SELECTION FOR LOW a-LINOLENIC ACID CONTENT IN SPRING TURNIP RAPE
I. Laakso1, S. Hovinen2, T. Seppänen-Laakso1 and R. Hiltunen1
1Department of Pharmacy, Biocenter 2, P.O. Box 56, FIN-00014 University of Helsinki,
2Boreal Plant Breeding, Myllytie 10, FIN-31600 Jokioinen, Finland
1e-mail: into.laakso@helsinki.fi; 2e-mail: simo.hovinen@mtt.fi
ABSTRACT
Individual plant selections have been continued from both free and hand-pollinated materials in the breeding for lower a-linolenic acid content in zero-erucic acid spring turnip rape (Brassica rapa ssp. oleifera). The main criterion has been the highest linoleic/a-linolenic acid (LA/a-LLA) ratio or the lowest a-LLA content. During the last 3 years, selections have resulted in a further decrease in a-LLA levels down to 7%, and an average increase in the LA/a-LLA ratio up to 4. This ratio is twice as high as that in control populations and cultivars. The mean LA level in the selected lines has varied between 27-31%. In 1996, most of the plants in an experimental plot derived from seeds of free pollinated, low a-LLA plants had pale-yellow flowers. The pedigrees of this mutant line have maintained their low a-LLA contents (7%) and flower colour also in the next two generations. The single plant yields with a-LLA contents of 4-5% that have been found continuously in breeding materials have markedly speeded up the effects of the selections.
KEYWORDS Brassica rapa ssp. oleifera, fatty acid breeding, linoleic/a-linolenic acid ratio
INTRODUCTION
Zero-erucic acid rapeseed oil has a unique fatty acid composition with high oleic acid (OA, 18:1n-9) and moderate essential linoleic (LA, 18:2n-6) and a-linolenic acid (a-LLA, 18:3n-3) contents. The proportions of a-LLA (10%) and LA (24%) are sufficient to compete over oleic acid (57%) and to cause an increase in all the major long-chain metabolites of the n-3 and n-6 fatty acid families in plasma phospholipids, especially when replacing saturated fat by rapeseed oil (Seppänen-Laakso et al., 1992). Competition ensures the preferential incorporation of essential fatty acids to be utilized by the body. The effects resulting from the high competitiveness of a-LLA make rapeseed oil a highly valuable source of n-3 fatty acids of vegetable oil origin.
Reducing the a-LLA content apparently decreases the nutritional value of rapeseed oil, but it also increases the stability to oxidation making it useful to the margarine industry. Most of the breeding efforts at lowering a-LLA have been made with rape (B. napus ssp. oleifera) using mutagenesis and transformation techniques (Scarth, 1995).
In Finland, where spring turnip rape (Brassica rapa ssp. oleifera) is best adapted to the short growing period, breeding programmes aimed at modified fatty acid compositions have been directed at altering the proportions of LA and a-LLA (Hovinen and Laakso, 1987; 1991; Laakso et al., 1995) and high oleic acid content (Vilkki and Tanhuanpää, 1995). In the first selections based on the highest LA/a-LLA ratio, a single plant yield was found with an a-LLA level of below 7% (Hovinen and Laakso, 1987). This prompted us to start selections for low a-LLA content, too. Since then, a marked increase in the variation towards lower a-LLA levels has been found. In the breeding lines that include materials from controlled crosses, the mean a-LLA content has ranged from 8 to 10%, and the LA/a-LLA ratio from 2 to 5 (Laakso et al., 1995). In this study, the results of the selection experiments carried out during 1996-98 are presented.
EXPERIMENTAL
During the last 12 years, the breeding programme has included single plants selections, pair-crosses in the field and fatty acid analyses (Hovinen and Laakso, 1987; 1991; Laakso et al., 1995). The seed materials were originally derived from two erucic acid-free, low glucosinolate and yellow-seeded spring turnip rape populations. About 350 fatty acid analyses have been performed each year. The selections have been based on the highest LA/a-LLA ratio or lowest a-LLA content. A third type of selection aimed at high LA levels only has been carried out simultaneously.
In 1996-98 the breeding material was grown on fields with clay soil at the Kotkaniemi Experimental Station (Kemira Agro Ltd), 40 km to the North-West of Helsinki. The seed materials obtained from selected, controlled pair-crosses, as well as from free pollinated plants, were sown on small plots bulking plant yields with desired fatty acid profile. Progenies from controlled pair-crosses formed small plots ('lines'). Selected plant progenies from free pollinated materials also formed their own lines. The plot size was 8-10 m2.
In new recurrent selection from the lines, about 15-20 free pollinated, vigorous and disease-free plants were selected each year. In addition, about 10-15 pairs of bag-isolated plants were reciprocally hand-pollinated within each plot. Fatty acid analysis was performed on pair-cross and free pollinated yields, control populations and cultivars during the winters. The seed materials sown in the field in 1998 had LA/a-LLA ratios of between 5-6, a-LLA levels lower than 6% and LA levels higher than 34%. This represented the basis for the three types of selections (a-c, Tables 1-2).
RESULTS AND DISCUSSION
The mean oleic (OA), linoleic and a-linolenic acid contents of the lines in each selection category, control populations and cultivars during the last three years are shown in Table 1.
Table 1. The proportions of main fatty acids in the free-pollinated breeding and control materials in 1996-98. Selection based on the highest LA/a-LLA ratio (a), the lowest a-LLA (b) or the highest LA content (c).
|
SELECTED LINES |
CONTROLS AND CULTIVARS |
||||||||||
|
Breeding materials |
N |
OA % Mean |
LA % Range Mean |
a-LLA % Range Mean |
LA/a-LLA ratio |
N
|
OA % Mean |
LA % Mean |
a-LLA % Mean |
||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Bor-96 a b c |
71 117 15 |
55.5 55.9 50.4 |
22.7-34.9 21.9-36.5 26.7-37.1 |
29.5 28.2 32.6 |
5.6-11.0 5.4-14.7 8.4-13.0 |
8.3 8.5 10.2 |
3.7 3.5 3.2 |
60 60* |
58.7 60.8
|
23.0 21.1
|
12.5 11.8
|
|
Bor-97 a b c |
77 101 30 |
59.6 59.5 56.7 |
22.2-34.6 21.2-32.4 21.7-33.7 |
27.4 26.0 27.5 |
4.1- 9.8 5.7-12.6 7.6-11.6 |
6.5 7.8 9.6 |
4.3 3.4 2.9 |
40 40* |
60.8 60.9
|
22.2 22.3
|
10.8 10.7
|
|
Bor-98 a b c |
61 32 32 |
57.3 59.0 52.2 |
23.3-35.2 21.5-31.7 22.6-38.4 |
29.3 26.5 31.4 |
5.4- 9.1 5.8-12.9 6.8-13.4 |
6.9 7.6 9.8 |
4.3 3.7 3.3 |
32 31* |
58.4 59.6 |
23.7 22.6
|
11.9 11.7 |
N = number of plants analysed; *spring turnip rape cultivars (‘Kelta’ and ‘Kulta’)
Comparisons between the levels of the three major fatty acids in the breeding lines and control populations show that the oleic acid (OA) content has been relatively unaffected by the selections based on the highest LA/a-LLA ratio or the lowest a-LLA content (Table 1; Bor-97-98 a,b). It is also clear that the sum of LA and a-LLA is within the range of 33-36% in both the selected and control lines. Since the decrease in a-LLA was compensated by LA, the selection seems to have a predominant effect on the desaturation step between LA and a-LLA. In contrast, the selection for higher LA content alone (c) mainly affected the corresponding step between OA and LA.
Table 2. The proportions of the main fatty acids and LA/a-LLA ratio in the hand-pollinated lines in 1996-98. Selection based on the highest LA/a-LLA ratio (a), the lowest a-LLA (b) or the highest LA content (c).
|
SELECTED LINES |
LA/a-LLA RATIO
|
|||||||
|
Breeding materials |
N
|
OA % Mean |
LA % Range Mean |
a-LLA % Range Mean |
Selections |
Control populations |
||
|
|
|
|
|
|
|
|
|
|
|
Bor-96 a b c |
32 7 9 |
50.0 50.9 46.2 |
21.2-37.7 26.2-37.6 29.4-37.2 |
31.6 31.2 33.9 |
5.1-12.7 5.9-11.8 8.2-13.9 |
8.3 7.7 10.9 |
4.0 4.2 3.2 |
1.8
|
|
Bor-97 a b c |
78 51 30 |
55.8 56.7 48.0 |
24.5-36.9 24.3-34.8 26.8-42.7 |
30.9 29.4 34.1 |
3.7-10.7 7.0- 9.2 7.4-17.3 |
6.3 7.0 11.2 |
5.1 4.2 3.2 |
2.1
|
|
Bor-98 a b c |
30 28 28 |
57.4 58.4 52.8 |
22.7-34.0 23.2-31.0 26.7-35.6 |
29.7 27.5 30.7 |
5.2- 9.3 5.9-12.3 6.9-13.5 |
6.8 7.2 10.5 |
4.5 4.0 3.0 |
2.0
|
N = number of individual plants
The decrease in the a-LLA content down to 7% during the three-year period (Tables 1-2; a,b) was mainly caused by its high variability in both the free and hand-pollinated materials, in which a-LLA contents of 4-5% and LA/a-LLA ratios of from 5 to 6 have been commonly found. This also indicated an average LA/a-LLA ratio of 4, which is twice as high as that in the control populations and turnip rape cultivars (Tables 1-2; a,b). The ratio, which tended to be slightly higher in the high LA/a-LLA lines, has remained stable during successive generations. Separate selections for higher LA content have produced LA levels of from 28 to 34% and LA/a-LLA ratios of 3 (Tables 1-2; c).
|
Fig.1. Effects of continuous selection on the a-LLA content. |
|
Fig. 2. a-LLA levels in the materials grown in the field in 1998. |
The long-term effects of selection on the a-LLA contents is presented in Fig. 1. In both types of selection the trends towards lower a-LLA levels closely followed each other, showing continuous response to selection. Despite the clearly lower population mean (Fig. 2), low a-LLA selections had a relative variation (C.V., 17%) that was twice as high as that in the control material.
In 1996, the majority of the plants on one experimental plot (Bor 93916) derived from the combined seeds of two free pollinated, low a-LLA plants unprecedently formed pale-yellow flowers. Since then both the free and hand-pollinated materials of this mutant line have had an average a-LLA content of below 7% and an LA/a-LLA ratio higher than 4. In addition, in 1998 the plot yield (Bor 98007) was yellow-seeded. Controlled pair-crosses within the plot also produced pedigrees with pale-yellow flowers in the following two generations, indicating that this colour is a heritable trait. Another major difference in these plants is that the stem appeared to be more rigid.
CONCLUSIONS
Further reduction of the a-LLA content in spring turnip rape by a few percent units to the levels typical of low a-LLA rape cultivars is feasible due to its high variability and continuous response to selection. Lines with a lower a-LLA content would be valuable in developing new cultivars for the fat industry. However, improved fatty acid compositions should be combined with desirable agronomic traits such as a higher seed yield and better disease and cold resistance.