Twenty bacterial artificial chromosome (BAC) clones that could produce bright signals and no or very low fluorescence in situ hybridization (FISH) background were identified from Gossypium arboreum cv. JLZM, and G. hirsutum accession (acc.) TM-1 and 0-613-2R. Combining with 45S and 5S rDNA, a 22-probe cocktail that could identify all 13 G. arboreum chromosomes simultaneously was developed. According to their homology with tetraploid cotton, the G. arboreum chromosomes were designated as A1–A13, and a standard karyotype analysis of G. arboreum was presented. These results demonstrated an application for multiple BAC–FISH in cotton cytogenetic studies and a technique to overcome the problem of simultaneous chromosome recognition in mitotic cotton cells.
COTTON is one of the most important natural fiber and edible oil crops in the world. Significant progress has been made in the development of genetic maps for the purpose of gene and quantitative trait loci (QTL) mapping (Reinisch et al. 1994; Ulloa et al. 2002; Zhang et al. 2002; Mei et al. 2004; Nguyen et al. 2004; Rong et al. 2004; Song et al. 2005; Han et al. 2006; Guo et al. 2007). However, due to the large number and small size of the chromosomes and especially the absence of suitable cytogenetic markers such as bands, chromosome identification has lagged significantly behind the development of linkage mapping. Although many genetic materials such as translocation, aneuploid lines, and the nomenclature for the tetraploid cotton chromosomes have been developed (Brown 1980; Endrizzi et al. 1985), the routine and unambiguous identification of individual chromosomes, especially in mitotic cells, is almost impossible in cotton. The bacterial artificial chromosome–fluorescence in situ hybridization (BAC–FISH) technique has been developed and used in physical mapping (Hanson et al. 1995), and individual mitotic chromosomes were easily identified in cotton (Wang et al. 2006). Then one set of chromosome-specific BACs was developed, and the comparison mapping between physical and genetic maps was conducted (Wang et al. 2007a,b). Therefore, BAC–FISH has shown a unique superiority for cotton chromosome identification and physical mapping.
The cotton genus Gossypium is composed of 45 diploid and 5 tetraploid species. Among them 4 are cultivated Gossypium species: G. hirsutum L. [n = 2x = 26, (AD)1], G. barbadense L. [n = 2x = 26, (AD)2], G. arboreum L. (n = x = 13, A2), and G. herbaceum L. (n = x = 13, A1). Tetraploid species (G. hirsutum L. and G. barbadense L.) dominate worldwide cotton production. G. arboreum, as an A-genome diploid cotton, has been domesticated and cultivated in China for almost 2000 years (Xiang and Shen 1989). Due to some of its superior agronomic traits, such as disease and insect resistance, high fiber strength, and excellent plasticity, which upland cotton cultivars lack, G. arboreum is still planted and is used worldwide as a germplasm resource in present-day cotton breeding programs. G. arboreum is generally regarded as one of the best exemplars of the A-subgenome progenitors (Endrizzi et al. 1985; Wendel et al. 1995). Therefore, the G. arboreum species is important for agricultural production and genomic and evolution research in cotton.
Here, we described the screening of one diploid cotton (G. arboreum) and two tetraploid cotton (G. hirsutum) BAC libraries to identify new chromosome-specific clones. Combining with previous BAC clones and 45S and 5S rDNA, a 22-probe cocktail was developed for the purpose of simultaneous chromosome identification and karyotyping in G. arboreum. This enabled us to identify all of the 13 G. arboreum chromosome pairs simultaneously. According to their homology with the A-subgenome (At) chromosomes in tetraploid cotton, the nomenclature and standard karyotype of G. arboreum chromosomes were developed.
MATERIALS AND METHODS
The BACs used in this study came from three genomic BAC libraries, of which two were constructed from the tetraploid cotton G. hirsutum accession (acc.) TM-1 (Y. Hu and T. Z. Zhang unpublished data), and G. hirsutum acc. 0-613-2R, a cytoplasm male fertility line (Wendel et al. 1995; Yin et al. 2006), and A-genome G. arboreum cv. JLZM (Y. Hu and T. Z. Zhang unpublished data). The corresponding SSR markers used to screen the BAC clones were selected from the tetraploid map (Guo et al. 2007) and a new A-genome genetic map (Ma et al. 2007). The library screening was conducted as described previously by Wang et al. (2006). The 45S and 5S rDNA derived from Arabidopsis thaliana were kindly supplied by X. E. Wang, Nanjing Agricultural University, China. All the BAC clones were initially evaluated in the FISH, and only BAC clones that could produce bright signals and no or very low background were selected as candidates for multiple FISH.
To reidentify their physical locations, all BAC candidates for which the corresponding genetic markers were derived from the same linkage group were hybridized simultaneously with the At-derived chromosome-specific BACs (Wang et al. 2007b) in FISH on both JLZM and TM-1 mitotic metaphase chromosomes. When different locations occurred, more At-derived chromosome-specific BACs were used to hybridize in the simultaneous FISH to find the new BAC locations. Root tip, slide preparation, and single- and dual-color FISH have been described previously (Wang et al. 2006). For the 22-probe cocktail FISH, some modifications were made as follows: All the BAC DNA or rDNA were labeled by biotin and digoxigenin, detected by anti-digoxigenin-rhodamine (red) and avidin-fluorescein (green), and the white or yellow color was conducted by labeling the probe with biotin and digoxigenin. For the probe-cocktail mixture, 100 ng of each of the labeled probe DNA, 5 μg Cot-1 DNA derived from TM-1 and JLZM, respectively, were mixed together, precipitated in ethanol, and then dissolved in 4 μl distilled water plus 10 μl formamide, 2 μl 20× SSC and 4 μl 50% (w/v) dextran sulfate. After the probe mixture was denatured at 97° for 10 min, it was annealed for 1 hr before being applied to the slide.
Following the posthybridization washes, chromosomes were counterstained with DAPI (4′,6-diamidino-2-phenylindole) (Sigma, St. Louis). The slides were examined under an Olympus BX51 fluorescence microscope. Chromosome and FISH signal images were captured using an Evolution VF CCD camera (Media Cybernetics, Bethesda, MD) and merged using Image-Pro Express software. For karyotyping images, DAPI-stained chromosomes were measured using the same Image-Pro Express software. The arm ratio (average long arm/short arm ratio), total chromosome length (short arm + long arm), and relative chromosome length (length of the individual chromosome/total length of all chromosomes in the genome) were calculated for each of the 20 mitotic chromosomes.
RESULTS AND DISCUSSION
Cytogenetic studies rely on accurate and consistent chromosome identification, which is always a challenge in plant species with small chromosomes. Several G. arboreum karyotypes have been reported on the basis of chromosome identification by length (Skovsted 1933; Skovsted 1934; Wu and Chen 1984; Gennur et al. 1988; Song et al. 1991; Wang et al. 1997). However, the use of different materials and measurement variability between laboratories can potentially result in the misidentification of chromosomes. Consequently, significant discrepancies exist among the previously published karyotypes. More importantly, as one of the ancestor species of tetraploid, none of these G. arboreum karyotypes are integrated with the tetraploid or genetic linkage maps. To remove these limitations, we isolated a set of 20 BAC clones from A and At genomes using genetically mapped SSR markers. Combining with the 45S and 5S rDNA, this 22-probe cocktail not only allowed the correct identification of all 26 mitotic metaphase chromosomes of G. arboreum cv. JLZM, but also enabled us to completely integrate our karyotype with the genetic linkage map of G. arboreum (Figures 1 and 2). Moreover, the identification of BAC clones, which contain no or less repetitive DNA and generate bright and low or no background, is critical for the success of BAC–FISH. BAC–FISH has not been widely used in cotton cytogenetic studies due to the high percentage of repetitive sequences. In this study, we identified twenty BACs from more than fifty BACs, and successfully demonstrated the use of multiple BAC–FISH for cytogenetic research in cotton.
Among this 22-probe cocktail, six BAC probes were derived from the A-genome species G. arboreum cv. JLZM BAC clones, and all the others were from the tetraploid TM-1 or 0-613-2R clones, including nine previously identified chromosome-specific BACs (Wang et al. 2007a,b; Wang et al. 2006). Through the cohybridization with one set of At chromosome-specific BACs, the physical locations of diploid BACs were identified and the chromosomal nomenclature of G. arboreum was aligned with the tetraploid cotton, in that the JLZM chromosomes were designated as A1–A13 according to the At chromosomes (Figures 1 and 2). Therefore, our nomenclature system for G. arboreum was integrated with the At chromosomes. The linkage groups of the G. arboreum genetic map could also be identified according to the A- and At-derived BACs and corresponding SSR markers. The uniform nomenclature between the diploid A- and tetraploid At-genomes in cotton will facilitate chromosome identification, ease communication between different research groups, and facilitate data usage and comparisons between A- and At-genome maps.
On the basis of the identification of metaphase chromosomes, a FISH-based karyotype of G. arboreum JLZM was developed (Table 1). Our results show that the lengths of the mitotic metaphase chromosomes in the current numbering system are not in descending order. The current descending order of the metaphase lengths is A5 (including the rDNA section), A7 (including the rDNA section), A3, A4, A6, A1, A12, A2, A10, A8, A13, A11, and A9. All the chromosomes could be grouped into two classes according to their total lengths. Seven chromosomes (A1, A3, A4, A5, A6, A7, and A12) are the longer chromosomes and the other six (A2, A8, A9, A10, A11, and A13) are the shorter ones. The longest chromosome was chromosome A5, which had an absolute length of 3.79 μm and a relative length of 8.85%. The shortest chromosome was A9, which had an absolute length of 2.80 μm and a relative length of 6.55%. None of the chromosomes were exceptionally long or short, and all of them were metacentric except for chromosome A12, which was the only pair of midsized but submetacentric chromosomes.
Through the exceptional appearance of secondary constriction on the chromosome, some researchers reported finding one or two nucleolar organizer regions (NORs) in different G. arboreum species using the classical dyeing method (Edwards 1977; Wu and Chen 1984; Gennur et al. 1988; Song et al. 1991). Two major NOR loci were found using rDNA FISH in different G. arboreum species (Hanson et al. 1996; Wang et al. 2001; Bie et al. 2004). At the same time, two major NOR loci were also found in G. herbaceum, a diploid species highly homologous with G. arboreum. Although the different G. arboreum species might contain different numbers of NORs resulting from long-term evolution (Stebbins 1952; Gennur et al. 1988), we expected the rDNA FISH technique to result in more accurate results.
We also located two major NORs in this study, using the 45S rDNA cohybridization technique with some chromosome-specific BAC clones in FISH. One NOR was located near the end in the short arm of chromosome A5 and another was located at the end in the short arm of chromosome A7. A very minor 45S rDNA signal was also found and located on the chromosome A9 when using the single 45S rDNA as a probe in FISH. Its signal locus had some overlap with that of the 5S rDNA (also located on chromosome A9) and was too weak to be detected in multiple FISH when we adjusted the photo contrast. Hanson et al. (1996) found two major NORs in the A-subgenome of G. hirsutum, then they found another major NOR and located all three NORs on chromosomes At5, At7, and At9 (Ji et al. 1999). All of these results support our findings and indicate that G. arboreum was one of the ancestors of G. hirsutum.
It is widely accepted that G. arboreum or G. herbaceum (A-genome species) and G. raimondii (D-genome species), the ancestors of the A- and D-subgenomes of tetraploid cotton, respectively, were derived from the same ancestor 6–11 million years ago (Wendel 1989). Cytogenetic studies have revealed that the A-genome species differed from each other by a single translocation, whereas the A-subgenome differed from A, D, and Dt by two reciprocal translocations (Brown and Menzel 1950; Gerstel 1953; Menzel and Brown 1954; Endrizzi et al. 1985). The molecular data derived from tetraploid cotton also showed two translocations between chromosomes At2 and At3, and At4 and At5 (Brubaker et al. 1999; Rong et al. 2004; Desai et al. 2006; Guo et al. 2007). Our FISH results confirmed these findings. Two BAC translocations involving two chromosome pairs of A2 and A3 and A4 and A5 between G. arboreum and the A-subgenome of G. hirsutum were found. As shown in Figure 3, one BAC 64F22 that located on At2 of G. hirsutum was located on chromosome A3 of G. arboreum. It appears that the chromosomal segment containing the BAC 64F22 was involved in a translocation between chromosomes A3 and At2. Another BAC translocation was found between the chromosomes At4 and A5 shown in Figure 3. These results are consistent with previous cytological (Brown and Menzel 1950; Menzel and Brown 1954) and molecular (Brubaker et al. 1999; Rong et al. 2004) data. But two reciprocal translocations were found in their research. Unfortunately, because of the few BACs identified here, no translocation BACs were found on the other two pairs of chromosomes, A2 and At3 and A4 and At5. Therefore, we could not identify these two cases as reciprocal translocations, but with effort, we believe complementary results will be achieved. Moreover, although only a few translocation BACs were found here, the large size of the DNA segment represented should be expected to provide more powerful translocation detection.
Additionally, the results did not present an integrated “cytogenomic” map between physical and genetic maps, mainly because of the few JLZM BACs available. However, relationships were established between the 13 chromosomes of G. arboreum and linkage groups of A-subgenome chromosomes of G. hirsutum, and, no doubt, these data will facilitate linkage map construction and chromosomal assignments in G. arboretum and assist in the isolation and construction of a BAC-by-BAC physical map and even in the sequencing of the individual chromosomes.
We thank Xiu E. Wang of NAU for technical assistance and for supplying the 45S and 5S rDNA. This work was supported by the National Natural Science Foundation of China (30700510), Youth Sci-tech Innovation Foundation of Nanjing Agricultural University of China (KJ07002), Changjiang Scholars and Innovative Research Team in University of MOE (IRT0432), and the Programme of Introducing Talents of Discipline to Universities, China.
↵1 These authors contributed equally to this work.
Communicating editor: J. A. Birchler
- Received October 20, 2007.
- Accepted November 23, 2007.
- Copyright © 2008 by the Genetics Society of America