Reproductive characterization of interspecific hybrids among Capsicum species

The objective of this study was the reproductive characterization of Capsicum accessions as well as of interspecific hybrids, based on pollen viability . Hybrids wer e obtained between Capsicum species. Pollen viability was high in most accessions, indicating that meiosis is normal, resulting in viable pollen grains. The pollen viability of species C. pubescens was the lowest (27 %). The interspecific hybrids had var ying degr ees of pollen viability , from fertile combinations ( C. chinense x C. frutescens and C. annuum x C. baccatum) to male sterile combinations. Pollen viability also varied within the hybrid combination according to accessions used in the cross. Results indicate that male sterility is one of the incompatibility barriers among C psicumspecies since hybrids can be established, but may be male sterile.


INTRODUCTION
Interspecific hybridization is important in plant breeding as a tool for gene transfer from one genotype, usually a wild species, to another, cultivated genotype that does not have that gene (Hajjar and Hodgkin 2007). However, the cross between species may be inconsistent or incompatible due to the existence of a set of pre-and post-zygotic barriers (Hogenboom 1973). Pre-zygotic barriers can be, e.g., the absence of pollen grain germination and the delay or inhibition of pollen tube growth. After fertilization, the main barriers are embryonic death due to endosperm degeneration and the total or partial sterility of hybrid plants. These barriers have hindered or prevented the use of the wild species carrying important genes that are absent in the cultivated form in breeding programs; however, tomato breeders have used this tool to develop cultivars as reported by Hajjar and Hodgkin (2007).
The species of the genus Capsicum can be grouped into complexes, each comprising genotypes whose crosses will result in fertile hybrids: the annuum complex, comprising the species C. annuum (varieties glabriusculum and annuum), C. frutescens, C. chinense, C. chacoense and C. galapagoensis; the baccatum complex composed of the species C. baccatum (varieties baccatum, pendulum and praetermissum) and C. tovari, and the pubescens complex consisting of the species C. cardenasii, C. eximium and C. pubescens (Pickersgill 1997, Moscone et al. 2006). More recently, other methodologies have been used, such as RAPD markers (Costa et al. 2006) and chloroplast DNA (Ryzhova and Kochieva 2004), where the results have validated the classification of species in gene complexes. The species of the genus are classified into two groups, CES Monteiro et al. according to the number of chromosomes: one with 2n=2x=24 chromosomes and another with 2n=2x=26 chromosomes. Domesticated species belong to the first group (Moscone et al. 2007. In Capsicum, usually no interspecific hybrids involving species belonging to different gene complexes have been obtained due to some aspects of incongruity, i.e., incompatibility between species, unilateral incompatibility, abortion of the embryo after fertilization and male sterility (Onus andPickersgill 2004, Pickersgill 1997). Moreover, the more distant two species are genetically, more sterile or unviable their hybrid will be (Singh 1993).
One possible post-zygotic barrier can be hybrid male sterility, which can be measured by the viability of pollen grains in the hybrid combination. The pollen viability is an important indicator of the ability of pollen grains to germinate on the flower stigma and fertilize the egg cell, which is a decisive stage in fertilization (Dafni 1992). Information on pollen viability is essential for studies on the reproductive biology of plant species, allowing the confirmation and understanding of results reported of certain crosses. Pollen viability can be estimated by a number of methods: Lugol solution, Fluorochromatic Reaction (FCR), vital dyes such as tetrazolium salt, germination test in vivo and in vitro and Alexander´s solution. According to Alexander (1969), the staining solution is a differential dye that distinguishes viable and unviable pollen grains in most angiosperms, and is considered a quick, cheap and easy technique (Dafni 1992). The distinction by the solution is based on the three dye components: malachite green, acid fuchsin and orange G (Alexander 1969). Malachite green reacts with the pollen grain wall, so the aborted pollen grains are stained green, because they consist of the pollen wall only. Acid fuchsin colors the cytoplasm and mitochondria and its role in this solution is to stain the protoplasm of fertile pollen grains in a color range from red to purple red. Orange G reinforces the differentiation of the above colors, resulting in a clearer distinction between viable and unviable pollen grains (Alexander 1969).
According to Pickersgill (1997) interspecific hybridization has been used frequently in Solanaceae, particularly regarding the transfer of genes for disease and pest resistance; in Capsicum however, the percentage of pollen viability of interspecific hybrids ranges from high to low (Kumar et al. 1987, Lanteri andPickersgill 1993). According to Shifriss (1997), the degree of male sterility in interspecific Capsicum hybrids depends on the accessions used for each cross. For breeding programs of interspecific hybridization it is therefore essential to evaluate the pollen viability of accessions and hybrids.
This study aimed to characterize the reproductive capacity of interspecific hybrid combinations, obtained by crosses between accessions from domesticated species of Capsicum as well as the reproductive performance of 36 representative accessions of Capsicum species (C. annuum, C. chinense, C. frutescens, C. pubescens, and C. baccatum and their botanical forms) based on the pollen viability.

MATERIAL AND METHODS
Hybrid combinations obtained from artificial pollination among 36 representative accessions of domesticated Capsicum species were evaluated. C. annuum was represented by 12 accessions (10 accessions of C. annuum var. annuum -sweet and hot peppers and two accessions of C. annuum var. glabriusculum -ornamental pepper) ( Table 1). The hybrid nature of combinations was validated using molecular as well as phenotypic markers. The accessions and interspecific hybrids were grown in 5 L pots in greenhouses; 16 hybrid combinations were grown in a greenhouse of a research support unit and 27 at the Experimental Station of Pesagro, Campos dos Goytacazes, RJ. Fertilization and pesticides to control pests and diseases were applied according to the recommendations for the crop (Filgueira 2003).
For the reproductive characterization by pollen viability analysis of the species and hybrids, the flower buds were collected at anthesis and stored in 70 % ethanol solution. Subsequently, the anthers were squashed in drops of Alexander´s solution (Alexander 1969). Slides were prepared and examined under an optical microscope (Olympus BX 60, bright field), and the number of viable (purple) and number of unviable pollen grains (green) were counted. Five slides were prepared for each genotype, each representing one anther, where 250 pollen grains/ slide were counted, resulting in a total of 1250 pollen grains per accession. The data were converted into percentage (%) and the original data were subjected to analysis of variance using software Genes (Cruz 2001) based on a completely randomized design where the slides, five slides/ accession, represented the replicates.

Pollen viability of the accessions
The distinction of viable (fertile) from nonviable (infertile) pollen grains was performed with high accuracy (Figure 1), which was confirmed by the low coefficient of variation ( Table 2). The analysis of variance of pollen viability data showed significant differences between accessions at 1 % probability.
The pollen viability of the botanical forms C. annuum var. annuum (sweet and hot peppers) and C. annuum var.
glabriusculum (ornamental pepper) was high, with averages ranging from 85.4 % to 97 % ( Table 3). The pollen viability of the species C. annuum var annuum, sweet pepper, was lower (85.4 %) than of the other two forms, mainly because of accession UENF 1569, with an average of 43 %, considered low. According to Pickersgill (2007) domesticated species of Capsicum have a wide range of types, from wild forms, over the cultivated to the completely domesticated forms, except for C. pubescens which is only known in the cultivated form. Thus, the low pollen viability of accession UENF 1569 may be due to a problem of adaptation or due to the presence of some male sterility gene, as is well documented in the literature for C. annuum (Shifriss 1997).
The pollen viability of C. annuum var. annuum (hot pepper) and C. annuum var. glabriusculum was high, on average 90.8 and 97 %, respectively. The pollen viability of C. chinense was also high (98 %). Capsicum frutescens also had high pollen viability (86.8 %), with the exception of the accessions UENF 1560 and UENF 1588, which had intermediate mean values of, respectively, 60 % and 77 % ( Table 3).
The pollen viability of Capsicum baccatum and its botanical forms (var. baccatum and var. pendulum) were high, on average 98.3 % and 96.1 %, respectively, and highest for accession UENF 1417, with 100 % viability, indicating that these accessions have a normal meiosis and are therefore well-adapted to the environment where they were grown.
As expected, the pollen viability of Capsicum pubescens was very low (27 %), since this species was originated in the Andean region, with a natural environment of high altitude and low temperatures (Pickersgill 1997(Pickersgill , 2007, in contrast to the conditions in the region where the study was conducted.
In general, the pollen viability of Capsicum accessions was high. This finding indicates that the non-formation of fruit or the formation of seedless fruits observed in some interspecific crosses established and described in the literature cannot be attributed to pollen viability.

Pollen viability of the hybrid combinations
The analysis of variance of the data of pollen grain viability differed significantly among interspecific hybrids (Table 4), which can best be observed by the means of pollen viability of each hybrid combination (Table 5).
The percentage of pollen viability of the hybrids between the species C. annuum var. annuum x C. chinense (Figure 1) was low (56 %). This was due to the fact that the pollen viability in some combinations (UENF 1503 x UENF 1555) was very low (34 %) while the reciprocal combination had a high mean viability (mean = 87 %). Kumar et al. (1987) observed a pollen viability of 42 % in a hybrid between C. annuum x C. chinense; a cytological analysis showed that the hybrid C. annuum differs from C. chinense by two translocations and some minor structural chromosome changes, leading to partial hybrid sterility. Lanteri and Pickersgill (1993) reported that pollen viability of C. annuum x C. chinense hybrids ranged from 13 to 16 % and observed that during meiosis the hybrid accessions used in the study differed by two translocations. They also found that the translocations involved three rather than four chromosome pairs, since hexavalent associations were observed in the hybrids, ie, the same chromosome was common to the two independent translocations. According to the authors it is interesting to identify chromosomes that are involved in more than one translocation since sterility associated with heterozygous translocation can prevent gene introgression in Capsicum breeding  programs (Tanksley 1984). These results were not expected since the species belong to the same gene complex. However, according to Carputo (1999), similar problems occur in interspecific crosses of tomato and potato, both of the Solanaceae, as Capsicum. These results are interesting because they suggest that in this type of combination, C. chinense should be used as female parent to obtain good hybrid fertility.
Good pollen viability (81.5 % -84 %) was observed in the hybrid combination of C. annuum var. annuum x C. frustescens, and when C. annuum var. annuum was used as female parent, the pollen viability was slightly lower than of the reciprocal combination, with 95 % of viable pollen grains.
Hybrids between the species C. chinense x C. frustescens resulted in high pollen viability (Figure 1), of Table 3. Mean pollen viability of the representative accessions of the five domesticated Capsicum species around 94 %, confirming that these two species are compatible and produce fertile hybrids, so a gene transfer between the two species is possible. Unfortunately, no reciprocal hybrids were obtained, but these two species are considered genetically very close, since 12 bivalent pairs were observed in interspecific hybrids (Shifriss 1997).  Reproductive characterization of interspecific hybrids among Capsicum species C. annuum and C. baccatum are considered species belonging to different gene complexes in view of the crossability. However, fertile hybrids were obtained between the species C. annuum var. annuum (sweet or hot pepper) and C. baccatum var pendulum with pollen viability exceeding 90 % (Table 4). On the other hand, the viability of the reciprocal combinations was low, around 35 %, and combinations between C. baccatum var. baccatum x C. annuum var. annuum resulted in hybrids with very low fertility as well (Table 5), with an average 35 % of viable pollen grains. C. annuum and C. baccatum belong to different gene complexes, and the results reported here were not expected. An interesting observation was the high pollen viability when C. annuum was used as female parent, since the pollen viability of such combinations was high. According to Shifriss (1997), hybrids between C. annuum and C. baccatum have varying degrees of male sterility, according to the accession used in the interspecific cross. Kumar et al. (1987) estimated a pollen viability of around 42.6 % for the interspecific hybrid between C. annuum var cerasiforme x C. baccatum var. pendulum, and concluded that the two species differed by two translocations, one inversion and some small chromosomal changes and that the barriers between the two species are expressed in the hybrid fragility and the occurrence of chromosome desynapsis in the hybrid, once the occurrence of desynaptic genes was reported in C. annuum var. cerasiforme . Egawa and Tanaka (1986) however reported very low pollen viability (around 6.5 %) of the hybrid combination between C. annuum var minimum x C. baccatum var. baccatum.
In the hybrid combinations C. baccatum var. pendulum x C. chinense and C. baccatum var. pendulum x C. annuum var glabriusculum the pollen viability was very low (25 % and 32 %, respectively). The combinations C. baccatum var. pendulum x C. frutescens varied in pollen viability (from 54.5 % to 46 %), which was considered low, indicating partial male sterility of these hybrids. Bapa Rao et al. (1992) estimated pollen viability ranging from 23.6 to 25.4 % in a hybrid combination involving the two above species. According to the authors the main factors of hybrid sterility were sterility genes, heterozygous translocations, and meiotic behavior with segregation anomalies. The same authors reported difficulty in the establishment of the reciprocal combination, which were not obtained even after repeated artificial pollinations. In this study the reciprocal combination between the two species was achieved, but the pollen viability was very low (Figure 1), indicating a male sterile combination, confirming the data of the other authors.
The pollen viability of the hybrid combination C. frutescens x C. annuum var. glabriusculum was also very low (Table 5), less than 30 %, indicating problems of chromosome homology between these species; this result was rather unexpected as the species belong to the same gene complex. However, C. frutescens carries mitochondrial genes for male sterility (Shifriss 1997), so it is possible that the hybrid will show varying degrees of male sterility in all combinations involving this species as female parent.
Generally speaking, the hybrid pollen viability ranged from very low values, suggesting complete male sterility, to intermediate and high values, indicating fertility of the combination. A pollen viability lower than 50 % can result from a lack of chromosome homology, probably, due to the genetic distance between species, or due to the presence of heterozygous translocations or due to the presence of cytoplasmic male sterility genes. These combinations should be carefully analyzed to isolate and study the nature and inheritance of male sterility genes, an interesting feature that can be used in breeding programs targeting hybrid production (Pickersgill 1997).
In conclusion, some combinations are possible and fertile, enabling the gene transference between species, whereas the fertility degree of other combinations can differ, indicating partial or complete male sterility, assessed here based on the pollen viability. The conclusion was also drawn that infertility is a postfertilization barrier since the establishment of hybrid combinations is possible, but breeders will have to apply strategies as described by Yoon et al. (2006), to continue the breeding program of Capsicum.