Isolation, separation and purification of rutin from Banana leaves (Musa balbisiana)

Industrial Crops & Products 149 (2020) 112307

Contents lists available at ScienceDirect

Industrial Crops & Products journal homepage: www.elsevier.com/locate/indcrop

Isolation, separation and purification of rutin from Banana leaves (Musa balbisiana)

T

Panida Yingyuena, Suchada Sukrongb, Muenduen Phisalaphonga,* a

Department of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Pathum Wan, Bangkok 10330, Thailand Research Unit of DNA Barcoding of Thai Medicinal Plants, Department of Pharmacognosy and Pharmaceutical Botany, Faculty of Pharmaceutical Sciences, Chulalongkorn University, Bangkok, 10330 Thailand

b

A R T I C LE I N FO

A B S T R A C T

Keywords: Rutin Banana leaves Separation Purification

In this study, rutin was detected as a main component after the phytochemical screening of flavonoids from an ethanolic extract of banana leaves (Musa balbisiana) using thin layer chromatography plates (TLC). The rutin from the crude extract was separated using a series of solvent partition separations and was purified by using Sephadex™ column chromatography and semi-preparative reverse phase-high performance liquid chromatography (RP-HPLC). The compounds were characterized and quantified using NMR, Mass Spectroscopy (MS) and HPLC. A crude ethanolic extract of 124 g, containing 5.3 % of rutin, was obtained from 1 kg of dried banana leaf powder. After a series of solvent partition separations, the product from partitions using hexane/dichloromethane and dichloromethane/water in sequence contained a rutin content of 40.6 %. Purification using a Sephadex column yielded fractions with purity of rutin up to 74–84 % and further purification using semipreparative RP-HPLC produced a final product with a yield of rutin of 32.4 mg/g of the crude extracts and a purity of 98.4 %. The results revealed that banana leaves (M. balbisiana), a food industry by-product and agricultural waste, has the potential for use as an inexpensive and new source of rutin.

1. Introduction Rutin (Fig. 1) is a flavonoid naturally present in many plants. It is also known as rutoside, quercetin-3-rutinoside and sophorin (Kreft et al., 1999). It is slightly soluble in water, but more soluble in alcohols. Rutin has several important pharmacological properties, which are beneficial to health (Habtemariam and George, 2015) and can potentially be used as an antioxidant, antimicrobial, antifungal, antiallergic and anticancer drug as well as for treating diabetes and hypertension (Shama and Patel, 2013; Frezza et al., 2018). Rutin is found in the groats, seeds, leaves, and flowers of buckwheat (Hussain et al., 2017), in the flowers and leaves of citrus plants and in other edible and inedible plants at about 2%10 % dry weight (Fabjan et al., 2003; Suzuki et al., 2005). The highest content of rutin, at 10.5 %, has been reported in the dried fruit of the inedible shrub, smoke tree (Rhus cotinus) (Atanassova and Bagdassarian, 2009). The banana is one of the most widely distributed and consumed fruit in tropical and subtropical countries (Laillyza et al., 2014). Nutritionally, it is one of the world’s leading food crops with its high contents of minerals, vitamins, carbohydrates, flavonoids, and phenolic compounds (Imam and Akter, 2011). M. balbisiana is a species of

⁎

banana, which is popular in Thailand. All parts of the banana plant can be used: the fruit and inflorescences can be used as food; and the roots and trunks can be used as herbal medicines. The trunk can be used to make fiber to weave ropes. Banana leaves also have a wide range of applications because they are large, flexible, and waterproof. So, they are used for cooking, wrapping and serving food. They are also used for decorative and symbolic purposes in Buddhist ceremonies. The leaves of M. balbisiana are an undervalued commodity with a limited commercial value, which can be considered as an agricultural industry byproduct and waste (Padam et al., 2014). Ban Klong Krachong, Sawankhalok District, Sukhothai province, is the largest area for the production of fresh banana leaves in Thailand, where banana leaves have been sent to other markets in Thailand and exported to international markets. The growth stage for the leaves harvested is about six months after planting. The banana leaf cuttings for sale are preferable to cut in 2 periods: 1) in the morning at around 8.0–10.0 AM and 2) in the evening time at 3.0–6.0 PM; because during those periods, banana leaves are quite dry and there is no latex or dew flowing out. Thus, it can reduce or prevent stains and dirt spots on the banana leaves. After that, petioles are removed from the leaves and then leaves are cleaned with water, air dried and folded into a bundle to be sent to markets. From cutting and

Corresponding author. E-mail address: [email protected] (M. Phisalaphong).

https://doi.org/10.1016/j.indcrop.2020.112307 Received 13 September 2019; Received in revised form 22 February 2020; Accepted 3 March 2020 0926-6690/ © 2020 Elsevier B.V. All rights reserved.

Industrial Crops & Products 149 (2020) 112307

P. Yingyuen, et al.

small pieces, approximately 2 cm × 2 cm, then air-dried at 50 °C for 12 h. The dried banana leaves were then ground into a powder (Fig. 3). The extraction was carried out by mixing 1 kg of banana leaf powder in 5 L of 95 % (v/v) ethanol for 10 min in a batch reactor at room temperature (≈30 °C). The maceration was prolonged for 5 day at room temperature with 10 min-daily mixing. The solution was then filtered to remove any residual powder and the ethanolic extract was concentrated using a rotary evaporator at 50 °C before drying in air at room temperature (approximately 30 °C) (Fig. 4). The conditions for drying and extraction conditions were modified from the previous studies (Chew et al., 2011; Müller and Beindl, 2006). 2.3. Solvent partition separations Fig. 1. Chemical structure of rutin.

The partition process involved partitioning in n-hexane and dichloromethane. For partition with n-hexane, the dried crude extract (15 g) was dissolved in 30 mL of 95 % (v/v) ethanol in a 250-mL flask then 100 mL of ultrapure water were added. All components were allowed to mix at room temperature (30 °C) and then transferred into a separation funnel. After adding n-hexane (200 mL), the mixture was shaken at room temperature and allowed to settle and separate into two layers. The lower layer (water layer) was collected as the extract for the second partition with dichloromethane while the upper (n-hexane) layer was discarded. For the second partition, dichloromethane (100 mL) was added to the extract then the mixture was shaken at room temperature, transferred into a separation funnel and allowed to settle and separate into two layers. The upper layer (water layer) was collected as the extracted solution while the lower layer was discarded. The extract was then concentrated using a rotary evaporator at 50 ○C (Abu et al., 2017).

preparation processes, abundance imperfect or damaged banana leaves are left over as a by-product, which can be used as a good source of high-value compounds. Because banana leaves are widely available in large quantities, they can be used as a source of raw materials for the green technology industry. Rutin has powerful antioxidant properties and many health benefits; however, rutin cannot be used efficiently because of a high price of the product. To reduce the cost of producing rutin, it is important to find a rich source, which is widely available and inexpensive. Rutin has been detected in the crude extract and fractions of Musa paradisiaca leaves (Coskun, 2016), which are a similar family to M. balbisiana. However, no further studies for separating and purifying rutin from banana leaves have been reported. Therefore, the present study aims to be the first to report a methodology for isolating, separating and purifying rutin from banana leaves (M. balbisiana) by using a series of solvent partition separations, Sephadex column chromatography and semi-preparative RP-HPLC.

2.4. Separation using a Sephadex™ column The Sephadex column is a liquid chromatography medium designed for the molecular sizing of natural products. Dextran resin Sephadex G100, Sigma-Aldrich Bangkok, Thailand with a particle size of 40,120 μm was used as the stationary phase in the glass column 2.5 cm × 150 cm. The selected fraction was isolated using as a mobile phase of 100 % methanol, which was allowed to flow down through the column by gravity. The mean volume of eluted solvent for each fraction was ≈27.5 mL. The fractions were collected at 30-min intervals for the next step of purification using semi-preparative RP-HPLC (Wu et al., 2013).

2. Materials and methods 2.1. Materials, Chemical and reagents All chemical reagents including ethanol (95 %), n-hexane, hydrochloric acid, diphenylorinic acid, 2-aminoethyl ester, ethyl acetate, methanol, water, ammonia, hydrochloric, magnesium, sulfuric acid, ferric chloride, Lead acetate and PGE 400, dichloromethane were purchased from S.M. Chemical Supplies Co. Ltd., Bangkok, Thailand (distributor of Sigma-Aldrich) and used as received. Banana leaves were purchased from the local market in Ban Klong Krachong, Sawankhalok District, Sukhothai province, Thailand (≈ 0.3 US$/Kg). The banana used in this study belongs to the genus Musa of the Musaceae, species of Musa balbisiana Colla (BB genome). The botanical identification was previously performed. (Swangpol et al., 2007).

2.5. Isolation by thin layer chromatography (TLC) In order to determine the mobile phase mixture and selected fractions for further purification by semi-preparative RP-HPLC, the fractions obtained from the Sephadex column were separated using thin layer chromatography (TLC) comprising silica gel 60 RP-18 coated with F254s fluorescent indicator, and aluminum TLC plates, 20 × 20 cm (Merck, Kenilworth, NJ, USA). The TLC analysis was performed using various solvent systems: dichloromethane, ethyl acetate, water and methanol at ratios of 4:1, 2:1 and 1:1 (v/v). The solvent system

2.2. Procedure for sample preparation and extraction Fresh banana leaves of M. balbisiana (8.6 kg) (Fig. 2) were cut into

Fig. 2. Fresh banana leaves (M. balbisiana). 2

Industrial Crops & Products 149 (2020) 112307

P. Yingyuen, et al.

Fig. 3. The stages in sample preparation: (A) fresh banana leaves from M. balbisiana cut into small pieces; (B) air-dried at 50 °C for 12 h; (C) dried banana leaves; and (D) final ground banana leaf powder.

Fig. 4. The procedures for extraction: (A) banana leaf powder extracted with 95 % (v/v) ethanol at room temperature (30 °C); (B) the ethanolic extract; and (C) the dried crude extract after filtration, concentration and air-drying.

min (Takahama et al., 2009; Lu et al., 2006).

exhibiting the best separation of compounds was determined. The plates were developed using natural product reagent (NP) and visualized under a UV lamp (365 nm). The Rf values of the separated bands were calculated. A reverse phase TLC method was used to determine the components of the extract containing rutin (Georgeta et al., 2016).

2.7. Flavonoids testing For the preliminary flavonoid tests, the crude extract was dissolved in ethanol (95 %), then filtered and subjected to the following tests:

2.6. Purification using a semi-preparative reverse phase high performance liquid chromatography column (RP- HPLC)

2.7.1. Shinoda test A few fragments of magnesium turnings were added to the extract solution then concentrated hydrochloric acid was added drop wise and the mixture heated. The appearance of a pink, orange, or red to purple color indicated the presence of flavonoids (Vimalkumar et al., 2014).

Two fractions (S4 and S5) obtained from the Sephadex column were purified using semi-preparative RP-HPLC column: 10 mm inner diameter and 250 mm length (Merck, Darmstadt, Germany). The determination was carried out at 365 nm using an ultraviolet-visible (UV–vis) diode array detector (DAD). The information obtained from TLC analysis was used to select the active fractions and the mobile phase mixture for the purification steps using semi-preparative RPHPLC. An isocratic solvent system consisting of methanol: water at a ratio of 1:1 (v/v) was used as a mobile phase at a flow rate of 1.5 mL/

2.7.2. Ammonia test Dilute ammonia solution (5 mL) was added to the crude extract followed by concentrated H2SO4. The formation of a yellow color that disappeared on the addition of a few drops of H2SO4 indicated the presence of flavonoids (Vimalkumar et al., 2014). 3

Industrial Crops & Products 149 (2020) 112307

P. Yingyuen, et al.

3. Results and discussion

Table 1 Results for phytochemical screening of flavonoids in crude extracts. Flavonoid tests

Observation

Inference

Shinoda test Ammonia test

Red to pink coloration The yellow coloration disappears after some time. Yellow precipitate formed Blackish green coloration

+ +

Lead acetate test Ferric chloride test

3.1. Ethanol extraction and phytochemical screening of flavonoids Ethanol is an effective solvent for extracting polar and some nonpolar compounds. Ethanol is preferred for the extraction processes for food and pharmaceutical products because it has a favorable toxicological profile, with an FDA limit of a residual level of 0.5 % (Chew et al., 2011). Ethanol also has a relatively low boiling point, which makes it easy to remove from the final product. It has also been reported to be effective for extracting flavonoids from plants (Bucar et al., 2013). For the extraction process in the present study, 1.0 kg of ground air-dried banana leaves (M. balbisiana) were added to 5 L of 95 % (v/v) ethanol in a batch reactor at room temperature (30 °C) for a period of 5 d with daily mixing. The extract was filtered by vacuum filtration to remove any residual powder then evaporated to yield 124 g of the crude extract (12.4 % of the dried banana leaves), which contained 5.3 % rutin. The flavonoids from the crude extracts were then phytochemically screened using the Shinoda, Ammonia, Lead acetate test and Ferric chloride tests. The results of phytochemical screening (Table 1) showed that the crude ethanoic extract of banana leaves (M. balbisiana) contained flavonoids. The concentration of rutin in the crude extract was quantified using standard RP-HPLC method. The HPLC chromatographic spectra of the crude extract was obtained from the diluted sample in the concentration range of 0.1–5.0 mg/L. The rutin concentration of the crude extract was calculated based on the calibration curve obtained with standard solutions of rutin in methanol (supplementary material), which shows that there is a good linear relationship between the concentrations of rutin and their corresponding peak areas. Isocratic RP18-HPLC provides a rapid and sensitive assay for quantitative determination of rutin concentration (Lu et al., 2006; Kuntic et al., 2007; Vachirapatama et al., 2011). A good linear relationship between the peak areas and the concentrations of rutin was observed in the range of 0.8−80 mg/L (Lu et al., 2006). Fig. 5. shows the HPLC chromatographic spectra of the crude extract compared with the rutin standard. The peak for rutin, detected at the specific wavelength of 365 nm, appearing 5.7 min after injection as the major chromatogram was clearly separated from the other peaks. From the HPLC results, the rutin content was calculated as 52.9 ± 3.6 mg/g crude extract.

+ +

Indications: present (+); absent (-).

2.7.3. Lead acetate test The extract was treated with a few drops of lead acetate solution. The formation of yellow precipitate indicated the presence of flavonoids. Orange to crimson color also indicated the presence of flavonoids (Vimalkumar et al., 2014). 2.7.4. Ferric chloride test A few drops of neutral ferric chloride solution were added to the extract. The formation of a blackish red color indicated the presence of flavonoids. (Vimalkumar et al., 2014). 2.8. Characterization 2.8.1. Mass spectrometry (MS) Mass spectra of the compounds were recorded by electro spray ionization (ESI) technique using Micro TOF-benchtop ESI TOF MS (Bruker Billerica, Billerica, MA, USA) and ion polarity of positive at a flow rate of 4.0 ml/min, 0.3 bar on C-18 column with total run time of 15 min. The sample used for recording the mass spectrum was prepared by dissolving 0.1 mg of compound in 10 mL of methanol. After 30 s of exposure to an ultrasound bath or until the sample was thoroughly dissolved, the solution was injected directly into the electrospray interface of a 1200 L/MS/MS (Varian) mass spectrometer. Molecular ions scanning range (m/z) was 100-1500. The analysis was performed at National Center for Genetic Engineering and Biotechnology (BIOTEC)), Pathum Thani, Thailand. 2.8.2. Nuclear magnetic resonance spectroscopy (NMR) The 1H NMR was recorded for the isolated compound (15 mg in 1000 μl, MeOH) at 300 MHz on BRUKER AQS NMR spectrometer (Bruker biospin, AG, Switzerland). The J-modulated spin-echo for C-13 nuclei coupled to proton to determine number of attached protons (SEFT) was recorded at 300 MHz. The spectral width for 1H NMR was 0−10 ppm and 0−200 ppm for 13C NMR. The deuterated NMR solvent used in this study was methanol-d4 (CD3OD) containing 0.03 % (v/v) TMS (trimethylsilane), ≥99.8 % atom D. The number of scans: 1H-NMR was 64 and 13C-NMR was 20,000. The acquisition time was 5 s, the relaxation delay was 3 s and the spectrometer frequency was 300 MHz.

3.2. Solvent partition separations The crude ethanolic extract was further fractionated using n-hexane, dichloromethane and water, from low to high polarity, respectively. The percentage yield of the fractionated extracts was based on the weight of crude ethanolic extract. Partition using n-hexane exhibited the highest solids yield (62.9 %) with considerably lower yields from partitions using dichloromethane (17.1 %) and water (18.8 %). Hexane has been widely used to extract non- or less polar compounds, such as triglycerides, phytosterols, and phospholipids. The compounds from the hexane fraction usually contain large amounts of chlorophyll as well as waxes (Levente, 2005). Therefore, hexane can be used as a solvent to remove non- or less polar compounds from the crude extract. In contrast, the compounds from the water fraction were polar, such as alcohols and carbohydrate groups (Tang et al., 2006). The concentrations of rutin in extracts from n-hexane, dichloromethane and water fractions were determined using HPLC. The yields of rutin in the dichloromethane part and water fractions were 7.2 and 39.6 mg/g crude extract with rutin contents of 11.5 % and 40.6 % dry weight, respectively. No rutin was detected in n-hexane fraction. Rutin is a diglycosidic flavonoid. Therefore, it is a polar compound, which is not soluble in non-polar solvents, such as hexane.

2.8.3. High performance liquid chromatography (HPLC) The samples of crude extract, extracts from solvent partition separation, all fractions obtained from Sephadex column and all fractions separated by semi-preparative RP-HPLC were used for HPLC analysis to determine the concentration of rutin in each sample. The samples in methanol were filtrated through membrane filler (Millipore, USA) and were injected (10 μL) through RP-18 HPLC column (5 μm particle size, L × I.D. 25 cm × 4.6 mm) (Merck, Darmstadt, Germany) at room temperature (30 °C). The mobile phase composed of methanol and water (1:1 v/v) was eluted at a flow rate of 1 mL/min and effluent was monitored at 365 nm by UV detector. The peak areas were detected and compared with the standard rutin. A standard calibration curve was constructed using rutin hydrate ≥95 % (HPLC) (Sigma-Aldrich, St. Louis, MO) in different concentrations varied from 0.0 to 5.0 mg/L.

3.3. Separation by Sephadex™ column The water partition extract of 2.82 g was further isolated using a 4

Industrial Crops & Products 149 (2020) 112307

P. Yingyuen, et al.

Fig. 5. HPLC chromatographic spectra of: (A) rutin standard; and (B) crude extracts from banana leaves (M. balbisiana).

Sephadex G-100 column (2.5 cm × 150 cm), with 100 % methanol as the mobile phase. The extract was isolated according to differences in molecular size as it passed through the column packed with porous spherical particles 40,120 μm. Table 2 shows the seven fractions collected from the Sephadex column labeled as S1, S2, S3, S4, S5, S6 and S7. The yields from fractions S1 to S7 were 0.2 %, 1.1 %, 7.9 %, 49.8 %, 35.1 %, 5.0 % and 0.5 %, respectively. Rutin was found in fractions of S4, S5, S6 and S7 with rutin contents of 74.1 %, 77.9 %, 84.4 % and 82.3 % dry weight, respectively. The extracts could be eluted isocratically using a mobile phase of 100 % methanol so that there was no need to use other buffers during separation (Swifta et al., 2017). Each fraction was later checked by TLC, revealing the color differences in green, yellow, orange and blue as shown in Fig. 6.

Table 2 Percentage yield of fractions separated by Sephadex column, percentage purity of rutin, and rutin content based on g dry weight of crude extract. Fractions

%Yield of separation by Sephadex column,

%Purity of rutin

Rutin content (mg/g crude extract)

S1 S2 S3 S4 S5 S6 S7

0.2 1.1 7.9 49.8 35.1 5.0 0.5

0.0 0.0 0.0 74.1 77.9 84.4 82.3

0.0 0.0 0.0 23.3 ± 1.6 13.9 ± 1.0 2.0 ± 0.1 0.2 ± 0.0

3.4. Purification by semi-preparative C18-reverse phase-high performance liquid chromatography (RP- HPLC) 3.4.1. Determining the mobile phase mixture and selecting fractions using TLC The separation conditions for rutin were determined using TLC. The samples were spotted at a distance of 0.5 cm from the edge of the TLC plates using a microcapillary tube. The migration was performed in a tightly-closed flask saturated by the vapors of the solvents. In most cases, development over a distance of 8 cm led to the separation of a single area. The separation of rutin as an orange spot in samples was based on the retention factor (Rf) values after highlighting the spots in UV light at λ =365 nm. Ultraviolet light was selected as the visualizing agent because the chromophore, C = C-C = C-C = C-C = O, is a part of the flavones ring. Rutin was revealed as a yellow to orange spot under natural light and as a dark fluorescence under UV light (Georgeta et al., 2016). The presence of rutin was confirmed using natural product reagent (NP). The compositions of the solvent mixtures and the Rf values of the orange spots are shown in Table 3. The result revealed that the solvent mixture M7 (methanol: water, 1:1, v/v) was most effective for

Fig. 6. RP-TLC of crude extract (C) and the fractions (S1 to S7) obtained from a Sephadex column using mobile phase methanol : water (1:1).

5

Industrial Crops & Products 149 (2020) 112307

P. Yingyuen, et al.

flavone-8-C-glycosides eluted faster than the corresponding flavone-6C-glycosides (Marston and Hostettmann, 2005). Flavonoids, including orientin, vitexin, rutin and their isomers (isoorientin, isovitexin, quercetin-3-O-robinobioside) were previously detected in the common buckwheat sprout extract (CSE) and were separated using HPLC (Nam et al., 2015; Jang et al., 2019). The results in this study showed that the RP-HPLC system could separate rutin effectively. The S4 and S5 fractions provided rutin of different purities before they entered the RP-HPLC column. However, after passing through the column, the concentration of rutin in both fractions increased and the peak purities were quite similar (98.1 % and 98.9 %, respectively). The column efficiency would also depend on the system of mobile phase and stationary phase selected. The results showed that the low polarity of the stationary phase and the high polarity of the mobile phase could effectively separate rutin, which has a rather high polarity with a molecular weight of ∼ 610 Da. Rutin is a rutinoside consisting of quercetin with the hydroxyl group at position C-3 substituted with a disaccharidic moiety (6-rhamnopyranosyl-glucopyranose) trivially called rutinose. Its solubility in non-polar solvents is low, so rutin can hardly be separated using a non-polar solvent. Thus, an effective mobile phase with a specific affinity for separating a particular substance must be used. The ratio between polar and non-polar components in the mobile phase is a key factor affecting the separation efficiency, which has to be optimized in order to obtain a high purity of product. Chromatography is a technology of choice for commercial scale purification processes of bioactive compounds from plants for supplementary medical application, because this process can produce very high purity product with low operating cost and it can be modified into continuous process. Rutin has been reported for many pharmacological activities. The price of high-purity rutin (98 %) is around 50–60 US $/ kg (Alibaba.com, 26/12/2019). Compared to other sources of rutin, this study demonstrates that banana leaves (≤ 0.3 US $/ kg) and its byproducts (≤ 0.1 US $/ kg) has a good potential to be used as an inexpensive source for rutin production. However, for a commercial production process, further studies for the optimization and improvement of each step should be continuously carried on.

Table 3 Solvent mixture and Rf values of the orange spot for thin layer chromatography (TLC) isolation of rutin. Solvent mixtures

Components

Ratio (v/ v)

Rf values of orange spots

M1 M2 M3 M4 M5 M6 M7

dichloromethane-methanol dichloromethane-methanol ethyl acetate-methanol ethyl acetate-methanol Methanol-water Methanol-water Methanol-water

4 :1 2 :1 4:1 2:1 4:1 2 :1 1:1

0.15 0.25 0.20 0.33 0.45 0.50 0.60

isolating rutin, with the orange spot revealed at an Rf value of 0.60. Therefore, this mixture was used for separating rutin by semi-preparative RP-HPLC. The results from RP-TLC for the crude extract and the fractions obtained from the Sephadex column using the M7 mobile phase (methanol: water at 1:1) are shown in Fig. 6. RP-TLC of the S1, S2 and S3 fractions (Table 2) indicated an almost total absence of rutin, whereas RP-TLC of the S4 and S5 fractions clearly showed orange spots of a higher intensity than those from the S6 and S7 fractions. RP-TLC from the S4 fraction also indicated some mixed compounds, which might have affected the purity of the rutin product.

3.4.2. Yield and purity of rutin from semi-preparative RP-HPLC The S4 and S5 fractions were further purified using a semi-preparative RP-HPLC column (10 mm × 250 mm) and eluted isocratically with a solvent mixture of methanol and water at a ratio of 1:1 (v/v) and a flow rate of 1.5 mL/min. The chromatograms were recorded at 365 nm. The fractions were collected following the peaks of the chromatogram. The compounds were later identified using NMR, HPLC and MS. Three fractions were collected (W1, W2 and W3) for both S4 and S5 (Table 4). The yields of solids from W1, W2 and W3, separated using the semi-preparative RP-HPLC column from the S4 fraction of the Sephadex column, were 1.6 %, 85.3 % and 5.2 %, respectively, while those from the S5 fraction were 0.6 %, 90.5 % and 3.5 %, respectively. Based on the weight of the crude ethanol extract, the rutin yields from W2 and W3 obtained from the S4 fraction were 19.8 and 1.1 mg/g of crude extract, respectively, while those from the S5 fraction were 12.6 and 0.7 mg/g of crude extracts, respectively. However, no rutin was detected in W1 from the S4 and S5 fractions. The rutin product obtained from the W2 fraction of S4 and S5 was 32.4 mg/g of crude extracts with the purity of 98.4 %. In the present study, a C18 column was used as a non-polar stationary phase in RP-HPLC, with a mixture of methanol and water at a ratio of 1:1 (v/v) as a strong polar mobile phase for the separation of the polar molecules of rutin. The retention of rutin in RP-HPLC depends on the polarity of the stationary phase, the mobile phase, and the chemical structure of rutin. In general, the retention time is proportional to the chain length and the number of double bonds present in the rutin (Smith, 1995; Turková, 1975), which affect the polarity of the compound. The geometric isomerization of rutin might also play an important role in semi-preparative RP-HPLC separation. For example,

3.5. Identifying rutin using mass spectrometry (MS) and nuclear magnetic resonance (NMR) The products of W2 from the mixture of S4 and S5 fractions collected from the semi-preparative RP-HPLC using methanol and water (1:1, v/v) as the mobile phase were further identified by electrospray ionization MS (ESI-MS) and NMR. The ESI-MS data (Fig. 7) showed a major parent molecular ion [M+23] peak at m/z 633.14. The molecular weight (MW) of the W2 compound was at m/z 633.14 ([M + Na]+) and at m/z 611.16 ([M+H]+) Elemental analysis of this compound revealed the presence of C (53.11 %), H (4.91 %) and O (41.96 %). The mass spectra data established the chemical formula as C27H30O16. These ESI-MS data corresponded with those for rutin (Nidhi et al., 2014). According to the MS analysis, the MW of W3 was 617.14 m/z ([M + Na]+) (Fig. 8) and its elemental analysis reveals the presence of C (54.54 %), H (5.05 %) and O (40.40 %). The mass spectra data indicated the chemical formula as C27H30O15. These ESI-MS data corresponds to those of kaempferol-3-O-rutinoside. However, the

Table 4 Percentage yield of fractions separated by semi-preparative RP-HPLC, percentage purity of rutin and rutin content based on g dry weight of crude extract. Fractions

W1 W2 W3

%Yield of RP-HPLC

%Purity of rutin

Rutin content (mg/g crude extract)

S4

S5

S4

S5

S4

S5

Total

1.6 85.3 5.2

0.6 90.5 3.5

– 98.1 1.9

– 98.9 1.0

– 19.8 ± 1.4 1.1 ± 0.1

– 12.6 ± 0.2 0.7 ± 0.2

– 32.4 ± 1.2 1.8 ± 0.1

6

Industrial Crops & Products 149 (2020) 112307

P. Yingyuen, et al.

Fig. 7. MS Spectra of W2 (A) validated with rutin (C27H30NaO16+) (B).

6.19 (1H, d, J = 2 Hz, H-8), 6.38(1H, d, J = 8 Hz, H-5′), 7.66 (1H, m, H-2′, H-6′). The result of the 13C-NMR analysis of the compound chemical shifts (Fig.9) compared with those of the rutin standard of reference (1) (Selvaraj et al., 2013) and reference (2) (Guvenalp et al., 2006) is shown in (Table 6). Instrument MHz and deuterated NMR solvents of reference (1), reference (2) and this study were: (600 MHz, CD3OD), (100 MHz, CD3OD), and (300 MHz, CD3OD), respectively. These were characterized as 3,3′,4′,5,7-pentahydroxy flavones-3-rutinoside (rutin) (Selvaraj et al., 2013). The results therefore, confirmed the chemical structure of rutin and demonstrated an acceptably high purity of W2 compound obtained from the separation using semi-preparative RP-

information from the MS data should be confirmed by comparing their retention times on HPLC and by NMR analysis. The results of the 1H-NMR spectral analysis of the main product (W2) are shown in Table 5, revealing spectra (300 MHz in CH3OD, δ ppm): 3.20–3.67 (m,12H of sugar moieties), 3.71(d, J = 1.15 Hz, 1HRham), 1.02 (3H, d, J = 6 Hz, CH3-Rham), 4.40(4H, d, J = 7.8 Hz, H-1 Glu), 4.99 (1H, d, J = 2 Hz H-6), 6.09 (1H, d, J = 2 Hz, H-8), 6.28(1H, d, J = 8 Hz, H-5′), 7.50–7.56 (1H, m, H-2′, H-6′), which are comparable with those reported of the rutin standard (Selvaraj et al., 2013), 1HNMR (600 MHz in CH3OD, δ ppm): 3.33–3.64 (m,12H of sugar moieties), 3.81(d, J = 1.15 Hz, 1H-Rham), 1.10(3H, d, J = 6 Hz, CH3Rham), 4.52(4H, d, J = 7.8 Hz, H-1 Glu), 5.11 (1H, d, J = 2 Hz H-6),

Fig. 8. MS Spectra of W3 (A) validated with kaeampferol-3-O-rutinoside (C27H30NaO15+) (B). 7

Industrial Crops & Products 149 (2020) 112307

P. Yingyuen, et al.

Table 5 1 H-NMR spectral data (δ, J) for rutin (W2) obtained in this study in comparison to those of the previous report (δref, Jref) (Selvaraj et al., 2013).

Table 6 Comparison of 13C-NMR chemical shift data between the rutin references and the W2 compound obtained from this study.

δref

δ

J

Jref

Assignment

Position

Reference (1)a

Reference (2)b

W2 (this study)c

1.10 3.33–3.64 3.81 4.52 5.11 6.19 6.38 7.66

1.02 3.20–3.67 3.71 4.40 4.99 6.09 6.28 7.50–7.56

6 Hz – 1.15 Hz 8 Hz 2 Hz 2 Hz 8 Hz –

6 Hz – 1.15 Hz 7.8 Hz 2 Hz 2 Hz 8 Hz –

d,3H, CH3 m,12H of sugar moieties d, 1H, CH-CH2O glucose d, 1H, OCHO glucose d, 1H, OCHO ramnose d, 1H, Ar-H d, 1H, Ar-H m-2H, Ar-H (6′and 2′)

2 3 4 5 6 7 8 9 10 1’ 2′ 3′ 4′ 5′ 6′ 1’’ 2′’ 3′’ 4′’ 5′’ 6′’ 1’’’ 2′’’ 3′’’ 4′’’ 5′’’ 6′’’

158.5 135.7 179.4 163.0 100.0 166.0 95.0 159.4 105.7 123.7 117.8 145.9 149.9 116.1 123.2 104.8 75.8 77.2 71.5 78.2 68.6 102.5 72.3 72.2 74.0 69.8 18.0

158.5 135.6 179.4 162.5 99.9 166.0 94.8 159.3 105.6 123.1 117.6 145.8 149.7 116.1 123.5 104.7 75.7 77.2 71.4 78.1 68.6 102.4 72.0 72.2 73.9 69.7 17.9

158.5 135.6 179.3 163.0 100.2 166.5 94.9 159.3 105.5 123.6 117.6 145.9 149.9 116.1 123.0 104.7 75.7 77.2 71.4 78.1 68.6 102.4 72.2 72.1 73.9 69.7 17.9

HPLC. 4. Conclusions The present study has demonstrated that banana leaves from M. balbisiana, an agricultural by-product or waste, can be exploited as a new potential source of rutin. A method has been successfully developed for separating rutin from banana leaf extracts and purifying the rutin product. Chopped dried banana leaves were macerated at room temperature with 95 % ethanol (v/v) for 5 d. The extract was filtered and evaporated to yield the crude extract of 12.4 % from dried banana leaves with a rutin content of 52.9 mg/g crude extracts. The rutin from the ethanolic crude extract was then separated using a series of solvent partition separations. The most suitable mobile phase system was determined by using TLC. Sephadex column chromatography with 100 % methanol and semi-preparative RP-HPLC using an isocratic solvent system consisting of methanol: water (1:1 v/v) were used to purify the rutin product. This process provided a high yield of rutin at 32.4 mg rutin/g crude extract with a purity of 98.4 %.

*a (Selvaraj et al., 2013) (600 MHz, CD3OD). *b (Guvenalp et al., 2006) (100 MHz, CD3OD). *c (This study) (300 MHz, CD3OD).

Methodology, Supervision, Validation, Writing - review & editing.

Declaration of competing interest Acknowledgements The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

This work is financially supported by the 90th Anniversary of Chulalongkorn University Fund (Ratchadaphiseksomphot Endowment Fund).

CRediT authorship contribution statement Appendix A. Supplementary data

Panida Yingyuen: Data curation, Formal analysis, Investigation, Methodology, Writing - original draft. Suchada Sukrong: Conceptualization, Investigation, Methodology, Supervision. Muenduen Phisalaphong: Conceptualization, Investigation,

Fig. 9.

13

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.indcrop.2020.112307.

C-NMR spectra of W2 (rutin). 8

Industrial Crops & Products 149 (2020) 112307

P. Yingyuen, et al.

References

by high-performance liquid chromatography. J. Pharm. Biomed. Anal. 41, 1186–1190. Marston, A., Hostettmann, K., 2005. Separation and quantification of flavonoids. In: Andersen, O.M., Markham, K.R. (Eds.), Flavonoids: Chemistry, Biochemistry, and Applications. CRC Press, Boca Raton, FL, USA, pp. 1–36. Müller, I., Beindl, A., 2006. Drying of medicinal plants. Med. Aromat. Plants (Los Angel) 237–252. Nam, T.G., Lee, S.M., Park, J.H., Kim, D.O., Baek, N.I., Eom, S.H., 2015. Flavonoid analysis of buckwheat sprouts. Food Chem. 170, 97–101. Nidhi, G., Chauhan, R.S., Pradhan, J.K., 2014. Handbook of Medicinal Plants and Their Bioactive Compounds. pp. 51–57. Padam, B.C., Tin, H.S., Chye, F.Y., Abdullah, M.I., 2014. Banana by-products: an underutilized renewable food biomass with great potential. J. Food Sci. Technol. 51 (12), 3527–3545. Selvaraj, K., Chowdhury, R., Bhattacharjee, C., 2013. Isolation and structural elucidation of flavonoids from aquatic fern azolla microphylia and evaluation of free radical scavenging activity. Int. J. Pharm. Pharm. Sci. 5, 743–749. Shama, V.J., Patel, P.M., 2013. Evaluation of antibacterial activity of methanloic extract of plant Rivea ornata. Phamacy. Int. Res. J. Pharm. 4, 233–234. Smith, R.M., 1995. Chapter 3 Retention index scales used in high-performance liquid chromatography. J. Chromatography library 57, 93–144. Suzuki, T., Honda, Y., Mukasa, Y., 2005. Effects of UV-B radiation, cold and desiccation stress on rutin concentration and rutin glucosidase activity in tartary buckwheat (Fagopyrum tataricum) leaves. Plant Sci. 168, 1303–1307. Swangpol, S., Sotto, R., Volkaert, H.A., Seelanan, T., 2007. Utility of selected non-coding chloroplast DNA sequences for lineage assessment of Musa interspecific hybrids. J. Biochem. Mol. Biol. 40 (4), 577–587. Swifta, T., Hoskinsa, R., Telforda, R., Plenderleithb, R., Pownalla, D., Rimmera, S., 2017. Analysis using size exclusion chromatography of poly(N-isopropylacrylamide) using methanol as an eluent. J. Chromatogr. A 1508, 16–23. Takahama, U., Tanaka, M., Hirota, S., Yamauchi, R., 2009. Formation of an oxathiolone compound from rutin in acidic mixture of saliva and buckwheat dough: possibility of its occurrence in the stomach. Food Chem. 116, 214–219. Tang, J.C., Zhang, J.N., Wu, Y.T., Li, Z.X., 2006. Effect of the water extract and ethanol extract from traditional Chinese medicines Angelica sinensis (Oliv.) Diels, Ligusticum chuanxiong Hort. And Rheum palmatum L. On rat liver cytochrome P450 activity. Phytother. Res. 20, 1046–1051. Turková, J., 1975. Chapter 7 Affinity chromatography. J. Chromatography Library. 3, 89–97. Vachirapatama, N., Chamnakid, B., Kachonpadungkitti, Y., 2011. Determination of rutin in buckwheat tea and Fagopyrum tataricum seeds by high performance liquid chromatography and capillary electrophoresis. J. Food Drug Anal. 19, 463–469. Vimalkumar, C.S., Hosagaudar, V.B., Suja, S.R., Vilash, V., Krishnakumar, N.M., Latha, P.G., 2014. Comparative preliminary phytochemical analysis of ethanolic extracts of leaves of Olea dioica Roxb., infected with the rust fungus Zaghouania oleae (E.J.Butler) Cummins and non-infected plants. J. Pharmacogn. Phytochem. 3, 69–72. Wu, C., Jia, S., Fan, G., Li, T., Ying, R., Yang, J., 2013. Purification and identification of novel antioxidant peptides from enzymatic hydrolysate of Ginkgo biloba seed proteins. Food Sci. Technol. Res. 19, 1029–1035.

Abu, F., Mat Taib, C.N., Mohd Moklas, M.A., Mohd Akhir, S., 2017. Antioxidant properties of crude extract, partition extract, and fermented medium of Dendrobium sabin flower. Evid. Based Complement. Altern. Med. 2017, 1–9. Atanassova, M., Bagdassarian, V., 2009. Rutin content in plant product. J. Univ. Chem. Technol. Metal. 44, 201–203. Bucar, F., Wube, A., Schmid, M., 2013. Natural product isolation-how to get from biological material to pure compounds. Nat. Prod. Rep. 1, 525–545. Chew, K.K., Ng, S.Y., Thoo, Y.Y., Khoo, M.Z., Wan Aida, W.M., Ho, C.W., 2011. Effect of ethanol concentration, extraction time and extraction temperature on the recovery of phenolic compounds and antioxidant capacity of Centella asiatica extracts. Int. Food Res. J. 18, 571–578. Coskun, O., 2016. Separation techniques: chromatography. North Clin Istanbul. 3, 156–160. Fabjan, N., Rode, N., Košir, I.J., Wang, Z., Zhang, Z., Kreft, I., 2003. Tartary Buckwheat (Fagopyrum tataricum Gaertn.) as a source of dietary rutin and quercitrin. J. Agric. Food Chem. 51, 6452–6455. Frezza, C., De Vita, D., Spinaci, G., Sarandrea, M., Venditti, A., Bianco, A., 2018. Secondary metabolites of Tilia tomentosa Moench inflorescences collected in Central Italy: chemotaxonomy relevance and phytochemical rationale of traditional use. Natural Poduct Research 2018. https://doi.org/10.1080/14786419.2018.1550487. Georgeta, S., Ana, P.I., Tunde, H., Sanda, B., 2016. The isolation and identification of rutin from pharmaceutical products. Synthesis of Biol. Active Natural Products. 15, 109–114. Guvenalp, Z., Nurcan, K., Kazaz, C., Yusuf, K., Omur, D.L., 2006. Chemical constituents of galium tortumense. Turk. J. Chem. 30, 515–523. Habtemariam, S., George, K.V., 2015. Extractability of rutin in herbal tea preparations of Moringa stenopetala leaves. Beverages. 1, 169–182. Hussain, J., Bassal, M., Sarhan, H., Issam, M., Aga, H., 2017. Qualitative and quantitative comparison of rutin, quercetin and gallic acid concentrations in Syrian Capparis spinosa. L Leaves. J. Pharmacogn. Phytochem. 6, 407–415. Imam, M.Z., Akter, S., 2011. Musa paradisiaca L. And Musa sapientum L.: a phytochemical and pharmacological review. J. Appl. Pharm. Sci. 1, 14–20. Jang, D., Jung, Y.S., Kim, M.S., Oh, S.E., Nam, T.G., Kim, D.O., 2019. Developing and validating a method for separating flavonoid isomers in common buckwheat sprouts using HPLC-PDA. Foods. 8, 549. https://doi.org/10.3390/foods8110549. Kreft, S., Knapp, M., Kreft, I., 1999. Extraction of rutin from buckwheat (Fagopyrum esculentum Moench) seeds and determination by capillary electrophoresis. J. Agric. Food Chem. 47, 4649–4652. Kuntic, V., Pejic, N., Ivkovic, B., Vujic, Z., Ilic, K., Micic, S., Vukojevic, V., 2007. Isocratic RP-HPLC method for rutin determination in solid oral dosage forms. J. Pharm. Biomed. Anal. 43, 718–721. Laillyza, M., Iskandar, A., Suhartono, E., 2014. Bioactive compound and antioxidant activity of methanol extract mauli bananas (Musa sp) Stem. Int. J. Biosci. Biochem. Bioinform. 4, 110–115. Levente, L.D., 2005. Chlorophyll removal from edible oils. Int. J. Appl. Sci. Eng. 3, 81–88. Lu, Y.H., Liu, Z.Y., Wang, Z.T., Wei, D.Z., 2006. Quality evaluation of Platycladus orientalis (L.) Franco through simultaneous determination of four bioactive flavonoids

9