Hydroxybenzoic Acids Are Significant Contributors to the Antioxidant Effect of Borututu Bark, Cochlospermum angolensis Welw. ex Oliv.

Borututu (Cochlospermum angolensis) is an African tree whose bark has recently emerged as a herbal dietary supplement with claims for antioxidant activity. In order to substantiate the claimed activity of borututu supplements, we performed an activity-guided fractionation of the total extract utilizing a 1,1-diphenyl-2-picrylhydrazyl (DPPH) free radical scavenging assay. Subsequent flash and centrifugal chromatography resulted in the isolation of gallic acid (1) and protocatechuic acid (2) as the main antioxidant constituents. Two apocarotenoids and one flavonoid were also isolated from the chloroform fraction and were identified as cochloxanthin (3), dihydrocochloxanthin (4), and 7,4′-dimethyltaxifolin (5), respectively. A High-performance liquid chromatography (HPLC) method was also developed for fingerprinting borututu samples, with Compounds 1–4 suggested as chemical markers for quality control purposes.

With the recent emergence of borututu bark as a new antioxidant herbal beverage and dietary supplement [1,6,18] and due to the scarcity of phytochemical investigation of this plant, the goal of this study was to utilize high-performance liquid chromatography (HPLC) and an activity-guided approach to determine the most active constituents of borututu bark in a 1,1-diphenyl-2-picrylhydrazyl (DPPH) free-radical scavenging assay. A qualitative HPLC method was developed for (i) fingerprinting borututu crude and commercial products; (ii) determining the major phytochemical markers of the bark; and (iii) verifying the purity of isolated compounds. Gallic acid (1) and protocatechuic acid (2) were the most active DPPH scavengers identified in the ethyl acetate fraction. The two apocarotenoids cochloxanthin (3) and dihydrocochloxanthin (4) were also isolated and identified in addition to the flavanoid taxifolin-7,4 -dimethyl ether (5). The five isolated compounds are reported in borututu for the first time.

General Procedures
Coarse shredded borututu bark was purchased from Mond Trading Corp. (Toronto, ON, Canada). Other herbal products and dietary supplements were purchased online. DPPH and Trolox were purchased from Sigma (St. Louis, MO, USA). All solvents were of reagent grade (Fisher Scientific, Fair Lawn, NJ, USA). Ultrapure RO water was generated in-house (Barnstead Nanopure, Thermo Scientific, Marietta, OH, USA). HPLC fingerprinting, compound purity, and molecular weight determination were performed on an HPLC system equipped with UV/Vis and single-quadrupole mass detectors (LCMS-2020, Shimadzu, Kyoto, Japan). NMR experiments were run on an ECS-400 spectrometer (JEOL, Tokyo, Japan) using CDCl 3 and acetone-d 6 as solvents.

Preparation of Total Extract and Solvent Fractions
Borututu bark was ground to a fine powder and 500 g was soaked overnight in methanol (3 L × 3). The three batches were filtered, combined, and concentrated under vacuum at 45 • C in a rotary evaporator (model R-215, Buchi, Flawil, Switzerland) to yield 102.8 g of a dark reddish brown residue of total methanolic extract (TME). Four aliquots of TME (57.2 g total, ca. 14.3 g per run) were separately triturated to homogeneity with dry silica, placed in a 100 mL stainless steel cylinder and fractionated using solvents of increasing polarity, viz. n-hexane, CHCl 3 , EtOAc, and MeOH, in an accelerated solvent extractor (Dionex ASE 150, Thermo Fisher Scientific, Waltham, MA, USA) with the following settings: temperature, 55 • C; static time, 15 min; rinse volume, 60%; purge time, 100 s; static extraction cycles, 3. Fractions extracted with each solvent were combined and concentrated in a rotavapor at 45 • C then transferred to pre-weighed vials labeled HF, CF, EF, and MF for n-hexane, CHCl 3 , EtOAc and MeOH fractions, respectively. Total extracts of reference herbals and spices were prepared by ultrasonicating finely powdered 5 g samples of green tea (GTE), milk thistle (SME), horse apple (BPE) and nutmeg (MFE) in methanol for 15 min followed by filtration and drying at 45 • C under vacuum.

DPPH Assay [17]
DPPH stock solution was prepared in a volumetric flask by dissolving 6.25 mg of DPPH in 25 mL of MeOH (25 mg/100 mL, 0.62 mM). Trolox standard solution was prepared by dissolving 4.5 mg of Trolox in 250 mL of methanol (18 µg/mL, 72 µM). A DPPH working solution (0.16 mM) was prepared by a 1:3 dilution of the stock solution. Trolox calibrators were prepared by serial dilution of the stock standard solution to generate 5 concentration levels (1.13-18 µg/mL). Sample stock solutions were prepared at 0.5 mg/mL concentration. Samples were tested at 1:10 dilution and further dilutions were subsequently prepared based on initial runs. Assays were performed in 96-well plates as follows: 100 µL of each blank (MeOH), sample and standard were transferred to specific wells followed by 100 µL of the DPPH working solution (all samples run in triplicates). Each plate was covered and kept in the dark for 30 min, after which it was scanned at 515 nm in an Epoch plate reader (BioTek, Winooski, VT, USA). The concentration of each sample was calculated from the generated calibration curve (Figure 1

Extraction, Fractionation, and Isolation of Active Compounds
Extraction of borututu bark powder (500 g) yielded 102.7 g of total methanolic extract (TME), of which 57.2 g were used for subsequent fractionation. The utilized weight (57.2 g) subsequently yielded four fractions obtained by successive solvent extraction of TME-coated silica (HF, 0.64 g; CF, 5.63 g; EF 6.58 g; MF, 29.45 g). The ethyl acetate fraction (EF) showed the highest DPPH scavenging activity followed by the methanol fraction (MF). This was qualitatively verified by a visible inspection of DPPH-dipped TLC plates and quantitatively determined by 96-well plate DPPH-scavenging assays as shown in Figures 2A and 3, respectively. Thus, EF was further investigated to identify active compounds. Flash chromatography of an aliquot of EF (1.40 g) resulted in five subfractions (SF1-5) with SF4 (0.17 g) and SF5 (0.14 g) showing the highest activity ( Figure 3). TLC analysis of SF4 and SF5 showed two major compounds that bleached the purple background after dipping in DPPH solution ( Figure 2B). Each subfraction was subjected to centrifugal preparative TLC (Chromatotron), resulting in one major compound per subfraction (SF4: Compound 2, 0.07 g; SF5: Compound 1, 0.03 g). Although the CHCl3 fraction (CF) exhibited lower DPPH scavenging activity, its TLC profile showed two orange spots and one faint yellow spot that were isolated by preparative flash chromatography (Isolera One ® ) of an aliquot (0.48 g) to yield Compounds 3 (0.03 g), 4 (0.03 g), and 5 (0.005 g).

Extraction, Fractionation, and Isolation of Active Compounds
Extraction of borututu bark powder (500 g) yielded 102.7 g of total methanolic extract (TME), of which 57.2 g were used for subsequent fractionation. The utilized weight (57.2 g) subsequently yielded four fractions obtained by successive solvent extraction of TME-coated silica (HF, 0.64 g; CF, 5.63 g; EF 6.58 g; MF, 29.45 g). The ethyl acetate fraction (EF) showed the highest DPPH scavenging activity followed by the methanol fraction (MF). This was qualitatively verified by a visible inspection of DPPH-dipped TLC plates and quantitatively determined by 96-well plate DPPH-scavenging assays as shown in Figures 2A and 3, respectively. Thus, EF was further investigated to identify active compounds. Flash chromatography of an aliquot of EF (1.40 g) resulted in five subfractions (SF1-5) with SF4 (0.17 g) and SF5 (0.14 g) showing the highest activity ( Figure 3). TLC analysis of SF4 and SF5 showed two major compounds that bleached the purple background after dipping in DPPH solution ( Figure 2B). Each subfraction was subjected to centrifugal preparative TLC (Chromatotron), resulting in one major compound per subfraction (SF4: Compound 2, 0.07 g; SF5: Compound 1, 0.03 g). Although the CHCl 3 fraction (CF) exhibited lower DPPH scavenging activity, its TLC profile showed two orange spots and one faint yellow spot that were isolated by preparative flash chromatography (Isolera One ® ) of an aliquot (0.48 g) to yield Compounds 3 (0.03 g), 4 (0.03 g), and 5 (0.005 g).

Evaluation of Antioxidant Activity
Compounds 1 and 2 had the highest free-radical scavenging activity of all tested samples ( Figure 3). It is to be noted that antioxidant ellagic acid (Figure 4) derivatives reported earlier by Ferreres et al. are dimeric analogs of 1 and 2 and are probably responsible for the observed free-radical scavenging activity of MF [5]. When compared to the antioxidant herbs-green tea (GTE), milk thistle (SME), and horse apple (BPE)-the DPPH scavenging activity, expressed as percent Trolox equivalence, of the total borututu extract was approximately 10%, 50%, and 100% that of GTE, SME, and BPE extracts, respectively (Figure 3). The DPPH scavenging activity of borututu was equivalent to that of nutmeg (MFE), which is also reported to possess antioxidant activity (Figure 3) [23]. As shown in Figure 3, the DPPH activity of borututu samples gradually increased with each level of fractionation, reaching the highest activity with pure compounds. Thus, by adopting free-radical scavenging-guided fractionation, the most active antioxidants of borututu bark were isolated and identified as gallic and protocatechuic acid (1 and 2, respectively).

HPLC Analysis of Extract, Fractions, and Pure Compounds
HPLC fingerprinting was initially performed on borututu TME in order to identify different phytochemical markers that may be used to characterize the extract and to guide the subsequent isolation of these markers. Each solvent fraction of the total extract (CF, EF, and MF) contained one or more of the markers identified in TME. Figure 5A shows the fingerprint of the total extract and the chromatographic profiles of active subfractions with major markers identified in each. Fraction

Evaluation of Antioxidant Activity
Compounds 1 and 2 had the highest free-radical scavenging activity of all tested samples (Figure 3). It is to be noted that antioxidant ellagic acid (Figure 4) derivatives reported earlier by Ferreres et al. are dimeric analogs of 1 and 2 and are probably responsible for the observed free-radical scavenging activity of MF [5]. When compared to the antioxidant herbs-green tea (GTE), milk thistle (SME), and horse apple (BPE)-the DPPH scavenging activity, expressed as percent Trolox equivalence, of the total borututu extract was approximately 10%, 50%, and 100% that of GTE, SME, and BPE extracts, respectively ( Figure 3). The DPPH scavenging activity of borututu was equivalent to that of nutmeg (MFE), which is also reported to possess antioxidant activity (Figure 3) [23]. As shown in Figure 3, the DPPH activity of borututu samples gradually increased with each level of fractionation, reaching the highest activity with pure compounds. Thus, by adopting free-radical scavenging-guided fractionation, the most active antioxidants of borututu bark were isolated and identified as gallic and protocatechuic acid (1 and 2, respectively).

HPLC Analysis of Extract, Fractions, and Pure Compounds
HPLC fingerprinting was initially performed on borututu TME in order to identify different phytochemical markers that may be used to characterize the extract and to guide the subsequent isolation of these markers. Each solvent fraction of the total extract (CF, EF, and MF) contained one or more of the markers identified in TME. Figure 5A shows the fingerprint of the total extract and the chromatographic profiles of active subfractions with major markers identified in each. Fraction CF contained 3, 4, and 5 (retention times: 21.1, 22.3, and 13.6 min, respectively). Fraction EF showed 1 and 2 as its major constituents (retention times: 3.5 and 5.3 min, respectively). Fraction MF showed a cluster of peaks eluting between 7.5 and 11.5 min that were not isolated as pure compounds. However, these peaks showed the pseudomolecular ions [M−H] − 301, 315, 433, 447, 461, and 477 corresponding to ellagic acid and its glycosides and/or methyl analogs reported by Ferreres and co-workers [5]. The hexane fraction, on the other hand, did not possess any antioxidant activity and did not contain any significant UV-active markers. The same HPLC method was subsequently utilized to further confirm the purity of isolated compounds and to demonstrate their potential use as quality markers for borututu products ( Figure 5B).
Antioxidants 2017, 6, 9 7 of 9 CF contained 3, 4, and 5 (retention times: 21.1, 22.3, and 13.6 min, respectively). Fraction EF showed 1 and 2 as its major constituents (retention times: 3.5 and 5.3 min, respectively). Fraction MF showed a cluster of peaks eluting between 7.5 and 11.5 min that were not isolated as pure compounds. However, these peaks showed the pseudomolecular ions [M−H] − 301, 315, 433, 447, 461, and 477 corresponding to ellagic acid and its glycosides and/or methyl analogs reported by Ferreres and co-workers [5]. The hexane fraction, on the other hand, did not possess any antioxidant activity and did not contain any significant UV-active markers. The same HPLC method was subsequently utilized to further confirm the purity of isolated compounds and to demonstrate their potential use as quality markers for borututu products ( Figure 5B).

Conclusions
Borututu bark possesses significant in vitro free-radical scavenging activity that supports its use as an antioxidant herbal dietary supplement. The main active constituents were identified as gallic acid (1) and protocatechuic acid (2). Additionally, the two apocarotenoids cochloxanthin (3) and dihydrocochloxanthin (4) and the flavanoid taxifolin-7,4 -dimethyl ether (5) were identified as significant constituents of the bark. All identified compounds are reported for the first time in C. angolensis. Compounds 1-4 can serve as analytical markers for quality control of borututu products in addition to previously reported ellagic acids [5]. Development of quantitative analytical methods and further investigation of the pharmacokinetics/dynamics of borututu bark are thus warranted.
Author Contributions: Hao Wen Fu performed antioxidant assays for all samples, purified the active compounds, and compiled all the data. Ehab A. Abourashed is the project leader/academic advisor of Hao Wen Fu. He developed the HPLC method, collected and interpreted NMR spectra of all compounds, and wrote the manuscript.

Conflicts of Interest:
The authors declare no conflicts of interest.