Introduction

Magnolia obovata is a broad-leaved tree that belongs to the family Magnoliaceae and is native to Japan. The trees themselves are valued by landscape gardeners and the wood is used for furniture and various industrial arts. In addition, the bark is a source of valuable medicinal compounds, such as magnolol and honokiol [1, 2]. Magnolia species can be propagated from seeds and rooted cuttings and by grafting. However, the rate of germination of M. obovata seeds is relatively low (approximately 35%) [3]. Therefore, we postulated that a tissue culture technique might be useful for the propagation of M. obovata.

There have been several reports of culture in vitro of a few species in the family Magnoliaceae. Merkle and Sommer [4] reported the regeneration of Liriodendron tulipifera by somatic embryogenesis from immature embryos, and somatic embryogenesis from immature seeds, with subsequent germination, has been reported for other members of Magnoliaceae, such as M. virginiana, M. fraseri, M. acuminata, M. macrophylla and M. pyramidata [57]. More recently, Martin et al. [8] established a plant-regeneration system that included somatic embryogenesis for M. dealbata.

In contrast, efforts at the regeneration of M. obovata using tissue culture techniques have met with limited success. Since explants of M. obovata contain many phenolic compounds and the rate of contamination is very high, their callus formation is frequently inhibited [2]. Nakamura et al. [9] reported that multiple shoots were induced during culture of shoot apices of M. obovata, but no roots were produced. Kim et al. [10] reported successful somatic embryogenesis and the regeneration of plants from immature seeds of M. obovata. However, to our knowledge, there are no published reports of somatic embryogenesis and plant regeneration from mature seeds of M. obovata.

The main aim of the present study was to induce the formation of proembryogenic masses (PEMs) and the regeneration of complete plantlets from mature seeds of M. obovata. Mature seeds can be stored for a longer time than immature seeds and, thus, establishment of a method for embryogenesis from mature seeds should have economical benefits. This is the first report, to our knowledge, of the regeneration of M. obovata by somatic embryogenesis from mature seeds.

Materials and methods

Plant materials

Mature seeds were collected from an approximately 60-year-old specimen of M. obovata on the campus of the Tokyo University of Agriculture and Technology in Fuchu, Tokyo, Japan, on 16 August 2007 (Fig. 1a, b). The surface of seeds was sterilized by washing seeds as follows: immersion in 70% ethanol for 2 min; immersion in a solution of 6% sodium hypochlorite that contained a few drops of TWEEN 20® per 300 mL for 30 min; immersion in sterile distilled water for 10 min (repeated a total of five times); immersion in 0.01 M HCl for 3 min; and immersion in sterile distilled water for 5 min (repeated a total of three times). Hard mature seeds were bisected longitudinally with nippers, and the halves were placed, cut surface downward, on the indicated medium in a 90-mm plastic Petri dish.

Fig. 1
figure 1

Regeneration of plants by somatic embryogenesis from mature seeds of Magnolia obovata. a Fruits collected on 16 August 2007. Scale bar 5 cm. b Mature seeds from a fruit. Scale bar 2 cm. c PEMs induced after initial culture on MS basal medium that contained 1 μM GA3 for 1 month and transfer to half-strength MS basal medium without GA3 for 1 month. Scale bar 2 mm. d Globular-stage somatic embryos (arrowheads). Scale bar 100 μm. e Heart-stage somatic embryo. Scale bar 100 μm. f Torpedo-stage somatic embryo. Scale bar 100 μm. g Torpedo- and cotyledon-stage somatic embryos generated from PEMs. Scale bar 2 mm. h Mature somatic embryos derived from the torpedo- and cotyledon-stage embryos shown in g after 2 further weeks in culture. Scale bar 2 cm. i Germination from cotyledon-stage somatic embryos shown in h after a further 2 weeks in culture, with cotyledons and roots. Scale bar 2 cm. j Generation of plantlets from somatic embryos. Scale bar 2 cm. k Plantlets with secondary embryos at the base of the stem and the end of the root (arrowheads). Scale bar 2 cm

Induction of PEMs

The halves of mature seeds were initially cultured on Murashige and Skoog (MS) basal medium [11] or on B5 basal medium [12] that contained 0, 1, 5 or 10 μM GA3 (gibberellic acid; Wako, Japan) for 1 month because MS basal medium plus GA3 was used to induce somatic embryos from root explants of Carica papaya [13] and immature seeds of Quercus robur [14]. Then, we transferred the mature half-seeds to half-strength MS basal or B5 basal medium for further culture in the absence of GA3 to induce PEMs.

All media were supplemented with 30 g/L sucrose (Wako) and solidified with 3 g/L gellan gum (Wako). In addition, MS but not B5 basal medium was supplemented with myo-inositol (100 mg/L). All media were adjusted to pH 5.7 with 1 N KOH and then autoclaved at 121°C for 20 min. GA3 was added to the media as a filter-sterilized solution. All cultures were incubated at 25°C in darkness.

Induction and germination of somatic embryos

For the induction of somatic embryos that had been developed from PEMs, they were cultured on half-strength MS basal or B5 basal medium plus 30 g/L sucrose and 3 g/L gellan gum with replacement by fresh medium at monthly intervals at 25°C in darkness.

For germination of somatic embryos, a piece of filter paper was placed on top of the same gellan gum-solidified medium. All cultures were maintained at 16-h photoperiod at 25°C. Plantlets were cultured for 4 months in vitro and then transferred to potting soil (peat moss, vermiculite, perlite; v/v, 1:1:1).

Results and discussion

Induction and maturation of PEMs

We observed PEMs 1 month after the mature half-seeds had been placed on half-strength MS basal medium or B5 basal medium without GA3 (Fig. 1c). PEMs were yellowish and compact with evidence of considerable proliferation of cells. The frequencies of induction of PEMs were different among the eight initial media tested, although we did not observe repeated times (Table 1). The frequencies of induction of PEMs were 28, 4 and 4% on initial basal MS medium supplemented with 1, 5 and 10 μM GA3, respectively. In the absence of GA3, no PEMs were induced on MS basal medium. On initial B5 basal medium without GA3 and on initial B5 basal medium supplemented with 5 μM GA3, the frequencies of induction of PEMs were 4 and 8%, respectively. No PEMs were induced on B5 basal medium supplemented with 1 and 10 μM GA3. Thus, initial MS basal medium with 1 μM GA3 appeared to be the most frequent formation of PEMs from mature half-seeds of M. obovata.

Table 1 Numbers of proembryogenic masses (PEMs) induced under various conditions and the frequencies of germination from mature seeds of Magnolia obovata

One month after PEMs had been transferred to half-strength MS basal medium or B5 basal medium without GA3, we observed somatic embryos at the globular, heart, and torpedo stages (Fig. 1d–f). One month later, we observed approximately 90 somatic embryos at the torpedo and cotyledon stages (Fig. 1g).

In contrast to our study with mature seeds, Merkle and Watson-Pauley [6] cultured immature seeds of M. macrophylla in the yellow-poplar induction medium that had been developed by Merkle and Wiecko [5] and contained 9.0 μM 2,4-dichlorophenoxyacetic acid (2,4-D) and 1.1 μM benzyl adenine (BA). They observed the induction of both PEMs and somatic embryos. Kim et al. [10] reported that addition of 1.0 mg/L (approximately 4.5 μM) 2,4-D alone or in combination with 0.01 mg/L (approximately 0.5 μM) thidiazuron to MS basal medium induced the formation of embryogenic calli from immature seeds of M. obovata. However, we failed to observe the formation of PEMs and of somatic embryos from mature seeds of M. obovata when we added the combination of 9.0 μM 2,4-D and 1.1 μM BA to half-strength MS basal medium or B5 basal medium (data not shown). Therefore, neither 2,4-D nor BA appeared to be necessary for the formation of PEMs and somatic embryos from mature seeds of M. obovata.

Germination

We transferred the somatic embryos individually to half-strength MS basal medium or B5 basal medium without GA3 and incubated them with 16-h photoperiod for germination. After 2 weeks, the formation to cotyledons was apparent (Fig. 1h). Within 1 month, the embryos formed roots and hypocotyls (Fig. 1i). Cotyledons elongated, reaching close to 2 cm in length. The somatic embryos developed more rapidly and turned green after they were exposed to 16-h photoperiod.

After 1 month in culture, almost all the somatic embryos had germinated, and we counted the numbers of germinated somatic embryos with normal primary leaves and roots (Table 1). The frequencies of normal germination were 82, 90 and 85% for the somatic embryos that had been cultured initially on MS basal medium with 1, 5 and 10 μM GA3, respectively. The frequencies of germination were 83 and 81% for somatic embryos that had been cultured initially on B5 basal medium without GA3 and with 5 μM GA3, respectively. The remaining plantlets with abnormal cotyledons formed well-developed roots. After 1 month of elongation of cotyledons and hypocotyls, most plantlets were approximately 5 cm long and had developed several pairs of leaves (Fig. 1j). We sometimes observed the formation of secondary embryos at the base of the stem and ends of roots during the germination stage (Fig. 1k). After 3 months, most plantlets had reached approximately 15 cm in height and had developed sizeable leaves (Fig. 2a). We transferred the plantlets to pots of soil for acclimatization and the plants continued to grow (Fig. 2b).

Fig. 2
figure 2

Plants obtained by somatic embryogenesis from mature seeds of Magnolia obovata. a Plantlets derived from somatic embryos. Scale bar 5 cm. b Acclimatized plants 4 months after the transfer of plantlets to potting soil

Abscisic acid (ABA) was used to induce the germination and normal maturation of embryos of Liriodendron tulipifera [15], and Kim et al. [10] observed the germination of somatic embryos derived from immature seeds of M. obovata on the half-strength MS basal medium plus 1.0 mg/L (approximately 3 μM) GA3. By contrast, the germination from somatic embryos was observed on yellow-poplar induction medium without growth regulators in the case of several species of Magnolia [57], such as M. virginiana, M. fraseri, M. acuminata, M. macrophylla, and M. pyramidata. Martin et al. [8] also observed that the germination rate of somatic embryos of M. dealbata in woody plant (WP) basal medium without growth regulators was close to 90%. Similarly, we found that more than 80% of somatic embryos germinated on half-strength MS basal medium or on B5 basal medium without growth regulators. We found that neither ABA nor GA3 was needed for the maturation and germination of somatic embryos from mature seeds of M. obovata.

Conclusion

We produced plantlets from mature seeds of M. obovata by somatic embryogenesis. Our results indicate that the addition of 1 μM GA3 to initial MS basal medium might promote the induction of PEMs from mature half-seeds of M. obovata. More than 80% of somatic embryos germinated normally. Therefore, we obtained approximately 70–80 plantlets with normal primary leaves and roots from one half-seed that had induced PEMs. Approximately 10 months were required for the regeneration, from mature seeds, of plantlets that could be transferred to soil. The culture system for somatic embryogenesis developed in the present study should be useful for the propagation of medically important M. obovata.