Optimization of Extraction Yield and Phytochemical Characterization of Crude Methanolic Extract and Its Fractions of Mitragyna Speciosa Leaves

In this study, Response surface methodology (RSM) was applied to optimize the yield of crude methanolic extract of Mitragyna speciosa leaves using Ultrasound-assisted extraction (UAE). The crude methanolic extract and its fractions were quantied in terms of total phenolic content and total avonoid content, along with characterized using Fourier-transform infrared and Gas chromatography–mass spectrometry. The results showed the maximum yield of 49.72% at the optimal conditions (temperature, 34 °C; time, 25 min; and volume of solvent, 166 mL). The recovery crude methanolic extract for TPC and TFC were 137.3 ± 15.7 mg GAE/g and 90.3 ± 15.3 mg RE/g, respectively.

because it can improve the quality of chemical characteristics [18] . This technique is favorable because it required a simple apparatus and procedure compared with the solid-phase extraction technique.
Mitragyna Speciosa Korth (Rubiaceae family) is a tropical plant that can be found in Southeast Asia countries. M. speciosa is commonly known in Malaysia as 'ketom' or 'biak-biak' and 'kratum' in Thailand. M. speciosa has traditionally used as an herbal remedy to relieve tiredness and muscle fatigue, and to treat some common illnesses such as diarrhea, coughing, muscle pain, diabetes, wound, and hypertension. In addition, it is also used as a substitute for opium or morphine in the treatment of drug addicts [19] . Numerous studies have revealed several biological activity of M. speciosa such as antiin ammation, antinociceptive, antioxidant, antimicrobial [20][21][22][23] that are vital in a good healing process.
Due to the great bene cial value of this plant, research is needed to optimize the extraction process to ensure high pharmaceutical quality. To the best of our knowledge, there have been no studies to optimize the extraction yield from the M. Speciosa leaves extract. The current study is to optimize the ultrasoundassisted extraction conditions of the M. speciosa leaves extract to achieve maximum extraction yield extract by using RSM. Under optimize parameters, the crude methanolic extract was fractionated by liquid-liquid extraction to obtain its fractions. The crude methanolic extract and its fractions were quanti ed in terms of total phenolic content (TPC) and total avonoid content (TPC), as well as characterized using Fourier-transform infrared (FTIR) and Gas chromatography-mass spectrometry (GC-MS).

Materials
All the chemicals and reagents used were of analytical grade. Methanol, Folin-Ciocalteu phenol reagents, AlCl 3 , potassium acetate, gallic acid and rutin were obtained from Sigma-Aldrich (Germany). Sodium carbonate (Na 2 CO 3 ) was purchased from Merck (Darmstadt, Germany). Distilled water was puri ed in our laboratory. Fresh leaves of M. speciosa were collected from Perak, Malaysia. The plant was identi ed by botanist and the voucher specimen MFI 0121/19 has been deposited in the herbarium of the Institute of Bioscience, Universiti Putra Malaysia.

Sample Preparation
Fresh leaves of M. speciosa were thoroughly washed with running tap water. The leaves were cut into small pieces and subjected to freeze-drying method by kept overnight under -80 °C after which the frozen leaves were lyophilized. The leaves were milled into a ne powder using a laboratory grinder and the ground leaves were stored in an airtight jar.
Ultrasound-Assisted Extraction About 50.0 g of nely powdered of M. speciosa was placed in a conical ask and mixed with methanol. The solution was transferred to an ultrasonic bath (Power Sonic 405, Hwashin Technology Co., Seoul, Korea). The extract was than ltered and concentrated using a rotary evaporator at approximately 40 °C until the excess solvent was completely removed. Percentage of extraction yields of each run were calculated based on total dry weight according to the Equation (1):

Experimental design
Response Surface Methodology (RSM) was used to optimize the crude extract from the extraction of M. speciosa. Based on Central Composite Design (CCD), 20 experimental runs, with three factors at ve levels including six central point were constructed for the optimization process. The CCD was generated by Design Expert Software (Version 11, Stat. Ease Inc., Minneapolis, USA). The in uence of independent variables such as the temperature (A), time (B) and volume of solvent (C) on the extraction process were evaluated towards the dependent response (extraction yield). Quadratic model was developed from RSM which describing the extraction process. Table 1 shows the independent variables with the range of its parameters.

Fractionation
Fractionation was conducted by using liquid-liquid extraction. The optimized conditions obtained from RSM were used to get the crude methanol extract. The crude methanol extract was partitioned between nhexane and water. The aqueous layer was fractionated by using different polarities based solvents such as n-hexane, dichoromethane, ethyl acetate and butanol successively. Four solvent fractions were collected and concentrated with vacuum rotary evaporator.
Quanti cation of total phenolic content (TPC) and total avonoid content (TFC) TPC By using the Folin-Ciocalteu assay, the total phenolic content of the extracts was determined using method by [24] with several modi cations. Each extract of 100 μL at a concentration of 1 mg mL -1 or gallic acid standard solution was mixed with 0.5 mL Folin-Ciocalteu reagent. The mixture was then incubated for 3 min at room temperature and afterwards 1 mL of 7.5% sodium carbonate was added. After the mixture was heated for 1 min at 95 °C and allowed to cool at room temperature, blue complex formed. Absorbance was measured at 765 nm using spectrophotometer. The total phenolic compound concentration in extract was expressed in mg of gallic acid equivalent per gram of dry weight (mg GAE/g) extract.
TFC Total avonoid content was ascertained based on the method by Sepulveda et al. [25] with some modi cation. In brief, 0.5 mL of sample was mixed with 0.1 mL of 10% AlCl 3 , 0.1 mL potassium acetate and 4.3 mL of distilled water, followed by 30 min incubation at room temperature. Absorbance was measured at 415 nm using a spectrophotometer. The total avonoid content was expressed as milligram of rutin equivalents (RE) per gram dry matter of extract.

Fourier-Transform Infrared Spectroscopy (FTIR)
Each extract were mixed with KBr salt, using mortar and pestle, and compressed into a thin pellet. The absorption spectrum of a chemical compound was determined by infrared spectroscopy. Infrared spectra were recorded on a Perkin-Elmer FTR SPECTRUM BX spectrophotometer between 4 000 -400 cm -1 .

Gas Chromatography-Mass Spectroscopy (GC-MS)
Crude methanolic extract and its fractions were subjected to GC-MS analysis using the model instrument, GCMS-QP2010 Ultra (Shimadzu Co., Japan) attached with a capillary column DB-1 (0.25 m lm × 0.25 mm I. d. × 30 m length). Analysis was performed by injecting 1 L of the sample with a split ratio of 20 : 1. Helium gas (99.9 %) was used as the carrier gas at a ow rate of 1 mL/min. The analysis was performed in the EI (electron impact) mode with 70 eV of ionization energy. The injector temperature was maintained at 250 °C (constant).The column oven temperature was set at 50 °C (held for 3 min), raised at 10 °C per min to 280 °C (held for 3 min), and nally held at 300 °C for 10 min. The compounds were identi ed after comparing the spectral con gurations obtained with that of available mass spectral database (NIST and WILEY libraries).

Results And Discussion
Optimization using Response Surface Methodology (RSM) RSM was used to optimize the experiment that was designed using Central Composite Design (CCD) via Design Expert 11 Software. RSM was not only optimized, RSM can also investigate the relationship between parameters and its response. There were 20 experiments (three factors ve levels) that have been carried out to determine the optimum condition that will produce the highest amount of yield. Table  2 shows the design matrices of the actual and predicted value for extraction of M. speciosa. The actual value for the response (extraction yield) was from 21.34 -50.73%. The predicted values were corresponding to the actual values since achieved from the model tting technique. Figure 1 illustrates the correlation between predicted and actual values of extraction yield. Most of the points were closed to the line where indicates that the optimum condition for achieving high percentage yield able to determine by the predictability of the model. The quadratic polynomial model can be expressed by the response after employing multiple regression analysis on the actual. The relationship between the extraction yield, Y with the three independent variables can be described in Equation (4).
Where A is extraction temperature (°C), B is extraction time (min), and C is volume of solvent (mL).
For the statistical analysis, the validity of the model was determined by using analysis of variance (ANOVA) as shown in Table 3. The model that was developed suggested as a quadratic model and signi cant due to the F-value and p-value were 146.46 and <0.0001 respectively. There is only 0.01% chance that the F-value can lead to the noise. F-value was obtained when an ANOVA test was run and to determine the means between the populations is signi cantly different, besides p-value also can be determined. The F-value must be used in combination with p-value in order to evaluate the result signi cantly. The value of the coe cient of determination (R 2 ) and the adjusted coe cient determination (Adj. R 2 ) were found high, which were 0.9947 and 0.9879, respectively, indicating a satisfactory correlation between the independent variables and response. The differences between both R 2 were only less than 0.2 which revealed the effectiveness of the model. A model that contains a high value of the coe cient of determination which is more than 0.9 represents the model was a good t with high correlation [26] . Adeq precision was used to measure the signal to noise ratio. In this study, adeq precision was found greater than 4, which was 47.0226 that indicated an adequate signal. Hence, the designed model can be used to plot the design space.
Apart from that, the "lack-of-t" of the model should also take into count. The lack-of-t of this model was not signi cant which veri ed the accuracy of the model. According to Quanhong & Caili [27] , they reported that the variation can be predicted accurately based on the lack-of-t of the model where the value of F-value lowers than p-value.  Figure 2, the volume of the solvent was being xed at 125 mL and interaction on the extraction yield of M. speciosa. When the extraction temperature of 40 °C and the extraction time at 15 min, the extraction yield was found to be the lowest yield (28.24%), Meanwhile when the time at the maximum of 60 min and temperature at 40 °C, the highest extraction yield (45.24%) was achieved. These indicated that as the extraction time increases with extraction temperature, a higher amount of production able to obtain due to having signi cant conditions that able to release all the metabolites from the plant. Cares et al. [28] has reported that UAE was able to disrupt the biological membranes which enhance the release of compounds from plant and also able to improve mass transfer. Furthermore, thermally unstable compounds can obtain from using UAE [29] .
Effects of temperature and volume of solvent As shown in Figure 3, the effects of temperature and volume of solvent towards the extraction yield of M. speciosa were evaluated. The response surface plot was generated with the xed extraction time at 38 min and the interaction on response. The highest yield of extraction (44.11%) was obtained when the temperature was at 30 °C and used 125 mL amount of solvent. However, when the higher temperature was applied which was 50 °C and used the same volume of solvent (125 mL), the percentage yield was 32.57%. Thus, from these ndings, it represented that the compounds in M. speciosa were heat-sensitive; hence will lowering the production yield as temperature increases. Another study by Sheng et al. [30] stated that some thermo-sensitive compounds such as avonoids will degrade as exposed to high temperatures. Meanwhile, at the temperature of 40 °C and 200 mL of solvent used, the yield was 38.73% which lower compared to a similar condition (40 °C) and has 150 mL of solvent used. In order to obtain the optimum condition for the extraction process, the desirability function was evaluated. The highest amount of yields able to achieve from the optimum condition by considering all of the independent variables. From the study, the maximum extraction yield (49.72%) was optimized at 34°C of extraction temperature, in 25 min and the volume of solvent was 166 mL. The RSE percentage of the optimum condition was below 5% and represented the model was in good concurrence. The optimum conditions for the extraction yield of M. speciosa were tabulated in Table 4. There is no work reported on % yield crude methanolic extract. However, Orio et al. [31] revealed that % crude extract in methanol: water, 1:1 was 24.8% where the conditions were at 25 °C for 1 hour.  In the present study, the TFC of crude methanolic extract was similar with the previous study done by Parthasarathy et al. [23] which was 90.3 mg RE/g and 91.1 mg RE/g respectively. However, variation between TPC in the M. speciosa with those reported by previous studies was observed. TPC obtained from crude methanolic extract was more (137.3 mg GAE /g) compared to previous studies reported by Parthasarathy et al. [23] which was 105.58 mg GAE/g while a lower TPC (24.02 mg GAE/g) was reported by Lee, S.T. et al. [32] . These differences might be due to the usage of modern extraction which was ultrasound-assisted extraction (UAE) and environment conditions where it in uences the amount of TPC [33] . A study by Zhou et al., [34] on Melastoma sanguineum shows that UAE improved the extraction e ciency and saved a lot of time compared with maceration extraction. Additionally, the present plant was collected from a different location. For most of the plants, change in one of environmental factors such as light, soil water, temperature, soil fertility, and salinity may affect the accumulation of secondary metabolites [35] . However there is no investigation has been conducted towards hexane, dichloromethane, ethyl acetate, and butanol extract of M. speciosa. Characterization of crude methanolic extract and its fractions

FTIR analysis
Variation of peaks shows in Figure 5 indicated the various functional groups are obtained in the crude methanolic extract and its fractions of M. speciosa leaves.

GC-MS analysis
The crude methanolic extract and its fractions of M. Speciosa leaves were further characterized using GC-MS. A total of 64 phytocompounds were found in the crude methanolic extract and its fraction contained more than 1% area (Table 7). These compounds belong to different chemical classes, including alcohol, esters of fatty acids, steroids/triterpenes, aldehydes, ketones, and amides, Among the crude fractions, hexane fraction have more phytocompounds same with crude methanolic extracts such as 3,7,11,15tetramethyl-2-hexadecen-1-ol, hexadecanoic acid, phytol, 9,12,15-octadecatrienoic acid, squalene, campesterol, and stigmasterol. Ethyl acetate and dichloromethane fractions have no phytocompound same with the crude methanolic extract.
The presence of various bioactive compounds detected by GC-MS such as stigmasterol, campesterol, phytol, squalene, hexadecanoic acid, benzenesulfonamide, and hydroquinone justi es the used of the M. speciosa for various skin ailments treatment. In addition, stigmasterol is one the types of sterol found in the most plant has been studied for its pharmacological potential, including cytotoxic, antimutagenic, antioxidant, antitumoral, among other herbal approaches to pathological states in principles and practice of phytotherapies [36] .   FTIR Spectum of M. speciosa crude methanolic extract and its fractions. 5a) crude methanolic extract. 5b) hexane fraction. 5c) dichloromethane fraction. 5d) ethyl acetate fraction. 5e) butanol fraction.