• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of plntcellLink to Publisher's site
Plant Cell. Feb 2001; 13(2): 255–272.
PMCID: PMC102241

Domain Swapping and Gene Shuffling Identify Sequences Required for Induction of an Avr-Dependent Hypersensitive Response by the Tomato Cf-4 and Cf-9 Proteins

Abstract

The tomato Cf-4 and Cf-9 genes confer resistance to infection by the biotrophic leaf mold pathogen Cladosporium. Their protein products induce a hypersensitive response (HR) upon recognition of the fungus-encoded Avr4 and Avr9 peptides. Cf-4 and Cf-9 share >91% sequence identity and are distinguished by sequences in their N-terminal domains A and B, their N-terminal leucine-rich repeats (LRRs) in domain C1, and their LRR copy number (25 and 27 LRRs, respectively). Analysis of Cf-4/Cf-9 chimeras, using several different bioassays, has identified sequences in Cf-4 and Cf-9 that are required for the Avr-dependent HR in tobacco and tomato. A 10–amino acid deletion within Cf-4 domain B relative to Cf-9 was required for full Avr4-dependent induction of an HR in most chimeras analyzed. Additional sequences required for Cf-4 function are located in LRRs 11 and 12, a region that contains only eight of the 67 amino acids that distinguish it from Cf-9. One chimera, with 25 LRRs that retained LRR 11 of Cf-4, induced an attenuated Avr4-dependent HR. The substitution of Cf-9 N-terminal LRRs 1 to 9 with the corresponding sequences from Cf-4 resulted in attenuation of the Avr9-induced HR, as did substitution of amino acid A433 in LRR 15. The amino acids L457 and K511 in Cf-9 LRRs 16 and 18 are essential for induction of the Avr9-dependent HR. Therefore, important sequence determinants of Cf-9 function are located in LRRs 10 to 18. This region contains 15 of the 67 amino acids that distinguish it from Cf-4, in addition to two extra LRRs. Our results demonstrate that sequence variation within the central LRRs of domain C1 and variation in LRR copy number in Cf-4 and Cf-9 play a major role in determining recognition specificity in these proteins.

INTRODUCTION

A major goal in plant pathology is to determine the molecular mechanism of pathogen perception by plants. In gene-for-gene interactions, it has been postulated that plant disease resistance (R) genes encode receptors for the products of pathogen-encoded avirulence (Avr) genes (reviewed in Ellis et al., 2000). The tomato Cf-2, Cf-4, Cf-5, and Cf-9 genes confer resistance to the biotrophic leaf mold pathogen Cladosporium through recognition of different fungal-encoded Avr proteins (Avr2, Avr4, Avr5, and Avr9, respectively). Cf genes encode extracytoplasmic membrane-anchored glycoproteins (Piedras et al., 2000) composed predominantly of leucine-rich repeats (LRRs) and a short cytoplasmic domain that lacks an obvious signaling function (Thomas et al., 1998).

Most plant R genes that have been characterized encode proteins that contain LRRs (Ellis et al., 2000). LRR proteins are thought to have evolved independently in many organisms and are involved in protein–protein interactions (Kobe and Deisenhofer, 1994; Kajava, 1998). It has been postulated that the solvent-exposed amino acids of a conserved β-strand/β-turn structural motif are the major determinants of recognition specificity in this class of proteins (Kobe and Deisenhofer, 1994; Jones and Jones, 1997; Kajava, 1998).

The analysis of Cf-9 paralogs (Hcr9s, for homologs of Cladosporium resistance gene 9) at several loci on the short arm of chromosome 1 (Parniske et al., 1997, 1999; Thomas et al., 1997; Parniske and Jones, 1999) has identified hypervariable sequences in these highly homologous genes. Consistent with their proposed role as determinants of recognition specificity, most variation between Hcr9s is in sequences encoding the putative solvent-exposed amino acids of the LRR β-strand/β-turn structural motif. A similar analysis of Cf-2 homologs (Hcr2s) at the Cf-2/Cf-5 locus on chromosome 6 was also performed (Dixon et al., 1998). From these analyses, it was concluded that sequences encoding the putative solvent-exposed amino acids of their N-terminal LRRs, together with variation in their N-terminal LRR copy number, determine recognition specificity in Cf proteins (Dixon et al., 1998; Thomas et al., 1998). However, the molecular mechanism of Avr protein perception has not been determined (Dixon et al., 2000).

We have shown previously that Cf-4 and Cf-9 share >91% amino acid identity (Thomas et al., 1997). Some of the amino acids that distinguish Cf-4 and Cf-9 are located in their N-terminal domains A and B, but most correspond to putative solvent-exposed residues of the LRR β-strand/ β-turn structural motif (see Figure 1). Cf-4 and Cf-9 are distinguished further by their LRR copy number (25 and 27 LRRs, respectively). To determine which of these structural differences are important for Cf-4 and Cf-9 function, we tested Cf-4/Cf-9 chimeras for their ability to induce an Avr-dependent hypersensitive response (HR).

Figure 1.
Predicted Amino Acid Sequence of the Tomato Cf-4 Protein (Thomas et al., 1997).

An analysis of sequences that determine recognition specificity in Cf-4 and Cf-9 is possible because they discriminate between two Cladosporium Avr determinants (Avr4 and Avr9) that have been cloned (Van den Ackerveken et al., 1992; Joosten et al., 1994). When Cf-4 or Cf-9 are expressed in tomato or tobacco, they induce an HR in the presence of the cognate Avr protein. This has facilitated the development of rapid and reliable assays of Cf-4/Cf-9 gene function in tomato and several tobacco species (Hammond-Kosack et al., 1994, 1995, 1998; Joosten et al., 1997; Thomas et al., 1997, 2000; Kamoun et al., 1999; Van der Hoorn et al., 2000).

We used a variety of these bioassays to analyze defined Cf-4/Cf-9 domain swaps and chimeras generated by polymerase chain reaction (PCR)–mediated gene shuffling (Stemmer, 1994; Crameri et al., 1998). Gene shuffling is a powerful technique for generating large numbers of chimeras whose protein products can exhibit dramatically increased enzymatic activity (Stemmer, 1994; Crameri et al., 1998). This report describes the use of gene shuffling for the functional analysis of plant genes, which has enabled us to identify important sequence determinants of Cf-4 and Cf-9 function. Our data demonstrate that sequence changes within the central LRRs of domain C1 together with variation in LRR copy number are important determinants of recognition specificity in this unique class of plant R proteins.

RESULTS

Construction and Analysis of Cf-4/Cf-9 Chimeras

Domains A and B and the N-terminal LRRs in domain C1 of Cf-4 and Cf-9 contain all of the amino acids that distinguish these proteins (Figure 1). The clones Cf-4DS and Cf-9DS, which encode the wild-type Cf-4 and Cf-9 proteins (see Methods), contain restriction sites that facilitated the exchange of sequences encoding domains A and B and variable numbers of N-terminal LRRs. Alternately, chimeras were prepared by PCR-mediated gene shuffling. The chimeras were tested for their ability to induce an Avr4- or Avr9-dependent HR by using one or more of the following bioassays:

  1. Cf-4/Cf-9 chimeras cloned in T-DNA vectors were tested for their ability to induce an Avr-dependent HR by Agrobacterium-mediated transient expression in leaves of tobacco or Nicotiana benthamiana expressing 35S:Avr4 or 35S:Avr9 (Thomas et al., 2000; Van der Hoorn et al., 2000).
  2. Chimeras were stably expressed in transgenic tobacco. Tobacco and tomato plants expressing Cf-4 or Cf-9 induce an F1 seedling lethal phenotype when crossed to lines expressing 35S:Avr4 or 35S:Avr9 (Hammond-Kosack et al., 1994, 1998; Thomas et al., 1997, 2000).
  3. Transgenic tobacco plants were tested for their ability to resist infection by recombinant potato virus X (PVX). Plants expressing Cf-4 or Cf-9 exhibit necrotic lesions on inoculated leaves when inoculated with PVX:Avr4 or PVX:Avr9, respectively, and are resistant to virus infection (Kamoun et al., 1999; Thomas et al., 2000).
  4. Transgenic tomato plants expressing specific chimeras were tested for resistance to infection by Cladosporium.
  5. Transgenic tomato plants were infected with recombinant PVX-expressing Avr4 or Avr9 (PVX:Avr4 and PVX:Avr9) that induce a systemic necrosis in plants expressing Cf-4 or Cf-9 (Hammond-Kosack et al., 1995; Joosten et al., 1997; Thomas et al., 1997).

The predicted amino acid sequences of the Cf-4/Cf-9 chimeras described below are shown in Figures 2 and and66 together with a summary of bioassay data that summarize their capacity to induce an Avr-dependent HR.

Figure 2.
Amino Acid Sequences of Chimeras Used to Identify Functional Domains in Cf-4 and Cf-9.
Figure 6.
Schematic Representation of Cf-4/Cf-9 Shuffling and Protein Sequences of Chimeras That Induce an Avr4- or Avr9-Dependent HR.

Sequences within Cf-4 Domains A and B or Amino Acid E84 in LRR 1 Are Required for Induction of Avr4-Dependent HR in Tobacco and Tomato

Transgenic tobacco plants expressing different chimeras were tested for their ability to induce a seedling lethal phenotype after crossing to lines expressing 35S:Avr4 or 35S:Avr9 (Table 1). The Cf-4[9CB] chimera contains sequences encoding domains A and B of Cf-9 and the putative solvent-exposed variant amino acid A94 in LRR 1 (Figure 2A). The F1 progeny of a Cf-4DS × 35S:Avr4 cross-segregated for the wild type and a seedling-lethal phenotype (Thomas et al., 2000). In the progeny of Cf-4[9CB] × 35S:Avr4 crosses, half appeared normal and were indistinguishable from progeny of a Cf-4[9CB] × 35S:Avr9 cross (Figure 3A). The remaining seedlings were stunted and produced only two to three chlorotic or necrotic leaves that eventually died (Figure 3A). The self-progeny of Cf-4[9CB] transgenic plants were also susceptible to PVX:Avr4 infection (Table 1). These data suggest the Avr4-dependent HR induced by this chimera is attenuated compared with Cf-4DS.

Figure 3.
Identification of Functional Domains in Cf-4.
Table 1.
Characterization of Cf-4/Cf-9 Chimeras in Transgenic Nicotiana tabacum Plants

Several Cf-4[9CB] transgenic tomato plants were also analyzed. In contrast with progeny from Cf-4DS × 35S:Avr4 crosses (Thomas et al., 1997), the progeny from Cf-4[9CB] × 35S:Avr4 crosses were phenotypically normal at the seedling stage and grew to maturity. Approximately half of the plants at the six- to seven-leaf stage exhibited chlorotic and necrotic sectors on their oldest leaves (Figures 3B and 3C). This phenotype was due to weak activation of an Avr4-dependent response because it was not observed in progeny from Cf-4[9CB] × 35S:Avr9 crosses (Figure 3B). Tomato Cf-4[9CB] transgenic plants were also susceptible to infection by Cladosporium race 5 that expresses Avr4 and Avr9 (Figure 3F).

When the progeny of Cf-4[9CB] transgenic plants were inoculated with PVX:Avr4 at the two-leaf stage, only a weak response to PVX:Avr4 was observed. This response was manifested as leaf epinasty and chlorosis in systemically infected leaves, and occasionally as small necrotic lesions (Figure 3D). No systemic necrosis was observed, as was observed in Cf-4DS controls (results not shown), and virus-infected plants grew to maturity (results not shown).

Together, these data suggest that Cf-4[9CB] triggers a greatly attenuated Avr4-dependent HR. Sequences within Cf-4 domain A, the putative signal peptide sequence (Jones and Jones, 1997), domain B, or amino acid E84 in LRR 1 are required for full induction of an Avr4-dependent HR in tobacco and tomato, and for resistance to Cladosporium infection in tomato.

10–Amino Acid Deletion in Cf-4 Domain B Is Required for Full Avr4-Dependent HR

Two additional chimeras were tested to determine which of the sequences described above are required for full induction of the Avr4-dependent HR. Sequences encoding the additional 10 amino acids in Cf-9 domain B were deleted from Cf-9DS (amino acids I56 to Y65), as described in Methods. Sequences encoding Cf-9 domain A, the modified domain B, and A94 from LRR 1 were inserted into the corresponding region in Cf-4 to generate Cf-4[Δ9B] (Figure 2A). In Cf-4[+10B], DNA encoding the additional 10 amino acids in Cf-9 domain B (IRTYVDIQSY) was inserted into Cf-4DS between sequences encoding amino acids D55 and R56 (Figure 2A).

Both chimeras were tested for their ability to induce an Avr4-dependent HR in Nicotiana benthamiana. Cf-4[Δ9B] induced an HR that was only slightly weaker than the Cf-4DS control (Figure 3H). Cf-4[+10B] induced a weak Avr4-dependent HR similar to that induced by Cf-4[9CB] (Figure 3I). Together, these data demonstrate that the four variant amino acids within Cf-4 domain B, the variant amino acids within domain A, and amino acid E84 in LRR 1 can be substituted simultaneously with the corresponding sequences from Cf-9 without compromising Cf-4 function. Therefore, at least in the chimeras tested here, the absence of a 10–amino acid sequence in Cf-4 domain B is required for full Cf-4 function.

Additional Sequences Required for Cf-4 Function Are Located in LRRs 8 to 16

Two other chimeras were tested that contained sequences encoding Cf-9 domains A and B and amino acid A94. The Cf-4[9CH] chimera also contained sequences encoding Cf-9 LRRs 1 to 3 and amino acid C162 from LRR 4 (Figure 2A). In Cf-4[9CP], the substituted region was more extensive and included sequences encoding the seven N-terminal LRRs of Cf-9 (Figure 2A). Neither of these chimeras induced an aberrant seedling-lethal phenotype when crossed to a 35S:Avr4 line (Table 1). To determine whether this was due to the presence of Cf-9 domain B, we tested two additional chimeras.

The new chimeras contained the native Cf-4 domains A and B, amino acid E84 from LRR 1, and sequences encoding N-terminal LRRs from Cf-9. A cassette encoding LRRs 1 to 3 and the single variant amino acid in LRR 4 of Cf-9 (C162) was substituted into the corresponding region of Cf-4DS to generate Cf-4[9BH] (Figure 2A). In Cf-4[9BP] sequences encoding Cf-9, LRRs 1 to 7 were substituted for the corresponding region in Cf-4DS (Figure 2A). Both Cf-4[9BH] and Cf-4[9BP] induced a strong Avr4-dependent HR that was comparable to the Cf-4 control when tested in transient expression assays (Figures 3J and 3K).

These data demonstrate that substitution of N-terminal LRRs 1 to 7 of Cf-4 with the corresponding sequences from Cf-9 does not compromise its ability to induce an Avr4-dependent HR in tobacco. These LRRs contain 32 of the 67 amino acids that distinguish Cf-4 from Cf-9 (Figures 1 and and22).

Several other chimeras tested contained sequences encoding Cf-4 domains A and B and amino acid E84 together with variable numbers of Cf-9 N-terminal LRRs (Table 1). The chimeras Cf-4[9PE], Cf-4[9HE], and Cf-4[9BE] contained substituted regions encoding Cf-9 LRRs 8 to 15, LRRs 4 to 15, and LRRs 1 to 15, respectively (Figure 2A). None of the transgenic tobacco plants expressing these chimeras induced an Avr4-dependent HR (Table 1). Together, these data demonstrate that additional sequences required for Cf-4 function must be located within LRRs 8 to 16.

Cf-9 Function Is Not Compromised by Substitution with Cf-4 Domains A and B and Amino Acid E84

In Cf-9[4CB] sequences encoding Cf-4 domains A and B and amino acid E84 from LRR 1 were substituted into the corresponding region of Cf-9DS (Figure 2B). When Cf-9[4CB] was tested in transient expression assays, a strong HR was observed similar to that induced by the Cf-9DS control (Figure 4A). When transgenic tobacco plants expressing Cf-9[4CB] were crossed to a 35S:Avr9 line, half of their F1 progeny exhibited a seedling-lethal phenotype as in Cf-9DS × 35S:Avr9 controls (Table 1 and Figure 4B). The self-progeny of Cf-9[4CB] transgenic plants were also resistant to PVX:Avr9 infection (Table 1).

Figure 4.
Identification of Functional Domains in Cf-9.

When the progeny of three Cf-9[4CB] transgenic tomato plants were inoculated with PVX:Avr9, the leaves of infected plants exhibited systemic necrosis that was indistinguishable from Cf-9DS controls (Figures 4C and 4D). Tomato Cf-9[4CB] transgenic plants were also resistant to Cladosporium race 5 infection (Figure 4E). To confirm that the resistance was due to recognition of Avr9, we inoculated T2 progeny with the isogenic Cladosporium race 5.9 in which Avr9 has been deleted (Marmeisse et al., 1993). No resistance to Cladosporium race 5.9 was observed, demonstrating that resistance to infection was Avr9 dependent (Figure 4E).

Therefore, in contrast to Cf-4, domain B of Cf-9 can be replaced with the corresponding sequence from Cf-4 without compromising its ability to induce an Avr9-dependent HR in tobacco and tomato or to confer resistance to Cladosporium infection.

Additional Sequences Required for Cf-9 Function Are Located in LRRs 8 to 18

In Cf-9[4CH] sequences encoding Cf-4 domains A and B, LRRs 1 to 3 and amino acid S152 from LRR 4 were inserted into the corresponding region in Cf-9DS (Figure 2B). A more extensive region was substituted in clone Cf-9[4CP], which included sequences encoding Cf-4 LRRs 5, 6, and 7 (Figure 2B). The progeny of transgenic tobacco plants expressing Cf-9[4CH] or Cf-9[4CP] crossed to a 35S:Avr9 line exhibited a seedling-lethal phenotype (Figure 4G and Table 1). The development of the seedling-lethal phenotype in progeny of the Cf-9[4CH] × 35S:Avr9 cross was indistinguishable from the Cf-9DS × 35S:Avr9 control (Thomas et al., 2000) but was delayed in progeny from the Cf-9[4CP] 35S:Avr9 cross (Figure 4K), suggesting that Cf-9 function is attenuated. However, the self-progeny from Cf-9[4CH] and Cf-9[4CP] transgenic plants were resistant to PVX:Avr9 infection (Table 1).

Cf-9[4CH] and Cf-9[4CP] also were tested in transient expression assays. Cf-9[4CH] induced a slightly attenuated HR compared with Cf-9DS (Figure 4F). Cf-9[4CP] induced a delayed Avr9-dependent HR compared with Cf-9DS and Cf-9[4CH] (Figure 4J), consistent with the data above. To identify sequences within Cf-9 LRRs 4 to 7 that are required for full Cf-9 function, we constructed additional chimeras that contained substitutions of 4, 5, or 6 N-terminal LRRs of Cf-4 (Cf-9[4/4LRR], Cf-9[4/5LRR], and Cf-9[4/6LRR], respectively, see Figure 2B). However, when these sequences were tested in N. benthamiana, all induced an Avr9-dependent HR comparable to Cf-9[4CP] (Figure 2B; results not shown).

When the progeny from three Cf-9[4CH] transgenic tomato plants were inoculated with PVX:Avr9, they exhibited a systemic necrotic reaction that was delayed compared with Cf-9DS control plants (Figure 4H). Therefore, the substituted N-terminal LRRs in this chimera have a slight attenuating effect on development of the Avr9-dependent HR in tomato, as was observed in tobacco. However, the progeny of these transgenic plants exhibited Avr9-dependent resistance to Cladosporium infection (Figure 4I).

The progeny of three Cf-9[4CP] transgenic plants infected with PVX:Avr9 showed a delayed systemic necrosis compared with Cf-9DS controls and Cf-9[4CH] progeny (Figure 4L). Also, when the progeny of Cf-9[4CP] transgenic plants were inoculated with Cladosporium race 5 (Figure 4M), some restricted areas of fungal sporulation were observed on the abaxial leaf surface compared with Cf4 and Cf9 controls. However, these plants were significantly less sensitive to infection than Cf0 controls and siblings inoculated with C. fulvum race 5.9 (Figure 4M). Therefore, a correlation was observed between the ability to induce an Avr9-dependent HR and the ability of transgenic plants to resist Cladosporium infection (Figure 4).

In conclusion, as was observed for Cf-4, sequences encoding LRRs 1 to 7 of Cf-9 can be substituted without abolishing the induction of an Avr9-dependent HR. However, LRRs 1 to 7 of Cf-9 contain sequences that are required for full induction of the Avr9-dependent HR in tobacco and tomato and for resistance to Cladosporium infection.

The Cf-9 Amino Acids L457 and K511 in LRRs 16 and 18 Are Essential for Function

Most Cf-4/Cf-9 chimeras contained the 3′ half of Cf-4 that encodes the two variant amino acids F400 and N454 from LRRs 14 and 16 (Figure 1). None of the transgenic plants expressing chimeras that contained these two amino acids were able to induce an Avr9-dependent HR, and none of their progeny were resistant to PVX:Avr9 infection (Table 1). This included plants transformed with Cf-9[FN] whose protein product contains all Cf-9 variant amino acids except L457 and K511 in LRRs 16 and 18 (Figure 2B). Therefore, one or both of these amino acids appear to be essential for Cf-9 function, and chimeras were constructed to test this possibility.

Two sequence variants were generated that encode single amino acid substitutions corresponding to the two amino acids present in Cf-4 (F400 and N454). These clones were used to substitute the 3′ terminal halves of Cf-9, and two chimeras containing single amino acid substitutions, Cf-9[L457F] and Cf-9[K511N], were generated (Figure 2B). When these chimeras were tested in transient expression assays, neither Cf-9[L457F] nor Cf-9[K511N] could induce an Avr9-dependent HR (Figures 5A and 5B). Therefore, neither of the amino acids F400 or N454 in LRRs 14 and 16 of Cf-4 could substitute functionally for L457 and K511 in LRRs 16 and 18 of Cf-9.

Figure 5.
Characterization of Single Amino Acid and LRR Copy Number Variants of Cf-4 and Cf-9.

In contrast, the two Cf-4 variants Cf-4[F400L] and Cf-4[N454K] could induce an HR that was indistinguishable from the Cf-4DS control (Figures 5C and 5D). When the variant Cf-4[LK] was tested (which contains both Cf-9 amino acids; see Figure 2A), a slight attenuation in the development of the HR was observed compared with the Cf-4DS control (results not shown). Also, when tobacco Cf-4[LK] transgenic plants were crossed to a 35S:Avr4 line, their progeny exhibited a seedling-lethal phenotype similar to that induced by Cf-4DS × 35S:Avr4 controls (Thomas et al., 2000), and the progeny of Cf-4[LK] transgenic tobacco plants were also resistant to PVX:Avr4 infection (Table 1). Therefore, substitution of either of the Cf-4 amino acids F400 and N454 with the corresponding sequences from Cf-9 does not compromise the ability of Cf-4 to induce an Avr4-dependent HR.

Cf-4 and Cf-9 LRR Copy Number Variants

Cf-4 and Cf-9 differ in the number of their N-terminal LRRs within domain C1 (Thomas et al., 1997). Cf-4 contains a deletion of 46 amino acids relative to Cf-9 that corresponds to two complete LRRs (Figure 1). To determine the effect of varying LRR copy number in Cf-4 and Cf-9, we constructed two sequence variants. In Cf-4[+2LRR], sequences from Cf-9 encoding the two additional LRRs were inserted into Cf-4 between sequences encoding amino acids G316 and P317 (Figure 2A). In the second construct, Cf-9 sequences encoding the same region (P328 to G373) were deleted to generate Cf-9[Δ2LRR] that encodes a Cf-9 variant lacking 2 LRRs (Figure 2B). Both sequences were tested for their ability to induce an Avr4- or Avr9-dependent HR in N. benthamiana leaves. Neither Cf-4[+2LRR] (Figure 5E) nor Cf-9[Δ2LRR] (Figure 5F) could induce an Avr-dependent HR, demonstrating that altering LRR copy number compromises Cf-4 and Cf-9 function.

Characterization of Cf-4 and Cf-9 Gene-Shuffled Clones

PCR-mediated gene shuffling (Stemmer, 1994) of Cf-4 and Cf-9 was performed to increase the number of Cf-4/Cf-9 chimeras that could be analyzed (Figure 6A). Agrobacterium strain GV3101 clones containing shuffled sequences in a T-DNA vector under control of the CaMV 35S promoter were selected at random and assayed for their ability to induce an Avr-dependent HR in transgenic N. benthamiana plants expressing 35S:Avr4 and 35S:Avr9.

Of the 364 shuffled clones tested 108 (29%) induced an HR on test plants. Seventy-two of these HR-inducing clones were tested on N. benthamiana plants expressing either 35S:Avr4 or 35S:Avr9. Fifty-seven of the clones induced an Avr4-dependent HR, and 15 induced an Avr9-dependent HR. None of the clones induced an HR on nontransformed N. benthamiana leaves, and no clones induced an HR in response to both Avr4 and Avr9.

The predicted amino acid sequences of 27 clones that induced an Avr4-dependent HR and 12 that induced an Avr9-dependent HR are shown in Figure 6B. Most of the clones encoded Cf-9 domain B, along with amino acid K511 in LRR 18 and amino acid A94 in LRR 1. This suggests that some Cf-9 sequences were preferentially incorporated during reassembly. No clones contained point mutations (zero mutations in 80 kb), sequence duplications, deletions, or insertions. On average, one template switch was observed every 225 bases (results not shown). Sequence analysis also demonstrated that ~50% of template switches were not separated by restriction sites for any of the enzymes used to fragment Cf-4 and Cf-9 templates before shuffling. Therefore, this strategy can be used for efficient shuffling of homologous sequences.

Cf-4 LRRs 11 and 12 Are Required for an Avr4-Dependent HR

Data from the analysis of clones that induced an Avr4-dependent HR were mostly consistent with the results reported above for Cf-4/Cf-9 domain swaps and enabled the identification of additional sequence determinants of Cf-4 function. However, several functional clones contained the additional 10 amino acids from Cf-9 domain B (A05, E18, and C07; see Figure 6B). This contrasts with the data above in which Cf-4[9CB] and Cf-4[+10B] in tomato and tobacco transgenics induced an attenuated Avr4-dependent HR (Figures 3G and 3I). For example, clone A05 induced a strong HR and encodes a protein similar to Cf-4[+10B] apart from four additional Cf-9 amino acids within its N-terminal LRRs (Figure 6B). Other clones containing Cf-9 domain B and various Cf-9 LRR substitutions induced a weak HR similar to that induced by Cf-4[9CB] and Cf-4[+10B] (Figures 3G and 3I).

All sequences that induced an Avr4-dependent HR lacked the two additional LRRs present in Cf-9, consistent with the data reported above (Figure 5E). Most of the clones retained sequences encoding LRRs 11 to 14 of Cf-4. The analysis of other clones in this class (C04, A02, A01, E02, V05, E24, and V12) suggests that amino acids T376 and P387 in LRR 13 can be substituted with the corresponding amino acids A433 and R444 from LRR 15 of Cf-9 (Figure 6B). Also, because amino acids F400 and N454 in LRRs 14 and 16 can be substituted (Figures 5C and 5D), this suggests that all of the variant amino acids in LRRs 13, 14, and 16 of Cf-4 can be substituted by the corresponding Cf-9 amino acids without compromising Cf-4 function.

In clone A11, Cf-4 LRRs 1, 3, 4, 6, 7, 8, 9, and 10 were substituted with the corresponding sequences from Cf-9 (Figure 6B). It is possible, therefore, that the only variant amino acids required for Cf-4 function are located in LRRs 11 and 12. This region contains eight variant amino acids, of which five are putative solvent-exposed residues (Figure 6B). In one other clone (V12), two variant amino acids, I353 and G354, in LRR 12 were substituted by the Cf-9 amino acids V410 and E411 (Figure 6B). A weak but reproducible HR was induced by this chimera. The weak HR induced by this chimera may be due to the presence of Cf-9 domain B, or alternately, to substitution of amino acids I353 and G354 in Cf-4 LRR 12. In the former case, this would localize the variant amino acids that are important for Cf-4 function to LRR 11, a region containing only six amino acids (V322, R326, Q329, I330, W332, and N337) that distinguish it from Cf-9.

Important Sequences for Cf-9 Function Are Located in LRRs 10 to 18

All shuffled clones that induced an Avr9-dependent HR contained the amino acids L457 and K511 in LRRs 16 and 18 and sequences encoding the two additional LRRs of Cf-9, as predicted from the data above (Figures 5A, 5B, and 5F). We also demonstrated that LRRs 1 to 7 of Cf-9 could be substituted without abolishing the Avr9-dependent HR (Figures 4J to 4M). One other clone (G24) retains a strong HR-inducing activity 6 days after infiltration and contains a substituted region encompassing LRRs 1 to 9 of Cf-4 (Figure 6B). Therefore, important sequences for Cf-9 function must be located in a region between LRRs 10 to 18. This region encompasses the two additional LRRs in Cf-9 and 15 of the 67 amino acids that distinguish it from Cf-4.

Several clones were identified (A16, E16, E19, E23, C03, and C20) that encode chimeras in which Cf-9 amino acid A433 in LRR 15 was substituted by T376 from Cf-4 (Figure 6B). This substitution results in a weak but reproducible Avr9-induced HR 6 days after infiltration. These chimeras also contain additional sequence substitutions in LRRs 1 to 4, but as shown above, these substitutions do not significantly affect Cf-9 function in tobacco. Therefore, in contrast with amino acids L457 and K511, which cannot be substituted with the corresponding Cf-4 amino acids, A433 in LRR 15 of Cf-9 is only required for full induction of the Avr9-dependent HR in common with other sequences within LRRs 1 to 9 that have not been identified.

DISCUSSION

Domain swapping was previously used to identify sequences in tomato Pto that are required for avrPto recognition (Scofield et al., 1996; Tang et al., 1996), and domains in flax L alleles that confer resistance to specific races of flax rust (Ellis et al., 2000). Here, we report our analyses of Cf-4/Cf-9 chimeras that show that only a fraction of the variant amino acids that distinguish Cf-4 and Cf-9 are required for induction of an Avr-dependent HR in tobacco and tomato. Variant amino acids within the central LRRs in domain C1 of Cf-4 and Cf-9, together with variation in their LRR copy number, make a major contribution to recognition specificity in these proteins.

Cf-4 and Cf-9 Domains A and B

Some of the amino acids that distinguish Cf-4 and Cf-9 are located in domains A and B (Figure 1). Domain A is a predicted signal peptide that is cleaved during the maturation of Cf proteins (Jones and Jones, 1997), and sequences within this domain are unlikely to determine recognition specificity. Our experiments with Cf-4/Cf-9 domain A and B chimeras demonstrated that the variant amino acids within domain A could be substituted without compromising Cf-4 or Cf-9 function (Figures 3H and and4A4A).

Domain B of Cf proteins, the predicted mature N terminus, shows no homology with the plant extracellular LRR consensus sequence. A similar domain is present in other plant LRR proteins such as polygalacturonase-inhibiting proteins (PGIPs) and LRR receptor-kinases (Jones et al., 1994; Jones and Jones, 1997). Structural modeling of domain B suggests that it may adopt an α-helical fold (results not shown), and a 10–amino acid deletion in Cf-4 would remove approximately two α-helical turns. This may be important for the recognition specificity of ligand binding, subsequent signaling functions, or alternatively for Cf-4 protein stability.

The results of our analysis of Cf-4/Cf-9 chimeras expressed in transgenic tobacco and tomato plants were mostly consistent with the analysis of shuffled clones identified on the basis of their ability to induce an Avr-dependent HR in N. benthamiana leaves. Our analysis of domain B variants suggested that the native Cf-4 domain B is required for full Cf-4 function (Figure 3). Most shuffled clones that induced an Avr4-dependent HR comparable to Cf-4 did contain Cf-4 domain B (Figure 6B). However, several Cf-4/Cf-9 shuffled clones containing Cf-9 domain B, and other amino acids from domain C1 of Cf-9, did induce a strong Avr4-dependent HR (Figure 6B). For example, clone A05 encodes a protein highly homologous with that encoded by Cf-4[9CB] apart from four Cf-9 amino acids in LRRs 1, 8, 9, and 18 (Figure 6B). In fact, most shuffled clones that were analyzed contained the additional 10 amino acids of Cf-9 domain B. Some Cf-9 sequences may have been preferentially incorporated during the reassembly reaction (Figure 6B). Consequently, our analysis may have been biased toward the identification of chimeras containing the additional 10 amino acids of Cf-9 domain B that retained the capacity to induce an Avr4-dependent HR.

In contrast, the Cf-9[4CB] chimera induced a strong Avr9-dependent HR in transgenic tobacco, and tomato plants demonstrating Cf-9 domain B can be substituted without significantly affecting Cf-9 function (Figure 3). However, some exceptions were observed. For example, clone C09 (which contains the native Cf-9 domain B) induced a stronger Avr9-dependent HR 6 days after infiltration than the near identical clone Cf-9[4CP] that contains Cf-4 domain B (Figure 6B). It is possible that minor sequence variation in Cf-4/Cf-9 chimeras may affect their tertiary structure or stability, the ability to recognize their cognate ligands, or their ability to interact with signaling proteins that activate the plant defense response. Despite the fact that Cf-4 and Cf-9 are >91% identical, our analysis cannot exclude the possibility that null phenotypes (Table 1) or variations in the severity of the HR are due to an inherent instability or reduced expression of some chimeras. Future experiments with epitope-tagged versions of these proteins (Piedras et al., 2000) will determine whether differential protein accumulation can account for null phenotypes or variation in the severity of the HR.

LRR Sequences Required for Cf-4 and Cf-9 Function

The proposed structure of a plant extracellular LRR protein, PGIP, has been reported recently (Leckie et al., 1999). PGIPs and Cf proteins have similar LRR consensus sequences, and they may adopt similar tertiary structures (Jones and Jones, 1997). As in porcine ribonuclease inhibitor (Kobe and Deisenhofer, 1994), the LRRs in PGIPs were proposed to form a curvilinear β-sheet. Also, the putative solvent-exposed flanking residues of the β-strand/β-turn structural motif form an extensive ligand binding surface (Leckie et al., 1999). It was also shown that substitution of a single putative solvent-exposed amino acid in PGIP-2 enabled it to bind a novel substrate (Leckie et al., 1999). Molecular analysis of Hcr9s has revealed that sequences encoding these solvent-exposed amino acids have undergone diversifying selection consistent with their proposed role in determining recognition specificity in these proteins (Parniske et al., 1997, 1999; Thomas et al., 1997; Parniske and Jones, 1999).

Most of the amino acids that distinguish Cf-4 and Cf-9 are putative solvent-exposed amino acids of the β-strand/β-turn structural motif (Figure 1). Our analysis has shown that only a fraction of these variant amino acids are required for Cf-4 and Cf-9 function. The analysis of chimeras that induced an Avr4-dependent HR demonstrated that important variant amino acids required for Cf-4 function are located in LRRs 11 and 12 (Figure 6B). Four times as many Cf-4/Cf-9 shuffled clones that could induce an HR recognized Avr4 (Figure 6B). This may reflect the fact that important sequence determinants of Cf-4 function are located on two contiguous LRRs. Analysis of clone V12 suggests that the most important variant amino acids for Cf-4 function may be located in LRR 11 (Figure 6B). This will be tested in the future by substituting variant amino acids of Cf-4 LRRs 11 and 12 with the corresponding amino acids from Cf-9. Cf-4 LRRs 11 and 12 contain only eight amino acids that distinguish it from Cf-9, five of which are putative solvent-exposed residues (Q329, I330, and W332 in LRR 11, I353 and G354 in LRR 12). Amino acids I330 and W332 in LRR 11 and G354 in LRR 12 of Cf-4 are located where maximum variation in amino acid composition (seven in total for each position) was observed in a comparison of 18 Hcr9 sequences (Figure 6B). This is consistent with a previous suggestion that the hypervariable sequences within the central LRRs of domain C1 are important determinants of recognition specificity in Hcr9 proteins (Parniske et al., 1997).

In Cf-9, some variant amino acids required for full Avr9-dependent induction of an HR are located in LRRs 1 to 9 (Figures 4 and and6B).6B). Therefore, in Cf-9 the sequences required for function may be dispersed over a greater number of LRRs. Additional sequences required for Cf-9 function may also be located in the region delimited by LRRs 10 to 15, and these may be identified from the analysis of additional Cf-4/Cf-9 chimeras. This region contains the two additional LRRs in Cf-9 and 13 of the 67 variant amino acids that distinguish it from Cf-4, nine of which are putative solvent-exposed residues of the β-strand/β-turn structural motif. Two solvent-exposed amino acids, L457 and K511, in LRRs 16 and 18 could not be substituted with the corresponding Cf-4 amino acids F400 and N454 (Figures 5A and 5B), and these residues appear to be absolutely required for Cf-9 function. In the comparison of 18 Hcr9 sequences (Figure 6B), five different amino acids were found at the position occupied by L457 in Cf-9 LRR 16, demonstrating that this position is also hypervariable, whereas only two were found at the position occupied by K511 in LRR 18. In contrast with amino acid A433 in LRR15 and other sequences in LRRs 1 to 9 of Cf-9, L457 and K511 may be essential residues that contact with the Cf-9 ligand or a signaling partner protein that activates the plant defense response (Dixon et al., 2000).

Although our analysis has identified functional variant amino acids in Cf-4 and Cf-9, it could not address the potential role of conserved sequences in ligand recognition or signaling. The substitution of LRRs 1 to 10 of Cf-4 with sequences from Cf-9 did not compromise its function, suggesting that none of the variant amino acids in this region are required for Cf-4 function. Alternately, these amino acids may be effectively substituted with the corresponding sequences from Cf-9, or additional functional sequences in this region might be conserved between Cf-4 and Cf-9. The identification of additional functional sequences will be addressed in the future by analyzing functional chimeras produced by shuffling Cf-4 and Cf-9 together with many polymorphic Hcr9s (Parniske et al., 1997, 1999; Thomas et al., 1997; Parniske and Jones, 1999).

LRR Copy Number Variation in Cf Proteins

If the only important sequence determinants of Cf-4 function are located in Cf-4 LRRs 11 and 12, the insertion of additional LRRs distal to this region might not compromise its function. However, both the Cf-4 and Cf-9 LRR copy number variants tested were unable to induce an HR in tobacco leaves (Figures 5E and 5F). Also, all of the Cf-4/Cf-9 shuffled clones that induced an Avr4-dependent HR contained 25 LRRs, as in Cf-4 (Figure 6B). DNA sequence analysis of randomly selected clones revealed several chimeras that contained Cf-4 LRRs 11 to 16 plus insertions of the two additional Cf-9 LRRs (data not shown). Therefore, the absence of clones that recognize Avr4 and that contained 27 LRRs, as in Cf-9, was not due to their absence from the population of Cf-4/Cf-9 shuffled clones. Similarly, all chimeras that induced an Avr9-dependent HR contained 27 LRRs, as in Cf-9 (Figure 6B). As in other LRR proteins, such as porcine ribonuclease inhibitor (Kobe and Deisenhofer, 1994), it is likely that the spacing between specific functional sequences that contact with the ligand (as determined by the number of intervening LRRs) is critical for Cf protein function. Whether this critical spacing is required between LRRs 11 and 12 of Cf-4, and LRRs 16 and 18 of Cf-9, and sequences located proximally (such as domain B) or distally in each protein is unknown.

Unequal crossing-over in R genes resulting in variation in LRR copy number in R proteins is thought to be an important process for generating novel recognition specificities (Ellis et al., 2000). In Hcr2 proteins, more extensive variation in LRR copy number was observed in the LRRs of domain C1 than was observed in Hcr9s (Parniske et al., 1997; Dixon et al., 1998; Thomas et al., 1998). For example, the Cf-5 and Hcr2-5D proteins differ only in the number of their N-terminal LRRs (32 and 34 LRRs, respectively), but only Cf-5 determines Avr5-dependent resistance to C. fulvum infection (Dixon et al., 1998). Successive rounds of duplication and deletion events involving sequences encoding LRRs also has also been proposed to account for the generation of novel recognition specificities at the RPP5 locus in Arabidopsis and the flax L and M loci (Ellis et al., 2000).

We have demonstrated that sequence variation in the central LRRs of domain C1 and variation in LRR copy number make a crucial contribution to recognition specificity in the tomato Cf-4 and Cf-9 proteins. This report describes the use of gene shuffling to characterize plant genes. Further analysis of gene chimeras should facilitate the identification of additional functional sequences in these proteins. Gene shuffling of Hcr9s could also be used to generate large libraries of chimeras with a potentially vast spectrum of recognition specificities. As a result, it may be possible to identify R proteins that recognize defined pathogen-encoded proteins as a means to engineer novel disease resistance specificities in plants.

METHODS

Construction of Clones Cf-4DS and Cf-9DS

To facilitate the construction of Cf-4/Cf-9 chimeras, we generated two clones (Cf-4DS and Cf-9DS) by oligonucleotide mutagenesis that contained sites for ClaI, BglII, HindIII, and PvuII in the 5′ halves of each gene (Figure 1). Recognition sequences for these enzymes were removed from their 3′ halves. The novel restriction sites were introduced in sequences at the 5′ end of the open reading frames (ClaI) or in sequences that encode identical amino acids in each protein (BglII, HindIII, and PvuII). None of the nucleotide substitutions affect the predicted amino acid composition of Cf-4 and Cf-9.

Cf-4DS and Cf-9DS were constructed as follows. A 6.0-kb PstI fragment encompassing Cf-4 and 3.0 kb of 5′ flanking DNA (Thomas et al., 1997) was isolated from clone p129P6 (Thomas et al., 1997). A 7.5-kb PstI fragment encompassing Cf-9 was isolated from pSLJ8146 (Hammond-Kosack et al., 1998). Both fragments were cloned into pSLJ8131, a pUC119 derivative that contains sites for XbaI, SalI, PstI, and BamHI only. Single-stranded DNA containing the sense (+) strand of Cf-4 or Cf-9 was prepared as template for oligonucleotide mutagenesis as described by Jones et al. (1992). Pairs of mutagenic primers were used sequentially to introduce novel sites for the restriction enzymes ClaI, BglII, HindIII, and PvuII into the 5′ halves of Cf-4 and Cf-9 and to remove existing sites for these enzymes from their 3′ halves. Details of all oligonucleotides used in this study are shown in Table 2. The coding sequences of the mutagenized Cf-4 and Cf-9 clones were verified by DNA sequence analysis on an Applied Biosystems 377 sequencer (Foster City, CA). The mutagenized Cf-4 and Cf-9 clones that contained the 3′ untranslated region, an intron, and transcription termination sequences (Thomas et al., 1997) were excised as ClaI-BamHI fragments. The ClaI-BamHI cassettes were cloned into a modified version of pBluescript KS+ that lacked sites for EcoRI, HindIII, and PvuII to facilitate the reciprocal exchange of DNA fragments; these clones were designated Cf-4DS and Cf-9DS.

Table 2.
Oligonucleotide Primers Used in This Study

Chimeras Generated by Polymerase Chain Reaction–Mediated Mutagenesis

Additional constructs containing sequence insertions/deletions, variable numbers of leucine-rich repeats (LRRs), reciprocal exchanges of LRRs, or defined single amino acid exchanges (Figure 2) were generated by polymerase chain reaction (PCR)–mediated mutagenesis. PCR amplification (20 cycles of 94°C for 15 sec, 55°C for 15 sec, and 72°C for 45 sec) was performed with a mixture of Amplitaq DNA polymerase (Gibco BRL, Paisley, UK) and native Pfu DNA polymerase (Promega) at a unit ratio of 160:1. Suitable cassettes for cloning were produced by PCR amplification of two products that were digested with restriction enzymes to leave blunt termini that were ligated with T4 DNA ligase (Gibco BRL). The ligation products were reamplified with two of the original 5′ flanking primers and digested with two restriction enzymes for cloning into Cf-9DS or Cf-4DS.

Recombinant domain B fragments were synthesized as follows. For clone Cf-4[Δ9B], two fragments were amplified from a clone containing 35S:Cf-9DS by using primer pairs 35S1/M12 and M13/CTOM8. The fragments were digested with EcoRV and StuI, respectively, ligated, and then reamplified with 35S1/CTOM8. The final amplification product was digested with ClaI and BglII and cloned into the corresponding region in Cf-4DS. For clone Cf-4[+10B], two DNA fragments were PCR amplified from a clone containing 35S:Cf-4DS with primer pairs 35S1/M15 and M14/CTOM7. The products were digested with SspI and EcoRV, respectively, ligated, and then reamplified with 35S1 and CTOM7. The PCR product was digested with ClaI and BglII and cloned into the corresponding region of Cf-4DS (Figure 2).

In construct Cf-9[Δ2LRR], two LRRs were deleted from Cf-9DS. Two fragments were amplified from Cf-9DS by using primer pairs F19/M4 and M5/F5. The PCR products were digested with HindIII-EcoRV and HpaI-EcoRI, respectively, and ligated directly into HindIII-EcoRI–digested Cf-9DS. To insert two LRRs from Cf-9 into Cf-4, we first introduced an ApaI site (GGGCCC) into Cf-4DS at sequences encoding G316 and P317 of Cf-4. This was achieved by PCR amplification with primer pairs 35S1/M17 and M16/F5 using Cf-4DS as a template. The PCR products were digested with PvuII-ApaI and ApaI-EcoRI and cloned into PvuII-EcoRI–digested Cf-4DS. A PCR product encoding two LRRs from Cf-9 between P328 and G373, and delimited by ApaI sites, was amplified from Cf-9DS using primers M18 and M19. The DNA was digested with ApaI and cloned into the ApaI-digested vector described above after treatment with shrimp alkaline phosphatase. Clones with ApaI inserts in the desired orientation were identified by PCR analysis with primers 35S1/M19 and designated Cf-4[+2LRR].

The clones Cf-9[4/4LRR], Cf-9[4/5LRR], and Cf-9[4/6LRR] were generated by a similar strategy. For clone Cf-9[4/4LRR], two fragments were amplified using primers M14/M20 and M21/F5 that were digested with SspI and EcoICRI, respectively. For clone Cf-9[4/5LRR], fragments were amplified with M14/M22 and M23/F5 and digested with PmlI and EcoICRI; for clone Cf-9[4/6LRR], fragments were generated with M14/M24 and M25/F5 and also digested with PmlI and EcoICRI. Appropriate fragment pairs were ligated, reamplified with M14 and F5, digested with BglII and PvuII, and cloned into the corresponding region of Cf-9[4CP] (see Figure 2).

Two clones containing single amino acid substitutions in the 3′ half of Cf-9 were generated as follows. Primers M1 and F2 were used to amplify a product from a wild-type Cf-9 clone. The PCR product was digested with EcoRI and HindIII. This fragment was used to replace the corresponding region in a wild-type Cf-9 clone. The 3′ half of this clone was excised as an EcoRI-BamHI fragment and used to replace the corresponding region in Cf-9DS and Cf-4DS to generate clones Cf-9[L457F] and Cf-4[N454K]. A similar strategy was used to generate Cf-9[K511N]. Two fragments were PCR amplified using primers F6/M3 and M2/F2. The products were digested with EcoRI-BstYI and HindIII-BstYI, respectively, and cloned directly into the corresponding region of the wild-type Cf-9 clone. The EcoRI-BamHI fragment was subcloned as described above into Cf-9DS and Cf-4DS to generate clones Cf-9[K511N] and Cf-4[F400L].

Cf-4 and Cf-9 Gene Shuffling

Two XhoI-SacII fragments from pSLJ12574 and pSLJ12575 that contained sequences extending from the 5′end of the Cf-4 and Cf-9 reading frames to their internal HindIII sites were prepared from agarose gels. These fragments were flanked by pBluescript KS+ polylinker sequences and contained all of the polymorphic sequences that distinguish Cf-4 from Cf-9.

The purified fragments were fragmented further by restriction enzyme digestion to increase the frequency of template switching during reassembly (Kikuchi et al., 1999). Four micrograms of Cf-9 DNA was digested with Tsp509I or a combination of Tru9I and RsaI. Four micrograms of the Cf-4 fragment was digested with VIJI* (Chimerx, Milwaukee, WI) or a combination of AluI and XmnI. The DNAs were electrophoresed in a nondenaturing 6.5% (w/v) polyacrylamide gel, and fragments in the length range 20 to 400 bp were eluted overnight at 37°C in a buffer containing 0.5 M ammonium acetate, 10 mM magnesium acetate, 1 mM EDTA, pH 8.0, and 0.1% (w/v) SDS. The DNA was recovered by phenol:chloroform extraction and ethanol precipitation.

Equimolar amounts of Cf-4 and Cf-9 restriction fragments were mixed in different combinations (see Figure 6A) and used in a primer-free amplification reaction at a final concentration of 10 ng/mL. All amplification steps were performed with native Pfu DNA polymerase from Pyrococcus furiosus strain Vc1 DSM3638 (Promega, Madison, WI) to reduce the frequency of point mutations (Zhao and Arnold, 1997). Reactions were incubated at 95°C for 2 min and then 15 cycles of 95°C for 30 sec, 55°C for 30 sec, and 73°C for 1 min (+10 sec per cycle). The reaction products were then treated at 95°C for 30 sec and 55°C for 30 sec. The four mixtures were combined (see Figure 6A), and a second round of primer-free PCR was performed as follows: 73°C for 3 min 40 sec followed by 25 cycles of 95°C for 30 sec, 55°C for 30 sec, and 73°C for 3 min 40 sec followed by a final extension phase of 7 min at 73°C. To recover shuffled Cf-4 and Cf-9 sequences, we amplified aliquots of this mixture with pBluescript primers SR1 and SF2 (Table 2) as follows: 95°C for 2 min and 31 cycles of 95°C for 15 sec, 64°C for 30 sec, and 73°C for 3 min 30 sec. The reaction was completed by a further incubation at 73°C for 7 min.

Expression Vectors for Domain Swap Constructs and Cf-4/Cf-9 Shuffled Clones

For stable expression in tomato (Lycopersicon esculentum) or tobacco (Nicotiana tabacum), chimeras were excised as ClaI-BamHI fragments and cloned into a derivative of the T-DNA binary vector pSLJ7291 containing the native Cf-4 promoter (Thomas et al., 1997). For Agrobacterium tumefaciens–mediated transient expression, the ClaI-BamHI cassettes were cloned into pBin19 containing a 1.4-kb DNA fragment derived from vector pSLJ4K1 (Jones et al., 1992) that contains the cauliflower mosaic virus (CaMV) 35S promoter.

The products of Cf-4/Cf-9 gene shuffling were digested with ClaI and HindIII and ligated into the vector pSLJ12904 that had been digested with ClaI and HindIII and gel purified. This pBin19-derived vector contains the CaMV 35S promoter and a ClaI-HindIII cassette encoding part of the jellyfish green fluorescent protein fused to a HindIII-BamHI fragment that contains the 3′ terminal coding sequences and 3′ untranslated region of the Cf-9 gene. The ligation products were electroporated into Agrobacterium strain GV3101, and kanamycin-resistant clones were picked into 384-well microtiter plates.

Agrobacterium-Mediated Transient Gene Expression Assays

Initially, Cf-4/Cf-9 chimeras were cloned into pBin19 and electroporated into Agrobacterium strain C58C1 containing the helper plasmid pCH32 that overexpresses the virD2 and virE genes (Hamilton et al., 1996). Transformants were selected on nutrient agar plates containing tetracycline (2 μg mL−1) and kanamycin (40 μg mL−1). Single colonies subsequently were streaked on minimal medium agar plates containing tetracycline and kanamycin. These clones and Cf-4/Cf-9 shuffled clones in Agrobacterium strain GV3101 were prepared for transient expression in plants as follows. Stationary phase bacterial cultures were suspended in a solution containing 10 mM 2-[N-morpholino]ethanesulfonic acid, pH 6.0, 10 mM MgCl2, and 150 μM acetosyringone for 3 hr at room temperature.

Bacterial suspensions usually were infiltrated into N. benthamiana leaves by using a syringe as described previously (Thomas et al., 2000). Reproducible results also were obtained after infiltration into tobacco. In most experiments reported here, Agrobacterium suspensions were infiltrated into leaves of transgenic N. benthamiana or tobacco plants expressing either 35S:Avr4 or 35S:Avr9. For preliminary screening experiments, F1 hybrids expressing both transgenes were analyzed. Occasionally, mixtures of Agrobacterium suspensions expressing a test construct and Agrobacteria expressing 35S:Avr4 or 35S:Avr9 were coinfiltrated into leaves of wild-type N. benthamiana plants. However, the development of the hypersensitive response (HR) in these assays was slower than in the assays described above.

Inoculations with Cladosporium fulvum

Three- to 4-week-old plants were inoculated with C. fulvum race 5 or race 5.9 spore suspensions and scored for resistance or disease sensitivity 14 days after inoculation as described previously (Thomas et al., 1997).

Transformation of Cf0 Tomato

Binary vector plasmids were mobilized from Escherichia coli DH5α into Agrobacterium strain LBA4404 as described by Jones et al. (1992). Transformation of tomato Cf0 Moneymaker cotyledons and plant regeneration were performed as described previously (Thomas et al., 1997).

Acknowledgments

The Sainsbury Laboratory is funded by the Gatsby Charitable Foundation. Part of this research was funded by a grant from the European Community (EC Biotech BIO4 CT96 0515). B.B.H.W received funding from the Danish Research Academy. We thank Andrew Davis for plant photography, Sara Perkins and Justine Campling for their excellent horticultural assistance, and David Baker and Patrick Bovill for DNA sequencing. We are grateful to Saijun Tang for providing p8131 and several oligonucleotides, Julia Krueger and Leif Schauser for constructive comments on the manuscript, and Martin Parniske and Paul Schulze-Lefert for useful discussions on gene shuffling. We are also grateful to three reviewers for their suggestions on improving the manuscript.

References

  • Crameri, A., Raillard, S.-A., Bermudez, E., and Stemmer, W.P.C. (1998). DNA shuffling of a family of genes from diverse species accelerates directed evolution. Nature 391, 288–291. [PubMed]
  • Dixon, M.S., Hatzixanthis, K., Jones, D.A., Harrison, K., and Jones, J.D.G. (1998). The tomato Cf-5 disease resistance gene and six homologs show pronounced allelic variation in leucine-rich repeat copy number. Plant Cell 10, 1915–1925. [PMC free article] [PubMed]
  • Dixon, M.S., Golstein, C., Thomas, C.M., Van der Biezen, E.A., and Jones, J.D.G. (2000). Genetic complexity of pathogen perception by plants: The example of Rcr3, a tomato gene required specifically by Cf-2. Proc. Natl. Acad. Sci. USA 97, 8807–8814. [PMC free article] [PubMed]
  • Ellis, J., Dodds, P., and Pryor, T. (2000). Structure, function and evolution of plant disease resistance genes. Curr. Opin. Plant Biol. 3, 278–284. [PubMed]
  • Hamilton, C.M., Frary, A., Lewis, C., and Tanksley, S.D. (1996). Stable transfer of intact high-molecular-weight DNA into plant chromosomes. Proc. Natl. Acad. Sci. USA 93, 9975–9979. [PMC free article] [PubMed]
  • Hammond-Kosack, K.E., Harrison, K., and Jones, J.D.G. (1994). Developmentally regulated cell death on expression of the fungal avirulence gene Avr9 in tomato seedlings carrying the disease-resistance gene Cf-9. Proc. Natl. Acad. Sci. USA 91, 10445–10449. [PMC free article] [PubMed]
  • Hammond-Kosack, K.E., Staskawicz, B.J., Jones, J.D.G., and Baulcombe, D.C. (1995). Functional expression of a fungal avirulence gene from a modified Potato Virus X. Mol. Plant-Microbe Interact. 8, 181–185.
  • Hammond-Kosack, K.E., Tang, S.J., Harrison, K., and Jones, J.D.G. (1998). The tomato Cf-9 disease resistance gene functions in tomato and potato to confer responsiveness to the fungal avirulence gene product Avr9. Plant Cell 10, 1251–1266. [PMC free article] [PubMed]
  • Jones, D.A., and Jones, J.D.G. (1997). The role of leucine-rich repeat proteins in plant defences. Adv. Bot. Res. Inc. Adv. Plant Pathol. 24, 89–167.
  • Jones, D.A., Thomas, C.M., Hammond-Kosack, K.E., Balint-Kurti, P.J., and Jones, J.D.G. (1994). Isolation of the tomato Cf-9 gene for resistance to Cladosporium fulvum by transposon tagging. Science 266, 789–793. [PubMed]
  • Jones, J.D.G., Shlumukov, L., Carland, F.J., Scofield, S., Bishop, G., and Harrison, K. (1992). Effective vectors for transformation, expression of heterologous genes, and assaying transposon excision in transgenic plants. Transgenic Res. 1, 285–297. [PubMed]
  • Joosten, M.H.A.J., Cozijnsen, T.J., and De Wit, P.J.G.M. (1994). Host resistance to a fungal tomato pathogen lost by a single base-pair change in an avirulence gene. Nature 367, 384–386. [PubMed]
  • Joosten, M.H.A.J., Vogelsang, R., Cozijnsen, T.J., Verberne, M.C., and De Wit, P.J.G.M. (1997). The biotrophic fungus Cladosporium fulvum circumvents Cf-4-mediated resistance by producing unstable AVR4 elicitors. Plant Cell 9, 367–379. [PMC free article] [PubMed]
  • Kajava, A.V. (1998). Structural diversity of leucine-rich repeat proteins. J. Mol. Biol. 277, 519–527. [PubMed]
  • Kamoun, S., Honee, G., Weide, R., Lauge, R., Kooman-Gersmann, M., De Groot, K., Govers, F., and De Wit, P.J.G.M. (1999). The fungal gene Avr9 and the oomycete gene inf1 confer avirulence to potato virus X on tobacco. Mol. Plant-Microbe Interact. 12, 459–462.
  • Kikuchi, M., Ohnishi, K., and Harayama, S. (1999). Novel family shuffling methods for the in vitro evolution of enzymes. Gene 236, 159–167. [PubMed]
  • Kobe, B., and Deisenhofer, J. (1994). The leucine-rich repeat: A versatile binding motif. Trends Biochem. Sci. 19, 415–421. [PubMed]
  • Leckie, F., Mattei, B., Capodicasa, C., Hemmings, A., Nuss, L., Aracri, B., De Lorenzo, G., and Cervone, F. (1999). The specificity of polygalacturonase-inhibiting protein (PGIP): A single amino acid substitution in the solvent-exposed β-strand/β-turn region of the leucine-rich repeats (LRRs) confers a new recognition capability. EMBO J. 18, 2352–2363. [PMC free article] [PubMed]
  • Marmeisse, R., Van den Ackervecken, G.F.J.M., Goosen, T., De Wit, P.J.G.M., and Van den Broek, H.W.J. (1993). Disruption of the avirulence gene Avr9 in 2 races of the tomato pathogen Cladosporium fulvum causes virulence on tomato genotypes with the complementary resistance gene Cf-9. Mol. Plant-Microbe Interact. 6, 412–417.
  • Parniske, M., and Jones, J.D.G. (1999). Recombination between diverged clusters of the tomato Cf-9 plant disease resistance gene family. Proc. Natl. Acad. Sci. USA 96, 5850–5855. [PMC free article] [PubMed]
  • Parniske, M., Hammond-Kosack, K.E., Golstein, C., Thomas, C.M., Jones, D.A., Harrison, K., Wulff, B.B.H., and Jones, J.D.G. (1997). Novel disease resistance specificities result from sequence exchange between tandemly repeated genes at the Cf-4/9 locus. Cell 91, 821–832. [PubMed]
  • Parniske, M., Wulff, B.B.H., Bonnema, G., Thomas, C.M., Jones, D.A., and Jones, J.D.G. (1999). Homologues of the Cf-9 disease resistance gene (Hcr9s) are present at multiple loci on the short arm of tomato chromosome 1. Mol. Plant-Microbe Interact. 12, 93–102. [PubMed]
  • Piedras, P., Rivas, S., Droge, S., Hillmer, S., and Jones, J.D.G. (2000). Functional, c-myc-tagged Cf-9 resistance gene products are plasma-membrane localized and glycosylated. Plant J. 21, 529–536. [PubMed]
  • Scofield, S.R., Tobias, C.M., Rathjen, J.P., Chang, J.H., Lavalle, D.T., Michelmore, R.W., and Staskawicz, B.J. (1996). Molecular basis of gene-for-gene specificity in bacterial speck disease of tomato. Science 274, 2063–2065. [PubMed]
  • Stemmer, W.P.C. (1994). Rapid evolution of a protein in vitro by DNA shuffling. Nature 370, 389–391. [PubMed]
  • Tang, X., Frederick, R.D., Zhou, J., Halterman, D.A., Jia, Y., and Martin, G.B. (1996). Initiation of plant disease resistance by physical interaction of AvrPto and Pto kinase. Science 274, 2060–2063. [PubMed]
  • Thomas, C.M., Jones, D.A., Parniske, M., Harrison, K., Balint-Kurti, P.J., Hatzixanthis, K., and Jones, J.D.G. (1997). Characterization of the tomato Cf-4 gene for resistance to Cladosporium fulvum identifies sequences that determine recognitional specificity in Cf-4 and Cf-9. Plant Cell 9, 2209–2224. [PMC free article] [PubMed]
  • Thomas, C.M., Dixon, M.S., Parniske, M., Golstein, C., and Jones, J.D.G. (1998). Genetic and molecular analysis of tomato Cf genes for resistance to Cladosporium fulvum. Philos. Trans. R. Soc. Lond. B Biol. Sci. 353, 1413–1424. [PMC free article] [PubMed]
  • Thomas, C.M., Tang, S., Hammond-Kosack, K.E., and Jones, J.D.G. (2000). Comparison of the hypersensitive response induced by the tomato Cf-4 and Cf-9 genes in Nicotiana spp. Mol. Plant-Microbe Interact. 13, 465–469. [PubMed]
  • Van den Ackerveken, G.F.J.M., Van Kan, J.A.L., and De Wit, P.J.G.M. (1992). Molecular analysis of the avirulence gene Avr9 of the fungal pathogen Cladosporium fulvum fully supports the gene-for-gene hypothesis. Plant J. 2, 359–366. [PubMed]
  • Van der Hoorn, R.A.L., Laurent, F., Roth, R., and De Wit, P.J.G.M. (2000). Agroinfiltration is a versatile tool that facilitates comparative analyses of Avr9/Cf-9-induced and Avr4/Cf-4-induced necrosis. Mol. Plant-Microbe Interact. 13, 439–446. [PubMed]
  • Zhao, H., and Arnold, F.H.. (1997). Optimization of DNA shuffling for high fidelity recombination. Nucleic Acids Res. 25, 1307–1308. [PMC free article] [PubMed]

Articles from The Plant Cell are provided here courtesy of American Society of Plant Biologists
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...