Phytochemical, biological and toxicity studies of terminalia sericea burch. (Combretaceae)

Abstract

Terminalia sericea Burch. ex. DC (Combretaceae) is one of the 50 most popular medicinal plants in Africa. The fruit, leaves, stems and roots are commonly used for the treatment of cough, skin infections, diabetes, diarrhoea, venereal diseases and tuberculosis. However, the roots are most commonly used in the preparation of traditional medicines. Pharmacological studies have revealed that the crude root extracts display antibacterial activity against several Gram-positive and Gram-negative bacteria. Anolignan b, termilignan b and arjunic acid are reported to be the major antibacterial constituents present in the roots. Other compounds isolated from the roots include resveratrol-3-rutinoside, sericic acid, sericoside and arjunglucoside I. Authorities worldwide, including the Medicines Control Council of South Africa, have begun to regulate herbal drugs sold in the form of commercial formulations. Quality control of herbal drugs is challenging, since the chemical profiles of the raw materials may vary, depending on the origin of the plant material and the way that it was handled and processed. The chemistry, in turn, impacts on the safety and efficacy of the plant material. To date, there are no available data on parameters that can be used to standardise the quality of T. sericea raw materials. The aim of this study was therefore to provide information on the variation of the chemical constituents that contribute to the biological effects of the roots of T. sericea and also establish its safety. Since the compounds previously isolated from the roots were not commercially available, isolation of the major constituents of the roots was undertaken to obtain analytical standards. A crude dichloromethane:methanol (1:1) extract was initially fractionated using silica gel column chromatography, where after, some of the fractions were further purified using silica gel and Sephadex LH-20 column chromatography. Final purification of the enriched fractions was achieved using preparative high performance liquid chromatography coupled with mass spectrometry (prep-HPLC-MS). The structures of these compounds were subsequently elucidated using one- and two- dimensional nuclear magnetic

resonance (NMR) spectroscopy and mass spectrometry and identified as sericic acid (340 g), sericoside (500 g), resveratrol-3-rutinoside (240 mg) and arjunglucoside I (74 mg). The chemical variation within the crude root extracts of samples (n = 42) from ten populations in the Limpopo Province of South Africa, was determined using ultra performance liquid chromatography-mass spectrometry (UPLC-MS). A method was developed for the simultaneous determination of sericic acid, resveratrol-3-rutinoside, sericoside and arjungluicoside I in the extracts using UPLC with photodiode array detection (PDA). The method was validated according to the guidelines of the International Council for Harmonisation (ICH). A regression coefficient (R2) of 0.998 was obtained for sericic acid, resveratrol-3- rutinoside and arjunglucoside I, while the R2 value for sericoside was 0.999, indicating a linear relationship between the concentration and the detector response. Satisfactory limits of detection for sericic acid (25.2 ng/mL), sericoside (11.6 ng/mL), resveratrol-3-rutinoside (23.3 ng/mL) and arjunglucoside I (8.81 ng/mL) were determined. Recoveries of 98 % and 80% were obtained for samples spiked with 12.5 μg/mL and 25 μg/mL of resveratrol-3-rutinoside, respectively, indicating that the method is accurate. The intra- and inter-day variation in resveratrol-3-rutinoside concentration, measured over three days, indicated excellent analytical precision, since all the relative standard deviations were below 0.70 %. The quantitative data revealed that sericic acid (1.59 to 8.45 mg/g), sericoside (2.07 to 20.17 mg/g), resveratrol-3-rutinoside (0.65 to 29.82 mg/g) and arjunglucoside I (0.86 to 8.44 mg/g dry weight) were the major constituents of the root samples, but their concentrations were highly variable. Chemometric analysis of the aligned UPLC-MS data was used to investigate similarities and differences in the chemical profiles of the samples using an untargeted approach. A principal component analysis (PCA) model was constructed and subsequently hierarchical cluster analysis (HCA) indicated the presence of two main groups, which were found to be independent of the populations to which the samples belong. Classes, based on the HCA class identifiers, were subsequently assigned to the samples, and an orthogonal projection to latent structures-discriminant analysis (OPLS-DA) model was then constructed, (R2 cum = 0.996 and Q2 cum = 0.967). The corresponding loadings

plot allowed sericic acid, sericoside and resveratrol-3-rutinoside to be identified as biomarkers associated with the first group. Quantitative, rather than qualitative differences were responsible for the observed clustering pattern. Techniques that could be applied in quality control protocols for T. sericea root were investigated. High performance thin layer chromatography (HPTLC) analysis of the root extracts was optimised by testing different developing solvents and visualization reagents. The presence of the sericic acid (Rf = 0.80), sericoside (Rf = 0.49) and resveratrol-3-rutinoside (Rf = 0.36) were clearly visible on the plates. There were visible variations in the concentrations of resveratrol-3-rutinoside in representative samples from the 10 populations, corresponding to the UPLC results. The powdered samples were then analysed by mid-(MIR) infrared spectroscopy. Chemometric analysis of the data revealed no definitive clustering pattern. Partial least squares-discriminant analysis (PLS-DA) calibration models were established from the MIR spectral data combined with the accurate UPLC-values, for the prediction of the sericoside (R2Y = 0.848, Q2 = 0.757, RMSEP = 2.70 mg/g) and resveratrol-3-rutinoside (R2Y = 0.794, Q2 = 0.695, RMSEP = 4.37 mg/g) concentrations in powdered root samples. The antibacterial activities of the root extracts, column fractions and isolated compounds were determined using three Gram-positive and five Gram-negative bacteria, all selected due to their ability to cause intestinal and skin disorders. Extracts and fractions containing high concentrations of sericic acid exhibited the highest activities against Klebsiella pneumoniae (ATCC 13883), Pseudomonas aeruginosa (ATCC 27858), Salmonella typhimurium (ATCC 14028), Staphylococcus aureus (ATCC 25923), Staphylococcus epidermidis (ATCC 12223) and Shigella sonnei (ATCC 9292). The pure compound (sericic acid) was highly active against S. sonnei (MIC 0.078 μg/mL), a Gram- negative bacterium. There were no variations in the activity of the crude extracts against B. cereus and P. aeruginosa, while the MIC values obtained against S. typhi were variable and ranged from 0.25 to 1.0 mg/mL. Sericoside and resveratrol-3-rutinoside did not display any activity.

The anti-oxidant activities were evaluated using DPPH (2,2-diphenyl-1-picrylhydrazyl) and reducing power assays. The anti-oxidant assays revealed that resveratrol-3- rutinoside exhibited lower activity (DPPH = 186 μg/mL; RP = 184 μg/mL) compared to the crude extract (DPPH = 22.3 μg/mL; RP = 24.4 μg/mL) and ascorbic acid (DPPH = 11.3 μg/mL, RP = 145 μg/mL). Sericic acid and sericoside did not display any anti- oxidant activities. The variation in the anti-oxidant activities (4.58 to 26.0 μg/mL) of the samples from different populations was an indication of chemical variability. A toxicity study of the raw powdered plant material was conducted using vervet monkeys (Chlorocebus pygerythrus). Biochemical analysis (liver function tests, kidney function tests and hematology), physical and physiological examinations were conducted. The subjects were fed a normal diet supplemented with T. sericea root powder (2.14 g/kg per day) for 120 days, where after the diet was returned to normal (washout) for another 30 days. The treatment groups presented with elevated serum enzymes at Week 4, followed by the reduction of the elevated serum enzymes levels at Week 12. These results indicate short-term hepatotoxic effects, followed by hepatoprotective activity. Reduction of the serum glucose at Week 4 suggests hypoglycemic potential. However, elevated serum creatinine levels indicated possible nephrotoxicity. In conclusion, this study has indicated the variability in the chemical constituents of the roots of T. sericea, which affects the antibacterial and anti-oxidant activities. Sericic acid, resveratrol-3-rutinoside, and sericoside were, for the first time, identified as biomarkers that can be used for the quality control of raw root material to be used in herbal products. Sericic acid was also found to be the main antibacterial constituent of the roots. The hepatoprotective, nephrotoxic and hematotoxic effects observed in monkeys to which the root powder had been administered is cause for concern.