- Research article
- Open Access
Substantial genome synteny preservation among woody angiosperm species: comparative genomics of Chinese chestnut (Castanea mollissima) and plant reference genomes
© Staton et al. 2015
- Received: 19 February 2015
- Accepted: 19 September 2015
- Published: 5 October 2015
Chinese chestnut (Castanea mollissima) has emerged as a model species for the Fagaceae family with extensive genomic resources including a physical map, a dense genetic map and quantitative trait loci (QTLs) for chestnut blight resistance. These resources enable comparative genomics analyses relative to model plants. We assessed the degree of conservation between the chestnut genome and other well annotated and assembled plant genomic sequences, focusing on the QTL regions of most interest to the chestnut breeding community.
The integrated physical and genetic map of Chinese chestnut has been improved to now include 858 shared sequence-based markers. The utility of the integrated map has also been improved through the addition of 42,970 BAC (bacterial artificial chromosome) end sequences spanning over 26 million bases of the estimated 800 Mb chestnut genome. Synteny between chestnut and ten model plant species was conducted on a macro-syntenic scale using sequences from both individual probes and BAC end sequences across the chestnut physical map. Blocks of synteny with chestnut were found in all ten reference species, with the percent of the chestnut physical map that could be aligned ranging from 10 to 39 %.
The integrated genetic and physical map was utilized to identify BACs that spanned the three previously identified QTL regions conferring blight resistance. The clones were pooled and sequenced, yielding 396 sequence scaffolds covering 13.9 Mbp. Comparative genomic analysis on a microsytenic scale, using the QTL-associated genomic sequence, identified synteny from chestnut to other plant genomes ranging from 5.4 to 12.9 % of the genome sequences aligning.
On both the macro- and micro-synteny levels, the peach, grape and poplar genomes were found to be the most structurally conserved with chestnut. Interestingly, these results did not strictly follow the expectation that decreased phylogenetic distance would correspond to increased levels of genome preservation, but rather suggest the additional influence of life-history traits on preservation of synteny. The regions of synteny that were detected provide an important tool for defining and cataloging genes in the QTL regions for advancing chestnut blight resistance research.
- Comparative genomics
- Blight resistance
- Plant genomics
- Physical map
- Genetic map
Chestnuts are members of the Fagaceae family, comprised of about 900 species of trees and shrubs including other dominant Northern Hemisphere temperate forest trees such as oaks and beeches . Many members of this family are subject to significant biotic stresses, particularly from fungal and fungal-like diseases. Examples include Phytophthora cinnamomi, an oomycete that attacks hundreds of plant species around the world including both oaks and chestnuts [2, 3], and Phytophthora ramorum, the cause of sudden oak death, a disease currently devastating oaks in western North America . Rapid climate change and alien pathogen introduction have substantially increased the need for genetic tools to insure conservation and improvement of tree resources to meet future environmental challenges. A case in point, the American chestnut (Castanea dentata), was eliminated as a dominant eastern US forest tree species during the first half of the 20th century due an introduced fungal pathogen (Cryphonectria parasitica) . Utilizing the natural resistance of Chinese chestnut (Castanea mollissima) to the blight, chestnut breeding programs have for decades been working to produce a tree of primarily American chestnut genetic background with blight resistance . To assist the breeding efforts and to enable future genetic research in chestnut, an extensive genomic tool development initiative for Chinese chestnut was undertaken, resulting in an integrated set of resources including transcriptome sequences , a dense genetic map , BAC libraries and a genetically anchored physical map . These genomic tools create a powerful platform for trait mapping, applied breeding and comparative genomics.
Comparative genomics affords the opportunity to increase our understanding of the common underlying gene networks and molecular mechanisms of forest tree defense responses that may be conserved among plant species, especially close relatives suffering from similar diseases. However, in deciduous trees, we have limited knowledge of the degree of preservation of the structural organization of the genomes among either close or distantly related species. With the increase in genetic and genomic resources in some key forest and fruit trees species, we are now positioned to comparatively examine the extent of genome preservation at both low and high resolution among these species. These comparisons provide the opportunity to assess the ability to leverage the genomic information from highly characterized plant genomes for application in other less well-characterized species.
The physically mapped, genetically anchored EST sequence markers from the Chinese chestnut genetic map and the Chinese chestnut BAC end sequences (BES) spanning the physical map provide ideal substrates for genome comparisons with other tree species with highly characterized genomes at the sequence level, such as peach (Prunus persica) and poplar (Populus trichocarpa). Identified conserved syntenic blocks in these genomes could assist in candidate gene discovery when coupled with chestnut trait mapping efforts. Furthermore, these genomic comparisons provide a first low-resolution glimpse of the extent of genome change that has proceeded since the split of the families. However, to utilize comparative approaches with completely sequenced and assembled tree reference genomes as a means to identify potential candidate genes in marked intervals of less well sequence characterized species, we need to access the degree of preservation among tree genomes at a high resolution level, e.g. within 1 centiMorgan (cM) windows.
In order to gain both a low and a high resolution comparative picture of the extent genome change between Chestnut and other completely sequenced and assembled plant species genomes, we carried out comparative genome analyses utilizing BAC end sequences representing the physical map minimal tiling path and genetically anchored EST sequences in chestnut against completely sequenced and assembled genomes of angiosperm plant species of both phylogenetically close and distant relationships to chestnut. After completing this low-resolution comparative analysis, we were interested in assessing whether at high resolution the genome structural comparative relationships would be similar or different. Therefore, we carried out a comparative analysis between sequences of three regions of the Chinese chestnut genome that contain QTLs for chestnut blight resistance  with other plant genomes. These sequences were mapped to a number of completely sequenced plant genomes to examine the extent of genome preservation for these regions among diverse plant species. In addition to enabling us to evaluate the extent of preservation of the defined regions of the chestnut genome to other species, these sequences also provided genes to advance the study of the molecular underpinnings of blight resistance in Chinese chestnut as well as a set of markers for use in marker assisted breeding efforts.
In both the low and high resolution comparative analyses, we demonstrate that the peach genome is the best current reference model for chestnut, however surprisingly the grape genome is the second most highly similar genome in this study. Additionally, the comparative sequence analyses for the chestnut blight QTLs enabled us to define in high resolution the degree of gene preservation within the two chestnut QTLs that were previously found to overlap with mapped QTLs for powdery mildew resistance in peach, as reported by , suggesting that these regions might be important in generalized resistance to fungal pathogens. The combination of low and high resolution mapping of genomic sequence data provided in this report suggests that key high quality sequenced woody species genomes could serve as repositories of gene information for candidate gene studies in other related tree species with less well characterized genomes.
Alignments of integrated physical and genetic map to reference genomes
Significant improvements have been made to the integrated physical and genetic map since their publication . An additional 364 overgo probes have been successfully placed on the physical map, bringing the total overgo probes to 1,390. The total number of physical map contigs is 1300, a slight reduction from the previous version. The new probes were designed from Chinese Chestnut EST sequences that contain genetic map markers in order to create more links between the physical and genetic maps. This version of the physical map is available on the Hardwood Genomics Web as version 4 . The new probes hybridized to an average of 16.6 BAC clones each, slightly lower than the previously reported average of 17.5. Because the BACs included in the physical map represent 18X genome equivalents, this suggests that the majority of probes are hybridizing to a single genomic location. Less than 3 % of probes (41) hybridized to more than 50 clones; those that did are likely repetitive sequences in the genome. More than 73 % of probes (1,024) hybridized to one or two assembled contigs in the physical map. This confirms that most unique genomic locations have been assembled into physical map contigs, yielding accurate links from a genetic map location to a physical map location.
Of the 1,390 overgo probes, 1,013 originated from an EST contig also used to develop one or more genetic map markers, thereby allowing an association from the overgo probe to a linkage group. Of these overgo probes, 858 are associated with both a linkage group and a cM position from the consensus genetic map. The latter set of markers was uploaded into the FPC software as framework markers and placed using the FPC ctg - > chr algorithm. This resulted in the placement of 342 physical map contigs to linkage group locations. The 342 anchored physical map contigs span an estimated 522 Mb and represent 41.9 % of the total physical map span. A further 174 contigs are anchored to a single genetic marker by only 1 BAC hybridization, and the BAC marker had stronger hybridization signals to a different physical map contig. These contigs are not considered anchored but are noted in the FPC file. The algorithm found 129 contigs to contain strong overgo links to more than one genetic map location, and these also were not considered anchored. These contigs may be incorrectly assembled or the probes may actually reside at multiple locations in the genome.
In addition to the use of overgo probes, the physical map has been enhanced through BAC-end sequencing. Forward and reverse end-sequencing was carried out on a subset of the clones in the map assembly, resulting in 42,970 sequences [GenBank:HN270092-HN275251, JY172573-187037]. Part of this BES subset was the 1,132 BACs spanning the QTL regions of the genome; the rest were selected based on physical map spacing, with an effort to obtain evenly spaced sequences spanning all contigs. The sequences include 26,873,995 bases. This represents 3.4 % of the chestnut genome, estimated to be 794 Mb . The BAC library CMCMBb contained 19,064 of the BAC end sequences, of which 8,803 BACs had both uccessful forward and reverse sequences, creating a linked pair. The library CMCMBd had 23,872 BES and 11,187 successful BAC end sequence pairs.
The integrated physical and genetic map is a powerful genomic tool for targeted sequencing as well as comparison to other plant genomes. Using the BES spanning the physical map and the genetically-anchored EST-derived overgo probes, the software Symap was used to align the FPC contigs to the sequenced genomes of peach, strawberry, soybean, Medicago, poplar, papaya, Arabidopsis, Eucalyptus, grape and tomato. Peach and strawberry are members of the Rosales order, and they are the phylogenetically the closest examined relatives to the chestnut, which is in the Fagales order. Soybean, Medicago and poplar are members of the Fabidae/Rosid I clade, in orders more distant from Fagales. Arabidopsis, Eucalyptus and papaya fall into the sister Malvidae/Rosid II group. Grape is a core eudicot that is classified as an outgroup to all rosids. The furthest species from chestnut phylogenetically is tomato, a member of the asterids .
Some physical map contigs consistently aligned. Twelve physical map contigs aligned to all ten comparison species. Six of these 12 contigs corresponded to the QTL contigs, which were preferentially saturated with overgo markers and BES (discussed further below). The 12 contigs that aligned to all of the reference genomes studied were both significantly longer than the average physical map contig (1466 CBs versus 823 CBs) and contained more overgo markers (18 versus 3 average overgos/contig). Due to the overgo probes being predominantly EST-based, these contigs may represent gene rich regions and contain less repetitive sequence. It is possible that these regions retained conserved genomic structure on a macrosytentic scale across dicots.
The chestnut genetic linkage map was initially constructed by Kubisiak et al.  from an F2 family of an American x Chinese chestnut cross using RFLPs and RAPDs . Inoculation and phenotyping of the plants were carried out to identify QTLs conferring resistance to blight. Three main regions in the genome associated with variation in phenotype were found, one each on linkage groups B, F and G. The same population was genotyped using SNPs from the updated high resolution genetic map, and the QTL analysis was recalculated with the new markers . The locations of the three QTL regions were refined to LG-B from 40.9 to 50.4 cM, LG-F from 38.1 to 46.8 cM and LG-G from 35.7 to 39.5 cM. The three QTLs are referred to as Cbr1, Cbr2 and Cbr3, respectively.
Assembly of sequences spanning the three blight resistance QTL regions in the Chinese chestnut genome
Genetic map location
LGB (40.9-50.4 cM)
LGF (38.1-46.8 cM)
LGG (35.7-39.5 cM)
The ends of BACs from the QTL minimum tiling paths were forward and reverse end sequenced using Sanger technology. These sequences were compared to the next generation sequence data. The majority of the BAC ends were found in the sequence assemblies: 92 % for cbr1, 86 % for cbr2 and 92 % for cbr3. This affirmed that end sequence was produced from the majority of the BACs in each pool, and that the coverage across the minimum tiling paths were high.
The assembled genomic sequence scaffolds from the BAC pools were annotated for repeat elements and genes. Over 10 % of the sequence was identified as interspersed repeats, comprising 8.45 % retroelements and 1.26 % DNA transposons. LTR elements, common in many plant genomes, encompassed 7.1 % of the bases with slightly more Gypsy elements (4.06 %) identified than Copia elements (2.69 %). Over 9,000 microsatellite sequences were identified in the scaffolds, yielding an average of one microsatellite every 1,492 bases.
Genes in the QTL sequences annotated in silico with the gene ontology term “defense response”
Closest matching NCBI nr protein
Transcription factor tga1
Transcription factor TGA1 (Vitis vinifera)
cc-nbs-lrr resistance protein
Putative disease resistance protein RGA3 (Vitis vinifera)
rna recognition motif-containing protein
PREDICTED: DAZ-associated protein 1-like (Vitis vinifera)
Predicted protein (Populus trichocarpa)
Transcription factor tga1
TGA transcription factor 1 (Populus tremula x Populus alba)
14-3-3-like protein gf14 lambda
Hypothetical protein ARALYDRAFT_496774 [Arabidopsis lyrata subsp. lyrata)
Multicatalytic endopeptidase complex
Proteasome subunit alpha type-7 (Vitis vinifera)
Disease resistance protein at4g27190-like
PREDICTED: disease resistance protein At4g27190-like (Vitis vinifera)
cc-nbs-lrr resistance protein
cc-nbs-lrr resistance protein (Populus trichocarpa)
Feronia receptor-like kinase
Serine/threonine-protein kinase PBS1, putative (Ricinus communis)
PREDICTED: MLO protein homolog 1-like (Glycine max)
Transferase, transferring glycosyl groups, putative (Ricinus communis)
Histone-lysine n-methyltransferase ashh2-like
PREDICTED: uncharacterized protein LOC100245350 (Vitis vinifera)
Set domain protein
PREDICTED: uncharacterized protein LOC100245350 (Vitis vinifera)
Cysteine proteinase rd19a
Cysteine proteinase RD19a (Arabidopsis thaliana)
QTL sequences and symap analysis
The sequence data obtained for the QTL provides an opportunity to assess microsynteny from the chestnut genome to reference plant genomes on a base pair level, and to ask how microsynteny results compare with the macrosynteny results from the physical map alignment analysis. Peach provided the best reference genome in this analysis, with the highest percent of chestnut bases aligned in total and the highest percent of chestnut bases aligned uniquely. This suggests fewer rearrangements between chestnut and peach than between chestnut and the other genomes studied. The known genome duplication events since last divergence between chestnut and poplar as well as between chestnut and soybean are evident by the percent of chestnut BESs aligning to multiple locations in those genomes.
Alignments of chestnut cbr1 clustered in two main regions of the peach genome - scaffold 6 from 15.6 to 18.7 Mb and scaffold 7 from 16.7 to 17.7 Mb. The homology to peach scaffold 6 (from 16.5 Mb to 18.6 Mb) was previously identified in the recent genetic linkage map paper . The two previously reported regions of alignment to peach scaffold 7 (16.7-16.8 Mb and 17.6-17.7 Mb) now appear to be part of one single syntenic block on peach scaffold 7. Chestnut sequence scaffolds containing 4.8 Mb were mapped, representing 70 % of the cbr1 bases. Of the annotated cbr1 genes, 52.1 % match genes most closely in these two peach genome regions.
The previously identified syntenic region for cbr2 was against scaffold 2 from 17.9 to 20.0 Mb in peach. This block was confirmed in the current analysis, but two other regions also emerged as putative homologous areas. The three main syntenic regions found on peach for cbr2 in this analysis are scaffold 2 from 18.2 to 19.4 Mb, scaffold 3 from 13.7 to 14.1 Mb, and scaffold 5 from .7 to 1.0 Mb. The scaffold 2 syntenic region was not predicted by Symap, but the other two were predicted with very similar borders: scaffold 3 from 13.7 to 14.1 Mb (Fig. 3c) and scaffold 5 from .6 to 1.2 Mb (Fig. 3d). Symap did predict a region of synteny for cbr2 on peach scaffold 8 from 17.3 to 17.9 Mb that was not entirely supported by sequence evidence (Fig. 3g). One sequence scaffold, cbr2_scaffold22, aligned near position 17.3 Mb, but no other sequence agreed with the Symap prediction. The scaffolds placed in the three major syntenic regions in peach represent 56.9 % of the cbr2 assembly and 59.4 % of the cbr2 genes had a best match in peach within those three regions.
Cbr3 had the most clearly placed homologous region in peach. The sequence alignments and the Symap alignment agreed that it aligns to peach scaffold 1 from 39.7 to 40.0 Mb (Fig. 3b). Of all the bases in cbr3, 66.6 % were in scaffolds aligning to this region and 42.0 % of cbr3 genes aligned best to that region of peach.
Overall, Symap was effective at identifying most of the regions predicted by sequence alignments. It predicted syntenies in 5 of 6 regions, although of significantly increased size in one instance. It also predicted 1 region of synteny missed from the partial sequence coverage of the QTL regions. Had the comparative Symap results not been available for analyses, the absence of this gene information in the QTL sequences may not have been identified. Syntenic analysis provides both an avenue for transferring information across genomes and for validating new information.
Two main QTLs in peach have been identified for resistance to powdery mildew (Sphaerotheca pannosa var. persicae), located on scaffold 6 from 16.0 to 22.94 Mb and scaffold 8 from 11.48 to 17.21 Mb . As discussed in , two of the chestnut QTLs overlap these two QTLs from peach. The cbr1 QTL from chestnut aligns to a significant portion of the same region of scaffold 6 (15.6 to 18.7 Mb). Cbr2 aligns to the end of the peach powdery mildew QTL on chromosome 8.
Chinese chestnut has emerged as a model species for the family Fagaceae. Naturally resistant to the chestnut blight (Cryphonectria parasitica), it has become particularly important for restoration of American chestnut (Castanea dentata). A backcross breeding strategy utilizing Chinese chestnut as the source of natural resistance has been ongoing since the 1960s. Development of molecular tools to reduce time and space requirements for breeding are vital to quickly advancing American chestnut restoration efforts and for advancing our genetic understanding of resistance to other chestnut pathogens such as Phytophthora cinnamomi. Here we report on new Chinese chestnut genomic data sets including an improved integrated genetic/physical map and genomic sequence across QTL regions. These resources allow the chestnut genome to be compared on both a micro- and macro-synteny scale to other genomic model plant species.
The integrated physical and genetic map enabled a targeted sequencing approach for the QTL regions conferring resistance to chestnut blight. Utilizing overgo hybridization probes corresponding to genetic map markers, six physical map contigs from the QTL regions were selected. From these a minimum tiling path of BACs was pooled and sequenced. The resulting assembled genomic sequence scaffolds provide a list of candidate genes to pursue for further understanding of the molecular underpinnings of blight resistance. These candidate genes have been tentatively prioritized through functional profiling and assignment of Gene Ontology terms. Downstream analysis to explore the role of these genes is ongoing. Comparative genomic approaches as well as sequencing of the corresponding homologous sequences in American chestnut may help to further elucidate the source of resistance to blight in Chinese chestnut.
Identifying and sequencing BACs across a region of interest is a cost-effective strategy for obtaining information on the genes and genomic organization in these important regions without undertaking an expensive and complex whole genome sequencing effort. American chestnut is plagued by biotic challenges beyond chestnut blight. The oomycete Phytophthora cinnamomi causes root rot with a highly mortality rate and is widespread through the native range of American chestnut, presenting a significant barrier to any American chestnut reintroduction . Chinese chestnut can serve as a source of resistance to this pathogen as well as to chestnut blight . If QTL regions of resistance are found using a genetic mapping approach, the same BAC pooling and sequencing approaches could be used once again. The sequenced regions for blight resistance as well as any future desirable QTL have an additional utility beyond candidate gene discovery; they will enable discovery of markers for fine mapping and marker assisted selection. These markers will be essential for large scale backcross breeding efforts to produce a tree of primarily American chestnut origin but with a combination of Chinese chestnut chromosomal segments that confer necessary biotic resistance loci.
The genomic resources for Chinese chestnut also facilitate a survey of macrosynteny retained between chestnut and other species with sequenced genomes. Analysis of peach, strawberry, soybean, Medicago, poplar, papaya, Arabidopsis, Eucalyptus, grape and tomato found putative syntenic blocks in all species. Peach, the phylogenetically-closest hardwood tree from this list, showed the highest amount of conservation with chestnut. The percent of duplicated blocks between the two genomes indicates that no whole genome duplication events (WGDs) have taken place since the divergence of chestnut and peach from their last common ancestor. Poplar and soybean have both undergone recent whole genome duplications, and these events can be detected by the significantly larger amount of chestnut genomic sequence that can be aligned to two separate locations. In poplar, the WGD occurred 60 to 65 Mya , and over half of the chestnut syntenic blocks can be detected in two locations. For soybean, WGDs have occurred twice, at 13 and 59 Mya . The majority of chestnut blocks were detected in two locations, but interestingly, they were not found in three or four locations as might be expected from the occurrence of two WGDs. This is likely due to the ongoing diploidization of the soybean genome, where the majority of homologous genes (61.4 %) were found to exist in blocks involving only two locations while a much smaller percentage were found in blocks in three or four locations (5.63 and 21.53 %, respectively) . The maintenance of multiple regions of synteny with chestnut is not detectable in the other genomes that have undergone a WGD. Very few chestnut regions are aligned to multiple locations in the eucalyptus genome, which has evidence of a WGD 110 million years ago (Mya) , or to the tomato genome, which underwent a whole genome triplication around 71Mya . The lack of evidence of these events may be due to the large phylogenetic distance between chestnut and these species or the rate at which these genomes have undergone diploidization.
While the detection of WGDs does appear to depend on phylogenetic distance, the overall number and length of conserved genomic blocks between species did not always correspond to the phylogenetic distance. For eight species, greater than 87 % of the genomic sequence anchored to pseudo-chromosomes and genomic sequence comprised of less than 7 % N’s (unknown nucleotides). The quality of genomic alignments is likely to be comparable across these eight. The Medicago genome is slightly more fragmented with 71 % of the genomic sequence in eight pseudo-chromosomes and 12 % Ns. The papaya sequence is highly fragmented, with 29 % Ns and no pseudochromosomes. Interestingly, the number of putative syntenic blocks between chestnut and papaya was higher than the blocks found against eucalyptus, another Malvidae with a more contiguous genome reference sequence.
The two best conserved genomes with 39 % of chestnut mapping to the other genome in syntenic blocks were peach and grape. Although the results with the relatively closely-related peach clearly represent the expected close genomic relationship with Chestnut, grape showed greater extent of genome structure preservation than would fit our expectations based on phylogeny. No WGDs have occurred in either chestnut or grape since their last common ancestor. The grape genome is known to be structurally conserved with very few rearrangements since the paleo-hexaploid ancester of all rosids . The slow rate of structural rearrangements in grape that has maintained an ancestral state of organization is a possible reason for the extensive structural homology. However, papaya and strawberry also have not experienced WGDs since divergence from grape and the same level of conservation is not present. This may indicate that chestnut has a slower rate of structural reorganization over time, or another possible hypothesis is that grape is misplaced phylogenetically. In this last regard, it is interesting to note that grape (Vitaceae) was formerly botanically placed by Cronquist near the family Rhamnaceae in order Rhamnales , much closer to peach and chestnut than the other sequenced genomes in our study. Molecular studies led to its placement as an outgroup to the rosid I and II clades but this placement did not have strong support [22, 23]. This former phylogenetic position would support our findings concerning the degree of genome preservation among the study species.
If in fact grape is not misplaced phylogenetically, then it is interesting to note a trend toward woody plant species maintaining a higher level of structural synteny with each other than with herbaceous species in the same family. In this regard, poplar is an additional example. Poplar has maintained a greater degree of genome similarity to chestnut than strawberry, soybean and Medicago, which are all closer to chestnut phylogenetically. This could possibly suggest that woody plant species evolution is somehow more constrained, or much slower, than the herbaceous species within the same family. However, for strawberry in particular, the lack of syntenic blocks could be due to the fact that the strawberry genome assembly is slightly more fragmented than the peach, with 7 % of the sequence comprised of unknown nucleotide gaps versus 1 % for peach. For the species at a further phylogenetic distance, chestnut has double the quantity of conserved genomic blocks with papaya and eucalyptus versus Arabidopsis. However, the Arabidopsis thaliana genome may not be representative of other herbaceous species, as it has undergone recent (<10Mya) large-scale rearrangements and DNA loss [20, 24]. Additional whole genome references from other woody and herbaceous angiosperm species is needed to conduct future comparative work and draw conclusions regarding the relationship of genome preservation to plant growth and habit. A contiguous whole genome sequence from chestnut and other members of the Fagaceae (e.g. oak) would also help to resolve rosid structural synteny on a finer scale.
Alignments of integrated physical and genetic map to reference genomes
Overgo probes were hybridized to the BAC clones and anchored to the physical map using the 3-dimensional pooling procedure described in . Clones selected for BAC end sequencing were re-arrayed and cultured into 96-well plates. DNA was prepared with the ABI Dye Terminator chemistry and universal T7 and Sp6 primers. The results were resolved on an ABI 3730XL sequencer. A pipeline instituted for quality control performed base calling, filtering and vector/low-quality trimming with Phred, cross_match  and lucy . The trimmed sequences with greater than 5 % uncalled bases (N’s) or less than 100 total bases were discarded. To identify conserved regions, the chestnut physical map, BES sequences, and overgo probe sequences were input to the Symap software  and aligned to reference plant genomes at default parameter settings. The genomes for peach (Prunus persica, version 139), strawberry (Fragaria vesca, version 226), soybean (Glycine max, version 189), Medicago (Medicago truncatula, version 198), poplar (Populus trichocarpa, version 210), papaya (Carica papaya, version 113), Arabidopsis (Arabidopsis thaliana, version 167), Eucalyptus (Eucalyptus grandis, version 201), grape (Vitis vinifera) and tomato (Solanum lycopersicum) were downloaded from Phytozome .
Cbr1 includes 20 genetic map markers, and 16 have successful corresponding overgo probes on the physical map. All 16 hybridize to one or more of the following four physical map contigs: ctg3279, ctg4269, ctg9166, ctg11956. A sequencing pool of 50 minimum tiling path (MTP) BACs was constructed representing ctg4269 and ctg9166 (pool C). A second sequencing pool of 49 MTP BACs from ctg11956 and 3279 was created (pool D).
For Cbr2, seven of eight genetic map markers have a successful overgo probe. All of the probes except one corresponded to physical map ctg403. A MTP of 51 BACs was picked from this contig and pooled for sequencing (pool B).
For Cbr3, alleles conferring resistance inherited from Chinese chestnut and LOD ±1.0 support intervals were localized to the region within 35.7–39.5 cM on LG-E of the Chinese chestnut consensus genetic map . In total 26 genetic markers mapped into this interval were used for developing overgo probes for hybridizations. Of these, 18 probes were successfully hybridized. In total 9 physical contigs were pulled out for further consideration. The most associated markers CmSNP00127 and CmSNP01226 mapped to 37.4 cM on the consensus map were localized in physical map contig 7039. Forty BACs spanning contig 7039 were selected and pooled for sequencing (pool A). The genetic map markers in the QTL regions, their originating EST sequence names, and their physical map contig anchor points are in Additional file 5.
The minimum tiling path of each of the 6 physical map contigs was selected with FPC software . After being picked, the clones were fingerprinted and compared to the original clone fingerprints used during map construction. The fingerprinting procedure is described in . Four of the clones were found to be contaminated or had different fingerprints and were discarded. Six new clones that spanned the same areas of the physical map were selected to replace those four. Four total pools of DNA were created for sequencing. BAC addresses included in each pool can be found in Additional file 6.
Each of the four pools was prepared as a single end library and sequenced with ½ slide of 454 single end pyrosequencing on a Roche 454 GS FLX. Reads were screened for E. coli and the BAC vector pIndigoBAC5 using cross_match . Additional sequencing was performed on a MiSeq instrument (Illumina); paired end libraries (2x250) were prepared from each DNA pool and run twice. The reads were trimmed with Trimmomatic . The 454 data and the first run from the MiSeq were assembled with gsAssembler (Roche) using an overlap length of 30 and heterozygote mode. Using the second MiSeq data set, the contigs from the assembler were placed in scaffolds with the software SSPACE .
The resulting genomic sequence scaffolds were annotated for repeats and genes with the Maker pipeline . Additional annotation was performed with Augustus , PASA , Gmap , and RepeatMasker  using RepBase . Annotations from all software packages are available to be viewed on the hardwood genomics website  as tracks in Jbrowse . The final gene calls from Maker are used as the final gene annotations for analysis; Blast2GO  was used to functionally annotate these genes.
Using CoGe’s SynMap tool , the QTL sequences were aligned to ten reference genome species. The ten genomes downloaded from Phytozome were: peach (Prunus persica), strawberry (Fragaria vesca), soybean (Glycine max), Medicago (Medicago truncatula), poplar (Populus trichocarpa), papaya (Carica papaya), Arabidopsis (Arabidopsis thaliana), Eucalyptus (Eucalyptus grandis), grape (Vitis vinifera) and tomato (Solanum lycopersicum). The DAGChainer output from SynMap was filtered to retain only the alignments scoring over 300.
The data sets supporting the results of this article are available in the GenBank repository, (accessions HN270092-HN275251 and JY172573-18703 at http://www.ncbi.nlm.nih.gov/genbank/) and in NCBI’s short reach archive (accessions SRR1646589-SRR1646592 and SRR1646621-SRR1646628 at http://www.ncbi.nlm.nih.gov/sra).
The authors acknowledge with thanks the Clemson University Genomics Institute (CUGI), whose laboratory facilities were used for the BAC end sequencing, overgo hybridizations and BAC DNA preparation.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Watson L, Dallwitz MJ. The Families of Flowering Plants: Descriptions. Illustrations: Identification, and Information Retrieval. University of New Orleans; 1999.Google Scholar
- Crandall BS, Gravatt GF, Ryan MM. Root disease of Castanea species and some coniferous and broadleaf nursery stocks, caused by Phytophthora cinnamomi. Phytopathology. 1945;35:162–80.Google Scholar
- Global invasive species database [http://www.issg.org/database/welcome/]
- Rizzo DM, Garbelotto M. Sudden oak death: endangering California and Oregon forest ecosystems. Front Ecol Environ. 2003;1:197–204.View ArticleGoogle Scholar
- Anagnostakis SL. Chestnut blight: the classical problem of an introduced pathogen. Mycologia. 1987;23–37.Google Scholar
- Hebard FV. The backcross breeding program of the American chestnut foundation. J Am Chestnut Found. 2006;19:55–77.Google Scholar
- Barakat A, DiLoreto D, Zhang Y, Smith C, Baier K, Powell W, et al. Comparison of the transcriptomes of American chestnut (Castanea dentata) and Chinese chestnut (Castanea mollissima) in response to the chestnut blight infection. BMC Plant Biol. 2009;9:51.PubMed CentralView ArticlePubMedGoogle Scholar
- Kubisiak TL, Nelson CD, Staton ME, Zhebentyayeva T, Smith C, Olukolu BA, et al. A transcriptome-based genetic map of Chinese chestnut (Castanea mollissima) and identification of regions of segmental homology with peach (Prunus persica). Tree Genet Genomes. 2013;1–15.Google Scholar
- Fang G-C, Blackmon BP, Staton ME, Nelson CD, Kubisiak TL, Olukolu BA, et al. A physical map of the Chinese chestnut (Castanea mollissima) genome and its integration with the genetic map. Tree Genet Genomes. 2012;1–13.Google Scholar
- Chinese chestnut physical map [http://www.hardwoodgenomics.org/content/chinese-chestnut-physical-map]
- Kremer A, Casasoli M, Barreneche T, Bodénès C, Sisco P, Kubisiak T, et al. Fagaceae trees. In Forest Trees Springer; 2007;161–187.Google Scholar
- Angiosperm Phylogeny Website, Version 13. http://www.mobot.org/MOBOT/research/APweb/.
- Tuskan GA, DiFazio S, Jansson S, Bohlmann J, Grigoriev I, Hellsten U, et al. The Genome of Black Cottonwood, Populus trichocarpa (Torr. & Gray). Science. 2006;313:1596–604.View ArticlePubMedGoogle Scholar
- Kubisiak TL, Hebard FV, Nelson CD, Zhang J, Bernatzky R, Huang H, et al. Molecular mapping of resistance to blight in an interspecific cross in the genus Castanea. Phytopathology. 1997;87:751–9.View ArticlePubMedGoogle Scholar
- Foulongne M, Pascal T, Pfeiffer F, Kervella J. QTLs for powdery mildew resistance in peach x Prunus davidiana crosses: consistency across generations and environments. Mol Breed. 2003;12:33–50.View ArticleGoogle Scholar
- Anagnostakis SL. The effect of multiple importations of pests and pathogens on a native tree. Biol Invasions. 2001;3:245–54.View ArticleGoogle Scholar
- Schmutz J, Cannon SB, Schlueter J, Ma J, Mitros T, Nelson W, et al. Genome sequence of the palaeopolyploid soybean. Nature. 2010;463:178–83.View ArticlePubMedGoogle Scholar
- Myburg A, Grattapaglia D, Tuskan G, Jenkins J, Schmutz J, Mizrachi E, et al. The Eucalyptus grandis Genome Project: Genome and transcriptome resources for comparative analysis of woody plant biology. BMC Proceedings. 2011;5:I20. BioMed Central Ltd.PubMed CentralView ArticleGoogle Scholar
- Sato S, Tabata S, Hirakawa H, Asamizu E, Shirasawa K, Isobe S, et al. The tomato genome sequence provides insights into fleshy fruit evolution. Nature. 2012;485:635–41.View ArticleGoogle Scholar
- Jaillon O, Aury J-M, Noel B, Policriti A, Clepet C, Casagrande A, et al. The grapevine genome sequence suggests ancestral hexaploidization in major angiosperm phyla. Nature. 2007;449:463–7.View ArticlePubMedGoogle Scholar
- Cronquist A. An integrated system of classification of flowering plants. Columbia University Press; New York, NY: 1981.Google Scholar
- Soltis DE, Gitzendanner MA, Soltis PS. A 567-taxon data set for angiosperms: the challenges posed by Bayesian analyses of large data sets. Int J Plant Sci. 2007;168:137–57.View ArticleGoogle Scholar
- Wang H, Moore MJ, Soltis PS, Bell CD, Brockington SF, Alexandre R, et al. Rosid radiation and the rapid rise of angiosperm-dominated forests. Proc Natl Acad Sci. 2009;106:3853–8.PubMed CentralView ArticlePubMedGoogle Scholar
- Hu TT, Pattyn P, Bakker EG, Cao J, Cheng J-F, Clark RM, et al. The Arabidopsis lyrata genome sequence and the basis of rapid genome size change. Nat Genet. 2011;43:476–81.PubMed CentralView ArticlePubMedGoogle Scholar
- Gordon D, Abajian C, Green P. Consed: a graphical tool for sequence finishing. Genome Res. 1998;8:195–202.View ArticlePubMedGoogle Scholar
- Chou H-H, Holmes MH. DNA sequence quality trimming and vector removal. Bioinformatics. 2001;17:1093–104.View ArticlePubMedGoogle Scholar
- Soderlund C, Bomhoff M, Nelson WM. SyMAP v3. 4: a turnkey synteny system with application to plant genomes. Nucleic Acids Res. 2011:gkr123.Google Scholar
- Goodstein DM, Shu S, Howson R, Neupane R, Hayes RD, Fazo J, et al. Phytozome: a comparative platform for green plant genomics. Nucleic Acids Res. 2012;40:D1178–86.PubMed CentralView ArticlePubMedGoogle Scholar
- Soderlund C, Longden I, Mott R. FPC: a system for building contigs from restriction fingerprinted clones. Comput Appl Biosci CABIOS. 1997;13:523–35.PubMedGoogle Scholar
- Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014:btu170.Google Scholar
- Boetzer M, Henkel CV, Jansen HJ, Butler D, Pirovano W. Scaffolding pre-assembled contigs using SSPACE. Bioinformatics. 2011;27:578–9.View ArticlePubMedGoogle Scholar
- Holt C, Yandell M. MAKER2: an annotation pipeline and genome-database management tool for second-generation genome projects. BMC Bioinformatics. 2011;12:491.PubMed CentralView ArticlePubMedGoogle Scholar
- Stanke M, Steinkamp R, Waack S, Morgenstern B. AUGUSTUS: a web server for gene finding in eukaryotes. Nucleic Acids Res. 2004;32 suppl 2:W309–12.PubMed CentralView ArticlePubMedGoogle Scholar
- Haas BJ, Delcher AL, Mount SM, Wortman JR, Smith Jr RK, Hannick LI, et al. Improving the Arabidopsis genome annotation using maximal transcript alignment assemblies. Nucleic Acids Res. 2003;31:5654–66.PubMed CentralView ArticlePubMedGoogle Scholar
- Wu TD, Watanabe CK. GMAP: a genomic mapping and alignment program for mRNA and EST sequences. Bioinformatics. 2005;21:1859–75.View ArticlePubMedGoogle Scholar
- Smit AF, Hubley R, Green P. RepeatMasker Open-3.0. 1996.Google Scholar
- Jurka J, Kapitonov VV, Pavlicek A, Klonowski P, Kohany O, Walichiewicz J. Repbase update, a database of eukaryotic repetitive elements. Cytogenet Genome Res. 2005;110:462–7.View ArticlePubMedGoogle Scholar
- Hardwood Genomics Web [hardwoodgenomics.org]
- Skinner ME, Uzilov AV, Stein LD, Mungall CJ, Holmes IH. JBrowse: a next-generation genome browser. Genome Res. 2009;19:1630–8.PubMed CentralView ArticlePubMedGoogle Scholar
- Conesa A, Götz S, García-Gómez JM, Terol J, Talón M, Robles M. Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics. 2005;21:3674–6.View ArticlePubMedGoogle Scholar
- Lyons E, Pedersen B, Kane J, Freeling M. The value of nonmodel genomes and an example using SynMap within CoGe to dissect the hexaploidy that predates the rosids. Trop Plant Biol. 2008;1:181–90.View ArticleGoogle Scholar
- Consortium TPGS. Genome sequence and analysis of the tuber crop potato. Nature. 2011;475:189–95.View ArticleGoogle Scholar
- Bowers JE, Chapman BA, Rong J, Paterson AH. Unravelling angiosperm genome evolution by phylogenetic analysis of chromosomal duplication events. Nature. 2003;422:433–8.View ArticlePubMedGoogle Scholar
- Ming R, Hou S, Feng Y, Yu Q, Dionne-Laporte A, Saw JH, et al. The draft genome of the transgenic tropical fruit tree papaya (Carica papaya Linnaeus). Nature. 2008;452:991–6.PubMed CentralView ArticlePubMedGoogle Scholar
- Soltis DE, Albert VA, Leebens-Mack J, Bell CD, Paterson AH, Zheng C, et al. Polyploidy and angiosperm diversification. Am J Bot. 2009;96:336–48.View ArticlePubMedGoogle Scholar
- Young ND, Debellé F, Oldroyd GED, Geurts R, Cannon SB, Udvardi MK, et al. The Medicago genome provides insight into the evolution of rhizobial symbioses. Nature. 2011;480:520–4.PubMed CentralView ArticlePubMedGoogle Scholar