One of several genes coding for the major pea storage protein, legumin, has been completely sequenced. The sequence covers the whole of the transcribed region, plus 5' and 3' untranscribed sequences. The predicted protein sequence starts with a signal peptide and is followed by the legumin alpha polypeptide sequence of 36. 44kd and the beta polypeptide sequence of 20. 19kd . Compared to other legume storage proteins, the alpha and beta polypeptide sequences encoded by this legumin gene, which contain 3 met and 5 cys residues, are relatively rich in the sulphur amino acids. The coding sequence is interrupted by three introns which show boundary sequences typical of higher plant genes. The 5' end of the gene sequence contains a 'TATA box', a ' CAAT box' and a sequence showing some homology to an ' AGGA box'. An extra sequence, identical to the normal polyadenylation signal of the legumin message is seen in the 3' untranscribed region. The structure of the gene and the possible significance of secondary structures in the nascent RNA transcript in affecting the choice of polyadenylation site is discussed.

Maturing pea cotyledons accumulate large quantities of storage proteins at a specific time in seed development. To examine the sequences responsible for this regulated expression, a series of deletion mutants of the legA major seed storage protein gene were made and transferred to tobacco using the Bin19 disarmed Agrobacterium vector system. A promoter sequence of 97 bp including the CAAT and TATA boxes was insufficient for expression. Expression was first detected in a construct with 549 bp of upstream flanking sequence which contained the the leg box element, a 28 bp conserved sequence found in the legumin-type genes of several legume species. Constructs containing -833 and -1203 bp of promoter sequence significantly increased levels of expression. All expressing constructs preserved seed specificity and temporal regulation. The results indicate that promoter sequences between positions -97 and -549 bp are responsible for promoter activity, seed specificity, and temporal regulation of the pea legA gene. Sequences between positions -549 and -1203 bp appear to function as enhancer-like elements, to increase expression.

A single-copy extensin gene (atExt1) has been isolated from Arabidopsis thaliana (L.) Heynh. The deduced amino acid sequence consists of 374 amino acids which are organised into highly ordered repeating blocks in which Ser(Pro)4 and Ser(Pro)3 motifs alternate. Two copies of the Tyr-X-Tyr-Lys motif and 13 copies of the Val-Tyr-Lys motif are present, showing that this extensin may be highly cross-linked, possessing the capacity for both intra and inter-molecular bond formation. The gene atExt1 is normally expressed in the root and is silent in the leaf; wounding reverses this pattern, turning on the gene in the leaf and repressing it in the root. The promoter contains motifs which have been found to activate plant defence genes in response to salicylic acid, abscisic acid and methyl jasmonate; when these compounds are applied to the roots, the atExt1 gene is activated in the leaf.

We detail the expression of the Arabidopsis thaliana (L.) Heynh. atExt1 extensin gene. atExt1 is normally expressed in roots and inflorescences, and is induced by wounding, exogenously supplied salicylic acid, methyl jasmonate, auxins and brassinosteroids. Northern assays and histochemical analysis of transgenics expressing an atExt1:: gus fusion show that this gene is also induced by the brassica pathogen Xanthomonas campestris pv. campestris and that this induction is restricted to tissues close to the site of infection. Expression at regions of abscission and senescence also implicates atExt1 in these important developmental processes.

SUMMARY The function of the cell wall protein extensin has been the subject of much speculation since it was first isolated over 40 years ago. In order to investigate the role of extensins in plant defence, we used the gain-of-function strategy to generate transgenic Arabidopsis plants over-expressing the EXT1 extensin gene. These were infected with the virulent bacterial pathogen Pseudomonas syringae DC3000 and symptom development was monitored. Lesions on the transgenics were on average five-fold smaller than those on the wild-type, did not increase in area over the time period of infection, accumulated a small bacterial load and showed very little chlorosis outside the lesion boundary. By contrast, lesions on the wild-type were large, spread to over 50% of the leaf area, continued to increase in size over the time course of the infection, accumulated a bacterial load 100-fold higher than that found in the transgenics, and showed a large chlorotic area outside the lesion boundary. SEM of lesions showed no evidence of bacteria at the lesion boundary in the extensin-over-expressing transgenics, whereas bacteria were always seen at the lesion boundary on the wild-type. Analysis of transgenics carrying an EXT1-GUS promoter-reporter fusion showed expression of GUS in a ring around the boundary of the lesion. Basal defences and signal transduction pathways involved in plant defence were not perturbed in the transgenics, as shown by the analysis of the expression of PR1 and PDF1.2 genes. These results show that extensin over-expression limits pathogen invasiveness.