The Plant Cell, Vol. 5, 657-666, June 1993 O 1993 American Society of Plant Physiologists
An Alternative Transcript of the S Locus
Glycoprotein Gene in a Class II Pollen-Recessive
Self-lncompatibility Haplotype of Brassica oleracea
Encodes a Membrane-Anchored Protein
Titima Tantikanjana, Mikhail E. Nasrallah, Joshua C. Stein, Che-Hong Chen,’ and June B. Nasrallah2
Section of Plant Biology, Division of Biological Sciences, Cornell University, Ithaca, New York 14853
Recent reports have shown that SLG, one of two genes linked to the S locus of Brassica, encodes a secreted glycoprotein. We have used RNA gel blot analysis, genomic and cDNA clone analysis, expression in transgenic plants, and
immunodetection to characterize SLG2, the SLG gene derived from the S2 haplotype. This haplotype belongs to the class
II group of S haplotypes that exhibit a weak incompatibility phenotype and are pollen recessive. We showed that SLG2
produces two transcript forms: the expected 1.6-kb transcript that predicts a secreted glycoprotein and an alternative
1.8-kb transcript that predicts a membrane-anchored protein. Stigmas of the S2 haplotype and pistils of transgenic
tobacco plants transformed with the SLG2 gene produce a membrane-associated 62-kD protein as well as soluble 57and 58-kD glycoforms. Because of the sequence similarity between SLG2 and the extracellular domain of the S Locus
Receptor Kinase (SRK2) gene, the membrane-anchoredform of SLG2may be viewed as a naturally occurring truncated
form of the receptor that lacks the kinase catalytic domain. The occurrence of this protein has potential implications
for the activity of the full-length receptor. Furthermore, the underlying structure of the SLG2 gene suggests the evolution of SLG from an ancestral SRK-like gene.
INTRODUCTION
The self-incompatibility response of Brassica involves cell-tocell interactions between pollen and the papillar cells of the
stigma that result in the recognitionand inhibition of self-pollen
at the stigma surface. More than 50 naturally occurring variants, classically designated as S alleles, have been identified
at the self-incompatibility (S) locus in Brassica oleracea
(Ockendon, 1974). Self-incompatibility in this genus is sporophytically controlled (Bateman, 1955), and the activity of S
alleles in the stigma and pollen of heterozygous plants is dependent on the outcome of complex dominanthecessiveallelic
interactions (Thompson and Taylor, 1966). Based on their behavior relative to other alleles in the stigma and pollen of
heterozygous plants, S alleles have been arranged in a dominance series (Thompson and Taylor, 1966).Two major classes
of S alleles have been recognized(Nasrallah et al., 1991). Highactivity (class I) alleles are placed relatively high on the dominance scale and exhibit a strong incompatibility phenotype,
whereas low-activity (class II) alleles have a weak or leaky incompatibility phenotypeand exhibit recessive and competitive
interactions in%pollen.
.Molecular analysis of the Brassica self-incompatibility system has led to the view that the S locus is a complex locus
i
Current address: Sogetal Incorporated, Hayward, CA 95455.
To whom correspondence should be addressed.
that consists of at least two physically linked genes (Boyes and
Nasrallah, 1993): the S Locus Glycoprotein(SLG) gene encodes
a secreted glycoprotein (Nasrallah et al., 1985, 1987; Kandasamy
et al., 1989), and the S Receptor Kinase (SRK) gene encodes
a transmembrane protein kinase with an extracellular domain
that shares sequence similarity with SLG (Stein et al., 1991).
In view of this complexity, the naturally occurring S locus variants, historically called “S alleles,” should be viewed as “S
haplotypes” (Nasrallah and Nasrallah, 1993).
The high degree of S locus polymorphism is reflected in the
sequencevariability of S locus genes. In comparisonsbetween
different S haplotypes, extensive sequence polymorphisms are
observed between alleles of the SLG gene (Nasrallah et al.,
1987; Trick and Flavell, 1989; Chen and Nasrallah, 1990; Dwyer
et al., 1991; Goring et al., 1992a, 1992b; Gaude et al., 1993)
and SRKgene (Stein et al., 1991; Goring and Rothstein, 1992).
A particularly high leve1of polymorphism has been noted between class I and class II haplotypes:the SLG and SRKgenes
in the class II S, haplotype share only -70% sequence identity with class I alleles of these genes (Chen and Nasrallah,
1990; Stein et al., 1991).
Whether derived from class I or class II haploiypes, S locus
genes share certain features in common. In both classes, the
SRK gene encodes a 3.0-kb transcriptand consists of six exons
658
The Plant Cell
separated by seven introns, with its extracellular SLGhomologous S domain encoded in its entirety by the first exon
of the gene (Stein et al., 1991). In contrast, all SLG alleles
reported to date encode a 1.6-kb transcript (Nasrallah et al.,
1985,1987,1988; Trick and Flavell, 1989; Goring et al., 1992a,
1992b; Gaude et al., 1993). Where genomic clones were available, this transcript was shown to be colinear with the genomic
sequence, leading to the conclusion that the gene is intronless (Nasrallah et al., 1988; Chen and Nasrallah, 1990). SLG
and the S domain of SRK isolated from the same S haplotype
share a high level (>90%) of sequence identity, suggesting
that the two genes evolve in concert (Stein et al., 1991). The
origin of the SLG/SRK gene pair and the processes that allow
the apparent coevolution of these genes within a haplotype
while generating a high degree of polymorphism between
haplotypes are not understood. Nevertheless, the sequence
conservation exhibited by the SLG/SRK gene pair and the fact
that both genes are expressed specifically in stigmas and anthers (Sato et al., 1991; Stein et al., 1991) suggest a functional
interaction between their products. Together, the findings that
SRK exhibits intrinsic protein kinase activity (Goring and
Rothstein, 1992; Stein and Nasrallah, 1993) and that SLG is
necessary for the operation of self-incompatibility (Toriyama
et al., 1991; Nasrallah et al., 1992) suggested a model of selfincompatibility in which SRK, acting in combination with SLG,
couples the recognition event at the pollen-stigma interface
to a cytoplasmic phosphorylation cascade that leads to pollen rejection (Stein et al., 1991).
To investigate the molecular basis of the genetic interactions
that occur between S haplotypes in heterozygotes and of the
variable activity exhibited by different haplotypes, we have focused on the analysis of two representative haplotypes: the
class IS6 haplotype and the class II S2 haplotype (Chen and
Nasrallah, 1990; Stein et al., 1991). In this report, we show,
through the isolation of SLG2 cDNA clones and protein analysis, that unlike class I SLG genes which encode only secreted
glycoproteins, the SLG2 gene encodes a membrane-anchored
SLG protein as well as a secreted SLG form. This feature is
discussed in reference to the genetic and phenotypic properties of the S2 haplotype. The unusual structure of the SLG2
gene reported here provides clues to the evolution of the
SLG/SRK gene pair and suggests that the SLG gene was derived by duplication of the SRK gene.
RESULTS
SLG2 Gene Produces Two Classes of Transcripts
The analysis of SLG genes from several class I haplotypes has
shown that SLG is expressed specifically in cells of the pistil
and anther of Brassica (Nasrallah et al., 1988; Sato et al., 1991;
Goring et al., 1992a, 1992b) and produces a transcript that
is ~1.6 kb in length. In the stigmatic papillar cells, this transcript was shown to encode a secreted glycoprotein that
Br To
kb
Br To
kb
3.0>>
3.0>-
1.8 >
1.6 »*
A
B
Figure 1. RNA Gel Blot Analysis of SLG Homologous Transcripts in
an.S2 Homozygous Haplotype and in Transgenic Tobacco Transformed with the SLG2 Gene.
(A) Poly(A)* RNA from Brassica stigmas (lane Br) and transgenic
tobacco pistils (lane To).
(B) Poly(A)+ RNA from Brassica anthers (lane Br) and transgenic
tobacco anthers (lane To).
The RNA blots were hybridized with an SLG2 cDNA probe. The mo-
lecular length of the hybridizing bands is indicated in kilobases at left.
accumulates in the cell wall (Kandasamy et al., 1989). The analysis of SLG2, the SLG gene derived from the S2 haplotype, had
also identified an open reading frame that could potentially
encode a secreted glycoprotein (Chen and Nasrallah, 1990).
However, in the course of further characterizing the S2 haplotype, we noted an unexpected complexity on RNA gel blot
analysis. Figure 1A (lane Br) shows that an SLG2 cDNA probe
identifies three classes of transcripts in stigma RNA: 3.0-kb
transcripts corresponding to SRK mRNA and two smaller transcripts of ~1.6 and 1.8 kb. The occurrence of the latter two
transcripts suggested that either the SLG2 or the SRK2 gene
may encode alternative transcript forms.
To distinguish between these two possibilities, we examined
the transcript profiles in tobacco plants transformed with the
SLG2 gene. Tobacco provides a convenient heterologous system for the analysis of SLG genes: tobacco DNA and RNA do
not hybridize to Brassica SLG probes under standard hybridization conditions, and earlier work has shown that SLG transgenes are expressed in pistils and produce the expected
transcript in transgenic tobacco (Moore and Nasrallah, 1990).
A Ti vector carrying the SLG2 gene was introduced into
tobacco by Agrobacterium-mediated transformation. Several
kanamycin-resistant plants were generated and analyzed for
expression of the transgene. Figure 1A (lane To) shows that
the pistils of transgenic plants produced two SLG2 transcripts
A Membrane-AnchoredForm of SLG
659
translation initiation codon of the S region in each gene into
the second exon of the SRK2 gene. The alignment of the sequences 3' of the S domain is shown in Figure 2, with the
nucleotides numbered from the ATG initiation codon in each
gene. The sequence similarity was lost 11 bp before the donor splice site of the second intron of SRK2. Sequences
immediately downstream of the TAG codon (at positions 1312
to 1314) previously reported to function as a stop codon in
SLG2 (Chen and Nasrallah, 1990) exhibited significant sequence similarity to the first intron of the SRK2 gene. These
3'sequences were not entirely colinear, however. In SRK,, the
first S domain-encoding exon is separated from the second
transmembrane-encoding exon by a 2.2-kb intron, whereas
the corresponding region in SLG2 is only 833 bp in length.
This differenceis largely due to the fact that a 1.4-kb segment
corresponding to the region from nucleotides 1363 to 2785 in
the first intron of SRK2 is not represented in SLG2. In addition,
severa1smaller deletions/insertions were noted in this region.
which exhibited the same apparent sizes as the 1.6- and 1.8-kb
Brassica transcripts. Based on these results, we concluded
that the two transcripts are derived from the SLG2 gene and
are not truncated SRK2 transcripts.
Figure 1B shows that the 1.6- and 1.8-kb SLGrderived transcripts were also produced in anthers of Brassica (lane Br)
and transgenic tobacco (lane To). However, the relative steady
state levels of the two transcripts differed in stigmas and anthers. In Brassica, the two transcripts accumulated to
equivalent levels in stigmas, while the 1.6-kb transcript was
found at higher levels than the 1.8-kb transcript in anthers. A
similar difference in transcript leve1was observed in pistils and
anthers of transgenic tobacco (compare lanes To in Figures
1A and 16). The additionaltranscripts detected in tobacco anthers but not in Brassica anthers are attributed to aberrant
transcription and/or processing of the SLG2transcripts in this
tissue.
SLG2 Gene Shares Unexpected Nucleotide Sequence
Similarity with the SRK2 Gene 3' of the S Domain
Two SLG2 Transcripts Differ at Their 3' Ends
When the SLG2 genomic sequence was compared to that of
SRK2, the region of sequence similarity between the two
genes was found to extend from ~ 3 0 bp
0 upstream of the
A
G T G T G T T T T G G A C C G G A G A G C T C G T T G A ~ T C A G ~ T T T G C C G T C G G T G G T C M G A T1219
C
The sequence similarities between SLG2and SRK2 3'of their
S domains suggested that the 833-bp region extending from
A
c a i g g g ~ t q t - a t a c g t g a a c t c g a c c a a a t a t t t t t t a c t c g a c c a a a t 1913
~
1339
1351
tttt
1399
2833
aagacatt
: :t: : : ::
----
at::--:::t:::a
1573
gaa 2999
TTTTTCCTCGGA
:A:::::::::G
-------------aagtagatccaaaa
T G T T T M G T A C T T G ~ G A M T T T T G C G M G ~ C M C T T T T T A T T A T T ~ M G 2303
AG
atagtagatataaaatcg-
G A T T T G T T A T T A C T A G M T G A T A C ~ T G ~ G A G ~ T G M T G T ~ T T T A G M C C t2443
cca
aagtacaaaatattttaatattttaatctaaatctaaaaa~~tctacttaattaaacatc
2503
1:a:a:::t:--
Figure 2. Nucleotide Sequence Comparison of the SLGZ and SRK2 Genes 3' of the S Domain.
(A) These rows represent SLGz.
. . ..
(6) These rows represent SRK2.
Colons represent nucleotide identities and dashes represent gaps,introduced to optimize nucleotide sequence alignment. Exon sequences are
shown by uppercase letters and intron sequences by lowercase letters. The 3' untranslated sequence of the 1.6-kb SLGz mRNA species is underlined. Boxes indicate the stop codon in each sequence. The intron-exon junctions are indicated by arrows, and the putative polyadenylation
sites are marked by lines above the sequence. Regions of the SRKz intron sequences that do not exhibit sequence similarity to the SLG2 sequences are omitted, and the numbers of the omitted nucleotides are indicated. The SRK2 sequence after nucleotide 3703 (marked by [>]) does
not exhibit sequence similarity to the SLG2 sequence and is not shown. Nucleotides are numbered from the ATG initiation codon. Ninety nucleotides of the first exon are shown.
660
The Plant Cell
-
1.6 kb mRNA
tm
1.8 kb mRNA
tm
3.0 kb mRNA
mbp
Figure 3. Schematic Representationof SLGz Alternative Transcripts and of SLG2 and SRK2 Gene Structures.
Partia1 restriction maps of the two genes with their flanking regions are shown. 8, BamHI; H, Hindlll; X, Xhol; E, EcoRI; P, Pstl; Xb, Xbal. The
regions of sequence similarity shared by the two genes are indicated by boxes: the stippled boxes correspond to the protein coding regions,
and the white boxes delineate the regions of shared sequence located 5' of the first exon and in the first intron. The mRNAs encoded by the
genes are shown as filled black boxes with the putative signal sequences indicated by striped boxes, and the transmembrane dornains are indicated by the abbreviation "tm." The regions correspondingto the probes specific for the 1.6-kb SLGzmRNA (probe a) and the 1.8-kbSLG2 mRNA
(probe b) are indicated by bold lines above the SLGz gene.
nucleotides 1310 to 2142 in SLG2 (shown in lowercase letters
in Figure 2) corresponds to an intron and that the two SLG2
transcripts identified on RNA gel blots may differ at their 3'
ends. The rapid'amplificationof cDNA ends method was used
to identify the 3' limits of the SLG2 transcriptional unit. Amplification of reverse-transcribed stigma poly(A)+ RNA was
performed by the polymerase chain reaction with a 21-bp oligonucleotide primer (5' amplimer) located 742 bp downstream
of the translation initiation codon of SLG2 and a 3' oligo(dT)
primer. The amplified products were cloned into the pCR1000
vector (Invitrogen), and six positive clones were isolated by
hybridization with an SLG2 cDNA probe. Among these positive clones, two different classes of polymerase chain reaction
products were identified. One clone contained a 700-bp insert,
whereas five clones contained 900-bp inserts.
DNA sequence analysis of the cDNAs showed that the 700bp insert extended from the 5' amplimer to nucleotide 1451
of SLG2. The sequence of this cDNA was colinear with the
genomic SLG2 sequence over this region. In contrast, the sequence of the 900-bp cDNA inserts was colinear with the
SLG2 genomic sequence for only 568 bp (nucleotides 742 to
1309 in Figure 2). This 568-bp sequence was joined in the
cDNA to a 297-bp sequence identical to the region located 833
bp downstream of the S region and extending from nucleotides 2143 to 2439. These data demonstrated that the SLG2
gene consists of two exons separated by an 833-bp intron. Figure 3 diagrams the structures of the SLG2 and SRK2 genes,
the regions of shared sequence similarity, and the transcripts
they encode.
Based on the results of the cDNA analysis, the 1.6- and 1.8kb SLGPtranscripts are predicted to terminate at nucleotides
1451 and 2439, respectively. To confirm this prediction, we
hybridized pistil RNA from tobacco transformed with SLG2 to
probes expected to be specific for each transcript (probes a
and b in Figure 3). Figure 4 shows that, of the two transccipts
detected by the SLG2 cDNA (lane l), the 1.6-kb transcript hybridized specifically with a probe from a genomic clone (probe
a) extending from nucleotides 1275 to 1864 (lane 2), while the
1.8-kb transcript hybridized specifically with a probe from a
cDNA clone (probe b) containing nucleotides 1275 to 1309 in
the first exon and from nucleotides 2143 to 2273 in the second
exon (lane 3).
Because these results were unexpected, we rescreened a
hgtlO cDNA library constructed from stigma RNA of plants
homozygous for the se haplotype. Fifty SLGs cDNA clones
were isolated, and the ends of the cDNA inserts in these clones
were sequenced in an attempt to find alternative transcripts
of the SLG6 gene: The cDNA clones differed in the length of
their poly(A) tails. However, all50 cDNAs terminated at approximately the same position, which was 120 bp downstream of
the stop codon in the SLG6 gene, consistent with previous observations that this gene lacks an intron and encodes a
secreted glycoprotein product.
SLG2 Gene Sequence Predicts Secreted and
MembraneAnchored Protein Products
The two SLG2 transcripts are predicted to encode different
polypeptides. The 1.6-kb transcript includes the TAG stop codon
at positions 1312 to 1314 and is predicted to encode a primary
translational product of 437 amino acids. On the other hand,
the 1.8-kb transcript represents a spliced transcript that terminates at nucleotide2439. The TAG codon at positions 1312
A Membrane-Anchored Form of SLG
to 1314 is excluded from this transcript; as a result, a 494-amino
acid open reading frame that includes the S domain and 58
additional amino acids is generated.
Hydrophilicity analysis (Hopp and Woods, 1981) was performed on the sequences of the two predicted SLG2
polypeptides. Figure 5 shows that a hydrophobic region of 31
amino acids is present at the N terminus of both polypeptides.
A corresponding region in other SLG genes was previously
shown to function as a cleavable signal for targeting the protein to the endoplasmic reticulum and secretion. An additional
hydrophobic region of 22 amino acids is present near the C
terminus of the polypeptide predicted from the 1.8-kb transcript.
This hydrophobic sequence would stop the translocation of
the polypeptide and form a transmembrane helix that would
anchor the polypeptide to the plasma membrane. The mature
protein encoded by the 1.8-kb transcript would thus represent
a membrane-anchored form of SLG, whereas that encoded
by the 1.6-kb transcript represents secreted SLG.
A comparison of the predicted amino acid sequences at the
C terminus of the secreted and membrane-anchored forms
of SLG2 and the corresponding region in SRK2 is shown in
Figure 6. The region of sequence identity between the two
SLG2 polypeptides, which extends over the entire S domain
kb
Figure 4. Poly(A)* RNA Gel Blot Analysis of Pistils Obtained from
Tobacco Transformed with the S/.G2 Gene.
The same filter was probed sequentially as follows: lane 1, hybridization with the SLG2 cDNA probe; lane 2, hybridization with a probe
specific for the 1.6-kb mRNA (probe a in Figure 3); lane 3, hybridization with a probe specific for the 1.8-kb mRNA (probe b in Figure 3).
The lengths of the transcripts are indicated in kilobases at the left.
661
from residues 1 to 436, ends at the C terminus of the secreted
form one amino acid before the stop codon that generates this
form. In the membrane-anchored polypeptide, this region is
followed by 58 residues (Asp-437 to Val-494) encoded by the
second exon and not found in secreted polypeptide. Of these
58 residues, 22 amino acids (lle-450 to Trp-471) constitute the
hydrophobic region that corresponds to the transmembrane
domain of SRK2. Only four of these 22 residues diverge between SLG2 and SRK2. The remaining 23 amino acids are
predicted to form a short intracellular domain in the membraneanchored SLG2. The region of sequence similarity between
SLG2 and SRK2 extends into this cytoplasmic domain over
nine residues, only one of which diverges between the two
sequences. This nine-amino acid sequence, which in SLG2
contains five basic residues, is typical of sequences that
immediately follow the transmembrane domain in membranespanning proteins (von Heijne and Gavel, 1988). This juxtamembrane sequence is followed by 14 amino acids that are
unique to the membrane-anchored form of SLG2.
Immunodetection of Secreted and Membrane-Anchored
Forms of SLG2 in Stigmas
Protein immunoblot analysis was performed to determine if
the SLG2 gene does in fact produce soluble and membraneanchored protein forms. For this analysis, we used the polyclonal rabbit antiserum PR2, which identifies stigma antigens
corresponding to SLG glycoforms from class I and class II
haplotypes as well as the glycoprotein products of SLR genes
(T. Tantikanjana, M. E. Nasrallah, and J. B. Nasrallah, manuscript in preparation). Proteins were extracted from transgenic
and control tobacco pistils and from Brassica stigmas either
in a Tris-HCI buffer or in SDS-PAGE sample buffer. Figure 7
shows the immunoblot patterns obtained following SDS-PAGE
of the protein extracts. The soluble SLG2 proteins are identified as a protein doublet of ~57 and 58 kD evident in transgenic
tobacco pistils (Figure 7, lanes T) and in Brassica S2 stigmas
(lanes 2), but which are not in control untransformed tobacco
pistils (lanes C). Extraction of transgenic tobacco pistils or S2
stigmas in SDS-PAGE sample buffer solubilized an additional
SLG2 protein of ~62 kD (asterisks). Such a higher molecular
mass form is predicted from the 1.8-kb SLG2 transcript. The
low level of this 62-kD protein detected in Tris-HCI extracts is
ascribed to the presence of contaminating membrane vesicles. In contrast to these results, only soluble SLG proteins
were detected in stigmas from plants homozygous for the two
class I haplotypes analyzed, S6 (lanes 6) and S,3 (lanes 13+).
No higher molecular mass immunoreactive proteins were solubilized by SDS-PAGE sample buffer from the stigmas of these
class I haplotypes, which is consistent with predictions based
on the cDNA analysis reported earlier. It should be noted that
approximately equal amounts of soluble SLG protein from TrisHCI and SDS-PAGE sample buffer extracts from each tissue
sample were analyzed. However, the 50- and 51-kD products
of the SLR1 gene were more abundant in the SDS sample
buffer extracts than in the Tris-HCI extracts. This difference
662
The Plant Cell
A
4
3
2
1
O
-1
-2
-3
-4
I
1
Y
51
101
151
201
#
,
,
251
301
35 1
401
451
,
,
4
3
2
1
O
-1
-2
-3
-4
,Y,
-0.45
.
,
,
,
Figure 5. Hydrophilicity of the SLG2 Gene Products.
(A) Hydrophilicity profile of the secreted SLGp form predicted from the 1.6-kb transcript.
(B) Hydrophilicity profile of the membrane-bound SLG2form predicted from the 1.8-kb transcript.
The hydrophobic regions are marked by bars.
is attributed to the differential solubility of this protein in the
two buffers.
DISCUSSION
We have shown that the SLG2 gene contains an intron and
produces two transcripts that differ at their 3’ends. One
transcript contains sequences from the first exon and terminates within the first intron. This transcript, like the transcripts
previously reported for other alleles of SLG, is predicted to encode a secreted glycoprotein. The second and larger transcript
contains sequences from the first and second exons and is
predicted to encode a membrane-anchoredform of SLG. This
form consists of an extracellular domain that contains the 436
amino acids of the S domain plus an additional 13 residues,
a 22-amino acid transmembrane domain, followed by a short
A Membrane-Anchored Form of SLG
23-amino acid cytoplasmic domain. This membrane-anchored
form is expected to exceed the molecular mass of the secreted
form by ~6 kD. In fact, immunoblot analysis of native Sj
stigma proteins and of pistil proteins from transgenic tobacco
plants that express the SLG2 gene has revealed the presence
of soluble immunoreactive proteins of 57 and 58 kD and of
an immunoreactive protein of 62 kD that is extracted at appreciable levels only in detergent-containing buffer.
The features we have described for the SLG2 are unusual.
Previous analyses of several class I SLG genomic sequences
in two Brassica species (Nasrallah et al., 1988; Dwyer et al.,
1991) had indicated that these genes are intronless. Only one
transcript form is produced by these genes, as has been determined from the analysis of endogenous SLG mRNA in
Brassica tissues (Nasrallah et al., 1985; Trick and Flavell, 1989;
Goring et al., 1992a, 1992b) and of SLG-encoded transcripts
in transgenic plants (Moore and Nasrallah, 1990). These observations were verified in a thorough analysis of SLG6 cDNAs
performed in this study: we found no evidence for alternative
SLG transcripts among 50 cDNAs screened. In addition, the
extraction of stigmas with detergent-containing buffer failed
to reveal the occurrence of a membrane-bound form of SLG
in plants homozygous for class I haplotypes such as SB and
Sra. It remains possible, however, that the SLG genes from
class II haplotypes, such as those corresponding to the classic S5 and Sr5 alleles (Thompson and Taylor, 1966), which are
similar to S2 in their genetic behavior, are structurally similar
to SLG2. Analysis of these haplotypes should determine if the
production of a membrane-anchored SLG is a general feature
of the leaky and pollen-recessive class II haplotypes.
The two SLG2 transcripts are expressed in stigmas and anthers. In stigmas, they occur in approximately equal levels,
whereas in anthers, the smaller secreted glycoprotein-encoding
transcript is more abundant and may be more stable than the
SLG2 s
SLG2 m
MSCSGDGFLRLNNMNLPDTKTATVDRIIDVKKCEERCLSDCNCTSFAIAD
===============*=========»=================*»*=*==
399
399
SRK2
==========================TM===«=================
401
SLG2 s
VRNGELGCVFWTGELVEIRKFAVGGQDLYVRLNAADLG
437
SLG2 m
=====================================DFSSDEKRHRTGK
449
SRK2
====G===========A=====================I==G===D====
451
SLG2 m
IIGWSIGVSVMLILSVLVLCFMKRROKOAKAYAVGMKMFVEMEKV
SRK2
=======S========ILF===R========DATPIVGNQVLMNEWLPR.....
S domain
M
TH domain
kinase
494
domains
Figure 6. Sequence Alignment of the C Termini of the SLG? and SRK2
Gene Products.
Comparison of the predicted amino acid sequences at the C terminus
of the secreted (s) and the membrane-anchored (m) forms of SLG2
and of the S domain of the SRK2 protein. The 3' end of the S domain
is indicated by the arrowhead, and the transmembrane (TM) domain
is underlined. Only the beginning of the kinase domain of SRK2 is
shown. The numbers correspond to the amino acid residues in the
predicted amino acid sequences.
Tobacco
C
T
663
_____________
Brassica
2
2 * 6 1 3 *
a b a D a b a b a b a b
kD
97-
66-
43-
Figure 7. Immunoblot Analysis of Protein Extracts from Brassica
Stigmas and from Transgenic or Control Tobacco Pistils.
Stigmas or pistils were extracted in Tris-HCI buffer (lanes a) and SDSPAGE sample buffer (lanes b), and the amount of protein loaded was
adjusted to give equivalent staining of the soluble SLG bands. Total
protein (10 to 20 ng) for the Brassica sample or total protein (20 to 35
ng) for the tobacco sample was analyzed by 10% SDS-PAGE, transferred to nitrocellulose, and analyzed with the PR2 antiserum. C, control
untransformed tobacco pistils; T, transgenic tobacco pistils; 2, stigmas
from Brassica SjS? homozygotes; 6, stigmas from a Brassica SeS6
homozygote; 13+, stigmas from a Brassica Sr3S)3 homozygote. Lanes
2 and 6 represent sibling plants from one F2 population, and lanes
2+ and 13* represent sibling plants from another F2 population (see
Methods). The soluble and polymorphic SLG bands detected in TrisHCI and SDS buffer extracts are indicated by dots; the 62-kD SLG2
form detected predominantly in SDS buffer extracts of transgenic
tobacco pistils and Brassica S2 stigmas is indicated by asterisks. The
SLR1 protein doublet detected in all Brassica stigma extracts is indicated by open circles. Molecular mass standards in kilodaltons are
shown at the left.
larger transcript in this organ. The SLG promoter has been
shown to be active in the papillar cells of the stigma, in the
anther tapetum, and in microspores (Sato et aJ., 1991). At this
time, it is not known whether the two SLG2 transcripts are expressed both in tapetum and microspores or if the two
transcripts are translated in anther tissues. Nevertheless, at
least three classes of S locus-encoded molecules are predicted
to be displayed at the papillar cell-pollen interface in the S2
haplotype. As diagrammed in Figure 8, this interface would
exhibit a transmembrane SRK receptor protein kinase, a
secreted SLG glycoprotein that would accumulate in the papillar
cell wall (Kandasamy et al., 1989), and a membrane-anchored
SLG glycoprotein. In contrast, in class I haplotypes, such as
S6, the papillar cell-pollen interface would contain only the
first two classes of proteins (Figure 8).
The unique properties of the SLG2 gene described in this
paper provide a possible molecular explanation for the differences in the activities and genetic properties of S haplotypes.
The leakiness and/or recessive nature of the S2 haplotype
may be associated with the production of an anomalous
664
The Plant Cell
s2
Figure 8. Schematic Representation of Secreted and MembraneAnchored Proteins in the S2 and Haplotypes.
The mSLG2represents the 62-kD membrane-anchoredprotein found
in the S, haplotype but not in the s6 haplotype. The S domains of the
SLG-derived gene products are shown as stippled boxes, and the S
domains of the SRK-derivedgene products are shown as white boxes.
The transmembrane domains are indicatedby striped boxes, and the
black boxes represent the kinase domains of the SRK-derived gene
products.
membrane-bound form of SLG. To evaluate the possible effect of membrane-anchored SLG on papillar cell-pollen
signaling, it is useful to search for analogous situations in animal systems. To our knowledge, naturally occurringtruncated
forms of receptor protein kinases that lack the kinase catalytic domain and that are expressed in the same cell types
as the wild-type receptor have not been reported in other
systems. Such truncated receptors have, however, been engineered in vitro, reintroducedinto appropriatecells, and shown
to interfere with the function of the wild-type receptor (Amaya
et al., 1991; Hemmati-Brivanlouand Melton, 1992; Redemann
et al., 1992). In these studies, signaling through the receptor
was inhibited completely by a 200-fold excess of the truncated
receptor and only partially by lower levels of the mutant receptor. The activation of receptor protein kinases is thought to
depend on ligand-inducedreceptor dimerization mediated in
large part by the extracellular domain of the receptor (reviewed
in Ullrich and Schlessinger, 1990; Massague, 1992). The truncated receptorsthat lack the kinase domain apparently interact
with the wild-type receptor through their extracellular domain,
thus creating unproductivereceptor heterodimers and suppressing signal transduction. Because SLG and the extracellular
domain of SRKfromthe same S haplotype, although not identical, share a high degree of sequence identity, the SLG gene
may be viewed as corresponding to the extracellular domain
of SRK. In the Sz haplotype, the membrane-anchoredform of
SLG would simulate a truncated kinase-negativeform of SRK
and interferewith receptor function. Because SLGptranscripts
are only modestly more abundant than SRKz transcripts in Sz
tissues, only partial interference-and therefore a leaky
self-incompatibility phenotype-would be achieved. Such a
partially active receptor system could also account for the
pollen-recessive nature of the Sz haplotype in heterozygotes,
if competitive interactions between the proteins encoded by
different S haplotypes is assumed to occur in pollen-stigma
recognition.
In addition to suggesting a molecular basis for the genetic
properties of the Sz haplotype, the structure of the SLGp gene
has yielded important information relating to the molecular
evolutionary history of the SLG/SRK gene pair. The high degree of sequence similarity between SLG and the S domain
of SRK noted in severa1S haplotypes points to the occurrence
of an ancestral duplication event and the operation of a
homogenizing process, perhaps involving gene conversion,
that maintains the present-day high degree of sequence identity between the two genes. The sequence similarity we
observed in the introns and transmembrane domains of SLGz
and SRKz argues that SLG is derived from SRK by duplication of the extracellular domain. Based on the data from the
Sz haplotype, the duplication event may have been partial and
involved only the first two exons of SRK. Alternatively, a duplication of the entire SRK gene may have been followed by the
deletion of the exons that encode the kinase catalytic domain.
The subsequent loss of the second exon would have generated the SLG gene structure found in S haplotypes like S,. It
is also possible that the duplication event that generated the
SLG/SRKgene pair may have occurred more than once in the
evolutionary history of the S locus.
In any event, the work reported in this paper illustratesthat
it is possible, through the analysis of the intrinsic variability
of S haplotypes, to uncover molecular footprints in the evolutionary history of the Brassica S locus and to obtain clues to
the processes of gene duplication and diversificationas they
relate to function and evolution of a sporophytic genetic selfincompatibility system.
METHODS
Plant Material
The Brassica oleraceavar a/bog/abra(Chinese kale) inbredlhe homozygous for Sa self-incompatibility haplotype and the 6.oleracea var
acephala (kale) s6 and ST3homozygotes were derived from plant material initiallyobtained from the Gene Bank Facility at Wellesbourne,
U.K., courtesy of D. J. Ockendon. The B. oleracea plants homozygous
for the S,, Ss,
and S I 3haplotypes used in the protein analysis were
Fpplants derived from S& x s,& and S& x SI& families.
lsolation of Genomlc and cDNA Clones
The isolation of genomic clones containingthe S Locus Glycoprotein
(SLG2) and S Receptor Kinase (SRK,) genes from an S2genomic library in the vector XGEMll (Promega)was previouslydescribed (Chen
A Membrane-Anchored Form of S l G
and Nasrallah, 1990; Stein et al., 1991). The construction of cDNA
libraries from stigma poly(A)+ RNA in the vector XgtlO was described
previously (Nasrallah et al., 1987).
Plant Ransformation
A DNA fragment containing the SLGPgene was inserted into the Ti
vector pBIN19 (Bevan, 1984).The recombinantplasmid was introduced
into the &mbacterium tumefaciens pCIB54ZA136, a strain derived from
pEHAlOl (Hood et al., 1986), and the resulting bacterial strains were
used to transform leaf discs of Nicotiana tabacum cv Petit Havana as
described previously(Horsch et al., 1988). Kanamycin-resistant plantlets
resulting from independenttransformationevents were propagatedand
analyzed. The presence of inserted DNA sequences in the putative
transgenic plants was verified by DNA blot analysis.
RNA Gel Blot Analysis
For the analysis of endogenous Brassicatranscripts,stigmas from flower
buds at 1 day before anthesis and anthers at the binucleate microspore stage were harvested. For the analysis of transgene-encoded
transcripts, pistils and anthers of transgenic tobacco plants were harvested from 45-mm flower buds. Poly(A)+RNA was isolated by using
the FastTrack mRNA isolation kit (Invitrogen, San Diego, CA). The RNA
was subjected to electrophoresis on 1% (whr) agarose gels in the presente of formaldehyde, transferred to GeneScreen membrane (Du Fbnt),
hybridized, and processed as described previously(Stein et al., 1991).
3' End Amplification and Cioning of SLGz cDNAs
Poly(A)+ RNA isolated from S, stigmas was reverse transcribed and
amplified with the mRNA GeneAmp kit (Perkin-Elmer-Cetus). The
amplification was primed with a 5' 21-bp oligonucleotide primer (5'AATTATACGGAGAACAGTGAG3)and an oligo(dT) 3'primer. The resulting amplified cDNA was cloned into the vector pCRlOOO using the TA
cloning kit (Invitrogen).
DNA Sequence Analysis
Nucleotide sequence analysis of the amplified cDNA inserts and of
the SLGa and SRK, clones was performed by double-stranded sequencing (De1Sal et al., 1989)of a series of nested deletions generated
with the Erase-a-Base kit (Promega).
Proteln lmmunoblot Analysis
Brassica stigmas and transgenic tobacco pistils were collected from
open flowers and frozen in liquid NP.The tissue was extracted either
in 10 mM Tris-HCI, pH 7.2, buffer as described previously (Nasrallah
et al., 1985) or in SDS-PAGE sample buffer consisting of 80 mM TrisHCI, pH 6.8, 1% (whr) SDS, 15% (whr) dithiothreitol, and 10Y0 (v/v)
glycerol. For each tissue sample, the amount of total protein loaded
per lane was adjusted to produce equivalent amounts of soluble SLG
protein in the Tris-HCI and SDS sample buffer extracts. The proteins
were separated on 10% (whr) SDS-polyacrylamide gels and transferred
to nitrocellulose membranesby electroblotting as described previously
665
(Umbach et al., 1990). The membraneswere probed with the PR2antiserum, an antiserum raised to a trpE-SLR2 fusion protein (T.
Tantikanjana, M. E. Nasrallah, and J. B. Nasrallah, manuscript in prep
aration). The protein blots were then treated with goat anti-rabbit IgG
conjugated to alkaline phosphatase(Promega), and the immunoconjugates were detected according to the manufacturer's instructions.
ACKNOWLEDGMENTS
We wish to thank Bruce Howlett and Barr Ticknorfor their help in preparing Figures 2 and 5A. This work was supported by a grant from the
U.S. Department of Energy. lT.was supported in part by a Fulbright
predoctoralfellowship. J.C.S. was the recipient of a predoctoral fellowship from the Cornell National Science Foundation Plant Science
Center, a unit in the U.S. Department of Agriculture-Department of
Energy-NationalScience Foundation Plant Science Centers Program.
Received March 30, 1993; accepted April 27, 1993.
REFERENCES
Amaya, E., Musci, T.J., and Kirschner, M.W. (1991). Expression of
a dominant negative mutant of the FGF receptor disrupts mesoderm
formation in Xenopus embryos. Cell 66, 257-270.
Bateman, A.J. (1955). Self-incompatibility systems in angiosperms.
111. Cruciferae. Heredity 9, 52-68.
Bevan, M. (1984). Binary Agmbacterium vectors for plant transformation. Nucl. Acids Res. 12, 8711-8721.
Boyes, D.C., and Nasrallah, J.B. (1993). Physical linkage of the SLG
and SRK genes at the self-incompatibility locus of Brassica oleracea. MOI. Gen. Genet. 236, 369-373.
Chen, C.H., and Nasrallah, J.B. (1990). A new class of S sequences
defined by a pollen recessive self-incompatibility allele of Brassica
oleracea. MOI. Gen. Genet. 222, 241-248.
De1 Sal, G., Manfioletti, G., and Schneider, C. (1989). The CTABDNA precipitation method: A common mini-scale preparation of template DNA from phagemids, phages or plasmids suitable for
sequencing. BioTechniques 7, 514-519.
Dwyer, K.G., Balent, MA., Nasrallah, J.B., and Nasrallah, M.E.
(1991). DNA sequences of self-incompatibility genes from Brassica
campestris and 6.oleracea: Polymorphism predating speciation.
Plant MOI. Biol. 16, 481-486.
Gaude, T., Frlry, A., Heizmann, P., Marlac, C., Rougler, M., Fobis,
I., and Dumas, C. (1993). Expression of a self-incompatibility gene
in a self-compatible line of Brassica oleracea. Plant Cell 5, 75-86.
Goring, D.R., and Rothstein, S.J. (1992). The S-locus receptor kinase gene in a self-imcompatible Brassica napus line encodes a
functional serindthreonine kinase. Plant Cell 4, 1273-1281.
Goring, DA., Banks, P., Beversdorf, W.D., and Rothstein, S.J.
(1992a). Use of the polymerase chain reaction to isolate an Slocus
glycoprotein cDNA introgressed from B. campestris into 6.napus
ssp. oleifera. MOI. Gen. Genet. 234, 185-192.
666
The Plant Cell
Goring, D.R., Banks, P., Fallls, L., Baszczynski, C.L., Beversdorf,
W.D., and Rothstein, S.J. (1992b). ldentificationof an S-locus glycoprotein allele introgressedfrom 6. napus ssp. rapifera to 6.napus
ssp. oleifera. Plant J. 2, 983-989.
Hemmati-Brivanlou, A., and Melton, D.A. (1992). A truncated activin receptor inhibits mesoderm induction and formation of axial
structures in Xenopus embryos. Nature 359, 609-614.
Hood, E.E., Helmer, G.L.,Fraley, R.T., and Chilton, M.D. (1986). The
hypervirulence of Agmbacterium tumefaciens A281 is encoded in
a region of pTiBo542 outside of T-DNA. J. Bacteriol. 168,1291-1301.
Hopp, T.P., and Woods, K.R. (1981). Prediction of protein antigenic
determinants from amino acid sequences. Proc. Natl. Acad. Sci.
USA 78, 3824-3828.
Horsch, R.B., Fry, J., Hoffmann, N., Neidermeyer, J., Rogers, S.G.,
and Fraley, R.T. (1988). Leaf disc transformation. In Plant Molecular Biology Manual, S.B. Gelvin and R.A. Schilperoort, eds
(Dordrecht: Kluwer Academic Publishers), pp. A5:l-9.
Kandasamy, M.K., Paolillo, D.J., Faraday, C.D., Nasrallah, J.B., and
Nasrallah, M.E. (1989). The S-locus specific proteins of Brassica
accumulate in the cell wall of developing stigmapapillae. Dev. Biol.
134, 462-472.
Massague, J. (1992). Receptors for the TGFP family. Cell69,1067-1070.
Moore, H.M., and Nasrallah, J.B. (1990). A Brassicaself-incompatibility
gene is expressed in the stylar transmitting tissue of transgenic
tobacco. Plant Cell 2, 29-38.
Nasrallah, J.B., and Nasrallah, M.E. (1993). Pollen-stigma signaling in the sporophytic self-incompatibilityresponseof Brassica. Plant
Cell 5, in press.
Nasrallah, J.B., Kao, T.H., Goldberg, M.L., and Nasrallah, M.E.
(1985). A cDNAclone encoding an S-locus specific glycoprotein from
Brassica oleracea. Nature 318, 263-267.
Nasrallah, J.B.,
Kao, T.H.,
Chen, C.H., Goldberg, M.L., and
Nasrallah, M.E. (1987). Amino-acid sequence of glycoproteins en-
coded by three alleles of the S locus of Brassica oleracea. Nature
326, 617-619.
ablation of floral tissues. Annu. Rev. Plant Physiol. Plant MOI. Biol.
42, 393-422.
Nasrallah, M.E., Kandasamy, M.K., and Nasrallah, J.B. (1992). A
genetically defined trans-acting locus regulates S-locus function in
Brassica. Plant J. 2, 497-506.
Ockendon, D.J. (1974). Distributionof self-incompatibility alleles and
breeding structure of open-pollinated cultivars of brussels sprouts.
Heredity 33, 159-171.
Redemann, N., Holzmann, B., von Ruden, T., Wagner, E.F.,
Schlessinger, J., and Ullrich, A. (1992). Anti-oncogenic activity of
signalling-defective epidermal growth factor receptor mutants. MOI.
Cell Biol. 12, 491-498.
Sato, T., Thorsness, M.K., Kandasamy, M.K., Nishio, T., Hirai, M.,
Nasrallah, J.B., and Nasrallah, M.E. (1991). Activity of an S locus
gene promoter in pistils and anthers of transgenic Brassica. Plant
Cell 3, 867-876.
Stein, J.C., and Nasrallah, J.B. (1993). A plant receptor-likegene, the
S-locus receptor kinase of Brassica oleracea L., encodes a functional serinehhreonine kinase. Plant Physiol. 101, 1103-1106.
Stein, J.C., Howlett, B., Boyes, D.C., Nasrallah, M.E., and Nasrallah,
J.B. (1991). Molecular cloning of a putative receptor protein kinase
gene encoded at the self-incompatibility locus of Brassica oleracea.
Proc. Natl. Acad. Sci. USA. 88, 8816-8820.
Thompson, K.F., and Taylor, J.P. (1966). Non-linear dominance relationships between S alleles. Heredity 21, 345-362.
Toriyama, K., Stein, JC, Nasrallah, M.E., and Nasrallah, J.B. (1991).
Transformation of Brassica oleracea with an S-locus gene from 6.
campestrischanges the self-incompatibilityphenotype. Theor. Appl.
Genet. 81, 769-776.
lkick, M., and Flavell, R.B. (1989). A homozygous S-genotype of Brassica oleracea expresses two S-like genes. MOI.Gen. Genet. 218,
112-1 17.
Ullrich, A., and Schlessinger, J. (1990). Signal transduction by recep
tors with tyrosine kinase activity. Cell 61, 203-212.
Nasrallah, J.B., Yu, S.M., and Nasrallah, M.E. (1988). Selfincompatibility genes of Brassica oleracea: Expression, isolation
and structure. Proc. Natl. Acad. Sci. USA 85, 5551-5555.
Umbach, A.L., Lalonde, B.A., Kandasamy, M.K., Nasrallah, J.B.,
and Nasrallah, M.E. (1990). lmmunodetectionof protein glycoforms
encoded by two independent genes of the self-incompatibility multigene family of Brassica. Plant Physiol. 93, 739-747.
Nasrallah, J.B., Nishio, T., and Nasrallah, M.E. (1991). The selfincompatibility genes of Brassica: Expression and use in genetic
von Heijne, G., and Gavel, Y. (1988). Topogenic signals in integral
membrane proteins. Eur. J. Biochem. 174, 671-678.
An alternative transcript of the S locus glycoprotein gene in a class II pollen-recessive
self-incompatibility haplotype of Brassica oleracea encodes a membrane-anchored protein.
T Tantikanjana, M E Nasrallah, J C Stein, C H Chen and J B Nasrallah
Plant Cell 1993;5;657-666
DOI 10.1105/tpc.5.6.657
This information is current as of June 15, 2017
Permissions
https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532
298X
eTOCs
Sign up for eTOCs at:
http://www.plantcell.org/cgi/alerts/ctmain
CiteTrack Alerts
Sign up for CiteTrack Alerts at:
http://www.plantcell.org/cgi/alerts/ctmain
Subscription Information
Subscription Information for The Plant Cell and Plant Physiology is available at:
http://www.aspb.org/publications/subscriptions.cfm
© American Society of Plant Biologists
ADVANCING THE SCIENCE OF PLANT BIOLOGY