Identification of the Er 1 resistence gene and RNase S-alleles in Malus prunifolia var .

Woolly apple aphid (WAA; Eriosoma lanigerum Hausm.) is a major insect pest that has significant economic impact on apple growers worldwide. Modern breeding technologies rely on several molecular tools to help breeders select genetic determinants for traits of interest. Consequently, there is a need for specific markers linked to the genes of interest. Apple scions and rootstocks have an additional barrier to the introduction of pest resistance genes due to the presence of self-incompatibility S-RNase alleles. The genetic characterization and early identification of these alleles can amplify the contribution of a breeding program to the selection of resistant genitors that are as compatible as possible. In this study, we identified the Er1 gene involved in the resistance to WAA in Malus prunifolia var. ringo, also known as ‘Maruba Kaido’ rootstock, and we analyzed the inheritance pattern of the WAA resistance Er1 gene in a segregant population derived from Malus pumila ‘M.9’ and ‘Maruba Kaido’ rootstocks. The self-incompatibility of S-RNase alleles S6S26 of ‘Maruba Kaido’ were also identified along with their inheritance pattern. We also confirmed the identification of the S1S3 alleles in the ‘M.9’ rootstock. To the best of our knowledge, this is the first study to characterize WAA resistance and RNase S-alleles in ‘Maruba Kaido’. Furthermore, we discuss the potential use of the genetic markers for these genes and their potential impact on apple breeding programs.


Introduction
Current breeding technologies have included a wide range of traits or selection strategies to satisfy the needs of growers and markets worldwide in developing pest and disease resistant varieties, as well as varieties that meet fruit quality, precocity and yield standards (Evans et al., 2010).
Many breeding efforts over the last few years have focused on developing techniques of marker-assisted selection (MAS) that indirectly select genotypes for one or more traits of interest.Although improving the efficiency of selection by making it a more precise system, the technique relies on the identification of a sufficient number of markers.Further complicating breeding practices among apple species is the fact that most apple species exhibit a gametophytic, self-incompatibility (SI) mechanism, which prevents fertilization following selfpollination.The simple genetics of the gametophytic SI system imply that the cross-pollination pattern of an apple variety may be predicted from its RNase S-allele constitution (Broothaerts, 2003).
Considering the paucity of genetic information available about the WAA resistance source 'Maruba Kaido' and its SI systems, this study aimed to identify S-RNase alleles and a WAA resistant gene present in this variety by using a segregant population derived from a 'M.9' × 'Maruba Kaido'.In addition, we also confirmed the identification of the S 1 S 3 alleles in the 'M.9' rootstock used by Brazilian breeders.

Plant material and phenotypic analysis for resistance to WAA
A segregant population derived from the controlled cross between 'M.9' (Malus pumila) and 'Maruba Kaido' (Malus prunifolia) apple rootstocks was developed in order to investigate WAA resistance and the selfincompatibility alleles present in 'Maruba Kaido'.The cross was performed in the spring of 2002 in São Joaquim, Santa Catarina, Brazil.The population was composed of 49 F 1 individuals using 'Maruba Kaido' as the pollen donor.The seedlings obtained were planted in pots and maintained in a greenhouse.After one year, the seedlings were transferred to the field in Caçador, Santa Catarina, Brazil (26° 46' S; 51° 00' W), where they were multiplied using the stool layering technique.
In the following years the progeny was repeatedly inoculated with WAA by attaching heavily infested root pieces to each seedling in the field (Cummins and Aldwinckle, 1983).WAA samples were randomly collected from São Joaquim orchards and colonies were maintained on susceptible plant material.There was no evidence of biotypes being present in these inoculums that could overcome the WAA resistance under investigation.WAA infestations were assessed at various intervals, from 2 to 3 months after inoculation up to four years after transplanting to the field.
The average number of shoots infested per seedling during the evaluations was used in order to classify the susceptibility level using a three point scale adapted from Bus et al. (2008): Resistant = score 1, no infestation; Medium infestation = score 2, up to 2 shoots infested; and Susceptible = score 3, more than 2 shoots infested; scores of 2 and 3 were considered susceptible.Overwintering colonies of WAA are usually found in old pruning scars and only a few survivors were recorded on shoot galls from the previous year's infestation.Therefore, shoots infested in the previous year do not count in the scoring in the following year.Statistical analyses were based on the score of resistance.The c 2 test was applied to determine the inheritance of the genetic resistance derived from 'Maruba Kaido'.

Bulk segregant analysis using molecular markers linked to the resistance genes Er1, Er2 and Er3
Resistant and susceptible bulks were composed of extreme phenotypes (Resistant = score 1; Susceptible = score 3) based on the phenotypic characterization of the WAA resistance in the 'M.9' × 'Maruba Kaido' population.Bulks were analyzed according to the Bulk Segregant Analysis strategy (BSA) (Michelmore et al., 1991).Each bulk was composed of equal amounts of DNA from 10 genotypes.The resistant and susceptible bulks, parental and positive controls were genotyped with molecular markers (sequence characterized amplified regions -SCARs; simple sequence repeat -SSR; and single nucleotide polymorphism -SNPs) (Table 1), previously described by Bus et al. (2008) as being linked to the Er1, Er2 and Er3 WAA resistant genes.
The SNP NZsn_O05 marker can amplify both Er1 and Er3 genes.In order to be able to discriminate Er1 and Er3 genes, we have applied two additional Er1 exclusive markers.Samples of the varieties 'Northern Spy' (M.× domestica), M. × robusta and M. sieboldii accessions collected at the Apple Germplasm Bank (Caçador, Santa Catarina, Brazil) were used as positive controls as they have been previously characterized as presenting Er1, Er2 and Er3 genes, respectively (Sandanayaka et al., 2003).Genomic DNA was extracted from young healthy leaves following the cetyl trimethyl ammonium bromide (CTAB) method (Doyle and Doyle, 1990) and quantified by spectrophotometer.Samples were stored at -20 o C un-til use.Polymerase chain reaction (PCR) conditions were put in place following Bus et al. (2008).PCR products were visualized by using 1.5 % and 3 % agarose gel electrophoresis and a fluorescent staining.

Validation of WAA resistance gene marker in 'M.9' × 'Maruba Kaido' progeny
The 49 F 1 individuals from the 'M.9' × 'Maruba Kaido' population were randomly screened in a doubleblind test coupling phenotypic evaluation with the results of genotype testing to validate the suitability of the NZsn_O05 marker for the Er1 gene.At least three replicates of PCR analysis were conducted.

S-locus genotyping
The same DNA samples for the WAA investigation were also used to examine the self-incompatibility RNase alleles present in this population.Sixteen RNase S-allele primers from various sources (Table 1) were used to screen the 'M.9' × 'Maruba Kaido' population.PCR, enzymatic digestion and gel electrophoresis conditions for the S 1 , S 2 , S 3 , S 4 , S 5 , S 6 , S 7 , S 9 , S 10 , S 16 , S 19 , S 20 , S 22 , S 23 , S 24 and S 26 were performed according to Albuquerque-Junior et al. ( 2011) with modifications.Small amplified fragments (< 300-bp) were separated on a 4 % denaturing polyacrylamide gel and visualized using silver-staining as described by Creste et al. (2001).
Product sizes were estimated by comparison with a 10-bp DNA ladder.Several scion cultivars of M. domestica (Apple Germplasm Bank) were used as positive (S-allele present) and negative (S-allele absent) controls.These cultivars were characterized by Albuquerque-Junior et al. (2011).The c 2 test was applied to determine the inheritance of the S-alleles derived from 'Maruba Kaido'.

Er resistance gene identification
The phenotypic characterization revealed that 'Maruba Kaido' WAA resistance is monogenic.Therefore, the BSA strategy was employed to identify the resistance gene (or locus) in the apple rootstock genome.To screen the bulks, four molecular markers previously described as being associated with WAA resistance were used.In the variety 'Northern Spy', the SCAR markers NZsc_GS327 and NZsc_O05 are linked to the Er1 resistance gene.The SNP marker NZsn_O05 is linked to the Er1 and Er3 Sci.Agric.v.72, n.1, p.62-68, January/February 2015 genes in 'Aotea 1'.Furthermore, the SSR marker NZms_ EB145764 is associated with the Er2 gene in 'Robusta 5'.Expected sizes for each of the four fragments were observed in the controls utilized in this study (Table 1).
The markers NZsc_GS327, NZsc_O05 and NZms_ EB145764 revealed the amplification of fragments for both parents and bulks.However, the marker NZsn_O05 (linked to Er1 and Er3) was polymorphic and only amplified the expected fragment (880bp) in the resistant genitor 'Maruba Kaido' and the resistant bulk (Figure 2).
Both NZsc_GS327 (Er1) and NZsc_O05 (Er1) amplified a fragment in susceptible parents and progenies that are different from the one associated with resistance.
However, it is difficult to differentiate susceptible and resistant alleles because of the similar size and the resistant progeny may also present the susceptible allele.The SSR marker NZms_EB14564 (Er2) is also expected to show amplification in 'M.9'.Moreover, NZsnO05 (Er1 and Er3) shows a complex set of alleles (Bus et al., 2008).With this in mind, we conducted another set of PCR reactions in order to detect the alleles in all four markers in a higher resolution gel (3 % agarose).From this analysis we were able to detect alleles in the NZsc_O05 (Er1) and NZsn_ O05 (Er1 and Er3) markers that were present only in the resistant parent individual and the resistant bulk (Figure 3).This evidence suggests that Er1 is present and not Er3, since we were able to detect the co-dominant bands in NZsc_O05 and the inheritance experiment showed monogenic resistance.Nevertheless, NZsn_O05 amplifies both genes.To the best of our knowledge, this is the first time that the Er1 gene has been described for 'Maruba Kaido'.

Validation of the NZsn_O05 marker
We used the segregant progeny from the cross 'M.9' × 'Maruba Kaido' to validate the link between marker NZsn_O05 and the Er1 gene which is present in the resistant parental 'Maruba Kaido'.Fourty nine seedlings were tested for the presence of this marker and the resistant phenotype in a double-blind analysis.Twenty one out of 29 seedlings that were resistant to WAA (phenotype classified as score 1, Resistant) were correctly predicted (Table 2).Therefore, eight plants with the resistant phenotype did not show the Er1 gene through PCR analysis.All 20 individuals identified as a susceptible phenotype (score 3) were correctly predicted by the absence of PCR fragment amplification of the NZsn_O05 marker.

Self-incompatibility RNase alleles
RNase S-alleles from apple cultivars have been previously cloned by several authors (Broothaerts et al., 1995;Janssens et al., 1995;Sassa et al., 1996;Verdoodt et al., 1998;Matsumoto et al., 1999;Matsumoto and Kitahara, 2000;Nerum et al., 2001;Schneider et al., 2001;Kitahara and Matsumoto, 2002;Broothaerts, 2003).Based on the nucleotide sequences of the S-alleles encoding S-RNases, a PCR on restriction fragment length polymorphism analysis method (PCR-RFLP) for S-allele identification was used (e.g.Albuquerque-Junior et al., 2011).The method consists of PCR amplification of each S-allele using specific primers for individual S-genes.We used 16 primer combinations to screen the most common apple rootstocks present in Brazilian orchards.
Our results confirm the presence of S 1 and S 3 haplotypes for the 'M.9' cultivar previously detected by Matsumoto et al. (2003).We also identified S 6 and S 26 in the 'Maruba Kaido' rootstock and we believe that this is the first time that these alleles have been observed and documented for this variety.S 26 is a rare allele which has been shown to differ from other S-alleles and it has been detected in the crabapple variety 'Baskatong' (M.domestica × niedzwetzkyana), M. floribunda 821 (Verdoodt et al., 1998;Broothaerts, 2003) and in M. × domestica old and M. sylvestris hybrids (Dreesen et al., 2010).

Discussion
Woolly apple aphid is a pest which can have major economic consequences for apple growers, particularly in Brazil and other countries in the Southern Hemisphere (Denardi, 2002).Despite the fact that resistance of 'Maruba Kaido' to WAA was described more than 30 years ago (Way et al., 1990), the resistance gene has not been utilized in Brazilian commercial apple breeding.The hesitation to include this resistant variety in breeding programs is due to the lack of knowledge of the inheritability of such genes, thus making commercially desirable resistant phenotypes difficult to obtain.
Bulked segregant analysis (Michelmore, 1991) in combination with SCARs markers (Paran and Michelmore, 1993) are powerful techniques for identifying markers tightly linked to, or co-segregating with, genes underlying monogenic traits (Malek et al., 2000).Using these techniques, coupled with phenotypic analysis, we were able to identify a new source of resistance to WAA from the 'Maruba Kaido' rootstock.
The first evidence of a bimodal segregation pattern of WAA resistance was provided by Bus et al. (2008), suggesting the presence of major genes for resistance in 'Northern Spy' (Er1), 'Robusta 5' (Er2) and 'Aotea 1' (Er3).However, significant segregation distortions towards either resistance or susceptibility were observed for all three genes.Resistance in other apple species is monogenic and located in the apple chromosome 8 for the Er1 and Er3 genes and chromosome 17 for the Er2 gene (Bus et al., 2008).In addition, the possibility of a potential link between Er1 resistance gene and the SI locus was discarded by Tobutt et al. (2000) after investigation with the aid of the Got-1 and Got-4 isozyme markers.
The successful transferability of Er1 gene marker to 'Maruba Kaido' rootstock and its validation are of ma-jor interest to apple breeders in Brazil.The availability of these molecular markers represents an additional benefit to the breeding program, as it is possible to screen for specific genes.No false positive genotypes were screened for the presence of the Er1 gene and therefore the marker combination assures a high degree of certainty that the gene will be present in subsequent progenies.
Our results accurately predicted both resistant genotypes and phenotypes at a rate of 72.41 % (PCR results in correlation to phenotypic characterization).Previous studies predicted 89.6 % of all resistant individuals from the progeny 'Royal Gala' × 'Northern Spy' segregating for the Er1 gene (Bus et al., 2008).Although the percentage of correct prediction of Er1 was not high (72.41%) considering the distance of 4 cM from the gene (Bus et al., 2008), the NZsn_O05 marker did not provide false positive results, which are more problematic to breeding programs.The lack of detection of the Er1 gene in some of the phenotypically resistant individuals might be related to either an incorrect phenotypic characterization or mismatched amplification of PCR.Moreover, the identification of the Er1 in 'Maruba Kaido' brings another genetic resource for developing WAA resistance.By pyramiding multiple genes, this resource can be used to create new rootstock cultivars with durable pest and disease resistance.
We were also able to confirm the S 1 S 3 self-incompatibility alleles for 'M.9' and identify the S 6 S 26 alleles in 'Maruba Kaido' for the first time.As expected for an S-compatible cross, the frequency of the four segregating S-alleles was similar (c 2 = 0.852; p = 0.64, df = 1); however, there was genotypic segregation distortion towards two types (S 3 S 26 and S 1 S 6 ) in the interspecific progeny.One explanation for segregation distortion is the presence of post-zygotic barriers.Fertile hybrids are often obtained from pollination between Malus species with compatible S-alleles.However, hybrid lethality is commonly observed.In this study, approximately 300 seeds were obtained from the interspecific cross, but only 100 were able to germinate and only 60 became established.Of the 60 seedlings, three were triploids and eight were produced through contamination by foreign pollen (data not shown).
Alternatively, sequence similarity of some Salleles, which would produce a partial incompatibility response, could lead to the segregation distortion.Heng et al. (2011) investigated the recognition speci-  and observed that RNase S-alleles present in both species have maintained the same recognition specificity after the divergence of the two species.Heng et al. (2011) also observed that amino acid substitutions found between PpS 8 -RNase and MsS 3 -RNase do not alter the recognition specificity.Therefore, further studies on the sequence similarity and incompatibility reaction must be performed in order to investigate possible partial incompatibility between parents.Furthermore, S 6 and S 26 sequences are not yet fully characterized in the literature.Previous studies have determined that mutations in the nucleotide sequence might change amino acid sites that determine specificity differences (Vieira et al., 2010).Because of self-incompatibility, most apple cultivars need cross-pollination in order to bear fruit.Discrimination of S-genotypes of apple cultivars is, therefore, important for selecting the proper pollen donor in fruit production (Kim et al., 2009;Li et al., 2011).
Brazilian apple orchards are in real need of new resistant rootstocks that include the traits of dwarfing and fruit quality.To date, the wide use of 'M.9' and 'M.26' rootstocks has increased production by improving fruit quality and enabling mechanized harvest; however, these rootstocks demand higher amounts of insecticides and fungicides.Therefore, integrating knowledge of the genetic identification of both WAA resistance and self-incompatibility in 'Maruba Kaido' could lead to the development of genetically resistant cultivars for use in traditional breeding.
The use of such cultivars would benefit not only Brazilian and worldwide apple breeders, but consumers as well.Nonetheless, further work should be done in order to characterize the genomic region related to WAA resistance by sequencing the genes already identified and to prospect other alleles.Since the malus genome is now available in public databases, denser genetic maps around these genes could be developed, which will then help the development of other molecular markers with greater potential to improve their effectiveness as marker assisted selection tools in other genetic backgrounds.

Table 1 −
The sequences and cultivar references of the PCR primers for the sequence characterized amplified region (SCAR), single nucleotide polymorphism (SNP), and simple sequence repeat (SSR) markers linked to woolly apple aphid (WAA; Eriosoma lanigerum) resistant genes used in this study.S-RNase specific primers used to identify the self-incompatibility alleles present in Malus prunifolia 'Maruba Kaido' apple rootstock are also included.