CYTOPLASMIC MALE STERILITY AND INTER AND INTRA SUBGENOMIC HETEROSIS STUDIES IN BRASSICA SPECIES : A REVIEW

Plants of the genus Brassica comprise a remarkably diverse group of crops and encompass varieties that are grown as oilseeds, vegetables, condiment mustards and forages. One of the basic requirements for developing hybrid varieties in oilseed Brassica is the availability of proven heterosis. The development of hybrid cultivars has been successful in many Brassica spp. Midparent heterosis and high-parent heterosis (heterobeltiosis) have extensively been explored and utilized for boosting various quantity and quality traits in rapeseed. Heterosis is commercially exploited in rapeseed and its potential use has been demonstrated in turnip rape (B. rapa L.) and Indian mustard (B. juncea L.) for seed yield and most of the agronomic traits. The oilseed rape plant, B. napus, possesses two endogenous male sterile cytoplasms, nap and pol. Ogura type of cytoplasmic male sterility was first discovered in Japanese wild radish and other male-sterile Brassicas (Ogura bearing cytoplasm) derived from interspecific crosses. Information concerning the allelic frequencies of restorers can be useful in trying to understand their evolutionary origins. The ogu, pol and nap cytoplasms of B. napus induce sterility in all, some, and only a few cultivars, respectively. In this study, different kinds of male sterility, combining ability and heterosis of qualitative and quantitative traits in different Brassica species will be reviеwed.


Introduction
The Brassica is an important genus of angiosperms consisting of over 3,200 species with highly diverse morphologies.They are used as nutritious vegetables, condiments and for producing oil (Nasrin et al., 2011).The development of hybrid genotypes/high-yielding varieties has an important role to fill the gap between food production and human population.Heterosis and combining ability studies in Brassica species lead to the development of the hybrid genotypes/varieties that will produce better seed yield, plant biomass, oil content, protein content and canola type traits for good health.Cytoplasmic male sterility (CMS) is one of the most examining the genetic control of grain yield in oilseed rape, both additive and nonadditive gene effects have been found to be involved (Yadav et al., 2005).

Relationship and origin of the Brassica species
Plants of the genus Brassica comprise a remarkably diverse group of crops and comprise varieties that are grown as oilseeds, vegetables, condiment mustards and forages.The cytogenetic and evolutionary associations between the major oilseed and vegetable species are commonly depicted as U's triangle, named after the Korean scientist who first formulated it (U, 1935).U speculated that B. carinata, B. juncea and B. napus are each allotetraploids formed by interspecific hybridization events between the parental diploid species B. nigra, B. rapa and B. oleracea.Thus, hybridization between B. nigra and B. rapa resulted in the formation of B. carinata, and between B. nigra and B. oleracea in the formation of B. juncea.B. napus is a hybrid between B. oleracea and B. rapa.The relationships among these species first assumed by U have since been proved by a large variety of genetic and molecular analyses.The haploid genomes of B. rapa, B. nigra and B. oleracea are assigned A, B and C respectively.Thus, diploid B. rapa has two copies of the A genome within 20 chromosomes (AA, n=10, 2n=20) and diploid B. napus has also two copies of both the A and C genomes within 38 chromosomes (AACC,n=19,2n=38).The diploid species have a long history of domestication and B. rapa was already cultivated during the Bronze Age in Northern Europe (Persson et al., 2001), towards the end of the sixteenth century in Holland and Belgium, and in the eighteenth century in Britain.Molecular and morphological studies have suggested that B. rapa originated from two independent centers: Europe and Asia (He et al., 2002;Zhao et al., 2005).The Asian types contain several subgroups of species which are mainly used as leafy vegetables, while the European types are used as oilseeds (Reiner and Ebermann, 1995).Based on vernalization requirement before flowering, B. rapa can be grouped into winter and spring types and presently for oilseed production in moderate and warm temperature regions, mainly spring type is cultivated.The Brassica species together are the second largest oilseed crop produced worldwide (FAO, 2012).The most important Brassica species is B. napus, but B. rapa is also of special interest as a progenitor of B. napus and B. juncea.The oil is presently processed, as natural substance in the petrochemical industry for biodiesel and over 3.9 million tons of biofuel was produced by the EU in 2005 (EC, 2006).The subspecies rapifera of B. rapa is cultivated either for its turnips or leaves.In Northern Spain, Portugal and Southern Italy (Padilla et al., 2005), it is used either as leafy vegetable for human consumption or as fodder for feeding animals, depending on the morphotype.The swollen root is consumed by both humans and animals.In China, different morphotypes of B. rapa are vegetable cultivars, which comprise Chinese cabbage (subsp.pekinensis) distinguished by its large leaves with wrinkled surfaces, and Pak choi (subsp.chinensis) which does not form heads (Zhao et al., 2005).

The concept of subgenome of Brassica
Long years of evolution and artificial selection have made the A and C genomes in B. napus to some extent different from the A genome in B. rapa and B. juncea, the C genome in B. oleracea and B. carinata (Li et al., 2010).To distinguish the difference, the concept of subgenome was introduced to genus Brassica.Thus, A r , A n and A j were used to characterize the A genome in B. rapa, B. napus and B. juncea, B b , B j and B c for the B genome of B. nigra (black mustard), B. juncea and B. carinata.The C o , C n and C c were used for the C genome of B. oleracea, B. napus and B. carinata (Li et al., 2004(Li et al., , 2005a(Li et al., , 2006b(Li et al., , 2007;;Qian et al., 2007) (Li et al., 2004).After that, most interspecific hybrids in Brassica could be considered as intersubgenomic hybrids, such as A r A n C n (B.napus × B. rapa) and A r B c C c (B. carinata × B. rapa).As shown in prior studies, the positive correlation was detected between genetic distance of parents and their F 1 mid-parent heterosis for seed yield in B. napus.The new-typed of B. napus including A r A r C c C c or A r A r C n C n with normal meiosis can be created by interspecific hybridization and molecular selection.The heterosis would be expected in the hybrid of A r C c C n C c or A r A n C n C n if hybridization were performed between the new typed B. napus with genome composition A r A r C c C c or A r A r C n C n and the natural B. napus with genome composition of A n A n C n C n .A r A n C n C n was generated by the hybridization between A r A r C c C c × A n A n C n C n and A r A r C n C n × A n A n C n C n .The method for the creation of new-typed B. napus with genome composition of A r A r C c C c and A r A r C n C n is as shown in Figure 2.

Gynodioecy
Gynodioecy is the co-existence of hermaphroditic and female (male sterile) individuals in populations of the same plant species such as Brassica.Male sterility may have multiple causes.It can result from adverse growth conditions, from diseases, or from mutations.Naturally occurring genetically male sterile plants in hermaphrodite species generally maintain fully normal female functions.The phenotypic expressions of male sterility are very diverse from the complete absence of male organs, the failure to develop normal sporogenous tissues (no meiosis), the abortion of pollen at any step of its development, the absence of stamen dehiscence or the failure of mature pollen to germinate on compatible stigma.Although multiple causes for pollen abortion exist, cytoplasmic male sterility, which depends on mitochondrial genes, is the determinant of gynodioecy most frequently met in natural populations.The molecular data collected over the last two decades by molecular biologists on sterility inducing mitochondrial genes and their expression allow a better understanding of the protection and the role of this sexual dimorphism in plant species in relation to theoretical models established by population geneticists (Budar and Pelletier, 2001).

Male sterility
Nuclear male sterilities are generally recessive mutations that affect a huge number of functions.These mutations may affect for example proteins involved in male meiosis (Glover et al., 1998), the metabolism of plant hormones, the biosynthesis of complex lipid molecules or the synthesis of secondary metabolites (Aarts, 1995).Cytoplasmic male-sterility (CMS) which has been observed in over 150 plant species is usually identified as "maternally inherited deficiency in producing viable pollen" and the mitochondria are responsible for this trait.This characterization leads us to call CMS any mitochondrial mutation, which impairs the proper functioning of mitochondria, leading to male sterility.For example, in Nicotiana sylvestris, mitochondrial mutants were obtained from in vitro culture (Gutierres et al., 1997).They exhibit male sterility, but also many other have abnormal characters throughout all their development.Such mutated mitochondrial genomes have not been clearly selected for in natural conditions.Therefore, a better definition for CMS would be: "maternally inherited male sterility resulting from a specific (mitochondrial) gene whose expression impairs the production of viable pollen without otherwise affecting the plant".The present information about CMS genes results from studies in several cultivated species, where this trait is used for F 1 hybrid production.The Ogura type of cytoplasmic male sterility (Ogura, 1968) was first discovered in Japanese wild radish.The male sterility inducing gene (orf138) was determined (Grelon et al., 1994) by studying the recombined mitochondrial DNA of 'cybrids' (cytoplasmic hybrids) obtained through protoplast fusion (Pelletier et al., 1983) between fertile Brassica (napus or oleracea) and male-sterile Brassicas (Ogura bearing cytoplasm) derived from interspecific crosses (Bannerot et al., 1974).The 'orf138 locus' encompasses two open reading frames coding respectively 138and 158-amino-acid polypeptides which are co-transcribed.The latter, also called orfB, communicates to the subunit 8 of the ATPase complex (Gray et al., 1998), and the orf138 gene has some similarities with different radish mitochondrial DNA sequences representing that it is related to intragenomic recombination.The particular structure of this locus determines the stability of the orf138 gene.In some of the plants which resulted by protoplat fusion, the male sterility was unstable due to a frequent loss of the CMS-inducing gene (Bellaoui et al., 1998).The instability of the CMS is also detected when a second copy of the orfB gene allows the loss of the copy related to the orf138 gene.

The possible role of CMS and restorer genes in evolution
In alloplasmic situations, the observed phenotype is often complex, resulting from the association of both chloroplastic and mitochondrial genomes with an alien nuclear counterpart.Some of the CMS induced by alloplasmic situations could also be a revival of ancient CMS in the cytoplasm donor species, which have been silenced by fixation of nuclear restorers in these species.A new nuclear background, which has never selected restorer genes for this CMS, allows the cytoplasm to induce sterility again.One can expect to detect the CMS inducing gene in such alloplasmic situations by looking for an ORF, present but silent in the cytoplasm donor species and expressed (and probably co-transcribed with an indispensable mitochondrial gene) in the species where CMS is expressed.This is observed in the case of the alloplasmic B. napus and B. juncea plants carrying B. tournefortii cytoplasm.In the B. tournefortii mitochondrial genome, a chimaeric ORF (orf263) is related to the atp6 gene, but it is not expressed.In alloplasmic lines of B. napus or B. juncea carrying the B. tournefortii cytoplasm, orf263 is co-transcribed with atp6 and the bicistronic mRNA is accumulated (Landgren et al., 1996).Although the responsibility of orf263 in the male sterile phenotype of alloplasmic plants remains to be proven, the expression behavior of this mitochondrial gene in different nuclear backgrounds conforms to what is expected from a 'silent' CMS gene reactivated in an alloplasmic context.One of the most intriguing features of CMS, still open to hypotheses, is the precise localisation in the development of the phenotype induced by a mitochondrial gene which, in all cases except the bean pvs gene (Abad et al., 1995), is constitutively expressed.This observation is not so surprising when one considers that these genes have recruited expression signals from other mitochondrial genes and are therefore submitted to the same controls.One can expect that CMS genes which have been maintained for a long time (and probably selected for) in natural populations, such as the Ogura CMS present in different Rhaphanus species (Yamagishi et al., 1996;1997) or the bean CMS determinant present in different Phaseolus species will have an effect strictly limited to pollen production.Hence, the limited phenotype induced by these genes must result from a specific 'activity' in male organs.It is likely that the possible means of impairing male gametogenesis via a mitochondrial defect without affecting the female fitness of the plant are limited.This remark might appear to be in disagreement with the observed absence of homology between CMS gene sequences.However, it is noteworthy that at least one hydrophobic domain and a possible or shown relationship to the mitochondrial membrane are shared between the encoded proteins (Schnable and Wise, 1998).Some stretches of amino acids are sometimes shared by these otherwise unrelated proteins (Tang et al., 1996;L'Homme et al., 1997), but an interpretation of this latter observation in terms of function of the encoded proteins is still premature.

Population genetics of restorers
Information concerning the allelic frequencies of restorers can be useful in trying to understand their evolutionary origins.The ogu, pol and nap cytoplasms of B. napus induce sterility in all, some, and only a few cultivars, respectively.Hence, one might realize that the ogu restorer (Rfo) is absent from B. napus germplasm, pol restorers are rare, and nap restorers are more common.Higher-plant mitochondrial genomes are thought to exist in the form of multiple circular DNA molecules capable of recombining with one another at high frequency across specific recombination repeat sequences generally greater than 2 kb in size.Infrequent recombination events across shorter repeat sequences can also happen, and rearrangements resulting from such rare events represent the major form of evolutionary change occurring in the molecule.In some cases these rare recombination events have created chimeric genes that are associated with cytoplasmic male sterility (CMS) (Hanson, 1991;Hanson and Folkerts, 1992;Bonen and Brown, 1993).CMS is a widespread, maternally inherited trait of higher plants that results in pollen abortion.The expression of these chimeric genes can often be influenced by specific nuclear genes, termed restorers of fertility (Rf) that suppress male sterility and allow for normal pollen development.It remains to be explained how the expression of these novel genes specially alters pollen development.CMS in B. napus can be given either by alien cytoplasms, such as the Ogura cytoplasm of radish, or by endogenous cytoplasms established within cultivated varieties (Shiga and Baba, 1980).The endogenous systems entail an unusual set of interactions between different nuclear and mitochondrial genotypes.Male fertility of most nuclear genotypes is unaffected by the cytoplasm assigned nap found in most B. napus cultivars.The nap cytoplasm can, however, give male sterility on a limited number of cultivars that lack corresponding Rf genes, such as the cultivar 'Bronowski' (Shiga and Baba, 1980).Male-fertile 'maintainers' of such cultivars possess a cytoplasm designated cam that is thought to be derived from B. campestris, one of the diploid progenitor species of the amphidiploid B. napus.A third cytoplasm, 'Polima' or pol, confers male sterility on most B. napus cultivars (Fan et al., 1986).A few genotypes can restore fertility to pol male-sterile cytoplasms, and restoration may take place through the action of either of two unlinked dominant alleles (Fang and McVetty, 1989).Both nap and pol cytoplasms can therefore induce male sterility in appropriate nuclear backgrounds, whereas the cam cytoplasm appears unable to induce male sterility in B. napus.Several lines of evidence suggest that the region of the pol mitochondrial genome surrounding the atp6 gene is associated with the CMS trait.It is the only region for which transcript differences have been identified among fertile pol CMS, and fertility restored pol plants (Singh and Brown, 1991;Witt et al., 1991).This region is also organized differently between the sterile pol and fertile cam mitochondrial genomes (L'Homme and Brown, 1993), and it is the only region that is genetically correlated with CMS in somatic hybrids (Wang et al., 1995).Singh and Brown (1991) found that the pol atp6 gene is co-transcribed with a chimeric gene, orf224.The first 58 codons and the 5¢ noncoding region of this chimeric gene resulted from a common mitochondrial gene of unknown function termed orfB.The pol mitochondrial genome contains a 4.5-kb DNA segment immediately upstream of atp6 that is missing in cam mtDNA and spans orf224.An apparently intact copy of this segment was also established near the atp6 gene in the nap mitochondrial genome.However, this copy is 5 kb further upstream of atp6, and nap atp6 transcripts are monocistronic.Since the genomic environments in which the 4.5-kb segment is located differ between pol and nap mtDNAs, it was suggested that the segment should be transposed since the divergence of lineages leads to the two mitochondrial genomes (L' Homme and Brown, 1993).Nothing associated with this transposition was detected in the pol mitochondrial genome, of approximately 1 kb of DNA from a region immediately upstream of atp6.The finding of an mtDNA segment containing a sequence related with pol CMS at a distinct site in the nap mitochondrial genome suggested that the sequence might also be associated with the nap CMS.To address this issue, we have sequenced the nap and pol 4.5-kb segments, the regions upstream of atp6 in the cam and nap mtDNAs that are missing from the pol mitochondrial genome, and the regions neighboring the sites of these rearrangements.An open reading frame (ORF) homologous to, but divergent from, orf224 was found at the matching site in the nap 4.5-kb segment.This open reading frame, orf222, is co-transcribed with the trans-spliced exon c of the nad5 gene and another ORF.We present results indicating that expression of the orf222 region is uniquely associated with the male sterility induced by the nap cytoplasm.This study reveals several rearrangements that have occurred during the evolution of the nap, cam and pol mitochondrial genomes in addition to those previously identified (L'Homme and Brown, 1993), all involving the transposed nap and pol 4.5-kb segments.The two segments, though highly similar in sequence over most of their length, each contain an internal region not found in the other.In addition, 603 bp of the pol segment most distal to the chimeric ORF are missing in the nap segment.The most striking difference between the nap and pol segments, however, is the extensive degree of sequence divergence at the 3¢ ends of the respective ORFs.The absence of the terminal region of the nap segment, and especially the extensive degree of change in the open reading frames, indicate that simple transposition of the 4.5-kb segment is not sufficient to account for the changes observed.Among the Brassica mtDNAs in general (L'Homme and Brown, 1993), it seems unlikely that the widespread divergence observed in the 3¢ ends of these sequences is the result of changes that occurred separately subsequent to the divergence of the nap and pol lineages.Under slightly reduced hybridization stringency conditions, a probe derived entirely from the non orfB-homologous portion of orf222 detects over ten different fragments in southern hybridizations to detect the cam, nap and pol Brassica mtDNAs, representing that the sequences from this region are present at multiple sites on the mitochondrial genome in each of these mtDNAs (Small et al., 1989).

Combining ability
Selection of parents for synthetic or hybrid breeding is based on their combining ability.Combining ability is the ability of a parent to generate superior progeny and it is divided into general and specific combining abilities (GCA and SCA, respectively).The GCA effect of a population is an indicator of the relative value of the population in terms of frequency of favorable genes as compared to the other populations.The SCA effects of two populations express the differences of gene frequencies between them and their divergence, as compared to the diallel populations (Viana, 2000).The mating designs often employed in the assessment of combining ability are line x tester and diallel crosses (Griffing, 1956;Gardner and Eberhart, 1966).These permit the selection of superior pure lines for hybridization and, in cross-pollinating species, screening of populations for use within and between population breeding programs.Studies on combining ability for traits such as yield and other agronomic traits are available in different Brassica species with diallel analysis.Most of the studies in Brassica napus showed significant GCA and SCA effects for yield and its component characters indicating that both additive and non-additive gene actions were important in the inheritance of these traits (Rameeh, 2010;2011a;2011b;2011c;2011d).Qian et al. (2003) evaluated intraspecific hybrid between B. rapa × B. napus for biomass yield in two years.A significant variation observed for both GCA and SCA indicates that both additive and non-additive effects influenced biomass yield production.The ratios of variance component for GCA to SCA were 89% in 1999 and 88% in 2000, showing that GCA had more important role though both were significant.Wang et al. (2007) studied combining ability for different traits in subspecies of Chinese B. rapa.They observed that yield per plant and length of main inflorescence were mainly controlled by SCA; plant height, number of primary branches, siliques of primary branches, seeds per silique and 1000-seed weight were controlled by both GCA and SCA; and number of secondary branches, siliques of secondary branches and siliques per plant were mainly controlled by GCA.Combining ability of 15 subspecies of B. rapa was estimated by using diallel including reciprocals for 12 characters related to yield and oil content (Singh and Murty, 1980).Gene action was predominantly controlled by SCA effects with GCA effects playing a minor role in the traits including oil content and 50% of flowering.Yadav et al. (1988) in nine inbred lines of brown sarson used as females and three other cultivars as males examined the combining ability of their 27 hybrids.Specific combining ability was observed to control all traits when the hybrids were evaluated for plant height, number of branches per plant, number of seeds per pot, 1000-seed weight and seed yield per plant (Rameeh, 2012a;2012b;2012c;2012e;2013a;2013b).

Heterosis
Heterosis, the term that followed "heterozygosis" which was first used at the beginning of the 20th century, was defined by Shull (1948) as "the increased size, the excessive kinetic energy, the increased productiveness, resistance to disease or to unfavorable conditions of the environment, the stimulating effects of hybridity which may be observed in cross-bred organisms when compared with corresponding inbred or comparatively more pure-bred organisms".In short, heterosis is "the increase in size or rate of growth of progeny over parents" (Duvick, 1999;Chen, 2010;Jahnke et al., 2010).This is a phenotypic description and heterosis is generally observed as a property of quantitative traits, therefore the first theoretical explanation of heterosis was given through quantitative genetics.
Heterosis can be described in different ways.One formulation is the difference between the hybrid and the mean of the two parents, known as mid-parent heterosis.Falconer and Mackay (1997) explained the theoretical background of mid-parent heterosis based on the relationship between genetic distance and dominance effect.Later, Lamkey and Edwards (1998;1999), based on a theoretical framework described by Willham and Pollak (1985), revisited and refined the theoretical relationship by differentiating heterosis at the population level and in an inbred line cross system.The previous could be derived from the genetic architecture of both parents, whether they were from random-mating or inbred populations.The concepts of baseline heterosis are panmictic mid-parent heterosis, and inbred mid-parent heterosis.The inbred mid-parent heterosis, which is the sum of baseline heterosis and panmictic mid-parent heterosis, has been generally exploited in the production of hybrid cultivars.Another particularly important point that appeared from the theoretical concerns of heterosis is that the performance of an F 1 hybrid is a function of dominance and unlinked dominance interacting via dominance epistasis at loci showing genetic divergence.The first hypothesis proposed as an explanation of heterosis was the theory of overdominance presented by East (1936).This idea assumed that "vigor is promoted when the genes at certain loci are unlike".On the other hand, Jones (1917) showed that heterosis "could result from normal gene action and be a phenomenon accompanying hybridity".This observation led to a second hypothesis named "dominance theory", although it can be better described as "avoidance of recessive harmful genes" since the idea is based on heterozygous loci that prevent deleterious effects brought about by recessive genes.Rasmusson (1933) proposed a gene interaction hypothesis, which was later called "epistasis theory".In a quantitative genetic sense, estimation of heterosis effects is actually a breakdown of genetic architecture into its components, dissecting phenotypic variation into additive genetic and dominance gene actions and their epistatic interactions (Comstock and Robinson, 1948;Griffing, 1956;Mather and Jinks, 1972).Within a properly set mating design to develop certain types of generations, the heterosis effect is often treated as an effect of dominance, or the "specific combining ability" effect, especially when epistasis is ignored in the model.
In Indian mustard (Brassica juncea), Yadava et al. (1974) reported heterosis of 239 per cent over better parent for seed yield per plant.A wide range of positive heterosis for the number of primary and secondary branches per plant, plant height and seeds per siliqua were also reported (Rawat, 1975).In another study of partial diallel analysis of mustard, significant heterosis for yield per plant, number of primary and secondary branches, length of siliqua, 1000-seed weight were reported (Ram et al., 1976).Banga and Labana (1984) reported up to 200 per cent heterosis in B. juncea.It was further verified that some of the crosses exhibited heterobeltiosis.Verma et al. (2000) reported a significant degree of heterobeltiosis in several varieties of Indian mustard.
Heterosis is the difference in performance between F 1 generation and midparent or high-parent and has been a major breeding tool for plant productivity improvement.Preferably, inbred lines with genetically different backgrounds are used as parents for F 1 production.It makes maximum use of heterosis by combining favorable alleles of the individual homozygous parents.In populations such as B. rapa, a part of heterosis is already utilized in base population due to their open pollination with plants being partly heterozygous.However, it can take advantage of the homozygous plants within the population for heterosis, and also heterotic increase which could result by crossing heterozygous plants.Parental populations with different genetic make-up such as cultivars (Shuler et al., 1992), synthetics (Falk et al., 1998), and subspecies (Wang et al., 2007) have been used in heterosis studies in B. rapa.For estimating heterosis in crosses between population, Lamkey and Edwards (1999) suggested the term panmictic mid-parent heterosis for the difference between the mean of two random mating populations and the mean of a hybrid population produced by crossing individual plants of the two populations.Dominance, overdominance and epistasis are the three principal genetic explanations for heterosis.The dominance hypothesis stipulated that heterosis is contributed by favorable alleles of both parents.Overdominance is a condition where loci in the heterozygous state are superior to parents and epistasis defined as complex interactions of favorable alleles of the two parents (Xiong et al., 1998;Crow, 1999;Tian and Dai, 2004;Sun et al., 2004).Heterosis can only occur when parental cultivars used for F 1 production differ in gene frequencies (Falconer and Mackay, 1996).Heterosis for different agronomic traits has been reported.Significant positive correlation of mean performances with heterosis and heterobeltiosis effects for yield components except 1000-seed weight in rapeseed indicated that selection of the superior crosses based on heterosis and heterobeltiosis effects will be effective for their mean performances improving for these traits except 1000-seed weight (Rameeh, 2012d).Schuler et al. (1992) in inter-cultivar F 1 s of B. rapa reported mid-parent heterosis (MPH) of 18% for seed yield.Falk et al. (1998) in cultivars of spring B. rapa reported 25% MPH in seed yield.Kaur et al. (2007) in B. rapa subspecies of toria, brown sarson and yellow sarson observed 31% heterosis in intra group crosses and 17% in inter group crosses for seed yield.Wang et al. (2007) in Chinese B. rapa vegetables reported MPH of 10% for plant leaves, 44% for petiole fresh weight and 17% for the length of the biggest leaf.One of the most expensive steps in heterosis utilization is the identification of parental combinations that produce F 1 with superior yield.Therefore, the prediction of F 1 performance with accuracy from morphology or molecular data is important.This could reduce the cost involved in evaluating parent and crosses in field trials to identify parental combinations that will give high F 1 performance.The predictions of heterosis from parental genetic distance have been widely studied in many crops though hardly utilized.It is estimated by calculating distances of molecular or phenotypic data for prediction of the heterosis from field experiments (Wang et al., 1995).Reports on the extent of correlation between genetic distance and heterosis have varied for traits and studies.Qian et al. (2003) in interspecific hybrids between B. rapa and B. napus reported a larger genetic distance based on molecular marker and a higher biomass yield.Qian et al. (2007) observed a weak correlation between genetic distance and heterosis for interspecific crosses of European spring and Chinese semi winter lines.Kaur et al. (2007) observed a negative correlation between genetic diversity and hybrid performance in diverse morphotypes of B. rapa.

The possible mechanism for production of heterosis in intersubgenomic hybrids
Interaction between different genomes of Brassicas might be the reason for production of heterosis.Liu et al. (2002) revealed that some DNA fragments of A r were significantly associated with biomass production in trigenomic hybrids (A r A n C n ), but those DNA fragments had no direct relationship with the heterosis of yield.Qian et al. (2007) detected that some DNA segments that introgressed from A r had positive effects on seed yield of intersubgenomic hybrids of A r A n C n C n .Li et al. (2006a) indicated that seed yield of intersubgenomic hybrids of A r A n C c C n was positively correlated with the genomic proportion of A r , C c and A r + C c in the new-typed B. napus.The above-mentioned phenomena suggested that the increasing subgenome portion of A r and C c in the new-typed B. napus might further strengthen the intersubgenomic heterosis for seed yield.Allelic combinations present in hybrids might result in the alteration of allele expression profiles, production of novel allelic interactions, genesis of beneficial adaptations in the hybrids and give rise to heterotic phenotypes (Springer and Stupar, 2007;Chen et al., 2008).Chen et al. (2008) also found that the introgression of A r and C c subgenomes of B. rapa (ArAr) and B. carinata (B c B c C c C c ) could lead to considerable differences in the gene expression profiles of the partial new-typed B. napus (A r/n A r/n C c/n C c/n ) compared with their parents.Chen et al. (2008) considered that dominance and overdominance effects were important in the intersubgenomic hybrids (Huang et al., 2006;Ju et al., 2006).About 15.04% and 0.66% of the transcript-derived fragments (TDFs) that were differentially expressed between the intersubgenomic hybrids and their parents showed a significant correlation with at least one or over two of analyzed traits of yield.This indicated that allelic variation introduced from A r /C c subgenome may lead to many positive allelic combinations in the intersubgenomic hybrids (Chen et al., 2008).Some TDFs, such as Copia-like TDF, were activated in new-typed B. napus.It indicated that the DNA methylation and chromatin remodeling might be involved in the production of intersubgenomic heterosis (Hirochika et al., 2000;Zilberman et al., 2007;Chen et al., 2008).Further research revealed that 12 TDF-markers were mapped to 12 different linkage groups within the one DH population developed by Qiu et al. (2006).Four of these TDFs were located within the confidence intervals of eight quantitative trait loci (QTLs) for yieldrelated traits, which could explain the phenotype variation from 4.41% to 13.45% in the TN DH population.

Conclusion
Studies on combining ability for traits such as yield and other agronomic traits are available for the different Brassica species with different genetic designs.Most of the studies in Brassica napus showed significant GCA and SCA effects for yield and its component characters indicating that both additive and non-additive gene actions were important in the inheritance of these traits and their performance heterosis.Heterosis is the difference in performance between F 1 generation and mid-parent or high-parent and has been a major breeding tool for Brassica species productivity improvement.Heterosis is commercially exploited in rapeseed and its potential use has been demonstrated in turnip rape (B.rapa L.) and Indian mustard (B.juncea L.) for seed yield and most of the agronomic traits.The oilseed rape plant, B. napus, possesses two endogenous male sterile cytoplasms, nap and pol.They were used along with their respected restorer genes for commercial hybrid production.Interaction between different genomes of Brassicas might be another reason for production of heterosis, therefore the concept of subgenome was introduced to the genus Brassica.

Figure 1 .
Figure 1.U's triangle representation of the relationships between the diploid Brassica species B.nigra, B. rapa and B. oleracea and the allotetraploid species B. carinata, B. juncea and B. napus.The haploid genomes of the diploid species of B. rapa, B. nigra and B. oleracea are referred to as A, B and C, respectively.

Figure 2 .
Figure 2. The letters of subgenome of Brassicas and their relationship.
(Figure 2).B. napus was extensively cultivated in the world.The limited geographical range of B. napus and its intensive breeding have led to a relatively narrow genetic basis in this species.Many efforts have focused on exploring novel B. napus breeding stocks by the hybridization of B. rapa × B. oleracea or B. napus × B. juncea, B. carinata × B. nigra