Somatic Hybrids Between Arabidopsis Thaliana and Cabbage (Brassica Oleracea L.) with All Chromosomes Derived from A

Home , Arabidopsis

Somatic Hybrids between Arabidopsis thaliana and Cabbage (Brassica oleracea L.) with all Chromosomes Derived from A. thaliana and Low Levels of Fertile Seed

Hiroshi Yamagishi*, Shinya Nakagawa, Daisuke Kinoshita, Atsushi Ishibashi and Yoko Yamashita

Department of Biotechnology, Kyoto Sangyo University, Kamigamo, Kyoto 603-8555, Japan

We obtained somatic hybrids between Arabidopsis thaliana Columbia, and a cabbage variety, ‘Chusei Succession’ (Brassica oleracea L.), in addition to our previous success using the cabbage variety ‘Fujiwase’. All hybrids of the two combinations possessed the chloroplast genome of cabbage, while the mitochondrial genome was recombined. PCR primers specific to each of the five chromosomes of A. thaliana were designed and used to analyze the hybrids. It was found that the hybrids had all five chromosomes of A. thaliana. Somatic hybrids between Columbia and ‘Chusei Succession’ flowered and showed extremely low pollen fertility, producing no seeds by spontaneous self-fertilization, whereas seeds were obtained by crosses between somatic hybrids and three cultivars of B. oleracea. The seeds could be useful materials for the transfer of genes of A. thaliana to cabbage and to establish cabbage lines with novel mitochondrial genomes.

Key Words: Arabidopsis thaliana, Brassica oleracea (cabbage), chromosomes, seed fertility, somatic hybrids.

and Mathias, 1995; Siemens and Sacristán, 1995), most Introduction failed to obtain offspring. The single exception was the Arabidopsis thaliana has been utilized worldwide as hybrids between UV-irradiated A. thaliana and non- a model plant in plant genetics and molecular biology. treated Brassica napus produced by Forsberg et al. Especially after the accomplishment of genome (1998a, b). Because the hybrids produced seeds by back- sequencing projects (The Arabidopsis Genome Initiative, crossing with B. napus, progenies were selected for 2000), genome information of A. thaliana accelerated resistance to Leptoshaeria maculans to use for breeding comparative genomics for genetic mapping and the B. napus (Bohman et al., 2002). One reason for the identification of various useful genes in related species scarcity of research work on cell fusion using A. thaliana (Lan et al., 2000; Sillito et al., 2000), whereas attempts is that the protoplast culture and regeneration of plants to use the genes clarified in the A. thaliana genome for are much more difficult in A. thaliana than in other breeding crop plants are restricted in spite of the fact plants. Another reason is that A. thaliana belongs to a that A. thaliana is a member of family Brassicaceae to different tribe (Sisymbrieae) from Brassiceae, which which many useful crops belong. One main reason is contains many crops cultivated around the world, even that A. thaliana can not be cross-hybridized with any though they are in the same family. crops. The only method to obtain hybrids between However, an efficient protoplast culture system for A. thaliana and crop plants is cell fusion. A. thaliana was established (Yamagishi et al., 2002), by Gleba and Hoffmann (1980) had the first success in which more than a 50% shoot regeneration rate was producing Arabidobrassica, somatic hybrids between achieved. Using this system, we produced successfully A. thaliana and Brassica campestris, but the hybrids somatic hybrids with B. napus (Yamagishi et al., 2002), neither established as plants nor produced progenies. Raphanus sativus (Yamagishi and Glimelius, 2003), and Though a few groups reported asymmetric somatic Brassica oleracea (Yamagishi and Nakagawa, 2004). hybrids thereafter (Bauer-Weston et al., 1993; O’Neill Among them, hybrids with R. sativus and B. oleracea were the first reports of these combinations. Furthermore, Received; December 21, 2007. Accepted; February 12, 2008. although UV treatment was used with B. napus to obtain * Corresponding author (E-mail: [email protected]). asymmetric hybrids in the cell fusion with B. napus

277 278 H. Yamagishi, S. Nakagawa, D. Kinoshita, A. Ishibashi and Y. Yamashita

(Yamagishi et al., 2002) for the purpose of overcoming cell fusion treatment were as described in the previous the phylogenetic remoteness, somatic hybrids were paper (Yamagishi and Nakagawa, 2004). The fused produced without such treatment in the two latter protoplasts were cultured under the same media and combinations (Yamagishi and Glimelius, 2003; culture conditions as mentioned in Yamagishi and Yamagishi and Nakagawa, 2004). Nakagawa (2004). In the previous paper we reported somatic hybrids between A. thaliana Columbia and cabbage variety Investigations of the genome structure of somatic hybrids B. oleracea ‘Fujiwase’ (Yamagishi and Nakagawa, Because various kinds of PCR primers to detect the 2004). Although the hybrids are growing in a glasshouse, chloroplast and mitochondrial genomes were designed they have not flowered yet, and progenies have not been and utilized previously (Yamagishi and Glimelius, 2003; obtained. Thus, in order to increase hybrid materials, Yamagishi and Nakagawa, 2004), we applied them also somatic hybrids using another cultivar of cabbage to the somatic hybrids obtained here. For the chloroplast ‘Chusei Succession’ were produced. The hybrids are genome, two intergenic regions of trnF and ndhJ, and expected to be useful as breeding materials for cabbage trnH and psbA were studied, while PCR patterns were by introducing genes from A. thaliana. DNA markers investigated in atp1 and coxI regions for the mitochon- for each of the five chromosomes of A. thaliana were drial genome. On the other hand, for the nucleus genome, designed and their presence in the hybrids was only polymorphisms in the Tpi gene were examined investigated by PCR. Furthermore, because the hybrids previously by PCR (Yamagishi and Nakagawa, 2004), of Columbia and ‘Chusei Succession’ flowered, flower and it was not clarified whether the hybrids had all five morphology and pollen fertility were observed. The chromosomes of A. thaliana. Thus, DNA markers hybrids were crossed with B. oleracea as pollen parents, specific to each of the five A. thaliana chromosomes and seed fertility was examined. were designed. The chromosomes, gene names deposited in the databases (EMBL etc.), and the sequences of the Materials and Methods primers for PCRs are shown in Table 1. Using these Plant materials for cell fusion and the regeneration of primers, the PCR patterns of hybrids were compared somatic hybrids with those of the parents. The ecotype Columbia of A. thaliana and the cabbage variety ‘Chusei Succession’ (Nihon Nosan, Japan) were Cross hybridization of somatic hybrids with B. oleracea used for cell fusion. Both parental species were cultured The somatic hybrids of Columbia and ‘Fujiwase’ aseptically and protoplasts were isolated from hypocotyls continued vegetative growth and did not flower. On the and leaves of 2-week-old seedlings of A. thaliana, other hand, somatic hybrids of Columbia and ‘Chusei whereas true leaves were used for protoplast isolation Succession’ were propagated by meristem culture, and in the cabbage. The enzymes, isolation procedure, and the hybrid plants grew to the flowering stage, thus, the

Table 1. PCR primers and their sequences used to detect the five chromosomes of A. thaliana.

Chromosome Gene name Primer Sequence (5'→3')

IF10A5.13F1 GGAGGAGAAATCCAAGTCAAG R2 CTGTGTTCCTGCAAGCTTGT IF17F8.14F2 GCTCAGCAACCATATCATGC R1 CAAGCCTACTCTCTGAGATCC II F16P2 F1 TGGATTCTCTCGTTGAATCCG R2 GTCGAGCAATGCATTTTCCCT II T20F21 F1 GGAACAAAGCAGAGTTCAACG R2 CCATTGAGTTATATGCCCCG III F15B8.130 F2 ACATACAACACCTGATGGACGT R1 CTGTTCCATTAGATCGTCGAG III MXC7.10 F1 CCAACTACTGTCACGAAGAAG R2 AAGCTCATAACTGACACCTC IV F18A5.110 F1 TGAAAGGGCTTCCTGGTATG R2 GTACCATCATCTGTCTGGAAC IV T12G13 F1 GGATATGCAGGCATGTACATC R2 GCTATGAGAAGACGGCAATG V MZN1.16 F1 AAGCGATGGCTGCTTATAAGG R2 TCACTTACACTATCGCTGACTC VMQD19.17F1 GATAGCTCTGAGACTCACTTAG R2 TGGTAAAACCCACACCATGA J. Japan. Soc. Hort. Sci. 77 (3): 277–282. 2008. 279 flower morphology and pollen fertility of the latter regeneration and hybrid plant establishment was hybrids were observed. Pollen fertility was determined improved. by counting the pollen stained with acetocarmine using a microscope. For each plant, more than 500 pollen Presence of the five chromosomes of A. thaliana in grains were observed; thereafter, the hybrid plants were somatic hybrids crossed with three cultivars of B. oleracea, ‘Fujiwase’ As mentioned previously, we developed PCR markers (Cabbage; B. oleracea var. capitata), ‘Kairan’ for various sites in the chloroplast and mitochondrial (B. oleracea var. alboglabra), and ‘Purple Bird’ genomes (Yamagishi and Nakagawa, 2004), and these (Kohlrabi; B. oleracea var. gongylodes). In the crosses, markers were tested in somatic hybrids, which showed B. oleracea plants were used as the pollen parents. The the identical characteristics to those of hybrids between number of seeds were counted after maturation. Columbia and ‘Fujiwase’. Namely, the two somatic hybrid plants of Columbia and ‘Chusei Succession’ Results possessed only the chloroplast genome derived from the Regeneration of somatic hybrids between Columbia and cabbage parent, while they had hybrid traits in the ‘Chusei Succession’ mitochondrial genome. The PCR patterns of the hybrids After fusion treatment of the protoplasts isolated from and their parents are shown in Table 3 and Figure 1. Columbia and ‘Chusei Succession’, protoplasts were By PCRs with the primers designed for genes specific cultured in modified 8P medium and the colonies were to each of the five chromosomes of A. thaliana, DNA transferred to the medium for callus development. After fragments with the expected molecular sizes were clearly proliferation on the medium, 127 calli were transplanted amplified from Columbia, whereas the DNA fragment onto the regeneration medium. Thirty-three calli pattern of the cabbage parents either had different differentiated to form shoots, but most were judged to fragment sizes or produced no amplification (Table 3). be shoots of A. thaliana from observation of their In the former case, the DNA marker was expected to morphology because the shoots showed bolting soon after regeneration, which is a typical feature of A. thaliana different from that of the cabbage. While Table 2. Development of calli, shoots, and hybrid plants by culture nine shoots had morphological traits of the somatic after cell fusion treatment. hybrid, only two developed roots and established as Developmental stages of culture Number % somatic hybrid plants (Table 2). The number of hybrid Calli transplanted to the regeneration medium 127 100 plants was the same as our previous report in which Calli with regenerated shoots 33 26.0 ‘Fujiwase’ was used as the fusion parent (Yamagishi Hybrid shoots regenerated 9 7.1 and Nakagawa, 2004), but the efficiency of shoot Hybrid plants established 2 1.6

Table 3. PCR patterns of somatic hybrids between A. thaliana and cabbage for DNA markers in the nuclear, chloroplast, and mitochondrial genomesz.

Parentsx A + Fx A + Sx Genome Sitey AFS 1 2 1 2 Chloroplast trnF/ndhJ SLL L L L L trnH/psbA LSS S S S S Mitochondria atp1 SLLS + LS + LS + LS + L coxI (5') −++ + + + + coxI (3') +−− + + + + Nucleus I F10A5.13 +−− + + + + I F17F8.14 +−− + + + + II F16P2 L S S L + SL + SL + SL + S II T20F21 +−− + + + + III F15B8.130 L S S L + SL + SL + SL + S III MXC7.10 +−− + + + + IV F18A5.110 S L L S + LS + LS + LS + L IV T12G13 S − LSSS + LS + L V MZN1.16 L S − L + SL + SL L V MQD19.17 +−− + + + +

z PCR pattern was classified into + (presence) or − (absence) of the DNA fragment, and L (large) or S (small) DNA size. y Numbers of nuclear genes (I–V) indicate the chromosomes where genes are present. x A, F, S, A + F, A + S: See Figure 1. 280 H. Yamagishi, S. Nakagawa, D. Kinoshita, A. Ishibashi and Y. Yamashita

Fig. 1. PCR patterns of somatic hybrids of A. thaliana and cabbage for the chloroplast, mitochondrial, and nuclear genomes. (A): Inter-genic region of trnF and ndhJ of chloroplast. (B): Atp1 gene of mitochondria. (C), (D): Nuclear genes present in chromosome I (C; F17F8.14) and chromosome V (D; MZN1.16), A, A. thaliana Columbia, F, Cabbage ‘Fujiwase’, S, Cabbage ‘Chusei Succession’, A + F, Somatic hybrids of Columbia and ‘Fujiwase’, A + S, Somatic hybrids of Columbia and ‘Chusei Succession’, M, Molecular weight marker (φX174/ HaeIIIdigest) work as a codominant marker, whereas in the latter it Table 4. Flower size and pollen fertility of the somatic hybrids A. thaliana may function as a dominant marker (Table 3). All the between Columbia and cabbage ‘Chusei- Succession’. somatic hybrids of the two combinations using ‘Fujiwase’ and ‘Chusei Succession’ produced DNA Hybrids and parental Petal length Petal width Pollen fertility fragments identical to Columbia (Table 3). The results species (mm) (mm) (%) indicated that somatic hybrids contain the whole genome A + S-1 8.6 5.4 4.0 of A. thaliana. Interestingly, a difference in DNA A + S-2 8.8 5.9 6.5 amplification between the two cabbage parents was A. thaliana Columbia 1.2 0.8 93.1 observed in the two DNA markers. For example, in PCR B. oleracea ‘Fujiwase’ 12.1 6.9 94.8 for the MZN1.16 gene on chromosome 5, ‘Fujiwase’ produced a DNA fragment of a different size to Columbia; however, ‘Chusei Succession’ failed to amplify any fragment. The differences between the parental varieties were exactly reflected in the PCR patterns of the hybrids (Table 3 and Fig. 1).

Flower morphology and pollen and seed fertility of the somatic hybrids between Columbia and ‘Chusei Succession’ Since somatic hybrids between Columbia and ‘Chusei Succession’ flowered in the greenhouse, their flower morphology and pollen fertility were observed. The flower size was quite different between the parental species (Table 4), and the two hybrid plants had petal Fig. 2. Flower morphology of somatic hybrid (center) between sizes that resembled cabbage more closely than A. thaliana (left) and ‘Chusei Succession’ (right). A. thaliana. On the other hand, the hybrid possessed white petals like A. thaliana (Fig. 2). Although flower species (Table 4). Whereas both the parental species morphology was intermediate between the two parents, showed pollen fertility levels higher than 90%, those of pollen fertility was drastically lower than the two parental the somatic hybrids were below 10%, indicating that the J. Japan. Soc. Hort. Sci. 77 (3): 277–282. 2008. 281

Table 5. Seed fertility of crosses between somatic hybrids and our report with the combination of A. thaliana and B. oleracea. R. sativus (Yamagishi and Glimelius, 2003). This is Hybrids as Pollen No. of flowers No. of B/A another example of somatic hybrids with all the female parents parents pollinated (A) seeds (B) (%) chromosomes of A. thaliana. We did not examine the A + S-1 Fujiwase 119 1 0.84 presence of the chromosomes of B. oleracea and we do A + S-2 Fujiwase 60 1 1.67 not know whether the hybrids are symmetric; however, A + S-1 Kairan 148 0 0 the hybrids possessed DNA bands of the same size as A + S-2 Kairan 313 3 0.96 those of cabbage, when the cabbage parents had a unique A + S-1 Kohlrabi 1671 9 0.54 fragment. Thus, it was suggested that somatic hybrids A + S-2 Kohlrabi 469 3 0.64 also contain cabbage chromosomes in which the counterpart genes of A. thaliana are present. Further- more, progenies of the hybrids will possess the whole hybrids were practically all male sterile. chromosome set of cabbage after recurrent back-crosses The somatic hybrids were crossed with three variaties with cabbage. By this method, we will obtain progenies of B. oleracea (Table 5). Except for the combination of that have part of the genome of A. thaliana in addition ‘A + S-1’ and ‘Kairan’, at least one seed was obtained to the complete genome of cabbage. We have obtained by pollination. The seed fertility expressed by the the first back-crossing generation, which is the basic proportion of the seed number to the number of flowers material to establish such chromosome addition lines by pollinated ranged from 0% to 1.67%. As a whole, the back-crosses undertaken hereafter. cabbage parent ‘Fujiwase’ provided slightly higher seed While seeds were obtained from crosses of somatic fertility than ‘Kairan’ or ‘Kohlrabi’ (Table 5). hybrids with cabbage etc., the hybrids generally showed low seed fertility with B. oleracea. The poor seed fertility Discussion could be derived either from low female fertility of the Following the first success in somatic hybridization hybrids or from remote crosses between the hybrids and between A. thaliana and cabbage (Yamagishi and B. oleracea, or both. Observation of the changes in seed Nakagawa, 2004), this second report used another variety fertility and the genome structure along with continuous of cabbage. Cabbage is one of the most important back-crossing with cabbage will help to clarify the vegetables in the world, having the oldest history of genetic mechanisms affecting seed fertility. cultivation (Chiang et al., 1993). Because of this, The most interesting trait of hybrids from the breeding work on cabbage has been conducted on a viewpoint of plant breeding is male sterility. The range of traits and by various breeding methods. Not frequency of pollen stained with acetocarmine was only intra-specific crosses but interspecific and around 5%, but the hybrids did not produce any seeds intergeneric hybridizations have been applied for genetic at all by spontaneous self-fertilization (data not shown). improvement of this vegetable. Recently, somatic cell Thus, hybrid plants were estimated to be practically male fusion, especially to obtain intergeneric hybrids, was sterile, even though the pollen function for fertilization attempted with Moricandia arvensis and R. sativus was not examined strictly. Furthermore, the genetic (Kameya et al., 1989; Motegi et al., 2003; Toriyama et causes of low pollen fertility are uncertain at present, al., 1987). This report is another example of intergeneric similar to low seed fertility. However, it could be and intertribal somatic hybridization. Somatic hybrids noteworthy that all somatic hybrids with two combina- are useful materials for genetic improvement of cabbage tions were recombinants or hybrids for the mitochondrial and other kinds of vegetables belonging to B. oleracea. genome (Table 3 and Fig. 1). It is well known that From another aspect, somatic hybrids are also useful for mitochondrial genes, especially chimeric genes, cause genetic and molecular studies on the effects of nuclear male sterility in various plants (Hanson and Bentolila, and organellar genes on various traits of cabbage, 2004); therefore, recombination in the mitochondrial because A. thaliana is a model plant for genetics and genomes between A. thaliana and cabbages may have molecular biology whose genome has been completely induced male sterility. Because the cytoplasmic male sequenced (The Arabidopsis Genome Initiative, 2000). sterility used in Brassica is restricted to that of Ogura By assessing somatic hybrids with DNA markers so far, the new mitochondrial genome produced by cell specific to each chromosome of A. thaliana, it was found fusion could provide us with a new male-sterile that all four hybrid plants of the two combinations cytoplasm for F1 hybrid breeding, which is highly contained all five A. thaliana chromosomes. Because of effective in Brassica and Raphanus. On the other hand, the phylogenetic remoteness of A. thaliana and Brassica the hybrids obtained by us contained solely the species, previous experiments sought to obtain asym- chloroplast genome of cabbage, indicating that after metric hybrids rather than symmetric hybrids (O’Neill back-crossing with cabbage, we could establish cabbage and Mathias, 1995; Siemens and Sacristán, 1995). As a plants with a recombined mitochondrial genome together result, no investigations have observed all the with the nuclear and chloroplast genomes of cabbage. chromosomes of A. thaliana in somatic hybrids except If male sterility is maintained in such plants, they could 282 H. Yamagishi, S. Nakagawa, D. Kinoshita, A. Ishibashi and Y. Yamashita be important materials for F1 hybrid breeding of cabbage Kameya, T., H. Hanzaki, S. Toki and T. Abe. 1989. Transfer of and other B. oleracea vegetables. radish (Raphanus sativus L.) chloroplast into cabbage Brassica oleracea In conclusion, we obtained new somatic hybrids ( L.) by protoplast fusion. Japan. J. Genet. 64: 27–34. between A. thaliana and cabbage ‘Chusei Succession’. Lan, T. H., T. A. Delmonte, K. P. Reischmann, S. Kowalski, J. Since the hybrids flowered, they were pollinated with McFerson, S. Kresovich and A. H. Paterson. 2000. EST- three cultivars of B. oleracea and produced seeds, enriched comparative map of Brassica oleracea and although the hybrids were practically male sterile. The Arabidopsis thaliana. Genome Res. 10: 776–788. progenies of the hybrids by back-crossing with cabbage Motegi, T., I. S. Nou, J. Zhou, A. Kanno, T. Kameya and Y. could be useful materials for breeding cabbage, Hirata. 2003. Obtaining an ogura-type CMS line from especially for cytoplasmic male sterility. asymmetrical protoplast fusion between cabbage (fertile) and radish (fertile). Euphytica 129: 319–323. Literature Cited O’Neill, C. M. and R. J. Mathias. 1995. Somatic fusion for the transfer of agronomically important traits into Arabidopsis Bauer-Weston, B., W. Keller, J. Webb and S. Gleddie. 1993. thaliana from Brassica spp. Proc. 9th Int. Rapeseed Cong: Production and characterization of asymmetric somatic 1100–1112. hybrids between Arabidopsis thaliana and Brassica napus. Siemens, J. and M. D. Sacristán. 1995. Production and Theor. Appl. Genet. 86: 150–158. characterization of somatic hybrids between Arabidopsis Bohman, S., M. Wang and C. Dixelius. 2002. Arabidopsis thaliana and Brassica napus. Plant Sci. 111: 95–106. thaliana-derived resistance against Leptosphaeria maculans Sillito, D., I. A. Parkin, R. Mayerhofer, D. J. Lydiate and A. G. in a Brassica napus genomic background. Theor. Appl. Genet. Good. 2000. Arabidopsis thaliana: a source of candidate 105: 498–504. disease-resistance gene for Brassica napus. Genome 43: 452– Chiang, M. S., C. Chong, B. S. Landry and R. Crête. 1993. 460. Cabbage. p. 113–155. In: G. Kalloo and B. O. Bergh (eds.). The Arabidopsis Genome Initiative. 2002. Analysis of the genome Genetic improvement of vegetable crops. Pergamon Press, sequence of the flowering plant Arabidopsis thaliana. Nature Oxford. 408: 796–815. Forsbeg, J., C. Dixelius, U. Lagercrantz and K. Glimelius. 1998a. Toriyama, K., K. Hinata and T. Kameya. 1987. Production of UV-dose dependent DNA elimination in asymmetric somatic somatic hybrid plants, ‘Brassicomoricandia’, through proto- hybrids between Brassica napus and Arabidopsis thaliana. plast fusion between Moricandia arvensis and B. oleracea. Plant Sci. 137: 65–76. Plant Sci. 48: 123–128. Forsbeg, J., U. Lagercrantz and K. Glimelius. 1998b. Comparison Yamagishi, H., M. Landgren, J. Forsberg and K. Glimelius. 2002. of UV-light, X-ray and restriction enzyme treatment as tools Production of asymmetric hybrids between Arabidopsis in production of asymmetric somatic hybrids between thaliana and Brassica napus utilizing an efficient protoplast Brassica napus and Arabidopsis thaliana. Theor. Appl. Genet. culture system. Theor. Appl. Genet. 104: 959–964. 96: 1178–1185. Yamagishi, H. and K. Glimelius. 2003. Somatic hybrids between Gleba, Y. Y. and F. Hoffmann. 1980. “Arabidobrassica”. a novel Arabidopsis thaliana and cytoplasmic male-sterile radish plant obtained by protoplast funcion. Planta 149: 112–117. (Raphanus sativus). Plant Cell Rep. 22: 52–58. Hanson, M. R. and S. Bentolila. 2004. Interactions of mitochondrial Yamagishi, H. and S. Nakagawa. 2004. Somatic hybrids between and nuclear genes that affect male gametophyte development. Arabidopsis thaliana and cabbage (Brassica oleracea). J. Plant Cell 16 (Suppl.): s154–s169. Japan. Soc. Hort. Sci. 73: 319–323.