Rhizobium meliloti Elicits Transient Expression of

The Plant Cell, Vol. 4, 1199-1211, October 1992 O 1992 American Society of Plant Physiologists

Rhizobium meliloti Elicits Transient Expression of the Early Nodulin Gene ENOD12 in the Differentiating Root Epidermis of Transgenic Alfalfa Magalie Pichon, Etienne-Pascal Journet, Annie Dedieu, Françoise de Billy, Georges Truchet, and David G. Barker’ Laboratoire de Biologie Moléculaire des Relations Plantes Microorganismes, INRA-CNRS, BP27, 31326 CastanetTolosan Ckdex, France

To study the molecular responses of the host legume during early stages of the symbiotic interaction with Rhbobium, we have cloned and characterized the infection-related early nodulin gene MtENOD12 from Medicago fjuncalula. In situ hybridization experiments have shown that, within the indeterminate Medicago nodule, transcription of the MfENOD12 gene begins i n cell layers of meristematic origin that lie ahead of the infection zone, suggesting that these cells are undergoing preparation for bacterial infection. Histochemical analysis of transgenic alfalfa plants that express an MfENOD12 promoter-p-glucuronidase gene fusion has confirmed this result and further revealed that MfENOD12 gene transcription occurs as early as 3 to 6 hr following inoculation with R. meliloti i n a zone of differentiating root epidermal cells which lies close to the growing root tip. It is likely that this transient, nodulation (no@ gene-dependent activation of the ENODl2 gene also corresponds to the preparation of the plant for bacterial infection. We anticlpate that this extremely precocious response to Rhizobium will provide a valuable molecular marker for studying early signal exchange between the two symbiotic organisms.

INTRODUCTION The symbiotic interaction between prokaryotic rhizobia and leguminous plants leads to the formation of nove1 plant organs known as root nodules. Within these organs, the microsymbiont uses photosynthate-derived energy to convert atmospheric nitrogen into ammonia, a form of fixed nitrogen that can be assimilated by the plant host. A complex interplay between the legume host and its bacterial partner is required to assure the induction and subsequent development of the nitrogenfixing root nodule. In the case of temperate legumes such as pea and alfalfa, the first morphological event marking the initial symbiotic interaction with Rhizobium is the characteristic curling of root hairs into so-called “shepherd’s crooks.” Specialized tubular structures known as infection threads then convey the microsymbiont from the root hair tip toward the root inner cortex, where the initiation of cell division has already led to the formation of the nodule primordium. As infection threads penetrate and ramify within the primordium, releasing bacteria into the central tissue, a zone of apical meristematic activity directs outward growth of the differentiating nodule. Mature pea and alfalfa nodules are cylindrical in shape because these nodules, known as indeterminate, possess persistent apical

To whom correspondence should be addressed.

meristems. A longitudinal section through such a nodule reveals the entire series of developmentalsteps that correspond to the concerted codifferentiation of the two symbiotic partners. The apical (or distal) meristematic zone I is followed consecutively by the prefixing zone II (zone of infection), the amyloplasi-rich interzone 111111, the nitrogen-fixing zone 111, and finally zone IV of senescence (Vasse et al., 1990). Recent research has shown that multiple signal exchange is essential for the correct recognition between the plant and the microsymbiont (for reviews, see Long, 1989; Fisher and Long, 1992; Verma, 1992). In particular, rhizobial nodulation (nod)genes, whose transcription requires plant flavonoids, are responsible for the synthesis of extracellular lipooligosaccharides that mediate the specific symbiotic interaction with the legume host (D6narié and Roche, 1992). In the case of R. meliloti, sulfated and acetylated lipooligosaccharideshave been purified from the supernatantsof luteolin-induced cell cultures (Lerouge et al., 1990), and it has been demonstrated that these so-called Nod factors are biologically active in specific root hair deformationassays on the host plant alfalfa. Furthermore, these same signaling molecules are also capable of eliciting cortical cell divisions and the formation of genuine nodules on alfalfa roots (Truchet et al., 1991). Interestingly, the sulfate group appears to be necessary for the expressionof host specificity because nonsulfated factors are not active on alfalfa,

1200

The Plant Cell

but are now able to elicit symbiotic responses on the roots of common vetch, a nonhost for R meliloti (Roche et al., 1991). A detailed analysis of the plant response to these bacterial signal molecules requires the identification of plant genes that can serve as molecular markers for the recognition, infection, and nodule organogenesis triggering processes. The pioneering work of Bisseling and coworkers (reviewed in Nap and Bisseling, 1990) has shown that certain plant genes are specifically expressed in a variety of plant tissues that are involved in early stages of rhizobial infection and nodule development. It was shown that transcription of the pea early nodulin gene EA/OD12, which encodes a (hydroxy)proline-rich protein, takes place within root cortical cells that either contain or lie ahead of the advancing infection thread (Scheres et al., 1990). Transcripts of EA/OD12 were also found within cells of the nodule primordium prior to infection thread penetration and within the infection zone of the mature root nodule. These same authors further showed that ENOD12 mRNA could be detected in total RNA extracts of pea root hairs 24 to 48 hr after infection with

RsatM.tr M.tr

M.tr M.tr M.sat

Eco Eco Hin

Eco Hin Eco kb -12-

sf

-6-

R leguminosarum and that this response was dependent on functional nod genes. The fact that EA/OD12 transcription could also be elicited in root hairs following treatment with cell-free supernatants of flavonoid-induced bacterial cultures provided further evidence for the role of Nod factors in the expression of this early plant symbiotic gene (Scheres et al., 1990). In recent years, research in our laboratory has focused on the nitrogen-fixing symbiosis between R. meliloti and a vari-

ety of plants from the genus Medicago, including alfalfa and the diploid autogamous M. truncatula (Barker et al., 1990). The availability of a wide range of R meliloti mutants that have altered symbiotic properties and the means of purifying the corresponding Nod factors from these mutant strains make this a most attractive system with which to study the regulation of plant genes in response to bacterial lipooligosaccharide signals. Furthermore, certain alfalfa genotypes are amenable to transformation and regeneration procedures (Deak et al., 1986; Chabaud et al., 1988), thus enabling us to construct transgenic plants expressing reporter genes under the regulatory control of the plant gene of interest. In this way, the transcriptional activity of the gene can be evaluated both at the cellular level and throughout the intact root system using a simple histochemical staining procedure (Jefferson et al., 1987). In this study, we describe the isolation and characterization of a single-copy gene, homologous to pea EA/OD12, that was obtained by screening a genomic library of M. truncatula. The tissue-specific transcription of this early nodulin gene has been studied by both in situ hybridization and the analysis of transgenic alfalfa plants that express a chimeric gene fusion between the M. truncatula MtENOD12 promoter and the (3-glucuronidase (gusA) reporter gene (previously known as uidA). The discovery of novel expression patterns for the E/VOD12 gene provides new insights into the molecular and cellular mechanisms involved in preparing plant tissues for subsequent infection by Rhizobium.

RESULTS

-1.5Identification of an At. truncatula Gene Homologous to Pea ENOD12

MfENOD12

PsENOD12 probes

Figure 1. Genomic DMA Gel Blot Analysis Using Pea and M. truncatula 0VOD12 Probes. Fifteen micrograms of pea genomic DNA (P. sat), 5 ng of M. truncatula DNA (M. tr), and 5 ng of alfalfa DNA (M. sat) were digested with either EcoRI (Eco) or Hindlll (Hin), electrophoresed on 0.8% agarose gels, and blotted onto GeneScreen membranes. Hybridization was carried out as described in Methods using either a Ps£WOD12 cDNA probe

(left-hand three lanes) or the 0.5-kb Sphl-BamHI fragment of the Mf£A/OD12 gene (right-hand three lanes).

Figure 1 shows that several hybridizing bands of variable intensity can be observed when a pea EA/OD12 cDNA fragment is used to probe restriction digests of M. truncatula genomic DNA under low-stringency conditions. Genomic clones corresponding to each of the major hybridizing bands were isolated by screening an M. truncatula gene library constructed in the phasmid vector pGY97 (Vincze and Kiss, 1990). By means of DNA-DNA hybridization and partial sequence analysis (see below and results not shown), we were able to conclude that only one of these clones (pMt12) contained a gene that is homologous to the pea EA/OD12 probe. Details of the four other M. truncatula genes that were isolated by this screening procedure will be presented elsewhere.

Medicago ENODl2 Gene Transcription

-A-

-

1201

probe

BIS

H

H B

H

Sp

B M I

C

H I

BIS I

-B5'

...GCATGCCATGCAATMTAATACTAACGGGTCTAGTGCCT~GCTATTTTTTATTTCMGCAMATTGAATTTATCCAAATTGAATATTACATTT

-227

TTTMATAATATTCCACGTACACCMTACGTTCTTTGTTAATATTCGACACAAATTTT~CATTACATAGAGTGATTAATTATTAGTTTATTGATCTATAGTMTAATAACCTATT~ -108 TCGGTCGAGACATTGCACTTCCATTGAGGCCCCTAACTACTATAAAACCTGATTATTCCCTMATCCCTATATGTTATATGCACTAACATT~TTACTACTTMA ATG GCT TCC Met A l a Ser

Y

9 (3)

TTT TCG CTG TCC ATA TTA GTG TTT TTC TTT TCT GCT CTT GTC CTT GTT CCT C A I GGC TTT GCT GAA TAT TAC CTT TAT CCT GCT TAT AGG

99

L e u Val L e u Val P m Gln G7y Phe Ala G l u T y r T y r Leu T y r Pro A l a T y r A r g

(33)

Phe Ser L e u Ser I l e Leu Val Phe Phe Phe Ser A l a

CCA CCA CAA ACG AAA CCA CCG GTG M T AAG CCA TCA CAC AAG GAA CCA CCG GTG M T AAG CCA CCA CAC AAG GAA CCG CCG GTC CAC M G

189

P r o P r o G l n T h r Lys P r o P r o V a l A s n Lys P r o Ser H i s Lys G l u P r o P r o V a l A s n Lys P r o P r o H i s Lys G l u P r o P r o V a l His LYS

(63)

CCA CCA CAC AAG GAT CCA CCG GTT AAT M G CCA CCA CAA AAG GAA TCA CCA GTG CAC AAG CCA CCA CGA AAG GAA TCA CCG ACG CAT AGG Ser P r o T h r H l s A r g P r o P r o H i s Lvs A s p P r o P r o V a l A s n Lys P r o Pro G l n Lys G l u Ser P r o V a l His LVS

(93)

CAT CCT CCA GCA GAA GAT AAC ATC CAT TTC TAA A G C A T G A A A A T A C A A C T T T A A G G C T T A T T T A T A T C T A C A A T T T G T T G T A T

387

His P r o P r o A l a G l u Asp Asn

Ile H l s P h e

279

*

.

TMTAATTTTTGTACAATGTATGGGCMGCMGTTGTATCTTGCCTTATCTAT~CA. . 3 '

Figure 2. Restriction Map and DNA Sequence of the M. truncatula fNODl2 Gene.

(A) Partia1 restriction map of the 8.8-kb insert of the genomic clone pMtl2 containing the MtfNOD12 gene (see Methods). The coding region of MtENODlP is representedby the thickened horizontal lhe, and the accompanying arrow shows the direction of transcription. A horizontal bar indicates the 0.5-kb Sphl-BamHl restriction fragment that was used for hybridizationstudies. The 2.3-kb promoter fragment that was amplified by PCR to generate the gusA transcriptional fusion (see Methods) is shown by the horizontal dashed line bordered by asterisks. Abbreviations for restriction sites are as follows: B, BamHI; C, Clal; H, Hindlll; M, Mscl; N, Ncol; Sp, Sphl; B/S, BamHIISau3A junction. (8) DNA sequence lying between the Sphl and Mscl restriction sites and covering the MtENOD12 coding region. The deduced amino acid sequence has been annotated to indicatethe putative signal peptide (italics),the peptide cleavage site (vertical arrow), and the proline-richpentapeptide repeat motifs (double-underlining).A potential TATA element within the promoter region has been underlined. Nucleotide numbering is relative to the A residue (+1) of the initiator ATG codon; there is no " O position. Amino acid numbering is shown in parentheses.

The partia1restriction map of the pMtl2 insert is presented in Figure 2A, showing both the location of the coding region and the direction of transcription of MtENOD12, as determined by sequence analysis (see below). Genomic gel blot hybridization of M. fruncatula DNA with the 0.5-kb Sphl-BamHI fragment, which covers part of the coding region and the 5' noncoding region of MtENOD12 (Figure 2), showed that this single-copy gene lies within the 12-kb EcoRl and 6-kb Hindlll genomic fragments (Figure 1) and, thus, corresponds to the most intense hybridization signal seen with the pea ENOD12 probe. Govers et al. (1991) have shown that, in the case of the pea (Pisurn safivum) genome, there are two closely related ENOD12 genes, PsENOD12A and PsENOD12B, and that both genes have the same organ-specific pattern of expression. The fact that MtENOD12 is a single-copy gene has greatly simplified the analysis of MtENOD12 transcription by means of

specific DNA and RNA hybridization probes (see below). In contrast, severa1 hybridizing bands can be seen when alfalfa genomic DNA is hybridized with the MfENOD12 probe (Figure 1). This probably reflects the allelic heterozygosity commonly found with the tetraploid allogamous alfalfa and serves to illustrate the advantage of using the diploid autogamous M. fruncafula for such molecular studies.

Sequence Analysis of the MtENOD12 Gene The nucleotide sequence of the coding strand of MtENOD12 and the deduced amino acid sequence are presented in Figure 28. No alternative open reading frame of significant length could be identified on either strand. As is the case for the pea ENODl2 proteins (Govers et al., 1991), the ATG initiation codon

The Plant Cell

1202

of MtENOD12 is followed by 23 amino acids that, according to the rules of Von Heijne (1983), are likely to serve as a membrane-translocation signal peptide. The remaining coding sequence (79 amino acids) is mainly composed of the repeating pentapeptide unit PPXXX, a structural feature that has also been found in a family of hydroxyproline-richcell wall proteins of soybean known as SbPRPs (Hong et al., 1990). The homologies between the Medicago and the two pea ENOD12 proteins are shown in Figure 3A. It is striking that only a single gap of 7 amino acids has to be introduced into the C-terminal region of MtENOD12 to optimize the alignment with PsENOD12A. The two regions of maximum homology correspond to the signal peptide sequence (83% nucleotide and 71% amino acid identities) and the proline-rich repeat region (76% nucleotide and 69% amino acid identities). The greater nucleotide sequence homology coupled with the near perfect alignment of the two sequences strongly suggest a common

-A-

v

MIENODIZ

----

PsENOD12A

MPSFSLSILVFFFSALVLVFWFA EYYL YPAYR P W T K PPVNK PSHKE PPVNK **** nu I." * e.. ***** * * LI * I ** *UD*" u na. >(" >, MPSFFLSSLVLFLAALILVFWLA QYHL NPVYE PPVNG PPVNK PFQKE TPVHK

PsENOD12B

MPSLFLSSLVLFLAALILVFWFA QYHL NPVDE PPVNE PTVNK P W K E TPVYK

Mf PSA PSB

nu*

* ****

*I*"*U**I*.*******%"

*

a**

e***

*

------PPHKE PPVHK PPHKO PPVNK PFQKE SPVHK PPRKE SPTH.. **

*nu*

um

*\- *

>?*

*

* ** **

I****

I

* ***** 0

*iy*

**o**

*l*

.I

.. .. .RHPPAEONIHF ** *

*i>**

P W K E TPVHK PFQKE PPRHK P W K E PPRHK PPHKK SHLHVTKPSYGKHPTEEHNIHF **** I) * * n * I; ** * *nu** P W K K SFUYM PEQKE , . . . .. . SHLHV.. .YOKHLTAEHNIHI

.. .. . . .

. .....

.

-B4.-

.-.-.-.-.____

*.-. **

m

MlENOO I2

GTTTATlGI\TCTATAGTMTMTAACCTATT~TC~lCGtG (-99) ***o I 11111

PsENOD12B

GI\TTATTGPIGPITATA.TAATAAATTAGTTTT

* *******

******

**

.....GPITTGPIG ( - 8 5 )

MI

ACATTKACTTCCATTGAGGCCCTAACTAC~ÃZZECCTG~TT~TTC ( - 50) .\,:, *e, *..IllZ** *' *O\'* III'*LI*Ll**ll* *****"Q

PsB

A C C T T C C A C T T C T A T T ~ G G P I T C C T T A C T A G ~ C C T T A T T A T T C( - 3 6 )

Mf

CCTAAATCCCTATAIGTTATATKACTAACATTAAAATTACTACTTMA

*-.-.-.-.

.-.* (lil

PSB

*

** **

I,"

CCCATAT..

,

**>.ir

-.-.-* &,l,n**

I.

evolutionary origin for these two genes. Closer inspection of the proline repeat region of MtENOD12 and PsENOD12A reveals five well-conserved repeats of the decameric sequence PPVNI~KPPHloKE. Despite the extensive deletion within the fsEN0012B coding region, it is clear that the two putative pea proteins are more closely related to each other than to the Medicago protein. A comparison between the 5' flanking sequence of MfENOD12 and that of fsENOD12B (the only upstream sequence currently available) reveals three stretches of quite striking homology within 140 bp upstream from the translation initiation site (Figure 38). The first stretch runs from positions -1 to -29 (76% conservation),the second from -43 to -105, including the putative TATA element ( ~ W O and ) , the third from -111 to -141 (68%). Numbering is based on the Medicago sequence. Interestingly,it has already been shown that upstream sequences of the fsENOD12B gene, which correspond to the central conserved stretch, can also be aligned with equivalent promoter regions of the small family of soybean genes that encode the SbPRPs (Govers et al., 1991). However, as these authors have pointed out, the considerable differences in both developmental and tissue-specific expression patterns shown by €NO012 and these three soybean f R P genes (Wyatt et al., 1992) make it unlikely that this conserved region could be a determinant of organ-specific regulation.

(-1 )

i)Ii.,lllll,I"

. .. . , , . , . .GTG.A~AAACACAAAAATCATTACTTAAA ( - 1 )

Figure 3. Homologybetweenthe Coding and 5' UpstreamSequences of the Medicago and Pea ENODl2 Genes. (A) Alignment of the deduced amino acidsequences of the MtENOD12 gene and the two peagenes,PsENOM2AandPsENOD12B.Gaps have been introduced into the MtENOD12 gene and the PsENOD12B gene (Govers et al., 1991)to maximizethe alignments, and amino acid identities are indicatedby asterisks. The putativesignal peptide is italicized, the peptide cleavage site is marked by a vertical arrow, and the prolinerich pentapeptide repeat elements are overlined. (6) Alignment of nucleotide sequences 5'to the initiation codons of the MtENOD12 and PsfNOD12B genes. Gaps have been introduced into the PsENOD12B promoter sequence to optimize the alignment, and dashed horizontallineswith arrowheads mark the three stretches of homology referredto in the text. PotentialTATA elements are underlined and overlined, and the transcription initiation site of the pea PsENOD12B gene is double underlined (Govers et al., 1991).Nucleotide numbering is the same as given in Figure 2.

Spatial-Temporal Expression of MtENODl2 during Nodule Development Studies on the variation in ENOD12 mRNA levels during nodule development in pea had revealed two particular characteristics. First, gene transcription could be detected earlier during nodule development as compared with previously described nodule-specific genes, and second, the abundance of €NO012 transcripts decreases as the young immature nodule develops into the mature nitrogen-fixing nodule (Scheres et al., 1990). When M. truncafula plants are grown in aeroponic conditions (see Methods), nodules first become visible on the root system between 3 and 4 days following inoculation with R. meliloti, and nitrogenase activity is first detectable 2 to 3 days later. Figure 4 shows the profile of MtENOD12 mRNA abundance in total RNA extracts prepared from nodules of M. fruncatulaharvested between 4 and 8 days after inoculation. The leve1 of MfENOD12 transcripts is relatively high at 4 to 5 days postinoculation and then drops rapidly by a factor of approximately 10-fold as the nodule continues to mature and begins to fix atmospheric nitrogen. As is the case for the pea nodule, this is in striking contrast to the profiles observed for transcripts encoding either the abundant oxygen-buffering protein leghemoglobin or the nodule parenchyma-specific protein ENOD2 (van de Wiel et al., 1990), both of which remain at constantly high levels as the nodule reaches maturity (Figure 4). A control hybridization with a human ubiquitin probe has been included to show that RNA loadings were approximately equal.

Medicago EA/OD12 Gene Transcription

days post-inoculation

4

5

6

7

8

-MfENOD12

ENOD2

- Leghemoglobin

- Ubiquitin

N 2 fixation Figure 4. Comparison of the Expression Patterns ot the Mf£WOD12 Gene and Other Symbiosis-Related Genes during Nodule Development. Total RNA was extracted trom M. truncatula nodules harvested between 4 and 8 days after inoculation with R. meliloti, and 10-ng samples were subjected to denaturing gel electrophoresis followed by transfer to GeneScreen membranes (see Methods). The same RNA gel blot was hybridized consecutively with the following 32P-labeled DMA probes: the 0.5-kb Sphl-BamHI fragment of /WEA/OD12 (see Figure 2 and text); a soybean ENOD2 cDNA fragment; an alfalfa leghemoglobin cDNA fragment; and a human ubiquitin gene fragment (see Figure 6 of Gallusci et al., 1991 for details regarding these heterologous probes). Hybridization and washing conditions were identical for all probes (see Methods), and the blot was totally stripped between hybridizations. Nitrogen-fixing activity within M. truncatula nodules can first be detected approximately 6 days postinoculation under the aeroponic growth conditions used in these experiments.

M. truncatula nodules harvested 2, 3, or 4 weeks after inoculation showed essentially the same pattern of mRNA abundance as did 8-day-old nodules (results not shown), which is consistent with the presence of the persistent apical meristem of the indeterminate Medicago nodule. Scheres et al. (1990) have shown that pea EA/OD12 mRNA can also be detected (albeit at low levels) in both stem and flower tissue, showing that these genes are not strictly nodule specific. Gel blot analyses with total RNA isolated from different M. truncatula tissues (uninoculated roots, stem, hypocotyl, cotyledon, leaf, petiole, flower, and dry seeds) have so far failed to reveal the presence of MfEA/OD12 mRNA. This does not, of course, rule out the possibility of either very low transcript levels or brief transient expression in such tissues. Having established that the pattern of MfEA/OD12 mRNA abundance during nodule development resembles that described for the homologous pea genes, we decided to examine

1203

the tissue-specific location of these transcripts within the nodule using in situ hybridization. Sections of M. truncatula nodules, harvested either 4 days or 3 weeks after inoculation, were hybridized with a 35S-labeled antisense RNA probe. Control experiments with sense probes routinely gave low, even backgrounds (results not shown). Figure 5A shows that a relativelf uniform hybridization signal is present within the central tissue of the 4-day-old nodule. At this early stage of development, the nodule is approximately spherical in shape with a central tissue in which ramifying intercellular infection threads penetrate the plant tissue. Our results showed that all cells within the central tissue of the 4-day-old nodule express the Mf£A/OD12 gene, which is in line with the results obtained for EA/OD12 mRNA localization in immature pea nodules (Scheres et al., 1990). In situ hybridization experiments carried out on sections of 3-week-old mature nitrogen-fixing nodules of M. truncatula showed that A/WEA/OD12 transcripts are present at the distal end of the prefixation zone II, corresponding to a region of the nodule where bacteria are being released from the infection threads (Figure 5B). When examining the apical region of the 3-week-old Medicago nodule at a higher magnification, it is possible to distinguish meristematic cells undergoing cell division (Figure 5C). While MfEA/OD12 transcripts are absent in these actively dividing cells, a hybridization signal is clearly visible in the two to three cell layers in which infection threads are not yet present. These results suggest that MtENOD~\2 transcription is initiated in the proximal cell layers of zone I that have ceased to divide, but that are not yet part of the prefixation zone II where infection takes place. The hybridization signal drops to background levels at the proximal end of prefixing zone II where bacteroids and plant cells are rapidly differentiating. This occurs prior to the amyloplast-rich interzone region (results not shown), which we have previously shown to be the site of leghemoglobin gene transcriptional activation (de Billy et al., 1991).

Expression of a Transcriptional MfEA/OO12 PromotergusA Fusion in Nodules of Transgenic Alfalfa To complement the in situ hybridization studies and to facilitate the analysis of /WE/VOD12 gene expression during earlier stages of the symbiotic interaction, we decided to introduce an MfEA/OD12 promoter-gus>4 fusion into transgenic Medicago plants. The 2.3 kb of DNA lying immediately upstream of the /WE/VOD12 ATG translation initiation codon was cloned in front of the Escherichia coli gusA coding region in such a way as to generate a precise transcriptional fusion (see Methods). This chimeric construction was then introduced into M. varia A2 plant tissue by means of an Agrobacterium fumefac/ens-leaf disc transformation protocol (Chabaud et al., 1988). Regeneration of whole plants via somatic embryogenesis led to the isolation of 20 kanamycin-resistant plants that were phenotypically indistinguishable from the nontransformed line. Genomic DNA gel blot analysis of 12 of these regenerated plants provided

1204

The Plant Cell

*•"

C

35

Figure 5. In Situ Hybridization of M. truncatula Nodule Sections (7 urn) with an Mf£WOD12 S-Labeled Antisense Probe. (A) Four-day-old immature nodule showing a uniform pattern of hybridization within the central tissue. Silver grains appear as bright spots when viewed by dark-field microscopy. Bar = 50 um(B) Three-week-old mature nodule. The hybridization signal is maximal at the distal end of the prefixation zone II (large asterisk), which corresponds to the infection zone of the nodule. Note the low level of hybridization at the proximal end of this zone (arrows). The meristematic zone is indicated by the small asterisk. Bar = 200 um. (C) Bright-field microscopic image of a 3-week-old nodule. Mf£WOD12 transcripts (silver grains now appear as dark spots) are present in the two to three cell layers of the "preinfection" zone (star, see Discussion), which lie proximal to actively dividing meristematic cells. The arrowhead indicates a cell in anaphase. Note that infection threads (arrows) are not yet present in the cells of the "preinfection" zone. Bar = 20 um.

direct evidence for the successful integration of between two and five copies of the reporter gene fusion (results not shown). To examine the expression pattern of the MtENOD12-gusA fusion following inoculation with Rhizobium, primary transformants were taken through several cycles of vegetative propagation and then 4- to 5-cm-long cuttings were grown for about 3 weeks in aeroponic conditions until the root systems were well developed. R. melilotiwas added to the liquid growth medium lacking combined nitrogen, and samples of the roots were examined for GUS activity at regular intervals after inoculation. Two of the 20 plants were scored negative for GUS activity throughout the experiment, and the remaining 18 responded in a qualitatively identical fashion. Nodules first became visible on the transgenic alfalfa root system approximately 3 to 4 days following Rhizobium inoculation, showing that M. varia and M. truncatula have very similar kinetics of nodule development. These immature nodules developed an intense indigo blue coloration when treated with the histochemical GUS substrate X-gluc, and subsequent sectioning showed that the GUS activity was distributed uniformly throughout the central tissue (results not shown). Figure 6A shows that at a slightly later stage of nodule development (4 to 5 days postinoculation) the blue staining zone had clearly

moved to the distal end of the central tissue. Expression of the gusA fusion is clearly visible in tissue that lies distal to the zone where infection threads are visible. This correlates well with the localization of Mf£7VOD12 mRNA by in situ hybridization (Figure 5C). The very pale coloration that is present in certain peripheral cells of the nodule may be due to limited diffusion from the intensely staining regions. The distal localization of GUS activity in nitrogen-fixing and elongating indeterminate alfalfa nodules can be clearly seen in stained whole root segments (Figure 6B). Thick sections (80 u,m) of GUS-stained mature nodules have also been stained with potassium iodide to reveal starch-containing cells (Figure 6C), showing that the MfEA/OD12 promoter fusion is no longer being expressed in the central cell layers that immediately precede the amyloplast-rich interzone region.

Expression of the Mt£NOD12 Promoter-gusA Fusion during the Earliest Stages of the Symbiotic Interaction Having established that the expression of the EA/OD12 promoter-guaA fusion in transgenic alfalfa nodules was very similar to the pattern of gene transcription in M. truncatula

Medicago OVOD12 Gene Transcription

nodules as determined by in situ hybridization, we decided to focus on reporter gene expression within the alfalfa root system at much earlier stages of the symbiotic interaction. The first response to the addition of R. meliloti could be detected in apical regions of the root system as little as 3 to 6 hr after inoculation. As shown in Figure 7A, a uniform pale blue staining covered a region that starts within the root elongation zone just behind the root tip, continues throughout the zone of root hair emergence and development, and terminates at the start of the mature root hair zone. GUS activity was present in all epidermal cells of this reactive region, including those

which had developed root hairs (Figures 7B and 7C), and no staining could be detected in internal cortical cell layers (results not shown) or in the region of the root with mature root hairs (Figure 7D). It should be emphasized that identical expression patterns were recorded for all 18 of the transgenic lines analyzed. During the following 18- to 24-hr period, we observed an increase in the blue coloration within this reactive zone, although the intensity of staining remained uniform throughout. Control experiments using these same transgenic alfalfa plants failed to detect GUS activity in either root hairs or root epidermal cells prior to inoculation or following a 6-day period

1205

of nitrogen starvation in the absence of Rhizobium. Furthermore, inoculation with a non-nodulating strain of R. meliloti carrying a mutation in the nodA gene failed to elicit any reporter gene expression within the apical region of the root (results not shown), demonstrating that this early response is indeed dependent on the activity of the bacterial nod genes. Histochemical staining of transgenic alfalfa root segments 48 to 72 hr after inoculation with R. meliloti revealed a number of discrete, dark blue-colored loci within the mature root hair region (Figure 7E). Preliminary observations have shown that certain early symbiotic events, such as root hair deformation, root hair infection, and cortical cell division, do indeed take place within these reactive loci, where GUS activity can now also be detected within the root cortex (results not shown). However, a detailed cytological study will be required to identify those cell types that express the gusA fusion and to correlate events occurring on the root surface with those that take place within the cortex. With the exception of these intensely staining loci, all the surrounding epidermal cells (including root hairs) were without detectable GUS activity. Harvesting and staining root segments at intermediate time points (results not shown) have confirmed that this region corresponds to the

* ,•

B Figure 6. GUS Activity in Root Nodules of Transgenic M. war/a Expressing the Mf£WOD12 Promoter-gusA Gene Fusion. (A) Section (1 to 2 urn thick) of a 5-day-old immature nodule. Reporter gene activity is present in distal cell layers (arrows) that lie ahead of the prefixation zone II (asterisk). The blue coloration is most intense within the distal part of the prefixation zone, where it is possible to visualize sections through infection threads (arrowheads). Bar = 100 urn. (B) Histochemical staining of a whole nodulated root segment harvested 7 days postinoculation. GUS activity is clearly visible at the distal end of the nodules. The arrowhead indicates a localized region of the root where earlier stages of the symbiotic interaction are visible (see also Figure 7E). Bar = 250 urn.

(C) A section (80 urn thick) of a mature nitrogen-fixing nodule harvested 3 weeks after inoculation. A gradient of GUS activity can be observed within prefixation zone II (star), decreasing from the distal end toward the proximal end. The section was cleared with sodium hypochlorite and then stained with potassium iodide (Vasse et al., 1990) to reveal the amyloplast-rich cells of interzone ll/lll (arrowheads). Bar = 200 urn.

1206

The Plant Cell

D Figure 7. Expression of the Mf£A/OD12 Promoter-gu&4 Fusion in Transgenic M. varia Roots during Early Stages of the Symbiotic Interaction. (A) Pattern of GUS activity throughout the zone of epidermal cells close to the growing root tip 20 hr after inoculation with R. meliloti. Bar = 250 |im. (B) Reactive epidermal cells of the root elongation zone (marked by a single arrowhead in [A]). Bar = 125 urn. (C) Reactive zone of root hair emergence (double arrowheads in [A]). Bar = 125 urn. (D) Zone of mature root hairs (triple arrowheads in [A]) in which GUS activity cannot be detected. Bar = 500 urn. (E) GUS activity within discrete loci of the mature root hair zone of the transgenic alfalfa root harvested 72 hr after inoculation. Note that, with the exception of the intensely stained regions, GUS activity is absent throughout the remaining root epidermis. Bar = 700 urn.

maturation of the zone which had previously stained uniformly for GUS activity close to the root tip (Figures 7A to 7C). We therefore concluded that, for the majority of differentiating root epidermal cells, this early transcription of the /Wf£A/OD12 gene is a transient phenomenon (compare Figures 7A and 7E).

DISCUSSION

Detailed molecular and cellular analyses of the events that occur during the earliest stages of the symbiotic interaction between the host legume and the corresponding rhizobial partner require the identification of genes that can act as markers for the plant response. With this goal in mind, and wishing to focus our studies on the Medicago-R meliloti symbiosis

for the reasons discussed earlier (see Introduction), we have cloned and characterized an M. truncatula gene, Mff A/OD12, that appears to be both structurally and functionally homologous to the pea early nodulin gene EA/OD12 (Scheres et al., 1990). These authors have proposed that the protein encoded by the pea EA/OD12 gene is most probably a (hydroxy)prolinerich cell wall protein. Their argument is based on the presence of a putative N-terminal transmembrane signal sequence and by drawing an analogy with the small family of SbPRPs, which are composed almost entirely of a proline-rich pentapeptide repeat motif (Hong et al., 1990). The fact that the homology between the deduced pea and Medicago ENOD12 amino acid sequences is greatest within both the signal peptide (71%) and the domain of proline-rich repeats (69%) (Figure 3A) further argues for the cell wall localization of this nodulationrelated protein.


By means of in situ hybridization experiments, we have shown that, in both immature and mature nitrogen-fixing nodules of M. truncatula, MtENOD12 transcription is maximal in the zone in which infection threads are spreading and releasing bacteria into host cells. Those cells that are “infected” will subsequently differentiate into the enlarged Rhizobium-filled cells of the nitrogen-fixingzone, whereas the uninfected cells will remain small in size and develop large vacuoles. By examining the apical meristematic region of the Medicago nodule in greater detail, we have been able to observe that MtENOD12 transcription is in fact initiated within a narrow zone, composed of two or three cell layers, which is immediately adjacent and proximal to the actively dividing meristematic cells, but clearly lacking infection threads (Figure 5C). We propose that the term “preinfection zone” be used to describe this narrow band of cells lying between the meristematic cells and the infection thread region. That MtfNOD12 transcription should be triggered within this zone is interesting in light of the fact that, during early stages of infection, pea ENODl2 transcripts have been found in cortical cells that lie ahead of the progressing infection thread (Scheres et al., 1990), thus leading to the hypothesis that “diffusible” signal molecules originating from the infection thread are responsible for fNOD12 gene activation at a distance. Our observations would suggest that a similar gene activation mechanism exists in the developing nodule, where the cells that lie immediately ahead of the infection thread region are also preparing for subsequent infection. This would also explain why all cells within this zone contain €NOD12 transcripts irrespective of whether they become infected or remain uninfected. One of the principle advantages of studying symbiosisrelated plant gene expression in species of the genus Medicago lies in the possibility of obtaining transgenic alfalfa via A. tumefaciens transformation and somatic embryogenesis. With the notable exception of Lotus (Petit et al., 1987), alfalfa is the only legume that is currently amenable to such routine transformation and regeneration procedures. Based on the assumption that gene regulatory mechanisms would be highly conserved between closely related species of the genus Medicago, we have introduced a chimeric gene composed of a 2.3-kb MtENOD12 promoter fragment fused to the coding region of the gusA reporter gene into the high-frequency regenerating line of alfalfa, M. varia A2. A qualitatively homogeneous response, in terms of reporter gene expression, was obtained for 18 of the 20 transgenic plants tested following inoculation with R. meliloti. More importantly, the distribution of GUS activity within the nodules that formed on the roots of these transgenic plants correlated remarkably well with the localization of MtENOD12 mRNA, as determined by in situ hybridization analyses on sections of M. truncatula nodules (Figure 5). The leve1 of GUS activity was found to be maximal within the invasion zone of both immature and mature nitrogenfixing nodules. Furthermore, expression of the chimeric reporter gene also appeared to be initiated in preinfection cell layers which precede the zone of infection thread proliferation (Figure 6A). These results provide convincing evidence

1207

that the 2.3-kb MtENOD12 promoter fragment contains all the information necessary for regulated expression of the Medicago fNOD12 gene and that our reporter gene assay in transgenic alfalfa provides avalid means of evaluating the expression patterns of this early symbiotic gene. When transgenic alfalfa plants were used to examine the expression of the MtENOD12 gene during the earliest stages of the symbiotic interaction, we discovered that reporter gene activity could first be detected in roots as little as 3 to 6 hr following inoculation with R. meliloti. Furthermore, GUS activity was present not only in young developing root hairs but throughout all epidermal cells of a region extending from just behind the growing root tip as far as the beginning of the mature root hair region (Figure 7A). The relatively uniform pattern of staining suggests that all cells on the outer surface of the root and lying within this zone respond to the presence of the bacterial symbiont. Because a nodulation-deficient mutant of R. meliloti carrying a Tn5 insertion in the nodA gene does not elicit this reaction, we can reasonably conclude that this corresponds to a nod gene-dependent symbiotic response. The fact that fNOD12 gene expression should be triggered in this actively differentiating region of the root is of considerable interest because it is now well established for alfalfa (and indeed for most other legumes so far examined) that the events that lead to subsequent nodule formation are generally initiated within the part of the root that lies between the elongating root tip and the zone of root hair emergence (Bhuvaneswari et al., 1981; CaetanoAnollésand Gresshoff, 1991). Even when successful infections are initiated within the more mature region of the root, these are usually restricted to a zone no greater than 1 cm distant from the point of first root hair emergence. The striking correlation with the pattern of early reporter gene expression suggests that the transcriptionof the ENODl2 gene parallels the differentiation of a zone that is undergoing preparation for subsequent Rhizobium infection. Epidermal cells that form root hairs are known as trichoblasts, and during this differentiation process, polar tip growth is established only after cells have initiated a round of cell division and have arrested in cytokinesis (for a review, see Kijne, 1991). We can speculate that, in response to rhizobial signals, specific symbiosis-related proteins, such as ENOD12, may be incorporated into the developing root hair cell wall, thus rendering the root hairs susceptible to Rhizobium infection. Modifications in the cell wall of the root hair could have an important role in severa1 stages of the infection process including bacterial attachment, localized cell wall degradation, and the development of the infection thread. In the case of pea, it has been shown that the mRNA population of root hairs is significantly modified following infection by R. leguminosarum, with the appearance of at least one new mRNA species and a significant enhancement in the levels of a second mRNA (Gloudemans et al., 1989). However, the identity and subcellular localization of the corresponding proteins have not yet been determined. Could the rhizobial signals that trigger this very early reaction in epidermal cells close to the root tip be the same

1208

The Plant Cell

symbiotic Nod factors recently identified as extracellular lipooligosaccharides? Interestingly, Nod factors are able to specifically induce root hair branching of the host legume (Lerouge et al., 1990). Morphologically, this branching process corresponds to the formation of a new growth tip on the side of the root hair and is, therefore, analogous in certain respects to root hair growth during trichoblast differentiation. Because it has been shown that Rhizobium Nod factors are able to stimulate root hair development (Roche et al., 1991) and also to induce ENOD12 gene expression in pea root hairs (Scheres et al., 1990), these molecules could well be responsible for initiating changes in the number and composition of the root hairs that will permit subsequent Rhizobium infection. The purification of sulfated lipooligosaccharides from the parenta1 strain of R. me/iloti(NodRm-IV [Ac,S]), which specifically elicit alfalfa root hair branching, and the corresponding nonsulfated derivatives purified from nodH mutants (NodRm-IV [Ac]), which have lost this capacity (Roche et al., 1991), will now enable us to examine directly how these molecules influence ENODl2 gene transcription in relation to the infection process. The zone of epidermal and root hair cells that stained uniformly for GUS activity continued to be clearly visible near the root tips of transgenic alfalfa plants until approximately 24 hr after inoculation with R. meliloti. During the period 24 to 72 hr postinoculation, as the distance between the reactive zone and the root tip gradually increased dueto root growth, a small percentage of root hairs within this zone began to stain very intensely for GUS activity, while the remaining root hairs and epidermal cells rapidly lost their blue coloration (Figure 7E). Preliminaryanalysis suggested that GUS activity was also present within inner regions of the root cortex at this stage (Figure 7E and results not shown), most probably corresponding to the development of the nodule primordium, as described for the pea ENOD12 gene (Scheres et al., 1990). Taken together, our findings suggest the following scenario. Within hours of the addition of Rhizobium, the Medicago ENOD12 gene is activated transiently within epidermal cells of a reactive zone close to the root tip. This rapid response leads to the differentiation of root hairs that are susceptible to infection by Rhizobium. Root hairs that are infectedcontinue to express the ENOD12 gene, and in this case, the leve1of transcription is enhanced. Such a series of events is interesting for the following reasons. First, we can draw a correlation between the expression of MtENOD12 in epidermal cells preparing for infection thread initiation (this study), the expression of the pea ENOD12 gene in cortical cells and cells of the nodule primordium that lie ahead of the infection thread (Scheres et al., 1990), and the expression of the Medicago gene in cells of the preinfection zone of the differentiating nodule prior to infection thread penetration (this study). Could Nod factors be involved in all of these processes? The fact that the appropriate factors can elicit both cortical cell divisions and nodule organogenesis in alfalfa roots (Truchet et al., 1991) and cortical cell divisions in Vicia roots (Spaink et al., 1991) provides indirect evidence that this may indeed be the case. Second, our findings suggest

that recognition of the lipooligosaccharide factor precedes differentiation of the root hair. If so, then it is easier to appreciate how severa1 different determinants of host specificity may be involved in the preliminary stages of the symbiotic interaction. The production and specific recognition of Nod factors would provide the initial trigger to the process, leading subsequently to the expression of other specificity determinants, such as lectins (Diaz et al., 1989), involved at later stages of the interaction (e.g., at the surface of the root hair). The identification and localization of Nod factor receptors will clearly be of considerable importance in understanding the role(s) played by these molecules during symbiosis. In addition to providing new insights into the nature of the plant response during the earliest stages of the Rhizobiumlegume interaction, transgenic alfalfa expressing the ENODl2-gusA fusion should also prove useful for monitoring the plant reaction to a variety of R. meliloti mutants with altered symbiotic characteristics.In particular, the very precocious expression of reporter gene activity in root epidermal cells should provide an excellent marker for studying the response to mutants that produce modified Nod factors, and may even lead to a convenient and sensitive assay for the lipooligosaccharide factors themselves.

METHODS

Plant Material and Growth Conditions Plants of Medicago truncatula cv Jemalong were grown aeroponically and inoculated with the wild-type Rhizobium meliloti RCR2011, as previously described (Gallusci et al., 1991). Cuttings of the M. varia

genotype A2 (Deak et al., 1986) were kindly provided by G. B. Kiss (Szeged, Hungary) and were propagated vegetatively in axenic culture on SH agar medium (Schenk and Hildebrandt, 1972) containing 1% sucrose. For nodulationexperiments, cuttings of transgenic M. varia plants were grown aeroponically using the same conditions as for M. truncatula. After 2 to 3 weeks, plants were inoculated with either R. meliloti RCR2011 or the control non-nodulatingnodA- mutant strain (GM15386; Debelléet al., 1986).Whole root fragments or noduleswere collected at various times after inoculation and treated as described below.

Purification of Nucleic Acids and Filter Hybridization The isolation and purificationof high molecular weight genomic DNA from leaves of M. truncatula cv Jemalong, M. sativa cv Gemini, and Pisum sativum cv Rondo were carried out as described in Barker et al. (1988). Total RNA was extracted from M. truncatula nodules harvested at various times after inoculation with R. melilotiaccording to Lullien et al. (1987). Electrophoresisof restricted genomic DNA on nondenaturing agarose gels and of denatured total RNA on 6% formaldehyde-agarose gels was performed according to Sambrook et al. (1989). Transfer to GeneScreen membranes (Du Pont-New England Nuclear) and subsequent hybridization at 37OC in the presence of 50% formamide and 10% dextran sulfate were carried out following the manufacturer's instructions. After hybridization, blots were


washed in 2 x SSC (1 x SSC is 0.15 M NaCI, 0.015 M sodium citrate), 0.1% SDS at temperaturesbetween 55 and 65OC. Radioactiveprobes were preparedby the oligolabeling procedure(Feinberg and Vogelstein, 1983) using U-~~P-~CTP, and unincorporated nucleotides were subsequently removed by spin dialysis through Sepharose CL 6B (Pharmacia, Sweden).

lsolation of a Genomic Clone Containing the M. truncatula ENOD12 Gene

The construction and screening protocol of the genomic library of M. fruncafula leaf DNA prepared in the phasmid vector pGY97 (Vincze and Kiss, 1990) has already been described (Gallusciet al., 1991).A pea ENOD12 cDNA probe (Scheres et al., 1990), kindly provided by T. Bisseling(Wageningen, The Netherlands), was used to identify positive clones within the library. Hybridization conditions were the same as those used for genomic DNA gel blot analysis (see above). One of these clones, pMtl2, was partially sequenced and found to contain a gene (MtENOD12) whose coding sequence is highly homologous to pea ENODl2 (see text). This clone has a total insert size of approximately 8.8 kb, and Figure 2A shows the partial restriction map of the insert and the location of the MtENOD12 gene.

DNA Sequenclng

The sequence of the Sphl-Mscl DNA fragment that covers the MtENOD12 coding region and immediateflankingsequences (Figure 2) was determined by subcloning short restriction fragments (200 to 300 bp) into the multipurpose vector pUC19. To obtain the sequence of both DNA strands, the dideoxy chain termination reaction (Sanger et al., 1977) was carried out using double-strandedtemplates (Murphy and Kavanagh, 1988) and polymerasepriming from both ends of the pUC19polylinker.The junction sequences betweenadjacent fragments were confirmed either by using overlapping clones or by sequencing from synthetic oligonucleotideprimers. The nucleotidesequence data reported in this paper has been submitted to EMBL, GenBank, and DDBJ as accession number X68032.

In Situ Hybridlzation

For the preparation of the single-stranded RNA probes, the 05-kb SphlBamHl restriction fragment from the MtENOD12 genomic clone was subcloned into pBlueScript SK+ (Stratagene). Synthesis and partial hydrolysisof radiolabeled sense and antisense RNA were carriedout as described in de Billy et al. (1991). In situ hybridizations on 7-bmthick sections using 35S-labeledRNA probes were performed as described previously (de Billy et al., 1991), except for the addition of a 24-hr prehybridizationstep using the standard hybridization buffer minus dextran sulfate.

1209

a template for a PCR, using the phosphorylated forward primer P1 (5’-TTAGGAAITC(EcoRI)ATATACATGGGGGAG-3’)and reverse primer P2 (5’-GGAAGCCATGG(Ncol)TAAGTAGTAATTTT-3’);the bases underlined in the primer sequences correspond to substitutions in the

genomic sequence. Twenty cycles of amplification were performed (94% for 1 min, 57OC for 1 min, 72°C for 3 min) under otherwise standard PCR conditions (Gelfand and White, 1990). The PCR product was then blunt-endedusing T4 DNA polymerase (Sambrook et al., 1989) and cloned into Smal-linearized pUC19 to obtain pUC19-prMtl2. Sequenceanalysis of five such clones failed to reveal a single error within the 300-nucleotidestretch that lies immediately upstream of the ATG codon, suggesting a very low error frequency for the Taq DNA polymeraseunder our experimentalconditions.Because of the presente of a second Ncol site within the MtENOD12 promoter (Figure 2), it was necessaryto carry out a partial digestionof pUC19-prMtl2after linearizationwith EcoRl to recover the full-length 23-kb EcoRI-Ncol fragment.This promoter fragmentwas then cloned betweenthe EcoRl and Ncol polylinker sites of pCCOGUS (Axelos et al., 1989) so that subsequent digestion with EcoRl and Pstl would yield a DNA fragment containing the MtENOD12 promoter fused to the gusA coding sequence with a 3‘ flanking polyadenylation signal (cauliflower mosaic virus 35s).After addingan Sstl site to the 3’terminus, the resultant fragment was cloned between the unique EcoRl and Sstl sites of the binary vector pLPl00 to give pLP100-prMt12. Plasmid pLPl00 (a kind gift of I?Ratet, Gif-sur-Yvette, France) is a derivative of pBinl9 (Bevan, 1984). As a result of the preceding manipulations,the MtfNOD12promoter sequence upstream of the P-glucuronidase(gusA)coding region is identicalto that in the MtfNOD12gene itself, except for the substitution of two C residues for the two A residues at positions -1 and -2. The resultantjunction sequence -ACCATG-, which has been verified by sequencing pLP100-prMt12,satisfies the preferencesshown by plant genes for the three nucleotidepositions that precede the ATG codon (Cavener and Ray, 1991).

Transformation of

M. varia and Recovery of Transgenic Plants

The binary vector pLP100-prMt12was mobilized into the Agmbacterium tumefaciens LBA4404(Hoekema et al., 1983) according to the procedure describedby Holsterset al. (1978),with three freeze-thaw cycles, and transformedA. tumefaciens colonies were selected by growth on 50 pglmL kanamycin. Leaf segments of M. varia A2 were transformed via A. tumefaciens, and somatic embryogenesis was induced on kanamycin-resistantcallus tissue as described by Chabaudet al. (1988). Embryos were matured on the modified UM medium of Stricklandet al. (1987),with the addition of 5 glL charcoal (T. Huguet,personalcommunication). The regenerated primary transformants were grown in axenic culture on SH medium (Schenk and Hildebrandt, 1972) with 1% sucrose and propagated by taking cuttings. Patterns of reporter gene expressionfrom the MtENOD12 promoter-gusA fusion were unaltered even after numerous cycles of vegetative propagation.

Histochemical Localization of GUS Activity Constructlon of a Transcriptional MtENOD12 Promoter-gusA Fuslon As afirststep, the 23kb DNA fragment lying upstream of theMt€NODl2 ATG translation initiation codon (Figure 2) was amplified usingthe polymerase chain reaction(PCR)to introduce appropriate restrictionsites at either end of the fragment. The genomic clone pMtl2 was used as

Histochemicalstaining for GUS activity was performed as previously described (Jefferson et al., 1987)with the following modifications.Whole root fragments or nodules were excised from the plant and prefixed byvacuuminfiltrationwith an ice-coldsolution of 0.3Vopformaldehyde in 0.1 M potassium phosphate buffer, pH 7.0, followed by incubation on ice for 45 min. After two washes in the phosphate buffer, whole

1210

The Plant Cell

organs were immersed in the GUS substrate solution containing 1 mM X-gluc (5-bromo-4-chloro-3-indolylglucuronide, cyclohexylammonium salt; Biosynth AG, Staad, Switzerland), 5 mM EDTA, 0.5 mM potassium ferricyanide, O5 mM potassiumferrocyanide,and 0.1 M potassium phosphatebuffer, pH 7.0. lncubationwas performedin the dark at 37% for periods of time between 4 and 24 hr, dependingupon the intensity of the coloration. After rinsing in phosphate buffer, stained tissues were observed either as whole specimens or as sections (80 pm thick; Microcut H 1200; Bio-Rad)with an Olympus Vanox light microscope using bright-fieldoptics. To improvethe contrast betweenstained and nonreactive tissues, the samples were briefly cleared with sodium hypochlorite (Boivin et al., 1990). Using these methods, endogenous GUS activity was never observed in either root or root nodule tissues from untransformed alfalfa plants. For the localization of GUS activity at the cellular level, noduleswere dissected from stained nodulated root segments and then, successively, postfixed for 1 hr in 2.5% glutaraldehyde buffered with 0.2 M sodium cacodylate, pH 7.2, rinsed in the same buffer, dehydrated in an alcohol series, and embedded in Epon resin (Merck, Darmstadt). Sections (1 to 2 pm thick) were counterstainedwith basic fuchsin (2% in distilled water) and observed by bright-field microscopy.

ACKNOWLEDGMENTS

We are very gratefulto Ton Bisseling,Wageningen, The Netherlands, for providing us with the pea ENOD12 probe and to Gyorgy 8. Kiss, Szeged, Hungary,for cuttings of the M. varia genotype A2. We would also like to thank colleagues in the laboratory for their constructive criticisms of the manuscript.

Received June 18, 1992; accepted August 17, 1992.

REFERENCES

Axelos, M., Bardet, C., Llboz, T., Le Van Thai, A., Curie, C., and Lescure, 8. (1989). The gene family encoding the Arabidopsis thaliana elongation factor EF-1a: Molecular cloning, characterization and expression. MOI.Gen. Genet. 219, 106-112. Barker, D.G., Gallusci, P., Lullien, V., Khan, H., Ghbrardi, M., and Huguet, T. (1988). ldentification of two groups of leghemoglobin genes in alfalfa (Medicago sativa) and a study of their expression during root nodule development. Plant MOI. Biol. 11, 761-772. Barker, D.G., Bianchi, S., Blondon, F., DattB, Y., Duc, G., Flament, P., Gallusci, R, Gbnier, R, Guy, R, Muel, X., Tourneur, J., DBnarib, J., and Huguet, T. (1990). Medicagotruncatula, a model plant for studying the molecular genetics of the Rhizobium-legumesymbiosis. Plant MOI. Biol. Rep. 8, 40-49. Bevan, M.W. (1984). Binary Agrobacterium vectors for plant transformation. Nucl. Acids Res. 12, 8711-6721.

Caetanodnoll6ss, G., and Gresshoff, P.M.(1991). Alfalfa controls nodulation during the onset of Rhizobium-inducedcortical cell division. Plant Physiol. 95, 366-373. Cavener, D.R., and Ray, S.C. (1991). Eukaryotic start and stop translation sites. Nucl. Acids Res. 19, 3185-3192. Chabaud, M., Passiatore, J.E., Cannon, F., and Buchanan-Wollaston, V. (1988). Parametersaffectingthe frequency of kanamycin-resistant alfalfa obtained by Agrobacterium tumefaciens-mediatedtransformation. Plant Cell Rep. 7, 512-516. Deak, M., Kiss, G.B., Koncz, C., and Dudits, D. (1986). Transformation of Medicago by Agrobacferium-mediatedgene transfer. Plant Cell Rep. 5, 97-100. Debellb, F., Rosenberg, C., Vasse, J., Maillet, F., Martinez, E., Dbnarib, J., and Truchet, G. (1986). Assignment of symbiotic developmental phenotypesto common and specific nodulation (nod) genetic loci of Rhizobium meliloti. J. Bacteriol. 168, 1075-1086. de Billy, F., Barker, D.G., Gallusci, t?, and Trochet, G. (1991). Leghaemoglobingene transcriptionis triggeredin a single cell layer of the indeterminatenitrogen-fixing root nodule of alfalfa. PlantJ. 1,27-35. Dbnarib, J., and Roche, P. (1992). Rhizobium nodulation signals. In Molecular Signals in Plant-Microbe Interactions, D.P.S. Verma, ed (Boca Raton, FL: CRC Press), pp. 295-324. Diaz, C.L., Melchers, L.S., Hooykaas, P.J.J., Lugtenberg, B.J.J., and Kijne, J.W. (1989). Root lectin as a determinant of host-plant specificity in the Rhizobium-legumesymbiosis. Nature 338,579-581. Feinberg,A R , and Vogelstein, 8. (1983). A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132, 6-13. Fisher, R.F., and Long, S.R. (1992).Rhizobium-plantsignal exchange. Nature 357, 655-660. Gallusci, P., Dedieu, A., Journet, E.P., Huguet, T., and Barker, D.G. (1991). Synchronousexpression of leghaemoglobingenes in Medícago truncatula during nitrogen-fixingroot nodule development and response to exogenously supplied nitrate. Plant MOI. Biol. 17, 335-349. Gelfand, D.H., and White, T.J. (1990). Thermostable DNA polymerases. In PCR Protocols: A Guide to Methodsand Applications, M.A. Innis, D.H. Gelfand, J.J. Sninsky, and T.J. White, eds (New York: Academic Press), pp. 129-141. Gloudemans, T., Bhuvaneswari, T.V., Moerman, M., van Brussel, T., van Kammen, A., and Bisseling, T. (1989). lnvolvement of Rhizobium leguminosarumnodulation genes in gene expression in pea root hairs. Plant MOI. Biol. 12, 157-167, Govers, F., Harmsen, H., Heldstra, R., Michlelsen, P., Prins, M., van Kammen, A., and Bissellng, T. (1991). Characterizationof the pea ENODl2Bgene and expression analyses of the two ENODlP genes in nodule, stem and flower tissue. MOI. Gen. Genet. 228, 160-166. Hoekema, A., Hirsch, P.R., Hooykaas, P.J.J., and Schilperoort, R.A. (1983). A binary plant vector strategy based on separation of virand T-regionsof the Agrobacterium tumefaciens Ti-plasmid. Nature 303, 179-180.

Bhuvaneswari, T.V., Bhagwat, A.A., and Bauer, W.D. (1961). Transient susceptibilityof root cells in four common legumesto nodulation by rhizobia. Plant Physiol. 68, 1144-1149.

Holsters, M., de Waele, D., Depicker, A., Messens, E., Van Montagu, M., and Schell, J. (1978). Transfection and transformation of A. tumefaciens. MOI.Gen. Genet. 163, 181-187.

Boivin, C., Camut, S., Malplca, C.A., Ruchet, G., and Rosenberg, C. (1990). Rhizobiummeliloti genes encoding catabolismof trigonelline are inducedunder symbiotic conditions. Plant Cell2,1157-1170.

Hong, J.C., Nagao, R.T., and Key, J.L. (1990). Characterization of a proline-richcell wall protein gene family of soybean. J. Biol. Chem. 265, 2470-2475.


Jetferson, R.A., Kavanagh, T.A., and Bevan, M.W. (1987). GUS fusions: P-Glucuronidaseas a sensitive and versatile gene fusion marker in higher plants. EM60 J. 6, 3901-3907. Kijne, J.W. (1991).The Rhizobiuminfection process. In Biological Nitrogen Fixation, G. Stacey, H. Evans, and R. Burris, eds (London: Chapman and Hall), pp. 348-397. Lerouge, P., Roche, P., Faucher, C., Malllet, F., Truchet, G., PromQ, J.-C., and DenariQ, J. (1990). Symbiotic host-specificity of Rhizobium melilotiis determined by a sulphated and acylatedglucosamine oligosaccharide signal. Nature 344, 781-784. Long, S.R. (1989).Rhizobium-legumenodulation: Life together in the underground. Cell 56, 203-214. Lullien, V., Barker, D.G., de Lajudle, P., and Huguet, T. (1987). Plant gene expression in effective and ineffectiveroot nodules of alfalfa (Medicago sativa). Plant MOI. Biol. 9, 469-478. Murphy, G., and Kavanagh, T. (1988).Speeding-upthe sequencing of double-stranded DNA. Nucl. Acids Res. 16, 5198. Nap, J.-P., and Bissellng, T. (1990). Developmental biology of a plant-prokaryotic symbiosis: The legume root nodule. Science 250, 948-954. Petlt, A., Stougaard, J., Kühle, A., Marcker, K.A., and TempQ, J. (1987). Transformation and regeneration of the legume Lotus corniculatus: A system for molecular studies of symbiotic nitrogen fixation. MOI. Gen. Genet. 207, 245-250. Roche, I?,DebellQ, F., Malllet, F., Lerouge, P., Faucher,C., Truchet, G., DQnarlQ, J., and PromQ,J.-C. (1991).Molecular basis of symbiotic host specificity in Rhizobium meliloti: nodH and nodPQ genes encode the sulfation of lipo-oligosaccharide signals. Cell 67, 1 131-1 143. Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989).MolecularCloning: A Laboratory Manual, 2nd ed. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press). Sanger, F., Nlcklen, S., and Coulson, A.R. (1977).DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci. USA 74, 5463-5467.

121 1

Schenk, R.V., and Hildebrandt, A.C. (1972).Medium and techniques for inductionand growth of monocotyledonousand dicotyledonous plant cell cultures. Can. J. Bot. 50, 199-204. Scheres, B., van de Wiel, C., Zalensky, A., Horvath, B., Spaink, H., van Eck, H., Zwartkruis, F., Wolters, A.-M., Gloudemans, T., van Kammen, A., and Bisseling, T. (1990).The fNOM2 gene product is involved in the infection process during the pea-Rhizobium interaction. Cell 60, 281-294. Spaink, H.P., Sheeley, D.M., van Brussel, A.A.N., Glushka, J., York, W.S., Tak, T., Geiger, O., Kennedy, E.P., Reinhold, V.N., and Lugtenberg, B.J.J. (1991).A nove1highly unsaturated fatty acid moiety of Iipo-oligosaccharide signals determines host specificity of Rhizobium. Nature 354, 125-130. Strickland, S.G., Nlchol, J.W., McCall, C.M., and Stuart, D.A. (1987). Effect of carbohydrate source on alfalfa somatic embryogenesis. Plant Sci. 48, 113-121. Truchet, O., Roche, P,. Lerouge, P., Vasse, J., Camut, S.,de Billy, F., PromQ, J.-6, and Denarib, J. (1991). Sulphated lipo-oligosaccharide signals of Rhizobium meliloti elicit root nodule organogenesis in alfalfa. Nature 351, 670-673. van de Wiel, C., Scheres, E., Franssen, H., van Lierop, M.J., van Lammeren, A., van Kammen, A., and Bisseling, T. (1990).The early nodulintranscript fNOD2 is locatedin the nodule-specific parenchyma (inner cortex)of peaand soybeannodules. EMBOJ. 9,l-7. Vasse, J., de Billy, F., Camut, S., and Truchet, G. (1990).Correlation between ultrastructuraldifferentiation of bacteroids and nitrogenfixation in alfalfa nodules. J. Bacteriol. 172, 4295-4306. Verma, D.P.S. (1992).Signals in root nodule organogenesisand endocytosis of Rhizobium. Plant Cell 4, 373-382. Vincze, E., and Kiss, G.B. (1990).A phosphategroup at the COS ends of phage lambda is not a prerequisite for in vitro packaging: An alternative method for constructing genomic libraries using a new phasmid vector, pGY97. Gene 96, 17-22. Von Heijne, G. (1983).Patterns of amino acids near signal-sequence cleavage sites. Eur. J. Biochem. 133, 17-21. Wyatt, R.E., Nagao, R.T., and Key, J.L. (1992).Patterns of soybean proline-richprotein gene expression. Plant Cell 4, 99-110.

Rhizobium meliloti elicits transient expression of the early nodulin gene ENOD12 in the differentiating root epidermis of transgenic alfalfa. M Pichon, E P Journet, A Dedieu, F de Billy, G Truchet and D G Barker Plant Cell 1992;4;1199-1211 DOI 10.1105/tpc.4.10.1199 This information is current as of March 7, 2018 Permissions

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298 X

eTOCs

Sign up for eTOCs at: http://www.plantcell.org/cgi/alerts/ctmain

CiteTrack Alerts

Sign up for CiteTrack Alerts at: http://www.plantcell.org/cgi/alerts/ctmain

Subscription Information

Subscription Information for The Plant Cell and Plant Physiology is available at: http://www.aspb.org/publications/subscriptions.cfm

© American Society of Plant Biologists ADVANCING THE SCIENCE OF PLANT BIOLOGY