Optimization and Comparative Study of Different Extraction Methods of Sixteen Fatty Acids of Potentilla anserina L. from Twelve Different Producing Areas of the Qinghai-Tibetan Plateau

In this study, supercritical fluid extraction (SFE), ultrasonic-assisted extraction (UAE), and microwave-assisted extraction (MAE) were applied to explore the most suitable extraction method for fatty acids of Potentilla anseris L. from 12 different producing areas of the Qinghai-Tibetan Plateau. Meanwhile, the important experimental parameters that influence the extraction process were investigated and optimized via a Box-Behnken design (BBD) for response surface methodology (RSM). Under optimal extraction conditions, 16 fatty acids of Potentilla anserina L. were analyzed via high-performance liquid chromatography (HPLC) with fluorescence detection, using 2-(4-amino)-phenyl-1-hydrogen-phenanthrene [9,10-d] imidazole as the fluorescence reagent. The results showed that the amounts of total fatty acids in sample 6 by applying SFE, UAE, and MAE were, respectively, 16.58 ± 0.14 mg/g, 18.11 ± 0.13 mg/g, and 15.09 ± 0.11 mg/g. As an environmental protection technology, SFE removed higher amounts of fatty acids than did MAE, but lower amounts of fatty acids than did UAE. In addition, the contents of the 16 fatty acids of Potentilla anserina L. from the 12 different producing areas Qinghai-Tibetan Plateau were significantly different. The differences were closely related to local altitudes and to climatic factors that corresponded to different altitudes (e.g., annual mean temperature, annual mean precipitation, annual evaporation, annual sunshine duration, annual solar radiation.). The temperature indices, photosynthetic radiation, ultraviolet radiation, soil factors, and other factors were different due to the different altitudes in the growing areas of Potentilla anserina L., which resulted in different nutrient contents.


Introduction
Potentilla anserina L. is a perennial herb plant that is dominant in alpine meadows [1,2]. As a folk food and a traditional Tibetan medicine, it was known as "ginseng fruit." It is mainly distributed in Qinghai Province, the Tibet Autonomous Region, Gansu Province (Gannan Prefecture), and Sichuan Province (Aba Prefecture and Garze Prefecture) in China [3,4]. According to existing literature, the roots of Potentilla anserina L. have high levels of polysaccharides [5], flavonoids, phenolic compounds [6,7], triterpenes, triterpene glycosides [8,9], ellagic acid glycosides [10], and amino acids [3]. In addition, the roots of this plant have a favorable nutritional value, with the characteristics of high protein, high dietary fiber, richness in fatty acids, various mineral elements, and low sodium content [11]. The abundant and diverse active substances of Potentilla anserina L. contribute to a wide variety of biological activities, including anticariogenic effects, hepatoprotective abilities [1,10], immunomodulatory effects [5], anti-inflammatory impact [12], and the ability to ameliorate acute hypobaric hypoxia-induced brain impairment [1].
Fatty acids, which are compounds composed of carbon, hydrogen, and oxygen, are the main components of neutral fat, phospholipid, and glycolipid. As important functional compounds, fatty acids play critical roles in improving insulin resistance [13,14], preventing and treating cancer [15], and treating cardiovascular diseases [16]. At present, the main methods for extracting fatty acids are supercritical fluid extraction (SFE), ultrasonic-assisted extraction (UAE), and microwave-assisted extraction (MAE) [17]. SFE, which is safe and reliable, does not require organic solvents to protect fatty acids from contamination or to improve their biological activity and purity. The UAE and MAE methods are simple and operable.
In 2008, the main contents of the free fatty acids in Potentilla anserina L. from Yushu Tibetan Autonomous Prefecture, Qinghai, obtained by distillation extraction with n-hexane, were α-linolenic acid (3490 µg/g), linolic acid (7351 µg/g), and palmitic acid (2262 µg/g) [18]. However, the content and composition of fatty acids extracted from Potentilla anseris L. by the SFE, UAE, and MAE methods have not been reported.
The purpose of this study is to identify the most suitable method for extracting fatty acids of Potentilla anseris L. from 12 different producing areas of the Qinghai-Tibetan Plateau. The pivotal extraction parameters that influenced the experimental process were investigated and optimized via a Box-Behnken design (BBD) for response surface methodology (RSM). On this basis, 16 fatty acids of Potentilla anserina L. were analyzed via high-performance liquid chromatography (HPLC) with fluorescence detection, using 2-(4amino)-phenyl-1-hydrogen-phenanthrene [9,10-d] imidazole as the fluorescence reagent.

Materials
Samples of Potentilla anserina L. from 12 different producing areas of the Qinghai-Tibetan Plateau were identified on the basis of plant appearance. The tuberous root of Potentilla anserine L. was used for the extraction of fatty acids in this study. All samples were propagated by cuttings in 2018; they were 3 years old at the time of collection. All relevant information about the Potentilla anserina L. samples, including sample codes, producing areas, coordinates, altitudes, and picking times, are displayed in Table 1.

Experimental Design and Data Analysis
The Box-Behnken design (BBD) for response surface methodology (RSM) was utilized as the observatory indicator to investigate the extraction rate of total fatty acids from Potentilla anserina L by the SFE, UAE, and MAE methods. Samples from Gannan Tibetan Autonomous Prefecture were selected as representative, and SFE, UAE and MAE were all developed in static mode. The key total fatty acids extraction parameters by SFE were optimized, as follows: X a1 , extraction temperature (values = 30, 45, and 60 • C); X a2 , extraction pressure (values = 30, 35, and 40 MPa); and X a3 , extraction time (values = 1, 2, and 3 h). The important parameters affecting the total fatty acids extraction by UAE were optimized as follows: X b1 , extraction temperature (values = 30, 40, and 50 • C); X b2 , extraction time (values = 30, 60, and 90 min); and X b3 , extraction volume (values = 50, 70, and 90 mL). The critical characteristics affecting the total fatty acids extraction by MAE were optimized as follows: X c1 , extraction power (values = 300, 400, and 500 W); X c2 , extraction time (values = 30, 60, and 90 s); and X c3 , extraction times (values = 3, 5, and 7). The dependent variable (Y) was the total peak area of the fatty acids. The experimental designs for the SFE, UAE, and MAE methods are shown in Tables 2-4.

SFE, UAE, and MAE
The SFE of the total fatty acids from Potentilla anserina L was conducted with supercritical fluid extraction equipment (Applied Separations Inc, Allentown, PA, USA) [17,20,21]. A 10 g powder of Potentilla anserina L was accurately weighed in supercritical carbon dioxide fluid extraction equipment for 2.1 h. The extraction temperature was 47 • C, the extraction pressure was 36 MPa, and the flow of carbon dioxide was 40 L/h. Then, the extract was collected.
The UAE of the total fatty acids from Potentilla anserina L was carried out with the MEC-200SAH ultrasonic extraction instrument (Wuxi Instrument Manufacturing Co., LTD, Wuxi, China) [22,23]. A 10 g powder of Potentilla anserina L was accurately weighed in a 250 mL triangular flask; then, a 73 mL methanol-methylene chloride solution (1:3, v/v) was added. The ultrasonic treatment power was 600 W, ultrasonic extraction was conducted for 67 min at 42 • C with pH 6.5, and the liquid/solid ratio was 7.3:1 mL/g.
For the MAE of the total fatty acids from Potentilla anserina L. [24,25], 10 g powder of Potentilla anserina L. was accurately weighed and inserted with 50 mL n-hexane into a triangle bottle. The microwave treatment power was 420 W, with intermittent radiation five times; each radiation time was 63 s. After each radiation, the triangle bottle was cooled to room temperature with cold water and then transferred into a microwave oven for radiation and filtered with an extraction bottle.
After SFE, UAE, and MAE, the extract of total fatty acids from Potentilla anserina L. filtrated and determined by HPLC.

Optimization of SFE, UAE, and MAE Conditions
In order to obtain the optimal extraction conditions, a series of extraction variables was designed to optimize and explore the interactions between these variables by the BBD. The experimental design and results of fatty acids were provided in Tables 2-4.
In optimizing the SFE conditions, the F-value of the model was 91.86, demonstrating that the model was significant. The "prob > F" values < 0.0500 implied that the model terms were significant, including X a1 , X a2 , X a3 , X a1 X a2 , X a1 X a3 , X a2 X a3 , X a1 2 , X a2 2 , and X a3 2 . The results indicated that extraction temperature, extraction pressure, and extraction time were the most important parameters affecting SFE efficiency, and the quadratic model could describe precisely the experimental response. The final equation (Equation (1)) for the SFE design is shown below: The three-dimensional response surfaces (Figure 1a-c) are represented on the basis of the optimal conditions, and the interaction between the variables was investigated to determine the optimization of the maximum content of the fatty acids. Figure 1a demonstrates the combined effects of extraction temperature and extraction pressure. Figure 1b highlights the combined effects of extraction temperature and extraction time. Figure 1c depicts the combined effects of extraction pressure and extraction time. Based on the overall results of the optimization study with the actual convenience of the experimental operation, the optimal SFE conditions were selected as follows: extraction temperature = 47 • C, extraction pressure = 36 MPa, and extraction time = 2.1 h. Under these optimal conditions, the peak area of the fatty acids was 3915.989.
In optimizing the UAE conditions, the F-value of the model was 88.40, demonstrating that the model was significant. The "prob > F" values < 0.0500 implied that the model terms were significant, including X b1 , X b2 , X b3 , X b1 X b2 , X b1 2 , X b2 2 , and X b3 2 . The results indicated that extraction temperature, extraction time, and extraction volume were the most important parameters affecting UAE efficiency, and the quadratic model could describe precisely the experimental response. The final equation (Equation (2)) for the UAE design is shown below: Y 2 = 4176.00 + 111.38 X b1 + 105.50 X b2 + 82.13 X b3 − 42.50 X b1 X b2 + 1.75 X b1 X b3 − 39.00 X b2 X b3 − 225.13 X b1 The three-dimensional response surfaces (Figure 1d-f) are represented on the basis of the optimal conditions, and the interaction between the variables was investigated to determine the optimization of the maximum content of the fatty acids. Figure 1d demonstrates the combined effects of extraction temperature and extraction time. Figure 1e highlighted the combined effects of extraction temperature and extraction volume. Figure 1f depicts the combined effects of extraction time and extraction volume. Based on the overall results of the optimization study with the actual convenience of the experimental operation, the optimal SFE conditions were selected as follows: extraction temperature = 42 • C, extraction time = 67 min, and extraction volume = 73 mL. Under these optimal conditions, the peak area of the fatty acids was 4206.907.
In optimizing the MAE conditions, the F-value of the model was 14.66, demonstrating that the model was significant. The "prob > F" values < 0.0500 implied that the model terms were significant, including Xc 1 , Xc 3 , Xc 1 2 , Xc 2 2 , and Xc 3 2 . The results indicated that extraction power and extraction volume were the most important parameters affecting MAE efficiency, and the quadratic model could describe precisely the experiment response. The final equation (Equation (3)) for the MAE design is shown below: Y 3 = 3607.00 + 82.25 X c1 + 52.13 X c2 + 65.63 X c3 − 28.50 X c1 X c2 − 53.00 X c1 X c3 − 23.25 X c2 X c3 − 193.38 X c1 2 − 220.6 2X c2 2 − 225.13 X c3 The three-dimensional response surfaces (Figure 1g-i) are represented on the basis of the optimal conditions, and the interaction between the variables was investigated to determine the optimization of the maximum content of the fatty acids. Figure 1g demonstrates the combined effects of extraction power and extraction time. Figure 1h highlights the combined effects of extraction power and extraction times. Figure 1f depicts the combined effects of extraction time and extraction times. Based on the overall results of the optimization study with the actual convenience of the experimental operation, the optimal MAE conditions were selected as follows: extraction power = 420 W, extraction time = 63 s, and extraction times = 5. Under these optimal conditions, the peak area of the fatty acids was 3621.258.
The three-dimensional response surfaces (Figure 1g-i) are represented on the basis of the optimal conditions, and the interaction between the variables was investigated to determine the optimization of the maximum content of the fatty acids. Figure 1g demonstrates the combined effects of extraction power and extraction time. Figure 1h highlights the combined effects of extraction power and extraction times. Figure 1f depicts the combined effects of extraction time and extraction times. Based on the overall results of the optimization study with the actual convenience of the experimental operation, the optimal MAE conditions were selected as follows: extraction power = 420 W, extraction time = 63 s, and extraction times = 5. Under these optimal conditions, the peak area of the fatty acids was 3621.258.

Validation of the Method
The optimized analysis method for the fatty acids was validated by a linear regression equation, limits of detection (LODs), limits of quantification (LOQs), and intra-day and inter-day precisions. The linearity relationships were provided by the plot of peak area versus the amounts of the sixteen fatty acids standards. As summarized in Table 5, the correlation coefficients of octanoic acid, capric acid, undecanoic acid, lauric acid, myristic acid, α-linolenic acid, linolic acid, pentadecanoic acid, palmitic acid, oleic acid, heptadecanoic acid, stearic acid, n-nonadecylic acid, arachidic acid, n-heneicosanoic acid and behenic acid were higher than 0.9960, with excellent linear responses. In addition, the LOD and LOQ ranges were from 0.14 ng/mL to 1.37 ng/mL and 1.18 ng/mL to 3.40 ng/mL, respectively, and the instrument precision of the intra-day and inter-day validations was < 2.07 and 2.19, respectively. As summarized in Table 6, the percentage recoveries ranged from 97.0% to 103.0%, calculated by the ratio of the spiked samples concentrations to the actual samples concentrations. These results clearly indicated that the optimized method from Gannan Tibetan Autonomous Prefecture by MAE (d). Peak labels are 1 for octanoic acid, 2 for capric acid, 3 for undecanoic acid, 4 for lauric acid, 5 for myristic acid, 6 for α-linolenic acid, 7 for linolic acid, 8 for pentadecanoic acid, 9 for palmitic acid, 10 for oleic acid, 11 for heptadecanoic acid, 12 for stearic acid, 13 for n-nonadecylic acid, 14 for arachidic acid, 15 for n-heneicosanoic acid, and 16 for behenic acid. PIA: 2-(4-amino)-phenyl-1-hydrogen-phenanthrene [9,10-d] imidazole.

Validation of the Method
The optimized analysis method for the fatty acids was validated by a linear regression equation, limits of detection (LODs), limits of quantification (LOQs), and intra-day and inter-day precisions. The linearity relationships were provided by the plot of peak area versus the amounts of the sixteen fatty acids standards. As summarized in Table 5, the correlation coefficients of octanoic acid, capric acid, undecanoic acid, lauric acid, myristic acid, α-linolenic acid, linolic acid, pentadecanoic acid, palmitic acid, oleic acid, heptadecanoic acid, stearic acid, n-nonadecylic acid, arachidic acid, n-heneicosanoic acid and behenic acid were higher than 0.9960, with excellent linear responses. In addition, the LOD and LOQ ranges were from 0.14 ng/mL to 1.37 ng/mL and 1.18 ng/mL to 3.40 ng/mL, respectively, and the instrument precision of the intra-day and inter-day validations was <2.07 and 2.19, respectively. As summarized in Table 6, the percentage recoveries ranged from 97.0% to 103.0%, calculated by the ratio of the spiked samples concentrations to the actual samples concentrations. These results clearly indicated that the optimized method was precise and suitable for analysis of the 16 fatty acids in Potentilla anserina L. from 12 different producing areas of the Qinghai-Tibetan Plateau.

Comparison of SFE, UAE and MAE
Considering Potentilla anserina L. from Gannan Tibetan Autonomous Prefecture as an example, the amounts of total fatty acids by SFE, UAE, and MAE were 16.58 ± 0.14 mg/g, 18.11 ± 0.13 mg/g, and 15.09 ± 0.11 mg/g, respectively. As Tables 7-9 show, the amounts of total fatty acids in samples from Yushu Tibetan Autonomous Prefecture Qinghai by SFE, UAE, and MAE were 14.08 ± 0.11 mg/g, 15.13 ± 0.11 mg/g, and 12.67 ± 0.10 mg/g, respectively. As a safety and environmental protection technology for extracting the 16 fatty acids of Potentilla anserina L. from 12 different producing areas of the Qinghai-Tibetan Plateau, SFE removed higher amounts of fatty acids than did MAE, but lower amounts of fatty acids than did UAE. Overall, there was no need to introduce organic solvents in the experimental process to protect fatty acids from pollution or to preserve their high biological activity and purity. UAE was operable and yielded the highest fatty acids with a simple test device. Compared with the amounts of fatty acids obtained by SFE and UAE, the amount of fatty acids obtained by MAE was lower, but required shorter time (the extraction times for MAE, UAE, and SFE were 63 s, 67 min, and 2.1 h, respectively).
In addition, in contrast with other studies that used different methods of free fatty acids extraction [18], the total content of the free fatty acids in Potentilla anserina L. from Yushu Tibetan Autonomous Prefecture Qinghai, by distillation extraction with n-hexane, was 14.427 mg/g. In this study, the total contents of the free fatty acids in same sample by SFE, UAE and MAE were 14.08 ± 0.11 mg/g, 15.13 ± 0.11 mg/g, and 12.67 ± 0.10 mg/g, respectively. These results show that there are certain differences in the contents of the fatty acids obtained from the same sample by distillation extraction with n-hexane, UAE, and MAE. It demonstrates that it is crucial to explore the most suitable method for extracting the fatty acids of Potentilla anseris L.   hydrogen-phenanthrene [9,10-d] imidazole as the fluorescence reagent. The results showed that the amounts of total fatty acids in sample 6 by applying SFE, UAE, and MAE, respectively, were 16.58 ± 0.14 mg/g, 18.11 ± 0.13 mg/g, and 15.09 ± 0.11 mg/g, and the amounts of total fatty acids in sample 1 by applying SFE, UAE, and MAE, respectively, were 14.08 ± 0.11 mg/g, 15.13 ± 0.11 mg/g, and 12.67 ± 0.10 mg/g. As an environmental protection technology, SFE removed higher amounts of fatty acids than did MAE, but lower amounts of fatty acids than did UAE. UAE was operable and had the highest fatty acids yield, with a simple testing device. Compared with the amounts of fatty acids obtained by SFE and UAE, the amount of fatty acids obtained by MAE was lower, but the method required shorter time. In addition, the contents of the 16 fatty acids of Potentilla anserina L. from the 12 different producing areas of the Qinghai-Tibetan Plateau were significantly different. Based on UAE, for example, the amount of total fatty acids from sample 6 was high, 18110.23 ± 128.92 µg/g, while the amount of total fatty acids from sample 11 was only 10110.06 ± 81.16 µg/g. The differences were closely related to local altitudes and to climatic factors that corresponded to different altitudes (e.g., annual mean temperature, annual mean precipitation, annual evaporation, annual sunshine duration, and annual solar radiation). Because of the different altitudes, the temperature indices, photosynthetic radiation, ultraviolet radiation, soil factors, and other factors were different in the growing area of Potentilla anserina L., resulting in different nutrient contents.