Based on paper I in this series, our goals in this paper were to determine the relationship between prebombardment pretreatments and temperatures, microspore cell cycle when bombarded, and the frequencies of homozygous and hemizygous transgenic progeny in barley ( Hordeum vulgare L.). Of the 104 fluorescent plants selected when using the GFP fluorescence transgene, 28 were albino and 76 plants were green. Thirty-one green plants were confirmed to be transgenic; the others were either transient green fluorescent protein expression or selected due to autofluorescence. Of the 31 plants, 23 came from embryos expressing a high level of fluorescence during selection and eight from 51 plants exhibiting a low level of fluorescence. Of the two pretreatments used to induce embryogenesis, 24 of 31 plants were from the cold pretreatment for 21 days (C) versus seven from the 4 day cold plus mannitol pretreatment. Following pretreatment, the microspores were subjected to a high-osmotic period (0.5 mol/L mannitol plus sorbitol) of 4 h prebombardment and 18 h postbombardment at either 25 or 4 °C. Of the 31 transgenic plants, 19 were produced following the 25 °C 4 h prebombardment. Sixteen of the 19 were doubled haploid plants (seven being homozygous for the transgene) and the other three plants were haploid. Of the remaining 12 plants recovered following the 4 h 4 °C prebombardment treatment, nine were haploid and three were doubled haploid plants, two of the latter being homozygous for the transgene. All 12 haploid plants obtained were treated with colchicine and produced homozygous transgenic doubled haploids. Of the two promoters compared, 30 plants had the actin promoter and only one had the 35S promoter. The use of arabinogalactan protein in the culture medium was very beneficial, giving rise to 29 of the 31 plants. The best procedure for obtaining transgenic barley plants from this study was pretreatment C, leaving the cultures at either 4 or 25 °C during the 4 h prebombardment high-osmotic period, using the actin promoter and having arabinogalactan protein in the microspore culture medium. With this procedure, the transgenic frequency was improved 8- to10-fold over previous reports on bombardment of microspores. It yielded about one transgenic plant per Petri dish and is comparable with Agrobacterium frequencies on structures derived from microspores.
