A Cluster of Four Receptor-Like Genes Resides in the Vf Locus That Confers Resistance to Apple Scab Disease

Corresponding author: Schuyler S. Korban, University of Illinois, 310 Madigan Bldg., 1201 W. Gregory Dr., Urbana, IL 61801., s-korban{at}uiuc.edu (E-mail)

Communicating editor: A. H. D. BROWN

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The Vf locus, derived from the crabapple species Malus floribunda 821, confers resistance to five races of the fungal pathogen Venturia inaequalis, the causal agent of apple scab disease. In our previous research, the Vf locus was restricted to a BAC contig of 290 kb covered by five overlapping BAC clones. Here, we report on cloning of the resistance gene(s) present in the Vf BAC contig using a highly reliable and straightforward approach. This approach relies on hybridization of labeled cDNAs to amplified inserts of subclones derived from BAC inserts, followed by recovery of full-size transcripts by rapid amplification of cDNA ends (RACE). A cluster of four resistance paralogs (Vfa1, Vfa2, Vfa3, and Vfa4) was identified in the Vf locus. Vfa1, Vfa2 and Vfa4 had no introns and are predicted to encode proteins characterized with extracellular leucine-rich repeats (LRRs) and transmembrane (TM) domains. However, Vfa3 contains an insertion of 780 bp at the end of the LRR motif, resulting in multiple truncated transcripts. Comparison of Vfa1, Vfa2, and Vfa4 paralogs revealed a high degree of overall homology in their deduced amino acid sequences, while divergences were mainly restricted within LRR domains, including variable LRR units, numerous amino acid substitutions, and several residue deletions/duplications. Differential expression profiles among the four paralogs were observed during leaf development. Vfa1, Vfa2, and Vfa3 were active in immature leaves, but slightly expressed in mature leaves, while Vfa4 was active in immature leaves and was highly expressed in mature leaves.

MORE than 30 disease resistance (R) genes have been isolated and characterized from a wide range of plant species (HULBERT et al. 2001 ). Numerous R genes are predicted to encode receptors that function in the recognition of corresponding "ligands" controlled by specific dominant avirulence (Avr) genes present in pathogens (HAMMOND-KOSACK and JONES 1997 ). This leads to an active plant defense response, previously described by FLOR 1956 as gene-for-gene interaction. Most R genes are organized as complex gene families clustered within particular chromosomal regions (WISE 2000 ). Although isolated R genes confer resistance to a wide range of pathogens (fungi, viruses, bacteria, and nematodes), they share various conserved functional domains and fall into several distinct classes (RICHTER and RONALD 2000 ; WISE 2000 ). The most prevalent class is the NBS-LRR family that contains a nucleotide-binding site (NBS) and multiple leucine-rich repeats (LRRs) (YOUNG 2000 ). The tobacco N gene; the Arabidopsis RPS2, RPM1, and RPP5 genes; and the flax L6 and M genes belong to this class (HAMMOND-KOSACK and JONES 1997 ). The tomato Cf genes represent the second class, which is characterized by extracytoplasmic LRRs and a C-terminal membrane anchor (DIXON et al. 1996 ; PARNISKE et al. 1997 ). The tomato Pto gene, conferring resistance to a bacterial pathogen, falls into a third class that encodes a serine/threonine protein kinase, but lacks LRRs (MARTIN et al. 1993 ) and requires an NBS-LRR gene (Prf) to function (SALMERON et al. 1996 ). The rice Xa21 gene, conferring resistance to a bacterial pathogen, is regarded as the fourth class, featuring both extracellular LRRs and a transmembrane protein kinase (SONG et al. 1995 ).

On the basis of the LRR protein structure model, a consensus sequence, xxLxLxx, within the LRR is predicted to form a ß-strand/ß-turn structure (JONES and JONES 1997 ). The aliphatic residues (L) project into the hydrophobic core, whereas the other residues (x) form a solvent-exposed surface that presumably makes direct contact with elicitors of a specific pathogen, leading to recognition of that pathogen (KOBE and DEISENHOFER 1995 ). Comparison of nucleotide sequences among closely related gene homologs has revealed that solvent-exposed residues are hypervariable (PARNISKE et al. 1997 ). The ratio of nonsynonymous to synonymous nucleotide substitutions (K_A:K_S ratio) of solvent-exposed residues is greater than one, thus indicating that these residues must have undergone adaptive evolution (MICHELMORE and MEYERS 1998 ). The ever-changing population of a pathogen imposes a selection pressure to continually alter recognition specificity (BERGELSON et al. 2001 ). In flax, domain swaps between alleles of the L6 gene have demonstrated that the LRR is an important determinant of specificity (ELLIS et al. 1997 ). Likewise, domain swaps between Pto and a closely related paralog, Fen, have identified a few amino acids in the ligand-binding domain that are critical for pathogen specificity (SCOFIELD et al. 1996 ). Recent experiments with the rice resistance gene Pi-ta and its Avr gene show specific binding between the LRR domain of the Pi-ta protein and the cognate Avr-Pi-ta protein (JIA et al. 2000 ). Active R genes with differential expression are involved in durable resistance of the Cf-9 locus to Cladosporium fulvum, whereby Cf-4 and Cf-9 genes are active from the seedling stage onward, while Hcr9s, -9E, and -9A or -9B are detectable in adult plants only (PARNISKE et al. 1997 ).

Apple scab, incited by the fungal pathogen Venturia inaequalis (Cke.) Wint., is one of the most serious diseases of apple (BENAOUF and PARISI 2000 ). Almost all apple cultivars grown commercially around the world are susceptible to this disease. Resistance to this disease was derived from a wild species of apple, Malus floribunda 821, and the chromosomal region containing a scab-resistance locus, Vf, has been widely introgressed into susceptible commercial apple cultivars (KORBAN 1998 ). A bacterial artificial chromosome (BAC) contig covering the Vf locus has been constructed and consists of a minimal tilling path of three BAC clones (XU and KORBAN 2002 ).

Identifying putative coding sequences in a BAC contig containing gene(s) of interest is critical in a map-based cloning strategy. Traditional available methods in pursuing this goal include the following: (1) direct screening of a cDNA library (ELVIN et al. 1990 ), (2) cDNA selection by hybridization (LOVETT et al. 1991 ; PARIMOO et al. 1991 ), (3) the CpG island approach (CROSS et al. 1999 ), and (4) the exon-trapping method (BUCKLER et al. 1991 ). More recently, sequencing an entire BAC insert and predicting open reading frames are becoming more common (ZHANG 1997 ). However, no single method is capable of identifying all transcribed sequences from a BAC contig. Frequently, a combination of methods is required to construct transcript maps for a defined chromosomal region (SOOD et al. 2001 ). We present herein an efficient and straightforward approach for identifying putative transcribed sequences contained within BAC inserts. This approach is based on hybridization of a labeled cDNA population to an array of amplified inserts of subclones constructed from BAC inserts containing Vf gene(s). A high signal-to-noise ratio in hybridization allows pinpointing positive subclones containing putative coding sequences. Sequencing inserts of positive subclones and comparing deduced amino acid sequences with those available in the GenBank database have revealed the existence of putative resistance genes within the Vf locus. A full-length cDNA sequence is then obtained, using rapid amplification of cDNA ends (RACE). This sequence information allows further cloning of genomic gene(s) from BAC clones containing Vf gene(s).

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Subcloning of BAC inserts into the plasmid vector pBlueScriptII-KS:
A BAC contig spanning the Vf locus contained five overlapping BAC clones (M4-P-11, M61-M-16, G11-J-23, G7-C-18, and G53-N-7; Fig 1). The BAC clone M61-M-16, located in the middle of the BAC contig, was selected to construct a subclone library. BAC DNA was extracted from a 2-liter culture using the Plasmid Maxi kit (QIAGEN, Valencia, CA), followed by purification using CsCl-gradient centrifugation (SAMBROOK et al. 1989 ). A total of 4 µg of purified BAC DNA was partially digested with the restriction enzyme Sau3AI. The reaction was performed at 37° for 15 min in a total volume of 150 µl, using 3 units of Sau3AI. Digested fragments of 2 kb in size were harvested from a 1% agarose gel using a QIAEX II gel extraction kit (QIAGEN) and then ligated into a BamHI-digested pBlueScriptII-KS vector. A total of 1 µl of ligation product was used to transform competent Escherichia coli cells by electroporation using a Bio-Rad (Richmond, CA) gene pulser. Transformed cells were grown onto Luria broth (LB) plates containing ampicillin (100 µg/ml), X-gal, and isopropyl thiogalactoside. White colonies were picked and placed in a grid onto fresh LB plates. Following an overnight incubation at 37°, each colony was dissolved in 30 µl TE buffer, heated at 100° for 10 min, and centrifuged at 4000 rpm for 20 min. A pair of PCR primers (T3 and T7 promoters), flanking the multiple cloning site of pBlueScriptII-KS, was introduced to amplify inserts of subclones using their DNA-containing supernatants as template DNAs. A total of 1 µl of the PCR product was denatured, using an equal volume of 1 N NaOH, and placed in a grid onto a nylon membrane for hybridization. A total of 96 (12 x 8) dots were arrayed on a single membrane.

View larger version (18K): In this window In a new window Download PPT slide   Figure 1. A 200-kb BAC contig of the Vf locus and schematic representation of the four Vf gene paralogs.

cDNA synthesis from apple leaf tissues:
Fresh leaf tissues of a Vf-containing apple cultivar, "GoldRush," were sprayed with 2 mM salicylic acid and ddH₂O, respectively. Salicylic acid is known to activate resistance gene(s) and its downstream defense system in plant tissues. Therefore, it is used in this study to induce transcripts for disease resistance in apple leaf tissues of GoldRush. Following a 24-hr incubation at room temperature, leaves from both treatments were harvested and stored at -80° in an ultra-freezer. Total RNA was extracted from 2 g of leaf tissue using the LiCl method (SAMBROOK et al. 1989 ), and mRNA was isolated using a Poly(A)Tract mRNA Isolation System III (Promega, Madison, WI). The isolated mRNA was then used to synthesize cDNA, using a Marathon cDNA amplification kit (CLONTECH, Palo Alto, CA).

Identification of subclones containing putative transcribed sequences:
cDNAs from both salicylic acid- and ddH₂O-treated leaf tissues were mixed and labeled with a ³²P isotope, using a Random Primers labeling system (GIBCO BRL, Gaithersburg, MD). Prehybridization, hybridization, and membrane washing were performed according to SAMBROOK et al. 1989 . Membranes were exposed to a hyperfilm (Amersham, Buckinghamshire, England) for 3 days at -80°. Positive subclones were then spotted onto the film, and their corresponding recombinant plasmids were extracted using the Mini Preparation kit (QIAGEN). Inserts of positive subclones were sequenced from both ends, using T3 and T7 promoter primers. Analysis of insert sequences was performed for all positive subclones to identify putative transcribed sequences present within the BAC insert. Resistance gene homologs were predicted after comparing deduced amino acid sequences with the GenBank database using BLAST.

PCR-based search for additional Vf homologous sequences:
A two-step PCR-based strategy was pursued. First, a pair of common primers was designed on the basis of conserved regions among the Vf gene homologs found in the BAC clone M61-M-16. The common primers were then used to amplify all five BAC clones spanning the Vf locus, along with their neighboring BAC clones, to search for additional Vf homologous sequences present in the Vf region. PCR products were cloned into pBlueScriptII-KS, followed by extensive investigation of fingerprints of cloned PCR products using an array of restriction enzymes, including EcoO109I, MspI, MobI, HinfI, HincII, Sau3AI, and TaqI. Several fingerprinting patterns were observed, and for each fingerprinting pattern, at least two clones were selected for sequencing. Second, a pair of nested common primers was designed on the basis of the newly identified conserved sequences and used to further amplify all BAC clones in the Vf region. Cloning and analysis of PCR products were performed as described in the first step. Using this two-step PCR-based search approach, all putative Vf homologous sequences in the Vf region were recovered.

It is highly likely that PCR artifacts are generated and cloned during the PCR-based search strategy. Thus, clones harboring PCR artifacts must be ruled out; otherwise, they will interfere with identifying authentic Vf homologous sequences. This can be achieved by a simple PCR reaction. A pair of specific primers for each putative Vf homologous sequence is designed on the basis of its polymorphic nucleotides and used to amplify three sources of template DNAs: the cognate PCR product clone, the cognate BAC clone, and the genomic DNA of GoldRush. A putative Vf gene homolog is deemed authentic only if uniform PCR products are present in all three sources of template DNAs, using its specific primers. In contrast, a putative Vf gene homolog is considered a PCR artifact when specific primers can amplify only the cognate PCR product clone, but not the cognate BAC clone and the genomic DNA of GoldRush.

Recovery of full-length cDNA using RACE:
The RACE method was utilized to construct and characterize cDNAs transcribed by Vf gene paralogs, including putative start and termination points for both transcription and translation, sizes of the open reading frames, and leader and trailer sequences. Sequence information from RACE analysis was further used to clone the full-length cDNA and genomic gene for each Vf paralog. For RACE analysis, both 3' and 5' primers were designed on the basis of conserved regions of the four Vf paralogs and with an overlapping region of 1.5 kb. A mixture of cDNAs from both immature and mature leaves of GoldRush was used as template DNA to amplify both 3' and 5' cDNA ends. RACE products of expected sizes were cloned into a pT-Adv vector. Therefore, both 3' and 5' RACE clones were a mixture of RACE clones including all active Vf gene paralogs. Identifying a RACE clone containing an individual Vf gene paralog was accomplished using Vfa-specific primers.

Expression profiles of Vf gene paralogs revealed by reverse transcription-PCR:
Three sources of cDNAs were synthesized from respective mature leaves of a scab-susceptible apple cultivar "Red Delicious" and immature and mature leaves of the Vf-containing resistance apple cultivar GoldRush. The nested common primers as well as Vfa-specific primers were used to amplify the three sources of cDNAs to determine expression profiles for each of the four Vf gene paralogs. Negative controls were also set up separately by using corresponding mRNAs (100 ng each) from these tissues in PCR reactions, utilizing both common and Vfa-specific primers, to rule out any genomic DNA contamination prior to running reverse transcription (RT)-PCR. Each of the RT-PCR analyses was repeated a second time to confirm repeatability of results.

Sequencing genomic Vf gene paralogs:
Genomic genes of four Vf paralogs were isolated from two BAC clones, M61-M-16 and G7-C-18. A total of 4 µg of purified BAC DNA was partially digested at 37° for 15 min with 0.5 units of Sau3AI in a total volume of 150 µl. The digested fragments of 10 kb in size were harvested from a 1% agarose gel using the QIAEX II gel extraction kit (QIAGEN) and then ligated into a BamHI-digested pCAMBIA2301 vector. A total of 1 µl of ligation product was used to transform competent E. coli cells by electroporation, using a Bio-Rad gene pulser. Both the nested common and Vfa-specific primers were used to screen all subclones to find those containing Vf gene paralogs, each with a full-length coding sequence, at least 2-kb promoter, and 3'-flanking region. Promoter and coding regions were sequenced for each of the four Vf gene paralogs.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of positive subclones containing putative transcribed sequences:
PCR amplification of subclone inserts showed that >90% of subclones derived from the BAC clone M61-M-16 contained inserts of 2 kb in size. A total of 190 amplified inserts were denatured and arrayed onto two hybridization membranes. The insert of the BAC clone M61-M-16 was estimated to be 115 kb in size (XU and KORBAN 2002 ). Therefore, all 190 subclones, with an average insert size of 2 kb, represented more than three times the BAC insert, thus providing a 95% probability of covering the entire BAC insert. Hybridization of dot-blot membranes with labeled cDNA revealed a total of 30 positive subclones with strong signals (Fig 2). Of these 30 positive subclones, 15 were randomly selected for sequencing. Analysis of insert sequences revealed at least four open reading frames. The deduced amino acid sequences were compared to those found in the GenBank database using BLAST. Two amino acid sequences highly matched some resistance genes. They showed 29% sequence identity and 31% sequence similarity to Cf2/Cf5 resistance genes in tomato and 31% sequence identity and 35% sequence similarity to some putative disease resistance genes in Arabidopsis, respectively. These two open reading frames were deemed likely Vf gene homologs and were designated as Vfa1 and Vfa2. The third open reading frame was similar to two unknown proteins in Arabidopsis, while the fourth open reading frame was likely to be a transposase. Only the two resistance-like genes were subjected to further analysis.

View larger version (41K): In this window In a new window Download PPT slide   Figure 2. Dot-blot hybridization to pinpoint subclones harboring putative coding sequences. A subclone library is constructed from a Vf-containing BAC clone, M61-M-16. Inserts of the subclone are amplified using common primers flanking the cloning sites of the pBluescript II-KS. PCR products are denatured and then dot blotted onto membranes (12 x 8 dots per membrane). Hybridization is carried out using labeled cDNAs derived from the Vf-containing apple cultivar GoldRush. Positive subclones with very high signal-to-noise ratios have been obtained, and inserts of positive subclones are sequenced to identify putative coding regions. The large dot (indicated by an arrow) corresponds to M61-M-16 BAC DNA used as control.

Recovery of all putative resistance gene homologs from the Vf region:
Sequence alignment between Vfa1 and Vfa2 revealed a deletion of 162 bp in the Vfa2, a number of polymorphic single nucleotides, and several conserved regions flanking the deletion. A pair of PCR primers was designed on the basis of flanking conserved regions to search for additional putative Vf gene homologs along the Vf locus (Table 1). Five overlapping BAC clones in the Vf locus, along with their neighboring BAC clones, were subjected to PCR amplification. As a result, all five BAC clones and one neighboring BAC clone, M24-B-18, gave rise to PCR products. Interestingly, PCR products from different BAC clones varied in both band numbers and sizes. From left to right, M4-P-11 had a single band of 1.2 kb; both M61-M-16 and G11-J-23 had two bands of 1.2 and 1.04 kb in size; G7-C-18 had a single band of 1.04 kb; and both G53-N-7 and M24-B-18 gave rise to two bands of 1.04 and 0.8 kb in size. The PCR products from different BAC clones were individually cloned into pBlueScriptII-KS, and restriction fingerprints of cloned PCR products were analyzed with a series of restriction enzymes. A total of five banding patterns were observed, indicating that at least five different PCR products had been generated. At least two clones for each banding pattern were selected for sequencing. Accordingly, five putative Vf gene homologous sequences were obtained. Sequence alignment of these five Vf homologs revealed some conserved regions that were available for designing a pair of nested common primers (Table 1). The nested common primers were used to amplify the same set of BAC clones. The resulting PCR products were cloned into pBlueScriptII-KS and analyzed using a series of restriction enzymes. Following sequencing and sequence alignment, three new putative Vf homologous sequences were identified. In total, eight putative Vf gene homologs (types 1–8) were recovered from the two-step PCR-based search strategy.

Sequence alignment showed some interesting features present in these eight putative Vf homologous sequences (Fig 3). Types 1 and 3 were derived from Vfa1 and Vfa2, respectively, while type 2 was a chimeric sequence between types 1 and 3. The remaining five types were all derived from BAC clones G53-N-7 and M24-B-18. Of these five putative Vf homologous sequences, types 4 and 7 were deemed novel. Types 5 and 6 were chimeric sequences between types 4 and 7, while type 8 shared the same sequence as type 7, except for a single-base-pair difference. Considering that only a small fraction of PCR product clones harbored chimeric sequences, and that all chimeric sequences were derived from those BAC clones containing at least two putative Vf homologous sequences (M61-M-16, G11-J-23, G7-C-18, and M24-B-18), it was speculated that such chimeric sequences were derived from PCR artifacts. To confirm this, a pair of specific primers was designed for each Vf homolog on the basis of its polymorphic nucleotides and used to amplify its cognate PCR product clones, cognate BAC clone, and genomic DNA of GoldRush. As a result, specific primers for types 1, 3, 4, and 7 amplified their corresponding three sources of template DNAs. This indicated that these four putative Vf homologous sequences were present in both BAC clones and genomic DNA of GoldRush. However, for the three chimeric sequences, PCR products were observed only in the cognate PCR product clones and not in cognate BAC clones or in genomic DNA of GoldRush. This indicated that these were derived from PCR artifacts. A single polymorphic nucleotide present in type 8 was likely introduced during the PCR reaction, as only one PCR product clone that harbored the sequence of type 8 was obtained.

View larger version (42K): In this window In a new window Download PPT slide   Figure 3. Alignment of eight types of Vf gene homologous sequences that have been recovered from the Vf locus by using a two-step PCR-based search strategy. Only polymorphic nucleotides are shown. Types 1, 3, 4, and 7 are confirmed to be derived from the Vf locus, whereas types 2, 5, 6 (chimeric), and 8 (point mutation) are deemed PCR artifacts. Numbers written vertically along the top of conserved sequences correspond to nucleotide positions. Dots (...) indicate identical nucleotides with the consensus sequence. Dashes (- - -) correspond to missing nucleotides. Solid and open rectangles show identical sections among different Vf homologous sequences.

Following this extensive PCR-based search and validation strategy, only four Vf gene homologous sequences were found clustered within the Vf locus. The two novel Vf homologs were designated Vfa3 and Vfa4. Since all four Vf gene homologous sequences were derived from the Vf locus, these were then deemed as Vf gene paralogs. The left-most BAC clone M4-P-11 and the subsequent two clones, M61-M-16 and G11-J-23, contained Vfa1, suggesting that Vfa1 was the left-most Vf paralog. This was followed by Vfa2, which was present in three BAC clones, M61-M-16, G11-J-23, and G7-C-18. The flanking region of Vfa3 shared some restriction fingerprints with the BAC clone G7-C-18, thus suggesting Vfa3 was the third Vf paralog (data not shown). The Vfa4 was then located as the right-most Vf paralog. Both Vfa3 and Vfa4 were present in G7-C-18 and M24-B-18 (Fig 1).

Characterization of cDNA sequences of the four Vf paralogs using RACE:
With partial sequences of the four Vf paralogs in hand, 5' and 3' Vf gene primers, with an overlapping region of 1.5 kb, were designed for RACE analysis (Table 1). Each primer was 25 bp in length and contained a high G + C content to enhance PCR specificity. Screening of RACE product clones was first carried out using the nested common primers and then using each of the four pairs of Vf paralog-specific primers (Table 1). Since the nested common primers and all Vfa-specific primers were located within the overlapping region, these primers were available for screening both 5' and 3' RACE clones. In this RACE analysis, both 5' and 3' RACE clones for each of the four Vf paralogs were obtained, suggesting that all four Vf paralogs were transcribed. For each Vf paralog, both 5' and 3' RACE products were sequenced. Putative start codons for the four Vf paralogs were identified following sequence analysis of 5'-end cDNA fragments having the largest sizes. Leader sequences, upstream of the putative start codon for the four Vf paralogs, were estimated to be 25 bp. All 3'-end cDNA fragments had typical polyadenylation signals, indicating the fidelity of the RACE reaction. Sizes of 3' untranslated regions (UTRs) varied not only among different Vf paralogs, but also among different 3' RACE clones of the same Vf paralog. For instance, we sequenced seven 3' RACE clones of Vfa4, and sizes of UTRs were found to be 84 bp for one clone, 250 bp for two clones, and 354 bp for the remaining four clones. Generally, sizes of UTRs ranged from 100 to 400 bp.

A full-length cDNA sequence was reconstructed for each of the four Vf paralogs by merging sequences of 5'- and 3'-end cDNA fragments on the basis of their overlapping regions. The sizes of coding sequences were estimated to be 3048 bp for Vfa1, 2943 bp for Vfa2, and 2748 bp for Vfa4, respectively. Vfa3 had multiple truncated transcripts resulting from premature termination of transcription. The shortest truncated transcript was only 1220 bp. Moreover, a nonsense mutation was present at 228 bp downstream of the start codon in the Vfa3 transcripts.

When compared with the database in the GenBank, the deduced amino acid sequences of Vfa1, Vfa2, and Vfa4 show high homology to Cf resistance genes in tomato and can be subdivided into similar functional domains (Fig 4). Domain A is a putative signal peptide of 23 amino acids and is likely predicted to be removed in the mature protein. Domain B is the presumed NH₂ terminus of the mature protein. Domain C contains multiple N-terminal LRRs. Domain D is of unknown function. Domain F is composed of hydrophobic residues and is likely predicted to be the transmembrane (TM) domain. Domains E and G are rich in acidic and basic residues, respectively, and presumably function in anchoring and directing the transmembrane domain to the cell membrane. Thus, the Vf paralogs belong to a LRR-TM resistance gene class and encode proteins characterized with extracellular LRRs and TM domains.

View larger version (58K): In this window In a new window Download PPT slide   Figure 4. Comparison of amino acid sequences of Vfa1, Vfa2, and Vfa4. The amino acid sequences of the three Vf paralogs are divided into seven domains, similar to those of Cf genes in tomato. Domain A is a signal peptide; B is the NH2 terminus of a mature protein; C is an LRR domain, and the core structure of xxLxLxx is highlighted; D is of unknown function; E is rich in acid residues; F is a transmembrane domain; and G is rich in basic residues. A high rate of divergence is present in the LRR domain, including deletion of LRR units, residue deletions/duplications, and amino acid substitutions.

At the amino acid level, a high degree of overall sequence homology was observed among Vfa1, Vfa2, and Vfa4. Sequence identity was 84% for the pair Vfa1/Vfa2, 78% for the pair Vfa1/Vfa4, and 82% for the pair Vfa2/Vfa3. Just like other resistance genes identified in various plant species, amino acid sequence divergence was mainly present in the LRR motifs, including LRR copy numbers, amino acid substitutions, deletions, and duplications (Fig 4). The LRR copy numbers for Vfa1, Vfa2, and Vfa4 were 30, 29, and 26, respectively. Several amino acid deletions/duplications were scattered along the LRR motifs. Pairwise comparisons of the three Vf paralogs within the LRR motifs revealed 26 sites with three different amino acid residues and 106 positions with two variable amino acid residues. Compared to LRR motifs, C-terminal domains (D, E, and F) were highly homologous among Vfa1, Vfa2, and Vfa4, and only a few amino acid substitutions were observed.

Upon closer examination of the amino acid sequences of the three full-length Vfa paralogs, Vfa1, Vfa2, and Vfa4, the 6 amino acids on either side of the first four core sequences (xxLxLxx) show distinct differences (Fig 4). Those amino acids on the N-terminal side are highly conserved (only 1 of 24 amino acids is polymorphic), whereas those on the C-terminal side are highly polymorphic (18 of 24 amino acids are polymorphic). By contrast, when considering the 3 amino acids on either side of core sequences 20–27 (counting from the N-terminal end), it is revealed that those on the N-terminal side contain 12 polymorphic amino acids of 24 residues while those on the C-terminal side contain only 1 polymorphic amino acid among 24 residues (Fig 4). Whether the high level of polymorphism in particular locations in the amino acid sequences of the Vf paralogs is the result of adaptive evolution (because the residues are solvent exposed) or is a result of mutations that survived by chance and have a neutral effect (because the altered residues are not solvent exposed) should be further explored.

Expression profiles of the four Vf paralogs:
RT-PCR was used to investigate expression patterns of the four Vf gene paralogs during leaf development. The nested common primer and Vf paralog-specific primers were used to amplify cDNAs and their corresponding mRNAs derived from mature leaves of a scab-susceptible cultivar, Red Delicious, as well as from immature and mature leaves of a scab-resistant cultivar, GoldRush. No PCR products were observed in standard PCR reactions (negative controls) using mRNAs as templates, thus indicating no genomic DNA contamination occurred during mRNA preparations.

The nested common primer amplified all three sources of cDNAs (Fig 5). As cDNA of Red Delicious gave rise to the largest PCR band, this indicated the presence of some active Vf homologous sequences of large sizes in the susceptible cultivar. Interestingly, cDNAs from both immature and mature leaves of GoldRush gave rise to variable PCR products. The bulk of RT-PCR products from immature leaves of GoldRush were estimated to be 1.2 kb in size and corresponded to Vfa1, whereas most RT-PCR products from mature leaves of GoldRush were 0.8 kb in size and corresponded to Vfa4. The four Vf paralog-specific primers further revealed expression profiles of the four Vf gene paralogs. Three Vf paralogs, Vfa1, Vfa2, and Vfa3, were expressed in immature leaves, but were only slightly detectable in mature leaves of GoldRush, whereas Vfa4 was active in immature leaves and was highly expressed in mature leaves (Fig 5). This indicated that Vf gene paralogs were differentially expressed during leaf development.

View larger version (99K): In this window In a new window Download PPT slide   Figure 5. Analysis of expression profiles of the four Vfa paralogs by RT-PCR. PCR was performed using either cDNAs (lanes 1, immature leaves of a scab-susceptible apple cultivar Red Delicious; lanes 2, immature leaves of a Vf-containing scab-resistant cultivar GoldRush; and lanes 3, mature leaves of GoldRush) or genomic DNA (lanes 4, Red Delicious; lanes 5, GoldRush; and lanes 6, M. floribunda 821, the original source of Vf). M, 1-kb DNA ladder marker. (A) PCR amplification using the nested common primers, designed on the basis of conserved regions of LRR domains. The sizes of PCR products were expected to be 1199 bp for Vfa1, 1037 bp for Vfa2, 1049 bp for Vfa3, and 836 bp for Vfa4. The bulk of RT-PCR products in immature leaf tissue of GoldRush corresponded to Vfa1 (lanes 2), while most RT-PCR products in mature leaf tissue of GoldRush corresponded to Vfa4 (lanes 3). RT-PCR products of large sizes were also observed in cDNAs from the scab-susceptible cultivar Red Delicious (lanes 1). (B) PCR amplification using Vfa-specific primers. (a) Vfa1-specific primers; (b) Vfa2-specific primers; (c) Vfa3-specific primers; and (d) Vfa4-specific primers. Vfa1, Vfa2, and Vfa3 are strongly expressed in immature leaf tissues (a–c, lanes 2), but are weakly expressed in mature leaf tissue (a–c, lanes 3) of GoldRush, whereas Vfa4 is expressed in immature leaf tissue and is strongly expressed in mature leaf tissues of GoldRush (d, lane 3).

Characterization of four Vf gene paralogs:
The four genomic Vf paralogs were cloned from their corresponding BAC clones, each including at least a 2-kb promoter region, the full-length coding sequence, and 5–12 kb of the 3'-flanking region. Sequencing of the coding and the promoter regions was performed for each Vf paralog. Alignment of coding regions between cDNA and genomic sequences revealed that Vfa1, Vfa2, and Vfa4 were intronless genes and exhibited a 100% match between cDNAs and genomic genes. Furthermore, putative start and termination codons as well as leader and trailer sequences derived from cDNA analysis were further confirmed using genomic sequences, whereas Vfa3 had a 780-bp insert with a high A + T content at the end of the LRR motif, and the insert was assumed to cause multiple premature terminations of RNA transcription. The insertion produced 16-bp (5'-AATTTATCGAATAATC-3') direct repeats of the target sequences flanking the insert, suggesting that the insert was a transposable-like element. No sequences were found to be homologous with this insert following BLAST search with the database in GenBank.

The promoter regions of the four Vf paralogs are high homologous for 300 bp upstream of the putative start codon, but beyond that they show wide divergence (Fig 6). A TATA box is clearly identified 63 bp upstream of the putative start codon. Several nucleotide deletions, duplications, and substitutions are scattered along the promoter regions. The Vfa1, Vfa2, and Vfa3 promoters are very similar in expression profiles, but exhibit a certain degree of nucleotide divergence. Two deletions in Vfa1 (216 and 147 bp upstream of the start codon, respectively), one nucleotide duplication in Vfa2 (186 bp upstream of the start codon), and several nucleotide substitutions for all Vfa1, Vfa2, and Vfa3 have been observed. It can be concluded that these sequence divergences in the promoter regions have no significant impacts on their functions. The most striking divergence is a 5'-TCCCT-3' direct duplication immediately upstream of the TATA box in Vfa4. This duplication may be strongly related to the high expression level of the Vfa4 promoter in mature leaves of GoldRush.

View larger version (21K): In this window In a new window Download PPT slide   Figure 6. Alignment of promoter sequences (300 bp upstream of the start codon) for the four Vfa gene paralogs. Start codon and TATA box are indicated by solid rectangles. A direct repeat of 5'-TCCCT-3' is present immediately upstream of the TATA box in Vfa4 and is presumed to enhance the expression level of Vfa4 at the mature stage of leaf development.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Search for putative coding sequences from BAC clones:
To pinpoint putative coding sequences within a BAC contig spanning the target gene(s) is a crucial step for map-based gene cloning. Traditional methods often fall short of this expectation and frequently result in delays in cloning of the target gene(s). Identification of positive cDNA clones from a cDNA library usually requires screening of a large number of cDNA clones using labeled BAC inserts. Often, it is difficult to obtain true positive cDNA clones due to low abundance of target cDNA clones or interference of homologous sequences (ELVIN et al. 1990 ; VINATZER et al. 2001 ). The exon-trapping approach is useful only when a gene contains introns that can be recognized by the host cell (BUCKLER et al. 1991 ). Gene discovery by isolating CpG islands is mainly used in conjunction with other methods as CpG islands are linked to 60% of human genes (CROSS et al. 1999 ). cDNA selection by hybridization is an expression-dependent method (LOVETT et al. 1991 ; PARIMOO et al. 1991 ) and therefore falls short in constructing a transcript map of a BAC insert. While sequencing the entire BAC insert to predict putative coding regions is an efficient method, it is costly and offers an indirect approach for gene discovery.

The strategy presented in this study is reliable, direct, and quite powerful to pinpoint all potential coding sequences within a BAC insert. A key factor in this strategy is the use of amplified inserts of subclones rather than colonies in hybridization. This modification dramatically reduces the strong background that is often detected in colony hybridization and results in positive clones with much higher signal-to-noise ratios. Sequencing inserts of positive subclones and comparing their deduced amino acid sequences with those in the GenBank database allow rapid identification of transcripts contained within BAC inserts. Like other expression-dependent methods, hybridization using labeled cDNAs involves similar challenges as gene discovery is strongly dependent on the sources of cDNA populations (LOVETT et al. 1991 ; PARIMOO et al. 1991 ). As our aim is to search for Vf resistance gene(s) that are expressed in leaf tissue, only cDNAs from the leaf tissue of a Vf-containing apple cultivar, GoldRush, are used. For the purpose of constructing a transcription map, all sources of cDNAs or normalized cDNA populations are required. Positive subclones with strong signals have indicated the presence of a large portion of a gene that is highly expressed in the leaf tissue. In contrast, subclones with faint signals contain either a short fragment of a strongly expressed gene or a large fragment of a weakly expressed gene.

To recover all potential Vf homologous sequences from the Vf locus, a two-step PCR-based screening approach is introduced. This PCR-based strategy is independent of gene expression and therefore provides a complementary approach for revealing all Vf homologous sequences. As demonstrated in this study, the PCR-based search strategy has proved to be quite powerful. Not only eight putative Vf homologous sequences are identified within the Vf locus, but also three other Vf haplotypes are recovered in the M. floribunda 821 genome, one of which is very close to the Vf locus (data not shown). One problem encountered in this PCR-based search is the occurrence of PCR artifacts, as these interfere with identification of authentic Vf homologous sequences. This is also a major problem encountered in a resistance gene analog (RGA) search strategy whereby degenerate PCR primers are utilized to identify all possible RGAs spread throughout a whole genome (LEISTER et al. 1996 ). Therefore, it is necessary to validate the authenticity of the Vf homologous sequences in the PCR-based search strategy to rule out all PCR artifacts. Our approach of using specific primers to amplify different sources of template DNAs has proved to be effective and highly reliable.

The RACE technique rapidly enables construction of a full-length cDNA sequence based on a partial transcribed sequence. In addition, the usefulness of this technique can be expanded to search for all active members of a gene family and to identify related genes on the basis of conserved sequence information. During cloning of 5' RACE products, multiple clones of different sizes have been recovered due to premature termination of the first-strand cDNA synthesis. This is an intrinsic limitation of the RACE technique (CHENCHIK et al. 1995 ). Therefore, several clones having large 5' RACE products must be sequenced before the putative start codon can be identified. All 3' RACE products have typical poly(A) signals; however, 3' UTRs are variable among the different Vf paralogs as well as among clones in the same paralog. Highly homologous sequences downstream of stop codons are present among the Vf paralogs, and several sequence blocks are rich in A + T content. It is likely that RNA terminal riboadenylate transferase binds to different A + T-rich blocks, thus resulting in the recovery of variable sizes of UTRs. The sequence information revealed by RACE analysis is then used to clone the full-length cDNA and its corresponding genomic gene.

Apart from resistance-like genes involved in the Vf locus, other genes are also present in this region. Whether or not these closely linked genes are related to the plant defense system is yet unknown and must be investigated. Our results indicate that the Vf locus is a gene-rich region. This is consistent with the finding of VINATZER et al. 2001 where >50 expressed sequences have been recovered during screening of positive cDNA clones using labeled BAC inserts.

Common and specific PCR primers:
We have taken full advantage of the sequence information of the four Vf paralogs to design both common and specific primers. In the first half of the LRR domain, deletion of complete LRR units is observed whereby both Vfa2 and Vfa3 lack two LRRs, while Vfa4 lacks five LRRs. Conserved sequences flanking the deletion region have been used to design three common primers (Table 1). These common primers not only have amplified the four Vf paralogs and other Vf homologous sequences, but also on the basis of the sizes of the PCR products have allowed determination of the identities of the Vf homologs. The polymorphic sequences of each Vf paralog have been used to design Vfa-specific primers, and this has allowed pinpointing of individual Vf paralogs. As demonstrated in this study, utilization of both common and Vfa-specific primers in PCR-based search, RACE, RT-PCR, and screening of genomic genes has dramatically accelerated our efforts to clone and characterize the Vf paralogs. Moreover, additional information has been unveiled by using these primers. During RACE analysis, a total of 48 positive RACE clones have been identified using the nested common primers. Of these, 38 clones are confirmed to be derived from the Vf locus by using the four Vfa-specific primers. This suggests that the Vf locus is very active in the resistance to apple scab. The remaining 10 positive RACE clones that could not be confirmed with Vfa-specific primers must then be derived from other chromosomal regions containing Vf homologous sequences. Likewise, during RT-PCR analysis, RT-PCR products have been generated from the susceptible apple cultivar Red Delicious by using the nested common primers. All these findings suggest that some other Vf homologous sequences, apart from the Vf paralogs, are also being expressed in the leaf tissue. Whether or not these expressed sequences also confer resistance to apple scab or to some other diseases must be investigated.

Four Vf paralogs are the only resistance genes present in the Vf locus:
After sequencing of 15 positive clones followed by an intensive PCR-based search, a total of four Vf paralogs were identified in the Vf locus. These four Vf paralogs belong to the LRR-TM resistance gene class and share many common features with the Cf genes found in tomato (DIXON et al. 1996 ; PARNISKE et al. 1997 ). Is there any other class of resistance gene(s) interspersed within the Vf locus? To answer this question, first we used degenerate primers derived from the major resistance gene class of NBS-LRR to amplify BAC clones covering the Vf locus. However, no NBS-LRR resistance-like sequences have been obtained. Second, the four Vfa-specific primers have been used to screen an apple-mapping population consisting of 468 individuals (XU and KORBAN 2000 ), and no recombinant was detected between Vf resistance and the four Vf paralogs. Third, for all R gene families reported thus far, only members of the same resistance gene class tend to cluster within the same locus (DIXON et al. 1996 , DIXON et al. 1998 ; PARNISKE et al. 1997 ; MEYERS et al. 1998 ; WISE 2000 ). All the above findings strongly support the fact that no genes from other resistance classes are interspersed within the Vf locus, and the four Vf paralogs are the only resistance genes clustered within the Vf locus. The coding sequences of Vfa1 and Vfa2 are identical with those of HcVf1 and HcVf2, previously reported by VINATZER et al. 2001 . The Vfa3 and Vfa4 paralogs are novel and have not been previously reported.

Possible roles of the Vf paralogs within the Vf locus:
A resistance gene belonging to the LRR-TM class first recognizes the elicitor of a specific pathogen and then passes down this signal to such elements as protein kinases to induce a plant's defense system. The observed high homology in the C-terminal region among Vfa1, Vfa2, and Vfa4 suggests that these three Vf paralogs activate the same defense system in apple.

Three active Vf paralogs, Vfa1, Vfa2, and Vfa4, in the Vf locus exhibited differential expression profiles during leaf development. This suggested that a natural pyramiding of genes must have occurred within the Vf locus. Transferring these Vf paralogs individually into apple via genetic transformation, followed by analysis of scab resistance profiles of transgenic lines, will elucidate the functionality of these Vf paralogs and assess their role(s) in the durability of scab resistance in apple leaf tissues.

All apple cultivars, regardless of their resistance or susceptibility to apple scab, show resistance to apple scab in older mature leaves, commonly referred to as ontogenic resistance (GESSLER and STUMM 1984 ). Thus, it can be speculated that some factors may contribute to scab resistance in older mature leaves. One such factor is the physical barrier, such as the strength of cell walls and/or cuticle layers of mature leaves. This will either prevent the pathogen from penetrating into cells of the cuticle layer or lead to accumulation of cutinase inhibitors, thus blocking development of subcuticular stromata of the fungal pathogen and preventing its spread (VALSANGIACOMO and GESSLER 1988 ; KOLLER et al. 1991 ). Another factor may involve the presence of some unknown Vf-like resistance genes that are active during late stages of leaf development. In this study, RACE and RT-PCR have demonstrated the presence of some Vf orthologs that are outside the Vf locus and are transcribed in mature leaves. It is likely that these Vf orthologs may contribute to resistance to apple scab in mature leaves of susceptible apple cultivars. However, this may be different in the wild crabapple species M. floribunda 821, the original source of the Vf locus. As the Vf locus confers sustained resistance to apple scab, it seems unlikely for M. floribunda 821 to carry other highly functional Vf orthologs. The Vf haplotypes detected in the PCR-based search may mainly serve as reservoirs for generating novel resistance specificities to counterattack the ever-changing populations of the pathogen. After the Vf locus is incorporated into susceptible apple cultivars via genetic transformation, it is anticipated that a more durable resistance will be present in both immature and mature leaves.

To date, no single recombination event among the three active Vf paralogs has been observed. This indicates that the Vf locus is usually inherited intact. However, individuals containing the Vf locus have exhibited a wide range of symptoms (CHEVALIER et al. 1991 ), implying that genetic backgrounds may play a significant role in scab resistance. This may be achieved through either modification in transcription and translation of the Vf paralogs or alteration in the expression of genes involved in the defense system in the apple genome.

	ACKNOWLEDGMENTS

This work has been supported by funds received from the Illinois Council for Food and Agricultural Research external program.

Manuscript received July 12, 2002; Accepted for publication September 25, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BÉNAOUF, G. and G. PARISI, 2000  Genetics of host-pathogen relationships between Venturia inaequalis races 6 and 7 and Malus species. Phytopathology 90:236-242.[Medline]

BERGELSON, J., M. KREITMAN, E. A. STAHL, and D. TIAN, 2001  Evolutionary dynamics of plant R-genes. Science 292:2281-2285.[Abstract/Free Full Text]

BUCKLER, A. J., D. D. CHANG, S. L. GRAW, J. D. BROOK, and D. A. HABER et al., 1991  Exon amplification: a strategy to isolate mammalian genes based on RNA splicing. Proc. Natl. Acad. Sci. USA 88:4005-4009.[Abstract/Free Full Text]

CHENCHIK, A., F. MOQADAM, and P. SIEBERT, 1995  Marathon cDNA amplification: a new method for cloning full-length cDNAs. CLONTECHniques 10:5-8.

CHEVALIER, M., Y. LESPINASSE, and S. RENAUDIN, 1991  A microscopic study of different classes of symptoms coded by the Vf gene in apple for resistance to scab (Venturia ineaqualis). Plant Pathol. 40:249-256.

CROSS, S. H., V. H. CLARK, and A. P. BIRD, 1999  Isolation of CpG islands from large genomic clones. Nucleic Acids Res. 27:2099-2107.[Abstract/Free Full Text]

DIXON, M. S., D. A. JONES, J. S. KEDDIE, C. M. THOMAS, and K. HARRISON et al., 1996  The tomato Cf-2 disease resistance locus comprises two functional genes encoding leucine-rich repeat proteins. Cell 84:451-459.[Medline]

DIXON, M. S., K. HATZIXANTHIS, D. A. JONES, K. HARRISON, and J. D. G. JONES, 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.[Abstract/Free Full Text]

ELLIS, J. G., G. J. LAWRENCE, M. AYLIFFE, P. ANDERSON, and N. COLLINS et al., 1997  Advances in the molecular genetic analysis of the flax-flax rust interaction. Annu. Rev. Phytopathol. 35:271-291.[Medline]

ELVIN, P., G. SLYNN, D. BLACK, A. GRAHAM, and R. BUTLER et al., 1990  Isolation of cDNA clones using yeast artificial chromosome probes. Nucleic Acids Res. 18:3913-3917.[Abstract/Free Full Text]

FLOR, H. H., 1956  The complementary systems in flax and flax rust. Adv. Genet. 8:29-54.

GESSLER, C. and D. STUMM, 1984  Infection and stroma formation by Venturia inaequalis on apple leaves with different degrees of susceptibility to scab. Phytopathol. Z. 110:119-126.

HAMMOND-KOSACK, K. E. and J. D. G. JONES, 1997  Plant disease resistance genes. Annu. Rev. Plant Physiol. Plant Mol. Biol. 48:575-607.

HULBERT, S. H., C. A. WEBB, S. M. SMITH, and Q. SUN, 2001  Resistance gene complexes: evolution and utilization. Annu. Rev. Phytopathol. 39:285-312.[Medline]

JIA, Y., S. A. MCADAMS, G. T. BRYAN, H. P. HERSHEY, and B. VALENT, 2000  Direct interaction of resistance gene and avirulence gene products confers rice blast resistance. EMBO J. 19:4004-4014.[Medline]

JONES, D. A. and J. D. G. JONES, 1997  The role of leucine-rich repeat proteins in plant defenses. Adv. Bot. Res. Adv. Plant Pathol. 24:89-167.

KOBE, B. and J. DEISENHOFER, 1995  A structural basis of the interactions between leucine-rich repeats and protein ligands. Nature 374:183-186.[Medline]

KÖLLER, W., D. M. PARKER, and C. M. BECKER, 1991  Role of cutinase in the penetration of apple leaves by Venturia inaequalis. Phytopathology 81:1375-1379.

KORBAN, S. S., 1998  What's new with disease-resistant apple cultivars. Proc. Trans. Ill. Hort. Soc. 131:74-76.

LEISTER, D., A. BALLVORA, F. SALAMINI, and C. GEBHARDT, 1996  A PCR-based approach for isolating pathogen resistance genes from potato with potential for wide application in plants. Nat. Genet. 14:421-429.[Medline]

LOVETT, M., J. KERE, and L. M. HINTON, 1991  Direct selection: method for the isolation of cDNAs encoded by large genomic regions. Proc. Natl. Acad. Sci. USA 80:9628-9632.

MARTIN, G. B., S. H. BROMMONSCHENKEL, J. CHUNWONGSE, A. FRARY, and M. W. GANAL et al., 1993  Map-based cloning of a protein kinase gene conferring disease resistance in tomato. Science 262:1432-1436.[Abstract/Free Full Text]

MEYERS, B. C., K. A. SHEN, P. ROHANI, B. S. GAUT, and R. W. MICHELMORE, 1998  Receptor-like genes in the major resistance locus of lettuce are subject to divergent selection. Plant Cell 10:1833-1846.[Abstract/Free Full Text]

MICHELMORE, R. W. and B. C. MEYERS, 1998  Clusters of resistance genes in plants evolve by divergent selection and a birth-and-death process. Genome Res. 8:1113-1130.[Abstract/Free Full Text]

PARIMOO, S., S. R. PATANJALI, H. SHUKLA, D. D. CHAPLIN, and S. M. WEISSMAN, 1991  cDNA selection: efficient PCR approach for the selection of cDNAs encoded in large chromosomal DNA fragments. Proc. Natl. Acad. Sci. USA 80:9623-9627.

PARNISKE, M., K. E. HAMMOND-KOSACK, C. GOLSTEIN, C. M. THOMAS, and D. A. JONES et al., 1997  Novel disease resistance specificities result from sequence exchange between tandemly repeated genes at the Cf-4/9 locus of tomato. Cell 91:821-832.[Medline]

RICHTER, T. E. and P. C. RONALD, 2000  The evolution of disease resistance genes. Plant Mol. Biol. 42:195-204.[Medline]

SALMERON, J. M., G. E. D. OLDROYD, C. M. T. ROMMENS, S. R. SCOFIELD, and H. S. KIM et al., 1996  Tomato Prf is a member of the leucine-rich repeat class of plant disease resistance genes and lies embedded within the Pto kinase gene cluster. Cell 86:123-133.[Medline]

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual, Ed. 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SCOFIELD, S. R., C. M. TOBIAS, J. P. RATHJEN, J. H. CHANG, and D. T. LAVELLE et al., 1996  Molecular basis of gene-for-gene specificity in bacterial speck disease of tomato. Science 274:2063-2065.[Abstract/Free Full Text]

SONG, W. Y., G. L. WANG, L. L. CHEN, H. S. KIM, and L. Y. PI et al., 1995  A receptor kinase-like protein encoded by the rice disease resistance gene, Xa21. Science 270:1804-1806.[Abstract/Free Full Text]

SOOD, R., T. I. BONNER, I. MAKALOWSKA, D. A. STEPHAN, and C. M. ROBBINS et al., 2001  Cloning and characterization of 13 novel transcripts and the human RGS8 gene from the 1q25 region encompassing the heredity prostate cancer (HPC1) locus. Genomics 73:211-222.[Medline]

VALSANGIACOMO, C. and C. GESSLER, 1988  Role of the cuticular membrane in ontogenic and Vf-resistance of apple leaves against Venturia inaequalis. Phytopathology 78:1066-1069.

VINATZER, B. A., A. PATOCCHI, L. GIANFRANCESCHI, S. TARTARINI, and H. B. ZHANG et al., 2001  Apple contains receptor-like genes homologous to the Cladosporium fulvum resistance gene family of tomato with a cluster of genes cosegregating with Vf apple scab resistance. Mol. Plant-Microbe Interact. 14:508-515.[Medline]

WISE, R. P., 2000  Disease resistance: what's brewing in barley genomics. Plant Dis. 84:1160-1170.

XU, M. L. and S. S. KORBAN, 2000  Saturation mapping of the apple scab resistance gene Vf using AFLP markers. Theor. Appl. Genet. 101:844-851.

XU, M. L. and S. S. KORBAN, 2002  AFLP-derived SCARs facilitate construction of a 1.1 Mb sequence-ready map of a region that spans the Vf locus in the apple genome. Plant Mol. Biol. 50:803-818.[Medline]

YOUNG, N. D., 2000  The genetic architecture of resistance. Curr. Opin. Plant Biol. 3:285-290.[Medline]

ZHANG, M. Q., 1997  Identification of protein coding regions in the human genome by quadratic discrimination analysis. Proc. Natl. Acad. Sci. USA 94:565-568.[Abstract/Free Full Text]

This article has been cited by other articles:

				V. Shulaev, S. S. Korban, B. Sosinski, A. G. Abbott, H. S. Aldwinckle, K. M. Folta, A. Iezzoni, D. Main, P. Arus, A. M. Dandekar, et al. Multiple Models for Rosaceae Genomics Plant Physiology, July 1, 2008; 147(3): 985 - 1003. [Abstract] [Full Text] [PDF]

				E. Belfanti, E. Silfverberg-Dilworth, S. Tartarini, A. Patocchi, M. Barbieri, J. Zhu, B. A. Vinatzer, L. Gianfranceschi, C. Gessler, and S. Sansavini The HcrVf2 gene from a wild apple confers scab resistance to a transgenic cultivated variety PNAS, January 20, 2004; 101(3): 886 - 890. [Abstract] [Full Text] [PDF]