Construction of physical maps for the sex-specific regions of papaya sex chromosomes

Physical mapping of HSY

Physical mapping started with screening the hermaphrodite BAC library using the sex co-segregating sequence characterized amplified region (SCAR) marker W11 by which four positive BACs were identified. The BAC end sequences (BES) of all four positive BACs were used for sorting out the configuration to select the two most extended ends as probes for the next round of screening of the BAC library. This stepwise chromosome walking process produced an HSY contig of about 900 Kb.

Additional sex co-segregating markers were used for screening the BAC library to establish multiple start points to extend the physical map, which greatly improved HSY chromosome walking that was hindered by highly repetitive sequences. Eighty five of the 225 sex co-segregating AFLP markers had sufficient distance from neighboring bands and fragment size (200 to 500 base pairs) to be suitable for Southern hybridization. These 85 markers were excised from AFLP gels to develop Carica papaya sex-liked markers (CPSM). Sixty three of the 85 fragments were re-amplified using the original AFLP primers. These 63 CPSMs were hybridized to the hermaphrodite BAC library, and 42 yielded positive hits. Among the 42 hybridized markers, 27 identified BACs that were either in the existing contig or formed new contigs. The remaining 15 probes hit numerous BACs indicating that they contained repetitive sequences. The library screening using CPSM resulted in five contigs with a combined size of 2.5 Mb [8].

Physical mapping was further assisted by sequencing selected seven BACs that were confirmed by FISH on the Y^h[14, 18]. These seven clones served as anchors, or start points, for chromosome walking to extend the physical map and to build the minimum tiling path. When the genome wide physical map based on fingerprints of the entire hermaphrodite BAC library became available [9], BACs on a minimum tiling path were selected and validated by multiple PCR amplification, and the BACs at the extreme ends of the existing contigs were used as seeds to search for contigs in the physical map. The genomewide physical map coupled with BAC end sequences (BES) accelerated construction of the HSY physical map. At this stage, higher stringency selection criteria were applied when adding new BAC clones: (1) primers from one of the BES of the candidate extension clone must amplify the anchor/seed BAC clone and vice versa; (2) at least two PCR product sequences show 100% match in the overlapping region between the candidate BAC and the established BAC in the physical map; (3) one or both BES of the candidate show male-specificity (not in all cases); (4) ultimately, candidate BAC clones have to be confirmed to be on the Y^h chromosome by FISH [16].

Four (Knobs 2 – 5) of the five heterochromatic knob structures that were Y^h-specific [16] were positioned on the physical map (Figure1a). Unmapped Knob 1, the only Knob structure shared between the Y^h and X chromosomes, located in the gap between Knob 2 and Border A. Chromosome walking to close this gap kept landing to the X chromosome and thus could not fill the gap. The last two BACs at the end of the contig bearing Knob 2, SH62E02 and SH61K24, were confirmed by FISH to be on the X chromosome (Figure1a).

The physical map of HSY reached about 8.5 Mb from four contigs with the remaining three gaps. Multiple attempts to close the gaps by re-hybridization of probes from the extreme ends of contigs to the hermaphrodite BAC library were unsuccessful. Two new BAC libraries were constructed using different restriction enzymes, one from male AU9 and the other from female SunUp [19] for the screening. Two of the three gaps were filled by screening the male AU9 BAC library, one with AM126G07 and the other with AM57M14. A third male AU9 BAC, AM125I09, extended the contig for the remaining gap. This BAC was confirmed to be on the Y^h chromosome, even though the seed BAC 61K24 was on the X chromosome. The final version of the HSY physical map was now about 8.5 Mb, excluding shared border sequences with X chromosome (Figure1a).

A total of 332 PCR-amplified probes was used for library screening and primers used for designing probes were used again for further confirmation of positive clones and cross amplification for configuration during chromosome walking (Additional file 1: Table S1). Unfortunately, a majority of those positive clones from library screening had been eliminated due to less than maximal extension, redundancy, or more likely being false positive as a result of repetitive elements or the initial low stringency library screening. A minimum BAC tiling path containing 77 anchored BACs (Figure1a) was built on the basis of the several lines of evidence during chromosome walking as described in the methods section.

Physical mapping of the HSY corresponding region in the X chromosome

Two X chromosome-specific BAC clones, SH53E18 (EF661026) and SH61H02 (EF661023), previously isolated through the presence of X-specific sequences, showed high sequence identity (>95%) to genic regions of two HSY BACs, SH 95B12 and SH85B24, respectively [14]. These two X BACs were used as seed BACs to start chromosome walking on the X counterpart on both directions simultaneously. New seed BACs were identified based on sex-co-segregating SSR markers from the draft genome of female SunUp [17, 20, 21]. For example, the source sequence of the sex-co-segregating SSR marker, P6K697Y0, aligned with the end sequences of SH30B21 with 99% identity, and the SH30B21 were used as new starting points for chromosome walking on both directions until overlapping BAC clones were identified or the gaps were filled. In addition to the basic chromosome walking steps applied for the HSY, we also applied in silico chromosome walking using sequence data from the draft genome of female papaya [17] for physical mapping of the X chromosome (in Material and Methods). BAC clones, SH12I03, SH12G20, SH99B01, and SH23D15 were all identified through in silico chromosome walking.

Two contigs were built with one gap remaining after exhausting the genomic resources, including the draft genome, physical map, and BES. Attempting to fill the gap ( Additional file 2: Supplemental note 1) was not successful. The remaining gap on the X physical map was closed on the HSY physical map. The gap of the HSY physical map between Border A and Knob 2 was filled on the X physical map. The BAC used to identify Knob 1 by FISH was mapped in the region near Border A of the X chromosome map.

The physical map of the X counterpart consists of 57 SH-BACs, including 13 border BACs shared between HSY and X. The minimum tiling path of the X-specific region contains 44 BACs, spanning about 5.4 Mb, and with a gap remaining between AM136D11 and SF08K16 (Figure1b).

Physical mapping of borders A and B

Mapping the borders of the non-recombining HSY ensures the completeness of the HSY and its corresponding X physical maps. Two BACs, SH85C03 and SH86B15, were identified first through chromosome walking for the HSY physical map. Soon after, we identified the same extended BACs in the X chromosome walking process, suggesting that the location of this shared BAC clones might be in the border region. Concurrently, sex-linked SSR markers were used to identify more clones at the two border regions. Two sex-linked SSR markers (CPM1055C0 and P3K2608C0), showed genetic distances of 2 and 3 cM from the non-recombining HSY, respectively, one on each side the HSY. These two markers were used to retrieve the source sequences in the draft genome. The associated BAC clone SH94G23 was designated as border A and SH68N07 was designated as border B of the HSY region. BAC SH68N07 at border B was physically about 1 Mb away from SH86B15 and in the same fingerprint map as SH86B15 ( Additional File 3: Figure S2). Therefore, the minimal tiling path between these two BACs was established by the fingerprint map being confirmed by cross amplification among the overlapped BACs (Figure S2b). BAC SH94G23 at border A was almost imbedded in SH69A15, another large clone about 345 kb in size. SH69A15 was used as a seed BAC for chromosome walking in both directions until SH54A09 was identified, which overlapped with the X BAC clone, SH84M10 during chromosome walking. However, a gap remained unfilled between border A and the HSY physical map, specifically between SH54A09 and AM125I09, (Figure1a). Attempt to fill the gap had identified clones that were all on X chromosome without any evidence of being Y^h-specific. This region contains Knob 1, the only Knob structure shared between X and HSY.

Defining the borders of HSY via fine mapping

Two of 43 SSR markers: 58 C24-25b and BorderB-2001a were polymorphic at border B. BorderB-2001a identified three recombinants among the 87 previously identified ones. These three recombinants were then mapped using the 58 C24-25b marker which located inside BorderB-2001a. Only one recombinant was detected (Table3). No recombinations were observed at the position of a more inner sex-co-segregating marker, P3K4642, suggesting that border B was located between 58 C24–25b and P3K4642, within BAC SH58C24 (Figure1b; Table3). Therefore, the total HSY region was defined genetically to lie between BACs SH84M10 and SH85C03.

BAC libraries

The three BAC libraries used for physical mapping were constructed from leaves of: (1) a hermaphrodite plant of the gynodioecious papaya variety SunUp (SH) [26], (2) a female plant of SunUp (FS), and (3) a male plant of the dioecious variety AU9 (AM) [19]. The hermaphrodite SunUp BAC library, as the primary one used in this study, included 39,168 BAC clones providing 13.7X papaya genome equivalents. This library was from Hin d III digestion with an average insert size of 132 kb from two batches of partial digestion; the first batch consisted of 18,432 BAC clones with an average insert size of 87 kb, The second batch consisted of 20,736 BAC clones with an average insert size of 174 kb [26]. The second batch of the hermaphrodite SunUp BAC library was particularly useful for physical mapping by providing many large-insert clones that were used to build the minimum tiling path for sequencing. Two additional BAC libraries were constructed using different restriction enzymes to fill the possible gaps of the primary BAC library. Bst Y I was used for the SunUp female BAC library which consisted of 36,864 clones with an average insert size of 104 kb, providing 10.3x genome equivalents. Eco R I was used for the AU9 male BAC library which consisted of 55,296 clones with an average insert size of 101 kb, providing 15.0x genome equivalents [19].

BAC end cloning and BAC end sequencing

A total of 67,179 BES of the hermaphrodite SunUp BAC library was obtained through high throughput BAC DNA isolation and sequencing of both ends of the BAC DNA with ABI 3700 x1 DNA Analyzers [21]. We applied both BAC end cloning and BAC DNA direct sequencing for the BAC clones with failed BES or without BES for all libraries including the hermaphrodite SunUp, female SunUp, and male AU9. In the BAC end cloning procedure, 30 ng of BAC DNA were isolated through standard miniprep procedure and digested with Not I and Dra I and then ligated to Not I and EcoRV sites of pPCR-script AMP SK(+) vector (Stratagene). The BAC ends were directly amplified by using KS (5′-TCGAGGTCGACGGTATCG-3′) and LF1 (5′-ACCTGCAGGCATGCA-AGC-3′) primers for left end, and KS and RR4 (5′-GGTGACACTATAGAATACTCAAGC-3′) primers for right end. The amplified BAC end with a size between 200–1000 bp was purified and sequenced by ABI 3730XL. For the BAC DNA direct sequencing approach, the BAC DNA was isolated using Pure Yield Plasmid Midiprep System (Promega, Madison, WI, USA) following the manufacturer’s instructions. The DNA concentration was adjusted to 200 ng/ul and then sequenced by the ABI 3730XL using T7 for the left BES and M13rev primers for the right.

Primer design

Primer3 (v. 0.4.0) [32] was used for designing primers used for chromosome walking. The default criteria for the primers designed from available BES was applied except that the product size range was adjusted to reach the maximum allowed by the size of the BES. About 60 bp at the start and end of the sequences were avoided for primer design due to possible poor sequence quality. For the primers designed from fully-sequenced BAC clones, about 5 kb of sequence from each end were submitted to Primer3 and three pairs of primers with product size between 800-1000 bp were designed at 1, 2, and 3 kb away from the ends. All primers were synthesized by Invitrogen Corp (Carlsbad, CA 92008 USA) and Bioneer Inc (Alameda, CA 94501).

BAC library screening through hybridization

The PCR products (1–5 ug) used as probes were purified with Wizard SV gel and PCR clean-up (Promega, Madison, WI, USA) and eluted to 16 μl with double distilled water. The probes labeled with digoxigenin-dUTP were used for hybridization, and the hybridization signal was detected by chemiluminescence method using DIG High Prime DNA Labeling and Detection Starter Kit II (Roche Applied Science, Indianapolis, IN 46250 USA). The procedure was mainly based on the manufacturer’s instructions with modification. Briefly, the 20 μl labeling reaction including the 16 μl denatured probe and 4 μl DIG-High Primer was incubated at 37°C for 20–24 hrs. The filter (23 x 23 cm) sets containing the complete BAC library gridded in duplicate were pre-hybridized overnight in Q-trays with 100 ml of DIG Easy Hyb buffer and then hybridized overnight in fresh buffer including denatured labeled probe and 50 ng labeled λ DNA for blocking background signals. After hybridization, the filters were subjected to high stringency washes and immunological detection according to the instruction and exposed to X-ray film for 3–12 hrs. Positive BAC clones were identified by BAC-DMS 2.1, a SAS coded program [33].

BAC library screening through PCR

Clones of the hermaphrodite SunUp BAC library were pooled following the Matrix and SuperPool pooling strategy by Amplicon Express (Pullman, WA 99163 USA). The pool screening was done in a two-round PCR procedure by first using super pools (12 plates) as templates and second using matrix pools as template following the User’s Manual version 3.2 (Amplicon Express, Pullman, WA 99163 USA). Identification of positive clones was accomplished by comparing the positive hits to the pool keys provided in the manual.

PCR confirmation of positive clones and configuration

The positive clones identified from the library screening were inoculated into 200 μl Luria broth (LB) medium with 12.5 μg/μl of chloramphenicol and grown overnight at 37°C on a rotary shaker. The glycerol stocks were made by mixing equal volumes of overnight culture and 30% glycerol. The stocks were used as templates for PCR to verify the positive clones from library screening and to configure the extent of overlap with the existing physical map. The 10 μl PCR reaction included 0.8 μl glycerol stock of the clone, 1X PCR buffer, 0.15 mM of each dNTP, 1.8 mM MgCl₂, 0.15 μM of reverse and forward primers, and 0.8 units of Taq polymerase. The PCR program was 94°C for 5 min, then 40 cycles of 30 s of denaturing at 94°C, 40 s of annealing at 55°C, and 30–60 s (depending on the expected size) of extension at 72°C, and then with a final extension at 72°C for 10 min. PCR products were separated on 1.0 % agarose gel and visualized on a gel doc system (BioRad, Hercules, CA 94547 USA).

Y-specificity test

For Y-specificity test of PCR-confirmed positive clones, PCR was run on all templates including glycerol stocks of the positive clones and four DNA samples from female SunUp, hermaphrodite SunUp, female AU9, and male AU9 as templates. If the size of the PCR products from the male or hermaphrodite were distinctively different from that of the females of SunUp and AU9, the marker was documented as Y-specific.

Fluorescence in situ hybridization

The HSY locations of the extended positive BAC clones were further validated by Fluorescence in situ hybridization (FISH). Slide preparation was described previously [16]. The FISH procedure was conducted according to Jiang et al. (1995)[34].

BAC insert size determination

The confirmed positive BAC clones were inoculated into 5 mL of LB medium with 12.5 μg/μl chloramphenicol and incubated overnight at 37°C on a rotary shaker. Plasmid DNAs were isolated from the culture using a standard miniprep procedure. After precipitation, the DNA pellet was disolved in 30 μl of TE buffer. 10 μl of the plasmid DNA was subjected to Not I enzyme digestion at 37°C for 3 hr and loaded to the Clamped homogenous electric fields (CHEF) gel with the MidRange Marker II (New England Biolab, USA) as a size standard.

Chromosome walking to identify the extended clones

Chromosome walking was conducted in both directions until an overlapped clone was identified or gaps between contigs were filled. The basic steps employed during chromosome walking to identify each extended BAC clones from the anchored clones are outlined in Additional file 4: Figure S1.

For chromosome walking on X counterpart, we also applied in silico chromosome walking using sequence data from the draft genome of female papaya [17]. We blasted the BESs of anchored BACs against the sequence contigs of the female papaya draft genome. The top hits with sequence identity above 98% were then blasted against the BESs of all the clones in the hermaphrodite BAC library [21]. The candidate overlapping BAC clones were identified and then subjected to PCR confirmation and cross amplification to sort out the configuration of the BAC contigs. The remaining steps were the same as for HSY chromosome walking to validate overlapping sequences ( Additional file 4: Figure S1). FISH was also used to examine the localization of each potential anchor BAC clone on the X chromosome.

Genomic resources

Relevant genomic resources include a draft genome of female SunUp [17], the BES of a hermaphrodite SunUp BAC library [21], a physical map of the papaya genome [9], and a collection of 16,362 EST unigenes [17].

Fine mapping of the sex determining region

The F₂ mapping populations of 1522 progenies were derived from four different crosses (Table1). These populations were grown at the Kunia substation on Oahu, Hawaii. Young leaves of each F₂ progeny were used for DNA isolation [35]. Among the 1460 individuals, 452 were phenotyped as female and 1008 as hermaphrodite by observing their flowers types, and 62 were excluded from the analysis due to unknown (not scored) sex type.

SSRs longer than 18 bp in length and the corresponding primers for each SSR were identified using SSR finder software (http://www.maizemap.org/bioinformatics.htm) and used for fine mapping. The source sequences for SSR primer design from the draft genome of female SunUp included HSY BACs (SH75H06 and SH58C24), X BACs (SH54M22, SH99B01, SH69D05), BACs near borders (SH12I03, SH84M10, and SH86B15), and 4 supercontigs (66, 177, 123, 39). All SSR primers ( Additional file 5: Table S2) were synthesized by Bioneer Inc (Alameda, CA 94501). The primer sets were screened for polymorphism using DNA samples of the parents and two DNA pools of female and hermaphrodite F₂ progenies. The PCR conditions were described previously [20]. The markers exhibiting sex-linked polymorphism were used for fine mapping.