Planta (Berl.) 113, 263--270 (1973) 9 by Springer-Verlag 1973
The Development of Haploid Embryoids from Anther Cultures of Atropa belladonna L. A. Rashid and H. E. Street Botanical Laboratories, University of Leicester, Leicester, LE1 7RH, U.K. l~eceived May 14 / June 5, 1973
Summary. Development of haploid embryoids from the microspores of Atropa belladonna occurs with relatively high frequency when anthers are excised from buds in which the petals are shorter than the sepals (at this stage microspores are predominantly uninucleate) and cultured on a medium containing iron as the ferric salt of ethylenediamine-di-O-hydroxyphenylacetic acid (FeEDDHA). Additions of combinations of kinetin, auxin and casamino-acids to the culture medium induce callusing in both haploid and diploid tissues, lead to tile origin of embryoids from somatic tissues of the anther and should be avoided. Simple techniques for the maintenance of haploid clones are described. Stages in early embryogenesis in the pollen grains have been observed and these indicate that embryogenesis is most frequently initiated by an equal division in the uninucleate spore. The frequency of grains showing embryoid formation is very low and it is estimated that plantlets are formed from up to 50% of these grains, Introduction Zenkteler (1971) has previously described the occurrence of embryogenesis in anthers of Atropa belladonna cultured on a medium containing an auxin and a cytokinin and reported the occurrence of haploid, diploid and triploid embryoids. Other workers have similarly reported the production of haploid and diploid plantlets by anther cultures of other Solanaceae cultured on media containing auxin and cytokinin or enriched with undefined natural supplements (Nishi and Mitsouka, 1969; Narayanaswamy and Chandy, 1971; Sunderland and Wicks, 1971; Iyer and l~aina, 1972). In a number of these eases callus development has been reported as arising either from the anther tissues or from the young embryoids. One objective of the present study was to define the nutritive conditions which are promotive of haploid embryoid development and suppressive of callus formation in Atropa belladonna. In Nieotiana spp. embryoid development has been reported from anthers with microspores at the tetrad stage (Nakata and Tanaka, 1968), at the uninucleate (Nitsch, 1969) and at the binucleate stages (Sunderland, 1971; Sunderland and Wicks, 1971). I n Nicotiana tabacum the uninucleate microspore divides to give a microspore containing a vegetative and a generative cell and when embryoids are formed from such
264
A. Rashid and H. E. Street
grains t h e y are derived from the vegetative cell. However, occasionally, the division of the u n i n u c l e a t e spore has been observed to be s y m m e t r i c a l a n d in these cases both cells have been considered to be i n v o l v e d i n e m b r y o i d f o r m a t i o n ( S u n d e r l a n d a n d Wicks, 1971). N a r a y a n a s w a m y a n d C h a n d y (1971) have described equal division in u n i n u c l e a t c micro. spores of cultured a n t h e r s of Datura metel, b u t i n c o n t r a s t I y e r a n d I~aina (1972), working with the same species, have described the first division as u n e q u a l a n d considered the embryoids to arise from the vegetative cell. A second objective of the present s t u d y was to follow the microspore divisions i n v o l v e d in e m b r y o i d f o r m a t i o n i n Atropa belladonna a n d to decide, if possible, the mierospore stage from which pollen-embryoids can be most readily obtained.
Materials and Methods A succession of flowering Atropa belladonna plants grown under long days was available at intervals of 4--5 weeks over the period June to October 1972. Flower buds with petals shorter than the sepals (Group I) and flower buds with petals equal or slightly longer than the sepals (Group II) were used. Pollen grains in anthers from Group II stained quickly and densely with acetocarmine and were rich in starch grains. Grains from Group I anthers stained only lightly with acetocarmine even after 30 min and did not contain starch. Both groups showed a single centrally-placed diffuse nucleus when stained with acetocarmine. Grains from Stage II buds stained in propionic orcein, usually showed the presence of a peripherally located, densely staining nucleus (generative nucleus); this was not seen in grains from Stage I buds. The grains from Stage ~ buds therefore appeared to be uninucleate, those from Stage II to contain a vegetative and generative nucleus.
For each flowering, buds were taken from during the second week of flowering until seed formation was occurring in the first formed flowers. However, embryoid development in culture was most abundant from anthers obtained early in each flowering period and it is the results of anther cultures from these harvests which are described. Sepals were removed from the flower buds and the buds then immersed in 90% ethanol for 2-3 min, washed with sterile water and dissected under aseptic conditions. The anthers, with most of the filament excised, were transferred to the surface of 10 ml medium solidified with 0.5% Difco bacto-agar and contained in a screw-capped McCartney bottle of 30 ml capacity. The anthers from any one bud (i.e. up to 5 anthers, though in some cases 4 or 3 due to individual anthers being dropped) were included in the same culture bottle evenly spaced over the medium surface. The basic medium contained 2% sucrose, 100 rag/1 meso-inositol, vitamins after White (1954) and the mineral salts either after Linsmaer and Skoog (1965) (LS medium) or after ~itsch (1969) (N medium). Iron was added as FeEDTA (10-4 iV[) (Shear et aI., 1959) or as FeEDDHA (the ferric salt of ethylenediaminedi-O-hydroxyphenylacetic acid). The media were variously supplemented with casamino acids 0.1 To w/v (CA), kinetin (kin) 10-5 M and 10-6 M, naphthaleneacetic acid (NAA) 10-6 M, and indolyl-aeetic acid (IAA) 10-5 and 10-6 M. pH of the medium was adjusted to pH 5.5 before autoclaving. Cultures were incubated at 25=[_0.5~ in continuous warm white fluorescent light (100 lumens per sq ft) for up to 12 weeks.
Haploid Embryoids from Anther Cultures
265
At intervals during culture, individual anthers were dissected and the pollen grains stained in acetocarmine and propionic orcein. The ploidy of embryoids and established plantlets was determined by treatment of root tips for 2 h in saturated paradichlorobenzene solution, fixation in acetic alcohol, hydrolysis for 12 rain at 56~ C in 1 N HC1, staining with Fculgen reagent and squashing in acetocarmine.
Results and Discussion
Cultural Conditions Conducive to the Development o/ Haploid Plantlets Very few anthers from Stage I I buds cultured on the basic LS medium gave embryoids ; none produced embryoids on LS medium supplemented with CA, Kin and NAA or with all three supplements but callus development occurred from the residual filament and from internal anther tissues. Anthers from Stage I I buds frequently dehisced soon after being placed in culture and showed no further development except occasional eMlus proliferation from the anther wall. Anthers from Stage I buds only rarely dehisced prematurely and produced embryoids with higher frequency on the LS-based media; in medium supplemented by CA, Kin and NAA, 50 % of the anthers gave rise to embryoids but the anthers also showed callus development from the residual filament and from the anther wall and when individual anthers were examined it was found that certain of the plantlets were arising from the connective. Anthers from Stage I I buds cultured in N medium failed to produce embryoids except in media supplemented with Kin and IAA; with 10-5 M Kin plus 10.5 M IAA 24% of the anthers produced embryoids but these anthers again showed callus development and evidence that at least some embryoids were arising from the connective. The use of IAA instead of NAA did not result in reduced callus formation (Thomas and Street, 1972). Anthers from Stage I buds again showed a higher incidence of embryoid formation in the N-based media and embryoids were obtained in the absence of eMlus formation on the basic medium and in this medium in which the FeEDTA was replaced by F e E D D H A at 10.5 M (4.35 mg/1). Plantlets from these anthers were all haploid. The use of F e E D D H A delayed the appearance of embryoids as compared with FeEDTA. A further experiment confined to Stage I buds and involving the N medium containing 1, 2 or 3 % sucrose and with F e E D D H A at various concentrations resulted in embryoid development in all the test media. There was no evidence of callus formation and all plantlets examined were haploid. In media with 2% sucrose and 5 mg/1 F e E D D I I A , 36% ; and with 3% sucrose and 15mg/1 FeEDDtIA, 40% of the anthers yielded embryoids.
266
A. Rashid and H. E. Street
The results indicated that Stage I buds gave rise to embryoids more readily than Stage I I buds; with Stage I I buds embryoid formation was associated with callus formation. Even with Stage I buds, there were always some cultures in which no anthers gave rise to embryoids and only rarely did all of the 4 or 5 anthers of each culture show embryoid development. In some cases embryoids only emerged from one half of the anther although microscopic examination then always showed the other half of the anther to contain embryoids at a less advanced stage of development. The number of emergent embryoids per anther was also variable. The majority of counts for cmbryoids per anther lay between 13-21 and the highest count was 30. With media inducing callus formation anthers usually burst within 4 weeks to reveal embryoids; with media not inducing callus formation anthers burst within 6-12 weeks of initiating the cultures. When a high number (3 or 4) of the anthers of the culture produced embryoids, these tended to show fasciation which disappeared when the plantlets were removed and cultured in separate culture bottles. This fasciation, therefore, m a y be a "crowding" effect possibly in response to a volatile factor accumulating within the confined space of the closed culture bottle. Irrespective of "crowding" some embryoids were polycotyledonary, some showed fusion of cotyledons and others showed branching of the embryoid axis. Some of the embryoids arising in media containing F e E D D t I A developed a corolla-like tubular structure which was pigmented like the corolla of the normal flowers. If embryoids were not detached and transferred to new culture bottles, they frequently showed the formation of daughter embryoids superficially on the embryoid axis. Similar embryo-like structures were occasionally observed during plantlet deve]opment (i.e. at a later stage when leaves were differentiated) on the plantlet axis or on young leaves. The main axis of developing plantlets always produced adventitious roots. Plantlets were cultured in medium lacking auxin and cytokinin since these growth regulators inhibit axis elongation and induced callusing of the embryoid tissues. Starting with individual haploid plantlets of anther origin, propagation and multiplication without loss of haploidy has been achieved by a combination of techniques: (1) By culture of 5 m m long stem apices in the N medium; (2) by culture of excised leaves which rooted in the N medium and gave rise to shoot buds at the cut base, particularly in the presence of 10 .7 M 6-y-y-dimethylallylaminopurine (DMAAP); (3) by culture of excised root tips in the root culture medium of Street and McGregor (1952) supplemented with 10 -~ rag/1 IAA and induction of bud formation from the callus formed at the basal end of the root by incorporating DMAAP into the medium.
Haploid Embryoids from Anther Cultures
267
Stages in Embryogenesis /tom Pollen Observations on embryogenesis were based upon studies of Stage I anthers cultured in N medium containing FeEDDHA. As indicated in Materials and Methods, microspores of Stage I anthers stained only lightly with acetoearmine. By day 28 of culture from 0 to 30 mierospores per anther were noted which were slightly enlarged and promptly and densely stained with acetocarmine (Fig. 1 A). Many other grains stained very slightly and appeared to be collapsed. The deeply staining grains appeared to be potentially embryogenic and on 7 occasions, in different anthers, such grains which had undergone an equal division were observed (Fig. 1B). Grains of this type stained with Feulgen have been observed with 2 equally staining nuclei (Fig. 1 C). On a single occasion a grain stained in Feulgen contained three nuclei, two staining diffusely and one densely (the latter possibly representing the nucleus of a generative cell) (Fig. 1D). Slightly later in cultures pollen grains showing early stages of embryogenesis were observed and usually all the cells appeared to be similar and could be readily visualised as deriving from grains in which the initial division had been equal. Fig. 1 E, F, G show 4, 5 and 6 cells of this kind within the pollen grain wall. In one case a grain was observed which could be interpreted as containing 5 cells of this type plus a generative cell ( / Fig. 1 H). On another occasion a grain was observed containing 6 "vegetative "-like cells and two distinct " generative "-like cells but this was not recorded photographically. In older cultures pairs of embryoids fused at their base were observed, and at the stage when young embryoids were present we have twice observed a grain bursting prematurely and apparently containing two pro-embryoids (Fig. 1 I). When anthers which failed to develop embryoids were examined, they always contained abnormal enlarged grains packed with starch grains (Fig. 1 J). These grains were often associated together in small aggregates. Occasionally grains were observed suggesting division of such grains into 2 equal, 2 unequal, and 4 cells (Fig. 1K). These abnormal grains have also been observed in anthers producing embryoids. As culture proceeded it was clear that there was no synehrony within each pollen sac in the initiation of embryogenesis, and embryoids in various stages of development were present. Pollen grains containing embryoids with larger numbers of cells were squashed to release their cells; all the cells appeared to be similar and no cells suggestive of degenerating generative-type cells were observed (Fig. 1 L, M). The exact stage at which the pollen grain wall ruptures to release the embryoids was variable. Normally free embryoids are observed at the stage shown in Fig. 1N; here the embryoid would still be described as globular
268
A. Rashid and H. E. Street
Fig. 1A--O. Stages in the development of pollen embryoids, A and B, Pollen grains from 4-week-old culture on N medium containing FeEDDttA; acetocarmine stained preparations. • 635. A. shows a slightly enlarged grain with dense contents which is considered to be an embryogenic grain. B. First division into two equal cells occurring in an embryogenic grain. C and D. Feulgen stained preparations of pollen grains. C. Pollen grain containing two equal cells and equally staining nuclei. • i008. D. Pollen grain with 3 nuclei, two diffusely staining (vegetative .*) nuclei and one densely staining (generative "?) nucleus. • i205. E--H. Pollen-
Haploid Embryoids from Anther Cultures
269
although one can now distinguish the root pole. The more mature embryoid shown in Fig. 1 0 shows the root pole and the two eotyledonary mounds. These observations indicate that very few of the population of microspores are induced to embark upon embryogenesis under the most favourable cultural regime established in this work. Our estimate is that in up to half of these grains, embryoid development may proceed to produce a plantlet. Iron has been implicated as a key compound for pollen to undergo embryogenesis (Nitsch, 1971) and this process was increased in the present work by replacing FeEDTA with FeEDDHA. The mode of action of F e E D D H A remains to be elucidated; it is a strong chelate of iron but has earlier been reported to simulate cytokinin-like effects (Maheshwari and Seth, 1966; Chopra and Rashid, 1969). Our evidence suggests that most frequently the first division leading to embryoid formation is an equal division but embryoids can also develop from grains in which the first division is unequal and gives rise to a vegetative and a generative cell. Our impression, also, is that large, starch-filled, swollen microspores are not embryogenie. To extend significantly studies on embryogenesis in Atropa belladonna pollen it is essential to further develop the precision with which anthers can be chosen at the correct developmental stage and to improve the cultural conditions so that embryogenesis is more prolific and more uniform in separate anthers. When such experimental material is available, electron microscopic study of pollen embryogenesis in this species will be rewarding. One of us (A. R.) is indebted to the University Grants Commission, New Delhi and the British Council, London for financial assistance. Technical assistance of Mr. K. Reader and Mr. C. Bargery is gratefully acknowledged.
embryoids in early stages of development; stained with acetocarmine. • 806. E, F, G. 4-, 5- and 6-celled emblToids respectively. H. 6-celled embryoid with presumptive generative cell (-->). I. Two embryoids developing from one pollen grain; note common exine. • 806. J. An abnormally swollen grain containing many starch grains. • 477. K. 4-celled aggregate probably derived from a grain of type shown in J. • 477. L and M. Squashed pollen embryoids at globular stage. Note all of cells similar and no evidence of any cells undergoing breakdown. • 1386. N and O. Embryoids with differentiated root pole only and with root pole and two cotyledons. • 248
270
A. Rashid and H. E. Street
References Chopra, R. N., Rashid, A. : Induction of shoot-buds in Anoectangium thomasonii Mitt. by a metal chelate, FeEDDHA. Z. Pflanzenphysiol. 61, i99-202 (1969) Iyer, R. D., Raina, S. K. : The early ontogeny of embryoids and callus from pollen and subsequent organogenesis in anther cultures oi Datura metel and Rice. Planta (Ber].) 104, 146-156 (1972) Linsmaer, E. M., Skoog, F. : Organic growth factor requirements of tobacco tissue cultures. Physiol. Plantarum (Cph.) 18, 100-127 (1965) Maheshwari, S. C., Seth, P. N. : Induction of flowering in Wol//ia microscopica by the iron salt of ethylenediamine-di-O-hydroxyphenylaeeticacid (FeEDDHA). Z. Pflanzenphysiol. 55, 89-91 (1966) Nakata, K., Tanaka, M. : Differentiation of embryoids from developing germ cells in anther cultures of tobacco. Jap. J. Genet. 43, 65-71 (1968) Narayanaswamy, S., Chandy, L. P.: In vitro production of haploid, diploid and triploid androgenic embryoids and plantlets in Datura metel L. Ann. Bot. 35, 535-542 (1971) Nishi, T., Mitsouka, S. : Occurrence of various ploidy plants from anther and ovary culture of rice plants. Jap. J. Genet. 44, 341-345 (1969) Nitseh, J. P.: Experimental androgenesis in Nicotiana. Phytomorpho]ogy 19, 389-404 (1969) Nitsch, J. P.: La production in vitro d'embryons haploides: l~6sultats et perspectives. In: Colloques Internationaux CNRS. Les Cultures de Tissus de Plantes. Strasbourg. 281-294 (1971) Shear, D. E. G., Fletcher, B. H., Street, H. E.: Studies on the growth of excised roots. VIII. The growth of excised tomato roots supplied with various inorganic sources of nitrogen. New Phytologist 58, 128-141 (1959) Street, H. E., McGregor, S. M.: The carbohydrate nutrition of excised tomato roots. II. The mechanism of sugar absorption by excised roots. Ann. Bot. 16, 185-205 (1952) Sunderland, N. : Anther culture: A progress report. Sci. Prog. (Oxf.) 59, 527-549 (1971) Sunderland, N., Wicks, F. M.: Embryoid formation in pollen grains of Hicotiana tabacum. J. exp. Bot. 22, 213-226 (1971) Thomas, E., Street, H. E.: Factors influencing morphogenesis in excised roots and suspension cultures of Atropa belladonna. Ann. Bot. 36, 239-247 (1972) White, P. R. : The cultivation of animal and plant cells. London: Thomas & Hudson 1954 Zenkteler, M. : In vitro production of haploid plants from pollen grains of Atropa belladonna L. Experientia (Basel) 27, 1087 (1971)
The Development of Haploid Embryoids from Anther Cultures of Atropa belladonna L. A. Rashid and H. E. Street Botanical Laboratories, University of Leicester, Leicester, LE1 7RH, U.K. l~eceived May 14 / June 5, 1973
Summary. Development of haploid embryoids from the microspores of Atropa belladonna occurs with relatively high frequency when anthers are excised from buds in which the petals are shorter than the sepals (at this stage microspores are predominantly uninucleate) and cultured on a medium containing iron as the ferric salt of ethylenediamine-di-O-hydroxyphenylacetic acid (FeEDDHA). Additions of combinations of kinetin, auxin and casamino-acids to the culture medium induce callusing in both haploid and diploid tissues, lead to tile origin of embryoids from somatic tissues of the anther and should be avoided. Simple techniques for the maintenance of haploid clones are described. Stages in early embryogenesis in the pollen grains have been observed and these indicate that embryogenesis is most frequently initiated by an equal division in the uninucleate spore. The frequency of grains showing embryoid formation is very low and it is estimated that plantlets are formed from up to 50% of these grains, Introduction Zenkteler (1971) has previously described the occurrence of embryogenesis in anthers of Atropa belladonna cultured on a medium containing an auxin and a cytokinin and reported the occurrence of haploid, diploid and triploid embryoids. Other workers have similarly reported the production of haploid and diploid plantlets by anther cultures of other Solanaceae cultured on media containing auxin and cytokinin or enriched with undefined natural supplements (Nishi and Mitsouka, 1969; Narayanaswamy and Chandy, 1971; Sunderland and Wicks, 1971; Iyer and l~aina, 1972). In a number of these eases callus development has been reported as arising either from the anther tissues or from the young embryoids. One objective of the present study was to define the nutritive conditions which are promotive of haploid embryoid development and suppressive of callus formation in Atropa belladonna. In Nieotiana spp. embryoid development has been reported from anthers with microspores at the tetrad stage (Nakata and Tanaka, 1968), at the uninucleate (Nitsch, 1969) and at the binucleate stages (Sunderland, 1971; Sunderland and Wicks, 1971). I n Nicotiana tabacum the uninucleate microspore divides to give a microspore containing a vegetative and a generative cell and when embryoids are formed from such
264
A. Rashid and H. E. Street
grains t h e y are derived from the vegetative cell. However, occasionally, the division of the u n i n u c l e a t e spore has been observed to be s y m m e t r i c a l a n d in these cases both cells have been considered to be i n v o l v e d i n e m b r y o i d f o r m a t i o n ( S u n d e r l a n d a n d Wicks, 1971). N a r a y a n a s w a m y a n d C h a n d y (1971) have described equal division in u n i n u c l e a t c micro. spores of cultured a n t h e r s of Datura metel, b u t i n c o n t r a s t I y e r a n d I~aina (1972), working with the same species, have described the first division as u n e q u a l a n d considered the embryoids to arise from the vegetative cell. A second objective of the present s t u d y was to follow the microspore divisions i n v o l v e d in e m b r y o i d f o r m a t i o n i n Atropa belladonna a n d to decide, if possible, the mierospore stage from which pollen-embryoids can be most readily obtained.
Materials and Methods A succession of flowering Atropa belladonna plants grown under long days was available at intervals of 4--5 weeks over the period June to October 1972. Flower buds with petals shorter than the sepals (Group I) and flower buds with petals equal or slightly longer than the sepals (Group II) were used. Pollen grains in anthers from Group II stained quickly and densely with acetocarmine and were rich in starch grains. Grains from Group I anthers stained only lightly with acetocarmine even after 30 min and did not contain starch. Both groups showed a single centrally-placed diffuse nucleus when stained with acetocarmine. Grains from Stage II buds stained in propionic orcein, usually showed the presence of a peripherally located, densely staining nucleus (generative nucleus); this was not seen in grains from Stage I buds. The grains from Stage ~ buds therefore appeared to be uninucleate, those from Stage II to contain a vegetative and generative nucleus.
For each flowering, buds were taken from during the second week of flowering until seed formation was occurring in the first formed flowers. However, embryoid development in culture was most abundant from anthers obtained early in each flowering period and it is the results of anther cultures from these harvests which are described. Sepals were removed from the flower buds and the buds then immersed in 90% ethanol for 2-3 min, washed with sterile water and dissected under aseptic conditions. The anthers, with most of the filament excised, were transferred to the surface of 10 ml medium solidified with 0.5% Difco bacto-agar and contained in a screw-capped McCartney bottle of 30 ml capacity. The anthers from any one bud (i.e. up to 5 anthers, though in some cases 4 or 3 due to individual anthers being dropped) were included in the same culture bottle evenly spaced over the medium surface. The basic medium contained 2% sucrose, 100 rag/1 meso-inositol, vitamins after White (1954) and the mineral salts either after Linsmaer and Skoog (1965) (LS medium) or after ~itsch (1969) (N medium). Iron was added as FeEDTA (10-4 iV[) (Shear et aI., 1959) or as FeEDDHA (the ferric salt of ethylenediaminedi-O-hydroxyphenylacetic acid). The media were variously supplemented with casamino acids 0.1 To w/v (CA), kinetin (kin) 10-5 M and 10-6 M, naphthaleneacetic acid (NAA) 10-6 M, and indolyl-aeetic acid (IAA) 10-5 and 10-6 M. pH of the medium was adjusted to pH 5.5 before autoclaving. Cultures were incubated at 25=[_0.5~ in continuous warm white fluorescent light (100 lumens per sq ft) for up to 12 weeks.
Haploid Embryoids from Anther Cultures
265
At intervals during culture, individual anthers were dissected and the pollen grains stained in acetocarmine and propionic orcein. The ploidy of embryoids and established plantlets was determined by treatment of root tips for 2 h in saturated paradichlorobenzene solution, fixation in acetic alcohol, hydrolysis for 12 rain at 56~ C in 1 N HC1, staining with Fculgen reagent and squashing in acetocarmine.
Results and Discussion
Cultural Conditions Conducive to the Development o/ Haploid Plantlets Very few anthers from Stage I I buds cultured on the basic LS medium gave embryoids ; none produced embryoids on LS medium supplemented with CA, Kin and NAA or with all three supplements but callus development occurred from the residual filament and from internal anther tissues. Anthers from Stage I I buds frequently dehisced soon after being placed in culture and showed no further development except occasional eMlus proliferation from the anther wall. Anthers from Stage I buds only rarely dehisced prematurely and produced embryoids with higher frequency on the LS-based media; in medium supplemented by CA, Kin and NAA, 50 % of the anthers gave rise to embryoids but the anthers also showed callus development from the residual filament and from the anther wall and when individual anthers were examined it was found that certain of the plantlets were arising from the connective. Anthers from Stage I I buds cultured in N medium failed to produce embryoids except in media supplemented with Kin and IAA; with 10-5 M Kin plus 10.5 M IAA 24% of the anthers produced embryoids but these anthers again showed callus development and evidence that at least some embryoids were arising from the connective. The use of IAA instead of NAA did not result in reduced callus formation (Thomas and Street, 1972). Anthers from Stage I buds again showed a higher incidence of embryoid formation in the N-based media and embryoids were obtained in the absence of eMlus formation on the basic medium and in this medium in which the FeEDTA was replaced by F e E D D H A at 10.5 M (4.35 mg/1). Plantlets from these anthers were all haploid. The use of F e E D D H A delayed the appearance of embryoids as compared with FeEDTA. A further experiment confined to Stage I buds and involving the N medium containing 1, 2 or 3 % sucrose and with F e E D D H A at various concentrations resulted in embryoid development in all the test media. There was no evidence of callus formation and all plantlets examined were haploid. In media with 2% sucrose and 5 mg/1 F e E D D I I A , 36% ; and with 3% sucrose and 15mg/1 FeEDDtIA, 40% of the anthers yielded embryoids.
266
A. Rashid and H. E. Street
The results indicated that Stage I buds gave rise to embryoids more readily than Stage I I buds; with Stage I I buds embryoid formation was associated with callus formation. Even with Stage I buds, there were always some cultures in which no anthers gave rise to embryoids and only rarely did all of the 4 or 5 anthers of each culture show embryoid development. In some cases embryoids only emerged from one half of the anther although microscopic examination then always showed the other half of the anther to contain embryoids at a less advanced stage of development. The number of emergent embryoids per anther was also variable. The majority of counts for cmbryoids per anther lay between 13-21 and the highest count was 30. With media inducing callus formation anthers usually burst within 4 weeks to reveal embryoids; with media not inducing callus formation anthers burst within 6-12 weeks of initiating the cultures. When a high number (3 or 4) of the anthers of the culture produced embryoids, these tended to show fasciation which disappeared when the plantlets were removed and cultured in separate culture bottles. This fasciation, therefore, m a y be a "crowding" effect possibly in response to a volatile factor accumulating within the confined space of the closed culture bottle. Irrespective of "crowding" some embryoids were polycotyledonary, some showed fusion of cotyledons and others showed branching of the embryoid axis. Some of the embryoids arising in media containing F e E D D t I A developed a corolla-like tubular structure which was pigmented like the corolla of the normal flowers. If embryoids were not detached and transferred to new culture bottles, they frequently showed the formation of daughter embryoids superficially on the embryoid axis. Similar embryo-like structures were occasionally observed during plantlet deve]opment (i.e. at a later stage when leaves were differentiated) on the plantlet axis or on young leaves. The main axis of developing plantlets always produced adventitious roots. Plantlets were cultured in medium lacking auxin and cytokinin since these growth regulators inhibit axis elongation and induced callusing of the embryoid tissues. Starting with individual haploid plantlets of anther origin, propagation and multiplication without loss of haploidy has been achieved by a combination of techniques: (1) By culture of 5 m m long stem apices in the N medium; (2) by culture of excised leaves which rooted in the N medium and gave rise to shoot buds at the cut base, particularly in the presence of 10 .7 M 6-y-y-dimethylallylaminopurine (DMAAP); (3) by culture of excised root tips in the root culture medium of Street and McGregor (1952) supplemented with 10 -~ rag/1 IAA and induction of bud formation from the callus formed at the basal end of the root by incorporating DMAAP into the medium.
Haploid Embryoids from Anther Cultures
267
Stages in Embryogenesis /tom Pollen Observations on embryogenesis were based upon studies of Stage I anthers cultured in N medium containing FeEDDHA. As indicated in Materials and Methods, microspores of Stage I anthers stained only lightly with acetoearmine. By day 28 of culture from 0 to 30 mierospores per anther were noted which were slightly enlarged and promptly and densely stained with acetocarmine (Fig. 1 A). Many other grains stained very slightly and appeared to be collapsed. The deeply staining grains appeared to be potentially embryogenic and on 7 occasions, in different anthers, such grains which had undergone an equal division were observed (Fig. 1B). Grains of this type stained with Feulgen have been observed with 2 equally staining nuclei (Fig. 1 C). On a single occasion a grain stained in Feulgen contained three nuclei, two staining diffusely and one densely (the latter possibly representing the nucleus of a generative cell) (Fig. 1D). Slightly later in cultures pollen grains showing early stages of embryogenesis were observed and usually all the cells appeared to be similar and could be readily visualised as deriving from grains in which the initial division had been equal. Fig. 1 E, F, G show 4, 5 and 6 cells of this kind within the pollen grain wall. In one case a grain was observed which could be interpreted as containing 5 cells of this type plus a generative cell ( / Fig. 1 H). On another occasion a grain was observed containing 6 "vegetative "-like cells and two distinct " generative "-like cells but this was not recorded photographically. In older cultures pairs of embryoids fused at their base were observed, and at the stage when young embryoids were present we have twice observed a grain bursting prematurely and apparently containing two pro-embryoids (Fig. 1 I). When anthers which failed to develop embryoids were examined, they always contained abnormal enlarged grains packed with starch grains (Fig. 1 J). These grains were often associated together in small aggregates. Occasionally grains were observed suggesting division of such grains into 2 equal, 2 unequal, and 4 cells (Fig. 1K). These abnormal grains have also been observed in anthers producing embryoids. As culture proceeded it was clear that there was no synehrony within each pollen sac in the initiation of embryogenesis, and embryoids in various stages of development were present. Pollen grains containing embryoids with larger numbers of cells were squashed to release their cells; all the cells appeared to be similar and no cells suggestive of degenerating generative-type cells were observed (Fig. 1 L, M). The exact stage at which the pollen grain wall ruptures to release the embryoids was variable. Normally free embryoids are observed at the stage shown in Fig. 1N; here the embryoid would still be described as globular
268
A. Rashid and H. E. Street
Fig. 1A--O. Stages in the development of pollen embryoids, A and B, Pollen grains from 4-week-old culture on N medium containing FeEDDttA; acetocarmine stained preparations. • 635. A. shows a slightly enlarged grain with dense contents which is considered to be an embryogenic grain. B. First division into two equal cells occurring in an embryogenic grain. C and D. Feulgen stained preparations of pollen grains. C. Pollen grain containing two equal cells and equally staining nuclei. • i008. D. Pollen grain with 3 nuclei, two diffusely staining (vegetative .*) nuclei and one densely staining (generative "?) nucleus. • i205. E--H. Pollen-
Haploid Embryoids from Anther Cultures
269
although one can now distinguish the root pole. The more mature embryoid shown in Fig. 1 0 shows the root pole and the two eotyledonary mounds. These observations indicate that very few of the population of microspores are induced to embark upon embryogenesis under the most favourable cultural regime established in this work. Our estimate is that in up to half of these grains, embryoid development may proceed to produce a plantlet. Iron has been implicated as a key compound for pollen to undergo embryogenesis (Nitsch, 1971) and this process was increased in the present work by replacing FeEDTA with FeEDDHA. The mode of action of F e E D D H A remains to be elucidated; it is a strong chelate of iron but has earlier been reported to simulate cytokinin-like effects (Maheshwari and Seth, 1966; Chopra and Rashid, 1969). Our evidence suggests that most frequently the first division leading to embryoid formation is an equal division but embryoids can also develop from grains in which the first division is unequal and gives rise to a vegetative and a generative cell. Our impression, also, is that large, starch-filled, swollen microspores are not embryogenie. To extend significantly studies on embryogenesis in Atropa belladonna pollen it is essential to further develop the precision with which anthers can be chosen at the correct developmental stage and to improve the cultural conditions so that embryogenesis is more prolific and more uniform in separate anthers. When such experimental material is available, electron microscopic study of pollen embryogenesis in this species will be rewarding. One of us (A. R.) is indebted to the University Grants Commission, New Delhi and the British Council, London for financial assistance. Technical assistance of Mr. K. Reader and Mr. C. Bargery is gratefully acknowledged.
embryoids in early stages of development; stained with acetocarmine. • 806. E, F, G. 4-, 5- and 6-celled emblToids respectively. H. 6-celled embryoid with presumptive generative cell (-->). I. Two embryoids developing from one pollen grain; note common exine. • 806. J. An abnormally swollen grain containing many starch grains. • 477. K. 4-celled aggregate probably derived from a grain of type shown in J. • 477. L and M. Squashed pollen embryoids at globular stage. Note all of cells similar and no evidence of any cells undergoing breakdown. • 1386. N and O. Embryoids with differentiated root pole only and with root pole and two cotyledons. • 248
270
A. Rashid and H. E. Street
References Chopra, R. N., Rashid, A. : Induction of shoot-buds in Anoectangium thomasonii Mitt. by a metal chelate, FeEDDHA. Z. Pflanzenphysiol. 61, i99-202 (1969) Iyer, R. D., Raina, S. K. : The early ontogeny of embryoids and callus from pollen and subsequent organogenesis in anther cultures oi Datura metel and Rice. Planta (Ber].) 104, 146-156 (1972) Linsmaer, E. M., Skoog, F. : Organic growth factor requirements of tobacco tissue cultures. Physiol. Plantarum (Cph.) 18, 100-127 (1965) Maheshwari, S. C., Seth, P. N. : Induction of flowering in Wol//ia microscopica by the iron salt of ethylenediamine-di-O-hydroxyphenylaeeticacid (FeEDDHA). Z. Pflanzenphysiol. 55, 89-91 (1966) Nakata, K., Tanaka, M. : Differentiation of embryoids from developing germ cells in anther cultures of tobacco. Jap. J. Genet. 43, 65-71 (1968) Narayanaswamy, S., Chandy, L. P.: In vitro production of haploid, diploid and triploid androgenic embryoids and plantlets in Datura metel L. Ann. Bot. 35, 535-542 (1971) Nishi, T., Mitsouka, S. : Occurrence of various ploidy plants from anther and ovary culture of rice plants. Jap. J. Genet. 44, 341-345 (1969) Nitseh, J. P.: Experimental androgenesis in Nicotiana. Phytomorpho]ogy 19, 389-404 (1969) Nitsch, J. P.: La production in vitro d'embryons haploides: l~6sultats et perspectives. In: Colloques Internationaux CNRS. Les Cultures de Tissus de Plantes. Strasbourg. 281-294 (1971) Shear, D. E. G., Fletcher, B. H., Street, H. E.: Studies on the growth of excised roots. VIII. The growth of excised tomato roots supplied with various inorganic sources of nitrogen. New Phytologist 58, 128-141 (1959) Street, H. E., McGregor, S. M.: The carbohydrate nutrition of excised tomato roots. II. The mechanism of sugar absorption by excised roots. Ann. Bot. 16, 185-205 (1952) Sunderland, N. : Anther culture: A progress report. Sci. Prog. (Oxf.) 59, 527-549 (1971) Sunderland, N., Wicks, F. M.: Embryoid formation in pollen grains of Hicotiana tabacum. J. exp. Bot. 22, 213-226 (1971) Thomas, E., Street, H. E.: Factors influencing morphogenesis in excised roots and suspension cultures of Atropa belladonna. Ann. Bot. 36, 239-247 (1972) White, P. R. : The cultivation of animal and plant cells. London: Thomas & Hudson 1954 Zenkteler, M. : In vitro production of haploid plants from pollen grains of Atropa belladonna L. Experientia (Basel) 27, 1087 (1971)
