Introduction

Stagonospora nodorum blotch caused by the necrotrophic fungus Stagonospora nodorum (E. Mull.) Hedjar (anamorph; Phaeosphaeria nodorum) is a major disease of common wheat (Triticum aestivum L., 2n = 6x = 42, AABBDD genomes) and durum wheat (T. turgidum L. ssp. durum, 2n = 4x = 28, AABB genomes), and it occurs in all major wheat growing areas of the world. The fungus has the ability to infect leaves causing Stagonospora nodorum leaf blotch (SNLB) and glumes resulting in Stagonospora nodorum glume blotch (SNGB). Therefore, it has the ability to cause significant yield losses (Eyal 1981; King et al. 1983) and negatively impact grain quality (Eyal et al. 1987).

Recently, a number of host–toxin interactions have been characterized in the wheat-S. nodorum pathosystem (Liu et al. 2004a, b, 2006; Friesen et al. 2006, 2007, 2008). Liu et al. (2004a, b) identified and partially characterized the first S. nodorum host-selective toxin (HST), designated SnTox1, and mapped the corresponding host sensitivity gene (Snn1) on the short arm of chromosome 1B. In QTL analysis, the Snn1 locus accounted for 58% of the variation in susceptibility to SNLB indicating that a compatible Snn1–SnTox1 interaction played an important role in causing disease.

The second S. nodorum HST was identified by Friesen et al. (2006) and designated SnToxA. This toxin was first identified in the tan spot fungus (Pyrenophora tritici-repentis) (Tomas and Bockus 1987; Lamari and Bernier 1989), another important necrotrophic wheat pathogen, and designated Ptr ToxA. Friesen et al. (2006) identified a gene (SnToxA) with 99.7% similarity to the Ptr ToxA gene in the S. nodorum genome and showed that the ToxA gene was horizontally transferred from S. nodorum to P. tritici-repentis prior to 1941. The Tsn1 gene, which maps to the long arm of wheat chromosome 5B, was previously known to confer sensitivity to Ptr ToxA (Faris et al. 1996; Haen et al. 2004; Lu et al. 2006). Using Tsn1-disrupted mutants and an intervarietal hard red spring wheat population of recombinant inbred (RI) lines derived from BR34 × Grandin (BG population), it was demonstrated that Tsn1 confers sensitivity to both Ptr ToxA and SnToxA and that Tsn1 accounts for as much as 68% of the phenotypic variation in susceptibility to SNLB (Friesen et al. 2006; Liu et al. 2006) indicating that a compatible Tsn1–SnToxA interaction plays a highly significant role in the development of SNLB.

The third S. nodorum HST to be identified was SnTox2, which causes necrosis on wheat genotypes harboring the Snn2 gene on the short arm of wheat chromosome 2D (Friesen et al. 2007). Spore inoculation experiments of the BG population with a S. nodorum isolate that produced both SnToxA and SnTox2 indicated that the Tsn1 and Snn2 loci accounted for 20 and 47% of the variation in SNLB susceptibility, respectively. In a multiple regression model, the two loci together accounted for 66% of the phenotypic variation indicating that the effects of the Tsn1–SnToxA and Snn2–SnTox2 interactions are additive and that they are highly important in disease development.

More recently, a fourth S. nodorum HST, designated SnTox3, was identified and partially characterized (Friesen et al. 2008). Sensitivity to SnTox3 was governed by the Snn3 gene, which was mapped to the short arm of chromosome 5B in the BG population. Evaluation of S. nodorum isolates that produced both SnTox2 and SnTox3 indicated that compatible Snn2–SnTox2 and Snn3–SnTox3 interactions were both important factors in the development of SNLB with the Snn2 and Snn3 loci explaining as much as 37 and 17% of the variation in disease, respectively.

Therefore, three toxin sensitivity loci—Tsn1, Snn2, and Snn3—have been mapped in the wheat BG population. We previously reported the construction of genetic linkage maps in the BG population consisting of 646 markers (Liu et al. 2005) with markers tightly linked to Snn3 (Friesen et al. 2008), but rather loosely linked to Snn2 (Friesen et al. 2007) and Tsn1 (Liu et al. 2006). The identification of additional markers, particularly ones closely linked to toxin sensitivity loci would be especially beneficial for marker-assisted selection (MAS) of toxin insensitivity alleles. The objectives of this research were to further saturate the genetic maps in the BG population by incorporating additional simple sequence repeat (SSR; microsatellite) markers, identify or develop markers tightly linked to the toxin sensitivity genes Tsn1 and Snn2, and to investigate the utility of the markers for MAS.

Materials and methods

Plant materials

The BG population of RI lines derived from the cross between the Brazilian hard red spring wheat variety BR34 and the North Dakota hard red spring wheat variety Grandin was developed and provided by Dr. James Anderson, University of Minnesota. A total of 118 F7:9 RI lines were used for mapping as described in Liu et al. (2005). A set of 88 accessions representing T. aestivum and T. turgidum subspecies (Table 1) was genotyped using markers developed for Tsn1, and a subset of 16 of the accessions from the same collection (Table 2) was used to evaluate markers closely linked to Snn2 to validate the utility of the markers for MAS. Most of the 88 lines were obtained from the USDA-ARS National Small Grains Collection, Aberdeen, ID, except that seed of Scoop 1, Mexicali, and Altar 84 were obtained from the International Maize and Wheat Improvement Center (CIMMYT), Mexico; Arina and Forno were obtained from Dr. Beat Keller, University of Zurich, Switzerland; and the T. turgidum ssp. dicoccoides accessions 16-1, 16-29, 18-1, A-33, A-35, B-16, B-6, C-19, C-36, I-50, L-1, L-10, L-33, and L-40 were obtained from Dr. E. Nevo, University of Haifa, Israel. The Tsn1 gene was originally mapped in an F2 population derived from W-7976, which is a synthetic hexaploid wheat developed at CIMMYT derived from the durum wheat cultivar Mexicali, and the hard red spring wheat variety Kulm (Faris et al. 1996; Haen et al. 2004). Therefore, the marker alleles possessed by these two genotypes served as a reference for those observed in the 88 accessions.

Table 1 Accessions of Triticum aestivum and T. turgidum subspecies tested for reaction to ToxA and genotyped with markers Xfcp1, Xfcp620 and Xfcp394
Full size table
Table 2 Hexaploid wheat cultivars infiltrated with SnTox2-containing cultures and genotyped with markers XTC253803 and Xcfd51, which flank the Snn2 locus on chromosome arm 2DS
Full size table

Phenotypic evaluations for reaction to ToxA and SnTox2

Cultures of ToxA were purified as described in Zhang et al. (1997) and provided by Dr. Steven W. Meinhardt, Department of Plant Pathology, North Dakota State University. Cultures of SnTox2 were partially purified as described in Friesen et al. (2007). The BG population was previously screened with both ToxA and SnTox2 and the results were presented in Liu et al. (2006) and Friesen et al. (2007), respectively. Here, we used ToxA to screen the 88 accessions listed in Table 1, and SnTox2-containing cultures for screening the 16 accessions listed in Table 2. More lines were not evaluated with the SnTox2-containing cultures because protocols for purifying SnTox2 have not yet been established, and it is possible that the partially purified cultures could contain additional, yet unidentified, toxins. Plants were infiltrated at the second leaf stage as described in Liu et al. (2006) and were scored as sensitive or insensitive 3 days later based on the presence or absence of necrosis, respectively. These experiments were repeated twice.

Molecular mapping

SSR markers not previously screened by Liu et al. (2005) were chosen from the SSR primer sets CFA, CFD (Sourdille et al. 2004), KSUM (Yu et al. 2004), GDM (Pestsova et al. 2000), GWM (Röder et al. 1998), and WMC (Somers et al. 2004). All PCR primer sequences are available in the Graingenes database (http://wheat.pw.usda.gov/GG2/index.shtml).

We are in the process of using a map-based cloning approach to isolate the Tsn1 gene and we have assembled a physical BAC contig spanning the locus (Lu and Faris 2006; Faris et al., unpublished). In the initial stages of BAC contig assembly, Lu et al. (2006) mined flanking BAC sequences for SSRs and developed two markers, Xfcp1 and Xfcp2 that delineated Tsn1 to a 0.8 cM interval in a tetraploid mapping population. Here, we mined new, more closely associated, BAC sequences flanking the Tsn1 locus for SSRs using the SSRIT software (Temnykh et al. 2001). Two SSRs flanking the Tsn1 candidate gene region were identified for which PCR primers were developed (Table 2). These two SSR markers designated Xfcp394 and Xfcp620 delineate the Tsn1 locus to a 0.07 cM genetic interval, which corresponds to a 344 kb segment (Faris et al., unpublished). Here, Xfcp394 and Xfcp1 were mapped in the BG population (Xfcp620 was monomorphic) and Xfcp1, Xfcp394, and Xfcp620 were all used to genotype the set of 88 wheat lines. The physical order of these markers relative to Tsn1 is known to be Xfcp1–Xfcp620–Tsn1–Xfcp394.

We also identified an SSR within the 3′ untranslated region of the 5B homoeoallele (q-B1) of the domestication gene Q on chromosome 5A (Zhang et al., unpublished) and developed the marker Xfcp621. All FCP SSR primer sets were designed using the MacVector v8.0 software (Accelrys, Inc. San Diego, CA).

To target additional markers to the Snn2 locus on chromosome arm 2DS, we downloaded the sequences for 147 ESTs mapped to the 2DS5-0.47-1.00 deletion bin by the NSF-funded EST project (http://wheat.pw.usda.gov/NSF/). All EST sequences were subjected to BLASTn searches of the Dana Farber Cancer Institute (DFCI) gene indices database (http://compbio.dfci.harvard.edu/tgi/) to identify corresponding tentative consensus (TC) sequences. PCR primers were designed from TC sequences (or EST sequences when no corresponding TC was identified) using the program PRIMER3 (Rozen and Skaletsky 2000) (Table 3).

Table 3 Newly developed PCR-based markers, their primer sequences and annealing temperatures, and their chromosomal locations in the population of recombinant inbred lines derived from the hard red spring wheat varieties BR34 and Grandin
Full size table

Protocols for DNA isolation, PCR, and genotyping of SSR and EST-STS markers were as described in Liu et al. (2005) and Lu et al. (2006), respectively. PCR products for all markers were either separated on 6% polyacylamide gels, stained with SYBR green II, and digitally scanned using a Typhoon 9410 variable mode imager (GE Healthcare, Waukesha, WI), or separated on 2% agarose gels, stained with ethidium bromide and photographed.

Markers were placed on the existing framework maps using the computer program Mapmaker v2.0 for Macintosh (Lander et al. 1987). The ‘GROUP’ command was used to identify the correct linkage group and the ‘TRY’ command was used to determine the best marker interval for new markers. The ‘RIPPLE’ command was then used to validate the marker order. Those markers mapping at an LOD < 3.0 were placed in the most likely intervals.

Results

Mapping additional SSR markers in the BG population

Of the 182 SSR primer sets screened for polymorphism between BR34 and Grandin, 75 (41%) detected polymorphisms and led to the identification of 106 SSR marker loci for an average of 1.4 loci per primer set (Table 4, Fig. 1). New SSR marker loci were detected on all chromosomes with the exception of chromosome 3D, which is non-recombinant in this population (Liu et al. 2005). The number of new SSR markers per chromosome ranged from one on chromosomes 3A and 4D to 21 on chromosome 5B for an average of 5.4 per chromosome (Fig. 1). The number of SSR loci detected per primer pair ranged from 1 to 3 (Table 4). Among the three subgenomes, 29, 52, and 25 SSR loci were assigned to the A, B, and D genomes, respectively. The 106 new SSR loci were added to the existing framework linkage maps of the BG population. A total of 75 of the new markers mapped at an LOD > 3.0, whereas 33 markers did not map at an LOD > 3.0 and were placed in the most likely marker intervals (Fig. 1). However, 23 previously mapped markers that did not map at an LOD > 3.0 now meet this criterion with the addition of the 75 new markers.

Table 4 Chromosomal locations of markers generated by 72 SSR primers derived from the CFA, CFD, GDM, GWM, KSUM, and WMC sets and mapped in the population of recombinant inbred lines derived from the hard red spring wheat varieties BR34 and Grandina
Full size table
Fig. 1
figure 1figure 1

Updated genetic linkage maps generated in the BG population with 142 new marker loci. New SSR markers are shown in red and EST-STS markers are shown in blue. Anchor markers, mapped at an LOD > 3.0, are shown with lines drawn across the chromosome. Of the markers mapping at an LOD < 3.0, only newly mapped markers are shown and are placed in the most likely positions. Double diagonal lines through chromosomes indicate grouping at an LOD < 3.0. Approximate positions of centromeres are indicated by black regions on the chromosomes

Full size image

Markers tightly linked to Tsn1

As stated above, Xfcp620 was monomorphic between BR34 and Grandin and thus was not mapped in the BG population. The marker Xfcp1 mapped 0.5 cM proximal to Tsn1 in the BG population and Xfcp394 mapped 0.5 cM distal to Tsn1 (Fig. 1). Therefore, these two SSR markers delineated Tsn1 to a 1.0 cM interval in the BG population.

Identification and development of markers for the Snn2 region

The initial mapping of Snn2 by Friesen et al. (2007) indicated that it resided on chromosome arm 2DS in a 13.5 cM interval flanked by markers Xgwm614 and Xbarc95. The mapping of the additional 106 markers in this work led to the identification of markers Xcfd56 and Xcfd51, which delineated Snn2 to an 8.1 cM interval with Xcfd51 being linked at only 0.4 cM (Fig. 1).

In an effort to target more markers to the Snn2 region, the wheat ESTs mapped to the 2DS5-0.47-1.00 deletion bin were used. Corresponding TC sequences were identified for 107 of the 147 ESTs, and primers were designed for these and the 40 ESTs for which no TCs were identified. Of the 147 primer pairs surveyed, 28 (19%) amplified fragments that were polymorphic between BR34 and Grandin (Table 3) and were mapped in the BG population (Fig. 1). The 28 primer pairs detected 36 loci of which 11, 14, and 8 resided on chromosomes 2A, 2B, and 2D, respectively. Three loci were detected on chromosomes 1B, 5B, and 6B. The EST-STS markers XTC253803, XTC240114, and XBE489611 were all linked to Snn2 on 2DS. Among these, XTC253803 was the closest at 3.6 cM on the distal side of Snn2 (Fig. 1). Therefore, Snn2 is delineated to a 4.0 cM interval by XTC253803 and Xcfd51.

Validation of markers tightly linked to Tsn1 and Snn2

The 88 accessions listed in Table 1 were evaluated for reaction to purified ToxA and genotyped with markers Xfcp1, Xfcp394 and Xfcp620 to evaluate the utility of the markers in predicting reaction to the ToxA. Among the 88 accessions, Xfcp1 detected two different alleles in addition to a null allele, and Xfcp394 and Xfcp620 detected two alleles each. Xfcp1 detected a 402 bp fragment most often associated with the ToxA sensitive (dominant Tsn1 allele) genotype and a 374 bp fragment most commonly associated with the ToxA insensitive (recessive tsn1 allele) genotype. One allele of Xfcp394 consisted of a 328 bp fragment, which was most commonly associated with the Tsn1 allele (ToxA sensitivity) (Fig. 2). The other allele detected by Xfcp394 consisted of a 383 bp fragment most commonly associated with the recessive tsn1 allele (ToxA insensitivity) (Fig. 2). Xfcp620 detected alleles of 226 and 252 bp, which were most commonly associated with the tsn1 and Tsn1 alleles, respectively.

Fig. 2
figure 2

Molecular profiles of 16 hexaploid wheat varieties revealed using markers XTC253803, Xcfd51, Xfcp1, Xfcp394, and Xfcp620. Fragments amplified by XTC253803 and Xcfd51 are separated on 8% polyacrylamide gels and those by Xfcp1, Xfcp394, and Xfcp620 are separated on 2.0% agarose gels. Cultivar names are shown across the top. Marker fragments of known size are indicated by arrows along the right of each image

Full size image

Among the 88 accessions, 56 (64%) harbored “sensitive” or “insensitive” alleles at all three marker loci in ToxA sensitive and insensitive accessions, respectively (Table 1). In other words, no recombination was observed within the segment harboring the three SSR markers and Tsn1 among these 56 accessions. Only four genotypes including two T. aestivum ssp. aestivum spring wheat varieties (BR34 and Selkirk), and two T. turgidum ssp. dicoccoides accessions (18-1 and B-6) had recombination events between Xfcp620 and Tsn1. Null alleles were present for Xfcp1 in 14 accessions. Disregarding these, eight accessions including one T. aestivum ssp. aestivum winter wheat cultivar (Atlas 66), one T. aestivum ssp. compactum cultivar (Premier), one T. aestivum ssp. sphaerococcum accession (971), one T. turgidum ssp. carthlicum accession (7282), one T. aestivum ssp. spelta accession (Steiners Roter Tiroler Dinkel), and three accessions of T. turgidum ssp. dicoccum (425b, 35900, and 372) showed recombination between Xfcp1 and Xfcp620. There were 15 genotypes with recombination events between Xfcp394 and Tsn1. These included five of the eight T. aestivum ssp. aestivum winter wheats evaluated (Cheyenne, Jagger, TAM105, Forno, and Norstar), two durum varieties (Ben and Langdon), one accession each of T. aestivum ssp. macha (6332) and T. turgidum ssp. dicoccum (372), and six accessions of T. turgidum ssp. dicoccoides (16-1,16-29, L-1, L-10, L-33, and L-40). No accession harbored recombination events immediately flanking Tsn1, but the T. turgidum ssp. dicoccum accession 372 harbored recombination events between Xfcp1 and Xfcp620 and between Tsn1 and Xfcp394.

Sixteen of the 88 lines were evaluated for reaction to partially purified cultures of SnTox2 and genotyped with markers Xcfd51 and XTC253803, which flank the Snn2 locus on chromosome arm 2DS (Table 2, Fig. 2). As mentioned above, it is possible that the partially purified SnTox2 cultures may contain additional, yet unidentified toxins. Therefore, sensitivity to the partially purified cultures is not proof that a line is sensitive to SnTox2. However, an insensitive reaction is indicative of SnTox2 insensitivity. The hexaploid wheat cultivars Atlas 66, BR34, Cheyenne, Chinese Spring, Jagger, Salamouni, TAM105, and Opata 85 were insensitive to the SnTox2-containing culture indicating they contain the recessive snn2 allele, whereas the remaining eight lines were sensitive and may carry the dominant Snn2 allele (Table 2).

XTC253803 detected four different alleles among the 16 cultivars (Table 2, Fig. 2). Among the eight cultivars that were sensitive to the SnTox2-containing culture, all had the same 197 bp allele with the exception of Sumai 3, which had a 196 bp allele. Six of the eight SnTox2 insensitive cultivars (Atlas 66, BR34, Cheyenne, Chinese Spring, Salamouni, and TAM 105) all had a 194 bp allele. Of the remaining two SnTox2 insensitive cultivars, Jagger had a 197 + 202 bp fragment pair and Opata 85 had a 197 bp fragment, which was the same as that observed in the SnTox2 sensitive cultivars.

Xcfd51 detected six different alleles, including a null-allele, among the 16 cultivars (Table 2, Fig. 2). The null and four other alleles were present among the SnTox2 insensitive lines Atlas 66, BR34, Cheyenne, Chinese Spring, Jagger, Salamouni, TAM105, and Opata 85. With the exception of Sumai 3, which had the same allele as Opata 85, Xcfd51 detected a common allele in all the cultivars that were sensitive to the SnTox2-containing culture, which differed from those observed among the insensitive lines.

Discussion

High-density genetic linkage maps developed in intervarietal or interspecific wheat populations are particularly useful because MAS is typically performed in such populations. The development of wheat intervarietal linkage maps in the past has been ever challenging due to the inherent low levels of polymorphism. However, with the development and employment of SSR markers, maps generated in several interspecific populations have recently been published (Sourdille et al. 2003; Suenaga et al. 2005; Liu et al. 2005; Torada et al. 2006). The BG population is particularly useful because it segregates for a number of important agronomic traits including disease resistance and end-use quality.

In the original mapping of the BG population, Liu et al. (2005) screened 600 SSR primer sets for polymorphism between BR34 and Grandin, of which 175 (30%) revealed polymorphisms and yielded 328 marker loci. Including SSR and target region amplified polymorphic (TRAP) markers, Liu et al. (2005) reported a total of 646 markers, of which 352 mapped at an LOD > 3.0 and were used to construct the initial linkage maps. The maps spanned a genetic distance of 3,045.8 cM and had an average density of one marker per 8.7 cM. In this work, we added 141 new marker loci to the maps. The framework maps in the BG population now consist of 480 marker loci that span 3595.8 cM and have an average marker density of one marker per 7.5 cM. This is an increase of 18% in map length and a 14% increase in marker density compared to the original BG maps published by Liu et al. (2005). The increase in map length was primarily due to the extension of linkage groups to include the short arms of chromosomes 1A, 2A, and 5D, which were previously uncovered by markers (Liu et al. 2005).

The wheat-S. nodorum system appears to rely largely on compatible interactions between host-selective toxins and corresponding host genes that confer sensitivity, and these components interact in an “inverse” gene-for-gene manner. In this scenario, a compatible host–toxin interaction relies on the direct or indirect recognition of the toxin by a dominant host gene product, which leads to toxin sensitivity and enhanced disease susceptibility. Absence of either the toxin or the dominant host gene precludes recognition and results in an incompatible, or resistant, response. The chromosomal locations of the genes Tsn1, Snn2, and Snn3, which confer sensitivity to the S. nodorum host-selective necrosis toxins ToxA, SnTox2, and SnTox3 were previously reported in the BG population and found to reside on chromosome arms 5BL, 2DS, and 5BS, respectively (Liu et al. 2006; Friesen et al. 2007, 2008). Friesen et al. (2008) tagged the Snn3 locus with the marker Xcfd20, demonstrated that this marker was linked to Snn3 at a distance of 1.4 cM, and indicated its usefulness for MAS.

Friesen et al. (2007) defined Snn2 to a 13.5 cM interval flanked by markers Xgwm614 and Xbarc95 on chromosome arm 2DS. Although these flanking markers are both user-friendly SSRs, the interval of 13.5 cM is rather large for efficient MAS. In the current work, the mapping of additional SSRs and 2DS bin-mapped ESTs led to the identification and development of PCR-based markers that reduced the Snn2 interval to a more desirable 4.0 cM.

The Tsn1 gene was previously delineated to a 15.3 cM interval in the BG population by markers Xfcp261 and Xfcp380 (Liu et al. 2006). Initial work toward the positional cloning of Tsn1 led to the development of SSR markers Xfcp1 and Xfcp2, which delineate Tsn1 to a 0.8 cM interval in a tetraploid wheat mapping population (Lu et al. 2006). Subsequent positional cloning work led to the development of two additional SSR markers, Xfcp620 and Xfcp394, which we validated in the current work. These markers delineate Tsn1 to a 0.07 cM interval, which corresponds to 344 kb, in a tetraploid mapping population (Faris et al., unpublished). We also validated the utility of Xfcp1 described by Lu et al. (2006) who also suggested that Xfcp2 would be suitable for MAS of Tsn1. Therefore, the availability of four user-friendly and effective SSR markers (Xfcp1, Xfcp2, Xfcp394, and Xfcp620) tightly linked to Tsn1 provide wheat researchers multiple options in the event one or more of the markers are monomorphic in given breeding materials. The importance of developing multiple user-friendly markers for a given gene was realized in this work due to the fact that Xfcp620 was not polymorphic in the BG population. Once the Tsn1 gene has been isolated, we will work to develop “perfect” allele specific markers for the gene itself.

Marker-assisted selection against toxin sensitivity loci in backcrossing schemes is particularly beneficial because sensitivity is dominant, and backcrosses to sensitive recurrent parents yield only sensitive plants. Heterozygous BC1 individuals can easily be selected using the codominant markers developed in this work, and used in subsequent backcrosses to the recurrent parent without progeny testing or test crossing. Given the importance of the Snn2–SnTox2 interaction in susceptibility to SNLB and of the Tsn1–ToxA interaction in susceptibility to both tan spot and SNLB, it would be desirable to remove the toxin sensitivity alleles from elite germplasm and cultivars, and the markers identified/developed in this work will be useful for removing the dominant Tsn1 and Snn2 alleles. However, among the 88 accessions tested, linkage disequilibrium was not as high as expected at the Tsn1 locus with 19 (22%) of the lines having undergone apparent recombination events between Tsn1 and either Xfcp620 or Xfcp394. It is interesting to note that, among the eight T. aestivum winter wheat cultivars evaluated, five (63%) showed recombination between Tsn1 and Xfcp394 (Table 1). Therefore, caution is needed when selecting based on these marker profiles, and it is important that the reaction to ToxA is known for breeding materials before the markers are employed.

For Snn2, numerous alleles were detected by both XTC253803 and Xcfd51 among the eight lines insensitive to SnTox2-containing cultures, but all except one of the eight sensitive lines had the same allele (Table 2, Fig. 2). The cultivar Sumai 3 was the one exception, and it had the same allele as the SnTox2 insensitive line Opata 85 at the Xcfd51 locus and a unique allele at the XTC253803 locus. Therefore, either Sumai 3 has undergone recombination events flanking Snn2, or it is actually insensitive to SnTox2 and the sensitive reaction was due to the presence of a yet unidentified toxin in the partially purified SnTox2 cultures. Among the cultivars with sensitivity to the SnTox2-containing culture, Grandin is the only cultivar proven to be sensitive to SnTox2, and it is possible that other “sensitive” cultivars with the same marker alleles as Grandin are not actually sensitive to SnTox2, but rather a yet unidentified toxin. However, the likelihood of that scenario is low. Once the SnTox2 protein has been purified, a more robust effort to characterize germplasm for presence of Snn2 can be undertaken. In the mean time, Xcfd51 and XTC253803 should be useful for selection of genotypes lacking the dominant Snn2 alleles and eliminating SnTox2 sensitivity.