Genetic basis of unstable expression of high gamma-tocopherol content in sunflower seeds
© Garcia-Moreno et al.; licensee BioMed Central Ltd. 2012
Received: 7 September 2011
Accepted: 1 May 2012
Published: 18 May 2012
Skip to main content
© Garcia-Moreno et al.; licensee BioMed Central Ltd. 2012
Received: 7 September 2011
Accepted: 1 May 2012
Published: 18 May 2012
Tocopherols are natural antioxidants with both in vivo (vitamin E) and in vitro activity. Sunflower seeds contain predominantly alpha-tocopherol (>90% of total tocopherols), with maximum vitamin E effect but lower in vitro antioxidant action than other tocopherol forms such as gamma-tocopherol. Sunflower germplasm with stable high levels of gamma-tocopherol (>85%) has been developed. The trait is controlled by recessive alleles at a single locus Tph2 underlying a gamma-tocopherol methyltransferase (gamma-TMT). Additionally, unstable expression of increased gamma-tocopherol content in the range from 5 to 85% has been reported. The objective of this research was to determine the genetic basis of unstable expression of high gamma-tocopherol content in sunflower seeds.
Male sterile plants of nuclear male sterile line nmsT2100, with stable high gamma-tocopherol content, were crossed with plants of line IAST-1, with stable high gamma-tocopherol content but derived from a population that exhibited unstable expression of the trait. F2 seeds showed continuous segregation for gamma-tocopherol content from 1.0 to 99.7%. Gamma-tocopherol content in F2 plants (average of 24 individual F3 seeds) segregated from 59.4 to 99.4%. A genetic linkage map comprising 17 linkage groups (LGs) was constructed from this population using 109 SSR and 20 INDEL marker loci, including INDEL markers for tocopherol biosynthesis genes. QTL analysis revealed a major QTL on LG 8 that corresponded to the gamma-TMT Tph2 locus, which suggested that high gamma-tocopherol lines nmsT2100 and IAST-1 possess different alleles at this locus. Modifying genes were identified at LGs 1, 9, 14 and 16, corresponding in most cases with gamma-TMT duplicated loci.
Unstable expression of high gamma-tocopherol content is produced by the effect of modifying genes on tph2 a allele at the gamma-TMT Tph2 gene. This allele is present in line IAST-1 and is different to allele tph2 present in line nmsT2100, which is not affected by modifying genes. No sequence differences at the gamma-TMT gene were found associated to allelic unstability. Our results suggested that modifying genes are mostly epistatically interacting gamma-TMT duplicated loci.
Tocopherols are the main antioxidants present in seed oils. They form a family of four fat-soluble compounds with vitamin E activity named alpha-, beta-, gamma-, and delta tocopherol. Tocopherols contain a 6-chromanol ring structure methylated to varying degrees at positions 5, 7, and 8, and an isoprenoid-derived C16 saturated side chain at position 2. The four tocopherols differ by the number and positions of the methyl groups on the 6-chromanol ring. Alpha-tocopherol is trimethylated, beta- and gamma-tocopherol are dimethylated, and delta-tocopherol is monomethylated .
Tocopherols are important antioxidants operating both in vivo, protecting cells from oxidative damage, as well as in vitro, protecting oils and oil-based products from oxidation . The relative in vivo and in vitro antioxidant properties of the specific tocopherols is determined by their chemical structure. The relative biological activity of the tocopherols is estimated as 100% for alpha-tocopherol, 15 to 27% for beta-tocopherol, 3 to 20% for gamma-tocopherol, and 0.3 to 2% for delta-tocopherol . However, there is a widespread confusion concerning their relative potency in vitro, though it is generally accepted that alpha-tocopherol shows better antioxidant activity in fats and oils at lower concentrations, but at higher concentrations gamma-tocopherol is a more active antioxidant . Studies conducted in sunflower seed oil, in which alpha-tocopherol accounts for more than >90% of the total tocopherols, concluded that substitution of alpha-tocopherol by gamma-tocopherol has a positive impact on the stability of the oil [6–9].
Four sunflower germplasms named LG-17, T2100, IAST-1, and IAST-540 in which alpha-tocopherol in the seeds was almost completely replaced by gamma-tocopherol have been developed [6, 10, 11]. Gamma-tocopherol in the novel germplasm accounts for more than 85% of the total seed tocopherols, compared with more than 90% alpha-tocopherol in conventional sunflower seeds. The increased gamma-tocopherol levels are produced by recessive alleles at the Tph2 locus [6, 12, 13], which encodes a gamma-tocopherol methyltransferase (gamma-TMT) enzyme . This enzyme catalyzes the methylation of delta- and gamma-tocopherol to yield beta- and alpha-tocopherol, respectively . Gamma-TMT mutation in sunflower disrupts the synthesis of alpha-tocopherol and causes the accumulation of gamma-tocopherol . In a detailed sequence analysis of the gamma-TMT gene in sunflower, Hass et al.  identified two gamma-TMT paralogs (gamma-TMT-1 and gamma-TMT-2) and five different haplotypes (haplotypes 4 and 5 corresponding to paralogs 1 and 2, respectively). Both gamma-TMT paralogs 1 and 2 cosegregated with Tph2 and were mapped to linkage group (LG) 8 of the sunflower linkage map. Even though the Tph2 mutation reduced or disrupted the expression of the two paralogs in developing sunflower seeds, none of the DNA polymorphisms found within the gamma-TMT Tph2 gene were associated with the high gamma-tocopherol phenotype . The authors suggested that the Tph2 mutation must be very tightly linked to the gamma-TMT locus on LG 8 and may disrupt regulatory sequences.
Phenotypic studies of Demurin et al.  and García-Moreno et al.  concluded that the four high gamma-tocopherol lines LG-17, T2100, IAST-1, and IAST-540 possess the same allele at Tph2, as no transgressive segregations were observed in crosses involving the four lines. Lines LG-17, T2100 and IAST-540 were isolated from germplasm or M1 mutant plants that segregated for high alpha-tocopherol (>90%) and high gamma-tocopherol (>85%), with no intermediate levels of both tocopherol forms being observed [6, 10–12]. Conversely, IAST-1 was isolated from an M2 plant that exhibited large variation for alpha- and gamma-tocopherol levels, with gamma-tocopherol content in M3 seeds showing a continuous variation from zero to 84.6% and M4 seeds from selected M3 plants showing a variation from 60.4 to 97.4% gamma-tocopherol, which was uniformly high (>90%) in M5 seeds from selected M4 plants . García-Moreno et al.  suggested that the intermediate gamma-tocopherol levels observed during the isolation of IAST-1 might be produced by the presence of modifying genes that determined unstable expression of mutated alleles at Tph2 locus. Modifying genes have been found to influence important traits in sunflower such as high oleic acid content  or broomrape resistance .
In this study, the genetic analysis of a population that showed segregation from low to high gamma-tocopherol values obtained from the cross between the two high gamma-tocopherol lines IAST-1 and nmsT2100 has been carried out. In the course of this analysis, we identified two different alleles at the gamma-TMT Tph2 locus at LG 8. The allele tph2 was present in line nmsT2100, whereas the allele tph2 a was identified in line IAST-1. Additionally, we found four modifying genes at LGs 1, 9, 14 and 16 that in most cases corresponded to duplicated gamma-TMT loci. Modifying genes influenced the expression of tph2 a alleles, but did not affect tph2 alleles.
nmsT2100 x IAST-1 linkage map coverage
Effect of gamma-TMT-F1/F2/R24 on seed gamma-tocopherol content
No. individuals within each marker class
Mean ± SD for gamma-tocopherol content (% total tocopherols)within each marker class
52.3a ± 29.6
72.2b ± 19.1
98.9c ± 0.8
86.8a ± 11.8
96.9b ± 2.6
98.6b ± 0.9
QTL affecting gamma-tocopherol content in the nmsT2100 x IAST-1 population
Significant gene effects5
ORS552 to Gamma_TMT_F9/R24c
ZVG34 to Gamma_TMT_F1/F2/R24
ORS887 to ORS176
ORS185 to ORS307
Gamma_TMT_F9/R24b to ORS700
Significant gene effects 5
Supp int. 3
Marker interval 4
R 2 (%)
Gamma_TMT_F1/F2/R24 to ORS70
ORS887 to ORS176
ORS185 to ORS307
ORS700 to ORS757
Besides Tph2-Gamma-TMT, four other QTL with a moderate effect on gamma-tocopherol content were observed at the F2 level. These QTL were located at LG 1 (GamT1.1), 9 (GamT9.1), 14 (GamT14.1), and 16 (GamT16.1). They individually accounted for 2.3-4.1% of trait variation (Table 3 and Figure 5). The QTL on LG 9, 14 and 16 were also detected in the F3 analyses at practically identical positions (Table 3 and Figure 5). The model with all the QTL explained 44.4% and 61.2% of the phenotypic variance for gamma-tocopherol content in the F2 and the F3 generations, respectively. The QTL peaks of GamT1.1 and GamT16.1 were centered on the gamma-TMT loci mapped at these linkage groups (gamma-TMT-F9/R24c in LG1 and gamma-TMT-F9/R24b in LG 16) (Figure 5). Additionally, the QTL peak of GamT14.1 was likely to correspond to the gamma-TMT-F9/R24d locus mapped to LG 14 in populations CAS-12 x IAST-540  and IAST-413 x HA-89 (Del Moral L, unpublished data), located 1.5 cM from the ZVG64 marker locus at this LG.
Effect of epistatically interacting marker loci on seed gamma-tocopherol content
Genotype ofγ-TMT-F1/F2/R24(LG 8)
Genotype of ORS578(LG 14)
Genotype of ORS126(LG 16)
Several loci amplified by the gamma-TMT F9/R24 INDEL marker in IAST-1 and nmsT2100 (Figure 4) were cloned and sequenced. A band of about 1200 bp from the non-polymorphic locus a and a band of about 375 bp from the dominant locus c that mapped to LG 1 were sequenced from IAST-1. A band of about 1200 bp from the non-polymorphic locus a and a band of about 750 bp from the dominant locus b that mapped to LG 16 were sequenced from nmsT2100. The locus d that mapped to LG 14 was sequenced from lines IAST-540 and HA-89.
The locus a fragment isolated from nmsT2100 harboured alleles from both gamma-TMT paralogs, which showed showed 100% sequence identity to gamma-TMT haplotype 4 (paralog 1) and 5 (paralog 2) from Hass et al. . For IAST-1, the locus a sequence showed 100% sequence identity to gamma-TMT haplotype 4. The locus b consensus sequence was 785 bp long and showed a significant homology to gamma-TMT haplotypes 2 and 3 (GenBank accessions nos. DQ229829 and DQ229830, 3e−58), with 94% maximum sequence identity spanning 61% of the locus b fragment coverage. The consensus sequence for locus c (388 bp long) showed a significant homology to gamma-TMT haplotypes 4 and 5 (GenBank accessions nos. DQ229831 to DQ229834, 4e-14), with 86% maximum sequence identity spanning 19% of the locus c fragment coverage. The sequences of locus d from IAST-540 (405 bp) and HA-89 (406 bp) were very similar, with a 96% of sequence identity between them. The consensus sequence obtained from this alignment showed significant homology to gamma-TMT haplotypes 1, 3, 4 and 5 (GenBank accession nos. DQ229828, DQ229830, and DQ229831 and DQ229834; 8e−18), with 89% maximum sequence identity spanning 18% of the fragment coverage.
Agarose gels from PCR products obtained with the F1 and R92 primers flanking the gamma-TMT gene mapped to LG 8  revealed two bands in nmsT2100 and IAST-1 lines that were identified as the two gamma-TMT paralogs 1 and 2 described by Hass et al. . Consensus nucleotide sequences for IAST-1 gamma-TMT paralog 1 (4126 bp) and paralog 2 (4280 bp) were identical to gamma-TMT haplotype 4 (GenBank accessions nos. DQ229831 and DQ229832) and 5 (GenBank accessions nos. DQ229833 to DQ229834), respectively, from Hass et al.  ( Additional file 1: Figure S1). Consensus nucleotide sequence for nmsT2100 gamma-TMT paralog 1 (4126 bp) was identical to gamma-TMT haplotype 4  ( Additional file 1: Figure S1). Consensus nucleotide sequence for nmsT2100 gamma-TMT paralog 2 (4281 bp) was almost identical to gamma-TMT haplotype 5 , with the exception of five SNPs, one in the 5´UTR (G to A), one in intron 1 (C to T), and three in intron 4 (A to T) ( Additional file 1: Figure S1).
The results of this research suggest that high gamma-tocopherol lines IAST-1 and nmsT2100 possess different mutated alleles at the gamma-TMT Tph2 locus on LG 8. The gamma-TMT catalyses the methylation step from gamma-tocopherol to alpha-tocopherol in sunflower seeds . Mutated tph2 alleles disrupt the activity of gamma-TMT, which results in accumulation of gamma-tocopherol . The allele at nmsT2100 (tph2) is stable and not affected by modifying genes, whereas the allele at IAST-1 (tph2 a ) is unstable and affected by modifying genes. Putative modifying genes located at LG 1, 9, 14 and 16 were identified in the present research, being the effect of those on LG 1, 14 and 16 highly significant in the epistatic interaction with the Tph2 locus on LG 8. Studies on expression of Tph2 mutations in plant tissues other than seeds also pointed to differences between T2100 and IAST-1 lines, as the latter showed lower gamma-tocopherol content in leaves, roots, and pollen .
High gamma-tocopherol lines T2100 and IAST-1 were isolated following different strategies. T2100 derived from an open-pollinated cultivar that showed variation for high gamma-tocopherol content (>85%) at the single-seed level . Genetic study of the trait in T2100 seeds indicated that the trait was controlled by recessive alleles at a single locus Tph2 in such a way that Tph2Tph2 and Tph2tph2 genotypes produced low levels of gamma-tocopherol (<5%) and tph2tph2 genotypes showed high levels of gamma-tocopherol (>85%), with no intermediate levels being observed . On the contrary, IAST-1 derived from a mutagenesis program in which an M2 seed with intermediate gamma-tocopherol content (19.2%) was identified. The trait showed continuous variation for gamma-tocopherol levels (zero to 84.5%) in the M3 generation, which allowed selection of genotypes with stable high gamma-tocopherol content in the M4 generation . A comparative genetic study between both lines concluded that they shared the same allele at Tph2, as no transgressive segregation was observed in the F1 and F2 generations from crosses between them . However, the evaluation of several F2 populations from crosses of IAST-1 with T2100 and the conventional sunflower line HA89 showed that intermediate gamma-tocopherol values (5 to 85%) occurred in some F2 populations, whereas other F2 populations showed uniformly high gamma-tocopherol content (IAST-1 x T2100) or segregated into well-defined low and high gamma-tocopherol classes (IAST-1 x HA89) . The results of this research suggest that intermediate gamma-tocopherol levels are produced by the effect of modifying genes on tph2 a alleles from IAST-1. We hypothesize that modifying genes were present in the mutagenized population from which IAST-1 was developed, where selection against negative alleles at modifying genes led to the isolation of IAST-1, and also in other lines such as HA89 and T2100. Modifying genes have no phenotypic effect in absence of alteration in the major gene . In this research, it was also found that modifying genes have no phenotypic effect on the mutated allele at T2100 (tph2). Accordingly, they are expected to be segregating in HA89 and T2100 lines, which produced different segregation patterns in crosses with IAST-1 . The occurrence of different segregation patterns depending on allelic configuration of modifying genes has been previously reported in sunflower for high oleic acid content [24, 25] and broomrape resistance .
Full length genomic gamma-TMT sequences corresponding to the Tph2 locus on LG 8 have been obtained in this study for both IAST-1 and nmsT2100. Both lines have two gamma-TMT paralogs. Paralog 1 is identical between the two lines and also to the gamma-TMT haplotype 4 from Hass et al. , which included sequences from both high and low gamma-tocopherol lines. Paralog 2 from IAST-1 is identical to gamma-TMT haplotype 5 from Hass et al. , which also included sequences from both high and low gamma-tocopherol lines, while paralog 2 from nmsT2100 carries slight SNP differences. Consequently, we have not found any sequence difference in the gamma-TMT gene from IAST-1 that would explain why the tph2 a allele present in this line is unstable and affected by modifying genes. Previous research did not identify sequence differences in the gamma-TMT gene related to the high gamma-tocopherol phenotype . The authors found identical gamma-TMT paralog 1 and 2 sequences in both high and low gamma-tocopherol lines and reported that mutation leading to the high gamma-tocopherol trait in the sunflower material derived from the high gamma-tocopherol line LG-17  may disrupt regulatory sequences of the gamma-TMT gene. Relating the nmsT2100 line, the slight sequence differences in gamma-TMT paralog 2 were changes to those nucleotides also present in paralog 1 sequences from IAST-1 and nmsT2100 (2 SNPs in the 5´UTR region and in intron 1) or found in a repetitive sequence within an intron (3 SNPs in intron 4) and were not likely to be involved in determining a more stable gamma-TMT allele.
Modifying genes affecting important traits for sunflower breeding such as high oleic acid content [24, 25] and broomrape resistance  have been reported. The discovery of modifying genes affecting high gamma-tocopherol content confirms that the occurrence of modifying genes is not a rare phenomenon in sunflower genetics. Thus far there was no indication on the nature and mode of action of modifying genes. The results of the present research suggest that Tph2 modifying genes are in most cases additional gamma-TMT loci duplicated in the sunflower genome. This was confirmed for modifying genes at LG 1 and 16, whereas there is also indication that the modifying gene at LG 14 might be an additional copy of a gamma-TMT. Duplicated gamma-TMT loci have also been found in safflower  and rapeseed . In fact, the two gamma-TMT loci from safflower (one of them co-segregating with the safflower Tph2 gene determining high gamma-tocopherol content in this crop) were identified by using the F9/R24 primer combination from the sunflower gamma-TMT , the same one used in this study to map different gamma-TMT loci. These results suggested that the F9 and R24 primer sequences are within a highly conserved region of the gamma-TMT gene. In sunflower, Hass et al.  mapped, in addition to gamma-TMT-Tph2 on LG 8, another gamma-TMT locus on LG 16, although the authors considered this locus a randomly amplified polymorphic DNA (RAPD). In this study, the locus on LG 16 (gamma-TMT locus b) showed significant homology to the sunflower gamma-TMT gene and underlay a gamma-tocopherol QTL. Our results indicated that interaction between duplicated gamma-TMT loci revert the high gamma-tocopherol phenotype to intermediate-low gamma-tocopherol values. This effect has also been reported for modifying genes affecting the high oleic acid phenotype in sunflower, where modifying genes suppressed the effect of the Ol-FAD2-1 allele that is essential for accumulating high oleic acid content [18, 28]. Also, epistatically interacting duplicated MPBQ/MSBQ-MT genes that modify beta-tocopherol levels have been reported in sunflower . It is well known that duplicate gene pairs can form negative epistasis due to their overlapping functions  and that suppression of a mutant phenotype can be altered by gene dosage . However, additional biochemical, functional, and sequence analyses are required to determine the mode of action of duplicated gamma-TMT loci in sunflower.
From a breeding perspective, the existence of several modifying genes influencing high gamma-tocopherol content should not be a constraint for the development of cultivars with seeds rich in gamma-tocopherol, as this research revealed that modifying genes do not alter the phenotypic expression of the mutated allele tph2 present at nmsT2100. On the other hand, the influence of modifying genes on expressivity of mutated tph2 a alleles opens up the possibility of selection for balanced levels of alpha- and gamma-tocopherol content, provided that stable combinations of modifying genes can be obtained. The development of mid oleic acid hybrids of sunflower was based on genetic stocks carrying modifying genes that limit the expression of high oleic acid content produced by mutated alleles at Ol locus (>80%) to the mid-range level (55–75%) . In this sense, determination of allelic diversity at gamma-TMT loci underlying gamma-tocopherol modifying genes present in hybrid parental lines might be a useful tool to predict the presence of adequate allelic combinations giving rise to the desired levels of gamma-tocopherol content. Further studies should elucidate whether the other two gamma-tocopherol sources developed thus far, IAST-540 and LG-17, possess stable or unstable alleles at Tph2. Demurin et al.  reported differences in the expressivity of the recessive homozygotes of the Tph2 gene from LG-17 in different genetic backgrounds, which might be indicative of allelic instability.
The results of this research suggest that high gamma-tocopherol in sunflower lines IAST-1 and nmsT2100 is determined by different mutated alleles at the gamma-TMT Tph2 locus on LG 8. The allele at nmsT2100 is not affected by modifying genes, whereas the allele at IAST-1 is unstable and affected by modifying genes. Putative modifying genes have been located at LG 1, 9, 14 and 16, being the effect of those on LG 1, 14 and 16 highly significant in their epistatic interaction with the Tph2 locus on LG 8. Finally, our results suggest that modifying genes correspond to gamma-TMT loci duplicated in the sunflower genome. Phenotypic effects of modifying genes altering the expression of important genes in crop plants are known , but in most cases the genetic basis for modification remains unclear. In this study, we shed light into the mode of action and nature of modifier genes in sunflower, and suggest the relevance of duplicated loci affecting the expression of seed quality specific mutations.
The study included the sunflower lines nmsT2100 and IAST-1, both with high gamma-tocopherol content (>85%). nmsT2100 is a NMS line developed by introgressing monogenic recessive NMS from line P21  into high gamma-tocopherol line T2100 . IAST-1 was isolated in the course of a chemical mutagenesis program on seeds of an accession of ‘Peredovik’ . Twenty-four half seeds of nmsT2100 and IAST-1 were nondestructively analyzed for tocopherol profile as described below, germinated and planted in pots under open air conditions in the spring of 2005. NMS plants of nmsT2100 were pollinated with pollen of IAST-1 plants. Half seeds of the parents as well as F1 half seeds were analysed for tocopherol profile. F1 and parent half seeds were sown in March 2006 and the corresponding plants were grown in pots under open air conditions. F1 plants were bagged before flowering to obtain the F2 generation. F2 seeds from several F1 plants were analysed for seed tocopherol profile. In most cases, F2 seeds had uniformly high gamma-tocopherol content (>90%). However, some F1 plants showed segregation for gamma-tocopherol content at the F2 seed level, which indicated the expression of modifier genes. One population of 192 F2 seeds from a single F1 plant that showed large segregation for gamma-tocopherol content was selected for the molecular analyses. F2 half seeds were germinated and the corresponding plants were grown in pots under open air conditions in 2007. F2 plants were bagged before flowering to produce the F3 seed. Five fully expanded leaves from each F2 plant (135 F2 plants) were cut, frozen at −80°C, lyophilised and ground to a fine powder in a laboratory mill. DNA was isolated from ground leaf tissue from each F2 plant and from five plants of nmsT2100 and IAST-1 parental lines as described in Berry et al. . Twenty four F3 seeds from each F2 plant that produced sufficient amount of seeds were analysed for tocopherol profile.
Two additional F2 populations were also used in this study with the objective of mapping tocopherol biosynthesis loci. One population, described in García-Moreno et al. , derived from a cross between CAS-12, with wild-type tocopherol profile mainly made up of alpha-tocopherol, and IAST-540, with high gamma-tocopherol content. The other population derived from a cross between lines IAST-413 and HA-89, both with wild-type tocopherol profile mainly made up of alpha-tocopherol, though IAST-413 is characterized by increased total tocopherol content .
The analysis of tocopherol profile was made for all analyzed generations and populations in half seeds. The half-seed technique is a common technique used in sunflower breeding consisting in cutting a small seed piece from the seed part distal to the embryo, which is used for nondestructive analysis of seed quality traits, as the remaining seed containing the embryo can be germinated after the corresponding analysis . Individual half seeds were analysed for tocopherol profile following the method of Goffman et al. . Half seeds were placed into 10-ml tubes with 2 ml iso-octane. The half seeds were then crushed with a stainless steel rod as fine as possible. The samples were stirred and extracted overnight at room temperature in darkness (extraction time about 16 h). After extraction, the samples were stirred again, centrifuged, and filtered. Twenty-five μl of the extract were analysed by HPLC using a fluorescence detector at 295 nm excitation and 330 nm emission and iso-octane/tert-butylmethylether (94:6) as eluent at an isocratic flow rate of 1 ml min-1. Chromatographic separation of the tocopherols was performed on a LiChrospher 100 diol column (250 mm x 2 mm I.D.) with 5-μm spherical particles, connected to a silica guard column (LiChrospher Si 60, 5 mm x 4 mm I.D.). The peak areas of the individual tocopherols were corrected according to their previously calculated response factors: alpha-tocopherol = 1.0; beta-tocopherol = 1.80; gamma-tocopherol = 1.85; delta-tocopherol = 2.30.
A complete linkage map for the nmsT2100 x IAST-1 population was constructed to scan the genome for modifier genes affecting the expression of the Tph2 gene. For this, the parental lines IAST-1 and nmsT2100 were initially screened for polymorphisms in two replicate samples together with four F2 individuals, using a genome-wide framework of 95 sunflower SSRs . A preliminary genetic linkage map from this population was constructed. A set of INDEL markers , identified by ZVG prefixes, and an additional set of SSR markers mapped by Tang et al.  and Yu et al. , identified by ORS and CRT prefixes, were additionally screened for polymorphisms between these parental lines to complete the linkage map. INDEL markers for the tocopherol biosynthesis genes gamma-TMT, MPBQ/MSBQ-MT and tocopherol cyclase described by Hass et al.  and Tang et al.  were also screened for polymorphisms between nmsT2100 and IAST-1. PCRs for SSRs analyses were performed as described by Pérez-Vich et al. . INDEL analyses were carried out following Yu et al.  and Hass et al. . SSR and INDEL amplification products were separated on 3% (w/v) Metaphor® (BMA, Rockland, ME, USA) and 1.5% agarose gels, respectively, in 1x TBE buffer with ethidium bromide incorporated in the gel. SSR and INDEL markers revealing polymorphisms were then genotyped in the nmsT2100 x IAST-1 F2 population, following the protocols mentioned above.
Chi-square statistics were computed on each genotyped locus to detect deviations from the expected Mendelian ratios for codominant (1:2:1) or dominant (3:1) markers. The nmsT2100 x IAST-1 linkage map was constructed using the software MAPMAKER/EXP version 3.0b (Whitehead Institute, Cambridge, MA, USA) . Two-point analysis was used to identify linkage groups (LGs) at a LOD score of 3 and a maximum recombination frequency of 0.40. Three-point and multi-point analyses were used to determine the order and interval distances between the markers at each LG. The Haldane mapping function was used to compute the map distances in centiMorgans (cM) from the recombination fractions. Multiple loci detected by a single maker were coded with the marker name plus the suffix a, b, c, or d to indicate each duplicate locus. Linkage group maps were drawn using the MapChart software .
Genetic analysis of modifier genes was performed in several stages. In the first stage, the significance of each marker’s association with the phenotypic trait [gamma-tocopherol content at the F2 seed and F2 plant (average value of 24 F3 seeds per F2 plant) generations] was determined by one-way analysis of variance (ANOVA) using the statistical package SPSS Statistics v. 19, with marker genotypes being classes. In this analysis, we identified an unexpected macromutation on LG 8 at the Tph2 locus. The effects of the macromutations, if ignored, could dramatically reduce the power for identifying other genes or QTL affecting the studied trait.
In a second stage, composite interval mapping (CIM) [43, 44] was used to scan the genome for QTL affecting gamma-tocopherol content, in order to strengthen and corroborate the results of the analyses of variance, evaluate the existence of additional QTL, and estimate the interaction and global effect of all the detected QTL. Computations were carried out using the software PLABQTL Version 1.1 . The phenotypic data consisted on gamma-tocopherol content in the F2 seed and F2 plant generations. Additional analyses were carried out by using other parameters calculated from the F3 seed data such as the minimum, maximum and the standard deviation of gamma-tocopherol content in F3 seeds per F2 plant, and the number of F3 seeds within each F2 plant with less than 90% of gamma-tocopherol content. Since these analyses gave similar results to those obtained with the mean gamma-tocopherol value per F3 family, only results based on the F3 mean value are shown. Analyses were made initially with the “first” statement to check the database for errors and outliers. Next, simple interval mapping (SIM) was carried out for an initial scan and detection of QTL with main effects. Finally, CIM was performed with markers closest to the main QTL as co-factors. Genome-wide threshold values (α = 0.05) for declaring the presence of QTL were estimated from 1000 permutations of each phenotypic trait . The thresholds of the LOD score (and their 0.95 confidence intervals) were 2.57 (2.48–2.70) and 2.65 (2.58–2.76) for gamma-tocopherol content in the F2 and the F3 generations, respectively. Estimates of QTL positions were obtained at the point where the LOD score reaches its maximum in the region under consideration. One-LOD support limits for the position of each QTL were also calculated . The proportion of phenotypic variance explained by each individual QTL was calculated as the square of the partial correlation coefficient (R 2 ). Estimates of the additive (ai) and dominance (di) effects, as defined by Falconer , for the ith putative QTL, the total LOD score, as well as the total proportion of the phenotypic variance explained by all QTL, were obtained by fitting a multiple regression model including all putative QTL for the respective trait simultaneously . The occurrence of QTL x QTL interactions was tested by adding digenic epistatic effects to the model.
QTL software such as PLABQTL estimate epistatic interactions among previously identified QTL. Since modifying genes are defined as genes having no known effect except to intensify or diminish the expression of a major gene , their effect as individual loci and subsequently their interaction with major loci may be undetectable with this type of analyses. In consequence, two-way interactions between the Tph2 major locus and all the marker loci genotyped in this study were also tested. Two-way interactions were analyzed at a significance threshold of P ≤ 0.05 by analysis of variance using the general linear model (GLM) of SPSS Statistics v. 19. Statistical significance of differences for gamma-tocopherol content in different genotypes combining two marker loci were also computed using Duncan’s multiple range test. The significant epistatic interaction terms were combined with those of the previously identified QTL in multiple locus models using the “seq” statement of PLABQTL.
In the course of the genetic analyses of the nmsT2100 x IAST-1 population, we identified different gamma-TMT loci amplified with INDEL marker gamma-TMT-F9/R24 associated to gamma-tocopherol QTL. In order to confirm their nature, these loci were sequenced as follows. F9/R24 INDEL fragments amplified from the IAST-1 and nmsT2100 parental lines were separated on a 1.5% agarose gel, excised and purified by means of the QIAquick gel extraction kit (Qiagen GmbH, Hilden, Germany). The purified fragments were ligated into the T/A vector (pCR2.1) and the recombinants were transformed to TOP10 Chemically Competent E. coli using the TOPO-TA cloning kit (Invitrogen, San Diego, CA, USA) as described by the manufacturer. Five recombinant bacterial colonies (white) per isolated band were picked from the plate containing ampicillin and X-gal as selective media and cultured overnight at 37°C. Plasmids were extracted and purified using QIAprep Spin Miniprep Kit (Qiagen GmbH, Hilden, Germany). PCR with M13 forward and reverse vector primers and F9 and R24 primers, and restriction enzyme digestion was performed to confirm the presence and size of the insert. Sequencing in both forward and reverse orientations of the cloned fragments (two clones per locus) was performed at GATC Biotechnology (Konstanz, Germany) using the M13 forward and reverse sequencing primers. Sequence analysis was conducted with the aid of the software Vector NTI Advance 10.3.0 (Invitrogen, San Diego, CA, USA).
Sequencing primers designed each 500–1000 bp into the gamma-TMT gene
nuclear male sterile
simple sequence repeat.
The research was funded by the Spanish Ministry of Science and Innovation and the European Union FEDER (research project AGL2007-62834) and Dow Agrosciences LLC. We acknowledge support of the publication fee by the CSIC Open Access Publication Support Initiative through its Unit of Information Resources for Research (URICI).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.