Nondestructive Assessment of Sinapic Acid Esters in Brassica Species: II. Evaluation of Germplasm and Identification of Phenotypes with Reduced Levels.
The concentration of phenolic compounds in the major oilseed crops of the genus Brassica, i.e., canola-rapeseed (B. napus L. and B. rapa L.) and oriental mustard [B. juncea (L.) Czern.], is much higher than in other oilseeds, accounting for about 30 times the amount found in soybean [Glycine max (L.) Merr.] (Shahidi and Naczk, 1992). The presence of phenolic compounds is one of the principal factors currently limiting the use of canola meal as a source of food-grade protein (Kozlowska et al., 1990).
Esterified phenolic acids (phenolic esters) are the predominant form of phenolic compounds in canola meal. Among them, the choline ester of sinapic acid (sinapine) is the most abundant, both in canola (Kozlowska et al., 1990) and in other Brassica spp. (Bouchereau et al., 1991). There are several studies on the variability of sinapine and other phenolic compounds in the seed of Brassica spp. (e.g., Kerber and Buchloh, 1980; Kozlowska et al., 1983; Bouchereau et al., 1991; Thies, 1994: Matthaus, 1997), although an overall characterization of existing variability for these compounds within the genus has not been performed. Furthermore, existing information on different species is scattered across the literature, and the use of different methods and different units of measurement makes quantitative comparisons difficult. According to the review of Shahidi and Naczk (1992), the total content of phenolic acids in rapeseed flour ranges from 6.2 to 12.8 g [kg.sup.-1] flour (dry weight basis). Bouchereau et al. (1991) analyzed samples of six cultivated species of Brassica and found a range from 14.9 [micro] mol eq sinapine [g.sup.-1] dry seed in B. juncea to 70.8 [micro] mol [g.sup.-1] in B. oleracea, with the average content in B. napus being around 35 [micro] mol [g.sup.-1] (range from 24.9-47.4). These values indicate a large variation within the genus for SAE content.
Breeding for reduced levels of phenolic compounds in canola varieties has been rarely attempted. The main reasons have been the lack of variability for reduced levels and the lack of adequate methods for large-scale screenings of breeding material and germplasm accessions. Kraling et al. (1991) reported sinapic acid esters (SAE) contents in breeding material of B. napus of 17.8 to 71.9 [micro]mol [g.sup.-1] of defatted meal. Their results demonstrated that breeding lines selected for various traits may contain wide variability for phenolic compounds.
In a companion paper (Velasco et al., 1998), we reported the development of an accurate near infrared reflectance spectroscopy (NIRS) calibration equation to measure the total content of SAE in intact seed of a wide range of Brassica spp. The use of NIRS allows a fast and nondestructive determination of SAE content, which is simultaneous to the analysis of other seed quality traits such as oil, protein and glucosinolate contents and the fatty acid composition of the seed oil. The objective of this study was to use the new methodology to screen a large number of B. napus breeding lines and a germplasm collection of 21 Brassica spp. for SAE content, and to identify genotypes with reduced levels of these compounds.
MATERIALS AND METHODS
The germplasm collection consisted of 1487 accessions from 21 species of Brassica, which was collected from a large number of research institutions worldwide. The samples were permanently stored at 4 [degrees] C.
A total of 1361 samples of winter B. napus were evaluated. Each sample corresponded to a single plant, selected from different breeding programs at the Institute for Agronomy and Plant Breeding, Gottingen, Germany. Most of the programs involved modifications in the fatty acid composition of the seed oil. Some mutant lines for high oleic (Rucker and Robbelen, 1995) and low linolenic acid content (Robbelen and Nitsch, 1975) were included. These samples represented the range of breeding materials at the Institute. Samples from plants grown at Gottingen from 1993-1994 to 1996-1997 were included.
A set of 115 samples of B. carinata derived from a mutation breeding program (Velasco et al., 1997) was used, because of its large variation for glucosinolate, oil, and protein contents, to study the relationship between SAE content and these seed quality traits. The seed corresponded to the M6 generation, with the M5 plants grown at Cordoba, Spain in 1995-1996 under severe drought conditions. The samples were analyzed for SAE content by the reference method used in this study (Thies, 1991), for glucosinolate content by high performance liquid chromatography (Kraling et al., 1990), and for oil and protein content by NIRS (Velasco et al., 1997).
A total of 196 samples from the germplasm collection were selected, analyzed for SAE content by the reference method, and used in the development of an NIRS calibration equation for SAE content, as described by Velasco et al. (1998). All the samples from the breeding material and all the samples from the germplasm collection were analyzed for SAE content by NIRS, by applying the mentioned calibration equation.
After these analyses, samples with reduced SAE content from the germplasm collection ([is less than] 7 g [kg.sup.-1], n = 112) and from breeding lines ([is less than] 8.5 g [kg.sup.-1], n = 75) were selected. These samples were analyzed for SAE content by the reference method to check and improve the accuracy of NIRS, as described by Velasco et al. (1998). All the SAE values reported in this paper were obtained through NIRS analysis and the aforementioned calibration equation. Data from the analyses by the reference method are given in only a few cases and are indicated in the text. The SAE content is expressed as grams per kilogram seed.
RESULTS AND DISCUSSION
Evaluation of the Germplasm Collection
The range of SAE content within the Brassica collection was very wide, from 1.7 g [kg.sup.-1] in a sample of B. tournefortii to 15.5 g [kg.sup.-1] in a sample of B. carinata (Table 1). Four species clearly showed lower SAE contents: B. tournefortii, B. souliei, B. oxyrrhina, and B. barrelieri, although the number of accessions available from the latter three was rather low (n = 2-5). In contrast, 31 accessions of B. tournefortii were analyzed, with an average SAE content of 4.1 g [kg.sup.-1] compared with the collection average of 9.4 g [kg.sup.-1]. Five entries of this species showed a SAE content below 3.0 g [kg.sup.-1].
Table 1. Total sinapic acid esters content in a germplasm collection consisting of 1487 samples of 21 species of Brassica.
Species n Mean Minimum Maximum g [kg.sup.-1] B. x amarifolia Narain et Prakash 3 6.7 6.2 7.2 B. balearica Pers. 3 8.6 5.6 10.7 B. barrelieri (L.) Janka 4 5.7 4.0 7.8 B. carinata A. Braun 292 12.5 6.8 15.5 B. elongata Ehrh. 9 8.9 6.3 10.4 B. fruticulosa Cyr. 15 7.8 5.6 9.6 B. gravinae Tenore 5 7.2 6.3 8.5 B. incana Tenore 4 8.3 7.1 9.5 B. juncea (L.) Czern. 250 8.1 4.5 12.9 B. maurorum Durieu 1 6.9 B. montana Pourret 2 9.3 8.9 9.8 B. napus L. 90 8.0 5.0 11.1 B. nigra (L.) Koch 128 8.9 4.9 11.2 B. oleracea L. 339 10.0 6.1 13.7 B. oxyrrhina Coss. 5 5.0 3.0 6.8 B. purpuraria (Bailey) Bailey 4 7.1 4.8 8.9 B. rapa L. 298 8.2 4.0 11.1 B. rupestris Raf. 1 9.2 B. souliei (Batt.) Batt. 2 5.0 4.3 5.7 B. tournefortii Gouan 31 4.1 1.7 8.0 B. villosa Biv. 1 10.6
The species B. tournefortii, B. oxyrrhina, and B. barrelieri originated from similar native habitats, i.e., sandy soils of coastal regions (Tsunoda, 1980). Bouchereau et al. (1991), analyzing different genera of Brassicaceae, found the lowest levels of choline esters in species having sand dunes as native habitat, such as Crambe maritima L., Diplotaxis tenuifolia (L.) DC., and Matthiola sinuata (L.) R.Br. Such associations between native habitat and phenolic contents deserve further investigation to establish habitat as a possible predictor of low SAE content in wild germplasm.
The highest average and maximum SAE contents were found in B. carinata, with an average value of 12.5 g [kg.sup.-1] (n = 292). Bouchereau et al. (1991) reported very low variability in this species, which they attributed to the lack of wild forms and the restricted area of cultivation. The low number of accessions analyzed may have contributed to the narrow variability. Our germplasm collection revealed a wide range of variation, from 6.8 to 15.5 g [kg.sup.-1]. Unfortunately this range did not include entries with low levels of SAE. B. carinata is an amphidiploid species originated from the diploid ancestors B. nigra and B. oleracea (Hemingway, 1979). Previous studies have reported very high SAE content in B. oleracea (Lamour et al., 1987; Bouchereau et al., 1991), which is confirmed in this study. Intraspecific differences at the subspecies, convariety or variety level were scarcely significant in all the species, and therefore a more detailed taxonomic description of the results is not reported.
A total of 112 samples from 18 species with SAE content below 7.0 g [kg.sup.-1] were selected and analyzed by the reference method. The SAE content in these samples ranged from 1.8 to 7.7 g [kg.sup.-1] (reference method), confirming the low levels of the accessions and also the accuracy of NIRS in the analysis of this trait.
Selection for Reduced SAE Content in Breeding Material of B. napus
The SAE content of the breeding lines of B. napus ranged from 5.0 to 17.7 g [kg.sup.-1]. This range was considerably greater than that found in germplasm material of this species (5.0-11.1 g [kg.sup.-1]). Such a difference could simply be caused by the higher number of samples analyzed from breeding lines, 1361 vs. 90. The lowest levels of SAE found in the breeding material were similar to those found for B. napus in the germplasm collection (about 5.0 g [kg.sup.-1]), and not very different from those reported by Shahidi and Naczk (1992) (6.2 g [kg.sup.-1], dry weight basis).
Figure 1a shows the SAE content (NIRS values) of the set of 1361 samples from breeding lines of B. napus. A total of 75 of them having SAE content below 8.5 g [kg.sup.-1] were selected and analyzed by the reference method to confirm the accuracy of NIRS. The results of these analyses are shown in Fig. 1b. The SAE content in the selected samples ranged from 4.9 to 9.4 g [kg.sup.-1], demonstrating the high reliability of NIRS analyses to perform a selection for reduced levels of this trait.
[Figure 1 ILLUSTRATION OMITTED]
Correlation of SAE Content with Other Traits
An important consideration while selecting for reduced levels of an antiquality trait is simultaneous variation in economically desirable traits. A set of 115 samples of B. carinata was selected because it included great variability for SAE content and other seed quality traits (oil, protein, and glucosinolate contents). Furthermore, all the plants derived from the same line after mutagenesis and selection (Velasco et al., 1997), and had been grown under the same environment, which eliminates the potential influence of the genetic background or the environmental conditions on the correlation coefficients. The SAE content was positively correlated with oil content and negatively correlated with protein and glucosinolate contents (Table 2). The coefficients of correlation were highly significant in all cases.
Table 2. Correlation coefficients among seed quality traits in a set of 115 samples of Brassica carinata. All coefficients are significant at P < 0.017.
SAE Oil Protein Glucosinolates SAE 0.68 -0.70 -0.51 Oil -0.92 -0.66 Protein 0.75 g [kg.sup.-1] [Micro] mol [g.sup.-1] Range 10.7-18.9 196-471 232-482 57.5-193.4
([dagger]) SAE (sinapic acid esters) content data are from the reference method (Thies 1991). Oil and protein contents were determined by NIRS. Glucosinolate content was analyzed by HPLC.
The set of samples of B. carinata used in the correlation study showed very high SAE contents, ranging from 10.7 to 18.9 g [kg.sup.-1] (Table 2). About one half of the samples had higher SAE contents than the maximum for this species found in the germplasm collection (Table 1). These high values could have been caused by environmental factors since the plants were cultivated under severe drought conditions. Bouchereau et al. (1996) found a significant increase of sinapine content in B. napus when subjected to water deficit stress during the reproductive period.
The correlation coefficients among seed quality traits were also studied in the set of 1361 samples from breeding lines of B. napus. In this case, the values of the four traits investigated (SAE, oil, protein and glucosinolate content) were obtained through NIRS. Correlation results obtained with B. carinata were confirmed by the significant positive correlation found between SAE and oil contents, and the negative correlation between SAE and protein contents (Table 3). However, there was no correlation between SAE and glucosinolate contents. The different results for this correlation could be caused by the different ranges for glucosinolate content.
Table 3. Correlation coefficients among seed quality traits in a set of 1361 samples of Brassica napus. All the traits were determined by near infrared reflectance spectroscopy (NIRS)([dagger]).
SAE Oil Protein Glucosinolates SAE 0.30(**) -0.18(**) -0.01(ns) Oil -0.68(**) -0.23(**) Protein 0.32(**) g [kg.sup.-1] [Micro] mol [g.sup.-1] Range 5.0-17.7 295-474 158-294 1.6-133.4
([dagger]) (*) P < 0.05; (**) P < 0.01; (ns), not significant.
The positive correlation between SAE and oil contents may explain the higher maximum values of SAE content in breeding material of B. napus than those found in germplasm material. The maximum oil content in samples of B. napus from the germplasm collection was 427 g [kg.sup.-1], compared with a maximum of 471 g [kg.sup.-1] in the breeding material. These values were obtained by applying the NIRS calibration equation for oil content in B. napus currently used at the institute.
The evaluation of the germplasm collection showed wide variation for the oilseed species of the genus (B. carinata, B. juncea, B. napus, and B. rapa). In the three latter species, entries with SAE content between 4.0 and 5.0 g [kg.sup.-1] were identified by NIRS and confirmed by the reference method. Most of the accessions of these species are landraces, old cultivars or wild material, on which selection for SAE content has not ever been performed. Further variability within each accession may be expected, and simple mass selection might lead to the isolation of genotypes of these species with lower SAE content.
Similar reduced levels of SAE were found in the germplasm collection and the breeding material of B. napus. Although selection from both sources will have to be performed over several generations under the same environment to check the real potential for reduced SAE content, some of the entries of the breeding material showed clear advantages. They were characterized by reduced glucosinolate and no erucic acid. For similar reduced levels of SAE, the use of this material would permit a rapid incorporation of reduced SAE content into canola-quality commercial varieties.
The significant correlations above mentioned suggest that a selection for reduced SAE content would reduce oil and increase glucosinolate content. The increase of oil content, because its economic value, and the reduction of glucosinolate content, because of their antinutritive effects, are important objectives in Brassica oilseed breeding. Therefore, indirect changes in these traits associated with selection for reduced SAE content should be avoided by simultaneous selection.
The lowest levels of SAE content found in B. carinata were considerably higher than those found in the other three Brassica species. B. carinata possesses an inherent capability to synthesize high levels of SAE, which might contribute to its exceptional resistance to a wide range of pests and diseases (Malik, 1990). Genotypes of this species with low glucosinolate content have not been developed (Getinet et al., 1997). Considering the negative correlation between SAE and glucosinolate contents within B. carinata (Table 2), an additional increase of SAE content is foreseeable during the selection for reduced levels of glucosinolates. It will be therefore necessary to perform a simultaneous selection for reduced levels of both traits.
The low levels of SAE found in some species, especially B. tournefortii, open up the possibility of reducing these compounds in cultivated species through interspecific hybridization and backcrossing. The best entry of this species showed a SAE content of 1.7 g [kg.sup.-1] by NIRS and 1.8 g [kg.sup.-1] by the reference method. In as much as the entries of this species were derived from wild material collected in diverse habitats, the determination of genetic, as opposed to environmental, control over the expression of SAE content becomes paramount. Understanding the genetic basis for low SAE content would allow one to predict the gains to be made by using B. tournefortii to improve meal quality of cultivated Brassica oilseeds by simple interspecific crossing (Nagpal et al., 1996).
Although most of the species examined showed a wide variation for SAE content, significantly lower levels were found in some wild species, especially B. tournefortii. The use of this species in interspecific breeding for reduced
SAE content in cultivated Brassica oilseeds appears to be the most efficient strategy to reduce these compounds. Further intraspecific variability in selected germplasm entries of cultivated species need to be examined. Simultaneous selection for low SAE, low glucosinolate and high oil contents should be performed to maintain previous gains made in seed quality traits. Simultaneous analysis of all these traits by NIRS is of great value for plant breeders in making further genetic improvement in Brassica oilseed quality.
Abbreviations: NIRS, near infrared reflectance spectroscopy: SAE, sinapic acid esters.
The authors are grateful to Dr. A. De Haro, Institute for Sustainable Agriculture of Cordoba, Spain, for providing, a set of samples of B. carinata, and to Christine Reuter and Uwe Ammermann for skillful assistance in the analyses by the reference method. The first author received a grant from the Direccion General de Ensenanza Superior (Ministerio de Educacion y Cultura, Spain).
Bouchereau, A., J. Hamelin, I. Lamour, M. Renard, and F. Larher. 1991. Distribution of sinapine and related compounds in seeds of Brassica and allied genera. Phytochemistry 30:1873-1881.
Bouchereau, A., N. Clossais-Besnard, A. Bensaoud, L. Leport, and M. Renard. 1996. Water stress effects on rapeseed quality. European J. Agron. 5:19-30.
Getinet, A., G. Rakow, J.P. Raney, and R.K. Downey. 1997. Glucosinolate content in interspecific crosses of Brassica carinata with B. juncea and B. napus. Plant Breeding 116:39-46.
Harbone, J.B. 1980. Plant phenolics, p. 329-402. In E.A. Bell and B.V. Charlwood (ed.) Secondary plant products. Encyclopedia of plant physiology, Volume 8. Springer-Verlag, Berlin-Heidelberg, Germany.
Hemingway, J.S. 1979. Mustards. Brassica spp. and Sinapis alba (Cruciferae), p. 56-59. In N.W. Simmonds (ed.) Evolution of crop plants. Longman, London.
Kerber, E., and G. Buchloh. 1980. Der Sinapingehalt in Cruciferensamen. Angew. Bot. 54:47-54.
Kozlowska, H., D.A. Rotkiewicz, R. Zadernowski, and F.W. Sosulski. 1983. Phenolic acids in rapeseed and mustard. J, Am. Oil Chem. Soc. 60:119-123.
Kozlowska, H., M. Naczk, F. Shahidi, and R. Zadernowski. 1990. Phenolic acids and tannins in rapeseed and canola, p. 193-210. In F. Shahidi (ed.) Canola and rapeseed. Production, chemistry, nutrition and processing technology. Van Nostrand Reinhold, New York.
Kraling, K., G. Robbelen, W. Thies, M. Herrmann, and M.R. Ahmadi. 1990. Variation of seed glucosinolates in lines of B. napus. Plant Breeding 105:33-39.
Kraling, K., G. Robbelen, and W. Thies. 1991. Genetic variation of the content of sinapoyl esters in seeds of rape, B. napus. Plant Breeding 106:254-257.
Lamour, J., M. Renard, and F. Laher. 1987. A survey of sinapine and related choline esters in seeds of B. napus and those of its genitors. p. 1475. In Groupe Consultatif International de Recherche sur le Colza (ed.) Proc. 7th Int. Rapeseed Cong. Poznan, Poland. 11-14 May 1987.
Malik, R.S. 1990. Prospects of Brassica carinata as an oilseed crop in India. Exp. Agric. 26:125-129.
Matthaus, B. 1997. Antinutritive compounds in different oilseeds. Fett/Lipid 99:170-174.
Nagpal, R., S.N. Raina, Y.S. Sodhi, A. Mukhopadhyay, N. Arumugam, A.K. Pradhan, and D. Pental. 1996. Transfer of Brassica tournefortii (TT) genes to allotetraploid oilseed Brassica species (B. juncea AABB, B. napus AACC, B. carinata BBCC): Homoeologous pairing is more pronounced in the three-genome hybrids (TACC, TBAA, TCAA, TCBB) as compared to allodiploids (TA, TB, TC). Theor. Appl. Genet. 92:566-571.
Robbelen, G., and A. Nitsch. 1975. Genetical and physiological investigations on mutants for polyenoic fatty acids in rapeseed, Brassica napus L. I. Selection and description of new mutants. Z. Pflanzenzuchtg. 75:93-105.
Rucker, B., and G. Robbelen. 1995. Development of high oleic acid rapeseed, p. 389-391. In Groupe Consultatif International de Recherche sur le Colza (ed.) Proc. 9th Int. Rapeseed Cong. Cambridge, U.K. 4-7 July 1995. Henry Ling Limited, Dorchester, UK.
Shahidi, F., and M. Naczk. 1992. An overview of the phenolics of canola and rapeseed: Chemical, sensory and nutritional significance. J. Am. Oil Chem. Soc. 69:917-924.
Thies, W. 1991. Determination of the phytic acid and sinapic acid esters in seeds of rapeseed and selection of genotypes with reduced concentrations of these compounds. Fat. Sci. Technol. 93:49-52.
Thies, W. 1994. Die wertbestimmenden Komponenten des Rapsschrotes. Vortr. Pflanzenzucht. 30:89-97.
Tsunoda, S. 1980. Eco-physiology of wild and cultivated forms in Brassica and allied genera, p. 109-120. In S. Tsunoda et al. (ed.) Brassica crops and wild allies. Japan Scientific Societies Press, Tokyo.
Velasco, L., J.M Fernandez-Martinez, and A. De Haro. 1997. Selection for high sum of oil and protein in Ethiopian mustard (Brassica carinata Braun). Eucarpia Cruciferae Newsl. 19:97-98.
Velasco, L., B. Matthaus, and C. Mollers. 1998. Nondestructive assessment of sinapic acid esters in species of Brassica: I. Analysis by near infrared reflectance spectroscopy. Crop Sci. 38:1645-1650 (this issue).
Leonardo Velasco(*) and Christian Mollers
Institut fur Pflanzenbau und Pflanzenzuchtung, Georg-August-Universitat, Von-Siebold-Str. 8, D-37075 Gottingen, Germany. Received 29 Dec. 1998. (*) Corresponding author (firstname.lastname@example.org).
Published in Crop Sci. 38:1650-1654 (1998).
|Printer friendly Cite/link Email Feedback|
|Author:||Velasco, Leonardo; Mollers, Christian|
|Date:||Nov 1, 1998|
|Previous Article:||Nondestructive Assessment of Sinapic Acid Esters in Brassica Species: I. Analysis by Near Infrared Reflectance Spectroscopy.|
|Next Article:||Molecular Characterization of Two Triticum speltoides Interstitial Translocations Carrying Leaf Rust and Greenbug Resistance Genes.|