A family of β-galactosidase cDNAs related to development of vegetative tissue in Cicer arietinum

A family of b-galactosidase cDNAs related to development of vegetative tissue in Cicer arietinum Rocı ´o Esteban, Emilia Labrador, Berta Dopico* Dpto. de Fisiologı ´a Vegetal, Facultad de Biologı ´a, Universidad de Salamanca, Centro Hispano-Luso de Investigaciones Agrarias, Pza Doctores de la Reina s/n, Salamanca 37007, Spain Received 26 April 2004; received in revised form 6 September 2004; accepted 7 September 2004 Available online 30 September 2004 Abstract In the cell wall of Cicer arietinum epicotyls, there are a family of b-galactosidases, one of them named bIII-Gal is a b-galactosidase able to degrade cell wall pectins. The role of the remainder b-galactosidases has not been established. In this paper, we describe the cloning and expression pattern of a family of three C. arietinum b-galactosidase cDNAs (named CanBGal-1, CanBGal-4 and CanBGal-5) and we compare these results with the previously characterized CanBGal-3 cDNA clone, which encode the bIII-Gal. The shared amino acid sequence identity among the four b-galactosidase deduced proteins (named b-Gal, bIII-Gal, bIV-Gal and bV-Gal) ranged from 63% to 81%. All display the putative active site of family 35 of the glycosyl hydrolases. An unusual characteristic of one of the chickpea b-galactosidases (bI-Gal) is the presence at the C-terminus of the enzyme of a galactose binding lectin domain. The CanBGals gene expression along seedlings and adult plant could suggests different roles of their corresponding protein throughout the chickpea plant. The expression of CanBGal-5 is related to young and meristematic stages with high cell division rate, such as the meristematic hook, very young epicotyls, and apical internodes. By contrast, CanBGal-1 and -4 seem to be more strongly related to advanced stages of epicotyl growth, increasing their expression along epicotyl age, and also in basal non-elongating stem internodes. In adult plants, CanBGal-1 shows its highest expression levels in leaves, while CanBGal-4 seems to be better represented in adult roots. This is the first report about several members of the genomic family of b-galactosidases acting during development of vegetative organs. #2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Cell wall; Cicer arietinum; Development; Epicotyl; b-Galactosidase 1. Introduction b-Galactosidases (EC 3.2.1.23), a widespread family of glycosyl hydrolases, are characterized by their ability to hydrolyze terminal, non-reducing b-D-galactosyl residues from b-D-galactosides. Several studies have shown that b-galactosidases catalyze the hydrolysis of terminal galactosyl residues from carbohydrates, glycoproteins and galactolipids. b-Galactosidase action has been reported to release free galactose during normal metabolic recycling of galactolipids, glycoproteins and cell wall components, and to degrade cell wall components during cell wall expansion, cell senescence and fruit ripening [1–5]. Furthermore, under favorable thermodynamic in vitro conditions many b-galactosidases have specific biosynthetic activities through both transglycosylation and reverse hydrolysis [6,7]. In higher plants, softening-related b-galactosidases have been purified from tomato [8,9], apple [3], mango [10] and papaya [11] among other fruits, and their biochemical properties have been characterized. More recently, these physiological and biochemical studies have been supported by the characterization of b-galactosidase genes, such as those described in tomato [12,13] or in Japanese pear [14]. Besides the role of pectins in fruit ripening, there are also several reports indicating that the loss of pectic galactose seems to play an important role in cell wall extensibility and hence pectic polymers, which are highly complex compounds, must be involved in the metabolism of expanding cell walls [15,16]. Fujino and Itoh [17] described a clear www.elsevier.com/locate/plantsci Plant Science 168 (2005) 457–466 * Corresponding author. Tel.: +34 9 2324500x1951; fax: +34 9 23294682. E-mail address: [email protected] (B. Dopico). 0168-9452/$ – see front matter #2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.plantsci.2004.09.009 difference in cell wall architecture between elongating and non-elongating regions, suggesting a modification in the molecular form of pectic polysaccharides during the elongation of epidermal cells. Those authors suggested that neutral sugar side-chains could be involved in the interaction with other cell wall components. b-Galactosidase is one of the enzymes that seems to play a role in the process of modification of the pectin structure, removing b-galactosyl linkages on the neutral sugar side-chains [18–20]. Several studies have focused on the biochemistry of b-galactosidases during cell wall modification during growth [5,21–24], but as far as we know, few studies have been carried out on the molecular characterization of b-galactosidase genes in vegetative elongating organs. The purification of a family of chickpea b-galactosidases in Cicer arietinum epicotyls was first reported by Dopico et al. [25]. One of them, bIII-Gal, was characterized as a protein involved in the cell wall autolytic process [18,23,26]. It was proposed that bIII-Gal might be important in the cell wall changes taking place during elongation. The cDNA corresponding to bIII-Gal, named CanBGal-3, was subsequently cloned [27] and its function in pectin degradation was supported by its expression in a potato system. The role of the remainder b-galactosidases has not been yet established. In this paper, we described the cloning and expression pattern of three C. arietinum b-galactosidase cDNAs (named CanBGal-1, CanBGal-4 and CanBGal-5) and we compared these results with the previously characterized CanBGal-3 cDNA clone, which encode the bIII-Gal. The members of this family of b-galactosidase genes, may be involved in galactose metabolism during cell wall turnover, elongation processes, or architectural pectin changes in C. arietinum plants. 2. Materials and methods 2.1. Plant material Chickpea seeds (C. arietinum L. cv. Pedrosillano) previously sterilized in 0.1% (w/v) sodium hypochlorite were germinated in water in the dark at 25 8C and 80% relative humidity on glass plates covered with filter paper. The growth period ranged from 2 to 8 days, after which the epicotyls were collected. Different plant tissues were collected for studies of gene expression: hooks, epicotyls, mesocotyls (root–epicotyl junction zone), cotyledons and roots from 4-day-old etiolated seedlings. Stems of 11-day-old plants were divided into 5 internodes, numbered 1–5 from base to apex. Roots from 11-day-old plants and folioles, pods, flowers and seeds from adult plants were also collected. Immature folioles (L1) were those coiled up and smaller than 4 mm whereas mature folioles (L2) were those fully expanded and larger than 4 mm. Immature pods (P1) were 1 cm green pods whereas mature pods (P2) were full-grown green pods (about 2.5 cm). Immature seeds (S1) were green seeds from mature pods. Finally, mature dry seeds (S2) were also studied. 2.2. RNA extraction Five grams of frozen material were ground in liquid nitrogen and resuspended in 10 ml of extraction buffer [10 mM Tris–HCl, pH 8.0, containing 1% (w/v) triisopropyl naphthalene sulphonic acid, 6% (w/v) 4-amino salicylic acid and 5% (v/v) phenol mixture (500 g phenol, 70 ml meta-cresol, 0.5 g 8-hydroxyquinoline, 150 ml distilled water)]. Samples were extracted as described in Mun ˜oz et al. [28]. 2.3. Screening for b-galactosidases in a chickpea cDNA library The cDNA library used was constructed from poly (A)+ RNA extracted from epicotyls of 5-day-old etiolated chickpea seedlings grown in water as described by Mun ˜oz et al. [28]. The screening of the library was performed as described in [27]. A heterologous screening of the library was performed using a Malus domestica b-galactosidase clone (pABG1, accession no. L29451, kindly provided by Dr. Ross, New Zealand) as probe. The heterologous screening afforded two different chickpea b-galactosidase cDNAs. A mixture of these two chickpea b-galactosidase clones was used as probe in further homologous screenings in order to isolate as many chickpea b-galactosidase cDNAs as possible. In these screenings two new b-galactosidase cDNAs were isolated. Thus, we finally collected four different chickpea b-galactosidase cDNAs, named CanBGal-1, CanBGal-3, CanBGal-4 and CanBGal-5. 2.4. DNA and deduced protein sequence analysis Plasmid DNA was prepared using the Wizard plus SV miniprep kit (Promega, Madison, WI, USA) based on the alkaline lysis method [29]. Sequence analysis was performed automatically using the dRhodamine terminator cycle sequencing ready reaction kit on an ABI Prism 377 genetic analyzer (both from PE Biosystems, Norwalk, CT, USA). The DNA and deduced protein sequences were compared with other sequences in the EMBL/GenBank and Swall databases, respectively, using the FASTA algorithm [30]. Alignment and the phylogenetic tree of the chickpea and other plant b-galactosidases was done using the Clustal method from the MegAlign program: part of the Lasergene sequence analysis software available from DNASTAR. Putative signal peptides were analyzed based on the SignalP 2.0 program [31] and putative N-glycosylation sites and the b-galactosidase active site [32] were analyzed at the Prosite web page (http://www.expasy.ch/prosite). Conserved domains of the proteins were analyzed using the NCBI R. Esteban et al. / Plant Science 168 (2005) 457–466458 Conserved Domain Search (http://www.ncbi.nlm.nuh.gov/) [33]. 2.5. Northern analysis Northern experiments were performed as described in Mun ˜oz et al. [28]. cDNA fragments from the 30non-coding end of the clones were used as specific probes to minimize any potential cross-hybridization. cDNA specific probes were amplifiedbyPCRinaGeneAmpPCRsystem9700 thermocycler (PE Biosystems, Norwalk, CT, USA), cleaned by the High Pure PCR product purification kit (Roche, Germany), and labelled with 32 Pusingthe Random primed kit (Roche, Germany). Total RNA (10 mg/lane) was electrophoresed, transferred onto Hybond N nylon membranes (Amersham Pharmacia Biotech, UK), and hybridized using the radiolabelled specific probes. After hybridization, membranes were washed twice with 2SSC, 0.1% SDS at 42 8Cfor5min each and twice with 0.1SSC, 0.1% SDS at 42 8Cfor 20 min each. The autoradiographs were analyzed on a Bioimage 60S Image Analyzer (Millipore, Bedford, MA, USA, Visage 4.6 K Software). Subsequent hybridization of the blots using a 500 bp fragment from 18S rRNA as probe was used as internal control to determine the relative amounts of RNA per lane. Northern experiments were performed at least two times each. 2.6. Genomic DNA extraction and Southern analysis Genomic DNA was isolated using the Plant DNA isolation kit (Roche, Germany) from 14-day-old chickpea plant folioles. DNA (10 mg) was digested using HindIII and EcoRI restriction enzymes, separated on agarose gel, and transferred onto a Hybond N nylon membrane (Amersham Pharmacia Biotech, UK) following the manufacturer’s instructions. Blots were hybridized at 65 8C for 12 h in 1% (w/v) SDS, 5SSPE, 5Denhards’solution using the [ 32 P]-labelled specific probe, constructed as described above. Membranes were washed twice with 2SSC, 0.1% SDS at 65 8C for 5 min each, twice with 0.5SSC, 0.1% SDS at 65 8C for 20 min each and twice with 0.1 SSC, 0.1% SDS at 65 8C for 20 min each. 3. Results 3.1. Chickpea b-galactosidase clones Screening of the C. arietinum cDNA library revealed the presence of at least four different cDNAs encoding bgalactosidases: CanBGal-1, CanBGal-3, CanBGal-4 and CanBGal-5. CanBGal-3, CanBGal-4 and CanBGal-5 are full-length clones whose sizes were 2600, 2900 and 2700 bp, respectively. CanBGal-1, whose size was 2656, is not a full-length clone. The mRNA length estimated for this clone by Northern blot was 3200 nt. The proteins deduced from the chickpea CanBGal clones, named bI-Gal, bIII-Gal, bIV-Gal and bV-Gal, had a significant level of shared amino acid sequence identity with one another and also with other published plant b-galactosidases. bIII-Gal and bIV-Gal are the closest at amino acid sequence level (81% identity). bI-Gal and bV-Gal show a lower degree of identity: about 63% between each other and with respect to the other two bgalactosidases. The proteins bIII-Gal, bIV-Gal and bV-Gal present a predicted N-terminal hydrophobic signal peptide, whose putative cleavage site is indicated by an arrow in Fig. 1,as determined by Signal P [31]. The hydrophobic signal peptides are 29, 23 and 26 amino acids long for bIII-Gal, bIV-Gal and bV-Gal, respectively. After the release of the signal peptide, the mature proteins have a molecular mass of 77.9, 78 and 80 kDa for the bIII-Gal, bIV-Gal and bV-Gal, respectively. bIII-Gal and bV-Gal show a high predicted isoelectric point (8.06 and 8.45), whereas the predicted isoelectric point observed for the bIV-Gal was 5.79. No conclusive data could be obtained for bI-Gal, owing to the lack of approximately 140 amino acids at the N terminus of this protein. The four chickpea b-galactosidase proteins display the active site of family 35 of the glycosyl hydrolases described by Henrissat [32] (marked by asterisks in Fig. 1). Position 6 of this active site was different in the four proteins, bIII-Gal and bIV-Gal have Met whereas bI-Gal and bV-Gal have Leu. According to Henrissat [32], Met and Leu are possible conserved susbstitutions. A characteristic of the chickpea b-galactosidases is that they have different length. In particular, bI-Gal is 115 amino acids longer at the C-terminal end than the other chickpea b-galactosidases (Fig. 1). It is interesting to note that this extra peptide of the bI-Gal comprises an unusual galactose binding lectin domain, as point out the analysis of the conserved domains of the proteins using the NCBI Conserved Domain Search [33]. No other chickpea b-galactosidases present this domain. All chickpea b-galactosidases show at least one putative N-glycosylation site. bI-Gal and bIII-Gal have only one, whereas bIV-Gal and bV-Gal have four and two, respectively. The putative N-glycosylation sites are highlighted in Fig. 1. 3.2. Phylogenetic tree The differences among chickpea b-galactosidases are clearly seen in the phylogenetic tree, compiled using the four deduced chickpea proteins and other plant b-galactosidases, based on protein sequence alignment (Fig. 2). The distribution of chickpea b-galactosidases in this tree suggests that there are subgroups of b-galactosidases distinguishable by their sequence. C. arietinum b-galactosidases appear R. Esteban et al. / Plant Science 168 (2005) 457–466 459 separated in three different phylogenetic branches. bI-Gal appears in group I, together with proteins with a molecular mass ranging from 92.1 to 93.6 kDa and 832 to 847 aminoacid residues. The larger size of those proteins was primarily due to an addition of approximately 100 amino acids at their carboxyl termini, as we have described above in bI-Gal (Fig. 1). All the proteins included in this group I present a galactose binding lectin domain. The bV-Gal protein appears in branch II, together with b-galactosidases from Phaseolus vulgaris,Vitis vinifera and Lycopersicon esculentum TBG6. This group is heterogeneous in their molecular mass, but shows several conserved regions R. Esteban et al. / Plant Science 168 (2005) 457–466460 Fig. 1. Alignment of the four deduced b-galactosidase amino acid sequences of the CanBGal clones. Sequences were aligned using the Clustal method in the MegAlign program, part of the Lasergene sequence analysis software program available from DNASTAR (Madison, WI). Boxes indicate amino acid identity with respect to the majority. An arrow indicates the putative signal sequence cleavage site. The active site of family 35 of glycosyl hydrolases is marked by asterisks. Putative N-glycosylation sites are highlighted. within the group (data not shown), such as Gln at the positions 272 and 361, whereas are Glu in the other groups. The proteins analyzed in this group have been related to early development processes in grape [34,35] and tomato [13] fruits. The bIII-Gal and bIV-Gal, together with b-galactosidases from other leguminosae such as Vigna radiata and Lupinus angustifolius are clustered in group IV. All the galactosidases in this group IV are between 721 and 731 amino acids long and have molecular weights ranging from 79 to 82.1. Most of them have been related to cell wall pectin degradation. Although sharing a molecular weight similar to group IV, the proteins clustered in group V present a clear difference in the active site of family 35 of the glycosyl hydrolases described by Henrissat [32]. All of them have Met and Ala at positions 6 and 7, respectively, instead of Leu and Ser, present in most of the plant b-galactosidases described to date. Some b-galactosidases have Met or Ala at those positions but never both together as in group V. The b-galactosidase from Dianthus caryophyllus, detectable in senescing carnation petals and under strict regulation by ethylene [1], appears on a separate branch. 3.3. Expression pattern during growth The temporal expression pattern in epicotyls of the four CanBGal was examined using RNA extracted from epicotyls ranging in age from 2 days (when the epicotyl emergence starts) to 8 days (when the rate of growth starts to fall) (Fig. 3). (Data previously published referred to CanBGal-3 [27] were included to a better discussion of the results). Specific probes were used for each cDNA in order to minimize any potential cross-hybridization. CanBGal-1 and CanBGal-4 mRNAs were detected throughout the growth of R. Esteban et al. / Plant Science 168 (2005) 457–466 461 Fig. 2. Phylogenetic tree representing the evolutionary relationship among the four deduced proteins from the CanBGal cDNAs of Cicer arietinum and several plant b-galactosidases. epicotyls and their mRNA levels increased with the length of the epicotyls. On the contrary, the CanBGal-5 mRNA levels decreased with the age of epicotyls. CanBGal-3 started to be expressed on 4 days and no transcript was observed in very young epicotyls (2 and 3 days) [27]. Expression along the stem internodes from 11-day-old plants (Fig. 4) showed a similar correlation to growth as that found along epicotyl age (Fig. 3). The mRNA levels of CanBGal-1 and CanBGal-4 increased from the apical young internode (5th) to the basal one (1st), whereas CanBGal-5 expression followed the opposite pattern, showing the highest mRNA levels in apical internodes and decreasing towards the basal ones, where almost no expression was detected. CanBGal-3 showed the highest level of transcripts in the 3rd internode and almost no expression was found in the youngest (5th) internode; i.e., the one closest to the meristematic tissue (Fig. 4). 3.4. Tissue specificity in seedlings and adult plants As expected, according to the screening of the library, all the CanBGals were expressed in epicotyls, showing their R. Esteban et al. / Plant Science 168 (2005) 457–466462 Fig. 3. Northern analysis of CanBGal mRNA expression levels in seedling epicotyls during elongation (A). Signals were quantified and normalized according to 18S rRNA hybridization and are presented as integrated optical density (B). Numbers refer to days after sowing. Fig. 4. Differential expression of CanBGals in the five internodes of 11-day-old plants (A). Signals were quantified and normalized according to 18S rRNA hybridization and are presented as integrated optical density (B). Internodes are numbered 1–5 from the basal to the apical one. highest level of expression in these seedling organs, except for CanBGal-4 (Fig. 5). CanBGals transcripts were also detected in other elongating seedlings organs, such as mesocotyls and radicles, but at lower levels than in epicotyls, except for CanBGal-4 where the highest transcript level was found in both mesocotyls and radicles. Only CanBGal-1 was significantly expressed in cotyledons, although a very low mRNA level of CanBGal-4 was also detected in those organs. Transcripts of CanBGal-1, -4 and -5 were found in apical hooks (meristematic tissue) but always at lower levels than in epicotyls. CanBGal-5 expression in hooks represented the second highest transcript level, after that of epicotyls. No CanBGal-3 transcripts were detected in hooks [27]. The expression pattern of CanBGal clones in adult plants was broad. Besides the stem internodes, in adult plants the highest expression of CanBGal-1 and CanBGal-3 was found in mature leaves. Of note was the higher transcript level in old leaves (L2) as compared with young ones (L1). CanBGal-4 and CanBGal-5 showed maximum transcript levels in young pods (P1) and in the case of CanBGal-4, also in adult roots (Fig. 6). All the clones showed very low or no expression in dry seeds and in mature pods. It should be noted that CanBGal-3 was poorly expressed in young leaves and not expressed in adult roots. CanBGal-4 had the broadest range of expression and CanBGal-5 was poorly expressed in mature leaves and flowers (Fig. 6). 4. Discussion Most studies on b-galactosidase genes refer to their expression in ripening fruit, where the degradation of galactan is associated with the ripening process (reviewed by R. Esteban et al. / Plant Science 168 (2005) 457–466 463 Fig. 5. Differential expression of CanBGals in different parts of C. arietinum 4-day-old seedlings (A). Signals were quantified and normalized according to 18S rRNA hybridization and are presented as integrated optical density (B). H: hook; E: epicotyl; M: mesocotyl; C: cotyledon; R: radicle. Fig. 6. Differential expression of CanBGals in different parts of C. arietinum adult plants (A). Signals were quantified and normalized according to 18S rRNA hybridization and are presented as integrated optical density (B). L1: young leaves; L2: mature leaves; P1: immature pods; P2: mature pods; F: flowers; S1: immature seed; S2: dry mature seed; r: adult roots. Brummell and Harpster [36]). However, in elongating vegetative organs, the function of the different b-galactosidase genes has been studied to a lesser extent, although changes in the neutral side chains of pectin that could be mediated by galactosidases during growth have been reported [15,23,37–40]. Kikuchi et al. [41] suggested that the neutral side chains of RG-I, such as galactan and arabinan, play an important role in linkages between pectins and other components, such as cell wall hemicelluloses, enhancing the rigidity of cell walls. In C. arietinum, we have previously reported several b-galactosidases purified from epicotyl cell walls [25], and one of them, bIII-Gal was proposed to be involved in the cell wall changes taking place during elongation [23]. During our search for cDNAs coding b-galactosidases, a total of four cDNA clones, named CanBGal-1, CanBGal-3, CanBGal-4 and CanBGal-5, all with a significant level of sequence identity, were found. These four genes would be members of a small family of b-galactosidase genes. All of them appear to be present in the chickpea genome as a single copy, as determined by Southern blot (data not shown). The high sequence homology of the four clones with one another and with other plant b-galactosidases, together with the presence of the glycosyl hydrolase family 35 consensus sequence [32] (Fig. 1), suggest that they may all have b-galactosidase activity. Other b-galactosidase gene families have been described in plants, such as in L. esculentum with seven members [13] or Carica papaya [11] with three members. In fully sequenced plant genomes such as Arabidopsis thaliana there are 43 database entries corresponding to b-galactosidases of the glycosyl hydrolase family 35, however several of them are just fragments. In Oryza sativa,sevenbgalactosidases appear in the databases. The characterization and analysis of the gene expression of several CanBGals could help to understand the putative role of the corresponding encoded proteins in chickpea cell wall metabolism. In a recent report, the CanBGal-3 was identified as the cDNA corresponding to bIII-Gal and its function in cell wall pectin degradation was supported by its expression in a potato system [27]. Here, we studied the gene expression at the mRNA level of the remainder CanBGals in order to find the possible role for each chickpea b-galactosidases. From the present results we can observed that the expression of CanBGal-5 was related to young and meristematic stages with a high cell division rate, such as meristematic hooks, very young epicotyls and apical internodes (Figs. 3–5). It would appear that the protein bV-Gal is involved in the initial stages of plant development. Few examples of b-galactosidases related to high cell division stages have been described. As far as we know, only Nunan et al. [35] in grape, and Smith and Gross [13] in tomato (TBG6), have reported b-galactosidases whose transcripts were accumulated specifically in the early stages of fruit development; a stage with an intensive cell division activity followed by cell expansion. It is interesting to note that these two proteins appeared together with chickpea bVGal in the group II of the phylogenetic tree, indicating a possible relationship between the structure and function of these proteins. In contrast with CanBGal-5, CanBGal-1 and -4 seem to be more expressed in advanced stages of epicotyl growth, increasing with age and also in basal, non-elongating stem internodes (Figs. 3 and 4). No b-galactosidase genes related to advanced stages of growth have been described, although the expression of several b-galactosidases has been reported to increase during senescence, as in carnation [1] and in asparragus [42,43]. However, the release of galactose in senescence is accompanied by cell wall disassembly, whereas the expression patterns of CanBGal-1 and -4 are highest when there is an increase in cell wall rigidity. We can, therefore, propose that the activity of these proteins could be to hydrolyze an unknown substrate whose modification causes cell wall rigidity, probably by compacting pectins. Even though CanBGal-1 and CanBGal-4 seem to share the same pattern of expression at the mRNA level in epicotyls and stems, it should be noted that the predicted isoelectric points for the bI-Gal and bIV-Gal proteins are quite different. This suggests the possibility that these proteins might be bound to a different substrate located at a different site in the cell wall. Furthermore, in adult plants CanBGal-1 shows its highest expression in leaves, while CanBGal-4 seems to be more important in adult roots. It should also be noted that the CanBGal-1 transcripts are the only ones abundant in cotyledons, whose cell walls could have galactose-rich polymers different from those found in epicotyls, indicating a putative substrate specificity. The presence of a galactose binding lectin domain in the protein bI-Gal suggests a particular role of this protein. Although this domain is far away from the putative active site, and therefore might not affect the catalytic properties of the enzyme, this domain, if proven to be functional could affect the biochemical characteristics of the enzyme containing it. The presence of this domain is not usual between b-galactosidases; according to databases only about 42 proteins among all the plant b-galactosidases described contain it [33]. Trainotti et al. [44] refer two b-galactosidases in strawberry with a similar lectin-like domain, suggesting their possible sugar-binding activity on the basis of the highly conserved structural identity between animal lectins and the C-terminal domain of these two strawberry proteins. The expression pattern of the CanBGal cDNAs studied in the present report was quite different from the expression of CanBGal-3 encoding the bIII-Gal able to hydrolyze cell wall pectins [27]. In that case, the highest level of transcript accompanied the stages of maximum elongation, such as 4-day-old epicotyls [27] or stem internodes with high growth ability (Fig. 4). No transcript was found in the regions with high cell division rate such as apical intenodes, hooks or very young epicotyls. R. Esteban et al. / Plant Science 168 (2005) 457–466464 The CanBGals gene transcription in seedlings and adult plant (Figs. 5 and 6) could suggests specific function and/or location of their corresponding protein throughout the chickpea plant. The idea that these four b-galactosidases from chickpea could be involved in different modifications of galactosyl residues in the cell wall in different phases of development or in different organs is based on the high number of specific natural substrates described for b-galactosidases. An example indicating such high specificity is provided by the b-galactosidase from Copaifera langsdorfficotyledons, with considerable specificity towards xyloglucan oligosaccharide [45]. For hydrolysis, this b-galactosidase recognises only the galactosyl substitution of the xylose adjacent to the non-reducing end glucose residue of xyloglucan oligosaccharides. In a similar way to C. arietinum, it is common to find b-galactosidase isoforms in the same species; these do not usually show the same ability to hydrolyze pectins or other cell wall substrates. In general, these isoforms present differences, indicating that they act in a different way or against different substrates with different degrees of branching [11,25,46]. The cloning of several b-galactosidase isoforms could help to clarify the role of each one in the same or different processes, although to date this point has only been studied in fruits such as tomato [13]. Several b-galactosidase clones have been described in Fragaria ananassa or Phaseolus aureus, but to our knowledge no data on these clones have been published. As far as we know, this is the first report of several members of the genomic family of b-galactosidases acting in vegetative organs. This group of b-galactosidases acts during the development and growth of vegetative organs in C. arietinum and could thus affect different processes and allow a high regulation of the processes involved. More work will be necessary to establish the specific substrate and action of each b-galactosidase. Acknowledgements The authors thank Dr. Gavin S. Ross for providing the cDNA clone pABG1 from Malus domestica. This work was supported by a Grant from the EC (CT972224) and a Grant from the Direccio ´n General de Ensen ˜anza Superior e Investigacio ´n Cientı ´fica, Spain (BOS2002-01900). References [1] K.G. Raghothama, K.A. Lawton, P.B. Goldsbrough, W.R. Woodson, Characterization of an ethylene-regulated flower senescence-related gene from carnation, Plant Mol. Biol. 17 (1991) 61–71. [2] E.J.I. De Veau, K.C. Gross, D.J. Huber, A.E. Watada, Degradation and solubilization of pectin by b-galactosidases purified from avocado mesocarp, Physiol. Plant. 87 (1993) 279–285. [3] G.S. Ross, T. Wegrzyn, E.A. MacRae, R.J. Redgwell, Apple bgalactosidase. Activity against cell wall polysaccharides and characterization of a related cDNA clone, Plant Physiol. 106 (1994) 521–528. [4] M.S. Buckeridge, J.S. Reid, Purification and properties of a novel b-galactosidase or exo-(1,4)-b-D-galactanase from the cotyledons of germinated Lupinus angustifolius L. seeds, Planta 192 (1994) 502–511. [5] M.B. Singh, R.B. Knox, b-Galactosidase of Lilium pollen, Phytochemistry 24 (1985) 1639–1643. [6] E. Bonning, M. Lahaye, J. Vigouroux, J.F. Thibault, Preliminary characterization of a new exo-b-(1,4)-galactanase with transferase activity, Int. J. Biol. Macromol. 17 (1995) 345–351. [7] J.H. Yoon, K. Ajisaka, The synthesis of galactopyranosyl derivatives with b-galactosidases of different origins, Carbohydr. Res. 292 (1996) 153–163. [8] R. Pressey, b-Galactosidases in ripening tomatoes, Plant Physiol. 71 (1983) 132–135. [9] A.T. Carey, K. Holt, S. Picard, R. Wilde, G.A. Tucker, C.R. Bird, W. Schuch, G.B. Seymour, Tomato exo-b(1-4)-D-galactanase: isolation, changes during ripening in normal and mutant tomato fruit, and characterization of a related cDNA clone, Plant Physiol. 108 (1995) 1099–1107. [10] Z.M. Ali, S. Armugam, H. Lazan, b-Galactosidase and its significance in ripening mango fruit, Phytochemistry 38 (1995) 1109– 1114. [11] Z.M. Ali, S.Y. Ng, R. Othman, L.Y. Goh, H. Lazan, Isolation, characterization and significance of papaya b-galactanases to cell wall modification and fruit softening during ripening, Physiol. Plant. 104 (1998) 105–115. [12] D.L. Smith, D.A. Starrett, K.C. Gross, A gene coding for tomato fruit b-galactosidase II is expressed during fruit ripening, Plant Physiol. 117 (1998) 417–423. [13] L.D. Smith, K.C. Gross, A family of at least seven b-galactosidase genes is expressed during tomato fruit development, Plant Physiol. 123 (2000) 1173–1183. [14] A. Tateishi, H. Inoue, H. Shiba, S. Yamaki, Molecular cloning of b-galactosidase from Japanese pear (Pyrus pyrifolia) and its gene expression with fruit ripening, Plant Cell Physiol. 42 (2001) 492– 498. [15] K. Kakegawa, Y. Edashige, T. Ishii, Metabolism of cell wall polysaccharides in cell suspension cultures of Populus alba in relation to cell growth, Physiol. Plant. 108 (2000) 420–425. [16] W.G.T. Willats, C.G. Steele-King, S.E. Marcus, J.P. Knox, Side chains of pectic polysaccharides are regulated in relation to cell proliferation and cell differentiation, Plant J. 20 (1999) 619–628. [17] T. Fujino, T. Itoh, Changes in pectin structure during epidermal cell elongation in pea (Pisum sativum) and its implications for cell wall architecture, Plant Cell Physiol. 39 (1998) 1315–1323. [18] B. Dopico, G. Nicola ´s, E. Labrador, Cell wall localization of the natural substrate of a b-galactosidase, the main enzyme reponsible for the autolytic process of Cicer arietinum epicotyl cell walls, Physiol. Plant. 80 (1990) 636–641. [19] P. Valero, E. Labrador, Inhibition of cell autolysis and auxin-induced elongation of Cicer arietinum epicotyls by b-galactosidase antibodies, Physiol. Plant. 89 (1993) 199–203. [20] H. Konno, H. Tsumuki, Purification of a b-galactosidase from rice shoots and its involvement in hydrolysis of the natural substrate in cell walls, Physiol. Plant. 89 (1993) 40–47. [21] H. Konno, Y. Yamasaki, K. Katoh, Characteristics of b-galactosidase purified from suspension cultures of carrot, Physiol. Plant. 68 (1986) 46–52. [22] H. Konno, Y. Yamasaki, K. Katoh, Enzymatic degradation of pectic substances and cell walls purified from carrot cell cultures, Phytochemistry 25 (1986) 623–627. [23] B. Dopico, G. Nicola ´s, E. Labrador, Changes during epicotyl growth of an autolysis-related b-galactosidase from the cell wall of Cicer arietinum, Plant Sci. 72 (1990) 45–51. [24] H. Konno, S. Nakashima, T. Nakato, K. Katoh, Pectin-bound b-galactosidase present in cell walls of carrot cells under the different calcium status, Physiol. Plant. 114 (2002) 213–222. R. Esteban et al. / Plant Science 168 (2005) 457–466 465