The letter “M” in the names of material lines refers to the origin of a particular microspore culture. For example, the line HwiM-2 is the second strain derived from the microspore culture of a leading cultivar Hwiparam. The line CR291M was selected for resistance against clubroot and a virus, following microspore culture of a hybrid with BR079 x ECD-4 (IT number: 04-33-3) to target clubroot and with 3M-291 to induce virus resistance. All lines incorporating CR291M in these experiments, such as CR291M-64, therefore have resistance to two diseases.

Three F₁ Brassica, C218M-13 x HagamM-50, CC507 x BulM-68 and CR291-7M-64 x HwiM-2 had sown for this trial with KB-68 x WY-25 Raphanus first. Female parent plants were B. rapa ssp. pekinensis. Because inbred intergeneric crosses may be sterile [27], F₁ crosses were sown to produce offspring similar BB#12. BB#12 was developed using both female and male F₁ hybrids. The seed of the combinations including CR291M-64 x HwiM-2 was produced in mutual crossing by bees in small net cages that were 2 m long × 1 m wide × 2 m high. Two cultivars of ssp. rapifera (turnip) were included in dominant investigation later.

The KB-68 x WY-25 radish has round, red, fleshy roots and purple leaf veins. It was employed to understand the segregation pattern of the flesh color in the intergeneric hybrid, because both parents have red flesh, although the female Chinese cabbage are F₁ hybrids. Ovule culture was conducted 10 days after hybridization between Brassica and Raphanus. However, a flower branch of CR291M-64 x HwiM-2 was unexpectedly maintained for approximately 25 days, and the seed pods appeared to grow well. Because many of the ovules had already been cultured at that time, ovule culture was stopped to assess mature seed production. Eventually, the CR291M-64 x HwiM-2 hybrid produced mature seed from its cross with Raphanus. To our knowledge, this is the first report of mature seed production in an F₁ Brassica. Thirty-six plants (from 64 hybrid seeds) appeared to have similar morphology, including purple veins and leaves in the growing point area, regardless of their production from the hybrid CR291M-64 x HwiM-2, with different traits in Brassica. A marker test for similarity was requested from Seoul National University.

The combination of CR291M-64 x HwiM-2 and its parental inbred, together with crosses of reciprocal hybrids of both parents and two accessions of Raphanus, KB-68 x WY-25 and locally inbred, were sown to investigate the dominance relationships and the effect of radish in mature seed. Because dominance and no effects were recognized in this experiment, we established 11 further plantings to confirm the dominance relationships and no influence of radish already observed; we also obtained additional information concerning the production of mature seed. These experiments deployed two crosses from each of the CR291M-64 and CR291M-96 strains, five fraternal lineages (CR291M-2, CR291M-5, CR291M-10, CR291M-66, and CR291M-96), Shogoin turnip (introduced from Japan), and Kangwha turnip (our line). Two F₁ combinations, KB-68 x WY-25 and TBM-48 x BDM-7, were sown together with an inbred Shogoin radish to investigate radish effects. The synthesis of TBM-48 x BDM-7 has white flesh and short, fat roots. The inbred Shogoin radish had completely white skin and flesh, as well as a slender root. This appearance was therefore quite different from the appearance of the KB-68 x WY-25 hybrid. Thirty-one radish varieties were sown on August 20 for an autumn test. These varieties were secured and hybridized with CR291M-64 to investigate their ability to produce mature seed.

In total, 432 hybrid non-dry seeds were sown by the end of December. Of these, 328 germinated plants grew by the end of the following June and were able to self-pollinate. The 305 grains of dry seed obtained from CR291M-64 x HwiM-2 were sown to examine the differences between non-dried and dry seeds. Some large seeds were observed during sowing, but they were excluded from analysis because non-dried seeds could not be distinguished by seed size and every plant exhibited purple central veins and leaves. Some green plants germinated from these large seeds primarily; because they were all similar in appearance (distinct from hybrids crossed with radish at the five-leaf stage), a marker test was requested from Seoul National University.

Seeds of Chinese cabbage and radish were sown in a fall crop on August 10 and 20, respectively. Some plants from this sowing were chosen for seed production at the adult stage. For analysis of absolute seed production only, seeds were sown and vernalized under natural conditions from September to February of the following year. If seeding was requested from March to June, the seedling tray was incubated in a refrigerator at 4˚C - 5˚C for at least 45 days to prevent de-vernalization in summer. Sowing in July and August was postponed until September seeding time.

Temperatures rose to 35˚C on some days in July and August, but generally fell below 25˚C at night. Beginning in mid-November, heating facilities were used to prevent exposure of plant materials to nighttime temperatures below 5˚C. The plants were therefore grown at temperatures of 5˚C - 25˚C from October until the following February, with minimum temperatures of about 15˚C from March to June. If vernalization was needed, nursery-stage plants were moved to unheated plastic greenhouses and maintained at minimum temperatures of 2.0˚C - 4.0˚C and a maximum temperature of 12.0˚C by ventilation. All stamens of Brassica were removed within 1 day of flowering. They were then hybridized immediately with radish pollen and covered with an oiled paper bag. The pollinated branches were usually harvested at 35 days after mating; however, they were harvested 45 days after mating during winter, regardless of heating system usage. All other management protocols followed the standard practices of the BioBreeding Institute.

The numbering systems of chromosomes 1 and 5 were used for marker investigation in radish. Thirty plants were collected from the BioBreeding Institute for marker investigation. Single nucleotide polymorphisms of KB68 (KB) and Wonyeon25 (WY) radish were discovered using the Genome Analysis Toolkit (version 3.6-0) HaplotypeCaller [35]. Genomic sequences of WK10039 in radish [36] were used as reference genomes. Variant filtration was performed manually based on read depth. For CAPS marker design, digestibility of flanking regions from filtered single nucleotide polymorphism sites of two cultivars was examined to identify candidate regions for CAPS markers. For each pair of sequences, if only one sequence was the target site of the restriction enzyme Hind III (AAGCTT), primer sets were designed around the target region. Two primer sets were designed to distinguish hybrids from KB68 and WY25 strains. Polymerase chain reaction was performed using 20-μl reactions containing 200 ng of genomic DNA, 4 units of Taq polymerase (Takara), 0.2 mM dNTPs, and 0.2 mM primers, combined with 1× polymerase chain reaction buffer (Takara). Polymerase chain reactions were performed under the following conditions: denaturation at 95˚C for 5 min followed by 30 cycles of amplification (95˚C for 30 sec, 60˚C for 30 sec, and 72˚C for 1 min) and a final extension at 72˚C for 10 min. Amplicons were digested with Hind III (New England Biolabs) and visualized with 2% agarose gel electrophoresis. This procedure was generally similar to the protocol used in the radish marker test on dry-seed-derived green plants of Chinese cabbage CR291M-64 and HwiM-2, although it used standards CR (CR291M-64), Hwi (Hwi-M2), BF1 (CR x Hwi), KB (KB-68), WY (WY-25), and RF1 (KB x WY). The line Chiifu-401 [37] was used as the reference genome. Four primer sets were designed to distinguish CR291M-64 and HwiM-2, using 3 purple and 9 green plants.

CR291M-64 x HwiM-2 hybrids produced pods that were 3.5 - 5.3 cm in length by day 37 and the others not produced seeds (Figure 1). Sixty-four seeds from 11 pods were placed on filter paper in Petri dishes (Table1). Seeds sown on filter paper generally germinate within 2 - 3 days at 25˚C; in this trial, germination began on day 9 and continued for 33 days (Supplementary TableS1). In addition, the germinated plants appeared fragile. They were therefore kept in the Petri dish for several more days before transplantation into soil. Retardation and extension of germination were presumably caused by seed dormancy that was found in the sowing of the dry seed.

Individuals subjected to marker analysis at Seoul National University were genetically differentiated and became morphologically dissimilar as they grew into adult plants. In a dominance trial, the combination of CR291M-64 x HwiM-2 generated mature seeds. The parent CR291M-64 plants also produced seeds, but the cross counterpart HwiM-2 did not produce seed. Mature seeds from the CR291M-64 line were definitively dominant. Because the line HwiM-2

was recessive, a CR291M-40 x HwiM-2 hybrid was used to generate dominant seeds. The strain CR291M-180 was uncertain because it was crossed with the dominant line CR291-64. The HwiM-2 x CR68M-107 hybrid failed to produce seeds; so, the CR68M-107 line was recessive, which derived from BulM-68 through microspore culture for virus resistant line after hybridization with 3M-291 (Table2). Four combinations and five inbred lines, including two turnip strains were mated with one, two, or all three radish cultivars; they produced mature, dominant seed. Since the line of C218M-13 is recessive, CR291M-96 is a dominance stain: that is as the line test (Table3). On the other hand, the CR291M-64 line produced intergeneric mature seed without exception when hybridized with 31 radish strains cultured for fall cropping (Supplementary TableS2). It has known that radish is no influence on mature seed production of intergeneric hybrid between Chinese cabbage and radish.

Two hundred and fifty-four of 305 dry seeds germinated without delay and continued growing for an extended duration. Among the germinated plants, 25 that originated from larger seeds were green and grew more vigorously than did purple plants that sprouted concurrently (Figure 2). The results of marker examination indicated that the green plants were induced from self-pollination or apomixis of the CR291M-64 strain. Three purple individuals were true hybrids

*(05-80-14B): Introduction number.

crossed between inbred CR291M-64 and a KB-68 x WY-25 radish cross. Therefore, every item mentioned so far has turned out to be false except that the inbred CR291M-64 produced an intergeneric hybrid seed with dominance. The hybrid is thought to have resulted from the two being in the same net cage with bees. Weak self-incompatibility may have been responsible. It was identified that another experiment carried out after the marker test (Supplementary TableS3). The 328 plants sown from non-dried seeds and 254 plants grown from dried seed were all purple and produced purplish flowers, a result not previously reported in cruciferous intergeneric hybrid crops (Figure 3). Unusually, they did not produce seeds within more than about 100 days of pollination after blooming. Some of the 36 plants initially seeded produced microspore-derived embryos; the offspring of approximately 150 plants did not generate seeds in self- or cross-pollination involving more than 100 flowers each [27].

Seoul National University prepared two CAPS markers of radish and tested 30 of 36 hybrids. Two markers of KB-68 were 562 and 786 base pairs long. In WY-25, markers were 414:148 and 261:525 base pairs long when cut with Hind III. The individual numbers of plants 1, 2, 6, and 8 of the two markers were KB (KB-68): WY (WY-25), WY: KB, KB: KB, and WY: WY, respectively (Table 4). These plants clearly all differed from each other. Seoul National University also investigated four CAPS markers prepared on chromosomes 4, 6, 8, and 10 in the maternal parent CR291M-64; these markers were analyzed in 25 green plants. The line of HwiM-2 was cut off at the marker of chromosome 6, and strain CR291M-64 was cut off at all other markers of three chromosomes (Table 5). Nine green and three purple plants had the same CR291M-64 genotype, including CR291M-64 and BF1 (CR291M-64 x HwiM-2) (Figure 4). However, HwiM-2 had different genotypes of the CR291M-64 x HwiM-2 hybrid. Two

*KB: KB-68, WY: WY-25, 0ne and 5: number of chromosomes, F: forward, R: reverse. *Two markers of KB-68 were 562 and 786 base pairs long. In WY-25, markers were 414:148 and 261:525 base pairs long when cut with Hind III.

previously identified radish markers were not present in the nine green plants, although they were expressed in the three purple individuals containing RF1 (KB-68 x WY-25) without BF1 (CR291M-64 x HwiM-2) (Figure 5).