Identification of Residues in a Hydrophilic Loop of the Papaver rhoeas S Protein That Play a Crucial Role in Recognition of Incompatible Pollen

Correspondence to: F. Christopher H. Franklin, F.C.H.Franklin{at}bham.ac.uk (E-mail), 44-121-414-5925 (fax).

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

The self-incompatibility response involves S allele–specific recognition between stigmatic S proteins and incompatible pollen. This response results in pollen inhibition. Defining the amino acid residues within the stigmatic S proteins that participate in S allele–specific inhibition of incompatible pollen is essential for the elucidation of the molecular basis of the self-incompatibility response. We have constructed mutant derivatives of the S₁ protein from Papaver rhoeas by using site-directed mutagenesis and have tested their biological activity. This has enabled us to identify amino acid residues in the stigmatic S proteins of P. rhoeas that are required for S-specific inhibition of incompatible pollen. We report here the identification of several amino acid residues in the predicted hydrophilic loop 6 of the P. rhoeas stigmatic S₁ protein that are involved in the inhibition of S₁ pollen. Mutation of the only hypervariable amino acid, which is situated in this loop, resulted in the complete loss of ability of the S protein to inhibit S₁ pollen. This clearly demonstrates that this residue plays a crucial role in pollen recognition and may also participate in defining allelic specificity. We have also established the importance of highly conserved amino acids adjacent to this hypervariable site. Our studies demonstrate that both variable and conserved amino acids in the region of the S protein corresponding to surface loop 6 are key elements that play a role in the recognition and inhibition of incompatible pollen in the pollen–pistil self-incompatibility reaction.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Self-incompatibility (SI) in the field poppy (Papaver rhoeas) is determined gametophytically by a multiallelic S locus (Lawrence et al. 1978 ). It is proposed that genes residing at the S locus encode a stigmatic S protein and a corresponding pollen S protein. Both proteins are responsible for a highly specific recognition event leading to rejection of self-pollen (Franklin et al. 1995 ). Previous studies of wild populations suggest that there are ~40 to 50 different S alleles present in each population, and comparative studies suggest that the overall figure for the species is ~65 (Lane and Lawrence 1993 ).

We have reported the cloning of cDNAs for the S₁, S₃, S₇ (partial sequence), and S₈ alleles of the stigmatic S gene, together with Sn₁, a putative stigmatic S gene from P. nudicaule (Foote et al. 1994 ; Walker et al. 1996 ; Kurup et al. 1998 ). Several of these genes have been expressed in Escherichia coli to produce recombinant S proteins. These have been isolated and demonstrated to be biologically active because they specifically inhibit pollen of the same S genotype in an in vitro bioassay. The S proteins are small (~15 kD) secreted proteins; some are modified by N-glycosylation. Comparison of the amino acid sequences derived from the cloned S gene cDNAs has revealed that they are highly polymorphic, sharing between 50 and 65% identity. Nevertheless, they are predicted to adopt virtually indistinguishable secondary structures (Walker et al. 1996 ; Kurup et al. 1998 ). These are comprised of a series of six ß strands with two -helical regions located at the C terminus and linked by seven hydrophilic loops, which exhibit high surface probability.

The product of the pollen S gene interacts with the stigmatic S protein; however, this product has yet to be identified, although it is believed to be a receptor of some type. We have recently identified a pollen plasma membrane proteoglycan that interacts with the S protein, although not in an S allele–specific manner. This S protein binding protein is thought to be an accessory receptor with a role in mediating the specific interaction between the stigma and pollen S proteins (Hearn et al. 1996 ). Studies have demonstrated that this recognition event activates a calcium-based signaling pathway (Franklin-Tong et al. 1993 , Franklin-Tong et al. 1995 , Franklin-Tong et al. 1997 ) and is associated with phosphorylation of a number of pollen proteins. These events lead to the rapid arrest of pollen tube growth typical of an incompatible pollination (Rudd et al. 1996 , Rudd et al. 1997 ).

Currently, an area of crucial importance in the study of SI is defining the molecular basis of S allele specificity together with identifying the pollen S gene product. Recently, a number of groups have addressed these areas of concern in the SI RNases (S RNases) of the Solanaceae. Although the control of the SI system in the Solanaceae is genetically identical to that of P. rhoeas, mechanistically the systems are entirely different. There is no detectable homology between the S RNases and the poppy S proteins. Sequence alignments across a range of S RNases have revealed that these proteins possess five highly conserved domains and two hypervariable domains. A number of studies have investigated the role of the hypervariable regions in the control of allele-specific recognition of pollen. In an elegant experiment, Matton et al. 1997 substituted the hypervariable domains of Solanum chacoense S₁₁ with those of S₁₃. The result of this substitution was that the transgenic recombinant allele behaved as a fully functional S₁₃ allele. This led these authors to conclude that allelic specificity can be determined by the hypervariable regions alone. Although their observations are consistent with this conclusion, as Verica et al. 1998 point out, this result may have been influenced by the particular combination of S alleles used in the study.

In comparison with all S alleles that have been characterized to date, S₁₁ and S₁₃ of S. chacoense are remarkably similar because they differ in only 10 of 190 amino acids. When similar domain swap experiments were conducted between more divergent pairs of alleles, they did not acquire the new specificity, even though the hybrid proteins retained ribonuclease activity and lost their specificity (Kao and McCubbin 1996 ; Zurek et al. 1997 ). Such observations indicate that swap experiments between more divergent pairs of alleles can result in disruption of recognition. This implies that allelic recognition may not be localized to any particular domain of the S RNase and has led to the suggestion that residues outside of the two designated hypervariable domains may also participate in S specificity (Verica et al. 1998 ).

To elucidate the molecular basis of the pollen–pistil interaction in poppy, which is the next stage, we are attempting to define the amino acid residues within the stigmatic S proteins that are essential for the recognition of incompatible pollen. We have constructed mutant derivatives of the S proteins and tested their biological activity. In this study, we demonstrate that certain amino acids are crucial for S-specific activity, because some of these mutant S proteins are defective in their inhibition of incompatible pollen. Our studies provide evidence that both variable and conserved amino acids are important in the pollen–pistil SI interaction.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Choice of Amino Acid Residues for Mutation in the S₁ Protein
Our studies to date indicate that variations in amino acid sequence are likely to be responsible for the different specificities of the alleles of the P. rhoeas S protein (Foote et al. 1994 ; Franklin-Tong et al. 1995 ; Walker et al. 1996 ). Therefore, we compared the available poppy S protein sequences to identify variable residues that might participate in S allele–specific recognition of incompatible pollen. These sequence alignments are shown in Figure 1. They reveal a number of blocks of complete sequence identity or similarity interspersed with more variable regions. Figure 1 also illustrates the secondary structure predictions for the S proteins based on these sequences. We were particularly interested in any variable amino acids present in the predicted hydrophilic surface loops, because these may be readily available for interaction with the pollen S gene product. It is evident from Figure 1 that despite the overall high level of sequence polymorphism between the S alleles, only one amino acid residue, corresponding to Asp at position 79 in loop 6 of S₁, is completely variable across these five sequences (indicated by gray shading). The sequence alignment also indicates that the S₃ protein possesses an additional amino acid in loop 5. Although loop 5 is not completely polymorphic across all five alleles, it does exhibit a significant degree of variability.

View larger version (0K): [in this window] [in a new window]   Figure 1. Alignment of the P. rhoeas S Alleles and Secondary Structure Predictions. Shown is an alignment of the deduced amino acid sequences of the mature polypeptides of all P. rhoeas S alleles analyzed to date, including the partial S7 sequence. Sequences for products of alleles S1, S3, S7, and S8 are from P. rhoeas; the Sn1 product is from P. nudicaule. The numbers at the end of each line of the sequences indicate the amino acid residue positions at the end of that line. Above the sequences, the predicted secondary structure of all of these proteins is outlined. The arrows indicate predicted ß strands, blocks indicate predicted helices, and numbered lines 1 to 7 indicate the predicted hydrophilic surface loops. The only amino acid residue (Asp-79 in loop 6 of S1) that is completely variable across these five sequences is indicated in the gray-shaded vertical block. The positions of mutations made in the S1 protein are also indicated in gray blocks.

On the basis of these observations, a program of mutagenesis, which is described below, was devised. We have indicated the positions of mutations made to the S₁ protein described here in gray blocks in Figure 1. We were aware of potential problems with activity if the mutations affected folding of the protein, and so all substitutions were, as far as possible, conservative, or they involved replacing the amino acid residues present in the S₁ sequence with those found at the corresponding position in the S₃ or S₈ sequences. Furthermore, because the majority of the mutations were targeted to the predicted hydrophilic surface loops, they were less likely to adversely interfere with the folding of the recombinant proteins. The prime candidate for mutation in the S₁ product was Asp-79 in loop 6. This is because it is the only hypervariable residue identified by the S allele sequence alignment; in addition, it was predicted to be in a surface loop. Subsequent mutations were incorporated in loop 5, which exhibited considerable variability, and in the remaining surface loops.

After generating the mutant sequences, the mutant S proteins were expressed in E. coli, and the proteins were purified, refolded, and tested to determine whether they had altered biological activity (see Methods). Their ability to specifically inhibit S₁ pollen was compared with that of the nonmutated wild-type S₁ (S₁wt) recombinant protein. Figure 2 illustrates the normal S-specific activity of the S₁wt recombinant protein. This analysis reveals that at concentrations of ~10 µg mL^-1 and above, complete S-specific inhibition of incompatible pollen can be observed. Furthermore, even at the highest concentration tested (150 µg mL^-1), no nonspecific inhibitory effect was detectable at these concentrations.

View larger version (0K): [in this window] [in a new window]   Figure 2. A Dose–Response Curve for the S-Specific Inhibition of S1 Pollen by the S1wt Protein. The S-specific inhibition activity of the recombinant S1wt protein was determined using S1S4 and S2S6 pollen. The biological activity of the S protein was tested over a range of concentrations from 10 to 150 µg mL-1. Pollen from plants of the S1S4 genotype was inhibited by ~50% (which is the theoretical maximum inhibition because the pollen used was heterozygous) at concentrations 10 µg mL-1. In contrast, fully compatible pollen (from plants of genotype S2S6) revealed no detectable inhibition, even at the higher concentrations of S1wt used. This clearly demonstrates that the S-specific inhibitory effect of the S1wt protein is effective in the range of 10 to 150 µg mL-1, with no nonspecific inhibitory effect at these concentrations.

Both Variable and Conserved Residues in Loop 6 Participate in Inhibition of Incompatible Pollen
Figure 3A shows the biological activity of the mutant S proteins generated in loop 6. The first mutation made was D79G (substitution of Gly for Asp at position 79 in the S₁ product). This is a substitution in the S₁ sequence, with the residue found at this position in the S₃ protein. As shown in Figure 3A, Figure D79G was unable to inhibit S₁ pollen, even when applied at 150 µg mL^-1 (which is ~15-fold the amount of protein normally required to inhibit incompatible pollen). Thus, this single alteration in the sequence resulted in the complete loss of S-specific activity. This substitution did not confer S₃ inhibitory activity on the mutant.

View larger version (0K): [in this window] [in a new window]   Figure 3. Effect of Mutations in the S1 Protein on S-Specific Pollen Inhibitory Activity. The mutant S proteins were tested for their ability to specifically inhibit S1 pollen. The normal S-specific activity of the wild-type S1 (S1wt) recombinant protein served as the standard against which the activity of the mutant S proteins was compared. The mutant proteins were routinely assayed at several concentrations in a randomized format, in at least three independent experiments, and against pollen carrying the S1 allele and fully compatible pollen of a different S genotype (see Methods). The activity of the mutant S proteins is expressed as the percentage of activity relative to S1wt. (A) Effect of loop 6 mutations. Shown is pollen inhibitory activity with substitutions to the predicted loop 6 region of the S1 amino acid sequence. Pollen inhibitory activity of the mutant proteins is shown for two concentrations (75 and 150 µg mL-1) and is expressed as a percentage of S1wt protein activity. Note that all mutations, except for R76H and D77E/D78E, result in complete loss of pollen inhibitory activity. (B) Effect of other mutations. Shown is pollen inhibitory activity with substitutions to the predicted hydrophilic loops of the S1 amino acid sequence. Pollen inhibitory activity of the mutant proteins is shown for 25 and 75 µg mL-1 and is expressed as a percentage of S1wt protein activity. Note that there is no significant reduction in pollen inhibitory activity, except for K63R/E64K/T65G and K63R/E64K/T65G/C82Y.

A virtually identical result was obtained with D79H, which is the corresponding residue from S₈. Inhibition of S₁ pollen by D79H was only 4.3% that of inhibition by S₁wt, even at a concentration of 150 µg mL^-1. This was not statistically significantly different from the germination medium control (P = 0.05). We have also made the reciprocal mutation to the S₈ sequence by substituting the hypervariable His-80 in S₈ with Asp, which is located at the same relative position in the S₁ sequence. This mutation to the S₈ allele (H80D) resulted in a complete loss of biological activity toward S₈ pollen, as shown in Figure 3A, and did not confer inhibitory activity toward S₁ pollen. These data clearly demonstrate that substitutions of Asp-79 in the S₁ protein and the corresponding His-80 in the S₈ protein have a critical effect on the S-specific biological activity of the stigmatic S protein. These mutant S proteins are completely unable to inhibit pollen carrying the S₁ allele, that is, their ability to induce an incompatible response has been removed. Therefore, the amino acid residues at this hypervariable site in loop 6 must play a critical role in recognition and inhibition in the SI interaction of the stigmatic protein with incompatible pollen.

In contrast to the hypervariability seen at position 79 in S₁, several of the remaining residues in loop 6 (residues 76 to 78) are either completely or highly conserved across the five alleles that we have sequenced thus far (as shown in Figure 1). However, because ~65 S alleles are predicted to occur in P. rhoeas (Lane and Lawrence 1993 ), it is quite possible that the amino acids that occupy these apparently conserved positions do, in fact, vary in certain alleles. Either way, their proximity to Asp-79 suggests that they may still contribute to the interaction between the S protein and the pollen component. On this basis, Arg-76 and Asp-77 and Asp-78 in S₁ were substituted in turn with His.

As illustrated in Figure 3A, Figure R76H had no detectable effect on inhibition of S₁ pollen. This indicates that this residue does not participate in the recognition of incompatible pollen. Importantly, it also implies that the protein can tolerate substitutions to nonessential residues with no apparent effect on biological activity. In contrast, both D77H and D78H dramatically reduced the inhibitory activity of the protein against S₁ pollen. At best, the activities of D77H and D78H were only 6 to 9% of the activity exhibited by S₁wt and were not significantly different (P = 0.05) from the germination medium control. The double mutant protein D77H/D78H was completely inactive against S₁ pollen, and no pollen inhibitory activity was detected, as shown in Figure 3A.

Although the substitution of Asp with His could be considered relatively conservative in terms of hydration potential, it removed a negative charge (two in the case of D77H/D78H). To explore the significance of a negatively charged carboxyl side chain, we constructed a second double mutant (D77E/D78E). Glu is charged negatively like Asp but has a greater side chain volume. As shown in Figure 3A, when tested against S₁ pollen, no statistically significant difference (P = 0.05) was detected between the biological activities of D77E/D78E and S₁wt. These data suggest that the negatively charged carboxyl side chains of these two conserved, acidic amino acids are an important element in the recognition of S₁ pollen.

Effect of Mutations in Surface Loop 5 of the S₁ Protein
The sequence alignment also revealed a number of variable amino acids in loop 5 (see Figure 1). In particular, S₃ possesses an extra amino acid residue at this point. Initially, Lys-63 in S₁ was substituted with the residue found in the corresponding position in S₃ to give construct K63R. As shown in Figure 3B, when this mutant was assayed for biological activity, it was fully active, displaying 102% activity of that exhibited by S₁wt when tested at 25 µg mL^-1. Because there was no significant reduction of biological activity, we substituted all three variable residues in loop 5 with corresponding S₃ residues. This triple mutant K63R/E64K/T65G retained close to full activity when applied to pollen at concentrations >75 µg mL^-1 (83% that of S₁wt at 75 µg mL^-1). However, as shown in Figure 3B, at 25 µg mL^-1, biological activity was significantly reduced (P = 0.05) to 58% of S₁wt. This reduction suggests that although residues in loop 5 of the S₁ protein may influence the activity of the protein, they are not solely responsible for determining S allele specificity of the protein. It is possible that the loss of activity is simply attributable to these substitutions having a nonspecific adverse influence on protein refolding. Although this explanation cannot be ruled out, it is considered unlikely for two reasons. First, the residues are substituted with corresponding residues from the S₃ protein sequence, and second, they are located in a hydrophilic surface loop, where they are less likely to have a detrimental effect on protein folding.

The Majority of Mutations in Surface Loops Have No Biological Effect
To further explore a role in pollen recognition for residues within the remaining predicted loops (loops 1, 2, 3, 4, and 7), we constructed a series of single and double mutations to the S₁ sequence, targeting each loop sequentially: in loop 1, N15H; in loop 2, D27H; and in loop 3, two double mutants were made. Thr-36, which is predicted to lie at the edge of the loop, was substituted for Ser, and Ser-37 was substituted for Asp (T36S/S37D); in addition, H39Q/D40E was created. Loop 4 differed from the other surface loops in that the secondary structure prediction suggested that the two conserved Phe residues located within this loop may produce a shallow hydrophobic pocket. To test the significance of this, we substituted Phe-49 for Met (F49M), which has similar hydrophobicity but differs significantly in structure, because it does not possess an aromatic side chain. In loop 7, Asp-99 was substituted for His (D99H).

The mutant S proteins with substitutions in loops 1, 2, 3, 4, and 7, when tested for their ability to specifically inhibit S₁ pollen, were demonstrated to be biologically active, as shown in Figure 3B. When tested at a concentration of 75 µg mL^-1, they all exhibited activity against S₁ pollen indistinguishable from S₁wt. With the exception of the loop 2 mutant (D27H), the result was essentially the same at a concentration of 25 µg mL^-1. Mutant D27H, however, did exhibit a significant reduction (P = 0.05) in activity to ~78% that of S₁wt. This observation is reminiscent of that found for the loop 5 triple mutation, and for the same reasons, its significance is unclear. Overall, the mutations presented in this section had a minor or no effect on the ability of the S₁ stigmatic protein to inhibit S₁ pollen. On the basis of our data, the S₁ protein can tolerate mutations in these remaining surface loops without detriment to its biological activity.

A Conserved Cys Residue Is Essential for Activity of S₁
A notable feature of the poppy S protein sequence alignment is the presence of four absolutely conserved cysteine residues, relative to positions 23, 57, 82, and 86 of the S₁ sequence (see Figure 1). We previously suggested that these may form two disulfide bridges that are essential in maintaining tertiary structure (Walker et al. 1996 ; J.P. Ride, unpublished results). To investigate the importance of these conserved Cys residues, we introduced a further mutation into the loop 5 triple mutant K63R/E64K/T65G/C82Y. As shown in Figure 3B, this substitution (C82Y) caused complete inactivation of the protein, and no S₁ pollen inhibition activity was detectable. Because this mutant has virtually normal activity, we interpret this result to implicate Cys-82 as being of vital importance to the activity of the protein. This is most probably due to its influence on the tertiary structure of the S protein.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

In the studies described here, we have begun to define the role of regions in the stigmatic S proteins of P. rhoeas that are required for allele-specific inhibition of incompatible pollen. Our studies indicate that variability in amino acid sequence between stigmatic S proteins is likely to be responsible for determining S allele specificity rather than glycosylation. This is based on two key observations. First, not all poppy S proteins, for example, S₃ (Walker et al. 1996 ), are glycosylated. Second, there is no detectable difference in biological activity between the S₁ protein extracted from poppy stigmas (which is glycosylated) and the nonglycosylated recombinant S protein expressed in E. coli (Foote et al. 1994 ; Franklin-Tong et al. 1995 ).

Residues Located in Predicted Hydrophilic Loop 6 Participate in Pollen Recognition
In contrast to the S RNases of the Solanaceae (Ioerger et al. 1990 , Ioerger et al. 1992 ), the sequence alignments between the poppy S proteins do not readily identify defined hypervariable domains. Instead, short blocks of identity are interspersed with variable residues. The available sequence information from the cloned Papaver S alleles (Foote et al. 1994 ; Walker et al. 1996 ; Kurup et al. 1998 ) suggests that it is highly probable that residues determining allele specificity will be scattered throughout the primary sequence.

Only one amino acid site has been identified that is polymorphic across available S alleles sequenced thus far. This site is Asp-79 in the S₁ protein and the corresponding His-80 in the S₈ protein (Foote et al. 1994 ; Walker et al. 1996 ; Kurup et al. 1998 ). This site is predicted to lie within a hydrophilic loop (loop 6) and, on the basis of the current evidence, appears to be flanked by conserved, charged amino acids. The studies described here demonstrate that mutation of this site resulted in the complete loss of biological activity. This clearly implies that the amino acid found in the different S proteins at this site plays a crucial role in pollen recognition in these alleles; because it is variable between S alleles, it may participate in defining allelic specificity. However, although the ability to reject S₁ pollen can be affected by mutation of this single residue, it cannot be the sole determinant of S allele specificity for two key reasons. First, the mutations in S₁, D79H (a substitution from S₁ to S₈), and H80D (a substitution from S₈ to S₁) did not confer a new allelic specificity on either mutant protein (data not shown). That is, simply switching the S₁ residue at this site for its S₈ counterpart, and vice versa, did not change specificity. Second, population genetics studies of SI in P. rhoeas predict that there are ~65 S alleles in the species (Lane and Lawrence 1993 ). Because there are only 20 amino acids, allelic variation cannot be specified by a single polymorphic site within the protein. It would seem likely that specificity is determined by several residues that, in combination, specify an allelic identity.

Despite introducing a series of single, double, and triple mutations into the remaining predicted surface loops, we have detected no dramatic loss of S protein inhibitory activity comparable with that seen with the loop 6 mutations. This clearly demonstrates that mutations to these residues do not affect S-specific biological activity. Therefore, we conclude that in the case of S₁, these residues do not contribute to the S allele–specific recognition of the pollen. Our results also illustrate that the S proteins can tolerate mutations within these regions and that alterations at these sites presumably do not have any significant effect on protein folding.

The importance of maintaining the integrity of the tertiary structure is clearly illustrated by the mutation to one of the four conserved cysteine residues. When Cys-82 was replaced with a Tyr residue (C82Y), the mutant protein was totally inactive. Although there is as yet no direct evidence, we believe that Cys-82 forms a disulfide bridge with Cys-86, whereas Cys-24 and Cys-57 form another disulfide bridge. This hypothesis is based on the sequence alignments of the recently identified gene family of S protein homologs (SPH) in Arabidopsis (J.P. Ride, F.C.H. Franklin, and D.F. Marshall, unpublished data). To date, 17 SPH sequences have been identified. These all encode small (~15 kD), basic, secreted proteins that are predicted to adopt secondary structures that are very similar to those of the S proteins. All SPH sequences possess two Cys residues that align with Cys-24 and Cys-57 of S₁. This strongly suggests that this pair of Cys residues forms one of the two predicted disulfide bridges in the poppy S proteins. If this is the case, then Cys-82 and Cys-86 potentially would form a short hairpin structure adjacent to loop 6. The loss of pollen inhibitory activity caused by the substitution of one of these residues (e.g., C82Y) suggests that this hairpin structure may itself participate in pollen recognition.

Further evidence that amino acid residues in loop 6 play an important role in pollen recognition and the interaction of the S proteins with incompatible pollen was revealed by the mutation of the Asp residues at positions 77 and 78 in loop 6. These residues are conserved across all poppy S alleles characterized to date. In the case of S₁ at least, it seems that the negative charges present in the carboxyl side chains of Asp-77 and Asp-78 are of particular importance. The evidence is that these residues can be replaced with equally acidic Glu residues with little or no effect on biological activity, whereas substitution with more basic His residues effectively abolishes function. It is conceivable that these charged amino acids stabilize the interaction with the pollen component or perhaps have a role in maintaining the structural integrity of the S₁ protein in the region of loop 6.

An important goal of future studies is to determine the tertiary structure of an S protein. This will enable the identification of amino acids located at the protein surface or at other positions that are available for interaction with the pollen component. It will also identify amino acids that may be at distant positions in the primary sequence but that are juxtaposed to the same region when the protein adopts its active conformation. Such information is likely to be essential if we are to define precisely which amino acid residues specify a particular allelic identity of an S protein. It may also provide some explanations for the minor reductions in activity that we have observed with some of the mutant proteins described. It is possible that in the tertiary form of the S protein, these residues are located close to key amino acids and that the mutations have had an indirect, adverse effect on the ability of crucial residues to interact with the pollen component.

These studies have enabled us to define several residues within the stigmatic S₁ protein that are required for allele-specific inhibition of S₁ pollen. Furthermore, they have enabled us to rule out the participation of several other residues. Taken together, these data strongly indicate that the region of the S₁ protein corresponding to surface loop 6 is a key element in the recognition of incompatible pollen.

	METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Plant Material and Cross-Classification of S Alleles
The families of the Shirley variety of Papaver rhoeas were as previously described by Foote et al. 1994 . Plants carrying S alleles distinct from S₁, S₃, and S₈ were identified using a combination of cross-classification by pollination (see Campbell and Lawrence 1981 and Lane and Lawrence 1993 for full experimental details) and protein gel blotting with immunodetection using antisera raised to recombinant S₁ and S₃ proteins.

Site-Directed Mutagenesis
The wild-type S₁ protein (S₁wt) was originally produced from cells carrying the vector pPRS100 as described by Foote et al. 1994 . Two methods were used to introduce amino acid substitutions into the S₁wt protein. Specific mutations were introduced into loop 6 of the S₁ amino acid sequence by using the polymerase chain reaction–based method described by Higuchi et al. 1988 . DNA amplification was performed using mismatched primers to incorporate the required mutation from a DNA template of the gene for the mature S₁ polypeptide ligated into pBluescript SK+ (Stratagene, La Jolla, CA). The mutated gene was then cloned into the expression vector pMS119 (Lessl et al. 1992 ) at the NdeI and HindIII restriction sites. Mutations to other loops were performed by the method described by Stanssens et al. 1989 using the S₁ gene cloned into the pMac5-8 phagemid vector via the BamHI and HindIII restriction sites. The mutations were incorporated using mismatched primers and the mutated gene cloned into the vector pMS119, as given above. To facilitate protein expression, a translation initiation codon was incorporated at the start of the gene sequences.

Nucleotide Sequencing
Nucleotide sequencing was performed both manually by the dideoxy method (Sanger et al. 1977 ) using the T7 Sequenase 2.0 kit (Amersham) and by automated sequencing (Alta BioScience, University of Birmingham, UK).

Database Searches and Structural Predictions
For preliminary sequence analysis, we used software from the Genetics Computer Group (Madison, WI). More detailed comparisons were performed with MACAW using BLOSUM 62 comparison tables (Schuler et al. 1991 ; Lawrence et al. 1993 ). The predicted hydrophilicity plots for S₁ and S₃ were determined according to the method of Kyte and Doolittle 1982 using a window of seven residues. Secondary structures of the S proteins were predicted using the PHD program (Rost and Sander 1993 , Rost and Sander 1994 ; Rost et al. 1994 ).

Expression of Recombinant Proteins
Cells containing wild-type and mutant gene constructs were induced with 1 mM isopropyl ß-D-thiogalactopyranoside for 4 hr, resulting in high-level expression of recombinant proteins that accumulated as insoluble inclusion bodies. Cells were harvested by centrifugation (5000g at 4°C for 10 min) and resuspended in lysis buffer containing 50 mM Tris-Cl, 100 mM NaCl, and 1 mM EDTA, pH 8.0. They were disrupted by using lysozyme and deoxycholate and by sonication: inclusion bodies were collected by centrifugation and washed in lysis buffer (Sambrook et al. 1989 ). The resulting proteins were dissolved in lysis buffer containing 6 M guanidine–HCl, 0.5 M 2-mercaptoethylamine, and 0.125 mM phenylmethylsulfonyl fluoride, pH 8.0, for 1 hr at room temperature and refolded by rapid dilution in 100 mM Tris containing 50 mM L-arginine, 2 mM EDTA, 10 mM cystamine, and 5% (v/v) glycerol, pH 8.0, at 14°C to a final concentration of 100 µg mL^-1. After incubation overnight, the solution was dialyzed against 50 mM Tris-HCl and 100 mM NaCl, pH 8.0, and concentrated against the same buffer containing 12.5% polyethylene glycol 6000. The final protein solution was stored at -80°C. Protein concentration was estimated using the Coomassie Brilliant Blue R 250 dye-binding method (Bradford 1976 ).

Testing the Biological Activity of the Mutant S Proteins
The biological activity of mutant proteins was determined using an adaptation of our in vitro bioassay (Franklin-Tong et al. 1988 ; Foote et al. 1994 ). Recombinant mutant proteins were assayed at several concentrations based on the dose–response curve. The concentrations used for testing the biological activity of the mutant protein were significantly above levels required to detect S-specific pollen inhibition activity to ensure that lack of activity could not be attributed to using a concentration that might be close to the threshold of activity. Protein activity was initially tested at 25 and 75 µg mL^-1 if reduced or no activity was detected. Additional tests were performed at 75 and 150 µg mL^-1, which are concentrations significantly above levels required to detect S-specific pollen inhibition by S₁wt. To detect any nonspecific inhibition, we also tested proteins against pollen of a fully compatible S genotype.

To assay for biological activity, we dialyzed proteins against liquid germination medium containing 1% (w/v) H₃BO₃, 1% (w/v) KNO₃, 1% (w/v) Mg(NO₃)₂·6H₂O, 3.1% (w/v) CaCl₂·2H₂O, and 13.5% (w/v) sucrose. Half-compatible pollen carrying the S₁ allele (e.g., S₁S₄ or S₁S₃) and fully compatible pollen carrying S₂S₄ or S₄S₆ alleles were allowed to prehydrate on multiwell slides for 30 min at 25°C. Aliquots (20 µL per well) of germination medium and appropriately diluted S proteins were applied to the wells, and pollen was allowed to germinate for 2 hr at 25°C. Experiments were randomized on the slides and included three replicates of each treatment. Wells were scored for germinated and nongerminated pollen (n >100 per well). The percentage of germination of pollen that was treated with germination medium (no protein) and S₁wt and the mutant S protein was calculated. Data plotted represent the relative activity of the mutant S proteins expressed as a percentage of S₁wt activity. Minitab 10.51 Xtra (Minitab Inc., State College, PA) was used to perform a one-way analysis of variance of the percentage of germination after angular transformation of the data.

	FOOTNOTES

¹ These authors contributed equally to this work.
² Current address: Faculty of Bioresources, Mie University, 1515 Kamihama, Tsu 514, Japan.

	ACKNOWLEDGMENTS

We thank Antony Jones for technical assistance and John Martin and colleagues for horticultural support. This research was supported by the Biotechnology and Biological Sciences Research Council (BBSRC) and Japan Society for the Promotion of Science (J.S.P.S.).

Received June 17, 1998; accepted August 5, 1998.

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Bradford, M.M. (1976) A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein–dye binding. Anal. Biochem. 72:248-254[CrossRef][ISI][Medline].

Campbell, J.M., and Lawrence, M.J. (1981) The population genetics of the self-incompatibility polymorphism in Papaver rhoeas L. The number and distribution of S-alleles in families from three localities. Heredity 46:69-79.

Foote, H.C.C., Ride, J.P., Franklin-Tong, V.E., Walker, E.A., Lawrence, M.J., and Franklin, F.C.H. (1994) Cloning and expression of a novel self-incompatibility (S-) gene from Papaver rhoeas L. Proc. Natl. Acad. Sci. USA 91:2265-2269[Abstract/Free Full Text].

Franklin, F.C.H., Lawrence, M.J., and Franklin-Tong, V.E. (1995) Cell and molecular biology of self-incompatibility in flowering plants. Int. Rev. Cytol. 158:1-64.

Franklin-Tong, V.E., Lawrence, M.J., and Franklin, F.C.H. (1988) An in vitro bioassay for the stigmatic product of the self-incompatibility gene in Papaver rhoeas L. New Phytol. 110:109-118[CrossRef].

Franklin-Tong, V.E., Ride, J.P., Read, N.D., Trewavas, A.J., and Franklin, F.C.H. (1993) The self-incompatibility response in Papaver rhoeas is mediated by cytosolic free calcium. Plant J. 4:163-177.

Franklin-Tong, V.E., Ride, J.P., and Franklin, F.C.H. (1995) Recombinant stigmatic self-incompatibility (S-) protein elicits a Ca²⁺ transient in pollen of Papaver rhoeas.. Plant J. 8:299-307[CrossRef][ISI].

Franklin-Tong, V.E., Hackett, G., and Hepler, P.K. (1997) Ratio-imaging of Ca²⁺_i in the self-incompatibility response in pollen tubes of Papaver rhoeas.. Plant J. 12:1375-1386[CrossRef][ISI].

Hearn, M.J., Franklin, F.C.H., and Ride, J.P. (1996) Identification of a membrane glycoprotein in pollen of Papaver rhoeas which binds stigmatic self-incompatibility (S-) proteins. Plant J. 9:467-475[CrossRef].

Higuchi, R., Krummel, B., and Saiki, R.K. (1988) A general method of in vitro preparation and specific mutagenesis of DNA fragments: Study of protein and DNA interactions. Nucleic Acids Res. 16:7351-7363[Abstract/Free Full Text].

Ioerger, T.R., Clark, A.G., and Kao, T.-h. (1990) Polymorphism at the self-incompatibility locus in Solanaceae predates speciation. Proc. Natl. Acad. Sci. USA 87:9732-9735[Abstract/Free Full Text].

Ioerger, T.R., Gohlke, J.R., Xu, B., and Kao, T.-h. (1992) Primary structural features of the self-incompatibility proteins in the Solanaceae. Sex. Plant Reprod. 4:81-87.

Kao, T.-h., and McCubbin, A.G. (1996) How flowering plants discriminate between self and non-self pollen to prevent inbreeding. Proc. Natl. Acad. Sci. USA 93:12059-12065[Abstract/Free Full Text].

Kurup, S., Ride, J.P., Jordan, N.D., Fletcher, E.G., Franklin-Tong, V.E., and Franklin, F.C.H. (1998) Identification and cloning of related self-incompatibility S-genes in Papaver rhoeas and Papaver nudicaule.. Sex. Plant Reprod. in press.

Kyte, J., and Doolittle, R.F. (1982) A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157:105-132[CrossRef][ISI][Medline].

Lane, M.D., and Lawrence, M.J. (1993) The population-genetics of the self-incompatibility polymorphism in Papaver rhoeas. VII. The number of S-alleles in the species. Heredity 71:596-602.

Lawrence, C.E., Altschul, S.F., Boguski, M.S., Liu, J.S., Neuwald, A.F., and Wooton, J.C. (1993) Detecting subtle sequence signals: A Gibbs sampling strategy for multiple alignment. Science 262:208-214[Abstract/Free Full Text].

Lawrence, M.J., Afzal, M., and Kenrick, J. (1978) The genetical control of self-incompatibility in Papaver rhoeas L. Heredity 40:239-285.

Lessl, M., Balzer, D., Lurz, R., Waters, V., Guiney, D.G., and Lanka, E. (1992) Dissection of IncP conjugative plasmid transfer: Definition of the transfer region Tra2 by mobilization of the Tra1 region in trans.. J. Bacteriol. 174:2493-2500[Abstract/Free Full Text].

Matton, D.P., Maes, O., Laublin, G., Xike, Q., Bertrand, C., Morse, D., and Cappadocia, M. (1997) Hypervariable domains of self-incompatibility RNases mediate allele-specific pollen recognition. Plant Cell 9:1757-1766[Abstract].

Rost, B., and Sander, C. (1993) Prediction of protein structure at better than 70% accuracy. J. Mol. Biol. 232:584-599[CrossRef][ISI][Medline].

Rost, B., and Sander, C. (1994) Combining evolutionary information and neural networks to predict protein secondary structure. Proteins 19:55-72[CrossRef][ISI][Medline].

Rost, B., Sander, C., and Schneider, R. (1994) PHD—An automatic mail server for protein secondary structure prediction. Comput. Appl. Biosci. 10:53-60[Abstract/Free Full Text].

Rudd, J.J., Franklin, F.C.H., Lord, J.M., and Franklin-Tong, V.E. (1996) Increased phosphorylation of a 26-kD pollen protein is induced by the self-incompatibility response in Papaver rhoeas.. Plant Cell 8:713-724[Abstract].

Rudd, J.J., Franklin, F.C.H., Lord, J.M., and Franklin-Tong, V.E. (1997) Ca²⁺-independent phosphorylation of a 68-kD protein is stim-ulated by the self-incompatibility response. Plant J. 12:507-514[CrossRef].

Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual, 2nd ed. (Plainview, NY: Cold Spring Harbor Laboratory Press).

Sanger, F., Nicklen, S., and Coulsen, A.R. (1977) DNA sequencing with chain termination inhibitors. Proc. Natl. Acad. Sci. USA 74:5463-5467[Abstract/Free Full Text].

Schuler, G.D., Altschul, S.F., and Lipman, D.G. (1991) A workbench for multiple alignment construction and analysis. Proteins Struct. Funct. Genet. 9:180-190[CrossRef][ISI][Medline].

Stanssens, P., Opsomer, C., McKearn, Y.M., Kramer, W., Zabeau, M., and Fritz, H.-J. (1989) Efficient oligonucleotide-directed construction of mutations in expression vectors by the gapped duplex DNA method using alternating selectable markers. Nucleic Acids Res. 17:4441-4454[Abstract/Free Full Text].

Verica, J.A., McCubbin, A.G., and Kao, T.-h. (1998) Are the hypervariable regions of S RNases sufficient for allele-specific recognition of pollen? Plant Cell 10:314-316.

Walker, E.A., Ride, J.P., Kurup, S., Franklin-Tong, V.E., Lawrence, M.J., and Franklin, F.C.H. (1996) Molecular analysis of two functional homologues of the S₃ allele of the Papaver rhoeas incompatibility gene isolated from different populations. Plant Mol. Biol. 30:983-994[CrossRef][ISI][Medline].

Zurek, D.M., Mou, B.Q., Beecher, B., and McClure, B.A. (1997) Exchanging sequence domains between S-RNases from Nicotiana alata disrupts pollen recognition. Plant J. 11:797-808[CrossRef][ISI][Medline].

This article has been cited by other articles:

				N. S. Poulter, S. Vatovec, and V. E. Franklin-Tong Microtubules Are a Target for Self-Incompatibility Signaling in Papaver Pollen Plant Physiology, March 1, 2008; 146(3): 1358 - 1367. [Abstract] [Full Text] [PDF]

				S. Li, J. Samaj, and V. E. Franklin-Tong A Mitogen-Activated Protein Kinase Signals to Programmed Cell Death Induced by Self-Incompatibility in Papaver Pollen Plant Physiology, September 1, 2007; 145(1): 236 - 245. [Abstract] [Full Text] [PDF]

				S. G. Thomas, S. Huang, S. Li, C. J. Staiger, and V. E. Franklin-Tong Actin depolymerization is sufficient to induce programmed cell death in self-incompatible pollen J. Cell Biol., July 17, 2006; 174(2): 221 - 229. [Abstract] [Full Text] [PDF]

				J. D. Higgins, E. Sanchez-Moran, S. J. Armstrong, G. H. Jones, and F. Chris. H. Franklin The Arabidopsis synaptonemal complex protein ZYP1 is required for chromosome synapsis and normal fidelity of crossing over Genes & Dev., October 15, 2005; 19(20): 2488 - 2500. [Abstract] [Full Text] [PDF]

				J. D. Higgins, S. J. Armstrong, F. C. H. Franklin, and G. H. Jones The Arabidopsis MutS homolog AtMSH4 functions at an early step in recombination: evidence for two classes of recombination in Arabidopsis Genes & Dev., October 15, 2004; 18(20): 2557 - 2570. [Abstract] [Full Text] [PDF]

				A. Geitmann, W. McConnaughey, I. Lang-Pauluzzi, V. E. Franklin-Tong, and A. M. C. Emons Cytomechanical Properties of Papaver Pollen Tubes Are Altered after Self-Incompatibility Challenge Biophys. J., May 1, 2004; 86(5): 3314 - 3323. [Abstract] [Full Text] [PDF]

				R. Mercier, Susan. J. Armstrong, C. Horlow, N. P. Jackson, C. A. Makaroff, D. Vezon, G. Pelletier, G. H. Jones, and F. C. H. Franklin The meiotic protein SWI1 is required for axial element formation and recombination initiation in Arabidopsis Development, July 15, 2003; 130(14): 3309 - 3318. [Abstract] [Full Text] [PDF]

				B. N. Snowman, D. R. Kovar, G. Shevchenko, V. E. Franklin-Tong, and C. J. Staiger Signal-Mediated Depolymerization of Actin in Pollen during the Self-Incompatibility Response PLANT CELL, October 1, 2002; 14(10): 2613 - 2626. [Abstract] [Full Text] [PDF]

				S. J. Armstrong, A. P. Caryl, G. H. Jones, and F. C. H. Franklin Asy1, a protein required for meiotic chromosome synapsis, localizes to axis-associated chromatin in Arabidopsis and Brassica J. Cell Sci., September 15, 2002; 115(18): 3645 - 3655. [Abstract] [Full Text] [PDF]

				M. K. Uyenoyama, Y. Zhang, and E. Newbigin On the Origin of Self-Incompatibility Haplotypes: Transition Through Self-Compatible Intermediates Genetics, April 1, 2001; 157(4): 1805 - 1817. [Abstract] [Full Text]

				R. Dixit and J. B. Nasrallah Recognizing Self in the Self-Incompatibility Response Plant Physiology, January 1, 2001; 125(1): 105 - 108. [Full Text]

				A. Geitmann, B. N. Snowman, A. M. C. Emons, and V. E. Franklin-Tong Alterations in the Actin Cytoskeleton of Pollen Tubes Are Induced by the Self-Incompatibility Reaction in Papaver rhoeas PLANT CELL, July 1, 2000; 12(7): 1239 - 1252. [Abstract] [Full Text]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via ISI Web of Science (40)