PLS221                                                                          Instructor:  C.F. Quiros

Brassica crops: part 2.



Applications of maps and molecular markers in Brassica research:

There are two major applications of this technology in Brassica crops. The first one is on the study of genome structure, origin and evolution, and the second one is for gene tagging for marker assisted selection and related activities.

ORIGIN AND STRUCTURE OF THE BRASSICA GENOMES:

Early hypotheses: A common assumption has been that the n=8, 9 and 10 cultivated species have evolved in an ascending dysploid series from a common primitive genome, `Urgenome' (Haga, 1938). Although there are no known Brassica species in nature with genomes of less than n=7 chromosomes, Sikka (1940) postulated that the ancestral genome consisted of five basic chromosomes, and that Brassica developed in the direction of tetraploidy, being B. rapa (x=10) an autotetraploid. On the other hand, Catcheside (1934) and Robbelen (1960) believed that this number was actually six, postulating the following genomic formulae

The ancestral genome presumably originated the n=8, 9 and 10 chromosome genomes by hybridization and polyploidy, resulting in aneuploidy and chromosomal rearrangements. (Sikka 1940). Thus, as a corollary of this hypothesis, the cultivated diploids are considered secondary polyploids or paleopolyploids (Catcheside 1934, Prakash and Hinata, 1980) .

General attributes of the Brassica genomes: A great deal of information has been gathered during the past few years on the genomic structure of all three cultivated genomes. Analysis of synteny maps for the B and C genomes, and linkage maps for the A, B and C genomes (Truco and Quiros 1994) reveal extensive sequence duplication. Translocations are of common occurrence in Brassica and have been reported by independent investigators as a widespread event in various species (Snogerup 1980; Quiros et al. 1988, Kianian and Quiros (1992b).  In addition to duplications and linkage rearrangements, deletions seem to be another important molding force of the Brassica genomes as explained before.

 Plasticity of the Brassica genomes:  The highly duplicated nature of the Brassica genomes have important implications in structural changes of the chromosomes. The homoeologous regions arising by duplications under certain situations, such as those imposed by hybridization, will facilitate intra-genomic and inter-genomic recombination events.

1) Intragenomic homoeologous recombination: We have discussed already the high frequency of intragenomic recombination detected mostly in aneuploids, such as multiple chromosome alien addition lines and newly synthesized amphidiploids.  Undoubtedly natural amphiploids have gone thorough this same process as indicated by the comparative mapping data for the C-genome (Cheung et al. 1997a) and for the A-genome  (Hoenecke and Chyi, 1991).

 2) Intergenomic homoeologous recombination: Occasionally a few of the diploid individuals derived from alien addition lines will result from inter-genomic recombination. For example, B. rapa-oleracea monosomic addition plants were found to carry a few C-genome-specific markers present in the alien chromosomes of the parental plant, indicating that intergenomic recombination had taken place. In synthetic B. napus recombination of this type was estimated to be approximately 10%.

Genomic relationships: Comparative mapping for all three basic genomes from the cultivated diploid species is slowly progressing in Brassica. Comparative mapping for the A, B and C genomes has been reported, allowing to draw inferences on the origin of the three genomes based on their relationships (Lagercrantz and Lydiate 1996; and Truco et al. 1996). All three Brassica species share regions of homology in their genomes.  Often a single linkage group showed regions of homology with more than one group of the other species.    It is evident that extensive gene reordering has taken place during the evolution of Brassica species, even though there is considerable conservation among certain chromosome regions within and among the three genomes.  This results in complex intra and intergenomic chromosomal relationships where colinearity is maintained for some segments, but broken up for other chromosomal regions.

Comparative mapping with Arabidopsis, used  as a model for a simpler genome: Taking advantage of the fact that A. thaliana (tribe Arabideae) is a related crucifer to Brassica species (tribe Brassiceae) (Price et al. 1994), comparative mapping between the two genera is taking place at an increasing pace (for review see Paterson 1997).  A. thaliana has a small (120 Mb) and only n=5 chromosomes. Approximately 50% of the A. thaliana genome is duplicated, indicating that this species might have had also a polyploid origin. This adds unexpected complexity when comparing the Arabidopsis genome to those of the Brassica species. Kowalski at al. (1994) found extensive rearrangements between RFLP maps of A. thaliana and B. oleracea, however, islands of conserved gene organization were identified. Further, single copy genes in Arabidopsis are found in general in multiple copies in BrassicaB. oleracea BAC sequencing  is just beginning. So far it indicates that gene content in both species is similar, however, Brassica has larger intergenic spacer and introns due to insertion of transposable elements. 

On the origin of the genomes:  It is uncertain what is the number of chromosomes of the Brassica ancestral genome although it is possible that this number is 4 or 5 chromosomes. It is unlikely that the Brassica genomes originated by polysomy or duplication of whole chromosomes.  Therefore, the complexity of the existing chromosomal relationships discards the possibility of autopolyploidy  as an explanation for the higher chromosome numbers observed today in the cultivated genomes.

Lagercrantz and Lydiate et al. (1996) concluded that an hexaploid ancestral species originated all three Brassica cultivated genomes. This conclusion was based on the observation that 22 to 35% of the loci detected by the RFLP probes used were triplicated, in spite of the fact that these probes detected on the average 1.9 loci. This hypothesis makes necessary the assumption of extensive chromosome loss following polyploidization to account for the B genome chromosome number (n=8), assuming an ancestral genome of x=4. Most recent evidence for this hypothesis comes from sequencing the B. rapa genome, where it has been estimated a gene content of 52 to 53,000 genes, 2x of the number reported for A. thaliana (Mun et al 2009). To account of this discrepancy loss of duplicated redundant genes has been proposed. Synteny with A. thaliana shown for 20 blocks some including whole chromosome arms.  Alignment of orthologous segments between a trascriptome map of B. oleracea and physical map of A. thaliana fails to disclosed triplicated segments. Actually, some A. thaliana segments, for example on chromosome 1, are represented in more than 6 B. oleracea chromosomal segments.

Cyclic amphiploidy and the origin and evolution of the Brassica species:
  Sikka (1940) mentioned two main forces molding the Brassica genomes departing from an ancestral genome of n=5. The first one is hybridization and the second one is chromosomal structural changes. Spontaneous interspecific hybridization in Brassica is well documented by the existence of natural amphidiploid species, driven mainly by unreduced gametes (Harlan and deWet 1975). During the formation of new amphiploids,  crosses of sesquidiploid individuals back to their diploid parents likely have produced in nature changes in genomic numbers.  Additionally, newly formed amphidiploids may have crossed to remnant sesquidiploids.  These crossing possibilities will produce a plethora of aneuploids, some of which will have chromosome segments duplicated in multiple copies, explaining the triplication observed for some loci. Previous to undergoing these hybridization events, the ancestral Brassica genome, containing probably   x=4 or 5 chromosomes, may have derived novel cytotypes of similar chromosome numbers. These would arise as a result of chromosomal structural modifications due to differential evolutionary forces caused by spatial isolation of the species containing the ancestral genome. Most likely reciprocal translocations, which are quite common in crucifers in general and Brassica species in particular, have been responsible to these structural modifications. Therefore, based on these events that we know they occur in Brassica species, it is not necessary to invoke hexaploidy to explain the origin of the Brassica genomes.  Following Sikka’s insight, based on common events observed in Brassica species, it is proposed that  the A, B and C diploid genomes are actually partial amphiploids derived by hybridization.  Because of their ancestral common origin, these three genomes have conserved chromosome segments and extensive duplications. After genomic stabilization these have generated by another cycle of hybridization the cultivated allotetraploid species we know today. (see model in motion)  

APPLICATIONS OF THE MAPS IN BREEDING

 The Brassica linkage maps are now extensively applied to tag genes of interest, including quantitative trait loci (QTL) of economic importance.

  Vernalization requirement for flower induction:  Two genomic regions determining  biennial  habit in B. rapa have been identified, by crossing biennial to annual types. These regions associate to two linkage groups in B. napus  carrying genes related to flowering time. (Teutonico and Osborn 1995, Ferreira et al. 1995a). co  gene homologues determining flowering time in A. thaliana were detected in B. nigra by Lagercrantz et al. (1996). Camargo and Osborn (1996) performed QTL analysis  of flowering time in F3 populations B. oleracea obtained by crossing cabbage and broccoli.  A total of five QTLs were detected. One of these was linked to the S locus locus slg6. Another QTL was associated to petiole length. No conservation for flowering time QTLs were detected between B. oleracea and B. napus and B. rapa.

 Freezing tolerance and winter survival in B. napus and B. rapa: QTL analysis of these traits have been carried out in both species. In the former species, no QTLs were detected. In B. rapa, however, four QTLs were observed. Two of these linked to A. thaliana  RFLPs for cold induced (COR6.6a) and stress related (DHS2) cDNAs  (Teutonico et al. 1995).

 Linolenic acid content in B. napus: This acid by oxidations lowers the quality of the oil, so the goal is to reduce it. Markers for this trait have been identified by several laboratories. One of the four markers detected by Tanhuanpaa et al. (1995a), corresponded to one marker detected by Hu et al. (1995), which accounted for 37% of the variation for this trait.  The markers disclosed by Jourdren et al. (1996a,b) and Thormann et al. (1996) were related the  fad3 (omega 3 desaturase) gene of A. thaliana (Arondel et al. 1992).

 Oleic acid content in B. rapa: A QTL in a linkage group of six markers associated to oleic acid content was detected by Tanhuanpaa et al. (1996). OPH-17, a codominant RAPD marker was the best for selection of this trait. It was converted to a SCAR marker for better reproducibility. The oleic acid QTL also affected content of palmitic and linoleic acid content, which indicates that it controls either chain elongation or desaturation steps.

 Palmitic acid content in B. rapa:  Tanhuanpaa et al. (1995b) detected a RAPD marker associated to this trait in the same linkage group associated to oleic acid content.

 Erucic acid content in B. napus: This acid is considered an antinutrient for animals, so the goal of the breeder is to eliminate it from the seeds. All variation for this trait is explained by genomic regions (Thormann et al. 1996, Ecke et al. 1996). These correspond to two of three QTLs detected for seed oil content. Markers for loci determining this trait were reported by Jourdren et al. (1996c).

 Glucosinolate reduction in B. napus : Traditionally, breeders have tried to reduce glucosinolate in rapeseed for animal feeding, since some of these compounds such as progoitrin, are goitrogenic in animals. Reducing glucosinolate and erucic acid by Canadian breeders has resulted in “double zero” Canola (Canadian oil) varieties. Four (Uzunova et al. 1995 ) to five QTLs Toroser et al. (1995) have been detected for this trait, two of which were major.  Campos de Quiros and Mithen  (1996) integrated both studies revealing RFLPs for low (GLS-1 marker) and high (GLS-2 marker) content after screening varieties with contrasting amounts of this compound. Varieties with intermediate content displayed both markers.

Glucosinolate enhancement in B. oleracea: Although some glucosinolates are considered antinutrients, others, such as the aliphatic glucosinolate glucoraphanin, are considered desirable. This is because these compounds produce cancer-protecting agents, such as sulfurophane in broccoli and other crops. Four or five major genes have been identified in the pathway of these compounds (see section below on glucosinolates).

 Clubroot (Plasmodiophora brassicae) resistance in B. oleracea : RAPD markers associated to this trait were detected by Grandclement et al. (1996). Previous work in B. napus by Figdore  et al. (1993) had identified three  QTLs for this trait. Landry et al. (1992)  found two QTLs associated to resistance to this disease in B. oleracea.

 Blackleg resistance (Leptosphaeria maculans) in B. napus: For cotyledon and  stem resistance, two to three common QTLs detected. For field resistance two other unrelated QTLs were observed (Ferreira et al. 1995b). Dion et al. (1995) also detected two QTLs for field resistance to this disease, one of which seems to correspond to a major resistance gene named LmFr1. This gene was flanked by two RFLP markers  at 5% recombination on each side.

 White rust (Albugo candida race 2) in B. napus: A single gene responsible for resistance to this disease was mapped (Ferreira et al. 1995c).  The locus responsible,  for the resistance, named ACA1,  has been located on linkage group 4, flanked by two markers at approximately 5 cM on each side (Kole et al. 1996).

 Black rot (Xanthomonas campestris) in B. oleracea: Two QTLs were found associated to this trait by Camargo et al. (1995). One of this was also associated to petiole length.

 Fertility restorer gene for cytoplasmic male sterility in B. napus: An isozyme and a  RAPD marker were found associated to the fertility restorer gene (Rfo) for the 'Ogura' radish system by Delourme et al. (1994). Jean et al. (1997) reported a series of DNA markers linked to the 'Polima' restorer gene (Rfp) in canola, which mapped on linkage group 18.

 Morphological traits: Genes or markers for 28 traits, some of which were associated to as many as five QTLs were determined  in a B. rapa progeny of Chinese cabbage ‘Michihili x Spring broccoli (Song et al. 1995b).  The same type of study was done by Kennard et al. (1994) for 22 traits in an F2 progeny B. oleracea, resulting of crossing broccoli by cabbage. Upadhyay et al. (1996) found a marker for yellow seed coat color in B. juncea.  Markers were also detected for six quantitative traits. An RFLP for seed coat content was detected by Teutonico and Osborn (1994) and van Deynze et al. (1995) in B. napus.

 The DNA-based makers used for the creation of these maps have also been used extensively for variety identification and fingerprinting (see for example Hu and Quiros, 1991).

 The sequencing of Arabidopsis genome is providing a very important resource to Brassica geneticist who relies on the high level of identity between the genes of the species in the two related genera. The relative close taxonomic proximity of A. thaliana to Brassica makes possible to transfer effectively this information for applied and basic research projects. This capability is enhancing our knowledge on the origin and evolution of the Brassica genomes. A good example of this is the on going research on the genetics of glucosinolates in Brassica.

Genetics of glucosinolates (GSL):

A number of studies suggest that consumption of vegetables, in particular crops such as broccoli [Brassica oleracea (Italica Group)] and other crucifers, reduce the incidence of cancer in humans and other mammals (Block et al., 1992; Fahey and Talalay 1995, Prochaska et al., 1992). This seems to be due to the presence of inducers of phase II enzymes that detoxify carcinogens and mutagens in various mammalian organs (Prestera et al., 1996; Prochaska et al., 1992; Talalay et al., 1995). In broccoli, the isothiocyanate sulfuraphane, derived from the GSL glucoraphanin by the action of the enzyme myrosinase, was identified as a potent inducer of these enzymes, conferring protection against mammary tumor growth in rats after treatment with dimethyl benzanthracene, a carcinogenic agent (Zhang et al., 1992; 1994). Glucoraphanin is one of the major GSL present in some crops of B. oleracea such as broccoli (Farnham et al., 2000), cauliflower [B. oleracea (Botrytis Group)], cabbage [B. oleracea (Capitata Group)] and brussels sprouts [B. oleracea (Gemmifera Group)] (Rosa et al., 1997).

 

GSL are a diverse class of thioglucosides that are synthesized by many species of the order Capparales, including Brassica and Arabidopsis Heynh. The GSL molecule consists of two parts; a common glycone moiety and a variable aglycone side chain (Fenwick et al., 1983; Rosa et al. 1997).  The aglycone part may contain aliphatic, indolyl, or aromatic side chains and is derived from a corresponding a-amino acid. The general GSL biosynthetic pathway proposed by Underhill (1980), Larsen (1981) and (Haughn et al., 1991) considers that aliphatic GSL are derived from methionine. Genetic studies in Arabidopsis thaliana (Mithen et al., 1995; Mithen and Campos 1996) and Brassica sp. (Magrath et al. 1993; 1994) support the biochemical pathway proposed for biosynthesis of aliphatic GSL. The synthesis of these compounds is determined by a simple genetic system containing two distinct sets of genes, one determining side-chain elongation and the second one chemical modification of the side-chains. Aliphatic GSL profiles vary considerably in A. thaliana ecotypes and Brassica sp. These GSL are synthesized in the following sequence: methylsulfinylalkyl, alkenyl and hydroxy-types, which can be divided into three-carbon (3C) four-carbon (4C) and five-carbon (5C) groups based on their side-chain length.

 

 

 

Presence of a functional allele for BoGSL-ALK results in synthesis of alkenyl glucosinolates, such as sinigrin (3C) and gluconapin (4C), which can be converted by hydroxylation into the anti-nutrient progoitrin (See diagram below). On the other hand, the recessive allele BoGSL-ALK will result in the accumulation of glucoraphanin (4C), a source for the anticarcinogen sulfurophane, and glucoiberin (3C). BoGLS-ALK has been cloned (Li and Quiros 2002) and correspond to a family of genes coding for 2-oxoglutarate-dependent dioxygenases (2-ODDs) located on chromosome 4 of Arabidopsis thaliana (Hall et al. 2001). 

 

Methionine

                      ¯

                                      GSL-ELONG

                      ¯           ®             ®             ®       ¯

GSL-PRO  ®

                      ¯                                           ¯

          3-Methylthiopropyl                 4-Methylthiobutyl

          (glucoibeverin)                     (glucoerucin)

                      ¯                                           ¯                       ¬ GSL-OXID

 

          3-Methylsulphinylpropyl       4-Methylsulphinylbutyl  ®     sulfuraphane

          (glucoiberin)                          (glucoraphanin)

                      ¯                                           ¯                       ¬GSL-ALK

                                                                                          

           2-Propenyl                             3-Butenyl

          (sinigrin)                                (gluconapin)

                                                                    ¯                       ¬GSL-OH

                                                   2-Hydroxy-3-butenyl

                                                       (progoitrin)

 

Aliphatic glucosinolate pathway and the major genes involved in each step.

GSL hydrolysis

GSL can be hydrolyzed by myrosinase into isothiocyanates (ITC), nitriles and epithionitriles.  Only ITC have biological activity, therefore having high GSL does not necessarily translate into high yield of nutraceuticals, such as sulforaphane. In A. thaliana two genes have been identified controlling the GSL hydrolysis process, which is also influenced by environment. : ESP (epithiospecifier protein) is a nitrile-enhancing myrosinase cofactor and ESM1, increases yield of ITC. Homologs to these genes have been found in B. oleracea. In order to maximize GSL conversion to ITC, ESP must be non-functional and ESM1 must be functional.

Mechanisms for controlling pollination

Genic male sterility (gms): A large number of mutants causing sterility have been described in Brassica species, both dominant and recessive. Their use is limited for commercial hybrid seed production due to the difficulty in maintaining the sterile plants (Delourme and Budar, 1999).

Cytoplasmic male sterility (cms)

Several cms systems have been reported in Brassica. Most of them are based on the conventional system of cytoplasm/nuclear interaction, where:

S msms = male sterile; N msms = maintainer

S/N Ms__ = restorer

Some of these are now being used for developing hybrid cultivars of rapeseed,  which can result in up to 40% heterosis.

At the present time there is not a perfect system that could be used reliably in hybrid seed production. There are problems of instability at different environmental conditions, or poor seed set in the female lines, lack of maintainers or restorers, chlorophyl deficiencies, deformed flowers of lack of nectaries.

Types of cms

A) Autoplasmic (spontaneous on a given species)

1) pol: found in China in  Polish B. napus cv Polima, quite temperature stable but not completely. (Hansen et al 1991).

2) ctr (Bronowsky): is the most recent system found in napus. It arose in the F2 generation of triazine resistant lines Tower x Bronowski.

3) nap (SHIGA): first to be discovered in napus cytoplasm in crosses of 2 Japanese varieties, Hokuriku x Isuzu. Restorers and maintainers are available. Also reported for B. rapa.

4) Ogura and kosena in radish.

B) Alloplasmic (cytoplasmic transter by hybridization)

1) nigra  Cytoplasm of this species transferred to B. oleracea,  broccoli and then to cauliflower and other oleracea vegetables and rapeseed via backcrossing. The anthers develop as petals (petaloid sterility pp)

cms: [b]pp (b=nigra cytoplasm)

Maintainer: [c]pp (c=oleracea cytoplasm), found in kales and cabbage

Restorer: [c]PP , found in kales and cabbage

2) tour: (Anand) Derived from B. tourneforiii.  Restorers have been recently found, leading to the development of the first B. juncea F1 hybrid varieties. Also transferred to B. napus, B. rapa and B. oleracea by protoplast fusion . Restorers from nigra and juncea.

3) mur: Found after transferring napus nuclei into Diplotaxis muralis cytoplasm. It has been transferred to turnips and rapeseed. The sterility is complete and stable, but lacks maintainers.

4) ogu: This cms found in a japanese var. of radish, by Ogura. After all, the Raphanus x Brassica, hybrid "Raphanobrassica", obtained early this century by Karpechencko has been very useful for the development of cytoplasmic male sterility. F1 hybrid seed production in B. oleracea has been achieved using this system.

a) napus nucleous in radish cytoplasm: Temperature stable, maintainers are found in all napus and Japanse radish varieties), and restorers found in European radishes. Plants grown at low temperature display chlorosis and lack of nectaries. This problems were overcame by substituting napus chloroplasts by radish chloroplasts using protoplast fusion and selection. 

b) oleracea nucleous into radish cytoplasm: Introduced by Bannerot in 1977.

S(ogu)msms = male sterile

Nms ms= male fertile, maintainer (found in napus)

NMsms , restorer (from radish)

The same drawback were observed: Brassica nucleus does not function normally in radish cytoplasm. Problem with chloroplast development.

Molecular causes of cms: Mitochondrial ATP synthase gene seems to be involved. Found in most cms systems across different species. Chimeric gene formed by recombination involving ATPsynthase is the common theme. Chimeric protein accumulates only in reproductive tissue, degraded in other tissue. Cms seems to be due to poor ATPsyn activity, or chimeric protein attachement to inner mitochondrial membrane disrupting it.  Mitochondria must have important role in tapetum necessary for pollen development. Chimeric protein seems to interfere also with transcription of floral development gene resulting in floral abnormalities.

 

Restorer genes

Dominant, often monogenic.

Ogura, no restorer in Brassica only in Raphanus, Rfo restorer  introduced via raphanobrassica  to napus and by cybridization to juncea and rapa

 

Kosena system, restorers Rfk1 and Rfk2 needed  in radish but Rfk1 is enough in B. napus.

 

A few restorers cloned, code for pentatricopeptide proteins (PPR) which influence RNA processing of mitoch transcripts. Function of restorer genes cloned is unknown. Restorers seem to affect transcript profile of loci involved in cms. PPR domain seems to bind RNA and other proteins, involved in organelle gene expression. Mitochondrially targeted proteins.

 

Restortion due to inhibition of male sterility gene often at tranlation and also protein levels.  It seems to be post transcriptional control related to translation efficiency or stability of protein. Similar mechanism for other systems with other genes.

 

For F1 hybrid production Ogu system is the most used in napus and olerac (tested in rapa) because of stability and lack of secondary effects on yield etc. Used to a lesser extent are pol and Bronoski in napus.

Cybridization and cms engineering

Protoplast fusion allows mixing of cytoplasms resulting in cytoplasmic recombinants or "cybrids". Its originality resides in new genetic combinations of mitochondria and chloroplast genomes never obtained in plants before.

Pelletier et al. pioneered the use of this approach to solve the problem of cytoplasm/nucleous incompatibility between radish and Brassica to obtain desirable ogu cms lines. He fused protoplasts of ogura cms napus (radish cytoplasm) with those of B. napus rapeseed variety 'Brutor'.

Cybridization has been quite successful in Brassica. It also included the combination of cms mitochondria with atrazine resistant plastids from campestris. A series of  intra- and inter-tribal somatic hybrids have been produced in crucifers. For a recent review see Glimelius, (1999).
 

Asymmetric hybrids: The protoplasts from one species is irradiated before fusion to transfer only part of its genome to resulting hybrid.
 Forsberg and Glimelius (1995) reported the production of "asymmetric hybrids" between B. napus and Arabidopsis thaliana. The task was accomplished by irradiating A. thaliana protoplasts with X-rays or UV light, or pretreatment of these protoplasts with a restriction enzyme.
 
 

Self-incompatibility (SI)

SI was first observed in radish in 1920, studies of this phenomenon started in the 50's in other crucifers.

Studied mostly in B. oleracea and more recently in B. rapa. Allotetraploids species tend to be self-fertile.

SI in Brassica is of the sporophytic type. In other words, the incompatibility phenotype of the pollen is not determined by its own genotype but by the genotype of the pollen producing plant. Sporophytic and gametophytic SI systems are probably polyphyletic, since they are mechanistically different.

The sporophytic SI system found in crucifers is the best understood and the only one where both male and female determinants have been characterized.  This system has complex dominance relationships of SI alleles not found in the gametophytic system. Heterozygous plants can be dominant, co-dominant or intermediate in SI allele activity, which might be different in the male or female side.

SI is genetically determined by the S-locus, which is actually a multi-gene complex inherited as a single segregating unit called the S-haplotype. The S-locus is highly variable, consisting of as many as 50 haplotypes. Possible allelic interactions:

Dominance S1>S2

codominance S1=S2

Mutual weakening. No action by either allele

Intermediate gradations 0-100% activity by each allele.

The site of pollen inhibition is at the stigma, that has elongated cells called papillae. Papillae appear in the stigma 5 to 7 days before anthesis. Thus pollination at this stage, or even 2 days before anthesis, results in successful pollination. This is what is called bud pollination.

The S-locus in Brassica oleracea corresponds to an approximately 200 kb region containing several transcriptional units co-segregating with the SI phenotype. This region contains the two essential components of the system, a gene coding for a stigmatic receptor called  SRK (S-receptor kinase, plasma membrane-spanning serine/threonine receptor kinase) and a gene coding for a ligand carried by the pollen called SP11 (also called SCR, for cysteine-rich protein) (Takayama and Isogai  2003). A  third gene, SLG gene (S-locus glycoprotein), encodes for abundant glycoproteins (S-locus specific glycoproteins, SLSG) that are secreted in the papillae cell wall of the stigma. SLG and the extra cellular or S domain of SRK is highly polymorphic. Thus the S domain probably provides allelic specificity, as it binds to the pollen ligand produced by SP11, also a highly polymorphic gene, in the tapetum of the anther.


SRK determines S phenotype specificity in the stigma and SP11 in pollen.

Both genes, SRK and SLG are coordinately regulated. When self-pollination takes place, SP11 (also SCR, cyteine rich) binds the SRK receptor inducing its autophosphorylation, which triggers a yet to be characterized signaling cascade leading to self pollen rejection (Takayama and Isogai 2003). The incompatibility reaction usually occurs in less than one hour following pollination.

 slocus

 

SI phenotypes are traditionally scored by two methods: by number of seeds after pollination, or pollen tube growth by microscopy by aniline blue and epifluorescence, now it is possible by allele specific PCR reactions.

Genetic analysis has been facilitated by the finding of extensive polymorphism of both, the SLG and SRK loci. Recently cDNA sequences coding for the SLG  and SRK loci have been isolated and primers constructed, serving to determine S genotypes. After the PCR reaction the products must be subjected to restriction analysis to detect polymorphism (Brace et al 1993, Nishio et al. 1994, 1996, 1997).  

Specificity of  S haplotypes: The S haplotyples, which include alleles for more than in gene, share different degrees of similarity to each other. They have been classified in two main groups, class I and class II, on the basis of SLG and SRK similarity of their amino acid sequences.  Dominance relationships in the stigma is different and might be mediated by SRK molecule interaction and competition for effectors. 

Self-compatible lines often has non-functional SRKs. Yellow ‘sarson’, a naturally self-compatible oil seed crop of B. rapa probably is due to a mutation in a gene producing an aquaporin-like protein involved in the signaling process.

Sequencing data of SLG alleles in oleracea and campestris indicates that SLG polymorphism are ancient an likely have predated speciation in Brassica. SLG8c (camp) and SLG13o (ol) may have diverged from a common ancestral SLG gene existing before the divergence of these two species.

Transformation by SLG and SRK of the same class causes co-suppression between the transgene and endogenous SLG or SRK genes causing breakdown of incompatibility.

Depending of the horticultural type, there are different levels of self compatibility. For example, kales are highly SI, while at the other end, cauliflower is quite self compatible.

Of 1200 lines of cauliflower tested in INRA Rennes, 95% were self-compatible.

The National Vegetable Research Stn at Wellesbourne has a collection known of S alleles useful for the identification of unknown alleles.

Transformation ability: This technique has been worked out for the most important cultivated species. It is mainly based on Agrobacterium although electroporation, PEG and microinjection techniques has been also reported. Floral dipping  does not work well in Brassica as it does in Arabidopsis.

The main problem is to standardize the transformation techniques due to the large polymorphism present within and among species. There is a recent review of this technique in B. oleracea by Puddephat et al. (1996). Applications: Insect resistance, Bt effective for diamond back moth

Herbicide resistance extensively used in Canola since 1995: 75% of Canola acreage in Canada is herbicide resistance:. 40% glyphosate, 20% imidazolinone and glufosinate.

Outlook:
There is substantial interest in Brassica among biotech companies because of its amenability to practically all aspects of genetic engineering. Introduction of herbicide resistance has been one of the major goals for rapeseed, as well manipulation of  fatty oil composition. Recently an important objective is the manipulation secondary metabolites, including antioxidants, vitamins and glucosinolates.  An important development in Brassica genetics is the sequencing of the Arabidopsis genome and now the sequencing of the Brassica genome. Being these species in the same family, transfer of information form this model plant will take place very rapidly in Brassica. Routinely, gene homologs are being characterized in Brassica species based on Arabidopsis.



Back to Brassica crops, first part
Back to syllabus



Last modified: April 27, 2010

© Carlos F Quiros, 1998