Search

Cyclamen purpurascens is considered suitable for horticultural breeding of cyclamens because it has an attractive fragrance that is not found in other wild species. To improve the commercial value of cyclamen flowers, this fragrance has been introduced into ornamental cultivars. However, variation in flower color is somewhat limited in these cultivars, and therefore understanding the genetic networks of flower coloration in C. purpurascens is required. We previously isolated DNA fragments of anthocyanin biosynthetic genes from C. purpurascens, broadening our understanding of the biosynthetic pathway of flavonols, which are co-pigments in flower coloration. In this study, we isolated complete open reading frames of flavonol synthase genes from C. purpurascens (CpurFLS1 and CpurFLS2) and analyzed the in planta functions of the genes by molecular complementation assay using the fls mutant of Arabidopsis thaliana. Expression patterns in several organs of C. purpurascens were also determined. The results strongly suggest that the CpurFLS genes participate in flavonol synthesis. We discuss the involvement of these two FLSs in flower coloration in C. purpurascens.

The value of plant provenance (plant origin) is well-known phenomena in woody plants, but less is known in herbaceous plants (perennials). This study with common cyclamen (Cyclamen purpurascens Mill.) was conducted to reveal the importance of specific environmental site properties of plant origin for plant growth and plant quality in the next years. The plants were observed in years 2013 and 2014, more than 10 years after removing and replanting them from the original sites. Morphological characteristics of plants were evaluated by measuring the length and the width of plant rosettes, whereby plants originated from different sites did not show any significant differences. Additionally, the pigment composition, flavonol and anthocyanin content of plant leaves were evaluated. Plants removed from sunny sites showed significantly lower chlorophyll values (total chlorophyll, chlorophyll a) in the both observed years; lower carotenoid and total pigment values were measured only in year 2013. The prevailing anthocyanin in cyclamen leaves was malvidin-3,5-diglucoside with 57.28 µg l−1 FW in the year 2013 and with 103.68 µg l−1 FW in the year 2014. Plants originated from the sunny sites accumulated in 2013 significantly more malvidin-3,5-diglucoside in comparison with plants from shady sites of origin. The major substances from the flavonol group were quercetin-3-O-rutinoside and quercetin-dirhamnosyl-glucoside in both analysed years. The cyclamen leaves originated from sunny sites contained in 2013 significant more quercetin-dirhamnosyl-glucoside than cyclamen leaves from shady sites. The results of the study show that different stress parameters (irradiation and water supply in specific year) have a significant impact on the morphological and also internal parameters of cyclamen leaves.

Anthers of a F1 hybrid (2n=41) between Cyclamen persicum (2n=2x=48) and C. purpurascens (2n=2x=34) were cultured to produce microspore-derived plants. Embryoids were produced when anthers, containing microspores at the early uninucleate stage of pollen development, were cultured in B5 medium containing sucrose (90 g l-1) and NAA (0.1, 1 mg l-1) or 2,4-D (0.1 mg l-1) in the dark at 5 °C for 4 days, then at 25 °C for 60 days. The embryoids usually developed into plantlets when cultured in B5 medium containing sucrose (30 g l-1) in the dark at 25 °C. At meiosis, the F1 hybrid, used as source for anther culture, formed some cells with restitution nuclei at telophase and dyads at the tetrad stage, which resulted in the production of viable pollen grains as unreduced gametes. Plants produced by anther culture were grouped into sterile plants with 2n=41 chromosomes and fertile plants with 2n=82 chromosomes. The present findings suggested that the sterile plants were polyhaploids originating from unreduced microspores (n=41) of the F1 hybrid and that the fertile plants were amphidiploids induced by a spontaneous doubling during culture of chromosomes of such unreduced microspores.

Interspecific crosses were made to introduce the scent of flowers of C. purpurascens into C. persicum cultivars and ovule culture was used to rescue the abortive hybrid embryos. Cultivars of C. persicum diploid (CPD, 2n=2×=48) and C. persicum tetraploid (CPT, 2n=4×=96) were the pistillate parents and wild species of C. purpurascens (CP, 2n=34) were staminate parents. After pollination, crossed ovaries were collected periodically and examined using paraffin sections. Histological observations suggested that both hybrid ovules of CPD x CP and CPT x CP should be transferred to culture medium 35 days after pollination. Based up on this observation, crossed ovaries were collected 28 days after pollination and ovules with placenta were transferred to MS (1962) medium containing 3% sucrose. These ovules were cultured in the dark at 25° C. The hybrids (2n=41) derived from CPD x CP had the scent of C. purpurascens, whereas the hybrids (2n=65) derived from CPT x CP had the scent of C. persicum. Although both hybrids had complete genomes from the parents and produced a few viable pollen grains, they failed to yield viable seeds by self- and cross-pollination with fertile pollen grains of C. persicum cultivars.

We cultured colchicine-treated hybrid ovules in vitro to produce fertile amphidiploids of C. persicum (2n=2x=48. referred to as AA) × C. purpurascens (2n=2x=34, referred to as BB). Seedlings and mature plants were obtained from the ovules without colchicine and those exposed to 50 mg/l colchicine for 5, 10 and 15 days, whereas they were not obtained from the ovules exposed to 50 mg/l colchicine for 20 days and 500 mg/l for 5, 10, 15 and 20 days. Although 8 mature hybrids derived from the ovules without colchicine produced a few fertile pollen grains, they failed to produce viable seeds by self-fertilization. The hybrids had 41 somatic chromosomes. Four and 3 mature plants were derived from ovules exposed to 50 mg/l colchicine for 10 and 15 days, respectively. One each among 4 and 3 mature plants showed a high frequency of pollen grain fertility, produced several seeds by self-fertilization, and had 82 somatic chromosomes which is twice the number of hybrid chromosomes (2n=41, AB). These findings indicated that these plants are amphidiploids (2n=82, AABB) between C. persicum and C. purpurascens. Three and 2 viable seeds were derived by the conventional crosses of diploid C. persicum × the amphidiploid and the amphidiploid × C. purpurascens, respectively. Flowering plants that developed from the seeds of diploid C. persicum × the amphidiploid were barely fertile and had 65 somatic chromosomes (2n=65, AAB), whereas those that developed from the seeds of the amphidiploid × C. purpurascens were barely fertile and had 58 somatic chromosomes (2n=58, ABB). The somatic chromosomes indicated that these plants are probably sesquidiploids between the amphidiploid and either C. persicum or C. purpurascens. The interspecific cross-breeding of cyclamen using the amphidiploids and the sesquidiploids is discussed.

Although theCyclamen subg.Cyclamen spp. are morphologically variable, previous studies suggest a superficial cytological uniformity. New chromosome numbers and an indication of karyotypic instability are reported inC. hederifolium andC. africanum that reveal that the cytology of the subgenus is more complicated than previous accounts suggest. The possible significance of these phenomena is discussed. The diploid status ofC. purpurascens has been confirmed and a distinction between the three diploid karyotypes has been described. The cytological variation may help to explain the well documented morphological variation exhibited in these species.