 28days in wild type (P<0.001) but not in Ppccd8. The diameter ofwild-type colonies was significantly higher at density 2 and 4colonies per plate, and reached a plateau at 39 days for the twohighest densities (Fig. 4E). These data show that in P. patens,strigolactones accumulate in the medium as the number ofgametophores increases to a threshold concentration to controlcolony extension in response to colony density.

DISCUSSIONConservation of strigolactone biosynthesis amongland plantsThe first steps in strigolactone biosynthesis involve the cleavageof carotenoids by CCD7 and CCD8 (Beveridge and Kyozuka,2010; Leyser, 2008). The moss genome contains one orthologuefor each corresponding gene and our phylogenetic analysissuggests that the CCD7 and CCD8 clades evolved before theseparation of the bryophytes and the vascular plants from acommon ancestor 420 million years ago (Kenrick and Crane,1997b). Mutations in either of these two genes are associatedwith strigolactone deficiency in pea and rice (Gomez-Roldan etal., 2008; Umehara et al., 2008). Our analysis of strigolactoneproduction in wild type and Ppccd8 demonstrates that PpCCD8is similarly involved in strigolactone biosynthesis in moss. Inseed plants, the cytochrome P450 MAX1 is necessary for thesynthesis of the shoot branching inhibitor, and graftingexperiments indicated that it acts downstream of the two CCDs(Booker et al., 2005). Although there is no MAX1 homologue inthe moss genome, P. patens produces strigolactones that werepreviously identified in the exudates of seed plants. Thissuggests that MAX1 functions in later steps of the activebranching inhibiting hormone biosynthesis, i.e. conversion ofstrigolactones to the shoot branching inhibitor. Hence, thebioactive signal in P. patens and seed plants may be different.Alternatively, moss may have other P450(s) able to replace thefunction of MAX1 for strigolactone biosynthesis.

The Ppccd8 mutant still produced strigolactones, whereas nostrigolactones were detected by LC-MS in the ccd8 mutants of peaand rice (Gomez-Roldan et al., 2008; Umehara et al., 2008),suggesting that another strigolactone biosynthesis pathway may bepresent in moss. Hence, seven CCD genes other than PpCCD8 arepresent in the genome of Physcomitrella (see Fig. S3 in thesupplementary material) that may be able to cleave carotenoids toproduce small amounts of strigolactones. Alternatively, as root

exudates from the ccd8 mutants of pea and rice did inducegermination of root parasitic weed seeds, it could suggest that thesemutants may also still produce small amounts of strigolactones(Gomez-Roldan et al., 2008; Umehara et al., 2008).

Strong upregulation of CCD7 and CCD8 transcription isgenerally observed in strigolactone-deficient mutants (Dun et al.,2009). Our work revealed that transcription levels of both PpCCD7and PpCCD8 genes were also strongly upregulated in Ppccd8,indicating that the transcriptional feedback regulation that controlsthis biosynthetic pathway is probably conserved among land plants.We demonstrated that expression of pea CCD8 fully complementsthe developmental phenotype of Ppccd8, which providesevidence for functional complementation of CCD8 proteins inthese distant species. These results showed that the first steps ofstrigolactone biosynthesis are conserved among Embryophytes.However, strigolactone biosynthetic pathway is not fullycharacterised, and even in seed plants the biochemical function ofMAX1, for example, is still unknown. A better understanding ofthe final steps of the strigolactone biosynthesis pathway shouldhighlight difference and similarity between land plants.

Developmental and colonial defects ofstrigolactone deficient P. patensDetailed analysis of the strigolactone-deficient mutant showed thatPpCCD8 controls moss development from the germination stagewith an earlier spore germination, higher number of primarychloronemata developing from the spore and higher branching ofchloronemata in Ppccd8. This indicates that the strigolactonepathway is already involved in the spore development to controlspecific cellular process. The apparent different phenotypes may

RESEARCH ARTICLE Development 138 (8)

Fig. 4. Strigolactones control colony extension in response tocolony density. (A-D)Thirty-nine-day-old wild-type (top panels) andPpccd8� (below panels) colonies grown at different densities. Scalebars: 1 cm. (E)Diameter of wild-type (blue lines) and Ppccd8� (greenlines) colonies grown in light at different densities, two colonies (white),four colonies (light grey), nine colonies (dark grey) and 15 colonies(black) (n12). Data are mean±s.e.m.

DEVELO

PMENT

result from the same cellular mechanism of higher rates of lateralcell differentiation and/or division and would need furtherinvestigation.

The subsequent development of colonies was similar in bothwild type and Ppccd8 until 3 weeks after germination, indicatingthat strigolactones are not required to establish the normaldevelopmental pattern of colonies. This is consistent with theabsence of detectable morphological alteration in the mutant.

Under our growth conditions, and for a density of 15 to 20colonies per Petri dish, wild-type colonies stop to extent and todifferentiate a branched protonema after 3 weeks. By contrast,Ppccd8 colonies maintain their extension with a continuousproduction of secondary chloronemata and buds on caulonemata.The decrease in gametophore size, in number of gametophorebearing a capsule and in spore number per capsule observed instrigolactone-deficient Ppccd8 should be seen as a trophicresponse to higher intra-colony competition for resourcesbetween gametophores. Similarly, in seed plants, the highbranching of strigolactone-deficient ccd8 mutants results in anincrease in the number of flowers per plant with a concomitantdecrease in flower size and seed number per plant (Snowden etal., 2005).

Strigolactone activity in moss are reminiscent ofquorum-sensing signal in bacteriaIn wild type, the arrest of colony extension is sudden and involvesa coordinated response from the whole colony. This response isdevelopmentally associated with a complete arrest of caulonemabranching and elongation. By contrast, despite producing a smallamount of strigolactones, Ppccd8 colonies continue to extend. Allthese results suggest that wild-type colonies released strigolactonesor derived molecules into the medium to control colony extension.These data also indicate that a minimal threshold must be reachedfor the colony to respond to strigolactones and imply thatstrigolactones are synthesised and degraded in media as thisthreshold is never reached by Ppccd8 colonies. Strigolactones arecharacterised by a tricyclic lactone (ABC part) connected, via anenol ether bridge, to a butyrolactone group (the D-ring) (Yoneyamaet al., 2009). They are known to be short-lived in the rhizosphereowing to this labile ether bond that spontaneously hydrolyses inwater (Akiyama and Hayashi, 2006; Akiyama et al., 2010). Thisinstability creates a steep concentration gradient that has beensuggested to be a reliable indicator of the proximity of a host root(Parniske, 2008).

Our analyses further demonstrate that strigolactones act asdiffusible signals that regulate colonial extension in response topopulation densities (Fig. 4A-D). In wild type, the final diameterof 45-day-old colonies is negatively correlated with the number ofcolonies per plate and this correlation completely disappears forPpccd8 colonies (Fig. 4E). Based on these features, as well as onthe fact that strigolactones are low molecular weight molecules thatdiffuse in the medium, we propose that strigolactones behave asquorum sensing-like molecules. Quorum sensing is a common, ifnot universal, cell-to-cell communication process in the microbialworld. It is mediated by low molecular weight molecules, the mostcommon signals in bacteria being acyl-homoserine-lactones (Ngand Bassler, 2009), that accumulate in the environment dependingon the density of the organisms synthesising these molecules(Fuqua and Greenberg, 2002). When a quorum-sensing signalmolecule reaches a threshold concentration, i.e. a threshold celldensity, the entire population responds, usually through thecoordinated expression of specific target genes. By analogy, we

propose that when a ‘quorum’ of strigolactone producinggametophores is reached, the whole colony and the neighbouringcolonies respond by halting their extension.

The idea of chemical interactions between mosses regulatingspreading of colonies is a long-standing observation for peopleworking with mosses. A diffusible signal called ‘Factor H’, firstidentified in Funaria hygrometrica, was postulated to account forthe maintenance of a clear zone between moss colonies or clonesthat did not involve competition for nutrients (Bopp, 1959; Klein,1967). Factor H, which may be a complex of substances, wasknown to inhibit caulonema growth and lateral spread and topromote bud formation (Bopp and Knoop, 1974; Klein, 1967).Watson (Watson, 1981) proposed that such diffusible substanceshave developmental as well as ecological implications and may bepossible determinant of moss community structure in nature.Strigolactone could be good candidates for the caulonema growthand lateral spread inhibition effect of the factor H. However,strigolactones have no inductive effect on bud formation and ratherfavour the simultaneous growth of gametophores alreadydifferentiated in the population. Thus, more extensive work isneeded to asses whether strigolactones could be one component ofthe factor H.

Possible role of strigolactone in early land plantevolutionThe colonisation of land by plants is associated with a largenumber of developmental, morphological, anatomical and chemicalinnovations (Kenrick and Crane, 1997b; Langdale, 2008).Arbuscular mycorrhizal symbiosis is an association between one ofthe oldest group of fungi, the Glomales, and plants. The fungalpartner provides nutrients, essentially phosphate, and water to theplant that brings carbohydrates to the fungi (Parniske, 2008). Thissymbiosis has been found in almost all major lineages of extantland plants and is considered as one of the major innovations thatcontributed to the colonisation of land by plants (Kenrick andCrane, 1997b; Renzaglia et al., 2000). Mycorrhization is tightlyassociated with strigolactone biosynthesis (Akiyama et al., 2005;Akiyama et al., 2010; Gomez-Roldan et al., 2008). Despite the factthat P. patens, like most mosses, is not mycorrhizal, otherbryophytes, liverworts that are considered the earliest-diverginglineage of land plants, and hornworts are mycorrhizal (Read et al.,2000; Wang and Qiu, 2006; Wang et al., 2010). The evolutionaryhomology of symbiotic interactions between gametophytes ofliverworts and hornworts and mycorrhizal fungi, and those betweensporophytes of vascular plants and the fungi has been demonstratedrecently (Wang et al., 2010). These data and our demonstration thatstrigolactones are signalling molecules in the bryophyte P. patenssupport the hypothesis that strigolactones are very ancientmolecules that have played a crucial role in the ecologicaladaptation of plants to terrestrial environment. Further studies withMarchantia, a mycorrhizal liverwort recently used in biologicalstudies (Bowman et al., 2007; Chiyoda et al., 2008; Ishizaki et al.,2008,) would confirm the role of strigolactones in the establishmentof arbuscular mycorrizal symbiosis in bryophytes.

Land plants (embryophytes) evolved from streptophyte greenalgae, a small group of freshwater algae ranging from unicellularflagellates (Mesostigma) to complex filamentous thalli withbranching (Charales) (Becker and Marin, 2009). No CCD7 orCCD8 homologues have been found in streptophyte green algaeEST databases, which could mean that strigolactones arose veryearly in the land plant lineage. However, no complete genome iscurrently available for streptophyte green algae (Becker and Marin,

1537RESEARCH ARTICLEStrigolactones function in moss

DEVELO

PMENT

1538

2009; Timme and Delwiche, 2010) and whether they are able tosynthesize strigolactones would give some clue on the earliestrole(s) of strigolactones during plant evolution.

In this work, we have shown that strigolactones repressprotonema branching in P. patens that is the differentiation of alateral cell from subapical cells. Branching in seed plants involvedseveral steps from the initiation of axillary meristems at the axilsof leaves, their subsequent development in axillary buds that willremain in a dormant state or will outgrow to produce a branch.Strigolactones control specifically the maintenance in a dormantstate of this particular organ, the axillary bud. It is quite intriguingthat strigolactones control branching of the gametophyte in mossand branching of the sporophyte in seed plants. P. patens genomeanalysis and its comparison with the genome of Arabidopsis havesuggested that genes controlling gametophyte development havebeen recruited for sporophyte development (Nishiyama et al., 2003;Rensing et al., 2008). Whether common molecular actors areinvolved to control gametophyte (protonema) or sporophytebranching in response to strigolactones will be revealed by futurestudies. However the term ‘branching’ can be misleading and, forexample, it is very likely that sporophyte branching due todichotomy observed in many early taxa, involved differentmechanisms from what is currently known in seed plants (Kenrickand Crane, 1997a; Langdale, 2008). Thus, more data will beneeded in different taxa in order to decipher branching mechanismsin plants (Boyce, 2010). Nevertheless, this work is a first step inthe understanding of the evolution of branching patterns andcontributes to the wonderful story of how a small family ofmolecules has been co-opted during evolution to control chemicalinteractions between organisms and whole-plant architecture.

AcknowledgementsWe thank F. D. Boyer for provided GR24, K. Mori for strigol, T. Asami for [2,3-13C2]strigol, and K. Akiyama for [6’-2H]orobanchol and [6’-2H]orobanchylacetate. H.P. was supported by a PhD fellowship from the Ministère del’Education Nationale, de la Recherche et de la Technologie (MENRT). Workfrom the C.R. group was supported by ANR (Agence Nationale de laRecherche, project ANR.05.BLAN.0262), and work from the K.Y. group wassupported by KAKANHI (18208010, 2109111) and Program for Promotion ofBasic and Applied Researches for Innovations in Bio-oriented Industry. Wethank Y. Dessaux for helpful discussions; H. Höfte, Y. Dessaux, G. Pelletier andS. Bonhomme for their comments on the manuscript; and M. Sawabe for herassistance.

Competing interests statementThe authors declare no competing financial interests.

Supplementary materialSupplementary material for this article is available athttp://dev.biologists.org/lookup/suppl/doi:10.1242/dev.058495/-/DC1

ReferencesAkiyama, K. and Hayashi, H. (2006). Strigolactones: chemical signals for fungal

symbionts and parasitic weeds in plant roots. Ann. Bot. (Lond.) 97, 925-931.Akiyama, K., Matsuzaki, K. and Hayashi, H. (2005). Plant sesquiterpenes

induce hyphal branching in arbuscular mycorrhizal fungi. Nature 435, 824-827.Akiyama, K., Ogasawara, S., Ito, S. and Hayashi, H. (2010). Structural

requirements of strigolactones for hyphal branching in AM fungi. Plant CellPhysiol. 51, 1104-1117.

Ashton, N. W. and Cove, D. J. (1977). The isolation and preliminarycharacterisation of auxotrophic and analogue resistant mutants of the moss,Physcomitrella patens. Mol. Gen. Genet. 154, 87-95.

Ashton, N. W., Grimsley, N. H. and Cove, D. J. (1979). Analysis of gametophytedevelopment in the moss, Physcomitrella patens, using auxin and cytokininresistant mutants. Planta 144, 427-435.

Auldridge, M. E., McCarty, D. R. and Klee, H. J. (2006). Plant carotenoidcleavage oxygenases and their apocarotenoid products. Curr. Opin. Plant Biol. 9,315-321.

Becker, B. and Marin, B. (2009). Streptophyte algae and the origin ofembryophytes. Ann. Bot. 103, 999-1004.

Besserer, A., Puech-Pages, V., Kiefer, P., Gomez-Roldan, V., Jauneau, A., Roy,S., Portais, J. C., Roux, C., Becard, G. and Sejalon-Delmas, N. (2006).Strigolactones stimulate arbuscular mycorrhizal fungi by activating mitochondria.PLoS Biol. 4, e226.

Beveridge, C. A. and Kyozuka, J. (2010). New genes in the strigolactone-relatedshoot branching pathway. Curr. Opin. Plant Biol. 13, 34-39.

Booker, J., Auldridge, M., Wills, S., McCarty, D., Klee, H. and Leyser, O.(2004). MAX3/CCD7 is a carotenoid cleavage dioxygenase required for thesynthesis of a novel plant signaling molecule. Curr. Biol. 14, 1232-1238.

Booker, J., Sieberer, T., Wright, W., Williamson, L., Willett, B., Stirnberg, P.,Turnbull, C., Srinivasan, M., Goddard, P. and Leyser, O. (2005). MAX1encodes a cytochrome P450 family member that acts downstream of MAX3/4 toproduce a carotenoid-derived branch-inhibiting hormone. Dev. Cell 8, 443-449.

Bopp, M. (1959). Versuche zur Analyse von Wachstum and Differenzierung desLaubmoosprotonemas. Planta 53, 178.

Bopp, M. and Knoop, B. (1974). Régulation de la différenciation chez leprotonéma des Mousses. Bull. Soc. Bot. France 121, 145.

Bouwmeester, H. J., Roux, C., Lopez-Raez, J. A. and Becard, G. (2007).Rhizosphere communication of plants, parasitic plants and AM fungi. TrendsPlant Sci. 12, 224-230.

Bowman, J. L., Floyd, S. K. and Sakakibara, K. (2007). Green genes-comparative genomics of the green branch of life. Cell 129, 229-234.

Boyce, C. K. (2010). The evolution of plant development in a paleontologicalcontext. Curr. Opin. Plant Biol. 13, 102-107.

Chiyoda, S., Ishizaki, K., Kataoka, H., Yamato, K. T. and Kohchi, T. (2008).Direct transformation of the liverwort Marchantia polymorpha L. by particlebombardment using immature thalli developing from spores. Plant Cell Rep. 27,1467-1473.

Cove, D., Bezanilla, M., Harries, P. and Quatrano, R. (2006). Mosses as modelsystems for the study of metabolism and development. Annu. Rev. Plant Biol.57, 497-520.

Dun, E. A., Brewer, P. B. and Beveridge, C. A. (2009). Strigolactones: discoveryof the elusive shoot branching hormone. Trends Plant Sci. 14, 364-372.

Fourgoux-Nicol, A., Drouaud, J., Haouazine, N., Pelletier, G. and Guerche, P.(1999). Isolation of rapeseed genes expressed early and specifically duringdevelopment of the male gametophyte. Plant Mol. Biol. 40, 857-872.

Fuqua, C. and Greenberg, E. P. (2002). Listening in on bacteria: acyl-homoserinelactone signalling. Nat. Rev. Mol. Cell Biol. 3, 685-695.

Gomez-Roldan, V., Fermas, S., Brewer, P. B., Puech-Pages, V., Dun, E. A.,Pillot, J. P., Letisse, F., Matusova, R., Danoun, S., Portais, J. C. et al. (2008).Strigolactone inhibition of shoot branching. Nature 455, 189-194.

Hajdukiewicz, P., Svab, Z. and Maliga, P. (1994). The small, versatile pPZP familyof Agrobacterium binary vectors for plant transformation. Plant Mol. Biol. 25,989-994.

Ishizaki, K., Chiyoda, S., Yamato, K. T. and Kohchi, T. (2008). Agrobacterium-mediated transformation of the haploid liverwort Marchantia polymorpha L., anemerging model for plant biology. Plant Cell Physiol. 49, 1084-1091.

Johnson, X., Brcich, T., Dun, E. A., Goussot, M., Haurogne, K., Beveridge, C.A. and Rameau, C. (2006). Branching genes are conserved across species.Genes controlling a novel signal in pea are coregulated by other long-distancesignals. Plant Physiol. 142, 1014-1026.

Kenrick, P. and Crane, P. R. (1997a). The Origin and Early Diversification of landPlants. Washington: Smithsonian Institution Press.

Kenrick, P. and Crane, P. R. (1997b). The origin and early evolution of plants onland. Nature 389, 33-39.

Klein, B. (1967). Versuche sur Analyse der Protonemaentwicklung der Laubmoose.IV. Mitteilung. Der endogene Faktor H und seine Rolle bei der Morphogenesevon Funaria hygrometrica. Planta 73, 12.

Lang, D., Zimmer, A. D., Rensing, S. A. and Reski, R. (2008). Exploring plantbiodiversity: the Physcomitrella genome and beyond. Trends Plant Sci. 13, 542-549.

Langdale, J. A. (2008). Evolution of developmental mechanisms in plants. Curr.Opin. Genet. Dev. 18, 368-373.

Leyser, O. (2008). Strigolactones and shoot branching: a new trick for a youngdog. Dev. Cell 15, 337-338.

Lin, H., Wang, R., Qian, Q., Yan, M., Meng, X., Fu, Z., Yan, C., Jiang, B., Su,Z., Li, J. et al. (2009). DWARF27, an iron-containing protein required for thebiosynthesis of strigolactones, regulates rice tiller bud outgrowth. Plant Cell 21,1512-1525.

Matusova, R., Rani, K., Verstappen, F. W., Franssen, M. C., Beale, M. H. andBouwmeester, H. J. (2005). The strigolactone germination stimulants of theplant-parasitic Striga and Orobanche spp. are derived from the carotenoidpathway. Plant Physiol. 139, 920-934.

McElroy, D., Blowers, A. D., Jenes, B. and Wu, R. (1991). Construction ofexpression vectors based on the rice actin 1 (Act1) 5� region for use in monocottransformation. Mol. Gen. Genet. 231, 150-160.

Ng, W. L. and Bassler, B. L. (2009). Bacterial quorum-sensing networkarchitectures. Annu. Rev. Genet. 43, 197-222.

Nishiyama, T., Fujita, T., Shin, I. T., Seki, M., Nishide, H., Uchiyama, I.,Kamiya, A., Carninci, P., Hayashizaki, Y., Shinozaki, K. et al. (2003).

RESEARCH ARTICLE Development 138 (8)

DEVELO

PMENT

Comparative genomics of Physcomitrella patens gametophytic transcriptomeand Arabidopsis thaliana: implication for land plant evolution. Proc. Natl. Acad.Sci. USA 100, 8007-8012.

Parniske, M. (2008). Arbuscular mycorrhiza: the mother of plant rootendosymbioses. Nat. Rev. Microbiol. 6, 763-775.

Pietrzak, M., Shillito, R. D., Hohn, T. and Potrykus, I. (1986). Expression inplants of two bacterial antibiotic resistance genes after protoplast transformationwith a new plant expression vector. Nucleic Acids Res. 14, 5857-5868.

Read, D. J., Ducket, J. G., Francis, R., Ligron, R. and Russell, A. (2000).Symbiotic fungal associations in ‘lower’ land plants. Philos. Trans. R. Soc. Lond. BBiol. Sci. 355, 815-831.

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

Rensing, S. A., Lang, D., Zimmer, A. D., Terry, A., Salamov, A., Shapiro, H.,Nishiyama, T., Perroud, P. F., Lindquist, E. A., Kamisugi, Y. et al. (2008). ThePhyscomitrella genome reveals evolutionary insights into the conquest of land byplants. Science 319, 64-69.

Renzaglia, K. S., Duff, R. J., Nickrent, D. L. and Garbary, D. J. (2000).Vegetative and reproductive innovations of early land plants: implications for aunified phylogeny. Philos. Trans. R. Soc. Lond. B Biol. Sci. 355, 769-793.

Saitou, N. and Nei, M. (1987). The neighbor-joining method: a new method forreconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406-425.

Schaefer, D. G. (2002). A new moss genetics: targeted mutagenesis inPhyscomitrella patens. Annu. Rev. Plant Biol. 53, 477-501.

Schaefer, D. G. and Zryd, J. P. (1997). Efficient gene targeting in the mossPhyscomitrella patens. Plant J. 11, 1195-1206.

Schaefer, D. G., Delacote, F., Charlot, F., Vrielynck, N., Guyon-Debast, A., LeGuin, S., Neuhaus, J. M., Doutriaux, M. P. and Nogue, F. (2010). RAD51 lossof function abolishes gene targeting and de-represses illegitimate integration inthe moss Physcomitrella patens. DNA Repair (Amst.) 9, 526-533.

Snowden, K. C., Simkin, A. J., Janssen, B. J., Templeton, K. R., Loucas, H. M.,Simons, J. L., Karunairetnam, S., Gleave, A. P., Clark, D. G. and Klee, H. J.(2005). The decreased apical dominance1/Petunia hybrida CAROTENOIDCLEAVAGE DIOXYGENASE8 gene affects branch production and plays a role inleaf senescence, root growth, and flower development. Plant Cell 17, 746-759.

Sorefan, K., Booker, J., Haurogne, K., Goussot, M., Bainbridge, K., Foo, E.,Chatfield, S., Ward, S., Beveridge, C., Rameau, C. et al. (2003). MAX4 andRMS1 are orthologous dioxygenase-like genes that regulate shoot branching inArabidopsis and pea. Genes Dev. 17, 1469-1474.

Tamura, K., Dudley, J., Nei, M. and Kumar, S. (2007). MEGA4: MolecularEvolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24,1596-1599.

Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. and Higgins, D.G. (1997). The CLUSTAL_X windows interface: flexible strategies for multiplesequence alignment aided by quality analysis tools. Nucleic Acids Res. 25, 4876-4882.

Timme, R. E. and Delwiche, C. F. (2010). Uncovering the evolutionary origin ofplant molecular processes: comparison of Coleochaete (Coleochaetales) andSpirogyra (Zygnematales) transcriptomes. BMC Plant Biol. 10, 96.

Trouiller, B., Schaefer, D. G., Charlot, F. and Nogue, F. (2006). MSH2 is essentialfor the preservation of genome integrity and prevents homeologousrecombination in the moss Physcomitrella patens. Nucleic Acids Res. 34, 232-242.

Umehara, M., Hanada, A., Yoshida, S., Akiyama, K., Arite, T., Takeda-Kamiya, N., Magome, H., Kamiya, Y., Shirasu, K., Yoneyama, K. et al.(2008). Inhibition of shoot branching by new terpenoid plant hormones. Nature455, 195-200.

Wang, B. and Qiu, Y. L. (2006). Phylogenetic distribution and evolution ofmycorrhizas in land plants. Mycorrhiza 16, 299-363.

Wang, B., Yeun, L. H., Xue, J. Y., Liu, Y., Ane, J. M. and Qiu, Y. L. (2010).Presence of three mycorrhizal genes in the common ancestor of land plantssuggests a key role of mycorrhizas in the colonization of land by plants. NewPhytol. 186, 514-525.

Watson, M. A. (1981). Chemically mediated interactions among juvenile mossesas possible determinants of their community structure. J. Chem. 7, 367-376.

Yoneyama, K., Xie, X., Sekimoto, H., Takeuchi, Y., Ogasawara, S., Akiyama,K. and Hayashi, H. (2008). Strigolactones, host recognition signals for rootparasitic plants and arbuscular mycorrhizal fungi, from Fabaceae plants. NewPhytol. 179, 484-494.

Yoneyama, K., Xie, X. and Takeuchi, Y. (2009). Strigolactones: structures andbiological activities. Pest Manag. Sci. 65, 467-470.

1539RESEARCH ARTICLEStrigolactones function in moss

DEVELO

PMENT