This article is mentioned in:

Introduction

Free radicals are compounds with unpaired electrons, capable of causing diseases such as cancer by searching for reactive electron pairs. The harmful effect of free radicals can be inhibited by antioxidant compounds derived from plants, which are usually phenolic compounds and their derivatives (l).

The conventional method used to extract plant crude drug in traditional medicine is maceration for 15 min, followed by pressing. Maceration is the process of soaking plant crude drug in extraction solvent at room temperature. Extraction efficiency is often affected by several factors, including solvent concentration, duration, temperature and the ratio of ingredients to solvent (2). Soursop (Annona muricata L.) is a plant that thrives in tropical regions. Soursop leaves are utilized as natural medicine due to their various health benefits and as a source of antioxidants that inhibit peroxide formation. Soursop can serve as an anticancer, antidiarrheal, and antidiabetic agent (3). The present study aimed to optimize extraction conditions (length of extraction, crude drug:solvent and solvent concentration) of soursop leaves, which are used in traditional medicine. These leaves contain flavonoids and tannins (3), which contribute to antioxidant activity.

The present study aimed to determine the optimal process variables by response surface methodology (RSM). RSM is a statistical technique that is often used to optimize and develop processes influenced by multiple factors (4,5). The analysis was conducted to achieve maximum levels of 2,2-diphenyl 1-picrylhydrazyl (DPPH), cupric ion reducing antioxidant capacity (CUPRAC) and ferric reducing antioxidant power (FRAP), as well as total phenolic content (TPC) and total flavonoid content (TFC), along with identification and determination of flavonoid compounds in the optimized extract.

Materials and methods

Reagents

Fresh leaves of Annona muricata L. were collected from Cimahi City, West Java, Indonesia. Reagents used included ethanol, distilled water, chloroform, toluene, magnesium powder, amyl alcohol, hydrochloric acid, gelatin, Stiasny's reagent (30% formaldehyde:hydrochloric acid, 2:1), iron (III) chloride, sodium acetate, sodium hydroxide, ether, Liebermann's-Burchard reagent (acetic anhydride:concentrated sulfuric acid, 2:1), ammonia, Dragendorff's reagent [bismuth (III) subnitrate, concentrated nitric acid and potassium iodide], Mayer's reagent [mercury (II) chloride and potassium chloride], sodium carbonate, Folin-Ciocalteu reagent, gallic acid, methanol analytical grade quercetin, methanol, aluminum chloride, 2.4.6-tripyridyl-s-triazine (TPTZ), 2.2-diphenyl 1-picrylhydrazyl (DPPH), ascorbic acid and neocuproine. Instruments included water bath, microscope (Olympus Corporation), UV lamp (Camag), UV-visible (vis) spectrophotometer (Thermo Fisher Scientific, Inc.; Trace 1300) and high-performance liquid chromatography (HPLC) device (Shimadzu Corporation).

Preparation of crude drug

Soursop leaves were washed, air-dried, cut into pieces, dried at 40-50˚C in an oven, then ground to crude drug powder, followed by storage in a dry closed container.

Phytochemical screening

Phytochemical screening was conducted to identify the presence of secondary metabolites such as flavonoids, phenols, tannins, saponins, alkaloids, coumarins, steroids/triterpenoids, and quinones. Previously, a stock solution was prepared by boiling 5 g of the soursop leaf crude drug with 100 ml of water for 15 min, followed by filtration. The filtrate was used for secondary metabolite analysis.

Flavonoid determination

For flavonoid determination, a 5 ml extract stock solution was prepared and placed in a test tube. Then, 100 mg Mg, 1 ml concentrated HCl, and 5 ml amyl alcohol were added to the test tube, which was gently shaken. The test tube was left for some time, and color changes were observed. Positive results for flavonoids were indicated by color changes to red, yellow, and orange in the amyl alcohol layer (6).

Tannin determination

To determine tannins, a 15 ml stock solution was divided into 3 test tubes (each containing 5 ml stock solution). A total of 2-3 drops of 5% FeCl3 were added to the first test tube, and color changes were observed. Positive results for tannins were indicated by color changes to green, blue, and black. In the second test tube, a few drops of 10% gelatin were added, and the formation of a white precipitate indicated a positive result for tannins. In the third test tube, Stiasny's reagent was added, heated in a water bath, and the formation of a pink precipitate indicated positive results for catechol tannins. The test results using Stiasny's reagent were filtered, and the filtrate was taken. The filtrate was then added with 1 M sodium acetate and 5% FeCl3, and a color change to blue indicated positive results for tannin gallate. Stiasny's reagent was prepared by mixing 10 ml 37% formaldehyde with 10 ml concentrated HCl (6).

Quinones determination

For quinone determination, a 5 ml stock solution was prepared and placed in a test tube. A few drops of 1 N NaOH were added, and a color change to red indicated positive results for quinones (6).

Saponin determination

To examine saponins, a 10 ml stock solution was shaken in a test tube until foam 1-10 cm high was formed. If the foam remained stable for 10 min, 1 drop of 2N HCl was added. Stable foam after adding 2N HCl indicated positive results for saponins (6).

Steroid/triterpenoid examination

For steroid/triterpenoid examination, 1 g of crude drug was mixed with 20 ml of ether, ground in a mortar, filtered, and evaporated to a residue. Liebermann-Burchard reagent was added to the residue, and color changes to green-blue indicated positive results for steroids, while red-purple indicated positive results for triterpenoids (6).

Alkaloid determination

To determine alkaloids, 2 g of crude drug was mixed with 10 ml of 2N HCl, filtered, and 5 ml of 25% ammonia was added to the filtrate. Liquid-liquid extraction with chloroform was performed, and the resulting chloroform layer was tested with Dragendorff's reagent. Orange color changes indicated positive results for alkaloids. Further liquid-liquid extraction with 2N HCl was conducted, and Mayer's reagent was used to test for alkaloids, with orange color changes indicating positive results and the formation of a white precipitate indicating positive results for alkaloids (6).

Coumarin determination

For coumarin determination, NaOH filter paper was prepared by soaking filter paper in 1 N NaOH and drying it. Then, 0.5 g of crude drug sample was weighed, and 5 ml of chloroform was added for extraction.

Extraction design with Box-Behnken model

Extraction parameters were designed using the RSM method with Box-Behnken model in Minitab 21 (dti.itb.ac.id/minitab/). The parameters included duration of extraction (10-40 min), crude drug:solvent ratio (1:3-1:10) and solvent concentration (70-96%). A 10 min extraction time was used as the lowest level because the natural product extract industry in Indonesia uses a maceration process for 15 min, followed by pressing. The use of a time <15 min allowed for a margin from the standard procedure to explore optimization possibilities. Meanwhile, the highest level was set at 40 min; industrial extraction times are <1 h due to inefficiency and increased costs. The crude drug:solvent ratio was 1:3 for the lowest level to ensure proper immersion, with the solvent level 5 cm above crude drug. For the highest level, a 1:10 ratio was used according to extraction standard (7). Additionally, 70 and 96% ethanol solvent concentrations were used at the lowest and highest levels, respectively. These concentrations are commonly applied in the natural product extract industry as universal solvents. Extracts were prepared as in Table I.

Table I Soursop leaf extraction parameters using Box-Behnken model.

Table I

Soursop leaf extraction parameters using Box-Behnken model.

Extract no.	Extraction duration, min	Crude drug: solvent ratio	Solvent concentration, %
1	10	1:3.0	83
2	40	1:3.0	83
3	10	1:10.0	83
4	40	1:10.0	83
5	10	1:6.5	70
6	40	1:6.5	70
7	10	1:6.5	96
8	40	1:6.5	96
9	25	1:3.0	70
10	25	1:10.0	70
11	25	1:3.0	96
12	25	1:10.0	96
13	25	1:6.5	83
14	25	1:6.5	83
15	25	1:6.5	83

Extraction and extract characterization

A total of 10 g soursop leaf crude drug was weighed and extracted using maceration, followed by pressing. The solvent used for extraction was ethanol and the extract was concentrated using a water bath. This was followed by determination of % yield of extract and the specific gravity. A 1% extract was prepared in 10 ml of each extraction solvent. The dissolved part and the insoluble part were transferred using a pipette into a 5 ml pycnometer. The weights of the empty pycnometer and the pycnometer containing the extract were measured. The specific gravity of water was assumed to be 1 g/ml. The formula used to calculate the extract specific gravity is: [(pycnometer weight + extract)-empty pycnometer weight]/pycnometer volume.

TPC

TPC was analyzed by adding 10% Folin-Ciocalteu to a series of standard and sample solutions. Gallic acid was used as the standard at a concentration range of 60-130 µg/ml to create a calibration curve. The 1,000 µg/ml gallic acid stock solution was diluted with methanol for analysis. Subsequently, 50 µl each concentration of gallic acid was added, followed by 500 µl 10% Folin-Ciocalteu reagent and 400 µl 1 M sodium carbonate. The mixture was incubated at room temperature for 30 min. The absorbance of soursop leaf extract was determined using a UV-Vis spectrophotometer at 765 nm. Absorption readings of standard solution were taken three times for each concentration and a calibration curve was created. Absorption measurement of each extract solution was performed six times and TPC was determined according to the gallic acid calibration curve as gallic acid equivalent (GAE) (8).

TFC

Analysis of TFC was carried out using a sample and a series of standard solutions. Quercetin (45-100 µg/ml) was used as standard: Stock solution (1,000 µg/ml) was diluted with methanol to generate a calibration curve. A total of ~100 µl each quercetin concentration was combined with 300 µl methanol analytical grade, 20 µl 1 M sodium acetate 20 µl 10% AlCl3 and 560 µl distilled water. After incubating the mixture for 30 min at room temperature. Absorbance of soursop leaf extract was determined using a UV-Vis spectrophotometer at a wavelength of 415 nm. Measurement was performed three times for each concentration to obtain a calibration curve. This was followed by the determination of the TFC of each sample. Leaf extract was dissolved in methanol pro analysis and subjected to the same procedure as the quercetin standard solution. The absorbance measurement of the extract solution was performed six times for each extract. TFC was determined using the quercetin calibration curve and denoted as quercetin equivalent (QE) (9).

Antioxidant activity. DPPH method

DPPH method was performed as outlined by Celep et al (10), with minor adjustments. Ascorbic acid (3-8 µg/ml)was used as a standard solution. Absorbance of 50 µg/ml DPPH solution was determined using a UV-vis spectrophotometer at a wavelength of 517 nm to obtain A0 value. The stock solution of ascorbic acid standard (200 µg/ml) was prepared by dissolving 20 mg of ascorbic acid in 100 ml methanol analytical grade. A total of 125 µl each concentration was mixed with 750 µl DPPH solution and incubated at room temperature for 30 min. The absorbance of each standard solution was measured at a wavelength of 517 nm using methanol as a blank. The concentration of each standard solution was measured three times and a calibration curve was constructed based on percentage of DPPH scavenging on the y-axis vs. various amounts of ascorbic acid on the x-axis (10). This was followed by the determination of antioxidant activity of each sample. Subsequently, each soursop leaf extract was dissolved in methanol analytical grade and subjected to the same procedure. The absorbance measurement of extract solution was repeated six times for each extract. Absorbance was the percentage of DPPH scavenging by soursop leaf extract. Antioxidant activity was calculated according to the calibration curve of ascorbic acid as ascorbic acid equivalent antioxidant capacity (AEAC) (10).

CUPRAC method. Antioxidant activity analysis using the CUPRAC method was performed as described by Özyürek et al (11), with minor adjustments. Ascorbic acid was used as a reference solution (3-8 µg/ml). CUPRAC stock solution was obtained by mixing CuCl2.H2O (1,705 µg/ml) with neocuproine solution (1,562 µg/ml). CUPRAC stock (100 µg/ml) was prepared using ammonium acetate buffer (pH, 7.0). The ascorbic acid solution (200 µg/ml) was diluted using 250 µl ammonium acetate buffer. A total of 750 µl CUPRAC solution was added to the diluted ascorbic acid, which was incubated at room temperature for 30 min in the dark. The absorbance was determined using a UV-Vis spectrophotometer at 450 nm using ammonium acetate buffer as a blank and the measurements were repeated three times. A calibration curve was created based on % increase in CUPRAC (11). Each soursop leaf extract was dissolved in methanol and subjected to the same procedure as the reference solution, with six repetitions. Antioxidant activity was determined as AEAC (11).

FRAP method. The FRAP method was used to test antioxidant activity, following modified procedures of Özyürek et al (11). Ascorbic acid was utilized as a standard solution (3-8 µg/ml). FRAP stock solution was obtained by mixing FeCl3.6H2O (0.02 M) with TPTZ solution (0.01 M). Subsequently, the stock solution was diluted with acetate buffer (pH, 3.6; 0.04 M) at a 1:1:10 ratio of FeCl3.6H2O:TPTZ:sodium acetate buffer at concentration of 710 µg/ml. The ascorbic acid solution (200 µg/ml) was diluted to 3-8 µg/ml) using distilled water. FRAP solution was mixed with 500 µl diluted ascorbic solution and left to incubate at room temperature in a dark room for 30 min. The absorbance of each reference solution was determined using a UV-Vis spectrophotometer at a wavelength of 595 nm, with sodium acetate buffer used as a blank. A calibration curve was constructed based on the % increase in the FRAP capacity (11). For each extract, soursop leaves were dissolved in methanol analytical grade and subjected to the same procedure as the reference solution, with six repetitions. Antioxidant activity was assessed by calculating using the calibration curve of ascorbic acid (11).

Optimized extract analysis

Antioxidant activity, TFC, and TPC were dependent variables used in the optimization process. Regression coefficient analysis was performed to obtain a polynomial model and final equation. This was followed by the construction of contour and surface plot for each independent and dependent variable. Based on this plot, optimization of extraction time, crude drug-solvent ratio, and solvent concentration was conducted using the Minitab to identify the optimal values. A lack of fit test is performed to evaluate how well a model represents the relationship between variables in an experiment and the response variable. To assess fit to the data, compare the P-value to the significance level, typically set at 0.05. If the P-value exceeds the significance level, it is not possible to conclude that the model does not fit the data adequately. Re-extraction was carried out under optimal conditions to calculate the error between the theoretical calculations (obtained from the Minitab application) and the actual calculations.

Determination of flavonoids by HPLC

Identification and determination of flavonoid content in optimized extract were performed using HPLC. The mobile phase contained methanol and water with a linear gradient of 40-60% methanol for 5 min, followed by 70% methanol for 5 min and 40% methanol for 15 min. The stationary phase used was LiChrospher® 100 RP-C18 (5 µm; 100x4 mm). A 20 µl injection volume was used with a temperature of 30˚C and a flow rate of 1 ml/min. The UV-vis detector with a wavelength of 360 nm was applied for detection. Flavonoid content was determined using the one-point method (12). The standard solutions used as a standard included rutin, quercetin, kaempferol, apigenin, and luteolin-7-O-glucoside dissolved in methanol. The optimized extract was dissolved in methanol to generate a concentration of 10,000 µg/ml before analyzing the flavonoid content compound using HPLC.

Statistical analysis

Data were analyzed using Minitab 21 and are presented as the mean ± standard deviation of 6 independent experimental repeats. Statistical analysis was performed applying one-way ANOVA followed by Tukey's post hoc test Correlation between TPC, TFC, and antioxidant activity of different test methods was determined using Pearson's method. P<0.05 was considered to indicate a statistically significant difference.

Results

Determination of plants

Characterization by analyzing plant morphology using specimens collected from Bandungense Herbarium, School of Life Sciences and Technology, Bandung Institute of Technology confirmed that the plant was A. muricata L.

Phytochemical constituents

The results of phytochemical screening showed that soursop leaves contained flavonoids, phenols and steroids/triterpenoids (Table II).

Table II Phytochemical constituents in Annona muricata L. leaf crude drug.

Table II

Phytochemical constituents in Annona muricata L. leaf crude drug.

Component	Detection
Flavonoid	Positive
Phenol	Positive
Tannin	Negative
Quinone	Negative
Saponin	Negative
Steroid/triterpenoid	Positive
Coumarin	Negative
Alkaloid	Negative

Extraction and extract characterization

Characterization involved measuring the specific gravity, with consistent values observed across all samples, ranging from 0.806 to 0.993 g/ml (Table III). A narrow specific gravity range indicates that the specific gravity did not have an influence on the activity. Extract 4 showed the highest yield of 42.30%. Extract 11 showed the lowest yield of 4.00%.

Table III Extract yield and specific gravity of soursop leaf extract.

Table III

Extract yield and specific gravity of soursop leaf extract.

Extract no.	Yield, %	Specific gravity, g/ml
1	5.100	0.973
2	5.700	0.990
3	15.000	0.993
4	42.300	0.843
5	13.500	0.883
6	10.100	0.871
7	8.100	0.814
8	11.100	0.806
9	5.900	0.874
10	17.500	0.884
11	4.000	0.850
12	18.100	0.811
13	12.270	0.977
14	12.270	0.977
15	12.270	0.977

Antioxidant activity

Table IV shows antioxidant activity of ethanol extract of soursop leaf determined by DPPH, CUPRAC and FRAP methods. The regression linear equation of ascorbic acid with DPPH method was y=9.5356x + 8.7503 with R2=0.99. The regression linear equation of ascorbic acid with CUPRAC method was y=7.606x + 16.517 with R2=0.991. The regression linear equation of ascorbic acid with FRAP method was y=7.9599 x + 31.911 with R2=0.9907. Based on DPPH method, the extract 7 exhibited the highest antioxidant activity at 189.695±14.673 mg AEAC/g sample. Conversely, the extract 3 showed the lowest activity at 87.969±12.584 mg AEAC/g sample. Using CUPRAC and FRAP, the extract 2 demonstrated the highest antioxidant activity at 171.074±8.614 mg AEAC/g sample and 226.835±25.057 mg AEAC/g sample, respectively. Conversely, the lowest results were observed with extract 9 & 10.

Table IV Antioxidant activity of ethanol extract of soursop leaf.

Table IV

Antioxidant activity of ethanol extract of soursop leaf.

Extract no.	DPPH, mg AEAC/g	CUPRAC, mg AEAC/g	FRAP, mg AEAC/g
1	108.681±8.180	136.967±11.064	145.126±16.437
2	158.426±22.242	171.074±8.614	226.835±25.057
3	87.968±12.584	131.880±4.993	140.254±20.714
4	118.294±6.332	139.386±9.885	146.261±34.490
5	98.775±4.672	126.350±12.808	151.118±14.585
6	137.804±11.568	145.081±7.665	141.887±17.917
7	189.695±14.673	162.121±11.076	208.803±13.513
8	116.776±13.624	126.559±14.482	118.968±23.190
9	99.229±8.612	94.179±13.117	142.441±11.330
10	122.240±9.361	143.413±12.258	130.644±22.285
11	157.812±7.962	139.580±11.452	165.184±31.459
12	158.470±5.428	128.141±8.691	140.773±17.117
13	140.770±7.471	148.054±11.347	160.971±17.454
14	140.770±7.471	148.054±11.347	160.971±17.454
15	140.770±7.471	148.054±11.347	160.971±17.454

[i] AEAC, ascorbic acid equivalent antioxidant capacity.

TPC and TFC

TPC and TFC are presented in Table V. TPC was calculated using the gallic acid calibration curve. Linear regression equation of gallic acid was y=0.0067x - 0.0374, with R2=0.9956. Similarly, in TFC method, quercetin was dissolved in methanol and used as a reference solution. The linear regression equation for quercetin was y=0.0052x + 0.0022, with R2 value of 0.9945 The extract with the highest phenol and flavonoid content was extract 7, yielding 119.388±14.057 mg GAE/g sample and 91.212±4.796 mg QE/g sample. The lowest phenol value was found in extract 10, resulting in 50.851±2.570 mg GAE/g sample. The lowest total flavonoid value was obtained from extract 5, yielding 20.095±1.510 mg QE/g sample.

Table V TPC and TFC of soursop leaf extract.

Table V

TPC and TFC of soursop leaf extract.

Extract no.	TPC, mg GAE/g	TFC, mg QE/g
1	74.799±10.171	35.292±1.083
2	100.134±3.561	32.741±8.609
3	88.716±4.359	52.703±1.870
4	74.114±3.462	47.549±1.069
5	80.333±6.545	20.095±1.510
6	88.891±9.098	31.002±1.777
7	119.388±14.057	91.212±4.796
8	77.821±3.457	85.060±3.938
9	118.393±4.757	22.204±1.296
10	50.851±2.570	24.486±1.544
11	95.383±3.967	67.590±3.182
12	78.365±4.929	58.106±5.511
13	86.038±3.449	45.433±4.226
14	86.038±3.449	45.433±4.226
15	86.038±3.449	45.433±4.226

[i] AEAC, ascorbic acid equivalent antioxidant capacity; TPC, total phenolic content; TFC, total flavonoid content.

Multiple linear regression and antioxidant activity response model

Table VI presents ANOVA for assessing the fit of the optimization model and the regression coefficients of the experimental variable. The suitable model for all antioxidant activity response data is a quadratic model. This model was obtained from the analysis using the Minitab application where X1=extraction time, X2=crude drug-solvent ratio, X3=Solvent concentration, as expressed in the following equations: TPC=206 + 5.91X1 - 17.8 X2 - 3.26 X3 + 0.0095 X1 x X1 - 0.304 X2 x X2 + 0.0203 X3 x X3 - 0.190 X1 x X2 - 0.0643 X1 x X3 + 0.278 X2 x X3; TFC=33 + 0.72 X1 + 15.6 X2 - 3.18 X3 + 0.0231 X1 x X1 - 0.698 X2 x X2 + 0.0368 X3 x X3 - 0.012 X1 x X2 - 0.0219 X1 x X3 - 0.065 X2 x X3; DPPH=-206 + 15.35 X1 + 25.0 X2 + 0.2 X3 - 0.0498 X1 x X1 - 0.92 X2 x X2 + 0.0363 X3 x X3 - 0.090 X1 x X2 - 0.1433 X1 x X3 - 0.148 X2 x X3; CUPRAC=-95 + 5.64 X1 + 39.9 X2 + 17.4 X3 + 0.0233 X1 x X1 - 0.691 X2 x X2 - 0.0785 X3 x X3 - 0.127 X1 x X2 - 0.0696 X1 x X3 - 0.333 X2 x X3 and FRAP=-697 + 9.15 X1 + 13.0 X2 + 16.6 X3 + 0.0323 X1 x X1 - 0.25 X2 x X2 - 0.078 X3 x X3 - 0.348 X1 x X2 - 0.1033 X1 x X3 - 0.065 X2 x X3.

Table VI Regression coefficients of the predicted model.

Table VI

Regression coefficients of the predicted model.

Model parameter	TPC, mg GAE/g	TFC, mg QE/g	DPPH, mg AEAC/g	CUPRAC, mg AEAC/g	FRAP, mg AEAC/g
Intercept	206.000	33.000	-206.000	-95.000	-697.000
Linear
X1	5.910	0.720	15.350	5.640	9.150
X2	-17.800	15.600	25.000	39.900	13.000
X3	-3.260	-3.180	0.200	17.400	16.600
Interaction
X1 X2	-0.190	-0.012	-0.090	-0.127	-0.348
X1 X3	-0.064	-0.022	-0.143	-0.070	-0.103
X2 X3	0.278	-0.065	-0.148	-0.333	-0.065
Quadratic
X12	0.010	0.0231	-0.050	0.023	0.032
X22	-0.304	-0.698	-0.92	-0.691	-0.250
X32	0.020	0.037	0.036	-0.079	-0.078
R2	0.765	0.925	0.773	0.738	0.568
P-value lack of fit	0.281	0.208	0.634	0.536	0.160

[i] X1, extraction duration; X₂, crude drug:solvent ratio; X₃, concentration solvent; GAE, gallic acid equivalent; QE, quercetin equivalent; AEAC, ascorbic acid equivalent antioxidant capacity; TPC, total phenolic content; TFC, total flavonoid content; DPPH, 2,2-diphenyl 1-picrylhydrazyl; CUPRAC, cupric ion reducing antioxidant capacity; FRAP, ferric reducing antioxidant power.

Analysis of regression coefficient and response surface plot

Surface contour plot facilitated visualization of statistical significance of the independent variables on the dependent variables (Figs. 1 and 2). Extraction time of 10 min and an ethanol concentration of 96% yielded the highest results for total phenol, total flavonoid, DPPH, CUPRAC, and FRAP tests. The total phenol test displayed a high response at a crude-drug ratio of 1:10, whereas flavonoid, DPPH, CUPRAC, and FRAP tests showed the highest response values at a crude-drug ratio of 1:6.5.

Figure 1 Response surface plot showing the interactions of factors affecting the yield of TPC and TFC. (A) TPC and (A-a) crude drug:solvent ratio to extraction duration, (A-b) solvent concentration to extraction duration and (A-c) solvent concentration to crude drug:solvent ratio. (B) TFC and (B-a) crude drug:solvent ratio to extraction duration, (B-b) solvent concentration to extraction duration and (B-c) solvent concentration to crude drug:solvent ratio. TPC, total phenolic content; TFC, total flavonoid content.

Figure 2 Response surface plot of interactions of various factors affecting antioxidant activity. (A) DPPH and (A-a) crude drug:solvent ratio to extraction duration, (A-b) solvent concentration to extraction duration and (A-c) solvent concentration to crude drug:solvent ratio. (B) CUPRAC and B-a) crude drug:solvent ratio to extraction duration, (B-b) solvent concentration to extraction duration and (B-c) solvent concentration to crude drug:solvent ratio. (C) FRAP and (C-a) crude drug:solvent ratio to extraction duration, (C-b) solvent concentration to extraction duration and (C-c) solvent concentration to crude drug:solvent ratio. DPPH, 2,2-diphenyl 1-picrylhydrazy; CUPRAC, cupric ion reducing antioxidant capacity; FRAP, ferric reducing antioxidant power.

Optimization model validation

The contour plot overlay showed that duration of 10 min, a crude drug:solvent ratio of 1:6.5 and a solvent concentration of 96% were the optimal extraction conditions with a high desirability level of 0.817 (Fig. 3). The closer the desirability value is to 1, the better model obtained. Re-extraction was performed with optimal conditions (Table VII) to verify the accuracy of the model. Subsequently, TPC, TFC, DPPH, CUPRAC and FRAP were assessed (Table VIII). Experimental values had an error range of 3.6-10.7% compared to the predicted values.

Figure 3 Overlay contour plot of TPC, TFC, DPPH, CUPRAC and FRAP. Lines indicate the values of TPC, TFC, DPPH, CUPRAC, and FRAP responses, and the white areas indicate the regions with the highest response. (A) Extraction duration to crude drug:solvent ratio. (B) Solvent concentration to extraction duration. (C) Solvent concentration to crude drug:solvent ratio. The key indicates the antioxidant activity value. DPPH, 2,2-diphenyl 1-picrylhydrazy; CUPRAC, cupric ion reducing antioxidant capacity; FRAP, ferric reducing antioxidant power; TPC, total phenolic content; TFC, total flavonoid content.

Table VII Optimal soursop leaf extraction conditions.

Table VII

Optimal soursop leaf extraction conditions.

Variable	Value
Extraction duration, min	10.000
Crude drug:solvent ratio	1:6.500
Solvent concentration, %	96.000
Desirability	0.817

Table VIII Model validation.

Table VIII

Model validation.

Antioxidant assessment	Experimental value	Predicted value	Error, %
TPC, mg GAE/g	119.388±14.057	110.878	7.675
TFC, mg QE/g	91.212±4.796	86.989	4.854
DPPH, mg AEAC/g	189.095±15.931	179.133	5.561
CUPRAC, mg AEAC/g	162.121±11.076	156.509	3.586
FRAP, mg AEAC/g	204.679±5.164	185.088	10.585

[i] TPC, total phenolic content; GAE, gallic acid equivalent; TFC, total flavonoid content; QE, quercetin equivalent; DPPH, 2,2-diphenyl 1-picrylhydrazyl; AEAC, ascorbic acid equivalent antioxidant capacity; CUPRAC, cupric ion reducing antioxidant capacity; FRAP, ferric reducing antioxidant power.

Quantitative correlation between TFC and TPC and antioxidant activity

Correlation between TFC and TPC and antioxidant activity is shown in Table IX. The correlation between DPPH, CUPRAC and FRAP is presented in Table X, while CUPRAC and FRAP are shown in Table XI. The correlation between TFC and TPC, as well as the three antioxidant test methods, is shown in Table XII. The correlation ranges from 0.6 to 0.9, indicating a moderate to very strong correlation.

Table IX Pearson correlation coefficient between TPC and TFC of soursop leaf extract with DPPH, CUPRAC and FRAP based on the Box-Behnken model.

Table IX

Pearson correlation coefficient between TPC and TFC of soursop leaf extract with DPPH, CUPRAC and FRAP based on the Box-Behnken model.

	DPPH correlation		CUPRAC correlation		FRAP correlation
Extract no.	TPC	TFC	TPC	TFC	TPC	TFC
1	0.892a	0.825a	0.944b	0.730a	0.923b	0.753a
2	0.885a	0.855a	0.797a	0.664c	0.922b	0.791a
3	0.752a	0.766a	0.887a	0.910b	0.718a	0.740a
4	0.965b	0.830a	0.896a	0.978b	0.888a	0.856a
5	0.910b	0.938b	0.971b	0.888a	0.842a	0.962b
6	0.659c	0.719a	0.974b	0.983b	0.611c	0.665c
7	0.902b	0.951b	0.996b	0.722a	0.628c	0.692c
8	0.939b	0.840a	0.952b	0.986b	0.933b	0.772a
9	0.884a	0.780a	0.929b	0.855a	0.914b	0.949b
10	0.993b	0.941b	0.847a	0.962b	0.981b	0.911b
11	0.993b	0.806a	0.790a	0.989b	0.981b	0.870a
12	0.971b	0.828a	0.989b	0.857a	0.862a	0.900b
13	0.988a	0.942b	0.988b	0.949b	0.913a	0.850a
14	0.988a	0.942b	0.988b	0.949b	0.913a	0.850a
15	0.988a	0.942b	0.988b	0.949b	0.913a	0.850a

[i] _aStrong, _bvery strong and _cmoderate correlation. TPC, total phenolic content; TFC, total flavonoid content; DPPH, 2,2-diphenyl 1-picrylhydrazyl; CUPRAC, cupric ion reducing antioxidant capacity; FRAP, ferric reducing antioxidant power.

Table X Pearson correlation coefficient between antioxidant activity of soursop leaf extract and CUPRAC and FRAP based on the Box-Behnken model.

Table X

Pearson correlation coefficient between antioxidant activity of soursop leaf extract and CUPRAC and FRAP based on the Box-Behnken model.

	Pearson's correlation coefficient
DPPH extract no.	CUPRAC	FRAP
1	0.956a	0.985a
2	0.583b	0.883c
3	0.841c	0.939a
4	0.779c	0.899a
5	0.951a	0.972a
6	0.784c	0.864c
7	0.890c	0.715c
8	0.900a	0.973a
9	0.991a	0.758c
10	0.818c	0.979a
11	0.782c	0.983a
12	0.973a	0.919a
13	0.977a	0.965a
14	0.977a	0.965a
15	0.977a	0.965a

[i] _aVery strong, _bmoderate and _cstrong correlation. CUPRAC, cupric ion reducing antioxidant capacity; FRAP, ferric reducing antioxidant power; DPPH, 2,2-diphenyl 1-picrylhydrazyl.

Table XI Pearson correlation coefficients between antioxidant activity of soursop leaf extract using CUPRAC and ferric reducing antioxidant power based on the Box-Behnken model.

Table XI

Pearson correlation coefficients between antioxidant activity of soursop leaf extract using CUPRAC and ferric reducing antioxidant power based on the Box-Behnken model.

CUPRAC extract no.	Pearson's correlation coefficient
1	0.953a
2	0.877b
3	0.917a
4	0.866b
5	0.930a
6	0.739b
7	0.563c
8	0.848b
9	0.825b
10	0.801b
11	0.858b
12	0.868b
13	0.932a
14	0.932a
15	0.932a

[i] _aVery strong, _bstrong and _cmoderate correlation. CUPRAC, cupric ion reducing antioxidant capacity.

Table XII Pearson correlation coefficients between TPC and TFC of optimized soursop leaf extract and antioxidant activity.

Table XII

Pearson correlation coefficients between TPC and TFC of optimized soursop leaf extract and antioxidant activity.

	Pearson's correlation coefficient
Antioxidant measure	TPC	TFC	DPPH	CUPRAC
DPPH	0.902a	0.951a	NA	NA
CUPRAC	0.996a	0.722b	0.890b	NA
FRAP	0.628c	0.692c	0.715b	0.563c

[i] _aVery strong, _bstrong and _cmoderate correlation. DPPH, 2,2-diphenyl 1-picrylhydrazyl; CUPRAC, cupric ion reducing antioxidant capacity; FRAP, ferric reducing antioxidant power; TPC, total phenolic content; TFC, total flavonoid content; NA, not applicable.

Determination of flavonoids by HPLC

Fig. 4 shows the HPLC chromatogram of the ethanol extract of soursop leaves alongside the standard. The flavonoid in optimized extract showed the same retention time as the standard. Table XIII shows flavonoid levels in the optimized extract. Based on the chromatogram, only one peak in the extract chromatogram overlapped with the peak of standard chromatogram. Chromatogram extract at 5.413 min showed similarity with rutin standard (extract 2) with retention time of 5.436 min. Determination of the rutin level in the extract was carried out using one point method, with the results showing a value of 11.52±1.06 mg/g.

Figure 4 High-performance liquid chromatogram of ethanolic soursop leaf extract and reference. Solid line, extract; dashed line, standards. (1) Luteolin-7-O-glucoside, (2) rutin, (3) quercetin, (4) kaempferol and (5) apigenin.

Table XIII Flavonoid levels in optimized soursop leaf ethanol extract.

Table XIII

Flavonoid levels in optimized soursop leaf ethanol extract.

	Retention time, min		AUC
Compound	Reference	Sample	Reference	Sample	Level, mg/g
Luteolin-7-O-glucoside	5.10	-	1,508,800	-	-
Rutin	5.45	5.41	1,496,903	3,448,142±318,308	11.52±1.06
Quercetin	8.48	-	3,155,894	-	-
Kaempferol	9.55	-	2,137,028	-	-
Apigenin	9.66	-	3,794,254	-	-

[i] AUC, area under the curve; -, no peak with the same retention time as the reference compound.

Discussion

The present phytochemical screening showed that soursop leaves contained flavonoids, phenols and steroids/triterpenoids. Hasmila et al (13), phytochemical screening of soursop leaf ethanol extract showed the presence of steroids/triterpenoids, alkaloids, flavonoids, phenols, and tannins. Qorina et al (14) stated that ethanol extract from soursop leaves contains flavonoids, steroids/triterpenoids, alkaloids, glycosides and tannins. The compounds produced by plants are influenced by various factors, including weather, soil, genetic factors and environments (15).

The principle of extraction is ‘like dissolves like’, showing that the compound will dissolve in a solvent with the same polarity. Here, maceration was used to soften and destroy plant cell walls to release phytochemical compounds that could dissolve (16).

Antioxidant activity of extract may be affected by extraction duration, crude drug:solvent ratio and solvent concentration, as well as assessment method (17). Therefore, the present study used three activity antioxidant assessment methods, namely DPPH, CUPRAC and FRAP. Antioxidant activity was carried out by measuring the reduction of the DPPH radical. DPPH free radical is purple in color, and when it is reduced, it changes to yellow (18).

CUPRAC antioxidant test assesses ability of antioxidant to convert Cu2+ to Cu+. Neocuproine is a frequently used ligand in CUPRAC assays (11,12). Cu2+-neocuproine complex is reduced by antioxidant to Cu+-neocuproine, which is a chromophore with the maximum absorbance at 450 nm (19).

Antioxidant activity was assessed by measuring the reduction of the iron ion ligand complex (Fe3+) to complex iron (Fe2+) using the FRAP method in acidic conditions. This was conducted by measuring the increase in absorbance at 593 nm (19).

The antioxidant activity showed a significant correlation with AEAC: Higher AEAC was correlated with greater antioxidant activity. Based on DPPH method, extract 7 with extraction duration of 10 min, a crude drug:solvent ratio of 1:6.5, and a solvent concentration of 96% ethanol showed the highest antioxidant activity. According to Justino et al (20), soursop leaf ethanol extract has a half-maximal inhibitory concentration of 28.1±4.4 µg/ml.

CUPRAC method showed that extract 2 had the highest antioxidant activity. Orak et al (2017) showed that antioxidant activity of soursop leaf methanol extract using the CUPRAC method is 2.4 mmol Trolox/g extract, while 679 µM AE/g was obtained in another investigation (21,22).

The FRAP method was used to test antioxidant activity of different extracts. Extract 2 showed the highest antioxidant activity. Orak et al (2017) stated that the methanol extract of soursop leaf exhibits 789.9±2.4 mmol Fe2+/g extract. Meanwhile, other investigations reported that antioxidant activity assessed by FRAP for ethanol leaf extract is 783.15 µM AE/g (21,22).

Extraction solvent significantly affected antioxidant activity, assessed by DPPH, CUPRAC and FRAP methods. These methods involve different mechanisms, leading to different results. DPPH involves hydrogen transfer, while CUPRAC and FRAP mechanisms are based on electron transfer (23).

Flavonoid compounds are the most abundant components of soursop leaf extract (3). Phenol and flavonoid compounds are typically the primary sources of antioxidant activity in plants, with determination of TPC based on the modified Pourmorad method (8). This method entails a reduction-oxidation reaction between phenol and Folin-Ciocalteu reagent in basic conditions. The addition of sodium carbonate allows phenolic compounds to dissociate into phenolic ions. Folin-Ciocalteu reagent facilitates conversion of phenolic compounds into phenolic ions, initiating a reaction where oxidized phenolic reduces heteropoly acid. This consists of phosphotungstate acid and phosphomolybdate, which form a blue-colored molybdenum-tungsten complex. The measurement of molybdenum-tungsten is carried out using a UV-light spectrophotometer with 765 nm wavelength (24).

TFC was measured using a modified version of the method outlined by Chang et al (9). In this method, the absorbance of flavonoid groups on samples was measured using a UV-light spectrophotometer with 415 nm wavelength based on colorimetry. The color formation reaction occurs due to aluminum chloride forming a complex with the hydroxyl group at the C-3 and C-5 atoms, as well as the carboxyl group at the C-4 atom, C-3' and C-4' (9).

In the TPC determination method, gallic acid was dissolved in methanol and used as a reference solution. The largest phenolic content was obtained in extract 7. Orak (21) showed that TPC value for methanol soursop leaf extract is 244.61±7.00 mg catechin Equivalent/g). Additionally, Nguyen et al (25) reported that TPC of soursop leaf ethanol extract is 609.08±5.82 µg GAE/mg extract.

Flavonoid content for the ethanol extract of soursop leaves showed that the highest value was obtained in extract 7. Orak et al (21) similarly obtained a flavonoid content of 81.32±3.45 mg QE/g.

TPC and TFC were used as response variables in the regression analysis of the experimental design using the Box-Behnken model. Flavonoids are classified as phenol when there were hydroxyl groups on rings A and B in the structure. However, flavonoid compounds with high methylation are not considered phenols. A high content of flavonoid in extract does not guarantee antioxidant activity (8). Flavonoid shows high activity when hydroxyl groups are at C3' and C4' on ring B, a hydroxyl group at C3, a ketone group at C4 and double bonds at C2 and C3(26). DPPH test relies on the scavenging of DPPH by antioxidant, which can donate hydrogen or contain OH in their structure. In the CUPRAC and FRAP methods, antioxidant converts Cu2+-neocuproine to Cu+-neocuproine or Fe3+-TPTZ to Fe2+-TPTZ. The E0 values for Cu2+/Cu+-neocuproine and Fe3/Fe2+-TPTZ are 0.60 and 0.77 V, respectively. A compound exhibits antioxidant activity when it acts as a reducer or has a lower E0 value than E0 Cu2+/Cu+-neocuproine and Fe3+/Fe2+-TPTZ. This indicates that E0 values of antioxidant must be lower than those of Cu2+/Cu+-neocuproine (0.6V) and Fe3/Fe2+-TPTZ (0.77V) to show reducing ability (27). Moreover, there is a tendency for extract to have low flavonoid content, not meeting the criteria of flavonoids with high antioxidant capacity, resulting in high antioxidant activity value when tested using DPPH, CUPRAC, or FRAP methods (27).

Solvent concentration significantly influenced TFC and DPPH response. Model suitability was evaluated with a lack of fit test and correlation. Le Man et al (28) reported that surface response model is considered suitable when R2>0.7. Regression model activity antioxidant for TPC, TFC, DPPH and CUPRAC had R2>70%, showing a good association between dependent variable and response. However, the regression model for FRAP showed R2 value of 56.8%, suggesting that variation in response was not properly explained by dependent variable. This discrepancy may be attributed to potential interference in the absorbance measurement (29), leading to variations in response. The lack of significant difference in the fit test showed that the regression and surface response model (P-value >0.05) was able to predict antioxidant compounds in soursop leaf extract (30).

The efficiency of extraction is improved by controlling extraction duration. Based on optimization, the highest antioxidant activity was observed at extraction of 10 min. Previous research on soursop leaf using microwave-assisted extraction obtained an optimal antioxidant activity at 9.84 min (31). Initially, extraction efficiency increases with duration. When solute equilibrium is reached inside and outside the cell, increasing extraction time does not have an impact (32). The present study suggested solute equilibrium was reached at 10 min.

Another parameter used to increase extraction efficiency is the crude drug:solvent ratio. At a ratio of 1:6.5, optimal antioxidant activity was obtained, as indicated by TFC, DPPH, CUPRAC and FRAP Increasing the ratio leads to a higher extraction yield. However, a higher ratio could cause excessive solvent extraction, requiring a longer concentration time (32).

The concentration of solvent affects efficiency of extraction. Zhang et al (32) reported that solvent with polarity values similar to the solute, based on the law of solubility, generally show improved performance. Ethanol concentrations ranging from 70 to 96% are as a universal solvent due to the ability to extract polar and non-polar compounds (33). Here, an ethanol concentration of 96% produced the maximum antioxidant activity. This showed that antioxidant compounds contained in soursop leaf crude drug were more soluble in 96% ethanol (lower polarity) compared with 70% ethanol (higher polarity).

The error obtained from each response theoretically was <25%. This suggested that results were good and value predictions could describe the actual condition (34). Therefore, the result between actual and theoretical response based on optimization using RSM with Box-Behnken model could be reproduced.

Correlation analysis between TPC and TFC of soursop leaf extract regarding DPPH, CUPRAC and FRAP was performed. According to Schober et al (35), Pearson correlation coefficient of 0.90-1.00 denotes very strong, 0.70-0.89 strong, 0.40-0.69 moderate, 0.10-0.39 weak and 0.00-0.10 negligible correlation. TPC and TFC were positively correlated with antioxidant activity. TPC consistently showed a very strong or strong correlation with antioxidant activity assessed by DPPH, CUPRAC and FRAP. This suggested that phenol and flavonoid groups contributed to antioxidant activity assessed by DPPH, CUPRAC and FRAP. According to previous research, TPC and TFC of soursop leaf ethanol extract show moderate correlation with antioxidant activity of CUPRAC, with values of 0.589 and 0.646, respectively. Soursop leaf extract also has strong and very strong correlations with FRAP, at values of 0.899 and 0.900, respectively (21).

Additionally, the correlation between DPPH and CUPRAC as well as DPPH and FRAP was strong, showing a linear antioxidant activity. Although the correlation between CUPRAC and FRAP was moderate, it could still be considered linear.

Optimal conditions were extraction duration of 10 min, crude drug:solvent ratio of 1:6.5, and a solvent concentration of 96%. The optimized extract was analyzed for flavonoid compound levels using HPLC method. The most common flavonoid in soursop plants is quercetin, although leaf extract contains rutin, quercetin, and kaempferol (3). In the genus Annona, the most common flavonoids are kaempferol-3-O-galactoside, luteolin-7-O-glucoside, quercetin-3-O-rhamnoside and their derivatives (36).

Balderrama-Carmona et al (37) used acidified ethanol extract from soursop leaves and ultra-performance liquid chromatography and reported rutin levels of 6.52±0.59 mg/g. The differences in results could be attributed to variation in extraction process and methods. In conclusion, extraction process using the maceration method to obtain optimal antioxidant activity can be optimized using Response Surface Methodology. Optimization analysis using the Box-Behnken method showed that the optimum extraction of soursop leaves through maceration and pressing was achieved with an extraction time of 10 min, a crude drug-solvent ratio of 1:6.5, and a solvent concentration of 96%. This optimal extraction yielded total phenol, total flavonoid, DPPH, CUPRAC, and FRAP. In general, phenols and flavonoids contribute to the antioxidant activity of DPPH, CUPRAC and FRAP.

Acknowledgements

The authors would like to thank the School of Pharmacy at Bandung Institute of Technology (Bandung, Indonesia) for providing the facilities to perform this research.

Funding

Funding: The present study was supported by the Institute for Research and Community Service, Bandung Institute of Technology (Penelitian, Pengabdian Masyarakat, dan Inovasi 2024 grant no. 31/IT1.C10/SK-KU/2024).

Availability of data and materials

The data generated in the present study may be requested from the corresponding author.

Authors' contributions

FMR performed experiments, analyzed and interpreted data and wrote the manuscript. IF conceived and designed the study. IF, RH and HP analyzed and interpreted data and revised the manuscript. IF and RH confirm the authenticity of all the raw data. All authors have read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References