Abstracts / Journal of Biotechnology 131S (2007) S23–S31 S29 the duration of the trial and through several rounds of vegetative propagation. Analysis of height, weight and sugar levels revealed no significant difference between transgenic and wild type lines. PHB production correlated with gene expression and the absence of toxicity, our next objective was to enhance gene expression. Detailed analysis of the trial data revealed that the temporospatial accumulation pattern was the same in all plants, only scaled relative to total production. Combined with a very sensitive assay, this allows us to screen plantlets prior to trans- fer to the greenhouse and discard plants below a threshold. Our pipeline now generates 200 transgenic plants per week or the same number generated in the initial 3 year trial. By a combi- nation of increased numbers and new construct designs (new promoters, codon optimised sequences, etc), we have success- fully identified more than 20 lines producing at 2–4 times the previous best rate. These plants are currently being evaluated in full trials. We have observed that PHB production is absent in meso- phyll cell in all our plastid targeted lines. Transient expression of plastid targeted GFP indicates that absence of production is not due a failure of the dicot Rubisco SSU leader to target mesophyll plastids in C4 plants. Full utilisation of the plastid potential may require metabolic engineering of the acetyl CoA pool in mesophyll plastids. A detailed mapping of organelles indicates that apart from in vascular bundles, stem plastid content is low and low PHB accumulation is consistent with this. A preferable stem target is mitochondria, which are abundant in the stem parenchyma cells. As indicated, however, mitochondrial targeted PHB genes did not lead to PHB production despite targeting apparently working for GFP. There is some evidence that mitochondrial targeting could be embryonic fatal, thus controlled rather than ubiquitous expression may be necessary. doi:10.1016/j.jbiotec.2007.07.047 12. EDTA and urease effects on Hg uptake by Lepidium sativum Beata Smolinska * , Krystyna Cedzynska Technical University of Lodz, Lodz, Poland The phytoextraction process was conducted under laboratory conditions with the use of garden cress plants (Lepidium sativum). The experiment was carried out in a model soil, which was characterized before conducting the process. Inor- ganic forms of mercury (HgCl 2 , HgSO 4 , Hg(NO 3 ) 2 ) were used for contamination of the soil. The phytoextraction process was conducted after edta application to the soil and after urease application. Also the influence of simultaneous addition of ethylenediaminetetraacetic acid (EDTA) and urease into the soil on phytoextraction process was measured. In all variants of phy- toextraction process the total mercury concentrations in roots, stems and leaves of garden cress were determined. The result showed that garden cress accumulated mercury from soil. The overall maximum concentration of mercury in its compounds was found in roots of the plant. In all cases, before addition of urease and EDTA, the translocation process and distribution of mercury in the plant tissues were limited. The addition of urease caused an increase of enzyme activity in the soil and at the same time caused an increase of mercury concen- tration in plant tissues. Application of EDTA increased solubility of mercury and caused an increase of metal uptake by plants. After simultaneous addition of EDTA and urease into the soil garden cress accumulated about 20% of total mercury concentra- tion in the soil. Most of mercury compounds were accumulated in leaves and stems of the plants (46,0–56,9% of total mercury concentration in the plant tissues). doi:10.1016/j.jbiotec.2007.07.048 13. Downregulation of structural lignin genes to improve digestibility and bioethanol production in maize David Caparr´ os-Ruiz * , Montserrat Capellades, Silvia Fornal´ e, Pere Puigdom` enech, Joan Rigau Consorci CSIC-IRTA, Barcelona, Spain Lignin biosynthesis is a complex process common to all vas- cular plants. Lignin is mainly composed of three subunits; p-hydroxyphenyl (H), guaiacil (G) and syringyl (S), whose pro- portion in the final polymer differs depending on the plant species (Lewis and Yamamoto, 1990). Lignin increases the strength and stiffness of fibres, improves the efficiency of water transport in the vascular system, and protects plants against pathogen attacks (Boerjan et al., 2003). However, lignin is a negative value both for maize forage digestibility (Ralph et al., 2004) and for bioethanol production (Torney et al., in press). Most of the functional knowledge known on lignin biosyn- thesis has been generated using dicotyledonous plants (Anterola and Lewis, 2002). Although the majority of the genes involved in lignin biosynthesis has been identified in maize (Collazo et al., 1992; de Obeso et al., 2003; Caparr´ os-Ruiz et al., 2006), at present very little is known concerning how lignification takes places in this plant. Only some spontaneous maize mutants hav- ing brown midrib (bm) pigmentation in lignifying tissues have been associated to lignin metabolism; the bm1, defective in cin- namyl alcohol dehydrogenase (CAD) (Halpin et al., 1998), and the bm3 in which the caffeic acid o-methyltransferase (COMT) gene is disrupted (Vignols et al., 1995). The bm1 mutant has a 90% reduction of CAD activity but the molecular causes of this mutation are unknown. There- fore, to gain knowledge on how lignin is produced in maize, we have generated a CAD-RNAi transgenic line. We produced 59 independent events, identified 6 of them as low-copy lines and achieved the homozygous stage. One of these homozygous CAD-RNAi lines shows 80% reduction of CAD activity. Trans- genic plants do not show the typical brown midrib pigmentation but present leaves and stalks more pigmented than the wild type plants. In addition, transgenic stalks present alterations on the vasculature architecture: they have smaller but more abundant