The Plant Cell, Vol. 15, 771–781, March 2003, www.plantcell.org © 2003 American Society of Plant Biologists
Structural and Transcriptional Analysis of the Self-Incompatibility
Locus of Almond: Identification of a Pollen-Expressed F-Box
Gene with Haplotype-Specific Polymorphism
and Hisashi Hirano
Graduate School of Agriculture, Kyoto University, Kyoto 606-8502, Japan
Kihara Institute for Biological Research and Graduate School of Integrated Science, Yokohama City University, Yokohama
Department of Pomology, University of California, Davis, California 95616
Abhaya M. Dandekar,
Thomas M. Gradziel,
Gametophytic self-incompatibility in Rosaceae, Solanaceae, and Scrophulariaceae is controlled by the
sists of an S-RNase gene and an unidentified “pollen
S haplotype–specific region containing the S-RNase gene, was sequenced completely. This region was found
to contain two pollen-expressed F-box genes that are likely candidates for pollen
lotype–specific F-box protein), was expressed specifically in pollen and showed a high level of
quence polymorphism, comparable to that of the S-RNases. The other is unlikely to determine the
cause it showed little allelic sequence polymorphism and was expressed also in pistil. Three other
cloned, and the pollen-expressed genes were physically mapped. In all four cases,
genes and were located at the
S haplotype–specific region, where recombination is believed to be suppressed, suggesting
that the two genes are inherited as a unit. These features are consistent with the hypothesis that
This hypothesis predicts the involvement of the ubiquitin/26S proteasome proteolytic pathway in the RNase-based gameto-
phytic self-incompatibility system.
S locus, which con-
S” gene. An
70-kb segment of the S locus of the rosaceous species
S genes. One of them, named SFB (S hap-
specificity of pollen be-
S haplotypes were
SFBs were linked physically to the S-RNase
SFB is the pollen S gene.
Self-incompatibility (SI) in flowering plants is a genetic system
that prevents self-fertilization by enabling the pistil to reject pol-
len from genetically related individuals, thus promoting out-
crossing. Genetic studies have shown that most SI systems are
controlled by a single multiallele locus called the
allele of pollen matches that of the pistil, the pollen is rec-
ognized as “self” and rejected by the pistil (de Nettancourt,
2001). However, recent studies on pistil- or pollen-specific self-
compatible mutants (Thompson et al., 1991; Sassa et al., 1997;
Golz et al., 1999) and transformation experiments with solana-
ceous species (Lee et al., 1994; Murfett et al., 1994; Dodds et
al., 1999) demonstrate that SI phenotypes of pistil and pollen
are determined by different genes, called pistil
genes, respectively. Based on the findings that the
multigene complex, the term “haplotype” has been adopted to
denote variants of the locus, and the term “allele” is used to
denote variants of a given polymorphic gene at the
(McCubbin and Kao, 2000).
The families Rosaceae, Solanaceae, and Scrophulariaceae
locus is a
display gametophytic self-incompatibility (GSI) and share the
gene product, called the S-RNase (for review, see
McCubbin and Kao, 2000). S-RNase is a basic glycoprotein with
RNase activity (McClure et al., 1989). RNase activity of S-RNase
is required for the pistil to reject self-pollen (Huang et al., 1994;
Royo et al., 1994; McCubbin et al., 1997). The rRNA of incom-
patible pollen of
sistent with a model postulating that the S-RNases act as intra-
cellular cytotoxins (McClure et al., 1990). On the other hand,
gene of the RNase-based GSI remains to be iden-
tified. Circumstantial evidence suggests two conceivable mod-
els for the action of pollen
product: the gatekeeper model
and the inhibitor model (Thompson and Kirch, 1992; McCubbin
and Kao, 2000). In the gatekeeper model, the pollen
is assumed to be a gatekeeper located at the plasma mem-
brane of the pollen tube. The gatekeeper interacts specifically
with cognate S-RNase, assisting in its uptake into the pollen
cytoplasm, leading to the degradation of pollen RNA and the
arrest of self-pollen tube elongation. In the inhibitor model, all
S-RNases enter the pollen tube regardless of their
types; however, the pollen
product inhibits the activity of all
S-RNases except for the cognate S-RNase. As a consequence,
the cognate S-RNase remains active in the pollen tube, leading
to the degradation of self-pollen RNA. Golz et al. (1999, 2001)
produced pollen-part self-compatible mutants of
compatible pollination with irradiated pollen and found that a
deletion mutant for the pollen
is degraded in the style, con-
gene could not be recovered
These authors contributed equally to this work.
To whom correspondence should be addressed. E-mail sassa@
yokohama-cu.ac.jp; fax 81-45-820-1901.
Article, publication date, and citation information can be found at
772 The Plant Cell
by this screen. This finding suggests that the pollen
is essential for pollen tube elongation, which is consistent with
the inhibitor model but in conflict with the gatekeeper model. Im-
munocytological experiments showing the entry of the S-RNase
into the compatible pollen tube support the inhibitor model (Luu
et al., 2000).
The identification and characterization of the pollen
essential to understanding the molecular mechanisms of the
RNase-based GSI. The pollen
following features: first, tight linkage to the S-RNase gene; sec-
ond, a high level of
haplotype–specific sequence polymor-
phism; third, pollen-specific expression. The tight genetic link-
age between the pollen
gene and the S-RNase gene
suggested that chromosome walking would be one of the feasi-
ble approaches for the identification of the pollen
ever, conventional chromosome walking in Solanaceae is diffi-
cult because the solanaceous
centromere, a region abundant with repetitive sequences
(Coleman and Kao, 1992; Matton et al., 1995; Bernacchi and
Tanksley, 1997; Entani et al., 1999). McCubbin et al. (2000)
screened a BAC library of the
molecular markers linked to the S-RNase gene and walked an
from multiple starting points. Although they
obtained 51 BAC clones spanning a
locus region, they did not construct a contig encompassing
the region that encodes
specificity, which is delimited by re-
combination breakpoints. Therefore, it is not clear whether the
genomic clones contain the pollen
We previously conducted chromosome walking on the
haplotype of almond, which belongs to the Rosaceae, and con-
structed the contigs covering an
-RNase gene (Ushijima et al., 2001). Genomic DNA gel blot
analyses revealed that the nucleotide sequence of the
region, which is defined by the two boundary markers NP79R
and NP182F and contains the S
verged and specific to each
haplotype. The regions outside
of the boundary markers were relatively conserved among dif-
haplotypes. The structural heteromorphism is the dis-
tinctive feature of loci consisting of coadapted gene complexes
in different genetic systems, including sporophytic SI of Brassi-
caceae (Ferris and Goodenough, 1994; Brown and Casselton,
2001; Kusaba et al., 2001). The heteromorphism of the region
specificity genes is believed to maintain the
tight association of the pistil
suggesting that the pollen
gene of almond is located in the
haplotype–specific region (Ushijima et al., 2001).
Here, we sequenced the
gion using a shotgun strategy. Based on the prediction of open
reading frames (ORFs), two pollen-expressed genes were iso-
lated. One of the two genes, named
F-box protein), showed a high level of allelic polymorphism and
pollen-specific expression. Comparative analysis of the four
haplotypes showed that all
kb of the S-RNase genes. Distances between genes and mark-
ers are highly variable among the
heteromorphism of the region, which might ensure the tight as-
and the S-RNase gene. These features of
gene is expected to have the
locus is located close to the
genome with 13
2-Mb region around the
gene (McCubbin et al.,
200-kb region around the
-RNase gene, was highly di-
gene and the pollen
alleles are located within
haplotypes, showing the
are consistent with those of the expected pollen
encodes an F-box protein that is a component of a class of
ubiquitin ligase (SCF complex) (Deshaies, 1999), raising the
possibility that the ubiquitin/26S proteasome proteolytic sys-
tem plays a central role in the discrimination of self/nonself pol-
len in the RNase-based GSI.
Sequence Analysis of the
We previously predicted the boundaries of the almond
and suggested that the pollen
70-kb area of the
haplotype–specific region that also con-
tains the S
-RNase gene (Ushijima et al., 2001). To identify the
haplotype–specific region was sequenced
completely using the shotgun strategy. Three genomic clones
covering the region, p79SCU (
36 kb), were used to construct the shotgun library
(Figure 1A). A total of
1000 sequences, which included PCR
products filling gaps between the shotgun contigs, were assem-
bled and gave a nucleotide sequence of 71,953 bp. The
sequence was annotated by RiceGAAS (Rice Genome Auto-
mated Annotation System [Sakata et al., 2002]; http://ricegaas.
dna.affrc.go.jp/index.html), which automatically analyzes large
sequences using several programs for the prediction and anal-
ysis of protein-coding sequences: BLAST (Basic Local Align-
ment Search Tool [Altschul et al., 1990]), AutoPredLTR (Sakata
et al., 2002), and GENSCAN (Burge and Karlin, 1997).
GENSCAN identified 12 ORFs (ORF1 to ORF12) in the se-
quenced region (Figure 1A). By this prediction, the three exons
of the S
-RNase gene were misidentified as different genes,
ORF1 and ORF2 (Table 1). AutoPredLTR search, which pre-
dicted a long terminal repeat (LTR) sequence, showed that the
70-kb region contained four pairs of LTRs (Figure 1A). Se-
quences between the LTRs were found to be similar to ret-
rotransposon-like sequences and were designated retro0 to
retro3 (Figure 1B, Table 1). All four retrotransposon-like se-
quences included several mutations (insertion or deletion),
causing their putative polyproteins to be truncated. Retro0 was
found to be disrupted by the insertion of retro1 to retro3 (Figure
1B), implying that retro0 was first inserted into the
type. Retro3 also was disrupted by the insertion of a 2.5-kb
DNA sequence. The inserted sequence showed high similarity
to the region at
18 kb (Figure 1A) but displayed no sig-
nificant homology with sequences in the public databases.
gene might be present in an
7 kb), pPdC55 (
35 kb), and
cDNA Cloning of the Pollen-Expressed Genes Located at
To clone the pollen-expressed genes at the sequenced
region, we designed 17 forward and 11 reverse primers based
on the prediction of ORFs. The primers were used for rapid am-
plification of cDNA ends (Frohman et al., 1988) with pollen
cDNA. Pollen from almond cv Nonpareil (
homozygous cultivar is not available. For most
ORFs, sequences of isolated cDNA clones were not identical
) was used, be-
Locus Gene of Almond 773
to, but were similar to, those of the corresponding ORFs, sug-
gesting that they constitute gene families and that those ORFs
are not expressed in pollen. However, sequences of two of
the cDNAs completely matched the corresponding ORF se-
quences (ORF1 and ORF3). The first exon of the predicted
ORF1 was isolated as a pollen-expressed gene and is called
ORF1 was located 6.7 kb upstream of the S
(Figure 1A). The transcriptional orientations of ORF1 and the
-RNase gene were the same. The deduced amino acid se-
quence of ORF1 showed 29.3% homology with that of
locus F-box) of
F-box motif (Lai et al., 2002). We named ORF1
cus F-box of Prunus dulcis. PdSLF of the Sc haplotype, desig-
nated PdSLFc, contained the sequence of the cosmid end
probe NP79R, which defined the boundary of the S haplotype–
specific region (Ushijima et al., 2001). NP79R sequence has
been shown to be conserved in different S haplotypes of al-
mond (Ushijima et al., 2001). PdSLF of the Sd haplotype, desig-
nated PdSLFd, was amplified by PCR from the cosmid clone
of the Sd haplotype. The deduced amino acid sequences of
locus region and encodes a protein with an
(Figure 2A, Table 1).
Figure 1. Sequence Analyses of the Sc Haplotype–Specific Region.
(A) Scheme of the Sc haplotype–specific region. The nucleotide sequence of the ?70-kb Sc haplotype–specific region was characterized by GEN-
SCAN, BLASTX, and AutoPredLTR search of RiceGAAS (Sakata et al., 2002). The A nucleotide of the putative initiation codon (ATG) of the Sc-RNase
gene (Ushijima et al., 1998) is positioned as ?1. Exons of the Sc-RNase gene and molecular markers NP79R and NP182F, which define the bound-
aries of the almond S locus (Ushijima et al., 2001), are represented by black boxes. Red, blue, and green boxes represent the results of GENSCAN,
BLAST, and AutoPredLTR searches, respectively. Each pair of LTRs is linked by lines.
(B) Scheme of the retrotransposon-like sequence-rich region (?29 to ?53 kb). Gray boxes represent LTR sequences. Hatched boxes represent the
region encoding the polyprotein of retro1, retro2, and retro3. White boxes represent the region for the retro0 polyprotein, into which retro1, retro2, and
retro3 were inserted.
774The Plant Cell
PdSLFc and PdSLFd were highly similar to each other (95.1%
identity; Figure 2B), as were AhSLF-S2 and AhSLF-S2L of Anti-
rrhinum (97.9% identity; Lai et al., 2002), suggesting that they
are unlikely to control the S specificity of pollen.
The ORF3 cDNA also encoded a protein with an F-box motif
(Figure 3). ORF3 and the Sc-RNase gene were located in in-
verse orientation and separated by ?1.6 kb (Figure 1). ORF3
contained one intron at the 5? untranslated region (Figure 3).
Except for the N-terminal region containing the F-box motif se-
quence, ORF3 showed no significant homology with known
Table 1. Summary of Homology Search (BLASTX) Results for the ORFs Predicted by GENSCAN
ORF Homolog (Accession No.) PollenStyle Description
ORF1 (1st exon)
ORF1 (7th exon)
ORF3 (1st and 2nd exons)
ORF3 (3rd exon)
Antirrhinum SLF-S2 (AJ297974)
Almond Sc-RNase (AB011470)
Almond Sc-RNase (AB011470)
Arabidopsis putative protein (AL138652)
Arabidopsis putative F-box protein (AC018907)
Oryza sativa putative protein (AC084763)
Zea mays retrotransposon (AF082133)
Nicotiana retrotransposon (CAA32025)
Arabidopsis retrotransposon (AC006248)
Arabidopsis Ser/Thr protein phosphatase (U80922)
PdSLF (Prunus dulcis S locus F-box protein)
N-terminal region of the Sc-RNase
C-terminal region of the Sc-RNase
NoSFB (S haplotype–specific F-box protein)
Designated retro1 in Figure 1
Designated retro2 in Figure 1
Designated retro3 in Figure 1
Figure 2. Amino Acid Sequence Comparisons between SLFs.
Deduced amino acid sequences were compared between PdSLFc and AhSLF-S2 (A) and between PdSLFc and PdSLFd (B). Sites that are conserved
or that have only conservative replacements (amino acid groups defined by Dayhoff et al. : C, STAPG, MILV, HRK, NDEQ, and FYW) are marked
with asterisks or dots, respectively.
Pollen-Expressed S Locus Gene of Almond775
proteins in the public databases. ORF3 has no known localiza-
tion signal, suggesting that it is a cytoplasmic protein. Genomic
DNA gel blot analysis with the ORF3 probe gave an Sc haplo-
type–specific signal in all cultivars that contained the Sc-RNase
gene (Figure 4A). Based on this result, ORF3 was designated
SFB (S haplotype–specific F-box protein). Phylogenetic analy-
sis showed that SFB is a member of the A3 subfamily in the
plant F-box superfamily defined by Gagne et al. (2002) (data
S Haplotype–Specific Sequence Polymorphism of SFB
To clone the alleles of SFB, PCR was conducted with the
primer pairs SC05MetRV/SCa05R or SC05MetRV/SFBTerHI
(Figure 3). Three clones were isolated from genomic DNA of cv
Mission (SaSb), Sauret No. 1 (SaSd), and Monterey (SbSd). Geno-
mic DNA gel blot analyses of the three clones showed the Sa,
Sb, and Sd haplotype–specific signals (Figure 4B), suggesting
that these SFB clones were derived from the respective S hap-
Because the three clones lacked both ends of the coding re-
gion, rapid amplification of cDNA ends was conducted using
pollen RNAs from different S haplotypes to obtain full-length
cDNA sequences of SFBa, SFBb, and SFBd. Deduced amino
acid sequences of the SFB alleles were aligned and compared
(Figure 5). The F-box motif was conserved at the N-terminal re-
gions of all SFBs. Sequence identities among SFBs were 68.4
to 76.4% (Table 2), which were comparable to those of almond
S-RNases (54.2 to 76.2%) (Ushijima et al., 1998; Tamura et al.,
2000). Although variable sites are dispersed in the primary
structure of the SFB, two regions at the C terminus are quite
variable, as they are in the hypervariable region of the S-RNase
(Ioerger et al., 1991; Ushijima et al., 1998).
Organ-Specific Expression of the S Locus Genes
To investigate the expression pattern of SFB and PdSLF, we
performed reverse transcriptase–mediated (RT) PCR analyses
using RNA from leaf and floral organs of Nonpareil as tem-
plates. RT-PCR products for the S-RNase gene and the actin
gene, which were used as controls, did not contain genomic
DNA–derived PCR products, which are expected to contain in-
trons and thus would be longer than the expected RT-PCR
products, suggesting that the templates were not contami-
nated with genomic DNA. In addition, the forward primer for
SFB was designed to include the junction of the two exons to
prevent amplification from a genomic DNA template (Figure 3).
The cDNA for S-RNase was amplified as expected from the
Figure 3. Nucleotide and Deduced Amino Acid Sequences of the cDNA for SFBc (ORF3).
The deduced amino acid sequence is shown under the nucleotide sequence of the cDNA. The arrowhead represents the position into which an 85-bp
intron sequence was inserted. The amino acid sequence of the F-box motif is underlined. Arrows represent the positions and directions of primers
that were used to amplify the alleles (see Figure 5) or to characterize organ-specific expression patterns (see Figure 6). The 5? region of primer
SC05MetRV (5?-TGATATCATTTTCTACAGGATGAC-3?) corresponds to the 3? end of the intron and is represented by a dotted line. Restriction sites
for EcoRV and BamHI were added to the 5? regions of primers SC05MetRV and SC05TerHI (5?-TGGATCCGTTTAATAATTATTGAG-3?), respectively.
776 The Plant Cell
style and at low amounts from the ovary (Figure 6). The cDNA
for PdSLF was amplified not only in the male organs (anther,
pollen grain, and germinated pollen) but also in the style, where
the pollen S gene is not expected to be expressed. On the
other hand, SFBc was expressed only in the male organs. Sig-
nal intensities of pollen and germinated pollen were higher than
that of anther, suggesting that SFB likely is expressed only in
Comparison of the Gene Organization of Four S Haplotypes
The cosmid or fosmid libraries were constructed from the ge-
nomic DNAs of almond cv Nonpareil (ScSd), Sauret No. 1 (SaSc),
and Monterey (SbSd). To construct the contigs encompassing
the Sa, Sb, and Sd haplotypes, we screened the libraries with
probes for SFBs, the S-RNase genes, and the S locus bound-
ary markers NP79R (?PdSLF) and NP182F (Ushijima et al.,
2001). The order and orientations of the genes and markers
were determined by long PCR. The S-RNase genes and SFBs
of all S haplotypes were located in the region between the two
boundary markers NP79R and NP182F. The relative order and
orientations of the genes and markers were conserved among
the four S haplotypes (Figure 7). However, the physical dis-
tances between them were significantly different. The physical
distance between SFB and the S-RNase gene was ?1.6 kb in
the Sc haplotype but was ?10 kb in the other three S haplo-
We previously scanned the S locus region of almond for se-
quence diversity by DNA gel blot analysis using probes located
at varying distances from the S-RNase gene. The results
showed that the region between two boundary markers,
NP79R and NP182F, is highly diverged among the different S
haplotypes, but the regions outside of the two markers are rela-
tively well conserved (Ushijima et al., 2001). In this study, con-
tigs encompassing the two markers were constructed for three
additional S haplotypes. Comparative analysis of the contigs
showed that the physical distances between genes and mark-
Figure 4. Genomic DNA Gel Blot Analyses for SFB.
Genomic DNAs from seven almond cultivars were digested with HindIII, separated on 0.8% agarose gels, and transferred to nylon membranes. The
blot was probed with the Sc haplotype genes (SFBc and the Sc-RNase gene in [A]) and the homologs of SFBc (SFBa, SFBb, and SFBd in [B]). The ar-
rows Sa, Sb, Sc, and Sd denote the respective S haplotype–specific signals. Abbreviations for almond cultivars are as follows: JF, Jeffries (ScmSd); NP,
Nonpareil (ScSd); MS, Mission (SaSb); #2, Sauret No. 2 (SaSc); MR, Merced (SbSc); #1, Sauret No. 1 (SaSd); MT, Monterey (SbSd).
Pollen-Expressed S Locus Gene of Almond777
ers are quite variable among different S haplotypes, supporting
our previous idea that the region is heteromorphic (Ushijima et
al., 2001). Sequence analysis of the Sc haplotype region re-
vealed that the region is rich in retrotransposon-like sequences,
which may have been causal to the region being heteromor-
phic. These findings indicate that the almond S locus has been
subjected to repeated rearrangements, deletions, and inser-
tions, like the S locus of Brassicaceae (Boyes et al., 1997;
Casselman et al., 2000; Kusaba et al., 2001). The structural het-
eromorphism is a distinctive feature of loci consisting of co-
adapted gene complexes in different genetic systems (Ferris
and Goodenough, 1994; Brown and Casselton, 2001). The het-
eromorphism of the S locus region is believed to maintain the
tight association of the two S specificity genes, the S-RNase
gene and the pollen S gene, in the RNase-based GSI and sug-
gests that the pollen S gene is located within the S haplotype–
specific region, which is defined by the markers NP79R and
NP182F in almond.
Sequence analysis of the Sc haplotype–specific region fol-
lowed by cDNA cloning revealed that two genes located in the
region, PdSLF and SFB, are expressed in pollen. PdSLF corre-
sponded to the S locus boundary marker NP79R and, as
expected, showed little sequence polymorphism between al-
leles. Considering the high degree of sequence diversity of the
S-RNases, the pollen S protein is expected to show a high level
of sequence diversity for specific interaction with the cognate
Figure 5. Alignment of the Amino Acid Sequences of SFBs and the F-Box Motif.
(A) Amino acid sequence alignment of four SFBs. Amino acid sequences of SFBs were aligned using CLUSTAL X (Thompson et al., 1997). The F-box
motif and two variable regions are boxed. Sites that are conserved or that have only conservative replacements are marked with asterisks or dots, re-
(B) Alignment of amino acid sequences of the F-box motif. Amino acid sequences of F-box motifs from 13 F-box proteins were aligned. Conserved
amino acid residues are shown as black boxes. Arrowheads mark the amino acid positions important for the Skp–F-box protein interactions between
human Skp1 and Skp2 (Schulman et al., 2000). The species from which each sequence is derived is denoted by the initials of its scientific name (i.e.,
Pd means Prunus dulcis [almond]). Sequence data included in the alignment are as follows: SLF-S2 Antirrhinum hispanicum; UFO, FBX8, ORE9, and
TIR1 of Arabidopsis thaliana; Fim of Antirrhinum majus; Stp of Pisum sativum; and Skp2 of Homo sapiens.
778 The Plant Cell
S-RNase. Therefore, it seems unlikely that PdSLF is the pollen
In contrast to PdSLF, SFB showed distinctive features of a
candidate for the pollen S gene: tight linkage to the S-RNase
gene; a high level of sequence polymorphism; and pollen-spe-
cific expression. All SFBs were observed to be located at the S
haplotype–specific region and to be linked physically (within
?30 kb) to the S-RNase gene. In addition, heteromorphism at the
S haplotype–specific region also could maintain the linkage of the
two genes by suppressing recombination at this region, suggest-
ing that SFB and the S-RNase gene are inherited as a unit.
SFB showed a high level of the S haplotype–specific se-
quence polymorphism, comparable to that observed for the
S-RNases of Prunus (Ushijima et al., 1998; Tamura et al., 2000).
Amino acid sequence alignment showed that SFB contained
two quite variable regions, as did S-RNase (Ioerger et al., 1991;
Ushijima et al., 1998). The hypervariable region of S-RNase is
exposed on the surface of the folded protein molecule and
plays a pivotal role in the recognition of self-pollen (Matton et
al., 1997, 1999; Ida et al., 2001; Matsuura et al., 2001). The two
variable regions of SFB at the C terminus were found to be hy-
drophilic enough to be exposed on the surface (data not
shown) and may have important recognition functions analo-
gous to those of the S-RNase.
The pollen SI phenotype of Rosaceae and Solanaceae is
controlled gametophytically, suggesting that the pollen S gene
is expressed only in pollen. RT-PCR analysis revealed that SFB
is expressed exclusively in the male reproductive organs and
that expression in pollen grains and in germinated pollen is
higher than that observed in whole anthers, suggesting that
SFB is expressed only in pollen. This result is consistent with
our finding that the 1.2-kb promoter region of SFBc directs the
expression of a ?-glucuronidase reporter gene in pollen but not
in anther cells of transgenic tobacco (H. Sassa, K. Ushijima, R.
Tao, and H. Hirano, unpublished results). Although we cannot
exclude the possibility that the anther cell also expresses a
small amount of SFB, it seems unlikely that the sporophytically
expressed SFB affects the pollen SI phenotype, because it
lacks signal peptide for secretion. In addition, the pollen tube
rejection of the RNase-based GSI is a slow reaction that occurs
inside the style tissue. This finding is in contrast to the sporo-
phytic SI of Brassica, which is a quick response on the stigma
surface, and suggests that exogenous SFB might have little ef-
fect on pollen. These findings suggest that SFB functions
strictly in a gametophytic manner. All of the features of SFB
discussed above are consistent with the hypothesis that SFB
controls pollen S specificity in almond, representing the candi-
date for the pollen S gene of the RNase-based GSI systems.
Further analyses for physical interaction between SFB and the
S-RNase, as well as functional analyses, will be required to
clarify the role of SFB in the RNase-based GSI.
The F-box proteins are known to be a component of a class
of ubiquitin ligase, the SCF complex (Deshaies, 1999). In the
ubiquitin/26S proteasome pathway of protein degradation, tar-
get proteins destined for degradation become modified by co-
valent attachment of multiple ubiquitins. The polyubiquitinated
proteins then are recognized by the 26S proteasome and de-
graded. The SCF complex catalyzes the final step of ubiquitina-
tion of target proteins. The F-box protein serves as a receptor to
recruit appropriate target proteins to the core complex for ubiq-
uitination. It associates with an SCF component, Skp1, through
the F-box motif at the N terminus, and its C terminus binds spe-
cifically to the target (Bai et al., 1996). Our finding of the F-box
protein SFB as a candidate for the pollen S gene product of the
RNase-based GSI suggests that the likely target of the SCF
complex containing the SFB, SCFSFB, is the S-RNase.
The idea that the S-RNase is degraded in pollen tubes
through the ubiquitin/26S proteasome pathway seems to be
compatible with a proposed model for SI, which hypothesizes
that the pollen S gene product is an RNase inhibitor. Despite
the apparent difficulty of accommodating the high degrees of
structural diversity of the S-RNases, the inhibitor model is sup-
ported by recent findings on S haplotype–independent uptake
of the S-RNase by the pollen tube (Luu et al., 2000) and the es-
sential role of pollen S products for pollen tube elongation (Golz
et al., 2001). As a possible interpretation of the peculiar nature
of the pollen S product in the model, it was hypothesized that a
pollen S protein contains an RNase inhibitor domain and an S
haplotype specificity domain and that it interacts differently
with self and nonself S-RNases (McCubbin and Kao, 2000).
SFB might function as a component of SCFSFB and ubiquitinate
all of the nonself S-RNases for degradation but interact specifi-
cally with the self S-RNase to leave it active, leading to the ar-
rest of self pollen tube growth. Recently, it was reported that
hnRNP-U functions as a pseudosubstrate that binds to the cor-
responding F-box protein but is not destined for degradation
by the SCF complex (Davis et al., 2002). Based on this example,
it is possible that self S-RNase serves as a pseudosubstrate for
Table 2. Amino Acid Sequence Identities (%) among SFBs
Figure 6. Organ-Specific Expression of S Locus Genes.
RNAs were derived from leaf and floral organs of Nonpareil. cDNAs
were synthesized from the RNAs and were used for RT-PCR with gene-
specific primers. L, leaf; Pd, pedicel; Sp, sepal; Pt, petal; St, style; O,
ovary; A, anther; P, pollen; gP, germinated pollen.
Pollen-Expressed S Locus Gene of Almond779
SCFSFB. The two variable regions located at the C terminus of
SFB are implicated in binding with the pseudosubstrate self
S-RNase. In plants, the F-box gene family shows a substantial
expansion relative to other eukaryotes (Gagne et al., 2002).
Identification of almost 700 probable F-box genes in Arabidop-
sis suggests that the F-box gene family is one of the largest in
plants and that the SCF complex–mediated ubiquitin/26S pro-
teasome pathway is a major route for cellular regulation. The
RNase-based GSI might represent one of the specialized roles
of this highly divergent proteolytic pathway in plants.
Seven cultivars of almond (Prunus dulcis)—Nonpareil (ScSd), Mission
(SaSb), Sauret No. 2 (SaSc), Merced (SbSc), Sauret No. 1 (SaSd), Monterey
(SbSd), and Jeffries (ScmSd)—were used. Jeffries is a naturally occurring
mutant found as a sport on a Nonpareil tree and lacks the Sc haplotype,
and the mutated haplotype is designated Scm (Kester et al., 1994; Tao et
al., 1997; Ushijima et al., 2001). Pollen was germinated in vitro as de-
scribed by Lush et al. (1997) with slight modifications.
Genomic clones (Figure 1A) were sheared physically by sonication,
blunt-ended, and phosphorylated with T4 DNA polymerase and T4 nu-
cleotide kinase (New England Biolabs, Beverly, MA). The resulting frag-
ments were separated on a 1% agarose gel, and the 1.5- to 2.0-kb frag-
ments were cloned into pBluescript II SK? (Stratagene, La Jolla, CA) or
pGEM-3Zf(?) (Promega, Madison, WI). The lengths of the inserted frag-
ments were confirmed by PCR. The PCR products were purified by Mul-
tiscreen PCR (Millipore, Bedford, MA) and sequenced with an autose-
quencer (model 4000L; Li-Cor, Lincoln, NE). Sequencing was continued
until a minimum of approximately fourfold sequence coverage was
achieved for each clone. Sequence alignments were generated using
SeqMan II software (DNASTAR, Madison, WI). Gaps between shotgun
contigs were amplified and filled by PCR with Pyrobest DNA polymerase
(TaKaRa, Otsu, Japan).
Isolation of Nucleic Acids
Freeze-dried leaves were ground into powder using a mortar and pestle.
Genomic DNAs were isolated from the ground leaves as described by
Ushijima et al. (2001). RNAs were isolated from leaves and floral organs
of Nonpareil as described by McClure et al. (1990).
Rapid Amplification of cDNA Ends
Total RNA from the pollen of Nonpareil was used for the synthesis of
first-strand cDNA essentially as described for the SMART RACE (rapid
amplification of cDNA ends) cDNA Amplification Kit (Clontech, Palo Alto,
CA) and by Schmidt and Mueller (1999). The reaction mixture contained
1 ? first-strand synthesis buffer (50 mM Tris-HCl, pH 8.3, 75 mM KCl,
and 6 mM MgCl2), 1 mM deoxynucleotide triphosphates, 5 mM DTT, 2
mM MnCl2, 1 ?M RACE-N (5?-AAGGCTCCGTCGGCATCGATCGCG-
CGACTCTTTTTTTTTTTTTTTTT-3?), 1 ?M SMART II (5?-AAGCAGTGG-
TAACAACGCAGAGTACGCGGG-3?), 4 ?g of total RNA, and 200 units of
SuperScript II reverse transcriptase (Invitrogen, Carlsbad, CA) and was
incubated at 42?C for 1.5 h. RACE-N and SMART II add the sequences
for adapter primers to the cDNA at the 3? and 5? ends, respectively. 3?
RACE used the resulting cDNA, ExTaq (TaKaRa), RACE II (5?-AAG-
GCTCCGTCGGCATCGATC-3?) as the adapter primer, and the gene-
specific primers designed from SFBs or the putative open reading
frames predicted by the GENSCAN program. 5? RACE used universal
primer mixture (UPM) as the adapter primer instead of RACE II; 1 ? UPM
contains two primers, 0.1 ?M UPM short (5?-CTAATACGACTCACTATA-
GGGC-3?) and 0.02 ?M UPM long (5?-CTAATACGACTCACTATA-
GGGCAAGCAGTGGTAACAACGCAGAGT-3?), and extends the 5? end of
the cDNA for step-out PCR (Matz et al., 1999). If necessary, second-
round PCR was conducted with gene-specific primers and adapter
primers RACE III (5?-CGGCATCGATCGCGCGACTC-3?) and NUP (5?-
AAGCAGTGGTAACAACGCAGAGT-3?) for 3? and 5? RACE, respectively.
RACE fragments were subjected to DNA gel blot analyses using the
three genomic clones for the Sc haplotype–specific region (Figure 1A) as
probes to detect cDNA species with low expression levels. The RACE
fragments showing signal(s) were cloned into a plasmid vector. The pos-
itive clones for each RACE fragment were selected by colony hybridiza-
tion with the probes of the genomic clones. For each open reading
frame, 6 to 11 independent cDNA clones derived from different primer
pairs were subjected to sequence analysis.
Figure 7. Gene Organization of Four S Haplotypes of Almond.
Boxes represents exons of the S-RNase genes, SFBs, and the S locus boundary markers NP79R and NP182F (Ushijima et al., 2001). NP79R is part of
the PdSLF gene (see text). Arrows denote the orientations of the genes. In the Sa and Sb haplotypes, the physical distance between NP79R and the
S-RNase gene was not determined accurately, although it is certain that NP79R links to the S-RNase gene.
780 The Plant Cell
Genomic DNA Gel Blot Analysis
Five micrograms of genomic DNAs digested with HindIII was separated
and blotted onto a nylon membrane. The membrane was probed with
digoxigenin-labeled cDNAs for genes expressed in pollen, washed, and
visualized as described by Ushijima et al. (2001).
Reverse Transcriptase–Mediated PCR for Characterization of the
Gene Expression Pattern
RNAs from leaves and floral organs of Nonpareil were treated with
DNaseI (Invitrogen). Their cDNAs were synthesized by PowerScript
(Clontech) with oligo-d(T) primer. The resulting cDNAs were used as
templates for PCR amplification with gene-specific primer sets SFB-
cFmet (5?-CCAACCGCAAAAGATGACATTC-3?) and SFB-C2R (5?-GGG-
TTCCATATKTGTATWGG-3?) for SFBc (Figure 3), SCa01F (5?-GATTGG-
TGGGGATGTGCTGTAG-3?) and SCa01R (5?-CTTCGGTAACCATAA-
GAATCTC-3?) for PdSLF, AS1II and Amy-C5R for the S-RNase gene
(Tamura et al., 2000), and ActF1 (5?-ATGGTGAGGATATTCAACCC-3?)
and ActR1 (5?-CTTCCTGTGGACAATGGATGG-3?) for the actin gene that
was used as an internal control. PCR was performed with ExTaq
(TaKaRa) using a program of 30 cycles at 94?C for 30 s, 53?C for 30 s,
and 72?C for 45 s, an initial denaturing at 94?C for 2.5 min, and a final ex-
tension at 72?C for 7 min. PCR products were separated on a 1.5% aga-
rose gel and stained with ethidium bromide.
Construction and Screening of the Cosmid Library
Cosmid or fosmid libraries were constructed from three almond culti-
vars—Nonpareil (ScSd), Sauret No.1 (SaSc), and Monterey (SbSd)—using
the pWEB::TNC Cosmid Cloning Kit or the CopyControl Fosmid Library
Production Kit (Epicentre, Madison, WI). The libraries were screened
with digoxigenin-labeled probes for the S-RNase gene SFB, NP79R and
NP182F, as described by Ushijima et al. (2001). Long PCR with ExTaq or
the Expand Long Template PCR System (Roche Diagnostics, Mann-
heim, Germany) was performed to determine the order and the orienta-
tions of the genes on the cosmid/fosmid clones.
Upon request, all novel materials described in this article will be made
available in a timely manner for noncommercial research purposes.
The accession numbers for the sequences mentioned in this article are
as follows: AB081587 (Sc haplotype–specific region), AB092966 (SFBa),
AB092967 (SFBb), AB079776 (SFBc), AB081648 (SFBd). Other acces-
sion numbers are AJ297974 (SLF-S2 of Antirrhinum hispanicum),
NM_102834, AC005825, AF305597, and AF327430 (UFO, FBX8, ORE9,
and TIR1 of Arabidopsis thaliana), S71192 (Fim of Antirrhinum majus),
AF004843 (Stp of Pisum sativum), and AB050979 (Skp2 of Homo sapiens).
We gratefully acknowledge S. Uratsu (University of California, Davis) for
collecting and providing plant materials and M. Kusaba (Institute of Ra-
diation Breeding, National Institute of Agrobiological Sciences, Ministry
of Agriculture, Forestry, and Fisheries, Japan), for helpful comments.
This work was supported by Grant-in-Aid 13660011 for Scientific Re-
search (C) to H.S. and Grant-in-Aid 13460014 for Scientific Research (B)
to R.T. from the Japan Society for the Promotion of Science and by
grants to A.M.D. from the Almond Board of California.
Received November 14, 2002; accepted December 24, 2002.
Altschul, S.F., Gish, W., Miller, W., Myers, E.W., and Lipman, D.J.
(1990). Basic local alignment search tool. J. Mol. Biol. 215, 403–410.
Bai, C., Sen, P., Hofmann, K., Ma, L., Goebl, M., Harper, J.W., and
Elledge, S.J. (1996). SKP1 connects cell cycle regulators to the ubiq-
uitin proteolysis machinery through a novel motif, the F-box. Cell 86,
Bernacchi, D., and Tanksley, S.D. (1997). An interspecific backcross
of Lycopersicon esculentum ? L. hirsutum: Linkage analysis and a
QTL study of sexual compatibility factors and floral traits. Genetics
Boyes, D.C., Nasrallah, M.E., Vrebalov, J., and Nasrallah, J.B. (1997).
The self-incompatibility (S) haplotypes of Brassica contain highly
divergent and rearranged sequences of ancient origin. Plant Cell 9,
Brown, A.J., and Casselton, L.A. (2001). Mating in mushrooms:
Increasing the chances but prolonging the affair. Trends Genet. 17,
Burge, C., and Karlin, S. (1997). Prediction of complete gene structures
in human genomic DNA. J. Mol. Biol. 268, 78–94.
Casselman, A.L., Vrebalov, J., Conner, J.A., Singhal, A., Giovannoni,
J., Nasrallah, M.E., and Nasrallah, J.B. (2000). Determining the
physical limits of the Brassica S locus by recombinational analysis.
Plant Cell 12, 23–33.
Coleman, C.E., and Kao, T.-H. (1992). The flanking regions of two
Petunia inflata S alleles are heterogeneous and repetitive sequences.
Plant Mol. Biol. 18, 499–511.
Davis, M., Hatzubai, A., Andersen, J.S., Ben-Shushan, E., Fisher,
G.Z., Yaron, A., Bauskin, A., Mercurio, F., Mann, M., and Ben-
Neriah, Y. (2002). Pseudosubstrate regulation of the SCF?-TrCP ubiq-
uitin ligase by hnRNP-U. Genes Dev. 16, 439–451.
Dayhoff, M.O., Schwartz, R.M., and Orcutt, B.C. (1979). A model of
evolutionary change in proteins. In Atlas of Protein Sequence and
Structure, Vol. 5, Suppl. 3, M.O. Dayhoff, ed (Washington, DC:
National Biomedical Research Foundation), pp. 345–352.
de Nettancourt, D. (2001). Incompatibility and Incongruity in Wild and
Cultivated Plants. (Berlin: Springer-Verlag).
Deshaies, R.J. (1999). SCF and cullin/RING H2-based ubiquitin ligases.
Annu. Rev. Cell Dev. Biol. 15, 435–467.
Dodds, P.N., Ferguson, C., Clarke, A.E., and Newbigin, E. (1999).
Pollen-expressed S-RNases are not involved in self-incompatibility in
Lycopersicon peruvianum. Sex. Plant Reprod. 12, 76–87.
Entani, T., Iwano, M., Shiba, H., Takayama, S., Fukui, K., and Isogai,
A. (1999). Centromeric localization of an S-RNase gene in Petunia
hybrida Vilm. Theor. Appl. Genet. 99, 391–397.
Ferris, P.J., and Goodenough, U.W. (1994). The mating-type locus
of Chlamydomonas reinhardtii contains highly rearranged DNA se-
quences. Cell 76, 1135–1145.
Frohman, M.A., Dush, M.K., and Martin, G.R. (1988). Rapid produc-
tion of full-length cDNAs from rare transcripts: Amplification using a
single gene-specific oligonucleotide primer. Proc. Natl. Acad. Sci.
USA 85, 8998–9002.
Gagne, J.M., Downes, B.P., Shiu, S.-H., Durski, A.M., and Vierstra,
R.D. (2002). The F-box subunit of the SCF E3 complex is encoded by
a diverse superfamily of genes in Arabidopsis. Proc. Natl. Acad. Sci.
USA 99, 11519–11524.
Golz, J.F., Su, V., Clarke, A.E., and Newbigin, E. (1999). A molecular
Pollen-Expressed S Locus Gene of Almond781
description of mutations affecting the pollen component of the Nicoti-
ana alata S locus. Genetics 152, 1123–1135.
Golz, J.F., Su, V., Oh, H.-Y., Kusaba, M., and Newbigin, E. (2001).
Genetic analysis of Nicotiana pollen-part mutants is consistent with
the presence of an S-ribonuclease inhibitor at the S locus. Proc. Natl.
Acad. Sci. USA 98, 15372–15376.
Huang, S., Lee, H.-S., Karunanandaa, B., and Kao, T.-H. (1994).
Ribonuclease activity of Petunia inflata S proteins is essential for
rejection of self-pollen. Plant Cell 6, 1021–1028.
Ida, K., Norioka, S., Yamamoto, M., Kumasaka, T., Yamashita, E.,
Newbigin, E., Clarke, A.E., Sakiyama, F., and Sato, M. (2001). The
1.55 Å resolution structure of Nicotiana alata SF11-RNase associated
with gametophytic self-incompatibility. J. Mol. Biol. 314, 103–112.
Ioerger, T.R., Gohlke, J.R., Xu, B., and Kao, T.-H. (1991). Primary
structural features of the self-incompatibility protein in Solanaceae.
Sex. Plant Reprod. 4, 81–87.
Kester, D.E., Micke, W.C., and Viveros, M. (1994). A mutation in ‘Non-
pareil’ almond conferring unilateral incompatibility. J. Am. Soc. Hor-
tic. Sci. 119, 1289–1292.
Kusaba, M., Dwyer, K., Hendershot, J., Vrebalov, J., Nasrallah, J.B.,
and Nasrallah, M.E. (2001). Self-Incompatibility in the genus Arabi-
dopsis: Characterization of the S locus in the outcrossing A. lyrata
and its autogamous relative A. thaliana. Plant Cell 13, 627–643.
Lai, Z., Ma, W., Han, B., Liang, L., Zhang, Y., Hong, G., and Xue, Y.
(2002). An F-box gene linked to the self-incompatibility (S) locus of
Antirrhinum is expressed specifically in pollen and tapetum. Plant
Mol. Biol. 50, 29–42.
Lee, H.-S., Huang, S., and Kao, T.-H. (1994). S proteins control rejec-
tion of incompatible pollen in Petunia inflata. Nature 367, 560–563.
Lush, W.M., Opat, A.S., Nie, F., and Clarke, A.E. (1997). An in vitro
assay for assessing the effects of growth factors on Nicotiana alata
pollen tubes. Sex. Plant Reprod. 10, 351–357.
Luu, D., Qin, X., Morse, D., and Cappadocia, M. (2000). S-RNase
uptake by compatible pollen tubes in gametophytic self-incompatibil-
ity. Nature 407, 649–651.
Matsuura, T., Sakai, H., Unno, M., Ida, K., Sato, M., Sakiyama, F.,
and Norioka, S. (2001). Crystal structure at 1.5 Å resolution of Pyrus
pyrifolia pistil ribonuclease responsible for gametophytic self-incom-
patibility. J. Biol. Chem. 276, 45261–45269.
Matton, D., Maes, O., Laublin, G., Xike, Q., Bertrand, C., Morse, D.,
and Cappadocia, M. (1997). Hypervariable domains of self-incom-
patibility RNases mediate allele-specific pollen recognition. Plant Cell
Matton, D.P., Luu, D.T., Xike, Q., Laublin, G., O’Brien, M., Maes, O.,
Morse, D., and Cappadocia, M. (1999). Production of an S RNase
with dual specificity suggests a novel hypothesis for the generation of
new S alleles. Plant Cell 11, 2087–2097.
Matton, D.P., Mau, S.L., Okamoto, S., Clarke, A.E., and Newbigin, E.
(1995). The S-locus of Nicotiana alata: Genomic organization and
sequence analysis of two S-RNase alleles. Plant Mol. Biol. 28, 847–858.
Matz, M., Shagin, D., Bogdanova, E., Britanova, O., Lukyanov, S.,
Ditachenco, L., and Chenchik, A. (1999). Amplification of cDNA
ends based on template-switching effect and step-out PCR. Nucleic
Acids Res. 27, 1558–1560.
McClure, B.A., Gray, J.E., Anderson, M.A., and Clarke, A.E. (1990).
Self-incompatibility in Nicotiana alata involves degradation of pollen
rRNA. Nature 347, 757–760.
McClure, B.A., Haring, V., Ebert, P.R., Anderson, M.A., Simpson,
R.J., Sakiyama, F., and Clarke, A.E. (1989). Style self-incompatibility
gene products of Nicotiana alata are ribonucleases. Nature 342, 955–957.
McCubbin, A.G., Chung, Y.-Y., and Kao, T.-H. (1997). A mutant S3
RNase of Petunia inflata lacking RNase activity has an allele-specific
dominant negative effect on self-incompatibility interactions. Plant
Cell 9, 85–95.
McCubbin, A.G., and Kao, T.-H. (2000). Molecular recognition and
response in pollen and pistil interactions. Annu. Rev. Cell Dev. Biol.
McCubbin, A.G., Zuniga, C., and Kao, T.-H. (2000). Construction of a
binary bacterial artificial chromosome library of Petunia inflata and the
isolation of large genomic fragments linked to the self-incompatibility
(S-) locus. Genome 43, 820–826.
Murfett, J., Atherton, T.L., Mou, B., Gasser, C.S., and McClure, B.A.
(1994). S-RNase expressed in transgenic Nicotiana causes S-allele-
specific pollen rejection. Nature 367, 563–566.
Royo, J., Kuntz, C., Kowyama, Y., Anderson, M., and Clarke, A.E.
(1994). Loss of a histidine residue at the active site of S-locus ribonu-
clease is associated with self-compatibility in Lycopersicon peruvi-
anum. Proc. Natl. Acad. Sci. USA 91, 6511–6514.
Sakata, K., Nagamura, Y., Numa, H., Antonio, B.A., Nagasaki, H.,
Idonuma, A., Watanabe, W., Shimizu, Y., Horiuchi, I., Matsumoto,
T., Sasaki, T., and Higo, K. (2002). RiceGAAS: An automated anno-
tation system and database for rice genome sequence. Nucleic Acids
Res. 30, 98–102.
Sassa, H., Hirano, H., Nishino, T., and Koba, T. (1997). Style-specific
self-compatible mutation caused by deletion of the S-RNase gene in
Japanese pear (Pyrus serotina). Plant J. 12, 223–227.
Schmidt, M.W., and Mueller, M.W. (1999). CapSelect: A highly sensi-
tive method for 5? CAP-dependent enrichment of full-length cDNA in
PCR-mediated analysis of mRNAs. Nucleic Acids Res. 27, e31.
Schulman, B.A., Carrano, A.C., Jeffrey, P.D., Bowen, Z., Kinnucan,
E.R.E., Finnin, M.S., Elledge, S.J., Harper, J.W., Pagano, M., and
Pavletich, N.P. (2000). Insights into SCF ubiquitin ligases from the
structure of the Skp1-Skp2 complex. Nature 408, 381–386.
Tamura, M., Ushijima, K., Sassa, H., Hirano, H., Tao, R., Gradziel,
T.M., and Dandekar, A.M. (2000). Identification of self-incompatibil-
ity genotypes of almond by allele-specific PCR analysis. Theor. Appl.
Genet. 101, 344–349.
Tao, R., Yamane, H., Sassa, H., Mori, H., Gradziel, T.M., Dandekar,
A.M., and Sugiura, A. (1997). Identification of stylar RNases associ-
ated with gametophytic self-incompatibility in almond (Prunus dulcis).
Plant Cell Physiol. 38, 304–311.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., and
Higgins, D.G. (1997). The CLUSTAL X Windows interface: Flexible
strategies for multiple sequence alignment aided by quality analysis
tools. Nucleic Acids Res. 25, 4876–4882.
Thompson, R.D., and Kirch, H.H. (1992). The S locus of flowering
plants: When self-rejection is self-interest. Trends Genet. 8, 381–387.
Thompson, R.D., Uhrig, H., Hermsen, J.G.T., Salamini, F., and
Kaufmann, H. (1991). Investigation of a self-compatible mutation in
Solanum tuberosum clones inhibiting S-allele activity in pollen differ-
entially. Mol. Gen. Genet. 226, 283–288.
Ushijima, K., Sassa, H., Kusaba, M., Tao, R., Tamura, M., Gradziel,
T.M., Dandekar, A.M., and Hirano, H. (2001). Characterization of the
S-locus region of almond (Prunus dulcis): Analysis of a somaclonal
mutant and a cosmid contig for an S haplotype. Genetics 158, 379–386.
Ushijima, K., Sassa, H., Tao, R., Yamane, H., Dandekar, A.M.,
Gradziel, T.M., and Hirano, H. (1998). Cloning and characterization
of cDNAs encoding S-RNases from almond (Prunus dulcis): Primary
structural features and sequence diversity of the S-RNases in
Rosaceae. Mol. Gen. Genet. 260, 261–268.