Induced Mutagenesis in Safflower (Carthamus tinctorius L.) Uncovers High-Oleic Acid Mutants Genetically Distinct from the Canonical CtFAD2-1 Allele

Abstract

The high-oleic acid content of the safflower (Carthamus tinctorius L.) oil, regulated by the fatty acid desaturase 2-1 (CtFAD2-1) gene, provides superior oxidative stability for applications. To explore alternative genetic sources for this trait, we employed induced mutagenesis with gamma irradiation and ethyl methane sulfonate (EMS) for two safflower cultivars, AKS 207 and PKV Pink. Screening of M₂ populations identified several mutants with significantly higher oleic acid content, reaching up to 36.86%. The mutagenized populations also exhibited a wide spectrum of variation for other agronomically important traits, including increased oil content (up to 35.19%), enhanced seed protein (up to 22.51%), and seed size and weight. Correlation and principal component analyses confirmed the antagonistic relationship between oleic and polyunsaturated fatty acids and the positive association among seed size parameters. Molecular profiling using an allele-specific PCR assay targeting the CtFAD2-1 locus revealed that high-oleic mutants did not carry known mutations, suggesting the involvement of alternative alleles, micro-mutations, or other genes regulating oleic acid accumulation. This study provides valuable pre-breeding germplasm with improved agronomic and quality traits and identifies novel genetic sources for high-oleic acid in safflower. These mutants form a new genetic basis for understanding fatty acid biosynthesis and developing next-generation high-stability oil cultivars.

1. Introduction

The global agricultural sector faces a critical challenge in the oilseed sector. While major producers like India are among the top globally, many developing nations remain significantly reliant on imports to satisfy domestic needs for edible oils, with imports constituting around 57–60% of total consumption [1]. This widening gap between domestic production and consumption creates a significant economic burden, costing national treasuries billions annually, and renders the market susceptible to global price swings, thereby undermining food security objectives [2]. To become self-sufficient in producing edible oil, it is crucial to improve the yield and profitability of oilseed crops grown in local conditions, particularly in the large rainfed farming areas around the world that make up more than 70% of the land used for oilseeds [3].

The fatty-acid composition and the balance between saturated and unsaturated fatty acids are critical parameters that determine both the nutritional and functional value of edible oils [4]. Vegetable oils provide a concentrated source of energy, serve as carriers of fat-soluble vitamins, and supply essential fatty acids such as linoleic (18:2, omega-6) and linolenic (18:3, omega-3), which are vital for growth and development [5]. Dietary intake of these polyunsaturated fatty acids (PUFAs) has been shown to increase high-density lipoprotein (HDL) cholesterol, while oleic acid (18:1) intake lowers LDL cholesterol without adversely affecting HDL cholesterol [6,7]. Thus, improving the fatty-acid profiles of oilseed crops is important for human health and nutrition.

Safflower (Carthamus tinctorius L.) is an important oilseed crop with significant potential that produces oil rich in unsaturated fatty acids. Its oil is characterized predominantly by linoleic acid (18:2) and oleic acid (18:1), which together constitute nearly 90% of the total fatty acids [8,9]. Conventional safflower cultivars globally typically contain high levels of linoleic acid (71–75%) and low levels of oleic acid (14–20%) [10], though ranges vary in specific germplasm [11]. Despite its agronomic resilience to drought and salinity, safflower cultivation has witnessed a drastic decline in the traditional growing belts. The primary reason for this decline is its poor economic competitiveness compared to alternative rabi crops, which offer farmers higher and more stable returns.

The economic non-viability of safflower is rooted in two interconnected genetic limitations. First, cultivars are often characterized by a low seed oil content, typically ranging from 28 to 30%, which is substantially lower than that of other commercial oilseeds. Second, the high concentration of linoleic acid (71–75%), while nutritionally beneficial, renders the oil oxidatively unstable and unsuitable for high-temperature frying [12,13,14]. By contrast, safflower lines with higher levels of monounsaturated oleic acid content (>75%) produce oil with superior stability, extended shelf life, and suitability for industrial applications like bio-lubricants and cosmetics [15,16,17]. While high-oleic varieties exist globally, including India’s ISF-1 released in 2018, these varieties often do not solve the problem of low oil content [18].

The genetic foundation for elevated oleic acid levels in widely used safflower germplasms has been identified as allelic variations within the CtFAD2-1 gene (‘ol’ allele), which encodes the principal microsomal oleate desaturase (Δ12-desaturase) enzyme [19,20]. However, the potential for inducing favorable mutations in regulatory elements or transcription factors remains a largely unexplored but promising avenue for crop improvement. Therefore, a key objective of this study was not only to induce high-oleic phenotypes but also to investigate whether such phenotypes could arise from genetic mechanisms independent of the canonical CtFAD2-1 mutation.

Considerable natural variation in safflower oil composition has already been documented, including 2–3% stearic acid, 16–20% oleic acid, 6–8% palmitic acid, and 71–75% linoleic acid [21,22]. Furthermore, environmental conditions strongly influence oil quality, with cooler environments favoring higher linoleic acid content compared to warmer climates [23]. Another useful nutritional indicator is the polyunsaturated-to-saturated fatty-acid ratio (P/S index). Oils with a P/S index greater than 1 are considered beneficial, with safflower oil typically exhibiting one of the highest values (10.5), compared to other major oilseeds [24,25]. Despite its nutritional advantages, safflower remains a less competitive oilseed crop in India, primarily due to its relatively low seed oil content (28–30%) [26].

Conventional breeding of these complex and polygenic traits is hindered by the narrow genetic base of safflower germplasm and the strong influence of genotype × environment interactions [27]. Therefore, induced mutagenesis presents a powerful strategy to create novel genetic variability that may not be available in the existing gene pool [9,28]. To effectively overcome the complicated variation generated in a mutagenized population, a robust analytical framework is required. Understanding the interrelationships among seed, oil, and fatty acid traits is essential for effective selection and trait improvement in safflower. Correlation analysis provides insight into the degree and direction of association between key traits, thereby enabling identification of indirect selection criteria for complex quality parameters such as oil stability and protein content [29], while principal component analysis (PCA) allows for the reduction in multidimensional trait data into a few principal components, thereby revealing patterns of variation and clustering of genotypes based on multiple traits [30].

Therefore, this work was conducted to generate genetic variability in safflower via chemical and physical mutagenesis. The primary objective was to isolate and characterize superior mutant lines exhibiting a beneficial combination of increased seed oil content and an enhanced fatty acid profile, with a specific focus on elevating oleic acid concentration. This is key research to provide a foundation for the development of high-yielding genotypes with the increased oleic acid content. This will help make safflower farming more sustainable in rainfed regions and support the goal of producing more edible oil through sustainable agriculture.

2. Materials and Methods

2.1. Plant Material and Mutagenic Treatment

The present investigation was conducted using seeds of two released safflower (Carthamus tinctorius L.) cultivars, AKS 207 and PKV Pink, which are high-yielding commercial cultivars but have normal oleic acid content (18.63, 19.03%, respectively), obtained from the Oilseeds Research Centre, Dr. Panjabrao Deshmukh Krishi Vidyapeeth, Akola, Maharashtra, India (20.69938° N, 77.02679° E). Both cultivars are widely cultivated under rainfed conditions in Maharashtra and represent important genetic resources for safflower improvement.

For mutagenic treatments, separate sets of 1000 dry and healthy seeds per treatment, with an average moisture content of 8–10%, were subjected to each individual gamma ray dose and EMS concentration. An additional set of 1000 untreated seeds served as the control for each cultivar. Gamma irradiation was carried out at the Bhabha Atomic Research Centre (BARC), Mumbai (19.02271° N, 72.92750° E), using a ⁶⁰Co gamma source at doses of 250, 300, and 350 Gy. For chemical mutagenesis, seeds were treated with 100 mL of a solution of ethyl methane sulfonate (EMS) at concentrations of 0.4%, 0.5%, and 0.6%. Before EMS treatment, seeds were presoaked in distilled water for 3 h to facilitate uniform imbibition, followed by immersion in EMS solution prepared in 0.1 M sodium phosphate buffer (pH 7.0) for 12 h in the dark with continuous shaking at room temperature to prevent stratification. This buffer system was used to minimize the rapid hydrolysis of EMS. After treatment, seeds were thoroughly washed under running tap water for 2 h to terminate the mutagenic effect. Control seeds were subjected to identical treatment with a buffer solution without EMS.

Following mutagen exposure, seeds were immediately sown in the experimental field at the Oilseeds Research Centre, Dr. Panjabrao Deshmukh Krishi Vidyapeeth, Akola, Maharashtra, India. The sowing was performed in an unreplicated trial using plots of 10 m × 9 m with an inter-row spacing of 45 cm and an intra-row spacing of 20 cm.

2.2. Estimation of Lethal Dose (LD₅₀)

The objective of this step was to identify the optimal mutagen doses capable of inducing a high frequency of mutations while minimizing excessive lethality and other detrimental biological damage in the M₁ generation. One hundred seeds of AKS 207 and PKV Pink varieties were subjected to the varying doses of 50, 100, 150, 200, 250, and 300 Gy for gamma rays and 0.1, 0.2, 0.3, 0.4, and 0.5% EMS in three replications to work out the LD₅₀ dose by probit analysis. The median lethal dose (LD₅₀) values, determined by probit analysis, were 288 Gy (AKS 207) and 325 Gy (PKV Pink) for gamma irradiation. For the chemical mutagen EMS, the LD₅₀ values were 0.46% for AKS 207 and 0.49% for PKV Pink. These LD₅₀ estimates provided a physiological basis for selecting the mutagen doses used in the present study. The use of multiple doses allowed assessment of dose-dependent mutagenic effects in both varieties.

2.3. Trait Evaluation and Analytical Procedures

The M₂ population was subjected to rigorous phenotypic screening to identify superior mutants based on the specific breeding objectives for each variety. Morphological observations were recorded on 20 randomly selected plants per treatment group. In the M₁ generation, all plants were harvested as a single plant selection across all the gamma and EMS treatments to raise the M₂ progeny, including the macro mutants. Selection strategies were tailored to the specific agronomic limitations of each cultivar. AKS 207, characterized by bold seeds (hundred-seed weight: ~6.3 g) but moderate oil content (~29.8%), was targeted for high oil potential. Conversely, PKV Pink, which possesses high oil content (~33.6%) but suffers from small seed size (hundred-seed weight: ~3.7 g), was subjected to selection prioritizing seed size enhancement to improve market acceptance. Only mutants exhibiting these specific phenotypes were advanced for detailed characterization. Seed length (mm) and width (mm) were measured on ten randomly selected seeds per mutant using digital Vernier calipers (0.01 mm precision). The length/width ratio was calculated from these values. The weight of 100 randomly counted, filled seeds was recorded in grams using a precision balance. Determined by filling a standard 100 mL measuring cylinder with seeds and recording the weight to assess bulk density. To minimize the confounding effects of environmental variability on these quantitative traits, all mutants were evaluated alongside their respective untreated controls under identical agronomic conditions. Following mutagenic treatment and field establishment of the M₂ generation, selected mutant lines were evaluated for key seed quality parameters, including oil content, fatty acid composition, desaturation indices, and molecular markers associated with variations in oleic acid. These assessments were designed to capture both phenotypic and genotypic changes induced by mutagenesis and to identify mutants with potential nutritional and industrial value.

2.3.1. Estimation of Oil Content (%)

Seed oil content was estimated using the modified method of Yadav and Murthy [31] using nuclear magnetic resonance spectroscopy (NMR, MQC-5 analyser (Oxford, London, Abingdon, UK) equipped with preloaded ‘EasyCal’ calibration software (v5.0). Before analysis, a minimum of 15 g of seeds per plant were dried in a hot-air oven at 103 °C for 2 h to reduce moisture content. The room temperature of the NMR facility was maintained at 25 ± 2 °C. Analyses were conducted using a 40 mm NMR probe at an operating frequency of 5 MHz, with sixteen scans performed at 40 °C.

2.3.2. Fatty Acid Composition and Profiling

In varieties AKS 207 and PKV Pink, a total of 55 and 32 plants, respectively, were selected in the M₂ generation, and 10 g of seeds per plant were used for the fatty acid composition analysis. The fatty-acid composition of selected M₂ mutants, along with their respective controls, was determined by gas chromatography (GC) at the ICAR–Indian Institute of Oilseeds Research (IIOR), Hyderabad, India (17.32168° N, 78.41427° E). Oil was extracted using a fully automated Soxhlet apparatus (Soxtherm, C. Gerhardt GmbH & Co. KG, Königswinter, Germany). 300 mg of oil was transesterified using 3 mL of 13% methanolic KOH for 70 min at 60 °C. Hexane was added to separate the organic phase. The upper phase was washed with water till it reached neutral pH. The mixture was dried over anhydrous sodium sulfate and concentrated with nitrogen to get fatty acid methyl esters [10]. The samples (0.2 μL) were injected in split mode (split: column ratio 30:1) into a gas chromatograph (Agilent 7890B, Santa Clara, California, USA). A capillary column (DB 225, 30 m × 0.320 mm × 0.25 μm) was used with nitrogen as the carrier gas at 1 mL/min, and detection was done with a flame ionisation detector (FID). The injector, detector, and oven were maintained at 210 °C, 220 °C, and 230 °C, respectively [32]. The fatty acid composition was determined by identifying and calculating relative peak areas using open lab software.

2.3.3. Oleic Desaturation Ratio (ODR)

The oleic desaturation ratio (ODR), an indirect measure of Δ12-desaturase activity responsible for the conversion of oleic acid to linoleic acid, was calculated according to Ntiamoah and Rowland [33]. Lower ODR values are indicative of reduced desaturase activity, which typically corresponds to higher oleic acid accumulation in the seed oil.

2.3.4. Polyunsaturated/Saturated Fatty Acid Index (P/S Index)

We calculated the P/S index, which is the ratio of polyunsaturated fatty acids (PUFA) to saturated fatty acids (SFA), to evaluate the nutritional quality of safflower oil. Oils with P/S values greater than 1 are generally considered desirable from a dietary perspective, and safflower typically demonstrates one of the highest values among oilseed crops.

2.4. Estimation of Mutagenic Effectiveness and Efficiency

Mutagenic effectiveness and efficiency were estimated following the procedure of Konzak et al. [34]. The frequencies of both chlorophyll and viable mutants were estimated using the formulae recommended by Gaul (1958) [35]. Mutation frequency was calculated as a percentage of the total population:

$$Mutation frequency (\%)={\displaystyle \frac{Number of visible mutants\times 100}{Total number of plants in a dose}}$$

Screening for mutations was performed in the M₂ generation. The determinants of mutation were defined as distinct phenotypic deviations from the untreated control. Mutagenic effectiveness was defined as the frequency of mutations induced per unit dose of mutagen. For chemical mutagens, effectiveness was calculated as:

$$Mutagenic Effectiveness ={\displaystyle \frac{Frequency of Mutation in M_2(\%)}{Concentration\times Duration of treatment}}$$

where:

Concentration = Concentration of the mutagen used (%).

Duration of treatment = constant at 12 h for all treatments.

For physical mutagens, effectiveness was calculated as:

$$Mutagenic Effectiveness={\displaystyle \frac{Frequency of Mutation in M_2 (\%)}{Dose \left(kR\right)}}$$

Mutagenic efficiency was calculated as the proportion of mutations in relation to the undesirable biological effects caused by the mutagen, expressed as:

$$Mutagenic Efficiency={\displaystyle \frac{Frequency of Mutation in M_2(\%)}{Percentage of Biological Damage}}$$

2.5. Statistical and Multivariate Analysis

To investigate the interrelationships among seed quality, oil content, and fatty acid composition traits, correlation analysis and PCA were carried out using RStudio software [36] (Version 2025.09). Trait-wise Pearson’s correlation coefficients were estimated with the ‘Hmisc’ package [37] (Version 5.1-0), and correlation matrices were visualized using the ‘corrplot’ package [38] (Version 0.95). PCA was performed to reduce dimensionality and identify major contributing traits to the observed variability. The PCA biplots were constructed using the ‘FactoMineR’ package [39] (Version 2.13), and visualizations were refined with the ‘factoextra’ package [40] (Version 1.0.7). Area-proportional Euler diagrams were constructed using the ‘eulerr’ R package [41] (Version 7.0.4) within the R statistical computing environment to visualize the unique mutant cardinality and pleiotropic relationships between traits in the mutant populations. To visualize the differential effectiveness of various mutagenic treatments on the resulting selection of unique mutant phenotypes, a Sankey plot was generated. The graph was generated using the ‘networkD3’ R package [42] (Version 0.4.1).

2.6. Genotyping and Molecular Profiling of Higher and Mid-Oleic Mutants

Genomic DNA was isolated from eleven mutant plants using the cetyltrimethylammonium bromide (CTAB) method [43]. DNA quality and integrity were assessed by agarose gel electrophoresis (0.8% agarose gel; 1× TAE). The M₂ mutants, along with reference checks including Montola-2000 (a high-oleic line) and A1 (a low-oleic line), were genotyped using an allele-specific PCR assay targeting the CtFAD2-1 gene [44,45]. The assay involved three allele-specific primers for CtFAD2-1, which are described below.

Forward 1: 5′GAAGGTGACCAAGTTCATGCTCAGGCGAAACGGTTGTAGGG3′

Forward 2: 5′GAAGGTCGGAGTCAACGGATTCAGGCGAAACGGTTGTAGGT3′

Common Reverse: 5′CAGCCTGTTCGCCACTCTCA3′

The allele-specific primers were designed following the protocol described by Jatayev et al. [46] for the Amplifluor assay, while the assay itself was performed using the Kompetitive Allele-Specific PCR (KASP) master mix. The KASP master mix was supplied by LGC Genomics, Hoddesdon, UK. The master mix contained FAM- and HEX-specific FRET (fluorescence resonance energy transfer) cassettes, a ROX passive reference dye, KASP Taq DNA polymerase (specially modified for allele-specific PCR), dNTPs, and MgCl₂ in an optimized buffer solution. The assay mix (20×) was prepared using 1 μL (100 pmol/μL) of allele-specific forward primer F1, 1 μL (100 pmol/μL) of allele-specific forward primer F2, 15 μL (100 pmol/μL) of common reverse primer R, and 183 μL of dH₂O. The PCR mix (5 μL) consisted of 2.5 μL KASP™ master mix, 0.25 μL assay mix, and 2.25 μL genomic DNA. The PCR programme was carried out using an Applied Biosystems Veriti™ 384-Well Thermal Cycler (Thermo Fisher Scientific, Waltham, MA, USA) with the following profile: 94 °C for 15 min (initial activation), followed by 10 touchdown cycles of 94 °C for 20 s and 61–55 °C for 60 s (decreasing by 0.6 °C per cycle), and finally 26 cycles at 94 °C for 20 s and 55 °C for 60 s [45]. Detailed information on the assay procedure is available in Kadirvel et al. [45]. Fluorescence signals were recorded using a plate reader, and allele clustering was visualized using ‘KlusterCaller’ software (Version 3.4.1.36; LGC Genomics, Hoddesdon, UK) and model (VICTORTMX3, PerkinElmer, Waltham, MA, USA). The raw fluorescence data (FAM/HEX values) for all genotyped samples is provided in Supplementary Table S4.

To investigate whether increased oleic acid levels were associated with the known mutation in the fatty acid desaturase gene CtFAD2-1, eight AKS 207 mutants and three PKV Pink mutants with higher oleic acid content were analyzed. SNP genotyping was employed to predict allelic status at the CtFAD2-1 locus, which is presumed to correspond to the ‘ol’ allele known to confer high oleic acid content. Single-marker genotyping was used to validate the association of the CtFAD2-1 locus with the high-oleic trait in these mutants, using Montola-2000 (high-oleic) and A1 (low-oleic) as reference controls.

3. Results

The morphological and quality parameters composition of the tested oils and fats for the individual M₂ mutants developed using variety AKS 207 are represented in Supplementary Table S1. Similarly, individual plant data for 32 mutants developed and selected from the variety PKV Pink were presented in Supplementary Table S2. The mutants in both varieties were selected based on morphological and quality traits, viz., seed yield, seed traits, number of branches, maturity, oil content, protein content, etc., as compared to the respective checks.

3.1. Estimates of Mutagenic Effectiveness and Efficiency

Mutagenic effectiveness and efficiency based on sterility are key parameters used to evaluate the potency of mutagens. The results of mutagenic effectiveness and efficiency for different doses of gamma rays and EMS in the M₂ generation are presented in

Table 1. Mutation frequency, mutagenic effectiveness, and efficiency of mutagens in inducing mutation in safflower varieties AKS 207 and PKV Pink in M₂ generation.

Variety/Dose	Plant Population	Frequency of Mutation	Pollen Sterility (%)	Mutagenic Effectiveness (%)	Mutagenic Efficiency (%) (MF/S)
AKS 207
250 Gy	1672	1.38(23)	11.63	0.06	0.08
300 Gy	1454	2.20(32)	16.47	0.07	0.09
350 Gy	1325	2.26(30)	24.32	0.06	0.07
0.4% EMS	1144	3.93(45)	16.73	0.82	0.20
0.5% EMS	790	4.18(33)	22.49	0.70	0.15
0.6% EMS	624	5.77(36)	27.46	0.80	0.18
Dry Control	252	NA	8.11	NA	NA
Wet Control	249	NA	5.69	NA	NA
PKV Pink
250 Gy	1437	0.56(8)	9.75	0.02	0.03
300 Gy	936	5.02(47)	14.69	0.17	0.29
350 Gy	1053	1.52(16)	22.63	0.04	0.06
0.4% EMS	895	1.01(9)	16.89	0.21	0.06
0.5% EMS	831	3.01(25)	18.69	0.50	0.12
0.6% EMS	607	1.65(10)	24.65	0.23	0.05
Dry Control	255	NA	7.41	NA	NA
Wet Control	261	NA	6.39	NA	NA

Note: Figures in parentheses are the total number of distinct phenotypic mutants observed, NA—Not applicable.

, Figure 1 and Figure 2. Across both safflower varieties, EMS treatments were more effective than gamma irradiation. The most effective dose of EMS in AKS-207 was 0.4% (0.82%), followed closely by 0.6% (0.80%). In PKV Pink, the most effective treatment was 0.5% EMS (0.50%), followed by 0.6% EMS (0.23%). Mutagenic efficiency exhibited unique varietal responses. The highest efficiency in AKS-207 was found to be 0.4% EMS (0.20%), and the second highest was at 0.6% EMS (0.18%). Conversely, in PKV Pink, the 300 Gy gamma-ray treatment (0.29%) was the most efficient, followed by 0.5% EMS (0.12%). The spectrum of viable mutations indicated that the mutagenic treatments with gamma rays and EMS induced mutations affecting chlorophyll, plant habit, leaf morphology, and seed and oil characteristics. A high spectrum of viable mutations was recorded at 0.4% EMS mutagenic treatment (45) followed by 0.6% EMS (36), 0.5% EMS (33) and 300 Gy dose (32) in the AKS 207 treated population. In the treated M₂ population of variety PKV Pink, the 300 Gy (47) dose induced the highest spectrum, followed by 0.5% EMS (25), 350 Gy (16), 0.6% EMS (10), 0.4% EMS (9), and 250 Gy (8) in the chemical mutagenic dose. These findings indicate that while EMS was generally more effective in inducing mutations, gamma irradiation demonstrated higher efficiency in certain conditions.

3.2. Seed Characteristics of Identified Mutants

Mutagenic treatments markedly influenced seed morphology, leading to the identification of several distinct mutant categories in both safflower varieties. The major seed-related variants were bold and long-seeded mutants, bold and short-seeded mutants, and high test-weight mutants (Figure 3 and Figure 4). These mutants displayed considerable variation relative to the controls, thereby broadening the available phenotypic diversity for potential use in genetic improvement programs.

In AKS 207, a total of eighteen bold and long-seeded mutants were identified, characterized by seed length >8.50 mm and seed width >4.11 mm. The mutants identified from AKS 207 were three each in 250 Gy (T1-5, T1-76, T1-79), 300 Gy (T2-3, T2-9, T2-18), and 350 Gy (T3-7, T3-8, T3-74); three in 0.4% EMS (T4-10, T4-27, T4-59); five in 0.5% EMS (T5-21, T5-22, T5-58, T5-63, T5-66); and one in 0.6% EMS (T6-57) (

Table 2. Seed oil content, fatty acid composition, and nutritional indices of selected M₂ safflower mutants exhibiting superior quality traits.

Mutant	Dose of Mutagen (Gy/%)	Oil %	Linoleic Acid	Oleic Acid	Stearic Acid	Palmitic Acid	Oleic Desaturation Ratio	P/S Index	Protein %
AKS 207
	Control	29.79	68.09	18.63	4.16	6.92	3.27	6.15	15.86
T1-5	250	27.36	67.91	21.21	3.52	7.31	3.2	6.27	20.23
T1-24	250	32.27	70.54	19.28	3.58	6.6	3.66	6.93	13.69
T1-79	250	25.27	62.33	24.15	4.72	8.82	2.58	4.6	14.77
T2-3	300	24.41	63.4	27.52	3.17	5.91	2.3	6.98	15.42
T2-69	300	29.28	60.67	28.8	2.65	7.88	2.11	5.76	12.81
T2-70	300	28.25	51.63	32.03	6.69	9.64	1.61	3.16	15.05
T2-72	300	28.58	74.74	18.07	1.35	5.84	4.14	10.39	11.48
T3-7	350	24.45	67.37	21.47	3.96	7.2	3.14	6.04	20.44
T3-8	350	24.27	66.25	22.56	3.94	7.24	2.94	5.93	20.16
T3-17	350	27.65	63.31	25.64	3.51	7.54	2.47	5.73	21.04
T4-14	0.4	31.36	71.23	18.08	3.04	7.66	3.94	6.66	14.84
T4-29	0.4	29.6	72.02	19.68	2.09	6.2	3.66	8.69	20.9
T4-30	0.4	28.4	69.49	20.65	2.69	7.17	3.37	7.05	22.12
T4-59	0.4	27.69	73.78	18.46	1.68	6.07	4	9.52	15.16
T4-62	0.4	31.07	58.18	29.21	2.86	9.75	1.99	4.61	12.08
T5-16	0.5	31.99	66.97	22.71	2.86	7.47	2.95	6.48	18.48
T5-23	0.5	29.94	66.8	23.39	3.22	6.6	2.86	6.8	21.49
T5-58	0.5	26.12	51.93	36.86	2.68	8.53	1.41	4.63	14.77
T5-63	0.5	26.76	66.95	20.57	4.24	8.23	3.25	5.37	14.35
T5-64	0.5	35.19	69.1	18.49	4.08	8.33	3.74	5.57	12.25
T5-65	0.5	25.69	63.1	27.37	2.09	7.44	2.31	6.62	15.16
T6-2	0.6	23.99	58.63	26.88	5.39	9.1	2.18	4.05	15.43
T6-13	0.6	32.16	65.42	23.2	4.1	7.28	2.82	5.75	17.36
T6-31	0.6	32.42	71.68	18.69	2.77	6.87	3.84	7.44	20.34
T6-55	0.6	24.79	73.51	19.18	1.01	6.3	3.83	10.06	15.47
PKV Pink
	Control	33.6	70.23	19.03	2.98	7.76	3.69	6.54	16.7
T1-85	250	32.01	78.13	13.84	1.14	6.89	5.65	9.73	17.61
T1-86	250	33.88	72.14	19.21	1.35	7.3	3.76	8.34	15.61
T2-41	300	34.71	70.75	18.46	2.74	8.05	3.83	6.56	15.26
T2-45	300	30.24	61.94	26.5	3.42	8.15	2.34	5.35	15.54
T2-87	300	34.13	69.46	20.42	1.72	8.4	3.4	6.86	13.79
T2-90	300	33.89	74.07	17.14	1.38	7.41	4.32	8.43	16.47
T2-91	300	34.64	74.18	17.25	1.68	6.89	4.3	8.66	12.11
T2-92	300	28.64	72.09	20.37	1.22	6.32	3.54	9.56	16.81
T2-93	300	34.16	72.66	18.42	1.95	6.97	3.94	8.15	14.81
T3-42	350	29.33	69.34	21.1	1.94	7.63	3.29	7.25	21.39
T3-43	350	31.18	60.22	28.96	2.19	8.58	2.08	5.59	17.89
T4-95	0.4	33.8	69.28	18.58	4	8.14	3.73	5.71	16.52
T5-51	0.5	26.79	69.59	20.42	2.25	7.74	3.41	6.97	22.51
T5-52	0.5	34.5	69.03	21.18	2.38	7.41	3.26	7.05	15.89
T5-99	0.5	32.74	61.94	27.35	2.58	8.13	2.26	5.78	20.16
T5-100	0.5	33.27	69.24	18.12	4.99	7.65	3.82	5.48	15.19
T5-103	0.5	33.76	74.26	15.65	3.19	6.91	4.75	7.35	15.33
T6-96	0.6	34.96	77.36	14.88	1.4	6.36	5.2	9.97	13.16
T6-97	0.6	33.66	73.31	15.81	3.75	7.12	4.64	6.74	15.96
T6-98	0.6	34.67	70.44	18.41	4.01	7.13	3.83	6.32	15.02

; Figure 3A,D,G; Supplementary Table S1). The best mutant was T5-22, with a seed length of 11.04 mm and a seed width of 5.74 mm. The difference for seed length was 3.35 mm, and 2.05 mm for seed width. In PKV Pink, seven bold and long-seeded mutants were observed, with two from 250 Gy (T1-85, T1-89), three from 350 Gy (T3-43, T3-47, T3-89), one from 0.4% EMS (T4-94), and one from 0.5% EMS (T5-51) (Table 2; Figure 4B,G; Supplementary Table S2). In this variety, bold and long-seeded mutants surpassed control values of 7.69 mm seed length and 3.69 mm seed width.

In addition, bold and short-seeded mutants were identified in both backgrounds. In AKS 207, ten such mutants (seed length <8.5 mm and seed width <4.5 mm) were recorded, comprising four at 300 Gy (T2-4, T2-69, T2-71, T2-72), one at 350 Gy (T3-75), four at 0.4% EMS (T4-28, T4-30, T4-61, T4-62), and one at 0.5% EMS (T5-15) (Figure 3H). In PKV Pink, four bold and short-seeded mutants were identified: one at 250 Gy (T1-84), two at 300 Gy (T2-49, T2-87), and one at 0.6% EMS (T6-98) (Figure 4I).

A particularly notable category comprised the high test-weight mutants, which represent large-seeded variants with greater seed density. In AKS 207, twenty-one such mutants were identified with test weight values exceeding 6.7 g, compared to 6.3 g of the control. These included one mutant from 250 Gy (T1-1), three from 300 Gy (T2-9, T2-20, T2-71), four from 350 Gy (T3-8, T3-73, T3-75, T3-81), five from 0.4% EMS (T4-10, T4-27, T4-35, T4-60, T4-61), seven from 0.5% EMS (T5-16, T5-21, T5-23, T5-58, T5-63, T5-66, T5-68), and one from 0.6% EMS (T6-55) (Figure 3C,I). In PKV Pink, fourteen high-test-weight mutants were found. Their weights were between 4.7 and 5.6 g, which is higher than the control value of 3.7 g. These were distributed as three mutants in 250 Gy (T1-84, T1-85, T1-86), five in 300 Gy (T2-121, T2-49, T2-122, T2-90, T2-91), three in 350 Gy (T3-42, T3-46, T3-47), one in 0.4% EMS (T4-94), and two in 0.5% EMS (T5-100, T5-103) (Figure 4C,G).

3.3. Variation in Oil Content and Fatty Acid Composition of Mutants

Mutagenic treatments induced significant quantitative variation in seed oil content and fatty acid profiles compared to the controls. We identified mutants exhibiting specific biochemical improvements, including increased oil content (>34%), elevated oleic acid (>25%), increased stearic acid, and reduced palmitic acid. The fatty acid profiles and oil content of these selected superior mutants are summarized in Table 2, while the complete dataset for the M₂ population is provided in Supplementary Tables S1 and S2.

3.3.1. High Oil Content Mutants

In the mutagenic population of AKS 207, seven mutants surpassed the high oil preliminary screening threshold of >30% (control: 29.79%). Most notably, mutant T5-64 recorded an oil content of 35.19%, representing a substantial relative increase of 18.1% over the parent variety (Table 2). Other promising mutants included T1-24 (32.27%) under 250 Gy irradiation, T4-14 (31.36%) and T4-62 (31.07%) under 0.4% EMS, T5-16 (31.99%) under 0.5% EMS, and T6-13 (32.16%) and T6-31 (32.42%) under 0.6% EMS. In PKV Pink, twelve high-oil mutants were identified exceeding the control level of 33.60%, with the highest increase observed in T6-96 (34.96%) and T2-41 (34.71%).

In PKV Pink, twelve high-oil mutants were identified, exceeding the control level of 33.60%. These included one mutant each under 250 Gy (T1-86, 33.88%) and 0.4% EMS (T4-95, 33.80%), three under 0.6% EMS (T6-96, 34.96%; T6-97, 33.66%; T6-98, 34.67%), two under 0.5% EMS (T5-52, 34.50%; T5-103, 33.76%), and five under 300 Gy (T2-41, 34.71%; T2-87, 34.13%; T2-90, 33.89%; T2-91, 34.64%; T2-93, 34.16%) (Table 2).

3.3.2. High Linoleic Acid Mutants

In AKS 207, nine mutants exhibited higher linoleic acid levels, surpassing the control value of 68.09%. The most prominent lines included T2-72 (74.74%), T4-59 (73.78%), T6-55 (73.51%), and T4-29 (72.02%). In PKV Pink, eleven high linoleic acid mutants were observed compared to the control (70.23%), with the highest values recorded in T1-85 (78.13% at 250 Gy), T6-96 (77.36% at 0.6% EMS), T5-103 (74.26% at 0.5% EMS), T2-91 (74.18% at 300 Gy), and T2-90 (74.07% at 300 Gy) (Table 2).

3.3.3. High Oleic Acid Mutants

In AKS 207, eight mutants were classified as high oleic (>25%) compared to the control (18.63%) (Table 2). Three mutants were isolated under 300 Gy (T2-3, 27.52%; T2-69, 28.80%; T2-70, 32.03%), one under 350 Gy (T3-17, 25.64%), one under 0.4% EMS (T4-62, 29.21%), two under 0.5% EMS (T5-58, 36.86%; T5-65, 27.37%), and one under 0.6% EMS (T6-2, 26.88%). Notably, mutant T2-70 not only accumulated higher oleic acid but also showed increased stearic (6.69%) and palmitic acid (9.64%) contents, along with a pronounced reduction in linoleic acid (51.63%) relative to the control (3.96%, 7.20%, and 67.37%, respectively). In PKV Pink, three high oleic mutants were detected: T2-45 (26.50%) at 300 Gy, T3-43 (28.96%) at 350 Gy, and T5-99 (27.35%) at 0.5% EMS, compared to 19.03% in the control.

3.3.4. High Stearic Acid Mutants

In AKS 207, four mutants exhibited higher stearic acid levels, surpassing the control value of 4.16%. These included T2-70, T6-2, T1-79, and T5-63. In PKV Pink, three high stearic acid mutants were observed compared to the control (2.98%): T5-100 (4.99% at 0.5% EMS), T6-98 (4.01% at 0.6% EMS), and T4-95 (4.00% at 0.4% EMS) (Table 2).

3.3.5. Low Palmitic Acid Mutants

Palmitic acid is the principal saturated fatty acid in safflower oil, and its reduction is a key objective for improving nutritional quality and reducing the risk of cardiovascular disease. Although bidirectional variation was observed in the mutant population, we prioritized the characterization of lines with reduced palmitic acid levels. In AKS 207, two low palmitic acid mutants were found at 300 Gy: T2-72 (5.84%) and T2-3 (5.91%). Both were lower than the control (6.92%). In PKV Pink, two low palmitic mutants were detected: T2-92 (6.32%) at 300 Gy and T6-96 (6.36%) at 0.6% EMS, compared to 7.76% in the control (Table 2).

3.3.6. Oleic Desaturation Ratio (ODR)

The ODR, which is an indirect way to measure Δ12-desaturase activity, was always lower in high oleic mutants. This confirms that there is an inverse relationship between oleic acid accumulation and desaturation efficiency. In AKS 207, eight mutants showed low ODR values (<2.5) compared to 3.92 in the control. These were three mutants at 300 Gy (T2-3, 2.30; T2-69, 2.11; T2-70, 1.61), one at 350 Gy (T3-17, 2.47), one at 0.4% EMS (T4-62, 1.99), two at 0.5% EMS (T5-58, 1.41; T5-65, 2.31), and one at 0.6% EMS (T6-2, 2.18). In PKV Pink, three low ODR mutants were identified compared to 3.69 in the control: T2-45 (2.34 at 300 Gy), T3-43 (2.08 at 350 Gy), and T5-99 (2.26 at 0.5% EMS) (Table 2).

3.3.7. Polyunsaturated-to-Saturated Fatty Acid Ratio (P/S Index)

The P/S index, an important indicator of nutritional value, was substantially improved in several mutants. In AKS 207, three mutants recorded values greater than 9.0 compared to 6.15 in the control: T2-72 (10.39 at 300 Gy), T4-59 (9.52 at 0.4% EMS), and T6-55 (10.06 at 0.6% EMS). In PKV Pink, three mutants also exceeded 9.0 compared to 6.54 in the control: T1-85 (9.73 at 250 Gy), T2-92 (9.56 at 300 Gy), and T6-96 (9.97 at 0.6% EMS) (Table 2).

3.4. Protein Content Variation in Mutants

In the present study, the mutants obtained from AKS 207 ranged from 11.48% to 22.12%, with the control at 15.86% (Supplementary Table S1). A total of eight mutants exceeded the 20% threshold. These included one mutant at 250 Gy (T1-5, 20.23%), three at 350 Gy (T3-7, 20.44%; T3-8, 20.16%; T3-17, 21.04%), two at 0.4% EMS (T4-29, 20.90%; T4-30, 20.12%), and one each at 0.5% EMS (T5-23, 21.46%) and 0.6% EMS (T6-31, 20.34%) (Table 2). These observations suggest that both physical and chemical mutagens were effective in enhancing seed protein concentration, with 350 Gy and moderate EMS doses showing consistent effects. In PKV Pink, protein content varied between 12.11% and 22.51%, compared to the control value of 16.59% (Supplementary Table S2). Three mutants surpassed the 20% threshold: one under 350 Gy (T3-42, 21.39%) and two under 0.5% EMS (T5-51, 22.51%; T5-99, 20.16%) (Table 2). The results indicate that PKV Pink also responded positively to mutagenic treatments, although the frequency of high-protein mutants was comparatively lower than AKS 207.

3.5. Integrated Analyses of Phenotypic Architecture and Pleiotropic Linkages

Correlation analysis revealed distinct positive and negative associations among seed morphological and biochemical traits (Figure 5; Supplementary Table S3). Strong positive correlations were observed between linolenic acid and oleic desaturation ratio (r = 0.94), as well as linolenic acid and P/S index (r = 0.80), indicating that higher levels of linolenic acid were closely associated with increased desaturation and polyunsaturate indices. Similarly, the oleic desaturation ratio and P/S index were highly correlated (r = 0.66), reflecting their common dependence on desaturase activity. Seed size-related traits also showed strong positive relationships. Seed length correlated positively with 100-seed weight (r = 0.70), while seed width was strongly associated with seed length and 100-seed weight. Palmitic acid and oleic acid displayed a significant positive correlation (r = 0.45), suggesting coordinated variation in saturated fatty-acid biosynthesis. In contrast, oleic acid was negatively correlated with the oleic desaturation ratio (r = –0.90), consistent with the biochemical role of Δ12-desaturase in converting oleic acid to linoleic acid. Oleic acid also exhibited strong negative associations with linoleic acid (r = –0.95) and P/S index (r = –0.59), reflecting the trade-off between oleic and polyunsaturated fatty acids. These relationships clearly separate mutants with enhanced oleic content from those with elevated polyunsaturates.

To further validate the correlation results, PCA was performed to investigate the interrelationships between seed morphological and oil quality traits in the mutant population (Figure 6). Of the 13 traits utilized in PC analysis, only the first five principal components with eigenvalues over 1 accounted for most of the variability, capturing approximately 85.39% of the variability among the traits in the examined accessions. The first three principal components (PC1, PC2, and PC3) explained a substantial proportion of the total phenotypic variation, capturing the major trends among traits. Traits such as oleic acid contributed most to the PC1; similarly, seed width and seed length and width ratio contributed to the PC2 and PC3, respectively. Linolenic acid, oleic desaturation ratio, and polyunsaturated-to-saturated fatty acid index clustered closely and loaded strongly on PC1, indicating a high degree of positive correlation among them. In contrast, oleic acid was oriented opposite to these traits, signifying a strong negative association with linoleic acid, the oleic desaturation ratio, and the P/S index. This pattern is consistent with the biochemical pathway in which oleic acid is desaturated to linoleic and linolenic acids by Δ12-desaturase, thereby reducing the oleic desaturation ratio while increasing PUFA-related indices. Seed size–related traits such as seed length, seed width, and 100-seed weight were grouped together and loaded positively on PC2, suggesting that they contribute collectively to seed size and density. Oil content had a moderate correlation with seed weight parameters, suggesting a possible connection between increased seed size and greater oil accumulation.

A set-theory analysis was performed for each variety across nine effective phenotypic traits to quantify the degree of pleiotropy, where a single mutant expressed multiple traits. A critical preprocessing observation confirmed the complete identity of the mutant sets for the traits ‘High oleic acid mutants’ and ‘Oleic desaturation ratio’ in both varieties. The ODR is a measurement fundamentally derived from the concentrations of oleic acid and related fatty acids; therefore, it is anticipated that mutations leading to an elevated oleic acid phenotype would similarly affect the ODR calculation. This biological and analytical redundancy permits the consolidation of the two traits into a single analytical set, effectively reducing the complexity of the Euler diagram structure from 10 dimensions to 9 independent, actionable trait categories. The resulting Euler diagrams (Figure 7A,B) visually map the unique mutant counts found in every trait intersection. AKS 207 exhibited the most complex architecture, showing four-trait pleiotropy (N = 2), notably in mutant T2-3 (linking seed size, high oleic acid, and low palmitic acid). PKV Pink displayed a maximum three-trait overlap (N = 2), such as mutant T6-96, which simultaneously optimized oil content, reduced palmitic acid, and improved the P/S index.

The differential efficacy of the gamma ray and EMS mutagens across the two genetic backgrounds was analyzed using a Sankey diagram (Figure 8). This visualization reveals the quantitative flow of unique mutants obtained from the varieties used in the study. AKS 207 exhibited the greatest response to gamma ray treatments. The 300 Gy dose generated the highest total number of selected mutants, showing successful induction across diverse categories, including bold and long-seeded, high test-weight, and key oil quality traits. Conversely, the majority of successful mutant lines in PKV Pink originated from EMS treatment. The 0.5% EMS concentration was the most effective, serving as the primary source for lines exhibiting enhanced quality traits, particularly the high oleic acid/ODR, low palmitic acid, and improved P/S index. This clear divergence in response highlights the importance of matching the mutagenic agent to the genetic background to maximize the recovery of desirable multi-trait phenotypes.

3.6. Molecular Profiling of Selected Mutants for Higher and Mid-Oleic Acid Content

Oleic acid is one of the most critical determinants of safflower oil quality, as higher concentrations not only improve oxidative stability but also enhance the oil’s nutritional and industrial utility. Since oleic acid levels are primarily regulated by the activity of the fatty acid desaturase 2 (FAD2) enzyme, molecular profiling of mutants with altered oleic acid composition provides valuable insight into the genetic basis of trait modification. In the present study, mutant lines with oleic acid content exceeding 25% were selected for genotyping. Before the molecular analysis, these lines were evaluated three times at monthly intervals to confirm the consistency of the trait expression.

Among the mutants analyzed, T2-3, T3-17, T6-2, and T4-62 of AKS 207 and T3-43 of PKV Pink consistently recorded oleic acid levels above 27% (Table 2 and Supplementary Tables S1 and S2). Amplification with the CtFAD2-1 specific marker was successful only in four mutants of AKS 207 (T2-3, T2-70, T3-17, and T6-2) (Table 2 and Figure 9). The remaining lines failed to amplify, suggesting that their high oleic acid phenotype may not be directly associated with mutations in the CtFAD2-1 locus.

Fluorescence signals generated by the KASP/Amplifluor assay were visualized as allele clustering plots (Figure 9). Samples with strong FAM signals, positioned along the x-axis, were classified as homozygous for the wild-type CtFAD2-1 allele (OLOL), corresponding to the low-oleic phenotype. In contrast, a sample with a strong HEX signal, clustering near the y-axis, was homozygous for the mutant allele (olol), which is associated with the high-oleic phenotype. Among the amplified samples, the M₂ mutants clustered with the homozygous wild-type allele (OL/OL), while only the high-oleic control line Montola-2000 amplified as homozygous for the mutant allele (ol/ol). The lack of amplification in the remaining high-oleic mutants suggests potential deletions or mutations affecting the primer binding sites at the CtFAD2-1 locus, further indicating a novel genetic basis for the trait in these lines.

4. Discussion

The present investigation successfully demonstrated the utility of induced mutagenesis in creating novel genetic variability for seed and oil quality traits in safflower. Both physical (gamma rays) and chemical (EMS) mutagens were effective in generating mutants with significant alterations in seed morphology, oil composition, and protein content. Importantly, several mutants displayed combinations of desirable traits, underscoring the potential of mutagenesis as a tool for crop improvement. Seed trait variability was clear in the M₂ population of both the varieties studied. Mutants with bold and long or short seeds, as well as lines with high test weight, were identified, reflecting the impact of mutagenic treatments on seed size and density. Such seed-associated traits are of direct relevance to yield potential and oil recovery efficiency, consistent with earlier reports in safflower and other crops [47,48,49]. The recovery of high-test-weight mutants suggests alterations in seed density, possibly linked to changes in kernel-to-hull ratio, which has been highlighted as a determinant of oil content and processing quality. Mutagenic effectiveness and efficiency are important to recognize that mutagenic efficiency can vary across plant tissues and individuals due to differential test conditions, which influence the expression of the mutagen’s true potential [34]. Similar findings have been reported previously by Rampure et al. [50] in safflower, Anbarasan et al. [51] in sesame, Julia et al. [52] in Indian mustard, and Saha and Paul [53] in sesame. In the current study, EMS demonstrated a higher mutagenic effectiveness and efficiency compared to gamma rays. This difference was also reflected in the mutation spectrum. EMS treatments proved particularly effective in inducing a broad range of desirable quality traits, including bold seed morphology (e.g., at 0.4% concentration) and biochemical improvements like high oil content (>30%). In contrast, gamma irradiation was more frequently associated with drastic morphological changes (such as small, constricted seeds at 250 Gy) and significant quantitative variation in fatty acid profiles, particularly linoleic acid. This suggests that while both mutagens effectively enhance genetic variability, EMS appears to be the more precise agent for generating agronomically superior mutants in this study.

Oil content and composition were substantially modified among the induced mutants. The isolation of mutants with oil content exceeding 35% (e.g., T5-64), representing an 18% increase over the control, parallels the findings of Sakr et al. [54], who similarly employed mutagenesis to successfully break yield barriers and enhance oil content in safflower. Regarding fatty acid quality, the significant elevation of oleic acid (>25%) observed in our study is consistent with Khadeer and Anwar [55], who reported that induced mutations could effectively alter the fatty acid desaturation pathway. The concomitant reduction in linoleic acid observed in these high-oleic lines reinforces the strong antagonistic relationship described by Yermanos et al. [56]. Furthermore, our identification of high stearic and low palmitic acid mutants mirrors the biochemical shifts reported by Subaşı et al. [22] in EMS-treated safflower populations. These specific alterations, particularly the optimization of ODR and P/S indices, validate the utility of mutagenesis in creating health-beneficial oil profiles, a trend also noted in broader oilseed systems. Protein content variation among mutants was also noteworthy, with several lines recording more than 20% seed protein compared to their respective controls. This trait holds importance not only for human consumption but also for improving the nutritional value of seed meal used in livestock feed. The coexistence of high oil and high protein content in certain mutants highlights the scope for simultaneous improvement of dual traits that are often considered antagonistic in oilseeds. The identified mutants demonstrate the reproducibility of mutagenic approaches in generating desirable seed-size variation, as evidenced by the bold and long-seeded and bold and short-seeded mutants found in the study. The occurrence of bold, long-seeded mutants and bold, short-seeded mutants was aligned with earlier findings of Gawande [57], Rampure et al. [50], and Julia et al. [52]. This study suggests that mutagenesis can generate both elongated and compact seed variants within the same genetic background.

Our observation of continuous variation in oil content across the mutant population supports the understanding that seed oil content is controlled by several genes without apparent dominance, as previously noted by Tonguç et al. [58]. Furthermore, the strong negative correlations we observed between oleic acid and linoleic acid (r = –0.95) align with the significant negative associations reported by Yermanos et al. [56]. This confirms that mutagenesis successfully altered the fatty acid balance while adhering to the fundamental biosynthetic constraints of the crop. Fatty acid compositions, such as oleic, linoleic, and oil contents, varied considerably among safflower varieties, and the highest negative correlation was found between linoleic and oleic acids [59]. Total oil content is the key yield trait for the oilseed crops. In this study, multiple high oil content mutants were identified from both the varieties and from several treatments. Comparable high oil variants in safflower have previously been reported by Sakr et al. [54], confirming the potential of mutagenesis in generating novel oil-rich lines. Similarly, higher oleic acid content in identified mutants is of particular significance for industrial applications, as higher oleic levels are associated with greater oil stability, extended shelf life, and reduced linoleic acid levels [60]. Oils rich in oleic acid are increasingly in demand for use in frying, infant formulations, and spray-based flavoring due to their thermal stability. Similar findings have been reported in safflower by Khadeer & Anwar [55]. The P/S index, an important indicator of nutritional value, was substantially improved in several mutants. Oils with values above 1 are considered beneficial, and safflower mutants frequently exceeded this threshold [25]. These results are aligned with earlier findings reporting safflower oil to have one of the highest P/S indices among vegetable oils, surpassing oils such as linseed, soybean, sunflower, and peanut [25,61].

Proteins obtained from oilseed crops are important for human consumption due to their high nutritional value and health benefits. Enhancing seed protein content is a critical objective in safflower breeding, as it directly increases the economic value of the de-oiled cake used for livestock feed [62,63]. The simultaneous improvement of both oil and protein content is a desirable objective in safflower breeding, as it enhances the dual utility of the crop. The protein content ranges observed in this study are in agreement with previous reports on safflower seed protein values, which range between 14.70% and 16.21% on a dry weight basis [64], while Nagaraj [65] noted a variation from 14% to 19%. The higher values recorded in the present study suggest that mutagenesis can extend the natural variability in protein content. Similar enhancements in seed protein following EMS or gamma irradiation have been documented in other oilseeds. For instance, Olorunmaiye et al. [66] observed significant improvement in crude protein in groundnut with EMS doses ranging from 0.75 to 1.00%, while Pavadai et al. [67] reported the highest protein enrichment in soybean at 0.5% EMS and 500 Gy gamma irradiation. These findings highlight the potential of induced mutagenesis to enrich the nutritional quality of safflower by broadening the genetic base for seed protein content.

The correlation coefficient in plant breeding is a statistical measure that quantifies the strength and direction of the relationship between two traits. It helps breeders understand how changes in one trait may predict changes in another [68,69]. In the current study, the correlation results confirm that seed dimensions contribute directly to the overall seed weight. In contrast, oleic acid was negatively correlated with the oleic desaturation ratio (r = –0.90), consistent with the biochemical role of Δ12-desaturase in converting oleic acid to linoleic acid [68,69]. Oleic acid also exhibited strong negative associations with linoleic acid (r = –0.95) and P/S index (r = –0.59), reflecting the trade-off between oleic and polyunsaturated fatty acids [70,71,72]. These findings are also supported by Rudolphi et al. [73]. These relationships clearly separate mutants with enhanced oleic content from those with elevated polyunsaturates. Similarly, the distribution of traits in the PCA biplot also confirmed the antagonistic relationship between oleic acid and polyunsaturated fatty acids and highlighted the positive contributions of seed morphological traits toward seed yield potential. These insights complement the correlation analysis and provide a multivariate perspective on trait interrelationships within the mutant population. Correlation analysis and PCA provided further clarity on the interrelationships among seed and oil quality traits. Positive correlations between seed size, test weight, and oil content confirm that selection for physical seed traits can indirectly improve oil yield. This observation is consistent with Kante et al. [7], who similarly reported that bolder seed types in safflower are often associated with higher test weights and oil recovery. PCA revealed that a limited number of principal components explained most of the variation, thereby identifying trait clusters that can be targeted simultaneously in selection programs. These multivariate patterns align with the selection indices reported by Cerrotta et al. [27], further validating the efficacy of PCA in isolating superior genotypes from complex breeding populations. Such multivariate approaches enhance the efficiency of breeding strategies by revealing the underlying structure of trait variability in mutagenized populations.

Molecular profiling of high- and mid-oleic acid mutants using CtFAD2-1-specific markers provided useful insights. While some mutants amplified with the marker, several high oleic lines did not, suggesting that their phenotype may be controlled by alternative genetic mechanisms or modifying genes rather than mutations in CtFAD2-1 alone. This lack of CtFAD2-1-based amplification in high-oleic mutants suggests the likely involvement of alternative regulatory loci or epistatic modifiers influencing fatty acid desaturation in safflower, which may indicate a novel genetic pathway for this highly valuable trait. This finding is in agreement with earlier studies [71,74], reporting the role of genetic background and environmental modulation in determining oleic acid levels. The inability of most mutants to amplify the CtFAD2-1 mutation emphasizes the need for deeper genomic studies, including resequencing approaches, regulatory sequences, or genome-wide mutation detection approaches to identify novel allelic variants and regulatory factors controlling fatty acid desaturation in safflower. This observation is consistent with Hamdan et al. [75], who demonstrated that high-oleic inheritance can be influenced by background or modifying genes, leading to segregants with lower oleic levels than their parental sources. Environmental modulation is also known to play a role, as shown by García-Inza et al. [76], who reported temperature-dependent shifts in the balance of oleic and linoleic acids. Thus, the higher oleic acid levels in non-amplified lines from this study may be attributed to modifying loci or environmental responsiveness rather than direct CtFAD2-1 mutations. The biochemical mechanism underlying these findings is well established. The FAD2 enzyme catalyzes the desaturation of oleic acid to linoleic acid during lipid biosynthesis [77]. Defects or mutations in CtFAD2-1 typically reduce desaturation activity, leading to oleic acid accumulation. Previous studies confirmed that the OL allele is associated with structural variation in the CtFAD2-1 gene, and KASP assays were successfully developed to differentiate wild-type OL from mutant ol alleles [19,45].

The outcomes of this study have both scientific and practical significance. From a scientific perspective, they highlight the capacity of induced mutagenesis to broaden the genetic base for complex traits such as oil composition and protein content. From a breeding perspective, the identified mutants provide valuable genetic resources for developing safflower varieties with improved oil quality, better nutritional indices, and enhanced adaptability to diverse end uses. The integration of mutagenesis, biochemical screening, multivariate analysis, and molecular validation represents a comprehensive framework for safflower improvement and can serve as a model for similar efforts in other oilseed crops.

5. Limitations and Future Directions

It is important to note that the KASP markers used in this study were designed to target the specific ol mutation (deletion of ‘C’ at position +603), which is standard in high-oleic breeding programs. Therefore, while our results confirm the absence of this specific allele, we cannot strictly rule out the presence of other, novel mutations or structural variations elsewhere within the CtFAD2-1 coding sequence that were not detected by this assay. Future research will involve targeted deep sequencing of the full CtFAD2-1 gene and promoter regions in these selected mutants to definitively characterize the molecular lesions responsible for the high-oleic phenotype.

6. Conclusions

The present study successfully demonstrated the utility of induced mutagenesis in generating novel genetic variability for seed and oil quality traits in safflower. Substantial variation was observed in seed morphological characteristics, including bold and long-seeded, bold and short-seeded, and high-test-weight mutants, reflecting the effectiveness of both gamma irradiation and EMS in altering seed size and density. Mutants with significantly improved seed test weight and protein content (>20%) were also identified, highlighting their dual value for oil and meal quality improvement. In addition to seed traits, mutagenesis generated considerable variability in oil content and fatty acid composition. Some mutants of AKS 207 and PKV Pink had more oil than their controls, while others had changes in their fatty acid profiles that were desirable, such as more oleic acid (>25%), more stearic acid, and less palmitic acid. These modifications are of particular importance as they enhance oil stability, nutritional value, and suitability for industrial applications. Mutants with a favorable polyunsaturated-to-saturated fatty acid ratio (P/S index >9.0) were also isolated, aligning with the requirements for healthier edible oils. Molecular profiling via KASP assays at the CtFAD2-1 locus indicated that none of the high oleic acid mutants possessed the known ol allele; all M₂ mutants were grouped with the wild-type (OLOL) genotype, whereas Montola-2000 was identified as homozygous for the mutant allele (olol). This suggests that the elevated oleic acid phenotype in the identified mutants may be governed by non-canonical allelic variants, by modifying loci, or by epistatic interactions beyond the CtFAD2-1 gene. These findings provide new insights into the genetic architecture of oleic acid regulation in safflower. Furthermore, the trait correlations and PCA further supported these findings by revealing significant associations among seed, oil, and fatty acid composition traits, demonstrating that mutants with superior oil and protein content often overlapped with those carrying improved fatty acid profiles. Together, these results highlight the potential of induced mutagenesis as an efficient tool for broadening the genetic base of safflower and developing ideotypes with enhanced oil quality, nutritional value, and agronomic performance.