Plant Cell Reports

Plant Cell Reports (1994) 13:601-606

9 Springer-Verlag1994

Production of vigorous, desiccation tolerant white spruce (Picea glauca [Moench.] Voss.) synthetic seeds in a bioreactor S . M . Attree 1, M.K. Pomeroy 2, ,, and L. C. Fowke 1 1 Department of Biology, University of Saskatchewan, Saskatoon, Sask, S7N 0W0, Canada 2 Plant Research Centre, Agriculture Canada, Ottawa, Ont K1A OC6, Canada Received 10 March 1994/Revised version received 7 April 1994 - Communicated by I. K. Vasil

Summary. This report describes a low-cost method for generating large numbers of high quality mature white spruce (Picea glauca [Moench.] Voss) somatic embryos which survived desiccation and grew to plantlets more vigorously than excised zygotic embryos cultured in vitro. Somatic embryos from suspension culture were supported within a culture chamber on a flat absorbent pad above the surface of a liquid culture medium containing 20-50/xM abscisic acid and 7.5 % polyethylene glycol. Throughout a 7 week culture period 3 L of fresh medium was pumped into one end of the chamber, while the spent medium exited by gravity from the opposite end. Over 6,300 cotyledonary stage white spruce somatic embryos were recovered after this time from a single culture chamber without manual manipulation. The somatic embryos were of excellent appearance with well developed cotyledons, and possessed high levels of storage lipids. They survived drying to about 8 % moisture content following treatment for 4 weeks at 63 % relative humidity, and following imbibition converted to normal plantlets at a frequency of 92 %, compared to 80 % for embryos grown in Petri dishes. Somatic embryos cultured within the bioreactor developed to plantlets that were 20 % longer than zygotic embryos excised from mature seed and grown in vitro, and were 38 % longer than somatic embryos cultured upon agar medium in Petri dishes.

Key words: Polyethylene glycol - Picea glauca Bioreactor Triacylglycerol Somatic embryo Desiccation

Introduction The use of somatic embryos for commercial plant production requires the inexpensive production of highquality somatic embryos. The ability of somatic embryos * Plant Research Centre contribution No. 1523 Correspondence to: S. M. Attree

to withstand drying to low moisture contents is important for storage, and plays a role in the developmental transition between maturation and germination (Bewley and Black 1985; Kermode 1990; Attree and Fowke 1993) leading to enhanced postgerminative vigour. Synthetic seed technology based on the use of somatic embryos, is expected to revolutionize propagation of conifers (Attree and Fowke 1993; Gupta et al. 1993). Recent protocols for the production of mature conifer somatic embryos include combinations of increased osmotica and abscisic acid (ABA) during development, followed either by partial drying (Roberts et al. 1990; Roberts 1991) in which 5-10 % moisture is removed, or full drying (Attree et al. 1991, 1992; Attree and Fowke 1993; Misra et al. 1993) to moisture contents below those recorded for zygotic embryos from mature dry seed. The somatic embryos contained high levels of storage reserves, but were less vigorous than their zygotic counterparts. In most instances, the cultures were maintained in Petri dishes containing either agar solidified media, or less commonly absorbent pads soaked with liquid development media (Boulay et al. 1988). Frequent transfer of developing embryos to fresh media was required. As a consequence, methods for large scale micropropagation are labour intensive and costly. Development of conifer embryogenic suspensions in mechanicaly stirred bioreactors would permit simple handling of large quantities of material, thereby reducing labour costs, but such bioreactors are expensive and there are no reports of successful maturation of conifer somatic embryos and plantlet recovery following any form of submerged liquid culture. We describe a simple procedure which provides a low-cost method for producing vigorous white spruce somatic embryos, using a large culture chamber, supported on a fiat absorbent bed above the surface of a flow of liquid culture medium, followed by late

602 maturation stage drying. The quality of the somatic embryos recovered using this bioreactor was compared with that of somatic embryos developed on agar-gelled medium within Petri-dishes and zygotic embryos from mature seed.

When the media requirements in three culture chambers were the same, costs could be reduced by using a single reservior carboy containing 10 L of medium, and one spent medium carboy. In this instance, three-way connectors were used to split the tubing to each culture chamber before it entered the pump, and to reconnect the tubing prior to joining the spent medium carboy.

Materials and methods

Culture media. The basal medium was a half strength Litvay et al. (1981) medium supplemented with 250 mg/L glutamine and 500 mg/L casein hydrolysate. Immature white spruce somatic embryos (line WS1) were maintained as suspension cultures in basal medium, 1 % sucrose, 9 /~M 2,4-dichlorophenoxyacetic acid and 4.5 /zM benzyladenine. For somatic embryo development, the basal medium contained 3 % sucrose, 7.5 % (w/v) polyethylene glycol (PEG) 4000 (Fluka Chemical Corp.) and 20-50 #M + racemic ABA (Sigma; product number A 2784). The ABA was added prior to autoclaving. The plantlet regeneration medium contained 2 % sucrose, 0.6 % agar and lacked plant growth regulators. Agar (Difco Bitek) was included in media where required at 0.4 %. The above media were adjusted to pH 5.7. Plastic Petri-dishes (10 cm) containing 15-20 mL agar medium were used for growth of recovered embryos. Dishes were sealed with Parafilm (American Can Co., Greenwich, USA) and all cultures were incubated at 25~ Culture apparatus. A culture chamber was fabricated using a high-density polypropylene tray with lid (Scienceware). This was 28 cm long, 16 cm wide and 15 cm high. The central section of the lid was replaced with glass to facilitate viewing of the cultures. Entrance and exit ports comprised of nylon bulkhead fittings (Cole Panner) were incorporated at diagonally opposite side comers of the base. The chamber base was lined with cotton wool over which was placed a 48 #m nylon mesh and then a filter paper support (Whatman no. 1) cut to fit the internal dimensions of the chamber base. The nylon mesh allowed the developed somatic embryos to be easily removed from the culture chamber without tearing the filter paper support. The culture pad was saturated with development medium and then the culture-chamber lid was secured with autoclave tape. Three to four litres of development medium was added to a 10 L carboy (Nalgene). The medium outlet from the carboy was connected to the culture chamber using a 2.5 m length of silicon tubing. The exit port of the culture chamber was connectedusing a 1.5 m length of silicone tubing to a second carboy which served to collect the spent medium. Both carboys were vented with filters (Millipore Millex-FGs0, 0.2 #m pore size). The complete apparatus was autoclaved for 45 min at 121 ~ During cooling the apparatus was placed in a laminar flow unit to reduce the chance of microbial contamination. The peristaltic pump had a multiple head which allowed up to three culture units to be run simultaneously.

Somatic embryo development. Somatic embryo development was carried out using methods modified from Attree et al. (1990, 1991). Suspension cultures were first precultured for one week in basal medium with 3 % sucrose and growth regulators reduced to 1/5 of the maintenance medium levels. After this time, 10 g of precultured immature somatic embryos were washed and resuspended (20 % w/v) in 50 mL of plant growth regulator-free medium which was inoculated as evenly as possible onto the filter-paper support to cover the entire surface. The culture chamber lid was then replaced and resealed with autoclave tape. The liquid development medium was pumped from the reservoir carboy into the culture chambers using a peristaltic pump at a rate of 60 mL per day. This was equivalent to one medium change per week and about 3 L per 7 week culture period. The spent-medium carboy was placed below the level of the culture chamber so that excess culture medium emptied from the culture chamber under the force of gravity. After development the filter paper support was cut into sections which were placed in empty Petri-dishes. The cotyledonary embryos were then dried in an environment of 63 % relative humidity as described previously (Attree et al. 1991). After drying for 4 weeks the desiccated somatic embryos were either analyzed for storage lipid or imbibed and plantlets subsequently recovered. For comparison, 0.75 mL of a 20 % (w/v) suspension of precultured somatic embryos were plated onto filter paper supports overlaying agar gelled development medium (Attree et al. 1991; 1992). Moisture content and lipid analysis. Moisture content was determined using 50-80 desiccated somatic embryos, repeated three times (Attree et al. 1991). Storage lipids in the somatic embryos were analyzed using methods outlined previously (Pomeroy et al. 1991; Attree et al. 1992). Sample sizes consisted of 50-180 somatic or zygotic embryos. Each sample was divided into 2-3 replicates for analysis, and experiments were repeated three times. Results shown are replicate means of one experiment. Plantlet regeneration. Following development and desiccation somatic embryos intended for further culture were imbibed by transferring the filter-paper support pieces onto plantlet regeneration medium. Somatic embryos were maintained under low light intensity (2 Wrn"2, 20 h photoperiod, 20 W cool-white fluorescent lamps). Two to three days later they were separated and

603

Fig. 1. Three bioreactor units. The apparatus consists of (from leg to fight): reservoirs of liquid nutrient medium; single peristaltic pump; culture chambers; spent medium containers (bar = 30 cm). Fig. 2. White spruce somatic embryo development after 7 weeks culture within a bioreactor chamber with 50/zM ABA and 7.5 % PEG (bar = 2 mm). Fig. 3. White spruce somatic embryos treated as described in Fig. 2, then dried for four weeks at 63 % relative humidity. The somatic embryos are shrunken and have a yellow waxy appearance (bar = 1 nun). Fig. 4. Portion of a dried culture treated as described in Fig. 3, showing high density of germinating somatic embryos. The embryos were imbibed with liquid culture medium and placed in the light for 5 days. All somatic embryos have turned green and commenced elongation (bar = 2 mm). Fig. 5. Somatic embryo-derived plantlets after 2 weeks growth. Note extensive root and shoot elongation (bar = 5 ram). Fig. 6. Zygotic embryo-derivedplants grown for 2 weeks from excised zygotic embryos (bar = 5 nun).

604 placed horizontally on fresh plantlet regeneration medium, and maintained at the same low light intensity. To determine the frequency of plantlet regeneration 400 embryos were scored for their ability to survive desiccation and regenerate normal looking plantlets. For comparisons with somatic embryos, whole seeds were surface sterilized and germimtted on plantlet regeneration medium, or the zygotic embryos were dissected from the seeds and germinated in vitro. These, together with whole seeds were grown in vitro on plantlet regeneration medium. Growth measurements are based on a total of 81 zygotic seedlings germinated from whole seed, 29 zygotic seedlings germinated from embryos excised from seed, and 160 somatic embryos.

6.0

E 0 ee--

5.0 4.0

E 3.0 0

x::

2.0

e.-

1.0

Results

Figure 1 shows three bioreactor units run from a single peristaltic pump. In a bioreactor culture chamber the white spruce somatic embryos underwent development and produced normal looking embryos with well developed cotyledons (Fig. 2). No manual manipulation of the developing embryos was required during a 7 week culture period. Each culture chamber typically yielded cotyledonary-stage somatic embryos almost covering the entire surface of the filter paper support. From one culture chamber 6,314 normal looking somatic embryos were recovered. The somatic embryos started to turn green during the later stages of development (after the 7th week of culture) when cultured in the presence of 20 tzM ABA, but precocous germination was inhibited when 30-50 #M ABA was used (Fig. 2). After drying, the somatic embryos within the bioreactors were hard and shrunken and had a translucent yellow waxy appearance (Fig. 3), with a moisture content (+ s.c.) of 7.56 + 0.03 %. Following imbibition they swelled and regained their well-developed predesiccated appearance, and over the next three to five days turned green at a frequency exceeding 99 % (Fig. 4). The somatic embryos converted to normal looking plantlets (Fig. 5) which compared favourably in appearance with excised zygotic embryos cultured in vitro (Fig. 6). Four weeks after imbibition the plantlet conversion frequency was 92.0 % (368/400), and somatic plantlets were 20 % longer than zygotic plants grown for 4 weeks in vitro from excised embryos (Fig. 7), and 38 % longer than somatic embryos cultured on agar medium in Petri dishes; however, the length of the bioreactor cultured somatic plantlets was not as great as for zygotic plants recovered from whole seed. Somatic embryos of white spruce developed on agar medium and then desiccated, converted to plantlets at frequencies of 79.5 % (318/400). During subsequent culture plantlets underwent sustained root elongation and epicotyl development. The storage lipid accumulation was investigated in order to determine whether the improved growth of the somatic embryos from the bioreactors was due to enhanced

0.0

embrYo type Fig. 7. Length ~ confidence limits, P = 0.05) of zygotic seedlings germinated from whole seed or excised zygotic embryos, and somatic plantlets grown from somatic embryos developed in Petri dishes or in a bioreactor. All plants were grown in vitro for four weeks on plant growth regulator-free germination medium. storage lipid accumulation. Table 1A shows that somatic embryos cultured within bioreactors or upon agar medium within Petri dishes possessed high levels of storage lipids. The values for total storage lipid per embryo were over 3 times higher than zygotic embryo levels. Fatty acid compositions of the TAG were similar in all cases (Table 1B).

Discussion

We have shown that white spruce somatic embryos can exceed the growth of in vitro grown excised zygotic embryos. The high frequency of somatic embryos that turned green and commenced growth following desiccation shows that they were extremely tolerant to late stage drying to about 8 % moisture content following development for 7 weeks within the bioreactor. The moisture content of desiccated white spruce somatic embryos cultured for 4 weeks was previously determined (Attree et al. 1991) to be between 20 to 30 %. The plantlet conversion frequency of over 90 % obtained during the present study was higher than germination frequencies often obtained from zygotic seed. Somatic embryos converted to plantlets at a frequency of 80 % following development on agar then desiccation, similar to results obtained previously (Attree et al. 1992). Maturation, desiccation tolerance, and postgerminative

605 Table 1. A, dry weights (DW), total lipid (TL) and tdacylglycerol(TAG) fatty-acidmethylesters (FAMES) content (~. s.c.), and; B, fatty acid compositionsof TAG (FAMES), of white spruce zygotic embryosexcised from mature seed, and somatic embryos developed with 20 pM ABA and 7 % PEG, either on agar medium in Petri dishes or over liquid medium in a bioreaetor. A Embryo DW (rag)

% DW of TAG

TAG (/.tg per embryo)

TAG/TL (%)

Somatic Petri dish

0.54+0.047

32.0+0.8

172.6+13.6

67.8+0.9

Bioreactor

0.56+0.004

27.3+ 1.9

150.0+9.6

76.0+4.3

Zygotic

0.17+0.004

30.8+ 1.1

52.3+ 1.1

63.5-4-5.6

B

Fatty acid compositionof TAG (FAMES) (%) 16:0

16:2

18:0

18:1

18:1~

18:2

18:2b

18:3

gC-20,22~

Somatic Petri dish

6.1

1.1

2.1

24.8

2.4

49.7

11.3

0.5

1.8

Bioreactor

7.1

1.0

2.7

23.7

3.3

48.4

12.5

0.7

0.5

Zygotic

5.4

1.1

2.1

22.1

2.2

48.3

16.8

0.7

1.6

"Double bond in the C-7 position instead of the C-9. bDoublebonds at the C-5 and C-9 positions instead of the C-9 and C-12 positions. *Representsthe sum of all identified C-20 and C-22 fatty acids. growth were all improved following culture in bioreactors compared to culture on agar medium in Petri dishes. However, somatic embryos produced by either method possessed similar lipid levels, which were over three times the level of zygotic embryos. Both culture methods also produced somatic embryos with fatty acid compositions which were comparable to the fatty acid compositions of zygotic embryos. The better growth from the bioreactor grown cultures may have been due to an enhanced supply of nutrients or the efficient removal of waste products, resulting from the medium flow. Additionally, the large air space in the culture chamber may have provided a more suitable gaseous environment encouraging normal development. A non-plasmolysing moisture stress effected by high molecular weight osmotica was instrumental in promoting normal development, storage reserve accumulation, and desiccation tolerance in white spruce somatic embryos, while plasmolysing osmotica were detrimental (Attree et al. 1991, 1992; Misra et al. 1993). The lipid contents and fatty acid compositions of the zygotic embryos reported here confirm our previous observations (Attree et al. 1992). The continuous-flow culture chamber type bioreactor described here is less sophisticated than a mechanically-stirred submerged liquid bioreactor, so is considerably cheaper. Running three units seven times in

a year could yield over 125,000 mature embryos, while running 25 units seven times per year could yield over one million mature somatic embryos. Embryo drying promotes normal germination of the somatic embryos and when combined with the bioreactor, should further reduce production costs by providing a means of storing somatic embryos. Thus, for optimal efficiency quantities of high quality somatic embryos can be produced throughout the year, then stored and pooled with somatic embryos from subsequent production runs. Large quantities of somatic embryos makes machine handling a viable option, and desiccated somatic embryos can be germinated synchronously to provide vigorous plants of uniform age and size for planting. The culture system described here allows further modifications to be made to the culture environment by, for example, controlling the gaseous composition of the air space in the culture chamber, or optimising the composition and/or flow rate of the culture medium, or making gradual changes in the concentrations of plant growth regulators and osmotica as the liquid medium is added. These parameters are being investigated and the method is being applied to embryogenic cultures of additional conifer species. Our recent studies led us to use a modified Litvay et al. (1981) medium for establishing and maintaining embryogenic cultures of a variety of

606 conifer species with least difficulty. During development in this medium white spruce somatic embryos required 3050 #M ABA to prevent precocious germination. In previous studies of desiccation of white spruce somatic embryos (Attree et al. 1991, 1992; Misra et al. 1993) LP medium (von Arnold and Eriksson 1981) was used, and the optimal ABA concentration that prevented precocious germination and promoted storage lipid accumulation was observed to be between 16 to 24 #M. The reasons for the tendency for late-stage greening with the modified Litvay medium compared to LP medium at the lower ABA levels are unclear, but suggests that media components have an influence on ABA action and, therefore, optimal ABA concentrations may vary with different media formulations. The high vigour of the somatic embryos described here suggests that the increased level of storage lipid within somatic embryos compared to zygotic embryos (Attree et al. 1992) has in part compensated for the absence of a nutritive megagametophyte. This should be of value in creating synthetic seeds of conifers capable of germinating ex vitro.

Acknowledgements. We are grateful to Dennis Dyck for assistance with the figures, Mei Zhang and Ken Stanley for technical assistance, and to Gordon Justy for advice concerning biorector construction. We acknowledge financial support through the Province of Saskatchewan Environmental Technology Development Program, Pacific Regeneration Technologies Inc., and the Natural Sciences and Engineering Research Council of Canada.

References Attree SM, Tautorus TE, Dunstan DI, Fowke LC (1990) Somatic embryo maturation, germination, and soil establishment of plants of black and white spruce (Picea mariana and Picea glauca). Can. J. Bet. 68:2583-2589 Attree SM, Moore D, Sawhney VK, Fowke LC (1991) Enhanced maturation and desiccation tolerance of white spruce (Picea glauca [Moench.] Voss) somatic embryos: Effects of a non-plasmolysing water stress and abscisic acid. Ann. Bet. 68: 51952 Attree SM, Pomeroy MK, Fowke LC (1992) Manipulation of conditions for the culture of somatic embryos of white spruce for improved triacylglycerol biosynthesis and desiccation tolerance. Planta 187: 395-404

Attree SM, Fowke LC (1993) Embryogeny of gymnosperms; advances in synthetic seed technology of conifers. Plant Cell Tiss. Org. Cult. 35:1-35 Bewley DJ, Black M (1985) Seeds: Physiology of Development and Germination. Plenum Press, New York' Boulay MP, Gupta PK, Krogstrup P, Durzan DJ (1988) Development of somatic embryos from cell suspension cultures of Norway spruce (Picea abies Karst.). Plant Cell Rep. 7:134-137 Gupta PK, Pullman G, Timmis R, Kreitinger M, Carlson WC, Grob J & Welty E (1993) Forestry in the 21st Century: The biotechnology of somatic embryogenesis. Bio/Technology 11" 454--459 Kermode AR (1990) Regulatory mechanisms involved in the transition from seed development to germination. CRC Crit. Rev. Plant Sci. 9" 155195 Litvay JD, Johnson MA, Verma D, Einspahr D, Weyrauch K (1981) Conifer suspension culture medium development using analytical data from developing seeds. Tech. Pap. Set. Inst. Paper Chem. 115:1-17 Misra S, Attree SM, Leal I, Fowke LC (1993) Effect of abscisic acid, osmoficum, and desiccation on synthesis of storage proteins during the development of white spruce somatic embryos. Ann. Bet. 71:11-22 Pomeroy MK, Kramer JKD, Hunt DJ, Keller WA (1991) Fatty acid changes during development of zygotic and microspore derived embryos of Brassica napus. Physiol. Plant. 81:447-454 Roberts DR, Sutton BCS, Flinn BS (1990) Synchronous and high frequency germination of interior spruce somatic embryos following partial drying at high relative humidity. Can. J. Bet. 68: 10861090 Roberts DR (1991) Abscisic acid and mannitol promote early development, maturation and storage protein accumulation in somatic embryos of interior spruce. Physiol. Plant. 83:247-254 von Arnold S, Eriksson T (1981) In vitro studies of adventitious shoot formation in Pinus contorta. Can. J. Bet. 59" 870-874