INTRODUCTION

Evaluation of natural products for their chemical components is challenging due to the inherent diversity of their chemical composition. Separation techniques and their application to evaluate specific chemical components of natural products is  an important  aspect which permits accurate characterization and quantification. It has been reported that almost 80% of known natural substances are non-volatile and thermolabile, which makes high performance liquid chromatography (HPLC) with ultra violet (UV) detection the most commonly used analytical technique for natural products (1). Photo diode array (PDA) detectors enhance the versatility of the analysis by allowing multi-wavelength detection of compounds based on their distinct chromophore   active   regions.   PDA   is   thus

 

reports on the use of HPLC for qualitative analysis of herbal medicines, more importantly for the fingerprinting of South African traditional medicines (3). In addition, chromatographic fingerprinting has been widely accepted and recommended by various regulatory authorities such as WHO (4), US-FDA (5), EMEA (6) and

MHRA (7) to assess the consistency of batch to batch dosage forms containing phytochemical components of the harvested plants.

In the current international regulatory scenario, qualitative  and  quantitative analytical methods are considered mandatory. Even though Sceletium plants have been relatively well- researched for their chemical constituents, validated methods for quantitative analysis of the various alkaloid components found in Sceletiumspecies have not yet been reported in the published literature.

particularly useful for multi-component sample                                                                                   

analysis since one sample can  usually provide

sufficient   data   to   assess   the   purity   of   the individual   components   (2).   There   are   many

 

Corresponding   Author:   Isadore   Kanfer,   Division   of

Pharmaceutics,  Faculty  of  Pharmacy,  Rhodes  University, Grahamstown, South Africa. Email: I.Kanfer@ru.ac.za

Smith et al. (8) reported a qualitative method using gas chromatography-mass spectrometry (GC-MS) attached to a nitrogen-phosphorus detector (NPD). The GC-MS analysis was carried out on a DB-5 capillary column of 0.25 mm internal diameter (i.d), which was temperature programmed from 230°C to 260°C at 1°C per minute. The detector was maintained at 350°C. The method described the separation of 4΄-O- demethylmesembrenol, mesembrine and mesembrenone.

Gericke et al. (9), in their patent application, reported three methods using GC and one using a HPLC method. In all three methods, the samples were prepared by dissolving the extracts in a minimum quantity of methanol. The first method is mentioned as a “fast system” for “large number of samples”. The method uses a DB-1 fused silica capillary column 30 m x 0.25 mm i.d, with helium as carrier gas. The column temperature was maintained isothermally at 200°C for 15 minutes and then programmed at 100°C/min to 300°C. The injector and the flame ionization detector were maintained at 230°C and 300°C respectively (9). A second method is mentioned as “high resolution analyses” for “selected samples”. The method also uses a DB-1 column and helium as carrier gas, similar to the first method. The column temperature was isothermally maintained at 150°C for 15 minutes and then programmed at 60°C/ min to 320°C. The injector and the NPD were maintained at 230°C and 300°C respectively (9). A third method mentioned is for GC-MS analysis, with the column,  carrier gas and the temperatures maintained similar to the second method. The MS conditions were not described (9). The patent also includes an HPLC method using a reversed phase system (C18 column). The mobile phase comprised 30% trimethylamine solution (1% in water) and 70% acetonitrile solution (60% in water) at a flow rate of 1 ml per minute. The PDA detector was set for two channels, one at 280±30 nm (Channel A) and the other at 292±10 nm (Channel B). The method suggests that “for concentrations below 0.05 mg per ml, channel A is more accurate” (9).

The  South  African  National  Biodiversity

Institute (SANBI), South African  Medical Research Council (MRC) and the University of Western Cape (UWC) published a collaborative paper describing another HPLC method on the internet as a part of a monograph project for herbal plants. It was suggested that the method could be applied to perform fingerprinting analysis of medicinal plant extracts. The HPLC

 

method which was obtained by personal communication from the authors, uses methanol and 1% acetic acid in water mixed over a 25 minute gradient program, passing through a C18 column and dual channel detection at 280 and 325nm. A typical HPLC chromatogram of a methanol extract and the retention times of the major compounds purported to elute at 2.38, 3.15,

4.89 and 7.84 was displayed on the chromatogram (10).

No validation data were reported in any of the

analytical methods described for Sceletium. Furthermore, the use of UV wavelengths in the range of 280 nm and 325 nm for detection are not optimum for most of the Sceletium alkaloids where,    except    for    Δ7mesembrenone    (λmax

=290nm), their λmax is at 228nm.

Hence the main objective was to develop a

simple, rapid, accurate, precise and reproducible HPLC method that can be applied for the qualitative and quantitative analysis for the quality control of Sceletium plant material and associated products containing such alkaloids.

METHODS

 

Reagents and Materials

HPLC grade methanol, acetonitrile (ACN) (Romil Ltd. Cambridge, Great Britain) and lab grade ammonium hydroxide 25% solution (Associated Chemical Enterprises (Pty) Ltd. Southdale, South Africa) was used. Water was purified in a Milli- Q® system and Millex HV®  hydrophilic PVDF

0.45 μm membrane filters (Millipore, Bedford, USA) were used for filtration. Sceletium reference compounds mesembrine, mesembrenone, Δ7mesembrenone were isolated from plant material and the compounds mesembranol and epimesembranol were prepared by hydrogenation of mesembrine. Mesembrine hydrochloride was prepared from mesembrine base. All compounds were qualified by 1H, 13C and 2D NMR and mass spectroscopy. Sceletium plant samples were obtained as gift samples from cultivators.

Instrumentation

Method validation was performed on  a HPLC SpectraSYSTEM P2000 pump connected to an AS1000 auto sampler and UV1000 variable- wavelength UV detector (Thermo Separation Products, Riviera Beach, FL, USA) was used to monitor the compounds at 228nm. The output was interfaced to LCQ software (Finnigan). NMR analysis was performed on a Bruker Advance DRX    400    MHz    NMR    Spectrophotometer (Rheinstetten, Germany). The method development was carried out on an Alliance 2690 HPLC connected to a PDA detector 2996 (Waters Corporation, Milford, MA, USA). The alkaloids were separated on two different HPLC columns, Luna® C18 (2), 5 μm, 150 mm x 4.6 mm i.d. and

Hypersil®     150   x   4.6   mm   i.d   C18     column

(Phenomenex®, Torrence, CA, USA). Analytical balances, Type AG 135 and a Micro Balance MX-

5 (Mettler Toledo, Switzerland) were used for weighing standards and samples. An electronic pipette model 71050XET (Biohit PLC, Helsenki, Finland) was used to transfer standard and sample solutions for dilutions.

Method development

Initially, the method described in the patent literature (9) was used to observe the profiles of Sceletium plant extract. The resulting chromatographic profile detected with a PDA detector in the range of 400 to 200 nm indicated many unresolved peaks which eluted with in 5 minutes and could not be identified since reference standards for each individual alkaloid were not available. Since the alkaloids are basic molecules, a mobile phase comprising of 100 mM solution of ammonium acetate in water and ACN 70:30 (v:v) was initially used   to separate the

 

alkaloid constituents. The methanol extract showed distinct peaks at RT 5.278, 7.054, 7.668 and 9.543 with UV spectra showing maxima at 229±5 nm, 278±2 nm and 284 nm (Figure 1).

Whereas, using a mobile phase consisting of relatively high salt concentration 100mM ammonium acetate solution facilitated identification of the relevant alkaloid components, the resulting  peaks eluted  with relatively poor peak shape. Hence, attempts were made to exclude the use of buffer salts and develop a mobile phase using a higher pH to resolve the various Sceletium alkaloids and which would be beneficial for use with LC-MS.

The alkaloids, mesembrine, mesembrenone and ∆7mesembrenone were isolated and their structures confirmed by NMR. The HPLC peaks

were identified based on the retention times of the prepared reference compounds.

Subsequently, the optimum conditions involved the use of a mobile phase comprising of water:ACN:ammonium hydroxide solution (25%) mixed in a ratio of 70:30:0.01 (v:v:v) and was tested under isocratic conditions using a Hypersil®   C18 column. The three alkaloids were

well resolved with ∆7mesembrenone at RT 5.242,

mesembrenone at RT 7.579 and mesembrine at RT 10.894 minutes (Figure 2).

 

 

 

The isocratic conditions when applied to the analysis of mesembranol and epimesembranol, posed two problems. Firstly, the epimesembranol peak was retained on the column for an extended period of over 35 minutes (Figure 3) and secondly, separation between mesembranol and mesembrine could only be achieved with an increase of water content to 72 parts. Even though separation was achieved, the resolution and the peak shape were not ideal due to peak tailing (Figure 4).

Sceletium plant powder was analyzed using the above mentioned isocratic conditions which

 

separated ∆7mesembrenone (RT 5.981 min), an unknown compound (RT 6.421 min), mesembrenone (RT 8.774 min), mesembranol (RT 10.502 min) and mesembrine (12.577 min) (Figure 5).

Even though the developed isocratic system resolved the alkaloidal components, the RTs and the peak shape prompted changes to be made to the mobile phase which involved the use of gradient system. In this manner the RTs of the identified alkaloids could be manipulated to optimize peak shapes and also to reduce the RT of epimesembranol.

 

Figure 3. HPLC-PDA chromatogram of mesembranol and epimesembranol using a mobile phase comprising of water:ACN:ammonium hydroxide solution mixed in a ratio of 70:30:0.01 (v:v:v). UV spectra for each peak obtained on-line using a PDA detector are depicted above the chromatogram.

 

 

 

 

 

 

RESULTS

Confirmation of peak identity

The identity of each relevant alkaloidal peak in the standard chromatogram (Figure 6) was confirmed by the respective spectral data ( 1H, 13C and 2D NMR and mass spectroscopy as well as x- ray crystallography for the crystalline compounds, mesembrine hydrochloride and mesembranol, the others being semi-solid/liquids) obtained for each of the mesembrine–alkaloids used as reference standards.

 

 

Linearity

Calibration curves were constructed by plotting the peak area of each alkaloid versus the concentration corresponding to that alkaloid on each of three days. The curves obtained were found to be linear with determination coefficients better than 0.99 and in one instance, 0.9858 (Table 1).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Epimesembranol

 

Day 1

Day 2

Day 3

 

y = 32.126x + 11333 y = 29.952x + 15173 y = 32.581x + 13702

 

R2 = 0.9925

R2 = 0.9921

R2 = 0.9964

 

 

 

 

 

 

 

 

 

 

Low                 (+1.76)1 3.962                              3.89 (±0.05)              96.20              2.96              (+1.84) 4.04               4.00 (±0.07)               99.00            1.75            (+1.92) 4.22                        4.31 (±0.05)                   102.13                       1.22                3.00

High                 (+8.80) 11.00               10.73 (±0.13)             97.50              1.50             (+9.20) 11.40             10.92 (±0.24)              95.80            2.20           (+9.60) 11.90                         11.69 (±0.12)               98.70                          1.10                1.50

Mesembranol content (μg/ml)

 

Observed content (μg/ml)

%

Recovery

 

% RSD (n=3)

Mesembranol content (μg/ml)

 

Observed content (μg/ml)

%

Recovery

 

% RSD (n=3)

Mesembranol content (μg/ml)

 

Observed content (μg/ml)

 

%

Recovery

 

% RSD (n=3)

 

Recovery

% RSD (n=3)