RNA as a signalling molecule - Intercellular Communication in Plants - page 67

Nia scions of Class I or Class II indicated that the presence of the Nia transgene
in the middle section was not required for transmission of the silenced state to the
top scion (Palauqui et al. , 1997). In fact, up to 20 cm of wt plant segment could
be interspaced without affecting the efficiency or the rate of silencing transmission
(Plate 3.1C). However, as this intermediate section carried the endogenous copy of
the Nia gene, it was still possible that long-distance transmission of the signal was
dependent upon homologous DNA or RNA. The fact that triple-grafting experiments
conducted with the GFP system gave similar results ruled out this possibility, as the
intermediate wt section had no GFP sequence at all (Voinnet et al. , 1998). Therefore,
these results demonstrated that the silencing signal could move over long distances
through cells in which there is no corresponding nuclear gene.
Grafting experiments were also used to address the molecular requirements for
the suppression effect of the phloem-transported signal in the Nia tobacco plants
(Palauqui et al. , 1997). It was found that wt scions grafted onto co-suppressed
Class-II plants were unable to undergo silencing (as assessed by normal accumu-
lation of the endogenous Nia transcript and the lack of chlorosis), despite the fact
that the same tissues were clearly competent to transport the signal over long dis-
tances (Plate 3.1B, C). Silencing occurred only if the scion carried a Nia trans-
gene (Class-I and Class-II plants). However, the presence of the target transgene
per se was not sufficient for silencing to occur because a Nia transgene that had
been transcriptionally silent was not competent to perceive the silencing signal
(Palauqui et al. , 1997). Further experiments established that non-transgenic mutant
plants in which the Nia mRNA levels exceeded those normally found in wt plants
(Class-III plants) also exhibited a silencing phenotype when grafted onto silenced
stocks of Class-II plants (Palauqui & Vaucheret, 1998). Thus, transgenes appear to
be dispensable for the perception of the signal and the mRNA degradation process
that follows (Plate 3.1B). However, high levels of target transcripts, whether of en-
dogenous or transgenic origin, seem to be absolutely required, at least in the Nia
system.
3.3.4.2 Cell-to-cell movement of RNA silencing
Cell-to cell movement upon phloem unloading. Upon phloem unloading, silencing
usually spreads across the whole leaf lamina, indicating cell-to-cell movement of
the silencing signal. The pathway for this cell-to-cell movement was elucidated by
observations made in the N. benthamiana inducible GFP silencing system. Hence,
there was no GFP silencing in the stomata guard cells of leaves that were already
expanding at the time of silencing initiation (Voinnet et al. , 1998). Stomata guard
cells progressively lose plasmodesmatal connections with neighbouring cells as
the age of the leaf increase; therefore, they would have been symplastically isolated
before the signal had moved into those mature leaves. However, in leaves developing
after the signal had spread to the apical growing point, GFP was uniformly silenced,
even in the stomata guard cells. Thus, guard cells are competent for gene silencing,
provided that the signal invades leaves early in their development, before their
symplastic isolation (Voinnet et al. , 1998). Therefore, the most straightforward