Abstracts

Two point deterministic model for acquisition of in vitro pollen grain androgenetic capacity based on wheat studies*

Magali Ferrari Grando¹ and Maria Irene B. de Moraes-Fernandes ²

*Part of a thesis presented by M.F.G. to the Graduate Course of Genetics of the

Universidade Federal do Rio Grande do Sul (UFRGS) in partial fulfillment

for the requirements of the Master’s degree.

¹Universidade de Passo Fundo, Caixa Postal 566, 99001-970 Passo Fundo, RS, Brasil.

²Centro Nacional de Pesquisa em Trigo, EMBRAPA, Caixa Postal 569,

99001-970 Passo Fundo, RS, Brasil. Send correspondence to M.I.B.M.-F.

ABSTRACT

This article discusses, from the standpoint of cellular biology, the deterministic and indeterministic androgenesis theories. The role of the vacuole and of various types of stresses on deviation of the microspore from normal development and the point where androgenetic competence is acquired are examined. Based on extensive literature review and data on wheat studies from our laboratory, a model for androgenetic capacity of pollen grain is proposed. A two point deterministic model for in vitro androgenesis is our proposal for acquisition of androgenetic potential of the pollen grain: the first switch point would be early meiosis and the second switch point the uninucleate pollen stage, because the elimination of cytoplasmatic sporophytic determinants takes place at those two strategic moments. Any abnormality in this process allowing the maintenance of sporophytic informational molecules results in the absence of establishment of a gametophytic program, allowing the reactivation of the embryogenic process.

INTRODUCTION

When anthers are cultured in appropriate culture media, some pollen grains produce, by successive mitosis, a haploid embryo which can be induced to regenerate a green plantlet that has only the male genome. By colchicine treatment, or sometimes spontaneously, the genome is duplicated, returning to diploidy, achieving complete homozygosity with restoration of fertility.

Besides the scientific interest in understanding the basic mechanism of pollen grain totipotency, the haploid technique is also valuable in crop breeding programs, since the required homozygosity for agronomic selection can be achieved in one generation. The conventional breeding procedure, on the other hand, demands several years of labor-intensive field observations on space-consuming segregating populations to achieve genetic uniformity (reviews in Moraes-Fernandes, 1987 and Picard et al., 1994). Genetic, physiological and environmental factors are involved in the change of the program from normal male gametophyte development towards pollen grain embryogenesis. Nitsch (1974a) proposed that the signal for deviation from the normal pollen development pathway to the embryogenic route was cellular division disturbance at the first mitosis, resulting in two cells with equal size nuclei. An alternative view, given by Heberle-Bors (1985), suggests that it is a special group of abnormal pollen grains, the p-pollen grains, which follow the embryogenic route of development. He proposed that the cytoplasm of p-pollen maintained the sporophytic determinant informational molecules, while the normal pollen eliminated such molecules, as it was able to follow the gametophytic route.

We have experienced difficulties in interpreting data from our laboratory with either of the above theories (Grando, 1990). The two point deterministic model, which better explains our experimental results, is proposed here. This model proposes two points where the androgenetic capacity of pollen grain may be acquired.

In situ development of pollen grain

A pollen mother cell (PMC) creates, after meiotic division, four haploid cells. These are microspores, which develop into the male gametophyte (pollen grain). The newly formed microspore has a centrally located nucleus and a large number of minute vacuoles in the cytoplasm. These vacuoles coalesce during pollen development to form a single large cell which displaces the nucleus from the central location to the opposite side of the pore (Figure 1a,b). The location of the vacuole also makes spindle formation for the first microspore mitosis off-center (Figure 1c), and results in two cells of unequal sizes: a large vegetative cell with a diffuse nucleus and a small generative cell with a condensed nucleus. The vegetative cell grows and displaces the vacuole that begin to disappear. The cells begin to adhere to the cell membrane and this structure becomes the pollen grain (Figure 1d).

Figure 1 - Normal development of a wheat pollen grain: a, uninucleated microspore with a central nucleus; b, vacuole microspore with peripheral nucleus; c, asymmetrical division of haploid nucleus; d, binucleate pollen grain with a big vegetative cell presenting diffuse nucleus and generative cell with a condensed nucleus; e, trinucleate mature pollen presenting one vegetative nucleus and two long spermatic nuclei with cytoplasm rich in plastids with starch grains.

Differences in developmental pathways of these two cells (nutritional role for the vegetative and fertilization for the generative) may result from the different amounts of cytoplasm. The vacuole is important for microspore differentiation because when it is absent the first pollen grain mitosis would be symmetrical, giving rise to two cells of equal size (Figure 2) (Sax and Edmonds, 1933; Sax, 1935; Brumfield, 1941; La Cour, 1949).

Figure 2 -In situ pollen grain with identical nuclei arising by: a, absence of vacuole indicated by central division; the division can occur at distinct angles, resulting in two identical nuclei found at several places, and b, wrong cell spindle axis in presence of vacuole. The right division angle is showed in Figure 1c.

When the gametophytic program is established, the nucleus of the vegetative cell maintains a metabolically active state. This is evidenced by its diffuse aspect (free for transcription), intensive RNA/protein synthesis, abundance of ribosomes and other organelles in the cytoplasm, and the intense stainability of the cell with acetocarmine (Sangwan and Camefort, 1982a; Raghavan, 1987). The nucleus of the generative cell, on the other hand, is primarily inactive metabolically. There is little cytoplasm and only a few organelles associated with it. Its chromatin is highly condensed with the bulk of the DNA associated with histones. Very little, if any, RNA protein synthesis can be detected (Sauter, 1973; Bederaska, 1981, 1984; Sangwan and Sangwan-Norreel, 1987a).

At the beginning of pollen tube germination there is a general decrease of genic activity and long life RNA starts to disappear. The energy supply for germination is derived from the abundant starch grains in the plastids of the vegetative cell (Goss, 1968; Nitsch, 1972; Sunderland, 1974; Sangwan and Camefort, 1982a).

Two elongated sperm nuclei are formed by division of the generative nucleus. One will fertilize the egg cell to form the zygote and the other fertilizes the central polar nucleus to form the triploid endosperm.

Androgenesis as a deviation of the gametophytic route

Two steps are necessary for androgenesis: the change of the gametophytic program to induce the embryogenetic route and the development of the embryoid (Sangwan, 1983). Androgenetic embryogenesis has only been observed experimentally in vitro when microspores are cultivated at a specific developmental stage (Bhojwani et al., 1973).

The androgenetic pathway in Hyoscyamus niger occurs in uninucleate pollen grains which synthesize mRNA during the first hour of culture. Raghavan (1979a,b) suggested that this synthesis is induced by the stress of anther excision and not by culture media. Sunderland (1980), on the other hand, proposed that both environmental and nutritional stresses promote slackness in the gametophytic control system.

Androgenesis as a mistake at first pollen grain mitosis

The frequency of in vitro embryoid production in Datura after 10 days of anther culture was found to be the same as the frequency of binucleate pollen grains having two equal-size nuclei. This led to the conclusion that the symmetrical first pollen mitosis is the signal recognized by the cell to reactivate the sporophytic program, even without fertilization (Nitsch, 1974a). Cold treatment was found to be capable of increasing the frequency of embryogenic anthers as well as the frequency of pollen grains with two equal-size nuclei (Nitsch and Norreel, 1972, 1973).

Temperature stresses were found to result in the absence of or with incomplete vacuole formation in Tradescantia. This led to the absence of nuclear migration from the central location to the periphery. Thus, two equal-size cells were produced after the first microspore mitosis, due to polarity disturbance (Sax, 1935, 1937; La Cour, 1949). Several other stresses applied as pre-treatments produced similar results: centrifugation (Sangwan-Norreel, 1977), irradiation (Stolarz, 1974), cut off spike tip (Picard, 1973), and water submersion (Wilson et al., 1978).

However, cold treatment increases androgenetic embryoid formation but does not increase the frequency of pollen grains with equal-size nuclei in Nicotiana tabacum (Duncan and Heberle-Bors, 1976; Sunderland and Roberts, 1979; Bajaj, 1983). These controversial results may be explained by taking into account that the production of microspores with equal-size nuclei may proceed by two different routes: no vacuole formation (Figure 2a) or modification of the cell division axis (Figure 2b). We postulate that only the first route leads to androgenesis, because the vacuole plays an important role in the degradation of cytoplasmatic contents that are autophaged before the gametophytic program is established (Sangwan and Sangwan-Norreel, 1987b). We may therefore presume that the latter does not give rise to the androgenetic route because of the maintenance of sporophytic determinants.

It was found, from our studies in wheat androgenesis, that there was no significant correlation between the frequencies of pollen grains with identical nuclei obtained from in situ anthers and androgenetic structures from cultivated ones (Grando, 1990). For example, the frequency of identical nuclei pollen of the 407DH x 355DH genotype was 2.32% and the embryo frequency was 1.0%, while in the 319DH genotype the frequencies were 1.35 and 2.85%, respectively. The occurrence of two different origins of pollen grains with identical nuclei partly explains frequency discrepancies among different studies.

In a study of in vitro pollen development (Grando, 1990), from the first to the 14th day after anther culture, it was observed that androgenetic structures can arise from vegetative, generative or both types of nuclei, and, at lower frequencies, from pollen grains with two identical nuclei (Figure 3).

Figure 3 - Segmentation pattern of in vitro wheat pollen observed from the 1st to 14th day of anther culture. The figure shows that a uninucleated microspore may degenerate before division, (a) or may give rise, by mitosis, to normal binucleated grain presenting vegetative and generative nuclei (b). Identical nuclei are formed in low frequency (c). The first mitotic division occurs at the 4th day of culture. The normal binucleated pollen may follow the normal in situ developmental pattern forming starch and then, degenerate (d). The second mitotic division takes place at the 6-8th and at the 10th day of culture. In the androgenetic pollen, the vegetative, generative or both nuclei are able to divide giving rise to an embryo (e, f, g). In pollen with identical nuclei, both cells contribute to androgenesis (h). Pollen degeneration can occur in any step of this process. At the 14th day, multicellular structures can be seen (i).

Pollen dimorphism and the deterministic model for androgenesis

Two classes of pollen grains have been found in Nicotiana tabacum anther cultures: normal and atypical pollen (Sunderland and Wicks, 1971). The normal grains stained intensely with acetocarmine, followed the gametophytic development pathway, and degenerated when cultured in vitro. The atypical grains were smaller, stained weakly, and followed the embryogenetic pathway. This second class of grains was named p-pollen (pre-mitotic pollen) by Heberle-Bors and Reinert (1979).

Such pollen dimorphism has also been observed in in situ mature anthers, in which the normal pollen grains are larger, full of starch, and the atypical ones smaller, weakly stained with acetocarmine and occur at a lower frequency (Sunderland, 1974). The intense cytoplasmic synthesis that follows the first meiotic division of the normal grains does not occur in the atypical grains (Sunderland, 1978). In addition, the atypical grains have a thinner exine, weaker cytoplasm, fewer ribosomes, condensed mitocondria and repressed gametophytic differentiation.

In situ pollen dimorphism in wheat has been observed in our laboratory (Grando, 1990) as well as by De Buyser and Picard (1975), Zhow (1980), He and Ouyang (1983) and Herberle-Bors and Odenbach (1985). The p-pollen observed in our laboratory in wheat mature anthers had either a single nucleus, two identical nuclei or vegetative and generative ones (Figures 4 and 5).

Figure 4 - Pollen dimorphism in mature wheat anthers in situ: a, normal trinucleate pollen grain, stained by acetocarmine and with stored starch; b, atypical pollen (p-pollen), smaller, weakly stained by acetocarmine, without starch, vegetative and generative.

Figure 5 - P-pollen with two identical nuclei.

In barley there is a correlation between the frequency of abnormal pollen in situ and the androgenetic structures obtained in vitro (Dale, 1975). This correlation led to speculation that the ability to form embryoids in vitro is pre-determinated at some early stage of microspore ontogeny with the formation of atypical p-pollens (Horner and Street, 1978; Horner and Mott, 1979).

When Nicotiana tabacum anther donnor plants are cultivated under short day and low temperature conditions, the frequency of p-pollen in situ and of haploid plants from isolated pollen culture in vitro increases. A strong correlation was found between the in situ p-pollen and in vitro pro-embryoid frequencies in this species by Heberle-Bors and Reinert (1979, 1980, 1981).

It was concluded by Heberle-Bors (1982a) that tobacco p-pollen is a particular form of male sterility. In accordance with this author, it does not deposit starch and does not express the gametophytic program which would lead to fertilization. Instead, it degenerates in situ as the anther becomes mature. This form of sterility is strongly correlated with the deviation of the sexual balance towards femaleness, when the tobacco plant (a long-day plant) has been grown under short day and low temperature regimes. P-pollen is induced just before or during PMC meiosis. Day length and temperature during meiosis determine the sexual balance and hence p-pollen frequency (Heberle-Bors, 1982a).

Androgenesis can be explained by a failure in the rigid genetic control for the determination of pollen gametophytic behavior. This is due to changes in environmental factors during the early ontogeny of pollen grain development, leading to female instead of male behavior (Vasil, 1980). In the two point deterministic model proposed by us early meiosis is the first switching point in which the pollen grain can acquire competence to form a haploid embryo.

The growing conditions for deviation of sexual balance can vary between species (Heberle-Bors, 1982a). In our studies (Grando, 1990), a wheat double-haploid line (PF 853003) cultivated in the greenhouse, under summer stress growing conditions, produced up to 50% sterile pollen, rudimentary anthers and multiple ovaries, indicating a change of sex balance favoring femaleness.

Application of feminization agents at the pre-meiosis stage increases the frequency of p-pollen and androgenetic embryos in tobacco (Heberle-Bors, 1983). The same was observed in Triticum aestivum anther culture, with the use of the gametocide potassium ferridazon (Picard et al., 1987). Genetic and physiological factors are involved in embryogenic pollen formation (Picard et al., 1990).

The requirement for shifting from the sporophytic to the gametophytic pathway

To answer questions such as how p-pollen is formed and why is it potentially embryogenic, it is necessary to understand the cytoplasmatic modifications which occur during meiosis of pollen-mother cells. A complete elimination of sporophytic determinants of the long life, ribosome-associated m-RNA is needed in the PMC in order to establish the new gametophytic genetic program and guarantee normal development of the male gametophyte (Dickinson and Heslop-Harrison, 1977). A dramatic decrease in cytoplasmic RNA contents in angiosperms occurs during the zygotene stage of meiotic prophase I, as well as a reduction in the ribosome population between leptotene and diakinesis (MacKenzie et al., 1967; Dickinson and Heslop-Harrison, 1970). Such material reduction from the PMC cytoplasm is probably associated with the elimination of sporophytic determinants.

When the process of elimination of the sporophytic determinants, the polysomes, is improperly carried out during meiosis, the normal gametophytic development of the pollen grain will not proceed. This blockage of normal gametophytic development is evident in the p-pollen by the absence of intensive RNA and protein synthesis that occurs in normal pollen grains after first mitosis (Heberle-Bors, 1982b).

In Pteriodophyta, development may sometimes deviate from the normal haploid/diploid cycles. Bell (1970, 1979) proposed that such deviation is due to the inheritance of sporophytic development informational particles by the haploid spores from spore mother cells, determining the development into a haploid plant. The same error may take place in the formation of p-pollen as suggested by Heberle-Bors (1982b).

The above viewpoint is named the deterministic model (Heberle-Bors, 1982b). It proposes that the embryogenic capacity of pollen grains is pre-determined early during PMC meiosis. In vitro culture conditions merely aid the expression of this capacity and do not alter the frequency of such totipotent pollen grains in a given anther.

The controversy about the shifting point

It is universally accepted that the androgenetic embryo results from a deviation of normal gametophytic development, but the moment at which this deviation takes place is controversial. The indeterministic theory assumes that the shift occurs after the detachment of the floral bud from the donor plant. The environmental conditions during the pre-culture and culture treatments are responsible for such a shift. Theoretically, every pollen grain cultured is potentially capable of becoming androgenic, if cultivated before switching off of gametophytic development-determining genes (Vasil, 1973).

The deterministic theory states that the environment, during PMC meiosis, affects male gamete differentiation. If meiosis is normal, the sporophytic determinants are eliminated in preparation for establishment of the gametophytic program. If the sporophytic determinants are maintained because of meiotic abnormalities, the cell maintains its embryogenetic capacity. Heberle-Bors (1985) states that only at meiosis can the pollen grain become androgenetic; after meiosis, it is only the viability of this pre-determinated pollen that can be affected. We propose that after meiosis the pollen grain has one more chance to become embryogenic: during uninucleated pollen stage.

Evidence for and against the deterministic theory

The evidence for the deterministic model includes:

a) the similarity between p-pollen observed in mature anthers in situ and the kind of pollen that produces embryos in vitro (Sunderland, 1974; Dale, 1975; Horner and Street, 1978; Heberle-Bors and Reinert, 1980). The additional nuclear division observed in p-pollen in situ resembles multicellular structures forming embryos in vitro (Figure 6) (Sunderland, 1974, 1978; De Buyser and Picard, 1975; Horner and Mott, 1979; Heberle-Bors, 1982a, Heberle-Bors and Odenbach, 1985; Grando, 1990).

Figure 6 - Wheat p-pollen in vivo showing several nuclei similar to multicellular structures in vitro.

b) the change in environmental conditions during meiosis modifies the frequencies of in situ p-pollen in the anthers of tobacco plants, along with that of embryoids in vitro (Heberle-Bors, 1982a).

c) The application of chemical agents during meiosis in tobacco (Heberle-Bors, 1983) and wheat (Picard et al., 1987) increases the frequency of the atypical pollens in situ, along with the frequency of embryoids obtained in vitro.

d) pollen culture ab initio (Heberle-Bors and Reinert, 1980; Rashid and Reinert, 1981a,b), extracting p-pollen from mature anthers with a centrifuge density gradient and its subsequent cultivation in vitro, provided the final proof that, at least in N. tabacum, p-pollen is really embryogenic.

The evidence against the deterministic model is based on the fact that in many cases a positive correlation between the p-pollen frequency in situ and androgenetic structures in vitro is not found. This lack of correlation may be partially due to the quick degeneration of p-pollen during anther maturation. As a result, a fraction of the cultured p-pollens becomes unresponsive (dead). The growth-inhibiting substances present in the culture medium may also throw off the correlation (Heberle-Bors, 1984).

In some cases the frequency of pollen that becomes embryogenic is higher than the frequency of atypical pollen observed in situ (Roberts-Ochlschlager and Dunwell, 1991). Dunwell (1978) used this fact to argue against the idea that p-pollen is the only source of embryoids.

In order to clarify the relation between wheat in situ p-pollen and in vitro embryoid frequencies, we cultivated anthers of three double-haploid lines from our program. No significant correlation was found. One anther of the line PF 853003 produced a very high frequency of p-pollen, but did not produce embryos. Subsequent studies, however, showed that this genotype presented 2.8% p-pollen in situ and 1.5% pro-embryoids on the 14th day of culture, but neither reached the final embryogenetic stage. It was concluded that this genotype does not have good androgenetic capacity because pro-embryoid development in culture media is prevented (Grando, 1990).

Low temperatures may induce an increase in p-pollen frequencies in tobacco, but sometimes it also stimulates production of inhibitory substances in the anther tissue of genotypes blocking in vitro embryogenetic development (Heberle-Bors, 1984). It is clear that p-pollen is a necessary but insufficient requirement for androgenesis, since the expression of its embryogenic potential depends on other factors. Because of that, we think that the use of p-pollen as a marker of androgenetic capacity is not a reliable measure. The anther donor plant genotype controls the frequency of p-pollen, the intensity of its change by environmental stresses, as well as the production of inhibitory substances by anther tissue (Heberle-Bors, 1984, 1985).

Two point deterministic model for acquisition of in vitro pollen grain androgenetic capacity

During the mitotic prophase of pollen grains in Datura inoxia, cytoplasmatic structures, mainly ribosomes, are autophaged by cytoplasmatic vacuoles that have lysozymic action (Sangwan, 1986). Therefore, there is evidence that the elimination of molecules carrying sporophytic information occurs not only at meiosis, but also seems to be reactivated at the first haploid mitosis because of the lytic activity in the big central vacuole in uninucleate pollen. The number of ribosomes at the early uninucleate stage is reduced when the vacuole is formed in Datura metel (Sangwan and Camefort, 1982b). So, it can be proposed that equal-size pollen nuclei resulting from a lack of vacuole formation are also competent for androgenesis because the final stage of elimination of sporophytic determinants does not take place.

The two point deterministic model for in vitro androgenesis is our proposal for the acquisition of androgenetic potential of pollen grains: the first switching point would be the early meiosis and the second switching point the uninucleate microspore stage, because the elimination of cytoplasmatic sporophytic determinants takes place at those two strategic points. Any abnormality in this process allowing the maintenance of sporophytic informational molecules results in the absence of establishment of a gametophytic program, allowing the reactivation of the embryogenic process.

Figure 7 explains in a schematic way the model proposed here, which puts together suggestions made mainly by Heberle-Bors (1982b, 1985) and Sangwan and Sangwan-Norreel (1987b), complemented by observations made in our work (Grando, 1990).

Figure 7 - Synthetic model for androgenesis. Stress operating at two "changing points" can modify the normal development of pollen grain, that became competent for embryogenesis. This potential would be dependent on sporophytic information kept in the cytoplasm, which normally are eliminated at meiosis and at the uninucleated vacuolate stage of pollen grain.

Evidence from stress effects on androgenesis favoring the two point deterministic model

Several types of stress, such as temperature, light, nutrition, water, chemicals, osmotic shock, cutting and centrifugation, when applied to each of these two switching points, increase the frequency of embryoids.

Low temperature is extensively applied to the spike or floral bud before or just after inoculation to increase embryoid production. In Datura innoxia this stress applied to uninucleated pollen increases the frequency of identical nuclei pollen (Nitsch and Norreel, 1973; Nitsch, 1974a,b; Sangwan-Norrell, 1977). Sangwan and Camefort (1984) observed in D. innoxia that cold treatment induces strong vacuolar modification. Gametophytic differentiation can be repressed by low temperatures as suggested by Rashid et al. (1981) and Bajaj (1983).

Anther excision and the environmental stress caused by anther inoculation onto culture medium can also promote vacuole modification. Therefore, the idea that embryogenesis induction takes place in the culture media is not totally wrong, since pollen has been cultivated in the uninucleated stage.

However, in N. tabacum, cold treatment at the uninucleate stage does not induce identical nuclei formation (Duncan and Heberle-Bors, 1976; Sunderland and Roberts, 1979), suggesting that this species is less susceptible to stress at this stage. Sunderland and Roberts (1979) reported that cold treatment is more effective at the binucleate stage, after vacuole disappearance, which is explained by increasing embryogenic pollen viability, preventing cytoplasmatic degeneration of p-polen (Rashid et al., 1981).

Schimid and Keller, 1986, reported by Picard et al. (1990) obtained good haploid embryo induction when wheat plants were sprayed with a gametocide before meiosis. They also found a positive effect when the gametocide was incorporated into the culture medium in which anthers were inoculated.

Another proposal is that species may differ in the main switching point utilized to produce embryogenic pollen. In the N. tabacum pathway, this main switching point is early meiosis, because stresses at this point affect p-pollen frequencies, since cold treatment does not prevent vacuole formation at the uninucleate stage, and the binucleate stage is optimal for inoculation.

In the D. innoxia pathway, the main switching point is the uninucleate stage, because cold treatment disturbs vacuole formation, leading to identical nuclei pollen. The susceptibility to stress at this point could explain why route I (first pollen grain symmetric division; Sunderland, 1974) is the main androgenetic route in this species and why there is better embryoid production after inoculation at the uninucleate stage.

ACKNOWLEDGMENTS

We thank Gelsi Galon, Rosana Gazola and Renati Fenner for help with some steps of the experimental work, to J.E.F. Figueiredo for helpful discussion, Dr. Ching Yeh Hu and Alice Kaliz de Oliveira for critical reading of the manuscript, and CNPq for a research grant for the first author.

RESUMO

Este artigo discute, do ponto de vista da biologia celular, as teorias determinística e indeterminística sobre a androgênese. São examinados o papel do vacúolo e dos diferentes estresses no desvio do grão de pólen de seu desenvolvimento normal e o ponto onde a competência androgenética é adquirida. Baseado numa extensa revisão da literatura e dados de estudos realizados em trigo pelo nosso laboratório, é proposto um modelo para a capacidade androgenética do grão de pólen. O modelo determinístico de dois pontos para androgênese in vitro é nossa proposta para explicar a aquisição da capacidade androgenética do grão de pólen: o primeiro ponto de mudança seria no início da meiose e o segundo ponto de mudança no estádio de pólen uninucleado, porque a eliminação dos determinantes esporofíticos citoplasmáticos ocorre nestes dois pontos estratégicos. Qualquer anormalidade neste processo permite a manutenção das moléculas de informação esporofítica, resultando na ausência do estabelecimento do programa gametofítico e permitindo a reativação do processo embrionário.

(Received July 13, 1995)

Publication Dates

History