RÉSUMÉ
Fruit ripening is associated with a number of physiological and biochemical changes. They include degradation of chlorophyll, synthesis of flavor compounds, carotenoid biosynthesis, conversion of starch to sugars, cell wall solublisation and fruit softening. These changes are brought about by the expression of specific genes. People are interested in the molecular mechanism involved in the regulation of gene transcription during fruit ripening. Many fruit-specific promoters such as PG, E4, E8, and 2A11 have been characterized and shown to direct ripening-specific expression of reporter genes. AGPase plays the key role in catalyzing the biosynthesis of starch in plants. It is a heterotetrameric enzyme with two small subunits and two large subunits, which are encoded by different genes. In higher plants, small subunits are highly conserved among plant species and expressed in all tissues. And the large subunits are present at multiple isoforms and expressed in a tissue-specific pattern. In fruits, the expression pattern of the large subunits varies with plant species. That made it important to study the transcriptional regulation of the large subunits of AGPase in different plant species. Northern-blot analysis indicates in watermelon, an isoform of the large subunits Wml1 expressed specifically in fruits, not in leaves. The 5' flanking region of Wml1, which covers 1573bp, has been isolated through the method of uneven PCR. And transient expression assay has shown that the 1573bp (named WSP) can direct fruit-specific expression of GUS gene. Our goal in this study was to scan the promoter region for main regulatory regions involved in fruit-specific expression. A chimaeric gene was constructed containing the WSP promoter, the beta-glucuronidase (GUS) structural sequence as a reporter gene and the nopaline synthase polyadenylation site (NOS-ter). The plasmid pSPA was digested with Hind III + Hinc II and promoter fragment of 1573bp (from 180bp to 1752bp) was cut out and cloned into Sma I sites of pBluescript SK(-), to produce pBSPA-16. The same insert was then cut out with Hind III + BamH I, and ligated with transient expression vector pBI426 digested by HindIII + Bgl II to produce pISPA-16. Three 5'-end deletions of the promoter were obtained and fused to GUS gene in plant transient expression vector pBI426: the 1201bp fragment (from 551bp to 1752bp) was generated by digestion of pBSPA-16 with BamH I + SnaB I, the 898bp fragment (from 854bp to 1752bp) by BamH I + EcoRV. Both fragments were ligated with pBluescript SK(-) digested by BamH I + Sma I, to produce pBSPA-12 and pBS-PA-9. The inserts were cut out with HindmIII + BamH I and ligated with pBI426 digested by Hind III + Bgl II, to produce pISPA-12 and pISPA-9. The 795bp fragment (from 957bp to 1752bp) was generated by digestion of pSPA with Hinc II + EcoR I, promoter fragment was cut out and cloned into Sma I sites of pBluescript SK(-), to produce pBSPA-8. The same insert were cut out with Hind III + BamH I, and ligated with transient expression vector pBI426 digested by Hind III + Bgl II. The 1573bp fragment and three 5'-end deletions were delivered into watermelon leaf, stem, flower and fruit of different development stages (5, 10, 20 days after pollination) via particle bombardment using a biolistic PDS-1000/He particle gun. Bombardment parameters were as follows: a helium pressure of 1200 psi, vacuum of 91432.23Pa, 7 cm between the stopping screen and the plate. Histochemical assay were done on all the tissues bombarded after incubation for 2 days. The 1573bp fragment had the strongest promoter activity, and can induce GUS expression in fruits of 5 and 20 days after anthesis and flowers, but not in fruits of 10 days after anthesis, leaves and stems. Fragments of 1201bp and 898bp can induce GUS expression only in fruits of 20 days after anthesis, and with lower expression levels than 1573bp. Fragment of 795bp was not able to direct GUS expression in any of the tissues bombarded (data not shown). It can be concluded that of the 1573bp, 1201 bp, 898bp Wml1 5'flanking regions include the necessary information directing fruit-specific expression. Deletion from 180bp to 551bp doesn't affect the fruit-specificity of the promoter, but lowered the expression level. There may be some cis-acting elements located in this region, which can enhance external gene expression in later stages of fruit development. Deletion from 854bp and 958bp led to loss of GUS expression. This region includes the necessary information needed for gene expression as well as the regulatory elements for fruit-specific transcription.