Variation among S-locus haplotypes and among stylar RNases in almond

In many plant species, self-incompatibility systems limit self-pollination and mating among relatives. This helps maintain genetic diversity in natural populations but imposes constraints in agriculture and plant breeding. In almond [Prunus dulcis (Mill.) D.A. Webb], the specificity of self-incompatibility is mainly determined by stylar ribonuclease (S-RNase) and S-haplotype-specific F-box (SFB) proteins, both encoded within a complex locus, S. Prior to this research, a nearly complete sequence was available for one S-locus haplotype. Here, we report complete sequences for four haplotypes and partial sequences for 11 haplotypes. Haplotypes vary in sequences of genes (particularly S-RNase and SFB), distances between genes and numbers and positions of long terminal repeat transposons. Haplotype variation outside of the S-RNase and SFB genes may help maintain functionally important associations between S-RNase and SFB alleles. Fluorescence-based assays were developed to distinguish among some S-RNase alleles. With three-dimensional modelling of five S-RNase proteins, conserved active sites were identified and variation was observed in electrostatic potential and in the numbers, characteristics and positions of secondary structural elements, loop anchoring points and glycosylation sites. A hypervariable region on the protein surface and differences in the number, location and types of glycosylation sites may contribute to determining S-RNase specificity.


Table S11
Results of one-way multivariate analysis of variance analysis (MANOVA) for the KASP assay results shown in the left-hand panels of Supplementary Fig. S7. Clusters (genotype calls) were defined based on normalised FAM and HEX intensities for each of 15 KASP assays applied to a panel of almond cultivars. For each assay, p values are reported for both the Wald-type statistic (WTS) and the asymptotic model-based 'parametric' statistic (PBS). In cases for which two clusters were defined (degrees of freedom = 2) significance of WTS and PBS indicates significant separation between the two clusters. In cases for which three or four clusters were defined (degrees of freedom = 4 or 6), p values are also shown for post hoc Tukey contrasts for each pairwise comparison between clusters.

Table S14
Ranges and means for the fruit set on bagged branches of trees for which self-fertility was predicted based on KASP marker results

Marker(s) used to distinguish between selfincompatible and selffertile trees
Trees predicted to be self-incompatible Trees predicted to be self-fertile

Figure S1
Open reading frames (ORFs) detected for the S7-, S1-, S8-and Sf-haplotype sequences. For each haplotype, the SLF gene (white) S-RNase gene (cyan), SFB gene (dark blue) and long-terminal-repeat retrotransposons (grey) are shown, along with the ORFs (arrows). The strands on which those ORFs were detected are indicated by forward and reverse arrows.

Figure S2
Sequence alignment of 15 almond S locus F-box proteins. Absolutely conserved amino acid residues are indicated by black arrowheads above the alignment.

Figure S4
Primer design for a marker assay to distinguish Sf from other almond S-RNase alleles. A 198 bp region of the S-RNase gene showing aligned sequences for the Sf self-fertility allele and eight SI alleles, with annealing sites for the WriPdSf-1 primers shown in white text on a black background.

Figure S6
Results obtained with four primer sets applied to samples of synthesized DNA representing almond S-RNase alleles and 1:1 mixtures representing heterozygous genotypes. Data shown are intensities of FAM and HEX fluorescence, each normalised against fluorescence from an internal ROX reference. The results in the upper panel are for pure DNA samples representing nine almond S-RNase alleles (S1, S3, S5, S7, S8, S9, S23, S25 and Sf). With each primer set, strong FAM fluorescence (blue data points) was detected for Sf, weak HEX fluorescence (grey data points) was detected for up to three SI alleles (S1, S7 and S8 for WriPdSf-2 and WriPdSf-5; S7 for WriPdSf-3; S1 and S8 for WriPdSf-4) (null alleles) and strong HEX fluorescence (red data points) was detected for all other SI alleles (HEX alleles). The results shown in the lower panel are for 1:1 mixtures representing the possible heterozygotes involving these alleles. Strong FAM fluorescence was detected for each mixture of Sf with a null allele. Both FAM and HEX fluorescence (green data points) were detected for each mixture of Sf with a HEX allele. Strong HEX fluorescence was detected for each mixture of two HEX alleles and for each mixture of a HEX allele with a null allele. Weak HEX fluorescence was detected for each mixture of two null alleles.

Figure S8
Primer design for two-primer assays to distinguish among S-RNase alleles. Aligned sequences of nine alleles in four regions (a, b, c and d) of the S-RNase gene, with positions of conserved regions (C1, C2, C3 and RC4) shown. Annealing sites for primers are shown in white text on a black background and with arrows, labelled with the names of the primer pairs to which they belong. Each primer pair consists of one untailed primer (simple arrow) and one primer to which a FAM tail was attached at the position shown by a circle.

Figure S9
Primer design for three-primer assays to distinguish among S-RNase alleles. Aligned sequences of nine alleles in two regions (a and b) of the S-RNase gene, with positions of conserved regions (C1, C2, C3 and RC4) shown. Annealing sites for primers are shown in white text on a black background and with arrows, labelled with the names of the primer sets to which they belong. Single arrows indicate common untailed primers. Pairs of arrows indicate pairs of allele-specific primers. For each pair of allele specific primers, a FAM tail was added to one primer and a HEX tail was added to the other primer, at the positions shown by white and black circles, respectively.