Plant Molecular Biology 31:517-527, 1996. Q 1996 KluwerAcademic Publishers. Printed in Belgium. 517 Isolation and characterisation of a pod dehiscence zone-specific polygalacturonase from Brassica napus Morten Petersen ~ , Lilli Sander 1 , Robin Child 2, Harry van Onckelen 3, Peter UlvskoC and Bernhard Borkhardt 1'* l The Biotechnology Group, DIPS, Lottenborgvej 2, DK-2800, Lyngby (*authorfor correspondence); 21ACR-Long Ashton Research Station, Department of Agricultural Sciences, University of Bristol, BS18 9AF Long Ashton, UK; 3Department of Biochemistry, University of Antwerp, 1 Universiteitsplein, Wilrijk, B-26 l O Antwerp, Belgium Received 10 November 1995; accepted in revised form 27 February 1996 Key words: polygalacturonase isoforms, Brassica napus, pod dehiscence, pod shatter Abstract Seven distinct partial cDNAs, similar in sequence to previously described polygalacturonases (PGs), were amplified from cDNA derived from rape pod wall, dehiscence zone and leaves by the polymerase chain reaction. Northern analysis showed that one clone, PG35-8, was expressed at low levels in the dehiscence zone during the first five weeks after anthesis but was very abundantly expressed at week 6. In contrast, no PG35-8-related RNA was detected in the pod wall. Our data suggest that there are temporal and spatial correlations between the breakdown of the middle lamella, of the dehiscence zone cells and the pattern of synthesis of PG35-8 transcripts which may indicate a role for this particular PG in rape pod dehiscence. PG35-8 was used to isolate five cDNA clones from a rape dehiscence zone cDNA library. Restriction enzyme analysis and partial sequencing revealed that they were derived from four highly homologous transcripts which are probably allelic forms of a single gene. One full-length clone, RDPG1, was completely sequenced. The predicted protein of RDPGI showed its highest identity with PG from apple fruit with an identity of 52%. Introduction Several physiological and developmental processes in plants involve the action of cell wall hydrolases (for review see [13, 25]). Well known examples include abscission, pollen development and fruit maturation [32, 11,3, 5, 6, 33, 31, 34]. Pod dehiscence in rape resembles the process of abscission as it occurs in a descrete cell layer and involves the breakdown of cell wall material [21]. These authors produced compelling evidence that the weakening of the rape pod which eventually leads to pod shatter in oilseed rape is accompanied by ultra- structural changes which culminate in the degradation of the middle lamella of the dehiscence zone cell walls. This is in accordance with work of Josefsson [ 16] and Picart and Morgan [26] who have presented histochem- The nucleotide sequence data reported appear in the EMBL Nucleotide Sequence Database under the accession number X95800. ical evidence that pectin is lost from dehiscence zones of maturing rape pods. Although biochemical analysis of maturing rape pods has shown a correlation, both temporally and positionally, between cellulase activity and pericarp separation, a similar positive correlation for polygalacturonase (PG) could not be established [22]. Correlation may be obscured, however, as cell wall hydrolases act in concert to degrade cell wall polymers and within each enzyme class overlapping expression of different isoforms, in particular of exo- and endo-acting forms, may be involved in mediating this disintegration [5, 19]. We report in this paper on the cloning of several rape PG isoforms by polymerase chain reaction (PCR) and, based on their identity with other plant PGs, dis- cuss their possible mode of action. A full-length cDNA clone of one PG isoform that is expressed exclusive- ly in the dehiscence zone late in pod development is described. This isoform possesses the characteristics
Transcript
Plant Molecular Biology 31:517-527 , 1996.
Q 1996 KluwerAcademic Publishers. Printed in Belgium. 517
I s o l a t i o n a n d c h a r a c t e r i s a t i o n o f a p o d d e h i s c e n c e z o n e - s p e c i f i c p o l y g a l a c t u r o n a s e f r o m Brassica napus
Morten Petersen ~ , Lilli Sander 1 , Robin Child 2, Harry van Onckelen 3, Peter UlvskoC and Bernhard Borkhardt 1'* l The Biotechnology Group, DIPS, Lottenborgvej 2, DK-2800, Lyngby (*author for correspondence); 21ACR-Long Ashton Research Station, Department of Agricultural Sciences, University of Bristol, BS18 9AF Long Ashton, UK; 3Department of Biochemistry, University of Antwerp, 1 Universiteitsplein, Wilrijk, B-26 l O Antwerp, Belgium
Received 10 November 1995; accepted in revised form 27 February 1996
Key words: polygalacturonase isoforms, Brassica napus, pod dehiscence, pod shatter
Abstract
Seven distinct partial cDNAs, similar in sequence to previously described polygalacturonases (PGs), were amplified from cDNA derived from rape pod wall, dehiscence zone and leaves by the polymerase chain reaction. Northern analysis showed that one clone, PG35-8, was expressed at low levels in the dehiscence zone during the first five weeks after anthesis but was very abundantly expressed at week 6. In contrast, no PG35-8-related RNA was detected in the pod wall. Our data suggest that there are temporal and spatial correlations between the breakdown of the middle lamella, of the dehiscence zone cells and the pattern of synthesis of PG35-8 transcripts which may indicate a role for this particular PG in rape pod dehiscence.
PG35-8 was used to isolate five cDNA clones from a rape dehiscence zone cDNA library. Restriction enzyme analysis and partial sequencing revealed that they were derived from four highly homologous transcripts which are probably allelic forms of a single gene. One full-length clone, RDPG1, was completely sequenced. The predicted protein of RDPGI showed its highest identity with PG from apple fruit with an identity of 52%.
Introduction
Several physiological and developmental processes in plants involve the action of cell wall hydrolases (for review see [13, 25]). Well known examples include abscission, pollen development and fruit maturation [32, 11,3, 5, 6, 33, 31, 34].
Pod dehiscence in rape resembles the process of abscission as it occurs in a descrete cell layer and involves the breakdown of cell wall material [21]. These authors produced compelling evidence that the weakening of the rape pod which eventually leads to pod shatter in oilseed rape is accompanied by ultra- structural changes which culminate in the degradation of the middle lamella of the dehiscence zone cell walls. This is in accordance with work of Josefsson [ 16] and Picart and Morgan [26] who have presented histochem-
The nucleotide sequence data reported appear in the EMBL Nucleotide Sequence Database under the accession number X95800.
ical evidence that pectin is lost from dehiscence zones of maturing rape pods. Although biochemical analysis of maturing rape pods has shown a correlation, both temporally and positionally, between cellulase activity and pericarp separation, a similar positive correlation for polygalacturonase (PG) could not be established [22]. Correlation may be obscured, however, as cell wall hydrolases act in concert to degrade cell wall polymers and within each enzyme class overlapping expression of different isoforms, in particular of exo- and endo-acting forms, may be involved in mediating this disintegration [5, 19].
We report in this paper on the cloning of several rape PG isoforms by polymerase chain reaction (PCR) and, based on their identity with other plant PGs, dis- cuss their possible mode of action. A full-length cDNA clone of one PG isoform that is expressed exclusive- ly in the dehiscence zone late in pod development is described. This isoform possesses the characteristics
518
neccessary for a PG involved in middle lamella break- down. Therefore it may play a determining role in causing pod shatter in rape.
Materials and methods
Plant material
Brassica napus cvs. Topas and Fido were grown in pots in glasshouses. Successive sowings were made to ensure that material was available for collection throughout the summer. Flowers were labelled at anthesis and pods harvested at known intervals.
Electron microscopy
The process of cell separation in the dehiscence zone was examined using the following standard procedures. Pods of known age were cut into 1.0 mm transverse slices and fixed in a mixture of 4% formaldehyde and 5% gluteraldehyde for 1 h. This was post-fixed in 1% osmium tetroxide for 1 h and embedded in epoxy (Spurr's) resin. Ultrathin sections were cut using a Reichert Ultramicrotome, stained in saturated uranyl acetate for 5 min and lead citrate for 5 min and exam- ined using a transmission electron microscope.
Molecular weight down-shift assay
Pods were harvested 6.5 weeks after anthesis and peri- carp tissue consisting of 2-3 layers of dehiscence zone cells, the replum (bearing the main vascular supply to the pods) and the septum was separated from the remaining pod wall tissue and the seeds. Dehiscence zones were ground in liquid nitrogen in a mortar, transferred to a new mortar and extracted with 4 ml per g fresh weight (FW) of 10 mM MOPS, pH 7, supplemented with 1 mM PMSF and centrifuged at 12 000 x 9 for 20 min. The supernatant was discarded and the pellet was resuspended in water (10 ml/g FW) and centrifuged as above. The supernatant was again discarded. The pellet was resuspended in 3 M LiCI (2 ml/g FW) and stirred for 30 rain at 0 °C. The super- natant, collected following centrifugation as above, was taken to represent primarily proteins which were strongly bound in the cell wall through ionic inter- actions. The LiCI extract was incubated overnight at room temperature with an equal volume of substrate (200 #g/ml polygalacturonic acid) (Sigma, P7276). Before use, the substrate was dialysed extensively
against 300 mM NaOAc, pH 5.0. Substrate incubated with boiled enzyme (control) and with native enzyme were analysed on a Pharmacia Superrose 6 column. Eluent was 300 mM NaOAc, flow 0.5 ml/h, and frac- tions of 660 #1 were analysed for uronic acids accord- ing to the method of Blumenkrantz and Asboe-Hansen [4].
RNA isolation
A 20 g portion each of leaf, dehiscence zone, pod wall, root or stem were ground in liquid nitrogen and homogenized for 30 s in a Waring blender with 100 ml of extraction buffer (4 M guanidinium thiocyanate, 25 mM sodium citrate pH 7.0, 0.5% sarkosyl, 0.1 M 2- mercaptoethanol). The homogenate was transferred to a fresh tube and 1/10 volume of 2 TEM sodium acetate pH 4.0, and i volume of phenol/chloroform saturat- ed with (10 mM Tris-HC1 pH 8, 1 mM EDTA) was added. The solution was shaken vigorously, cooled on ice for 15 min, centrifuged at 10 000 x 9 for 15 min at 4 °C, and the supernatant was re-extracted with the phenol/chloroform mixture. An equal volume of iso- propanol was added to the resultant supernatant and RNA was precipitated by an overnight incubation at - 2 0 °C. After centrifugation at 10 000 × 9 for 15 min, the RNA pellet was dissolved in 2 ml of extraction buffer. Fourteen ml of 4 M LiCL was added and the solution kept in an ice-bath overnight. The RNA was pelleted by centrifugation at 10 000 x 9 for 15 min, washed in 80% ethanol, dried and dissolved in 1 ml of sterilized water. Poly(A) + RNA was isolated on an oligo-d(T) Sepharose column according to manufac- turer's instructions (Boehringer, Mannheim).
Primer design
Four degenerated primers were designed based on conserved regions from published PG amino acid sequences from tomato [7, 14], maize [23], avocado [10] and evening primrose [6]. The sequence of the four degenerated primers were PG 1, CCAGGAATTCAAYACIGAYGGNRTNCA; PG3, GGACGAATTCACNGGNGAYGAYTGYAT; PG2, CGACGGATCCAIGTYTTDATNCKNA; PG5, CACAGGATCCSWIGTICCIYKDATRTT. Underlined nucleotides represent bases added for cloning purpos- es. The primer location in proportion to the toma- to PG cDNA clone are from amino acid 247-252 (PG1), 268-273 (PG3), 326-331 (PG2) and 384-389 (PG5). The predicted size of the four possible PCR
products is: 460 bp (PG1/PG5); 385 bp (PG3/PG5); 280 bp (PGI/PG2) and 210 bp (PG3/PG2). A restric- tion enzyme site for EcoRI was introduced at the 5' end of the two uptream primers PGI and PG3 and a BamHI site was introduced at the 5' end of the two downstream primers PG2 and PG5.
Reverse transcription and nested PCR
Random or oligo-dT primed first-strand cDNA synthe- sis was performed with total RNA from each tissue and M-MLV reverse transcriptase under conditions speci- fied by the manufacturer (Life Technologies/BRL).
All PCR reactions had the following final compo- sition: 50 mM KC1, 10 mM Tris-HC1 pH 8.3, 1.5 mM MgCI2, and 0.001% (w/v) gelatin, 100 pmoles of each degenerated primer and first-strand cDNA in 50 #1 reaction volume. After an initial denaturation of tem- plate cDNA at 95 °C for 3 min the PCR was initiated by adding 1 U of Taq DNA polymerase (hot start PCR) using the following conditions: 1 min at 95 °C, 1 min at 45 °C, 1 min at 72 °C for 35 cycles followed by 72 *C for 3 min. For hot-start nested PCR, 2 #1 of a PCR reaction was applied as template in a new PCR reaction.
PCR cloning
The PCR products were extracted with chloroform and ethanol-precipitated, redissolved in TE and digest- ed with the restriction enzymes BamHI and EcoRI. The restricted PCR products were purified from low- melting agarose and cloned into pGEM-7z cut with BamHI and EcoRI. DNA sequences were obtained by the dideoxy chain termination method using Sequenase version 2.0 (Pharmacia).
Northern analysis
Total RNA was separated by gel electrophoresis in 0.66 M formaldehyde/l% agarose [29]. RNA was transferred onto Hybond-N (Amersham) filters and fixed to the filter by UV irradiation. The filters were prehybridised for 4 h at 68 °C in 5 x Denhardt's solu- tion, 25 mM Na2HPO4, 25 mM NaHzPO4, 0.1% pyrophosphate, 750 mM NaC1, 5 mM EDTA and 100 #g/ml denatured herring sperm DNA. The labelled PCR products were heat-denatured and added direct- ly to the pre-hybridisation buffer and hybridisation was continued for 16 h at 68 °C. Filters were washed according to Sambrook et al. [29] with the final wash
519
carried out at 68 °C in 0.2× SSC, 0.1% SDS (high- stringency wash). The filters were autoradiographed at - 8 0 °C using an intensifying screen. Labelling of probe DNA was performed as described [ 12].
Construction and screening of a dehiscence zone-specific cDNA library
A cDNA library was constructed from poly(A) + RNA isolated from the dehiscence zone 6 weeks after anthe- sis. Gel-purified cDNA larger than 1 kb was ligated into the 1ZAP II vector following the manufacturer's instructions (Stratagene). The primary library consist- ed of 1.25 × 106 pfu with an averaged cDNA insert size of ca. 1.3 kb. Library screening was performed under high stringency according to standard procedures [29].
DNA sequencing and computer analysis
DNA sequencing of the full-length cDNA clone RDPG1 was performed on subclones using RNA pro- motor and cDNA sequence specific primers and Seque- nase version 2.0 (Pharmacia, Denmark). Comput- er analysis was performed with the GCG version 7 sequence analysis software package [9].
Southern hybridisation
Total DNA was extracted and blotted acording to Del- laporte et al. [8] and Sambrook et al. [29]. Prehybridi- sation was carried out at 68 °C (high stringency) in 5 x SSPE (1 x SSPE is 0.18 NaC1, I0 mM sodium phoso- hate pH 7.7, 1 mM EDTA), 0.1% SDS, 0.2% BSA, 0.2% Ficoll and 50 #g/ml carrier DNA. The final wash was performed at 68 °C with 0.1 × SSPE and 0.1% SDS (high-stringency wash).
Results
Anatomical and biochemical characterisation of the pod dehiscence zone
It is presumed that middle lamella dissolution can- not proceed without homogalacturonan degradation. In order to determine when middle lamella dissolu- tion, and hence galacturonan degradation proceeded at high rates, transverse sections of pod dehiscence zones were prepared for transmission electron microscopy throughout pod development. The changes in dehis- cence zone cell wall structure during pod senescence
520
Figure 1. Electron micrographs of the dehiscence zone during pod development. A (top left). Intersection of cell walls in the dehiscence zone 5 weeks after anthesis, showing parallel microfibrils and fenestrated structure of middle lamella (darkened). Note unbroken cytoplasm containing active organelles. B (top right). Intersection of cell walls in the dehiscence zone 7 weeks after anthesis showing separation along the middle lamella. The parallel microfibrils in the cellulose remain intact. Note fragmented cytoplasm. C (bottom). Separated cells in the dehiscence zone between thickened pericarp-edge cells (right) and replum cells (left), 8 weeks after anthesis. The magnification of the cell wall (before and after separation) was x 18 000 and that of the separating cells was x 1350.
are shown in Fig. 1. The pr imary cell wall and middle
lamella remained intact whilst the pod was still pho-
tosynthetic, 5 weeks after anthesis (WAA) (Fig. I A). Two weeks later, after pod wall senescence had begun, dissolut ion of the cell wall was clearly visible (Fig. 1B). By the t ime pod senescence was completed, 8 WAA, total separation of the cells in the dehiscence zone had taken place. It appeared that this was attributable to
break down of the middle lamella because the pr imary walls remained intact (Fig. 1C).
In an attempt to unders tand the mechanism of mid- dle lamella dissolution, a molecular weight down-shif t assay was carried out with polygalacturonate as sub- strate. Based on the anatomical studies, 6 WAA was chosen as a developmental stage likely to be associated with maximal polygalacturonase activity and extracts of wal l -bound protein were prepared from dehiscence
Figure 2. Depolymerisation of polygalacturonate by a crude extract of dehiscence zone wall proteins. Polygalacturonate was incubated with boiled enzyme (solid line, control) and with active extract (dotted line) and analysed on a Pharmacia Superrose 6 column.
zone cells at this stage. Molecular-weight down-shift assays are only semi-quantitative at best but this type of assay provides information on mode of action. A gel filtration chromatogram of substrate (homogalactur- onate) and the reaction products are shown in Fig. 2. A very prominent monosaccharide peak would be expect- ed if the reduction of molecular weight was to be accounted for by hydrolases acting in exo fashion. Note that only a minor proportion of products elutes at Vttot where free galacturonic acid should be indi- cated. The results indicate that a wall-bound endo- polygalacturonase (endo-PG) was responsible for the depolymerisation of the substrate.
PCR strategy
A nested RT-PCR strategy was adopted for the cloning of PG-encoding transcripts. A procedure was devel- oped that would ensure identification of as complete a set of PGs as possible while at the same time allow- ing the partial clones to be classified with respect to PG type. The primers were designed in such a way that all PCR products would contain a common cDNA region located between primer PG3 and PG2 because (1) this region encodes a highly conserved amino acid sequence that is present in all sequenced bacterial, fun- gal and plant PGs and (2) because the region also encodes a cysteine residue that is specific for pollen PGs and which we believed could differentiate exo- and endo-PG isoforms (see Discussion).
521
Table 1. Frequency in % of PG isoforms in dehiscence zone, pod walls and leaves determined with RT-PCR. A. PG analysis with the PG3/PG2 primer combination. 17 PG cDNA clones were analysed in both dehiscence zone and pod walls and 10 PG cDNA clones were anal- ysed in leaves. B. PG analysis with the PG3/PG5 primer combination. 24 and 25 PG cDNA clones were analysed from dehiscence zones and pod walls, respectively. Leaves were not analysed with this primer combination.
A type dehiscence zones pod walls leaves
PG32-24 12 17 0
PG32-25 12 0 0
PG32-32 23 59 0
PG32-16 53 17 90
PG35-8 0 0 0
PG32-8 0 6 0
PG32-26 0 0 10
B
type dehiscence zones pod walls
PG32-24 0 48
PG32-25 0 0
PG32-32 0 0
PG32-16 0 44
PG35-8 100 8
PG32-8 (I 0
PG32-26 0 0
As RNA material for the RT reaction, total RNA was isolated from dehiscence zones 6 weeks after anthesis. In addition, total RNA was isolated from pod walls and leaves as reference material. The first PCR reaction was performed using the PGI/PG5 primer combination and then nested PCR was performed with the PG3/PG2, PG1/PG2 or PG3/PG5 primer combi- nation using a small aliquot of the PG1/PG5 PCR reaction as template. The nested PCR products were purified, digested with the restriction enzymes BamHI and EcoRI, cloned into pGEM-7z, and sequenced (see Materials and methods).
Diversity of pod PG isoforms in oilseed rape
The deduced amino acid sequence of the cloned PCR products was scanned for the presence of the highly conserved amino acid motif shared by all sequenced plant PGs. Comparison of the identified rape PGs revealed seven highly divergent isoforms (Fig. 3). The tissue from which the seven partial PG cDNAs were obtained and their amino acid identity are shown in Tables 1 and 2. Three partial PG cDNAs were obtained
522
kiwi genomic SIVNGSRKVRV NDI'/
apple fruit SIVSGSQRVQA TDI~
PG35-8 S IEDGSQNVQI NDL']
tomato fruit SIVSGSQNVQA TNI ~]
avocado fruit SIESGSKMVIA TNI']
peach fruit SIGPGTSNLWI EGVI
tomato abscission SIGPGTSNLWI EGI.Z
PG32-26 SIQTGCSNVYV HNVI~
primrose pollen SLGDGSKNINI TNIq
tobacco pollen SVGDETEQLYI TRVq
cotten pollen SIGDGTKNMVI KEI'I
PG32-24 S IGGGTENLLV EGVfi
PG32-25 SVGDGMKNLLV ERVS
PG32-8 SIGDGTRDLLV ERV'I
rape pollen SVGDGMKNLLI EKV~
PG32-32 SVGDGMKNLL I EKV'9
PG32-I 6 SVGDGMKNLHV EKV'I
alfalfa pollen SIGDGSKQITV QGVI';
Arabidopsis pollen SIGDGTENLIV ENVI~
maize pollen SIGPGTSKVNI TGV'I
2GPGHG ISIGSLC
~GPGHG ISIGSLI
~GPGHG ISIGSLI
ZGPGHG ISIGSLI
CGPGHG ISIGSLC
CGPGHG ISIGSLC
CGPGHG ISIGSLC
CGPGHG ISIGSL(
CGPGHG ISVGSL(
CGPGHG ISVGSL(
CGPGHG ISIGSLG
CGPGHG LSIGSLG
CGPGHG ISIGSLG
CGPGHG ISIGSLG
CGPGHG ISVGSLC
CGPGHG ISVGSLG
CGPGHG ISVGSL(
CGPGHG LSVGSL(
CGPGHG ISIGSL(
CGPGHG ISIGSL(
{GN SEAHVSDVVV
EDG SEDHVSGVFV
DDN SKAYVSGIDV
3GN SEAYVSNVTV
DRN SEAHVSGVLV
KEQ EEAGVQNVTV
NKQ QELGVQNVTV
KDS TKACVSNITV
RYK NEESVVGIYV
GNP DEKPVVGVFV
KFQ NEEPVEGIKI
KYP NEQPVKGITI
LYG HEEDVTGVKV
LYV KEEDVTGIRV
RYG WEQDVTDITV
RYG WEQDVTDINV
RYG NEQDVSGIRV
KFT TEENVEGITV
RYP NEQPVKGVTV
RYK DEKDVTDINV
NGAKLCGTTNG--
NGAKLSGTSNG
DGATLSETDNG
NEAKIIGAENG
DGGNLFDTTNG__
KTVTFSGTQNG m
KTVTFSGTTNG
RDVVMHNTMTG
K~ rITGSQNGm
R~ PFTNTDNG
S~ ~ITNTSNG
RB IIKHTDNG
V5 TLRNTDNG
VK TLINTDNG
K~ TLEGTSNG
KK TLEGTDNG
I~ TLQQTDNG
K~ TLTATDNG
R~ LIKNTDNG
K[ TLKKTMFG
III
Figure 3. Alignment of the deduced amino acid sequences of plant PGs together with the partial rape PG clones. The sequences shown cover the region located between the primers PG3 and PG2. The numbering at the right indicates PG subclasses: I, fruit; II, abscission; III, pollen. The box marked with an asterix shows the cystein residue corresponding to amino acid 317 in the tomato fruit PG and the large box shows the conserved region used to identify the rape PCR clones as PGs. The PG protein sequences are from kiwifruit [1], apple [2], peach [20], maize [23], tobacco [33], evening primrose [6l, tomato [14, 17], rape [27], alfalfa (EMBL accession number U20431), cotton [15], Arabidopsis (EMBL accession number X72292) and avocado [ 18].
only from a single tissue namely PG32-25 from dehis- cence zones, PG32-8 from pod walls and PG32-26 from leaves. PG32-32, PG32-24 and PG35-8 were found only in the two pod tissues, whereas PG32-16 was obtained from all three tissues analysed.
Tissue-specific expression of the PGs
PCR is a very sensitive method that can amplify tran- scripts present at very low levels in a tissue. Therefore, PCR can not readily be used to compare transcript lev- els in different tissue types. In addition, when used in combination with degenerated primers the amount of the different PCR products is highly dependent on primer design. The expression of the different PG cDNA clones in the dehiscence zone and pod wall as well as in roots, stems, and leaves was therefore investi- gated by northern analysis. Only the expression of one of the seven partial PG cDNAs, namely PG35-8, was detected by this method (Fig. 4). PG35-8 hybridised to a 1.7 kb transcript expressed in the dehiscence zone during the initial stages of pod development. A large increase in the steady state level was apparent 6 weeks after anthesis in the dehiscence zone while no tran- scripts corresponding to PG35-8 were detected in pod walls, roots, stems or leaves.
Isolation and characterisation of full-length cDNA clones encoding PG35-8
A cDNA library was constructed in AZap II from mRNA isolated from the dehiscence zone 6 weeks after anthesis. Screening 300 000 plaques with the PG35- 8 cDNA clone as probe gave ca. 200 hybridisation signals. Five strongly hybridising plaques (RDPG1- 5) were purified to homogeneity. After excision of the plasmid DNA from the lambda vector, restriction enzyme analysis showed the cDNA inserts to be ca. 1.6 kb in all clones except one, which had an insert of 1.3 kb. Partial sequencing of the cDNA ends revealed that the cDNAs were derived from three highly similar RNA transcripts having 94-98% and 91-98% identity in the 5'- and Y-untranslated regions, respectively.
Due to the genetic status of Brassica napus (r~ = 19) as an amphidiploid hybrid between the two diploid species B. rapa (n = 9) and B. oleracea (n = 10) Southern analysis was undertaken to determine the copy number of the RDPG gene in the rape genome (Fig. 5). When the RDPG 1 cDNA clone was hybridised to XbaI or BglII-digested genomic DNA, two restric- tion enzymes with no corresponding sites in the three different cDNA clones, two hybridising bands were observed in each digest. The two bands in the BgllI
Table 2. PG amino acid identity (%).
Type PG32-25 PG32-32 PG32-24 PG35-8 PG32-8 PG32-26
PG32-16 75 81 60 54 71 44
PG32-25 - 75 58 50 81 44
PG32-32 - 56 52 67 40
PG32-24 - 44 62 48
PG35-8 54 52
PG32-8 - 46
523
Figure 4. Tissue-specific expression of PG35-8. Northern blots were performed on equal amount of total RNA (7/~g) under high- stringency conditions. WAA, week after anthesis.
digest are, however, only visible after short film expo- sure. These two hybridisation patterns indicate that at most two genes are present in the rape genome that are highly homologous to RDPG1. Since the RDPG cDNA clones contain three HindlII restriction sites this conclusion is also supported by the finding of very few hybridising bands in the HindlII-digested genomic DNA. The hybridisation pattern obtained after HindlI digestion is difficult to interpret with certainty as one of the cDNA clones contains a single restriction site for this enzyme.
The complete sequence of clone RDPG1 and the deduced amino acid sequence of the largest open read- ing frame with similarity to other PGs is shown in Fig. 6. The open reading frame encodes a protein, 433 amino acids in size, with an estimated molecu- lar weight of 46.6 kDa. The protein sequence contains two potential N-glycosylation sites at amino acid posi- tions 159-161 and 393-395, respectively. When the rape PG sequence was compared to sequences in the EMBL database it showed highest identity with apple fruit PG (52%). Like other secreted proteins, the dehis- cence zone specific PG is most probably cleaved post- translationally [20]. The most likely cleavage site is located between amino acids 23 and 24 and gives rise to a mature protein with an estimated molecular weight of 44.2 kDa.
Figure 5. Southern blot hybridisation of rape genomic DNA. Genom- ic DNA (10/~g) was digested with the restriction enzymes HindlII (lane 1 ), HindlI (lane 2), Xbal (lane 3) or BglII (lane 4), separated on a 1% agarose gel, and blotted onto a Hybond-N membrane. Hybridi- sation of the membrane was carried out under high stringency. The positions of molecular weight markers are indicated on the left.
Discussion
Electron microscopic examination of rape pod struc- ture during senescence showed that the dehiscence zone cells separated from each other at the final stage of pod development and that this was primarily attribut- ed to a dissolution of the middle lamella [21, 30, 26]. Our studies showed that all wall breakdown in the dehiscence zone was initiated by swelling of the middle lamella (Fig. 1B). Cell separation followed within one week and wall microfibrils were intact at this time (Fig. 1C). A biochemical investigation failed to substantiate a correlation between PG activity and pod dehiscence [22]. Total PG activity, measured as released reducing sugars from polygalacturonate, remained nearly constant during pod development and approximately at the same level in the dehiscence zone and in the pod walls [22]. This reducing sugar assay
524
detects exo-polysaccharide hydrolases with excellent sensitivity while endo-acting enzymes are detected less efficiently relative to the degree of depolymerisation attained during the reaction. It was thus speculated that endo-PGs might have gone unnoticed in the study of Meakin and Roberts [22]. Molecular-weight down- shift assays have the converse sensitivity and using this method we have been able to demonstrate endo- PG activity in the dehiscence zone cell walls late in pod development.
PG expressed in pollen and fruits has been cloned from a variety of plant species. The pollen types are assumed to be exo-acting but, as Tebbutt e t al. [33] point out, this has not been proved. PG cDNA and genomic clones have been isolated from pollen of maize, evening primrose, cotton, oilseed rape and tobacco [23, 28, 6, 15, 27, 33].
Genes expressed during fruit ripening are known to encode endo-acting PGs in two cases, tomato [31 ] and peach [20]. cDNA clones from apple and a genomic clone from kiwi [2, 1 ] were regarded as endo-PGs due to their sequence similarity with the tomato endo-PG cDNA clone. This assumption is furthermore corrobo- rated by measurements of endo-PG activity in ripening fruits of kiwi and apples [34, 35].
Alignment of the deduced amino acid sequence of pollen and fruit PG clones shows a high degree of sequence conservation [33]. A striking feature of this alignment is that pollen PGs carry a cysteine residue corresponding to amino acid 317 in the alignment shown in Fig. 3 which is not present in the fruit-related PGs. Cysteine is important in the folding of proteins due to its ability to create sulfur bridges, and the possi- bility therefore exists that this amino acid reflects a dif- ference in the tertiary structure between exo- and endo- PG. A putative tomato abscission PG cDNA clone has recently been cloned and characterised by Kalaitzis [17]. This PG is grouped together with a peach fruit PG cDNA clone isolated by Lester [20]. The mode of action of abscission specific loGs is unknown.
In our study we have identified a total of seven highly divergent partial PG cDNAs from oilseed rape of which six were obtained from pods. Four of the latter were found both in pod walls and dehiscence zone tissue while PG32-8 was found only in pod wall and PG32-25 only in tissue containing dehiscence zone cells. There are at least three possible reasons for the presence of multiple PG isoforms in the pod tissues. Exo-PG and endo-PG could act in a concerted man- ner during pod development as has been observed in peach fruit abscission zones [5]. Alternatively, our tis-
sue preparations consist of different cell types which may each express different isoforms. Thirdly, it may be that some PG genes assigned for physiological pro- cesses other than pod development have a low basal level of expression in this tissue, and will be detected due to the high sensitivity of PCR. Candidates of the latter category include PG32-25 which was identified only twice and PG32-8 which was identified only once.
Only two PG isoforms were identified in rape leaves. One of these, PG32-16, was found in all three tissues examined and the other, PG32-26, was only identified in leaves. This observation makes it unlikely that the finding of several PG types in the rape pod is due to genomic DNA contamination, as the RNA was isolated by the same method from leaves and pods.
The PGs identified by PCR are to a great extent defined by primer design. This is exemplified by PG35- 8, which was never cloned after amplification with the PG3/2 primer combination. An explanation for this was provided after sequencing of RDPG 1 which disclosed that a tryptophane present in the conserved region used to design primer PG2 was substituted by tyrosine in RDPG1. By inference, oilseed rape may contain addi- tional PGs which are not amplified by the primers used in this study.
PG32-16, PG32-24, PG32-25, PG32-32 and PG32- 8 are closely related to pollen PGs (Fig. 3). These five rape pod PGs contain the cysteine residue suggested by Tebbut et al. [33] as one character that separates pollen PGs from fruit PGs. However, only one, PG32-32, is identical to a previously described rape pollen PG [27]. We suggest that the grouping of the five rape pod PGs together with pollen PGs has more relevance to their mode of action than to the tissue from which they are derived. If pollen PGs are exo-acting then these five rape pod PGs could have this type of activity.
Of the two remaining rape PGs identified in this study, PG32-26 is grouped together with a tomato abscission PG [17] and a fruit l°G isolated from peach [20] while PG35-8 is grouped together with the remain- ing fruit PGs. Both PG32-26 and PG35-8 lack the unique cysteine present in the pollen PGs. These two clones are the only types for which circumstantial evi- dence could classify them as possible endo-PG iso- forms. PG32-26 was not detected in the dehiscence zone material and is therefore an unlikely candidate for an endo-PG involved in dehiscence. However, due to its high identity with the putative tomato abscission PG it will be interesting to determine its role in leaf abscission.
TGTTGTCTTGTATATGATTCACTGAAGTTGATTGCTTGTCCATGAATAAA TGAATAATATCATTTCTCT (A) n 1599
Figure 6, DNA and deduced amino acid sequence of the full-length cDNA clone RDPGI. The predicted signal peptide is underlined to the cleavage site after amino acid 23. Two potential N-glucosylation sites at position 159-161 and 393-395 are also underlined. A putative polyadenylation site is highlighted.
A quite unique cloning pattern emerged for PG35- 8 during the PCR analysis: PG35-8 was the only type identified in the pod dehiscence zone when the PG3/5 primer combination was used in a nested PCR (24 clones analysed). In contrast, when the same approach was used with pod wall material PG35-8 was found only twice (25 clones analysed). This strongly indi- cates that the transcript corresponding to PG35-8 is much more abundant in the dehiscence zone than in the pod walls. Northern analysis of the rape pod develop- ment showed this to be correct. PG35-8 was expressed
in the dehiscence zone throughout pod development and expression increased strongly 6 weeks after anthe- sis. It was not found in any other tissue. No transcripts corresponding to the other rape PG types were detect- ed in any of the tissues. The expression pattern of PG35-8 mRNA correlates well with the breakdown of dehiscence zone cell walls shown by our ultrastructural studies.
Work is now in progress to generate transgenic rape plants with the full-length RDPG1 cDNA in antisense
526
to establish the role of this putative endo-PG in rape pod shatter.
Acknowledgements
The authors wish to thank Mette Carlsen, Tine Jensen and Helle Munck Petersen for skilled technical assis- tance and Johan Botterman of Plant Genetic Systems for helpful comments on the manuscript. This study was supported by the Agricultural Research Council (SJVF 23-2400-2) and the European Commission (AIR CT93 0879).
References
1. Atkinson RG, Gardner RC : A polygalacturonase gene from kiwifruit (Actinidia deliciosa). Plant Physiol 103:669-670 (1993).
2. Atkinson RG: A cDNA clone for endopolygalacturonase from apple. Plant Physiol 105:1437-1438 (1994).
3. Baird LM, Reid PD: Cellulase activity and localization during induced abscission of Coleus blumei. J Plant Growth Regul 11 : 129-134 (1992).
4. Blumenkrantz N, Asboe-Hansen G: New method for quantita- tive determination of uronic acids Anal Biochem 54:484-489 (1973).
5. Bonghi C, Rascio N, Ramina A, Casadoro G: Cellulase and polygalacturonase involvement in the abscission of leaf and fruit explants of peach. Plant Mol Biol. 20:839-848 (1992).
6. Brown SM, Crouch ML: Characterization of a gene fami- ly abundantly expressed in Oenothera organensis pollen that shows sequence similarity to polygalacturonase. Plant Cell 2: 263-274 (1990).
7. DellaPenna D, Alexander DC, Bennett AB: Molecular clon- ning of tomato fruit polygalacturonase: Analysis of polygalac- turonase mRNA levels during ripening. Proc Natl Acad Sci USA 83:6420-6424 (1986).
8. Dellaporta SL, Wood J, Hicks JB: A plant DNA miniprepara- tion: Version II. Plant Mol Biol 1:19-21 (1983)
9. Devereux J, Haeberli P, Smithies O: A comprehensive set of sequence software programs for VAX. Nucl Acids Res 12: 387-395 (1984).
10. Dopico B, Lowe AL, Wilson ID, Merodio C, Grierson D: Cloning and characterization of avocado fruit mRNAs and their expression during ripening and low-temperature storage. Plant Mol Biol 21:437-449 (1993)
11. del Campillo E, Reid PD Sexton R, Lewis LN: Occurrence and localization of 9.5 cellulase in abscising and nonabscising tissues. Plant Cell 2:245-254 (1990).
12. Feinberg AP, Vogelstein B: A technique for radio-labeling DNA restriction endonuclease fragments to high specific activ- ity. Anal Biochem 132:6-13 (1983).
13. Fischer RL, Bennett AB: Role of cell wall hydrolases in fruit ripening, Annu Rev Plant Physiol Plant Mol Biol 42:675-703 (1991).
14. Grierson D, Tucker G A, Keen J, Ray J, Bird C R. Schuch W: Sequencing and identification of a cDNA clone for tomato polygalacturonase. Nucl Acids Res 14:8595-8603 (1986).
15. John EM, Petersen MW: Cotton (Gossypium hirsutum L) pollen-specific polygalacturonase mRNA: tissue and tempo- ral specificity of its promotor in transgenic tobacco. Plant Mol Biol 26:1989-1993 (1994).
16. Josefson E: Investigation into shattering resistance of crucifer- ous oil crops. Zeitschr Pflanzenzucht 59:384-396 (1968).
17. Kalaitzis P, Koehler SM, Tucker ML: Cloning of a tomato polygalacturonase expressed in abscission. Plant Mol Biol 28: 647-656 (1995).
19. Lashbrook CC, Gonzales-Bosch C, Bennett AB: Two divergent endo-b-1,4-glucanase genes exhibit overlapping expression in ripening fruit and abscising flowers. Plant Cell 6:1485-1493 (1994).
20. Lester DR, Speirs J, Orr G, Brady CJ: Peach (Prunus persica) endopolygalacturonase cDNA isolation and mRNA analysis in melting and nonmelting peach cultivars. Plant Physiol 105: 225-231 (1994).
21. Meakin PJ, Roberts JA. Dehiscence of fruit in oilseed rape (Brassica napus L) I. Anatomy of pod dehiscence. J Exp Bot 229:995-1002 (1990).
22. Meakin PJ, Roberts JA. Dehiscence of fruit in oilseed rape (Brassica napus L) II. The role of cell wall degrading enzymes and ethylene. J Exp Bot 229:1003-1011 (1990).
23. Niogret ME Dubald M, Mandaron P, Mache R: Characteri- zation of pollen polygalacturonase encoded by several cDNA clones in maize. Plant Mol Biol 17:1155-1167 (1991).
24. Orr G, Brady C: Relationship of endopolygalacturonase activi- ty to fruit softening in a freestone peach. Postharv Biol Technol 3:121-130 (1993).
26. Picart JA, Morgan DG: Preliminary studies on the effect of 4-chlorophenoxyacetic acid on the senescence and dehiscence of pods in oil seed rape (Brassica napus). Plant Growth Regul 4:169-175 (1986).
27. Robert LS, Allard S, Gerster JL, Cass L, Simmonds J: Isola- tion and characterization of a polygalacturonase gene highly expressed in Brassica napus pollen. Plant Mol Biol 23: 1273- 1278 (1993).
28. Rogers HJ, Allen RL, Hamilton WDO and Lonsdale DM: Pollen specific cDNA clone from Zea mays. Biochim Biophys Acta 1089:411-413 (1991).
29. Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY (1989).
30. Sexton R, Jamieson GGC, Hall MHIL: An ultrastructual study of abscission zone cells with special reference to the mecha- nism of enzyme secretion. Protoplasma 91 : 369-387 (1977).
31. Smith CJS, Watson CF, Ray J, Bird CR, Morris PC, Schuch W, Grierson D: Antisense RNA inhibition of polygalacturonase gene expression in ripening tomatoes. Nature 334:724-726 (1988).
32. Taylor JE, Tucker GA, Lasslett Y, Smith CJS, Arnold CM, Wat- son CF, Schuch W, Grierson D, Roberts JA: Polygalacturonase
33.
expression during leaf abscission of normal and transgenic tomato plants. Planta 133-138 (1990). Tebbutt SJ, Rogers HJ, Lonsdale DM: Characterization of a tobacco gene encoding a pollen-specific polygalacturonase. Plant Mol Biol 25:283-297 (1994).
5 2 7
34. Wegrzyn TF, MacRae EA: Pectinesterase, polygalacturonase and i%galactosidase during softening of ethylene-treated kiwifruit. Hort Science 27:900-902 (1992).
35. Wu Q, Szakacs-Dobozi M, Hennat M, Hrazdina G: Endo- poly- galacturonase in apples (Malus domestica) and its expression during fruit ripening. Plant Physiol 102:219-225 (1993).
