Box-Behnken experimental design for extraction optimization of cytotoxic activity from Curcuma aeruginosa rhizome

Curcuma aeruginosa is the common name of temu hitam in Indonesia, and the rhizome parts of the plant have several pharmacological activities. Generally, pharmacological activities are associated with bioactive content in the extract ofmedicinal plants. Several factors can in luence the bioactive extraction from medicinal plants such as solvent types, extraction time, extraction technique, and liquid-to-solid ratio. In this research, the extraction factors for extraction yield and cytotoxic activity of C. aeruginosa rhizomewere optimized using the Box-Behnken experimental design. Effect of ethanol concentration, the ratio of liquid to solid, and extraction time for the maceration process was studied. The cytotoxic activity was determined by the brine shrimp lethality test. The optimum value that maximizes the extraction yield was 70% ethanol, 300:15 ml/g liquid to solid ratio, and 1-day extraction time. The optimum value that maximizes the cytotoxic activity was 70% ethanol, 150:15ml/g liquid to solid ratio, and 2-day extraction time. The predicted extraction yield and cytotoxic activity at these projected values are 14.78% and 78.26 mg/l, respectively. In this model, Adeq Precision (10.35 and 4.16), R-Squared (0.86 and 0.79), and F-value (7.92 and2.04) is rational to it themodel for extraction yield and cytotoxic activity from C. aeruginosa rhizome.

The extraction is the crucial step for the recovery of metabolite compounds from medicinal plants. Various factors can affect the extraction effectiveness and extract yields such as solvent types (Qomaliyah et al., 2019), extraction time (Soós et al., 2019), extraction technique (Hmidani et al., 2019), and solvent-to-solid ratio (Sajid et al., 2019). The evaluate interaction in its extraction factors is proportionately important to enhance the extract yield. Response surface methodology, called RSM, is one alternative statistical method useful for improving complex extraction procedures. Several works have successfully reported for optimizing the extraction yield using RSM in the medicinal plant (Belwal et al., 2016;Pandey et al., 2018). Therefore, the study aims to optimize the extraction parameters, i.e., solid-toliquid ratio, extraction time, and ethanol concentration in C. aeruginosa rhizome using RSM. Also, extraction yield and cytotoxicity for potency pharmacology studied at maximum conditions are also investigated.

Plant material
The dry rhizome sample of C. aeruginosa was collected from Tropical Biopharmaca Research Center, Bogor Agricultural University, Indonesia, in February 2019.

Extraction process and determination of cytotoxic activity
The powdered rhizome sample (15 g) of C. aeruginosa were macerated with varying volume (75 -300 ml) of ethanol at a varying concentration (50 -96%) in 500 ml elementary lask. These samples were mixed and macerated with a shaker at 235 rpm for different time extraction (1 -3 days). The solid-to-liquid proportion, ethanol concentration, and time extraction were based on an experimental design created with response surface methodology and Box-Behnken design using Design-Expert version 11 (State Ease, Inc.). After extraction, the solution was iltered using Whatman no. 5 ilter paper and then evaporated at 50 • C using a rotary evaporator (BUCHI, R-250, Switzerland). Based on the extracted content, the extraction yield was determined (%, w/w). The extract directly applied to investigate the cytotoxic activity. Cytotoxic activity was performed according to the brine shrimp lethality test by (Meyer et al., 1982).

Experimental design
RSM and Box-Behnken design were used to evaluate the effect of extraction factors for obtaining the greatest extraction yield and cytotoxic activity with the maceration method from C. aeruginosa rhizome. The extraction parameter includes, ethanol concentration (%), liquid-to-solid ratio (ml/g), and time extraction (day). These variables were used for the optimization of extraction yield and cytotoxicity responses. The Box-Behnken experimental design consists of 15 experimental runs with the three replications at design center points of 150:15 ml/g liquid-to-solid ratio, 70% ethanol concentration, 2-day time extraction. (Table 1) showed the Box-Behnken experimental design and response of the extraction process. Optimization of the extraction process was performed using Design-Expert version 11.0 (Stat-Ease, Inc).

Model itting
The two-factor interaction and quadratic were suggested as a model statistic for extraction yield and cytotoxicity responses, respectively (Table 2).In these models, the R 2 value was 0.8559 of extraction yield and 0.7859 of cytotoxic activity responses. This study showed an R 2 value of less than 0.9, which indicated that the low tolerability of the model (Koocheki et al., 2009). (Table 3) presents the analysis of variance (ANOVA) employed in the 2FI and quadratic models for extraction yield and cytotoxic activity responses, respectively. The F values of 7.92 and 2.04 for extraction yield and cytotoxic activity, respectively, indicated the extraction yield model was signi icant but not signi icant for cytotoxicity response. The signi icant model indicated that the extraction parameters had a considerable in luence on extraction yields but not effect cytotoxic activity (Tan et al., 2016). This term, the extraction yield is signi icantly affected by the liquid-tosolid ratio (A), time extraction (C), and interaction between liquid-to-solid ratio and time extraction (AC). The second order of liquid-to-solid ratio was signi icantly affected by the cytotoxicity response. The Adeq Precision was 10.35 and 4.16 for extraction yield and cytotoxicity response, respectively. Because the Adeq Precision more than 4.0, this result indicated that the model was rationally acceptable to navigate the design optimize (Shahinuzzaman et al., 2019). (Figure 1) presents the predicted vs. actual value for extraction yield and cyto-

Extraction yield
As presented in (Figure 2), liquid-to-solid ratio, ethanol concentration and extraction time were understood in the ranges of 75:15 -300:15 ml/g, 50 -96%, and 1 -3 days for extraction yield response. Extraction yield was ranged 6.29% to 20.63% with maximum yield on the solid-to-liquid ratio of 300:15 (ml/g), ethanol concentration of 70%, and time extraction of 2 days (Table 1). At the constant extraction time (2 days), the maximum extraction yield (15.46%) was found at the solid-toliquid ratio of 300:15 (ml/g) and ethanol concentration of 96% (Figure 2a). (Figure 2b) illustrates the in luence of liquid-to-solid ratio and extraction time on the extraction yields. The response surface 3D plot was generated with the ethanol concentration ixed at 73%. The highest extraction yield (20.63%) was obtained at a liquid-to-solid ratio of 300:15 (ml/g) at the 1-day maceration process. The variation of extraction yield with extraction time and ethanol concentration at constant liquid-to-solid ratio (187.5:15 ml/g) is presented in (Figure 2c). The highest extraction yield was approximately 12.98% at an ethanol concentration of 73% and 2-day extraction. This work shows that the extraction yield increased linearly with extraction variables of ethanol concentration and liquidto-solid ratio. The effect of liquid-to-solid ratio, ethanol concentration, and extraction times on the cytotoxic activity was shown in (Figure 3). Cytotoxic activity (LC 50 value) was ranged 59.62 mg/l to 409.61 mg/l. The highest cytotoxicity (lower value of LC 50 ) was obtained at a liquid-to-solid ratio of 150:15 (ml/g), 70% ethanol, and 2-days extraction time. Because LC 50 < 1000 mg/l, all extracts were potency showed as the anticancer agent (Meyer et al., 1982). Cytotoxic activity under different ethanol concentrations and the liquid-to-solid ratio at a constant time of 2-days is seen in (Figure 3a). The lowest LC 50 value, highest cytotoxicity (113.65 mg/l), was obtained at a liquid-to-solid ratio of 75:15 ml/g and 96% ethanol. (Figure 3b) represents the interaction between ethanol concentration and extraction times on cytotoxic activity. The maximum cytotoxicity (59.62 mg/l) was observed at extraction situations 187.5:15 (ml/g) of liquid-to-solid ratio and 2-days extraction at a ixed of ethanol concentration (73%). (Figure 3c) depicts the cytotoxic activity with respect to ethanol concentration and extraction time at the liquid-to-solid ratio of 187.5:15 ml/g. The highest cytotoxic activity (59.62 mg/l) was recorded at 73% ethanol and 2-days extraction time.

CONCLUSIONS
RSM with Box-Behnken experimental design was effectively used to optimize the extraction factors from C. aeruginosa rhizome. The ethanol concentration, liquid-to-solid ratio, and extraction time were important parameters affecting the extraction yield and cytotoxic activity. For the extraction yield (20.63%), the maximize combination of parameters was 70% ethanol, 300:15 ml/g liquid to solid ratio, and 1-day extraction time. In cytotoxic activity (59.62 mg/l), the maximize combination of parameters was 70% ethanol, 150:15 ml/g liquid-to-solid ratio and 2-day extraction time.