Level of Sterility and Morphological Flowers Differentiation of Petaloid Male-sterile Plants of Carrot

,


Introduction
Breeding of carrot (Daucus carota) hybrid cultivars, very pronounced in commercial production, is based on the development of inbred lines and incorporation of cytoplasmic male sterile (CMS) lines crossed with male fertile pollinators.Two morphologically distinct phenotypes of cytoplasmic male sterility are known: brown anthers (Welch and Grimball, 1947) and petaloids (Thompson, 1961).Brown-anther sterility is characterized by formed but unrolled, shriveled filaments and brownish anthers which are a result of tapetum degeneration (Michalik, 1971).Brown anther CMS was the first type used for developing hybrid carrot varieties, but recently petaloidy, the second type of sterility, is the more widely employed in carrot breeding programs, as CMS lines of the petaloid type have been found to maintain their male sterility better than those of brown-anther type (Davey, 1999;Bach et al., 2002).Petaloid sterility is manifested as the replacement of stamens with a second additional whorl of petals or petal-like, bract-like or carpelloid structures (Erickson et al., 1982;Kitagawa et al., 1994).Extensive genetic studies on carrot male sterility demonstrated a nuclear-cytoplasmic interaction for both CMS types.Thompson (1961) first described the cytoplasmic inheritance of petaloidy in carrot lines suggesting that the dominant alleles of each of the three duplicate nuclear genes, Ms1, Ms2, and Ms3, were necessary to maintain sterility for both cytoplasms, and dominant alleles at one or more epistatic loci could restore fertility (Timin and Vasilevsky, 1997).Contrary hypotheses were postulated for the brown-anther CMS system.The results of Hansche andGabelman (1963), andBanga et al. (1964) suggested that expression of the brown-anther sterility was due to a homozygous recessive locus Ms5 or a dominant allele for Ms4, but dominant allele of either of the two complementary loci would restore the fertility.
Several authors report that both CMS systems can be influenced by specific environmental conditions, in particular high temperatures, which promote occurring fertile plants (Michalik, 1979;Mehring-Lemper, 1987;Wolyn and Chahal, 1998;Bőrner et al., 1995).The effects of restorers can also provide valuable information on the level of sterility, even if the genes involved are not formally identified (Budar and Pelletier, 2001).
In this study, evaluation of the phenotypic uniformity within the carrot backcross populations in relation to several morphological traits of petaloidy expression was assessed.

Plant material
Eight male-sterile lines (A1, A2, A3, A4, A5, A6, A7, A8) and corresponding maintainer lines (B) bred at the Research Institute of Horticulture (RIH), Skierniewice, Poland, were used in this study.Male sterile lines (A) in their initial stage (before backcrossing) originated from the USDA carrot genetics and breeding program at Madison, Wisconsin, USA.Fertile lines (B) were bred at the RIH and originated from the material that has never been examined with regards to frequency of genes maintaining male sterility.

Growth and culture
The investigations were performed in 2001-2009.Every other year of the study, seeds were planted at the RIH, Skierniewice, Poland, in mid-May and roots were harvested during the final week of October.After vernalization at 2ºC, carrot roots were planted out during May in field.The soil type was a pseudopodsolic over loamy sand (1.15% organic mater, pH 6.5).Fertilisation and plant protection against pests and diseases were provided according to current recommendations for carrot.For each of the studied populations of the eight male sterile lines (A), different number of plants in plots were planted at a distance of 25 x 50 cm.Five plants from each cross were planted in a greenhouse to perform consecutive backcross to the corresponding maintainer.

Evaluation of the flower phenotype
The carrot flowers were macroscopically observed every other day from the beginning of the first blooming umbel to the end of flower blooming in the rertiary umbels.A male-fertile flower of carrot has five petals and five stamens with cylindrical filaments, oval anthers on the top, and a pistil with two styles joined at the bottom (Fig. 1a).Male-sterile plants were characterized by lack of the stamens transformed into additional second whorl of petals with different shape, size, and color ranging from white to green (Fig. 1b-e, 2a-d).Flowers of all cytoplasmically male-sterile (petaloid) plants were classified also according to different transformation of petals into two subtypes: complete petaloidy, e.g.leafy petals (Fig. 1b), and other structures e.g.spoon-like, filamentous structure, lobed (Fig. 1c-e).Those genotypes which at the moment of flowering had several normally developed stamens on one or more umbels of the first and higher orders were classified as partially fertile.
To describe organ modifications and to permit accurate classification of the morphology, the carrot flowers were investigated under light microscope (x10) in some cases of the petaloids.

Results and Discussion
When the corresponding maintainers (B) of male-sterile cytoplasm were backcrossed to the eight petaloid (A) lines, a high variability in the sterility level within and between most of the progenies was observed and significant differences between them were noticed.
The highest number (100%) of male-sterile plants occurred in the progenies of the 2A x 2B cross and no male-fertile plants within all backcross populations (BC 1 -BC 3 ) were noted (Table 1).That may suggest a complementary action of sex determination genes in both 2A and 2B lines.The remaining seven crosses segregated in male fertile and male sterile plants in different backcross generations.The percentage of male sterility in their progenies was high, irrespective of the backcross generation.A significant progress in CMS lines development was also noticed in the second group of four crosses: 1Ax1B, 3Ax3B, 5Ax5B, and 6Ax6B.There were large shifts observed in the backcross ranks for the percentage of sterile plants as the backcross method progressed.The best results with 100% of sterile plants were recorded for the populations most advanced in backcrossing.It is noteworthy, that in their BC 1 populations, the highest number of fertile plants was observed, but it gradually decreased in the consecutive backcrosses, and reached none in final populations.For the third and the last group of crosses (4Ax4B, 7Ax7B, 8Ax8B), a high fluctuations existed regarding the level of sterility in individual backcross populations in individual backcross populations.(Table 1).The lack of complementary genes responsible for male sterility in both A and B lines probably persisted.
It is characteristic, that partially-fertile plants were not detected in majority of the populations.Within only five backcross generations the contribution of partially fertile plants was as low as 2.2, 3.7 (3Ax3B), 3.0 (6Ax6B), 3.5 and 14.3% (7Ax7B) (Table 1).In contrast to these results, Morelok (1974) indicated a frequent occurrence of partially-fertile plants in the brown anther type of cytoplasmically-strerile carrot populations.Various studies showed partial reversion of fertility for both cms systems (brown-anther and petaloid sterility).Possible explanations for this phenomenon have been provided, for example, specific environmental conditions (Michalik, 1979;Mehring-Lemper, 1987), incomplete penetrance of maintainer genes and undiscovered restorer alleles (Wolyn and Chahal, 1998), or rearrangements in the mitochondrial genome (Chahal et al., 1998) and rearrangements of genes (Bőrner et al., 1995;Nakajima et al., 2001;Scheike et al., 1992;Szklarczyk et al., 2000).Occurrence of the partially-fertile plants of carrot in this study appeared only at the beginning of backcrossing (BC 1 and BC 2 ), except for cross 7Ax7B, which was more labile and depending on the generation resulted in 3.5% (BC 1 ), 0% (BC 2 , BC 3 ) and 14.3% (BC 4 ) of partially-fertile plants.Those results may suggest that carrot should be classified not only to the group for which the different growing temperatures play significant role in occurrence of the partially-fertile plants, but also to the other type which are genetically conditioned.That also revealed different properties of carrot and onion sterility depending on the source of sterile cytoplasm (Michalik, 1979).
Most of the petaloid CMS lines showed large variation range in corolla color and resulted in the distribution of the male-sterile plants over two color groups: white and green (Table 2).The group of white flowers, besides these of pure white, consisted of those that were halfway between green and white, and white with green stripes (Fig. 2a-d).In 17 out of the 25 studied male sterile backcross populations, white color of corolla dominated (68%) over green color and in most crosses the highest number of white-flowered petaloid type was achieved for the populations most advanced in backcrossing as a result of selection towards white floral factor.
All male sterile flowers (petaloidy) of carrot lines were characterized by the lack of stamens that were replaced with petal-like organs, and spoon-shaped and/or filamentous structures, hereinafter called other structures (Fig. 1a-d).Petaloids with spoon-like structures were observed by others (Nothnagel et al., 1997;Wolyn and Chahal, 1998).No definite tendency could be found in the most advanced in backcrossing: BC 3 and BC 4 populations (Table 3).Taking into consideration the corolla color the intensity of the transformed petals into other structures predominated over petal-like organs in both in white-as well as in green-color flowers in most of the tested populations.In all male sterile flowers, irrespectively of the corolla color, the percentage of other structures and petaloid petals was very similar and reached 49% and 51%, respectively (data not shown).Within the other structure group, the spoon-shaped structures dominated and the filamentous shape was the least common of petaloidy observed (Fig. 1c-e).Those differences in morphological structures of male sterile flowers within individual BC 3 and BC 4 populations might be due to the environmental influence on the expression of petaloidy, which has also been reported previously (Nieuwhof, 1974;Chadha and Frese, 1981).Other authors have also found complete petaloidy (Chadha and Frese, 1981), with normal structures but smaller in size (Eisa and Wallace, 1969) or less ovate (Linke et al., 1999).
The studied petaloid male-sterile plants also showed different morphological abnormalities including petals, pistils and nectaries (Fig. 3a,b).Most of these abnormalities were restricted to pistils, especially the number of styli; instead of two, none or three to six often developed.In some of the male-sterile flowers the ovaries were missing; more flowers with missing ovaries were noticed on higher blooming umbels than on the main umbel.Ericson et al. (1982) has also observed morphological deformations ranging from slight twinning of styli (seen least frequently in white-corolla phenotypes) to multiple convoluted pistils (seen most frequently in the green-corolla phenotypes).In our study, the proportions of twinning of styli and twisting pistils were in almost the same ratio in green-as well as in white-corolla flowers (20-22%; data not shown).Very similar results and relationships were obtained when the morphological observations of nectaries were carried out.Previous observations (Ericson and Peterson, 1979;Ericson et al., 1982), noted a higher number of green-corolla phenotypes of petaloid CMS lines (which developed smaller nectaries producing little or no nectar) than the white-corolla phenotypes.Contrary to these data, in our study the smaller and shrunken nectaries were seen in green-corolla florets (20%) with almost the same intensity as those of white-corolla florets (29%).In the genetically male-sterile lines and wild carrots, similar variability among petal and petaloid organs according to shape, size and composition of structure has been also noted (McCollum, 1966;Kitagawa et al., 1994).Such changes make the flowers less attractive for pollinating insects which may result in decreased number of seeds set (Erickson et al., 1982;Dyki et al., 2010).Green-flowered petaloid types, which were late ripening, were also found to be associated with the reduced seed production (Litvinova, 1988).
Clear differentiation among the newly obtained sources of petaloid male-sterile lines was possible based on the phenotypic relationships.High stability of the male sterility was observed in most of the crosses.Only three crosses differentiated in sterility and the number of fertile/sterile flowers fluctuating in their progenies.Possibly, changing the growing conditions influenced the level of sterility in consecutive backcross populations each year.Instability of the male sterility under specific conditions, mainly high temperatures, was demonstrated by Mehring-Lemper (1987) and Kaul (1988).Majority of the populations were characterized by advantageous morphological flowers traits: white corolla, well developed nectaries (Fig. 1f) and styles.Those constitute an interesting genetic source material as parent-seed component for subsequent breeding of carrot F 1 hybrids.

Figure 1 .
Figure 1.Umbelets and flowers of fertile (a) and cytoplasmically male-sterile plants of carrots with differentially transformed of cms petals into complete petaloidy (leafy petals) (b), and other structures: spoon-like (c), lobed (d), filamentous (e), and well developed nectaries (f)

Table 1 .
Sterility level within various backcrosses of eight petaloid (A) lines and their corresponding maintainers (B)

Table 2 .
Corolla color differentiation in sterile plants of cms carrot lines

Table 3 .
Occurrence of petal-like and other structures in male sterile flowers of the latest backcrosses of carrot