Comparative Analysis Reveals Changes in Some Seed Properties in Amaranth Mutant Variety ‘Zobor’ (A. hypochondriacus A. hybridus)

The aim of our long-term research program is to improve the quality and quantity of amaranth production through mutation breeding using γ-radiation. In this paper, we present the characterization of the new variety ‘Zobor’ of A. hypochondriacus×A. hybridus developed by radiationinduced mutagenesis of hybrid K-433. Multiyear phenotypic characterization of an important yield parameter (1000-seed weight) showed that the studied mutant variety ‘Zobor’ has an advantage in seed weight over the nonirradiated control seeds of K-433 with predictable performance of this yield trait. ‘Zobor’ exhibited changes in seed morphometric parameters, starch particle size, and pasting properties with no change in amylose content and swelling power. Moreover, the seeds of ‘Zobor’ showed the significantly highest folate content among selected amaranth varieties. The mutant variety could, therefore, be interesting for the development of functional foods and as a low-management crop, attractive for cultivation in Europe.


Introduction
Amaranthus is a cosmopolitan genus of annual and perennial plants characterized by high stress tolerance, environmental adaptability, and ability to grow with minimal external inputs. The genus includes domesticated and endangered species, restricted endemics, and widespread weeds [1]. Amaranthus has more than 80 species [2], which are distributed in many parts of the world and can be grouped as grain (pseudocereals), vegetable, ornamental, and weed types [3]. Within the genus Amaranthus, there is a high genetic variation as a result of outcrossing ranging from 5% to 30% and plant hybridization, which promotes gene exchange. The ability of interspecific hybridization between the weedy and cultivated types, with the first acting as the "reservoir" of unique genes, as well as gene exchange, has increased crop adaptability to adverse environmental conditions [4].
Amaranthus spp. as underutilized species have promising economic and nutritional value. They are valuable plants with a high content of essential nutrients and potential health benefits for consumers, as well as usability for the preparation of healthy food and food additives, vegetables, beverages, animal feed, pharmaceutical products, and industrial nonfood production of byproducts. Many of the amaranth species are medicinally important and possess antiallergic, anticancer, antihypertensive, and antioxidant properties; they several authors [29][30][31]. Pseudocereal grains show different external and internal seed architecture with consequent differences in nutrient distribution [30]. This characterization can, therefore, reveal different and specific microvariations (in seed architecture, shape, color, and weight) within the Amaranthus genus and may be helpful for breeders to select suitable cultivars.
The aim of the study presented here was to analyze some seed properties of the amaranth mutant variety 'Zobor', which was developed by radiation-induced mutagenesis of the hybrid K-433 (A. hypochondriacus × A. hybridus). Phenotypic characterization, evaluation of seed morphological and morphometric traits, and evaluation of an important yield parameter, 1000-seed weight, were carried out. To determine the nutritional potential of the mutant variety, analyses of starch, some starch traits, and folate content were performed. The experiments were conducted over four growing seasons, and the data obtained for the mutant variety 'Zobor' were compared with the original nonirradiated hybrid K-433 and the other grain genotype Ficha (A. cruentus).

Plant Material, Phenotypic Traits, and 1000-Seed Weight
Two species of amaranth were studied in the present investigation: Amarantus cruentus L.-genotype Ficha and Amaranthus hypochondriacus × Amaranthus hybridus-accession K-433 and variety 'Zobor'. The Slovak mutant variety 'Zobor' was bred at the Institute of Plant Genetics and Biotechnology by radiation mutagenesis [19]. 'Zobor' was registered as a new variety with a long-term significantly increased 1000-seed weight over the nonirradiated control hybrid K-433 after DUS test examination. DUS testing is a method of determining whether a newly bred variety is clearly distinguishable from other existing varieties within the same species (distinctness), whether the individual plants of the new variety are uniform at the same stage of propagation (uniformity), and whether these characteristics do not change during subsequent generations (stability). The tests were conducted according to the International Union for the Protection of New Varieties of Plants (UPOV) Guidelines [32] over a standard period of 2 years (2014-2016) in an approved center, the Central Control and Testing Institute for Agriculture in Nové Zámky, Slovakia (122 m above sea level; 47 • 59 9.841" N 18 • 9 50.634" E). 'Zobor' as a candidate for a new variety was compared to a similar variety from the reference collection Koniz (Amaranthus hypochondriacus × Amaranthus hybridus).
The field experiments in the present study were conducted at the locality Nitra (290 m above sea level; 48 • 18 53.442 N 18 • 5 16.75 E). The soil type in Nitra is a Haplic Luvisol with pH of 7.4. The experiment was designed as a block in a split-plot arrangement with four replicates, and the plot size was 2.0 m × 1.5 m (2.5 m 2 ) for each experimental variant. Isolation was provided by a 1 m wide maize plot between each amaranth sample. No fertilizers or sprays were used during the growing season. Seeds were sown in May, and plants were harvested by hand in October each year. The panicles were cut and dried naturally, and the seeds were harvested by hand. Plants were cultivated during four consecutive years (2008-2011 as Y1-Y4, Figure S1).
The weight of 1000 seeds was calculated as the average of 10 independent measurements for each amaranth sample.

Morphological and Morphometric Analysis of the Seeds
For morphological analysis, amaranth seeds were prefixed in 4% (w/v) formaldehyde in 0.1 M potassium phosphate buffer (pH 7.0) for 2 h by vacuum filtration at room temperature. The fixative solution was then replaced with a new one, and seeds were fixed at 4 • C for an additional 24 h. The seeds were then washed (three times in 0.1 M potassium phosphate buffer). After washing, the seeds were dehydrated in a graded ethanol series (30%, 50%, 70%, 80%, 90%, 96%, and 100% ethanol). Ethanol was gradually replaced with xylene (3:1, 1:1, 1:3 v/v) up to 100% xylene and xylene/paraplast plus resin mixtures (3:1, 1:1, 1:3 v/v). Seeds were finally embedded in resin at 60 • C for 48 h. The 10 µm thick semithin sections were cut using a MicroTec Cut 4055 rotary microtome (Walldorf, Germany) and placed on glass slides treated with egg white glycerol adhesive. The sections were deparaffinized, hydrated, and stained with 0.25% (w/v) toluidine blue O in aqueous solution for a few seconds and mounted with Entellan™. The stained sections were observed with an Axioplan 2 light microscope (Carl Zeiss, Oberkochen, Germany), photographed with a Sony DXC-S500 digital color camera system, and analyzed with ImageJ software [33].
For morphometric analysis, 50 dried, undamaged seeds (Y4) typical of the genotype were randomly selected. Images of the selected seeds were acquired using a ZEISS Stereo Discovery V20 stereo microscope (Carl Zeiss, Oberkochen, Germany) equipped with a MRc5 camera. Seeds were placed on a dark carbon tape to prevent displacement, always in the same orientation. The dorsal side of the seed was imaged in a horizontal view, and the ventral side of the seed was imaged in a vertical view. The acquired images were then processed using ZEISS Axiovision software version 4.8 (Carl Zeiss, Oberkochen, Germany). Seeds were automatically segmented and then measured by the software. The following parameters were selected: area (mm 2 ), bound width (mm), bound height (mm), perimeter (mm), diameter (mm), and radius (mm) (Table S1; Figures 1-3). The collected data were statistically analyzed using SAS ® (Cary, NC, USA) Enterprise Guide 5.1 software.

Starch Isolation and Analysis
Amaranth starch was extracted following the acid procedure of Calzetta Resio et al. [34] with some modifications. Amaranth seeds of 0.1 g were soaked for 1.5 h in a solution with SO 2 concentration of 0.041% (w/v) at 53.9 • C. They were surface-dried for 20 min at the same temperature. All samples were homogenized with mortar and pestle and transferred to a microcentrifuge tube. For each 0.1 g of flour, 0.001 g of acarbose, 5 µL of acid protease, 5 µL of viscozyme, and 1 mL of acetate buffer (pH 5.0) were added. The samples were incubated overnight at 37 • C on a shaker. The suspension was centrifuged at 13,000× g for 5 min, the supernatant was discarded, and the pellet was resuspended in 1.0 mL of 2% SDS and centrifuged at 13,000× g for 5 min. The pellet was resuspended in 1.0 mL of 50% (w/v, density 1.6 g/mL) aqueous CsCl and centrifuged at 13,000× g for 5 min. The top layer was mechanically disrupted and centrifuged for 10 min. The supernatant was discarded, and the pellet was resuspended in 1.0 mL of fresh aqueous CsCl at 13,000× g for 10 min. For further purification, the pellet was washed twice with distilled water and twice with ethanol at 13,000× g for 10 min. The purified starch was dried overnight at room temperature.

Morphological and Morphometric Analysis of the Seeds
For morphological analysis, amaranth seeds were prefixed in 4% (w/v) formaldehyde in 0.1 M potassium phosphate buffer (pH 7.0) for 2 h by vacuum filtration at room temperature. The fixative solution was then replaced with a new one, and seeds were fixed at 4 °C for an additional 24 h. The seeds were then washed (three times in 0.1 M potassium phosphate buffer). After washing, the seeds were dehydrated in a graded ethanol series (30%, 50%, 70%, 80%, 90%, 96%, and 100% ethanol). Ethanol was gradually replaced with xylene (3:1, 1:1, 1:3 v/v) up to 100% xylene and xylene/paraplast plus resin mixtures (3:1, 1:1, 1:3 v/v). Seeds were finally embedded in resin at 60 °C for 48 h. The 10 µm thick semithin sections were cut using a MicroTec Cut 4055 rotary microtome (Walldorf, Germany) and placed on glass slides treated with egg white glycerol adhesive. The sections were deparaffinized, hydrated, and stained with 0.25% (w/v) toluidine blue O in aqueous solution for a few seconds and mounted with Entellan™. The stained sections were observed with an Axioplan 2 light microscope (Carl Zeiss, Oberkochen, Germany), photographed with a Sony DXC-S500 digital color camera system, and analyzed with ImageJ software [33].
For morphometric analysis, 50 dried, undamaged seeds (Y4) typical of the genotype were randomly selected. Images of the selected seeds were acquired using a ZEISS Stereo Discovery V20 stereo microscope (Carl Zeiss, Oberkochen, Germany) equipped with a MRc5 camera. Seeds were placed on a dark carbon tape to prevent displacement, always in the same orientation. The dorsal side of the seed was imaged in a horizontal view, and the ventral side of the seed was imaged in a vertical view. The acquired images were then processed using ZEISS Axiovision software version 4.8 (Carl Zeiss, Oberkochen, Germany). Seeds were automatically segmented and then measured by the software. The following parameters were selected: area (mm 2 ), bound width (mm), bound height (mm), perimeter (mm), diameter (mm), and radius (mm) (Table S1; Figures 1-3). The collected data were statistically analyzed using SAS ® (Cary, NC, USA) Enterprise Guide 5.1 software.

Starch Isolation and Analysis
Amaranth starch was extracted following the acid procedure of Calzetta Resio et al. [34] with some modifications. Amaranth seeds of 0.1 g were soaked for 1.5 h in a solution with SO2 concentration of 0.041% (w/v) at 53.9 °C. They were surface-dried for 20 min at the same temperature. All samples were homogenized with mortar and pestle and transferred to a microcentrifuge tube. For each 0.1 g of flour, 0.001 g of acarbose, 5 µL of acid protease, 5 µL of viscozyme, and 1 mL of acetate buffer (pH 5.0) were added. The samples were incubated overnight at 37 °C on a shaker. The suspension was centrifuged at 13,000× g for 5 min, the supernatant was discarded, and the pellet was resuspended in 1.0 mL of 2% SDS and centrifuged at 13,000× g for 5 min. The pellet was resuspended in 1.0 mL of 50% (w/v, density 1.6 g/mL) aqueous CsCl and centrifuged at 13,000× g for 5 min. The top layer was mechanically disrupted and centrifuged for 10 min. The supernatant was discarded, and the pellet was resuspended in 1.0 mL of fresh aqueous CsCl at 13,000× g for 10 min. For further purification, the pellet was washed twice with distilled water and twice with ethanol at 13,000× g for 10 min. The purified starch was dried overnight at room temperature.
The determination of amylose was performed according to the method of Mohammadkhani et al. [35]. Five milligrams of extracted starch including standard maize samples (27% amylose content and 100% maize amylopectin) were used. For each 1 mg of

Starch Isolation and Analysis
Amaranth starch was extracted following the acid procedure of Calzetta Resio et al. [34] with some modifications. Amaranth seeds of 0.1 g were soaked for 1.5 h in a solution with SO2 concentration of 0.041% (w/v) at 53.9 °C. They were surface-dried for 20 min at the same temperature. All samples were homogenized with mortar and pestle and transferred to a microcentrifuge tube. For each 0.1 g of flour, 0.001 g of acarbose, 5 µL of acid protease, 5 µL of viscozyme, and 1 mL of acetate buffer (pH 5.0) were added. The samples were incubated overnight at 37 °C on a shaker. The suspension was centrifuged at 13,000× g for 5 min, the supernatant was discarded, and the pellet was resuspended in 1.0 mL of 2% SDS and centrifuged at 13,000× g for 5 min. The pellet was resuspended in 1.0 mL of 50% (w/v, density 1.6 g/mL) aqueous CsCl and centrifuged at 13,000× g for 5 min. The top layer was mechanically disrupted and centrifuged for 10 min. The supernatant was discarded, and the pellet was resuspended in 1.0 mL of fresh aqueous CsCl at 13,000× g for 10 min. For further purification, the pellet was washed twice with distilled water and twice with ethanol at 13,000× g for 10 min. The purified starch was dried overnight at room temperature.
The determination of amylose was performed according to the method of Mohammadkhani et al. [35]. Five milligrams of extracted starch including standard maize samples (27% amylose content and 100% maize amylopectin) were used. For each 1 mg of The determination of amylose was performed according to the method of Mohammadkhani et al. [35]. Five milligrams of extracted starch including standard maize samples (27% amylose content and 100% maize amylopectin) were used. For each 1 mg of starch, 15 µL of 95% ethanol and 90 µL of 1 M aqueous NaOH were added. The samples were mixed well and shaken overnight at room temperature. The samples were diluted with distilled water to give 1 mg of starch per 200 µL. Subsamples of 200 µL of each starch solution were taken, neutralized with 1 mL of 0.05 M aqueous citric acid, stained with 800 µL of iodine solution, and diluted with 10 mL of distilled water to a final volume of 12 mL. The solutions were cooled to 18 • C for 20 min. The preparation of each sample was repeated four times, and absorbance was measured at 620 nm and 535 nm. The readings were converted to amylose contents in three ways. First, only the 620 nm absorbance curve was used. The absorbance values at 620 nm and 535 nm were evaluated according to ratio-based calculations. These calculations are intended to evaluate both the amylopectin content and the amylose content, thus compensating to some extent for the variations in the total starch content of the sample. where R is the ratio of absorbance at 620 nm to absorbance at 535 nm, AP 535 is the slope of the absorbance curve of amylopectin at 535 nm, AM 535 is the slope of the absorbance curve of amylose at 535 nm, and AP 620 is the slope of the absorbance curve of amylose at 620 nm [36]. The swelling test was developed from the swelling power test and the swelling volume test [37]. The analysis was performed in triplicate per sample. The samples (flour of 40 mg DW) were mixed with water and kept in a water bath at 92.5 • C for 30 min with regular gentle swirling (1 min 20×; 1.5 min 2×; 2 min 2×; 3 min 2×; 4 min 2×; 5 min 2×; 7.5 min 2×; 10 min 2×; 15 min 2×; 25 min 2×). The samples were cooled to room temperature in a water bath (20 • C) for 3 min, being gently inverted twice at the beginning and after 1.5 min. Samples were then centrifuged at 16,000× g for 10 min, and total starch swelling power was calculated as the ratio between the weight of the residue and the DW of the flour.
Particle size analysis was performed in a Coulter LS 230 (Beckman Coulter, Brea, CA, USA) using the entire starch sample in the flow through small-sample volume sampler, optimized for the refractive index of starch in water. The results are represented as weighted mean size and were calculated as particles with a diameter between 0.3 and 3 µm. The analyses of amylose, swelling test, and starch particle size were performed over four growing seasons (Y1-Y4).
Pasting properties of starch were determined using a Rapid Visco Analyzer (RVA, Model 3D, Newport Scientific, Warriewood, NSW, Australia), as described by Konik et al. [38], using Thermocline software for Windows. A flour sample of 3.5 g (14% moisture basis) was weighted into a canister filled with 25 mL of deionized water. The process started at 50 • C for 1 min, followed by a linear increase in temperature to 95 • C in 3 min 42 s, holding at 95 • C for 2 min 30 s, then cooling the system to 50 • C in 3 min 48 s, holding at 50 • C, and terminating the process after 13 min according to AACC method 76-21.02 [39]. The results were reported as temperature of pasting (PT), peak viscosity, viscosity at 95 • C holding (hold), viscosity at 50 • C (final), breakdown, and setback. The samples were measured in duplicate. The pasting properties of starch were made in duplicate from one growing season (Y4).

Folate
For folate analysis, 0.5 g of sample was weighed and extracted with 15 mL of extraction buffer (50 mM CHES/50 mM HEPES, 10 mM 2-mercaptoethanol, 2% sodium ascorbate; pH 7.85) in boiling water for 10 min, followed by a tri-enzyme treatment with α-amylase, hog kidney conjugase, and protease [40]. A blank sample was analyzed for each set of samples.
Total folate assay was carried out in 96-well microtiter plates using glycerol-cryoprotected Lactobacillus rhamnosus (ATCC 7469) as the growth organism and 5-formyltetrahydrofolate as the calibrant [41]. Two dilutions were prepared for each replicate.
Prior to individual folate vitamer determination by UHPLC, sample extracts were purified and concentrated using affinity chromatography with folate-binding protein [42]. Folate vitamers were separated on an HSS T3 column (1.8 µm, 2.1 × 150 mm; Waters, Etten-Leur, NB, The Netherlands) at 30 • C with a gradient of acetonitrile and 30 mM potassium phosphate buffer (pH 2.2) at a flow rate of 0.4 mL/min [43]. Folate vitamers were detected by photodiode array and fluorescence detectors and quantified using external calibration curves as described by Edelmann et al. [43]. Folate analyses were performed in duplicates for two growing seasons (Y3, Y4).

Statistical Analysis
Statistica 10 software (StatSoft Inc., Tulsa, OK, USA) was used for statistical analysis. Analysis of variance (ANOVA) and Tukey's HSD multiple comparison test were used to identify significant differences at p ≤ 0.01 and p ≤ 0.05.

Evaluation of Phenotypic Traits and Weight of 1000 Seeds
Crops can be bred for agronomic improvements such as yield, quality traits, and stress resistance. The differences in the tested phenotypic and morphological traits between the variety 'Zobor' and the reference variety Koniz are shown in Table S2. Distinctions were observed in six characteristics: color of the young leaf lower side, time of flowering, inflorescence color, time of plant maturity, anthocyanin coloration of stem base, and 1000seed weight.
Shape, size, and weight are the most easily identifiable characteristics of individual seeds [44]. In general, seed quality has a great influence on the economic production of agricultural crops. In amaranth, the most important physical quality trait and a critical factor in improving amaranth productivity is seed weight [12].
The 1000-seed weight was examined during four growing seasons (Y1-Y4). Although year, cultivar, and their interactions usually influence seed weight and yield, the γ-radiationderived variety 'Zobor' showed a significantly higher 1000-seed weight compared to the nonirradiated control K-433 in all periods studied ( Figure 4). However, the 1000-seed weight values obtained for the hybrid 'Zobor' were similar to those obtained for the genotype Ficha.

Statistical Analysis
Statistica 10 software (StatSoft Inc., Tulsa, OK, USA) was used for statistical analysis. Analysis of variance (ANOVA) and Tukey's HSD multiple comparison test were used to identify significant differences at p ≤ 0.01 and p ≤ 0.05.

Evaluation of Phenotypic Traits and Weight of 1000 Seeds
Crops can be bred for agronomic improvements such as yield, quality traits, and stress resistance. The differences in the tested phenotypic and morphological traits between the variety 'Zobor' and the reference variety Koniz are shown in Table S2. Distinctions were observed in six characteristics: color of the young leaf lower side, time of flowering, inflorescence color, time of plant maturity, anthocyanin coloration of stem base, and 1000-seed weight.
Shape, size, and weight are the most easily identifiable characteristics of individual seeds [44]. In general, seed quality has a great influence on the economic production of agricultural crops. In amaranth, the most important physical quality trait and a critical factor in improving amaranth productivity is seed weight [12].
The 1000-seed weight was examined during four growing seasons (Y1-Y4). Although year, cultivar, and their interactions usually influence seed weight and yield, the γ-radiation-derived variety 'Zobor' showed a significantly higher 1000-seed weight compared to the nonirradiated control K-433 in all periods studied ( Figure 4). However, the 1000-seed weight values obtained for the hybrid 'Zobor' were similar to those obtained for the genotype Ficha.   Table 1 shows the values of 1000-seed weight of the evaluated grain amaranths. The lowest value of 1000-seed weight was observed in the control hybrid K-433 (0.73 g) in growing season 2, while the highest weight was noted in Ficha and 'Zobor' harvested in seasons 1 and 3 (0.92 g and 0.88 g for Ficha; 0.91 g and 0.87 g for 'Zobor'). Gimplinger et al. [12] tested the genotypes of A. hypochondriacus and A. cruentus L. with similar seed weights as reported in our study. Singh et al. [45] demonstrated that the seed weight of A. hypochondriacus varies between 0.62-0.88 g, which is comparable to the K-433 values. However, Parveen et al. [28] estimated the average 1000-seed weight to be only 0.55 g for eight A. hybridus genotypes.
Pospišil et al. [46] reported an average 1000-seed weight of 0.65-0.73 g for two amaranth species, which was significantly increased by nitrogen fertilization in the dry season. Compared to the commercial grain varieties Plainsman (0.71 g) and Koniz (0.69 g), the hybrid variety 'Zobor' showed 1000-seed weight up to 18% and 21% higher, respectively [23]. From our results, it is evident that the growing season affects the studied yield parameter  (Table 1), but the mutant variety 'Zobor' showed consistent performance over several growing seasons. The increased seed size and yield could have a negative impact on the nutritional properties of amaranth seeds [47]. However, our previous studies [22,23,48] showed high nutritional quality of the amaranth germplasm obtained by γ-radiation with higher seed weight.

Morphological and Morphometric Characterization of the Seeds
The seeds of amaranth are lenticular and about 1 mm in diameter. They vary in color from white to creamish yellow, reddish, dark brown, and black [28][29][30]45,49].
Here, the morphology of amaranth seeds was studied by light microscopy and toluidine blue O staining ( Figure 5). The mature amaranth seed consists of a large, nutrient-rich perisperm (full of polygonal starch cells) surrounded by a peripheric embryo, which is consistent with previous observations described by other authors [30,31]. The seed coat, cotyledons, and shoot apical meristem are also clearly visible in the taken images. According to our observations, the morphology of the 'Zobor' seeds remained unchanged and corresponded to the morphology of nonirradiated amaranth seeds.  Seed size is another important characteristic for agricultural practice and the interest of crop breeders. Seed size may be used to distinguish some species within amaranth genotypes that are otherwise morphologically very similar. The present work shows that the Slovak mutant variety 'Zobor' with a long-term higher 1000-seed weight has visually larger seeds than the untreated control K-433 ( Figure 6). genotypes that are otherwise morphologically very similar. The present work shows that the Slovak mutant variety 'Zobor' with a long-term higher 1000-seed weight has visually larger seeds than the untreated control K-433 ( Figure 6). Seed size is another important characteristic for agricultural practice and the interest of crop breeders. Seed size may be used to distinguish some species within amaranth genotypes that are otherwise morphologically very similar. The present work shows that the Slovak mutant variety 'Zobor' with a long-term higher 1000-seed weight has visually larger seeds than the untreated control K-433 ( Figure 6). Statistical analysis of seed morphometric parameters confirmed that 'Zobor' had significantly greater seed area, bound height, radius and perimeter on the dorsal side of the whole seed, and bound height on the ventral side of the seed compared to the control K-433 (Table 2). Moreover, the variety 'Zobor' exhibited an increase in some seed morphometric parameters over genotype Ficha (Table 2). Statistical analysis of seed morphometric parameters confirmed that 'Zobor' had significantly greater seed area, bound height, radius and perimeter on the dorsal side of the whole seed, and bound height on the ventral side of the seed compared to the control K-433 (Table 2). Moreover, the variety 'Zobor' exhibited an increase in some seed morphometric parameters over genotype Ficha (Table 2). As for the perisperm, 'Zobor' had higher values in area, bound width, bound height, and perimeter compared to control K-433, and higher area, bound width, and perimeter than genotype Ficha (Table 3). There were no significant differences in pericarp area among the amaranths studied. Morphometric characterization of amaranth seeds has not been explored so much. Parveen et al. [28] observed wide variation (considerable diversity) in seed size, shape, color, and weight in eight amaranth genotypes. The measured seed length and width averaged 1.09 mm and 0.98 mm, respectively. Nevertheless, the study of seed morphology and anatomy could contribute and provide useful data for species differentiation and taxonomy classification.

Starch Analysis
Starch is the most abundant carbohydrate in amaranth grain. Amaranth starch has low amylose content, ranging from 2% to 12% depending on the genotype, while amylopectin content is about 90% to 98% [45]. The ratio of amylose and amylopectin affects the physical and chemical properties of starch. Furthermore, a very low amylose content indicates a waxy type of starch. Amaranth amylopectin consists of short-chain branched glucans with an average molecular weight of 11.8 × 10 6 g·mol −1 [50]. The small size of the amaranth starch grain, as well as its high amylopectin content, explains most of the physical properties of amaranth starch. The amylose content correlates significantly with the functional properties, including pasting, thermal properties, and texture, and it appears to be an important determinant of these properties. Considering these facts, amaranth starch shows good gelatinization properties and freeze/thaw stability, which are appreciated in the food industry [27,51,52].
The determination of amylose content, starch swelling power, particle size analysis, and pasting properties in seeds of genotype Ficha, mutant variety 'Zobor', and control K-433 was performed. The amylose content in the analyzed amaranth seeds ranged from 3.80% to 3.93% (Table 4). Similar values of amylose concentration were found by Hoover [53], Choi et al. [54], Chandla et al. [55], Condés et al. [56], and Yuan et al. [57]. Macrone [58] detected 4.1% amylose content in the commercially preferred variety Plainsman, but Baker et al. [59] reported that Plainsman starch does not contain amylose. Although we noted the effect of growing season on amylose content, no significant difference was observed between genotypes over the period tested. The size of starch grains and their geometric features differ between plant species and may influence the functional and physicochemical properties of starch [8]. As reported by Lindeboom et al. [60], there is limited information on the genetic control of starch granule size.
The particle analyzer was employed to study the size of amaranth starch granule size, and some differences in the granule size distribution were identified (Figure 7). The highest value of small starch granules (0-0.95 µm) was found in 'Zobor' (16.45%) compared to Ficha (11.53%) and control K-433 (4.11%). The volume-weighted mean size of investigated amaranth starch was 1.42 µm, which is in agreement with previous studies [52,61]. The mutant variety 'Zobor' had significantly smaller average granule size compared to the control hybrid K-433 and genotype Ficha (Table 4). Similarly, Wang et al. [62] investigated smaller starch granules in rice mutant lines compared to wild type. Macrone [58] examined a starch granule diameter of 1 µm in A. pumilus and A. hypochondriacus. We assume that the smaller size of starch granules of 'Zobor' compared to the nonirradiated control K-433 is compensated for by their higher content, but this needs to be verified by more detailed morphological analysis.
est value of small starch granules (0-0.95 µm) was found in 'Zobor' (16.45%) compared to Ficha (11.53%) and control K-433 (4.11%). The volume-weighted mean size of investigated amaranth starch was 1.42 µm, which is in agreement with previous studies [52,61]. The mutant variety 'Zobor' had significantly smaller average granule size compared to the control hybrid K-433 and genotype Ficha (Table 4). Similarly, Wang et al. [62] investigated smaller starch granules in rice mutant lines compared to wild type. Macrone [58] examined a starch granule diameter of 1 µm in A. pumilus and A. hypochondriacus. We assume that the smaller size of starch granules of 'Zobor' compared to the nonirradiated control K-433 is compensated for by their higher content, but this needs to be verified by more detailed morphological analysis. Gelatinization of starch is a process in which water disperses into the starch granule and causes swelling due to hydration of the amorphous phase, resulting in loss of crystallinity and molecular order. Depending on the origin of the starch, there is a specific temperature range for gelatinization related to the starch of each origin. The gelling ability of starch is of great importance in food production, and the textural properties of starch gels are used to evaluate the behavior of starch in food systems [63]. Gelatinization of starch is a process in which water disperses into the starch granule and causes swelling due to hydration of the amorphous phase, resulting in loss of crystallinity and molecular order. Depending on the origin of the starch, there is a specific temperature range for gelatinization related to the starch of each origin. The gelling ability of starch is of great importance in food production, and the textural properties of starch gels are used to evaluate the behavior of starch in food systems [63].
The starch swelling test is a simple, inexpensive laboratory method for measuring water absorption during gelatinization of starch [37]. Starch swelling power is related to grain structure and chemical composition, especially amylose and lipid content, rather than granule size [60,61].
In this case, no statistically significant difference in starch swelling power was found among the studied amaranth cultivars ( Table 4). The average value of starch swelling power of amaranth was 9.66 g·g −1 . Similar values were reported by Chandla et al. [55] who analyzed the starch swelling power of four different amaranth cultivars with values ranging from 8.10 to 10.29 g·g −1 . Choi et al. [54] compared the swelling power of amaranth starch with that of waxy sorghum and waxy millet starch, with the greatest difference observed at temperatures above 75 • C. It was confirmed that the swelling power of waxy starches has higher values than normal starches. Kong et al. [52] investigated this starch property in 15 amaranth cultivars at different temperatures. The authors reported a swelling power at 85 • C ranging from 9.1 to 20 g·g −1 . The swelling power of native and processed A. hypochondriacus flour studied by Siwatch et al. [64] ranged from 3.58 to 7.55 g·g −1 , with the swelling power decreasing after processing.
Starch gelatinization is influenced by amylose content and starch granule size. The gelatinization properties of amaranth starch may vary depending on the species and cultivars [65].
The pasting properties of three amaranth genotypes were determined using a Rapid Visco Analyzer (Table 5). The RVA profiles of 'Zobor', Ficha, and K-433 are shown in Figures S2-S4.
The mutant cultivar 'Zobor' had the significantly lowest pasting temperature, viscosity peak, holding strength, and final viscosity compared to Ficha and the control K-433 (Table 5). The pasting temperature of amaranth flour ranged from 70.18 to 74.25 • C. Chandla et al. [61] reported the pasting temperature and peak viscosity of A. hypochondriacus Druga as 74.31 • C and 1718.00 cP, respectively. Yuan et al. [57] investigated the pasting and gelling properties of amaranth flour and found similar average values for peak viscosity, holding strength, and final viscosity at 95 • C to those obtained in our study. The authors compared waxy and normal commercial flours at high heating temperatures, and the starch content of the flours was positively correlated with their peak and breakdown viscosities. Previous reports [66][67][68] suggested that starch properties can be affected by radiation as a processing technology. The γ-irradiation hydrolyzes the chemical bonds of starch, which leads to the decomposition of the polymer chain. The subsequently modified starch, characterized by increased solubility or decreased viscosity, can be used in the paper and textile industries. Kong et al. [68] studied the effect of γ-radiation on the rheological properties of amaranth starch and found that the viscosity of starch decreased with increasing radiation dose. Irradiation even affected the temperature properties and crystallization of the starch. The results achieved by Kong et al. [69] showed that mutation breeding based on γ-irradiation is an efficient way to obtain novel starch sources for various food and nonfood applications, which have the potential to replace chemically modified starches.
We can conclude that γ-irradiation could induce a change in starch particle size and pasting properties in the variety 'Zobor', although the amylose content and the swelling power were maintained.

Folate
The results of the determination of total folate and the distribution of the six folate vitamers in amaranth seeds are provided in Table 6. To obtain consistent data, two methods were used to determine the folate content-microbiological assay (MA) and ultrahighperformance liquid chromatography (UHPLC). The UHPLC values of total folate content could be lower compared to the microbiological values due to some unidentified folate compounds, whereas MA may give a response to non-folate compounds [42].  [70]. However, when comparing our results with older studies, higher values were obtained, such as those of Schoenlechner et al. [6]. The authors determined the total folate content from 528 to 730 ng·g −1 ng·g −1 DW in a flour of four amaranth varieties. Furthermore, amaranth samples in our study had higher folate content than described for oat grains [43], rye [42], and many cereal products [71], but lower than buckwheat and quinoa [72].
Folate is known to be an unstable vitamin, with 35-70% loss due to storage and processing [6]. Motta et al. [73] investigated the effects of processing on folate content in amaranth. The authors found that boiling and steaming reduced total folate content by 58% and 22%, respectively. Schoenlechner et al. [6] also determined the total folate loss in processed amaranth products such as bread, cookies, and pasta, and the folate loss in cookies was the lowest.
There is a lack of literature dealing with the distribution of folate vitamers. Here, we identified tetrahydrofolate (THF), 5-methyltetrahydrofolate (5-CH 3 -H 4 ), 10-formylfolic acid (10-CHO-PGA), 5-formyltetrahydrofolate (5-CHO-H 4 ), 5,10-methenyltetrahydrofolate (5,10-CH + -H 4 ), and folic acid (PGA) as naturally occurring folate derivatives in amaranth seeds. The major vitamer in all three amaranth samples was 5-CHO-H 4 , and the next abundant vitamers were 5-CH 3 -H 4 , PGA, and 10-CHO-PGA ( Figure 8). 5-CHO-H 4 , which is considered rather stable, is also the predominant vitamer in cereal products [71]. Motta et al. [73] identified 5-CH 3 -H 4 as the most abundant form of folate in amaranth, quinoa, and buckwheat. However, the authors detected only three folate vitamers. Folic acid is usually present in very small amounts in natural products. We suppose that the relatively high folic acid content determined here occurred mainly due to storage and sample preparation.  Our results show that the variety 'Zobor' has the highest total folate content among the tested amaranths. Therefore, fortification of cereal products with this amaranth flour is an interesting approach to enrich the final product and increase its folate content.

Conclusions
Comparative analysis of the Slovak mutant variety 'Zobor', nonirradiated hybrid K-433, and genotype Ficha revealed changes in some agronomic, morphometric, and nutritional seed traits. The irradiation developed variety 'Zobor' showed consistent performance in the important quantitative trait 1000-seed weight, which was significantly higher than in the control grain amaranth. Moreover, the seeds of the mutant variety were significantly larger compared to the control seeds of hybrid K-433 and genotype Ficha, as evidenced by several increased morphometric parameters. Changes in starch particle size Our results show that the variety 'Zobor' has the highest total folate content among the tested amaranths. Therefore, fortification of cereal products with this amaranth flour is an interesting approach to enrich the final product and increase its folate content.

Conclusions
Comparative analysis of the Slovak mutant variety 'Zobor', nonirradiated hybrid K-433, and genotype Ficha revealed changes in some agronomic, morphometric, and nutritional seed traits. The irradiation developed variety 'Zobor' showed consistent performance in the important quantitative trait 1000-seed weight, which was significantly higher than in the control grain amaranth. Moreover, the seeds of the mutant variety were significantly larger compared to the control seeds of hybrid K-433 and genotype Ficha, as evidenced by several increased morphometric parameters. Changes in starch particle size and pasting properties were observed in 'Zobor' seeds, although amylose content and swelling power were not altered. 'Zobor' had the highest folate contents, with similar vitamer distribution to the other two samples. To sum up, the current study showed that the mutant variety 'Zobor' represents stable, potentially high-yielding variety with high folate content.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/agronomy11122565/s1: Figure S1. Meteorological conditions during four amaranth cropping seasons in locality Nitra; Figure S2. Pasting properties of Slovak amaranth variety 'Zobor' obtained with Rapid Visco Analyzer (RVA, Model 3D, Newport Scientific); Figure S3. Pasting properties of genotype Ficha obtained with Rapid Visco Analyzer (RVA, Model 3D, Newport Scientific); Figure S4. Pasting properties of hybrid K-433 obtained with Rapid Visco Analyzer (RVA, Model 3D, Newport Scientific); Table S1. Parameters measured on the amaranth seeds; Table S2. Differences of phenotypic traits between reference variety Koniz and new variety 'Zobor'. Funding: This publication was supported by the Operational Program Integrated Infrastructure within the project: Demand-driven research for the sustainable and innovative food, Drive4SIFood 313011V336, co-financed by the European Regional Development Fund.