An oleate desaturase and a suppressor loci direct high oleic acid content of sunflower (Helianthus annuus L.) oil in the Pervenets mutantUn locus dˈoléate désaturase et un locus suppresseur gouvernent la haute teneur en acide oléique de lˈhuile de tournesol (Helianthus annuus L.) chez le mutant Pervenets.

André Adoutte
https://doi.org/10.1016/S0764-4469(01)01353-1Get rights and content

Abstract

All the 〚HOAC〛 lines derived from the Pervenets mutant carry a specific RFLP (oleHOS) revealed by an oleate desaturase cDNA used as a probe. The 〚LO〛 (linoleic) genotypes do not carry oleHOS, but another allele: oleLOR. We studied 〚HOAC〛 heredity in two segregating populations. In an F2 population, the 〚HOAC〛 trait co-segregated with oleHOS. In a recombinant inbred line F6 population, all 〚HOAC〛 RI lines carried oleHOS. The RI lines carrying oleHOS were either 〚LO〛 or 〚HOAC〛. The absence of 〚HOAC〛 RI lines with oleLOR eliminated the occurrence of a recombination event between the locus carrying oleHOS and the locus carrying Pervenets allele. The 〚HOAC〛 trait is due to 2 independent loci: the locus carrying oleHOS allele and another locus. One allele at this other locus may suppress the effect of oleHOS allele on the 〚HOAC〛 trait. The existence of this suppressor allele has only been suggested for sunflower.

Résumé

Toutes les lignées 〚HOAC〛 dérivées du mutant Pervenets portent un RFLP 〚HOAC〛 spécifique (oleHOS) révélé par un ADNc dˈoléate désaturase utilisé comme sonde. Les lignées normales ou 〚LO〛 (pour linoléique), ne portent pas oleHOS mais un autre allèle: oleLOR. Nous avons étudié lˈhérédité du caractère 〚HOAC〛 dans deux populations en ségrégation. Pour une population F2, le caractère 〚HOAC〛 co-ségrège avec oleHOS. Dans une population F6 de lignées recombinantes, toutes les lignées 〚HOAC〛 portent oleHOS. Les lignées recombinantes oleHOS sont soit 〚HOAC〛 soit 〚LO〛. Lˈabsence dˈindividu 〚HOAC〛 portant oleLOR élimine la possibilité dˈune recombinaison entre le locus portant oleHOS et celui portant lˈallèle Pervenets. Le phénotype 〚HOAC〛 est donc conditionné par deux loci indépendants: le locus portant lˈallèle oleHOS et un autre locus dont un allèle supprimerait lˈeffet de lˈallèle oleHOS sur le caractère 〚HOAC〛. Lˈexistence de cet allèle suppresseur (supHOAC) nˈa été suggérée que chez le tournesol.

Introduction

Sunflower (Helianthus annuus L.) is one of the major seed oil crops in Europe. Sunflower varieties usually produce an oil with about 60 % linoleic acid (C18:2), 20 % oleic acid (C18:1) and 10 % saturated fatty acid, both palmitic and stearic (C16:0 and C18:0). These varieties are henceforth referred to as 〚LO〛 for linoleic. The Pervenets high oleic acid content (〚HOAC〛) mutant has been obtained by chemical mutagenesis on a 〚LO〛 population VNIIMK8931 〚1〛. It carries the specific Pervenets allele induced by the mutagen treatment and leading to an oleic acid content (OAC) in seed oil higher than 65 %. Subsequently, varieties were derived from Pervenets and produce an oil with over 83 % C18:1 and about 7 % C18:2. 〚HOAC〛 sunflower oil presents advantages for long shelf-life and frying, and can be heated to higher temperatures than normal oil without smoking 〚2〛. 〚HOAC〛 sunflower oil has the same excellent effect on reducing low-density lipoprotein cholesterol as olive oil 〚3〛. Moreover, due to the absence of linolenic acid (C18:3), 〚HOAC〛 sunflower oil does not present the disadvantages of olive, rapeseed, soybean and peanut oils, which have about 10 % linolenic acid, and are thus more susceptible to oxidation at high temperatures 〚2〛.

To improve our knowledge about the Pervenets 〚HOAC〛 mutant, a simple model could be considered 〚4〛, 〚5〛, 〚6〛, 〚7〛. Oleic acid over-accumulation could be due to 1) over-activity of the stearoyl-ACP desaturase catalysing the desaturation of stearic acid to oleic acid or 2) a lack of oleoyl-PC desaturase (oleate desaturase) activity catalysing the desaturation of oleic acid to linoleic acid. This model has enabled us to reveal that in 〚HOAC〛 sunflower oleate desaturase activity is drastically reduced 〚5〛, and that there is less accumulation of oleate desaturase transcript in 〚HOAC〛 than in 〚LO〛 sunflower 〚8〛, 〚9〛, 〚10〛. This suggests that accumulation of oleate desaturase transcript and, consequently, oleate desaturase activity might be affected by the Pervenets allele. Moreover, all the 〚HOAC〛 genotypes carry a specific RFLP (oleHOS) revealed by the oleate desaturase cDNA used as a probe. The oleHOS allele is strictly correlated to the 〚HOAC〛 status of genotype, since thus far 〚LO〛 genotypes do not carry oleHOS but another allele: oleLOR 〚9〛, 〚11〛. Concerning the genetics of this mutation, several studies have been developed. However, they have led to discordant results concerning 1) the 〚HOAC〛 trait, which behaved as dominant, recessive or intermediate, 2) the different number of genes involved in the 〚HOAC〛 trait deduced by authors, and 3) some maternal effects (for review see 〚12〛). In all these studies, results concerning the Pervenets allele deduced by the authors were supported by the OAC distribution in their genetic material. Nonetheless, OAC depends on both the Pervenets allele and other unrelated factors. Indeed, the C18:1 / C18:2 ratio is susceptible to the environmental and genetic background of genotypes. The 〚HOAC〛 trait might also be influenced by such factors. This probably explains differences between results from genetic studies. Thus, the molecular and genetic nature of the 〚HOAC〛 character is still not well understood. This lack of understanding leads to several difficulties during breeding programs concerning the direct conversion process of 〚LO〛 lines into 〚HOAC〛 lines using Pervenets or 〚HOAC〛 lines already derived from Pervenets 〚12〛.

In the present study, we developed a genetic approach to the 〚HOAC〛 trait using two mapping populations. We proposed a model concerning the genetic determinism of the 〚HOAC〛 character. For this purpose, we studied the segregation of the 〚HOAC〛 trait and oleHOS:oleLOR alleles (= oleHL locus) in 〚HOAC〛:〚LO〛 F2 segregating population and in F6 segregating recombinant inbred lines (RI lines). The two segregation patterns were in agreement with one and two loci involved in the 〚HOAC〛 trait for the F2 and the RI lines populations, respectively. We proposed a model to reconcile the results between the F2 and the RI line populations. This model was in agreement with the results of the two different genetic studies and might explain most of the discordant results found in the literature.

Section snippets

Plant materials

F2 segregating population: The 〚LO〛 line BD40713 as female was crossed with the 〚HOAC〛 line BE78079 from Monsanto. One F1 plant was self-fertilised to produce the F2 progenies composed of 107 plants by 1998 in a greenhouse. They were used both to determine OAC on half-seed cotyledon and to determine oleate desaturase RFLP by genotyping with an oleate desaturase cDNA used as a probe. BD40713 and BE78079 parent lines and F1 plants of a similar cross were used in the same way as controls.

OAC content

〚LO〛 and 〚HOAC〛 parents of the F2 population displayed on average 30 and 83 % OAC, respectively. For sister F1 seeds of a similar cross, the OAC were found to be 83 %. The 〚HOAC〛 trait therefore appeared to be dominant. In the F2 population, OAC was quantified in half cotyledon of each F2 seed. The OAC histogram of the F2 offspring is shown in figure 1. The histogram is clearly bimodal. The OAC threshold for classification of plants in 〚LO〛 or 〚HOAC〛 should be located at 70 %. In the context of

Discussion

In the F2 segregating population, we observed co-segregation between the oleHOS allele and an OAC higher than 65 % (〚HOAC〛 trait). Thus, we revealed a strong genetic linkage between an oleHL locus carrying oleHOS / oleLOR alleles and the Pervenets allele (Rsq = 0.8642, p < 10-4). This suggests that the Pervenets allele is either in the oleHL locus or strongly linked to this locus. Moreover, the dominance / additivity ratio (90 %) reveals an almost complete dominance effect of the oleHOS region

Conclusion

We revealed a tight genetic linkage between oleHL locus and the locus carrying the Pervenets allele. Perez-Vich et al. 〚18〛 have also shown by QTL analyses that the QTL peak for OAC, explaining 84.5 % of total OAC variation, coincided with an oleate desaturase locus. Moreover, the genetic linkage explains the strong linkage disequilibrium between oleHOS and Pervenets alleles found previously 〚11〛. We can therefore conclude that it is the oleHOS allele at oleHL locus, which carries or is close

Acknowledgements

Séverine Lacombe’s work was supported by a CIFRE Contract between INRA and Monsanto. This joint work involved Y. Griveau, D. Varès and P. Lacombe (INRA - Montpellier), P. Jouve, S. Veillet, C. Millet, H. Guillot and W. Dioh (Monsanto), with the collaboration of A.G. Abbott (Clemson University, SC, USA).

References (18)

  • K.I. Soldatov

    Chemical mutagenesis in sunflower breeding

  • J.L. Sebedio et al.

    Utilization of sunflower oils in industrial frying operations, Grasas y aceites

    Special issue

    (1996)
  • B. Delplanque

    Intérêt nutritionnel des tournesols

  • J.B. Ohlrogge et al.

    Lipid biosynthesis

    Plant Cell

    (1995)
  • J. Shanklin et al.

    Desaturation and related modifications in fatty acids

    Ann. Rev. Plant Physiol. Plant Mol. Biol.

    (1998)
  • J.B. Ohlrogge et al.

    The genetics of plants lipids

    Biochem. Biophys. Acta

    (1991)
  • H. Fernandez et al.

    Inheritance of high oleic content in sunflower oil

    J. Genet. Bre

    (1999)
  • S. Lacombe et al.

    Analysis of desaturase transcript accumulation in normal and in high oleic oil sunflower development seeds

  • V. Hongtrakul et al.

    A seed specific D12 oleate desaturase gene is duplicated rearranged and weakly expressed in high oleic acid sunflower lines

    Crop Science

    (1998)
There are more references available in the full text version of this article.

Cited by (24)

View all citing articles on Scopus
View full text