A small gene family of broad bean codes for late nodulins
containing conserved cysteine clusters
Martin Fru¨hling, Ulrike Albus, Natalija Hohnjec, Gerhard Geise, Alfred Pu¨hler,
Andreas M. Perlick *
Uni6ersita¨t Bielefeld,Lehrstuhl fu¨r Genetik,Postfach100131,D-33501Bielefeld,Germany
Received 5 August 1999; received in revised form 25 October 1999; accepted 25 October 1999
Abstract
Five transcripts encoding different members of a nodulin family with conserved cysteine clusters (Cys-X4-Asp-Cys and
Cys-X4-Cys) were identified in broad bean root nodules. They displayed homologies to the early nodulins PsENOD3 and
PsENOD14 and the late nodulin PsNOD6 from pea. In addition to the occurence of putative secretory signal peptides, the spatial distribution of the cysteine residues was comparable in both the broad bean and the pea nodulins. Based on tissue print hybridizations, we found that the corresponding broad bean genes VfNOD-CCP1, VfNOD-CCP3 and VfNOD-CCP5 were expressed in the interzone II – III and the nitrogen fixing zone III of mature nodules whereas the gene VfNOD-CCP4 was first induced in the prefixing zone II. A strong expression of the VfNOD-CCP2 gene only could be detected the interzone II – III region. Sequence analysis of a genomic VfNOD-CCP1 clone isolated revealed the presence of one intron seperating a first exon encoding the signal peptide from a second exon encoding the cysteine cluster domain of this nodulin. Apart from the multiple presence of the common nodulin motifs AAAGAT and CTCTT on both DNA strands of the putative VfNOD-CCP1 promoter region a sequence element resembling the organ specific element of the soybean lbc3 gene promoter was identified. © 2000 Published by Elsevier Science Ireland Ltd. All rights reserved.
Keywords:Symbiotic nitrogen fixation; Tissue print hybridization; Metal binding;Vicia fabaL.
www.elsevier.com/locate/plantsci
1. Introduction
The interaction of soil bacteria belonging to the
genera Rhizobium, Bradyrhizobium and Azorhizo
-bium with leguminous plants leads to the
forma-tion of a novel plant organ — the root nodule [2,23,43]. In mature nodules, nitrogen-fixation is carried out by the microsymbiont.
The elicitation and development of an effective symbiosis involves a mutual exchange of specific
signals between both partners [37] and is accompa-nied by the expression of specific genes in both rhizobia and their host plants [23,45]. Initially, nodules develop from primordia by differentiation of specialized tissues. The nodules are infected by the symbiotic bacteria via infection threads [18]. After their release from these tubular structures, the bacteria are enclosed by a modified plant-derived membrane (peribacteriod membrane) and differentiate into bacteriods capable of reducing atmospheric nitrogen. The central tissues of fully developed indeterminate nodules formed on tem-perate legumes display a characteristic zonation. An apical meristem is followed by the prefixing
zone II, the interzone II/III rich in amyloplasts,
the nitrogen fixing zone III, and the senescence zone IV, according to the nomenclature of Vasse et al. [44].
The sequence data reported will appear in the EMBL database under the accession numbers AJ243461 and AJ243462 (VfNOD-CCP1), AJ243463 (VfNOD-CCP2), AJ243464 (VfNOD-CCP3), AJ243465 (VfNOD-CCP4), and AJ243466 (VfNOD-CCP5), respec-tively.
* Corresponding author. Tel.: +49-521-106-5631; fax: + 49-521-106-5626.
E-mail address: [email protected] (A.M. Perlick)
M.Fru¨hling et al./Plant Science152 (2000) 67 – 77
68
In recent years, the bacterial genetics of nodule formation and nitrogen fixation have been exten-sively studied [9,10]. An increasing number of plant genes specifically expressed in nodules, termed nodulin genes [42], have been identified and subdivided into early and late nodulin genes according to the time point of their expression [23,45]. Early nodulin genes (ENODs) are associ-ated with organogenesis and bacterial invasion of the root nodule. The sequences of many early nodulins (ENOD2 and ENOD12, for example) suggest that they represent proline-rich proteins most probably involved in cell wall biosynthesis. In general, late nodulin gene (NOD) products are thought to be involved in nodule function and include the oxygen transporter leghemoglobin,
en-zymes of carbon and nitrogen metabolism,
proteins located in the peribacteroid interface as well as a number of proteins, the functions of which remain to be identified [6].
To investigate the organ-specific gene expression
in broad bean (Vicia faba L.) root nodules, we
have constructed a nodule-specific cDNA library of approximately 700 independent cDNAs [28]. In addition to different leghemoglobins [12], tran-scripts homologous to the early nodulin genes
PsENOD2 [41], PsENOD5 and PsENOD12 [35]
were isolated [28]. We also detected a family of at least five nodule-specifically expressed broad bean
genes encoding glycine-rich proteins [20,36],
VfNOD28/32 the homologue ofMsNOD25 [19,21] and VfNOD32 encoding a narbonin-like nodulin with homologies to chitinases [29].
In this paper, we report on the characterization of five broad bean transcripts encoding a family of late nodulins with conserved cysteine-clusters. We describe the spatial and temporal expression of the transcripts in broad bean nodules and finally we report on the isolation and analysis of a corre-sponding genomic sequence.
2. Methods
2.1. Biological material, cDNA and genomic libraries
Broad bean plants (Vicia faba L. cv. Kleine
Thu¨ringer) were grown in the greenhouse in sterile clay granules (Seramis) using surface-sterilized seeds (saturated Ca-hypochlorite, 30 min).
Nodu-lated plants were obtained by inoculation of 2
day-old seedlings with Rhizobium leguminosarum
bv. 6iciae VF39 [30]. Flowers and seeds were analysed from field grown plants inoculated 3 days after sowing. The nodule cDNA library was
con-structed in lgt11 from poly(A)+ RNA isolated
from root nodules of V. faba L. cv. Kleine
Thu¨ringer [28]. A broad bean genomic library was
prepared in lEMBL3 [11] according to Sambrook
et al. [33].
2.2. Library screening, isolation of nucleic acids and recombinant DNA techniques
Recombinant lgt11 and lEMBL3 phages were
plated and screened for positive clones as de-scribed previously [20]. Three rounds of screening were performed to obtain single positive plaques. Isolation of phage DNA was carried out using standard protocols [33]. Plasmid DNA was
iso-lated from E. coli XL1-Blue using the ‘Plasmid
Mini Kit’ (Qiagen) according to the manufactur-er’s instructions. Probe DNA was extracted from agarose gels using the ‘QIAEX Gel Extraction Kit’ (Qiagen). For Northern blotting, RNA was iso-lated from nodules (32 days after sowing), unin-fected roots (32 days after sowing), leaves (32 days after sowing), seeds (90 days after sowing), epi-cotyls (8 days after sowing), stems (12 days after sowing) and flowers (60 days after sowing) of
broad beans using standard protocols [5].
Poly(A)+ RNA fractions were isolated by one
cycle of oligo(dT)-cellulose chromatography.
The ‘5%-AmpliFINDER RACE KIT’ (Clontech)
was used to perform 5% RACE reactions. The
cDNA synthesis was carried out according to the manufacturer’s instructions using gene-specific
primers. For subsequent PCR amplification of 5%
cDNA fragments gene-specific nested primers were used. To generate overlapping sequencing clones, exonuclease III digestions were carried out using the ‘Double Stranded Nested Deletion Kit’ (Phar-macia) according to the manufacturer’s instruc-tions. All other in vitro DNA manipulations were carried out using standard protocols [33].
2.3. DNA sequencing and analysis
gels were run on the ‘A.L.F. DNA Sequencer’ (Pharmacia) using sequencing gel mixes of stan-dard composition. All sequences reported here were determined from both strands. DNA se-quence data were read using the ‘A.L.F. MAN-AGER V3.0’ software (Pharmacia) and analysed using the programme ‘ANALYSEQ’ [38] and ‘LFASTA’ (based on [27]). Multiple sequence alignments were carried out using the program
‘CLUSTAL’ [14] from the PC/Gene software
package (IntelliGenetics, release 6.8). Predictions of signal peptides and their cleavage sites were carried out using the ‘SignalP world wide web server’ (release V1.1) [26].
2.4. Northern and cDNA-cDNA hybridizations
Northern blotting and hybridizations were car-ried out as described previously [28]. About 50 ng of probe DNA isolated from agarose gels were
labelled with 50 mCi of a32P-dATP according to
Feinberg and Vogelstein [8]. Stringent washes were
carried out at room temperature using 2×SSC,
0.1% (w/v) SDS (5 min) and at 68°C using 0.2×
SSC, 0.1% (w/v) SDS (twice for 30 min each).
For cDNA-cDNA hybridizations, 0.2 mg EcoRI
digested DNA from different cDNA clones was separated electrophoretically and blotted onto Hy-bond-N nylon membranes (Amersham) using
stan-dard protocols [33]. Radioactively labeled
first-strand cDNA probes were synthesized from 1
mg poly(A)+ RNA according to Fru¨hling et al.
[12] and used immediately for hybridization. Fil-ters were hybridized in a solution containing 50
mM Na phosphate pH 7.0, 5×SSC, 0.1% (w/v)
lauroylsarcosin, 2% (w/v) blocking reagent
(Boehringer), 7% (w/v) SDS and 50% (v/v)
for-mamide for 48 h at 42°C. Stringent washes were
carried out as described for Northern
hybridizations.
2.5. Tissue print hybridizations
Longitudinal sections of mature broad bean nodules (harvested 32 days after sowing) were printed on Hybond-N nylon membranes (Amer-sham) as described [36]. Hybridizations against digoxigenin-labeled antisense riboprobes were car-ried out as reported by Ku¨ster et al. [20]. Stringent washes and detection of hybridizing transcripts were carried out according to Kessler [17]. As a
control, prints were hybridized against the corre-sponding sense probes. In none of the cases ex-pression above background was observed.
To relate hybridizing regions to nodule zones, sections used for tissue-printing were stained for
starch in Lugol’s solution containing 1% (w/v) KI
and 1% (w/v) I2in distilled water. Stained sections
were photographed at the same magnification as the tissue-print filters.
3. Results
3.1. Fi6e broad bean cDNAs encode small
polypeptides characterized by conser6ed cysteine clusters
Preliminary sequence data indicated that incom-plete cDNAs from five different clone groups of a broad bean nodule-specific cDNA library [28] en-coded polypeptides with conserved cysteine clus-ters (cysteine cluster proteins: CCPs) [12]. The corresponding genes were designated CCP1, CCP2, CCP3, VfNOD-CCP4 and VfNOD-CCP5, respectively. To isolate and determine transcript sequences covering the entire CCP coding regions, we rescreened a nodule cDNA library with appropriate probes. In addi-tion the RACE-PCR technique was used to com-plete cDNAs which did not extend to the
5%-untranslated transcript regions. The sequences
Table 1
Comparison of five broad bean nodulins with conserved cyteine clustersa
Primary Amino acid sequence identity to (%) Transcript Processed
polypeptide polypeptide
VfNOD-kDa VfNOD- VfNOD- VfNOD- VfNOD- ENOD3 ENOD14 NOD6 aa
kDa aa
CCP2 CCP3 CCP5
CCP1 CCP4
¯ 34.8 31.9 28.6 28.4 27.8
40 4.45 23.6 28.2
7.44 65
VfNOD-CCP1
4.12
62 6.98 37 34.8 ¯ 35.7 43.9 24.6 36.2 35.3 34.3
VfNOD-CCP2
31.9 35.7 ¯ 43.7 22.1 36.8
40 27.4
7.57 4.63 35.2
VfNOD- 66 CCP3
28.6 43.9 43.7 ¯ 30.3 54.3 38.8
VfNOD- 64 7.42 40 4.67 33.8
CCP4
28.4 24.6 22.1 30.3 ¯ 29.6 32.4
33 3.69 24.2
VfNOD- 59 6.67 CCP5
CCPs meet all the criteria of a functional eukary-otic signal sequence. The cleavage sites were
pre-dicted to be between Ser25 and Thr26 for
VfNOD-CCP1, between Ala25 and Gln26 for
VfNOD-CCP2, between Gly26 and Glu27 for
VfNOD-CCP3, between Ala24 and Asn25 for
VfNOD-CCP4 and between Ala26 and Ser27 for
VfNOD-CCP5. The most striking feature of the C-terminal polypeptide domain were four con-served cysteine residues. They were arranged in
two clusters of the form Cys-X4-Asp-Cys and
Cys-X4-Cys, respectively. As is evident from Fig. 1, the
distance between the cysteine clusters vary be-tween 9 amino acids within VfNOD-CCP1 and
VfNOD-CCP3 and 14 amino acids within
VfNOD-CCP4. From these results we concluded that the five CCP transcripts encode members of a small broad bean polypeptide family.
Database searches revealed homologies of the different CCPs to the early nodulins PsENOD3
and PsENOD14 [35] and the late nodulin
PsNOD6 [16] from pea (Fig. 1). We could not detect any other significant homologies between the broad bean CCPs and protein sequences in the databases. The homologous pea nodulins dis-played the same characteristics as the CCPs. In addition to the occurence of putative signal pep-tides, the spatial distribution of the conserved
cysteine residues was comparable. In some cases the amino acid identities between the CCPs and the pea nodulins exceeded the identities between the members of the broad bean polypeptide fam-ily. For example VfNOD-CCP4 and PsENOD3 matched in 54.3% of all amino acid residues (see Table 1).
3.2. The CCP transcripts identified are exclusi6ely
expressed in root nodules
Our previous analysis showed that the CCP transcripts were expressed strongly in broad bean root nodules [12]. To analyse the expression pat-terns in greater detail, northern blot hybridizations were carried out (Fig. 2). These experiments re-vealed that the expression of the five CCP genes under investigation was restricted to nodules. No hybridizing transcripts could be detected in unin-fected roots, leaves, seeds, epicotyls, stems and flowers (see Fig. 2), even after overexposure (data not shown). The mRNAs identified in nodules were about 450 bases long, as judged by RNA markers. Considering the very short poly(A) tails in the cDNAs isolated, these transcript lengths were in accordance to the CCP sequences pre-sented. As determined by cDNA-cDNA hybridiza-tions (data not shown) CCP transcripts were first
M.Fru¨hling et al./Plant Science152 (2000) 67 – 77
72
Fig. 2. Expression of CCP genes in different broad bean tissues. Northern blots containing 30mg of total RNA from root nodules, uninfected roots, leaves, seeds, epicotyls, stems and flowers, were hybridized against different CCP cDNA probes. The probes and the deduced length of hybridizing transcripts are indicated on the right. Abbreviation: kb, kilo-bases.
3.3. The CCP genes are expressed in central nodule tissues
To elucidate the spatial distribution of the CCP transcripts in broad bean root nodules, tissue print hybridizations were performed. Longitudinal sec-tions of mature nodules were printed onto nylon membranes and hybridized to CCP sense and anti-sense riboprobes. Representative results of these experiments are shown in Fig. 3. To relate hy-bridizing regions to distinct nodule zones we
visu-alized the interzone II/III in the nodule sections
used for printing by staining for starch (see Fig. 3B) [44]. In general, hybridization signals occured exclusively with the antisense probes (data not shown) and were found to be restricted to the central tissues of the nodule (Fig. 3A). No hy-bridization was detected in peripheral tissues or in the nodule meristem. CCP1, VfNOD-CCP3, VfNOD-CCP4 and VfNOD-CCP5
tran-scripts were located in the interzone II/III and the
nitrogen fixing zone III. VfNOD-CCP1 transcripts were additionally detectable in the ineffective zone IV, whereas detection of VfNOD-CCP4 tran-scripts extended to the distal region of the nodule and comprised large areas of the prefixing zone II. VfNOD-CCP2 transcripts were predominantly
present in the interzone II/III. In much lower
amounts hybridizing VfNOD-CCP2 transcripts were also found to be dispersed in the distal region of the nitrogen fixing zone III.
3.4. Genomic organsisation of VfNOD-CCP1
As a step towards the characterization of the genomic organization of CCP nodulin genes, we isolated a 20 kb VfNOD-CCP1 fragment from a genomic broad bean library. Subcloning resulted
in the identification of a 4.4 kb EcoRI/SalI
frag-ment in clone c2 – 19, which hybridized to a
VfNOD-CCP1 cDNA. Fig. 4 shows the sequence
of a 2424 bp EcoRI/SphI subfragment of this
clone. Sequence analysis revealed that the
VfNOD-CCP1 gene consisted of two exons that showed sequence identity to the VfNOD-CCP1 transcript sequence determined. The two exons were interrupted by an 99 bp intron separating the first exon encoding the secretory signal peptide from the second exon encoding the cysteine cluster
domain (see Fig. 4). The exon/intron boundaries
identified were in good agreement with the
consen-detectable inRhizobium infected broad bean roots
sus sequences of splice junctions in genes of di-cotyledonous plants [13]. Upstream of the first base corresponding to the full length VfNOD-CCP1 transcript sequence, the sequence most simi-lar to a consensus of sequences surrounding nodulin gene TATA boxes could be identified at
position −37 to −25 (ACACTATAAATTG, 10
out of 13 bases matched the consensus reported by Joshi [15]). In addition, an analysis of the
com-plete −1759/ +1 putative promoter region of the
VfNOD-CCP1 gene revealed the multiple presence
of the sequence motifs AAAGAT and CTCTT, characteristic of leghemoglobin and other nodulin gene promoters [34,39], on both strands of the DNA sequence (see Fig. 4). The CTCTT sequence
located from position −173 to −169 was
iden-tified as part of a short element, which resembled the OSE (organ-specific element) of the soybean leghemoglobin lbc3 gene promoter [37] both in sequence and position (Fig. 5A). Apart from these motifs similar to published promoter motifs, a 12 bp inverted repeat and a 17 bp direct repeat were
M.Fru¨hling et al./Plant Science152 (2000) 67 – 77
74
Fig. 4. Sequence of a 2424 bpEcoRI/SphI fragment from the genomic VfNOD-CCP1 clone c2 – 19. The deduced amino acid sequence of VfNOD-CCP1 is printed in bold type above the DNA sequence with the extent of a secretory signal peptide being underlined. The putative signal peptide cleavage site is marked by an vertical arrow. An intron interrupting the VfNOD-CCP1 coding region is displayed in small letters. The first base corresponding to the full length VfNOD-CCP1 transcript sequence is indicated by an dot and was chosen as position +1. In the upstream region, the putative TATA box sequence is underlined twice. The sequence element resembling leghemoglobin OSE sequences are shadowed, while nodulin consensus sequences of the type AAAGAT and CTCTT on both strands of the DNA are underlined. A inverted repeat and a direct repeat are marked by arrows below the sequence.
located on the VfNOD-CCP1 putative promoter (Fig. 5B). The significance of these motifs for theactivity of the putative VfNOD-CCP1 pro-moter has to be demonstrated by the analysis of
promoter-reporter-gene fusions in transgenic
plants.
4. Discussion
exper-iments indicated that these transcripts occur exclu-sively in nodules. Hybridizing mRNAs were not detected in any other broad bean tissue tested. Taking also into account the first appearance of the transcripts in nodules 1 day after the onset of leghemoglobin expression the corresponding gene products can be regarded as late nodulins.
The amino acid sequences deduced from the broad bean transcript revealed homologies to a family of pea nodulins including the early nodulins PsENOD3 and PsENOD14 [35] and the late nodulin PsNOD6 [16]. Whereas in general the overall identities between the sequences are quite low, a bipartite domain structure of the broad bean and the pea nodulins was found to be well conserved, indicating a common ancestry of these nodulins. All of them are small polypeptides of about 7 kDa containing a N-terminal hydrophobic region probably serving as a secretory signal pep-tide for targeting to the plant endomembrane sys-tem. The common feature of the processed polypeptides representing the second domain are four cysteine residues which are arranged in two clusters. Considering these cysteine clusters we designated the five late nodulins VfNOD-CCP1,
VfNOD-CCP2, VfNOD-CCP3, VfNOD-CCP4
and VfNOD-CCP5 according to the nomenclature rules for nodulins [24,25,42]. It is obvious to spec-ulate that the different members of the CCP nodulin families from both pea and broad bean have similar functions in root nodules close related to their highly conserved cysteine residues. Since the spatial distribution of these residues was also
found in metal binding proteins [1], Scheres et al. [35] assumed a role for PsENOD3 and PsENOD14 in metal transport. A similar function was pro-posed for members of a soybean nodulin family characterized by the prescence of a putative signal peptide and the existence of two conserved
do-mains containing each two Cys-X7-Cys motifs
[32,34]. Apart from the correlation in structural features no sequence homologies were found
be-tween the soybean nodulins and the CCP
nodulins. However, whether or not the broad bean, pea and soybean nodulins mentioned are able to bind metal ions remains to be established biochemically. In any case, the supply of metal ions to the microsymbiont by the plant must be
part of the Rhizobium-legume interaction. The
bacteroids which for example require cobalt for the synthesis of vitamin B12 [7] and high amounts of molybdenum and iron for the synthesis of nitrogenase [22] totally depend on the plant for nutrition. Therefore, the existence of nodule-spe-cific metal transport proteins would be expected. They could be part of the peribacteroid space which constitute the symbiotic interface or could contribute to metal storage in the vacuoles of infected nodule cells. Interestingly, the early nodulin genes PsENOD3 and PsENOD14 and the late nodulin gene PsNOD6 are only expressed in infected cells [16,35]. In mature nodules the PsNOD6 mRNA is first detectable at the begin-ning of the interzone II – III and this transcript is present at a constant level in the older cell layers of the central tissue. A comparable spatial
M.Fru¨hling et al./Plant Science152 (2000) 67 – 77
76
sion pattern was observed for the broad bean nodulin genes VfNOD-CCP1, VfNOD-CCP3 and VfNOD-CCP5 whereas the VfNOD-CCP4 gene and the early nodulins genes PsENOD3 and PsENOD14 are first induced in the prefixing zone II [35]. The distribution of the CCP transcripts VfNOD-CCP1, VfNOD-CCP3 and VfNOD-CCP5 hints towards the requirement of the encoded nodulins in later stages of nodule development, whereas the VfNOD-CCP4 gene product might additionally be related to the infection process. In contrast to all other CCP nodulins tested a strong expression of the VfNOD-CCP2 gene only could be detected the interzone II – III region. Therefore the VfNOD-CCP2 gene could be regarded as a marker gene for the interzone II – III region in broad bean nodules like the previously described early nodulin gene VfENOD-GRP3 [20].
Considering the differences found in the spatial expression patterns of the CCP genes we assume that their regulation involves different mecha-nisms. The isolation of the VfNOD-CCP1 gene and its putative promoter region was a first step towards an analysis of the regulation of CCP gene expression. Sequence analysis of the genomic VfNOD-CCP1 fragment revealed the presence of an intron seperating two exons encoding the puta-tive signal peptide and the cysteine cluster domain of the VfNOD-CCP1 polypeptide. The same ge-nomic organisation resembling the bipartite do-main structure of the CCP nodulins was also found in the PsNOD6 gene [16], which further underlines the common ancestry of these nodulins. A number of sequence elements involved in medi-ating nodule specific expression have already been identified for several late nodulin gene promoters [3,4]. Computer searches for these motifs within the VfNOD-CCP1 promoter region identified mul-tiple copies of the common nodulin sequences AAAGAT and CTCTT. As is evident from Fig. 5A, the typical arrangement of AAAGAT and CTCTT subelements found in the soybean lbc3 OSE [39] and the Sesbania glb3 NICE elements [40] was fulfilled by one sequence motif, although the AAAGAT sequence itself was not perfectly conserved. Ramlov et al. [31] and Szczyglowski et al. [40] demonstrated that the CTCTT motif were imperative for the function of these elements, whereas mutations in the AAAGAT sequence had less pronounced effects. This observation could explain the lack of conservation of the AAAGAT
subsequence in the corresponding VfNOD-CCP1 promoter element. However, to assess the rele-vance of all sequence elements identified on the DNA level for the nodule specific expression of the VfNOD-CCP1 gene, further experiments using promoter-reporter-gene fusions in transgenic nod-ules have to be performed. To investigate the properties and functions of the broad bean CCP nodulins identified in more detail, we intend to characterize these polypeptides biochemically.
References
[1] J.M. Berg, Potential metal-binding domains in nucleic acid binding proteins, Science 232 (1986) 485 – 487. [2] N.J. Brewin, Development of the legume root nodule,
Annu. Rev. Cell. Biol. 7 (1991) 191 – 226.
[3] F.J. De Bruijn, J. Schell, Regulation of plant genes specifically induced in developing and mature nitrogen-fixing nodules:cis-acting elements and trans-acting fac-tors, in: D.P.S. Verrma (Ed.), Control of Plant Gene Expression, CRC Press, Boca Raton, FL, 1992, pp. 241 – 258.
[4] F.J. De Bruijn, R. Chen, S.Y. Fujimoto, A. Pinaev, D. Silver, K. Szczyglowski, Regulation of nodulin gene ex-pression, Plant Soil 161 (1994) 59 – 68.
[5] S.C. de Vries, J. Springer, J.G.H. Wessels, Diversity of abundant mRNA sequences and patterns of protein syn-thesis in etiolated and greened pea seedlings, Planta 156 (1982) 129 – 135.
[6] A.J. Delauney, D.P.S. Verma, Cloned nodulin genes for symbiotic nitrogen fixation, Plant Mol. Biol. Rep. 6 (1988) 279 – 285.
[7] R.O.D. Dixon, C.T. Wheeler, Biochemical, physiological and environmental aspects of symbiotic nitrogen fixation, in: J.C. Gordon, C.T. Wheeler (Eds.), Biological Nitro-gen Fixation in Forest Ecosystems: Foundations and Applications, Nijhoff/Junk, The Hague, The Nether-lands, 1983, pp. 107 – 171.
[8] A.P. Feinberg, B. Vogelstein, A technique for radiolabel-ing DNA restriction endonuclease fragments to high specific activity, Anal. Biochem. 132 (1983) 6 – 13. [9] H.M. Fischer, Genetic regulation of nitrogen fixation in
Rhizobia, Microbiol. Rev. 58 (1994) 352 – 386.
[10] R.F. Fisher, S.R. Long, Rhizobium-plant signal ex-change, Nature 357 (1992) 655 – 660.
[11] A.M. Frischauf, H. Lehrach, A. Poustka, N. Murray, Lambda replacement vectors carrying polylinker se-quences, J. Mol. Biol. 170 (1983) 827 – 842.
[12] A. Fru¨hling, H. Roussel, V. Gianinazzi-Pearson, A. Pu¨h-ler, A.M. Perlick, The Vicia faba leghemoglobin gene
VfLb29 is induced in root nodules and in roots colonized by the arbuscular mycorrhizal fungusGlomus fascicula
-tum, Mol. Plant Microbe Interact. 10 (1997) 124 – 131. [13] B.A. Hanley, M.A. Schuler, Plant intron sequences:
[14] D.G. Higgins, P.M. Sharp, CLUSTAL: a package for performing multiple sequence alignment on a microcom-puter, Gene 73 (1988) 237 – 244.
[15] C.P. Joshi, An inspection of the domain between putative TATA box and translation start site in 79 plant genes, Nucleic Acids Res. 15 (1987) 6643 – 6653.
[16] I. Kardailsky, W.-C. Yang, A. Zalensky, A. van Kammen, T. Bisseling, The pea late noduling gene PsNOD6 is homologous to the early nodulin genes PsENOD3/14 and is expressed after the leghaemoglobin genes, Plant Mol. Biol. 23 (1993) 1029 – 1037.
[17] C. Kessler, Nonradioactive Labeling and Detection of Biomolecules, Springer-Verlag, Berlin, 1992.
[18] J.W. Kijne, The Rhizobium infection process, in: G. Stacey, R.H. Burris, J. Hardd (Eds.), Biological Nitrogen Fixation, Chapman and Hall, New York, 1992, pp. 349 – 398.
[19] H. Ku¨ster, A.M. Perlick, A. Pu¨hler, Members of a broadbean nodulin family with partial homologies to the alfalfa nodulin 25 are composed of two types of amino acid repeats flanked by unique amino acid sequence termini, Plant Mol. Biol. 24 (1994) 143 – 157.
[20] H. Ku¨ster, G. Schro¨der, M. Fru¨hling, U. Pich, M. Rieping, I. Schubert, A.M. Perlick, A. Pu¨hler, The nod-ule-specificVfENOD-GRP3 gene encoding a glycine-rich early nodulin is located on chromosome I ofVicia fabaL. and is predominantly expressed in the interzone II – III of root nodules, Plant Mol. Biol. 28 (1995) 405 – 421. [21] H. Ku¨ster, M. Fru¨hling, A. Pu¨hler, A.M. Perlick, The
modular nodulins Nvf-28/32 of broad bean (Vicia faba
L.): alternative exon combinations account for different modular structures, Mol. Gen. Genet. 252 (1996) 648 – 657.
[22] T. Ljones, The enzyme system, in: A. Quispel (Ed.), Biology of Nitrogen Fixation, North-Holland Publishing Company, Amsterdam, 1974, pp. 617 – 639.
[23] P. Mylona, K. Pawloski, T. Bisseling, Symbiotic nitrogen fixation, Plant Cell 7 (1995) 869 – 885.
[24] J.P. Nap, T. Bisseling, Developmental biology of a plant-prokaryote symbiosis: the legume root nodule, Science 250 (1990) 948 – 954.
[25] J.P. Nap, T. Bisseling, The roots of nodulins, Physiol. Plant. 79 (1990) 404 – 414.
[26] H. Nielsen, J. Engelbrecht, S. Brunak, G. von Heijne, Identification of prokaryotic and eukaryotic signal pep-tides and prediction of their cleavage sites, Protein Eng. 10 (1997) 1 – 6.
[27] W.R. Pearson, D.J. Lipman, Improved tools for biologi-cal sequence comparison, Proc. Natl. Acad. Sci. USA 85 (1988) 2444 – 2448.
[28] A.M. Perlick, A. Pu¨hler, A survey of transcripts expressed specifically in root nodules of broadbean (Vicia fabaL.), Plant Mol. Biol. 22 (1993) 957 – 970.
[29] A.M. Perlick, M. Fru¨hling, G. Schro¨der, S.C. Frosch, A. Pu¨hler, The broad bean gene VfNOD32 encodes a nodulin with sequence similarities to chitinases that is homologous to (a/b)8-barrel-type seed proteins, Plant Physiol. 110
(1996) 147 – 154.
[30] U.B. Priefer, Genes involved in lipopolysaccharide pro-duction and symbiosis are clustered on the chromosome
ofRhizobium leguminosarumbiovar6iciaeVF39, J.
Bacte-riol. 171 (1989) 6161 – 6168.
[31] K.B. Ramlov, N.B. Laursen, J. Stougaard, K.A. Marcker, Site-directed mutagenesis of the organ-specific element in the soybean leghemoglobinlbc3 gene promoter, Plant J. 4 (1993) 577 – 580.
[32] H.E. Richter, N.N. Sandal, K.A. Marcker, C. Sengupta-Gopalan, Characterization and genomic organization of a highly expressed late nodulin subfamily in soybeans, Mol. Gen. Genet. 229 (1991) 445 – 452.
[33] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning. A Laboratory Manual, second ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 1989.
[34] N.N. Sandal, K. Bojsen, K.J. Marcker, A small family of nodule specific genes from soybean, Nucleic Acids Res. 15 (1987) 1507 – 1519.
[35] B. Scheres, F. van Engelen, E. van der Knaap, C. van de Wiel, A. van Kammen, T. Bisseling, Sequential induction of nodulin gene expression in the developing pea nodule, Plant Cell 2 (1990) 687 – 700.
[36] G. Schro¨der, M. Fru¨hling, A. Pu¨hler, A.M. Perlick, The temporal and spatial transcription pattern in root nodules ofVicia fabanodulin genes encoding glycine-rich proteins, Plant Mol. Biol. 33 (1997) 113 – 123.
[37] M. Schultze, A. Kondorosi, Regulation of symbiotic root nodule developments, Annu. Rev. Genet. 32 (1998) 33 – 57.
[38] R. Staden, The current status and probability of our sequence handling software, Nucleic Acids Res. 14 (1986) 217 – 231.
[39] J. Stougaard, N.N. Sandal, A. Groen, A. Ku¨hle, K.A. Marcker, 5% analysis of the soybean leghemoglobin lbc3
gene: regulatory elements required for promoter activity and organ specificity, EMBO J. 6 (1987) 3565 – 3569. [40] K. Szczyglowski, L. Szabados, S.Y. Fujimoto, D. Silver,
F.J. deBruijn, Site-specific mutagenesis of the nodule-in-fected cell expression (NICE) element and the AT-rich element ATRE-BS2 of the Sesbania rostrata
leghemoglobin glb3 promoter, Plant Cell 6 (1994) 317 – 332.
[41] C. van de Wiel, B. Scheres, H. Franssen, M.-J. van Lierop, A. van Lammeren, A. van Kammen, T. Bisseling, The early nodulin transcript ENOD2 is located in the nodule parenchyma (inner cortex) of pea and soybean root nodules, EMBO J. 9 (1990) 1 – 7.
[42] A. van Kammen, Suggested nomenclature for plant genes involved in nodulation and symbiosis, Plant Mol. Biol. Rep. 2 (1984) 43 – 45.
[43] P. van Rhijn, J. Vanderleyden, The Rhizobium-plant symbiosis, Microbiol. Rev. 59 (1995) 124 – 142.
[44] Vasse, F. de Billy, S. Camut, G. Truchet, Correlation between ultrastructural differentiation of bacteroids and nitrogen fixation in alfalfa nodules, J. Bacteriol. 172 (1990) 4295 – 4306.
[45] J.P.S. Verma, C.-A. Hu, M. Zhang, Root nodule develop-ment: origin, function and regulation of nodulin genes, Physiol. Plant. 85 (1992) 253 – 265.
[46] J. Zimmermann, H. Voss, C. Schwager, J. Stegemann, H. Erfle, K. Stucky, T. Kristensen, W. Ansorge, A simplified protocol for fast plasmid DNA sequencing, Nucleic Acids Res. 18 (1990) 1067.