The Plant Cell, Vol. 1, 483-491, May 1989 O 1989 American Society of Plant Physiologists

Sequence Variability of Three Alleles of the Self- lncompatibility Gene of Nicotiana alata

Marilyn A. Anderson,” Geoffrey 1. McFadden,” Robert Bernatzky,” Angela Atkinson,” Tirnothy Orpin,” Helen Dedman,” Geoffrey Tregear,b Ross Fernley,b and Adrienne E. Clarke”.’ ”Plant Cell Biology Research Centre, School of Botany, University of Melbourne, Parkville, Victoria 3052, Australia bHoward Florey lnstitute of Experimental Physiology and Medicine, University of Melbourne, Parkville, Victoria 3052, Australia

Three alleles of the self-incompatibility gene of Nicofiana alata have been cloned and sequenced. A comparison of the sequences shows a surprisingly low level of homology (56%) and the presence of defined regions of homology and variability. The homologous regions include the N-terminal sequence, most of the cysteine residues and glycosylation sites, as well as other blocks throughout the sequence. We interpret these conserved regions as “framework” and nonconserved regions as “hypervariable,” following the terminology used to describe analogous regions in the IgG supergene family. The low level of overall homology forms the basis of a general method for isolating S-allele-specific cDNAs. Allele-specific antibodies can be generated using synthetic peptides correspond- ing to one of the variable regions. When applied to sections of the pistil, these antibodies label the intercellular matrix in the stigma and transmitting tissue of the style and the cell walls in the epidermis of the placenta. Hindlll digestion of genomic DNA generates a characteristic pattern of S-gene fragments for each genotype. These restriction fragment length polymorphisms can be used to assign S-genotype to progeny arising from breeding experiments.

INTRODUCTION

The evolutionary success of flowering plants is attributed in part to the development of self-incompatibility, which acts to prevent self-fertilization and promote outcrossing (see de Nettancourt, 1977; Harris et al., 1984; Cornish, Pettitt, and Clarke, 1988, for reviews). In many plants with a genetically controlled self-incompatibility (SI) system, the specificity of the SI reaction is controlled by a single gene (S-gene), which has multiple alleles. Pollen is rejected if the same S-allele is expressed by both the pollen and the pistil. Self-incompatibility is divided into two main types: sporophytic, in which the pollen has the same S-phenotype as the parent, and gametophytic, in which pollen pheno- type is determined by the haploid genotype of the pollen. S-Allele-associated glycoproteins have been identified in the styles of a number of plants which exhibit either gametophytic or sporophytic incompatibility. cDNAs en- coding these glycoproteins have been isolated from Nico- tiana alata (gametophytic SI, Anderson et al., 1986) and Brassica oleracea (sphorophytic SI, Nasrallah et al., 1985, 1987).

In this paper, we describe the sequences of three alleles of the self-incompatibility gene of N. alata.

’ To whom correspondence should be addressed.

RESULTS

lsolation and Sequence of cDNA Clones Encoding the So- and Ss-Alleles of the SI Gene of N. alata

The first cDNA clone described for a gametophytic SI system corresponds to the S,-allele of N. alata and cross- hybridizes with RNA from styles homozygous for the S3- and S,-alleles (Clarke et al., 1987; Bernatzky, Anderson, and Clarke, 1988, Figure 1 A). This weak cross-hybridiza- tion was used as a strategy to isolate cDNAs encoding other S-alleles.

Differential screening of cDNA libraries prepared from RNA isolated from mature styles of the S3S3- or s&- genotype with cDNA prepared from SS,- or S6S6-RNA from mature styles showed that most plaques hybridized well with both probes. A few plaques in the S3S3 library (1.6%) hybridized strongly with the SS, probe and weakly with the probe, and a few plaques in the s6s6 library (1.6%) hybridized strongly with the && probe and weakly with the S3S3 probe. These plaques contained potential S3- cDNA and S,-cDNA, respectively. The S3-cDNA clones (50% of the clones obtained by differential screening) hybridized most strongly to an RNA from plants carrying an S,-allele (see Figure 1 B), but this RNA (about 1050 bases) appeared slightly larger than the related species in the S2- and S,-genotypes (900 bases). A minor hybridizing

484 The Plant Cell

A 13 33 22 66kb2-3 — ,..,*2-0 ——

0-56-

0-12——

S2-cDNAprobe

B2-3 ——2-0 ——

I 1

S3-cDNAprobe

0-12 —

2-3 ——2-0 ——

0-56——

0-12 ——

IS6-cDNA

probe

Figure 1. Gel Blot Analysis of RNA from Styles of Various S-Genotypes of N. a/afa.

Three identical gel blots were prepared containing RNA frommature styles of S,S3, 8383, 8282, and S6S6 genotypes.(A) Blot probed with 32P-labeled S2-cDNA.(B) Blot probed with S3-cDNA.(C) Blot probed with S6-cDNA.Equal amounts of RNA (5 ng) were applied to each lane andEcoRI/Hindlll restriction fragments of A DNA were used as sizemarkers. The autoradiograms were overexposed to show hybrid-ization between different alleles.

RNA species of about 650 nucleotides was also evident inthe S3-genotype and was detected with both the S3-cDNAprobe (see Figure 1 B) and the S6-cDNA probe (see Figure1C). The S3-cDNA shares 64% homology with the S2-cDNA at the amino acid level (Figure 2), and the amino-terminal sequence matches that of the purified S3-glyco-protein derived by N-terminal amino acid sequencing (Jah-nen et al., 1989). The signal sequence of the S3-glycopro-

tein is incomplete because the longest S3-clone (740 bp)terminated at an EcoRI site in the region coding for aminoacids at positions 25 to 27 (see Figure 2). The rest of thesequence was obtained by RNA sequencing (Hamlyn, Gait,and Milstein, 1981). The S6-cDNA clones bound moststrongly to an RNA species of about 900 bases fromplants bearing an S6-allele (see Figure 1 C). The predictedamino acid sequence of the longest cDNA clone (803 bp)is shown in Figure 2. Overall, the homology at the aminoacid level with the S2-glycoprotein is 62% and the predictedamino acid sequence of the S6-cDNA contains the amino-terminal sequence of the mature S6-associated glycopro-tein (Jahnen et al., 1989). Alignment of the amino acidsequences derived from the S2-, S3-, and S6-cDNA se-quences (see Figure 2) shows that there is about 56%homology overall. Although there are changes scatteredthroughout the sequence, the variation is mainly associ-ated with four hypervariable regions (see Figure 2) whichare predicted to be hydrophilic (see Figure 4A, regions A,B, C, and D). The conserved regions are associated withthe amino terminus, the cysteine residues, the glycosyla-tion sites, and the carboxy terminus. The 9 cysteine resi-dues of the mature S2-glycoprotein are all conserved in theS3- and S6-glycoproteins, but the S3- and S6-glycoproteinshave an additional cysteine residue at position 79. TheS2-, S3-, and S6-glycoproteins contain 4, 5, and 4 consen-sus sequences for A/-glycosylation, respectively. Four ofthe W-glycosylation sites are in conserved positions in thethree alleles, and none are included in the hypervariableregions apart from the fifth site, unique to the S3-glycopro-tein (see Figure 2, position 102).

|W PI I T [Fq RW P A A F d H T T P S

Iw PJ T A IF d H T

3A F E Y M Q L V LV 1

K H [Cl E R T [S T WFT I H G L W P D M H T [71 MIL Nl YpS P q K R I B N J M F T I H G L W P D M V S T M L N V <T P U K N I la S IBF T I H n I w p n f j V SUJ T ll_[J fb

A I[T]G[G|F FWT T [els R L a [TTlKpl iA I T Q G Y PWL S C T K R QME L L E IT v [TJK Lcyv p Q L s \c]i K G Q - IE Uwy \

Figure 2. Predicted Amino Acid Sequences of S2-, S3-, and S6-Glycoproteins.The amino acid sequence is numbered beginning with 1 for thefirst amino acid of the mature protein. The putative signal se-quence is indicated by negative numbers. Common sequencesare boxed. The most variable regions are underlined. The potentialN-glycosylation sites are circled. Gaps were included at positions19, 20, and 160 in the S2- and S6-sequence and 188 in the S2-sequence to maximize similarity. The amino acid sequences werepredicted from the nucleic acid sequences of the cDNA clones.

Self-Incompatibility Alleles of N. alata 485

DMA Gel Blot Analysis and Breeding Experiments

The question of whether these sequences represent allelicproducts of the same S-locus or products of distinct genesis addressed by DMA gel blot analyses and breedingexperiments. Figure 3A shows a DMA gel blot analysis ofa Hindlll digest of 8787, S6S6, 8383, and 8282 genomic DMAusing the Ss-cDNA as probe. The probe hybridized to singlefragments from the S/S?-, S6S6-, and SsSa-genomic DMAand to two fragments from S^-genomic DMA. The S3-and S6-cDNA probes revealed the same restriction frag-ments, except that each cDNA clone bound strongly tofragments from homologous DNA and weakly to fragmentsfrom heterologous DNA (see Figure 3A). The weak cross-hybridization is consistent with the variation distributedthroughout the cDNA sequences.

Breeding experiments were performed to determinewhether the restriction fragments identified by the threecDNAs segregated with S-genotype in the F,-generation.Two plants with no common S-alleles (S,S3 9 and S6S7 6")were crossed to produce 27 offspring, which were dividedinto incompatibility classes according to their ability toaccept or reject SsSa-, S6S6-, or S^-pollen. Four incom-patibility classes S,S6 (7 plants), S,S7 (4 plants), S;jS6 (8plants), and SaS? (8 plants) were obtained as predictedfrom the one locus hypothesis (de Nettancourt, 1977;Lawrence, Afzal, and Kenrick, 1978). A second populationwas produced by crossing two plants with one S-allele incommon (S,S3 $ and 8283 6); this population contained 21plants, which fell into the two predicted self-incompatibilityclasses, S,S2 (14 plants) and 8283 (7 plants). DNA gel blotanalyses of these two populations (see Figures 3B and3C) showed that, for each plant, the S-genotype assignedby breeding behavior was identical with that predicted byanalysis of the restriction fragment length polymorphisms(RFLPs). Although the patterns described are for the majorhybridizing bands, there are fainter bands that becomeapparent after longer exposure. These are indicated byarrows in Figures 3A and 3C.

Production of an Antibody Specific to theS2-Glycoprotein

An antibody was raised to a synthetic peptide correspond-ing to a stretch of the major hypervariable region B, whichis hydrophilic (see Figure 4A, Region B) and thus predictedto lie on the surface of the glycoprotein (Hopp and Woods,1981). This antibody bound specifically to the S2-glycopro-tein in protein gel blots under conditions where there wasno detectable binding to the other buffer-soluble styleproteins or to other S-allele glycoproteins (see Figure 4B).

Immunolocalization of the S2-Associated Glycoprotein

Light microscope localization of antibody binding is shownin Figure 5. When applied to longitudinal sections of the

77 66 33 22 77 66 33 22 77 66 33 22

83+

kbI 23

9-4i 6-7

I 2-3i 2-0

S2-cDNA

PROBE

SycDNA

PROBE

V3 « ¥7

Sg-cDNA

PROBE

¥737 16 17 36 16 37 36 16 37 36 36 17 3ft 16 16 37 37 37 37

12 23 12 12 12

*s*

Figure 3. DNA Gel Blot Analysis of S-Genotypes of N. alata.

(A) Comparison of the relative binding efficiency of the S2-, S3-,and S6-cDNA probes on genomic DNA from different S-genotypesof N. alata. Three filters containing Hindlll digests of S-,S7-, SeSe-,8383-, and S^-genomic DNA were probed separately with (a)32P-S2-cDNA, 32P-S3-cDNA, and 32PS6-cDNA. In addition to themajor hybridizing species, there is a fragment "a" common to allgenotypes that hybridizes weakly to all probes.(B) and (C) The use of RFLPs to identify the S-genotype ofindividuals in segregating populations. DNA gel blot analysis ofgenomic DNA from individual plants generated from crosses S,S3x S$S7 (B) and S,S3 x SA (C). Each lane contains DNA from aseparate plant and is labeled with the S-genotype assigned bybreeding behavior. The distinct Hindlll restriction fragments as-sociated with the S2-, S3-, S6-, and S7-alleles are indicated byarrows. The probe was a combination of S2-, S3-, and S6-cDNAclones labeled by random priming. No S!-genotype-specific frag-ments were detected with this probe. S,-Genotype was thereforeassigned when fragments corresponding to the other genotypeswere absent. Two fragments (a and b) hybridized weakly to theprobe.

486 The Plant Cell

3.33 -\

I,

- 2.93- _____Peptide

-22 1 30 60 90 120 150 180

Amino acid residue number

pistil, the antibody bound primarily to the extracellularmaterial in the stigma and transmitting tract of the style.In the ovary, the antibody was localized on the epidermisof the placenta and in a thin layer surrounding each ovule.Negligible binding occurred when the antibody to the syn-thetic S2-peptide was omitted or replaced with an antibodydirected to the synthetic arginine vasopressin-peptide(data not shown).

The electron micrographs of the style transmitting tissuein Figure 6 showed specific localization of the antibody tothe extracellular matrix. There was a little binding to thewall of the transmitting tissue cells and very little binding

13 66 33 23 22 13 66 33 23 22

_94

67

43

-30

20

Figure 4. An Antibody Raised to a Peptide Sequence of One ofthe Hypervariable Regions of the S2-Glycoprotein Binds Specifi-cally to the S2-Glycoprotein.

(A) Hydrophobicity plot of the deduced amino acid sequence ofthe S2-glycoprotein. Values above the line denote hydrophobicregions and values below the line represent hydrophilic regions.The putative hydrophobic signal peptide (amino acids -22 to -1)is stippled. The potential W-glycosylation sites are marked witharrows. The regions of the S2-glycoprotein with greatest sequencediversity from the S3- and S6-glycoproteins are shaded (A to D).Region B contains the peptide sequence used to generate the S2-specific antibody. The hydrophobicity profile was generated usingthe predictive rules of Kyte and Doolittle (1982) and a span of 9consecutive residues.(B) Specificity of antibodies raised against the synthetic S2-peptide.(a) Protein gel blot of style extracts probed with antibody raisedto the synthetic S2-peptide. The antibody specifically binds to the32-kD S2-associated glycoprotein. There is no detectable bindingto other style proteins or to S-glycoproteins of other genotypes.(b) Silver-stained SDS-polyacrylamide gel (12.5%) displaying ex-tracts from mature styles of various N. a/ata S-genotypes. TheS,-, S2-, S3-, and S6-allele-associated proteins are indicated byarrows. The S3-associated protein is present in lower amountsthan the other S-allele-associated glycoproteins and is oftenmasked by a protein of the same mobility, which is present in allgenotypes (Jahnen et al., 1989).

Figure 5. Binding of Anti-S2-peptide Antibody to Ovary and Style.

(A) Longitudinal section of stigma labeled with anti-S2-peptideantibody and immunofluorescent marker. The walls of the stigmacells, including the surface papillae, are labeled. The corticaltissues (CO) and epidermis of the style (arrows) are not labeled.(Scale bar = 100 ̂ m.)(B) Longitudinal section of the style showing cortical tissues (CO)and transmitting tract (TT). The intercellular matrix in the trans-mitting tract is labeled. (Scale bar = 100 ^m.)(C) Longitudinal section of locule showing labeling of placentalepidermis (arrows) and the outer surface of the ovules (OV). (Scalebar = 50 ̂ m.)

Figure 6. Electron Micrographs of Style and Ovary Showing Immunogold Localization of Anti-S2-Peptide Antibody (Scale bars = 500 nm.)

(A) Transverse section of style through transmitting tract region labeled with anti-S2-peptide antibody. The intercellular matrix (IM) isheavily labeled with immunogold markers. In addition, there is some labeling of the cytoplasm of the transmitting tract cells. Amyloplasts(AM), mitochondria (Ml), and vacuoles (VA) are not labeled.(B) Similar section to (A) labeled with control antibody. There is no specific labeling in any region.(C) Longitudinal section through locule (LO) labeled with S2-peptide antibody showing two cells of placental epidermis. The cell wall isheavily labeled.

488 The Plant Cell

to the transmitting tissue cells (see Figure 6A). The control anti-arginine vasopressin antibody did not label any struc- tures in the style (see Figure 6B). In the locule, the labeling is concentrated over the primary wall of the placenta1 epidermal cells (see Figure 6C).

DISCUSSION

A major finding of this study is the extent of variation between the alleles. Alignment of the amino acid se- quences derived from the S2-, S3-, and S6-cDNA sequences shows that, although there is more than 50% homology, there are also regions of considerable diversity. We de- scribe these nonconserved regions as “hypervariable” and the conserved regions as “framework” regions following the terminology described by Wu and Kabat (1970) for immunoglobulin structure. This feature of hypervariable and framework sequences in related proteins is common to a number of animal proteins encoded by the immuno- globulin gene superfamily (Williams and Barclay, 1988). Each of these genes is concerned with some aspect of cell recognition, and this seems to be a characteristic feature of this class of molecules. The products of the S alleles are presumed to have a recognition function, as the growth of pollen tubes bearing a particular allele is arrested within the style of plants bearing the same allele (de Nettancourt, 1977).

It is possible that the conserved cysteine residues act to form disulfide bonds and present the hypervariable regions in binding domains as for the immunoglobulin supergene family. The variation in the amino acid sequence of the S-allele products is associated with hydrophilic regions, which are predicted to be on the surface of the molecule (Hopp and Woods, 1981), and would be acces- sible for binding to appropriate receptors in the pollen or pollen tube. Antibodies raised to a synthetic peptide cor- responding to a stretch of the major hypervariable region bound to the native S2-glycoprotein in fresh and fixed tissue sections as well as in enzyme-linked immunosorbent assays (data not shown), suggesting that this peptide is exposed at the surface of the glycoprotein. The use of the synthetic peptide as an antigen obviated the difficulties previously encountered using either polyclonal or monoclo- na1 antibodies raised to purified Sn-glycoprotein. These difficulties of cross-reactivity with other style glycoproteins were due to the glycosyl substituents on the Sn-glycopro- tein, which have saccharide sequences common to many plant glycoproteins and which are immunodominant (How- lett and Clarke, 1981; Anderson, Sandrin, and Clarke, 1984). lmmunocytochemical studies with the antipeptide antibody showed that the S2-glycoprotein is localized ex- tracellulary in the mucilage that surrounds the transmitting tissue cells and the wall of the ovule. This corresponds to the expression of the S,-gene in transmitting tissue cells

and the epidermis of the placenta demonstrated by in situ hybridization (Cornish et al., 1987). The secretion of the S,-glycoprotein is consistent with the presence of signal sequences detected in two of the cDNA clones. The extra- cellular location in the mucilage of the transmitting cells and the ovule coincides with the pathway of pollen tube growth so that the pollen tubes would be in contact with the S2-glycoprotein as they grow through the pistil.

The carbohydrate chains are another source of potential variation and hence of potential allelic specificity. Although the glycosylation sites are conserved, the numbers of chains present vary and the sequences within these chains may also vary. Detailed structural analysis of the individual chains is required to clarify this point.

Together, the DNA gel blot analyses of genomic DNA and the breeding data indicate that the three cDNAs de- scribed are indeed allelic products of the same gene. Each S-allele has a characteristic set of Hindlll restriction frag- ments (Bernatzky et al., 1988) which hybridize most strongly to cDNA encoded by that allele and weakly with cDNA encoded by other S-alleles. This indicates that the S2-, S3-, and S6-cDNAs are derived from the same locus. If they were from different loci, the DNA gel blot analyses would be expected to show bands corresponding to all genotypes in each track, unless there was specific deletion of genes corresponding to all but the expressed allele. Breeding experiments demonstrated that the allele-specific RFLPs segregated with S-genotype in plants produced by crossing two individuals with no alleles in common or with one allele in common, consistent with the interpretation that the S2-, S3-, and S6-cDNAs represent alleles of the same locus or closely linked genes. Faint bands common to each genotype (see Figure 3A) were detected in DNA gel blot analyses; the significance of these bands cannot be interpreted until the corresponding DNA is cloned and sequenced. Comparison of the amino acid sequences of the three N. alata S-glycoproteins with the amino acid sequences of S-glycoproteins from the sporophytically controlled self-incompatibility of Brassica spp. (Isogai et al., 1987; Nasrallah et al., 1987) shows that there is no homology between the sequences of the two groups of S- glycoproteins, suggesting that the two systems of self- incompatibility evolved separately. The allelic S-glycopro- teins from Brassica species have a higher leve1 of se- quence homology than the three allelic S-glycoproteins from N. alata; like the N. alata S-glycoproteins, the cysteine residues are conserved, but there is apparently more variation in the numbers of potential glycosylation sites.

In the absence of information on the corresponding S- allelic products for pollen, it is difficult to interpret the amino acid sequence data in functional terms. We are also cautious of drawing general conclusions from a relatively narrow data base; in some systems such as clover (Att- wood, 1944; Williams, 1947), severa1 hundred alleles exist. At present we have information on only a few of the many alleles of two self-incompatibility systems; as more se-

Self-lncompatibility Alleles of N. alata 489

quence data from related systems become available, the structure-function relationship will become clearer.

METHODS

Plant Material

Plants were maintained in a pollinator-proof glasshouse. S-Geno- type was assigned by monitoring for seedset after hand pollination with pollen from tester plants homozygous for the S7-, SS-, S3-, and S,-alleles.

cDNA Library Construction and Screening

Poly-A+ RNA was isolated from mature S& and s& styles, and double-stranded cDNA was prepared using the Ribonuclease H method and Amersham Kit RPN1256. The cDNA was cloned into XgtlO (Huynh, Young, and Davis, 1984), and SB- and Ss-clones were selected from the libraries by differential screening with single-stranded 32P-cDNAs prepared from style poly-A’ RNA from S3S3 and s& genotypes. Prehybridization (2 hr at 68OC) was in 1.5 x SSPE, 1 .O% SDS, 0.5% BLOTTO, and 0.5 mg/ml denatured herring sperm DNA (Bio-Rad Bulletin 1234). Hybridization was in prehybridization solution with added dextran sulfate (1 0% w/v final concentration). The cDNA probes were used at 3 x 106 cpm ml-’. Prehybridization was for 2 hr at 68°C and hybridization for 16 hr at 68OC. Filters were washed twice in 2 x SSC, 0.1% SDS at 68”C, and once in 0.1 x SSC, 0.1% SDS at 68OC. Clones from the S S 3 library that hybridized with the SS3-cDNA but not the &&-CDNA were rescreened with 3ZP-labeled S2-cDNA clone NA- 2-2 (1 x 106 cpm/ml; Anderson et al., 1986). The clones selected in this way hybridized strongly with total SS3-cDNA, weakly with the S,-cDNA clone, and not detectably with the total S&-CDNA, and were therefore the most likely candidates for cDNAs encoding the S3-alleles. The S&&-CDNA library was screened in the same manner except that clones were selected which hybridized to S,&-CDNA and not SS,-cDNA. One candidate clone, with the longest insert, was selected for each of the S3- and S,-alleles and used for subsequent work (NA-3-1, NA-6-1).

The cDNA inserts from clones NA-3-1 and NA-6-1 were ligated into either the bacteriophage vector M13mpl8RF DNA for single- strand sequencing or into pGEM-3 (Promega) for double-strand sequencing using the chain termination method.

Sequence data obtained in this way were analyzed as described previously (Anderson et al., 1986).

DNA and RNA Gel Blots

RNA gel blot analysis was performed using 1.2% agarose-form- aldehyde gels as described previously (Anderson et al., 1986). The RNA was transferred to hybond-N (Amersham) and probed with either 32P-labeled S2-cDNA (NA-2-2), S3-cDNA (NA-3-l), or S6-cDNA (NA-6-1) (3 x 10’ cpm pg-’, 1 X 106 cpm ml-’).

Prehybridization (2 hr, 68°C) was in 1.5 x SSPE, 1 .O% SDS, 0.5% BLOTTO, and 0.5 mg/ml denatured herring sperm DNA (Bio-Rad Bulletin 1234). Hybridization was in prehybridization solution with added dextran sulfate (1 0% w/v final concentration).

Filters were washed twice in 2 X SSC, 0.1% SDS at 68OC, and once in 0.1 x SSC, 0.1% SDS at 68OC.

DNA for gel blots was prepared from 10 g of leaf tissue according to the method of Bernatzky and Tanksley (1986a). The genomic DNA (5 pg) was digested with Hindlll and fractionated on 0.9% w/v agarose gels before transfer onto a Zetaprobe membrane (Bio-Rad). The transfer and hybridization procedure was as described by Bernatzky and Tanksley (1986b) except that the DNA was transferred to the membrane by “wet blotting” in 0.5 M NaOH. The membrane was then neutralized with 0.5 M Tris-HCI, 0.1% SDS, pH 7.5, followed by 2 x SSC, 0.1% SDS before hybridization at 68°C. Filters were washed in 1 x SSC, 0.1% SDS at 68OC and exposed to x-ray film at -7OOC for 4 days. A Hindlll digest of X DNA was used as fragment-size markers.

Peptide Synthesis, Antibody Production, and Purification

The 22-amino acid peptide C-D-R-S-K-P-Y-N-M-F-T-D-G-K-K-K- N-D-L-D-E-R, starting from position 47 (Figure 2) was synthesized by the BOC-polystyrene solid-phase procedure using an Applied Biosystems Model 430A peptide synthesizer. The assembled peptide was cleaved from the resin with anhydrous hydrogen fluoride and the crude peptide purified by preparative reverse- phase HPLC using a pBondapak C18 column (buffer A, 0.1% TFA-water, buffer B, 0.1 o/o TFA-water-60% acetonitrile; 10% to 80% B over 60 min). The purified peptide was found to be homogeneous, as assessed by analytical HPLC and amino acid analysis. The yield from crude was 24%. The purified peptide (5 mg) in the reduced, monomeric form was conjugated to bovine thyroglobulin (0.5 mg) (Skowsky and Fisher, 1972). The conju- gated peptide (1 mg/ml in PBS) was mixed with an equal volume of Freund’s complete adjuvant. Three sheep were injected intra- muscularly with 2 ml. After 4 weeks, booster injections containing 0.5 mg of peptide conjugated to 0.1 mg of thyroglobulin and emulsified into 1 ml of Freund’s incomplete adjuvant were given to each of the three sheep. After 6 weeks, the sheep were bled. The antibody levels were assessed by radioimmunoassay using lZ51-peptide as antigen (Skelly, Brown, and Besch, 1973).

The antibody was affinity purified on an S,-peptide Sepharose column. The column was prepared by covalent attachment of 5 mg of S,-peptide to 5 g of activated CH-Sepharose (Pharmacia LKB Biotechnology Inc.) according to the manufacturer’s instruc- tions. Sheep serum (5 ml) was applied to the column. Bound antibody was eluted using citric acid-phosphate buffer and was pooled and dialyzed against phosphate-buffered saline (yield, 7 ml, 400 pg/ml).

Protein Extraction and lmmunoblotting

Style extracts were prepared by crushing three mature styles in 150 pI of extraction buffer (50 mM Tris-HCI, pH 8.5, 1 O mM EDTA, 1 mM CaCI,, 1 mM phenylmethylsulfonyl fluoride, 1 mM dithio- threitol). The extracts were microcentrifuged for 10 min at 4OC and 5 pl of the supernatant was removed and applied to an SDS- polyacrylamide gel (1 2.5%) in Laemmli sample buffer (Laemmli, 1970).

When electrophoresis was complete, the gel was electroblotted and the nitrocellulose was incubated with affinity-purified, antiS2-

490 The Plant Cell

peptide antibody (0.8 pg/ml in 1% gelatin in TBS [20 mM Tris- HCI, pH 7.5, 500 mM NaCl]). The nitrocellulose was washed with 0.05% Tween in TBS, and the bound antibody was visualized by incubation with rabbit anti-sheep IgG horseradish peroxidase con- jugate (0.1 pg/ml in 1% gelatin in TBS, Bio-Rad), followed by washing in 0.05% Tween in TBS and incubation with horseradish peroxidase color development solution (Bio-Rad).

lmmunocytochemistry

Styles (S&-genotype) were fixed as previously described (Mc- Fadden et al., 1988). For immunofluorescence, 1-pm sections were incubated with anti3,-peptide antibody (30 pg/ml) in SC buffer (McFadden et al., 1988) for 1 hr at 20°C. Sections were washed in SC, and then incubated with rabbit anti-sheep IgG conjugated to fluorescein isothiocyanate (1 :50 dilution, Sigma) for 1 hr at 20°C. Ultrathin sections for EM were incubated with anti- S,-peptide antibody (30 pg/ml) for 1 hr at 2OoC, washed, and then labeled with rabbit anti-sheep IgG antibody conjugated to 15 nm colloidal gold (EY Laboratories). The control antibody, raised in sheep against the synthetic peptide arginine vasopressin and purified by affinity chromatography, was a kind gift from Dr. Mario Congin.

ACKNOWLEDGMENTS

We thank Dr. M. Altschuler, Dr. R. Kalla, Dr. W. Jahnen, and Dr. R. Simpson for valuable discussions; Mr. B. McGinness and Ms. S. Mau for providing plant material; and Ms. M. Strode, Ms. A. Shuttleworth, and Ms. J. Noble for typing the manuscript.

Received March 13, 1989.

REFERENCES

Anderson, M.A., Sandrin, M.S., and Clarke, A.E. (1984). A high proportion of hybridomas raised to a plant extract secrete antibody to arabinose or galactose. Plant Physiol. 75, 1013- 1016.

Anderson, M.A., Cornish, E.C., Mau, S-L., Williams, E.G., Hog- gart, R., Atkinson, A., Bonig, I., Grego, B., Simpson, R., Roche, P.J., Haley, J.D., Penshow, J.D., Niall, H.D., Tregear, G.W., Coghlan, J.P., Crawford, R.J., and Clarke, A.E. (1986). Cloning of cDNA for a stylar glycoprotein associated with expression of self-incompatibility in Nicotiana alata. Nature 321,

Attwood, S.S. (1 944). Oppositional alleles in natural populations of Trifolium repens. Genetics 29, 428-435.

Bernatrky, R., and Tanksley, S.D. (1986a). Genetics of actin- related sequences in tomato. Theor. Appl. Genet. 72, 314-321.

38-44.

Bernatzky, R., and Tanksley, S.D. (1 986b). Methods for detec- tion of single or low copy sequences in tomato on Southern blots. Plant MOI. Biol. Rep. 4, 37-41.

Bernatzky, R., Anderson, M.A., and Clarke, A.E. (1988). Molec- ular genetics of self-incompatibility in flowering plants. Dev. Genet. 9, 1-12.

Clarke, A.E., Anderson, M.A., Bacic, A., Cornish, E.C., Harris, P.J., Mau, S.-L., and Woodward, J.R. (1987). Molecular as- pects of self-incompatibility in flowering plants. In Plant Gene Systems and Their Biology, J.L. Key and L. Mclntosh, eds (New York: Alan R. Liss Inc.), pp 53-64.

Cornish, E.C., Pettitt, J.M., Bonig, I., and Clarke, A.E. (1987). Developmentally controlled expression of a gene associated with self-incompatibility in Nicotiana alata. Nature 326, 99-1 02.

Cornish, E.C., Pettitt, J.M., and Clarke, A.E. (1988). Self-incom- patibility genes in flowering plants. In Temporal and Spatial Regulation of Plant Genes, Vol. 5, D.P.S. Verma and R. Gold- berg, eds (New York: Springer-Verlag), pp. 11 7-1 30.

de Nettancourt, D. (1 977). lncompatibility in angiosperms. In Monographs on Theoretical and Applied Genetics, Vol. 3, R. Frankel, G.A.E. Gall, and H.F. Linskens, eds (Berlin: Springer- Verlag) .

Hamlyn, P.H., Gait, M.J., and Milstein, C. (1981). Complete sequence of an immunoglobulin mRNA using specific priming and the dideoxynucleotide method of RNA sequencing. Nucl. Acids Res. 9,4485-4494.

Harris, P.J., Anderson, M.A., Bacic, A., and Clarke, A.E. (1984). Cell-cell recognition in plants with special reference to the pollen-stigma interaction. In Oxford Surveys of Plant Molecular and Cell Biology, Vol. 1, B.J. Miflin, ed (Oxford: Oxford Univer- sity Press), pp. 161-203.

Hopp, T.P., and Woods, R.R. (1981). Prediction of protein anti- genic determinants from amino acid sequences. Proc. Natl. Acad. Sci. USA. 78, 3824-3828.

Howlett, B.J., and Clarke, A.E. (1981). Role of carbohydrate as an antigenic determinant of a glycoprotein from rye-grass (Lol- ium perenne) pollen. Biochem. J. 197, 707-714.

Huynh, T.V., Young, R.A., and Davis, R.W. (1984). Constructing and screening cDNA libraries in XgtlO and Agtll. In DNA Cloning: A Practical Approach, D. Glover, ed (Oxford: IRL).

Isogai, A., Takayama, S., Tsukamoto, C., Ueda, Y., Shiozawa, H., Hinata, K., Okazaki, K., and Suzuki, A. (1987). S-Locus- specific glycoproteins associated with self-incompatibility in Brassica campestris. Plant Cell Physiol. 28, 1279-1291.

Jahnen, W., Batterham, M.P., Clarke, A.E., Moritr, R.L., and Simpson, R.J. (1 989). Identification, isolation, and N-terminal sequencing of style glycoproteins associated with self-incom- patibility in Nicotiana alata. Plant Cell 1, 493-499.

Kyte, J., and Doolittle, R.F. (1 982). A simple method for displaying the hydropathic character of a protein. J. MOI. Biol. 157, 105- 132.

Laemmli, U.K. (1 970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680- 685.

Lawrence, M.J., Afzal, M., and Kenrick, J. (1978). The genetical control of self-incompatibility in Papaver rhoeas. Heredity 40,

McFadden, G.I., Bonig, I., Cornish, E.C., and Clarke, A.E. 239-252.

Self-lncompatibility Alleles of N. alata 491

(1988). A simple fixation and embedding method for use in hybridization histochemistry on plant tissues. Histochem. J. 20,

Nasrallah, J.B., Kao, T.H., Goldberg, M.L., and Nasrallah, M.E. (1 985). A cDNA clone 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. (1 987). Amino-acid sequence of glycoproteins encoded by three alleles of the S-locus of Brassica oleracea. Nature 326, 61 7-61 9.

Skelly, D.S., Brown, L.P., and Besch, P.K:(1973). Radioimmuno- assay. Clin. Chem. 19, 146-186.

575-586.

Skowsky, W.R., and Fisher, D.A. (1972). The use of thyroglobulin to induce antigenicity to small molecules. J. Lab. Clin. Med. 80,

Williams, W. (1 947). Genetics of red clover (Trifolium pratense L.) compatibility. 111. J. Genet. 48, 69-79.

Williams, A.F., and Barclay, A.N. (1 988). The immunoglobulin superfamily-domains for cell surface recognition. Annu. Rev. Immunol. 6,381-405.

Wu, T.T., and Kabat, E.A. (1 970). An analysis of the sequences of the variable regions of Bence Jones proteins and myeloma light chains and their implications for antibody complementarity. J. Exp. Med. 132,211-250.

134-1 44.

DOI 10.1105/tpc.1.5.483 1989;1;483-491Plant Cell

and A E ClarkeM A Anderson, G I McFadden, R Bernatzky, A Atkinson, T Orpin, H Dedman, G Tregear, R Fernley

Sequence variability of three alleles of the self-incompatibility gene of Nicotiana alata.

 This information is current as of January 11, 2016

 

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists