Lessons on signalling in plant self-incompatibility systems - Intercellular Communication in Plants - page 248

'competitive interaction', is in agreement with the previously proposed 'inhibitor
model' (McCubbin & Kao, 2000).
According to this model, pollen S functions to modulate RNase activity, such
that RNase activity is active in self-pollinations, but inhibited in cross-pollinations.
This model had been previously favoured over one based on selective entry of
S-RNases into pollen tubes, as S-RNases appear to enter the cytoplasm of both
cross- and self-pollen tubes (Luu et al. , 2000). According to the inhibitor model,
each pollen S -gene product is capable of inhibiting all S-RNases except self, which
remains functional to inhibit self-pollen tube growth. The self-compatibility seen
in 'competitive interaction' results from the possession of two different pollen
S -alleles, each capable of inhibiting all S-RNases except self. In this situation, each
of the two S-RNases present in the pistil are neutralised by the pollen S -inhibitor
from the other S -haplotype.
The fact that SLF is an F-box-containing protein suggests a potential mecha-
nism for the inhibition - or more correctly neutralisation of S-RNase activity. F-box
proteins are components of SCF (Skip1/Cullin/F-box) complexes, which target pro-
teins for degradation through the ubiquitin ligase pathway. Clearly, degradation
S-RNases would be an effective method of neutralising them. In this scenario all
cross S-RNases would be targeted for degradation, while self-S-RNase would not.
An interesting implication of this is that SLF must interact with all S-RNases, pre-
sumably in a conserved region of these molecules (which must also be conserved
in the self-S-RNase); hence, there must also be a specific interaction between SLF
and its cognate self-S-RNase that somehow overrides the general interaction, so
preventing ubiquitination.
It is tempting, though highly speculative, to propose that the other S-RNase
interacting proteins that have been identified might also be involved in ubiquitin-
mediated degradation of S-RNases. A RING-HC-containing protein, PhSBP1 (Sims
& Ordanic, 2001), is a putative E3 ubiquitin ligase that interacts with S-RNase in
a non-allele-specific manner, and may well be a component of the SCF complex
involved in ubiquitinating S-RNases. In addition, there is also a potential role for
the calcium-dependent protein kinase (CDPK) that has previously been shown to
phosphorylate S-RNases (Kunz et al. , 1996). The SCF substrates that have been
characterised to date are recognised in a strictly phosphorylation-dependent manner,
usually by phosphorylation of serine residues on the target protein (for review, see
Craig & Tyers, 1999). CDPK has been shown to phosphorylate S-RNases on serine
residues in a non-allele-specific manner, and this phosphorylation could potentially
be the signal for SCF to bind to S-RNase.
The model shown in Fig. 10.2 draws together both the established data and spec-
ulation made above. S-RNase complexed with HT enters both cross- and self-pollen
tubes. CDPK phosphorylates S-RNase, signalling the binding of an SCF complex,
two components of which are SLF (which provides specificity) and PhSBP1 (which
provides ubiquitin ligase activity). In a compatible interaction, S-RNase is ubiq-
uitinated and subsequently degraded by the 26S proteosome. In an incompatible
interaction some alternative 'high-specificity' interaction occurs between SLF and