Mutations affecting spatial and temporal regulation of a beta‐phaseolin gene encoding the major storage protein of bean (Phaseolus vulgaris) were analyzed by stable and transient transformation approaches. The results substantiate the value of transient assays for rapid determination of the functionality of cis‐acting sequences and the importance of stable transformation to identify tissue‐specific determinants. Spatial information is specified primarily by two upstream activating sequences (UAS). UAS1 (−295 to −109) was sufficient for seed‐specific expression from both homologous and heterologous (CaMV 35S) promoters. In situ localization of GUS expression in tobacco embryos demonstrated that UAS1 activity was restricted to the cotyledons and shoot meristem. A second positive domain, UAS2 (−468 to −391), extended gene activity to the hypocotyl. Temporal control of GUS expression was found to involve two negative regulatory sequences, NRS1 (−391 to −295) and NRS2 (−518 to −418), as well as the positive domain UAS1. The deletion of either negative element caused premature onset of GUS expression. These findings indicate combinatorial interactions between multiple sequence motifs specifying spatial information, and provide the first example of the involvement of negative elements in the temporal control of gene expression in higher plants.

The cauliflower mosaic virus promoter is commonly used to drive transcription of chimeric genes in transgenic plants, including the cereals. To determine the tissue and cell types of cereal plants that the promoter functions in, transgenic rice plants containing a CaMV 35S promoter/GUS chimeric gene were analyzed for G U S activity. Insertion of a 35 S/GU S chimeric gene at low copy number into chromosomal DNA of plants regenerated from electroporated protoplasts was confirmed by gel blot hybridization analysis of uncut and endonuclease-digested DNA. Quantitative measurement showed that G U S activity was some tenfold higher in rice leaves than in tobacco leaves [ 8 ] whereas activities obtained for rice roots were similar to those reported for tobacco roots. Histochemical localization of GUS activity confirmed that the CaMV 35S promoter functions in cells of the leaf epidermis, mesophyll and vascular bundle. It is also active in the cortex and vascular cylinder of the root, but only marginally active in the root epidermis. The generally similar distribution and levels of GUS activity obtained in differentiated tissue of stably transformed rice plants indicates the value of the CaMV 35S promoter as a positive control for studies in gene activity in transgenic monocots and dicots.

Parameters influencing the stable transformation of Sorghum bicolor protoplasts with a chimeric neomycin phosphotransferase II (NPT II) gene by electroporation were investigated. The mean number of kanamycin-resistant calli produced increased in direct proportion to the concentration of DNA used for transformation. Linearization of the plasmid doubled the mean number of kanamycin-resistant calli produced, while the addition of carrier DNA had no effect. The copy number (1-4) of integrated genes was low compared with that frequently reported for PEG-mediated transformation. Two strategies for transforming protoplasts with a nonselectable, β-glucuronidase (GUS) gene were compared. One utilized a plasmid containing a CaMV 35S-NPT II gene covalently linked to a CaMV 35S-GUS gene, and the other strategy utilized the two genes on separate plasmids. DNA from all 77 kanamycin-resistant calli analyzed contained restriction fragments hybridizing to the NPT II probe; approximately 70% of the clones from all transformation treatments contained a 1.7-kb EcoRI/HindIII restriction fragment corresponding to the full-length gene. Of the kanamycin-resistant calli, 38-63% (depending on the transformation treatment) contained GUS-hybridizing fragments, and 8-19% contained the full-length gene. The addition of NPT II and GUS genes on a single plasmid or on separate plasmids did not appear to lead to an appreciable difference in the frequency of cointegration of these genes, although an increased proportion of the plasmid bearing the nonselectable (GUS) gene appeared to favor its cointegration.

No abstract

No abstract