Planta
Planta 144, 241-248 (1979)
9 by Springer-Verlag i979
A Structural Study of Germination in Celery with Emphasis on Endosperm Breakdown
(Apium graveolens L.)
Seed
J.V. Jacobsen 1 and E. Pressman 2 1 Division of Plant Industry, CSIRO, P.O. Box 1600, Canberra, A.C.T. 2601, Australia, and 2 Volcani Center, Agricultural Research Organization, Bet Dagan, Israel
Abstract. Germination of celery seed occurred after
6 d of imbibition in light. During this time the embryo enlarged at the expense of the adjacent endosperm cells and at the time of germination was 2-3 times as long as in the dry seed. Breakdown of the endosperm cells near the root cap preceeded radicle emergence. None of these changes occurred in darkness. Endosperm digestion began adjacent to the embryo and spread radially. In degrading cells, the aleurone grains often became larger and fewer in number. The cell walls were modified and appeared to undergo partial degradation. Ultimately the cells seemed to lose their contents. In cells adjacent to the root cap, similar changes occurred except there was a transient appearance of starch grains. Radial progression of endosperm breakdown also occurred in isolated endosperm treated with gibberellin A4 + 7. The results indicate that (1) the stimulus for breakdown of celery endosperm emanates from the embryo in response to light; (2) the stimulus may be a gibberellin because changes in endosperm cells and the sequence of endosperm digestion during germination resemble the responses of isolated endosperm to gibberellin; and (3) the radial progression of endosperm breakdown during germination may be the result of a sequential response of cells to a uniformly applied stimulus rather than the result of gradual embryo expansion. Key words: Apium - Endosperm breakdown - Germi-
nation (seeds).
Introduction
Seed germination involves processes and control mechanisms most of which are only partially understood. Although there has been considerable effort
applied to the study of structure and function in embryogeny (see the book by Maheshwari, 1950) there has been relatively little work on similar aspects of seeds during germination. Most of the work done on the cellular structure and physiology of germinating seed has been concerned with the embryo and with the exception of cereal grains, the endosperm or other non-embryonic nutritive tissue has usually been of peripheral interest and studied intensively only in a few cases such as castor bean (Ricinus communis), carob bean (Ceratonia siliqua), and guar (Cya-
mopsis tetragonoloba). Extra-embryonic reserves in seeds can occur in a variety of tissues. In many dicotyledonous and m o n o c o t y l e d o n o u s seeds the reserves occur in endosperm which can vary in quantity from non-existent at maturity (exalbuminous) to occupying the larger part of the seed. Endosperm can consist of all living cells or of some living and some dead cells as in the cereals. In some angiosperms, the storage reserves occur largely in perisperm (persistent nucellus) as for example in Yucca (Horner and Arnott, 1965) and in gymnosperms the storage tissue is the gametophyte tissue (for review see Brink and Cooper, 1947). Both perisperm and gametophyte tissues are living. All of these tissues are digested as the embryo enlarges and germination and seedling growth occur. In the cereals, the structure and composition of the endosperm and control of its mobilization during germination are fairly well understood (see reviews of Yomo and Varher, 1971, and Varner and Ho, 1976) but there is little understanding of the controls operating during germination in other plant families. The seeds of the Umbelliferae contain relatively large amounts of living endosperm which completely surrounds a small embryo located at one end of the seed (for review see Martin, 1946). In a previous study, we described how isolated endosperm of celery was induced to autolyse by gibberellin (Jacobsen
0032-0935/79/0144/0241/$ 01.60
242 et al., 1976). M a j o r changes occurred in the cell contents a n d the cell walls were partially h y d r o l y z e d so that the cells became separated f r o m each other. This led us to propose that d u r i n g g e r m i n a t i o n o f celery seed, e n d o s p e r m d e g r a d a t i o n m a y be c o n t r o l l e d by a h o r m o n e , or by h o r m o n e s , released from the e m b r y o rather t h a n by enzymes released by the e m b r y o , a n d in this study we have sought evidence to s u p p o r t this hypothesis.
Material and Methods Celery seed (Apium graveolens L. cv. Florida 683; Ferry Morse Comp., Mountain View, Cal., USA) was placed on moist filter paper in Petri dishes, and incubated in darkness or constant light at 20~ C. Light is required for germination of celery seed at this temperature (Morinaga, 1926; Taylor, 1949). After the required times, a slice was cut longitudinally from each side of the seed to facilitate entry of fixative and embedding medium. The seed was fixed usually for 2-3 d in 3% glutaraldehyde in 0.025 M phosphate buffer, pH 7.2, in the refrigerator, and processed for embedding in paraffin as described in Jensen (1962, p. 80-83). Paraffin was used because large numbers of seed had to be sectioned serially and for the purpose of this study maximal resolution of fine cellular structure was not required. Sections were cut 10 pm thick and stained in 0.05% toluidin blue in water, or in periodic acid-Schiff reagent to detect insoluble polysaccharides (Jensen, I962, p. 198-199).
J.V. Jacobsen and E. Pressman: Germination in Celery Seed De-embryonated seeds (half-seeds) were sterilized and incubated with and without 10-s M GA4+7 (from Prof. J. MacMillan, University of Bristol, U.K.) as previously described (Jacobsen et al., 1976). After the required times, half-seeds were fixed (with no further cutting), embedded, sectioned and stained as described above.
Results A t 20 ~ C in light celery seeds begin to germinate o n the 6th day after the c o m m e n c e m e n t of i m b i b i t i o n (Pressman et al., 1977). T h e structural changes occurring in the seed d u r i n g this time a n d up to 9 d are s h o w n in Fig. 1-5.
Emb~o I n dry seed (Fig. t A, B) the e m b r y o is small, centrally located a n d s u r r o u n d e d by e n d o s p e r m a l t h o u g h the e n d o s p e r m a d j a c e n t to the radicle is only 3 or 4 cells thick. The zone i m m e d i a t e l y a d j a c e n t to the e m b r y o ( D C in Fig. 1 B) appears to c o n t a i n degraded cells. Two d f r o m the b e g i n n i n g of i m b i b i t i o n , the e m b r y o begins to elongate within the seed a n d as time progresses the cotyledons extend into the e n d o s p e r m
Fig. 1 A-C. Longitudinal sections of dry celery seed. A The whole seed. x 51. B A higher magnification of the embryo zone of A. x 110. C The endosperm cells at the side of the embryo. The arrows point to aleurone grain inclusions staining red in toluidin blue. • 440. All sections were stained with toluidin blue. En=endosperm; Em=embryo; SC=seed coat; DC = depleted cells
J.V. Jacobsen and E. Pressman: Germination in Celery Seed
243
Fig. 2A-D. Longitudinal sections of celery seed after 4 d incubation in water. A Whole seed. x 51. B The endosperm at the radicle tip. x 275. C Endosperm adjacent to the cotyledons, x 275. D A zone similar to that shown in C but stained with PAS to demonstrate that some cell-wall material remains around depIeted cells adjacent to the embryo, x 275. Sections in A, B and C were stained in toluidin blue. R=radicle; AG=aleurone grain; A=embryo axis; C=cotyledon
(Fig. 2A) and the radicle is appressed to the endosperm (Fig. 2B). At 5 d, the embryo occupies about 2/3 the length of the endosperm (Fig. 3A) but the radicle is still within the endosperm (Fig. 3B). On the 6th day the radicle penetrates the endosperm and the seed coats, and emerges (Fig. 4A). As the radicle elongates further, so do the cotyledons and progressively more of the endosperm is consumed (Fig. 5A, B). Measurement of embryo growth (Fig. 6) shows that elongation is comprised of concurrent increases in the axis and the cotyledons. In darkness, the embryo does not elongate within the seed and there are no changes in the endosperm cells.
Endosperm 1. Changes during germination. The dry endosperm tissue consists largely of angular cells which have thick (3-5 ~tln) walls and which contain m a n y aleurone grains (Fig. 1 B, C and Jacobsen et al., 1976).
The aleurone grains of some cells contain round inclusions (see arrows Fig. 1 C) which stain red with toluidin blue and presumable contain phytin (Jacobsen et al., 1971) while the grains of other cells contain angular inclusions which do not stain and which are not visible in the photographs. The two types of inclusions do not occur in the same cell. The cells adjacent to the radicle appear similar to, although much smaller than, those in the bulk of the endosperm. As the embryo expands the adjacent endosperm cells are transformed. Some aleurone grains become larger (see A G in Fig. 2C) and the cell contents become less intensely stained by toluidin blue and less discernible. Ultimately the cell contents are either lost, or fail completely to be stained by toluidin blue (see cells adjacent to embryonic axis in Fig. 2C). The ceil walls become irregular in form and stain purple in toluidin blue, as opposed to the usual blue colour. The cell-wall material which does persist is strongly PAS-positive (Fig. 2D). These changes occur all around the embryo. The early stages of change occur 3-5 cells away from the embryo while advanced stages
Fig. 3 A - C . Longitudinal sections o f celery seed after 5 d incubation in water. A Whole seed. x 51. B Endosperm ceils adjacent to the root tip. x 275. C The root tip zone stained with PAS showing starch grains in the cells of both the root tip and the adjacent endosperm. • 275. Sections in A and B were stained with toluidin blue. S = starch grains
Fig. 4 A - D . Longitudinal sections of celery seed after 6 d incubation in water. A Whole seed. x 51. B, C Endosperm adjacent to the cotyledons, x 110 and x 275 respectively. D Endosperm beside the embryo, showing the depleted cells adjacent to the emerged radicle, x 110. All sections were stained with toluidin blue. A p = embryo apex
J.V. Jacobsen and E. Pressman: Germination in Celery Seed
245
Fig. 5A and B. Longitudinal sections of celery seed after 7 d (A) and 9 d (B) of incubation in water. Sections were stained with toluidin blue. • 51
1200
C). The transition from normal to apparently depleted cells is clearly quite rapid. Cells with no stainable contents exist next to cells with many aleurone grains (Figs. 2C, D and 4C). The depleted cells at the radicle tip, which are penetrated by the radicle, are continuous with the degrading band of cells around the embryo (Fig. 4 D). The band of degrading endosperm proceeds centrifugally as the embryo expands (Fig. 5A, B).
1000 Whole emb
800 9
•
Colyledons
O.
600 (.9 Z
9
4OO
/
200
0
•
0
i
I
i
J
I
I
J
1
2
3
4
5
6
7
DRY SEED
GERMINATION
INCUBATION TIME (days)
Fig. 6. Growth of the embryo in celery seed in the first 7 d of imbibition. Germination (radicle protrusion) occurred on day 6
occur immediately adjacent to the embryo (Fig. 2C, D) where the cells become compressed by the enlarging embryo. The cells adjacent to the radicle tip seem to undergo similar changes (Fig. 3 B) but there is one notable difference. Four to 5 d after the beginning of imbibition, while these cells are changing, inclusions which seem to be starch grains appear (Fig. 3 C) but their presence is transient and they are not present when germination occurs on day 6. The grains were identified as starch grains because they were PASpositive (Fig. 3 C) and stained intensely in iodine-potassium iodide solution (not shown). At the time of germination the cells adjacent to the radicle are depleted of stainable contents (Fig. 4D). As the embryo expands it is always surrounded by the same, narrow band of degrading cells (Fig. 4B,
2. Hormone-induced changes in excised tissue." The progress of endosperm degradation gives the impression of being controlled by the advancing embryo, however, such a progression of endosperm modification could occur if endosperm cells were programmed to respond to a stimulus sequentially. Evidence for this was obtained by following the course of endosperm degradation triggered by gibberellin, the only treatment known to replace control of endosperm breakdown by the embryo in celery (Jacobsen et al., 1976). Celery seeds were de-embryonated and cut at the end distal to the embryo to ensure the G A 4 + 7 reached all parts of the endosperm quickly. The endosperm pieces were incubated for various times in GA4+7 and then fixed and sectioned longitudinally (Fig. 7). The 3-6 layers of cells next to a cut surface did not respond to GA4 +7, presumably because they were damaged in some way. Fig. 7A shows endosperm from dry seed. Some cells (darkly stained) contain globoids in the aleurone grains and some (lightly stained) do not. The aleurone grains of the outer layer of endosperm cells are notable by not containing any globoids. The tissue maintains this appearance until on the 4th day (Fig. 7B) a group of cells in the core of the endosperm begins to show signs of responding to GA4+ 7. These changes have been described previously (Jacobsen et al., 1976) and include loss of contents stainable with toluidin blue and a
246
J.V. Jacobsen and E. Pressman: Germination in Celery Seed
Fig. 7 A - F . Changes occurring in endosperm of embryo-less half-seeds of celery incubated in GA,+7. Where it is found the modified zone is outlined with a dashed line. A At the beginning of incubation. B After 4 d incubation with GA4+7. C After 5 d incubation with G A , + > In this section some of the degraded cells have been lost from the endosperm core but this does not always occur. D After 6 d incubation with GA4+7. E After 6 d incubation in water only, showing that endosperm is still intact; control for D. F After l0 d incubation in GAb+ ~. All sections were stained with toluidin blue. x 91
J.V. Jacobsen and E. Pressman: Germination in Celery Seed
change in the colour of staining of the cell walls in toluidin blue from blue to purple. As time progresses, the zone of cell breakdown extends radially (Fig. 7 C0 D) until after 10 d incubation (Fig. 7F), essentially all cells of the endosperm, except those adjacent to a cut surface, have been modified. At all times there is an abrupt transition between unaffected and modified cells. None of the described changes occur in the absence of GA4 +v (Fig. 7 E).
Discussion
Germination in celery is preceeded by major changes in the endosperm and embryo. The embryo more than doubles in length and consumes a large part of the core of the endosperm before radicle protrusion occurs. Such pre-germination enlargement of the embryo has been described in a number of plants, e.g., Fraxinus excelsior (Oleaceae) (Steinbauer, 1937; Villiers and Wareing, 1964), Heracleum sphondylium (Umbelliferae) (Stokes, 1952), Viburnum spp. (Caprifoliaceae) (Giersbach, 1937), and Anemone coronaria (Ranunculaceae) (Bullowa et al., 1975). The seeds of all these species consist largely of living endosperm with a very small embryo at one end. Growth of the embryo within the seed appears to signify that germination must be preceeded by development of greater expansive force in the embryo and-or that the embryo is confined within the endosperm until the cells at the root tip weaken. The results of our study indicate that endosperm breakdown in celery seed is dependent on a stimulus from the embryo the release of which is light dependent. Two possibilities concerning the nature of the stimulus produced by the embryo and causing endosperm mobilization have been considered. One is that the embryo releases enzymes which degrade the endosperm. This has been thought to occur in germinating lettuce seed (Ikuma and Thimann, 1963), and has been accepted as being the likely mechanism for storage-tissue breakdown in other cases (Chen and Thimann, 1964; Homer and Arnott, 1966). The second possibility is that the embryo releases gibberellin which causes the endosperm cells to produce the enzymes necessary for their own modification (Jacobsen et al., 1976). The results in this paper provide evidence, although so far indirect that, for celery seed, the second alternative is more likely to be true. This evidence consists primarily of the demonstration that the changes occurring in endosperm cells adjacent to the embryo during germination are similar to those induced by GA4+7 in isolated endosperm (Jacobsen et al., 1976) although clearly the structural changes require examination with techniques have
247
considerably higher resolving power than those used in this study. Gibberellin is well-known to be released from the cereal embryo (MacLeod and Pahner, 1966; Yomo and Iinuma, 1966; Radley, 1967; 1969) and to initiate changes in the aleurone cells (Jones, 1969; Taiz and Jones, 1970, 1973) many of which are similar to the changes occurring in celery endosperm. The significance of the transient appearance of starch grains in the endosperm near the root tip is unknown and we know of no other reports of such a phenomenon. However, starch formation is known to occur in similar cells of other storage tissues. Cells of the endosperm of castor bean (Vigil, 1970), the female gametophyte of Douglas fir (Ching, 1965) and the perisperm of some species of Yucca (Homer and Arnott, 1966) also contain aleurone grains and lipid droplets and during seed germination, some of the sucrose formed from the lipid (for review see Ching, 1972) is channelled temporarily into starch (Ching, 1966) which may accumulate as grains (Vigil, 1970; Horner and Arnott, 1966). Presumably cells of celery endosperm also metabolize lipid to glucose some of which accumulates temporarily as starch grains, perhaps under the influence of the root cap. Although it is convenient to think of the expanding embryo causing progressive endosperm breakdown, the results of this study show that gibberellinpromoted endosperm modification occurring in the absence of the embryo follows a similar progression from the inner to the outer cells. This raises the possibility that the course of endosperm breakdown during germination is a function of the response time of individual endosperm cells to a stimulus rather than the result of limited diffusion of the stimulus from the embryo into the endosperm. It also raises the possibility that embryo expansion is controlled not by events within the embryo but by the rate at while the endosperm degrades itself and permits embryo enlargement. The authors wish to thank Dr. Menachem Sachs, Department of Vegetable Crops, Volcani Center for obtaining financial assistance for J.V.J., thus making this study possible. We also wish to thank Professor Moshe Negbi, Hebrew University of Jerusalem, Rehovot, Israel, for helpful suggestions and discussions during the work, Mr. M. Gottreich, Division of Subtropical Horticulture, Volcani Center for his guidance in the sectioning work, and Miss Rosemary Metcalf, Plant Physiology Section, CSIRO, Canberra, for her able technical assistance.
References Brink, R.A., Cooper, D.C.: The endosperm in seed-development. Bot. Rev. 13, 423-541 (1947) Bullowa, S., Negbi, M., Ozeri, Y. : Role of temperature, light and growth regulators in germination of Anemone coronaria L. Aust. J. Plant Physiol. 2, 91-100 (1975)
248 Chen, S.S.C., Thimann, K.V. : Studies on the germination of lightinhibited seeds of Phacelia tanacetifolia. Israel J. Bot. 13, 57 73 (1964) Ching, T.M. : Metabolic and ultrastructural changes in germinating Douglas fir seeds. (Abstr.) Plant Physiol. 40, Suppl., 8 (1965) Ching, T.M. : Compositional changes of Douglas fir seeds during germination. Plant Physiol. 41, 1313-1319 (1966) Ching, T.M. : Metabolism of germinating seeds. In: Seed Biology, vol. II, Germination control, metabolism and pathology, pp. 103518, Kozlowski, T.T., ed. New York: Academic Press 1972 Giersbach, J. : Germination and seedling production of species of Viburnum. Contrib. Boyce Thompson Inst. 9, 79 90 (1937) Homer, H.T., Arnott, H.J.: A histochemicaI and ultrastructural study of Yucca seed proteins. Am. J. Bot. 52, 1027-1038 (1965) Horner, H.T., Arnott, H.J.: A histochemical and ultrastructural study of pre- and post-germinated Yucca seeds. Bot Gaz. 127, 48 64 (1966) Ikuma, H., Thimann, K.V. : The role of seed coats in germination of photosensitive lettuce seeds. Plant Cell Physiol. 4, 169 185 (1963) Jacobsen, J.V., Knox, R.B., Pyliotis, N.A. : The structure and composition of aleurone grains in the barley aleurone layer. Planta 101, 189 209 (1971) Jacobsen, J.V., Pressman, E., Pyliotis, N.A.: Gibberellin-induced separation of cells in isolated endosperm of celery seed. Planta 129, 113-122 (1976) Jensen, W.A.: Botanical histochemistry. San Francisco: Freeman 1962 Jones, R.L. : Gibberellic acid and the fine structure of barley aleurone cells. II. Changes during the synthesis and secretion of c~-amylase. Planta 88, 73-86 (1969) MacLeod, A.M., Palmer, G.H. : The embryo of barley in relation to modification of the endosperm. J. Inst. Brew. (Lond.) 72, 580 589 (1966) Maheshwari, P.: An introduction to the embryology of angiosperms. New York: McGraw-Hill 1950 Martin, A.C. : The comparative internal morphology of seeds. Am. Midland Naturalist 36, 513-660 (1946)
J.V. Jacobsen and E. Pressman: Germination in Celery Seed Morinaga, T. : Effect of alternating temperatures upon the germination of seeds. Am. J. Bot. 13, 141 158 (1926) Pressman, E., Negbi, M., Sachs, M., Jacobsen, J.V. : Varietal differences in light-requirements for germination of celery (Apium graveolens L.) seeds and the effects of thermal and solute stress. Aust. J. Plant Physiol. 4, 821 831 (1977) Radley, M.: Site of production of gibberellin-like substances in germinating barley embryos. Planta 75, 164-171 (1967) Radley, M.: The effect of the endosperm on the formation of gibberellin by barley embryos. Planta 86, 218-223 (1969) Steinbauer, G.P.: Dormancy and germination of Fraxinus seeds. Plant Physiol. 12, 813-824 (1937) Stokes, P. : A physiological study of embryo development in Heracleum sphondylium L. I. The effect of temperature on embryo development. Ann. Bot. 16, 441-447 (1952) Taiz, L., Jones, R.L.: Gibberellic acid, fl-l,3-glucanase and the cell walls of barley aleurone layers. Planta 92, 73 84 (1970) Taiz, L., Jones, R.L.: Plasmodesmata and an associated cell wall component in barley aleurone tissue. Am. J. Bot. 60, 67 75 (1973) Taylor, C.A.: Some factors affecting germination of celery seed. Plant Physiol. 24, 93-102 (1949) Varner, J.E., Ho, D.T.: The role of hormones in the integration of seedling growth. In: The molecular biology of hormone action, pp. 173-194, Papaconstantinou, J., ed. New York: Academic Press 1976 Vigil, E.L.: Cytochemical and developmental changes in microbodies (glyoxysomes) and related organelles of castor bean endosperm. J. Cell Biol. 46, 435454 (1970) Villiers, T.A., Wareing, P.F. : Dormancy in fruit of Fraxinus excelsior L. J. Exp. Bot. 44, 359-367 (1964) Yomo, H., Iinuma, H.: Production of gibberellin-like substance in the embryo of barley during germination. Planta 71, 113 118 (1966) Yomo, H., Varner, J.E.: Hormonal control of a secretory tissue. Curr. Topics Devel. Biol. 6, 111 144 (1971)
Received 1 June; accepted 4 October 1978
Planta 144, 241-248 (1979)
9 by Springer-Verlag i979
A Structural Study of Germination in Celery with Emphasis on Endosperm Breakdown
(Apium graveolens L.)
Seed
J.V. Jacobsen 1 and E. Pressman 2 1 Division of Plant Industry, CSIRO, P.O. Box 1600, Canberra, A.C.T. 2601, Australia, and 2 Volcani Center, Agricultural Research Organization, Bet Dagan, Israel
Abstract. Germination of celery seed occurred after
6 d of imbibition in light. During this time the embryo enlarged at the expense of the adjacent endosperm cells and at the time of germination was 2-3 times as long as in the dry seed. Breakdown of the endosperm cells near the root cap preceeded radicle emergence. None of these changes occurred in darkness. Endosperm digestion began adjacent to the embryo and spread radially. In degrading cells, the aleurone grains often became larger and fewer in number. The cell walls were modified and appeared to undergo partial degradation. Ultimately the cells seemed to lose their contents. In cells adjacent to the root cap, similar changes occurred except there was a transient appearance of starch grains. Radial progression of endosperm breakdown also occurred in isolated endosperm treated with gibberellin A4 + 7. The results indicate that (1) the stimulus for breakdown of celery endosperm emanates from the embryo in response to light; (2) the stimulus may be a gibberellin because changes in endosperm cells and the sequence of endosperm digestion during germination resemble the responses of isolated endosperm to gibberellin; and (3) the radial progression of endosperm breakdown during germination may be the result of a sequential response of cells to a uniformly applied stimulus rather than the result of gradual embryo expansion. Key words: Apium - Endosperm breakdown - Germi-
nation (seeds).
Introduction
Seed germination involves processes and control mechanisms most of which are only partially understood. Although there has been considerable effort
applied to the study of structure and function in embryogeny (see the book by Maheshwari, 1950) there has been relatively little work on similar aspects of seeds during germination. Most of the work done on the cellular structure and physiology of germinating seed has been concerned with the embryo and with the exception of cereal grains, the endosperm or other non-embryonic nutritive tissue has usually been of peripheral interest and studied intensively only in a few cases such as castor bean (Ricinus communis), carob bean (Ceratonia siliqua), and guar (Cya-
mopsis tetragonoloba). Extra-embryonic reserves in seeds can occur in a variety of tissues. In many dicotyledonous and m o n o c o t y l e d o n o u s seeds the reserves occur in endosperm which can vary in quantity from non-existent at maturity (exalbuminous) to occupying the larger part of the seed. Endosperm can consist of all living cells or of some living and some dead cells as in the cereals. In some angiosperms, the storage reserves occur largely in perisperm (persistent nucellus) as for example in Yucca (Horner and Arnott, 1965) and in gymnosperms the storage tissue is the gametophyte tissue (for review see Brink and Cooper, 1947). Both perisperm and gametophyte tissues are living. All of these tissues are digested as the embryo enlarges and germination and seedling growth occur. In the cereals, the structure and composition of the endosperm and control of its mobilization during germination are fairly well understood (see reviews of Yomo and Varher, 1971, and Varner and Ho, 1976) but there is little understanding of the controls operating during germination in other plant families. The seeds of the Umbelliferae contain relatively large amounts of living endosperm which completely surrounds a small embryo located at one end of the seed (for review see Martin, 1946). In a previous study, we described how isolated endosperm of celery was induced to autolyse by gibberellin (Jacobsen
0032-0935/79/0144/0241/$ 01.60
242 et al., 1976). M a j o r changes occurred in the cell contents a n d the cell walls were partially h y d r o l y z e d so that the cells became separated f r o m each other. This led us to propose that d u r i n g g e r m i n a t i o n o f celery seed, e n d o s p e r m d e g r a d a t i o n m a y be c o n t r o l l e d by a h o r m o n e , or by h o r m o n e s , released from the e m b r y o rather t h a n by enzymes released by the e m b r y o , a n d in this study we have sought evidence to s u p p o r t this hypothesis.
Material and Methods Celery seed (Apium graveolens L. cv. Florida 683; Ferry Morse Comp., Mountain View, Cal., USA) was placed on moist filter paper in Petri dishes, and incubated in darkness or constant light at 20~ C. Light is required for germination of celery seed at this temperature (Morinaga, 1926; Taylor, 1949). After the required times, a slice was cut longitudinally from each side of the seed to facilitate entry of fixative and embedding medium. The seed was fixed usually for 2-3 d in 3% glutaraldehyde in 0.025 M phosphate buffer, pH 7.2, in the refrigerator, and processed for embedding in paraffin as described in Jensen (1962, p. 80-83). Paraffin was used because large numbers of seed had to be sectioned serially and for the purpose of this study maximal resolution of fine cellular structure was not required. Sections were cut 10 pm thick and stained in 0.05% toluidin blue in water, or in periodic acid-Schiff reagent to detect insoluble polysaccharides (Jensen, I962, p. 198-199).
J.V. Jacobsen and E. Pressman: Germination in Celery Seed De-embryonated seeds (half-seeds) were sterilized and incubated with and without 10-s M GA4+7 (from Prof. J. MacMillan, University of Bristol, U.K.) as previously described (Jacobsen et al., 1976). After the required times, half-seeds were fixed (with no further cutting), embedded, sectioned and stained as described above.
Results A t 20 ~ C in light celery seeds begin to germinate o n the 6th day after the c o m m e n c e m e n t of i m b i b i t i o n (Pressman et al., 1977). T h e structural changes occurring in the seed d u r i n g this time a n d up to 9 d are s h o w n in Fig. 1-5.
Emb~o I n dry seed (Fig. t A, B) the e m b r y o is small, centrally located a n d s u r r o u n d e d by e n d o s p e r m a l t h o u g h the e n d o s p e r m a d j a c e n t to the radicle is only 3 or 4 cells thick. The zone i m m e d i a t e l y a d j a c e n t to the e m b r y o ( D C in Fig. 1 B) appears to c o n t a i n degraded cells. Two d f r o m the b e g i n n i n g of i m b i b i t i o n , the e m b r y o begins to elongate within the seed a n d as time progresses the cotyledons extend into the e n d o s p e r m
Fig. 1 A-C. Longitudinal sections of dry celery seed. A The whole seed. x 51. B A higher magnification of the embryo zone of A. x 110. C The endosperm cells at the side of the embryo. The arrows point to aleurone grain inclusions staining red in toluidin blue. • 440. All sections were stained with toluidin blue. En=endosperm; Em=embryo; SC=seed coat; DC = depleted cells
J.V. Jacobsen and E. Pressman: Germination in Celery Seed
243
Fig. 2A-D. Longitudinal sections of celery seed after 4 d incubation in water. A Whole seed. x 51. B The endosperm at the radicle tip. x 275. C Endosperm adjacent to the cotyledons, x 275. D A zone similar to that shown in C but stained with PAS to demonstrate that some cell-wall material remains around depIeted cells adjacent to the embryo, x 275. Sections in A, B and C were stained in toluidin blue. R=radicle; AG=aleurone grain; A=embryo axis; C=cotyledon
(Fig. 2A) and the radicle is appressed to the endosperm (Fig. 2B). At 5 d, the embryo occupies about 2/3 the length of the endosperm (Fig. 3A) but the radicle is still within the endosperm (Fig. 3B). On the 6th day the radicle penetrates the endosperm and the seed coats, and emerges (Fig. 4A). As the radicle elongates further, so do the cotyledons and progressively more of the endosperm is consumed (Fig. 5A, B). Measurement of embryo growth (Fig. 6) shows that elongation is comprised of concurrent increases in the axis and the cotyledons. In darkness, the embryo does not elongate within the seed and there are no changes in the endosperm cells.
Endosperm 1. Changes during germination. The dry endosperm tissue consists largely of angular cells which have thick (3-5 ~tln) walls and which contain m a n y aleurone grains (Fig. 1 B, C and Jacobsen et al., 1976).
The aleurone grains of some cells contain round inclusions (see arrows Fig. 1 C) which stain red with toluidin blue and presumable contain phytin (Jacobsen et al., 1971) while the grains of other cells contain angular inclusions which do not stain and which are not visible in the photographs. The two types of inclusions do not occur in the same cell. The cells adjacent to the radicle appear similar to, although much smaller than, those in the bulk of the endosperm. As the embryo expands the adjacent endosperm cells are transformed. Some aleurone grains become larger (see A G in Fig. 2C) and the cell contents become less intensely stained by toluidin blue and less discernible. Ultimately the cell contents are either lost, or fail completely to be stained by toluidin blue (see cells adjacent to embryonic axis in Fig. 2C). The ceil walls become irregular in form and stain purple in toluidin blue, as opposed to the usual blue colour. The cell-wall material which does persist is strongly PAS-positive (Fig. 2D). These changes occur all around the embryo. The early stages of change occur 3-5 cells away from the embryo while advanced stages
Fig. 3 A - C . Longitudinal sections o f celery seed after 5 d incubation in water. A Whole seed. x 51. B Endosperm ceils adjacent to the root tip. x 275. C The root tip zone stained with PAS showing starch grains in the cells of both the root tip and the adjacent endosperm. • 275. Sections in A and B were stained with toluidin blue. S = starch grains
Fig. 4 A - D . Longitudinal sections of celery seed after 6 d incubation in water. A Whole seed. x 51. B, C Endosperm adjacent to the cotyledons, x 110 and x 275 respectively. D Endosperm beside the embryo, showing the depleted cells adjacent to the emerged radicle, x 110. All sections were stained with toluidin blue. A p = embryo apex
J.V. Jacobsen and E. Pressman: Germination in Celery Seed
245
Fig. 5A and B. Longitudinal sections of celery seed after 7 d (A) and 9 d (B) of incubation in water. Sections were stained with toluidin blue. • 51
1200
C). The transition from normal to apparently depleted cells is clearly quite rapid. Cells with no stainable contents exist next to cells with many aleurone grains (Figs. 2C, D and 4C). The depleted cells at the radicle tip, which are penetrated by the radicle, are continuous with the degrading band of cells around the embryo (Fig. 4 D). The band of degrading endosperm proceeds centrifugally as the embryo expands (Fig. 5A, B).
1000 Whole emb
800 9
•
Colyledons
O.
600 (.9 Z
9
4OO
/
200
0
•
0
i
I
i
J
I
I
J
1
2
3
4
5
6
7
DRY SEED
GERMINATION
INCUBATION TIME (days)
Fig. 6. Growth of the embryo in celery seed in the first 7 d of imbibition. Germination (radicle protrusion) occurred on day 6
occur immediately adjacent to the embryo (Fig. 2C, D) where the cells become compressed by the enlarging embryo. The cells adjacent to the radicle tip seem to undergo similar changes (Fig. 3 B) but there is one notable difference. Four to 5 d after the beginning of imbibition, while these cells are changing, inclusions which seem to be starch grains appear (Fig. 3 C) but their presence is transient and they are not present when germination occurs on day 6. The grains were identified as starch grains because they were PASpositive (Fig. 3 C) and stained intensely in iodine-potassium iodide solution (not shown). At the time of germination the cells adjacent to the radicle are depleted of stainable contents (Fig. 4D). As the embryo expands it is always surrounded by the same, narrow band of degrading cells (Fig. 4B,
2. Hormone-induced changes in excised tissue." The progress of endosperm degradation gives the impression of being controlled by the advancing embryo, however, such a progression of endosperm modification could occur if endosperm cells were programmed to respond to a stimulus sequentially. Evidence for this was obtained by following the course of endosperm degradation triggered by gibberellin, the only treatment known to replace control of endosperm breakdown by the embryo in celery (Jacobsen et al., 1976). Celery seeds were de-embryonated and cut at the end distal to the embryo to ensure the G A 4 + 7 reached all parts of the endosperm quickly. The endosperm pieces were incubated for various times in GA4+7 and then fixed and sectioned longitudinally (Fig. 7). The 3-6 layers of cells next to a cut surface did not respond to GA4 +7, presumably because they were damaged in some way. Fig. 7A shows endosperm from dry seed. Some cells (darkly stained) contain globoids in the aleurone grains and some (lightly stained) do not. The aleurone grains of the outer layer of endosperm cells are notable by not containing any globoids. The tissue maintains this appearance until on the 4th day (Fig. 7B) a group of cells in the core of the endosperm begins to show signs of responding to GA4+ 7. These changes have been described previously (Jacobsen et al., 1976) and include loss of contents stainable with toluidin blue and a
246
J.V. Jacobsen and E. Pressman: Germination in Celery Seed
Fig. 7 A - F . Changes occurring in endosperm of embryo-less half-seeds of celery incubated in GA,+7. Where it is found the modified zone is outlined with a dashed line. A At the beginning of incubation. B After 4 d incubation with GA4+7. C After 5 d incubation with G A , + > In this section some of the degraded cells have been lost from the endosperm core but this does not always occur. D After 6 d incubation with GA4+7. E After 6 d incubation in water only, showing that endosperm is still intact; control for D. F After l0 d incubation in GAb+ ~. All sections were stained with toluidin blue. x 91
J.V. Jacobsen and E. Pressman: Germination in Celery Seed
change in the colour of staining of the cell walls in toluidin blue from blue to purple. As time progresses, the zone of cell breakdown extends radially (Fig. 7 C0 D) until after 10 d incubation (Fig. 7F), essentially all cells of the endosperm, except those adjacent to a cut surface, have been modified. At all times there is an abrupt transition between unaffected and modified cells. None of the described changes occur in the absence of GA4 +v (Fig. 7 E).
Discussion
Germination in celery is preceeded by major changes in the endosperm and embryo. The embryo more than doubles in length and consumes a large part of the core of the endosperm before radicle protrusion occurs. Such pre-germination enlargement of the embryo has been described in a number of plants, e.g., Fraxinus excelsior (Oleaceae) (Steinbauer, 1937; Villiers and Wareing, 1964), Heracleum sphondylium (Umbelliferae) (Stokes, 1952), Viburnum spp. (Caprifoliaceae) (Giersbach, 1937), and Anemone coronaria (Ranunculaceae) (Bullowa et al., 1975). The seeds of all these species consist largely of living endosperm with a very small embryo at one end. Growth of the embryo within the seed appears to signify that germination must be preceeded by development of greater expansive force in the embryo and-or that the embryo is confined within the endosperm until the cells at the root tip weaken. The results of our study indicate that endosperm breakdown in celery seed is dependent on a stimulus from the embryo the release of which is light dependent. Two possibilities concerning the nature of the stimulus produced by the embryo and causing endosperm mobilization have been considered. One is that the embryo releases enzymes which degrade the endosperm. This has been thought to occur in germinating lettuce seed (Ikuma and Thimann, 1963), and has been accepted as being the likely mechanism for storage-tissue breakdown in other cases (Chen and Thimann, 1964; Homer and Arnott, 1966). The second possibility is that the embryo releases gibberellin which causes the endosperm cells to produce the enzymes necessary for their own modification (Jacobsen et al., 1976). The results in this paper provide evidence, although so far indirect that, for celery seed, the second alternative is more likely to be true. This evidence consists primarily of the demonstration that the changes occurring in endosperm cells adjacent to the embryo during germination are similar to those induced by GA4+7 in isolated endosperm (Jacobsen et al., 1976) although clearly the structural changes require examination with techniques have
247
considerably higher resolving power than those used in this study. Gibberellin is well-known to be released from the cereal embryo (MacLeod and Pahner, 1966; Yomo and Iinuma, 1966; Radley, 1967; 1969) and to initiate changes in the aleurone cells (Jones, 1969; Taiz and Jones, 1970, 1973) many of which are similar to the changes occurring in celery endosperm. The significance of the transient appearance of starch grains in the endosperm near the root tip is unknown and we know of no other reports of such a phenomenon. However, starch formation is known to occur in similar cells of other storage tissues. Cells of the endosperm of castor bean (Vigil, 1970), the female gametophyte of Douglas fir (Ching, 1965) and the perisperm of some species of Yucca (Homer and Arnott, 1966) also contain aleurone grains and lipid droplets and during seed germination, some of the sucrose formed from the lipid (for review see Ching, 1972) is channelled temporarily into starch (Ching, 1966) which may accumulate as grains (Vigil, 1970; Horner and Arnott, 1966). Presumably cells of celery endosperm also metabolize lipid to glucose some of which accumulates temporarily as starch grains, perhaps under the influence of the root cap. Although it is convenient to think of the expanding embryo causing progressive endosperm breakdown, the results of this study show that gibberellinpromoted endosperm modification occurring in the absence of the embryo follows a similar progression from the inner to the outer cells. This raises the possibility that the course of endosperm breakdown during germination is a function of the response time of individual endosperm cells to a stimulus rather than the result of limited diffusion of the stimulus from the embryo into the endosperm. It also raises the possibility that embryo expansion is controlled not by events within the embryo but by the rate at while the endosperm degrades itself and permits embryo enlargement. The authors wish to thank Dr. Menachem Sachs, Department of Vegetable Crops, Volcani Center for obtaining financial assistance for J.V.J., thus making this study possible. We also wish to thank Professor Moshe Negbi, Hebrew University of Jerusalem, Rehovot, Israel, for helpful suggestions and discussions during the work, Mr. M. Gottreich, Division of Subtropical Horticulture, Volcani Center for his guidance in the sectioning work, and Miss Rosemary Metcalf, Plant Physiology Section, CSIRO, Canberra, for her able technical assistance.
References Brink, R.A., Cooper, D.C.: The endosperm in seed-development. Bot. Rev. 13, 423-541 (1947) Bullowa, S., Negbi, M., Ozeri, Y. : Role of temperature, light and growth regulators in germination of Anemone coronaria L. Aust. J. Plant Physiol. 2, 91-100 (1975)
248 Chen, S.S.C., Thimann, K.V. : Studies on the germination of lightinhibited seeds of Phacelia tanacetifolia. Israel J. Bot. 13, 57 73 (1964) Ching, T.M. : Metabolic and ultrastructural changes in germinating Douglas fir seeds. (Abstr.) Plant Physiol. 40, Suppl., 8 (1965) Ching, T.M. : Compositional changes of Douglas fir seeds during germination. Plant Physiol. 41, 1313-1319 (1966) Ching, T.M. : Metabolism of germinating seeds. In: Seed Biology, vol. II, Germination control, metabolism and pathology, pp. 103518, Kozlowski, T.T., ed. New York: Academic Press 1972 Giersbach, J. : Germination and seedling production of species of Viburnum. Contrib. Boyce Thompson Inst. 9, 79 90 (1937) Homer, H.T., Arnott, H.J.: A histochemicaI and ultrastructural study of Yucca seed proteins. Am. J. Bot. 52, 1027-1038 (1965) Horner, H.T., Arnott, H.J.: A histochemical and ultrastructural study of pre- and post-germinated Yucca seeds. Bot Gaz. 127, 48 64 (1966) Ikuma, H., Thimann, K.V. : The role of seed coats in germination of photosensitive lettuce seeds. Plant Cell Physiol. 4, 169 185 (1963) Jacobsen, J.V., Knox, R.B., Pyliotis, N.A. : The structure and composition of aleurone grains in the barley aleurone layer. Planta 101, 189 209 (1971) Jacobsen, J.V., Pressman, E., Pyliotis, N.A.: Gibberellin-induced separation of cells in isolated endosperm of celery seed. Planta 129, 113-122 (1976) Jensen, W.A.: Botanical histochemistry. San Francisco: Freeman 1962 Jones, R.L. : Gibberellic acid and the fine structure of barley aleurone cells. II. Changes during the synthesis and secretion of c~-amylase. Planta 88, 73-86 (1969) MacLeod, A.M., Palmer, G.H. : The embryo of barley in relation to modification of the endosperm. J. Inst. Brew. (Lond.) 72, 580 589 (1966) Maheshwari, P.: An introduction to the embryology of angiosperms. New York: McGraw-Hill 1950 Martin, A.C. : The comparative internal morphology of seeds. Am. Midland Naturalist 36, 513-660 (1946)
J.V. Jacobsen and E. Pressman: Germination in Celery Seed Morinaga, T. : Effect of alternating temperatures upon the germination of seeds. Am. J. Bot. 13, 141 158 (1926) Pressman, E., Negbi, M., Sachs, M., Jacobsen, J.V. : Varietal differences in light-requirements for germination of celery (Apium graveolens L.) seeds and the effects of thermal and solute stress. Aust. J. Plant Physiol. 4, 821 831 (1977) Radley, M.: Site of production of gibberellin-like substances in germinating barley embryos. Planta 75, 164-171 (1967) Radley, M.: The effect of the endosperm on the formation of gibberellin by barley embryos. Planta 86, 218-223 (1969) Steinbauer, G.P.: Dormancy and germination of Fraxinus seeds. Plant Physiol. 12, 813-824 (1937) Stokes, P. : A physiological study of embryo development in Heracleum sphondylium L. I. The effect of temperature on embryo development. Ann. Bot. 16, 441-447 (1952) Taiz, L., Jones, R.L.: Gibberellic acid, fl-l,3-glucanase and the cell walls of barley aleurone layers. Planta 92, 73 84 (1970) Taiz, L., Jones, R.L.: Plasmodesmata and an associated cell wall component in barley aleurone tissue. Am. J. Bot. 60, 67 75 (1973) Taylor, C.A.: Some factors affecting germination of celery seed. Plant Physiol. 24, 93-102 (1949) Varner, J.E., Ho, D.T.: The role of hormones in the integration of seedling growth. In: The molecular biology of hormone action, pp. 173-194, Papaconstantinou, J., ed. New York: Academic Press 1976 Vigil, E.L.: Cytochemical and developmental changes in microbodies (glyoxysomes) and related organelles of castor bean endosperm. J. Cell Biol. 46, 435454 (1970) Villiers, T.A., Wareing, P.F. : Dormancy in fruit of Fraxinus excelsior L. J. Exp. Bot. 44, 359-367 (1964) Yomo, H., Iinuma, H.: Production of gibberellin-like substance in the embryo of barley during germination. Planta 71, 113 118 (1966) Yomo, H., Varner, J.E.: Hormonal control of a secretory tissue. Curr. Topics Devel. Biol. 6, 111 144 (1971)
Received 1 June; accepted 4 October 1978
