Rice Genetics IV - IRRI books - International Rice Research Institute

Akb4xHpaIIJKA52 (R 0)52-6 (R 1)52-9* (R 1)52-10* (R 1)JKA52 (R 0)MspI52-6 (R 1)52-9* (R 1)52-10* (R 1)Bkb2xwHpaIIJKA 5252-652-15*52-9*AzaC52-10*52-10-8*52-10-16JKA 52MspI52-652-15*52-9*52-10*12.2 –7.1 –5.0 –4.0 –3.0 –2.0 –1.6 –1.0 –0.5 –0.3 –132151927612.2 –7.1 6.1 5.0 4.0 –3.0 –2.0 –1.6 –1.0 –7560.5 –0.4 –263Fig. 2. Methylation and transgene expression in biolistically generated rice. Genomic DNA fromnonsilenced parent (JKA52) and progeny (JKA52-6) lines and silenced lines (JKA52-9 and JKA52-10) was digested with HpaII or MspI and hybridized with (A) a Btt CryIIIA coding region probe or(B) an mUbi1 promoter probe. 2x = 2-copy and 4x = 4-copy reconstruction of p35S/CryIIIADNA; arrows indicate locations of expected fragments. * denotes a silenced line. Modified fromKumpatla et al (1997) and Kumpatla and Hall (1999), with permission.Although linearized plasmid was found to yield twice the number of resistant calli ascircular plasmid, the additional expense and effort of linearization and purificationwere not deemed worthwhile. However, insertion of the linear transgene requiresthat fragmentation occur at some point. The relative physical fragility of sequenceswithin the plasmid, or susceptibility to plant nucleases, or both, are probably importantparameters. Since such sequences may be within the transgene element ratherthan in the vector elements, it is not surprising that direct (biolistic and electroporation)transformation approaches rarely yield discrete inserts if circular DNA is used. The35S promoter has been widely used for plant transformation, but evidence is accumulatingthat it contains several sites that are very susceptible to fragmentation (Kohli etal 1999, Kumpatla and Hall 1999). Other DNA features, such as the possiblesequence or structural differences between eukaryotic and prokaryotic DNA, havebeen suggested as signals that could lead to transgene sequences being recognized byhost surveillance systems as intrusive (Kumpatla et al 1998). Interestingly, Fu et al(2000a) have recently shown that linear transgene constructs lacking vector backbonesare very effective for rice transformation and reduce the occurrence of transgenesilencing.The above considerations indicate that many parameters contribute to variabilityin efficacy of expression from any transgene construct. However, it is tempting toconsider the possibility that some sequences are especially prone to silencing. Wehave accumulated substantial evidence that the rice RCg2 promoter (Xu et al 1995)may be such a sequence. The use of a strong root-specific promoter is especiallyGene silencing and its reactivation in transgenic rice 469