RNA as a signalling molecule - Intercellular Communication in Plants - page 64

polyploidysation resulting from repeated rounds of DNA replication in the absence
of intervening mitoses (Sugimoto-Shirasu & Roberts, 2003). It is conceivable that
the duplication of chromosome sets in those cells may have contributed to increase
in the pool of aberrant RNAs evoked above. The well-established random nature of
the endoreduplication process in leaves (Sugimoto-Shirasu & Roberts, 2003) would
also explain the stochastic distribution of silencing foci observed in source leaves
of the Class-II plants (Palauqui et al. , 1996).
3.3.3.2 Exogenously induced systemic silencing
Unlike the Class-II Nia plants, the transgenic GFP lines described previously did
not have the intrinsic capacity to spontaneously activate systemic silencing (Ruiz
et al. , 1998). Initiation and propagation of silencing occurred only if the plants
were provided with excess doses of the GFP transgene via Agrobacterium -mediated
transient expression of ectopic constructs (Voinnet & Baulcombe, 1997). A biolistic
procedure was developed as an alternative method to deliver the silencing triggers in
those plants and was exploited to show that systemic GFP silencing could be initiated
from as little as 8-10 randomly bombarded cells within one single leaf (Voinnet et al. ,
1998). Upon bombardment, GFP silencing progressively radiated around those cells
and formed macroscopically detectable red fluorescent foci resembling the chlorotic
spots observed on mature leaves of the Class-II Nia co-suppressed plants (Voinnet
et al. , 1998). Silencing eventually reached a vein from which it was transmitted
to remote parts of the plants that, eventually, became uniformly silenced for GFP.
Biolistic delivery of Nia DNA constructs was also shown to trigger localized and
systemic silencing in non-silenced Class-II tobacco plants (Palauqui & Balzergue,
1999).
The use of this technique allowed the rapid assessment of the requirements for
triggering systemic silencing in both transgenic systems. Various DNA constructs
were tested (homologous to the Nia and GFP coding sequence, respectively), rang-
ing from plasmids containing promoter-driven and full-length cDNA to gel-purified
PCR-amplified fragments. The results in both systems were remarkably similar
(Voinnet et al. , 1998; Palauqui & Balzergue, 1999) and indicated that sense, an-
tisense and promoterless constructs could trigger systemic silencing, although to
varying degrees. For instance, promoterless full-length cDNA constructs were con-
sistently less efficient than equivalent constructs with a 35S promoter, which was
also observed with biolistically induced systemic silencing of chitinase transgenes
in tobacco. Both in the GFP and Nia systems, the proportion of plants exhibiting
silencing upon bombardment also increased as the length of homology between
bombarded and stably integrated sequences was higher.
More recently, the same GFP system and a novel GUS -based silencing reporter
system were used to investigate the effect of bombarding homologous RNA, rather
than DNA (Klahre et al. , 2002). It was found that full-length sense and antisense
RNA could trigger systemic silencing at a low frequency. The frequency was greatly
enhanced (up to 75%) if the sense and antisense RNAs were pre-annealed to form
dsRNA prior to bombardment. Shorter-than-full-length dsRNA was less efficient in