MAPPING QTL(s) FOR RESISTANCE TO THE BEET CYST NEMATODE (Heterodera schachtii Schm.) IN RADISH (Raphanus sativus L.)

Mousa M. A. A, N. M. Kandeel, M. F. Mohamed and M. H. Aboul Nasr

Department of Horticulture, Faculty of Agriculture, University of Assiut

INTRODUCTION:

                The Beet cyst nematode (BCN), Heterodera schachtii Schm., is a serious threat in beet cultivating areas of the world. The BCNis a widespread parasite and the host-range includes many arable plant species among them horticultural crops belonging to several families (Cruciferae, e.g Brassica, Raphanus and Sinapis species; Leguminosae and Chenopodiaceae, e.g. Beta species) (Baukolh, 1976; Lange et al., 1990; Raski, 1950). In Egypt, one of the major obstacles facing the agricultural production is the damage caused by the pests including nematode (El- Sherif, 1992). The genus Heterodera is becoming an important pest in Egyptian agriculture. It has been detected in most vegetable and field crops (e.g. legume crops and maize). The virtual damage and the yield losses are not assessed. However, in some individual fields and pot experiments it was observed that the yield losses reached up to 25% (El Sherif 1992).

Oil-radish (Raphanus sativus L.) and white mustard (Sinapis alba) are considered to be useful source of the resistance to BCN (Heterodera schachtii Schm) (Baukolh, 1976). Therefore, many breeding programs have used oil radish cultivars to transfer the resistance to the commercial cultivars of the related Brassica species, e.g. oil-seed rape (Lelivelt et al., 1991; Levlivelt and Krens, 1992; Voss et al., 1999). Nevertheless, there are few number of investigations focused on the genetic behaviour of this pathogen in oil radish and white mustard. Baukolh (1976) studied the inheritance of the beet cyst nematode resistance in oil-seed radish. He reported that a major dominant gene controls the resistance trait. The resistance inheritance in white mustard (Sinapis alba) has not been described.

Recently, many PCR-based marker techniques were developed. These technologies encouraged and supported the geneticists and plant breeders to construct high-density genetic maps of different crop species. Moreover, there are numerous established studies for tagging and locating gene(s) controlling important traits of most economically important crops. Otherwise, DNA polymorphisms are now used extensively to elucidate the genetic nature of agriculturally important traits in a wide range of crops. For example, RAPDs, AFLPs were used efficiently to clarify the genetic behaviour of the fruit quality in tomato (Causse et al., 2002)

Bulked segregant analysis (BSA) is a powerful novel technique being used for identifying marker candidates linked to genes of agronomic treats (Michelmore et al., 1991). The method involves comparing two DNA pools of specific individuals from a segregating population. The number of individuals in each bulk depends on the type of marker techniques and the type of population used to construct the bulks. BSA is used successfully for identifying genes of qualitatively and quantitatively inherited traits (Ajisaka et al., 1999; Bryan et al., 2002). Our objective in the current study was to use this previously studied population to identify QTL(s) associated with the BCN resistance trait. Through the BSA we were able to identify RAPD and AFLP markers associated with the BCN resistance gene(s).

MATERIALS AND METHODS:

Plant material:

The F₂ segregating population was produced by self-pollinating single F₁ plant originating from the cross between two oil radish (Raphanus sativus L.) cultivars 'Pegletta' (P₁,resistant cultivar) and 'SilettaNova' (P₂, susceptible cultivar) (Kandeel et al., 2004).

BCN resistance testing:

                The resistance test was done as described by Toxopeus and Lubberts (1979). Seeds from the P₁ (resistant parent), P₂ (susceptible parent), F₁ hybrid, F₂ and the Brassica napus cultivar‘Madora’ populations were germinated in petri dishes on moist filter papers. Three days later, a total of 290, 30, 80, 80 and 240 from the total 720 seedlings of the F₂, F₁ hybrid, P₁, P₂ and the Brassica napus control, respectively, were transplanted individually into polyethylene folding boxes (40x20x120 mm) filled with 100 cm³ of an autoclaved soil and sand mixed medium (1:1 by volume). Soil medium was mixed by ratio of 7:1 with clay growing stone (Leni, Gebr. Lenz GmbH, 51691 Bergneustadt, Germany). Because of the restricted seeds of the F₁ hybrid, 30 seeds were sown only in the first three containers. The boxes were subsequently arranged in 8 containers and placed into two growth chambers. The growth chamber programmed to provide 16 h lighting (5000 Lux) and 8 h darkness. The lighting (day) and the darkness (night) temperatures were 22°C and 18°C, respectively.

                One week after transplanting, each plant was inoculated twice with a suspension of pre-hatched Juvenile II (J₂) larvae from Heterodera schachtii Schm. The first inoculation was performed one week after transplanting and each plant was inoculated with 1000 J₂ larvae (1 ml of the larvae suspension) on the right side of the plant place (approximately 2 cm from the plant base) at depth of 5 cm from the soil surface. Each plant was inoculated again at the next day with 1000 J₂ larvae on the left side of the plant and at the same depth. The infection was performed by 10 ml pipette.

Quantitative measurement of the visually resistance:

                About six weeks after the inoculation, the plants were transferred with soil from the boxes into dishes filled with water and the plant development was observed. Poor developed plants were excluded from further analysis. Thereafter, the soil was carefully discarded from the roots. Then the plants were transferred separately into other dishes filled with water. The number of cysts on the periphery and the main roots was counted. Otherwise, the cysts number in the water was also counted and added to the cyst number on the roots to calculate the total number of cysts per plant. A binocular microscope was used to estimate number of cysts on the roots and in the water.

                Generation means, standard error and variances were determined from individual plants within each experimental unit and pooled over replications.

Molecular marker analysis:

1- Bulked Segregant Analysis (BSA):

                Bulk segregant analysis (Michelmore et al., 1991) was used to identify RAPD and AFLP markers linked to the nematode resistance gene(s). The 24 F₂ individuals showing the lowest and highest number of cysts was selected to compose the resistant and susceptible bulks, respectively. The bulks were constructed by mixing equal DNA amount from each selected F₂ individual. The pair of bulks was screened with 60 RAPD primers and 72 AFLP EcoRI+3/MseI+3 primer combinations (PCs) to identify the candidate markers revealing clear qualitative polymorphisms (presence or absence of bands) and/or revealing quantitative polymorphisms (band intensity differences) between the bulks.

2- Map construction:

                The existed RAPD and AFLP integrated map for the same mapping population (Mousa, 2004) was used. The map was reconstructed after excluding RAPD and AFLP markers that were generated using a sub-set of the mapping population (64 F2 individuals) and/or has shown more missing genotypes. Therefore, this map consisted of 391 markers (66 RAPDs and 325 AFLPs) distributed over 9 LGs. The map covered a total of 1070 cM of the radish genome with a mean distance 2.73 cM between adjacent markers.

3- QTL analysis:

      MAPQTL 4.0 software (Van Ooijen and Maliepaard, 2001) was used to perform the quantitative trait loci analysis of the Beet cyst nematode resistance. The program uses the trait means and marker data in the non-parametric Kruskal-Wallis test (Lehmann, 1975) in order to search each locus separately for segregant QTL(s). Marker-QTL(s) associations are detected based on the differences level in average rank between the marker genotype classes. Stepwise, MAPQTL 4.0 applied Single-QTL Analysis Model using interval mapping (IM) function to the trait means and marker data. This method calculates the QTL likelihood (LOD score) for each position on the genome for the presence of a segregate QTL (Van Ooijen, 1992). The position on the target linkage group that revealed largest LOD score value is the estimated position of the QTL(s). Finally, Multiple-QTL analysis Model (MQM) (Jansen, 1994; Jansen and Stam, 1994) is applied for precisely detection putative QTL(s). MQM consists of two steps: 1) called 'restricted MQM mapping' and comprises the selection of markers to be used as cofactors and 2) called 'MQM mapping' and consists of the estimation of putative QTL(s) effects on the quantitative trait throughout the genome with correlation to the cofactor markers effects (Voorrips et al., 1997). The cofactor markers are selected based primarily on the results from Kruskal-Wallis test and the Interval mapping. It is advisable to select the markers that are located nearby the most closely linked markers to the putative QTL(s) as cofactors (Van Ooijen and Maliepaard, 2001). As a last step, we selected two markers left and right of the putative QTL position (with at least 2.0 LOD less than the highest LOD value) as support intervals. This has been done to obtain a 0.95 confidence interval around the putative QTL location.

RESULTS:

Resistance test:

                Tables (1, 2) show the reaction of the four oil radish (Raphanus sativus L.) populations P₁, P₂, F₁ and F₂ and one B.n.cv. 'Madora' (susceptible control), to the BCN parasitism. The mean number of cysts on P₁, F₁, P₂, F₂ and 'Madora' root systems was 4.12±0.534, 15.73±1.68, 135.53±7.22, 24.57±2.02 and 157.3±3.23, respectively. The number of cysts ranged from 0 to29, 56 to 349, 3 to 34, 0 to 217 and 39 to 298 for the P₁, P₂, F₁, F₂ and the B. n.cv. 'Madora', respectively.

                Frequency distributions for the number of cysts are presented in Fig. (1A, B and C). The majority of the resistant parent (Pegletta) plants (93%, 68 plants from total 73) belonged to the cysts class 0-10, 4 plants (5.5%) belonged to the cysts class 11-20 per plant and only 1 plant (1.4%) contained 29 cysts (Fig. 1A). On the contrary, 71.8 percent of the susceptible parent plants (46 plants of total 64 plants) showed more than 100 cysts/plants and 28.2 percent had from 50-100 cysts/plant. All F₁ plants were in the resistant parent classes. With respect to the distribution of the F₂ population, a bimodal distribution was observed (Fig. 1C). The results showed that 183 F₂ plants (74.7% of the total 245 individuals) had from 0 to 27 cysts/plant, 25 plants (10.2%) exhibited from 28 to 50 cysts/plant and 37 plants (15.1%) exhibited from 51 to 250 cysts/plant. Obviously, small portion of the F₂ individuals contained a cysts number comparable to the susceptible parent.

Bulked segregant analysis:

                DNA samples from the resistant parent, the susceptible parent and both the resistant and susceptible F₂ bulks were screened with 60 and 40 pre-selected RAPD and AFLP primers and PCs, respectively. The results showed that 15 RAPD and 6 AFLP markers generated polymorphic patterns differentiating the two bulks on the base of bands intensity (Fig. 2). We did not observe either RAPD or AFLP candidates that qualitatively distinguished the resistance and susceptible bulks (present in one bulk only). However, the RAPD candidate OPD-04-568 and the AFLP candidate E39M48-268 revealed the clearest band intensities differences between the two bulks, suggesting that both candidates were the nearest markers from the BCN resistance locus.

Table (1): The number of the used radish (Raphanus sativus L.) plants for different populations (P₁, P₂, F₁, F₂ and cultivar ‘Madora’) in determining the resistance to the Beet Cyst Nematode (BCN).

^a P₁= ‘Pegletta’, P₂ = ‘Siletta nova’, F₁= ‘Pegletta x Siletta nova’, F₂ = the segregating population (obtained from selfing F₁), and ‘Madora’ = Brassica napus susceptible control.

^b omitted due to poor development of root system.

Table (2): Range of cysts, standard error, average number of cysts and variance of generations P₁, P₂, F₁, F₂ and susceptible control Brassica napus cv. 'Madora'.

^a P₁= ‘Pegletta’, P₂ = ‘Siletta nova’, F₁= ‘Pegletta x Siletta nova’, F₂ = the segregating population (obtained from selfing F₁), and ‘Madora’ = Brassica napus susceptible control.

^b standard error

A

B

C

Fig. (1): The distribution of the reaction of Raphanus populations to the BCN parasitism: A) the parental 'Pegletta' and Siletta nova' cultivars, B) the F₁ hybrid from the intra-specific cross 'Pegletta x ‘Silleta nove’

 and C) the segregating F₂ population. Short arrows refer to the distribution peaks

while the long arrow points to the position of the threshold class.

Fig. (2): DNA amplification patterns obtained with RAPD primer OPD-04 on DNA samples of the resistant

and susceptible parents (P₁ and P₂), F₁ hybrid (F₁), resistant and susceptible bulks (R and S) and a sub set

of F₂ population. The locus OPD-04-568 revealed a quantitative differences (band intensity differences)

 between both bulks and segregated in the F2 population. The molecular weight is evaluated

 with the 100-bp DNA ladder from Gibco-BRL (Bethesda, MD, USA).

QTL analysis:

Mapping for target linkage group:

                For QTL analysis, only RAPD and AFLP markers that generated by the whole mapping population were used. This was done to either prevent or reduce false marker-QTL associations. The results obtained from the non-parametric Kurskal-Wallis test suggested the existence of QTLs for the resistance to Heteodera schachtii in LG6 (Table 3). The range of probability values suggested the existence of three QTLs distributed over a great number of markers on LG6. The first one is located near the AFLP marker E42M49-250, the second is located nearby the AFLP marker E41M60-484 at the centromeric region and the third appeared to be spread over a large number of markers that located at the telomeric region of LG6 starting from the AFLP marker E40M50-410 to the RAPD marker OPD-04-568. The 3 QTLs spanned more than 74% of the length of LG 6. Otherwise, 13 unmapped markers showed significant differences in the number of cysts between each marker genotypes (P<0.005). From these markers, the AFLP marker E39M50-142 and the RAPD marker OPI-17-644 showed the greatest Kruskal-Wallis values (14.41 and 19.62, respectively). These results suggested that both markers perhaps associated with the BCN resistant QTL.

Table (3): Testing the differences between marker phenotypic classes for linkage group 6 using the non-parametric Kurskal-Wallis test as a first indication for markers-QTL(s) accessions.

^a = Mean number of cysts for the resistance genotypes

^b = Mean number of cysts for susceptible genotypes; *, **, ***, ****, *****, ******, ******* significant at P<0.05, P<0.01, P<0.001, P<0.0001, P<0.00001, P<0.000001 and P<0.0000001, respectively.                                                                                                             

Mapping the nematode resistance gene (HsI^Raph):

                In order to detect more accurate and rigorous QTL(s) for the BCN resistance trait, a Single-QTL Model analysis was applied for each single locus using the Interval Mapping (IM) option of the MAPQTL 4.0 software. The interval mapping analysis is searching for markers-QTL associations based on the algorithm of odd (LOD score) values of each single locus. The results presented in Fig. (3) show that the highest LOD score values were observed only for the markers assigned to LG6. The LOD values and the explained variance of the markers on LG6 suggested the existence of single QTL for the BCN resistance. This QTL spread over great number of markers that located at one side of the chromosome. The AFLP marker E33M60-141 showed the largest LOD value (LOD=7.14) and explained 14.9% of the total phenotypic variations. On the other hand, all unmapped markers (that were suggested to associate with the BCN resistance QTL in Kurskal-Wallis test), revealed no association phenomena with the QTL of the resistance trait. Only one exception was observed for the unmapped RAPD marker OPI-17-644 which shown a LOD score value of 5.18 and explained 10.9% of the total phenotypic variance suggesting the reliable association between this marker and the BCN resistance gene.

Fig. (3): Interval mapping analysis (Single QTL model) applied on linkage groups of the Raphanus re- constructed combined map to detect markers-QTL(s) associations based on the LOD score values. X axes represent

 the map distance in centiMorgen (cM) and the Y axes represent the LOD score values.

                Multiple-QTL Model (MQM) analysis was applied to the trait of cysts count and the marker data in order to identify stable QTL for the BCN resistance. In addition, to facilitate rediscovering of other QTLs in LG6 and/or in other LGs using powerful selected cofactor markers. The MQM analysis was carried out by MAPQTL 4.0 software during two major steps. First, 2 makers were selected as a cofactors; E39M48-268(D) and OPD-04-568, which were located on LG6. Second, the LOD score values of possible QTLs throughout the genome were determined. The MQM results were highly consistent with that obtained from the IM analysis with respect to the number of discovered QTLs. As presented in Fig. (4) one major QTL located at the telomeric region of LG6 was distinguished. The target markers E39M48-268(D) and OPD-04-568 explained 48% and 62.2% of the total phenotypic variations, respectively (Table 4). Therefore, both markers were suggested to be tightly associated with the QTL of the BCN resistance in the repulsion phase. This QTL explained a total of 83.6% of the phenotypic variance.

Fig. (4): Localization of putative QTL(s) for BCN resistance gene(s) on Raphanus genome using approximate Multiple-QTL Model (MQM) analysis. OPD-04-568 and E39M48-268(D) and E40M50-410 markers at linkage group 6 are used as a cofactors. The X and Y axes represent the Map distance (cM) and the LOD score values, respectively.

Table (4): RAPD and AFLP markers detected to flank with the HsI^Raph gene for resistance to the BCN

* = Markers selected as cofactor.

^a = Total phenotypic variances explained by both markers as calculated by MapQTL program.

                On the other hand, using tow markers as cofactors permitted to discover another minor QTL at LG7 nearby the AFLP markers E41M48-260, E39M48-260 and E35M59-192 (Fig. 5). The reliability and stability of the minor QTL on LG7 were extensively investigated. The results suggested a random association between these markers and the QTL for the resistance to BCN (Table 5).

Fig. (5): Approximate Multiple-QTL Model (MQM) for analysis of target linkage group 6 to estimate the percentage

of phenotpic variations explained by the linked markers with QTL of BCN resistance trait.

OPD-04-563 and E39M48-268(D) markers were selected as cofactores.

The asterisks represented the significance level between marker

genotype classes revealed by Kruskal-Wallis test.

Table (5): Analysis of the probability of second QTL in linkage group7 detectedusing 2 cofactor markers.

^a = Mean number of cysts for the resistance genotypes.

^b =Mean number of cysts for the susceptible genotypes.

^c The homozygous genotypes for the resistance parent.

^d The heterozygous genotypes.

 ^e The homozygous genotypes for the susceptible parent.

** significant at P < 0.01

Fig. (6): Probability of second minor QTL located on linkage group7 for the BCN resistance.

DISCUSSION:

BCN resistance trait:

                The average cysts count for ‘Pegletta’ was 4.12, with a standard error of +0.53, whereas the average cysts count for ‘Silettanova’ was 135.53, with a standard error of +7.2. These results reflected the distinct genetic background between the two parents in the BCN resistance which is considered to be important criteria for gene mapping (Paterson et al., 1991). The F₁ plants were quite comparable to the resistant parent both in mean and phenotypic variance. All F₁ plants were appeared in the resistance parent classes. These results indicate the presence of dominance for the BCN resistance gene which was consistent with that reported by Baukolh (1976). The F₂ population showed a clear bimodal shape distribution, suggesting the distribution of at least one gene with major effects, rather than strict quantitative control. Moreover, the phenotypic analysis of the F₂ population indicated that the resistance to the BCN is governed by single major gene with dominance of the allele contributed by the resistance parent (Baukolh, 1976; Lelivelt and Hoogendoorn, 1993).

Efficiency of Bulked Segregant Analysis (BSA):

The resistant and susceptible bulks were constructed by mixing equal amounts of DNA from the individuals with the extreme resistant and susceptible phenotypes. The usage of BSA technique facilitates the identification of molecular markers linked to the BCN resistance gene. It permitted to remove the randomized genetic background of the unlinked loci and, therefore, allowed to detect the genomic region(s) associated with the BCN resistance (polymorphic DNA fragments between the bulks) (Mansur et al., 1993; Michelmor et al., 1991). In the present study, the resistance and the susceptible bulks were constructed from the 24 F₂ plants with the extreme phenotype of the resistant and the susceptible parents. We decided to use a large number of the F₂ individuals in each bulk to decrease the probability of false results (false linkage) that will arise from using small number of individuals. Moreover, the large number of individuals may increase the probability that the two pools will not differ for alleles other than the target once of the trait of interest (Chague et al., 1996). Nevertheless, we observed neither RAPD nor AFLP candidates clearly distinguished the two bulks (i.e. generated bands present in one bulk only). This perhaps can be attributed to two reasons: 1) one of the two DNA bulks or may be both comprised some recombinant F₂ plants (Ling et al., 2000), and 2) the distance between the RAPD and AFLP markers and the target locus was more than 25 cM, therefore, they differentiated the resistant and susceptible bulks based on the band intensity. Michelmore et al. (1991) studied the effectiveness of bulked segregant analysis to detect RAPD and RFLP markers linked within 30% window of recombination frequency. The results showed that RAPD markers were always detectable between the bulks when they were 10 cM around the target locus. In addition, RAPD markers were also detected as unequal band intensity by bulk segregant analysis when they were 30 cM around the target locus. For RFLP markers, bulked segregant analysis detected RFLP markers as unequal band intensity when they linked at 20 cM around the gene, whereas the RFLP markers linked at 30 cM around the gene have not been detected by bulked segregant analysis.

The suitable number of individuals to be used in constructing the bulks differs based on two criteria: 1) the type of marker technique (dominant or co-dominant) and 2) the mapping population (e.g., F₂, backcross, DH, RIL population) (Michelmor et al., 1991). For instance, a small number of individuals for each bulk are recommended to use when a dominant markers are used with an F₂ population. However, the frequency of false positive increases as the number of individuals used to construct the bulks decreased (Michelmor et al., 1991). Ling et al. (2000) observed that the RAPD markers OPW14800 and OPC17750 were present in the resistant bulk and absent in the susceptible bulk, when both bulks were constructed from 6 plants of each. By increasing the number of individuals to 15 plants in each bulk, OPW14800 and OPC17750 generated much weaker polymorphic bands from the susceptible bulks than that produced by the resistant bulk. They attributed these results to the recombinant individuals presented in the susceptible bulk as a consequent of increasing the number of individuals in each bulk.

                In the present study, two RAPD (OPD-04-568) and AFLP E39M48-268 markers were distinguished between the susceptible and the resistant bulks (based on the band intensity) at a fragment size of 586 and 268 bp, respectively. The results showed that both markers were co-segregated according to the expected segregation ratio of dominant monogenic inheritance. We suggest that these markers are linked to a component of the BCN resistance. In order to clarify the effects of these markers on the phenotypic variance of the resistant trait and to study precisely the genetics of the resistance, a QTL analysis was carried out using the molecular and the quantitative data.

QTL analysis:

      To our knowledge, this is the first report attempt to map the BCN resistance gene(s) in oil radish. The QTL analysis was done using the quantitative data collected from the resistance test and the combined map that re-constructed after the exclusion of the RAPD and AFLP markers with much missing genotypes. The missing genotype markers were reported to increase the false marker-QTL associations (Causse et al., 2002; Van Ooijen and Maliepaard, 2001). The non-parametric Kruskal-Wallis test was used for primarily detection of marker-QTL(s) associations. Because of the fact that Kruskal-Wallis test uses the linked and unlinked markers, it is preferable to increase the significant threshold to reduce as possible the false positive (Bryan et al., 2002; Van Ooijen and Maliepaard, 2001). Therefore, we used a significant threshold at P< 0.005 for marker-QTL associations in order to obtain a confidence level of 0.95 throughout the genome. The Kruskal-Wallis test revealed at least 3 different QTLs all are located at the arm of LG6 of the reconstructed combined map. Moreover, there were significant differences between the resistance and the susceptible genotypes of 13 unmapped markers that supported the effect of these markers in the resistance trait. The Kruskal-Wallis test is a simple helpful to screen all marker sets using the background of the quantitative data to perform a primarily knowledge about the marker-QTL(s) associations. However, the test results are unreliable and further investigations using rigorous linkage analysis systems are required, in order to obtain a confidence level of 0.95 throughout the genome.

                A more accurate QTL analysis was carried out using the Interval mapping (IM) option of the MAPQTL 4.0 software. The IM is a single-QTL model, searching each single locus (based on the LOD score values) for association with the QTL(s) of the BCN resistance. The results of the IM analysis were in contrast with the Kruskal-Wallis test. The IM analysis suggested the existence of one major QTL spread over a large number of markers assigned at the telomeric region of LG6. The greatest LOD score values were found at one arm of LG6. Moreover, only one unmapped marker (OPI-17-644) out of 13 (revealed from Kruskal-Wallis test) was observed to associate with the QTL of the BCN resistance. The MQM results were similar to that obtained from the IM mapping analysis. As presented in Fig. (IV-5) the markers mapped at the arm of LG6 showed the greatest LOD values. The LOD values of all remained linkage groups were too low in comparison to the LOD values of LG6. These results suggested the existence of one major QTL on LG6 at the telomeric region. The OPD-04-568 and E39M48-268(D) that were previously appointed out from the BSA (have shown differences on the band intensity between the resistance and the susceptible bulks) were observed to explain 48% (E39M48-268(D) and 62.2% (OPD-04-568) from the total phenotypic variance in the MQM analysis. These results reflect the contribution of BSA for mapping the BCN resistance gene (HsI^Raph). The two markers (E39M48-268(D) and OPD-04-568) that located near the major QTL showed the most significant differences between the homozygous resistance and susceptible genotypes, 16.6 and 93.8 (77.44 cysts) for E39M48-268(D) and 17.8 and 110.0 (92.2 cyst) for OPD-04-568, respectively. Moreover, E39M48-268(D) and OPD-04-568 segregated according to the expected 1:3 ratio, 65:176:4 (Pegletta : Sillta nova : missing, X² = 1.9, P>0.1) and 53:188:4 (Pegletta : Sillta nova : missing, X² = 1.7, P>0.1), respectively. These results suggested that both markers flanked the QTL of the BCN resistance in the repulsion phase. This QTL explained 83% of the total phenotypic variations.

      The stability of the minor QTL at LG7 was tested. The differences between marker genotypes were less significant than our threshold for the marker-QTL associations (P<0.005). The LOD values of E41M48-260, E39M48-260 and E35M59-1920 (the markers that associated with the minor QTL at LG7) were 0.64, 0.64 and 0.61, respectively. MQM analysis showed that the variance explained by E41M48-460, E39M48-460 and E35M59-192 were 6.6%, 6.7% and 6.8%, respectively. All these results suggested a random association. Instable QTLs were observed for the blackleg resistance in oilseed rape (Pilet et al., 2001). The authors attributed the instability of the detected QTLs to: 1) the genetic background of the crossed parent, 2) the infestation level of the pathogen and 3) the genome coverage by the constructed genetic map.

      The present study described the characterization and mapping of BCN resistance gene in oil radish. It represents a step towards isolation and integration of this gene to the economically important related crops. Two DNA markers OPD-04-568 and E39M48-268(D) flanked the resistance gene in the repulsion phase. Both markers can be usefully utilized in marker assisted selection for the BCN resistance in radish. Additional work is required to test the reproducibility of these markers by confirming these markers to a more efficient single locus PCR-based marker (e.g. SCAR markers).

REFERENCES:

Ajisaka H, Kuginuki Y, Shiratori M, Ishiguro K, Enomoto S and Hirai M (1999) Mapping of loci affecting the cultural efficiency of microspore culture of Brassica rapa L-syn. campestris L. using DNA polymorrphism. Breed Sci 49: 187-192.

Baukolh, H (1976). Untersuchung zur Wirtspflanzeneignunng der Kruziferen gegenüber dem Rübennematoden, Heterodera schachtii (Schmdt). Unter besondere berücksichtigung der Resistenzenzzüchtung. Dissertation Georg-August-Universität Göttingen. 72pp.

Bryan GJ, Mclean K, Bradshaw JE, De Jong WS, Phillips M, Castelli L and Waugh R (2002) Mapping QTLs for resistance to the cyst nematode Globodera pallida derived from the wild potato species Solanum vernei. Theor Appl Genet 105> 68-77.

Causse m, Saliba-Colombani V, Lecomte L, Duffe P, Rousselle P and Buret M (2002) QTL analysis of fruit quality in fresh market tomato: a few chromosome regions control the variation of sensory and instrumental traits. J Exper Botany, 377: 2089-2098.

Chauge V, Mercier JC, Guenard M, de Courcel A and Vedel F (1996) Identification and mapping on chromosome 9 of RAPD markers linked to sw-5 in tomato by bulked segregant analysis. Theor Appl Genet 92: 1045-1151.

El-Sherif, M (1992). Plant Nematode Problems and their Control in the Near East Region (FAO Plant Production and Protection Paper - 144). Proc Expert Consult Plant Nematode Problems and their Control in the Near East Region Karachi, Pakistan.

Jansen, RC (1994). Controling the Type I and Type II Errores in Mapping Quantative Trait Loci. Genetics 138: 871-881.

Jansen, RC and Stam P (1994). High resolution of quantative traits into multiple loci via interval mapping. Genetics 136: 1447-1455.

 Kandeel NM, Mohamed MF, Mohamed HA and Mousa MA (2004) An intra-specific genetic linkage map of oil radish (Raphanus satibvus L.) using random amplified polymorphic DNA (RAPD) markers. Proc of the Agric Develop in Arab Nations Workshop, 233-252

Lange, W, Jung Chr and Heijbroek W (1990). Transfer of beet cyst nematode resistance from Beta species of the section patellares to cultivated beet. Proc 53th Winter Congr Intern Inst Suger Beet

Lehmann, EL (1975). Nonparametrics. McGraw-Hill, New York.

Lelivelt, CLC (1991). Introduction of beet cyst nematode from Sinapis alba L. and Raphanus sativus L. into Brassica napus L. (oil-seed rape) through sexual and somatic hybridization. Ph.D. dissertation University of Wageningen, The Netherlands.

Lelivelt, CLC and Hoogendoorn J (1993). The development of juveniles of Heterodera schachtii in roots of resistant and suceptible genotypes of Sinappis alba, Brassica napus, Raphanus sativus and hybrids. Neth J Pl Path 99: 13-22.

Lelivelt, CLC and Krens FA (1992). Transfer of resistance to the beet cyst nematode (Heterodera schachtii Sch.) into the Brassica napus L. gene pool through intergeneric somatic hybridization with Raphanus sativus L. Theo Appl Genet 83: 887-894

Ling P, Duncan LW, Deng Z, Dunn D, Hu X, Huang S and Jr Gmitter FG (2000) Inheritance of citrus nematode resistance and its linkage with molecular markers. Theor Appl Genet, 100: 1010-1017.

Mansur LM, Orf J and Lark KG (1997) Molecular markers in mapping of plant genomes. Mol Biol 31: 163-171.

Michelmore, RW, Paran I and Kesseli RV (1991). Identification of markers linked to disease-resistance genes by bulked segregant analysis: A rapid method to detect markers in specitic genomic regions by using segreganting populations. Proc Natl Acad Sci USA 88: 9828-9832.

Mousa, MAA (2004) Polymerase chain reaction (PCR)-based marker analyses for the genome and nematode resistance QTL(s) of Raphanus sativus L. PhD dissertation, Assiut University, 71526 Assiut, Egypt.

Paterson, AH, Tankslley SD and Sorrells ME (1991). DNA markers in plant improvement. Adv Agron 46: 39-90.

Pilet ML, Duplan G, Archipiano M, Barret P, Barson C, Horvais R, Tanguy X, Lucas MO, Renard M and Delourme R (2001) Stability of QTL for field resistance to blackleg across two genetic backgrounds in oilseed rape. Crop Sci, 41: 197-205.

Raski, DJ (1950). The life history and morphology of the sugar beet nematode, Heterodera schachtii Schmidt. Phyatopathology 40: 135-152.

Toxopeus, H and Lubberts JH (1979). Breeding for resistance to the sugar beet nematode (Heterodera schachtii Schm.) in cruciferous crops. Pro. Eucarpia Cruciferae Conferance, Wageningen (Post Conferance Edition), p.151.

Van Ooijen, JW (1992). Accuracy of mapping quantitative trait loci in autogamous species. Theor Appl Genet 84: 803-811.

Van Ooijen, JW and Maliepaard C (2001). MapQTL version 4.0: Software for the calculation of QTL positions on genetic maps. CPRO-DELO, Wageningen. The Netherlands.

Voorrips, RE, Jongerius MC and Kanne HJ (1997). Mapping of two genes for resistance to clubroot (Plasmodiophora brassicae) in a population of double haploid lines of Brassica oleracea by means of RFLP and AFLP markers. Theo Appl Genet 94: 75-82.

Voss, A, Lühs WW, Snowdon RJ and Friedt W (1999). Development and molecular characterisation of nematode-resistant rapseed (Brassica napus L.). Dev Plant Breed 8:195-202.