Cassava brown streak disease (CBSD), caused by the Ipomoviruses Cassava brown streak virus (CBSV) and Ugandan Cassava brown streak virus (UCBSV), is considered to be an imminent threat to food security in tropical Africa. Cassava plants were transgenically modified to generate small interfering RNAs (siRNAs) from truncated full-length (894-bp) and N-terminal (402-bp) portions of the UCBSV coat protein (ΔCP) sequence. Seven siRNA-producing lines from each gene construct were tested under confined field trials at Namulonge, Uganda. All nontransgenic control plants (n = 60) developed CBSD symptoms on aerial tissues by 6 months after planting, whereas plants transgenic for the full-length ΔCP sequence showed a 3-month delay in disease development, with 98% of clonal replicates within line 718-001 remaining symptom free over the 11-month trial. Reverse transcriptase-polymerase chain reaction (RT-PCR) diagnostics indicated the presence of UCBSV within the leaves of 57% of the nontransgenic controls, but in only two of 413 plants tested (0.5%) across the 14 transgenic lines. All transgenic plants showing CBSD were PCR positive for the presence of CBSV, except for line 781-001, in which 93% of plants were confirmed to be free of both pathogens. At harvest, 90% of storage roots from nontransgenic plants were severely affected by CBSD-induced necrosis. However, transgenic lines 718-005 and 718-001 showed significant suppression of disease, with 95% of roots from the latter line remaining free from necrosis and RT-PCR negative for the presence of both viral pathogens. Cross-protection against CBSV by siRNAs generated from the full-length UCBSV ΔCP confirms a previous report in tobacco. The information presented provides proof of principle for the control of CBSD by RNA interference-mediated technology, and progress towards the potential control of this damaging disease.

grains, a number greatly in excess of that needed for fertilization (Uribelarrea et al., 2002). However, the pol-Development of improved genetic traits in maize (Zea mays L.) relen of maize is among the largest and heaviest of the quires robust measures to prevent pollen-mediated gene flow (PMGF) grasses, with a diameter of about 90 m. For compariand assure isolation of new traits, whether these traits are the result son, the pollen grains of ragweed (Ambrosia spp.) and of conventional breeding or of modern genetic techniques. Studies were conducted in California and Washington to evaluate the relation-Timothy (Phleum pratense L.) are 20 and 34 m in ship of distance and temporal separation for isolation from PMGF. diameter, respectively (Raynor et al., 1972). Most maize Kernel color was used to detect outcrossing from source plots of 0.4 pollen is dispersed by gravity downward from the tassel, to 1.2 ha in size to receptor plots planted at distances up to 750 m falling in the vicinity of the originating plant. Bateman and planting intervals of up to 3 wk from the pollen source. Outcross- 1947) found pollen deposition at 27 m was Ͻ1% of ing from source to receptor plots was observable to 0.0002% (1 kernel that close to the source plants. Raynor et al. (1972) in ≈500 000 kernels). Increasing temporal separation reduced the disestimated that 98% of maize pollen remains close to tance required to achieve genetic isolation. Outcrossing was Ͻ0.01% the originating plant, and that Ͻ1% would be found at 500 m when source and receptors flowered at the same time, whereas beyond 60 m. Jarosz et al. (2003) estimated that 95% this level of confinement was achieved at 62 m or less when 2 wk of of the pollen produced was deposited within 10 m of temporal separation was used. No outcrossing was detected at 750 m and 2 wk of temporal separation. This is the first practical evaluation M. Qualls, Qualls Ag Labs, Ephrata, WA 98823; S.A. Berberich, Chesterfield, MO 63017. Received 12 Dec. 2003. *Corresponding au-color inside yellow maize fields. The farthest distance thor (Philip.j.eppard@monsanto.com).studied was 50 m from the edge of the source plot, and a low but undefined level of outcrossing was observed

Cassava (Manihot esculenta Crantz) is a vitally important food source for many people in developing tropical countries. There are significant opportunities for improving the compositional qualities and pest resistance of cassava, and modern biotechnology is expected to play an important role in these improvements. The testing and development of genetically modified cassava will of course be subject to regulatory review, and experimental field trials must be performed in a fashion that prevents gene flow from the regulated plants. Methods to ensure reproductive isolation will be derived from a fundamental understanding of the biology of the crop. A current and comprehensive document on cassava reproductive biology is not yet available but is essential to guide regulators and scientists in planning and evaluating measures for reproductive isolation of confined field trials. This paper compiles a current view of the reproductive biology of cassava for use in experimental design and regulation of confined field trials. With the current state of knowledge on gene flow and seed dormancy in cassava, three methods for reproductive isolation of regulated experimental plots may currently be recommended: (i) removal of flower buds before flowering, (ii) destruction of plants before flowering, and (iii) floral bagging to contain pollen and seed. Areas for further research in cassava biology and biosafety are suggested.

Cassava brown streak disease (CBSD) presents a serious threat to cassava production in East and Central Africa. Currently, no cultivars with high levels of resistance to CBSD are available to farmers. Transgenic RNAi technology was employed to combat CBSD by fusing coat protein (CP) sequences from Ugandan cassava brown streak virus (UCBSV) and Cassava brown streak virus (CBSV) to create an inverted repeat construct (p5001) driven by the constitutive Cassava vein mosaic virus promoter. Twenty-five plant lines of cultivar TME 204 expressing varying levels of small interfering RNAs (siRNAs) were established in confined field trials (CFTs) in Uganda and Kenya. Within an initial CFT at Namulonge, Uganda, non-transgenic TME 204 plants developed foliar and storage root CBSD incidences at 96–100% by 12 months after planting. In contrast, 16 of the 25 p5001 transgenic lines showed no foliar symptoms and had less than 8% of their storage roots symptomatic for CBSD. A direct positive correlation was seen between levels of resistance to CBSD and expression of transgenic CP-derived siRNAs. A subsequent CFT was established at Namulonge using stem cuttings from the initial trial. All transgenic lines established remained asymptomatic for CBSD, while 98% of the non-transgenic TME 204 stake-derived plants developed storage roots symptomatic for CBSD. Similarly, very high levels of resistance to CBSD were demonstrated by TME 204 p5001 RNAi lines grown within a CFT over a full cropping cycle at Mtwapa, coastal Kenya. Sequence analysis of CBSD causal viruses present at the trial sites showed that the transgenic lines were exposed to both CBSV and UCBSV, and that the sequenced isolates shared >90% CP identity with transgenic CP sequences expressed by the p5001 inverted repeat expression cassette. These results demonstrate very high levels of field resistance to CBSD conferred by the p5001 RNAi construct at diverse agro-ecological locations, and across the vegetative cropping cycle.