• The wild apple (Malus sieversii) is a large-fruited species from Central Asia, which is used as a source of scab resistance in cultivar breeding.
• Phytopathological tests with races of Venturia inaequalis were performed to differentiate scab-resistance genes in Malus as well as an avirulence gene in the pathogen.
• A novel gene-for-gene interaction between V. inaequalis and Malus was identified. The locus of the scab-resistance gene Vh8 is linked with, or possibly allelic to, that of the Vh2 gene in Malus pumila Russian apple R12740-7A, at the lower end of linkage group 2 of Malus. Race 8 isolate NZ188B.2 is compatible with Vh8, suggesting the loss or modification of the complementary AvrVh8 gene, while isolate 1639 overcomes both Vh2 and Vh8, but is incompatible with at least one other gene not detected by any of the other race isolates tested.
• Our research is the first to differentiate scab-resistance genes in a putative gene cluster in apple with the aid of races of V. inaequalis.
The wild apple Malus sieversii (Lebed.) Roem. is recognized as a major progenitor of the domesticated apple (Malus ×domestica Borkh.) (Roach, 1988; Harris et al., 2002). Seed and wood collections have been made in the wild apple forests of Kazakhstan (Hokanson et al., 1997) to broaden the genetic base available for apple cultivar improvement (Noiton & Shelbourne, 1992; Noiton & Alspach, 1996). These collections included accessions that were free of diseases prevalent in the natural habitat of the wild apple, such as apple scab [Venturia inaequalis (Cke) Wint.] and powdery mildew (Podosphaera leucotricha) (Luby et al., 2001). The use of plant resistance is widely regarded as the most successful and cost-effective alternative means to the use of fungicides to control diseases. Evaluation in New Zealand of M. sieversii germplasm collected in 1993 indicated that the levels of field resistance to scab (Bus et al., 2002) and fire blight (Erwinia amylovora) (Luby et al., 2002) were high. Glasshouse evaluations of half-sib accessions, collected in 1995 and 1996 and inoculated with V. inaequalis from different geographical origins, suggested that some still unnamed scab-resistance genes exist in this germplasm (Luby et al., 2001). While many scab-resistance genes have been identified, mostly in crabapples (Korban & Chen, 1992), the genetics of scab resistance in M. sieversii and its significance in the evolutionary interaction with V. inaequalis remains unexplored.
The V. inaequalis–Malus interaction is one of the first examples for which gene-for-gene interactions were shown (Williams & Shay, 1957; Boone, 1971). Gene-for-gene interactions are common to many specialist (hemi-)biotrophic parasites, and were first proposed by Flor (1956, 1971) for the flax (Linum usitatissimum)–flax rust (Melampsora lini) relationship. In light of a better understanding of the molecular nature of genes, the gene-for-gene relationship was reinterpreted in the 1990s to involve the specific recognition governed by an avirulence (Avr) gene in the pathogen and the corresponding resistance (R) gene in the host (Dangl & Jones, 2001). The simplest explanation for the recognition event is the receptor–ligand model, in which the defence reactions are assumed to be the result of direct interaction between the Avr protein from the pathogen and the matching R receptor protein in the host (Bonas & Lahaye, 2002). However, as extensive research has provided only a few examples of direct R/Avr interaction, variations on this model, such as the coreceptor and, particularly, the guard protein models, have gained more acceptance recently (Van der Biezen & Jones, 1998; Luderer & Joosten, 2001; Bonas & Lahaye, 2002).
The hypersensitive response (HR) is often associated with gene-for-gene interactions, and is characterized by a rapid reaction by the host, which results in localized cell death often showing as pinpoint necrosis (Goodman & Novacky, 1994). A classic example in apple is the Vm gene, the gene-for-gene interaction of which conditions HR that becomes macroscopically visible about 2 d after infection (Hernandez Castillo, 1990). The Vfh gene from Malus floribunda 821 (Bénaouf & Parisi, 2000) also conforms to this model, but the pinpoint necrosis takes about 6 d to become macroscopically visible (Shay & Hough, 1952). While a gene-for-gene relationship has also been shown for Vf (Bénaouf & Parisi, 2000), this gene conditions nonHR in the form of chlorotic resistance reactions with limited sporulation, which take about 12 d to develop under optimal conditions for disease (Chevalier et al., 1991). Other scab-resistance genes, such as Vh2 (Bus et al., 2005), condition intermediate reactions in the form of distinctive stellate necrosis (SN) (Shay & Hough, 1952) after about 4–6 d. An apparent delay in the onset of the resistance reaction allows for some stroma development by V. inaequalis, where strands of subcuticular mycelium radiate out from the initial point of infection, resulting in growth in the form of a star.
In this paper we describe the discovery of a new resistance gene in M. sieversii, which we have named Vh8, as well as the gene-for-gene relationship with its complementary avirulence gene discovered in V. inaequalis. We show that the Vh8 locus differs from, but is linked with, the locus containing the Vh2 gene from scab differential host 2, TSR34T15 (an F2 descendant of the progenitor Malus pumila Russian apple R12740-7A), which has been overcome by race 2 of V. inaequalis (Shay & Williams, 1956). We also provide evidence for another putative gene-for-gene interaction between V. inaequalis and a second accession of M. sieversii, which was identified in the process of differentiating Vh2 and Vh8.
Materials and Methods
Plant material and V. inaequalis isolates
The genetic and host–pathogen interaction studies were performed on M. sieversii accession W193B, and a Royal Gala × W193B progeny of 152 seedlings. Accession W193B was selected from a GMAL3631 family grown from open-pollinated seed collected from a tree in the Tarbagatai mountain range in Kazakhstan (Luby et al., 2001).
The initial inoculation on the Royal Gala × W193B family was performed with a multi-isolate conidial suspension of V. inaequalis (3.5 × 105 conidia ml−1). The original mixture comprised 10 single-spore isolates (at that stage all defined as race 1) in a blend made in 1994. In the intervening years, the isolate mixture was maintained on air-dried leaves collected annually from susceptible seedlings in the breeding programme and stored at −20°C for use in the following year. At the time of phenotyping, several single-spore cultures were prepared from compatible scab lesions on Royal Gala × GMAL4190-W188B and Royal Gala × W193B seedlings showing resistant and susceptible (R + S) symptoms. Two of these compatible isolates, NZ188B.2 and NZ193B.2, were selected for subsequent experiments, together with six single-spore isolates of race 1 and one of race 2 (Table 1). Initial inocula were prepared from conidia harvested from cultures grown on cellophane overlaid onto potato dextrose agar (PDA) (Parker et al., 1995) and subsequently maintained on infected dried leaves.
Table 1. Origin of monoconidial isolates of Venturia inaequalis used in this study
Nelson, New Zealand
Nelson, New Zealand
Leaf of unknown cultivar
Nelson, New Zealand
Granny Smith leaf
Auckland, New Zealand
Schone van Boskoop
Elst, the Netherlands
Royal Gala × M.sieversii GMAL4190-W188B seedling leaf
Havelock North, New Zealand
Royal Gala × M.sieversii GMAL3631-W193B seedling leaf
Havelock North, New Zealand
General plant maintenance, inoculation and phenotyping
Seeds were planted in root-trainer baskets with seed-raising mix containing a general slow-release fertilizer. The seed-raising mix was saturated with water, and the baskets were held in a bin lined with plastic for a stratification period of 6–8 wk at 0–1°C in a cool store, followed by germination in the glasshouse. Clonal trees of accessions were grown on M.793 rootstocks in 3 l plastic bags with potting mix. Both seedlings and clonal trees were incubated under optimal conditions for infection as described by Gardiner et al. (1996), and phenotyping was performed in the third week after inoculation. Symptoms were scored as follows: R, resistant; S, susceptible; 0, no visible symptoms. Where useful, distinction in the resistance reactions was made between stellate necrotic (SN) and chlorotic (Chl) reactions. Leaf samples of symptoms for microscopic observations were cleared and stained according to Bruzzese & Hasan (1983), and mounted in an arabic gum solution (Cunningham, 1972). Both brightfield and autofluorescence observations were made on a Nikon Optiphot microscope equipped with epifluorescence. The autofluorescence of the resistance reactions was examined in the interference blue range (excitation filter 450–490 nm, dichroic mirror 505 nm, barrier filter 515 nm). Digital images were taken with a CoolSnap camera (Coherent Scientific, Adelaide, Australia) and digitally adjusted for contrast in photoshop ver. 6.2 (Adobe, San José, CA, USA).
Host–pathogen interaction studies
Five experiments were performed to study the resistances of M. sieversii accessions, segregating seedling progenies, and scab differential hosts with various V. inaequalis isolates.
Experiment 1 The Royal Gala × W193B seedlings used in the routine scab screening were repotted into 0.5 l plastic bags and cut back for the shoots to regrow in order to provide leaves suitable for inoculation. Many of the susceptible plants had died by that stage as a result of the first scab inoculation, hence the resistant and surviving susceptible seedlings were separately divided into three groups of about equal numbers. Each group of seedlings was spray-inoculated with a suspension of either isolate NZ110.1 (race 1) or NZ188B.2 (race 8) at an effective concentration (spore concentration × germination rate) of ≈1 × 105 conidia ml−1 or water. A reference population of newly raised Royal Gala × TSR34T15 (scab differential host 2) seedlings was split into five groups and inoculated with the same three inocula, as well as with additional conidial suspensions of isolates J243 (race 1) and NZ193B.2 (race 8) at the same effective concentration (1 × 105 conidia ml−1).
Experiment 2 Isolate NZ188B.2 was inoculated on a defined set of scab differential hosts (Lespinasse, 1994; Bénaouf & Parisi, 2000) in order to determine the avirulence gene combination of this isolate. Trees of Royal Gala (host 1), TSR34T15 (host 2), Geneva (host 3), TSR33T239 (host 4), 9-AR2T196 (host 5), Prima (host 6), M. floribunda 821 (host 7a), and Golden Delicious (host 7b) were inoculated using a droplet inoculation technique. The method consisted of placing 100 µl inoculum (concentration ≈ 4 × 104 conidia ml−1) into small inoculation chambers (8 mm diameter), which were clipped onto the youngest fully expanded leaf of actively growing shoots. These chambers were made from microtube lids with a 7 mm hole drilled in the top, which were glued onto hair clips. The chambers were removed 48 h after inoculation. The inoculations were replicated three times.
Experiment 3 Three families segregating for the Vh2 gene and four families derived from different accessions selected from the Kazakh M. sieversii germplasm were raised. The accessions were: GMAL3634-X391 from the Tarbagatai region; and GMAL3683-X59, GMAL4302-X8, and GMAL4038-X257 from the Djungarsky region. In a preliminary study, all four accessions were shown to contain the SCAR markers OPL19-433bp, OPB18-628bp and OPB18-799bp, except for GMAL3634-X391, which did not contain OPB18-799bp. The Vh2 families were derived from TSR34T15 (Bus et al., 2005). The susceptible parents were A20R01T289, A20R02T032 and A22R11T137 selected from crosses between Braeburn and Royal Gala, and Pacific Rose‘ from a Gala × Splendour cross. Half the seedlings from each family were spray-inoculated with isolate J222 (race 1) and the other half with isolate NZ188B.2 (race 8) at effective concentrations of ≈ 2 × 105 conidia ml−1.
Experiment 4 A Royal Gala × TSR34T15 (host 2) and a Pacific Beauty™ × GMAL4302-X8 family were raised at Plant Research International in Wageningen, the Netherlands. Pacific Beauty™ is a Gala × Splendour selection. The 2-wk-old seedlings were transplanted into 8 × 8 × 10 cm pots and grown in an ebb-and-flow fertigation system until the end of the experiment. Isolates EU-NL19 (race 1), EU-B05 (race 1), 1639 (race 2), and NZ188B.2 were used (Table 1). Inoculum was prepared from previously infected, dried leaves and was applied at a concentration of 2 × 104 conidia ml−1 in chambers as described in experiment 2. The plants were incubated in dark plastic tents at ambient temperature and 100% humidity. After 48 h the plants were taken out of the tents and the inoculation chambers removed. Symptoms were assessed after 2 and 4 wk and classed as susceptible (S), chlorotic (Chl), or necrotic (N). SN symptoms were not common with the droplet inoculation technique, but when observed, they were classed as such.
Experiment 5 Clonal trees of seven M. sieversii accessions collected as seed in Kazakhstan were evaluated for their susceptibility to race 8: GMAL3627 (W189A) and GMAL3631 (W193B, Y415) from the Tarbagatai region; and GMAL3650 (W180A, Y409), GMAL4027 (W203B), and GMAL4190 (W188B) from the Djungarsky region. Trees of Royal Gala (susceptible) and TSR34T15 (host 2) were used as controls. Leaves were inoculated (effective conidia concentrations are given in parentheses) with one of the race 8 isolates NZ188B.2 (6 × 104) or NZ193B.2 (10 × 104); or one of the race 1 isolates NZ110.1 (15 × 104) or J243 (21 × 104). The trees were inoculated by placing several 20 µl droplets of inoculum with a loop on the youngest unfolded leaf of actively growing shoots. The number of isolates inoculated depended on the availability of actively growing shoots of each accession, but each host–isolate combination was replicated at least three times.
Genetics of avirulence in race 8
Isolates J222 (race 1) and NZ188B.2 (race 8) were mated in vitro using a protocol adapted from Ross & Hamlin (1965). Mature leaves of Royal Gala were washed under running tap water to remove dust and debris, and cut into 2 × 2 cm squares avoiding the midrib of the leaf. The leaf pieces were autoclaved and placed in Petri dishes with 2% water agar. A plug of mycelium of each isolate was scraped from the edge of 3-wk-old colonies on PDA and mixed onto the leaf pieces with a scalpel. The plates were sealed with a double layer of parafilm and placed in a sealed plastic bag. The cultures were incubated at room temperature under ambient night/day conditions for 2 wk, after which they were maintained inverted in the dark at 8°C. The cultures were checked for the production of pseudothecia for the first time after 4 months. Once the pseudothecia produced asci with mature ascospores, spores were collected at random and mono-ascospore cultures were grown on PDA. The virulence of 62 progeny was assessed in four replicates on accession W193B with the inoculation chamber technique described above. Royal Gala trees were also inoculated to test the pathogenicity of the isolates.
DNA extraction, markers, PCR amplification and map construction
The genetic marker study was performed on the Royal Gala × W193B family. Leaves of the seedlings for DNA extraction were sampled before inoculation. DNA extraction was performed according to Gardiner et al. (1996).
The PCR primers and their specific annealing temperatures are given in Table 2. Both OPL19SCAR and OPB18SCAR are linked to the SN-conditioning Vh2 gene from Russian apple R12740-7A (Hemmat et al., 2002; Bus et al., 2005). A Royal Gala × TSR34T15 (host 2) family was used as reference for this gene in the marker studies. Microsatellite markers (SSR) of LG2 (Table 2) from a reference genetic linkage map of apple (Liebhard et al., 2002) were used to align our map of W193B around Vh8 to this map.
Table 2. Primers and specific annealing temperatures used in this study
Td = touchdown annealing temperature range; Ta = set annealing temperature.
The polymerase chain reaction (PCR) mixture for both SCAR primers was a 15 µl reaction mix containing 1.5 ng apple genomic DNA, 20 mm Tris–HCl pH 8.4 and 50 mm KCl buffer, 1.3 mm MgCl2, 1% formamide in distilled H2O, 0.1 mm of each dNTP, 0.1 µm of each primer, and 0.44 units of Platinum Taq DNA polymerase (Invitrogen, Carlsbad, CA, USA), overlaid with 15 µl paraffin oil. DNA was amplified in a Hybaid PCR express thermal cycler with the following programme for OPL19SCAR: 1 cycle of 2 min 45 s at 94°C; 40 cycles of 55 s at 94°C, 55 s at 55°C, 1 min 39 s at 72°C; and a final extension of 10 min at 72°C. For OPB18SCAR, the programme was as follows: 1 cycle of 2 min 45 s at 94°C; 20 cycles of 55 s at 94°C, 55 s at 65–55°C (touchdown annealing temperature dropping 0.5°C per cycle), 1 min 39 s at 72°C; 20 cycles of 55 s at 94°C, 55 s at 55°C, 1 min 39 s at 72°C; and a final extension of 10 min at 72°C. The PCR products were separated on 0.45% agarose (USB Corp., Cleveland, OH, USA) + 0.45% Low Range Ultra Agarose Certified™ (Bio-Rad, Richmond, CA, USA) gels.
The SSR markers were analysed with all the markers in the same multiplex reaction. Their PCR amplification solution was a 20 µl reaction mix containing 10 ng DNA; 20 mm Tris–HCl pH 8.4 and 50 mm KCl buffer; 1.5 mm MgCl2; 0.2 mm of dNTP; the same amount of forward and reverse primer specific to the marker: 0.07 µm for CH02b10, 0.13 µm for CH02c02a, 0.06 µm for CH03d01, 0.08 µm for CH05e03; 8.54 µl sterile distilled H2O; and 0.3 units of Platinum Taq DNA polymerase (Promega Corp., Madison, WI). DNA was amplified in a Techne Genius thermal cycler (Techne Cambridge Ltd, Cambridge, UK) using the following PCR conditions: 1 cycle of 2 min 30 s at 94°C; 33 cycles of 30 s at 94°C, 30 s at 55°C and 1 min at 72°C; and a final extension of 5 min at 72°C. The PCR products were loaded onto an ABI PRISM‘ 377XL sequencer (Applied Biosystems, Foster City, CA, USA) with an internal 50–500 bp size standard. Data were collected with ABI PRISM data collection software ver. 2.6, and analysed with ABI PRISM genescan ver. 3.1.2 and ABI PRISM genotyper ver. 2.1 software. The sizes of all the microsatellite marker alleles were standardized to those presented by Liebhard et al. (2002).
The genetic maps around the resistance genes were constructed using the programme joinmap (van Ooijen & Voorrips, 2001). Initial genetic linkage analysis was performed with a critical LOD score of 5.0 to associate loci into linkage groups. The second stage of linkage analysis established marker order within these linkage groups using the default settings of joinmap 3.0 for LOD and REC thresholds (1.0 and 0.4, respectively). The final locus order was determined by extensive proofreading and by minimizing the number of double crossovers flanking single loci. Recombination frequencies were converted to map distances using Kosambi's mapping function (Kosambi, 1944) and χ2 tests were used to identify any segregation distortion of parental alleles.
Genetic marker sequencing
PCR products from the following SCAR marker/accession combinations were sequenced: the OPL19SCAR from M. sieversii W193B and TSR34T15 (host 2); and the OPB18SCAR from W193B, Royal Gala, Russian apple R12740-7A, and one resistant seedling (AU311) from the Royal Gala × W193B family. The PCR products were purified using Qiagen Qiaquick PCR purification kits (Qiagen, Hilden, Germany). The fragments were then ligated into pGEM-T easy vector using the TA cloning protocol, and transformed into Escherichia coli TG1 using the Promega pGEM-Teasy protocol (GEM-T Easy Vector Systems, Promega Corp. 1993–97). Selection for transformants was performed on LB/ampicillin/IPTG/X-Gal agar plates, where the white colonies contained the inserted fragment in the vector. Plasmids were isolated using the Qiaprep Spin miniprep kit (Qiagen) and three replicates of each fragment were sequenced (DNA Sequencing Facility, School of Medicine, University of Auckland, New Zealand).
Phenotype and genetics of a new scab-resistance gene
In a routine evaluation of back-cross progenies developed for the introduction of new sources of scab resistance, it was observed that about half of the seedlings from a Royal Gala × M. sieversii W193B progeny gave a combination of resistant (R) and susceptible (S) symptoms (R + S) on the same leaf following inoculation with a multi-isolate inoculum of V. inaequalis (Fig. 1a). These isolates were selected at random from 150 isolates that had been collected from around New Zealand. All had previously been identified as race 1 (Lespinasse, 1994), that is, they were capable of infecting the universally susceptible cultivar Royal Gala but were avirulent on scab differential hosts h2 to h6 (V.G.M.B., unpublished data). Therefore this unusual combination of (R + S) symptoms suggested the presence of a new scab-resistance gene in accession W193B, as well as the presence of one or more isolates virulent to this gene in the multi-isolate inoculum. The resistance gene conditioned typical SN resistance reactions in incompatible interactions on the W193B host (Fig. 1b) and the seedlings, which were very similar to the symptoms on Vh2 hosts (Fig. 1d,e) (Aldwinckle et al., 1976; Hemmat et al., 2002). However, the combination of (R + S) symptoms had not been observed in earlier back-cross progenies of accessions carrying Vh2 inoculated with the V. inaequalis mixture (Bus et al., 2000, 2005).
The new resistance segregated as a single major gene, with 77 of the Royal Gala × W193B seedling progeny classified as (R + S) and 75 seedlings classified as S following the inoculation with the isolate mix. The segregation of the progeny did not differ significantly [P(χ2 > 0.03) = 0.88] from 1 : 1. Two single-spore V. inaequalis strains, NZ188B.2 and NZ193B.2, which were isolated from (R + S) seedlings, both caused lesions on W193B (Fig. 1c), which confirmed their ability to overcome the new gene. Indeed, one subset of the Royal Gala × W193B progeny originally classed as (R + S) after inoculation with the isolate mix showed S symptoms only when reinoculated with isolate NZ188B.2 (Table 3), while another subset showed R symptoms when reinoculated with race 1 isolate NZ110.1 (Table 3).
Table 3. Phenotypes on subsets of Royal Gala × Malus sieversii W193B seedlings after reinoculation with two single-spore isolates of Venturia inaequalis following a routine scab resistance screening with a multi-isolate inoculum
(R + S) = resistant and susceptible symptoms; R = resistant symptoms only; S = susceptible symptoms only; 0 = no symptoms.
Numbers of seedlings evaluated in parentheses.
Differential host–pathogen interactions of race 8
Isolate NZ188B.2 was not virulent on the differential hosts TSR34T15 (host 2), Geneva (host 3), TSR33T239 (host 4), 9-AR2T196 (host 5), Prima (host 6) and M. floribunda 821 (host 7a). It was compatible with Royal Gala (the universally susceptible host 1) and Golden Delicious (host 7b), which carries the Vg gene. The fact that race 1 isolates can infect Golden Delicious, but not W193B, supports our conclusion that isolate NZ188B.2 is virulent on W193B and overcomes the specific resistance gene in that host. Hence the new gene is different from any of the identified resistance genes of the differential hosts. We have named the gene of this new host–pathogen interaction Vh8, and designated the host that carries it (W193B) as host 8, and the associated strain of V. inaequalis race 8. A culture of isolate NZ188B.2 has been deposited at the International Collection of Micro-organisms from Plants at Landcare Research, Auckland, New Zealand (#14899).
As TSR34T15 (host 2), which carries the Vh2 gene (Bus et al., 2005), conditions the same type of SN symptoms (Williams & Shay, 1957) as Vh8, the relationship between the two scab-resistance genes was investigated further. The incompatible interaction between the Vh2 host and race 8 was confirmed in a Royal Gala × TSR34T15 progeny with isolates NZ188B.2 and NZ193B.2 (Table 4), and in a further three TSR34T15 progenies with isolate NZ188B.2 alone (Table 5). In all cases the race 8 isolates, as well as the race 1 isolates NZ110.1, J243, and J222, produced R : S = 1 : 1 segregations. However, the inoculation of subsets of seedlings with different isolates is still not proof of a single gene segregating, as it is possible for the current scab differentials to carry other, still unidentified genes conditioning the same phenotype, for which NZ188B.2 has the corresponding avirulence determinants. To exclude this possibility for the Vh2 differential host TSR34T15, a second batch of 105 seedlings of the Royal Gala × TSR34T15 progeny was tested, where each seedling was inoculated with both isolates NZ188B.2 and EU-NL19 (race 1). The isolates gave R : S = 1 : 1 segregations [P(χ2 > 0.01) = 0.88] with the same seedlings showing resistant or susceptible symptoms for each isolate. This confirmed that the observed resistance of TSR34T15 to NZ188B.2 was indeed caused by Vh2, and that this gene is different from Vh8.
Table 4. Segregations for subsets of a Royal Gala × TSR34T15 seedling family inoculated with four single-spore isolates of Venturia inaequalis
The Vh8 gene, like Vh2, confers distinctive SN reactions. The necrosis of the palisade mesophyll is macroscopically visible on the adaxial side of the leaves in the shape of a star that follows the pattern of stroma formation by V. inaequalis. The extent of the host response was up to several millimetres in diameter and radiated out from the penetration point of the infecting conidium. Microscopic observations of resistance reactions on hosts 2 and 8 revealed that the necrosis of the mesophyll was overlaid with a halo of necrotic epidermal cells (Fig. 1d,e). The necrosis of the palisade mesophyll of advanced reactions was about 20–30 cells wide in the areas under the extended mycelium growth of the pathogen, while 10–15 cells of the much larger epidermal cells determined the width of the halo. In the smaller resistance reactions very similar to HR, the mesophyll affected was limited to a core 30–50 cells wide. The pattern of the SN symptoms on the abaxial side of the leaves was very similar – intense autofluorescence on excitation by interference blue light in the spongy mesophyll of the leaves surrounded by a halo of necrotic epidermal cells. Some plants carrying either the Vh8 or Vh2 resistance gene showed limited sporulation in the necrotic area only, while on others subcuticular mycelium with very limited sporulation had advanced well beyond the necrotic areas. The necrotic areas showed strong autofluorescence, which was particularly intense in the palisade mesophyll (Fig. 1f,g). The mycelium of the fungus itself also showed some autofluorescence within the necrotic areas. There was no autofluorescence in leaves inoculated with water, or in susceptible leaves infected with V. inaequalis, except for constitutive autofluorescence of the vascular tissues.
Segregation of the avirulence gene
The progeny from the cross between V. inaequalis isolates J222 and NZ188B.2 produced about 10–15 pseudothecia per leaf piece, each with numerous asci containing eight ascospores (Fig. 1h). Of the subsample of 62 fungal progeny tested, 34 isolates caused resistance symptoms, and 28 isolates caused susceptible symptoms on host W193B (Fig. 1i). This segregation did not differ significantly [P(χ2 > 0.58) = 0.45] from an AvrVh8 : avrVh8 = 1 : 1 ratio. All isolates were tested on Royal Gala to confirm that they were pathogenic.
Genetic markers and sequences
The OPL19SCAR marker product originally identified for the Vh2 gene in a Royal Gala × TSR34T15 family (Bus et al., 2000, 2005) was also linked to the Vh8 gene in the Royal Gala × W193B family (Fig. 2). The PCR products from both hosts were cloned and sequenced (GenBank accession numbers: AY626823 for W193B; AY626824 for TSR34T15). The completely identical sequences of the 433 bp fragments confirmed that they are alleles of the same OPL19SCAR marker locus, and that both scab-resistance genes are alleles of the same locus, closely linked loci, or homologous loci.
The OPB18SCAR marker previously identified for the Vr gene (Hemmat et al., 2002) produced one PCR product of 628 bp, which was present in Royal Gala, W193B and Russian apple R12740-7A, and a second product of 799 bp, which was present only in W193B. The 628 bp product is in repulsion, and the 799 bp is in coupling phase with the Vh8 allele in the W193B parent of the Royal Gala × W193B family (Fig. 2). The 628 bp fragment did not segregate in the Royal Gala × TSR34T15 family (Fig. 2), indicating that the resistant parent is homozygous for this marker. Sequencing of the 628 bp products (GenBank accession numbers of the sequences are given in parentheses) from W193B (AY642927), Russian apple R12740-7A (AY642928), and Royal Gala (AY642929) showed that they were 100% identical. The sequences of the 799 bp fragments from W193B (AY642930) and its resistant progeny AU311 from a cross with Royal Gala are nearly identical to that of the 628 bp fragment, except for a 171 bp insert after base pair 304 (Fig. 3). Because of the presence of a further three SNPs, the homology of the part that the 799 bp sequences shared with the 628 bp allele of OPB18SCAR was 99.5%. This high degree of similarity again confirms that the SCAR markers in these two families are allelic.
The SCAR markers cosegregated with the Vh8 gene and the SSR marker CH03d01 (Fig. 4), which was previously mapped to the lower end of LG2 (Liebhard et al., 2002). Vh8 is therefore located on LG2, as is the Vh2 gene (Hemmat et al., 2002; Bus et al., 2005). Two other SSR markers, CH02b10 and CH05e03, which are known to be linked to Vh2 (Hemmat et al., 2002; Bus et al., 2005), were tested but did not segregate with Vh8 in the Royal Gala × W193B family, as the W193B parent was homozygous for both markers. Marker CH02c02a at the other end of LG2 did not cosegregate with Vh8, which confirms its distant location from the gene cluster at the lower end of LG2 (Liebhard et al., 2002).
Scab resistance in other M. sieversii accessions
Results of a preliminary genotyping study for scab resistance of M. sieversii germplasm collected in Central Asia demonstrated that the 799 bp product of the OPB18SCAR marker is prevalent in the M. sieversii germplasm. It is present in progeny sampled from 41 out of 45 open-pollinated families tested, while progeny from 42 out of the 45 families amplified the OPL19SCAR marker. The families that did not possess either of these two markers had either none or very few scab-resistant seedlings (data not presented), which suggests that the markers are in linkage disequilibrium with the Vh8 gene. To confirm the prevalence of Vh8 in this germplasm, the resistances in another 11 accessions, including GMAL4190-W188B, from whose progeny isolate NZ188B.2 originated, were evaluated with race 1 and race 8 isolates. Four accessions were evaluated in the form of seedling families. Resistance was observed on neither the accessions (Table 6) nor the seedling progenies (Table 7) when inoculated with race 8, but was observed when inoculated with race 1. This result indeed suggests that the resistance gene present in this germplasm is Vh8. The segregations of the four progenies did not differ significantly from R : S = 1 : 1 in the seedlings inoculated with race 1, except for one family (Table 7). However, there was a significant heterogeneity effect among these families – the segregations tended to be skewed towards either resistance or susceptibility, suggesting segregation distortions have occurred in the genomic region around Vh8. This agrees with earlier molecular marker mapping studies, which showed that the lower part of LG2 shows different levels of segregation distortion: none in Jonathan × Prima (W.E.W., unpublished data), some in Fiesta × Discovery (α = 0.05; Liebhard et al., 2003), and considerable distortion in Prima × Fiesta (α = 0.001; Maliepaard et al., 1998).
Table 6. Leaf reactions on clonal trees of seven Malus sieversii accessions and TSR34T15 (host 2) following droplet inoculation with two race 1 and two race 8 single-spore isolates of Venturia inaequalis
Race 1 isolates
Race 8 isolates
Weak expression of symptoms.
GMAL accessions 3627 and 3631 originated from the Tarbagatai; 3650, 4027 and 4190 from the Djungarsky region of Kazakhstan.
GMAL accessions 3683, 4038 and 4302 originated from the Tarbagatai; 3634 from the Djungarsky region of Kazakhstan.
χ2 for heterogeneity, 3 df.
Although only one gene conditioning SN was segregating in each of the progenies derived from the five different M. sieversii accessions when inoculated with New Zealand race 1 isolates, the possibility that they interacted with different genes could not be excluded, as the isolates were not tested on the same seedlings. To investigate this, a second seedling family of GMAL4302-X8 crossed with Pacific Beauty‘ was raised at Plant Research International in the Netherlands. The reason for performing this study here was because it involved a race of V. inaequalis that is not present in New Zealand. It was determined that GMAL4302-X8 carries Vh8 based on its SN phenotype, its susceptibility to race 8, and the presence of the OPL19SCAR and B18SCAR markers (Table 7), with OPL19SCAR segregating with resistance in the A22R11T137 × GMAL4302-X8 family (data not presented). Each seedling was inoculated with V. inaequalis race 1 isolate EU-B05; race 2 isolate 1639; and race 8 isolate NZ188B.2. Isolate 1639 had previously been shown to be compatible with Vh2 hosts, whereas isolate NZ188B.2 is incompatible with the same hosts (Bus et al., 2005). The findings for races 1 and 8 were as expected: race 1 identified the presence of a single major gene conditioning SN, while all seedlings were susceptible to race 8. However, isolate 1639 was compatible with the seedlings carrying the SN conditioning gene (Fig. 5a,b), but incompatible with seedlings carrying at least one, but possibly two previously undetected resistance genes conditioning chlorotic resistance reactions. Neither the race 1 nor the race 8 isolate carried avirulence alleles to these chlorotic reaction conditioning resistance gene(s) (Fig. 5c,d).
We have identified a new race-specific major scab-resistance gene in apple, as well as its complementary avirulence gene in a new race of the fungus. Inheritance studies showed the monogenic basis of both genes, confirming the generally assumed gene-for-gene relationship for the V. inaequalis–Malus pathosystem (Sidhu, 1987). Such R–Avr interactions have been suggested for a number of relationships based on inheritance studies of both avirulence genes in the pathogen (Williams & Shay, 1957; Bagga & Boone, 1968a) and resistance genes in crabapple hosts (Bagga & Boone, 1968b). Other gene-for-gene interactions have been proposed for Vm (Hernandez Castillo, 1990), and for Vf and Vg (Bénaouf & Parisi, 2000).
We have named the new resistance gene Vh8 and the corresponding avirulence gene AvrVh8. Accession M. sieversii W193B and V. inaequalis isolate NZ188B.2 are the primary representatives of the apple scab differential host 8 and race 8, respectively, of this new host–pathogen interaction. The naming of the new gene follows on from the seven differential hosts that have been defined to date (Lespinasse, 1994; Parisi & Lespinasse, 1996; Bénaouf & Parisi, 2000). Evaluation of isolate NZ188B.2 on this differential host set showed that the isolate lacks or has modified the AvrVh8 gene as well as the AvrVg gene corresponding to Vg from Golden Delicious (Bénaouf & Parisi, 2000). However, the latter avirulence gene is not common in V. inaequalis populations worldwide and hence has rendered Golden Delicious susceptible to scab.
It has not been established to date where race 8 originated, and how widely dispersed it is in New Zealand. However, the observations that M. sieversii did not appear to be present in New Zealand before 1993 (MacKay, 1993), when HortResearch introduced the first seed, and that the Vh8 gene is a novel scab-resistance gene, suggests that the race is not likely to be common in this country. Malus sieversii has indeed shown good resistance to scab in the field (Bus et al., 2002), which agrees with the generally accepted rule that primary and secondary gene centres of cultivated plants are good sources of genuine resistance (Leppik, 1970). To what extent the resistance in this germplasm is caused by major effect or by quantitatively inherited resistance genes is the subject of further investigations.
Microscopic observations of the distinctive SN resistance reactions confirmed that the macroscopically indistinguishable symptoms conditioned by both Vh8 and the Vh2 genes also share a high degree of similarity at the cellular level. At this level the symptoms also bear a high degree of resemblance to those conditioned by the Vm gene, where resistance reactions do not develop beyond a circular core of necrotic palisade mesophyll with a halo of necrotic epidermal cells (Win et al., 2003). The Vm gene conditions the HR resulting in pinpoint lesions within 2–3 d of infection. While the SN symptoms take up to 1 wk to develop (Shay & Hough, 1952), those conditioned by both Vh8 and Vh2 are as strong as those from the Vm gene, and show the same intensity in autofluorescence. In both cases the resistance reaction involves many more cells than those in direct contact with the appressorium (Vm) and/or mycelium (Vh2 and Vh8) of the fungus. Hence it is reasonable to assume that extensive cell-to-cell signalling of the resistance response takes place in an attempt to contain the infection. The similarities in the reactions at the cellular level of the HR and SN genes suggest that, while they differ in the development rates of their responses, some of the underlying resistance mechanisms may be the same. In the course of our experiments we confirmed earlier findings of V. inaequalis being able to survive the resistance reaction of the genes conditioning SN and to produce limited sporulation under favourable conditions (Aldwinckle et al., 1976).
We have shown that the Vh8 locus maps to the same region on LG2 as that of Vh2. Further research may show that the loci are in the same cluster, or even that the genes are allelic. The proximity of both genes to the OPL19SCAR marker at approximately the same distance, supported by the wider maps for Vh2 (Hemmat et al., 2002; Bus et al., 2005) and Vh8 (Fig. 4) suggests that they may map to the same (complex) locus. While genetic marker research increasingly suggests the presence of scab-resistance gene clusters in apple (Vinatzer et al., 2001; Xu & Korban, 2002; Gardiner et al., 2003; Hemmat et al., 2003), and possibly different resistance alleles at the same locus, this study is the first in which two linked major scab-resistance genes have been separated on the basis of differential host–pathogen interactions. In fact, the Vh8 gene would not have been recognized as a potential new gene in the absence of race 8. In our case, the (R + S) symptoms were caused by a V. inaequalis multi-isolate inoculum comprising at least one virulent isolate among a number of avirulent isolates. While the SN symptoms conditioned by the Vh8 gene are very similar to those conditioned by the Vh2 gene from Russian apple R12740-7A, their presence in combination with S symptoms suggested a new gene was involved, since this phenomenon had not been observed previously with the same inoculum in seedling families segregating for the Vh2 gene. The presence of only the SN instead of the (R + S) phenotype in the W193B family would have suggested the scab-resistance gene to be the same as the one from Russian apple, which would have been substantiated (incorrectly as we now know) by the OPL19SCAR, the primary marker used for the Vh2 gene in marker-assisted selection (Bus et al., 2000).
We have also identified another gene in GMAL4302-X8 that does not appear to be linked to Vh2 and Vh8. Isolate 1639, which was previously shown to overcome Vh2 (Bus et al., 2005), also overcomes Vh8, but not an as-yet unnamed gene conditioning chlorotic resistance reactions in the Pacific Beauty‘ × GMAL4302-X8 family. These chlorotic resistance reactions might not have been visible due to epistatic effects if a segregating family of GMAL4302-X8 had not been used. This example illustrates that a degree of caution is required when identifying resistance genes by a combination of markers and symptom development in a single accession. While genetic markers are a significant improvement over the use of symptom development alone, they may not always distinguish different alleles of the same locus or closely linked loci from each other, as they can be identified only in segregating populations. It also demonstrates that it is important to follow up marker screens of germplasm with segregation analysis in a progeny set to ascertain whether the presence of a marker known to be linked to other resistance genes associates with the presence of the resistance identified in the new germplasm.
The finding that the Vh8 gene may be predominantly responsible for the scab resistance identified in the M. sieversii germplasm sampled is particularly interesting in its own right. The two SCAR markers linked to this gene were both present in over 90% of the families sampled from seven different regions of Kazakhstan spanning over 1000 km (Luby et al., 2001). We tested the resistance of 11 of these families from the Tarbagatai and Djungarski regions directly by using the specific race 8 that can overcome Vh8 resistance, and demonstrated that, in each case, the resistance is overcome by this race and therefore may be traced to the same gene.
Linkage of the Vh8 locus to Vh2 adds another gene to a number of single scab-resistance loci (Bus et al., 2005; Gygax et al., 2004) as well as QTLs (Calenge et al., 2004) that have already been mapped to the same region on LG2 of apple. In addition, Vh4 (Bus et al., 2005) and Vr2 (Patocchi et al., 2004), which still may prove to be the same gene, are located at the top end of LG2. Such a large number of genes mapping to LG2 has important consequences for the breeding of scab-resistant varieties, and a structured approach is required to pyramid these genes. As the beneficial alleles of these loci often reside in different genetic backgrounds, they would be difficult to combine by a traditional breeding strategy. The positions of these loci relative to each other will make more directed breeding strategies feasible, as linked markers could be used to identify particular (rare) recombination events that couple beneficial alleles or uncouple a linkage between a beneficial and a deleterious allele. Detailed (molecular) genetic analysis of resistance gene clusters and complex loci may also answer the question of whether quantitative resistances are, in some cases, caused by residual effects of ‘defeated’ major gene resistances (Pedersen & Leath, 1988), as has been suggested for the Vf locus (Durel et al., 2003). In these cases the major gene is no longer effective as a single gene barrier against host infection by races able to overcome it, but offers partial protection that can, in combination with other quantitative loci, still confer a degree of pathogen resistance. The identification of resistance gene clusters may also have consequences for the interpretation of findings from older genetic studies on resistance genes, as new genes may be identified in these clusters, provided that specific virulent pathotypes exist in order to differentiate them (Hulbert et al., 1997). The different R : S ratios found between evaluations of subsamples of the same M. sieversii families (Luby et al., 2001) suggest that the pathogen populations in different parts of the world differ in their pathotypes. Every new resistance gene discovered doubles the number of possible avirulence/virulence combinations in V. inaequalis. However, although the number of isolates in this study was small, there is the suggestion that pathotypes may possess contrasting virulence patterns as a result of virulence association and dissociation (Vanderplank, 1984). While isolate 1639 can overcome the Vh8 gene in GMAL4302-X8 as well as Vh2 (Bus et al., 2005), and therefore can be identified as neither race 8 nor race 2, it is incompatible with genes that condition chlorosis in GMAL4302-X8 as well as A68R03T057 (Bus et al., 2005). Conversely, the race 1 isolates in our study are compatible with hosts carrying these minor effect genes, but incompatible with hosts carrying the major effect genes. Hence they are, strictly speaking, not race 1 which, by definition, is avirulent to all scab-resistance genes. However, as has been shown with host–pathogen interactions between minor genes in a number of ‘susceptible’ cultivars and virulence in V. inaequalis orchard populations (Sierotzki & Gessler, 1998; Koch et al., 2000; MacHardy et al., 2001), it is not unexpected that race 1 isolates are virulent to resistance genes that are only effective against a (small) proportion of the pathogen population. A similar situation to that of isolate 1639 in our study was encountered with race 7, which is able to overcome the major genes Vf and Vfh, but not the minor Vg gene (Bénaouf & Parisi, 2000). This may indicate that, while individual V. inaequalis strains can lose or modify several avirulence genes, there is a fitness penalty to losing (combinations of) avirulence genes associated with certain major effect genes for resistance (Bai et al., 2000; Leach et al., 2001). Research has been initiated to investigate this further in V. inaequalis.
We thank Luciana Parisi for supplying the European V. inaequalis isolates; Remmelt Groenwolt and Mandy Pedersen for assisting in the pathological tests; Sunita Khatkar and Michael Cook in assisting with the genetic marker studies; and Chandra Ranatunga, Jos Kanne, Adrie Kooijman, Ad Hermsen and Gerrit Stunnenberg for seed raising and maintenance of the plant material. This research was funded in part by the Foundation for Science, Research and Technology (C06X0212) and supported by a Fellowship of the Netherlands Ministry of Agriculture, Nature Management and Fisheries.