Lessons on signalling in plant self-incompatibility systems - Intercellular Communication in Plants - page 244

S 13 -RNase resulted in an S-RNase that exhibited S 13 -specificity but not S 11 -
specificity in transgenic plants. These results have been interpreted as suggesting
that HVa and HVb together are sufficient for S -haplotype specificity. Clearly, how-
ever, domain-swapping experiments assess only functions of amino acids that differ
between the two proteins under study. Combined, all three experiments do suggest
that HVa and HVb regions play a key role in encoding allelic specificity in S-RNases,
but that in addition, amino acids outside HVa and HVb (conserved between S 11 -
RNase and S 13 -RNase) are also likely to be involved in encoding allelic specificity
in S-RNases (Verica et al. , 1998).

10.2.3 The pollen S-gene

The nature of SI reactions dictates that there is recognition of pollen by the pistil;
hence, recognition molecules are required to be present in both tissues. Early models
of the mechanism of gametophytic SI were based upon a single gene. The product
of this gene was envisaged to either act via a dimerisation event within the pollen
tube, or alternatively different pistil and pollen products would be generated by
differential processing of a single gene or operon (Lewis, 1954). Since that time,
a considerable amount of data has been amassed which suggests that the pistil
and pollen S -components are in fact separate genes, culminating with the recent
identification of the pollen S -gene (Sijacic et al. , 2004). Searches for pollen proteins
that interact with S-RNases have identified a calcium-dependent protein-kinase-
like activity, which phosphorylate S-RNases in vitro (Kunz et al. , 1996), and, more
recently, a protein (PhSBP1) containing a RING-HC domain, which potentially may
be involved in the ubiquitin-ligase-mediated protein degradation pathway (Sims &
Ordanic, 2001). Neither of these interactions is S -allele specific, and recent data
suggest that although these molecules might participate in the SI response, neither
is the pollen S -specificity-encoding gene.
Identification of the pollen S -gene in this system has resulted from extensive phys-
ical and genetic mapping of the S -locus. S -loci are inherently regions of low or no
recombination, and thus they tend to accumulate repetitive elements. The extremely
repetitive nature of the regions flanking the S-RNase gene dissuaded attempts at
chromosome walking in this region until very recently (Coleman & Kao, 1992).
Novel technologies, in particular the development of Bacterial Artificial Chromo-
some (BAC) libraries, greatly improved the feasibility of this task by increasing the
upper size limit of genomic clones, thereby reducing the number of 'walking steps'
required. BAC clones containing S-RNase genes have been identified in Petunia
inflata (McCubbin et al. , 2000) and Antirrhinum hispanicum (Lai et al. , 2002).
A 63.7-kb BAC clone of the S 2 -locus from Antirrhinum , which contains the
S-RNase , has been fully sequenced (Lai et al. , 2002). Subsequent analysis of
this sequence revealed six putative genes with homology to previously reported
sequences, and of these four encode retrotransposons. More interestingly, a gene
termed SLF ( S -locus F-box), which encodes an F-box-containing protein and is
located approximately 9 kb downstream of the S-RNase gene, was identified. This