Perspective Chapter: Creation and Evolution of Intergeneric Hybrids between Brassica rapa and Raphanus sativus

Abstract

Although research has been conducted on intergeneric hybridization between Brassica and Raphanus, much of it remains unpublished. We have acquired numerous Brassica rapa ssp. pekinensis (kimchi cabbage) and R. sativus var. major (“big root radish”) hybrids, originally classified as intergeneric hybrids and named “baechumu” in 1995. A cultivar was identified BB#12, (renamed BB#1 for registration) in baemoochae following stabilization via a microspore mutation in 2006. Numerous hybrids were created for various purposes; some were sterile when self-pollinated but fertile in crosses with other cultivars. Microspore mutation also produced, BB#12x is a novel intergeneric hybrid. A new stable plant variety, BB#5, was selected from numerous inbred lines and produced via microspore culture; it has a very late bolting time and is cultivated in spring. The cultivar purple BB#10 was developed by adding radish chromosomes to turnip, including one providing the purple color, and double-crossing with BB#12, CMS BB#12, and normal BB#12. Now that the hybrid between ssp. pekinensis and radish has produced mature seeds as a dominant property, intergeneric hybrid cultivars can be bred in the future.

1. Introduction

There are two main kinds of intergeneric hybrids involving Raphanus: Raphanus and Brassica, and Brassica and Raphanus. Hybrids between Raphanus (radish) and Brassica oleracea var. (cabbage) were bred by Sageret [1] (cited from Prakash et al. [2]), Karpechenko [3], and McNaughton [4] and stabilized by Chen and Wu [5].

Our laboratory began conducting studies on intergeneric hybridization between kimchi (Chinese) cabbage (B. rapa ssp. pekinensis) and radish (R. sativus var. major) in 1986 [6], and research has been ongoing ever since (i.e., for 36 years). Originally, this was classified as an intergeneric hybrid and was named “baechumu” after a successful microspore culture in 1995 [7]. In 1997, the crossbreed was renamed again as “baemoochae” (where “bae” is from “baechu”, [kimchi cabbage], “moo” is from the Korean for radish [i.e., “mu”], and chae is from “chaeso” [vegetable in Korean]) [8]. Baemoochae was stabilized in 2006 [9]. In total, 25 papers on baemoochae have been published or accepted, 17 of which are either from our laboratory or list me as an author (four of these are in Korean). However, these papers do not represent all of the work that we have done, and some are available only in Korean.

Terasawa [10] was the first to report intergeneric hybridization between Brassica and Raphanus, followed by Takeshita et al. [11], Dolstra [12], Lange et al. [13], Lee et al. (1986–2022), and members of the laboratories of Professor Hyo Guen Park [14, 15], Professor Jongkee Kim [16, 17, 18], Professor Il-Sup Noh [19], Professor Jun Gu Lee [20], Professor Hyun Hee Kim [21, 22], and Professor Jin Hue Huh [23, 24]. Tonosaki et al. [25] in Japan, and Lou et al. [26], Zhang et al. [27], and Jin et al. [28] in China, have published papers on hybridization between Brassica and Raphanus. These studies written in English are briefly introduced and the Korean papers are discussed in detail; previously unpublished photographs and tables are also provided.

2. Development and application of an ovule culture system

To obtain a hybrid between Brassica and Raphanus, Brassica should be the maternal parent [14, 29]. The first successful Brassica-Raphanus hybrid seeds were acquired by Terasawa [10]. However, the hybrid seeds were between B. rapa ssp. chinensis and R. sativus, not ssp. pekinensis. Dolstra collected a wide range of varieties and used B. rapa, ssp. rapifera (turnip), ssp. oleifera (turnip-rape), ssp. chinensis (Chinese mustard), and ssp. pekinensis (Chinese cabbage) as female parents [12]. However, he could not obtain seeds from ssp. Pekinensis, but did obtain them from the other three subspecies. B. rapa ssp. pekinensis seemed to have a characteristic preventing the development of mature seeds. Lange et al. published a paper on Brassica-Raphanus hybrids in 1989. Most of Dolstra’s research focused on the creation of Brassica-Raphanus hybrids, with little discussion of future directions.

Takeshita et al. [11] attempted to germinate a hybrid seed between ssp. pekinensis and R. sativus using a culture of young ovules, but was not successful. Successful germination was reported by Been and Park [14], but the mature plant did not produce any seeds. Subsequently, a student of Professor Hyo G. Park studied an ovule culture to increase the number of germinating ovules [15]. Several sprouts were observed from one ovule, and the addition of 0.1 mg each of benzyl adenine (BA) and naphthalene acetic acid (NAA) to 1 L B₅ medium increased the number of plants (Table 1).

Amount (mg/L) of NAA and BA add to 1 L B5 medium	Cultured ovules	Calluses	Shoots	Roots and shoots	Plantlets established
0.1	93	5	24	4	27
0.5	92	0	2	0	0
1.0	98	5	0	0	0

Table 1.

Effects of NAA and BA on excised intergeneric ovules in hybridization between Kenshin (Brassica) and Jinju daepyung (Raphanus) performed in 1985.

On the basis of these results, the Horticultural Experiment Station (HES) cultured ovules of intergeneric hybrids between ssp. pekinensis and Raphanus on the modified B5 medium in 1986 and 1987 [30]. All three cultivars of kimchi cabbage and radish produced intergeneric hybrids, and BA, NAA, and 2, 4-dichlorophenoxy acetic acid (24-D) combined with 8- or 24-h dark treatment per day did not greatly influence the ovule culture. Of the 676 ovules cultured, 102 (15.1%) produced plants, including 22 multi-shoot embryos. Of the successfully germinated plants, 22 were harvested and 439 seeds were collected. This study was the first to harvest F₂ seeds from Brassica-Raphnus hybrids [6].

Two F₁ cultivars, Jeonsueng (Brassica) and Taeback (Raphanus), which were included in the above-described study, were used for subsequent experiments conducted in 1987–1990 [8]. In total, 1893 ovules were cultured: 1250 for 2n × 2n and 643 for combinations of 2n × 4n, 4n × 2n and 4n × 4n (Table 2). Once the seeds had been harvested, 4–5 seeds were sown according to their ploidy level. They were germinated evenly with 3–5 plants; 2n × 2n resulted in plants with a normal appearance, while 2n × 4n, 4n × 2n and 4n × 4n plants were mostly albino, for reasons that remain unclear. Notably, subsequent intergeneric hybrids of Brassica and Raphanus were all obtained by ovule culture.

Crop	Cultured ovules	Germinated ovules	Plants (including multi-shoot ovules	Pollenating plants	seeded
Kimchi cabbage	1.893	391 (20.7%)	Total 591 (31.1%); Tested 466 (24.6%)	98 (5.2%)	65 (3.4%)

Table 2.

Data from the experiments of the HES conducted in 1987–1990.

Medium: B5 + NAA 0.1 ppm + 24-D or BA 0.1 ppm.

(125.6%) ovules or plants died.

3. General characteristics of intergeneric hybrids

The morphology of the hybrid plants is intermediate between the two parents. The seed pod has a narrow septum at the center dividing the top and bottom parts (Figure 1). The lower part that attaches to the stem (bottom) is entirely kimchi cabbage, and the upper part (top) is entirely radish. This also applies inside the pod [8]. However, the seed appearance is indistinguishable between the kimchi cabbage and radish portions [13, 31]. The siliqua morphology seems to be a distinctive characteristic of the intergeneric hybrid between Brassica and Raphanus.

Figure 1.

Leaf and pod morphology of an intergeneric hybrid (shown between its parents).

Seeds were harvested from plants cultivated in the fall. The general characteristics of baemoochae plants are as follows ([31], Table 3). The leaves resemble those of a radish and have many lobules and a robust appearance. Baemoochae plants have only a few leaves (~30), with very large petioles (~10 cm in circumference). The petiole is white and round, differing from the parent cultivar; kimchi cabbage has a broad white petiole, and radish has a thin, green, circular petiole. The heading ability is very low. The roots are small, but the midportion bulges (similar to radish) before stabilization [31].

Line code ^Z	Head (g)	Plant weight (kg)	Leaf length (cm)	No. of leaves (each)	No. of lobules (each)	Petiole circum-ference (cm)	Root length (cm)	Root width (cm)	Root (g)
BB#1	900 loose	4.7	45	30	22	10.0	18.0	3.5	210

Table 3.

Characteristics of baemoochae grown in the fall.1

¹

Direct sowing on August 14, observed on November 7.

The flower of intergeneric hybrids is typical of Cruciferous plants, i.e., it is generally white [31]. The plants with yellow flowers presented in the cultivation of BB#6 (a novel unstable line). The yellow stock has not been registered.

Progeny (F₂) of the ssp. pekinensis and R. sativus hybrids were attained first. The average number of seeds produced was less than one per pod, similar to the results of Dolstra [12]. However, every pod reached maturity despite the empty husk [31]. The seed size differed even within the same year. The random weight of 1000 seeds of kimchi cabbage is approximately 3.5 g, compared with 7.0 g for the intergeneric hybrid and 14.0 g for radish. The number of seeds per milliliter is ~200 for kimchi cabbage, ~120 for the intergeneric hybrid, and ~50 for radish; seed vigor also differs [31].

The total number of ovules produced by the hybrid is 10–12, which is less than the average of 25 produced by kimchi cabbage and more than the 5–7 produced by radish. Of the seeds produced, 7–8 form the bottom kimchi cabbage part; the remaining 3–4 form the top radish part [31].

4. Development of a dihaploid production system

Eleven plants were randomly chosen to produce pollen, including OV115C, which was obtained by ovule culture and colchicine treatment; it was exposed to anther culture in 1986 and embryos appeared in six plants [6]. Eighty-four embryos from 20 anthers germinated. Fifty anther-derived plants flowered and 14 produced self-fertilized seeds. The number of chromosomes in the pollen of the mother cells of these 50 individuals indicated that there were 5, 30, 15 allodiploid, (n = 2x = 19), allotetraploid, (2n = 4x = 38), and allooctaploid (4n = 8x = 76) plants, respectively.

When individuals with different ploidy levels were cultured, more embryos appeared in allooctaploid (38II, 2n = 8x = 76) than allotetraploid (2n = 4x = 38) [pp. 56–67 of a 1998 report from the HES [32] (Table 4)]. These results can serve as a useful reference for future anther or microspore cultures.

Line code	Ploidy	No. of anthers		No. of embryos
Code		Inoculation	Germination
OA 15	19 I	300	0	0
OA 20	19 II	300	2	4
OA 31	19 IV	300	14	131

Table 4.

Numbers of embryos formed from anther cultures derived from OV115C with different ploidy levels.1

¹

OV: ovule culture after hybridization. C: colchicine treatment. OA: anther culture.

Microspores were cultured from the OA-20 line, which originated from the OV115C anther culture [7]. Of the 114 embryos, only 14 plants germinated, although a lot of calluses were present on these plants. The microspore culture successfully produced an intergeneric hybrid between kimchi cabbage and radish. A washing solution composed of B5 and NLN medium was more effective than a reduced half-concentration mixture, as was a density of 100,000 microspores over 72 hours than 50,000 over 24 hours or 200,000 over 120 hours (at 32.5°C).

To determine the correlation between ploidy and germination, the chromosomes at the root tip of plants and chloroplasts within the guard cell were counted. Haploid plants have about 10 chloroplasts, while diploids have about 14, tetraploids about 22, and octoploids about 36. Diploids were the most commonly germinated embryo, accounting for nine of the 14 plants.

Results from unpublished work showed that adding a liter of NLN medium containing 0.1 mg of BA to the microspore culture reduced the incidence of callus formation and increased embryo yields by an average of 6.8 per Petri dish. To acclimate the derived hybrid, B₅M₂-II medium, which comprises 400 mg/L of KCl and 600 mg/L of CaCl₂2H₂O, was added to the B₅ basal medium, resulting in an increase in the plant survival rate from 24% in the MS2 medium to 75% (Table 5). Comparing the stable and unstable lines, the BB#12-stabilized line generated many more embryos, demonstrating that stabilization is an important factor in the production of embryos in microspore cultures (Table 6). In another unpublished experiment, acclimation using a microponic system was performed repeatedly, and a plant survival rate of almost 100% was obtained when plants were acclimated for 20–30 days. This represents an important finding with respect to the microspore culture process (Figure 2). These improved microspore culture techniques have been providing excellent results for the baemoochae experiments conducted in our laboratory.

Medium	Numbers of embryos		Regenerated plants
	Transplanted	Died	Abnormal	Transplantable
MS2	102	40 (40%)	38 (37%)	24 (24%)
MS4N	90	53 (59%)	21 (23%)	16 (18%)
MS4K1	93	37 (40%)	12 (13%)	45 (48%)
B5M2-II	186	4 (2%)	43 (23%)	140 (75%)

Table 5.

Effects of plant medium on the regeneration of microspore-derived embryos of baemoochae (xBrassicoraphanus koranhort).1

¹

Cultivar: BB#12 (stabilized).

MS2: 2% sucrose in MS medium, as recommended by Keller and Armstrong (1979).

MS₄N: NH₄NO₃ was reduced to 550 mg/L from 1900 mg/L in the MS medium, with 2% sucrose.

MS₄K₁: BA (0.1 mg/L) and NAA (0.2 mg/L) were added to the NH₄NO₃-free MS medium, with 2% sucrose.

B₅M₂-II: 400 mg/L of KCl and 600 mg/L of CaCl₂2H₂O were added to the B5 basal medium, with 2% sucrose.

Line code	Genetical status	Embryos/Petri dish
		Average of 30	Maximum of 10
BB#4	Unstable	0.8	2.6
BB#12	Stable	27.0	56.8

Table 6.

Genotype specificity for embryo induction in microspore culture for unstable and stable baemoochae lines (xBrassicoraphanus koranhort).1

¹

Each culture was conducted three times in April, in 10 Falcon Petri dishes each time.

Medium: NLN13 (13% sucrose in NLN medium) and BA (0.1 mg/L).

Figure 2.

Baemoochae nurseries acclimated in a microponic system (Dr. Yoon presented). The box of the microponic system was 60 cm long, 37.6 cm wide, and 18.2 cm high, and the top was covered with cellophane with 12 small holes. Thin tubes were connected to a machine to create air bubbles in the bottom of the medium.

5. Taste and nutrition of baemoochae

The nutritional composition of baemoochae derived from microspore culture was analyzed at the Dietary Life Improvement Research Institute, Rural Development Administration, by request of Mr. Moo Kyoung Yoon based on standards for kimchi cabbage and radish (1996, Table 7). Baemoochae showed high nutritional value in both the fresh top-part and underground roots. In total, 14 of 20 elements were overrepresented in the fresh parts and roots [energy, moisture, protein, fat, sugar, fiber (cellulose), phosphorus, natrium (sodium), kalium (potassium), zinc, magnesium, vitamin B1, vitamin B2, and vitamin C]; the remaining six components (ash, calcium, iron, vitamin A and its precursors, β-carotene, and niacin) were genetically dominant [31].

Crop & part		E	M %	P g	F g	S g	C g	A g	c ㎎	P ㎎	I ㎎	n ㎎	k ㎎	z ㎎	m ㎎	A	β-c μg	B1 ㎎	B2 ㎎	Ni ㎎	VC ㎎
BR	R1	46	85	3.7	0.2	8.9	1.2	1.0	33	106	0.2	44	504	0.4	28	0	0	.12	.05	0.6	51
	L	37	88	3.2	0.5	6.1	1.1	1.2	55	59	1.6	52	504	0.1	15	60	360	.14	.13	0.5	99
CC		13	94	1.3	0.2	2.4	0.7	1.5	51	29	0.3	5	230	—	—	9	56	.05	.06	0.3	46
RS	R1	18	94	0.8	0.1	3.8	0.6	—	26	23	0.7	13	213	—	—	8	46	.03	0.02	0.4	15
	L	19	92	2.0	0.2	3.2	1.0	—	249	35	3.0	36	273	—	—	368	2210	.05	.10	0.6	75

Table 7.

General nutritional information (provided by Moo K. Yoon [33]).

BR: baemoochae, CC: kimchi cabbage, RS: radish, R1: root, L: leaf, E: energy (Kcal), M: moiture, P: protein, F: fat, S: sugar, C: cellulose, A: ash, c: calcium, p: phosphate I: ion, n: natrium, k: kalium, z: zinc, m: magnesium, R.E: vitamins A, ß-c, B1, and B2, ni: niacin, VC: vitamin C.

Baemoochae has a pleasant texture (crisp and juicy) and a unique flavor similar to wasabi. The component responsible for the spicy and sweet taste is sulforaphene, which has a very similar chemical structure to sulforaphane; however, sulforaphene has an additional double bond. Jongkee Kim [34] found that sulforaphene had the same anticancer effects as sulforaphane. Additionally, baemoochae juice had the same ability to eradicate Staphylococcus aureus as sulforaphene. These results have encouraged other scientists to investigate the baemoochae glucosinolate to sulforaphene via saliva.

Analysis of contents by part showed that baemoochae BB#6 had 294 μg sulforaphene per g of fresh root when harvested in November ([31], Table 8).

Outer leaf		Middle leaf		Inner leaf		Root
Midrib	Fresh	Midrib	Fresh	Midrib	Fresh
30.8	36.0	191.4	150.3	137.3	268.2	294.3

Table 8.

Content of sulforaphene in various parts of baemoochae BB#4 grown for 80 days in the fall (μg/g FW).1

¹

Mean of three replicates.

The location and cause of the high sulforaphene content in baemoochae were investigated. A small amount was found in the kimchi cabbage portion, but the majority was in the Taebaek radish portion (the male parent of baemoochae cross) [17]. The dry weight (DW) of baemoochae is about 55 μmol/g, which is less than that of Taebaek radish (75 μmol/g DW) but more than that of kimchi cabbage (28 μmol/g DW) [16, 18]. In another experiment that examined nine different crops (cauliflower, cabbage, broccoli, radish, baemuchae, pakchoi, Chinese cabbage, leaf mustard, and kale), the glucosinolate concentrations were lowest in radish tissues and differed widely among varieties [20].

In addition to the high concentrations of anticancer and antibacterial glucosinolate, baemoochae also has a high content of flavonoids, which have antiviral, antihistamine, and antioxidant effects [27].

6. Stabilization and evolution of baemoochae

When broccoli microspores were cultured with 0.01 μmole of n-nitroso n-methyl urethane (NMU), the embryo yield fell to 53%. However, the treatment increased the proportion of pollen-producing individuals to 73% (129 plants) and induced sterility in male plants [35]. Thus, whether NMU treatment increased the number of pollen-producing and sterile male intergeneric hybrid plants was investigated. In an experiment in which the microspore culture was treated with 0.01 μmole NMU, only nine plants produced pollen, with a seed yield of less than 0.5 seeds per pollination in 2005. When every strain was sown again for seed production, all seeds from the pods of four strains of nine lines matured into a greenish color, with the exception of one or two degenerates and the BB#1, BB#4, and OV115C cultivars, which turned brown; only one or two seeds matured in 2006. Unfortunately, after root harvest in the fall of 2007, the tap roots did not have swollen parts like a radish and had rotted inside (Figure 3). Seeds were produced again in 2008 and cultivated in autumn, like the previous year. The tap root was retained, but the root was not rigid and the inside did not rot, as it had the previous year. All lines were the same as before. The root rotting was probably a physiological disorder in the earlier generation of the mutation, although the cause was not clear.

Figure 3.

The appearance of Mi2-generation plants from stabilized seeds of baemoochae in 2007.

Seeds were obtained by hybridization of a reciprocal cross between BB#12 and a new intergeneric hybrid that had not produced any self-seeds, i.e., Jombaechu × Jeku Gaetmoo (04-33-81 × 04-80-8, 9) [36]. Among them, BB#12 was used as a mother line and applied for microspore mutation. One plant was selected and stabilized as a new late-bolting variety, i.e., BB#5 ([21]: [37]).

According to international regulations, the genus name should be xBrassicoraphanus, irrespective of the female and male parents, to commemorate Saqeret, who was the first to announce successful hybridization between Raphanus and Brassica. Species can be named according to future needs [38]. For example, the species name for baemoochae BB#5 was announced as koranhort [37], since this crop originated at the HES in Korea in 1986 and was developed as a stable cultivar at Chung-Ang University in Anseong in 2006. The scientific name of baemoochae is xBrassicoraphanus koranhort Saqeret & Lee (Soo-Seong).

To develop a new baemoochae line having a swollen root like a radish, the cultivar nidomi turnip (07-80-166. Brassica) was first hybridized with coastal south Gaetmoo radish (05-80-45. Raphanus) and subsequently crossed with baemoochae BB#12 to create this hybrid (166 × 45). Although the coastal south Gaetmoo radish had a purple vein, it was ignored since the leaf was green, and it was not clear whether the purple vein would become a purple leaf after several generations. Two plants used for multiplying seeds of turnip × radish were crossed to achieve cytoplasmic male sterility (CMS) of BB#12, i.e., to breed a CMS hybrid. This resulted in a CMS turnip × radish combination hybridized to a normal BB#12 to induce another CMS line of BB#12. Two plants crossed with normal BB#12 of CMSBB#12-11 × (166 × 45) produced only one or two seeds. The other two plants, crossed with BB#12 of CMSBB#12-7 × (166 × 45), produced seeds in amounts of 21.7% (81 seeds) and 20.0% (72 seeds) relative to that produced by the standard BB#12. Some purple plants (16 of 22 cultured) were produced from the 81 seeds mentioned above [39].

Since the seed production of BB#12 × {CMSBB#12-11 × (166x 45)} was unusual, a “marker test” was performed by Seoul National University, along with chromosome detection by Sahmyook University, which showed that the two lines were not genetically crossed. Radish chromosomes were intact before hybridization with the turnip, even though the radish chromosomes had “sandwiched” turnip chromosomes after hybridization. Turnip chromosomes are already intercalated with B genome chromosomes. Therefore, it was inferred that the turnip was intercalated with the radish chromosomes, including those responsible for the purple color. Therefore, the resulting purple cultivar was produced by chance [39].

7. Lack of a commercial F₁ hybrid in baemoochae

As the F₁ hybrid cannot be copied by other companies and plant breeders, growers must purchase these excellent seed varieties every year. To guarantee profit, the company sells a finite amount of seeds per year. Before the baemoochae was stable, the F₁ hybrid was developed for breeding. A male sterile line of mustard (Ogura) was received from Professor Il-Seop Noh of Sunchon National University and used (2004); an attempt was made to induce a baemoochae CMS line. An F₁ plant of baemoochae hybridized with “CMS mustard.” Seven out of 43 BC₁F₁ hybrids were fertile and produced pollen during the next year (Table 9). In total, 38 of 46 plants were fertile, with pollen at BC₂F₁. It was thought that the CMS of mustard would be recovered in baemoochae. Therefore, two generations advanced more, it was stopped (unpublished data). We then attempted to breed a line of baemoochae from a kimchi cabbage with CMS, in a further attempt to induce CMS. However, some plants remained fertile, similar to B. juncea.

Generation	Numbers of
	Plants investigated	Male sterile plants	Fertile male plants
BC1F1	43	7	36
BC2F1	46	38	8
BC3F1	30	18	12
BC4F1	40	22	18
Total	159	85	74

Table 9.

Results of backcrossing experiments with Brassica juncea to induce CMS in xBrassicoraphanus koranhort.

Among crucifers, Brassica napus, B. juncea, and B. carinata are self-compatible amphidiploids. The above pollination results show that this is also the case for baemoochae. Professor Il-Seop Noh, an expert in molecular biology specializing in self-incompatibility, investigated the reason for the self-compatibility of baemoochae. The self-incompatibility factors of both kimchi cabbage (BrRsSRK-1) and radish (BrRsSRK-2) pistils, as well as the pollen from kimchi cabbage (BrRsSP11-1), were functioning in baemoochae. However, radish pollen (BrRsSP11-2) did not function in baemoochae [19]. Therefore, the self-compatibility in baemoochae is due to the pollen from the radish portion of the hybrid. Since CMS is recovered and self-incompatibility does not function in baemoochae, there is no way to produce the F₁ hybrid seed.

8. Chromosome configuration of baemoochae

Investigations of the correlation between low seed yields and multivalent chromosomes in the original intergeneric hybrid from 1986 were conducted; OV115C plants produced abundant pollen and pods via self-pollination, but seed yields remained very low (< 1.0 per pollinated pod). Observations of chromosomes in the pachytene stage from the OV115C plant showed that four units formed a diamond shape, and five chromosome pieces were stacked with the three chromosomes in the cells. Thus, the meiosis itself was abnormal, resulting in lower seed productivity.

A report was published on the breeding of an unstable heading rapeseed, called “Hakuran,” at the Japanese Vegetable and Tea Industry Station in 1968. Our laboratory requested seeds and received a line in 1972. When pollinated for multiplication, the seed yield was low. The low fertility of this interspecific hybrid could be due to multivalent chromosomes, although McNaughton [40] and Dolstra [12] theorized that the low seed yield is due to a “genic imbalance” and “breeding barrier” between radish and cabbage, and Chinese cabbage and radish, respectively.

In 1999, Seon-Jung Lim used genomic in situ hybridization (GISH) to observe homologous chromosomes in baemoochae and seek a crossover. Since the homologous chromosomes were tetravalent (composed of two genera), crossovers in chromosomes were anticipated, but not discovered. However, it was demonstrated that baemoochae is a true hybrid of kimchi cabbage and radish in terms of chromosome constitution. GISH revealed 20 kimchi cabbage and 18 radish chromosomes. The results of this master’s thesis, including the first photograph of baemoochae, were presented at the first Chromosomal Conference, which was held in Shanghai, China, in 2000, and had attendees from Korea, Japan, and China. The paper was subsequently published in a journal [41].

Crossovers occur more frequently in tetraploids (4n = 2n = 38) than diploids (2n = n = 19) of xBrassicoraphanus, but less frequently than in stabilized lines such as BB#12. It is possible that the chromosomal dissimilarity of B. rapa and R. sativus prevents nonhomologous interactions between the parental chromosomes in allotetraploid xBrassicoraphanus, allowing normal diploid-like meiosis when homologous pairing partners are present [24].

Professor Hyun Hee Kim, a cytologist at Sahmyook University, was asked to analyze the chromosome composition of a stable BB#5 in meiosis. Miss Hadassah Roa Belandres, working in Professor Dr. Kim’s laboratory, performed fluorescence in situ hybridization (FISH) using 5S and 45S rDNA, along with GISH using a B. rapa genomic DNA probe [21]. According to the somatic chromosome complement revealed by FISH, baemoochae has 2n = 38, consisting of 17 metacentric and 2 submetacentric chromosome pairs. According to the GISH analysis, 19 bivalents were present in 42% of 100 meiotic cells, while a commination of 1 tetravalent and 17 bivalents was present in 28%, a commination of two tetravalents and 15 bivalents in 24%, and a commination of three tetravalents and 13 bivalents in 6%. She concluded that the 19 bivalents present in 42% of the meiotic cells were the main cause of the stability.

Professor Kim extended the study to investigate abnormal meiosis of pollen within a mother cell from the unstable line BB#4. In total, 500 pollen mother cells of the unstable line BB#4 and stable lines BB#12 and BB#5 were assessed in all five stages of meiosis, from diakinesis to anaphase. In the unstable line BB#4, 57.4% (n = 287) of the dividing cells were abnormal, compared with only 10–12% (n = 50–60) of the dividing cells of two stable hybrids, BB#12 and BB#5 (Table 10).

Line code	Diakinesis			Metaphase		Anaphase		Total (%) (500 cells)
				I	II	I	II
	Sticky	Rod and ring	Laggard	Laggard		Bridge and laggard
BB#4 (unstable)	35	18	22	154	8	49	1	287 (57.4%)
BB#12 (stable)	4	8	12	15	6	5	0	50 (10%)
BB#5 (stable)	6	0	5	17	5	27	0	60 (12%)

Table 10.

Numbers of abnormal meiotic cells in various baemoochae strains.

9. Properties of self-sterile but cross-fertile intergeneric hybrids

Five different hybrids of the intergeneric cross between kimchi cabbage (Brassica) and radish (Raphanus), Jombaechu × Jeju Gaetmoo (04-80-8 or 9), CR291M-64 × Shogoin, Taiwan Baiyu × Shogoin, Taiwan Baiyu × 40 days, and Chibu × Woenkyo#39, were developed for various purposes and none produced self-seeds. However, cross-seeds were produced in four combinations with existing baemoochae or mooyangchae lines. Combinations with Taiwan Baiyu × 40 days, which has no hair on its leaves, are expected; all other existing cultivars have hair on their leaves [36]. Pollen samples from a selection of 30 plants that did not produce self-seeds were stained and observed. Pollen from one plant did not stain at all, while <10% of the pollen of 18 plants stained; > 30% of the pollen of seven other plants was stained, while for two plants each >70% and 90% of the pollen was stained. There appears to be no correlation between pollen dyeing and self-seed production.

BioBreeding has four stable cultivars: BB#12, BB#5, purple BB#10, and mooyangchae. Despite a lack of understanding of the non-self-fertilization seen in new intergeneric hybrids, new baemoochae varieties can be bred by crossing with existing cultivars, such as BB#5 [37]. This could lead to the development of intergeneric hybrids between kimchi cabbage and radish.

10. Production of mature seeds by baemoochae

Hybrids between kimchi cabbage B. rapa ssp. pekinensis and radish initially failed to produce mature seeds ([11]: [12]). Thus, an ovule culture was developed ([14]: [15]: [30]). Cultivars BB#12 and BB#5 were bred by applying this ovule culture technique [9, 37]. Ovule culture is a powerful means of acquiring intergeneric hybrids between kimchi cabbage and radish. Notably, an inbreed of kimchi cabbage did create mature seeds as a dominant property. Radish cultivars and lines have little effect on this property; therefore, the kimchi cabbage is mainly responsible for the production of the mature seeds [39].

The major subspecies of B. rapa, i.e., ssp. rapifera (turnip) [12, 28], ssp. oleifera (turnip-rape) [12], and ssp. chinensis (pakchoi) ([10]: [12]) have also produced mature seeds. It is important to note that the subspecies of Brassica can be dominant or recessive when crossed with Raphanus first. When they are dominant, all combinations can be acquired between and within subspecies, including ssp. pekinensis, and numerous types of intergeneric hybrids can be produced. Ovule cultures older than about 10 days after hybridization are difficult to grow; it is possible that mature seeds are produced in Brassica only when needed.