Gene-Editing-Mediated Enhancement of Carotenoid Compound Accumulation in Common Wheat Grains

Abstract

Common wheat (Triticum aestivum L.) is a staple food crop for humans, yet it primarily accumulates the non-provitamin A carotenoid lutein and exhibits limited natural variation in provitamin A β-carotene among its various accessions. This characteristic necessitates the development of alternative strategies for provitamin A biofortification. To address this challenge, we targeted key control points in the carotenoid biosynthetic pathway using the CRISPR-Cas9 system in a wheat cultivar Fielder. Specifically, we knocked out the gene encoding lycopene ε-cyclase (LCYE), an enzyme that acts as a gatekeeper opposing the production of β-branch carotenoids. Biochemical analysis of homozygous transgene-free mutant endosperms at 20 days post-anthesis (DPA) revealed marked metabolic rerouting of carotenoid biosynthesis, characterized by differential, line-specific accumulation patterns. Provitamin A carotenoids—specifically β-carotene—increased by 26.1–34.5% relative to wild-type controls, concomitant with elevated 22.9–125.4% for zeaxanthin, 41.6–73.9% for violaxanthin, and 26.2–186.5% for antheraxanthin. However, these gains were offset by drastic lutein reduction in lines 1–4 and 5–1. Consequently, total carotenoid levels displayed non-uniform responses, with line 5–1 exhibiting a modest decrease relative to wild-type. Moreover, the mutant lines exhibited elevated levels of amylose and soluble sugar, and the seed coats and endosperms of the triple homozygous transgene-free mutant lines exhibited an orange-yellow hue. In conclusion, we have successfully developed novel carotenoids biofortified wheat lines through a gene-editing approach. This study demonstrates targeted redirection of carotenoid biosynthesis via gene editing as an effective strategy to enhance the nutritional value of commercial wheat and mitigate micronutrient deficiencies in modern food systems.

1. Introduction

Carotenoids serve as precursors for the synthesis of vitamin A, a nutrient that humans cannot synthesize de novo. Consequently, vitamin A must be obtained through dietary intake. Vitamin A deficiency (VAD) can make serious health consequences, with mild cases manifesting as night blindness and severe cases potentially leading to permanent blindness or even death [1,2,3]. However, these vital compounds remain chronically under-consumed in contemporary diets. Biofortification offers a durable, pragmatic strategy with untapped potential to combat vitamin A deficiency in low-income countries that depend heavily on staple crops. In 2025, researchers demonstrated that marker-assisted selection is an efficient route to provitamin-A-enriched maize, drastically cutting breeding costs and accelerating cultivar release [4]. Among carotenoids, β-carotene is the principal plant-derived precursor of retinol (vitamin A), essential for a wide range of physiological function [5]. These functions include night vision, cell proliferation, reproduction, embryonic development, brain activity, and immune system regulation [6,7]. Other carotenoids, such as zeaxanthin, antheraxanthin, and violaxanthin, also play crucial roles as antioxidants, neutralizing harmful reactive oxygen species [2].

Carotenoids constitute a diverse group of naturally occurring lipophilic compounds that are extensively distributed in green plants, algae, fungi, and bacteria. These compounds belong to the C40 isoprenoid class and typically consist of eight isoprenoid units [8]. The biosynthetic pathway of carotenoids begins with geranylgeranyl pyrophosphate (GGPP) as the precursor [9]. The first committed step in this pathway is that GGPP is catalyzed by phytoene synthase (PSY), which facilitates the condensation of two GGPP molecules to produce 15-cis-phytoene. Subsequently, phytoene undergoes a successive of desaturation reactions to form lycopene [10]. These desaturation steps are mediated by phytoene desaturase (PDS), ζ-carotene desaturase (ZDS), ζ-carotene isomerase (ZISO), and carotenoid isomerase (CRTISO) enzymes, respectively [11].

The cyclization of lycopene is a critical branching point in carotenoid biosynthesis, diverging into two distinct pathways. One pathway leads to the formation of α-carotene, and the other results in the production of γ-carotene and β-carotene. Plants contain two types of cyclase enzymes which mediate the process of ε-cyclase (LCYE) and β-cyclase (LCYB), respectively [12,13]. The LCYE and LCYB catalyzes the cyclization of one end of lycopene to form an ε-ring, yielding α-carotene. In contrast, LCYB cyclizes both ends of lycopene to form β-rings, thereby producing γ-carotene and β-carotene [10,12,14].

Further downstream in the plant carotenoid biosynthetic pathway, α-carotene and β-carotene undergo additional modifications to generate more complex oxygenated derivatives. These modifications are catalyzed by carotenoid hydroxylase (HYD). Specifically, α-carotene is predominantly converted into α-cryptoxanthin and lutein [14]. Meanwhile, β-carotene is transformed into a variety of xanthophylls, including β-cryptoxanthin, zeaxanthin, antheraxanthin, violaxanthin, and neoxanthin (Figure 1) [15].

The gene Lcye encoding ε-cyclase is a key determinant in the carotenoid biosynthetic pathway, exerting significant influence over the content and compositional ratios of carotenoids in plants. Subsequent studies have demonstrated its pivotal role in modulating carotenoid profiles across various plant species. For instance, the deficiency of LCYE enzyme in Arabidopsis results in the failure to synthesize lutein and a concomitant increase in β-carotene levels [16]; in the tomato plants with silenced Lcye, total carotenoids and β-carotene levels were elevated, while lutein content was reduced [17]; downregulation of the Lcye gene in rapeseed (Brassica napus) showed markedly elevated carotenoid levels in seeds [18]. Similarly, targeted editing of the Lcye gene in bananas led to a six-fold increase in β-carotene content in fruits, albeit with concurrent reductions in α-carotene and lutein levels [19]. In potatoes, downregulation of Lcye led to a 14-fold increase in β-carotene content and a 2.5-fold rise in total carotenoid content, with the potato callus tissue exhibiting a deep yellow hue [20,21]. Moreover, biochemical analyses of durum wheat mutant lines revealed a 75% increase in β-carotene content in grain in the complete mutant line compared to the control [22]. Collectively, these findings underscored the potential of Lcye as a valuable genetic tool for enhancing β-carotene content in plants, thereby holding significant promise for breeding programs aimed at improving nutritional quality.

Wheat is one of the largest cultivated cereals grown on almost 17% of the world’s agricultural land and serves as a staple crop globally; it serves as a crucial source of plant-based protein, minerals, and vitamins for human nutrition. The quality of wheat can be broadly categorized into processing quality and nutritional quality [23]. With the continuous improvement of living standards and the deepening awareness of healthy eating, the cultivation of wheat varieties with enhanced nutritional quality has gained increasing attention. Among these nutrients, vitamin A content is one of the key indicators for assessing the nutritional quality of wheat, playing a critical role in human health. While non-provitamin A carotenoids such as lutein predominantly accumulate in wheat endosperm, provitamin A species β-carotene remain remarkably scarce, ranging from merely 0.03 to 0.16 μg g⁻¹ in diverse Indian winter wheat varieties [24,25,26,27]. It is necessary to develop the wheat germplasm with enhanced provitamin A carotenoids by gene editing technology.

In this research, we employed the CRISPR gene editing system to target the TaLcye gene in wheat, aiming to enhance the nutritional profile controlled by this gene. Results showed that the carotenoid contents—β-carotene, zeaxanthin, antheraxanthin and violaxanthin—were substantially elevated in the grains at 20 DPA. Our results highlight the potential of gene editing as an effective approach for biofortification, providing a viable solution to address micronutrient deficiencies in major crop.

2. Materials and Methods

2.1. Plant Materials and Cultivation Conditions

A wheat cultivar Fielder was acquired from the Crop Germplasm Bank of China. Wheat seeds were grown in pots (20 cm × 30 cm) in a growth chamber maintained at 24 °C, with 16/8 h light/dark cycle, 300 μmol m⁻² s⁻¹ light intensity, and 45% humidity from February 2025 to May 2025. At 15–17 DPA, immature wheat grains were collected for Agrobacterium-mediated transformation (the Agrobacterium strain was C58C1), following the method described by Wang et al. (2017) [28].

2.2. Construction of the Vectors

The expression vector pWMB110-Cas9 for gene editing was constructed in our previous work by the homologous recombination method and inserting the Cas9 gene into the multiple cloning site (MCS) of pWMB110, which contains the bar gene as a selection marker for generating transgenic plants and the maize ubiquitin (UBI) promoter for driving the expression of Cas9 gene [29]. Moreover, a guide RNA sequence 5′-GTATGGGAGGACGAATTCAA-3′ was designed to target the three homologous TaLcye genes based on the sequences of TaLcye-3A (GeneBank accession ACF42349.1), TaLcye-3B (GeneBank accession ACF42350.1), and TaLcye-3D (GeneBank accession ACF42351.1) [22]. This guide RNA was driven by the TaU3 promoter and inserted into the multiple cloning site (MCS) of the pWMB110-Cas9 plasmid through homologous recombination, resulting in the final construct named TaLcye-CRISPR (Figure S1A,B).

2.3. Detection Analysis of the TaLcye-Edited Mutations

Genomic DNA was extracted from the candidate T0 transgenic wheat mutants, in which the targeted genes, TaLcye-3A (5′-GTCTGGTAGCATCCTTCAAGTAG-3′ and 5′-TATCATGCTGCCTGATGCAACTG-3′), TaLcye-3B (5′-GTCTGGTAGCATCCTTCAAGTAG-3′ and 5′-ACTAGAAACAATCATATATA-3′), and TaLcye-3D (5′-GTCTGGTAGCATCCTTCAAGTAG-3′ and 5′-GCTGCCTGACGCGACTGAATT-3′), were amplified using their respective specific primers. The bar gene was screened using the primer pairs 5′-ACCATCGTCAACCACTACATCG-3′ and 5′-GCTGCCAGAAACCACGTCATG-3′. The amplification program for the TaLcye-3A, TaLcye-3B, TaLcye-3D, and bar gene started with an initial denaturation at 95 °C for 10 min, followed by 30 cycles at 94 °C for 30 s, 64 °C for 30 s, and 72 °C for 30 s, followed by a final extension step at 72 °C for 5 min. The PCR products were digested with the restriction enzyme EcoR I (1 U) in a 10 μL reaction buffer containing 3 μL PCR product for 2 h at 37 °C. The resultant products were separated on a 2% agarose gel and visualized using a GelDoc XR System (BioRad, Hercules, CA, USA). The presence of edited mutations was inferred from the altered banding patterns compared to the wild-type controls. To distinguish different mutant types, the PCR products were sequenced and aligned to the wild-type sequences.

2.4. Quantitative Reverse-Transcriptase PCR (qRT-PCR)

Samples were collected from the leaves and immature grains at 15 DPA of the TaLcye-edited wheat plants and their wild-type Fielder, which were grown in a semi-controlled greenhouse environment. Total RNA was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s protocol. Reverse transcription was performed using HiScript III RT SuperMix (Vazyme, Nanjing, China) according to the manufacturer’s instructions. The qRT-PCR was performed using SevenFast SYBR Green Mix (Seven, Beijing, China) in a 7500 Fast Real-Time PCR system (Applied Biosystems, Foster City, CA, USA). Each reaction was carried out in a final volume of 20 μL, with 10 μL SevenFast SYBR Green Mix, 0.4 μM of each primer, and 1 μL cDNA. The thermal cycling program at 95 °C for 5 min, followed by 40 cycles of amplification (95 °C for 5 s, 60 °C for 15 s, 72 °C for 15 s). The primers for the key genes involved in the carotenoid biosynthetic pathway are listed in Table S1. Transcript abundance was expressed relative to that of TaActin using the 2^−ΔΔCT method, which allows for the quantification of gene expression levels relative to a reference gene; all samples were analyzed with three biological replicates (independently grown plant batches) and three technical replicates (three injections per extraction) [30].

2.5. Ultra-Performance Liquid Chromatography-Tandem Mass Spectrometry (UPLC-MS/MS)

Leaves and grains at 20 DPA were collected from the TaLcye-edited plants and wild-type, then 50 mg of each sample were weighed, and grounded into a fine powder under liquid nitrogen. Subsequently, added 5 mL solvent mixture composed of n-hexane, acetone, and ethanol (1:1:1, v/v/v), as well as 0.01% butylated hydroxytoluene (BHT, w/v) into the powdered samples. Following the extraction process, the samples were centrifuged to eliminate insoluble debris. The resulting supernatant was filtered through a syringe filter to yield a clear extract. This extract was subsequently injected into an ultra-performance liquid chromatography (UPLC) system (ExionLC™ AD, AB SCIEX, Framingham, MA, USA). Carotenoids were separated on a C30 column (YMC C30, 2.1 mm × 100 mm, 3 μm, BioRad, Hercules, CA, USA) at 28 °C using a gradient elution program. The mobile phase consisted of acetonitrile (1:3, v/v) with 0.01% BHT and 0.1% formic acid (A) and methyl tert-butyl ether with 0.01% BHT (B). The gradient program was: 0–3 min, 0% B; 3–5 min, 70% B; 5–9 min, 95% B; 10–11 min, 0% B; flow rate, 0.8 mL min⁻¹; injection volume, 2 mL [31,32,33].

The separated compounds were then detected via tandem mass spectrometry (QTRAP 6500+, AB SCIEX, Framingham, MA, USA). Linear ion trap (LIT) and triple quadrupole (QQQ) scans were acquired on a triple quadrupole-linear ion trap mass spectrometer (QTRAP), QTRAP^® 6500+ LC-MS/MS System, equipped with an APCI Heated Nebulizer, operating in positive ion mode and controlled by Analyst 1.6.3 software (Sciex). The APCI source operation parameters were as follows: ion source, APCI+; source temperature 350 °C; curtain gas (CUR) were set at 25.0 psi. Carotenoids were analyzed using scheduled multiple reaction monitoring (MRM) [31]. A series of carotenoid standard solutions were prepared at concentrations of 0.001, 0.005, 0.01, 0.05, 0.1, 0.5, 1, 5, 10, and 25 μg ml⁻¹. Chromatographic peak area data corresponding to the quantitative signals were recorded for each concentration. Standard curves were generated by plotting standard concentration (x-axis) versus peak area (y-axis) for each analyte, enabling subsequent quantitative calibration. The calibration curves showed good linearity with correlation coefficients (R²) > 0.995 for all analytes.

The Metware Database (MWDB) was established using reference standards, enabling qualitative annotation of the acquired LC-MS/MS datasets. The retention time, parent ion, and product ion of main metabolite are listed in Table S2. Then, the data were processed using MultiQuant 3.0.3 (Sciex) software. The chromatographic peaks corresponding to the analytes in different samples were integrated and corrected based on the retention times and peak shape information of the reference standards. Ultimately, the concentrations of the target compounds were determined using calibration curves derived from standard solutions. All samples were analyzed with three biological replicates (independently grown plant batches) and three technical replicates (three injections per extraction).

To facilitate the visualization of metabolite-level trends, each metabolite was mean-centered and scaled to unit variance (UV scaling, synonymous with Z-score or auto-scaling). Because this row-wise normalization is applied solely for heat-map rendering, it leaves all downstream statistical results unchanged. The calculation method is as follows: after centering the original data, each variable was divided by its standard deviation; heat maps were generated with the R software (R 4.0.x).

2.6. Grain Protein, Wet Gluten, and Starch Content Analysis

Grain protein and wet gluten content in the TaLcye-edited lines and wild-type were measured using the DA7200 near-infrared instrument (Perten, Hägersten, Sweden). Starch and amylose in the whole-grain flour was extracted and analyzed using the Solarbio Starch Content Assay Kit and the Amylose Content Assay Kit (Solarbio, Beijing, China). In each analysis, three biological replicates and three technical replicates were prepared for both each TaLcye-edited line and the wild-type.

2.7. Grain Soluble Sugar Analysis

About 0.1 g of fine sample powder from each TaLcye-edited line and the wild-type was weighed into a 15 mL centrifuge tube, and then 4 mL of 80% (v/v) ethanol was added. The centrifuge tube was then placed in a water bath at 80 °C for 40 min. After cooling, the sample was centrifuged at 4500 rpm for 10 min. The supernatant was collected and filtered through a 0.45 μm membrane disk filter (Sangon, Beijing, China).

Next, 1 mL of the filtered supernatant was taken, mixed with 5 mL of anthrone reagent, and then placed in a boiling water bath for 10 min. After cooling to room temperature, the absorbance (OD) value at 625 nm was measured using a spectrophotometer, and the soluble sugar content was calculated based on the standard curve prepared with known concentrations of soluble sugar standards. Each sample was subjected to three biological replicates and three technical replicates.

2.8. Scanning Electron Microscopy Analysis of Wheat Grains

Mature wheat grains from the TaLcye-edited lines and wild-type were fixed in 2.5% glutaraldehyde solution for 2–4 h. Then the grains underwent dehydration in a graded ethanol series (30%, 50%, 70%, 90%, and 100% ethanol), maintaining 15–30 min in each concentration. Subsequently, critical-point drying was performed using carbon dioxide as the displacement medium to remove ethanol while preserving sample integrity. The dried grains were sectioned to expose internal structures. The sections were then mounted on a sample holder and sputter-coated with a thin gold layer using an ion sputtering device to enhance conductivity. Finally, the samples were placed in the scanning electron microscopy (SEM) chamber for observing the surface and internal structures of the wheat grains.

2.9. Examination of the Main Agronomic Traits and Statistical Analysis

At wheat maturity stage, ten plants were randomly picked from each replicate plot for evaluating main agronomic traits including tiller number per plant, plant height, spike length, spikelet number per spike, grains per spike, grain length, grain width, and 1000-kernel weight (TKW).

Data analysis of the carotenoid content, grain protein content, wet gluten content, starch content, soluble sugar content, and main agronomic traits were performed using SPSS (v28.0.1.1); statistical significance was determined by one-way ANOVA followed by Tukey’s honestly significant difference (HSD) post hoc test. Different letters indicate statistically significant differences (p < 0.05).

3. Results

3.1. Identification of Wheat TaLcye Loss-of-Function Mutants

To assess the roles of TaLcye homoeologs in carotenoid metabolism and plant growth in wheat, we edited TaLcye homoeologs in a wheat cultivar Fielder and obtained candidate mutant plants. In the offsprings of TaLcye-edited plants, three transgenic-free homozygous mutant lines (1–4, 2–8, and 5–1) were selected after detailed molecular analysis (Figure S1C). Sanger sequencing revealed that line 5–1 exhibited a 2 bp, 4 bp, and 11 bp deletions in exon 2 of the TaLcye-A, TaLcye-B and TaLcye-D homoeologs, in order. Line 2–8 had a 1 bp and 2 bp deletions at the 3 bp upstream of the PAM sites of TaLcye-B and TaLcye-D homoeologs, respectively. Line 1–4 contained a 2 bp deletion at position 469–470 of the TaLcye-A ORF, and a single T insertion at position 476 of the TaLcye-B ORF, and a 15 bp deletion in the TaLcye-D homoeologs (Figure S2).

3.2. Expression Profiling of the Key Genes Involved in the Carotenoid Biosynthetic Pathway in the TaLcye Mutants

To investigate the effects of TaLcye-edited homeoalleles on carotenoid biosynthesis in wheat, we assessed the expression levels of other key genes including TaPsy, TaPds, TaZds, TaLcyb, TaHyd1, and TaHyd2 involved in carotenoid biosynthesis pathway in TaLcye mutant lines using qRT-PCR and compared them with those in the wild-type.

In leaves, the relative expression levels of TaPsy, TaPds, and TaLcyb showed no significant differences among the mutant lines 1–4, 2–8, and 5–1. In contrast, the transcript levels of TaHyd2 were downregulated, while TaZds and TaHyd1 were upregulated in mutant line 5–1 (Figure S3).

In grains, none of the six genes (TaPsy, TaPds, TaZds, TaLcyb, TaHyd1, TaHyd2) differed in expression between wild-type and mutant 2–8; only the triple homozygous mutant line 5–1 showed significant repression expression of TaHyd2 and induction of TaPds and TaZds. And the expression levels of TaLcyb and TaHyd1 were upregulated in the mutant lines 1–4 and 5–1 (Figure 2).

3.3. Carotenoid Profiles in the Leaves and Grains of the Wheat TaLcye Mutants

Vitamin A is an essential nutrient for humans. Provitamin A carotenoids, such as β-carotene, can be converted into vitamin A in the human body, with β-carotene being the most effective [34]. Lutein, another type of carotenoid, cannot be converted into vitamin A [35]. To elucidate the role of TaLcye-edited in carotenoid biosynthesis in the leaves and grains of the mutant lines, we analyzed and compared the carotenoid content in the mutant lines and their wild-type using LC-MS/MS (Figure 3, Figures S4 and S5). A total of nineteen metabolites were measured in the leaves (Table S3). Compared with wild-type, the mutant lines exhibited a significant decrease in total carotenoid content. And the mutant lines 1–4 and 5–1 experienced a substantial reduction in lutein content (Figure S6F; Tables S3 and S5), which is a major compound in the downstream branch of the TaLcye gene. While in another branch, antheraxanthin, violaxanthin, and zeaxanthin levels were significantly elevated relative to the wild-type (Figure S6C,E; Tables S3 and S5). Additionally, mutant lines 1–4 and 5–1 also showed a slight increase in the content of γ-carotene, β-citraurin, phytoene, and β-cryptoxanthin (Table S3).

The carotenoid profiles of fourteen metabolites in total were measured in the grains at 20 DPA of the mutant lines, and results revealed lower average values than those observed in their leaves (Figure 4; Table S4). The content of lutein exhibited a significant decrease in the mutant lines 1–4 and 5–1 compared with the wild-type (Figure 4E). While the β-carotene increased to 33.2–34.5%, zeaxanthin increased to 112.1–125.4%, violaxanthin increased to 73.5–73.8%, and antheraxanthin increased to 164.1–186.5% in the mutant lines 1–4 and 5–1 (Figure 4A–D; Table S5). Notably, total carotenoid content did not increase uniformly across the mutant lines; line 5–1 showed a 6.0% decrease compared to wild-type (Tables S4 and S5), despite the specific enhancement in provitamin A carotenoids. However, the mutant line 2–8 showed no significant difference in lutein content in comparison with wild-type (Figure 4E), and β-carotene content in the mutant line was increased by 26.1%, so the total carotenoid content was enhanced to 59.71 μg g⁻¹ (Figure 4A; Tables S4 and S5).

Retinol activity equivalents (RAE) constitute the international standard for quantifying vitamin A bio-efficacy, enabling uniform assessment of vitamin A derived from diverse dietary precursors. Per this convention, 12 µg of dietary β-carotene correspond to 1 µg RAE, while 24 µg of dietary α-carotene or β-cryptoxanthin each correspond to 1 µg RAE. With the World Health Organization recommending a daily intake of 500–600 µg RAE for adults, enhancing β-carotene concentrations in staple crops has become an increasingly important breeding objective. Using the RAE conversion formula [β-carotene/12 + (α-carotene + β-cryptoxanthin)/24], the calculated retinol activity equivalents were 0.2035, 0.2730, 0.2557, and 0.2712 μg g⁻¹ for wild-type, mutant lines 1–4, 2–8, and 5–1 seeds, respectively. This represents a 25.7–34.2% increase in vitamin A potential for the mutant lines compared to wild-type (Table S4).

3.4. Comparison of Several Main Agronomic Traits Between TaLcye Mutants and Their Wild Type

To elucidate the impact of TaLcye mutations on the physiology and productivity of wheat plants, we evaluated some growth and yield related traits of mutant lines and their wild-type, which were cultivated in a semi-controlled greenhouse setting. Our findings revealed that, despite reductions in plant height, spike length, spikelet number per spike, and grains per spike in mutant lines 1–4, 2–8, and 5–1 relative to the wild-type, these mutants exhibited significant increases in grain length, and the thousand-kernel weight had no significant difference between wild-type and mutant lines (Figure 5A–D;

Table 1. Main agronomic traits of wild-type and TaLcye-edited plants.

Line	Tiller Number	Plant Height (cm)	Spike Length (cm)	Spikelet Number	Grains per Spike	Grain Length (mm)	Grain Width (mm)	TKW (g)
Fielder	15.6 ± 1.06 a	102.9 ± 1.02 a	10.9 ± 0.22 a	18.1 ± 0.22 a	56.0 ± 0.51 a	6.75 ± 0.01 b	3.52 ± 0.01 a	45.48 ± 0.25 a
1–4	15.7 ± 0.93 a	81.5 ± 0.45 b	9.2 ± 0.13 b	16.5 ± 0.43 b	51.9 ± 1.05 b	7.26 ± 0.07 a	3.29 ± 0.04 b	43.51 ± 0.57 a
2–8	16.0 ± 0.53 a	84.2 ± 0.20 b	9.6 ± 0.22 b	16.8 ± 0.45 b	52.4 ± 1.00 b	7.16 ± 0.01 a	3.35 ± 0.03 b	43.52 ± 0.69 a
5–1	13.9 ± 0.27 a	81.7 ± 0.28 b	9.4 ± 0.21 b	16.3 ± 0.41 b	51.3 ± 1.00 b	7.27 ± 0.07 a	3.34 ± 0.02 b	43.36 ± 1.17 a

Comparison between the lines 1–4, 2–8, 5–1, and Fielder. Statistical significance was determined by one-way ANOVA followed by Tukey’s honestly significant difference (HSD) post hoc test; values are presented as mean ± SD; significantly different (p < 0.05) values within the same column are indicated with different superscript letters.

). Furthermore, the seed coat and endosperm of the triple homozygous mutant lines 1–4 and 5–1 displayed an orange-yellow hue (Figure 5E,F).

3.5. Comprehensive Analysis of Grain Quality in Wheat TaLcye Mutants

To systematically evaluate the grain quality of TaLcye-edited plants, we conducted a comprehensive analysis of several key parameters, including protein and wet gluten content, grain hardness, starch and amylose content, as well as soluble sugar content. The results revealed that the mutant lines (1–4, 2–8, and 5–1) exhibited no significant differences in protein, wet gluten content, and grain hardness compared to their wild-type counterparts. However, significant improvements were observed in the contents of soluble sugar and amylose in lines 1–4 and 5–1. Conversely, the total starch content was significantly reduced in these mutant lines (

Table 2. An analysis of the major quality traits of wild-type and TaLcye-edited grains.

Line	Starch (mg/g)	Amylose (mg/g)	Soluble Sugar (mg/g)	Protein (%)	Wet Gluten (%)	Hardness (%)
WT	549.312 ± 2.453 a	138.736 ± 2.253 c	50.242 ± 0.222 b	11.27 ± 0.05 a	25.30 ± 0.08 a	60.70 ± 0.18 a
1–4	439.223 ± 1.727 d	180.690 ± 0.624 a	66.114 ± 0.726 a	11.50 ± 0.09 a	25.44 ± 0.07 a	60.24 ± 0.07 a
2–8	517.222 ± 2.025 b	148.452 ± 0.250 b	62.004 ± 0.796 a	11.63 ± 0.08 a	25.48 ± 0.09 a	60.01 ± 0.16 a
5–1	474.519 ± 1.556 c	178.137 ± 1.556 a	65.610 ± 0.869 a	11.61 ± 0.10 a	25.44 ± 0.13 a	60.05 ± 0.18 a

Comparison between the lines 1–4, 2–8, 5–1, and Fielder. Statistical significance was determined by one-way ANOVA followed by Tukey’s honestly significant difference (HSD) post hoc test; values are presented as mean ± SD; significantly different (p < 0.05) values within the same column are indicated with different superscript letters.

). Additionally, scanning electron microscopy (SEM) analysis indicated an increase in A-type starch granules in the mutant lines, which may be attributed to the genetic modifications introduced by the TaLcye editing (Figure 6).

4. Discussion

Given the crucial role of β-carotene in human nutrition, significant efforts have been devoted to enhance its levels in staple crops. For instance, researchers successfully biofortified rice with improved β-carotene by reconstructing its entire biosynthetic pathway in the endosperm, called as “Golden Rice”. This was achieved through the introduction of several key genes from the carotenoid biosynthetic pathway, thereby significantly increasing the β-carotene content in rice grains [36,37]. In another study, the durum wheat cultivar Kronos was subjected to ethyl methanesulfonate (EMS) mutagenesis to create a TaLcye-silenced mutant, in which the metabolic pathway responsible for the synthesis of α-carotene and lutein was effectively blocked, while the β-carotene content in grains was increased by 75% compared to the wild-type control [22]. However, the EMS mutagenesis introduces mutations at random positions across the genome, increasing the risk of off-target genetic alterations. While the CRISPR-Cas9 gene editing system offers higher precision than EMS mutagenesis, directing modifications exclusively to predetermined genomic loci. For example, the CRISPR–Cas9-mediated knockout of Lcye in lettuce increased β-carotene abundance 2.7-fold [38]. It is reported that in the USA, Japan, Australia and several Latin-American countries, some transgene-free edited plants have been exempted from GM regulations and can be cultivated and sold without additional labeling [39]. Wheat, a staple food source for a significant portion of the global population, is characterized by its low β-carotene content in the grains, which limits the bioavailability of essential retinol (vitamin A) [23].

In the present study, we generated homozygous double and triple mutants of the TaLcye gene in wheat via the CRISPR-Cas9 genome editing system. And experimental results showed that the expression levels of TaPds and TaLcyb exhibited tissue-specific divergence between leaves and grains, which is likely attributable to the functional specialization of TaLcye in distinct tissue types. Analysis using UPLC-MS/MS revealed that although the β-carotene content in the mutant leaves was reduced, it was increased in all the grains at 20 DPA of mutant lines (Figure 3; Tables S3–S5). Specifically, line 5–1 exhibited a 34.5% increase in β-carotene content compared to the wild-type (Figure 4A; Tables S4 and S5). Providing evidence that simultaneous knockout of TaLcye on chromosomes 3A, 3B, and 3D is superior to double mutants in boosting grain β-carotene content. Additionally, line 2–8 demonstrated a slight increase in total carotenoid content in the grains at 20 DPA (Tables S4 and S5). Zeaxanthin, as a downstream metabolite of β-carotene, is highly concentrated in the macula of the eye’s retina, where it functions as a shield against blue light and oxidative stress, thereby reducing the risk of age-related macular degeneration (AMD) [40]. Accumulating evidence had established that among macular carotenoids, zeaxanthin—rather than lutein—serves as the principal protective agent against AMD pathogenesis [41,42,43]. In addition, antheraxanthin and violaxanthin was the metabolite of zeaxanthin, have been shown to possess specific anti-cancer properties and may contribute to slow the progression of atherosclerosis [44]. In plants, antheraxanthin and violaxanthin participate in the xanthophyll cycle and protect the photosynthetic system from light-induced damage under high light conditions [45,46]. Through knocking out the TaLcye, our present results by UPLC-MS/MS analysis revealed significant increases in the levels of these carotenoids in the leaves of line 5–1, with zeaxanthin increasing 2.6-fold, antheraxanthin increasing 2.8-fold, and violaxanthin increasing 1.5-fold compared to the wild-type (Figure S6; Tables S3 and S5). In the grains of line 1–4, the levels of zeaxanthin, antheraxanthin, and violaxanthin were elevated to 2.1 µg g⁻¹, 13.6 µg g⁻¹, and 29.9 µg g⁻¹, in order (Figure 4; Tables S4 and S5). The near-abolition of lutein in triple-mutant lines 1–4 and 5–1 reflects the complete loss of LCYE catalytic function (Figure S2), which efficiently diverted metabolic flux toward the β-carotene biosynthetic pathway while concomitantly disrupting ε-cyclization activity essential for lutein production. This metabolic trade-off eliminates a xanthophyll with established retinoprotective properties against age-related macular degeneration (AMD), highlighting the nutritional compromise inherent in maximizing provitamin A content. Future breeding strategies may address this nutritional compromise through introgression with high-lutein germplasm or multiplex CRISPR approaches that balance flux between the α- and β-carotenoid branches.

While the mutant lines exhibited 25.7–34.2% higher RAE values (0.2557–0.273 vs. 0.2035 μg g⁻¹) compared to wild-type, their realistic contribution to daily vitamin A requirements remains modest. Accounting for approximately 70% milling recovery and the typically low bioavailability of cereal carotenoids (5–15% after cooking), 100 g of whole grain flour from the highest-performing line 1–4 and 5–1 would provide only ~1–3 μg bioavailable RAE. This represents merely 0.2–0.6% of the WHO recommended daily intake for adults (500–600 μg RAE), underscoring the gap between biochemical potential and nutritional impact. Future work must quantify retention rates across the value chain and explore mitigation strategies, such as antioxidant encapsulation or optimized milling protocols, to ensure that enhanced grain content translates to dietary vitamin A delivery.

Lutein is a yellow carotenoid that typically imparts a pale-yellow hue to cereal grains. In the endosperm of mature tetraploid wheat grains, lutein is the predominant carotenoid component, contributing the seeds with a significantly yellow appearance [47]. In contrast, carotenes, particularly β-carotene, generally impart an orange color to grains. For instance, the accumulation of β-carotene in grains results in an orange or orange-yellow phenotype [48]. Here, the seed coat and endosperm of the wheat triple-homozygous mutant lines 1–4 and 5–1 displayed an orange-yellow hue, whereas line 2–8 (double mutant) remained indistinguishable from the wild-type, likely due to the different content of β-carotene and lutein in the seeds (Figure 5E,F).

Previous research found that the mutant of TaLcye in tetraploid wheat did not show differences in growth relative to control under the greenhouse growth condition [49]. In common wheat, homozygous TaLcye mutants exhibited significant reductions in plant height, spike length, spikelet number, and grains per spike. Conversely, grain length was increased, whereas thousand-kernel weight remained statistically unaltered (Figure 5A–D; Table 1). Such phenotypic alterations are characteristic of first-generation gene-edited lines and do not preclude breeding utility, provided subsequent backcrossing with elite high-yielding varieties is performed to recover agronomic fitness while retaining the enhanced provitamin A trait. Additionally, beyond nutritional enhancement, the 5.8–20.0% reduction in total starch coupled with elevated amylose content (Table 2) carries significant food science implications. Such altered amylopectin/amylose ratios typically increase dough elasticity and may reduce loaf volume—critical considerations for consumer acceptance in traditional bread/steamed bread preparations.

The altered carotenoid profile (increased β-carotene, decreased lutein) and modified starch architecture raise specific concerns for wheat functionality. The hydrophobic nature of enhanced carotenoids may interfere with gluten network hydration, potentially affecting dough extensibility in noodle production. Conversely, the elevated amylose content could improve resistant starch formation, offering glycemic control benefits but potentially compromising softness in fermented products. Comprehensive evaluation of processing functionality, including mixograph parameters and starch digestibility profiles, remains essential to establish whether the nutritional gains justify potential trade-offs in consumer sensory acceptance. Additionally, reconciling provitamin A enhancement with agronomic performance and balanced nutrition requires moving beyond single-gene knockout approaches. Tissue-specific promoters (endosperm-specific TaLcye silencing) could minimize pleiotropic effects on photosynthetic tissues where lutein serves photoprotection. Alternatively, weak alleles or RNAi constructs offering partial rather than complete TaLcye suppression might preserve trace lutein levels (for AMD protection) while still enhancing β-carotene. Marker-assisted stacking of PSY overexpression with TaLcye downregulation could further optimize flux without compromising yield. Finally, multi-trait selection indices incorporating grain yield, starch quality, and carotenoid balance will be essential to develop varieties acceptable to both farmers and consumers.

5. Conclusions

This study establishes proof-of-concept that CRISPR-mediated knockout of TaLcye homoeologs can redirect carotenoid flux toward provitamin A compounds in wheat endosperm, demonstrating the feasibility of metabolic engineering for biofortification in a staple cereal. While these findings demonstrate the metabolic potential of TaLcye editing, translation to agricultural practice requires addressing three critical gaps: (i) mitigating the lutein trade-off through allele mining or multiplexed editing of carotenoid pathway genes to restore balanced carotenoid profiles; (ii) agronomic optimization via marker-assisted backcrossing to introgression mutant alleles into elite cultivars without yield penalty; and (iii) rigorous assessment of carotenoid retention during milling, storage, and thermal processing to ensure nutritional efficacy in consumed food products.