The recently developed microprojectile method for gene transfer to intact cells has been successfully used to transform plant species including some which previously resisted attempts using Agrobacterium and protoplast mediated techniques. In addition, microprojectile bombardment has already proved uniquely suitable for other applications including direct transformation of organelle genomes and rapid assessment of transient expression of genetic constructs introduced into cells of intact tissues. Here we describe various microprojectile acceleration devices and the steps necessary to develop an effective microprojectile mediated transformation system for any plant species. We ernphasise the need to optimise the delivery of DNA into cells, and to tailor strategies for generating stably transformed plants based on the nature of the target tissue, behaviour in tissue culture, and available marker genes. Patterns of cotransformation and coexpression of introduced genes in stable nuclear transformants generated with microprojectiles are summarised, and other applications including organelle transformation are briefly described. We mention technical limitations to the application of microprojectile-mediated gene transfer which need to be overcome if the method is to achieve its full potential as a near-universal gene transfer technique with exciting applications in basic plant molecular biology and practical plant improvement.

A microprojectile accelerator has been constructed and used to bombard cultured sugarcane tissues with GUS reporter gene constructs. Design features useful to minimise target tissue damage and variation between shots are described. Transient expression of GUS occurred in pEmuGN-bombarded cells of nonregenerable suspension culture as well as in regenerable embryogenic callus of commercial sugarcane cultivar 463, and in suspension cultures capable of regeneration to plants. Parameters yielding transient GUS expression in up to 1055 cells per bombardment in homogeneous suspension cultures of sugarcane have been established with a mean of 206 expressing cells per bombardment over a series of 8 independent experiments. Approximately 4% of these transiently expressing cells continued to express GUS for extended periods, indicating probable stable transformation of intact cells of the commercial sugarcane cultivar. Microprojectile bombardment appears the most promising of the available gene transfer techniques for practical genetic transformation of sugarcane because most commercial cultivars readily form regenerable callus suitable for bombardment.

Excised leaflets from mature embryos of peanut cultivars Robut, Gajah, McCubbin and NC-7 produced multiple shoot primordia via organogenesis on Murashige and Skoog medium supplemented with 3 mg L-1 BAP and 1 mg L-1 NAA. Plants were regenerated by elongation of shoot primordia on transfer to MS medium supplemented with 5 mg L-1 BAP. The regeneration frequency of five plants per embryonic leaflet was 10-40-fold higher than previously reported from seedling leaflets. Microprojectile bombardment parameters were optimised for high transient expression rates of the β-glucuronidase (uidA or gus) reporter gene in peanut embryos (> 2000 cells per bombardment) and in excised embryonic leaflets (> 1200 cells per bombardment). The firefly luciferase reporter gene (luc) allowed repeated non-toxic assays, revealing the transition from hundreds of transiently expressing cells soon after bombardment, to a few stably expressing regions in callus grown without selection for 8 weeks after bombardment of embryonic leaflets.

Eight Australian barley cultivars were tested for efficiency of embryonic callus initiation and plant regeneration, from immature embryo explants in tissue culture. Optimisation of tissue culture conditions was performed for cultivars Bandulla, Clipper, Schooner and Tallon in an attempt to increase regeneration frequencies to levels suitable for genetic engineering of barley. Variables tested were 2,4-D concentration, salt composition, carbon source and immature embryo explant. Optimal culture medium composition varied between cultivars. Shoot regeneration rates from culture of isolated scutellar tissues were low for all four cultivars. Halved, immature embryos produced most shoots for cultivars Clipper, Schooner and Tallon, whereas Bandulla performed best with entire immature embryo explants. Clipper (a malting barley) and Bandulla (a feed barley) are suggested as model Australian cultivars for transformation studies. Immature embryos of Bandulla produced an average of 5.3 shoots and Clipper 10.1 shoots per embryo under optimal conditions. Our results show that rates of somatic embryo and plant regeneration sufficient for use in transformation studies can be achieved for diverse Australian Barley cultivars, through systematic testing of a range of key variables including explant type and medium composition.

No abstract