RepRoductive studies on ipecAc (Cephaelis ipecacuanha (BRot.) A. Rich; RuBiAceAe): meiotic BehAvioR And pollen viABility

Reproductive studies were carried out on Brazilian accessions of ipecac, Cephaelis ipecacuanha. meiotic behavior was studied using the squashing technique. irregular chromosome segregation in meiosis i and ii, many sets of chromosomes in telophase ii, micronuclei, incorrect cytoplasm division, incomplete cytokinesis and anomalous post-meiotic products, mainly polyads, were observed. the mean meiotic index was lower than 72%. pollen viability was analyzed using Alexander solution, and the percentages ranged between brevistylous and longistylous floral morphs (85.3 to 93.1%), and among different localities (82.5 to 92.6%) analyzed. the size of pollen ranged between viable and sterile, and empty and shrunken sterile. in its natural habitat, this species is known to propagate by vegetative multiplication, but sexual reproduction seems to be as important as the vegetative propagation to this species.

According to Akerele et al. (1991), C. ipecacuanha is considered one of the world's most important medicinal plants and Brazilian ipecac is considered the most valuable because it shows the highest emetine content (Assis, 1992). however, like other Brazilian native species that are disappearing due to the reduction of their habitats in areas of natural vegetation, ipecac is now considered a threatened species, (oliveira & martins, 1998). many wild species have undergone a narrowing of their genetic base due to predatory collection or destruction of vegetation cover (mcKeown, 1996), possibly endangering them through the action of natural selection (Brown, 1988).
vegetative propagation is considered the main mode of reproduction for C. ipecacuanha (sick, 1993), but sexual reproduction which depends on meiosis to produce gametes usually plays an important role in heterostylous species. sexual reproduction involves two important events that lead to genetic diversity: fusion of gametes with different gene information and exchange between genetic materials (crossing over). meiosis is a highly dynamic process and the main cell mechanism enabling the occurrence of sexual reproduction in angiosperms. the complex orchestration of a normal cell division includes chromosome pairing in zygotene, genetic exchange in pachytene, chiasmata formation in diplotene and chromosome segregation in anaphase i and ii (Franklin et al., 1999). meiosis is genetically controlled (Gottschalk & Kaul, 1974;Golubovskaya, 1979). Although the meiocyte is a highly specialized cell capable of producing four haploid cells, mutations, hybridizations, environmental stress, endogamy and other factors may alter the constitution or the expression of genes that act during meiosis (utsunomiya et al., 2002). more than 20 genes that affect meiosis have been described for maize, most of which are known to cause sterility (Golubovskaya, 1989(Golubovskaya, , quoted by defani-scoarize et al., 1995. micropropagation has served as a tool for rescuing endangered plants (palomino et al., 1999). studies on C. ipecacuanha to obtain whole plants in vitro have been successful (Jha & Jha, 1989;ideda et al., 1989), but in situ germplasm preservation can also be an important strategy to prevent narrowing of the genetic base. however, studies to ascertain the importance of sexual reproduction of this species have not been conducted. therefore, this study aimed to document and report valuable information on meiotic behavior and pollen viability, thereby contributing toward a better understanding of the reproductive aspects and the importance of sexual reproduction of this species, in view of its considerable medicinal importance.

mAterIAl And methods
Brazilian accessions of C. ipecacuanha collected in the municipalities of caratinga and carangolas (minas Gerais) and itaperuna (Rio de Janeiro) were analyzed. the species were kept in pots in a greenhouse in the Research support unit at uenF.
For meiotic studies, flower buds were fixed in ethanol -acetic acid (3:1) at room temperature for 24 h, transferred to 70% alcohol and stored under refrigeration until use. temporary slides were prepared by the squashing technique and the cells were stained with 1% acetic carmine. Although the flowers are very small and difficult to analyze, at least 25 cells were examined at each meiotic phase. the numbers of monads, dyads, triads, tetrads and polyads were recorded for calculation of the meiotic index (% im = [number of normal tetrads x 100] ÷ total of post-meiotic products counted), according to love (1951). the meiotic index considered in this study was the mean of four randomly sampled slides.
to ascertain viability, only the heterostyly of plants collected in itaperuna and caratinga was taken into account. Fixed pollen grains of flowers in anthesis were stained with Alexander solution (Alexander, 1969) for wall/cytoplasm reactivity, and were considered viable when their cytoplasm became stained and remained intact. the anthers were squashed in a drop of stain and, after two minutes at room temperature, the pollen grains were checked, counted and measured with the aid of an eyepiece micrometer. mechanically damaged grains were easily distinguishable by their normal size and traces of cytoplasm inside and outside the pollen wall. the number of viable and sterile pollen grains considered in this study was the mean of the count of these cells on five randomly sampled slides (repetitions) for both brevistylous and longistylous morph and collection site.
An optical microscope was used for the observations and photographs. the statistical analysis was performed using the Genes program (cruz, 2001).

Meiosis
due to the small size of buds and chromosomes, the prophase i stages could not be clearly analyzed, so an accurate cytological analysis was impossible. A conventional cytological analysis was employed to verify the meiotic behavior, confirming that chromosomes of ipecac paired as 11 bivalents in metaphase i. several meiotic irregularities were observed, whose type and frequency are presented in table 1.
chromosomes showing irregular chromosome segregation (ascending and laggard; Fig. 1a-c) were the main problems found in meiosis i. meiosis ii showed the same type of irregularity, but telophase ii displayed different numbers of chromosome sets (Fig. 1d-f) due to non-oriented bivalents on the equatorial plate. this fact was corroborated by the formation of abnormal post-meiotic products, mainly polyads, which showed irregularly shared cytoplasm and incomplete cytokinesis (Fig. 1g-h). the imbalanced distribution of chromosomes in anaphase ii persisted up to telophase ii. linear tetrads were observed. As indicated by the data in table 2, the meiotic index mean was 71.26%, with the quantity of polyads on average 72% greater than other irregularities. micronuclei were observed (Fig. 1i) as a consequence of irregular chromosome segregation, and incomplete cytokinesis was also found in mature pollen grains (Fig. 2a).

Pollen size and viability
pollen viability was found at three sites, ranging on average from 82.5 to 92.6% (table 3). pollen from two sites was analyzed in terms of heterostyly, with longistylous populations showing lower pollen viability than brevistylous populations (table 4). Application of the F test (table 5) revealed that: 1) the three sites analyzed showed significant differences in the percentage of pollen viability (p < 0.01); 2) caratinga and itaperuna showed no significant difference in the percentage of sterile pollen for the same floral morph; and 3) brevistylous and longistylous morphs from both caratinga and itaperuna showed significant differences with respect to their percentage of sterile pollen (p < 0.01). Although pollen from the carangolas (mG) population was not analyzed in terms of heterostyly, its values were similar to those of the longistylous populations.
two types of sterile pollen grains were observed: a) empty (Fig. 2b), measuring from 8 to 40 µm (27.92 ± 7.68 µm), with only the wall reacting to the stain and indicating absence or only traces of cytoplasm; and b) shrunken (Fig. 2c-d), measuring from 24 to 56 µm (38.95 ± 8.46 µm), with both the wall and the cytoplasm reacting to the stain, although a space was found between the plasmatic membrane and the cell wall, accompanied by cytoplasm contraction. the presence of micronuclei led to the formation of microcytes and hence, anomalous pollen grains (Fig. 2e). the size of the strongly stained pollen grains varied from 36 to 80 µm (55.56 + 8.71 µm), and some gametes       were considered giant cells (Fig. 2f), being up to 43.98% larger than the mean viable pollen grain size. Fig. 3 shows the sterile pollen grain frequencies in relation to their size.

dIscussIon
C. ipecacuanha is a diploid species with 11 bivalents. the basic number of the genus Cephaelis (= Psychotria l.) is x = 11 (Kiehn, 1986). According to Assis (1992), the karyotype of C. ipecacuanha is symmetric, displaying metacentric chromosomes, and is included among the more primitive types. however, a meiotic analysis revealed abnormal behavior. irregular chromosome segregation in meiosis i and ii could be the result of the non-oriented chromosomes, which show an inability to congregate on the equatorial plate, resulting in the grouping of many sets of chromosomes in telophase ii, as well as micronuclei, incorrect cytoplasm division, incomplete cytokinesis and polyads.
According to nicklas & Ward (1994), non-oriented bivalents may be related to impaired attachment of kinetochores to the spindle fibers. pagliarini (1990) reported that laggards may result from late chiasma terminalization. Ascending chromosomes are the result of precocious migration and, according to utsunomiya et al. (2002), generally consist of univalent chromosomes formed during late prophase i stages by precocious chiasma terminalization in early metaphase i, or may even result from low chiasma frequency or from the presence of asynaptic or desynaptic genes (pagliarini, 2000). laggards and non-oriented chromosomes may produce micronuclei if they fail to reach the poles in time to be included in the main telophase nucleus (Koduru & Rao, 1981;utsunomiya et al., 2002), leading to the formation of micro-pollen and, probably, to gametes with unbalanced chromosome numbers (mansuelli et al., 1995), such as aneuploids (defani-scoarize et al., 1995). despite meiotic irregularities, Kenton et al. (1988) stated that a period of chromosome instability can lead to the establishment of a new equilibrium and may represent an important source of variability in species with vegetative reproduction.
Abnormal cytokinesis was observed in C. ipecacuanha, resulting in the formation of polyads. two types of microsporogenesis are generally recognized, i.e., successive or simultaneous. in the same anther of C. ipecacuanha, we found linear and t-shaped tetrads that are characteristic of successive micropsporogenesis, and tetrahedral tetrads that are the result of simultaneous microsporogenesis (see Furness & Rudall, 1999). however, an intermediate type called "modified simultaneous" may have occurred in C. ipecacuanha, whereby an ephemeral cell plate is laid down after the first meiotic division, which then disappears and cytokinesis occurs simultaneously; sometimes the second division rapidly follows the first, before the cell plate is completely formed (murty, 1964;Furnes & Rudall, 1999). this may explain why some meiocytes presented cells with partial cytokinesis. in mutants, some genes have been identified as responsible for abnormal cytokinesis. in maize, the va (variable sterile; va1 and va2 alleles; Beadle, 1932) and el (Rhoades & dempsey, 1966) genes were responsible for the absence or partial formation of the cell plate, forming irregular post-meiotic products. not only genetic factors but also several environmental factors can account for the frequency of anomalous post-meiotic products due to differences in the genetic background, as in the case of T. latifolia (Berdnikov et al., 2002).
Giant cells were observed in C. ipecacuanha. these are imbalanced gametes probably formed by the absence of cytokinesis and/or irregular chromosome segregation. these imbalanced gametes exert a strong influence on sexual propagation, so that chromosomal instability can produce gametes with the addition or loss of chromosomes. sometimes, such abnormal gametes are tolerant and participate in fertilization, but the union of these gametes can produce considerable variations in chromosome numbers in subsequent generations (caetano-pereira et al., 1998). in maize, imbalanced gametes are unable to compete with normal gametes (defani-scoarize et al., 1995). in Typha latifolia, nullisomy was lethal to a developing gametophyte while disomy was tolerable, resulting in non-disjunction in the second division (Berdnikov et al., 2002).
C. ipecacuanha presented different pollen viability percentages in relation to floral morph and population locality. Assis (1992) studied C. ipecacuanha populations from the states of minas Gerais (mG), espírito santo (es), mato Grosso (mt) and Rondonia (Ro) Brazilian without taking into consideration the heterostyly. only the mG population showed a very low pollen viability of 47.74%, while plants in es showed 81.46% and indexes exceeding 91% were found in mG. the latter findings are congruent with those found in this study. the populations under study showed higher pollen viability when the floral morph was brevistylous, i.e., above 90%, while longistylous presented values between 85.2 to 87%. carangolas (mG) presented 82.58% pollen viability, and was probably a longistylous population. the meiotic index was 71.26%, probably indicating that the many types of irregularities found during meiosis affected the percentage of sterile pollen. According to love (1951), plants with a meiotic index of 90% to 100% may be considered quite cytologically stable, but it is actually impossible to ensure stability in plants with indices of 88% to 92%, since plants with a meiotic index of less than 90% are likely to encounter difficulties in outbreeding.
two types of sterile pollen were observed: empty and shrunken. C. ipecacuanha showed smaller empty pollen grains, indicating that the malformation probably occurred during chromosome segregation (microsporogenesis), while the shrunken pollen grains were larger and practically the same size as the viable pollen grains, which may have suffered genetic action during microgametogenesis (see twell, 1995). Gamete development can be divided into three sequential and distinct stages: pre-meiosis, meiosis and post-meiosis (caetano-pereira et al., 1997). the normal and harmonious course of meiosis (microsporogenesis) would ensure viability of the gamete, but the post-meiotic genes, i.e., genes that transcribe in microgametogenesis, act in the interaction between the generative and vegetative cell. A break in cell-to-cell communication can lead to gamete malformation (twell, 1995), making it equally unviable. Also according to mascarenhas (1990), genetic expression during pollen grain development is divided into two different stages, with transcripts that correspond to the expression of the early genes in the development of the microspores, but a decline after anthesis and late gene transcripts appearing shortly after mitosis of the microspores, accumulating until anthesis.
in some species, abnormal meiotic behavior is considered responsible for pollen sterility and low seed production (Golubovskaya, 1979;Koul, 1990;pagliarini et al., 1992consolaro et al., 1996). the presence of meiotic irregularities has significantly affected the meiotic index and pollen viability of C. ipecacuanha and may imply that the process of sexual reproduction is not totally effective. however, in this species, sexual reproduction seems to be as important as vegetative propagation. the existence of differences in the breeding mechanism (heterostyly) among populations strengthens this idea. heterostyly is thought to have evolved because it renders cross-pollination efficient (Björkman, 1995) and favors allogamy (Robbrecht, 1988). According to vuilleumier (1967), heterostylous species can maintain a balance between inbreeding and outbreeding; this flexibility makes it a highly adaptable reproductive system, which can change under different environmental pressures to divergent breeding systems. in Salvia brandegeei, the evolution of distyly may have been associated with an ecological shift to a new environment in which protandry failed to prevent increased levels of geitonogamy; heterostyly was then selected because it increased the efficiency of cross-pollination (Barrett et al., 2000).