Crop Science 41:1072-1079 (2001)
© 2001 Crop Science Society of America
CROP BREEDING, GENETICS & CYTOLOGY
Diallel Analysis for Tocopherol Contents in Seeds of Rapeseed
Fernando D. Goffman*,a and
Heiko C. Beckerb
a FAUBA, Institute of Industrial Crops, Av. San Martín 4453, 1417 Buenos Aires, Argentina
b Institute of Agronomy and Plant Breeding, Georg-August-Univ., Von-Siebold-Str. 8, D-37075 Göttingen, Germany
* Corresponding author (fgoffman{at}mail.agro.uba.ar)
 |
ABSTRACT
|
|---|
Because of their nutritional and antioxidative properties, tocopherol production is an interesting trait for the lipid quality of oil crops. Total tocopherol content in rapeseed (Brassica napus L.) is medium to low, and therefore, higher levels of tocopherol are desirable in this species. The objective of the present study was to determine the inheritance of 
-, 
-, and total tocopherol content and the -/-tocopherol ratio in seed of rapeseed. Two diallel mating designs with six parents each were used. In Diallel I, the parents selected were high or low for total tocopherol content and in Diallel II, the parents were high or low for the -/-tocopherol ratio. Parents and F1 hybrids were tested in a screenhouse in 1998 and under field conditions in 1999 by means of a completely randomized design with two replications. In addition, 10 selected F2 populations were grown along with their respective parents. Compared with the parents, the F1 hybrids showed a significantly higher -tocopherol content of about 6 mg kg-1 seed for Diallel I and 24 mg kg-1 seed for Diallel II. General combining ability effects in both diallels were highly significant (P < 0.01) and much larger than specific combining ability effects for all traits studied. Reciprocal effects were not statistically significant. -Tocopherol was not correlated with -tocopherol. The results indicate that tocopherol content and composition inheritances are strongly associated with additive gene action in rapeseed.
Abbreviations: GCA, general combining ability • 
ex, excitation wavelength • em, emission wavelength • MAT, maternal effect • NONM, nonmaternal effect • REC, reciprocal effect • SCA, specific combining ability • SD, standard deviation • total-T, total tocopherol content
|
INTRODUCTION
|
|---|
AN IMPORTANT GROUP of natural antioxidants with biological activity in vegetable oils is the tocopherols. They occur in four derivatives (-, ß-, -, and 
-tocopherol, the -form is known as vitamin E), differing in the methylation of the tocol head group (Larson, 1988). The main biochemical function of the tocopherols is believed to be the protection of polyunsaturated fatty acids against peroxidation (Kamal-Eldin and Appelqvist, 1996). The vitamin effectiveness of the tocopherols is very different, with -tocopherol being the most effective among all tocopherol derivatives. However, as antioxidant in vitro, the -tocopherol derivative is the most efficient of the fours with the -form being the least effective (Pongracz et al., 1995).
In rapeseed, total tocopherol content ranges from 300 to 800 mg kg-1 oil (Appelqvist, 1972; Goffman and Becker, 1998). These values are medium to low compared wtih those of other oil plants. Rapeseed oil contains, on average, 64% -tocopherol, 35% -tocopherol, and a very low percentage (<1%) of -tocopherol (Appelqvist, 1972; Goffman and Becker, 1998). The ratio of the content of - to -tocopherol can be used to describe the tocopherol composition in rapeseed. Goffman and Becker (1998) found this ratio varied from 0.32 to 1.40. Since -tocopherol exerts lower biological activity than does -tocopherol by ten-fold (Pongracz et al., 1995), an increase in the -tocopherol fraction could improve the antioxidant and vitamin activities.
To capitalize on the positive effect of tocopherols in rapeseed breeding programs, investigation of the genetic control of these compounds is required. The genetics of tocopherols was investigated in sunflower (Helianthus annuus L.) seeds by Demurin (1993), where two nonallelic genes controlling the tocopherol composition were identified. In oilseed rape, the genetic control of tocopherols is unknown. The objective of the present study was to determine the inheritance of -, -, and total tocopherol contents and the -/-tocopherol ratio in seeds of rapeseed.
|
MATERIALS AND METHODS
|
|---|
In 1997, two complete diallel crosses including reciprocals were produced using two sets of six genotypes each. In the first diallel (Diallel I), the parents selected were high or low for total tocopherol contents and in the second diallel (Diallel II), the parents were high or low for the -/-tocopherol ratio (Table 1). The hybrid seed was produced by hand pollination, and each diallel cross was carried out twice with two independent pairs of parental plants. The progenies of these two independent crosses were treated separately as replications in the screenhouse and field experiments. In 1998, the parental lines and the resulting F1 plants of the two diallels were tested in a screenhouse by means of a completely randomized design with two replications, each experimental unit consisted of four plants. F1 crosses and parents were planted in multipots and after emergence vernalized at 5 to 10°C for 8 wk. They were then transplanted to 11-cm pots in a screenhouse. Four self-pollinated plants of each F1 cross were harvested and seed samples were analyzed for tocopherols. In 1998–1999, a field experiment was conducted. The parents and the F1 hybrids were cultivated at the Reinshof experimental station near Göttingen, Germany, in plots in a completely randomized design with two replications. Each plot consisted of an 8-plant row (1.6 m). Four plants of each row were selfed and sampled for tocopherol contents. The mean values were used for the statistical analyses.
In addition, on the basis of the results of the parental analyses of the 1998 screenhouse experiment, 10 F1 plants (4 high x low total tocopherol contents, 4 high x low -/-tocopherol ratio, 1 high x high total tocopherol content and 1 high x high -/-tocopherol ratio) were selected for evaluation in the F2 generation. From each F2 population, one hundred F2 plants, along with 10 plants from each original parent, were cultivated under field conditions in 1998–1999 in single blocks (one block per F2 population along with their respective parents). Each F2 or parental plant was selfed using plastic bags and its seeds analysed for tocopherol content.
Within each diallel, parent means were compared by the Scheffe's test. An additional analysis of variance was done for comparing parents with hybrids. Griffing's (1956) method III, model I (all F1 hybrids including reciprocals without parents, with parents treated as fixed) was used to partition the sum of squares for the crosses into general (GCA) and specific (SCA) combining abilities and to ascertain their interactions with the environment. The F1 reciprocal effect (REC) was partitioned into maternal (MAT) and nonmaternal (NONM) sources according to Cockerham (1963). The analysis of variance was carried out by the DIALLEL-SAS program developed by Zhang and Kang (1997). GCA:SCA ratios with a theoretical maximum of unity were computed according to Baker (1978) as follows:
where gi = the GCA effect of the i parent and sij = the SCA effect of the Cross i x j.
Individual tocopherol contents were determined by HPLC as described by Thies (1997), using a fluorescence detector (ex = 295 nm and em = 330 nm), a C-18 diol column (250 x 3 mm I.D.), and isooctane/tert-butyl-methylether (94/6) as mobile phase at a flow rate of 0.7 mL/min. Total tocopherol content represented the sum of -, -, and -tocopherol contents. Tocopherol composition was expressed as the ratio of - to -tocopherol contents. Oil and protein contents were determined in 2 g intact seed samples by near-infrared reflectance spectroscopy (NIRS).
|
RESULTS AND DISCUSSION
|
|---|
Diallel Analysis
The parents from Diallel I differed for total tocopherol content, confirming their relative ranking presented in Table 1, in which ‘Lirabon’, ‘NPZ 04’, and ‘Samourai’ were classified as high-tocopherol and ‘Sv 0565’, ‘8980-1110/96’, and ‘H 176’ as low-tocopherol genotypes (Table 2). In Diallel II, differences for the -/-tocopherol ratio were consistently observed according to prior classification (Table 1), ‘2076-5397/94’, ‘7288-520/94’, and ‘NPZ 02’ showing an -/ tocopherol ratio above 1, and ‘H 111/2’, ‘B 005’, and ‘1485-14-21/95’ an -/ tocopherol ratio below 1. The analysis of variance revealed that crosses showed significantly higher -tocopherol contents than parents in both diallels (on average, 6 mg kg-1 seed for Diallel I and 24 mg kg-1 seed for Diallel II) (Table 2). That indicates the presence of some nonadditive genetic effects for this trait.
View this table:
[in this window]
[in a new window]
|
Table 2. Means for tocopherol traits of the rapeseed parents used for the diallel experiments and comparison between parents and F1 hybrids means, averaged over two environments.
|
|
Large environmental effects were observed for all traits in both diallels (Table 3), which can be explained by the extremely different environmental conditions in which the experiments were conducted (screenhouse vs. field). Variation in GCA was highly significant (P < 0.01) for all characters in both diallels, indicating that additive genetic effects are of major importance in the inheritance of both tocopherol content and composition. However, since the parents were not randomly selected, but rather selected for extreme values of total tocopherol content and composition, the relative amount of GCA variance may have been overestimated. Significant SCA effects were detected for -tocopherol content in both diallels and for total tocopherol and the -/-tocopherol ratio in Diallel II. REC estimates were only significant in Diallel I for -tocopherol and the -/-tocopherol ratio. For both variables, the REC effects were due to MAT and the NONM effects were not significant. Interactions of F1 hybrids and GCA effects with the environment were also significant in both diallels. Hence, genotypic response regarding the syntheses of - and -tocopherol was dependent on environmental conditions.
View this table:
[in this window]
[in a new window]
|
Table 3. Mean squares of the analysis of variance for alpha-, gamma- and total tocopherol contents (mg kg-1 seed) and the alpha-/gamma-tocopherol ratio for two diallels among rapeseed parents.
|
|
In Diallel I, the GCA effects revealed that the parents behaved genetically as expected, with the high-tocopherol parents exhibiting large positive GCA effects for total tocopherol and the low-tocopherol parents exhibiting negative GCA effects (Table 4). Similar results regarding the -/-tocopherol ratio were observed in Diallel II, where the parents with high -/-tocopherol ratio showed positive GCA effects and the ones with low -/-tocopherol ratio negative GCA effects. Both negative and positive GCA effects for each tocopherol-derivative are of interest since they may allow the development of lines containing different combinations of tocopherol content and composition. The GCA:SCA ratios near the theoretical maximum of unity, except for total tocopherol content in Diallel II, provided further evidence that tocopherol content and composition are predominantly controlled by loci with additive gene action.
View this table:
[in this window]
[in a new window]
|
Table 4. General combining ability effects with standard error (SE) for alpha-, gamma- and total tocopherol contents and the alpha-/gamma-tocopherol ratio for two diallels among rapeseed parents.
|
|
In both diallels, only few crosses exhibited significant SCA effects, which were mainly observed for -tocopherol and, therefore, for total tocopherol. The cross 1485-14-21 x H 111/2 showed a highly negative SCA effect, whereas the cross H 111/2 x B 005 showed a highly positive SCA effect for - and -tocopherol (data not shown). The results indicate that non-additive gene effects may be important for -tocopherol, particularly in some cross combinations.
GCA effects for -tocopherol were highly negatively correlated with GCA effects for -tocopherol in Diallel II material, but not in Diallel I (Table 5), where correlation was even positive but not statistically significant. These opposite correlations in both diallels are due to the different signs for GCA values for genotypes in each diallel. In Diallel I, plants exhibiting positive GCA effects for -tocopherol, also showed positive GCA effects for -tocopherol, and vice versa, except in the line 8980-1110/96, which showed positive GCA effects for -tocopherol, but negative for -tocopherol (Table 4). Contrarily, in Diallel II, each genotype exhibited either negative GCA effects for -tocopherol and positive GCA effects for -tocopherol, or vice versa. GCA effects for -, -, and total tocopherol, and for the -/-tocopherol ratio were highly correlated in both diallels with mean -, -, and total tocopherol contents, and the /-tocopherol ratio, respectively. These results indicate that GCA effects for tocopherol traits can be predicted from their respective means. While -tocopherol GCA effects were positively correlated with GCA effects for -/-tocopherol ratio, those for -tocopherol were negatively correlated in Diallel II in both cases. A highly significant correlation was detected between GCA effects for -tocopherol and mean -/-tocopherol ratio (opposite in sign to that found for -tocopherol), suggesting that if -tocopherol content is increased, its proportion in total tocopherol will also be increased. GCA effects for total tocopherol were correlated with mean total tocopherol in Diallel I but not in Diallel II, which indicate that predictability of total tocopherol content is genotype dependent. Compared with -tocopherol, mean -tocopherol was better correlated with mean total tocopherol, indicating that an increase in total tocopherol content could be easier achieved by increasing -tocopherol contents.
F2 Distribution
In general, F2 plants showed a continuous distribution for total tocopherol content with no segregation into discrete classes (Fig. 1 and 2). The number of F2 plants equal to the total tocopherol levels of the parents was high, which suggests that total tocopherol contents are controlled by a few genes. The parents 8990-1110/96 and H 176 displayed higher total tocopherol contents with larger confidence intervals in the evaluation of the F2 populations presented in Fig. 1B, and Fig. 2B, than in that of the F2 populations presented in Fig. 1A, and Fig. 2A. A failure in seed production of these parents from those evaluations probably caused the differences for this trait. Total tocopherol content for 8990-1110/96 and H 176 was overestimated in these evaluations. A continuous distribution was observed for high x high total tocopherol with a high number of F2 plants falling into the parent's range (Fig. 3), which indicates that the characters are controlled by a few genes. All F2 populations appeared to be normally distributed (Fig. 4 and 5), with the exception of that derived from the cross 7288-520/95 x B 005 (Fig. 4B). The distribution of the cross combination high x low -/-tocopherol ratio was continuous and no discrete classes were observed, showing a large number of plants being between both parents (Fig. 6). A more detailed analysis of the distribution of the F2 population derived from the cross 7288-520/95 x B 005 was done by representing its distributions for - and -tocopherol (Fig. 7). Alpha-tocopherol appears to be distributed normally, but -tocopherol distribution is truncated, with about 70% of the plants showing values above the midparent's mean. Although the F1 from this cross did not show significant SCA effect for -tocopherol, F2 data suggest that dominance effects were present for -tocopherol (Fig. 7).
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 1. Frequency distribution of total tocopherol contents in the F2 rapeseed populations of the crosses: (A) NPZ 04 x 8980-1110/96, (B) Samourai x 8980-1110/96. The 95% confidence intervals are represented for each parent.
|
|
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 2. Frequency distribution of total tocopherol contents in the F2 rapeseed populations of the crosses: (A) NPZ 04 x H 176, and (B) Samourai x H 176. The 95% confidence intervals are represented for each parent.
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 3. Frequency distribution of total tocopherol contents in the F2 rapeseed population of NPZ 04 x Samourai. The 95% confidence intervals are represented for each parent.
|
|
In sunflower seeds, two unlinked genes controlling the tocopherol composition, designated tph-1 and tph-2, were identified by Demurin (1993). The tph-1 gene controls the ratio of - and ß-tocopherol, whereas the tph-2 controls that of - and -tocopherol. In addition, by overexpressing the -tocopherol methyltransferase in Arabidopsis thaliana, an enzyme which catalizes the methylation of - into -tocopherol (Shintani and DellaPena, 1998), the proportion of -tocopherol in seeds was ninefold higher than that in the wild type, with no changes in total tocopherol contents. That suggests that only one gene controls the -/ tocopherol ratio in Arabidopsis. Our results are, therefore, in good agreement with Demurin (1993), and Shintani and DellaPena (1998), indicating that few genes regulate tocopherol composition in seeds.
Alpha- and gamma-tocopherol were not correlated in eight of 10 F2 populations (Table 6). They were positively correlated in NPZ 04 x Samourai and negatively correlated in 7288-520/94 x B 005. However, a negative correlation between both tocopherols was expected because -tocopherol has been identified as a precursor in the synthesis of -tocopherol (Soll and Schultz, 1980). But investigations about the biosynthetic pathway of tocopherols were only carried out in chloroplasts and not in seeds. The absence of correlation between - and -tocopherol contents suggests the possibility of breeding for higher - or -tocopherol content without expecting a decrease in the other tocopherol-derivative. Total tocopherol content was significantly negatively correlated with the -/-tocopherol ratio in six of the 10 F2 populations. In general, phenotypic correlations in the F2 populations agree with those given in Table 5.
Oil was not significantly correlated (P < 0.05) with -tocopherol content, with one exception (Table 7). In contrast, -tocopherol content was positively correlated (P < 0.01) with oil in five of 10 F2 populations, and positively but not significantly correlated in four of the five remaining crosses. That indicates that if tocopherol contents were increased by selection, no negative effects on oil content would be expected. The positive correlation between oil and total-tocopherol contents and the negative correlation between oil and -/-tocopherol ratio appear to be a consequence of the positive relationship observed between oil and -tocopherol. Because oil and protein were negatively correlated (P < 0.05) in all populations, the correlation coefficients between tocopherol traits and protein contents were similar but opposite in sign to those found with oil contents.
|
CONCLUSIONS
|
|---|
Both tocopherol content and composition are primarily controlled by additive genetic effects. Inferences must be limited to the respective populations of the six-parents diallel experiments, because parents were not selected at random. The large magnitude of GCA effects compared wtih the SCA effects strongly indicates that additive gene action is the primary determinant of tocopherol content and composition in this sample of parents. The small variation due to reciprocal effects indicates that tocopherol content and composition are controlled by nuclear genes. In addition, since the crosses involved were high and low for total and individual tocopherol contents and -/-tocopherol ratio, normal distributions in the F2 was further evidence for additive gene action for these traits.
The range of variation for tocopherol content and composition observed in this study does not necessarily represent the extremes. Other genotypes may contain higher or lower levels of tocopherols, which might be used in a breeding program. Molecular techniques such as those used by Shintani and DellaPena (1998) in Arabidopsis could be successfully applied in rapeseed to change its tocopherol composition. But an increase in total-tocopherol content attained by means of these techniques must be aimed at overexpressing or modifying enzymes involved in the synthesis of tocopherol precursors like homogentisate and phytylpyrophosphate. Breeding systems, such as simple recurrent selection, which make use of additive genetic variance would be effective for increasing -, -, or total tocopherol content or for modification of tocopherol composition in oilseed rape in a population derived from the material examined.
|
ACKNOWLEDGMENTS
|
|---|
The authors thank Michaela Grote, Rebecca Schütte, and Uwe Ammermann for excellent technical assistance. Financial support was provided by Niedersächsisches Ministerium für Wissenschaft und Kultur, Forschungsstelle für Biologische Rohstoffe, Göttingen. F.D. Goffman was funded by Deutscher Akademischer Austauschdienst (DAAD, Germany).
|
NOTES
|
|---|
Part of a dissertation submitted by F.D. Goffman in partial fulfillment of the requirements for a Ph.D.
Received for publication April 10, 2000.
|
REFERENCES
|
|---|
- Appelqvist, L.Å. 1972. Other lipids. p. 145–147. In Rapeseed, cultivation, composition, processing and utilization. Elsevier Publishing Co., Amsterdam, the Netherlands.
- Baker, R.J. 1978. Issues in diallel analysis. Crop. Sci. 18:533–536.[Abstract/Free Full Text]
- Cockerham, C.C. 1963. Estimation of genetic variances. p. 53–94. In W.D. Hanson and H.F. Robinson (ed.) Statistical genetics and plant breeding. Nat. Acad. Sci. Nat. Counc. Publ. 982. Nat. Acad. Sci., Washington, DC.
- Demurin, Y. 1993. Genetic variability of tocopherol composition in sunflower seeds. Helia 16:59–62.
- Goffman, F.D., and H.C. Becker, 1998. Phenotypic variability for tocopherol content and composition in seeds of winter rapeseed (Brassica napus. L) (In German). Vortr. Pflanzenzüchtung 42:105–106.
- Griffing, B. 1956. Concept of general and specific combining ability in relation to diallel crossing systems. Aust. J. Biol. Sci. 9:463–493.
- Kamal-Eldin, A., and L.Å. Appelqvist. 1996. The chemistry and antioxidant properties of tocopherols and tocotrienols. Lipids 31:671–701.[ISI][Medline]
- Larson, R. 1988. The antioxidants in higher plants. Phytochem. 27:969–978.
- Pongracz, G., H. Weiser, and D. Matzinger. 1995. Tocopherole- Antioxidantien der Natur. Fat Sci. Technol. 97:90–104.
- Shintani, D., and D. DellaPena. 1998. Elevating the vitamin E content of plants through metabolic engineering. Science 282:2098–2100.[Abstract/Free Full Text]
- Soll, J., and G. Schultz. 1980. 2-Methyl-6-phytylquinol and 2,3-dimethyl-5-phytylquinol as precursors of tocopherol synthesis in spinach chloroplasts. Phytochem. 19:215–218.
- Thies, W., 1997. Entwicklung von Ausgangsmaterial mit erhöhten alpha- oder gamma-Tocopherol-Gehalten im Samenöl für die Körnerraps-Züchtung. I. Quantitative Bestimmung der Tocopherole durch HPLC. Angew. Bot. 71:62–67.
- Zhang, Y., and M.S. Kang. 1997. Diallel-SAS: A SAS program for Griffing's diallel analyses. Agron. J. 89:176–182.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
Y. Zhang, M. S. Kang, and K. R. Lamkey
DIALLEL-SAS05: A Comprehensive Program for Griffing's and Gardner-Eberhart Analyses
Agron. J.,
June 17, 2005;
97(4):
1097 - 1106.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
V. Marwede, A. Schierholt, C. Mollers, and H. C. Becker
Genotype x Environment Interactions and Heritability of Tocopherol Contents in Canola
Crop Sci.,
May 1, 2004;
44(3):
728 - 731.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
among general combining ability (GCA) effects for various traits and means for 
and the