2025 Volume 75 Issue 2 Pages 93-101
Strong yellow color, caused by carotenoid accumulation, in semolina flour made from durum wheat (Triticum turgidum L. subsp. durum (Desf.)) is one of the most important traits for pasta production. The first step in the carotenoid biosynthesis pathway, which is catalyzed by phytoene synthase (PSY), is a bottleneck, and allelic variation of Psy-A1 in durum wheat produces different yellow pigment contents (YPC) in seeds. Durum wheat carrying leaf rust resistance gene Lr19, which was translocated from wheat relative Thinopyrum ponticum chromosome 7E to durum wheat chromosome 7A, is known to produce high YPC, and the causal gene is presumed to be Psy-E1, which is tightly linked to Lr19. In this study, Psy-E1 produced higher YPC than Psy-A1 alleles, such as Psy-A1k, Psy-A1l and Psy-A1o, in durum wheat. Segregation analysis demonstrated that Psy-E1 is located at the Psy-A1 locus on chromosome 7A. In a 2-year field test of near-isogenic materials, Psy-E1 was accompanied by yield loss with decreases in grain number per spike, test weight and thousand-kernel weight under moisture conditions typical of wheat-growing areas of Japan. Thus, Psy-E1 has the potential to contribute high YPC in durum wheat breeding programs, although the applicable cultivation environments are limited.
For pasta products made from semolina flour of durum wheat (Triticum turgidum L. subsp. durum (Desf.)), yellowness and elastic texture with high protein contents and gluten strength are desirable traits. In particular, a bright yellow color, which is caused by carotenoid yellow pigment content (YPC), is preferred for pasta products (Hentschel et al. 2002, Lachman et al. 2017). In addition to their role of a visual quality of pasta products, carotenoids have important roles in human nutrition and health owing to their provitamin A activity and antioxidant properties (Abdel-Aal et al. 2007, Ficco et al. 2014).
YPC of wheat is a stable phenotype influenced by genetics and showing high heritability (Clarke et al. 2006, Digesù et al. 2009). Recently, many genetic studies have examined carotenoid biosynthesis in plants. This pigment is a mixture of carotenoids, mainly lutein (Digesù et al. 2009, Prins et al. 1997). Lutein is biosynthesized from geranylgeranyl diphosphate. In Triticum species, including durum wheat, lutein accumulates as a yellow pigment in the endosperm. It is converted to phytoene by phytoene synthase (PSY) as the initial step, followed by the downstream reactions of the terpenoid biosynthesis pathway (Ramachandran et al. 2010). Phytoene is converted to lycopene in a desaturation step catalyzed by phytoene desaturase (Pds) (Bartley et al. 1999) and zeta-carotene desaturase (Zds) (Albrecht et al. 1995) and an isomerization step catalyzed by carotenoid isomerase (CRTiso) (Matthews et al. 2003). Lycopene is then converted to lutein through cyclization by lycopene-β-cyclase (βLCY) and lycopene-ε-cyclase (εLCY) and hydroxylation by β-hydroxylase (βOHase) and ε-hydroxylase (εOHase) (Fraser and Bramley 2004). In this pathway, three bottlenecks are known in several plants: phytoene synthesis by PSY, lycopene cyclization and carotene hydroxylation (Lindgren et al. 2003, Zhou et al. 2022). Among these bottlenecks, phytoene synthesis by PSY have been reported as rate-limiting step in wheat (Cong et al. 2009).
There are three PSY paralogs in wheat: PSY1 in leaf, which is involved in carotenoid accumulation; PSY2 in green tissues, which is involved in photosynthesis; and PSY3 in root, which regulates ABA biosynthesis under abiotic stress (Dibari et al. 2012, Li et al. 2008, Pozniak et al. 2007, Zhang and Dubcovsky 2008, Zhou et al. 2022). Zhang and Dubcovsky (2008) also reported two QTLs related to yellow pigment accumulation, which named Psy-A1 and Psy-B1 on chromosome group 7 in durum wheat. Allelic variations in PSY1 genes were later found in durum wheat and compared their effect for YPC, the effect order the YPC from high to low was Psy-A1o, Psy-A1l and Psy-A1a by recombination inbred lines (Singh et al. 2009). In other studies, Psy-A1o and Psy-A1l had the same effect on YPC and both are stronger than Psy-A1a (Campos et al. 2016).
Wild relatives of wheat, especially Thinopyrum ponticum (syn: Th. elongatum, Agropyron elongatum and Lophopyrum elongatum), have been shown to be a resource for leaf rust resistance in breeding programs (McCallum et al. 2016, Prasad et al. 2020). One leaf rust resistance gene, Lr19, has been transferred from A. elongatum (Th. ponticum) on chromosome 7E to the long arm of wheat chromosome 7D. A translocated segment containing Lr19 was associated with higher biomass and grain yield under irrigated conditions and with yellow endosperm designated causal gene as Y (Monneveux et al. 2003, Singh et al. 1998). Zhang et al. (2005) and Gennaro et al. (2009) generated durum wheat–Th. ponticum recombinant lines in which a segment containing Lr19 with Y was translocated to chromosome 7A in durum wheat.
Psy-E1, the orthologue of Psy-1 was presumed as candidate gene of Y locus derived from Th. ponticum, and it was involved in YPC (Zhang and Dubcovsky 2008). The genetic distance between Lr19 and Psy-E1 was estimated to be 9.8 cM (Xu et al. 2023). Also, unidentified Psy paralog adjacent to Psy-E1 was speculated by the apparent widespread conserved duplication on chromosome group 7 in the Poaceae (Ceoloni et al. 2014, Gallagher et al. 2004, Singh et al. 2009).
It has been reported that Lr19-linked Psy-E1 confers high YPC in durum wheat (Xu et al. 2023). However, effectiveness of Psy-E1 on durum wheat breeding program has not been reported so far. In this study, we evaluated Psy-E1’s contribution for high YPC and its effect on agronomic traits under Japanese climate condition comparison with the original Psy-A1 alleles in durum wheat, and possibility of Psy-E1 and Psy-A1 pyramiding to increase YPC.
DNA was extracted by the potassium acetate method (Dellaporta et al. 1983). DNA templates were amplified in a PCR Thermal Cycler Dice (TaKaRa Bio, Shiga, Japan) with LA Taq and 10× LA PCR Buffer II (TaKaRa Bio) and primers listed in Supplemental Table 1. PCR amplification of the open reading frame of Psy-A1 was an initial 94°C for 1 min, followed by 35 cycles of 95°C for 10 s and 72°C for 10 min. The amplified PCR fragments were separated by agarose gel electrophoresis, visualized with Gel Red stain (Biotium, Fremont, CA, USA) and extracted with a Qiaquick Gel Extraction Kit (Qiagen, Hilden, Germany). Cycle sequencing reactions were performed in a PCR Thermal Cycler Dice with BigDye Terminator v. 3.1 with sequencing primers listed in Supplemental Table 1 and purified with BigDye Terminator (Applied Biosystems, Waltham, MA, USA) following the manufacturer’s instructions. DNA was sequenced on a 3130xl Genetic Analyzer (Applied Biosystems).
Plant materials and cultivation conditionsTo develop near-isogenic lines (NILs), we used the durum wheat cultivar ‘Setodure’ (possessing Psy-A1k, see Results) as the recurrent parent (Yanaka et al. 2018). Durum wheat cultivars ‘AC Navigator’ (Clarke et al. 2000) and ‘Enterprise’ (Singh et al. 2010), and experimental line Ap1-22 (Zhang et al. 2005), which possess Psy-A1l, Psy-A1o and Lr19, respectively, were used as donors. Experimental line Ap1-22 had translocated chromosome 7E derived from Th. ponticum including Psy-E1 and Lr19 which approximately 31.7 Mb apart and existed estimated 511 genes (Wang et al. 2020). By backcrossing each of the donors to ‘Setodure’, we developed BC6F2 plants containing each gene of interest by DNA marker-assisted selection. And NIL possessing Lr19 was designated name as Psy-E1 NIL.
The three NILs and ‘Setodure’ were grown in a research field at the Western Region Agricultural Research Center (WARC), Fukuyama, Hiroshima, Japan (lat. 34°30′4″N, lng. 133°23′12″E). Their 756 seeds were sown in a strip 0.7 m wide × 7.2 m long on November 11th and harvested in June. There were 4 replicates of materials in two seasons, 2020–21 and 2021–22 (later referred to as harvest years 2021 and 2022, respectively).
Psy-A1, Psy-E1 and Lr19 genotypingDNA was extracted as described above. All PCR reactions used Thermal Cycler Dice with Quick Taq HS (Toyobo, Osaka, Japan), and the amplified PCR fragments were separated and visualized as above. PCR amplification of Psy-A1 alleles (Psy-A1k, Psy-A1l and Psy-A1o) was an initial 95°C for 2 min, followed by 35 cycles of 94°C for 30 s, 65°C for 30 s, and 68°C for 2 min. Amplification of Psy-E1 was an initial 95°C for 2 min; 10 cycles with a touchdown step of 94°C for 30 s, 65–56°C for 30 s with a decrease of 1°C per cycle, and 68°C for 1.5 min; and then 25 cycles of 94°C for 30 s, 55°C for 30 s and 68°C for 1.5 min. Amplification of Lr19 was an initial 95°C for 2 min; 8 cycles with a touchdown step of 94°C for 30 s, 68–61°C for 30 s with a decrease of 1°C per cycle, and 68°C for 1 min; and then 22 cycles of 94°C for 30 s, 60°C for 30 s and 68°C for 1 min. The multiplex PCR of Psy-A1 and Psy-E1 with HotStarTaq Plus (Qiagen) was an initial 95°C for 5 min, followed by 35 cycles of 94°C for 1 min, 65°C for 1 min, and 72°C for 1 min. Lr19 with Waxy-B1 as an internal control was amplified as for genotyping of Lr19 alone. Primers were listed in Supplemental Table 1, and DNA and deduced amino acid sequences among 3 Psy-A1 alleles and Psy-E1 were aligned and showed in Supplemental Figs. 1 and 2.
Segregation analysisSegregation ratios of Psy-A1 and Psy-E1 were calculated in three F2 populations: Psy-E1 NIL × ‘Setodure’ (Psy-A1k), Psy-E1 NIL × Psy-A1l NIL and Psy-E1 NIL × Psy-A1o NIL. Genotyping was performed by multiplex PCR of Psy-A1 and Psy-E1.
Whole grain, straight flour and semolina flour productionTo produce whole-grain, 30 g of dried seeds were pulverized in a cyclone mill (Cyclotec 1093; Foss, Hillerød, Denmark) with a 0.5-mm sieve.
To produce straight flour by small-scale milling, 150 g of dried seeds were milled using a Quadrumat Jr. (C.W. Brabender Instruments, Inc., South Hackensack, NJ, USA) without a sieve. The milled samples were sieved in a rotary sifter with a 355-μm sieve for 5 min at 360 rpm, and the material that passed through the sieve was used as straight flour.
To produce semolina flour by large-scale milling, seed from 4 replicates was combined, and 2 kg of seeds were milled in a Bühler Laboratory Mill (Bühler, Uzwil, Switzerland) following Cereal & Grains Association method AACC 26-41.01 (AACC Approved Methods of Analysis 1999b).
Pasta productionPasta processing using semolina flour in 2021 and the boiling test were performed following AACC 66-42.01 (AACC Approved Methods of Analysis 1999c) and AACC 66-50.01 (AACC Approved Methods of Analysis 1999d), respectively.
Yellow pigment content (YPC) quantification and b* value of CIELAB measurementYPC was extracted from whole grain, straight flour, and semolina flour by AACC 14-50.01 (AACC Approved Methods of Analysis 1999a) and quantified on a Nanodrop 2000c spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). The b* value of the CIELAB color space, indicative of yellow color intensity, was measured on a CM3500d spectrophotometer (Konica Minolta, Inc., Tokyo, Japan) with light source C and a 2° field for whole grain, straight flour, and semolina flour, and spectrophotometer CM-5 (Konica Minolta) with light source C and a 2° field for dried and boiled spaghetti. The dried spaghetti pieces were lined up without gaps and measured. Spaghetti that had been boiled for 5 min was packed in a transparent plastic bag and measured the next day.
Measurement of agronomic traitsAgronomic traits were assessed in 2020 and 2021 following the International Union for the Protection of New Varieties of Plants (UPOV) guidelines (https://www.upov.int/edocs/tgdocs/en/tg003.pdf). Heading date and maturity date were determined as the accumulated days after sowing. Culm length from the base of the plant to the panicle base was measured by ruler and averaged from 10 culms in each of the 4 replicates. Yield was measured from yielding of 0.7 m wide × 7.2 m long and calculated as kg/a. Spike number per square meter was counted manually and averaged from two areas within each of the 4 replicates. Spikes and grains were harvested from each plot. Awn and spike length were measured with a ruler, and spikelet and grains number per spike were counted manually. Spikelet density was calculated based on spikelet number and spike length. These spike traits were averaged from 10 spikes in each of the 4 replicates. Grain protein content and test weight were analyzed with a near-infrared spectrophotometer (IM9500; Perten Instruments, Stockholm, Sweden). Grain diameter, thousand-kernel weight and grain hardness were measured with a single-kernel characterization system (SKCS4100; Perten Instruments).
Statistical analysisThe YPC and b* values of ‘Setodure’ and NILs were analyzed by ANOVA and Tukey’s HSD at P < 0.05. For segregation analysis of Psy-E1 and Psy-A1, chi-squared analysis was applied with two models. By the same-locus model, the expected segregation ratio for homozygous Psy-E1 : heterozygous Psy-E1/Psy-A1 : homozygous Psy-A1 was 1:2:1. By the different-locus model, the expected segregation ratio for was both of them : only Psy-E1 : only Psy-A1 : null of them as 9:3:3:1. All statistics were analyzed in R version 4.2.0 software (R Core Team 2023).
When the genotype of Psy-A1 in ‘Setodure’ was examined by DNA sequencing with primer walking, it was a complete match to Psy-A1k (FJ393522) in T. turgidum subsp. dicoccoides and to FJ393527 in T. aestivum subsp. Spelta. Thus, the sequence in ‘Setodure’ is classified as Psy-A1k (registered as LC778135).
The Psy1 genotypes of ‘Setodure’ (Psy-A1k), Psy-A1l NIL and Psy-A1o NIL were clearly distinguished by a codominant DNA marker for Psy-A1 alleles, but not Psy-E1 (Fig. 1a). Psy-E1 NIL was clearly distinguished by a dominant DNA marker for Lr19 (Fig. 1d) and amplified with a novel Psy-E1 gene-specific DNA marker (referred to as EU096095 in Th. ponticum), but not Psy-A1, under both simplex and multiplex PCR conditions (Fig. 1b, 1c). F1 progenies crossing Psy-E1 NIL with ‘Setodure’, Psy-A1l NIL or Psy-A1o NIL were amplified same fragments with each parent (Fig. 1a–1d).
Genotyping of materials containing Psy-A1, Psy-E1 and Lr19. Templates were ‘Setodure’, 3 NILs, and 3 F1 hybrids of Psy-E1 NIL × ‘Setodure’ and of Psy-E1 NIL × Psy-A1 NILs. a Codominant marker detecting Psy-A1 alleles. b Dominant Psy-E1–specific marker. c Codominant marker detecting Psy-A1 and Psy-E1. d Dominant Lr19–specific marker with Waxy-B1 as internal control.
To assess the effect of Psy genotype on yellowness of flour and pasta, we measured YPC and the b* value of the CIELAB color space in ‘Setodure’ and the three NILs. YPC and b* were the highest in Lr19 NIL and lowest in ‘Setodure’ (Figs. 2, 3). For both YPC and b*, the genotype effects in order from high to low were Psy-E1 NIL > Psy-A1o NIL > Psy-A1l NIL > ‘Setodure’ (Psy-A1k), and there were significant differences among the genotypes regardless of sample type (whole grain or straight flour) and for both single years and 2-year averages (Figs. 2a, 2b, 3a, 3b).
Yellow pigment content (YPC) of ‘Setodure’ and 3 NILs in 2020, 2021 and 2-year average. Sample type was (a) whole grain flour, (b) straight flour and (c) semolina flour. Values were means ± SE (n = 4 in 2021 and 2022, n = 8 in 2-year average). Same letters in a and b indicate no significant differences by Tukey’s HSD at P < 0.05.
b*value of ‘Setodure’ and 3 NILs in 2020, 2021 and 2-year average. Sample type was (a) whole grain, (b) straight flour, (c) semolina flour, (d) dried spaghetti and (e) boiled spaghetti. Values are means ± SE (a–c: n = 4 in 2021 and 2022, n = 8 in 2-year average) and mean (d, e: 4 replicates were mixed together in). Same letters in a and b indicate no significant differences by Tukey’s HSD at P < 0.05.
YPC and b* in semolina flour in 2021, 2022 and the 2-year average (Figs. 2c, 3c, Supplemental Fig. 3c) and b* in dried and boiled spaghetti in 2021 (Fig. 3d, 3e, Supplemental Fig. 3d) were obtained without replication because large amounts of seed were needed for large-scale milling to obtain semolina flour for making pasta. The orders of the four genotypes were the same for YPC and b* in semolina flour and for b* in dry and boiled spaghetti as those for whole grain and straight flour.
Segregation analysisTo test our hypothesis that Psy-E1 lies at the same locus as Psy-A1, we analyzed segregation ratios in three F2 populations produced by crossing Psy-E1 NIL with ‘Setodure’ (Psy-A1k), Psy-A1l NIL or Psy-A1o NIL by chi-squared analysis with same- and different-locus models. Psy-A1- and/or Psy-E1-specific fragments were amplified in all F2 individuals (Tables 1, 2), none of any F2 individuals amplified neither Psy-A1- nor Psy-E1-specific fragments (Table 2). In three F2 populations produced by crossing Psy-E1 NIL with ‘Setodure’ (Psy-A1k), Psy-A1l NIL or Psy-A1o NIL, both fragments of Psy-E1 and Psy-A1 were amplified only on Psy-E1/A1 heterozygous (Fig. 1c). No significant difference from the expected 1:2:1 ratio was shown for any F2 population in the same-locus model (Table 1), whereas significant differences from the expected 9:3:3:1 ratio were shown for all three F2 populations in the different-locus model (Table 2). These results show that Psy-E1 and Psy-A1 lie at the same locus. Lr19 could be amplified in individuals heterozygous for Psy-A1 and Psy-E1 or homozygous for Psy-E1, but not those homozygous for Psy-A1. This result is consistent with earlier studies showing that Lr19 is closely linked to Psy-E1, and the linkage was unbroken in this analysis (data not shown) although other studies have shown that these genes can be separated (see Discussion).
| Cross | N | Segregation | Chi-squared value* | P-value | ||
|---|---|---|---|---|---|---|
| Psy-E1 homozygous | Psy-E1/Psy-A1 heterozygous | Psy-A1 homozygous | ||||
| Psy-E1 NIL × Setodure (Psy-A1k) | 267 | 71 | 122 | 74 | 2.049 | 0.359 |
| Psy-E1 NIL × Psy-A1l NIL | 237 | 59 | 114 | 60 | 0.116 | 0.944 |
| Psy-E1 NIL × Psy-A1o NIL | 258 | 56 | 135 | 67 | 1.412 | 0.493 |
* Expected ratio of Psy-E1 homozygous : Psy-E1/Psy-A1 heterozygous : Psy-E1 homozygous = 1:2:1 was tested by chi-squared test.
| Cross | N | Segregation | Chi-squared value* | P-value | |||
|---|---|---|---|---|---|---|---|
| Both Psy-E1 and A1 | Only Psy-E1 | Only Psy-A1 | Neither Psy-E1 nor A1 | ||||
| Psy-E1 NIL × Setodure (Psy-A1k) | 267 | 122 | 71 | 74 | 0 | 42.180 | 3.67E-09 |
| Psy-E1 NIL × Psy-A1l NIL | 237 | 114 | 59 | 60 | 0 | 28.242 | 3.23E-06 |
| Psy-E1 NIL × Psy-A1o NIL | 258 | 135 | 56 | 67 | 0 | 25.204 | 1.40E-05 |
* Expected ratio of both Psy-E1 and Psy-A1 : only Psy-E1 : only Psy-A1 : neither Psy-E1 nor Psy-A1 = 9:3:3:1 was tested by chi-squared test.
Our study was conducted in Japan, where the period of wheat maturation to harvest overlaps with the rainy season with moisture stress.
To evaluate the effects of Psy genotype on phenotypes other than yellowness, we tested the characteristics and performance of ‘Setodure’ and the NILs in a 2-year field trial. No significant differences were seen among ‘Setodure’, Psy-A1l NIL and Psy-A1o NIL in any of the investigated agronomic traits (Table 3). Yield, grain number, test weight and thousand-kernel weight were significantly lower in Psy-E1 NIL than in the other lines (Table 3), whereas culm length and protein content of Psy-E1 NIL were significantly higher. Other traits were not significantly different between Psy-E1 NIL and the other NILs (Supplemental Fig. 3a, 3b).
| Heading datea (Days) | Maturity datea (Days) | Spike number (/m2) | Yield (kg/10a) | Culm length (cm) | |
|---|---|---|---|---|---|
| 2021 | |||||
| Setodure | 154 ± 0.3 a | 203 ± 0.4 a | 453 ± 16.0 a | 55.1 ± 0.61 a | 82 ± 0.6 a |
| Psy-A1l NIL | 153 ± 0.3 a | 203 ± 0.3 a | 454 ± 19.3 a | 61.1 ± 2.66 a | 80 ± 0.8 a |
| Psy-A1o NIL | 153 ± 0.3 a | 203 ± 0.3 a | 471 ± 37.0 a | 54.5 ± 2.35 a | 81 ± 0.8 a |
| Psy-E1 NIL | 154 ± 0.0 a | 204 ± 0.3 a | 447 ± 23.2 a | 48.0 ± 1.00 b | 85 ± 1.1 b |
| 2022 | |||||
| Setodure | 147 ± 0.5 a | 202 ± 0.3 a | 414 ± 11.9 a | 51.7 ± 0.99 a | 84 ± 0.9 a |
| Psy-A1l NIL | 145 ± 0.6 a | 201 ± 0.5 a | 401 ± 13.3 a | 56.2 ± 1.04 a | 83 ± 0.8 a |
| Psy-A1o NIL | 146 ± 0.3 a | 201 ± 0.3 a | 433 ± 20.8 a | 54.5 ± 2.31 a | 85 ± 1.2 a |
| Psy-E1 NIL | 147 ± 0.5 a | 201 ± 0.3 a | 415 ± 33.1 a | 43.3 ± 1.26 b | 91 ± 0.8 b |
| Spike length (cm) | Awn length (cm) | Spikelet number (/spike) | Spikelet densityb | Grain number (/spike) | |
|---|---|---|---|---|---|
| 2021 | |||||
| Setodure | 9.2 ± 0.22 a | 6.9 ± 0.44 a | 23 ± 0.5 a | 2.5 ± 0.08 a | 74 ± 2.4 a |
| Psy-A1l NIL | 9.1 ± 0.11 a | 6.0 ± 0.25 a | 25 ± 0.6 a | 2.7 ± 0.08 a | 73 ± 2.0 a |
| Psy-A1o NIL | 9.0 ± 0.31 a | 7.5 ± 0.42 a | 22 ± 0.6 a | 2.4 ± 0.07 a | 74 ± 2.3 a |
| Psy-E1 NIL | 9.3 ± 0.35 a | 6.6 ± 0.39 a | 24 ± 0.7 a | 2.6 ± 0.13 a | 63 ± 1.3 b |
| 2022 | |||||
| Setodure | 9.2 ± 0.36 a | 6.9 ± 0.54 a | 23 ± 0.4 a | 2.6 ± 0.11 a | 71 ± 3.1 a |
| Psy-A1l NIL | 9.5 ± 0.23 a | 6.4 ± 0.56 a | 24 ± 0.5 a | 2.6 ± 0.06 a | 73 ± 1.8 a |
| Psy-A1o NIL | 9.2 ± 0.32 a | 7.3 ± 0.35 a | 22 ± 0.3 a | 2.4 ± 0.12 a | 73 ± 2.0 a |
| Psy-E1 NIL | 9.0 ± 0.22 a | 6.2 ± 0.48 a | 24 ± 0.5 a | 2.6 ± 0.08 a | 60 ± 2.3 b |
| Protein content (%) | Test weight (g/L) | Diameter (mm) | Thousand kernel weight (g) | Hardness (HI) | |
|---|---|---|---|---|---|
| 2021 | |||||
| Setodure | 11.3 ± 0.12 a | 863 ± 2.2 a | 3.27 ± 0.02 a | 52.5 ± 0.84 a | 93 ± 1.2 a |
| Psy-A1l NIL | 11.0 ± 0.14 a | 864 ± 2.2 a | 3.23 ± 0.02 a | 51.6 ± 0.51 a | 94 ± 0.8 a |
| Psy-A1o NIL | 11.3 ± 0.18 a | 862 ± 2.1 a | 3.25 ± 0.02 a | 52.1 ± 1.15 a | 93 ± 0.6 a |
| Psy-E1 NIL | 12.1 ± 0.22 b | 853 ± 2.3 b | 3.19 ± 0.01 a | 45.7 ± 0.56 b | 93 ± 1.0 a |
| 2022 | |||||
| Setodure | 12.2 ± 0.21 a | 846 ± 3.4 a | 3.24 ± 0.02 a | 51.9 ± 0.68 a | 91 ± 1.3 a |
| Psy-A1l NIL | 11.9 ± 0.26 a | 852 ± 2.9 a | 3.26 ± 0.00 a | 51.7 ± 0.22 a | 90 ± 0.4 a |
| Psy-A1o NIL | 11.8 ± 0.24 a | 849 ± 2.5 a | 3.22 ± 0.02 a | 51.2 ± 0.78 a | 90 ± 1.0 a |
| Psy-E1 NIL | 13.2 ± 0.25 b | 831 ± 2.8 b | 3.19 ± 0.02 a | 45.3 ± 0.91 b | 91 ± 0.6 a |
a Sowing day was Nov. 11th in each years. Heading and maturity date were calculated as the accumulated days after sowing.
b Spikelet density = Spikelet number/Spike length
Values are mean ± SE at 4 replicates (averages of 2 areas for spike number, 10 culms for culm length, and 8 spikes for spike traits). Within the same column, data marked with the same letter are not significant difference by Tukey’s HSD at P < 0.05.
The chromosome segment containing Lr19, derived from the wild relative Th. ponticum, confers resistance to leaf rust in wheat (Gennaro et al. 2009) and increases the yellowness in seeds, but the degree of its effect on yellowness in durum wheat was unknown. In this study, yellowness of flour and pasta as measured by YPC and b* values in multiple years and in different sample types decreased in the order of Psy-E1 NIL > Psy-A1o NIL > Psy-A1l NIL > ‘Setodure’ (Psy-A1k), (Figs. 2, 3). Singh et al. (2009) reported that b* value decreased in the order of Psy-A1l > Psy-A1o > Psy-A1a in two panels as modern cultivars and landraces in durum wheat, but no significant difference between Psy-A1l and Psy-A1o under some conditions. Lr19-possessing lines were generated in several studies, and their seeds and flour showed strong yellowness (Prins et al. 1997, Rai et al. 2019, Zhang et al. 2005). This is the first report that the yellowness conferred by Lr19 is higher than that conferred by any of the Psy-A1 alleles in durum wheat.
Lr19 is linked with Y on chromosome 7E from Th. ponticum, which confers yellow pigmentation in endosperm (Ayala-Navarrete et al. 2013, Sharma and Knott 1966, Zhang and Dubcovsky 2008). Here, Psy-E1 NIL had high YPC and carries Lr19 and Psy-E1 (Figs. 1–3), consistent with previous research. Interestingly, Psy-E1 NIL did not show amplification of any Psy-A1–specific amplicons and only F1 crossing Psy-E1 NIL with ‘Setodure’ (Psy-A1k) or Psy-A1 NILs show both Psy-A1–specific and Psy-E1–specific amplicons (Fig. 1). In addition, both fragments of Psy-E1 and Psy-A1 were detected only on Psy-E1/A1 heterozygous in F2 populations. Taking these facts into consideration, we clarified by using segregation analysis that Psy-E1 is not tandem arrangement with Psy-A1 but is located at the Psy-A1 locus (Table 1). These results provide further evidence that Psy-E1 is a candidate for the Y gene linked with Lr19.
We observed strong linkage between Lr19 and Psy-E1, consistent with previous research (Knott 1980, Marais and Marais 1990). This linkage block was not broken in this study (Table 1). However, Prins et al. (1997) revealed that white-endosperm lines with Lr19 lacked the Y locus from a translocated segment. Xu et al. (2023) also succeeded in breaking this linkage with homoeologous recombination and established a white-endosperm line lacking Psy-E1 but carrying Lr19. Wheat breeders would have to choose Psy-E1 or Psy-A1 alleles for controlling yellowness, as Psy-E1 is high likely to locate at same locus as Psy-A1.
Here, Psy-E1 NIL had both higher culm length and protein content, and lower yield than the Psy-A1–containing lines, with lower grain number per spike, test weight and thousand-kernel weight than the recurrent durum wheat parent ‘Setodure’ and Psy-A1 NILs in the ‘Setodure’ genetic background (Table 3). The results of negative correlation of yield and protein content in seeds was consistent with its relationship in cereals (Simmonds 1995). In bread wheat, yield was increased in Lr19-containing NILs in various backgrounds in India (Rai et al. 2019) and Mexico (Reynolds et al. 2001). The yield and number of grains per m2 were increased in Lr19-containing NILs under appropriate irrigated conditions compared to under drought conditions (Monneveux et al. 2003) and yield increase with biomass and number of spikes per m2 elevate was observed under non-moisture conditions with appropriate irrigations compared to under moisture conditions (Singh et al. 1998). They concluded that the translocation of chromosome 7E may be useful for enhancing yield under favorable conditions (Monneveux et al. 2003, Singh et al. 1998). Monneveux et al. (2003) pointed out that agronomic phenotypes of the effect of the chromosome 7E translocation depend on the recipient genotype. Thus, the agronomic potential and effectiveness of the 7E segment including Lr19 and Psy-E1 will need to be assessed under appropriate cultivation conditions and in various durum wheat backgrounds.
Functional characterization of PSY in carotenogenesis has revealed various regulatory stages acting at the transcriptional, post-transcriptional, post-translational and epigenetic levels (Zhou et al. 2022). The effects of engineering the carotenoid biosynthetic pathway through manipulation of Psy1 have been researched in many plants, because PSY is a highly conserved enzyme that catalyzes the first step in the carotenoid biosynthesis pathway and is a major rate-limiting enzyme (Cong et al. 2009, Lindgren et al. 2003, Zhou et al. 2022).
Alternative splicing occurs in Psy-A1 owing to duplication in an exon, producing aberrant transcripts and decreasing the yellow color of endosperm in bread wheat (Howitt et al. 2009). Cultivars of durum and bread wheat with strong yellow endosperm had higher Psy1 gene expression (Qin et al. 2016). Tissue-specific expression and overexpression of bacterial PSY genes dramatically increased provitamin A in bread wheat (Wang et al. 2014). However, functional characterization of Psy-A1 alleles and Psy-E1 in wheat is still incomplete, and the effects of these alleles need to be better understood to lead to effective durum wheat breeding in the future.
In this study, Lr19-linked Psy-E1 conferred higher yellowness in endosperm, a trait desirable for pasta production. However, it was accompanied by yield loss under the field conditions in this study, which were typical for wheat-growing areas of Japan. Among the future challenges for durum wheat breeding will be to break the linkage between Psy-E1 and yield loss to enable application of Psy-E1 in breeding programs, and to explore the favorable combination of Psy-A1 and Psy-B1 in durum wheat to produce yellowness as strong as that conferred by Psy-E1. Although Psy-B1 allele was still unknown in ‘Setodure’, the effect of Psy-B1 alleles may increase YPC. For Psy-B1 alleles in durum wheat, Psy-B1a and Psy-B1b had the same effect on YPC (Campos et al. 2016), and Psy-B1f had a stronger effect on YPC than Psy-B1g (He et al. 2009).
KK: Conceptualization, data curation, investigation, resources, validation, visualization, writing original draft. YB and MI: Data curation, investigation, validation. MY: Data curation, MY, WF and KT: Resources. HO, TT, HK and MY: Data curation, validation. All authors read and approved the final manuscript.
This study was funded by NIPPN Corporation.
