Plant Cell Reports

Plant Cell Reports (1993) 12:501-505

9 Springer-Verlag1993

Embryogenesis from microspores of Ginkgo biloba L., a medicinal woody species Dominique Laurain, Jocelyne Tr6mouillaux-Guiger, and Jean-Claude Ch~nieux Biotogie Cellulaire et Biochimic V6gbtale EA 1370 DRED, Facult6 de Pharmacie, 2bis Boulevard Tonnell6, 37042 Tours c6dex, France Received January 26, 1993/Revised version received April 22, 1993 - Communicated by I. Potrykus

SUMMARY

The present work establishes that isolated microspores of Ginkgo biloba L. cultured at densities of 1.5 to 5-104 per milliliter in Bourgin and Nitsch (1967)liquid medium are able to divide, both in the presence and in the absence of exogenous growth regulators, and to germinate by growing a pollen tube. In all experiments the microspores exhibited various modes of division leading to embryo formation in the liquid medium. Four weeks later, the microspores which had been previously submitted to various electrical stresses showed pro-embryo development earlier than those which had not. After ten weeks the number of embryos was found to be 300 to 5300 m1-1 following the experiments. When the embryos exhibited a slower growth in liquid medium, they were transferred onto various solid media for maturation. Two months later, embryos had proliferated visibly. ABBREVIATIONS : BN : Bourgin and Nitsch (1967) medium ; IAA : Indole-3-acetic acid ; KIN : Kinetin ; GS : Growth substances KEY WORDS : Embryogenesis, microspores, woody species, Ginkgo biloba

INTRODUCTION Ginkgo biloba L., a dioecious species, is an oviparous tree considered today as a living fossil. G. biloba is highly important because of the medicinal compounds present in the leaves. The ginkgolides have strongly antagonist effects on Platelet Activating Factors (Dupont et aL 1986) and are efficient against cardiovascular disorders and inflammatory reactions. Numerous publications have appeared on the pharmacology of these compounds (Van Beeket al. 1990). By contrast works on in vitro cultures of G. biloba are rare.

Correspondence to: L Tr6mouillaux-Guiller

However, calluses were formed from pollen cultured in vitro (Tulecke 1953 ; Tulecke 1957) and from zygotic embryos (Yates 1986 ; Carrier eta/. 1990). In addition, microscopic studies were made on the pollen development of G. bi/oba by Rohr (1980). Gametic embryogenesis from isolated microspores permits the production of isogenic diploids by chromosome diploidization and the isolation of mutants (Reinert and Bajaj 1977) or variants which present therapeutic properties (Sangwann - Norreel et al. 1986). In addition, androgenic embryos can be efficiently used in synthetic seed technology : artificial seeds obtained from microspore-derived embryos of barley maintained their germination capacity for at least six months (Datta and Potrykus 1989). Moreover, microspore cultures can be utilized as a cell culture system for gene transfer, as has been shown to obtain transgenic Indica-rice (Datta et al. 1990). The aim of our investigation is to develop a protocol for producing embryos and regenerating plantlets that would be able to synthesize metabolites at higher levels or to reveal new compounds with therapeutic properties. For the first time, we report the development of embryos in liquid medium from isolated microspores of G. biloba. MATERIALS AND METHODS Microspore isolation

Mesoblasts bearing male inflorescences were collected from a G.bi/oba - tree in the Botanical Garden of Tours. Harvests were made during a short period of time (from 23 rd March to the 3 rd April 1992) to obtain microsporophylls containing uninucleate microspores. The correct developmental stage was determined by nuclear staining with acetocarmine. Fresh microsporophylls introduced into 7.5% (w/v) calcium hypochlorite solution, were vigorously shaken for 5 rain to sterilize (2200 vibrations min -1). After washing 3 times with sterilized distilled water, each

502 microsporophyll was gently burst with forceps, liberating large quantities of microspores into 1 ml of Bourgin and Nitsch (1967) liquid media lacking growth regulators (Exp. I), or supplemented with hormones (Exp. I1-111). The different liquid media used in all experiments are shown Table 1. Five to eleven microsporophylls were necessary to achieve 1.5.104 to 5.104 microspores m1-1.

shock, t00 p.I of the microspore suspension were diluted with 1 ml of BN liquid medium supplemented with 11.40 p.M IAA and 0.93 I~M KIN. The electrostimulated microspore cultures were kept in the dark (as previously described). Table 2 : Various parameters used to electropulse microspore suspensions of G. biloba. Elec~'ie Fidd

Experimental methods of culture The basal BN (Bourgin and Nitsch 1967) medium was used with supplements as indicated (Table 1). In our study, three experiments were conducted as follows : microspores were isolated at various densities (1.5.104 to 105ml 1 ) and cultured either in free-hormone BN liquid medium (experiment I), or in BN liquid medium

(V cm -1 )

100

250

750

1000

20-30-100 20-30-100

20-30-100

20-30-100

3

3

3

Pulse durafion

(p.s)

20-30-100

Pulse number

3

3

1

3

1

3

3

1

3

1

3

1

RESULTS Microspore culture Freshly isolated rnicrospores in liquid medium, generally exhibited a spherical shape and displayed a dense cytoplasm with a central nucleus and small plastids (Fig. 1A ). Most microspores were 50 to 80 I~m in diameter. After 24 hours, the viability of isolated microspores of Ginkgo bi/oba at a uninucleate stage, was found to be respectively : 45% (Exp. I) without growth factors, 54% (Exp. II) with IAA and KIN, and 36 to 52% for electropulsed microspores (Exp. III). Within 48-72 hours the cellular volume increased and the diameter reached

supplemented with tl.40 ~M indole-3-acetic acid (IAA) and 0.93 I~M kinetin (KIN) (experiments II - III). In experiment III, before being cultured, microspores were electrostimulated according to the conditions described (Table 2). Two milliliters of final suspension were plated in 55 mm Petri dishes (Exp. I - II) and 1 ml in 35 mm Coming dishes (Exp. III). These microspore cultures were sealed with Parafilm, stored at 24 _+ 1~ in the dark for 8 days and subsequently exposed to light (1000 - 1500 lux) at the same temperature. After two months, microspore cultures were routinely diluted by 1-2 ml (Exp. I - II) and 100 - 200 ~d (Exp. III) of fresh liquid media at monthly intervals. After 4-5 months the embryos obtained from microspores were transferred onto various solid media with or without growth regulator supplements. Sucrose was added to some of the media at various concentrations (Table 1). All media were previously adjusted to pH 6 (before autoclaving).

102 to 125 I~m. After a week of incubation 25 to 36% of microspores started dividing. In all three experiments, the microspores of G. biloba exhibited the same various modes of development leading to a direct embryogenesis. Both in the presence and in the absence of growth hormones, the beginnings of germination were characterized by the formation of asymmetrical celled microspores (Fig. 1B). Pollen tubes (Fig. 1C) appeared 2 to 3 weeks later in 11 to 24 % of cultured microspores (with or without growth regulators).

Table 1 : Culture media used for microspores and for embryos derived from Ginkgo bi/oba liquid medium

50

=,,

BN solid medium

2IN

B~GS

A

B

C

D

E

F

G

BN basal medium

+

+

+

+

+

+

+

+

+

M sucrose

).057

0.057

0.014

0.029

0,140

0,260

0.029

0,029

0.029

gM 11.40 0.93

KIN ouconN milk

1.14

120

120

120

120

120

120

4,60

9.30

9.30

120

120

120

*BN = Mineral salts and organic additions of Bourgin and Nitsch medium (1967)

Electrostimulation conditions Isolated microspores of G. bi/oba were mixed (1.5 to 5 104ml -1) in buffer solution containing 5 mM MES (N morpholinoethane sulphonic acid), 3 mM MgCI2 and 57,8 mM sucrose. From this rnicrospore mixture samples of 100 Izl were dispensed into Corning dishes (35 mm in diameter) and were pulsed at various electric fields, pulse-durations and pulse-numbers (Table 2) using the Electropulsator ATEIM/CNRS (France). After the electric

Fig. 1 :/n vitro cultures of Ginkgo biloba microspores. A) Freshly isolated microspores in BN liquid medium - B) microspores showing asymmetrical divisions (A) with pollen-tube formation (4) - C) Growth of one pollen tube (PT) with rhizoid-structures (R) and an embryogenic cell (EC) on the tip. Bar = 100 p.m

503 From the outset of culture, through a phenomenon that we have named "inverted endocytosis", some microspores were observed to eject either a vegetative nucleus or a generative nucleus or a second prothallial nucleus into the extracellular medium, often observed during the protoplast culture in woody species (Fig. 2and Fig.3A). "Inverted endocytosis" was also observed in microspores that had not germinated (Fig. 2). These nuclear expulsions generated real egg cell-like structures. These in turn, by sustained endomytosis, produced unicellular microclones, which then gave rise to globular embryos (Fig. 3 B-C-D). On the other hand, without germination, some microspores directly evolved into microclones by endomitosis, others formed clusters by isodiametric cellular divisions, and all could generate embryos and embryo-clusters (Fig. 2). r~

Unequal =h- Cellular Divisions ~

Fourfold

asymmevical

N~:~ear Expulsions by =IE" Via Pollen Tube

I="" Pollen Tube Growlh

I~" celled miccospore

I

I

Second Prottlalllal

I

Generative NUcleus

'eL, J j

I

Unln~ cleate 9

Micro

Nuclear

Unicellular

by "IE"

by end~mitosis

Microspore rowlh ~ Size

~

Fig. 3 : "Nucleus-expulsion" and embryo-formation from G. biloba microspores. A) "Nucleus-expulsion" via a phenomenon called "invertedendocytosis" (IE). B-C-D Through numerous endomitosesa small mlcrospore derived cell led to a microclone (MC), which gave a proembryo (PRO), then an embryo (E).Bar = 100 ~m.

Globular embryo

NuClear Divisions ~

Isedlametric Cellular Divisions

I

vegetallve Nucleus

Unicellular Microclone ~ by endornltosls

I=

I c,u,~,,

Globular embryo inner cell wall of rnicrospore

~ r"'~-=oo'=o j

EMBRYCGENESIS PATI-r~AY GERMINATION arr TUBE GROWTH PATHWAY

Fig. 2 : Summary of responses observed in uninucleate microspores of Ginkgo biloba cultured in various conditions into Bourgin and Nitsch (1967) liquid media "IE" : Inverted Endocytosis

Embryo development in liquid medium The proembryos were characterized by voluminous amyloplastids before developing into globular, oblong, heart and torpedo embryos (Fig. 4 A-B-C-D). Embryo formation was completed six weeks later in liquid medium in the presence as well as absence of growth regulators. After ten weeks, the embryo number was found to be respectively 300 m1-1 (Exp. I), 350 m1-1 (Exp. II) and 3650 mY1 (standard Exp. Ill) (Table 3). After four weeks the embryos appeared earlier when the microspores were previously electrostimulated. Six to ten weeks later the number of embryos obtained in experiment III was respectively 250-600 ml "1 and 2500-5300 m1-1 (Table 3). Figure 5 shows the large numbers of embryos at this time. Embryogenesis was observed with an efficiency (the number of embryos divided by the number of microspores originally plated) of either 1.8 % (Exp. I) when the

Table 3 9 Embryo formation and embryo efficiency from cultured microspores of Ginkgo biloba for 4, 6 and 10 weeks. Experiments Formation of pro-embryo and densities of Efficiency (I- II- III) embryo ; quantities of embryos isolated of embryo (m1-1) after 4, 6 and 10 weeks microspore., formation (%) 10-4 ml-1 after 10 week,, 4 6 10

*PE -

**E

3OO

1.60

1.80

PE -

E

35O

1.90

1.80

III Microspores electrostimulated (Volt cm-t)

50 100 250 500 75O 1000

PE;E ! 400 PE;E 250 PE;E 600 PE;E 26O PE;E 4O0 PE;E 38O

4000 3500 2500 53OO 49OO 417O

2.40 1.60 1.50 2.40 2.10 2.10

16.70 21.90 16.70 22.00 23.30 19.90

Standard microspores

PE-

3650

1.70

21.50

150

* pro E (proembryos) ; ** E (embryos) - PE + E (Non evaluated) Microspores of G. biloba were cultured in hormone free BN medium (Exp.I) and in BN medium supplemented with 11,40 p.M IAA and 0,93 p.M KIN (Exp. II - III).

504

Fig. 6 : Stereomicroscopicobservations(x 10). A) Embryogeniccluster on solid mediumafter five months(x 1) - B) The same undermagnifiedview (x 6 ). DISCUSSION

Fig. 4 : Various stages of embryo-developmentfrom microspores of G. biloba. A) Oblong embryo after one month. B) Young heart-shaped embryo inside the microsporewalt. C) Heart-shapedembryo. D) Young torpedo-shapedembryothree to four monthslater (Co = Cotyledon)Bar = 100 pro.

Fig. 5 : Stereomicroscopic observation (x 10). After 10 weeks, large numbers of embryos(isolatedembryosand an embryo-cluster : EC) have been observedin liquid culturesof microsporesof G. biloba(x 7.5). microspores were cultured without any hormonal supply or of 1.8 % (Exp. II) in the presence of IAA and KIN. The highest frequency of embryo formation occurred (Exp. III), when the microspores were cultured in BN medium supplemented with IAA and KIN, with the following results : 21.7 % for standard microspores and 16.70 to 23.30 % for electrostimulated microspores (Table 3). After four to five months in liquid culture, the embryos had reached different stages of development. Those showing slowed growth were transferred onto various solid media (Table 1). The transfer was not detrimental to the embryos. In spite of this fact only the embryo-clusters showed signs of growth particularly on agar medium containing 11.40 I~M KIN, while the isolated embryos apparently developed no further. One month later, the clusters of embryos showed individual develo3ment, a few of them exhibited a club shape (Fig. 6A-B).

- CONCLUSION

The androgenic ability of microspores is influenced by various factors, thus the microspore density may be a significant factor leading to culture success (Keller and Stringam 1978). Cultures of G. bi/oba developed more favourably at microspore densities ranging from 1.5 to 5 104ml -1. Using a higher density (105ml -1) the embryoinitiation was slowed down (data not shown). Equally, microspores of Brassica napus showed an optimal density ranging from 104 to 4-104ml-1, but an inhibitory effect appeared for 105ml "1 (Huang et al. 1990). The microspores of G. biloba were isolated at the uninucleate stage regarded as the best favorable for microspore development in some other species (Keller and Stringam 1978). For example, microspores of Zea mays were most responsive at this stage (Coumans et al. 1989). On the other hand, the best favourable period to initiate embryogenesis in microspores of Nicotiana tabacum was found to be just before or at the time of the first mitosis (Nitsch and Nitsch 1969). Our results show that microspores of G. biloba did not absolutely require exogenous growth regulators. Similar results were observed during maize androgenesis (Mitchell and Petolino 1991) and in other species where the endogenous hormonal level was sufficient to produce embryos (Coumans et al. 1989). However, the presence of coconut milk in Bourgin and Nitsch medium (1967) may be considered as a supply of cytokinins, able to initiate the androgenesis from Ginkgo microspores. As shown by Keller and Stringam (1978) the cytokinins were found to be essential for a maximum of microspore response in Datura and Potato. On the other hand, we observe that the presence of IAA and KIN in medium was not detrimental for the embryo formation of G. biloba. Equally, microspores of Brassica campestris were able to produce large numbers of embryos in the presence as well as absence of growth hormones (Sato et al. 1989). Our study indicates a good resistance of microspores at high voltages. The electrical stresses seem to be responsible for early embryo developments from electropulsed microspores. It is well known that the application of electrical currents stimulates cellular division and differentiation, in particular for protoplasts of woody species regarded as recalcitrant to culture (Chand et al. 1988). In experiment III (Table 3), the highest yields of

505 embryos were obtained both from electrostimulated and unelectrostimulated (standard) microspores of G. biloba. Both of them offered large embryo numbers which were increased by a factor of ten as compared to experiments I-II carried out ten days earlier (with y o u n g e r microspores). These increased yields of embryos cannot be due to the electrostimulation, but may be due to a particular phase of the uninucleate stage. Keller and Stringam (1978) reported that the uninucleate stage has been subdivided, and that the best response could be obtained at specific substages. To c o n c l u d e : the present e x p e r i m e n t s have unequivocally shown that isolated microspores of G. biloba were directly able to develop androgenic embryos in liquid medium with or without auxin supplementation. Moreover, from microspores cultured at a uninucleate stage, the results seem to reveal the presence of a substage favourable to the embryogenesis. An extensive investigation of the progress of the uninucleate stage should assist the improvement of androgenesis in G. biloba species.

REFERENCES Bourgin JP, Nitsch JP (1967)Ann. Physiol. Veg. 9:377-382 Carrier DJ, Consentino G, Neufeld R, Rho D, Weber M, Archambault J (1990) Plant Cell Rep. 8:635-638

Chand PK, Ochatt SJ, Rech EL, Power JB, Davey MR (1988) J. Expt. Bet. 206:1267-1274 Coumans MP, Sukhinder S, Swanson EB (1989) Plant Cell Rep. 7:618-621 Datta SK, Potrykus I (1989) Theor. Appl. Genet. 77:820- 824 Datta SK, Peterhans A, Datta K, Potrykus I.(1990) Bio/Technology 8:736-740 Dupont L, Dideberg O, Germain G, Braquet P (1986) Acta Cryst. 42:1759-1762 Huang B, Bird S, Kemble R, Simmonds D, Keller W, Miki B (1990) Plant Celt Rep. 8:594-597 Keller WA, Stringam GR (1978) In Thorpe TA (ed) Frontiers of Plant Tissue Culture, The Bookstore, Canada, pp 113-122 Mitchell JC, Petolino JF (1991) J. Plant Physiol. 137:530 - 536 Nitsch JP, Nitsch C (1969) Science 163:85-87 Reinert J, Bajaj YPS (1977) in Reinert J, Bajaj YPS (eds) Plant Cell "tissue and Organ Culture, Springer - Verlag,Berlin, Heidelberg, New York, pp 251-267 Rohr R (1980) Cytologia 45:481-495 Sangwan-Norreel BS, Sangwan RS, Pare J (1986) Bull. Soc. Bet. Fr. 133:7-39 Sate T, Nishio T, Hirai M (1989) Plant Ceil Rep. 8:486-488 Tulecke WR (1953) Science 117:599-600 Tulecke WR (1957) Amer. J. Bet. 44:602-608 Van Beek TA, Scheeren HA, Rantio T, Griepink FC, Melger WC (1990) Planta Med. 56:509 Yates W. (1986) in somers DA, Gengenbach BG Biesboer DD, Hackett WP, Green CE (eds) VI Inter.Congr. Plant Tissue Cell Cult., ed. University of Minesota p.43