|Home | About | Journals | Submit | Contact Us | Français|
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
In view of the immense value of Brassica rapa in the fields of agriculture and molecular biology, the multinational Brassica rapa Genome Sequencing Project (BrGSP) was launched in 2003 by five countries. The developing BrGSP has valuable resources for the community, including a reference genetic map and seed BAC sequences. Although the initial B. rapa linkage map served as a reference for the BrGSP, there was ambiguity in reconciling the linkage groups with the ten chromosomes of B. rapa. Consequently, the BrGSP assigned each of the linkage groups to the project members as chromosome substitutes for sequencing.
We identified simple sequence repeat (SSR) motifs in the B. rapa genome with the sequences of seed BACs used for the BrGSP. By testing 749 amplicons containing SSR motifs, we identified polymorphisms that enabled the anchoring of 188 BACs onto the B. rapa reference linkage map consisting of 719 loci in the 10 linkage groups with an average distance of 1.6 cM between adjacent loci. The anchored BAC sequences enabled the identification of 30 blocks of conserved synteny, totaling 534.9 cM in length, between the genomes of B. rapa and Arabidopsis thaliana. Most of these were consistent with previously reported duplication and rearrangement events that differentiate these genomes. However, we were able to identify the collinear regions for seven additional previously uncharacterized sections of the A genome. Integration of the linkage map with the B. rapa cytogenetic map was accomplished by FISH with probes representing 20 BAC clones, along with probes for rDNA and centromeric repeat sequences. This integration enabled unambiguous alignment and orientation of the maps representing the 10 B. rapa chromosomes.
We developed a second generation reference linkage map for B. rapa, which was aligned unambiguously to the B. rapa cytogenetic map. Furthermore, using our data, we confirmed and extended the comparative genome analysis between B. rapa and A. thaliana. This work will serve as a basis for integrating the genetic, physical, and chromosome maps of the BrGSP, as well as for studies on polyploidization, speciation, and genome duplication in the genus Brassica.
Brassica is a model system for studying polyploidization and speciation since all the species in this genus have descended from a common hexaploid ancestor [1-6]. In addition, Brassica species share an ancestor with Arabidopsis, implying a similar basic genome, and thereby providing sequence-level colinearity between the two genera, particularly in euchromatic regions [7-9]. This relationship highlights the feasibility of utilizing the accumulated Arabidopsis information for the study of Brassica species. The genus Brassica includes economically important crop taxa with a wide range of morphologies, such as Chinese cabbage, mustard, cabbage, broccoli, oilseed rape, and other leafy vegetables. These taxa are classified into six genome types (AA, n = 10; BB, n = 8; CC, n = 9; AABB, n = 18; AACC, n = 19; BBCC, n = 17) according to the six representative species (AA, B. rapa; BB, B. nigra; CC, B. oleracea; AABB, B. juncea; AACC, B. napus; BBCC, B. carinata), and the genomic relationships of these taxa are well defined in U's triangle .
In view of the enomous value of Brassica in the fields of agriculture and molecular biology, genome sequencing projects have been proposed for each of the three diploid genomes . In order to better understand the A genome of Brassica and to take advantage of the colinearity between this genome and the Arabidopsis genome sequence, the multinational Brassica rapa Genome Sequencing Project (BrGSP) was launched in 2003 by scientists from five countries (Korea, Canada, the United Kingdom, China, and Australia) for sequencing B. rapa ssp. pekinensis cv. Chiifu-401-42 using a BAC-by-BAC approach . The initial objective of the BrGSP was to sequence the gene space of B. rapa, which represents approximately 330 Mb of its genome , at a Phase II quality level, whereby BACs would be sequenced to a single ordered and oriented contig, but with allowance for some gaps and ambiguous bases . The Korean group of the BrGSP [Korean Brassica Genome Project (KBGP)] sequenced 521 B. rapa BACs selected to represent genomic regions collinear with the majority of the euchromatic regions of the A. thaliana genome . These clones serve as "seed" BACs for the BrGSP, from which chromosome-scale sequencing is being initiated. The developing BrGSP has valuable resources for the community, including three BAC libraries [2,14], 200,017 BAC end sequences , 129,928 EST sequences (by April 2008), and an initial-version reference genetic map . To date, 631 BAC sequences, representing approximately 75.3 Mb, have been made public .
The initial reference linkage map of B. rapa was constructed with 556 markers (278 AFLPs; 235 SSRs; 25 RAPDs; and a total of 18 ESTPs, STSs, and CAPSs) based on 78 doubled haploid lines (CKDH line) derived from an anther culture of the F1 of a cross between diverse Chinese cabbage (B. rapa ssp. pekinensis) inbred lines; "Chiifu-401-42" (C) and "Kenshin-402-43" (K) . Ten linkage groups, designated as A1-A10 according to the common nomenclature of the B. napus reference linkage maps , served as a reference for the BrGSP. However, there remained ambiguity in reconciling the linkage groups with the 10 B. rapa chromosomes characterized using cytogenetic approaches.
The B. rapa chromosomes have been extensively studied by karyotyping based on morphometric measurements of mitotic metaphase chromosomes [18-21]. The definitive identification of each of the individual chromosomes has been problematic because some are small-sized or similar. Recently, using fluorescence in situ hybridization (FISH), six of the 10 chromosomes were distinguished unambiguously based on the chromosomal position of repetitive sequences, such as 45S rDNA, 25S rDNA, 5S rDNA, and centromeric repeats. However, this technique can be impractical in that multiple FISHs are required to distinguish these six chromosomes, and it is unable to distinguish the remaining four chromosomes [14,22-25]. Consequently, the BrGSP assigned each of the linkage groups (A1-A10) to the project members as chromosomal substitutes for sequencing .
In this study, we developed a second generation B. rapa reference linkage map, aligned unambiguously with the cytogenetic map of B. rapa. We also used our data to confirm and extend the comparative genome analysis of B. rapa and A. thaliana.
Of the 749 SSR motif-containing amplicons designed from 367 sequenced BACs, 311 (41.5%) showed polymorphism between Chiifu and Kenshin (Additional file 1). Many amplicons designed from Brassica sequences produce multiple bands as a consequence of the extensive genome duplication observed in these species. Anchorage of sequenced BAC clones to linkage maps therefore requires matching of the size of the PCR product from the BAC (derived from Chiifu genomic DNA) with the size of the polymorphic band amplified from genomic DNA (the Chiifu allele). This validation was conducted for all polymorphic markers. The polymorphic PCR product from the genomic DNA was found to match the BAC-derived product for 191 markers (61.4%), all of which were designed from 188 different BAC clones. There were several instances of the experimentally determined band being greater in size than was expected from the BAC sequences. We attribute these differences to sequencing errors. The 191 BAC-anchoring markers were added to the initial version of the B. rapa linkage map to produce the second generation linkage map (Fig. (Fig.1,1, Fig. Fig.2).2). This map now anchors 188 sequenced BAC clones to the B. rapa linkage map.
The linkage map contained 719 loci, comprising 267 AFLP, 411 SSR, 24 RAPD, 8 STS, 7 ESTP, and 2 CAPS markers, assigned to the 10 A genome linkage groups (A1-A10) of Brassica species following established nomenclature . The total length of the map was 1,123.3 cM with an average distance of 1.6 cM between adjacent loci (Table (Table1).1). The length of the linkage groups ranged from 91.3 cM (A10) to 138.5 cM (A06), and the number of markers in the 10 linkage groups ranged from 38 (A4) to 104 (A3). In total, 24.3% of the markers were mapped at the same loci or at less than 1-cM intervals. The average distance between adjacent loci was greatest for A2 (2.8 cM) and least for A7 (1.0 cM). Two gaps of greater than 10 cM were present (A2), measuring 13.5 cM and 17.7 cM, respectively.
In order to estimate the coverage of the linkage map with respect to the Arabidopsis genome, we aligned the sequences of the anchored BACs to the Arabidopsis genome sequence (Fig. (Fig.1,1, Fig. Fig.2).2). Of the 188 anchored BAC clones, 184 (representing 187 loci) could be aligned. The aligned regions represent 534.9 cM of the B. rapa linkage map and 60.5 Mb of the Arabidopsis genome sequence (Fig. (Fig.1,1, Fig. Fig.2,2, Table Table2,2, Additional file 2).
We detected 30 blocks of conserved synteny, as defined by two or more adjacent anchored B. rapa BACs aligned to the corresponding region of the Arabidopsis genome identified in the earlier maps for A genomes [27,28] (Fig. (Fig.1,1, Fig. Fig.2,2, Table Table3).3). If the extended positions were detected within our defined blocks of the B. rapa map (Table (Table3),3), they were considered extensions of the previously reported blocks (A-X). Blocks G, S, and H (in blocks 9, 25, and 28) were newly identified in the B. rapa linkage groups A3, A9, and A9, respectively, compared with the A genome map of B. napus [27,28]. Blocks G and H in the linkage groups of A3 and A9 were recognized in the A genome of B. juncea , whereas block S in A9 (in block 25) appeared to be unique to the B. rapa genome. The longest blocks were 59.9-98.7 cM on A5, aligned to 10.8 Mb of Arabidopsis chromosome 3, and 2.7-52.7 cM on A1, aligned to 6.2 Mb of Arabidopsis chromosome 4. On average (based on each of the 30 blocks of conserved synteny), we estimate that 1 cM of the B. rapa reference genetic map aligns to 341 kb of the Arabidopsis genome sequence. However, the large standard deviation (513 kb) indicates that this relationship varies greatly across the aligned genome segments. This alignment updated the initial version of the B. rapa linkage map .
Most of the alignments between the linkage map and the Arabidopsis genome sequence were consistent with collinearity blocks previously inferred using sequenced markers with homology to single Arabidopsis genes [27,28]. However, we identified the collinear regions for seven additional previously uncharacterized sections of the A genome by a single aligned BAC, and confirmed three of these by mapping a second marker. The number and ranges of the Arabidopsis gene models for which collinear homologous sequences could be identified were determined by using BAC annotations of the BrGSP , and these are shown in Table Table44.
Based on the 30 syntenic blocks, we detected 17 and 8 regions that were represented two and three times, respectively, within the B. rapa genome (Table (Table5).5). All redundantly represented regions were detected between different linkage groups, except one region represented three times, which was identified from two regions in A6 and one in A8 aligned to a 6.8-7.3 Mb region of Arabidopsis chromosome 1. The total sizes of the regions represented two and three times in the B. rapa genome were 18.9 Mb and 10.8 Mb, respectively, based on the sizes of the corresponding Arabidopsis regions. This result indicates that 29.7 Mb of the Arabidopsis genome corresponds to 70.2 Mb of the B. rapa genome. The detected redundant regions were smaller than 3.0 Mb (range: 0.1-3.0 Mb), indicating the scale of the maintained regions after hexapolyploidization. These genetically redundant regions were detected by the same cutoff value of sequence similarity, demonstrating that the regions had emerged simultaneously and undergone the same evolutionary events.
The B. rapa chromosomes are too small and compact to be distinguished by either morphological characteristics or the chromosome arm length ratio. In order to develop markers for complete karyotyping of these chromosomes, BAC clones genetically anchored on our reference genetic map were searched for repetitiveness by sequence similarities or FISH analysis. Consequently, 10 sets of nonrepetitive BAC clones (20 BACs) were selected to distinguish the 10 chromosomes and were fluorescence in situ hybridized on metaphase chromosomes (Table (Table6).6). The fluorescence signals of each set of BAC clones (red and green) were detected from each pair of 10 chromosomes (Additional file 3), confirming the utility of BACs in distinguishing the chromosomes. Six chromosomes hybridized by the selected BAC clones from A1, A3, A5, A6, A9, and A10 were recognized to be chromosomes 7, 2, 4, 5, 1, and 10, respectively, based on the following previously reported remarks: (1) the largest chromosome for chromosome 1, (2) the NOR-bearing chromosome for chromosome 2, (3) the 45S rDNA- and CentBr2-hybridized chromosome for chromosome 4, (4) the 45S rDNA-only-hybridized chromosome for chromosome 5, (5) the 45S rDNA- and 5S rDNA-hybridized chromosome for chromosome 7, and (6) the 5S rDNA-only-hybridized chromosome for chromosome 10 (Additional file 3). The four remaining chromosomes hybridized by BACs from A2, A4, A7, and A8 were assigned to chromosomes 3, 9, 6, and 8, respectively. This designation was based solely on the sizes of the four chromosomes, which were numbered from the largest to the smallest in accordance with chromosome morphology [24,31].
The 10 sets of BAC clones distinguishing each chromosome were defined as the definite karyotyping markers for the 10 chromosomes of B. rapa. In order to ensure the reliability of the markers, one BAC marker from each of the eight chromosome sets (chr 3-10; chromosomes 1 and 2 are obvious from their morphological remarks), was rehybridized onto one set of B. rapa chromosomes (Fig. (Fig.3).3). Consequently, all eight BACs were hybridized onto individual chromosome pairs, and chromosomes 1 and 2 were recognized by the morphological remarks (the largest chromosome for chromosome 1 and the NOR-bearing chromosome for chromosome 2), thus demonstrating the utility of the developed markers. The current orientations of four linkage groups (A1, A2, A3, A8) [16,27] are incorrect based on the orientation of chromosomes deduced from FISH signals (Fig. (Fig.3),3), and should be reversed.
A high density linkage map of B. rapa was constructed, resulting in an average marker density of one marker per 1.6 cM and the anchoring of 188 sequenced BAC clones. The SSR markers developed from these BAC sequences were validated by matching the sizes of the Chiifu alleles of the polymorphic band with the PCR product from the corresponding BAC clone. Approximately 57% of the markers on the map were of the SSR type, which are likely to be transferable to studies involving other Brassica species and populations. Moreover, the 188 sequenced BAC clones anchored on the reference genetic map will serve as a basis for integrating the physical map of B. rapa  with the reference genetic map.
Anchoring of the sequenced BACs onto the genetic map provided opportunities for aligning the reference map onto the Arabidopsis genome sequence on the basis of sequence similarity. Although Arabidopsis and B. rapa diverged approximately 20 million years ago , and underwent reshuffling during their respective evolutions, extensive collinearity was maintained between their genomes. On the basis of the 30 blocks of conserved synteny detected in this study, we illustrated possible contractions in the B. rapa genome. For example, in blocks 7 and 25, we estimated that 1 cM of the B. rapa map aligns with 1.2 Mb and 2.8 Mb of the Arabidopsis genome, respectively. In addition to the identified regions represented twice within the B. rapa genome (Fig. (Fig.1,1, Fig. Fig.2,2, Table Table3),3), the detected regions represented three times in the genome (Table (Table5)5) were similarly divergent in each case in that they were distinguished by the same level of sequence similarity. This supports the hypothesis that proposes the existence of a hexaploid ancestor in the evolution of Brassica species [1-6].
Most of the alignments between the linkage map and the Arabidopsis genome sequence are consistent with previously identified collinearity blocks [27,28]. These were initially inferred for B. napus using sequenced markers with homology to single Arabidopsis genes , and subsequently extended across a number of Brassicaceae . However, our alignments are considerably more robust, as they exploit sequences of whole BAC clones containing homologies to multiple collinear Arabidopsis gene models. Further, the alleles present in the BAC clones can be matched to the mapped bands in the marker assays. This advantage overcomes the "noise" encountered in comparative genomics as a consequence of homologies to transduplicated gene fragments  and enabled us to identify seven previously unreported collinearity blocks. Two of the newly discovered collinearity blocks, represented by BAC clones KBrS008C11/KBrB058B22 and KBrH014M07, correspond to parts of blocks G and H, respectively, of the proposed Brassicaceae ancestral karyotype linkage group AK3 . To date, these have been identified in only a single copy, and these newly discovered copies appear to represent one more of the three proposed paralogs of each that are expected to have arisen as a result of the proposed hexaploidy event in the ancestry of the Brassica species [1-6]. Block K, represented by KBrB072E02, also represents a third copy of the paralogs. Block M, not recognized in the A genome of B. napus , was identified. A newly discovered collinearity block represented by BAC clones KBrB080J22, KBrS012D09, and KBrH097M21 correspond to parts of blocks C, N, and S, respectively. These represent the fourth copies of the blocks to be identified, only three copies having being described by Schranz et al. . We interpret the discovery of these blocks as an indication that supernumerary segmental genome duplications, as described for the genomic regions containing the B. rapa orthologs of the Arabidopsis gene FLC , could be common. Five of these blocks (D, G, H, H, and N) were recognized in the A genome of B. juncea , supporting our new findings.
The twenty BAC clones used in this study were nonrepetitive and genetically anchored, thus allowing integration of the physical and genetic maps with each chromosome. As reported previously, comparative genetic mapping between Arabidopsis and B. rapa has revealed that collinear regions of the genomes are represented two or three times in B. rapa [5,34-36]. Jackson et al.  reported multiple duplications of the unique locus on Arabidopsis chromosome 2 (79 cM region) in the B. rapa genome by showing BAC FISH signals from four to six B. rapa chromosomes. In our study, some BACs were located at the loci represented twice (KBrB006A15, KBrB012O13, KBrB013N08, KBrB056I08, KBrB076B03, KBrH003E08, KBrH003N18, KBrH004D08) or three times (KBrH003F06, KBrH005C21, KBrB084H08) based on the Arabidopsis alignment data (Fig. (Fig.1,1, Fig. Fig.2,2, Table Table5).5). However, they were sufficiently divergent, showing unique hybridization and a single BAC FISH signal.
When the BrGSP was launched in 2003, the genome was assigned to participating countries by linkage groups, because at that time the chromosomes were not reconciled with the linkage groups. The previously reported karyotyping of B. rapa was based on chromosome length, and/or FISH patterns of repetitive DNAs [14,22-25]. From these studies, six chromosomes were distinguished unambiguously based on the chromosomal position of 45S rDNA, 25S rDNA, 5S rDNA, and centromeric repeats, whereas the remaining four chromosomes were ambiguous. Fukui et al. , Snowdon et al. , and Koo et al.  designated the recognized chromosomes to chromosomes 1, 2, 4, 5, 7, and 10 in association with chromosomes of the Brassica species, corresponding to chromosomes 1, 2, 5, 4, 3, and 10 of Lim et al. [14,25], respectively. These discrepancies in assigning chromosome numbers are due to the specific characteristics of Brassica chromosomes, which are small and compact, causing different decisions on chromosome length order in the different studies, particularly with metaphase chromosomes. As described earlier, matching chromosomes in different studies in terms of chromosome length is sometimes difficult, and thus emphasizes the importance of the karyotype markers used in this study, which standardize the karyotype. Our chromosome nomenclature is consistent with earlier reports by Fukui et al. , and these have been reinforced by Snowdon et al.  and Koo et al. .
FISH has been used not only for chromosome identification  but also for merging genetic, physical, and chromosomal maps [38-42], thereby providing information about genome organization. The first integration of the cytogenetic and genetic linkage maps in Brassica species was achieved in CC genome species by FISH using 22 probes representing 19 loci of nine chromosomes . In the present study, we assigned all 10 linkage groups of the B. rapa reference genetic map to each of the 10 chromosomes and corrected the orientation of the linkage groups by FISH. The 20 cytologically mapped loci will serve as the basis for integration of the genetic, physical, and chromosomal maps of B. rapa in the multinational BrGSP.
We constructed a second generation reference linkage map of B. rapa, which has an average marker density of 1 marker per 1.6 cM and to which 188 sequenced BAC clones are anchored. Anchoring of the sequenced BACs onto the reference genetic map provided opportunities for aligning the map onto the Arabidopsis genome sequence on the basis of sequence similarity. We detected 30 blocks of conserved synteny between the B. rapa and Arabidopsis genomes, illustrating rearrangement events with a trace of hexapolyploidy differentiating these genomes. Most of these were consistent with previously reported collinear blocks; however, we were able to identify seven regions as individual BAC clones or a pair of overlapping BAC clones, representing previously unreported collinearity blocks. One of these represents a previously "missing" block under the hypothesis of whole genome triplication, and three of the remaining blocks represent a supernumerary segment under the same hypothesis.
Nonrepetitive and genetically anchored BAC clones allowed integration of the genetic map with the cytogenetic map by developing definite karyotype markers. This is the first unambiguous alignment of the linkage map with the 10 B. rapa chromosomes. We envision that the genetic map and analysis presented here will serve as a basis for integrating the genetic, physical, and chromosomal maps of B. rapa in the multinational BrGSP as well as for studies on polyploidization and speciation in the genus Brassica.
Five hundred and twenty one sequenced BACs of B. rapa ssp. pekinensis generated by the KBGP  were collected from GenBank . SSRs were identified from the annotated BACs with SPUTNIK  using the parameters described previously by Hong et al. : (i) SSRs were determined to be positive if the repeats were ≥ 12 bp for mononucleotide, dinucleotide, and trinucleotide repeats, ≥ 16 bp for tetranucleotide repeats, and ≥ 20 bp for pentanucleotide repeats; and (ii) no variations (mutations) in repeat motifs were permitted
SSR loci of ≥ 18 bp, as described in the preceding, were chosen for marker development. One to three SSR primer sets per BAC were designed using Primer3 software  from the flanking sequences of the targeted SSR loci. The primer design criteria were as follows: 100-400 bp of amplicon size, 55-63°C of Tm, >35% of GC contents, and >18 bp of primer length. Totally of 749 primer sets were designed from 367 BAC sequences (Additional file 1). To develop SSR markers, polymorphisms between the parental lines (Chiifu-401-42 and Kenshin-402-43)  were evaluated by PCR amplification of the parental genomic DNA using the designed primer sets. PCR products were separated on 6% polyacrylamide gels and visualized by using a silver staining kit (Bioneer, Daejeon, Korea). Allele matching was performed by analysis of PCR products from genomic DNA of Chiifu-401-42 and the BAC clone from which primers were designed. Mapping was conducted only for markers where the size of the polymorphic band corresponding to the Chiifu allele matched the size of the PCR product from the BAC.
PCR amplification was carried out in 10 μl volumes, containing 0.5 units of Taq polymerase (Bioneer, Daejeon, Korea), 5 pmol of each primer, 250 μM dNTPs, 2.0 mM MgCl2, 1× Taq buffer, and 10 ng of genomic template DNA. The PCR profile was as follows: 5 min at 95°C, followed by 30-35 cycles with 30 s of DNA denaturation at 94°C, 30 s of annealing at the appropriate temperature, and 60 s of extension at 72°C, and final extension at 72°C for 7 min. PCR was carried out in a Bioneer thermal-cycler (Bioneer, Daejeon, Korea). We followed the locus nomenclature form used in previous reports by Choi et al.  which is based on the De Vicente et al.  format. The institute codes refer to the laboratory where each primer was screened; those used in this study were 'cnu' for Chungnam National University and 'nia' for the National Institute of Agricultural Biotechnology.
Previously reported plant materials, including two parental inbred lines and a 78-line DH population (CKDH) , were used for genetic mapping. Plant genomic DNA was isolated from fresh leaf material according to the procedure used by Guillemaut and Maréchal-Drouard .
Markers that were reproducibly polymorphic between parent lines were genotyped in the DH population. Linkage analysis and map construction were performed using JoinMap version 4.0 [48,49]. Linked loci were grouped in the LOD grouping threshold range of 3.8-5.0, and linkage groups were assigned as A1 to A10, corresponding to the previously reported initial version map  of this species. Locus order within the LOD grouping was generated for each linkage group using a recombination frequency below 0.45 and an LOD score above 0.5 for all marker pairs within each linkage group. A "ripple" procedure was performed after the addition of each marker and the "jump" thresholds were set to 5. Recombination frequencies were converted to centiMorgans (cM) with Kosambi's method for map-distance calculation .
All 188 sequenced seed BACs  anchored to the genetic linkage groups were used for the genome alignment. Repeat sequences in both the B. rapa BACs and the A. thaliana genomic sequences  were masked by RepeatMasker  and cross-matched  with our repeat database collected from the TIGR plant repeat database  and Repbase . Each of the B. rapa BACs were aligned to the A. thaliana genome sequence using BLASTZ  with parameters of [B = 0, C = 2, K = 2200]. The most highly conserved regions (HCRs) for the alignments between the species were identified by the following criteria: (i) total length of the aligned regions in HCRs between the species should be ≥ 5 kb and (ii) the number of aligned regions in HCRs should be more than three. To reinforce the method for confirming HCRs between the two species, B. rapa genes in the BAC sequences predicted by GenScan  and were compared with A. thaliana genes by BLAST X  with a cutoff value of 1e-10.
Seeds from B. rapa ssp. pekinensis cv. Chiifu-401-42 were germinated on moist filter paper in petridishes at 25°C for 48 hr. Root tips were sampled and prepared for FISH as previously described by Koo et al.  with the following modifications: root tips were excised and incubated in 2% cellulase (Sigma, St. Louis, USA), 1.5% macerozyme (Sigma, St. Louis, USA), 0.3% pectolyase (Sigma, St. Louis, USA), and 1 mM EDTA (pH 4.2).
Two BACs from each of the 10 linkage groups (Table (Table6)6) were selected as FISH probes to anchor the linkage groups to the chromosomes as well as to distinguish individual chromosomes of B. rapa ssp. pekinensis cv. Chiifu-401-42. The BAC clone, KBrH077I01, containing 176 bp of centromeric tandem repeats, was used as the CentBr2 probe (unpublished data). BAC DNA was extracted by a modified alkaline lysis method  with the following changes: BAC clones were cultured for 18 hr in 3 ml of LB media containing 12.5 μg/ml of chloramphenicol and harvested by centrifugation followed by decanting of the media. Cell pellets were resuspended in 200 μl of solution I (50 mM Tris·HCl, 10 mM EDTA [pH 8.0], 100 μg/ml RNase A) followed by shaking using a vortex. Lysis and neutralization were achieved by adding 200 μl of solution II (0.2 M NaOH, 1% SDS) and 200 μl of solution III (3.0 M potassium acetate [pH 5.5]) into the sample plates without a lysis incubation time. The lysates were then cleared and precipitated, followed by 70% ethanol washing. Each precipitated DNA pellet was re-suspended in 25 μl of 1 mM Tris (pH 8.0). BAC DNAs were labeled with biotin-16-dUTP using nick translation kits (Roche, Basel, Switzerland). The reaction was carried out at 15°C for 90 min followed by reaction blocking by the addition of 2 μl of 0.5 M EDTA. The reaction products were purified by ethanol precipitation.
The 45S and 5S rDNAs were amplified from the total genomic DNA of B. rapa by PCR using the following primer sets: 5'-TACCTGGTTGATCCTGCCAG-3' (forward) and 5'-TTGTCACTACCTCCCCGTGT-3' (reverse) for 45S rDNA ; 5'-GATCCCATCAGAACTTC-3' (forward) and 5'-GGTGCTTTAGTGCTGGTAT-3' (reverse) for 5S rDNA . The PCR cycling reaction was carried out in 100 μl volumes containing 2.5 units of Taq polymerase (Takara, Kyoto, Japan), 5 pmol of each primer, 250 μM dNTPs, 1× PCR buffer, and 10 ng of genomic template DNA. The amplifying PCR cycle was as follows: 35 cycles with 1 min of DNA denaturation at 94°C, 1 min of annealing at 55°C, and 1 min of extension at 72°C, followed by a 10 min final extension at 72°C. The amplified products were purified using a QIAquick gel extraction kit (Qiagen, Hilden, Germany). Labeling of the 45S and 5S rDNA was performed by a PCR cycling reaction in a 100 μl volume containing 2.5 units of Taq polymerase (Takara, Kyoto, Japan), 5 pmol of primer, 200 μM dATP, dCTP, dGTP, 140 μM dTTP, 60 μM digoxigenin-dUTP (Roche, Basel, Switzerland), 1× PCR buffer, and 10 ng of template rDNA. The labeling PCR cycle was identical to the amplifying cycle described previously.
FISH was carried out as described by Koo et al.  using 50 ng of labeled probes (BAC DNA, 45S rDNA, 5S rDNA, and CentBr) per slide. Briefly, chromosomal DNA on the slides was denatured with 70% formamide at 70°C for 2 min, followed by dehydration in 70, 85, 95, and 100% ethanol at -20°C for 3 min each. The probe mixture, containing 50% formamide (v/v), 10% dextran sulfate (w/v), 5 ng/μl salmon sperm DNA, and 500 ng/μl of labeled probe DNA, was denatured at 90°C for 10 min and kept on ice for 5 min. A 20 μl volume of the probe mixture was applied to the denatured chromosomal DNA and covered with a glass coverslip. Slides were then hybridized in a humid chamber at 37°C for 18 hr followed by standard washing (50% formamide for 10 min at 42°C, 2× SSC for 5 min at 42°C, and 4 × SSC/0.2% Tween-20 for 5 min at 42°C) and blocking (5% BSA/4 × SSC/0.2% Tween-20 for 10 min at RT). Probes were detected with avidin-FITC and anti-digoxigenin Cy3 (Roche, Basel, Switzerland). Chromosomes were counterstained using 1 μg/ml of DAPI (Vector Lab, Burlingame, USA). The signals were detected with a Cooled CCD Camera, CoolSNAP (Photometrics, Tucson, USA), and images were processed with software (Meta Imaging Series, version 4.6; Molecular Devices, Downingtown, USA) using a Leica epi-fluorescence microscope equipped with FITC-DAPI two-way or FITC-Rhodamine-DAPI three-way filter sets (Leica, Wetzlar, Gemany). The final printed images were prepared with Adobe Photoshop, version 8.0.
Reprobing wash was performed with 4 × SSC/0.2% Tween-20 for 30 min at 37°C, followed by dehydration in 70, 85, 95, and 100% ethanol at -20°C for 2 min each.
HRK designed the study, carried out the marker development, participated in the FISH analysis, analyzed and interpreted all data, and drafted the manuscript. SRC carried out genetic mapping, participated in the comparative anlaysis, and drafted the manuscript. JB performed the FISH experiment and participated in drafting the manuscript. CPH carried out the SSR analysis, participated in the comparative analysis, and drafted the manuscript. SYL, MJH, DVN, and MJ participated in the marker survey and genotyping. BSP and JWB participated in the design of the study and helped to draft the manuscript. IB participated in the synteny analysis and interpretation of alignment data, and drafted the manuscript. YPL conceived the study, participated in its coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.
Details of the 367 BAC sequence derived SSR primers including 191 newly mapped SSR markers.
Details of the alignment between the Brassica rapa reference genetic map and the Arabidopsis genome sequence.
Fluorescence in situ hybridization (FISH) mapping of 45S rDNA, 5S rDNA, CentBr2, and karyotyping markers (BACs) on the mitotic metaphase chromosomes of Brassica rapa. White arrows indicate the pair of chromosomes hybridized by each set of karyotyping markers (BACs). (O) and (X) indicate 'hybridized' and 'not hybridized' respectively, on the chromosome pair recognized by each set of karyotyping markers. Scale bar = 5 μm.
We thank the Korean Brassica Genome Project (KBGP) for providing the 521 seed BAC sequences to the community, and Jing Yue Cai and Da Un Han for technical support. This work was supported by grants from the Technology Development Program for Agricultural and Forestry Ministry of Agriculture, Forestry and Fisheries (grant no. 607002-05), and the Rural Development Administration (BioGreen 21 Program, grant no. 04-1-12-2), Republic of Korea.