Comparative Analysis of Polyphenolic Compounds in Different Amaranthus Species: Influence of Genotypes and Harvesting Year

Amaranth is a nutritionally valuable crop, as it contains phenolic acids and flavonoids, yielding diverse plant secondary metabolites (PSMs) like phytosterol, tocopherols, and carotenoids. This study explored the variations in the contents of seventeen polyphenolic compounds within the leaves of one hundred twenty Amaranthus accessions representing nine Amaranthus species. The investigation entailed the analysis of phenolic content across nine Amaranthus species, specifically A. hypochondriacus, A. cruentus, A. caudatus, A. tricolor, A. dubius, A. blitum, A. crispus, A. hybridus, and A. viridis, utilizing ultra performance liquid chromatography with photodiode array detection (UPLC-PDA). The results revealed significant differences in polyphenolic compounds among accessions in which rutin content was predominant in all Amaranthus species in both 2018 and 2019. Among the nine Amaranthus species, the rutin content ranged from 95.72 ± 199.17 μg g−1 (A. dubius) to 1485.09 ± 679.51 μg g−1 (A. viridis) in 2018 and from 821.59 ± 709.95 μg g−1 (A. tricolor) to 3166.52 ± 1317.38 μg g−1 (A. hypochondriacus) in 2019. Correlation analysis revealed, significant positive correlations between rutin and kaempferol-3-O-β-rutinoside (r = 0.93), benzoic acid and ferulic acid (r = 0.76), and benzoic acid and kaempferol-3-O-β-rutinoside (r = 0.76), whereas gallic acid showed consistently negative correlations with each of the 16 phenolic compounds. Wide variations were identified among accessions and between plants grown in the two years. The nine species and one hundred twenty Amaranthus accessions were clustered into six groups based on their seventeen phenolic compounds in each year. These findings contribute to expanding our understanding of the phytochemical traits of accessions within nine Amaranthus species, which serve as valuable resources for Amaranthus component breeding and functional material development.


Introduction
Polyphenols or phenolic compounds are present throughout the plant as secondary metabolites that play important roles in the plant's defense mechanisms against stress, pathogens, and UV radiation [1].In food, polyphenols contribute to taste, color, and stability and various studies suggest that diets high in plant polyphenols offer protection against numerous diseases [2].More than 10,000 types of polyphenolic compounds have been identified so far, and they are classified into four types according to their structural characteristics: phenolic acids, flavonoids, stilbenes, and lignans.The determination of polyphenol content in diverse plant species with significant utility is facilitated by studies employing analytical instruments like spectrophotometers, liquid chromatography (LC), and gas chromatography (GC).These investigations, conducted across a spectrum of grain, vegetable, and fruit crops, aim to characterize the specific types and concentrations of polyphenols present in each plant, providing valuable insights into their compositional variations [3][4][5][6].
Amaranth has attracted worldwide attention due to its identification as a source of diverse secondary metabolites, encompassing phytosterols, tocopherols, carotenoids, phenolic acids, and flavonoids [7].Amaranth, a C 4 plant classified within the Amaranthus genus, encompasses around 70 species.Although its primary cultivation occurs in Central and South America, its notable adaptability enables successful growth in diverse environments, extending to temperate-tropical regions globally [8,9].Three prominent grain Amaranthus species extensively cultivated are A. caudatus, A. cruentus, and A. hypochondriacus.In addition to these, 17 other species are specifically cultivated for their leaves [10].During the flowering stage, the leaves of amaranth are abundant in vitamins, minerals, and dietary fiber [11].They exhibit notable moisture and protein content, rendering them well-suited for various applications such as salads, green vegetables, animal feed, and other culinary uses [12,13].Additionally, amaranth leaves have been found to possess pharmacological values, such as cholesterol reduction, anticancer, and anti-inflammatory properties [14], which can be demonstrated through the high polyphenol content in amaranth [3].
In Amaranthus species, both leaves and flowers are characterized by a high content of flavonoids, with rutin emerging as the predominant compound [15,16].Analyzing the polyphenol content of Amaranthus plants across various growth stages reveals distinct patterns.Phenolic acids, specifically feruloylquinic acid and hydroxycinnamic acid, are notably high during the vegetative growth phase.In contrast, flavonoids such as rutin and quercetin become abundant during the flowering stage [17].Furthermore, it has been observed that the polyphenol content in amaranth leaves is subject to variations influenced by both biotic and abiotic stresses, even when derived from the same resource throughout the growth period [18][19][20].Hence, recognizing the significant differences in polyphenol content among accessions, evaluating different species of Amaranthus is crucial to identify potential variations among them.
The characteristics and concentrations of polyphenols in amaranth leaves can vary depending on the growth environment, necessitating a comparative analysis of findings from different studies.Previous research has focused on quantifying polyphenol content in specific amaranth species and resources [17].However, discrepancies in sampling procedures, analysis equipment, and methodologies across studies pose challenges in evaluating and comparing new resources based on prior results [21].This study aims to identify polyphenolic compounds in accessions from nine Amaranthus species.Additionally, we analyze data variations from plants grown in two different years and assess correlations between polyphenolic compounds.The outcomes of this research will contribute valuable insights to breeding initiatives targeting the development of amaranth accessions with enhanced nutritive value.

Plant Materials and Growth Conditions
A total of 120 accessions from 9 Amaranthus species were obtained from the National Agrobiodiversity Center (NAS; http://genebank.rda.go.kr; accessed on 10 January 2022) of the Rural Development Administration (RDA) in the Republic of Korea (Supplementary Table S1).The n9 Amaranthus species included 3 A. blitum, 18 A. caudatus, 11 A. crispus, 7 A. cruentus, 6 A. dubius, 7 A. hybridus, 31 A. hypochondriacus, 30 A. tricolor, and 7 A. viridis accessions.Two weeks after germination, these accessions were transplanted and cultivated in silt loam soil at the experimental field of Chungbuk National University in Korea (36 • 37 ′ 27.7 ′′ N 127 • 27 ′ 15.3 ′′ E) in 2018 and 2019.To enhance the integrity of each accession, six individual plants were established through the transplantation of individual Amaranthus seedlings at 20 cm intervals within the furrows of designated rows, with weekly irrigation applied.Fertilizer was not applied during the experiment to determine the actual genotypic differences throughout the harvesting year.Temperature and precipitation data recorded in Cheongju during the amaranth flowering period (May to August) in 2018 and 2019 were obtained from the Korea Meteorological Administration (Figure 1).

Plant Materials and Growth Conditions
A total of 120 accessions from 9 Amaranthus species were obtained from the National Agrobiodiversity Center (NAS; http://genebank.rda.go.kr; accessed on 10 January 2022) of the Rural Development Administration (RDA) in the Republic of Korea (Supplementary Table S1).The n9 Amaranthus species included 3 A. blitum, 18 A. caudatus, 11 A. crispus, 7 A. cruentus, 6 A. dubius, 7 A. hybridus, 31 A. hypochondriacus, 30 A. tricolor, and 7 A. viridis accessions.Two weeks after germination, these accessions were transplanted and cultivated in silt loam soil at the experimental field of Chungbuk National University in Korea (36°37′27.7″N 127°27′15.3″E) in 2018 and 2019.To enhance the integrity of each accession, six individual plants were established through the transplantation of individual Amaranthus seedlings at 20 cm intervals within the furrows of designated rows, with weekly irrigation applied.Fertilizer was not applied during the experiment to determine the actual genotypic differences throughout the harvesting year.Temperature and precipitation data recorded in Cheongju during the amaranth flowering period (May to August) in 2018 and 2019 were obtained from the Korea Meteorological Administration (Figure 1).

Sampling and Content Extraction
The sampling was conducted three months after planting, wherein the leaves of six plants per accession were freeze-dried using a FreeZone Freeze Dry System (Labconco, Kansas City, MO, USA), ground into powder, and homogenized; only this conjugated sample was gathered for UPLC-PDA analysis.To extract polyphenols, 100 mg of powdered leaf tissue was combined with 1 mL of 75% methanol and subjected to one hour of sonication.Afterward, the mixture underwent centrifugation at 12,000 rpm for 10 min.The resulting clear supernatant was filtered through a 0.2 µm filter into a clean tube, serving as the sample for determining the polyphenol content.

Sample Analysis by UHPLC-PDA
An ultra performance liquid chromatography (UPLC) system (Waters, Milford, MA, USA) equipped with a binary solvent delivery pump, auto-sampler, and a photodiode array (PDA) detector was used to identify individual polyphenols, as described previously [22].The UPLC settings used in this study are summarized in Table 1.The mobile phase consisted of a binary solvent system comprising water (Solvent A) and acetonitrile (Solvent B) supplemented with 0.1% formic acid.A total of 17 phenolic compounds such as, gallic acid, 3,4-dihydroxybenzoic acid, 4-hydroxybenzoic acid, 2,4-dihydroxybenzoic

Sampling and Content Extraction
The sampling was conducted three months after planting, wherein the leaves of six plants per accession were freeze-dried using a FreeZone Freeze Dry System (Labconco, Kansas City, MO, USA), ground into powder, and homogenized; only this conjugated sample was gathered for UPLC-PDA analysis.To extract polyphenols, 100 mg of powdered leaf tissue was combined with 1 mL of 75% methanol and subjected to one hour of sonication.Afterward, the mixture underwent centrifugation at 12,000 rpm for 10 min.The resulting clear supernatant was filtered through a 0.2 µm filter into a clean tube, serving as the sample for determining the polyphenol content.

Sample Analysis by UHPLC-PDA
An ultra performance liquid chromatography (UPLC) system (Waters, Milford, MA, USA) equipped with a binary solvent delivery pump, auto-sampler, and a photodiode array (PDA) detector was used to identify individual polyphenols, as described previously [22].The UPLC settings used in this study are summarized in Table 1.The mobile phase consisted of a binary solvent system comprising water (Solvent A) and acetonitrile (Solvent B) supplemented with 0.1% formic acid.A total of 17 phenolic compounds such as, gallic acid, 3,4-dihydroxybenzoic acid, 4-hydroxybenzoic acid, 2,4-dihydroxybenzoic acid, vanillic acid, caffeic acid, syringic acid, p-coumaric acid, ferulic acid, sinapic acid, rutin, quercetin 3-β-D-glucoside, benzoic acid, kaempferol 3-O-β-rutinoside, quercetin, cinnamic acid, and kaempferol were purchased from Sigma-Aldrich (St. Louis, MO, USA) and used as standards.For each standard compound, a methanol solution was prepared, resulting in a final concentration of 1000 ng/mL.A mixture of all standard compounds was then utilized to establish the calibration curve.The polyphenol content of each sample was calculated using the method described previously [22] with minor modification.

Statistical Analysis
To enable comparisons across various phenolic compounds, an analysis of variance (ANOVA) for significance p < 0.05 and Duncan's multiple range test were carried out.Additionally, for correlation analysis among these compounds, PAST3 software v4.03 [23] was used for principal component analyses (PCA) and hierarchical clustering was performed using R statistical software (Version 4.2.1).The data presented in the figures and tables are represented as mean ± standard deviation.The Relative Polyphenol Content Index (RPCI) was used to compare the levels of 17 polyphenols in the samples using the following formula: Standard score = (Absorbance − mean)/standard deviation RPCI = Average of standard score.
Furthermore, a comprehensive assessment was conducted to ascertain the statistical significance of the influence exerted by the genotype, year, and their interaction (genotype × year) on the levels of 17 polyphenolic compounds.

Hydroxybenzoic Acid
In the analysis conducted across one hundred twenty Amaranthus accessions spanning the years 2018 and 2019, examination of seven hydroxybenzoic acids revealed notable variations.Gallic acid (HBA1) content was observed to be at its lowest in 2018, ranging from 1.8 to 4.5 µg g −1 with an average of 2.0 µg g −1 , while 3,4-dihydroxybenzoic acid (HBA2) content was observed to be at its lowest in 2019, ranged from 2.9 to 14.8 µg g −1 with an average of 6.5 µg g −1 .Conversely, benzoic acid (HBA7) exhibited the highest content across both years, with concentrations of 142.5 µg g −1 in 2018 and 173.5 µg g −1 in 2019, yielding an average of 93.6 µg g −1 across the assessed Amaranthus species (Table 2).Notably, the content of 3,4-dihydroxybenzoic acid (HBA2) remained undetectable in A. blitum accessions throughout the cultivation periods under scrutiny.

Hydroxybenzoic Acid
In the analysis conducted across one hundred twenty Amaranthus accessions spanning the years 2018 and 2019, examination of seven hydroxybenzoic acids revealed notable variations.Gallic acid (HBA1) content was observed to be at its lowest in 2018, ranging from 1.8 to 4.5 µg g −1 with an average of 2.0 µg g −1 , while 3,4-dihydroxybenzoic acid (HBA2) content was observed to be at its lowest in 2019, ranged from 2.9 to 14.8 µg g −1 with an average of 6.5 µg g −1 .Conversely, benzoic acid (HBA7) exhibited the highest content across both years, with concentrations of 142.5 µg g −1 in 2018 and 173.5 µg g −1 in 2019, yielding an average of 93.6 µg g −1 across the assessed Amaranthus species (Table 2).Notably, the content of 3,4-dihydroxybenzoic acid (HBA2) remained undetectable in A. blitum accessions throughout the cultivation periods under scrutiny.

Hydroxycinnamic Acid
In the comprehensive analysis conducted across one hundred twenty Amaranthus accessions spanning the years 2018 and 2019, scrutiny of five hydroxycinnamic acids showcased significant variability.Cinnamic acid (HCA5) content displayed its nadir ranging from 2.0 to 3.4 µg g −1 in 2018, averaging 1.6 µg g −1 , and fluctuating between 1.8 to 6.3 µg g −1 in 2019, averaging 3.1 µg g −1 .Conversely, sinapic acid (HCA4) consistently exhibited the highest content across both years, registering concentrations of 55.4 µg g −1 in 2018 and 47.8 µg g −1 in 2019, resulting in an average of 24.3 µg g −1 across the evaluated Amaranthus species (Table 2).The UPLC-PDA chromatogram revealed the presence of five flavonoid compounds, which showed significant variability during the assessment period.Kaempferol (FLA5) content ranged from 2.2 to 9.3 µg g −1 in 2018 with an average of 3.5 µg g −1 and from 12.9 to 16.6 µg g −1 in 2019 with an average of 14.7 µg g −1 .Remarkably, rutin (FLA1) consistently exhibited the highest content across both years and ranged from 95.7 to 1485.1 µg g −1 in 2018 with an average of 910.4 µg g −1 and from 821.6 to 3166.5 µg g −1 in 2019 with an average of 1869.9 µg g −1 .

Flavonoid
The UPLC-PDA chromatogram revealed the presence of five flavonoid compounds, which showed significant variability during the assessment period.Kaempferol (FLA5) content ranged from 2.2 to 9.3 µg g −1 in 2018 with an average of 3.5 µg g −1 and from 12.9 to 16.6 µg g −1 in 2019 with an average of 14.7 µg g −1 .Remarkably, rutin (FLA1) consistently exhibited the highest content across both years and ranged from 95.7 to 1485.1 µg g −1 in 2018 with an average of 910.4 µg g −1 and from 821.6 to 3166.5 µg g −1 in 2019 with an average of 1869.9 µg g −1 .

Principal Component Analysis
PCA was conducted to compare the contents of seventeen polyphenols among one hundred twenty Amaranthus accessions.The analysis revealed eight principal components (PCs), each with eigenvalues > 1.0, collectively explaining 71.69% of the total variance in

Statistical Analysis of Genotype, Year, and Genotype × Year across 17 Polyphenols
In this study, we conducted an examination of the statistical significance of genotype, year, and genotype × year variables for 17 types of polyphenols.The findings revealed very high statistical significance for all variables across all polyphenols (Table 5).However, sinapic acid and kaempferol exhibited differences from the other 15 polyphenols, demonstrating comparatively lower significance in both year and genotype, respectively.These results suggest that sinapic acid and kaempferol may possess unique characteristics compared to the rest of the polyphenols.The study underscores the potential variation in the impact of genotype and year based on the type of polyphenol, indicating the need for further investigation into the distinctive properties and roles of sinapic acid and kaempferol in the context of polyphenol metabolism and biological activity.In this study, we conducted an examination of the statistical significance of genotype, year, and genotype × year variables for 17 types of polyphenols.The findings revealed very high statistical significance for all variables across all polyphenols (Table 5).However, sinapic acid and kaempferol exhibited differences from the other 15 polyphenols, demonstrating comparatively lower significance in both year and genotype, respectively.These results suggest that sinapic acid and kaempferol may possess unique characteristics compared to the rest of the polyphenols.The study underscores the potential variation in the impact of genotype and year based on the type of polyphenol, indicating the need for further investigation into the distinctive properties and roles of sinapic acid and kaempferol in the context of polyphenol metabolism and biological activity.

Discussion
The study found that all genotypes have similar polyphenol profiles, but there were significant differences in polyphenol concentrations between years and genotypes.During the flowering period from the month of May to August, total precipitation decreased by approximately 50% in 2019 (388.1 mm) when compared to 2018 (728.1 mm), while the total temperature remained relatively stable, with no significant change between 2018 and 2019 with an average of 25 and 24.2 • C, respectively (Figure 1).Sarker and Oba, (2018) reported increased antioxidant activities and 16 polyphenol contents with an increase in drought stress [20], and Barba de la Rosa et al. (2019) mentioned that, in addition to drought stress, external factors like insect damage, light limitation, and nutrient limitation can alter the polyphenol content of a given accession [18].
Numerous scientific reports have addressed the influence of factors such as total phenolic and anthocyanin content, maturity, and diverse plant species on antioxidant capacity [10,24].Phenolic compounds emerge as the predominant antioxidant components, contributing to robust antioxidant activity and stress response in various tested plants [25].To harness the potential of these substantial natural antioxidant sources, additional characterization of the phenolic composition is essential [3].In this study, we observed a broad variation in the content of seven hydroxybenzoic acids, five hydroxycinnamic acids, and five flavonoids across one hundred twenty accessions representing nine different Amaranthus species.
Plants accumulate phenolic compounds in response to various stress and climatic conditions, leading to increased production of reactive oxygen and nitrogen species within the host plant's body [26].Due to the variations in type and content among different plant species, studies measuring content are being conducted on various plant species to comprehensively understand and quantify these differences [27][28][29].Previous studies have identified similarities between quinoa and Amaranthus species in terms of several phenolic acids, flavonoids, and their glycosides [7,30].Similarly, in a study by Khanam et al. (2012), higher contents of gallic acid, vanillic acid, 4-hydroxybenzoic acid, caffeic acid, ferulic acid, sinapic acid, cinnamic acid, and quercetin-3-glucoside were reported in Amaranthus leaves compared to other leafy greens, including komatsuna, mizuna, pok choi, mitsuba, salad spinach, and lettuce [3].This emphasizes the substantial polyphenol content of Amaranthus and its potential applications.
The levels of various polyphenols in A. hypochondriacus and A. tricolor in this study were found to be higher than those reported by previous studies [3,18,31].Furthermore, in comparison with previous studies, the levels of 14 types of polyphenols identified in this study were 2-5 times higher, with rutin's level exceeding 10 times the previously reported amounts.This study facilitates a comprehensive and simultaneous comparison of polyphenol contents among various amaranth species and accessions, offering crucial insights for material development.This study's findings also confirmed notably high rutin content in Amaranthus compared to other polyphenols.Recognizing the substantial differences in polyphenol contents among accessions, adjustments were essential to mitigate potential impacts on the overall analysis.Similar to the antioxidant assay utilizing the Relative Antioxidant Capacity Index (RACI) to account for systematic differences in various antioxidant experiments [32], we introduced the concept of the Relative Polyphenol Content Index (RPCI).The highest RPCI values were observed in A. viridis (1.65) in 2018 and A. hypochondriacus (0.98) in 2019.The overall RPCI was elevated in A. viridis and three grain Amaranthus species.Despite a decrease in A. viridis' polyphenol content in 2019 compared to 2018, its total RPCI value remained the highest among the nine species, indicating that its polyphenol content surpassed that of other species in both years.Seasonal variation significantly influencing flavonoid biosynthesis in Tetrastigma hemsleyanum Diels & Gilg was reported [33].
In this study, a notable and statistically robust positive correlation was observed between rutin and kaempferol-3-O-β-rutinoside, with a correlation coefficient (r) of 0.93.Rutin and kaempferol-3-O-β-rutinoside are rutinosides of quercetin and kaempferol, respectively, which have independent pathways but go through the same biosynthetic mechanism [34].Interestingly, a strong positive correlation (r = 0.98) of rutin and kaempferol-3-O-β-rutinoside content was observed in A. hybridus and A. caudatus in the present study.Furthermore, a previous study by Chen et al. (2018) reported high accumulations of rutin and kaempferol 3-O-rutinoside in the Wuyi Rock tea cultivar, grown in the same environmental conditions subjected to the same cultivation practices, further supporting the observed correlation in our study [35].However, it is noteworthy that Gallic acid and benzoic acid displayed the strongest negative correlation (r = −0.43) in our study.Gallic acid also exhibited negative correlations with other polyphenols, likely due to its less frequent detection in 2019 compared to 2018 (Supplementary Table S4).Numerous investigations have explored the impact of seasonal variations on the production of plant secondary metabolites, specifically focusing on the accumulation of specific compounds in plants [36][37][38].
Hierarchical clustering analysis and PCA confirmed annual variation, dividing the accessions into six groups, independent of their species (Figure 4).Grain Amaranthus accessions were evenly distributed in groups 1, 2, 3, and 5, while all A. viridis and A. biltum accessions were clustered in groups 3 and 6, respectively.Groups 4 and 6 included only vegetable Amaranthus accessions.Group 5 was the most diverse, containing accessions belonging to six of the nine species.Compared with 2018, the polyphenol contents of groups 3, 4, and 6 were lower in 2019, while those of groups 1, 2, and 5 were higher in 2019.In contrast, flavonoid contents were higher in 2019 than in 2018 for all groups (Table 4).The inheritability of chemical traits in plants generally surpasses that of morphological, phenological, and life-history traits [39].This high heritability in chemical traits likely indicates high evolvability, despite potential sensitivity to environmental variation in heritability measures [40].Studies consistently demonstrate high or moderate estimates of heritability for PSMs production across various plant tissues, including leaves [41].The significant variability between the year and genotypes variables for 17 types of polyphenols may contribute to these high heritabilities, suggesting ample genetic variation between species.Similarly, in our study, clustering and PCA analysis confirmed variation, dividing the accessions into six groups independent of their species.

Conclusions
The findings of this study underscore the potential of one hundred twenty Amaranthus accessions spanning nine different species as valuable sources of polyphenols.Despite the high protein and nutrient contents of various Amaranthus species, the genus remains underutilized and often overlooked.Considering that the Amaranthaceae family comprises approximately 70 Amaranthus species, with 20 producing edible leaves and/or grains [31], our study provides a comprehensive evaluation of polyphenol content across different Amaranthus species under varying environmental conditions.These findings unveil new potential applications of Amaranthus spp.For instance, our study highlights that A. viridis and A. hypochondriacus exhibited higher polyphenol content in 2018 and 2019, respectively, compared to other Amaranthus species.Similarly, based on the overall harvest years (2018-2019) among the nine species, A. viridis showcased the highest RPCI, followed by A. hypochondriacus, suggesting their potential for the development of new health-promoting materials.Overall, the results of this study provide essential insights that can guide decision-making processes in Amaranthus breeding programs.

Figure 1 .
Figure 1.Temperature and precipitation data in Cheongju over two years (2018 and 2019).Total represents the sum of temperature and precipitation during the Amaranthus growing season (May to August).Tem: temperature; Pptn: precipitation.

Figure 1 .
Figure 1.Temperature and precipitation data in Cheongju over two years (2018 and 2019).Total represents the sum of temperature and precipitation during the Amaranthus growing season (May to August).Tem: temperature; Pptn: precipitation.

Table 2 .
Descriptive statistics of the contents of 17 polyphenols in leaf extracts of 9 Amaranthus species in 2018 and 2019.

Table 3 .
Eigenvalue and component matrix of the principal component (PC) axes and total variation explained by each PC.
PC: principal component.

Table 4 .
Average cluster values of 17 polyphenol contents of 9 Amaranthus species.

Table 5 .
Exploring the differential impact of genotype and year on polyphenol variability.

Table 5 .
Exploring the differential impact of genotype and year on polyphenol variability.