Maintenance of normal vegetative growth via attenuation of barnase as an agent of floral tissue ablation in Populus

Graduation date: 2004To reduce the environmental impacts caused by gene flow from transgenic trees, reproductive sterility genes have been developed that use a floral regulatory element linked to a cell toxin. This results in dysfunction or ablation of floral tissues, causing sterility. However, floral promoters often permit low levels of expression in vegetative tissues, which can impair plant health. We therefore developed an attenuation system to avoid the deleterious effects from unintended cytotoxin gene expression in vegetative tissues, and tested it in transgenic poplars. We used the promoter from the poplar ortholog of LEAFY (PTLF) to drive the barnase (ribonuclease) cytotoxin, and three heterologous promoters to drive the attenutation transgene, barstar. Barstar is a specific 1:1 inhibitor of barnase, and thus should protect vegetative tissues from low levels of barnase expression. The LEAFY gene was chosen is known to be an effective floral ablation agent, however, it also shows expression in vegetative tissues. The heterologous promoters included the cauliflower mosaic virus (CaMV) 35S basal promoter +5 to -72 fragment (35SBP), the CaMV 35S basal promoter fused to the 68 bp TMV omega element (35SBP omega), and the nopline synthase (NOS) promoter. We first studied the expression properties of the promoters by evaluating promoter::GUS gene fusions in transgenic poplar (Populus tremula x alba) via fluorometric GUS assays. In leaves, the NOS promoter imparted the highest expression with a mean expression level five-fold higher than PTLF, and 9- and 14-fold higher than 35SBP omega and 35SBP, respectively. Only the NOS and PTLF promoter showed tissue-specific expression patterns with respect to shoots, leaves, stems, and roots. The NOS promoter exhibited the strongest expression in roots. The PTLF promoter specified highest expression in shoot tips. Directed by the poplar floral promoter PTLF, the barnase gene was assembled into constructs harboring barstar driven by either 35SBP, 355BP omega, or NOS. All constructs also contained flanking matrix-attachment regions (rb7) from tobacco to\ud increase and stabilize transgene expression. An unattenuated PTLF::barnase construct (lacking barstar) failed to give rise to any transgenic plants, whereas the attenuated constructs had transformation efficiencies above four percent. However, their transformation rates were still significantly below that of constructs lacking barnase, which averaged 6.1%. Both absolute and relative growth rates of plants measured during a several month greenhouse trial were not statistically significant different between attenuated and transgenic or non-transgenic controls. However, four independent attenuation lines (7% of the attenuation lines), from two constructs, were identified that had very poor growth and significantly lower barstar:barnase RNA ratios than the other attenuated lines. Although the attenuation lines containing NOS::barstar had significantly higher barstar:barnase RNA ratios than the 35SBP::barstar or 35SBP Omega::barstar containing lines, no significant differences in relative growth rates were found. A statistically significant positive linear association was found between relative growth rate and barstar:barnase ratio in the attenuated lines. Graphical analysis suggested a threshold for barstar to attenuate barnase, above which additional levels of barstar did not provide further attenuation nor impact plant growth. By enabling transformation and normal growth of transgenic plants, the attenuation system developed may be a valuable means to produce healthy, sterile trees