Phenolic Composition and Antioxidant Activity of Purple Sweet Potato (Ipomoea batatas (L.) Lam.): Varietal Comparisons and Physical Distribution

The outer layer of purple sweet potato is removed during processing; however, this layer serves as a potential source of phenolics, especially anthocyanins. Herein, the phenolic composition and antioxidant activity were determined for the inner and outer layers of five purple sweet potato cultivars (‘Sinjami’, ‘Jami’, ‘Danjami’, ‘Yeonjami’, and ‘Borami’) harvested in Korea. Anthocyanins were identified using ultra-high-performance liquid chromatography coupled with electrospray ionization quadrupole time-of-flight mass spectrometer (UHPLC-(ESI)-qTOF-MS) and ultra-high-performance liquid chromatography-linear ion trap mass spectrometer (UHPLC-Ion trap-MS), and their composition was quantified using HPLC-coupled with diode array detector (DAD). Non-anthocyanin phenolic compounds (phenolic acids and flavonols) were quantified using UHPLC-(ESI)-triple quadrupole (QqQ). A total of 20 anthocyanins, including non-acylated or acylated peonidin, cyanidin, and pelargonidin glycosides, were identified. Peonidin 3-caffeoyl-p-hydroxybenzoyl sophoroside-5-glucoside was the major anthocyanin, with the highest level in the ‘Sinjami’ cultivar (outer; 12,366 mg/kg DW, inner; 14,832 mg/kg DW). Additionally, 12 phenolic acids and 6 flavonols (quercetin derivatives) were identified, with the outer layers of all cultivars displaying higher total levels than the inner layers. ‘Sinjami’ and ‘Jami’ had higher phenolic acid and quercetin derivative content and antioxidant activities than the other three cultivars (p < 0.05). Thus, the outer layers of ‘Sinjami’ and ‘Jami’ cultivars could be potential sources of anthocyanins and other phenolics.


Introduction
Anthocyanins are natural pigments responsible for the blue, red, or purple color in fruits and vegetables [1,2]. Furthermore, anthocyanins have protective effects in conditions such as cancer, inflammation, and cardiovascular diseases [3]. Safety concerns, such as adverse behavioral and neurological effects, of synthetic colorants commonly used in the food industry, have been raised over the past few years. In response to increasing consumer and food manufacturer demand for natural colorants, anthocyanins, having possible health benefits, could be promising alternatives to synthetic colorants as they have been consumed for centuries without any adverse effects [4,5].
Sweet potato (Ipomoea batatas L.) belongs to the family Convolvulaceae. It is believed to have originated from Latin America. In the 16th century, sweet potato was introduced to Asia, Africa, and Europe. Worldwide sweet potato production is estimated to be 105 million tons, with Asia accounting for 74.7% (79 million tons) of global production in 2016 [6,7]. Purple sweet potato (Ipomoea batatas (L.) Lam.) is a rich source of anthocyanins (515-1747 mg/kg fresh weight (FW)) [8]. The major anthocyanins in purple sweet potato are peonidin 3-sophoroside-5-glucoside and cyanidin 3-sophoroside-5-glucoside, which are mono-or di-acylated with caffeic, ferulic, and p-hydroxybenzoic acids [9]. Acylated anthocyanins account for more than 98% of total anthocyanin contents in purple sweet potato [10,11]. Although anthocyanins are natural colorants preferred by consumers over synthetic colorants, they are not stable during food processing. Previous studies showed that the stability of acylated anthocyanins during processing increases with an increasing number of acyl groups [12,13].
Anthocyanin composition is dependent on the food source and cultivar used. Varietal differences of anthocyanins were previously reported for 'Borami', 'Jami', 'Sinjami', and 'Yeonjami' purple sweet potato cultivars [21,22]; however, anthocyanin composition of other commercially important purple sweet potato cultivars in Korea, including 'Danjami', is unknown. There is also limited information on other phenolic compounds in purple sweet potato. More importantly, the physical distribution of phenolic compounds in purple sweet potato is unknown. Most studies have used a whole purple sweet potato to determine anthocyanins and other phenolic compounds [23,24]. When a purple sweet potato is cut horizontally, a white round band divides the outer and inner layers; the outer layer is often discarded during processing [25]. However, the outer layer is a possible commercial source of phenolics, especially anthocyanin pigments.
To address these issues, varietal differences and physical distribution (inner and outer layers) of anthocyanins and non-anthocyanin phenolic compounds were investigated in five commercially important purple sweet potato cultivars; their antioxidant activity was also determined.

Chemicals and Reagents
Anthocyanin standards (cyanidin-3-O-glucoside, peonidin-3-O-glucoside, and pelargonidin-3-O-glucoside) were purchased from Extrasynthese (Lyon Nord, France). Gallic and formic acid, Trolox, and 2,2-diphenyl-1-picrylhydrazyl (DPPH) were obtained from Sigma-Aldrich (S4t. Louis, MO, USA). HPLC-grade methanol and acetonitrile were purchased from Burdick & Jackson (Muskegon, MI, USA). The water used in this study was obtained from a water purification system (Milli-Q Direct 8, Merck Millipore, Billerica, MA, USA).  Figure 1. All samples used were peeled and cut into 1.5 cm slices. Inner and outer layers were divided by the jalapin layer of the cross-section. Samples were freeze-dried and stored at −80 • C until analysis. Before analysis, freeze-dried samples were ground into a fine powder using a high-speed blender (Nutri Ninja Duo Auto-iQ, SharkNinja, Shenzhen, China), and the powder was passed through a 0.85-mm sieve to obtain uniform size.

Extrtaction of Anthocyanin and Non-Anthocyanin Phenolic Compounds
Five milliliters of 0.2% HCl in methanol was added to 500 mg lyophilized purple sweet potato powder. The mixture was sonicated for 15 min using a sonicator (CPX3800, Branson Ultrasonic Corporation, Danbury, CT, USA) and centrifuged at 11,124× g for 20 min at 4 °C. This procedure was repeated four times until the purple color of the supernatant disappeared. All supernatant extracts were transferred into a volumetric flask and made up to 30 mL with extraction solvents. An aliquot (10 mL) of the extract was evaporated under nitrogen using a nitrogen concentrator (MG-2200, EYELA, Tokyo, Japan) and re-dissolved in aqueous 0.2% HCl (1.5 mL). Solid-phase extraction (SPE) was performed for further anthocyanin purification, as described previously with slight modifications [26]. After SPE, cartridges were preconditioned with 6 mL methanol and 6 mL aqueous 0.2% HCl, and the extract was loaded onto 1 g (6 cc) Sep-Pak C18 cartridges (Waters, Milford, MA, USA). After loading, the cartridge was washed with 6 mL aqueous 0.2% HCl. Anthocyanins were eluted with 8 mL 0.2% HCl in methanol. The eluted anthocyanin fraction was concentrated to dryness using the nitrogen concentrator (MG-2200), re-dissolved in 1 mL water and methanol (1:1, v/v), and filtered using a 0.22-μm polyvinylidene fluoride (PVDF) filter before HPLC and LC-MS analyses.
For the analysis of non-anthocyanin phenolic compounds, 500 mg of lyophilized purple sweet potato powder was mixed with 10 mL of 80% aqueous methanol solution, as described previously [27]. The mixture was sonicated for 15 min using the sonicator (CPX3800) and centrifuged at 11,124× g for 15 min at 4 °C. Supernatants were concentrated using a Speedvac concentrator (SPD 2010, Thermo Scientific, Waltham, MA, USA) and redissolved in 1 mL aqueous 0.1% formic acid. The re-dissolved concentrates were filtered through a 0.22-μm PVDF filter. Anthocyanin and non-anthocyanin analyses were performed within 1-2 days after extraction. All extraction procedures were performed in triplicate (n = 3).

Extrtaction of Anthocyanin and Non-Anthocyanin Phenolic Compounds
Five milliliters of 0.2% HCl in methanol was added to 500 mg lyophilized purple sweet potato powder. The mixture was sonicated for 15 min using a sonicator (CPX3800, Branson Ultrasonic Corporation, Danbury, CT, USA) and centrifuged at 11,124× g for 20 min at 4 • C. This procedure was repeated four times until the purple color of the supernatant disappeared. All supernatant extracts were transferred into a volumetric flask and made up to 30 mL with extraction solvents. An aliquot (10 mL) of the extract was evaporated under nitrogen using a nitrogen concentrator (MG-2200, EYELA, Tokyo, Japan) and re-dissolved in aqueous 0.2% HCl (1.5 mL). Solid-phase extraction (SPE) was performed for further anthocyanin purification, as described previously with slight modifications [26]. After SPE, cartridges were preconditioned with 6 mL methanol and 6 mL aqueous 0.2% HCl, and the extract was loaded onto 1 g (6 cc) Sep-Pak C18 cartridges (Waters, Milford, MA, USA). After loading, the cartridge was washed with 6 mL aqueous 0.2% HCl. Anthocyanins were eluted with 8 mL 0.2% HCl in methanol. The eluted anthocyanin fraction was concentrated to dryness using the nitrogen concentrator (MG-2200), re-dissolved in 1 mL water and methanol (1:1, v/v), and filtered using a 0.22-µm polyvinylidene fluoride (PVDF) filter before HPLC and LC-MS analyses.
For the analysis of non-anthocyanin phenolic compounds, 500 mg of lyophilized purple sweet potato powder was mixed with 10 mL of 80% aqueous methanol solution, as described previously [27]. The mixture was sonicated for 15 min using the sonicator (CPX3800) and centrifuged at 11,124× g for 15 min at 4 • C. Supernatants were concentrated using a Speedvac concentrator (SPD 2010, Thermo Scientific, Waltham, MA, USA) and re-dissolved in 1 mL aqueous 0.1% formic acid. The re-dissolved concentrates were filtered through a 0.22-µm PVDF filter. Anthocyanin and non-anthocyanin analyses were performed within 1-2 days after extraction. All extraction procedures were performed in triplicate (n = 3).
For anthocyanin identification, an in-house accurate mass database of all potential anthocyanins was built, and the database included previously reported anthocyanins in purple sweet potato cultivars and possible anthocyanins present in samples. The theoretical mass of each possible anthocyanin was calculated based on the molecular formula. Potential anthocyanins in extracts were identified based on a comparison of measured accurate masses to the calculated theoretical masses, using Waters MassLynx MS software version 4.1. Identification was assisted by MS n data-dependent mode of LTQ Velos Pro (ThermoFisher Scientific, Austin, TX, USA) linear ion trap mass spectrometer (UHPLC-Ion trap-MS) to determine fragments of the identified anthocyanins. For UHPLC-Ion trap-MS, mobile phases were 0.5% formic acid in water (A) and 0.5% formic acid in acetonitrile (B). Anthocyanins were separated on a Zorbax Eclipse XDB-C18 column (4.6 mm × 250 mm, 5 µm; Agilent Technologies, Santa Clara, CA, USA). The mobile phase gradient program used was as follows: 0-6.5 min, 10-12% (B); 6.5-10.5 min, 12-13% (B); 10.5-33 min, 13-17% (B); 33-60 min, 17%-65% (B); and 60-70 min, 65-95% (B). The injection volume was 20 µL, and the column oven temperature was maintained at 30 • C. The flow rate was 1.0 mL/min; half of the flow was diverted to the MS, and the rest was moved to waste. Capillary and source heater temperatures were 275 • C and 250 • C, respectively. Source voltage was 5.0 kV, and the S-lens RF level was 60.00%. Pseudomolecular ions were scanned in positive mode full scan and MS n data-dependent mode of UHPLC-Ion trap-MS with a mass range of 50-2000 m/z.

Quantification of Anthocyanins Using HPLC-DAD
Anthocyanin content was quantified using an Agilent 1260 Infinity II HPLC system (Agilent Technologies) with a diode array detector (DAD) [28]. Anthocyanin compounds were separated on a Zorbax Eclipse XDB-C 18 column (4.6 × 250 mm, 5 µm; Agilent Technologies). Mobile phases were 1.5% formic acid in water (A) and 1.5% formic acid in acetonitrile (B). The same gradient program as that for the UHPLC-Ion trap-MS method was used for quantifying anthocyanins using HPLC-DAD. The injection volume was 20 µL, and the column oven temperature was maintained at 30 • C. The flow rate was 1.0 mL/min, and detection was conducted at 520 nm.
The multiple reaction monitoring (MRM) mode was used to determine the phenolic compounds. The fragment voltage (FV) and collision energy (CE) were optimized for each phenolic compound to detect product ions at maximum intensity. MRM transitions from the precursor ion to the product ion were selected as follows: caffeic acid, m/z 179 > . Absolute quantification of caffeic acid, ferulic acid, chlorogenic acid, p-coumaric acid, quercetin 3-O-galactoside, and quercetin-3-O-glucoside was achieved using authentic standards. Quercetin diglucosides and cis-ferulic acid were quantified using standard curves of quercetin 3-O-glucoside and trans-ferulic acid, respectively. Caffeoylquinic acid and dicaffeoylquinic acids were quantified using a chlorogenic acid standard curve. Quantification results are expressed as mg/kg dry weight (DW).

Total Phenolic Content Measurement
Total phenolic content (TPC) was measured using the Folin-Ciocalteu reagent, as described previously [28]. Diluted sample extracts (40 µL) were added to 50 µL of 1 N Folin-Ciocalteu reagent and allowed to react for 1 min. Approximately 160 µL of 2% sodium carbonate (Na 2 CO 3 ) solution was added to the mixture. After 30 min of incubation at 25 • C, absorbance at 700 nm was measured using a spectrophotometer (Multiskan GO, Thermo Scientific, Waltham, MA, USA). TPC was expressed as mg gallic acid equivalent/100 g dried weight (mg GAE/100 g DW).
An ABTS assay was conducted as described previously, with some modifications [30]. ABTS radical cation (ABTS •+ ) solution was produced by reacting a 7 mM ABTS stock solution with 2.45 mM potassium persulfate (1:1, v/v) for 16 h before use, and the solution was diluted with methanol to obtain a starting absorbance of 0.77 at 734 nm. Approximately 300 µL ABTS •+ solution was mixed with 20 µL of sample extracts, and the absorbance was measured at 734 nm after 10 min. Results of the DPPH and ABTS assays were expressed as mg Trolox equivalent antioxidant capacity (TEAC)/g DW.

Statistical Analysis
Statistical analyses, including one-way analysis of variance (ANOVA), followed by Duncan's post hoc test and Pearson's correlation analysis, were performed using IBM SPSS Statistics 25 (v. 25.0, SPSS, Inc., Chicago, IL, USA). Principal component analysis (PCA) and heatmap of anthocyanins and non-anthocyanins, determined by HPLC-DAD and UHPLC-QqQ, were performed using XLSTAT (Microsoft Excel Add-in Software, New York, NY, USA).

Quantification of Anthocyanins in Purple Sweet Potato Cultivars
The total content of peonidin-based anthocyanins ranged from 6544 to 26,483 mg/kg DW, and that of cyanidin-based anthocyanins ranged from 943 to 3962 mg/kg DW. In each cultivar, peonidin-based anthocyanins showed much higher contents than cyanidinbased anthocyanins, except for the 'Borami' cultivar. The total peonidin-based anthocyanin content in the 'Sinjami' cultivar was approximately three-fold higher than the total cyanidin-based anthocyanin content. The total pelargonidin-based anthocyanin content was represented only in the outer and inner layer of the 'Borami' cultivar, ranging from 1242 to 2181 mg/kg DW. Among identified anthocyanins, 3-sophoroside-5-glucoside derivatives of peonidin and cyanidin, acylated with caffeic, ferulic, and p-hydroxybenzoic acids, predominated the anthocyanin composition of purple sweet potato. Relative proportions of cyanidin-, peonidin-, and pelargonidin-based anthocyanins in the 'Borami' cultivar were 1:6:1 in the outer layer and 1:7:1 in the inner layer. The ratio of cyanidin-, peonidin-, and pelargonidin-based anthocyanins was reported as 1:3:2 in the 'Borami' cultivar (980, 2810, and 2170 mg/kg DW, respectively) [22].
Anthocyanidin stability depends on the number of hydroxyl or methoxyl groups bonded with the B ring of aglycone. Their presence can decrease the aglycone stability; therefore, pelargonidin could contribute to higher stability than cyanidin and peonidin [34]. In a previous study, peonidin acylated with two p-coumaric acids showed higher stability than cyanidin acylated with two p-coumaric acids [35]. However, as mentioned above, anthocyanin stability can be further affected by the degree of acylation with phenolic acids. Therefore, it is necessary to study the relationship between changes in anthocyanin stability with the type of aglycone and the degree of acylation further.
In the outer and inner layers of all five cultivars, the most predominant anthocyanin component was peonidin 3-caffeoyl-p-hydroxybenzoyl sophoroside-5-glucoside (peak 17), as reported by previous studies [21,31]. The most abundant anthocyanin component in the 'Jami' cultivar was cyanidin 3-caffeoyl-p-hydroxybenzoyl sophoroside-5-glucoside, expressed as mg cyanidin-3-glucoside equivalent/100 g DW purple sweet potato [22]. To our knowledge, this is the first study to report the anthocyanin composition of the 'Danjami' cultivar. Table 3 shows the non-anthocyanin phenolic content in purple sweet potato cultivars analyzed using UHPLC-(ESI)-QqQ. The UHPLC-(ESI)-QqQ extracted ion chromatograms of non-anthocyanin phenolic compounds are shown in Figure S2. A total of 18 phenolic compounds belonging to hydroxycinnamic acid and flavonol classes were quantified in the MRM mode. Among the phenolic classes evaluated, chlorogenic acid (3-O-caffeoylquinic acid), two caffeoylquinic acid isomers, and five dicaffeoylquinic acid isomers were the predominant phenolic compounds in both outer and inner layers of the purple sweet cultivars. Table 3. Phenolic content in the outer and inner layers of five purple sweet potato cultivars by ultra-high-performance liquid chromatography-positive electrospray-triple quadrupole (UHPLC-(ESI)-QqQ) (mg/kg DW). In all cultivars, the major non-anthocyanin phenolic compound was caffeoylquinic acid isomer 2, followed by chlorogenic acid, regardless of the number of layers. In the outer layer, caffeoylquinic acid isomer 2 ranged from 30,433 to 86,062 mg/kg DW, and chlorogenic acid ranged from 12,818 to 20,650 mg/kg DW. In the inner layer, caffeoylquinic acid isomer 2 ranged from 24,029 to 56,801 mg/kg DW and chlorogenic acid from 6714 to 13,310 mg/kg DW. Chlorogenic acid content was significantly higher in the outer layer than in the inner layer in all cultivars (p < 0.05). 'Jami' contained a higher chlorogenic acid content (20,650 and 13,310 mg/kg DW for the outer and inner layer, respectively) than the other cultivars (p < 0.05), as with a previous study reporting higher chlorogenic acid concentration of 'Jami' than that of 'Danjami', 'Sinjami', and 'Yeonjami' [36]. 'Danjami' had significantly higher caffeoylquinic acid isomer 2 content (86,062 and 56,801 mg/kg DW for the outer and inner layer, respectively) than the other cultivars (p < 0.05).

Non-Anthocyanin Phenolic Composition of Purple Sweet Potato Cultivars
Hydroxycinnamic acid derivatives consisted of p-coumaric, caffeic, trans-ferulic, and cis-ferulic acids and showed significantly higher concentrations in the outer layer than in the inner layer regardless of the cultivar (p < 0.05). The most abundant hydroxycinnamic acid derivative was caffeic acid ranging from 124 to 478 mg/kg DW in the outer layer and from 44 to 70 mg/kg DW in the inner layer. The highest concentration of caffeic acid was in the outer layer of 'Danjami' (478 mg/kg DW), corresponding to a previous study [36].
The predominant quercetin derivative was quercetin-O-diglucoside, ranging from 6 to 30 mg/kg DW in the outer layer and from 6 to 27 mg/kg DW in the inner layer. 'Jami' had significantly higher quercetin-O-diglucoside content (30 and 27 mg/kg DW for outer and inner layers, respectively) than the other cultivars (p < 0.05). Quercetin 3-O-galactoside was only quantified in the outer layers of 'Sinjami' (1 mg/kg DW) but was not detected in 'Borami' and 'Danjami'. To the best of our knowledge, this is the first study to quantify quercetin derivatives in Korean purple sweet potato cultivars.

Total Phenolic Content and Antioxidant Activity
Total phenolic content (TPC) and antioxidant activity (determined by DPPH and ABTS assays) of the five cultivars are presented in Figure 2. The TPC value ranged from 1.80 to 7.37 mg GAE/g DW. DPPH radical scavenging activity ranged from 11.63 to 12.23 mg TEAC/g DW, and the ABTS radical scavenging activity ranged from 6.10 to 7.66 mg TEAC/g DW. In a previous study, TPC and DPPH values in the flesh sample of a Chinese purple sweet potato cultivar ranged from 1.83 to 13.85 mg GAE/g DW and 1.90 to 14.54 TEAC/g DW, respectively [37]. In this study, all purple sweet potato cultivars showed similar TPC and DPPH values as the previous study. Regarding varietal difference, the whole tuber of 'Jami' showed the highest DPPH and ABTS values (36.29 and 62.20 mg TEAC/g extract residue), followed by that of 'Sinjami', 'Danjami' and 'Yeonjami' [35]. This study did not compare the antioxidant activities of the outer and inner layers; therefore, the contents could not be directly compared. The outer layers showed higher TPC values than the inner layers, regardless of the cultivar (p < 0.05). The outer layer of 'Jami' had the highest TPC value (7.37 mg GAE/g DW), and the inner layer of 'Jami' and 'Sinjami' showed significantly higher TPC value than the other cultivars (p < 0.05). This study showed higher DPPH values in the outer layer of 'Sinjami', 'Jami', and 'Yeonjami' than in those of the other cultivars. DPPH values of the inner layer were the highest in 'Sinjami' (12.21 mg TEAC/g DW) and the lowest in 'Borami' (11.63 mg TEAC/g DW), similar to the anthocyanin content, whereas the highest ABTS value was in the inner layer of 'Jami' (7.66 mg TEAC/g DW) and the lowest in 'Yeonjami' (6.10 mg TEAC/g DW). The higher TPC and similar antioxidant activity (DPPH and ABTS values) in the outer layer than in the inner layer suggests the newly identified value of the outer layer as a rich source of phenolic compounds and antioxidants.  Figure 3a,b, the first two principal components (PC1 and PC2) explained 56.58% of the total variables, with values of 32.86% and 23.72% for PC1 and PC2, respectively. The five cultivars were well-separated into four clusters (cluster 1, 'Sinjami' (SO, SI); cluster 2, 'Jami' (JO, JI); cluster 3, 'Borami' (BO, BI); and cluster 4, 'Yeonjami' (YO, YI) and 'Danjami' (DO, DI). The outer (O) and inner (I) layers of each cultivar were classified under the same cluster. Most of the di-acylated anthocyanins, chlorogenic acid, trans-ferulic acid, and quercetin hexosides were correlated with 'Jami', whereas 'Sinjami' correlated with mono-acylated anthocyanins. Cluster 3 samples were highly correlated with pelargonidin-based anthocyanins, caffeic acid, and p-coumaric acid. Figure 3c presents a heatmap of identified anthocyanins and non-anthocyanin phenolic compounds in purple sweet potato samples. The data showed that the 'Sinjami' and 'Jami' cultivar contained higher cyanidin-and peonidin-based anthocyanin contents than the other cultivars. In addition, the two cultivars showed higher antioxidant activities (DPPH and ABTS). Among the five cultivars, 'Borami' was the only cultivar that exhibited high levels of pelargonidin-based anthocyanins. The 'Yeonjami' cultivar had high levels of non-acylated anthocyanins (i.e., cyanidin 3-sophoroside-5-glucoside and peonidin 3sophoroside-5-glucoside). The 'Jami' cultivar showed higher levels of most quercetin derivatives than other cultivars. and 'Jami' cultivar contained higher cyanidin-and peonidin-based anthocyanin contents than the other cultivars. In addition, the two cultivars showed higher antioxidant activities (DPPH and ABTS). Among the five cultivars, 'Borami' was the only cultivar that exhibited high levels of pelargonidin-based anthocyanins. The 'Yeonjami' cultivar had high levels of non-acylated anthocyanins (i.e., cyanidin 3-sophoroside-5-glucoside and peonidin 3sophoroside-5-glucoside). The 'Jami' cultivar showed higher levels of most quercetin derivatives than other cultivars. Figure 3d displays Pearson's correlation among anthocyanin contents, nonanthocyanin phenolic compound contents, TPC, and antioxidant activities (DPPH and ABTS). For simplification of the data presentation, the identified 20 individual anthocyanin compounds were grouped into three categories based on their aglycone (i.e., cyanidin, peonidin, and pelargonidin), and non-anthocyanin phenolic compounds were grouped into two categories (i.e., hydroxycinnamic acid and flavonol). Pearson's correlations among individual phenolic compounds, total phenolic content (TPC), and antioxidant activities (ABTS and DPPH) of the outer and inner layers are shown in Table  S2a-d. In the outer layer, peonidin content was positively correlated with antioxidant activities (DPPH and ABTS) (p < 0.05), and cyanidin contents positively correlated with only TPC (p < 0.05). Previous studies also showed a positive correlation between anthocyanin contents and antioxidant activities [38,39]. In the inner layer, cyanidin contents showed a significantly high correlation with flavonol contents (p < 0.05), with TPC positively correlated with DPPH activity (p < 0.05). Figure 3d displays Pearson's correlation among anthocyanin contents, non-anthocyanin phenolic compound contents, TPC, and antioxidant activities (DPPH and ABTS). For simplification of the data presentation, the identified 20 individual anthocyanin compounds were grouped into three categories based on their aglycone (i.e., cyanidin, peonidin, and pelargonidin), and non-anthocyanin phenolic compounds were grouped into two categories (i.e., hydroxycinnamic acid and flavonol). Pearson's correlations among individual phenolic compounds, total phenolic content (TPC), and antioxidant activities (ABTS and DPPH) of the outer and inner layers are shown in Table S2a-d. In the outer layer, peonidin content was positively correlated with antioxidant activities (DPPH and ABTS) (p < 0.05), and cyanidin contents positively correlated with only TPC (p < 0.05). Previous studies also showed a positive correlation between anthocyanin contents and antioxidant activities [38,39]. In the inner layer, cyanidin contents showed a significantly high correlation with flavonol contents (p < 0.05), with TPC positively correlated with DPPH activity (p < 0.05).
This study is the first systemic investigation of anthocyanin and non-anthocyanin phenolic compounds and antioxidant activities in various Korean purple sweet potato cultivars and can provide a new health value for the outer layer of purple sweet potato.