Abstract
Apple scab, caused by Venturia inaequalis (Cke.) Wint., is a destructive fungal disease of major apple cultivars worldwide, most of which are moderately to highly susceptible. Thus, development of scab resistant cultivars is one of the highest priorities of apple breeding programs. The principal source of resistance for breeding programs has been the scab resistance gene Rvi6 that originated from the Japanese crabapple Malus floribunda (Sieb.) sel. 821. Isolates of V. inaequalis able to overcome Rvi6 have been identified in Europe, but have not yet been reported on the American continents. We recently discovered scab infection on M. floribunda 821 trees in a research orchard at Geneva, New York, USA, where approximately 10% of the leaves bore profusely sporulating apple scab lesions, many of which had coalesced to cover entire leaves. Chlorosis and pinpoint pitting symptoms typical of failed infections by V. inaequalis on hosts bearing the Rvi6 and Rvi7 genes were also observed. We assessed genetic diversity and population genetic structure of six V. inaequalis isolates collected from M. floribunda 821, one isolate from ‘Nova Easygro’, one isolate from ‘Golden Delicious’ and two isolates from Europe (11 isolates in total) using 16,321 genome-wide SNPs. Population genetic structure and PCA separated the isolates into distinct European and USA groups. The forgoing suggests that the new Rvi6 virulent isolates emerged within USA populations, rather than being transported from Europe. The overcoming of resistance in M. floribunda 821 but not in descendant cultivars suggests that durable resistance to apple scab will require a more comprehensive understanding of Rvi6 mediated resistance in diverse genetic backgrounds.
Introduction
Apple scab, caused by the Ascomycete fungus Venturia inaequalis (Cke.) Wint., is the economically most important fungal disease of apples. In nearly all apple growing regions of the world, growers rely on fungicide applications each year to limit quality and yield loss due to scab. Scab-resistant cultivars require fewer fungicide applications, and their importance is high in organic production systems and home gardens (Brown and Maloney 2008). The commercial success of scab-resistant cultivars has been limited by less desirable fruit quality. The need to suppress scab without total reliance upon fungicides has therefore made development of scab resistant cultivars with high fruit quality one of the highest priorities of apple breeding programs.
The vast majority of the modern scab resistant cultivars originate from a Japanese crabapple clone, Malus floribunda (Sieb.) sel. 821, which was found to be completely free of scab in the beginning of the twentieth century. At the University of Illinois, C. S. Crandall built a collection of crabapples and, among other crossings, crossed two siblings from a M. floribunda 821 × ‘Rome Beauty’ progeny (Crandall, 1926). Two accessions from this cross progeny were later selected for their exceptional scab resistance yet large fruit size, and were used as resistance donors by the PRI (Purdue University, Rutgers University, and the University of Illinois) cooperative breeding program between three universities. The first PRI cultivar, ‘Prima’ was approved in 1967, and since then several breeding centers have developed resistant cultivars by crossing PRI materials (Crosby et al. 1990). Currently, there are more than one hundred modern scab resistant apple cultivars, with approximately 90% of them carrying the resistance derived from M. floribunda 821 (Brown and Maloney 2008, 2013).
To date, 18 V. inaequalis resistance genes (Rvi genes) including Rvi6 and Rvi7 from M. floribunda 821 and genes from other Malus accessions have been described. The gene-for-gene relationship has also been confirmed for the majority (Khajuria et al. 2018; Broggini et al. 2011; Bus et al. 2011). Based on the segregation, the resistance of M. floribunda 821 was also thought to rely on a single dominant gene, previously called Vf (syn. Rvi6), exhibiting resistance symptoms from hypersensitive reaction to chlorosis (occasionally necrosis or slight sporulation). However, segregation analysis of the resistance symptoms suggested the presence of an additional gene for scab resistance (Parisi et al. 1993; Williams and Kuc 1969). A large-scale pathogenicity study on a ‘Golden Delicious’ × M. floribunda 821 progeny demonstrated the presence of the Vfh (syn. Rvi7) gene responsible for the hypersensitive reaction in M. floribunda 821, that was distinct from the original Rvi6 (Bénaouf and Parisi 2000). The Vfh gene is hypothesized to be lost early on during the breeding process, or has been overcome by the pathogen (Bénaouf and Parisi 2000; Gessler, 1989; Parisi and Lespinasse 1996).
Although the resistance granted by a single major gene can be easily inherited, as preferred by breeders, the breakdown of major genes as a result of the fast evolution of a pathogen has been documented many times in several host-pathogen systems (Pink and Hand 2003). For instance, slight sporulating symptoms were observed on ‘Prima’ (Rvi6) in the greenhouse trails even before its release, whereas Rvi6 cultivars remained resistant under the field conditions. Moreover, the 1:1 segregation ratio of resistance expected from a single major gene is only observed if progeny expressing necrotic, slightly sporulating symptoms were still classified resistant (Crosby et al. 1990; MacHardy, 1996). In the European continent (Ahrensburg, Germany, Europe) sporulating lesions were detected on the seedlings of ‘Prima’ (Rvi6) for the first time in 1984, and later in 1988 on ‘Prima’, Coop 7, 9, and 10. The same inoculum was used for testing various Rvi6 cultivars/selections, all of which were susceptible to the new isolates, but the original source of the resistance, M. floribunda 821, remained intact (Parisi et al. 1993). The breakdown of the resistance of M. floribunda 821 (Rvi6, Rvi7), was reported later in East Malling, England (Roberts and Crute 1994). Interestingly, the isolates overcoming the resistance of M. floribunda 821 was avirulent on ‘Golden Delicious’ and on its resistant descendants (e.g., ‘Prima’ and ‘Florina’) harboring the ephemeral Rvi1 gene. However, it is still unclear whether the isolates were unable to overcome the resistance of these cultivars due to the Rvi1 gene, or other genetic factors prohibited the adaptation of the pathogen (Gessler and Pertot 2012; Roberts and Crute 1994). It was reported that the background resistance of M. floribunda 821 is low, and its severe infection happens when the two major genes Rvi6 and Rvi7 have been overcome. It further suggests that the Rvi6 resistant cultivars infected by less aggressive isolates might possess more complex resistance than that of M. floribunda 821 (Caffier et al. 2010; Parisi et al. 2004).
Further population genetic studies suggested that European Rvi6 virulent isolates did not emerge from a recent mutation in the AvrRvi6 gene, but have pre-existed for thousands of years in wild crabapple reservoirs, without relevant gene-flow to populations infecting domesticated apples. The virulent isolates on Rvi6 cultivars are genetically less variable, but distinct from those collected from susceptible cultivars suggesting a genetic bottleneck in their recent history (Guerin et al. 2004; Guerin et al. 2007; Guerin and Le Cam 2004; Lemaire et al. 2015; Michalecka et al. 2018). Although the Rvi6 virulent and non-virulent lineages might have been separated for thousands of years, the emergence of Rvi6 cultivars in Europe allowed gene flow between the two lineages, which might have increased the speed of the pathogen’s adaptation, especially in orchards where susceptible and resistant cultivars coexist (Michalecka et al. 2018). Human facilitated transport might represent another factor for the spread of Rvi6 virulent strains, and a possible cause of the inconsistency between the geographical origin and genetic polymorphism of isolates in several studies (Kaymak et al. 2016; Gladieux et al. 2010; Guerin et al. 2007; Michalecka et al. 2018).
New isolates that can overcome the resistance of M. floribunda 821 (Rvi6 and Rvi7 genes) have not yet been reported on the American continent. In a 33 year-long assay carried out in the Secrest Arboretum (Ohio), heavy sporulation was observed on a Japanese crabapple tree in 2003 and 2005, although its resistance stayed intact for decades. The authors however did not confirm whether the accession was identical to M. floribunda 821 (Beckerman et al. 2009). In this study, we report new V. inaequalis isolates collected from the infected M. floribunda 821 trees grown in an apple orchard, their origin and genetic diversity, and their pathogenicity on Rvi6 resistant ‘Macfree’.
Materials and methods
Field assessment of apple scab symptoms
In June 2019, apple scab infection was observed on approximately 20-year old trees of M. floribunda 821 and scab resistant apple cultivars ‘Prima’ and ‘Nova Easygro’ grown in the research orchard Darrow Farm (Geneva, USA). In addition, this orchard has ‘Gala’, ‘Macintosh’, ‘Golden Delicious’, ‘Florina’, and four PRI cultivars. The trees were present in four replications that were distributed randomly across the entire orchard space. The orchard was maintained without any fungicide spray since 2017 (Table 2).
Apple scab incidence was assessed using the 0-9 grade scale of Lateur and Populer (1994) as follows: 0 - no observation (missing plant), 1 - no visible lesions, 2 - one or very few lesions detectable on close scrutiny of the tree (0-1%), 3 - Immediately apparent lesions in general clustered in few parts of the tree (1-5%), 4 - intermediate, 5 - numerous lesions widespread over a large part of the tree (±25%), 6 - intermediate, 7 - severe infection with half of the leaves badly infected by multiple lesions (±50%), 8 - intermediate (±75%), or 9 - tree completely affected with (nearly) all the leaves badly infected by multiple lesions (>90%). Reaction type was determined according to the scale used by the PRI breeding program (Crosby et al. 1990) as follows: 0 - no symptom, 1 - pin point pits, 2 - irregular chlorotic or necrotic lesions, 3 - few restricted sporulation (M) mixture of necrotic and chlorotic and sparsely sporulating lesions, or 4 - abundant sporulation.
Sample collection
Scab infected leaves were collected from M. floribunda 821 and ‘Nova Easygro’ trees located in Darrow Farm (Geneva, New York) (Table 2). The fungus was isolated by sticking infected leaves to the inner top side of Petri dishes containing 1.2% water agar. The dishes were sealed and incubated overnight to let the spores fall on the agar and germinate. Germinating spores were observed under microscope and were carefully placed to new PDA plates under laminar flow to avoid the transfer of any contaminant. One scab isolate (‘VI-1797-9’) shared by H. Aldwinckle, Cornell University, Geneva, NY collected from host 4 (unspecified) in Ohio, USA (Schnabel et al. 1999) was also included. The well-characterized reference V. inaequalis isolates of European origin EU_B05 and EU_NL24 (Caffier et al. 2015) were obtained from INRA, France.
Artificial inoculation of the Vf resistant commercial, cultivar ‘Macfree’
One year-old grafted plants of the Rvi6 scab resistant ‘Macfree’ were inoculated with the V. inaequalis inoculum obtained from scab-infected M. floribunda 821 trees. Leaves were collected from each of the four M. floribunda 821 trees. The fresh sporulating lesions were cut, placed into a beaker containing distilled water and 0,5 μl/ml Tween 20, and shaken for 20 minutes. The suspension was filtered and the concentration was estimated by a hemocytometer, and set to at least 3×104 conidia/ml by a centrifuge.
Four ‘Macfree’ scions were grafted on ‘M9’ rootstocks and maintained in a greenhouse under controlled conditions. Before inoculation, grafted plants were pruned to acquire new growth and young leaves. The spore inoculum was used to spray the first 4-6 actively growing intact leaves from the top. The inoculated plants were held in a moist chamber under 18 °C and 12h photoperiod. To maintain 100% humidity, the sprayed leaves were covered first by wet paper towels, and then by resealable plastic bags. The incubated shoot part was tagged by marking the shoots below the bags to ease future evaluation. After 48 hours of incubation, the bags were removed, and the plants were further maintained under low light, 18 °C temperature, and 70% humidity with no irrigation for 2-3 weeks (Peil et al. 2018). Similar to the field assessment, reaction types that were used by the PRI program (Crosby, 1990) were determined. Light microscopy was used to evaluate the intensity of sporulation on the leaves. Spores from the infected leaves were suspended in a minimal amount of water and their morphology was observed under microscope to confirm the identity of V. inaequalis. Images were taken using Toupview software with a digital camera for microscope (OMAX, Gyeonggi-do, South Korea) for demonstration.
DNA extraction and genome sequencing of V. inaequalis isolates
DNA was extracted from 8 V. inaequalis isolates, 6 of which were obtained from M. floribunda 821, one from ‘Golden Delicious’, and one from ‘Nova Easygro’. In case of M. floribunda 821, the total 6 samples were collected from 3 different trees by sampling more than one region on two of the trees. In addition, two European isolates, EU-NL24 and EU-B05, and one U.S. isolate VI-1797-9 from Purdue were also used to extract DNA (Table 1). Cultures were grown on potato dextrose agar for four weeks. No more than 0.2g wet mycelia was cut and cleared from the agar by a sterile scalpel. Samples were ground in Eppendorf tubes with a disposable homogenizer pestle under liquid nitrogen. DNA was isolated with the Wizard Genomic DNA Purification Kit (Promega) as described by Singh and Khan (2019). The quality of DNA was checked using agarose gel (1%) electrophoresis and quantified with Nanodrop™ and UV-Vis Spectrophotometer (Thermo Fisher Scientific, USA).
A total of 50ng DNA for each sample was used to prepare genome sequencing libraries for Illumina Nextera skim sequencing at Institute of Biotechnology, Cornell University, Ithaca, NY. Agilent Bioanalyzer (Agilent; www.agilent.com) was used to assess the quality and quantity of genomic libraries. Individual V. inaequalis samples were barcoded and sequenced in a single lane of Illumina Mi-Seq platform to generate 2×250 base pair paired-end read data.
Sequence analysis and variant detection
Raw sequences were separated into sample-specific reads with the barcode information and used to assess the sequence quality with fastqc program (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). The program Trimmomatic (Bolger et al. 2014) was used to remove the sequencing adaptors and low-quality reads with quality score below 20. The genome assembly of V. inaequalis strain EU-B04 VeinUTG001 (Le Cam et al. 2019) was retrieved from National Center for Biotechnology Information (NCBI) genome database and used to align the high-quality reads from the previous step. The alignments were filtered to eliminate PCR duplicated sequences and a sorted binary alignment format (BAM) was obtained for each sample using SAMtools (Li et al. 2009).
Variants across the V. inaequalis genome were identified using Genome Analysis Toolkit (GATK version 3.8.0; McKenna et al. 2010) with sample-specific BAM files as input. The HaplotypeCaller plugin in GATK was used to obtain genotype variant call format (gVCF), and individual sample gVCF files were merged into a single variant file with GenotypeGVCF plugin. The raw variants were processed through “hard filtering” criteria as per GATK best practices. The resulting variants were used for base recalibration to remove false positives. The variant dataset was separated into single nucleotide polymorphisms (SNPs) and insertions/deletions (INDELs).
Genetic diversity and population structure analysis
Genetic diversity and population genetic analysis was conducted with SNP data only. The SNPs were filtered using minimum read depth criteria of four. Furthermore, we kept SNPs with no missing data even in a single sample. Remaining high-quality SNPs were used to calculate the genome-wide statistics of diversity, population structure and selection indices in V. inaequalis. The population level nucleotide variation was examined through number of variants, nucleotide diversity (π), and principal component analysis (PCA) in Tassel v5 (Bradbury et al. 2007). The structure in V. inaequalis population was further examined using BYM admixture model and haploid genome option in the TESS software (Caye et al. 2016) with default settings. In addition, spatial coordinates were provided for the 11 isolates to infer the accurate structure in V. inaequalis isolates. Convergence in the population structure results was checked by running twenty independent runs for individual K values ranging from two to eight. DIC and Kmax were plotted to identify number of clusters/groups in the samples. Figures showing assignment of isolates to different clusters were generated using Clumpak - Cluster Markov Packager Across K (Kopelman et al. 2015).
Results
Apple scab infection in the orchard and under controlled conditions
In the beginning of June, scab symptoms were observed on M. floribunda 821 in the orchard at Darrow Farm, Geneva, NY, USA on each of the four replicated trees. The initial sparse symptoms spread to approximately 10% (incidence grade 4) of the leaves later in July and August (Table 2). The heavy infection resulted in abundant sporulation (class 4), which often covered the whole leaf surface area (Figure 1). Chlorosis and pin point pit type resistance reactions typical to the Rvi6 and Rvi7 genes, respectively, which are hypothetically the result of the large amount of inoculum from avirulent strains, were also frequent on the leaves.
The scab-susceptible commercial cultivars, ‘Gala’, ‘Wijcik McIntosh’, ‘Marshall McIntosh’, and ‘Golden Delicious’ all showed severe scab symptoms (Symptom class 4). Scab symptoms were more severe on ‘Golden Delicious’ than ‘Gala’ (incidence 7 and 5, respectively). Many Rvi6 cultivars originating from M. floribunda 821, such as ‘Dayton’, ‘Florina’, ‘Jonafree’, ‘Priscilla’ and ‘Redfree’ were totally free of scab during the whole season. Weak symptoms were also detected on the descendants of M. floribunda 821, ‘Prima’, and ‘Nova Easygro’; however, on these cultivars the infection was barely detectable. No more than one or two leaves per tree were infected in these cultivars (grade 1). In case of ‘Nova Easygro’, both chlorosis and necrosis were detected around the slight sporulation as the sign of induced defense response (M), while the weak sporulating lesions on ‘Prima’ were all surrounded by unusually strong chlorosis (3).
Abundant sporulating scab symptoms were observed under microscope on the leaves of ‘Macfree’ after 15 days of inoculation in controlled environment (Figure 2). The sporulation covered a large part of the leaf surface area without the trace of restriction by chlorotic or necrotic response, hence we consider this reaction to represent the full compatibility infection class (4).
Genomic diversity in the Venturia inaequalis isolates
To understand the genetic diversity, population structure and admixture, we sequenced and analyzed the 11 V. inaequalis genomes in this study (Table 1). A total of 34.5 million raw sequencing reads were obtained, out of which 0.38% low quality reads were discarded for further analysis (Table 3). The filtered sequence data constitutes about 3.1 million average reads per sample, and provided 9.2X coverage of the V. inaequalis genome in NCBI. The read alignment rate against the V. inaequalis genome varied from 97.7% to 99.2%, with an average of 98.7% (Table 3).
After variant detection and GATK hard filtration, total of 199,607 SNPs were identified with no missing data and representing one SNP every 360 base pairs of the V. inaequalis genome. About 38.1% (n=76,130) SNPs were detected in European isolates and 25.2% (n=50,292) were present in the two isolates collected from U.S. locations other than Darrow farm (Supporting Figure S1). The seven isolates from Darrow farm constitute about 68.3% (n=136,406) of total SNPs identified. Approximately 4.4% of SNPs were shared between these three groups of V. inaequalis isolates (Supporting Figure S1). The 6 isolates from M. floribunda 8221 had 119,109 SNPs representing about 59.6% of total genetic diversity in the population. There were 55,130 SNPs identified in the three replicated isolates collected from the same M. floribunda 821 tree. We further observed about 0.29 nucleotide diversity (π) in this V. inaequalis collection. The seven isolates collected from Darrow farm had a π value of 0.39 and 6 isolates from M. floribunda 821 exhibited π value of 0.43, thus representing slightly higher nucleotide diversity than the entire collection. The nucleotide diversity of the two U.S. isolates outside of Darrow farm and the two European isolates was not estimated due to limited sample size.
Population Genetic Structure Analysis
The population structure was examined using 16,321 high-quality SNPs to perform principal component analysis (PCA) and BYM admixture estimation of V. inaequalis genomes. PCA clearly separated the 11 isolates into two distinct European and U.S. groups (Figure 3). The two European isolates from Belgium and Netherlands also showed clear distinction from each other. The U.S. groups also exhibited slightly dispersed pattern (Figure 3). For instance, two V. inaequalis isolates from outside the Darrow farm had moderate separation from the remaining isolates. Five of the seven isolates collected from Darrow farm displayed a close clustering, while two of them had a distinct placement on PCA biplot. The latter two isolates represent two of the three replications that were collected from the same tree of M. floribunda 821. The remaining isolate co-localize with the other isolates from M. floribunda 821. The other subgroup of 5 U.S. isolates from Darrow farm had diverse host range including ‘Golden Delicious’, ‘Nova Easygro’, and M. floribunda 821.
A genome admixture analysis reflected four main clusters across the 11 V. inaequalis isolates (Figure 4). Two of the clusters represent the two European isolates and the other two clusters were specific to the U.S. isolates. A major proportion of genomic composition was different between the two European isolates, which also differed from the remaining isolates from U.S. regions. A U.S. isolate VI-NY05-GD obtained from ‘Golden Delicious’ exhibited partially different genomic composition than the other U.S. isolates from within and outside Darrow farm. The six isolates from M. floribunda 821 mostly showed similar genome composition except VI-NY03-MF, which had some admixture from European genomes (Figure 4). The main differences observed in genomic composition at K=4 remained consistent as K value changed from 3 to 5. At K=5, the few U.S. isolates and two European isolates showed a small proportion of distinct genome other than the four groups observed at K = 4.
Discussion
We observed severe apple scab symptoms on trees of the Japanese crabapple M. floribunda 821, which carries the scab resistance genes Rvi6 and Rvi7, in a research orchard at Geneva, New York, USA in 2019. The breakdown of Rvi6 resistance was reported within a decade of its deployment from several European locations (Bénaouf and Parisi 1997; Parisi et al. 1993), but until now resistance has remained effective elsewhere. This is the first confirmed loss of the resistance of M. floribunda 821 in North America. Seven cultivars descended from M. floribunda 821 (‘Dayton’, ‘Jonafree’, ‘Prima’, ‘Priscilla’, and ‘Redfree’, ‘Florina’, and ‘Nova Easygro’) within the same orchard developed levels of scab much lower than observed on Japanese crabapple, despite their common ancestry with Japanese crabapple and possession of Rvi6 (Table 2). The high incidence of scab across a range of known susceptible cultivars, and the uniform infection among all replicate trees of M. floribunda 821, would indicate that low scab incidence among these remaining resistant cultivars was not due to escape, but because of the presence of Rvi6 in a genetic background still contributing additional resistance to V. inaequalis. The foregoing would be consistent with the report by Roberts and Crute (1994) that many Rvi6 cultivars remained resistant to scab in Europe after the resistance of M. floribunda 821 had eroded there. Parisi et al. (2004) also reported that ‘Prima’ remained more resistant to scab than M. floribunda 821 when challenged by isolates able to overcome Rvi6. It has been proposed that Rvi1, derived from ‘Golden Delicious’, might contribute additional resistance when paired with Rvi6. However, we observed continued resistance to scab among many Rvi6 cultivars that are not known to carry Rvi1, meanwhile slight sporulation was observed on ‘Prima’ carrying Rvi1 alongside Rvi6 (Table 2). Gessler and Pertot (2012) speculated that Rvi1 might occasionally act indirectly against Rvi6 virulent isolates, and Caffier et al. (2010) hypothesized a fitness cost to virulence towards Rvi6 conferred by loss of avrRvi6. However, the ability to overcome a larger number of R genes (excluding Rvi6) has not correlated with low fitness in past studies (Parisi et al. 2004; Peil et al. 2018).
It is important to distinguish between mild chlorotic flecking typical of greenhouse inoculations and the loss of field resistance to apple scab, wherein profusely sporulating lesions are macroscopically visible in orchard assessments. The chlorotic and necrotic symptoms seen on the field on ‘Prima’ and ‘Nova Easygro’ cannot be considered as evidence that their field resistance has been overcome. Likewise, chlorotic lesions supporting sparse sporulation have been reported on ‘Prima’ following greenhouse inoculations (Crosby et al. 1990). Even in cases of compatible reaction between the host and the pathogen, scientific agreement on whether resistance can be considered broken might depend on the severity and economic impact (Delmotte et al. 2016). Sparse chlorotic sporulation on commercial Rvi6 cultivars is not generally considered a sign of resistance breakdown (Bus et al. 2011). On the other hand, Rvi6 cultivars including ‘Macfree’ often appear to be susceptible under greenhouse conditions, indicating the breakdown of Rvi6 resistance, but remains highly resistant in field plantings (Crosby et al. 1990; Roberts and Crute 1994). We observed similar mild and chlorotic symptoms on several scab resistant cultivars in the field (Table 2), but these are easily distinguished both qualitatively and quantitatively from the symptoms recorded on M. floribunda 821.
The earlier discovery of Rvi6 and Rvi7-virulent isolates in Europe creates the possibility of transportation of such isolates to North America and subsequent spread. However, our comparison of genetic polymorphism of the European and American virulent isolates suggested that both evolved or were selected from genetically distinct and separate populations. In Europe, the Rvi6 virulent lineage has existed separately in non-agricultural areas for thousands of years (Lemaire et al. 2015). While there is considerable genetic distinction between the isolates from ‘Golden Delicious’ and the new virulent isolates, we do not have sufficient data to entirely exclude the possibility of gene flow and human transportation of isolates.
The considerable genetic polymorphism among the M. floribunda 821 virulent American isolates makes them less likely to be clonal variants. Interbreeding between Rvi6 virulent and non-virulent isolates has already been documented in mixed orchards (Michalecka et al. 2018). Even though our isolates were collected within an unsprayed experimental orchard comprised of several cultivars, genetic dissimilarity detected between the groups suggests that intense hybridization has not yet occurred between the M. floribunda 821 virulent and avirulent isolates studied here.
To successfully manage the disease resistance of cultivars, the distribution and evolutionary processes of pathogen isolates must be understood. The breakdown of resistance from M. floribunda 821, but not in descendant cultivars, suggests that the overall genetic background into which a single major R gene is incorporated can substantially affect outcomes with respect to durability of resistance over time. Broad-spectrum and partial polygenic resistance is considered a durable alternative to major resistance genes (Corwin and Kliebenstein 2017; Robinson, 1996; Khan and Chao 2017). Breeders prefer to choose the most promising disease resistance genes and gene combinations to achieve commercially successful cultivars with durable disease resistance. However, due to the time needed to select and breed for polygenic resistance and the possibility of linkage drag of undesirable alleles, breeding programs have prioritized major monogenic resistances. In-depth understanding of the factors resulting in the emergence of M. floribunda 821 resistance-breaking isolates, and the breakdown of resistance from M. floribunda 821 but not in descendant cultivars, may ultimately reduce the risk of eroding the resistance of Rvi6 cultivars and enhance their durability.
Conflict of interests
The authors declare that they have no competing interests.
Competing interests
All authors read and approved the manuscript. The authors declare that they have no competing interests.
Supporting Figure S1. Venn diagram showing the distribution of total SNPs in Venturia inaequalis isolates from Europe and USA (both Darrow farm and non-Darrow farm).
Acknowledgments
This research was supported by the New York Apple Research and Development Program (ARDP). We thank David Strickland and Liqiang Gao for culturing and DNA extraction of some of the isolates used in the study and Della Cobb-Smith for technical support.
Footnotes
mak427{at}cornell.edu