Introduction

Hessian fly [Mayetiola destructor (Say)] is one of the most destructive insects in common wheat (Triticum aestivum L.) and durum wheat (Triticum turgidum L. var. durum) in the world (Berzonsky et al. 2003). Larval attack of the seedling causes plant death or stunting of growth while larval attack of the plant during stem elongation causes broken stems and shriveled seeds. Both result in significant economic losses. Chemical control is rarely used because timing of application is difficult and few effective insecticides are available. The most common cultural control practices are to delay fall seeding and to destroy the volunteer wheat. Biological control with parasites provides little protection for the current crop. The use of resistant cultivars is the most effective and economical approach (Berzonsky et al. 2003).

To date, 32 Hessian fly resistance genes have been identified in common and durum wheat and their relatives, designated H1 through H32 (Williams et al. 2003; Liu et al. 2005; Sardesai et al. 2005; McIntosh et al. 2008). Deployment of some of the resistance genes has provided effective control of the insect in North America. However, wide use of resistant cultivars leads to the emergence of new virulent genotypes of Hessian fly (Gallun et al. 1961). Co-evolution of Hessian fly and its hosts, including wheat and its relatives, generates genetic variations of Hessian fly and then results in various biotypes (Kudagamage et al. 1990; Ratcliffe et al. 1994). One resistance gene usually confers resistance to one or a few biotypes of Hessian fly.

Sixteen biotypes of Hessian fly have been identified according to their virulence to four differentials, H3 (in ‘Monon’), H5 (in ‘Magnum’), H6 (in ‘Caldwell’), and H7H8 (‘Seneca’) (Ratcliffe and Hatchett 1997). The 16 biotypes are designated Great Plains (GP), and A through O. The biotype L is virulent to all four differentials, while GP is avirulent to all of them. However, more biotypes, including vH9 and vH13, have been found (Formusoh et al. 1996; Zantoko and Shukle 1997), and more will be found in future. Some of them may have overcome the resistance from these four genes or other resistant genes. Although H9 and H13 confer resistance to the most virulent biotype L of the 16 biotypes, they do not confer resistance to the biotype vH9 and vH13, respectively. The complexity of the interaction between Hessian fly and resistance genes in wheat and its relatives, and variability of Hessian fly in the virulence necessitate the deployment of newly identified resistant genes.

The gene H26, derived from Ae. tauschii (Cox and Hatchett 1994), confers resistance to Hessian fly populations that are currently difficult or impossible to control with other available H genes (Cox and Hatchett 1994; Wang et al. 2006; Xu et al. 2006). H26 is one of only two genes that are highly effective against a Hessian fly population recently detected in Oklahoma (Ming-Shun Chen, personal communication). However, H26 has not been commercially deployed. H26 was previously assigned to chromosome 4D using monosomic analysis (Cox and Hatchett 1994). Recently, this gene was mapped to the deletion bin 3DL3-0.81-1.00 on chromosome 3D using molecular markers (Wang et al. 2006). Because fewer molecular markers have been assigned to this distal deletion bin on 3DL than the homologous regions on chromosomes 3A and 3B (Somers et al. 2004), H26 locus was only loosely mapped with a few SSR markers and closely linked PCR-based markers have not been identified.

Wheat EST (expressed sequence tag) sequences can be used to develop user-friendly molecular markers such as STS (sequence tagged site) and SSR (simple sequence repeat) (Peng and Capitan 2005; Zhang et al. 2005; Perugini et al. 2008). There are 1,050,314 wheat ESTs available (http://www.ncbi.nlm.nih.gov/sites/entrez). A total of 16,000 of the ESTs were mapped to wheat deletion bins (Qi et al. 2004). These deletion-mapped ESTs are particularly useful for developing PCR-based DNA markers for saturation and fine mapping of a chromosomal interval harboring the genes of interest and gene cloning. So far, 120 wheat ESTs have been mapped to 3DL3-0.81-1.0 (http://wheat.pw.usda.gov/cgi-bin/westsql/map_locus.cgi) and they are excellent resource for developing new PCR-based markers for saturation mapping of the genes within this chromosomal interval.

In addition to wheat EST, genomic sequence information from model species, such as rice (Oryza sativa L.) whose genome has been sequenced, has been used for molecular mapping and gene isolation through comparative analysis in wheat (Liu and Anderson 2003; Distelfeld et al. 2004; Francki et al. 2004). However, due to the many disruptions in collinearity between rice and the genomes of wheat and barley, Brachypodium distachyon has been proposed as another model species of cereals. The bacterial artificial chromosome (BAC) libraries (Foote et al. 2004; Huo et al. 2006, 2008; Hasterok et al. 2006), ESTs (Vogel et al. 2006), and partial assembled genomic sequences have become available in B. distachyon. Limited data suggested that Brachypodium is likely more closely related to wheat than rice (Vogel et al. 2006; Bossolini et al. 2007). Thus, the genomic sequences and ESTs of Brachypodium could be another invaluable resource for molecular mapping and gene cloning in wheat.

The objectives of this study were to saturate the chromosomal region harboring the H26 locus using newly developed PCR-based STS (sequence tagged site) markers, which will facilitate not only deploying the H26 gene in wheat cultivars but also the genomic study of this chromosomal interval; and to determine the collinearity of this wheat genomic region with rice and Brachypodium genome.

Materials and methods

Plant materials and STS marker analysis

The mapping population of 96 F2 individuals derived from the cross between the resistant synthetic hexaploid wheat (SHW) line SW8 (Langdon/Ae. tauschii CIae 25) and the susceptible SHW line SW11 (Langdon/Ae. tauschii H-80-114-1) (Wang et al. 2006) was used for saturation mapping in the present study. DNA was extracted from the preserved (−80°C) young leaf tissues of the population as described by Dellaporta et al. (1983). One hundred and twenty wheat ESTs, assigned to the chromosomal bin 3DL3-0.81-1.0 where H26 resides (http://wheat.pw.usda.gov/cgi-bin/westsql/map_locus.cgi), were used to design primers with the computer program Primer3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) (Rozen and Skaletsky 2000) to detect STS marker loci near the H26 locus. The primers of the STS markers mapped to this chromosomal region are listed in Table 1. Four SSR markers (Xcfd223, Xgwm3, Xcfd211, and Xbarc71) that linked to H26 locus (Wang et al. 2006) were used as the anchors in this study and their primer sequences were obtained from the GrainGenes Database (http://wheat.pw.usda.gov/GG2/index.shtml).

Table 1 STS markers developed from wheat ESTs, rice genomic sequences, and Brachypodium genomic sequences
Full size table

Bulked segregant analysis was performed to identify marker loci closely linked to H26. Two bulks of DNA were prepared by pooling equal amounts of DNA from eight homozygous resistant and eight homozygous susceptible F2 individuals, respectively. The STSs were amplified at optimized PCR conditions. The annealing temperature for PCR was determined based on the melting temperature (T m) of the primer pair. It was calculated by subtracting 5 from the T m of the primer with lower T m value (Innis and Gelfand 1990) and was adjusted based on the relative intensity of target band among the all amplified ones for a primer pair (Table 1). The SSRs were amplified as described by Röder et al. (1998). PCR products were separated on 6% non-denaturing polyacrylamide gels in 0.5 XTBE buffer at 120 W for 1 h. The gels were scanned with a Typhoon 9410 variable mode imager (Molecular Dynamics, Ithaca, NY, USA) after staining with GelRed (Sigma, St. Louis, MO, USA).

Linkage of molecular markers with H26 in the mapping population was analyzed using MAPMAKER 2.0 (Lander et al. 1987) for Macintosh at LOD 6.0 with the Kosambi mapping function (Kosambi 1944).

Comparative analysis

For comparative analysis with the rice or Brachypodium genome, tentative consensus (TC) or EST sequences were subjected to BLASTn searches of the rice genomic sequences in the Gramene database (Ware et al. 2002; http://www.gramene.org/Multi/blastview) or to search of Brachypodium super contigs (http://www.brachypodium.org). For BLASTn searches, the threshold limits for significant hits were at least 80% nucleotide identity for at least 60 bases. We set e-value <e−7 in order to include maximum number of EST or TC hits with lowest e-value. When several significant hits were found, only the best hit was adopted.

To search wheat TCs corresponding to rice PAC AP003238, which is hit by the EST of the closet STS marker (Xrwgs12)–H26, the sequence of the rice PAC was subjected to BLASTn (Altschul et al. 1997) against wheat EST clusters (TIGR gene indices). A significant match for a TC was declared on the basis of e-value of <e−31 which was used to reduce the similarity among the significant TCs. The TC sequences were obtained from the Annotator Search of DFCI (http://compbio.dfci.harvard.edu/tgi/cgi-bin/tgi/tc_ann.pl?gudb=wheat). The TC sequences were then subjected to tBLASTx (Ware et al. 2002; http://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE=Translations&PROGRAM=tblastx&BLAST_PROGRAMS=tblastx&PAGE_TYPE=BlastSearch&SHOW_DEFAULTS=on&LINK_LOC=blasthome) searches of the NCBI nucleotide collection (nr/nt) to identify the putative corresponding protein. A significant match for a protein was declared on the basis of a minimum 80% amino acid identity for at least 50 amino acid residues of the protein sequence and an e-value of <e−7. The hit with the lowest e-value was considered the putative protein of a TC when several matches were found. The known genes within the rice genomic region AP003238 were obtained through TIGR v5 in Gramene (http://www.gramene.org/Oryza_sativa_japonica/index.html).

Results

More than 500 STS primer pairs were designed based on the sequences of the 120 wheat ESTs assigned to the deletion bin 3DL3-0.81-1.00. These primer pairs were then tested for polymorphisms between the two parents of the mapping population. Fifty-one pairs of primers were found to amplify polymorphic bands between two parents (Fig. 1). Bulked segregant analysis identified 24 STS co-dominant markers linked to H26 and they were mapped to the deletion bin 3DL3-0.81-1.00 (Table 1; Fig. 1). Thus, 20% of the ESTs within this chromosomal interval were converted to STS markers.

Fig. 1
figure 1

Four examples of STS primers that generated polymorphic bands as co-dominant markers between two parents. P1 and P2 are the parents SW8 and SW11, S and R are homozygous susceptible and resistant F2 plants, respectively

Full size image

In order to exploit the genomic sequences of the model species, rice and Brachypodium, the collinearities of the wheat genomic region harboring the H26 locus with rice and Brachypodium genomes were studied to develop additional markers for wheat. We designed 48 and 46 pairs of STS primers from the sequences of the rice and Brachypodium genomic regions which are collinear with the wheat genomic region harboring H26, respectively. One STS marker (Xrwgs17) developed from rice and one (Xrwgs10) from Brachypodium genomic sequences were mapped to the H26 region (Table 2). A total of 26 STS markers were mapped to the distal region of 3DL that resides within the deletion bin 3DL3-0.81-1.00 (Fig. 2).

Table 2 STS markers, their corresponding ESTs and TCs, and similarity to the Brachypodium and rice genomic sequences
Full size table
Fig. 2
figure 2

Comparative analysis of the saturated genetic map of 3DL terminal region, the consensus SSR genetic map of 3D, and 3D deletion map. Marker loci are listed to the right and centiMorgan (cM) distances to the left. Dashed lines link the same loci in the three maps. a Saturated genetic map of H26 with 26 new STS loci (Xrwgs1–Xrwgs26). b The consensus SSR map of 3D (Somers et al. 2004) showing gaps in the region. c Deletion map of 3D (Sourdille et al. 2004), indicating the chromosomal bin (shaded terminal region on the long arm) harboring Hessian fly resistance genes H26, H32, and H24. The fraction-length of the chromosomal bins (within parentheses), the breakpoints, and the centromere of Chinese Spring chromosome 3D are shown to the right

Full size image

The genetic map of partial 3DL, which spans a genetic distance of 42.5 cM with 30 molecular markers, was constructed in the population. This linkage map represents an average density of one marker per 1.4 cM. All markers were mapped at LOD ≥ 6.0. Two co-segregating STS markers, Xrwgs11 and Xrwgs12, have distances of 1 cM from the H26 locus. Another marker, Xrwgs10, is 3.2 cM proximal to H26. Several STS markers, such as Xrwgs4, Xrwgs5, and Xrwgs6, were co-segregating in the mapping population (Fig. 2). However, they were developed from different ESTs or TCs.

To evaluate the local collinearity of the deletion bin 3DL3-0.81-1.00 with rice genome we blasted the rice genome using 24 ESTs or corresponding TCs from which the mapped STS markers were developed. Fifteen of the ESTs or TCs hit the distal region of rice chromosome 1, two hit chromosome 3, one hit chromosomes 8 and 10, whereas five of the ESTs or TCs did not hit any rice genomic region under the significance threshold (Table 2). In general, there is good collinearity between the distal region of wheat 3DL and rice chromosome 1. However, wheat EST BE426418 detected a locus close to the one that EST BE444864 corresponded to in rice (Fig. 3).

Fig. 3
figure 3

Collinearity of the H26 region of wheat (left) with corresponding genomic region of rice chromosome 1 (right). Physical locations corresponding to the EST markers on the genetic map of 3DL are indicated as base pairs on the genomic region of rice. Genetic distances between the markers were indicated as cM to the left of the genetic map. The STS marker Xrwgs17 was developed directly from the rice genomic sequence

Full size image

To determine the collinearity between the deletion bin 3DL3-0.81-1.00 and Brachypodium genomic region(s), the 24 mapped ESTs or corresponding TCs were used as queries to BLAST against Brachypodium genomic sequences. Fifteen of the ESTs or corresponding TCs hit the Brachypodium Super contig 13 (Table 2). Two TCs (TC265775 and TC252434) hit the Super contig 3 and one (TC276760) hit the Super contig 2. Six of them did not hit any Brachypodium contig under the significance threshold (at least 60 bases, and an e-value of <e−7) (Table 2). Fourteen of the 15 ESTs or TCs and the marker Xrwgs10 are perfectly collinear between the distal region of 42.5 cM (from marker Xrwgs1–Xrwgs26) of wheat chromosome 3DL and Super13 contig of Brachypodium genome, but, as was seen in rice, EST BE426418 identified a locus close to the one that EST BE444864 corresponded to (Fig. 3).

Since the STS marker Xrgws12 derived from the EST BE426418 is 1 cM from H26 locus we blasted rice genomic sequences using BE426418 as a query (Table 3). The rice genomic sequence collinear with BE426418 is derived from the PAC AP003238. Within this rice genomic region 14 known genes were found (Table 3). Among them are genes for a membrane attach component, lipase, DNA binding, Leucine rich repeat (LRR), and fungal lignin peroxidase (Table 3).

Table 3 Genes within the rice PAC AP003238
Full size table

To search for putative genes corresponding to the wheat TCs that were hit by the rice PAC AP003238 we blasted the wheat EST clusters with AP003238 (Table 4). Twelve significant TCs were identified. Using tBLASTx, we blasted the NCBI nucleotide collection (nr/nt) with these TC sequences and obtained eleven predicted proteins, including one integral membrane protein-2B and one similar to ETS domain (DNA-binding domain). However, functions for rest of the predicted proteins are unknown (Table 4).

Table 4 Wheat TCs and their predicted proteins
Full size table

Discussion

Hessian fly resistance gene H26 was previously mapped to the 3DL distal region 3DL3-0.81-1 (Wang et al. 2006). Because this chromosomal interval has fewer PCR-based markers than other regions according to the consensus genetic map (Somers et al. 2004) the H26 locus was roughly mapped with only eight SSR markers and one TRAP marker in the previous mapping endeavor (Wang et al. 2006). The molecular markers tightly linked to H26 have not been identified previously. Through saturation mapping in the present study we successfully developed 26 new STS markers and mapped them onto this region. Two of the STS markers (Xrwgs11 and Xrwgs12) are 1 cM away from H26 locus. Another marker, Xrwgs10, is 3.2 cM proximal to H26. Since H26 conditions resistance to multiple biotypes of Hessian fly, including vH13 (Cox and Hatchett 1994; Wang et al. 2006; Xu et al. 2006), these three STS markers will be useful for marker-assisted selection in wheat breeding and germplasm development.

In addition to H26, several other economically and genetically important genes reside within the 3DL distal region 3DL3-0.81-1, including H24 (Ma et al. 1993) and H32 (Sardesai et al. 2005) for resistance to Hessian fly, Lr24 for resistance to leaf rust (Puccinia triticina Erikss) (Boyko et al. 1999), R1 for red kernel color (Nelson et al. 1995), Chl2 for hybrid chlorosis (Koba and Tsunewaki 1978; Erayman et al. 2004), and the genes for Esterase-5 (Devos and Gale 1993) and β-(1-3)-Glucanase (Li et al. 2001). Thus, the 26 newly developed STS markers in our study will facilitate the genetic study of a number of important traits or genes in this region.

The homoeologous group 3 of wheat is considered to be collinear to chromosome 1 of rice (Ahn et al. 1993; Kurata et al. 1994; Munkvold et al. 2004; Dilbirligi et al. 2006), but detailed studies on the microcollinearity particularly between 3DL and rice chromosome 1 have not been reported. The results from our study showed that 15 of the 24 ESTs mapped to the deletion bin 3DL3-0.81-1 (63%) had homology to the sequences on rice chromosome 1, suggesting collinearity between the distal region of 3DL and the distal region on the long arm of rice chromosome 1. This result also agrees with the study by Munkvold et al. (2004) in which the distal half of group 3 chromosomes (3L3-0.81-1) has better homology with rice chromosome 1 than the proximal half.

Although the result from this study showed good conservation in the deletion bin 3DL3-0.81-1.00 in terms of the homology with rice chromosome 1 disruptions in the collinearity between 3DL3-0.81-1.00 and rice chromosome 1 were observed. For example, the local inversion and distant translocation were observed for wheat in this study (Fig. 3). These rearrangements were also reported by Munkvold et al. (2004). Yet, the difference between 3DL3-0.81-1.00 and corresponding region on rice chromosome 1 is still too large to efficiently develop STS markers for wheat chromosome 3D using rice genomic sequence based on the collinearity. We designed 48 pairs of STS primers, but only one was mapped to the collinear region on 3DL (Table 2; Fig. 2). Therefore, a cautious approach should be taken when utilizing the rice genomic sequence for fine mapping in wheat.

Due to the limited degree of collinearity observed in wheat–rice comparative studies, B. distachyon has been proposed as a new model grass to study genomics of large-genome cereals (Draper et al. 2001). However, little is known about the collinearity between wheat and Brachypodium at the chromosome level. In the comparative mapping of Lr34 orthologous regions (Bossolini et al. 2007), six of 11 wheat markers detected the collinearity between Brachypodium and wheat chromosome 7A, and six of 11 markers revealed the collinearity between Brachypodium and wheat chromosome 7D. The results from our study revealed that 15 of 24 (63%) wheat ESTs evaluated had similar sequences as rice chromosome 1. Fourteen of the 15 ESTs and one marker are collinear between the distal region of wheat 3DL and Brachypodium Super contig 13 (Table 2). This points to the utility of B. distachyon as a model for the understanding of the large and complex genomes of wheat.

On the other hand, the collinearity between wheat 3DL3-0.81-1.00 and Brachypodium Super contig 13 is not perfect. The discrepancies between wheat 3DL3-0.81-1.00 and Brachypodium Super contig 13 limit the development of STS markers for wheat chromosome 3D from the genomic sequences of the Brachypodium Super contig 13 based on the collinearity. We designed 46 pairs of STS primers with Brachypodium Super contig 13 sequences based on the collinearity, but only one was mapped to wheat chromosome 3D (Table 2; Fig. 2). Therefore, as seen with rice, cautions need to be taken when employing the Brachypodium genomic sequence for molecular mapping and gene cloning in wheat.

Brachypodium was reported to be more closely related to wheat than to rice (Vogel et al. 2006; Bossolini et al. 2007). The genomic regions of Brachypodium and rice corresponding to the wheat 3DL3-0.81-1.00 were found to be similar instead in this study. Both Brachypodium and rice had 63% of evaluated ESTs collinear with wheat (Table 2). The wheat EST BE426418 detected a locus close to the ones that EST BE444864 corresponded to in both rice and Brachypodium (Fig. 3). These results indicate complexity of the collinearity of wheat genomes with Brachypodium genome. Therefore, local comparative mapping is suggested before using Brachypodium genomic sequences for fine mapping or gene cloning in wheat.

The majority of the cloned resistance genes, including those for disease and insect resistance, encode proteins with a nucleotide-binding site (NBS) domain and a LRR domain (Hammond-Kosack and Jones 1997). NBS domains have been shown to bind and hydrolyze ATP in plants and animals, and ATP binding appears to be essential for signal transduction. The LRR domain is a key determinant of protein–protein interactions (Ellis et al. 2000). For examples, the Xanthomosnas resistance gene, Xa1, in rice (Yoshimura et al. 1998) and the nematode resistance gene, Cre3, in wheat (Lagudah et al. 1997) have been shown to be NBS-LRR-like genes. NBS-LRR-like genes can be viable candidates for genes conditioning resistance to pest. In the present study, three genes with DNA binding, or ATP binding, or LRR were found within the rice PAC AP003238. Since the H26 is 1 cM from the marker Xrwgs12 derived from the wheat EST BE426418, and the AP003238 is the corresponding PAC of the wheat EST BE426418 in rice, these three genes could be candidate genes for resistance to Hessian fly. On the wheat side, one gene corresponding to TC237663 that was hit by the rice PAC AP003238, was found to have DNA-binding site (ETS domain) (Table 4). However, fine mapping as well as efforts toward cloning are needed to confirm these results.