We have used 16 diverse genotypes of wheat selected from the core set developed for improved grain quality at the ICAR-Indian Agricultural Research Institute (ICARI), New Delhi. The list of wheat genotypes used in the present investigation has been presented in the Supplementary Material ( Supplementary Table S1 ). The pre-treated seeds were sown in pots (22.5 cm diameter) at the Nanaji Deshmukh Plant Phenomics Centre, Pusa, ICAR-IARI, New Delhi, and recommended intercultural operations were followed along with the timely application of fertilizers and irrigations. The pots were kept in a regulated glasshouse at a temperature of 22 ± 3°C (daytime) and 18 ± 3°C (nighttime) during the vegetative stage, and at 28 ± 3°C (daytime) and 32 ± 3°C (nighttime) during the pollination and grain-filling stages. Other parameters, such as light and RH, were maintained, as mentioned in our previous publication (Kumar et al., 2022). Only healthy plants were tagged and further used for sample collection at a specific stage based on the Feekes scale. Samples were collected in three biological replicates from the tagged plants and were packed in an airtight container and stored at -20°C for further downstream analysis.

The freshly harvested grains from different genotypes of wheat were further subjected to fine grinding using a mechanized mini-grinder machine (Kolar Mill Stores Pvt. Ltd. Karnataka, India), keeping the temperature constant in order to protect them from the adverse effects of milling temperature on biochemical constituents. We used Ellman’s test to estimate the free thiols in the ground sample (Ellman, 1959). In brief, 10 µL of powder extract (prepared by passing the slurry prepared in green solvent through a chromatographic column) was placed in a test tube along with 50 µL of Ellman’s reagent [50 mM of sodium acetate (NaoAc), 2 mM of 5’5-Dithio-bis-(2-nitobenzoic acid) in water, and 1M of Tris solution (pH 8.0)], 100 µL of Tris solution, and 840 µL of distilled water. The absorbance of the solution was taken at 412 nm using a UV-Vis spectrophotometer. The thiol content was calculated using the following equations:

The thiol content was expressed as µmol/g of the sample.

The total RNA was extracted from the developing seeds (high thiol content - Halna, HD2985; Low thiol content - NIAW34, PBW443) collected during the button (G₀), milky-ripe (G₁), and seed hardening (G₂) stages using a RaFlex Total RNA Isolation Kit (GeNei, USA). The quality of the isolated total RNA was checked using a BioAnalyzer (Agilent, USA), and an OD 260 to 280 ratio of > than 1.9 was used for the expression analysis. We have selected the thiol-reducing genes [Thioredoxin (TRX, acc. no. AF438359.1), Glutaredoxin (GRX, acc. no. AF542185.1), and Glutathione reductase (GR, acc. no. XM_044564013)] for the oligo design using the Genefisher2 Primer design software (https://bibiserv.cebitec.uni-bielefeld.de/genefisher2/) and it was synthesized commercially ( Table 1 ). cDNA was synthesized using the Revert^Aid H Minus First Strand cDNA Synthesis Kit (Thermo Fisher Scientific) following the protocol, as given by the manufacturer. The expression analysis was performed using the Bio-Rad CFX96 platform, and the steps were followed as mentioned in our previous publication (Kumar et al., 2022). We used the β-actin gene (acc. no. JQ004803.1) as the endogenous control gene for normalizing the C_t value. The relative fold expression (2^^-ddCt value) was calculated using the Pfaffl method (Pfaffl et al., 2002).

In order to know the abundance of cysteine and methionine (Met) amino acids residues in regulatory genes linked with different metabolic pathways associated with the biosynthesis of yellow pigment and other proximate composition (macro- and micronutrients), we retrieved the amino acid sequences of the following genes from NCBI that are involved in different biosynthesis pathways and have been cloned from contrasting wheat cvs.: phytoene synthase gene (PSY, NCBI acc. no. FJ393546), which is involved in yellow pigment synthesis; ADP glucopyrophosphorylase gene (AGPase-L, acc. no. DQ406820), Triticum aestivum ADP glucopyrophosphorylase-1 gene (TaAGPase-1, acc. no. KC347594), starch synthase (SS, acc. no. AY050174), Starch synthase-III (SS-III, acc. no. AF258608), and pullulanase gene (acc. no. XM_044583776), which are involved in the starch biosynthesis pathway; phytoene desaturase (PDS, acc. no. FJ517553) and β-Carotene hydroxylase (BCH, acc. no. JX171675), which are involved in the tannin biosynthesis pathway; phenylalanine ammonia lyase gene (PAL, acc. no. XM_044557325), which is involved in the biosynthesis of phenolic compounds; chalcone synthase 2-like (CHS, acc. no. XM_044543986) and inositol hexakisphosphate kinase gene (IP6K1, acc. no. XM_044517289), which are involved in the biosynthesis of phytic acid. Further, the amino acid sequences were characterized using the Prot pi tool available online (https://www.protpi.ch/Calculator/ProteinTool). The abundance of Cys/Met was presented as a percentage of the amino acids.

We estimated the Cys/Met in the grains following the standardized method at our Nutrition Analysis Lab, ICAR-IARI, New Delhi. In brief, 50 µg of the sample was taken and transferred into a clean glass vial. To 50 mL broth tubes, 20 mL of 6N HCl was added, and the glass vial containing the sample was dipped into the tube and sealed with parafilm. The tube was placed in a dry bath at 60°C under N₂ gas for 15 min to maintain inertness. Further, the temperature was increased to 110°C and it was incubated for 24 h for complete hydrolysis of the protein sample (Marino et al., 2010). The filtrate was then evaporated in a vacuum flash evaporator and was made acid free by repeated washing with distilled water and subsequent evaporation. The hydrolyzed sample was further subjected to pre-column derivatization using O-phthalaldehyde in the presence of βxA7B5;-mercaptoethanol. Further, 80 µL of the derivatized sample was injected in high-performance liquid chromatograph (HPLC) (Shimadzu, Model CBM 20 A) equipped with a C18 reverse phase (RP) column and a fluorescence detector. The amino acids were identified and quantified by comparing the retention times and peak areas with those of the Sigma standard. The concentration of amino acids was presented as the percentage of a mole.

The estimation of yellow pigment (β-carotene) in the sample was carried out as per the modified method of Kumar et al (2022), published in the technical bulletin of ICAR-Indian Agricultural Research Institute, New Delhi, India. We used 1 kg of treated wheat grains (conditioned to 13% moisture and debranned to 6% using a fabricated peeling machine available at the Agricultural Engineering Workshop at ICAR-IARI, New Delhi, India). We used 40 g of fine powdered (100 mesh size particles) from each genotype prepared using a mini-miller machine (Kolar Mill Stores Pvt. Ltd. Karnataka, India). The fine powder was stored in air-tight bags in two groups for further downstream characterization. The estimation of total yellow pigment was carried out by using the cold percolation method. The fine powdered flour (40 g) was loaded onto the chromatograph column packed with cotton and sodium sulphate. Further, 75 mL of n-butanol was used for the extraction of the eluent from the bottom in a fresh test tube. The extract was packed in dark colored bottle and was used for the estimation of yellow pigment by measuring the absorbance at 412 nm. The yellow pigment was calculated using the extinction coefficient of 13,000 mM cm^-1.

The collected samples (0.1 g) were fractionated in 80% ethanol (hot) as per the protocol of Thayumanavan and Sadasivam (1984). In brief, the residue retained after centrifugation was washed repeatedly until the absence of color formation by the anthrone reagent. The dried residue was mixed with 5.0 mL of water and 6.5 mL of perchloric acid (52%). The mixture was incubated at 0°C for 20 min followed by centrifugation at 12,000 rpm for 10 min. The extraction using fresh perchloric acid was repeated twice, and further volume was made up to 100 mL. An aliquot of 0.1 mL was taken, and the volume was made up to 1 mL with water. Further, 4 mL of anthrone reagent was added, and the reaction mixture was heated for 8 min in a boiling water bath, followed by rapid cooling. The intensity of the green to dark green color was read by measuring OD at 630 nm. The glucose standard was used to estimate the concentration of starch by multiplying it by a factor of 0.9.

RS in the grain was estimated using the Resistant Starch Assay kit (Megazyme) (Englyst et al., 1992). In brief, 50 g of grains were used for the estimation following the protocols as provided by the manufacturers. A mixture of 0.1 mL of aliquot and 3.0 mL of glucose oxidase/peroxidase (GOPOD) reagent was incubated at 50°C for 20 min, and the OD was read at 510 nm against the reagent blank. The RS was calculated using the D-glucose solution standard provided in the kit.

The total phenolic content was estimated using the method of Singleton et al. (1999). In brief, 0.5 mL of supernatant (extract from the slurry by passing the mixture through a chromatograph column) was mixed with 5 mL of 1 M Folin–Ciocalteu reagent (FCR). Further, the solution was neutralized by adding 4 mL of 75 g/L saturated sodium carbonate and the solution was kept at room temperature (RT) for at least 2 h. The solution was further used for measuring the absorbance at 765 nm. The total phenolic content (TPC) was measured using Gallic acid as a standard curve, and the result was presented in terms of Gallic acid equivalents (mg GAE/g DW).

The tannin content was estimated using the Folin–Ciocalteu reagent method (Everette et al., 2010). We used 0.1 mL of extract and a further reaction mixture was prepared by adding 0.5 mL of FCR, 1 mL of 35% sodium carbonate (Na₂CO₃), and 7.5 mL of double-distilled water, followed by incubation at RT for 30 min. The RM was further used to measure the absorbance at 700 nm against the blank. The total tannin content was calculated using the standard curve of tannic acid and expressed as mg/mL.

The finely ground grains (1.0 g) were digested using 20 mL of the digestion mixture (consisting of HNO₃ and HClO₄ in the ratio of 9:4) using the digestion chamber. The mixture was evaporated until the volume was reduced to 3 to 5 mL. The flasks were then cooled at room temperature, and double-distilled water was added to make the volume 50 mL. The clear, digested solution in each flask was filtered using Whatman filter paper No. 42 and used to analyze the micronutrients. The micronutrients Zn and Fe were analyzed using an atomic absorption spectrophotometer (AAS, Electronics Corporation of India Limited, Hyderabad, India) by measuring the absorption at 213.9 nm and 248.3 nm for Zn and Fe, respectively. The concentration was estimated using the pure standard and the content was expressed as μg/g DW.

The finely ground grains (1 g) were used for the phytic acid estimation by digesting the powder using 20 mL of hydrochloric acid (0.66 M) overnight. Further, 0.5 mL of the extract was neutralized by adding 0.5 mL of NaOH (0.75 M) and used for the enzymatic dephosphorylation reaction using the phytic acid assay kit (Megazyme). The supernatant was used for the colorimetric determination of phosphorus by measuring the absorbance at 655 nm and the calculation was conducted as given in the instruction manuals.

The crude nitrogen present in the grain was estimated using the micro-Kjeldahl method (Miller and Houghton, 1945). In brief, the flour (100 mg) was digested in strong sulfuric acid (10 mL) along with 2 g of K₂SO₄ and CuSO₄ (in a ratio of 10:1) for 90 min at 420°C, followed by distillation using 40% NaOH and 4% Boric acid. The trapped ammonia was determined through titration with 0.1 N HCl and the burette reading was used to calculate the percentage of nitrogen using the formula -

The percentage of nitrogen calculated was used for the calculation of total protein content using the conversion factor of 5.81.

The gluten was estimated in the finely ground flour following the traditional method of washing the dough with water. We used 2 g of flour and added 1–2 mL of water for dough preparation. The prepared dough was wrapped in two layers of chicken cloth and washed with distilled water until a fine elastic layer was left over. The gluten leftover was dried in an oven at 72°C for overnight, and further weighed for the calculation of the percentage of gluten in the flour.

The samples were three biological replicates for the biochemical analysis, and we used the statistical tool of one-way ANOVA to analyze the variance of different parameters for statistically significant differences (P ≤ 0.05). The correlation between the biochemical traits was analysed using Pearson’s correlation coefficient (r, p<0.01).

Thiol is a very important component of storage proteins, providing stability and strong networking in dough and many desirable rheological characteristics in flour. It is considered one of the desirable traits for analyzing the quality of flour. We have analyzed the thiol content in the grains of diverse genotypes of wheat and observed the highest thiol content in the grain of wheat cvs. Halna followed by HD2985 and Raj3765 ( Figure 1a ). The thiol content was observed to be th elowest in wheat cv. NIAW34.

In order to understand the effect of the changes in the thiol content on developing endosperm tissues, we collected samples at different grain-filling stages: button, milky-ripe, seed hardening, and mature grains, and used them for the thiol estimation ( Figure 1b ). Wheat cv. Halna had the highest thiol content at different sub-stages of endosperm development, as compared to other cvs., whereas the lowest was observed in PBW343 and NIAW34 during different sub-stages. We observed significant variations in the flour thiol content across the genotypes (F-values are significant at p ≤ 0.05). Most of the wheat cvs., which are popular among consumers for flour and chapatti-making quality (HD2985, Raj3765, etc.), had very high thiol content, as compared to other genotypes. Manu and Rao (2008) reported that high thiol content in flour decreases the toughness of chapatti.

Disulfide bonds (SS) are formed by the oxidation of sulfhydryl groups (SH) and have been reported to be involved in stabilizing the folded conformation of gliadin (monomeric proteins). The polymeric structure of dough is also stabilized due to the presence of disulfide bonds. We analyzed the disulfide bonds in the grains of diverse genotypes of wheat. The disulfide content was observed to be the highest in the grains of wheat cv. HI1544 followed by Halna and PBW443 ( Figure 2c ). The lowest disulfide content was observed in wheat cv. UP-2506. Disulfide bonds are linked with the empirical rheology of flour and are considered one of the important traits.

We analyzed the pattern of increase in the disulfide content in the developing endosperm tissue in diverse cvs. and observed the highest disulfide content in wheat cv. HI1544 and the lowest in UP2506 at different sub-stages of endosperm growth and development ( Figure 2b ). Manu and Rao (2008) observed that protein disulfide content showed a positive correlation with dough hardness and the texture of chapatti. The thiol content has been observed to influence the formation of disulfide bonds in flour and helps in improving the rheological characteristics of dough. Various studies have shown that GSH-dependent protein-disulfide oxidoreductase (TPDO) present in the developing wheat kernels manipulate the quality matrix of dough, causing stiffness. Disulfide bond dynamics have been reported to enhance the dough extensibility in wheat (Osipova et al., 2007).

We selected three genes—TRX (acc. no. AF438359.1), GRX (acc. no. AF542185.1), and GR (acc. no. XM_044564013)—to assess the thiol-based redox sensing in developing endosperm (sub-stages G₀, G₁, and G₂) of Halna and HD2985 (high-thiol content genotypes) and NIAW34 and PBW443 (low-thiol content genotypes). TRX expression was observed to be higher (2.4-fold) in wheat cv. Halna during the G₂ stage compared to the G₀ stage ( Figure 3a ). The expression was observed to be the lowest in wheat cv. PBW443 compared to other genotypes. Overall, we observed an abundance of TRX transcripts in the developing grains of genotypes with high thiol content (Halna and HD2985) compared to low thiol-containing genotypes (NIAW34 and PBW443).

We observed similar patterns of GRX expression in contrasting wheat genotypes ( Figure 3b ). The relative fold expression was observed to be the highest (4.9-fold) in wheat cv. Halna during the G₂ stage, whereas it was observed to be the lowest (3.1-fold) in cv. NIAW34 compared to G_o. Sainz et al. (2022) reported the differential expression of 12 Trx and 12 Grx genes in response to water deficit stress in soybean.

Similarly, expression analysis of GR showed a relatively higher expression (9.9-fold) in wheat cv. HD2985 during the G₂ stage compared to G₀ ( Figure 3c ). The GR expression was observed to be the lowest (4.1-fold) in wheat cv. PBW443 during the G₂ stage compared to G_o. We observed significant variations (p ≤ 0.05) in the expression of TRX, GRX, and GR during different sub-stages of endosperm development in contrasting wheat cvs., with an abundance of transcripts in Halna and HD2985 compared to NIAW34 and PBW443. He et al. (2023) identified 85 GRX genes in common wheat using a bioinformatic method and reported differential upregulation of these genes under biotic and abiotic stresses in wheat.

The ratio of Cys: Met has been reported to play very important role in deciding the structural and thermal stability of protein under oxidizing conditions (Bhopatkar et al., 2020). In order to analyze the Cys/Met ratio, we retrieved the amino acid sequences of some of the regulatory genes (cloned from Triticum aestivum) linked with different metabolic pathways associated with the biosynthesis of macro- and micronutrients in the grains, i.e., yellow pigment biosynthesis (PSY, acc. no. FJ393546), starch biosynthesis [AGPase-L, (acc. no. DQ406820), TaAGPase-1 (acc. no. KC347594), SS (acc. no. AY050174), SS-III (acc. no. AF258608), and pullulanase gene (acc. no. XM_044583776)], tannin biosynthesis [PDS (acc. no. FJ517553) and BCH (acc. no. JX171675)], phenolic compound biosynthesis (PAL, acc. no. XM_044557325), and phytic acid biosynthesis [CHS (acc. no. XM_044543986) and IP6K1, (acc. no. XM_044517289)]. The amino acid sequences were characterized for the estimation of Cys/Met in the amino acid pool ( Figure 4A ). GBSS is considered one of the important enzymes involved in the starch biosynthesis pathway. We observed 2.25% Cys and 2.95% Met in the amino acid sequence of GBSS cloned from wheat ( Figure 4A ). Similarly, in the case of pullulanase, which is involved in RS synthesis, it had 0.98% Cys and 2.7% Met in the amino acid pool. The Cys content was observed to be the highest (2.25%) in GBSS and Met was observed to be the highest (4.04%) in the BCH gene involved in tannin synthesis. The ratio of Cys: Met was observed to be the highest in GBSS (0.77) and the lowest in BCH (0.08). Other genes also showed the same pattern of Cys and Met in their amino acid sequences. All these key genes were observed to have a high percentage of Cys and Met, which provide stability to the proteins against various factors. Walker et al. (2022) reported that S-containing amino acids protect the proteins from oxidation and correlated it with the folding stabilities of the proteins under stress.

We estimated the Cys and Met concentration in the grains of diverse cvs. of wheat in order to understand its effect on different metabolic pathways linked with various macro- and micronutrients. The cysteine content was observed to be the highest (3.56% mole) in wheat cv. Halna (thermotolerant) followed by HD2985 (3.40% mole), whereas it was observed to be the lowest (1.25% mole) in NIAW34 (thermosusceptible) ( Figure 4B ). Similarly, the methionine content was observed to be the highest (2.20% mole) in wheat cv. HD2985 (thermotolerant) and to be the lowest (0.33% mole) in cv. HD1914 (thermosusceptible). We observed significant (p ≤ 0.05) variations in the Cys and Met content in the grains among the diverse genotypes. Most of the genotypes that have been reported to be thermotolerant, such as HD2985, Halna, and HD2932, showed very high Cys and Met contents, whereas thermosensitive cvs., such as BT-Schomburgk, HD1914, and PBW343, showed low Cys and Met contents in the amino acid pool of the grains. The abundance of Cys and Met may be the reason behind the high stability of the proteins and tolerance level of the plant under heat stress, since it behaves as a strong antioxidant, enhancing the stability of the enzymes involved in various pathways responsible for the synthesis of many macro- and micromolecules. Liu et al. (2020) reported that higher concentrations of glutathione maintain the redox balance and enhance the accumulation of protein and S-containing amino acids in the grains of maize, which in turn improves the grain quality.

Yellow pigment is essentially carotenoids such as lutein and α/βxA7B5;-carotene, and these are considered potential natural antioxidants with many medical benefits. It also provides the shining color of the grains and is one of the traits used to decide the export value of wheat grains. This is one of the deciding factors in the popularization of wheat grains and consumer preference. Many processed products require wheat grains to have a very high content of yellow pigment. We also analyzed the yellow pigment in the grains of diverse genotypes of wheat to decide their quality. The yellow pigment content was observed to be the highest in the grains of wheat cv. NIAW34 (6.08 µg/g dry matter), followed by BT-Schomburgk (5.71 µg/g dry matter), whereas the content was observed to be the lowest in wheat cv. UP-2506 (1.01 µg/g dry matter) ( Figure 5a ). We observed significant variations (p ≤ 0.05) in the total yellow pigment content in the diverse genotypes selected for the present study. Butler and Ghugre (2020) reported that β-carotene acts as a potential iron enhancer and limits the phytate-chelating mechanism. It causes an increase in the availability of ionizable iron, which is considered an indicator of the bioavailability index. Wheat genotypes with a high yellow pigment are recommended to be used in the pasta industry, and it is one of the desirable traits.