Construction of a BAC library and mapping BAC clones to the linkage map of Barramundi, Lates calcarifer

Background Barramundi (Lates calcarifer) is an important farmed marine food fish species. Its first generation linkage map has been applied to map QTL for growth traits. To identify genes located in QTL responsible for specific traits, genomic large insert libraries are of crucial importance. We reported herein a bacterial artificial chromosome (BAC) library and the mapping of BAC clones to the linkage map. Results This BAC library consisted of 49,152 clones with an average insert size of 98 kb, representing 6.9-fold haploid genome coverage. Screening the library with 24 microsatellites and 15 ESTs/genes demonstrated that the library had good genome coverage. In addition, 62 novel microsatellites each isolated from 62 BAC clones were mapped onto the first generation linkage map. A total of 86 BAC clones were anchored on the linkage map with at least one BAC clone on each linkage group. Conclusion We have constructed the first BAC library for L. calcarifer and mapped 86 BAC clones to the first generation linkage map. This BAC library and the improved linkage map with 302 DNA markers not only supply an indispensable tool to the integration of physical and linkage maps, the fine mapping of QTL and map based cloning genes located in QTL of commercial importance, but also contribute to comparative genomic studies and eventually whole genome sequencing.


Background
Barramundi (Lates calcarifer), also called Asian seabass or the giant sea perch, belonging to the family Latidae is widely distributed in the coast and freshwater of the tropical Indo-west Pacific, from the Persian Gulf to India and Northern Australia [1,2]. Because of good meat quality and relatively high market value of L. calcarifer, it has become an attractive commodity of both large and small-scale aquaculture enterprises. It is commercially cultivated in Thailand, Malaysia, Singapore, Indonesia, Hong Kong, China and Australia in both brackishwater and freshwater ponds, as well as in cages in coastal waters. The global annual production of L. calcarifer was 400,000 MT according to FAO statistics. However, detailed breeding programs for genetic improvement of Asian seabass are still quite rare [3]. Identification of genomic regions and genes responsible for economically important traits could facilitate genetic improvement through marker-assisted selection [4], which is of importance for future aquaculture of L. calcarifer.
Linkage and physical maps are indispensable tools needed to identify genomic regions responsible for traits of interest. The genome of L. calcarifer is very compact (only 700 Mb) consisting of 24 chromosome pairs [5]. The first linkage map for L. calcarifer containing 240 microsatellite markers and genes on 24 linkage groups [6] was applied to mapping QTL for growth traits [7]. Libraries with large genomic DNA inserts are essential for physical mapping and positional cloning, particularly for higher eukaryotes [8]. The BAC (bacterial artificial chromosome) cloning system has become an invaluable tool in genomics studies because of its ability to stably maintain large DNA fragments and its ease of manipulation [9]. Genomic inserts in BAC clones have been shown to be very stable in E. coli and thus serve as ideal templates in generating whole-genome physical maps by DNA fingerprinting, developing sequence-tagged connectors and shotgun sequencing [10][11][12]. These features make the BAC cloning system a popular choice for high-throughput genomics studies [13]. BAC libraries have been developed for many economically important animal species such as cattle [14], pig [15], and sheep [16] and the highly endangered giant panda [17]. Only currently, BAC libraries were produced for some commercially important fish species such as salmon [18], catfish [12], rainbow trout, carp and tilapia [19].
Here, we describe the construction and characterization of a BAC library covering 6.9 times L. calcarifer haploid genome and mapping of 86 BAC clones to the linkage map. The BAC library and the improved linkage map of L. calcarifer will facilitate the integration of physical and linkage maps, fine mapping of QTL and identification of genes located in QTL of interest, maker-assisted selection and genome research.

Library construction
A BAC library of L. calcarifer was constructed using the HindIII cloning site in commercially prepared pCC1BAC vector (Epicentre, MD, USA). The BAC library consisted of a total of 49,152 clones, which were manually arrayed into 128 384-well plates.

Insert size distribution
To examine the quality of the BAC library, the sizes of 212 BAC clones randomly picked from the library were determined. All the 212 clones contained inserts. The insert size distribution of these 212 clones is shown in Figure 1 and 2. The average insert size was 98 kb, ranging from 45 to 200 kb. The insert size of over 80% of the BAC clones in this library was larger than 80 kb, and the insert size of 50% clones was smaller than 100 kb. This BAC library provides 6.9 time haploid genome equivalent based on a genome size of 700 Mb [6].

BAC library screening
To further assess the quality of this BAC library, 24 microsatellite markers randomly selected from each of the 24 linkage groups were used for hierarchical screening. PCRscreenings with the 24 microsatellite markers resulted in the number of positive clones varying from 3 to 15 with an average of 6.6 (  Figure 3). PCR screening with these 15 ESTs/ genes (PVALB-1, 5-HT, PROL-A, 14KDA-AP, AMY-A, MX, AP, LECT2, LYSO-G, IGF-1, TUB1A, TUB2B, GT7, CYP19A2 and AFPII) revealed that the number of positive BAC clones varied from 3 to 14 with an average of 7.3 ( Table 2). The actual average number of positive BAC clones detected by screening with microsatellites and ESTs/genes was near to theoretically calculated number of genome coverage (i.e. 6.9 time coverage of the haploid genome). At least 3 positive BAC clones for each of the ESTs/genes were identified from the library.

Microsatellite isolation from BAC clones and linkage mapping
In order to map BAC clones to the linkage map for future integrating linkage and physical maps, we isolated microsatellites from 300 BAC clones through enrichment of microsatellites. 864 clones collected from the microsatellite-enriched library were sequenced in both directions. 451 clones contained microsatellites (CA > 7 or GA > 7), yielding 229 unique sequences containing microsatellites. Among the 229 sequences, 218 had enough flanking regions for primer design. Primers were designed for 218 microsatellites, among which 63 within 63 BAC clones were informative in the mapping panel used for linkage mapping. 62 were mapped to 24 LG (Figure 4, 5, 6 and 7) while one marker LcaB044 remained unlinked, making the total number of mapped markers on the L. calcarifer linkage map to be 302. At least one BAC based marker was mapped on each LG. Thus, together with 24 microsatellites located in different BAC clones and used for screening the BAC library, a total of 86 BAC based markers have been anchored on the linkage map with at least one on each LG. The male maps of LG14 and LG18, which were split to two LG on the first generation map respectively [6] due to the limited number of markers, were merged to one LG respectively by integrating new microsatellites located in single BAC clones ( Figure 6). Details about primer sequences, GenBank accession number, annealing temperature for PCR, PCR product size and locations of the 63 markers located in different BAC clones are summarized in Table 3.

Discussion
A critical tool for genomic studies in fish species is the availability of deep-coverage large-insert genomic libraries, such as BAC libraries that can be used for physical mapping, integration of linkage and physical maps, positional cloning, comparative genomic studies and genome sequencing [13]. We constructed the first BAC library for L. calcarifer containing 49,152 clones with an average insert size of 98 kb ranging from 45 to 200 kb, indicating that this BAC library provided 6.9 × coverage of the L. calcarifer haploid genome. We have noticed that 50% of the inserts in our BAC library were under 100 kb. It is common that insert size of 50% of BAC clones was smaller than the size of DNA fragments recovered from gels. This phenomenon has been seen in several BAC libraries, such as the BAC library of tomato [13]. The reason for this is that smaller fragments could be included in larger fragments during electrophoresis, and during ligation, the relatively smaller fragments were preferentially ligated to the vectors. PCR screening of the library with 24 markers each from one of 24 LG and 15 randomly selected ESTs/genes demonstrated that the BAC library provided good coverage of the L. calcarifer genome. Whether the BAC clones with large inserts were of hybrid origin remains to be examined.
A second generation linkage map of L. calcarifer is under construction by integrating new markers including microsatellites, ESTs and candidate genes onto the first generation map. Low polymorphism of ESTs and candidate genes was a bottleneck to map them to the linkage map [6]. Using highly polymorphic microsatellites located in BAC clones harboring interesting genes and ESTs, these interesting genes and ESTs could be mapped onto the linkage map as shown in this experiment. By employing highly polymorphic microsatellites in BAC clones, we have mapped 86 BAC clones to the linkage map of L. calcarifer. At least one BAC clone has been anchored on each LG, which can be used to integrate linkage and physical maps in the future. The number of markers on the linkage map of L. calcarifer increased to 302 by mapping 62 novel microsatellites located in 62 BAC clones onto the map. The two male linkage groups (i.e. LG14 and LG18) which were split to two LG on the first generation map respectively [6] due to the limited number of markers on these LG, were merged to one LG respectively by integrating  The BAC library of L. calcarifer could be also used in constructing a physical map by BAC fingerprinting [12,20], sequencing BAC ends and positional cloning of QTL of commercial interests [4] so as to facilitate selective breeding of L. calcarifer. Eventually, the BAC library can be used in whole genome shotgun sequencing when it becomes necessary.

Conclusion
A first L. calcarifer BAC library with 6.9 × coverage of the haploid genome has been constructed and characterized. Screening the library with 24 markers and 15 ESTs/genes demonstrated good genome coverage of this library. Eighty-six BAC clones were mapped to the first generation linkage map, improving the marker density of the linkage map of L. calcarifer. This BAC library together with the improved linkage map not only supplies an indispensable tool to physical mapping, integration of physical and linkage maps, and positional cloning for genes of importance, but also contributes to comparative genomic studies and eventually genome sequencing.

Preparation of high-molecular-weight DNA
Five hundred microliters of blood was collected from a male individual of L. calcarifer with a heparinized syringe. The concentration of leucocytes was quantified to be approximately 10 9 cells/ml. Quantities corresponding to 2.14 × 10 7 cells were embedded in 40 µl of 2% InCert agarose (in PBS) for DNA extraction. The mixture was then transferred into ice-cold plug moulds (Bio-Rad, SG, Singapore). Individual plugs were released into cell lysis solution [1% lithiumdodecyl sulfate, 10 mM Tris (pH 8), 100 mM EDTA (pH 8)] that was incubated at 37°C for 1 hr with occasional swirling. The cell lysis solution was replaced with 50 ml of new cell lysis solution and incubated overnight at 37°C. The cell lysis solution was sup-planted with 50 ml of 20% NDS. Two ml of proteinase K (20 mg/ml) was added to each 50 ml of 20% NDS consisting of 0.2% N-laurylsarcosine, 2 mM Tris-HCL (pH9.0), 0.14 M EDTA. The solution was incubated at 37°C overnight. Plugs were washed three times with TE50 and 0.05 M EDTA for one hour at room temperature. The plugs were put into a fresh Falcon tube, and washed twice with 50 ml TE50 and 50 µl PMSF (100 mM) at 37°C for 20 min to inactivate proteinase K. The plugs were then washed twice with 50 ml of TE50 in the Falcon tube at room temperature for 30 min to get rid of the PMSF.

Partial digestion of high molecular weight DNA and size selection
Digestion with restriction enzyme HindIII, pulse field gel electrophoresis (PFGE), isolation and purification of high molecular weight (HMW) DNA were performed using the method described previously [21]. Briefly, after displacement of the plugs by 1 × TE buffer, the agarose plugs were soaked in 800 µl of HindIII digestion buffer [0.015% bovine serum albumin (BSA), 75 mM NaCl, 12 mM Tris-HCl (pH 7.50)] and 3 U of HindIII for 16 hours at 4°C, after which, 100 µl of MgCl 2 (100 mM) was added and the Size selection was carried out as described [22], with minor modifications. In brief, partially digested DNA was separated by PFGE in 0.5 × TBE on a CHEF-DRII apparatus (Bio-Rad, SG, Singapore) under the following conditions: 14°C, 6.0 v/cm, angle = 120°, initial switch time = 5 sec, final switch time = 15 sec, run time = 16 hours and ramping = linear. At the end of this electrophoresis step, the gel portion containing DNA of 50 kb or less in size together with the portion of the gel containing the original plugs was removed. 1% fresh agarose was added to the remain-ing gel followed by a second electrophoresis step using the same conditions for 18 hours. Gel slices containing size fractionated DNA were obtained by cutting horizontally at 0.5 cm intervals in the size range of 100-250 kb. Each excised gel slice was subsequently inverted and buried in 1% low-melting-point agarose gel. A third electrophoresis step using the same conditions for 18 hours was carried out to concentrate the widely spread DNA fragments in each gel slice into a sharp single band. The band of size selected genomic DNA was then excised and dialyzed in 1 × TAE at 4°C overnight.

Ligation and electroporation
Size fractionated DNA was recovered from each gel band by electroelution in Spectra/Por 7 dialysis bags (Spectrum Laboratories, CA, USA) as described [23].   USA). Electroporation was carried out using a BioRad Gene Pulser (BioRad, CA, USA) at 4 kΩ and 350 V. Cells were incubated in 1 ml SOC medium at 37°C for one hour with shaking and later spread on LB plates containing 12.5 µg/ml Chloramphenicol, 40 µg/ml X-gal and 100 µg/ml IPTG and incubated at 37°C for 24 hours to allow the blue color to develop sufficiently.

Isolation of BAC DNA and estimation of insert size
We isolated BAC DNA from 212 BAC clones randomly chosen using a QIAwell 8 Plasmid Kit (Qiagen, HRB, Germany) following the protocol of the manufacturer. Isolated BAC DNA were digested with the restriction enzyme NotI and then subjected to PFGE for 16 hours using the same PFGE conditions as those for high molecular weight DNA isolation.

Library pooling and PCR screening
White recombinant colonies were manually picked and arrayed to plates (Genetix, Hampshire, UK) of 384-well each containing of 60 µl of LB media and 25% glycerol. Plates were incubated overnight at 37°C and stored at -80°C. The frozen stocks of the primary clones in 384 well plates were recovered and transferred to 4 96-well PCR plates containing 100 µl LB medium supplemented with 15% glycerol and 12.5 µg/ml chloramphenicol, then incubated overnight at 37°C to make a copy of the BAC library.
To establish a hierarchical PCR screening system, the library was divided into 11 superpools each consisting of 12 plates of 384-wells. Each superpool was divided into 48 pools each consisting of one 96-well plate of BAC clones. Cultures from 48 pools were combined to make superpool DNA for the first step PCR screening. Cultures from 48 plates of 96-well BAC clones were combined to make pool DNA for the second step PCR screening. In each pool, cultures from each well of 96 clones from a 96well plate were used for the third step screening.