Androgenesis revisited.
Abstract: Androgenesis can be defined as the set of biological processes leading to an individual genetically coming exclusively from a male nucleus. Androgenesis was traditionally considered as the spontaneous, in vivo development of a male-derived haploid embryo from a fertilized egg where the female nucleus is eliminated. However, at present it is also possible to generate androgenic haploids/doubled haploids through in vitro microspore embryogenesis and by in vitro meiocyte-derived callogenesis. These three androgenic alternatives clearly differ in the inducible stage, but lead to the same final haploid or doubled haploid product. Whereas microspore embryogenesis is widely studied and applied, the other two routes are much less known. In this paper, the evidence accounting for the existence of these three alternative pathways is revised, as well as the mechanisms potentially involved in their induction. Their differences and similarities are discussed from a biological perspective, leading to the notion that they might represent an ancient survival mechanism triggered by similar factors.

La androgenesis se define como el conjunto de vias biologicas que dan lugar a un individuo cuyo fondo genetico proviene exclusivamente de un nucleo de origen masculino. Tradicionalmente, el concepto de androgenesis estaba restringido al desarrollo espontaneo, in vivo, de un embrion haploide o doble haploide a partir de una celula huevo fecundada en la que el material gendtico de origen femenino era inactivado o eliminado. Sin embargo, hoy en dia sabemos que existen otras vias para conseguir haploides o doble haploides androgenicos, mediante embriogenesis de microsporas y mediante la formacion de callos derivados de meiocitos. Estas tres alternativas androgdnicas difieren claramente en la etapa en la que es posible la induccion, pero dan lugar al mismo producto haploide o doble haploide final. Mientras que la embriogenesis de microsporas es un fenomeno ampliamente estudiado y de clara aplicacion practica, las otras dos rutas son mucho menos conocidas. En este trabajo se revisan las evidencias existentes al respecto de estas tres alternativas androgenicas, asi como los mecanismos potencialmente implicados en su induccion. Tambien se discuten sus diferencias y semejanzas, desde un punto de vista biologico, para llegar a la hipotesis de que estas rutas podrian representar un mecanismo atfivico de supervivencia activado por factores similares.

Keywords Microspore Embryogenesis * Meiocyte-Derived Callogenesis * Male-Specific Parthenogenesis * Plant Reproduction * Haploid * Doubled Haploid
Article Type: Report
Subject: Embryonic development (Research)
Parthenogenesis (Research)
Plants (Reproduction)
Plants (Methods)
Author: Segui-Simarro, Jose M.
Pub Date: 09/01/2010
Publication: Name: The Botanical Review Publisher: New York Botanical Garden Audience: Academic Format: Magazine/Journal Subject: Biological sciences Copyright: COPYRIGHT 2010 New York Botanical Garden ISSN: 0006-8101
Issue: Date: Sept, 2010 Source Volume: 76 Source Issue: 3
Topic: Event Code: 310 Science & research
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 237601496
Full Text: Introduction

During evolution, plant cells have retained the capacity to express virtually any part of their coding genome. This remarkable feature, called totipotency, allows for a differentiated cell to dedifferentiate and adopt a proliferative growth pattern (callus), or to deviate towards a developmental program different from the original. These programs include the regeneration of a new individual either through the successive regeneration of all of its vegetative organs (organogenesis) or directly by the entry into a new embryogenic pathway (embryogenesis). Using adequate experimental conditions, it is possible to induce organogenesis and/or embryogenesis from cells of a variety of plant tissues, including leaves, cotyledons, root tips, hypocotils, anthers, ovaries, etc. (Razdan, 2003; Vasil & Thorpe, 1994). From an applied point of view, one of the most interesting cell types to regenerate individuals is haploid (reduced) gamete precursors, due to the possibility to regenerate haploid or doubled haploid (DH) individuals.

Haploids are interesting for fundamental research, but their main utility is to produce DH lines, which are extremely useful tools for basic and applied research, including breeding programs (Chupeau et al., 1998; Dunwell, 2010; Forster et al., 2007; Touraev et al., 2001). From the standpoint of plant breeding, the DH alternative reduces the typical 7-8 inbreeding generations necessary to stabilize a hybrid genotype to only one. It is therefore much faster and cheaper, and obviously this is the main advantage of DH technology in plant breeding. Still within the context of plant breeding, DHs are also essential for genetic mapping of complex characters such as production or quality, the most interesting from an agronomical point of view, but also the most difficult to be addressed by other approaches. DHs are also a powerful tool in transgenesis, to avoid hemizygotes and save time and resources in the production of plants transformed with the transgene in both homologous chromosomes. Moreover, from a basic scientific point of view, these lines are also very useful for basic studies of linkage and estimation of recombination fractions. Although these studies can also be made conventionally (backcrosses or F2), DH lines have the advantage of being auto-perpetuable, meaning that they can be perpetuated simply by selling. They are also an extremely useful tool for genetic selection and screening of recessive mutants, because the phenotype of the resulting plants is not affected by the effects of dominance, and the characters determined by recessive genes can be easily identified. Another advantage is the ability to serve as a model system for studying embryo development in vitro, without the interference of maternal tissue, due to the large number of similarities of microspore-derived embryogenesis compared with zygotic embryogenesis (Segui-Simarro & Nuez, 2008a; Supena et al., 2008).

It is possible to obtain haploid/DH individuals through different developmental pathways, involving both the female and the male gametophytes (Dunwell, 2010). From the unpollinated female gametophyte, it is possible to obtain haploids through a pathway known as gynogenesis (Bohanec, 2009). Gynogenesis is used in species such as sugar-beet and onion, where other DH-inducing techniques have proven unsuccessful. Through this route, immature ovules are cultured up to maturation of embryo sac, where a haploid embryo is developed. For few species, this embryo is thought to be originated from antipodal or synergid cells, but in the vast majority of cases, the gynogenic embryo is derived from the egg cell (reviewed in Bohanec, 2009). Gynogenic embryos are predominantly haploid, which implies that in order to obtain the desired doubled haploid, additional treatments for chromosome doubling must be considered. A pathway similar to that induced by culturing immature ovules can be induced by wide, interspecific or even intergeneric crosses. By crossing two sexually incompatible species, it is possible to induce the development of a haploid plant coming exclusively from the female gamete, therefore excluding any genetic background from the male parental. This is the case for the "Hordeum bulbosum" method, initially discovered in 1970 (Kasha & Kao, 1970), and now widely used for DH production in barley breeding programs for the delivery of commercial varieties (Devaux, 2003; Hayes et al., 2003; Wedzony et al., 2009). In this method, a hybrid embryo is formed by the fusion of a H. bulbosum male gamete with a H. vulgare egg cell. However, post-zigotic incompatibility barriers lead to the progressive elimination of the chromosomes from the H. bulbosum male parental (Devaux, 2003), so that the resulting (haploid) embryo has a H. vulgate background. Incompatibility at the level of the endosperm makes also mandatory the isolation and in vitro rescue of the haploid embryo. As for gynogenesis, haploid embryos/plants must be exposed to chromosome doubling agents in order to become DH. Similarly, interegeneric hybridization has also been described as capable of inducing haploid plant development. For example, through intergeneric crosses between wheat and maize (Zenkteler & Nitzsche, 1984). The use of maize pollen as a mentor to induce the development of haploid embryos in other genera has been widely exploited in the last 20 years, mostly for cereals (reviewed in Wedzony et al., 2009). Intergeneric crosses have also been used for ploidy reductions, for example to reduce octoploids to tetrahaploids in strawberry (Janick & Hughes, 1974). Other strategies to induce haploids of female origin include the induction of parthenogenesis with irradiated pollen. In irradiated pollen, the sperm cells are inactivated and therefore unable to fuse with the egg cell, but capable to trigger haploid embryo development. This approach was used to generate female-derived parthenogenic haploids in Nicotiana rustica (Ivanov, 1938) and tomato (Ecochard et al., 1974) among others, or to reduce tetraploids to dihaploids in blackberry (Naess et al., 1998). Inactivation of pollen to be used as mentor for haploid parthenogenesis in Populus tremula was also achieved by toluidine blue treatments (Illies, 1974). A similar approach, in terms of using non-fertilizing pollen, consists of pollinating with pollen of relative species. This is the case of potato (De Maine, 2003), where dihaploids of Solarium tuberosum (4x) are obtained by pollination with Solanum phureja (2x). In this case, the two sperm cells (1x) from S. phureja fuse with the secondary nucleus of S. tuberosum (4x) to give rise to a 6x endosperm, which is able to support the development of a dihaploid (2x) embryo from the unfertilized S. tuberosum egg cell.

On the other hand, haploidy or doubled haploidy can also be achieved from the male gametophyte or its precursors through a pathway known as androgenesis. According to its initial definition, the concept "androgenesis" was invariably linked to "fertilization." It was originally coined to define a route involving sexual reproduction, egg fertilization, and subsequent inactivation of the female nucleus, so that a male-derived, haploid embryo is formed within the embryo sac (Rieger et al., 1968). This androgenic pathway, although present in nature, is extremely rare and little is known about it, likely due to its null potential for commercial exploitation. This situation contrasts with the enormous possibilities of microspore embryogenesis for both applied and basic research. Microspore embryogenesis, although described more than 40 years ago, has become of great practical importance for the agronomic industry in the last decade (Dunwell, 2010; Forster et al., 2007) due to its convenience for producing pure, homozygous DH lines much faster than the other methods above mentioned. In this route, a male-derived haploid individual is also obtained, but through a different developmental pathway, not involving egg fertilization. In addition, there is a third alternative to produce a haploid/DH individual from a male nucleus: plant regeneration from meiocyte-derived haploid callus. Examples on the existence of this route have been published during the last 40 years in highly recalcitrant species (Bal & Abak, 2007; Corduan & Spix, 1975; Gresshoff & Doy, 1972b, 1974; Segui-Simarro & Nuez, 2007), where other androgenic procedures have proven unsuccessful.

In the Proceedings of the 1st International Conference on "Haploids in Higher Plants," in 1974, the basis for a broad, inclusive use of the term "androgenesis" were established (de Fossard, 1974b). However, some researchers still want to preserve the original meaning, restricted only to in vivo androgenesis, despite the limited number of studies in plants reporting examples of egg fertilization and inactivation of the female nucleus. In addition, all of these examples date from more than 40 years ago. In parallel, other researchers are already using "androgenesis" as a synonym for microspore embryogenesis (Tai, 2005), which obviously does not include the original, in vivo pathway. And few people know that in some species it is also possible to obtain haploids/DHs by meiocyte-derived callogenesis. It is thus desirable to approach the different routes leading to male-derived haploidy from an integrative point of view. In this review, I revisit the concept of androgenesis from such an integrative perspective, focusing on these different experimental (or spontaneous) biological pathways to obtain a haploid (or potentially DH) individual from the male gametophyte or its precursors, and exploring the cellular and/or molecular triggers that may potentially induce each of them. Given its outstanding potential for plant breeding and interest in economic terms, nearly all of the research on male-derived haploidy has been focused on microspore embryogenesis. As a consequence, in the last years excellent and comprehensive reviews and books about microspore embryogenesis have been published elsewhere (Boutilier et al., 2005; Dunwell, 2010; Hosp et al., 2007a; Maraschin et al., 2005; Palmer et al., 2005; Pauls et al., 2006; Segui-Simarro & Nuez, 2008a, b; Touraev et al., 2009). For this reason, in this review I will only outline the main features of this process and the most relevant findings of this "new age" in the study of microspore embryogenesis. For more detailed information I will refer the reader to those reviews and the corresponding original research articles. However, and as a difference with recent reviews, I will pay special attention in this review to the other two, less known alternatives: male-derived haploid embryogenesis within the embryo sac, and meiocyte-derived callogenesis. I revise the most relevant literature documenting these phenomena from a biological point of view, in order to show that despite their low relevance, they constitute androgenic events. Finally, I relate these different androgenic events mentioning their differences, but also highlighting their commonalities. These common features strengthen the concept of androgenesis as a process whereby a haploid/DH individual is created from a male nucleus (of a gamete, gamete precursor or gametophyte precursor), regardless of the developmental route involved.

Male-Derived Haploid Embryogenesis in the Embryo Sac

The term androgenesis was initially confined to a male-specific form of parthenogenesis by which an embryo is believed to originate from a fertilized egg where the female nucleus is somehow inactivated or eliminated (Fig. 1, Route 1) and only those genes coming from the male parental are present (de Fossard, 1974b). In nature, this seems to be a rare alternative to sexual reproduction, only used by several clam families, a Saharan cypress tree, the Mexican axolotl Siredon mexicanum (reviewed in McKone & Halpern, 2003; Tulecke, 1965), some interspecific hybrids between Sicilian stick insects of the genus Bacillus (Mantovani & Scali, 1992), and in the particular case of a Drosophila mutant (Komma & Endow, 1995). This situation contrasts with that of teleost fishes, where this route has never been documented to occur spontaneously, but can be successfully induced and used to obtain DHs for practical purposes (Komen & Thorgaard, 2007).

In angiosperms, this spontaneous developmental pathway was described long ago, being the first reports as old as of 1929 (Clausen & Lammerts, 1929; Kostoff, 1929). However, 80 years after, published examples of this process are still scarce, probably due to the low rate of occurrence of this phenomenon in nature. Besides, most of these examples date from more than 40 years ago, with very scarce new contributions to the knowledge of this phenomenon in these four decades. In tobacco (Nicotiana tabacum) a mean rate of one androgenic haploid every 2,500 individuals was reported by Burk (1962). In maize (Zea mays), the highest mean frequency of spontaneous androgenesis reported is 1/80,000 in a sample size of 400,000 individuals from ordinary strains (Chase, 1969; Goodsell, 1961). Previous studies reported even lower frequencies of 1/187,500 and 1/ 214,000 over populations of 750,000 and 429,300 individuals, respectively (Gerrish, 1956; Seaney, 1955). As a reference, natural parthenogenesis, a more common and known event, occurred also in maize at a 1/1,000 rate (Chase, 1969). This sort of natural androgenesis was also reported to occur in some species of Capsicum (C. frutescens, Campos & Morgan, 1958), and in certain interspecific crosses of the genus Nicotiana. In particular, in crosses N. digluta [female] (2n=72) x N. tabacum [male] (2n=48), where a haploid N. tabacum (2n=24) descendant was obtained (Clausen & Lammerts, 1929), in crosses N. tabacum var. macrophylla (3n=72) x N. langsdorfii [male] (2n=18) where a small, haploid N. langsdorfii plant (2n=9) was obtained (Kostoff, 1929), in crosses N. glutinosa (2n=24) x N. repanda (2n=48), where a haploid N. glutinosa and a haploid N. repanda were obtained, among others (Kehr, 1951), and in crosses N. sylvestris [male] x F1 hybrids N. tabacum x N. sylvestris [female], where a haploid plant expressing only characters from N. sylvestris was observed (Kostoff, 1942). In addition to the spontaneous occurrence of this phenomenon, there are two documented cases of induced occurrence, in Crepis tectorum and Antirrhinum majus, using an experimental approach opposite to that above described for production of female-derived embryos with irradiated pollen. In this case, a male-derived embryo was obtained after emasculation of the female parental, irradiation of the pollen sac, and pollination with untreated pollen (Ehrensberger, 1948; Gerassimova, 1936). In all of these cases, the identification of the male origin of the haploid was based on a phenotypic characterization of the descendants expressing only those characters pertaining to the male parental. In the case of Kermicle experiments with maize, the system used allowed also for the identification of spontaneous genome duplication (Kermicle, 1974). Kermicle crossed a male parent heterozygous for a nuclear restorer ([Rf.sub.1] [rf.sub.1]) of Texas cytoplasmic male sterility (cms-T) with cms-T females ([rf.sub.1] [rf.sub.1]). He obtained 6 diploid, fertile plants, which gave 100% fertile descendants, indicating Rf Rf constitution, and one plant giving only sterile offpring, indicating rf rf.

[FIGURE 1 OMITTED]

The knowledge of the genetic control of this process is still at its infancy, with the exception of maize, where a spontaneous mutation seems to enhance the occurrence of this androgenic alternative. According to Kermicle experiments (Kermicle, 1969, 1971, 1994), spontaneous androgenesis in maize seems to be under genetic control. As opposed to the low frequency reported for wild type maize plants (see above), Kermicle observed a high incidence (up to ~3%) of spontaneous male-derived, androgenic haploids in a mutant inbred line carrying a mutant allele of the indeterminate gametophytel (ig1) gene. As opposed to many other mutants affecting the haploid gametophytes, the ig1 mutants are viable, being the ig mutant allele dominant and maternally sex-limited, since ig male plants did not show any difference in the frequency of androgenesis. This mutation is located on chromosome arm 3L and has as a direct effect an increased number of nuclear divisions before cellularization of the embryo sac, which generates in the embryo sac an indeterminate, extra number of micropylar and synergids cells, egg cells, central cells, and polar nuclei within central cells (Evans, 2007; Guo et al., 2004; Huang & Sheridan, 1996; Lin, 1978, 1981). This mutation also affects nuclear migration and cellular differentiation (Lin, 1978, 1981).This array of pleiotropic effects at the level of the female gametophyte gives rise to polyembryony and elevated ploidy levels of the endosperm after fertilization, derived from fertilization by different pollen tubes, together with miniature endosperms, early abortion of seeds and the occurrence of androgenic and gynogenic haploids in a 2:1 ratio (Kermicle, 1969, 1971, 1994; Lin, 1984). Other effects include sporophytic male sterility in some genetic backgrounds (Kermicle, 1994) or abnormal leaf morphology (Evans, 2007). Consistent with this, it was suggested (Lin, 1981) that the ig1 locus does not control a single process but acts as a regulator of other downstream genes involved in female gametophyte development. Some of the genes regulated by the ig1 locus could be those involved on the switch from proliferation to differentiation in the embryo sac, since in mutant embryo sacs the proliferative phase is prolonged (Huang & Sheridan, 1996; Lin, 1978). But most importantly, ig1 is thought to be involved in the repression of the embryogenic program in those cells that lack one of the two parental genomes. In other words, ig1 would be preventing the uncontrolled trigger of proliferation in non-gametic cells, and of embryogenesis in egg or sperm cells prior to fertilization. Once fertilization and karyogamy occur, ig1 would allow for embryogenesis to take place in the unicellular zygote. Thus in ig1 mutants, the embryogenic development of the haploid egg cell or the two haploid sperm cells after pollen tube discharge would not be repressed. This would be consistent with the above mentioned 2:1 rate of androgenic versus gynogenic embryo production observed by Kermicle.

The cellular aspects of this process are also largely unknown. In angiosperms, the haploid androgenic embryo is believed to originate from a reduced sperm nucleus. Although it is clear that the genetic material of the androgenic embryo comes from the male parental, it was generally assumed that androgenic haploids originate exclusively from divisions of the sperm nucleus, subsequent to fertilization. However, there does not seem to be enough supporting evidence for this. As pointed out by Pandey (1973), it might well be the vegetative nucleus of the pollen grain that is responsible for proliferative growth. This hypothesis is consistent with the demonstrated role that the vegetative nucleus has in a much better studied process as is microspore embryogenesis (Maraschin et al., 2005). The contribution of the female gametophyte remains obscure as well. Although it was shown that the resulting haploid individual inherits cytoplasm of female origin (Chase, 1963; Goodsell, 1961), it is not clearly established that such cytoplasm derives from the egg cell (Lacadena, 1974), so the possibility of a female gametophytic origin other than the egg cell should not be ruled out. This possibility should be carefully considered in the case of maize androgenic embryos derived from ig1 mutants. According to the recent knowledge above reviewed about the role of the ig1 gene, the androgenic mechanism triggered by the ig1 mutation would be different from that generally assumed, since the sperm-derived haploid embryo would not imply fertilization and elimination or inactivation of the female genome. In turn, the androgenic embryo would emerge directly from a sperm cell, without the need for fertilization as a prerequisite, and having a cytoplasmic background contributed by the sperm cell, and/or perhaps the central cell, but not the egg cell. Another intriguing aspect is the elimination of the maternal genome. Similar processes of chromosome elimination occur either in a spontaneous or in an induced manner during protoplast fusion (Liu et al., 2005), but unfortunately, as in androgenic elimination, very little is known about the underlying mechanism. These two processes do not seem to share a common trigger, since elimination in the fused protoplast nucleus is strongly affected by factors such as phylogenetic distance, different ploidy levels of parents, or irradiated dosage for induced elimination (Liu et al., 2005). Some of these factors, which may also be present in other types of haploid-inducing techniques (gynogenesis through interspecific or intergeneric hybridization) are not present in most of the reported cases of in vivo androgenesis. For example, it could be argued that phylogenetic distance might have an effect in the induction of androgenic haploids by crosses between the Nicotiana species mentioned above. However, in all of the examples of spontaneous, intraspecific crosses documented in maize, tobacco, Capsicum, or even in the maize ig1 mutants, phylogenetic distance cannot be used as an argument. In summary, it seems clear that in vivo androgenic plants inherit a male genetic background, but the origin of their cytoplasm and the mechanisms by which the female genome is inactivate still await to be unambiguously determined.

The practical application of this natural phenomenon is rather limited. In the particular case of the ig1 gene in maize, the side effect of an increased frequency of androgenic embryos has been used in breeding to transfer germplasm (e.g., a male sterility conditioning cytoplasm) from one variety of maize to the cytoplasm of another variety (Kindiger & Hamann, 1993). The use of the inbred line W23 (Kermicle, 1969) carrying the ig1 mutation, in combination with microsatellite (SSR) fingerprinting, has been proposed as a strategy for maize hybrid development in tropical germplasm (Belicuas et al., 2007). Recently, two patents have been developed in order to exploit the androgenic potential of the ig1 mutation by introducing this mutated gene in other plant species (Evans, 2005, 2009). However, apart from these particular examples, the low efficiency of this sort of in vivo androgenic development has limited its practical, routine use in maize. Beyond maize, the ig1 mutation has not been described in any other crop, and to the best of my knowledge, there is no other gene known to have a direct or indirect role in the in vivo induction of androgenic haploids. Thus, in vivo androgenesis as a system for producing natural androgenic haploids is not expected to have a significant impact in plant biotechnology in the next future.

In summary, 80 years after the first reports, little new is known about the cellular, molecular, or genetic basis of this intriguing process. There are many aspects still to be elucidated about this unusual developmental pathway that stands as a biological rarity. Unfortunately, although this pathway has been known for a long time, new findings are not likely to be published in the next few years. This is in part due to the difficulties imposed by the low frequency, but mostly due to the lack of applied interest of this spontaneous process.

Microspore/Pollen Embryogenesis

Besides those above mentioned, a male-derived haploid plant can be obtained through other different ways. The most common and best studied route is microspore/pollen-derived haploid embryogenesis. In this route, a male-derived haploid plant is formed when microspores, typically at the vacuolate stage, or young pollen grains are experimentally deviated from their original gamete-producing pathway towards embryogenesis (Fig. 1, Route 2). In 1964, Guha & Maheshwari demonstrated that this pathway can be induced through the in vitro culture of microspore/pollen-containing anthers of Datura innoxia. After Guha and Maheshwaft, numerous researchers realized the practical significance of such a discovery and explored this pathway in many different species and genera, and now it is considered the most powerful tool to obtain DH plants for practical breeding purposes or genetic mapping, among others. In addition to the method proposed by Guha and Maheshwari for anther excision and inoculation into solid medium, further research has evidenced that microspore embryogenesis can also be achieved through direct microspore isolation from the anther locule and inoculation into liquid medium. This latter method is more technically demanding, which limits its application to a range of species narrower than anther culture. However, in those species where it has been well set up, it is the method of choice due to its higher efficiency and reduced time to obtain haploid and/or DH plants.

In 2003, successful induction of microspore/pollen embryogenesis through in vitro culture of anthers or isolated microspores was documented for more than 250 plant species (Maluszynski et al., 2003b), including many species of agronomic interest, from herbaceous crops such as wheat, barley, rice, rapeseed (canola), tobacco or corn (reviewed in Maluszynski et al., 2003a), to trees such as mandarin, bitter orange or cork, among others (reviewed in Germana, 2006; Srivastava & Chaturvedi, 2008). A list of species and varieties of agronomic interest where protocolos for microspore embryogenesis are published and available can be found in http://www.scri.ac.uk/assoc/COST851/DHtable2005.xls. Among all of these species potentially inducible to microspore embryogenesis, some of them exhibit an excellent androgenic response, namely rapeseed (Brassica napus), tobacco (Nicotiana tabacum), and barley (Hordeum vulgate). But not all of the interesting crops respond efficiently to embryogenesis induction. Only in few of them the androgenic potential is high enough to obtain a number of embryos sufficient to be routinely used in DH production for breeding programs. Species like rapeseed, tobacco, barley or wheat are responsive enough to be considered as model systems. However, other scientifically or economically interesting species like Arabidopsis or tomato still remain extremely recalcitrant to haploid embryogenesis induction. In between from model and extremely recalcitrant species, many others are still far from an acceptable response. That is the case of woody species, where induction of microspore embryogenesis would be especially advantageous for breeding programs, but successful reports are still very limited (reviewed in Srivastava & Chaturvedi, 2008). In gymnosperms, for example, embryos have been obtained in some instances, but no plant regeneration has been reported (Andersen, 2005). It is thus clear that the genotype is one of the key factors controlling the embryogenic response of the microspores. In virtually all of the species studied so far, a constancy on the extremely different response between different cultivars has been observed. Even within model species the genotype has a strong influence, as there exist high and null response cultivars within the same species (Ferrie et al., 1995; Malik et al., 2008; Touraev et al., 2001). This fact together with the demonstration that androgenic competence can be inherited by descendants reveals a genetic basis and thus the possibility of breeding for this trait (Beckert, 1998; Rudolf et al., 1999).

As said, the late microspore--young pollen are the sensitive stages for embryogenesis induction. As a general rule, it is accepted that young microspores or mature pollen cannot be induced. For most species the only period of microspore sensitivity to inductive pretreatments revolves around the transition from the uninucleate, vacuolate microspore to early, bicellular pollen, i.e. around the first pollen mitosis (Maraschin et al., 2005; Raghavan, 1986; Touraev et al., 2001). In fact, there seems to be an association between the ability to undergo embryogenesis and the polarity of the first pollen division (Reynolds, 1997; Twell & Howden, 1998). But in non-model species, where induction is more difficult to achieve, narrower timeframes have been reported. Most studies support the notion that the inducible stage is the late, vacuolate microspore, where the microspore is ready to divide asymmetrically. On the other hand, some laboratories report higher efficiencies from early, just divided pollen. Capsicum annuum is a good example to illustrate such a discrepancy (Kim et al., 2004, 2008). In addition, in several crops, including model species such as rapeseed, induction has also been achieved at even later stages with acceptable yields (Binarova et al., 1997). Thus, although there is not a universal consensus about the exact start and end point between which reprogramming can be universally achieved, it seems that vacuolate microspores and in some cases young bicellular pollen are specially suitable to be induced. This is likely due to their proliferative, not yet fully differentiated transcriptional status (Malik et al., 2007), as opposed to mature pollen, where a specific expression program for pollen maturation genes is activated (Honys & Twell, 2004). It is also likely that advances in the specific particularities of recalcitrant species will lead to a widening of the inducible range of stages.

Besides the genotype and the developmental stage of the microspore, the third critical factor is culture conditions, and in particular the physical and/or chemical treatment necessary to trigger microspore embryogenesis. Once the microspore is at the fight stage, an inducing treatment must be applied. At present there is a wide consensus about the type of treatment potentially capable of induction: a stressing treatment. Microspores must be submitted to a set of physicochemical factors that promote a stress response, which in turn trigger the embryogenic response. Useful stressing agents are diverse. The most used stressors are heat and starvation, but different species may require particular combinations. For more detailed information, the reader is referred to an excellent review about the different stressing treatments proved to be valid in the most important androgenic systems (Shariatpanahi et al., 2006). As opposed to stress, hormones seem to be less critical for successful induction. According to Aionesei et al. (2005), the relatively low importance of hormones in microspore embryogenesis relies on the fact that hormonal autotrophy is a condition sine qua non for true embryogenesis.

Once exposed to the inductive treatment, many microspores immediately arrest and/or die. Some microspores follow a pollen-like type of development until they reach the mature pollen stage. Then, they arrest and die, typically at the 4th-5th day (Hosp et al., 2007a). Some others are effectively induced by the stress treatment and undergo a plethora of changes at different levels, from the whole cell architecture to gene expression. Large-scale changes include nuclear repositioning to the cell center, rearrangements of nuclear and cytoskeletal elements and breakdown of the large central vacuole. Gene expression changes can be grouped into three principal categories (Pauls et al., 2006): cellular response to the stress, suppression of the gametophytic program, and expression of the embryogenic program. An important number of genes, proteins and metabolites have been identified as directly or indirectly related to the trigger of each of these stages. They have been the subject of recent and excellent research articles and reviews (Boutilier et al., 2005; Dunwell, 2010; Hosp et al., 2007a, b; Joosen et al., 2007; Malik et al., 2007, 2008; Maraschin et al., 2005, 2006; Munoz-Amatriain et al., 2006; Pauls et al., 2006; Segui-Simarro & Nuez, 2008a; Stasolla et al., 2008). Since the amount of information largely exceeds the scope of this review, we will only outline some of their most relevant findings. The reader is referred to them for further details.

It is clear that in order to generate a stress response, the stress signal must be internalized first. Although little is known in this respect, abscisic acid (Maraschin et al., 2005, 2006; Reynolds & Crawford, 1996; Tsuwamoto et al., 2007; van Bergen et al., 1999) and some plant extracellular signal-regulated kinase homologues (Coronado et al., 2002; Segui-Simarro et al., 2005) have been proposed as intermediate transducers of extracellular stress signals for the activation of specific gene expression programs (reviewed in Segui-Simarro & Nuez, 2008a). Related to the cellular response to stress, changes in the levels of different heat-shock proteins (HSPs) have been described in several species during stress-induced microspore embryogenesis after exposure to heat, colchicine, or starvation stress (Cordewener et al., 1995; Segui-Simarro et al., 2003; Smykal, 2000; Zarsky et al., 1995). It is currently believed that HSPs have a cytoprotective role related to stress tolerance (Segui-Simarro & Nuez, 2008a). Instead, upregulation of catalase and glutathione S-transferase genes has been interpreted as a protective response against the oxidative stress generated by the in vitro culture conditions (Maraschin et al., 2005).

In addition to triggering the embryogenic program, the cell must cancel the gametophytic program. The blockage of starch synthesis (a marker of pollen maturation) and the elimination of starch reserves are key events in this process. Indeed, downregulation of a number of genes involved in starch biosynthesis and accumulation has been observed in barley induced microspores, together with the induction of genes for starch and sucrose breakdown (Maraschin et al., 2006). In parallel, a sort of dedifferentiation program seems to operate in order to clean the cytoplasm from microgametogenesis-related molecules (reviewed in Maraschin et al., 2005; Segui-Simarro & Nuez, 2008a).

The first indication for the onset of the embryogenic program is a symmetric division of the microspore (Simmonds & Keller, 1999; Smykal, 2000; Zaki & Dickinson, 1991), as opposed to the asymmetric division that defines the first pollen mitosis (compare pollen figures of 'Route 0' with the symmetrically dividing microspores of the green rectangle of 'Route 2' in Fig. 1). In most androgenic systems, such division is followed by continuous proliferative, randomly-oriented divisions, leading to an undifferentiated, globular mass of embryonic cells, as opposed to the well ordered division pattern of zygotic embryos (Segui-Simarro & Nuez, 2008a). From this stage on, the microspore-derived embryo (MDE) progressively approaches the morphology of its zygotic counterpart. Exceptionally, in a model system such as rapeseed it is possible to reproduce the ordered sequence of stages of zygotic embryos even from the very first stages, including the transversal divisions that give rise to a filamentous suspensor structure, and in parallel to an embryo proper at the end of the suspensor opposite to the basal cell (Joosen et al., 2007; Malik et al., 2007; Segui-Simarro & Nuez, 2008a; Supena et al., 2008). Other aspects of the embryogenic program are also similar between MDEs and zygotic embryos, but not equal. That is the case for hormonal regulation. Growth regulators like ethylene, IAA or ABA have been shown to be important for particular aspects of MDE development, as it happens for zygotic embryos, but either the temporal profiles or the absolute levels of these plant hormones differ from zygotic embryogenesis (Belmonte et al., 2006; Evans & Batty, 1994; Hays et al., 1999, 2000, 2001, 2002; Ramesar-Fortner & Yeung, 2006; Rudolf et al., 1999). These differences can be attributed to the absence of endosperm as a source of regulatory cues (Segui-Simarro & Nuez, 2008a). Other substances, including arabinogalactan proteins (AGPs), hordeins and chitinases, beneficial for the promotion of the zygotic embryo, are also synthesized by the MDE (Borderies et al., 2004; Boutilier et al., 2005; Letarte et al., 2006; Pulido et al., 2006; Tang et al., 2006) likely to mimic the zygotic scenario (Segui-Simarro & Nuez, 2008a), where the embryo-surrounding tissues account for the synthesis of these substances. Thus, MDEs may adopt certain endosperm-specific functions in order to overcome some of these deficiencies, as suggested by the upregulation of a number of endosperm-specific genes in MDEs of different species (Magnard et al., 2000; Massonneau et al., 2005), and by the ability of proteins secreted to the medium by maize, barley and wheat MDEs to promote and sustain in vitro zygotic development (Holm et al., 1994; Kumlehn et al., 1998; Paire et al., 2003).

The absence of regulatory cues coming from the endosperm or other seed tissues may also be behind the callogenic response observed in some species where indifferentiated, callus-like proliferative structures have been described to be formed upon in vitro culture of microspores and pollen grains. Haploid and DH plants have been regenerated from these calli, which has made this method a source of DHs in many species. This phenomenon has been described in a wide range of species, including coffee (Silva et al., 2009), trees as loquat (Li et al., 2008) or different poplar species (Baldursson et al., 1993), cereals such as rye (Ma et al., 2004) or wild barley relatives (Piccirilli & Arcioni, 1991), and

many ornamentals such as lily (Han et al., 2000), narcissus (Chen et al., 2005), coneflower (Zhao et al., 2006), Anemone (Ari et al., 2007), Dianthus (Nontaswatsri et al., 2008) or chrysanthemum (Yang et al., 2005). In cucurbits, callus proliferation seems the predominant or only way observed (Gdmes Juhasz & Jakse, 2005; Song et al., 2007). In oat, the genotype seems responsible for the callus or embryo-type of response. From 38 different wild and cultivated oat genotypes, 31 of them produced callus whereas 7 produced embryos, when exposed to the same culture conditions (Kiviharju et al., 1998). In other species, culture conditions have been found critical for callus or embryo generation. For example, in eggplant, the pionnering work of Dumas de Vaulx and Chambonnet described the generation of DH plants from embryos developed within anthers cultured in vitro (Dumas de Vaulx & Chambonnet, 1982). However, nearly 15 years after, Miyoshi described a method for eggplant DH production based on isolated microspore culture, where microspore gave rise exclusively to callus-like structures (Miyoshi, 1996). In our group, we have confirmed both results, but in addition we have found that by changing some conditions in the culture medium it is possible to induce embryo development up to a certain point. From then on, the eggplant embryo seems not to have all the elements needed in the culture medium, and therefore reverts to a proliferative callus status (Corral-Martinez et al., in press). Similarly, in pepper microspore cultures, different conditions may lead to different efficiencies in embryogenesis induction, different qualities in the embryos produced, or even the parallel production of good and bad quality embryos together with proliferating callus (Corral-Martinez et al., in press; Kim et al., 2004, 2008; Supena et al., 2006a, b). These data strongly point to the notion that callus proliferation is not an alternative, genotype-dependent pathway for haploid development and regeneration. Instead, it seems the consequence of suboptimal experimental conditions, which preclude the embryo to develop in a proper manner. It is widely known that successful induction of embryogenesis is more complex and demanding than induction of callogenesis. It is likely that most, if not all of the species where microspore-derived callus have been described, would produce embryos as soon as the particular needs for embryo development in these species are known and applied. Two evidences give further support to this notion. First, some of the species currently considered as highly responsive, embryo-producing model systems (e.g. barley, wheat or rice) were first considered as recalcitrant, at a time when calli were observed in cultures, the rate of embryo production was low or null, and obviously culture conditions were far from being optimal (Bouharmont, 1977; Gonzalez-Medina & Bouharmont, 1978; Myint & de Fossard, 1974; Picard et al., 1974). Second, in maize ig1 mutants, in vivo male-derived androgenic haploids are though to come from the embryogenic development of a sperm cell, a type of cell where in vitro embryogenesis induction has never been described. It is likely that being in the best environment possible (the embryo sac of its same species) with a perfect, in vivo control of all of the nutrients, growth factors and regulatory cues needed, can allow for a sperm cell to proceed through embryogenesis with a high probability of Success.

In summary, there are still many different, general and specific questions to be answered about microspore embryogenesis. Among them, the search for the factors and ultimately the genes governing the androgenic switch has concentrated most of the research efforts. During more than 40 years, numerous research groups have devoted their work to the search of the master key that switches microspores towards embryogenesis. Although numerous progresses have been made due to the development of advanced and sophisticated genomic, transcriptomic, proteomic and imaging tools, at present the search still continues. Several candidates have been postulated during these years, but the attempts to clearly identify the genes or gene groups unambiguously conferring the androgenic competence have been unsuccessful to date. However, in the last decade, a common landscape of cellular, molecular and genetic changes that define the inductive process is beginning to be depicted. It seems that we are slowly approaching a global understanding of this process. Instead of a monogenic genetic control, it seems that a number of inducing factors, different for each species, must concur at the same time and place for microspore embryogenesis to initiate. The MDE expression profiles of model species such as rapeseed (Joosen et al., 2007; Malik et al., 2007; Stasolla et al., 2008), barley (Maraschin et al., 2006), and tobacco (Hosp et al., 2007b) is remarkably similar, which highlights the importance of some of those common genes in this process. An interesting research line for the future would be to find whether the genes up or downregulated during microspore embryogenesis also exhibit expression changes in other androgenic pathways.

Meiocyte-Derived Callogenesis

Male-derived haploid and DH plants can also be regenerated from meiocyte-derived callus.

This is a rare route, less frequent and documented than microspore embryogenesis, where under the appropriate in vitro conditions, post-recombination meiocytes are induced to proliferate into calli. Then, haploid and DH plants can be obtained, either through indirect embryogenesis or through organogenesis over the meiocyte-derived callus (Fig. 1, Route 3). Callus induction from immature meiocytes has been reported for Arabidopsis thaliana, Vitis vinifera, and Digitalis purpurea (Corduan & Spix, 1975; Gresshoff & Doy, 1972b, 1974), although these pioneering studies were never continued. A similar situation was occasionally reported in barley, where nuclear fusion between adjacent microspores, suggested to be meiotic products not well separated, gave rise to multicellular structures made of diploid (potentially heterozygous), proliferating cells (Chen et al., 1984). In addition to in vitro induction, a similar process has also been documented to occur in vivo in anthers of Narcissus biflorus (Koul & Karihaloo, 1977) and of male-sterile interspecific hybrids (Solanum chacoense x Solanum tuberosum; Ramanna & Hermsen, 1974). Nevertheless, most of the knowledge about this process comes from research on tomato (Solanum lycopersicum) anther cultures. Given the outstanding importance of this crop all over the world, many laboratories have worked in the last four decades trying to induce haploids from tomato anthers (reviewed in Bal & Abak, 2007; Corral-Martinez et al., in press). The extreme recalcitrancy of this species has precluded the generation of haploid or DH tomato plants by the more conventional route of microspore embryogenesis, and only limited success has been reported in the induction of the first sporophytic divisions from isolated microspores (Dao & Shamina, 1978; Segui-Simarro & Nuez, 2007; Varghese & Gulshan, 1986). However, several reports have demonstrated that it is possible to regenerate haploid and DH plants from tomato anthers under certain experimental conditions (Gresshoff & Doy, 1972a; Segui-Simarro & Nuez, 2007; Zagorska et al., 1998, 2004).

Among these conditions, the most critical seem to be the genotype and the developmental stage. As in the other two androgenic ways to haploidy, the genotype plays an important role. In particular, male-sterile mutant lines have been shown to be especially sensitive to being induced (Segui-Simarro & Nuez, 2007; Zagorska et al., 1998; Zamir et al., 1980). Interestingly, male-sterile phenotypes are usually manifested at the late meiocyte stage, which in most cases overlap with the inductive window of meiocyte-derived callogenesis. The issue of the inducible developmental stage for tomato has been a matter of debate for decades. Pioneering studies proposed both the meiocyte (Gresshoff & Doy, 1972a, 1974; Zamir et al., 1980) and young microspores just released from the tetrad (Dao & Shamina, 1978; Gulshan & Sharma, 1981; Levenko et al., 1977; Varghese & Gulshan, 1986) as the stages inducible to callogenesis. It is worth to mention that most of the works supporting the microspore as the appropriate stage were unable to obtain fully regenerated haploid or DH plants. However, in the last decade and coinciding with the advent of new and sophisticated molecular and cellular techniques, all of the studies on this subject pointed to the meiocyte as the developmental stage to interfere with (Shtereva et al., 1998; Summers et al., 1992; Zagorska et al., 1998). More recently, it was found that for efficient induction, meiocytes must have passed meiotic prophase I, but tetrads of microspores must not be formed yet (Segui-Simarro & Nuez, 2005). Thus, culture of tetrad and microspore-carrying anthers exposed to the same experimental conditions consistently failed to develop callus (Segui-Simarro & Nuez, 2007). This developmental window implies that recombination must be successfully finished without disruption, but microspore formation (tetrad walling) has to be prevented.

After induction, meiocyte-derived calli originate by two different ways: (1) from haploid products still enclosed within the tetrad, that stop their gametophytic program and start proliferation, or (2) when diploid cells, coming from the fusion of two haploid products separated by defective, incomplete, or absent cell walls, start proliferation. In the first scenario, since callus originates from a tetrad-enclosed meiotic product (a microspore), one could argue that microspores are the true callus precursors. Nevertheless, it has to be remarked that induction must take place prior to microspore formation, i.e. tetrad compartmentalization. Then, haploid callus cells duplicate their haploid genome by nuclear fusion, giving rise to true DH cells which may regenerate a DH plant (Segui-Simarro & Nuez, 2007).

The second scenario would not give rise to a DH, since fusion of two reduced meiotic products generates new allele combinations not necessarily homozygous. Our group has recently confirmed this notion by characterizing 19 meiocyte-derived calli with five independent microsatellite (SSR) molecular markers proven polymorphic for the parental donor plants (J. M. Segui-Simarro, unpubl.). Homozygosity for all of the five SSR was found in 14 out of the 19 calli tested. However, five calli were found heterozygous for at least one SSR. These results are consistent with the notion of fusion between two haploid, different meiotic products. In addition to meiocyte-derived calli, one cannot rule out the possibility that some diploid calli originate from the anther connective or filament tissue induced to proliferate. Indeed, filament tissues typically exhibit a high proliferative response when cultured in vitro (Segui-Simarro & Nuez, 2006), and it is believed that tomato anther tissues at meiotic stages are more sensitive to tissue culture than those of older stages (Bal & Abak, 2007). All of these collateral, undesirable events make mandatory the analysis of every single regenerant, which clearly compromises the usefulness of this method for practical purposes.

As mentioned above, the number of examples of this route is limited. So the question arises as to why this process has not been extensively studied and documented in other, more interesting species. First, it must be noted that meiosis is an extremely sensitive developmental stage where the cell purges its cytoplasm from the expression of a diploid genome, adjusts it to the expression of a haploid genome, and readapts for the expression of the gametophytic potential. Conceivably, any physiological disturbance may collapse the whole process. This is why in vitro culture of meiocytes has been historically reported as "largely unsuccessful" (Shivanna & Johri, 1985). In fact, we have repeatedly tried to cultivate isolated meiocytes, but they never progressed in culture, which has precluded further cell tracking experiments for additional support of the meiocytic origin of callus. This has been the reason why the only way to dissect the cellular origin of haploid callus has been serial sectioning of in vitro cultured anthers (Segui-Simarro & Nuez, 2007). Anyway, the simplest explanation relies on the fact that most of the examples of this pathway come from some of the most recalcitrant species to be induced towards microspore embryogenesis. Given the economic impact of these species and/or their suitability for both fundamental and applied research, parallel efforts have been made to circumvent their extreme recalcitrancy by exploring alternative pathways to androgenic haploidy, whereas in species more easily inducible to microspore embryogenesis, focusing on alternatives is unnecessary in practice, and resources can be devoted to investigate more promising ways to doubled haploidy. Indeed, we do not exclude at all that meiocyte-derived callogenesis can also be induced in other species, especially in those where microspore embryogenesis is easily achieved. Examination of this possibility in a broad range of different species would provide additional support to this notion.

It could be argued that this pathway is not sufficiently different from microspore embryogenesis to be considered as an independent one, since callus may be formed in both cases upon in vitro induction, and it is likely that a better knowledge of the process and its particularities would make it even more similar to microspore embryogenesis. However, the marked differences in the inducible stages in each case have profound consequences in the final result of each process. In particular, the fact that the meiocyte is the inducible stage implies that through this route, androgenic haploids or DHs are not the only final product. Instead, androgenic diploids with different heterozygous combinations are also obtained. Thus it seems reasonable to consider at present, with the current knowledge about this process, this route as different from microspore embryogenesis. In light of current data, it is clear that from an applied or economic point of view, the usefulness of this pathway is far away from that of microspore embryogenesis, mostly due to its low efficiency in terms of DH yield. However, it is relevant in biological terms, since it reflects that totipotency of male gametophytes would not be restricted to the stage of vacuolate microspore-early bicellular pollen, as is widely accepted. Instead, it shows that under the adequate conditions, meiocytes are also able to deviate from their original program. This fact, together with other evidences that demonstrate the potential to produce MDEs of older pollen stages (Binarova et al., 1997) or even sperm cells, as in the maize ig1 mutants under in vivo conditions (see "Male-Derived Haploid Embryogenesis in the Embryo Sac"), suggests that further research in this and other stages could extend the current knowledge of male germ line totipotency and developmental plasticity.

A Revision of Pandey's Hypothesis

In this review we have dealt with three routes to haploidy, alternative to natural, zygotic embryogenesis and male gamete development in angiosperms (Fig. 1, Route 0): (1) in vivo haploid embryogenesis in the embryo sac (Fig. 1, Route 1), (2) microspore embryogenesis (Fig. 1, Route 2), and (3) meiocyte-derived callogenesis (Fig. 1, Route 3). These routes differ in several aspects (summarized in Table 1), being the most evident aspect the stage when the original program is deviated (the sperm cell/ fertilized egg, the vacuolate microspore/young pollen grain, and the post-recombination meiocyte respectively). However, most of the differences are observed between route 1 and the other two routes. Route 1 starts with a cell type which represents the end of a developmental pathway, the male gamete-producing pathway, whereas routes 2 and 3 are induced from cell types that represent intermediate stages in the male gametophytic pathway. Thus, as opposed to route 1, routes 2 and 3 imply a deviation from the original gametophytic program towards mature pollen and gamete formation. In both cases, it seems necessary to "block" the original program prior to the induction of a new one. It was previously postulated (Pandey, 1973) that the young microspore has four possible kinds of potencies: to develop normally into a male gametophyte, to develop into a female gametophyte (giant embryo sac-like pollen grains from certain species, see Pandey, 1973 for a review), to develop into a sporophyte (embryo), and to dedifferentiate into a callus. According to Pandey, in normal conditions the microspore displays a strong bias towards maleness. The strength of this bias would be a genotype-specific, particular feature of each species, and would be determined by the molecular environment of the cell, i.e. the developmental program being expressed at a given time. Therefore, it would account for the different recalcitrancy of a species to be deviated from maleness. Under certain natural or induced conditions, the balance of potencies can change by changing the molecular environment. Two kinds of factors are compulsory for such a shift: first, factors that neutralize the original program, reverting the cell to a totipotent status, and second, factors that provide the necessary environment to trigger a new developmental program (e.g. microspore embryogenesis). When such a specific environment is not present, the cell enters an undifferentiated, proliferative callus phase. Callus, in turn, may give rise to embryos, organs, or full plantlets provided that proper conditions for indirect embryogenesis, organogenesis, or plant regeneration are found and applied.

The last decade has witnessed a new way to approach the study of the molecular basis of androgenesis, focusing especially on the most useful alternative: microspore embryogenesis. The parallel work of several laboratories has boosted the search for the gene(s) responsible for the androgenic switch and considerable progress has been made. From a holistic point of view, these studies (see "Microspore/Pollen Embryogenesis") have confirmed the validity of Pandey's hypothesis, more than 35 years after. These studies have proven that induction of androgenesis must be preceded by a repression of the pollen-specific gametophytic program, thus keeping the microspore in an undifferentiated, totipotent status. These works also confirm that Pandey's hypothesis remains useful to explain the different alternatives observed to microsporogenesis (reviewed in Segui-Simarro & Nuez, 2008a). It seems relatively easy to reprogram towards embryogenesis in an increasing number of species by applying stress treatments to the microspore. In other species, more reluctant to microspore embryogenesis, microspore dedifferentiation can be achieved but conditions are not yet optimal for embryogenesis. This would be the case of microspore cultures in cucurbits, some solanaceae and many ornamentals, where callus proliferation has been extensively documented (see "Microspore/Pollen Embryogenesis"). A third option would be that of the most recalcitrant species (tomato, grape, Arabidopsis), where much more severe stresses have to be applied at earlier developmental stages to promote dedifferentiation and callus proliferation, but where the button to switch direct microspore embryogenesis on still awaits to be pushed.

Similar Triggers for All of the Androgenic Alternatives?

In spite of their specific differences, the three ways have in common a strong genotypic dependence and the final product: an androgenic haploid or DH, i.e. an individual whose genetic background is contributed exclusively by a male donor. Since the inducible stage in each route is markedly different, the most logical reasoning would lead to assume that the triggering factors and the mechanisms involved in each case are different. However, if we put the current knowledge about these three routes together, some coincidences between them may arise in terms of inducing factors. As seen in "Microspore/Pollen Embryogenesis", it is currently believed that in order to induce a microspore towards embryogenesis, a number of external factors must concur in a timely manner to generate metabolic alterations, which in turn promote changes of higher order in chromatin structure (Hosp et al., 2007a). This leads to the activation of certain genes, involved in proliferation and in the expression of a embryogenic program, which finally gives rise to the haploid embryo. The general molecular triggers of the process of in vivo androgenesis, supposedly from a fertilized egg, are largely unknown. Only some data are available in the particular case of androgenesis induced by the presence of the ig1 mutant gene in maize. In this case, it is thought that the wild type ig1 gene is a repressor of proliferation and embryogenesis in the embryo sac (Evans, 2007). Thus, it is likely that in ig1 mutants, the non-repressed expression of these genes generate the metabolic alterations needed to promote proliferation in cell types such as synergid or antipodal cells, and embryogenesis in embryo-producing cells such as the egg and the sperm cells after pollen discharge. In other words, the lack of ig1 function would create in the maize embryo sac, and therefore in the sperm cells present after pollen discharge, an embryogenesis-promoting, molecular environment similar to that created by the stress treatments in the case of microspore embryogenesis or meiocyte-derived callogenesis. In support of this hypothesis counts the fact that the ig1 mutation affects the nucleus and the microtubular cytoskeleton of the cells in the embryo sac (Huang & Sheridan, 1996) in a way similar to that described for stress treatments in induced microspores (reviewed in Maraschin et al., 2005; Segui-Simarro & Nuez, 2008a). For example, in both cases the nucleus fails to migrate to the position in the cell needed to follow the normal gametophytie program. Besides, all of the microtubular structures involved in cell division (spindle, phragmoplast) have an altered position, orientation and behaviour in both cases. Further support for this speculation may come from the studies of Komma and Endow in androgenic Drosophila ncd mutants (Komma & Endow, 1995), where they found an increased androgenic frequency due to the loss of function of NCD, a microtubular motor protein from the kinesin family.

In this context, it is tempting to speculate that in the other documented examples of in vivo androgenesis, where no ig1 mutant genes are involved, the external factors that govern the occurrence of androgenesis would promote effects similar to those promoted by the ig1 mutation of maize or the ncd mutation in Drosophila. For example, among these factors one could mention wide hybridization. The fact of hybridizing different species, as those mentioned in "Male-Derived Haploid Embryogenesis in the Embryo Sac", might have an effect similar to that described for the ig1 mutant gene in terms of creating a similar haploid embryo-promoting molecular environment, suitable for the sperm cell to develop as an embryo. Additionally, it might well be possible that androgenesis would be induced directly over the sperm cell, before fertilization and karyogamy, as described for maize ig1 mutants. Indeed, this is the best studied example of in vivo androgenesis, and as mentioned in "Male-Derived Haploid Embryogenesis in the Embryo Sac", several authors have expressed their reasonable doubts about the assumed, but not unambiguously proven origin of the male-derived embryo in this route as coming from a fertilized egg which losses or inactivates the female chromosomes. It was shown that wide hybridization between two chickpea relatives (Cicer arietinum x C. pinnatifidum) was able to induce, upon exposure to zeatin, sporophytic divisions in microspores within the anthers of the hybrid, (Mallikarjuna et al., 2005). This result would be reinforcing the notion of a role of interspecific hybridization as a trigger for haploid embryogenesis, and would also be extending this role beyond the sperm up to a different cell type, the microspore.

Finally, it would be interesting to pose the following question: what is the biological significance of androgenesis? In a review from 2000, Wang et al. proposed a very attractive hypothesis: "The spore development in mosses and ferns shows clear parallels with the androgenesis process. In both cases the haploid spore produced by the sporophyte divides mitotically and develops into a (haploid) multicellular structure, and both processes start with an increase of the cell volume. So we may hypothesize that during the evolution of plants, the spore development pathway into multicellular structures was greatly shortened in favor of direct gamete formation, but that this pathway is still present and can be activated as is shown in androgenesis". Perhaps, behind androgenesis there is just an alternative survival mechanism based on the totipotency of plant cells and on the capacity of primitive plants to develop a haploid multicellular structure during the extended haploid phase of their life cycle. Perhaps, androgenesis is just a developmental pathway based on this capacity of ancient plant relatives, and currently displaced by the evolutionary advantages of sexual reproduction. Thus, when we block or disturb the normal process of sexual reproduction by isolating anthers, meiocytes, microspores or pollen grains, or by hybridizing two distant species having severe pre-zygotic or post-zygotic reproductive barriers, we may be forcing the plant to activate this ancient, silent up to now mechanism in order to ensure reproduction by any means. However, the particular conditions of in vivo or in vitro androgenesis would favour the development of a haploid embryo or callus, and not a small multicellular gametophyte (protonema) as in ferns, or even a haploid plant as is the case for mosses.

Concluding Remarks

In this review I have summarized the main aspects of the three androgenic routes known so far in plants, in vivo haploid embryogenesis in the embryo sac, in vitro microspore/pollen embryogenesis, and in vitro meiocyte-derived callogenesis. The three of them exhibit clear differences, being the most important one the stage where the gamete or gamete precursor can be deviated from the original program. But most importantly, they present a number of common features, including the essential one in order to be considered as androgenic routes: an exclusively male-derived origin for the genome of the androgenic plant produced.

From an applied point of view, the main limitation for practical application in DH production of the three routes revised hereby relies on their low efficiency in terms of generation of androgenic DH plants. Efficiency has been largely improved, mostly in the case of microspore embryogenesis, thanks to the remarkable efforts of many research groups during the last 40 years. It is expected that in the next future, new advances and more sophisticated technology will lead to an even greater advance in the knowledge of this process and the increase of its efficiency. However, it has to be always kept in mind that efficiency in these kind of routes, involving the generation of haploid and DH individuals, will always be limited due to the unmasking of recessive lethal genes (de Fossard, 1974a), usually masked in heterozygous individuals, and unmasked and therefore abortive in haploid or recessive homozygous DH embryos.

From a more fundamental, biological point of view, these three routes share a number of features that point towards a possible common origin as an atavism carried through from the ancestors of the original plant species, and as a possible survival mechanism when the sexual way is blocked by any reason. In addition, in light of the current knowledge they might be triggered by similar mechanisms. All of this makes them not as different as they seem. Nevertheless, much more work is still needed to confirm many of these attractive hypothesis. Unfortunately, such work may not be possible to be developed in the case of in vivo androgenesis from the embryo sac, due to the difficulties imposed by its in vivo nature, within the embryo sac. This, together with the low efficiency, has been for sure one of the main drawbacks that has prevented an increased number of more profound studies of this process in the last 40 years.

Acknowledgements This work was supported by grant 'AGL2006-06678' from the Spanish Ministry of Education and Science.

DOI 10.1007/s12229-010-9056-6

Literature Cited

Aionesei, T., A. Touraev & E. Heberle-Bors. 2005. Pathways to microspore embryogenesis. Pp 11-34. In: C. E. Palmer, W. A. Keller, & K. J. Kasha (eds). Haploids in crop improvement II, vol.56. Springer-Verlag, Berlin Heidelberg.

Andersen, S. B. 2005. Haploids in the improvement of woody species. Pp 243-257. In: C. E. Palmer, W. A. Keller, & K. J. Kasha (eds). Haploids in crop improvement II, vol. 56. Springer-Verlag, Berlin Heidelberg.

Ari, E., S. Buyukalaca, K. Abak & S. Cetiner. 2007. Callus initiation for indirect pollen embryogenesis in Anemone coronaria. Proceedings of the 22nd International Eucarpia Symposium Section Ornamentals: Breeding for Beauty, Pt II 743: 87-90.

Bal, U. & K. Abak. 2007. Haploidy in tomato (Lycopersicon esculentum Mill.): A critical review. Euphytica 158: 1-9.

Baldursson, S., P. Krogstrup, J. V. Norgaard & S. B. Andersen. 1993. Microspore embryogenesis in anther culture of 3 species of Populus and regeneration of dihaploid plants of Populus trichocarpa. Canadian Journal of Forest Research-Revue Canadienne de Recherche Forestiere 23: 1821-1825.

Beckert, M. 1998. Genetic analysis of in vitro androgenetic response in maize. Pp 24-37. In: Y. Chupeau, M. Caboche, & Y. Henry (eds). Androgenesis and haploid plants. Springer-Verlag, Berlin, Heidelberg.

Belieuas, P. R., C. T. Guimaraes, L. V. Paiva, J. M. Dnarte, W. R. Maluf & E. Paiva. 2007. Androgenetic haploids and SSR markers as tools for the development of tropical maize hybrids. Euphytica 156: 95-102.

Belmonte, M. F., S. J. Ambrose, A. R. S. Ross, S. R. Abrams & C. Stasolla. 2006. Improved development of microspore-derived embryo cultures of Brassica napus cv Topas following changes in glutathione metabolism. Physiologia Plantarum 127: 690-700.

Binarova, P., G. Hause, V. Cenklova, J. H. G. Cordewener & M. M. van Lookeren-Campagne. 1997. A short severe heat shock is required to induce embryogenesis in late bicellular pollen of Brassica napus L. Sexual Plant Reproduction 10: 200-208.

Bohanec, B. 2009. Doubled haploids via gynogenesis. Pp 35-46. In: A. Touraev, B. P. Forster, & S. M. Jain (eds). Advances in haploid production in higher plants. Springer, Dordrecht.

Borderies, G., M. le Bechec, M. Rossignol, C. Lafitte, E. Le Deunff, M. Beckert, C. Dumas & E. Matthys-Roehon. 2004. Characterization of proteins secreted during maize microspore culture: Arabinogalactan proteins (AGPs) stimulate embryo development. European Journal of Cell Biology 83: 205-212.

Buuharmont, J. 1977. Cytology of microspores and calli after anther culture in Hordeum vulgare. Caryologia 30: 351-360.

Boutilier, K., M. Fiers, C. M. Liu & A. H. M. Van der Geest. 2005. Biochemical and molecular aspects of haploid embryogenesis. Pp 73 95. In: C. E. Palmer, W. A. Keller, & K. J. Kasha (eds). Haploids in crop improvement II, vol. 56. Springer-Verlag, Berlin Heidelberg.

Burk, L. G. 1962. Haploids in genetically marked progenies of tobacco. Journal of Heredity 53: 222-225.

Campus, F. F. & D. T. J. Morgan. 1958. Haploid pepper from a sperm. Journal of Heredity 49: 135-137.

Chase, S. S. 1963. Androgenesis--Its use for transfer of maize cytoplasm. Journal of Heredity 54: 152-158.

--. 1969. Monoploids and monoploid--derivatives of maize (Zea mays L.). Botanical Review 35: 117-167.

Chen, C. C., K. J. Kasha & A. Marsolais. 1984. Segmentation patterns and mechanisms of genome multiplication in cultured microspores of barley. Canadian Journal of Genetics and Cytology 26: 475-483.

Chen, L. J., X. Y. Zhu, L. Gu & B. Wu. 2005. Efficient callus induction and plant regeneration from anther of Chinese narcissus (Narcissus tazetta L. var. chinensis Roem). Plant Cell Reports 24: 401-407.

Chupeau, Y., M. Cabnche & Y. Henry. 1998. Androgenesis and haploid plants. Springer-Verlag, Berlin, Heidelberg.

Clausen, R. E. & W. E. Lammerts. 1929. Interspecific bybridisation in Nicotiana. X. Haploid and diploid merogony. The American Naturalist. 43: 279-282.

Cordewener, J. H. G., G. Hause, E. Gorgen, R. Busink, B. Hause, H. J. M. Dons, A. A. M. van Lammeren, M. M. van Lookeren-Campagne & P. Pechan. 1995. Changes in synthesis and localization of members of the 70-kDa class of heat-shock proteins accompany the induction of embryogenesis in Brassica napus L. microspores. Planta 196: 747-755.

Corduan, G. & C. Spix. 1975. Haploid callus and regeneration of plants from anthers of Digitalis purpurea L. Planta 124: 1-11.

Coronado, M. J., P. Gonzalez-Melendi, J. M. Segui, C. Ramirez, I. Barany, P. S. Testillano & M. C. Risueno. 2002. MAPKs entry into the nucleus at specific interchromatin domains in plant differentiation and proliferation processes. Journal of Structural Biology 140: 200-213.

Corral-Martinez, P., F. Nuez & J. M. Segui-Simarro. In press. In: D. Thangadurai (ed). Molecular agrobiology. Bentham Science Publishers, USA.

Dao, N. T. & Z. B. Shamina. 1978. Cultivation of isolated tomato anthers. Soviet Plant Physiology 25: 120-126.

de Fossard, R. A. 1974a. Methods for producing haploids. Pp 145-150. In: K. J. Kasha (ed). Haploids in higher plants: Advances and potential. University of Guelph, Guelph, Canada.

--. 1974b. Terminology in "Haploid" research. Pp 403-410. In: K. J. Kasha (ed). Haploids in higher plants: Advances and potential. University of Guelph, Guelph, Canada.

De Maine, M. J. 2003. Potato haploid technologies. Pp 241-247. In: M. Maluszynski, K. J. Kasha, B. P. Forster, & I. Szarejko (eds). Doubled haploid production in crop plants. A manual. Kluwer Academic Publishers, Dordretch, the Netherlands.

Devaux, P. 2003. The Hordeum bulbosum (L.) method. Pp 15-19. In: M. Maluszynski, K. J. Kasha, B. P. Forster, & I. Szarejko (eds). Doubled haploid production in crop plants. A manual. Kluwer Academic Publishers, Dordretch, the Netherlands.

Dumas de Vaulx, R. & D. Chambonnet. 1982. Culture in vitro d'antheres d'aubergine (Solanum melongena L.): Stimulation de la production de plantes au moyen de traitements a 35[degrees]C associes a de faibles teneurs en substances de croissance. Agronomie 2: 983-988.

Dunwell, J. M. 2010. Haploids in flowering plants: origins and exploitation. Plant Biotechnology Journal 8:377-424.

Ecochard, R., G. Merkx & M. Matteoli. 1974. Parthenogenesis induced by specific radioinactivation of the male gametes. Pp 136. In: K. J. Kasha (ed). Haploids in higher plants: Advances and potential. University of Guelph, Guelph, Canada.

Ehrensberger, R. 1948. Versuche zur auslosung von haploidie bei Blutenpflanzen. Biologisches Zentralblatt 67: 537-546.

Evans, M. 2005. New nucleic acid molecule encoding indeterminate gametophyte 1, useful for producing male sterility and for generating androgenetic progeny in plant species. Carnegie Inst Washington, USA, Patent no. US2005198711-A1.

Evans, M. M. S. 2007. The indeterminate gametophytel gene of maize encodes a LOB domain protein required for embryo sac and leaf development. Plant Cell 19: 46-62.

Evans, M. 2009. New isolated polypeptide or protein comprises the amino acid sequence of indeterminate gametophyte 1 (IG1) from Zea mays, useful for identifying and isolating an ortholog of ig1 in a non-Zea mays plant species. Carnegie Inst Washington, USA, Patent no. US2009151025-A1.

Evans, J. M. & N. P. Batty. 1994. Ethylene precursors and antagonists increase embryogenesis of Hordeum vulgate L anther culture. Plant Cell Reports 13: 676-678.

Ferrie, A. M. R., D. J. Epp & W. A. Keller. 1995. Evaluation of Brassica rapa L. genotypes for microspore culture response and identification of a highly embryogenic line. Plant Cell Reports 14: 580-584.

Forster, B. P., E. Heberle-Bnrs, K. J. Kasha & A. Touraev. 2007. The resurgence of haploids in higher plants. Trends in Plant Science 12: 368-375.

Gemes Juhasz, A. & M. Jakse. 2005. Haploids in the improvement of miscellaneous crop species (Cucurbitaceae, Liliaceae, Asparagaceae, Chenopodiaceae, Araceae and Umbelliferae). Pp 259-276. In: C. E. Palmer, W. A. Keller, & K. J. Kasha (eds). Haploids in crop improvement II. Springer-Verlag, Berlin Heidelberg.

Gerassimova, H. 1936. Experimentally produced haploid plant in Crepis tectorum. Biologitcheski Journal 5: 895-900.

Germana, M. 2006. Doubled haploid production in fruit crops. Plant Cell Tissue and Organ Culture 86: 131-146.

Gerrish, E. E. 1956. Studies od the monoploid method of producing homozigous diploids in Zea mays. Ph.D. Thesis. University of Minnesota, Minneapolis.

Gonzalez-Medina, M. & J. Bnuharmont. 1978. Experiments on anther culture in barley--Influence of culture methods on cell proliferation and organ differentiation. Euphytica 27: 553359.

Goodsell, S. F. 1961. Male sterility in corn by androgenesis. Crop Science 1: 227-228.

Gresshoff, P. M. & C. H. Doy. 1972a. Development and differentiation of haploid Lycopersicon esculentum (tomato). Planta 107: 161-170.

-- & --. 1972b. Haploid Arabidopsis thaliana callus and plants from anther culture. Australian Journal of Biological Sciences 25: 259.

-- & --. 1974. Derivation of a haploid cell line from Vitis vinifera and importance of stage of meiotic development of anthers for haploid culture of this and other genera. Zeitschrifi Fur Pflanzenphysiologie 73: 132-141.

Guha, S. & S. C. Maheshwari. 1964. In vitro production of embryos from anthers of Datura. Nature 204: 497.

Gulshan, T. M. V. & D. R. Sharma. 1981. Studies on anther cultures of tomato--Lycopersicon esculentum Mill. Biologia Ptantarum 23: 414-420.

Guo, F. L., B. Q. Huang, Y. Z. Han & S. Y. Zee. 2004. Fertilization in maize indeterminate gametophytel mutant. Protoplasma 223: 111-120.

Han, D, S., Y. Niimi & M. Nakann. 2000. Formation of calli from isolated microspore cultures of Asiatic hybrid lily 'Connecticut King'. Journal of the Japanese Society for Horticultural Science 69: 52-56.

Hayes, P., A. Corey & J. DeNoma. 2003. Doubled haploid production in barley using the Hordeum bulbosum (L.) technique. Pp 5-14. In: M. Maluszynski, K. J. Kasha, B. P. Forster, & I. Szarejko (eds). Doubled haploid production in crop plants. A manual. Kluwer Academic Publishers, Dordretch, the Netherlands.

Hays, D. B., R. W. Wilen, C. Sheng, M. M. Moloney & R. P. Pharis. 1999. Embryo-specific gene expression in microspore-derived embryos of Brassica napus. An interaction between abscisic acid and jasmonic acid. Plant Physiology 119: 1065-1072.

--, D. M. Reid, E. C. Yeung & R. P. Pharis. 2000. Role of ethylene in cotyledon development of microspore-derived embryos of Brassica napus. Journal of Experimental Botany 51 : 1851-1859.

--, R. M. Mandel & R. P. Pharis. 2001. Hormones in zygotic and microspore embryos of Brassica napus. Plant Growth Regulation 35: 47-58.

--, E. C. Yeung & R. P. Pharis. 2002. The role of gibberellins in embryo axis development. Journal of Experimental Botany 53: 1747-1751.

Holm, P. B., S. Knudsen, P. Mouritzen, D. Negri, F. L. Olsen & C. Roue. 1994. Regeneration of fertile barley plants from mechanically isolated protoplasts of the fertilized egg cell Plant Cell 6: 531-543.

Honys, D. & D. Twell. 2004. Transcriptome analysis of haploid male gametophyte development in Arabidopsis. Genome Biology 5: R85.

Hosp, J., S. D. E Maraschin, A. Touraev & K. Boutilier. 2007a. Functional genomics of microspore embryogenesis. Euphytica 158: 275-285.

--, A. Tashpulatov, U. Roessner, E. Barsova, H. Katholnigg, R. Steinborn, B. Melikant, S. Lukyanov, E. Heberle-Bors & A. Touraev. 2007b. Transcriptional and metabolic profiles of stress-induced, embryogenic tobacco microspores. Plant Molecular Biology 63: 137-149.

Huang, B. Q. & W. E Sheridan. 1996. Embryo sac development in the maize indeterminate gametophytel mutant: Abnormal nuclear behavior and defective microtubule organization. Plant Cell 8: 1391-1407.

Illies, Z. M. 1974. Induction of haploid parthenogenesis in Populus tremula by male gametes incativated with toluidine blue. Pp 136. In: K. J. Kasha (ed). Haploids in higher plants: Advances and potential. University of Guelph, Guelph, Canada.

Ivanov, M. A. 1938. Experimental production of haploids in Nicotiana rustica L. Genetica 20: 295-386.

Janick, J. & H. G. Hughes. 1974. Production of strawberry tetrahaploids from intergeneric crosses. Pp 137. In: K. J. Kasha (ed). Haploids in higher plants: Advances and potential. University of Guelph, Guelph, Canada.

Joosen, R., J. Cordewener, E. D. J. Supena, O. Vorst, M. Lammers, C. Maliepaard, T. Zeilmaker, B. Miki, T. America, J. Custers & K. Boutilier. 2007. Combined transcriptome and proteome analysis identifies pathways and markers associated with the establishment of rapeseed microspore-derived embryo development. Plant Physiology 144: 155-172.

Kasha, K. J. & K. N. Kao. 1970. High frequency haploid production in barley (Hordeum vulgate L.). Nature 225: 874-876.

Kehr, A. E. 1951. Monoploidy in Nicotiana. Journal of Heredity 42: 107-112.

Kermicle, J. L. 1969. Androgenesis conditioned by a mutation in maize. Science 166: 1422-1424.

--. 1971. Pleiotropic effects on seed development of the indeterminate gametophyte gene in maize. American Journal of Botany 58: 1-7.

--. 1974. Origin of androgenetic haploids and diploids induced by the indeterminate gametophyte (ig) mutation in maize. Pp 137. In: K. J. Kasha (ed). Haploids in higher plants: Advances and potential. University of Guelph, Guelph, Canada.

--. 1994. Indeterminate gametophyte (ig): Biology and use. Pp 388-393. In: M. Freeling & V. Walbot (eds). The maize handbook. Springer-Verlag, New York.

Kim, M., J. Kim, M. Yoon, D.-I. Choi & K.-M. Lee. 2004. Origin of multicellular pollen and pollen embryos in cultured anthers of pepper (Capsicum annuum). Plant Cell, Tissue and Organ Culture 77: 63 72.

--, L-C. Jang, J.-A. Kim, E.-J. Park, M. Yoon & Y. Lee. 2008. Embryogenesis and plant regeneration of hot pepper (Capsicum annuum L.) through isolated microspore culture. Plant Cell Reports 27: 425-434.

Kindiger, B. & S. Hamann. 1993. Generation of haploids in maize A modification of the indeterminate gametophyte (ig) system. Crop Science 33: 342-344.

Kiviharju, E., M. Puolimatka, M. Saastamoinen, S. Hovinen & E. Pehu. 1998. The effect of genotype on anther culture response of cultivated and wild oats. Agricultural and Food Science in Finland 7: 409-422.

Komen, H. & G. H. Thorgaard. 2007. Androgenesis, gynogenesis and the production of clones in fishes: A review. Aquaculture 269: 150-173.

Komma, D. J. & S. A. Endow. 1995. Haploidy and androgenesis in Drosophila. Proceedings of the National Academy of Sciences of the United States of America 92: 11884-11888.

Kostoff, D. 1929. An androgenic Nicotiana haploid. Zeitschrift fur Zellforschung 9: 640-642.

--. 1942. The problem of haploidy (Cytogenetic studies in Nicotiana haploids and their bearing on some other cytogenetic problems). Bibliographia Genetica 13: 1-148.

Koul, A. K. & J. L. Karihaioo. 1977. In vivo embryoids from anthers of Narcissus biflorus Curt. Euphytica 26: 97-102.

Kumlehn, J., H. Lorz & E. Kranz. 1998. Differentiation of isolated wheat zygotes into embryos and normal plants. Planta 205: 327-333.

Laeadena, J. R. 1974. Spontaneous and induced parthenogenesis and androgenesis. Pp 13-32. In: K. J. Kasha (ed). Haploids in higher plants: Advances and potential. University of Guelph, Guelph, Canada.

Letarte, J., E. Simion, M. Miner & K. J. Kasha. 2006. Arabinogalactans and arabinogalactan-proteins induce embryogenesis in wheat (Triticum aestivum L.) microspore culture. Plant Cell Reports 24: 691-698.

Levenko, B. A., V. A. Kunakh & G. N. Yurkova. 1977. Studies on callus tissue from anthers.l. Tomato. Phytomorphology 27: 377-383.

Li, J. Q., y. Q. Wang, L. H. Lin, L. J. Zhou, N. Luo, Q. X. Deng, J. R. Xian, C. X. Hou & Y. Qiu. 2008 Embryogenesis and plant regeneration from anther culture in loquat (Eriobotrya japonica L.). Scientia Horticulturae 115: 329-336.

Lin, B. Y. 1978. Structural modifications of female gametophyte associated with indeterminate-gametophyte (ig) mutant in maize. Canadian Journal of Genetics and Cytology 20: 249-257.

--. 1981. Megagametogenetic alterations associated with the indeterminate gametophyte (ig) mutation in maize. Revista Brasileira de Biologia 41: 557-563.

--. 1984. Ploidy barrier to endosperm development in maize. Genetics 107: 103-115.

Liu, J., X. Xu & X. Deng. 2005. Intergeneric somatic hybridization and its application to crop genetic improvement. Plant Cell, Tissue and Organ Culture 82: 19-44.

Ma, R., Y. D. Guu & S. Pulli. 2004. Comparison of anther and microspore culture in the embryogenesis and regeneration of rye (Secale cereale). Plant Cell Tissue and Organ Culture 76: 147-157.

Magnard, J. L., E. Le Deunff, J. Domeneeh, P. M. Rogowsky, P. S. Testillano, M. Rougier, M. C. Risueno, P. Vergne & C. Dumas. 2000. Genes normally expressed in the endosperm are expressed at early stages of microspore embryogenesis in maize. Plant Molecular Biology 44: 559-574.

Malik, M. R., F. Wang, J. M. Dirpaul, N. Zhou, P. L. Poiowick, A. M. R. Ferrie & J. E. Krochko. 2007. Transcript profiling and identification of molecular markers for early microspore embryogenesis in Brassica napus. Plant Physiology 144: 134-154.

--, --, J. Dirpaul, N. Zhou, J. Hammerlindl, W. Keller, S. R. Abrams, A. M. R. Ferrie & J. E. Krochko. 2008. Isolation of an embryogenic line from non-embryogenic Brassica napus cv. Westar through microspore embryogenesis Journal of Experimental Botany 59: 2857-2873

Mallikarjuna, N., D. Jadhav, H. Clarke, C. Coyne & F. Muehlbauer. 2005. Induction of androgenesis as a consequence of wide crossing in chickpea, international Chickpea and Pigeonpea Newsletter 12: 12-15.

Maluszynski, M., K. J. Kasha, B. P. Forster & I. Szarejko. 2003a. Doubled haploid production in crop plants. A manual. Kluwer Academic Publishers, Dordrecht.

--, -- & I. Szarejko. 2003b. Published doubled haploid protocols in plant species. Pp 309-335. In: M. Maluszynski, K. J. Kasha, B. P. Forster, & I. Szarejko (eds). Doubled haploid production in crop plants. A manual. Kluwer Academic Publishers, Dordretch, the Netherlands.

Mantovani, B. & V. Scali. 1992. Hybridogenesis and androgenesis in the stick insect Bacillus rossius grandii benazzi (Insecta, Phasmatoidea). Evolution 46: 783-796.

Maraschin, S. F., W. de Priester, H. P. Spaink & M. Wang. 2005. Androgenic switch: an example of plant embryogenesis from the male gametophyte perspective. Journal of Experimental Botany 56: 1711-1726.

Marasehin, S. D. F., M. Caspers, E. Potokina, F. Wulfert, A. Graner, H. P. Spaink & M. Wang. 2006. CDNA array analysis of stress-induced gene expression in barley androgenesis. Physiologia Plantarum 127: 535-550.

Massonneau, A., M. J. Coronado, A. Audran, A. Bagniewska, R. Mol, P. S. Testillano, G. Goralski, C. Dumas, M. C. Risueno & E. Matthys-Roehon. 2005. Multicellular structures developing during maize microspore culture express endosperm and embryo-specific genes and show different embryogenic potentialities. European Journal of Cell Biology 84: 663-675.

McKone, M. J. & S. L. Halpern. 2003. The evolution of androgenesis. American Naturalist 161: 641-656.

Miyoshi, K. 1996. Callus induction and plantlet formation through culture of isolated microspores of eggplant (Solanum melongena L). Plant Cell Reports 15: 391-395.

Munoz-Amatriain, M., J. T. Svensson, A. M. Castillo, L. Cistue, T. J. Close & M. P. Valles. 2006. Transcriptome analysis of barley anthers: Effect of mannitol treatment on microspore embryogenesis Physiologia Plantarum 127: 551-560.

Myint, A. & R. A. de Fossard. 1974. Induction of haploid callus from rice anthers and regeneration of plants. Pp 139. In: K. J. Kasha (ed). Haploids in higher plants: Advances and potential. University of Guelph, Guelph, Canada.

Naess, S. K., H. J. Swartz & G. R. Bauchan. 1998. Ploidy reduction in blackberry. Euphytica 99: 57-73.

Nontaswatsri, C., S. Ruamrungsri & S. Fukai. 2008. Callus induction and plant regeneration of Dianthus chinensis L. and Dianthus barbatus L. via Anther Culture. Proceedings of the International Workshop on Ornamental Plants 788: 109-114.

Paire, A., P. Devaux, C. Lafitte, C. Dumas & E. Matthys-Rochon. 2003. Proteins produced by barley microspores and their derived androgenic structures promote in vitro zygotic maize embryo formation. Plant Cell Tissue and Organ Culture 73: 167-176.

Palmer, C. E., W. A. Keller & K. J. Kasha. 2005. Haploids in crop improvement II. Springer-Verlag, Berlin.

Pandey, K. K. 1973. Theory and practice of induced androgenesis. New Phytologist 72: 1129-1140.

Pauls, K. P., J. Chart, G. Woronuk, D. Schulze & J. Brazolot. 2006. When microspores decide to become embryos--cellular and molecular changes. Canadian Journal of Botany-Revue Canadienne De Botanique 84: 668-678.

Picard, E., J. De Buyser & J. Bozza. 1974. Production of haploid plants by anther culture of wheat. Pp 143. In: K. J. Kasha (ed). Haploids in higher plants: Advances and potential. University of Guelph, Guelph, Canada.

Piccirilli, M. & S. Arcioni. 1991. Haploid plants regenerated via anther culture in wild barley (Hordeum spontaneum c-kock). Plant Cell Reports 10: 273-276.

Pulido, A., A. Hernando, F. Bakos, E. Mendez, M. Devie, B. Barnabas & A. Olmedilla. 2006. Hordeins are expressed in microspore-derived embryos and also during male gametophytic and very early stages of seed development. Journal of Experimental Botany 57: 2837-2846.

Raghavan, V. 1986. Polen embryogenesis. Pp 152-189. In: P. W. Barlow, P. V. Green, & C. C. Wylie (eds). Embryogenesis in angiosperms, a developmental and experimental study. Cambridge University Press, London.

Ramanna, M. S. & J. G. T. Hermsen. 1974. Embryoid formation in anthers of some interspecific hybrids in Solanurn. Euphytica 23: 423-427.

Ramesar-Fortner, N. S. & E. C. Yeung. 2006. Physiological influences in the development and function of the shoot apical meristem of microspore-derived embryos of Brassica napus 'Topas'. Canadian Journal of Botany-Revue Canadienne De Botanique 84: 371-383.

Razdan, M. K. 2003. Introduction to plant tissue culture. Science Publishers, Inc., Enfield, New Hampshire, USA.

Reynolds, T. L. 1997. Pollen embryogenesis. Plant Molecular Biology 33: 1-10.

-- & R. L. Crawford. 1996. Changes in abundance of an abscisic acid-responsive, early cysteine-labeled metallothionein transcript during pollen embryogenesis in bread wheat (Triticum aestivum). Plant Molecular Biology 32: 823-829.

Rieger, R. A., A. Michaelis & M. M. Green. 1968. A glossary of genetics and cytogenetics. Springer-Verlag, New York.

Rudolf, K., B. Bohanec & M. Hansen. 1999. Microspore culture of white cabbage, Brassica oleracea var. capitata L.: Genetic improvement of non-responsive cultivars and effect of genome doubling agents. Plant Breeding 118: 237-241.

Seaney, R. R. 1955. Studies on monoploidy in maize. Ph.D. Thesis. Cornell University, Ithaca, New York.

Segui-Simarro, J. M. & F. Nuez. 2005. Meiotic metaphase I to telophase II is the most responsive stage of microspore development for induction of androgenesis in tomato (Solanum Lycopersicum). Acta Physiologiae Plantarum 27: 675-685.

-- & --. 2006. Androgenesis induction from tomato anther cultures: Callus characterization. Pp 855-861. In: M. G. Fari, I. Holb, & G. D. Bisztray (eds). Acta Horticulturae, no. 725. ISHS, Debrecen, Hungary.

-- & --. 2007. Embryogenesis induction, callogenesis, and plant regeneration by in vitro culture of tomato isolated microspores and whole anthers. Journal of Experimental Botany 58: 1119-1132.

-- & --. 2008a. How microspores transform into haploid embryos: changes associated with embryogenesis induction and microspore-derived embryogenesis. Physiologia Plantarnm 134: 1-12.

-- & --. 2008b. Pathways to doubled haploidy: chromosome doubling during androgenesis. Cytogenetic and Genome Research 120: 358-369.

--, P. S. Testillano & M. C. Risueno. 2003. Hsp70 and Hsp90 change their expression and subcellular localization after microspore embryogenesis induction in Brassica napus L. cv Topas. Journal of Structural Biology 142: 379-391.

--, -- & --. 2005. MAP kinases are developmentally regulated during stress-induced microspore embryogenesis in Brassica napus L. Histochemistry and Cell Biology 123: 541-551.

Shariatpanahi, M. E., U. Bal, E. Heberle-Bors & A. Touraev. 2006. Stresses applied for the reprogramming of plant microspores towards in vitro embryogenesis. Physiologia Plantarum 127: 519-534.

Shivanna, K. R. & B. M. Johri. 1985. The angiosperm pollen. Structure and function. Wiley Eastern Limited, New Delhi, India.

Shtereva, L. A., N. A. Zagorska, B. D. Dimitrov, M. M. Kruleva & It. K. Oanh. 1998. Induced androgenesis in tomato (Lycopersicon esculentum Mill). II. Factors affecting induction of androgenesis. Plant Cell Reports 18: 312-317.

Silva, A. S., J. M. Q. Luz, T. M. Rodrigues, S. V. Marques, R. V. Marques & M. Pasqual. 2009. Coffea arabica L. anther callus and pro-embryoid induction by different growth regulators. Bioscience Journal 25: 19-27.

Simmonds, D. H. & W. A. Keller. 1999. Significance of preprophase bands of microtubules in the induction of microspore embryogenesis of Brassica napus. Planta 208: 383-391.

Smykal, P. 2000. Pollen emhryogenesis--the stress mediated switch from gametophytic to sporophytic development. Current status and future prospects. Biologia Plantamm. 43: 481-489.

Song, H., Q. F. Lou, X. D. Luo, J. N. Wolukau, W. P. Diao, C. T. Qian & J. F. Chen. 2007. Regeneration of doubled haploid plants by androgenesis of cucumber (Cucumis sativus L.). Plant Cell Tissue and Organ Culture 90: 245-254.

Srivastava, P. & R. Chaturvedi. 2008. In vitro androgenesis in tree species: An update and prospect for further research. Biotechnology Advances 26: 482-491.

Stasolla, C., M. F. Belmonte, T. Muhammad, E. Mohamed, K. Khalil, R. Joosen, C. Maliepaard, A. G. Sharpe, B. Gjetvaj & K. Boutilier. 2008. Buthionine sulfoximine (BSO)-mediated improvement in cultured embryo quality in vitro entails changes in ascorbate metabolism, meristem development and embryo maturation. Planta 228: 255-272.

Summers, W. L., J. Jaramillo & T. Bailey. 1992. Microspore developmental stage and anther length influence the induction of tomato anther callus. Hortscience 27: 838-840

Supeoa, E. D. J., W. Muswita, S. Suharsono & J. B. M. Custers. 2006a. Evaluation of crucial factors for implementing shed-microspore culture of Indonesian hot pepper (Capsicum annuum L.) cultivars. Scientia Horticultarae 107: 226-232.

--, S. Suharsouo, E. Jacobsen & J. B. M. Custers. 2006b. Successful development of a shedmicrospore culture protocol for doubled haploid production in Indonesian hot pepper (Capsicum annuum L.). Plant Cell Reports 25: 1-10.

--, B. Winarto, T. Rikseo, E. Dubas, A. van Lammeren, R. Offringa, K. Boutilier & J. Custers. 2008. Regeneration of zygotic-like microspore-derived embryos suggests an important role for the suspensor in early embryo patterning. Journal of Experimental Botany 59: 803-814.

Tai, G. C. C. 2005. Haploids in the improvement of solanaceous species. Pp 173-190. In: C. E. Palmer, W. A. Keller, & K. J. Kasha (eds). Haploids in crop improvement II, vol. 56. Springer-Verlag, Berlin Heidelberg.

Tang, X. C., Y. Q. He, Y. Wang & M. X. Sun. 2006. The role of arabinogalactan proteins binding to Yariv reagents in the initiation, cell developmental fate, and maintenance of microspore embryogenesis in Brassica napus L. cv. Topas. Journal of Experimental Botany 57: 2639-2650.

Touraev, A., M. Pfosser & E. Heberle-Bors. 2001. The microspore: A haploid multipurpose cell. Advances in Botanical Research 35: 53-109.

--, B. P. Forster & S. M. Jain. 2009. Advances in haploid production in higher plants. Springer, New York.

Tsuwamoto, R., It. Fukuoka & Y. Takahata. 2007. Identification and characterization of genes expressed in early embryogenesis from microspores of Brassica napus. Planta 225: 641-652.

Tuleeke, W. 1965. Haploid versus Diploidy in the reproduction of cell type. Pp 217-241. In: M. Locke (ed). Reproduction: Molecular, subcellular and cellular. Academic Press, New York.

Twell, D. & R. Howden. 1998. Mechanisms of asymmetric division and cell fate determination in developing pollen. Pp 69-103. In: Y. Chupeau, M. Caboche, & Y. Henry (eds). Androgenesis and haploid plants. INRA-Springer-Verlag, Berlin Heidelberg.

van Bergen, S., M. J. Kottenhagen, R. M. van der Meulen & M. Wang. 1999. The role of abscisic acid in induction of androgenesis: A comparative study between Hordeum vulgare L-cvs Igri and Digger. Journal of Plant Growth Regulation 18: 135-143.

Varghese, T. M. & Y. Gulshan. 1986. Production of embryoids and calli from isolated microspores of tomato (Lycopersicon esculentum Mill.) in liquid media. Biologia Plantarum 28: 126-129.

Vasil, I. K. & T. A. Thorpe. 1994. Plant cell and tissue culture. Kluwer Academic Publishers, Dordretch, The Netherlands.

Wang, M., S. van Bergen & B. Van Duijn. 2000. Insights into a key developmental switch and its importance for efficient plant breeding. Plant Physiology 124: 523-530.

Wedzony, M., B. P. Forster, I. Zur, E. Golemiec, M. Szechynska-Hebda, E. Dubas & G. Gotebiowska. 2009. Progress in doubled haploid technology in higher plants. Pp 1-33. In: A. Touraev, B. P. Forster, & S. M. Jain (eds). Advances in haploid production in higher plants. Springer, Dordrecht.

Yang, J. S., M. B. Endo & L Inada. 2005. Anther and microspore culture of chrysanthemum (Dendranthema grandiflorum (Ramat.) Kitam.). Journal of the Japanese Society for Horticultural Science 74: 78-86.

Zagorska, N. A., A. Shtereva, B. D. Dimitrov & M. M. Kruleva. 1998. Induced androgenesis in tomato (Lycopersicon esculentum Mill.)--I. Influence of genotype on androgenetic ability. Plant Cell Reports 17: 968-973.

--, L. A. Shtereva, M. M. Kruleva, V. G. Sotirova, D. L. Baralieva & B. D. Dimitrov. 2004. Induced androgenesis in tomato (Lycopersicon esculentum Mill.). III. Characterization of the regenerants. Plant Cell Reports 22: 449-456.

Zaki, M. A. M. & H. G. Dickinson. 1991. Microspore-derived embryos in Brassica: The significance of division symmetry in pollen mitosis I to embryogenic development. Sexual Plant Reproduction 4: 48-55.

Zamir, D., R. A. Jones & N. Kedar. 1980. Anther culture of male sterile tomato (Lycopersicon esculentum Mill.) mutants. Plant Science Letters 17: 353-361.

Zarsky, V., D. Garrido, N. Eller, J. Tupy, O. Vicente, F. Schiffl & E. Heberle-Bors. 1995 The expression of a small heat shock gene is activated during induction of tobacco pollen embryogenesis by starvation. Plant Cell and Environment 18: 139-147.

Zenkteler, M. & W. Nitzsche. 1984. Wide hybridization experiments in cereals. Theoretical and Applied Genetics 68: 311-315.

Zhao, F. C., D. Nilanthi, Y. S. Yang & H. Wu. 2006. Anther culture and haploid plant regeneration in purple coneflower (Echinacea purpurea L.). Plant Cell Tissue and Organ Culture 86: 55-62.

Published online: 13 May 2010

Jose M. Segui-Simarro (1,2)

(1) Instituto para la Conservacion y Mejora de la Agrodiversidad Valenciana (COMAV), Universidad Politecnica de Valencia, Ciudad Politecnica de la Innovacion (CPI), Edificio 8E--Escalera I. Camino de vera, s/n, 46022 Valencia, Spain

(2) Author for Correspondence; e-mail: seguisim@btc.upv.es
Table 1 Differences Between the Three Different Androgenic Routes

             Route 1: male-derived        Route 2: microspore
             haploid embryogenesis from   embryogenesis
             a fertilized egg

Program      Zygotic embryogenesis        microsporogenesis
deviated

Program      Haploid embryogenesis        Haploid embryogenesis/
triggered                                 callogenesis

Sensitive    One-celled zygote            Vacuolate microspore-
stages                                    young pollen

Occurrence   Natural                      Induced in vitro

Genotype     High                         High
dependence

Need for     No                           Essential
inductive
stress

Need for     No                           Relative
hormones

Applied      Null                         High
relevance

Documented   Few                          Abundant
examples

             Route 3: meiocyte-derived
             callogenesis

Program
deviated

Program      Callogenesis
triggered

Sensitive    Meiocyte from metaphase I
stages       to telophase II

Occurrence

Genotype     High
dependence

Need for     Relative
inductive
stress

Need for     Essential
hormones

Applied      Low
relevance

Documented   Few
examples
Gale Copyright: Copyright 2010 Gale, Cengage Learning. All rights reserved.