Quality Evaluation of the Oil of Camellia spp.

The oil of Camellia spp. has become a well-known high-quality edible oil because of its rich nutrition. It is of great significance to breed fine varieties of Camellia spp. for the sustainable growth of the Camellia spp. industry. This study mainly evaluated the quality and antioxidant capacity of the camellia seed from several sources. The fatty acid composition and main active components of 40 kinds of C. oleifera, C. vietnamensis, C. osmantha, and C. gigantocarpa seeds, and so on, from different regions, were tested using GC–MS and HPLC. The quality of different Camellia spp. germplasm resources was comprehensively evaluated using multiple indices. The unsaturated fatty acid content and the antioxidant capacity of C. vietnamensis from Hainan were higher than those of C. oleifera Abel. In addition, there were a few differences in the fatty acid compositions of Camellia spp. oil from different species. Correlation analysis confirmed that rutin, total saponin, total flavonoids, squalene, and vitamin E were strongly correlated to the antioxidant capacity of Camellia spp. In the comprehensive evaluation, the best quality and strongest antioxidant activity were found for Chengmai Dafeng (C. vietnamensis). These methods in the study were applied for the first time for the quality evaluation of the Camellia spp. species. This study provided new insights into the quality evaluation of the Camellia spp. species, thus facilitating further development of variety breeding along with quality evaluation.


Introduction
Camellia spp., also known as "Oriental olive oil" [1,2], is an evergreen shrub or small tree of the family Theaceae. Because of its high oil content, Camellia spp. is also known as one of the four woody oil plants, which also includes Olea europaea, Cocos nucifera L. and Elaeis guineensis Jacq. Camellia spp. has a long history of cultivation in its native country China, where it is mainly distributed in the Yangtze River Basin [3,4]. In addition, it is also scattered in Vietnam, Myanmar, Thailand, Malaysia, and Japan [5]. The main cultivated varieties are C. oleifera Abel., C. vietnamensis Huang, C. gigantocarpa, and C. chekiangoleosa [3]. Camellia spp. is a typical plant resource in China. It has high nutritional and medicinal value, and it has a long history of production and development [6].
Camellia spp. seeds are rich in unsaturated fatty acids, peptides, minerals, and vitamins, of which more than 80% of the constituents are unsaturated fatty acids, comprising mainly oleic acid, linoleic acid, and linolenic acid. Oleic acid, which can bate blood vessels and prevent cardiovascular and cerebrovascular diseases, accounts for more than 50% of the whole fatty acid content [7]. Furthermore, Camellia spp. oil contains squalene, vitamin E, tea saponin, tea polyphenols, sterol, and β-amyrin as well as other substances that are beneficial to the human body [8][9][10]. Squalene, used as a precursor to synthesize

Treatment of Camellia spp. Seeds
Camellia spp. seeds were dried in the oven at 50 • C, then shelled and crushed. The processed seeds were loaded into the filter paper package (Soxhlet extractor) with petroleum ether (boiling point: 60-90 • C) for extraction at 50 • C for 5 h to obtain Camellia spp. oil and to obtain a deoiled powder [23][24][25].

Fatty Acid Composition and Content Analysis
Fatty acid methyl esterification of Camellia spp. oil was carried out with reference to GB/T 17376-2008. Briefly, 40 µL of an oil sample was put into a 5 mL centrifuge tube, to which 2 mL of 1 mol·L −1 KOH/CH3OH solution and 2 mL of 5% sulfuric acid solution in a cold bath were added. The centrifuge tube was inverted, and the mixture was incubated in a water bath set to a constant temperature of 55 • C for 1 h. The tube was manually shaken every 10 min for 5 s and cooled in a cold bath. Thereafter, 2 mL of n-hexane was added to the mixture, which was then vortexed for 5 min. The sample was centrifuged for 5 min at 994× g [26]. The supernatant was filtered through a membrane solution filter (0.22 µm) and subjected to GC analysis.
The mixture was analyzed using an Agilent 7890B-7000B gas chromatography machine equipped with an Agilent122-1032G column (30 m × 0.25 mm × 0.25 µm) under the following temperature conditions: 100 • C for 1 min, followed by ramping of 6 • C·min −1 to 240 • C, and then maintenance at 240 • C for 12 min. The detector temperature was set to 270 • C. The flow rate of air was 450 mL·min −1 , the flow rate of hydrogen was 40 mL·min −1 , and the flow rate of the tail blow was 45 mL·min −1 . The injection volume was 1.0 µL. The levels of fatty acids are reported as relative proportions.  [27]. The mixture was analyzed using an Agilent 7890B-7000B gas chromatograph equipped with an HP-5MS column (30 m × 0.25 mm × 0.25 µm) under the following temperature conditions: 60 • C for 1 min, followed by ramping of 6 • C·min −1 to 270 • C, and then maintenance at 270 • C for 2 min. The transfer line and ion resource temperatures were set to 250 • C. The injected volume was 1 µL, splitless. The flow rate of pure helium (99.99%), the carrier gas, was 1.00 mL·min −1 . The mass spectrometry conditions were as follows. The ion source was EI and the temperature was 230 • C. The quadrupole temperature was 150 • C, ionization voltage was 70 eV, and the emission current was 34.60 µA. The multiplier voltage was 2000 V. Data were obtained continuously in the full-scan mode in the mass range of 50-450 (m/z). On the HPMSD chemical workstation, compounds were tentatively identified using the NIST2005 MS and WILEY275 MS libraries, and their relative contents were calculated by normalization of chromatographic peak area [28].
Camellia spp. de-oiled powder (0.1 g) was loaded into a 5 mL centrifuge tube containing 1.5 mL of 50% methanol solution, ultrasonically extracted for 30 min, and centrifuged at 2030× g for 5 min at 20 • C. The supernatant was removed and loaded into an elasticquartz capillary column and then filtered through the injection filter. The HPLC conditions were as follows: the separation column was ODS C18 (0.46 mm × 150 mm), and the detection wavelength was 280 nm. A dual pump system was used for mobile phases A and B. Mobile phase A was 0.2% ice/acetic acid, and mobile phase B was acetonitrile. The gradient program was as follows: 0-10 min, isocratic 10% B; 10-18 min, isocratic 20% B; 18-28 min, isocratic 35% B; 36-42 min, isocratic 65% B; 42-49 min, isocratic 100% B; 49-56 min, isocratic 10% B. The chromatography column temperature was 35 • C, and the flow rate was 1.0 mL·min −1 .

Determination of Total Phenolic and Total Flavonoid Content
The 0.3 g Camellia spp. degreasing powder was extracted by 4.5 mL 50% methanol extraction solution, then treated by ultrasonic wave (water bath 60 • C, power 100 W, frequency 40 kHz) for 30 min, filter, repeat the above steps for 3 times, combine the filtrate, vacuum concentrate at 45 • C, add equal volume of extraction solution, centrifuge at 1697× g for 15 min, suck the supernatant, and store at 20 • C for use [27].
The total phenol content in Camellia spp. was detected using the Folin-phenol method with gallic acid as the standard. The method of Ye [27] was used to prepare the standard curve and sample liquid as well as to determine the total phenol content.
Rutin was used as the standard material to detect the total flavonoid content, and the method of Ye [27] was used to prepare the standard curve and to determine the total flavonoid content.

Determination of Antioxidant Activity
The 1 g Camellia spp. degreasing powder was extracted by 15 mL 50% methanol extraction solution, then treated by ultrasonic wave (water bath 60 • C, power 100 W, frequency 40 kHz) for 30 min, filter, repeat the above steps for 3 times, combine the filtrate, vacuum concentrate at 45 • C, add equal volume of extraction solution, centrifuge at 1697× g for 15 min, suck the supernatant, and store at 20 • C for use [27].
The antioxidant capacity was determined using 1,1-diphenyl-2-picrylhydrazyl (DPPH) following the experimental method of Zantar et al. [29]. The diluted Trolox standard (100 µL) and sample solution were added to 3 mL of DPPH solution, and the same volume of 50% methanol was used as the blank control. The absorbance values were measured at 1 h at the wavelength of 517 nm. The average absorbance values of triplicate measurements were used for data analysis. The results were converted into the Trolox equivalent antioxidant capacity (unit of mmol·L −1 Trolox·g −1 DW), and the standard curve was plotted.

ABTS (3-ethylbenzothiazoline-6-sulfonicacid) Radical Scavenging Assay
The antioxidant capacity was determined using 2,20-azinobis-(3-ethylbenzothiazoline-6-sulfonicacid) (ABTS) following the experimental method of Ye [27]. The diluted Trolox standard (150 µL) and sample solution (50 µL) were added to 3 mL of the ABTS working solution, and the same volume of 50% methanol was used as the blank control. The mixtures were left in the dark for 1 h, and the absorbance values were measured at 734 nm. The average absorbance values of triplicate measurements were used for data analysis. The results were converted into the Trolox equivalent antioxidant capacity (unit of mmol·L −1 Trolox·g −1 DW), and the standard curve was plotted.

Ferric Reducing Antioxidant Power (FRAP) Assay
The antioxidant capacity in terms of the ferric reducing antioxidant power (FRAP) of Camellia spp. extracts was determined following the method described by Stojanovic et al. [30]. The diluted FeSO 4 standard solution (2 mL) and sample solution (40 µL) were mixed with 3 mL of the FRAP working solution, and the same volume of 50% methanol was used as the blank control. The mixtures were left in the dark for 1 h, and the absorbance values were measured at 593 nm. The average absorbance values of triplicate measurements were used for data analysis. The results were converted into the Trolox equivalent antioxidant capacity (unit of mmol·L −1 FeSO 4 ·g −1 DW), and the standard curve was drawn.

Data Statistical Analysis
All analyses were conducted in triplicate, with results expressed as average values ± standard deviation (AVG ± SD). Correlation analysis and the main component of Camellia spp. were analyzed using the SPSS software. Camellia spp. from Hainan Island and inland China were evaluated by assessing the quality and antioxidant capacity.

The Comparative Analysis of the Oil Content
As shown in Table 2, the study found that most of the oil content (w/w) of the samples was concentrated between 40% and 55%, and the average oil content was 46.87%. CMDF (C. vietnamensis) had the highest oil content (58.96%), and BT (C. sinensis) had the lowest oil content (30.22%). The oil content of the seven C. oleifera Abel samples was between 40% and 45%, while 10 C. vietnamensis Huang had more than 50%, namely QZ1 (C. vietnamensis), QZ4 (C. vietnamensis), QZ8 (C. vietnamensis), NC1 (C. vietnamensis), ND3 (C. vietnamensis), HS1 (C. vietnamensis), HS2 (C. vietnamensis), FS2 (C. vietnamensis), CMDF (C. vietnamensis), and BWL (C. vietnamensis). It showed that this C. vietnamensis had good characteristics in terms of oil content. In addition, the oil content of the three samples of ND3 (C. vietnamensis), FS2 (C. vietnamensis), and CMDF (C. vietnamensis) are higher than 55%, which can be used for further research and analysis. There are great differences in the oil content of dried Camellia spp. seeds in different regions. Each region should choose the varieties suitable for the region according to their own conditions, such as soil, temperature, cultivation methods, etc., to breed good varieties.

Standard Samples
The fatty acid composition of the standard samples was determined using a 7890N gas chromatograph. As shown in Figure S1, the fatty acid component was based on the retention time of the peak and their relative contents were calculated by normalization of the chromatographic peak area.

Camellia spp. Oil Samples
In all studied samples, the main fatty acids identified were oleic acid, linoleic acid, palmitic acid, and stearic acid. These four fatty acids accounted for more than 98% of the fatty acid content. Oleic acid was the primary unsaturated fatty acid of Camellia spp. oil, accounting for 86.23% (QZ2, C. vietnamensis) to 1.23% (BT, C. sinensis) of the total fatty acid content. By analyzing Table 3, we determined that the unsaturated fatty acids (UFA) contained in Camellia spp. seed oil included oleic acid, linoleic acid, linolenic acid, and palmitoleic acid, and the saturated fatty acids included myristic acid, arachidic acid, palmitic acid, and stearic acid.

Correlation between the Four Major Fatty Acids
Pearson correlation analysis was performed on four fatty acids constituting more than 98% of the total fatty acid content using the SPSS software, and the result is shown in Table 4. As shown in Table 4, the oleic acid content had a significant negative correlation with the linoleic acid and palmitic acid contents. However, the linoleic acid content had a significant positive correlation with the palmitic acid content and a significant negative correlation with the stearic acid content. It was concluded that oleic acid content in camellia oil could affect linoleic acid and palmitic acid contents in camellia oil, and linoleic acid content could affect stearic acid and palmitic acid contents.

The Polar Compounds from the Extraction Meal Determined Using HPLC
The peak areas of HPLC chromatograms were used to determine the contents of the active components in the standards, as shown in Figure S2. The total saponin content included tea saponins A and B. The standard curve for each standard was made according to the peak area and known concentration of the standard, as shown in Table S2.
Concentrations (C; mg·L −1 ) of the active components were obtained through peak area normalization, and the content of each active component was calculated from these concentrations, as shown in Figure 1B. Content (%) = (C × 1.5 mL/100 mg) × 100%.
As shown in Figure 1B, the content of tea saponin in the test substance was the highest, and the highest average value was 208.50 mg·g −1 . QZ3 (C. vietnamensis) had the highest tea saponin content (374.86 mg·g −1 ), and FS2 (C. vietnamensis) had the lowest tea saponin content. There were common peaks in the chromatograms of 40 samples, although slightly different peaks were observed for some special samples. For example, the HPLC chromatogram of BB (C. gigantocarpa) differed greatly from those of the other samples, and the target active component was not detected for BB (C. gigantocarpa). Therefore, BB (C. gigantocarpa) was considered to belong to a special species, and its specific composition was unknown.

Total Phenol and Flavonoid Contents
The contents of total phenols and flavonoids in 40 Camellia spp. samples were determined using a 752N UV spectrophotometer. The linear fit of the standard curve using gallic acid as the standard was y = 7.7093x (R2 = 0.9983), while the linear fit of the standard curve using rutin as the standard was y = 11.451x (R2 = 0.9946). The measured absorption values and the standard curve were used to calculate the total phenol and total flavonoid concentrations (g·L −1 ) in the sample, which were used to calculate the total phenol and total flavonoid contents, as shown in Figure 2A. As shown in Figure 2A, among the 40 kinds of Camellia spp. tested, the total phenol content was between 6.23% and 10.87%, and the average total phenol content was 8.51%. In particular, the highest total phenol content (10.87%) was found for CR3 (C. oleifera), and the lowest (6.23%) for GZTR (C. oleifera). The total was between 2.84% and 8.68%, and the average was 5.12%. The highest flavonoid content (8.68%) was found for CMDF (C. vietnamensis), and the lowest (2.84%) for XY (C. oleifera).

Antioxidant Capacity
The free radical scavenging capacities of Camellia spp. seeds were determined using three methods (DPPH, ABTS, and FRAP). The standard linear equations are shown in Table  S2. The measured light absorption values and the linear equations were used to obtain the antioxidant values of different varieties of Camellia spp. The results are shown in Figure 2B.
As shown in Figure 2B, the range of antioxidant capacities from the DPPH test was relatively large, between 0.20 and 1.53 mmol·L −1 Trolox·g −1 DW, with an average of 1.35 mmol·L −1 Trolox·g −1 DW. The highest antioxidant capacities were found for CMDF (C. vietnamensis), FS2 (C. vietnamensis), BT (C. sinensis), and GZ (C. gauchowensis), and the lowest for XY (C. oleifera). From the ABTS assay, the antioxidant values were between 0.27 and 1.64 mmol·L −1 Trolox·g −1 DW, with an average of 1.30 mmol·L −1 Trolox·g −1 DW. QZ1 (C. vietnamensis) had the highest antioxidant capacity, followed by BT (C. sinensis), and BB (C. gigantocarpa) had the lowest antioxidant capacity. Overall, obvious differences were not observed in the ABTS clearance ability among the varieties from different areas. Using the FRAP method, the antioxidant value of Camellia spp. varieties from 40 different producing areas was between 1.41 and 3.90 mmol·L −1 Trolox·g −1 DW. CMDF (C. vietnamensis) had the highest value, followed by GZ (C. gauchowensis), which was relatively high, and XY (C. oleifera) had the lowest value. According to these measurements, more pronounced differences in the antioxidant capacity among different cultivars were observed with the FRAP method than with the ABTS and DPPH methods.

Correlations between Bioactive Components and Free Radical Scavenging Capacity
The correlation between active components and antioxidant capacity was analyzed using SPSS software (26.0, SPSS Inc., MD-NC119 Armonk, NY, USA). As shown in Figure 2C, rutin had a significant negative correlation with DPPH clearance, ABTS clearance, and FRAP reduction. There was a significant positive correlation between total tea saponin and DPPH clearance, ABTS clearance, and FRAP reduction. Squalene and α-tocopherol showed a significant positive correlation with FRAP reduction. DPPH clearance, ABTS clearance, and FRAP reduction all showed a significant positive correlation with each other. Among them, the total flavonoids were highly correlated with DPPH and ABTS clearances. Therefore, the results obtained using the DPPH and ABTS methods indicated that the antioxidant capacity of Camellia spp. was dependent on the contents of active substances, such as rutin, tea saponin, and total flavonoids in Camellia spp. It was found that the DPPH clearance capacity, ABTS clearance capacity, and Fe 3+ reduction capacity all had significant positive correlations with the content of total flavonoids. For example, the total flavonoid content in XY (C. oleifera) was the lowest among the 40 Camellia spp. samples, and the corresponding DPPH and ABTS clearance abilities and Fe 3+ reduction ability were relatively weak, indicating low antioxidant activity.

Cluster Analysis
Data on the fatty acid content of each sample were standardized and used for cluster analysis and PCA analysis ( Figure 3A,B). A clear clustering tendency was observed for Camellia spp. samples containing similar profiles. C. vietnamensis and C. oleifera Abel. were clustered together, and thus regional divisions were more obvious. There were exceptions, for example, JX (C. oleifera), GX (C. oleifera), and GZTR (C. oleifera) were clustered with C. vietnamensis, which likely originated in mis-sampling. K13 (C. vietnamensis) was clustered with C. oleifera Abel., indicating that environmental conditions in mainland China were different from Hainan, thus affecting the fatty acid content. Data on the contents of bioactive components for each sample were standardized and used for cluster analysis and PCA analysis ( Figure 3C,D). The clustering results were similar to those of fatty acids.

Principal Component Analysis
The 15 component indicators of the 40 Camellia spp. samples were standardized for principal component analysis. As shown in Table S3, the characteristic value of the first principal component was 4.852, the variance was 32.345%, and the cumulative variance was 32.345%. The most important principal component was the first principal component, and the importance decreased from the second to the fifth principal component. The cumulative variance contribution rate of the first five principal components reached 77.262%, and thus the five main components reflected most of the information for all 15 components of Camellia spp. Therefore, these five principal components were selected as the comprehensive evaluation index.
The principal component load matrix reflected the extent to which each quality indicator impacted this principal component, as shown in Table S4. Using the 0.5 principle, the first main component included α-tocopherol, total tea saponins, rutin, Camellianin A, total flavonoids, ABTS, DPPH, and FRAP. The second major component included squalene, α-tocopherol, β-sitosterol, and β-amyrin. The third major component included total phenols and total flavonoids. The fourth major component included quercetin. The fifth main component included α-tocopherol, total tea saponins, rutin, Camellianin A, total flavonoids, ABTS, DPPH, and FRAP.
The function expressions of five main components were calculated from the initial factor load matrix and the characteristic value of each main component. Using the function expression of each main component, the score values and rankings of the main components of the 40 Camellia spp. samples were calculated, and then the comprehensive score and comprehensive ranking of the nutritional quality of the 40 Camellia spp. samples were calculated from the main component comprehensive score model (Table 5). As shown in Table 5, among the 40 kinds of Camellia spp., the three samples with high comprehensive scores were CMDF (C. vietnamensis), GZ (C. gauchowensis), and QZ1 (C. vietnamensis), and the samples with low rankings were WH (C. chekiangoleosa) and XY (C. oleifera).

Discussion
According to the GC analysis results, four of the five samples with high unsaturated fatty acids contents were C. vietnamensis, and thus we considered that the unsaturated fatty acid content of C. vietnamensis was higher than that of C. oleifera Abel. The unsaturated fatty acid content of QZJ (C. vietnamensis) was higher than 90%, and the squalene and α-tocopherol contents of CMDF (C. vietnamensis) and QZ1 (C. vietnamensis) were high. Therefore, these samples are useful for further studies. Docosa-13-enoic acid is a kind of super long chain fatty acid [31]. The studies showed that docosa-13-enoic acid may affect the digestion of rapeseed oil in humans, cause myocardial damage, make the cholesterol level of adrenal tissue rise and cause fat accumulation in the heart tissue [32]. Long-term consumption of rapeseed oil with high levels of docosa-13-enoic acid can increase the risk of cardiovascular diseases [33,34]. In summary, excluding docosa-13-enoic acid, which is damaging to the heart muscle, Camellia seed oil is considered highly nutritious because its unsaturated fatty acid content reaches 88%, a value far higher than those of vegetable oil, peanut oil, and bean oil. This study found that the fatty acid composition and content in varieties from different regions were different, which is consistent with previous results [20,35,36]. This is likely related to the effect of different environmental conditions, especially climatic conditions, such as temperature and humidity, on the composition of fatty acids [37][38][39]. In addition, other factors, such as harvest maturity and processing methods, can affect the composition of fatty acids [40]. This study found that the composition and content of fatty acids in samples from Hainan were significantly different from those in the samples of other producing regions, which is likely attributable to the unique geographical environment, including the climate, of Hainan [6,36].
The human body contains a large number of free radicals, including reactive oxygen species, of which excessive amounts are implicated in cardiovascular and cerebrovascular diseases, diabetes, tumors, and other diseases. Camellia spp. contains active substances that can remove free radicals, which can delay aging to a certain extent and protect the skin from the adverse effects of ultraviolet radiation. Therefore, it is a good, natural raw material for makeup and skincare products. The antioxidant activities among these 40 Camellia spp. species differed greatly, even for Camellia spp. cultivated in the same region. Therefore, differences in cultivation and management, including the use of light, water, and fertilizer, also have significant effects on the antioxidant activity of the oil. In this study, CMDF (C. vietnamensis) had the strongest DPPH clearance ability, while QZ1 (C. vietnamensis) had the strongest ABTS clearance ability, and CMDF (C. vietnamensis) had the strongest Fe 3+ reduction ability. Through comprehensive SPSS and PCA analyses, CMDF (C. vietnamensis) was found to have the strongest antioxidant capacity. The clustering results indicated that C. gauchowensis and C. vietnamensis clustered together, consistent with the results of Qi et al. [41], who found that the leaves, flowers, fruits, and seeds of C. vietnamensis from Hainan are similar to those of C. gauchowensis, which is mainly distributed in Guangdong and Hainan and considered a native species. This specie is important because it is the most widely distributed variety of Camellia spp., extending into the southern regions [41]. Therefore, we considered that the genetic relationship between C. gauchowensis and C. vietnamensis is very close.
The antioxidant capacity of C. vietnamensis from Hainan was found to be higher than that of C. oleifera Abel. Therefore, it is more beneficial to produce Camellia spp. in areas with high temperatures and sufficient light, which are the climatic conditions in Hainan. Moreover, the accumulation of secondary metabolites and the presence of other substances with antioxidant activity, such as flavonoids, in C. vietnamensis from Hainan may lead to an antioxidant activity that is higher than that of C. oleifera Abel. Additionally, although K13 belongs to C. vietnamensis, its antioxidant activity was relatively low. The inland climatic conditions under which K13 (C. vietnamensis) grows may result in the large difference in antioxidant activity between K13 and C. vietnamensis from Hainan. Differences in the antioxidant activity of Camellia spp. varieties from Hainan were not obvious, but may cause the small difference between total phenols and total flavonoids, although the contents of other active substances may also differ; further analysis using high-performance liquid-phase methods may be required.

Conclusions
The present study showed that the content of unsaturated fatty acids and the antioxidant capacity of C. vietnamensis from Hainan were generally higher than those of C. oleifera Abel. Additionally, this study also found that there were significant differences in the fatty acid compositions of Camellia spp. oil from different species. CMDF (C. vietnamensis) had the best quality in the comprehensive evaluation, and its antioxidant capacity was the strongest. Correlation analysis confirmed that rutin, total saponin, total flavonoids, squalene, and α-tocopherol were strongly correlated to the antioxidant capacity of Camellia spp. Camellia spp. oil is a high-quality vegetable oil and has broad application prospects. However, basic research and application of Camellia spp. are not comprehensive and systematic, and more functional properties of Camellia spp. have yet to be studied and developed. This study provides a theoretical basis for the breeding of improved varieties of Camellia spp. and the development and application of functional components.