Terraternal.it

Journal of Chromatographic Science, Vol. 39, November 2001 Determination of Astragaloside IV in Radix Astragali
(Astragalus membranaceus var. monghulicus) Using
High-Performance Liquid Chromatography with
Evaporative Light-Scattering Detection
Wenkui Li and John F. Fitzloff
Program for Collaborative Research in the Pharmaceutical Sciences, Department of Medicinal Chemistry and Pharmacognosy, College of
Pharmacy, University of Illinois at Chicago, 833 South Wood Street, Chicago, IL 60612
for AGS-IV on the column (6). Nevertheless, there is a drawback Abstract
in the application of the derivatization method (i.e., extensive A reverse-phase high-performance liquid chromatographic method
sample preparation and analysis time). It would be useful to is developed for the determination of astragaloside IV, a
develop an alternative method for the routine analysis of Radix characteristic constituent in Radix Astragali. Samples are analyzed
Astragali, which is more efficient.
by means of a reverse-phase column (Zorbax Eclipse XDB C18) using
As a mass detection method, evaporative light-scattering detec- acetonitrile and water under gradient conditions as the mobile
tion (ELSD) is based on the nebulization of liquid chromatog- phase for 30 min. An evaporative light-scattering detector is used
raphy (LC) column effluent into droplets by the nebulizing gas and set at an evaporating temperature of 43°C with a nebulizing gas
and the entrance of the resulting vapor into a temperature-con- (compressed air) pressure of 3.4 bar. The detection limit (signal-to-
trolled evaporator tube in which the evaporation of mobile phase noise ratio > 5) of astragaloside IV is 40 ng on-column.
takes place. The resulting “clouds” of solid microparticles arethen directed towards a narrow light beam. As a result, light isscattered by these microparticles and measured using a photo- Introduction
multiplier or photodiode. A plot of detector response versus ana-lyte concentration is sigmoidal, and the peak area (I) is related tothe sample size and shape (but not the chemical identity of the Radix Astragali is prepared from the roots of certain species of plants from the genus Astragalus (Leguminosae) and has beenused as a tonic, analgesic, antiseptic, and antisudorific in Chinesetraditional medicine (1). It possesses hepatoprotective, antioxida-tive, antiviral, antihypertensive, and immunostimulant activity,and these activities are well-documented and reviewed (2–4).
Astragaloside IV (AGS-IV) (Figure 1), a cycloartane-type triter-pene glycoside, has been regarded as one of its characteristic andactive constituents. The identification and quantitation of AGS-IVhas been carried out by thin-layer chromatography (TLC) (5).
Although the TLC method has been extensively employed in thequality assurance (QA)/quality control (QC) of Radix Astragaliand the products containing it (5), the method presents a chal-lenge for sensitivity and selectivity. Direct reverse-phase (RP)high-performance liquid chromatography (HPLC) with UV detec-tion could be used for the analysis of AGS-IV in Radix Astragalisamples. However, the detection of AGS-IV using UV is well-known for its insensitivity because of the weak chromophoricfunctionality of AGS-IV in the 200–210-nm region. A precolumnderivatization method has been proposed that allows an efficient Figure 1. Structure of AGS-IV.
detection using HPLC–UV with a detection limit (LOD) of 40 ng Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission.
Journal of Chromatographic Science, Vol. 39, November 2001 residual particles passing through the light beam) by the fol- sample vial. Methanol (Fisher, HPLC grade, 18 mL) was added, and the mixture was shaken and then sonicated at 25–30°C for 30min. After cooling, the mixture was filtered through filter paper (Whatman #40, Whatman Inc., Clifton, NJ) into a 250-mL round- where b is the slope of the response line, m is the mass of the com- bottom flask, and the residue was returned to the sample vial.
pound injected, and a is the response factor. As a result, plots of Another 18 mL of methanol was added, and the mixture was son- the peak area versus the analyte concentration with logarithmic icated at 25–30°C for 30 min. The extract was filtered through coordinates are linear. ELSD has been applied to a wide range of filter paper (Whatman #40) into the same round-bottom flask.
UV-transparent analytes including lipids (7,8), peptides (9), carbo- This extraction procedure was repeated once more before hydrates (10), and botanical bioactive compounds (11). We washing the residue with methanol (3 × 15 mL) while on the recently reported the determination of a marker compound, 24 filter. The combined methanol extracts were evaporated under reduced pressure at 35°C. The residue was redissolved and trans- HPLC–ELSD (12). It is interesting that the ELSD is not a tech- ferred with methanol to a 10-mL volumetric flask and diluted to nique that is widely known nor used in the QA/QC of herbal prod- volume with methanol. The sample solution was filtered through ucts. Based on a project with the aim to evaluate the application a 0.2-µm Whatman hydrophilic membrane filter into an HPLC of the ELSD in the QA/QC of dietary supplements, this study sample vial just before HPLC–UV–ELSD analysis.
describes the quantitative analysis of AGS-IV in Radix Astragali in Chromatographic conditions
a single run by HPLC–ELSD using gradient elution.
The chromatographic separations were carried out on a Zorbax Eclipse XDB C18 column (250- × 4.6-mm i.d., 5-µm particle size,PN#990967.902) (Agilent, Palo Alto, CA) protected by a Waters Experimental
Delta-Pak C18 guard column (Waters Technologies Ireland, Ltd.,Wexford, Ireland) and set at 20°C. The mobile phase used for the Materials and reagents
separation consisted of solvent A (water, deionized) and solvent B Pharmacopoeial Radix Astragali was obtained from the National (acetonitrile). The elution profile was: 0 → 30 min, 20% → 58% B; Institute for the Control of Pharmaceutical and Biological 30 → 31 min, 58% → 90% B; 31 → 35 min, 90% B (washing out); Products (Beijing, China). An AGS-IV reference standard was pro- 36 → 37 min, 90% → 20% B; 37 → 40 min, 20% B (recondi- vided by the Institute for the Advancement of Chinese Medicine, tioning). All gradient steps were linear. The flow rate was set at 1.6 Hong Kong Baptist University (Kowloon, Hong Kong), which was mL/min, the column temperature was fixed at 20°C, and the injec- originally purchased from the National Institute for the Control of tion volume was chosen to be 10 µL. The peak identification was Pharmaceutical and Biological Products. Methanol and acetoni- based on retention time and comparison with the injected trile (HPLC grade) were purchased from Fisher Scientific (Fair authentic reference standard. The peak was detected consecutively Lawn, NJ). Deionized water was obtained with an in-house Nano- with the UV and the ELSD. In UV, the detection wavelength was set pure water system (Barnstead, Newton, MA).
at 203 nm. After the UV detector, the eluent was transferred to the Apparatus
ELSD with a gain of 11, the evaporating temperature was at 43°C, A Waters 2690 Alliance HPLC system (Waters Corporation, and the nebulizing gas pressure at 3.4 bar. Prior to each run, the Milford, MA) equipped with a 996 photodiode-array UV detector, HPLC–UV–ELSD system was allowed to warm up for 20–30 min an online degasser, and an autosampler were used for solvent and the pumps were primed using the protocol suggested by the delivery and detection. After the UV detector, the column effluent manufacturer. Using a freshly prepared mobile phase, the baseline was directed to a Sedex (Alfortville, France) 75 ELSD. The was monitored until stable before the samples were run.
detector output was interfaced using a SATIN box to the Waters Reproducibility
Millennium 2000 chromatographic manager system loaded on a The precision and accuracy of the method were assessed by intra- Compaq (Houston, TX) 6400X/10000/CDS computer for data and interday validations. The variation was evaluated by injecting handling and chromatogram generation.
three sets of controls (50, 120, and 180 µg/mL, n = 3) on three sep- Preparation of standard solution
arate days. By substituting the peak area into the calibration curve In a clean, dry 10-mL volumetric flask, the AGS-IV reference equation from the same run, the measured concentrations were standard (2.0 mg) was accurately weighed and dissolved in obtained. By comparing calculated and theoretical concentrations, methanol to make a stock solution. Calibration working standard the relative errors were obtained. The coefficient of variance was solutions (4–200 µg/mL) were prepared by diluting the stock calculated by comparing the measured concentrations.
solution with methanol in appropriate quantities. Three controls The recovery was assessed by adding 100 µg of the standard to were also positioned to be in the lowest, middle, and highest 500 mg of Radix Astragali samples and extracted in a similar way regions of the calibration curve (i.e., 50, 120, and 180 µg/mL). All as the previously mentioned sample. Triplicate analysis provided working solutions were stored at –20°C and brought to room total AGS-IV, which was used to determine recovery.
Preparation of sample solution
Results and Discussion
Finely pulverized Radix Astragali roots (Astragalus mem- branaceus (Fish) Bge. var monghulicus (Bge.) Hsiao) were Optimization of the ELSD parameters
weighed (0.5 g) into a polytetrafluoroethylene-stopped 20-mL In ELSD, a constant nebulization process is important for sat- Journal of Chromatographic Science, Vol. 39, November 2001 isfactory repeatability. Several factors have influence on the standard solutions for three consecutive days. The calibration average diameter of the droplets and their distribution, which curve, log-transformed peak area (y) versus log-transformed con- include density, viscosity, and liquid surface tension. Among these centration (x), was calculated according to the least-squares factors, the nebulizer gas flow rate affects the signal response methods for AGS-IV tested with regression more than 0.998 most significantly. When the gas flow rate is too low, large droplets are formed, resulting in spikes and random noise.
However, when the gas flow rate is too high, the droplets decrease in size, which results in a decreased signal response. The The intra- and interday reproducibility of the method was eval- optimum nebulizer gas (compressed air) pressure in this workwas determined to be 3.4 bar.
The evaporating temperature is also an important parameter affecting the signal response. At low temperature, solvent evapo-ration is not complete; at high temperature, the detectorresponse is decreased, owing to the decrease in particle size by animproper vaporization of the nebulized analytes in the drift tube.
The signal-to-noise ratio (S/N) was improved when the tempera-ture was lowered to 43°C. Also, the gain in ELSD was set at 11 inorder to obtain the best sensitivity.
Chromatography
Figure 2 shows the typical LC–UV and LC–ELSD chro- matograms of AGS-IV with a retention time of approximately 16min within a 30-min gradient elution. The LOD (S/N > 5) of thedescribed method was observed for AGS-IV at 40 ng on-column inthe current assay. Because the compound has no double bond in Figure 3. Log-transformed calibration curve of AGS-IV.
the molecule, it shows poor UV absorption with an LOD of 200 ng(S/N > 5) on-column. This verifies the advantage of ELSD overnormal UV in the detection of UV transparent compounds.
Table I. Reproducibility for Three Consecutive Days
Method validation
Spiked Measured
Linearity was examined by applying the calibration working concentration concentration*
%Coefficient %Relative
(µg/mL)
(µg/mL)
of varience
* Mean ± standard deviation, n = 3.
Figure 2. Typical chromatograms of AGS-IV: (A) HPLC–ELSD with 40 ng (LOD)
of the AGS-IV reference standard on-column and (B) HPLC–UV with 40 ng of
Figure 4. Typical HPLC–ELSD chromatogram of the methanolic extract of the
the AGS-IV reference standard on-column.
sample with a retention time of 15.6 min for AGS-IV.
Journal of Chromatographic Science, Vol. 39, November 2001 uated by analyzing a set of three controls (50, 120, and 180 References
µg/mL, n = 3) on three separate days (n = 3) and calculating the %RSD and relative errors. As shown in Table I, the %RSD 1. H. Hikino, S. Fanayama, and K. Endo. Hypotensive principle of and the relative errors were found to be less than 1.48% and Astragalus and Hedysarum. Planta Med. 30: 297–302 (1976).
8.95%, respectively. The average recovery of AGS-IV was 95.8% 2. J.L. Rios and P.G. Waterman. A review of the pharmacology and tox- icology of Astragalus. Phytother. Res. 11: 411–18 (1997).
3. E. Bedir, N. Pugh, I. Calis, D.S. Pasco, and I.A. Khan. Immunostim- ulatory effects of cycloartane-type triterpene glycosides from Sample analysis
Astragalus species. Biol. Pharm. Bull. 23: 834–37 (2000).
As shown in Figure 4, three sets of Pharmacopoeial Radix 4. S. Toda and Y. Shirataki. Inhibitory effects of Astragali Radix, a crude Astragali samples were analyzed according to the method drug in Oriental medicines, on lipid peroxidation and protein oxida- described previously. The average content of AGS-IV was found tive modification by copper. J. Ethnopharmacol. 68: 331–33 (1999).
to be 0.016%. This compares favorably with a previous report (6) 5. The Chinese Pharmacopoeia. Radix Astragali. 2: 271–72 (1995).
6. M. Yao, Y. Qi, K. Bi, X. Wang, X. Luo, and C. Che. A precolumn
for A. monghulicus (Bge.) Hsiao from eight samples all grown derivatization high-performance liquid chromatographic method in different locations with content ranging between 0.03% to with improved sensitivity and specificity for the determination of astragaloside IV in Radix Astragali. J. Chromatogr. Sci. 38: 325–28
(2000).
7. T. Seppanen-Laskso, I. Laskso, H. Vanhanen, K. Kiviranta, T. Lehtimaki, and R. Hiltunen. Major human plasma lipid classes Conclusion
determined by quantitative high-performance liquid chromatog-
raphy, their variation and association with phospholipid fatty acids.
J. Chromatogr. B 754: 437–45 (2001).
An HPLC method has been developed for the determination 8. G.A. Picchioni, A.E. Watada, and B.D. Whitaker. Quantitative high- and quantitation of AGS-IV in Radix Astragali using an ELSD.
performance liquid chromatography analysis of plant phospholipids With the method, AGS-IV was successfully quantitated using and glycolipids using light scattering detection. Lipids 31: 217–21
the calibration curve with a detection limit of 40 ng on the 9. J. Bongers and T.K. Chen. RP-HPLC peptide mapping of proteins with an evaporative light scattering detector. J. Liq. Chrom. & Rel. Validation of the HPLC–ELSD method included inter- and Technol. 23: 925–33 (2000).
intraday precision and accuracy. All of the validation parameters 10. Y. Wei and M.Y. Ding. Analysis of carbohydrates in drinks by high- studied were found to have RSDs less than 8.95% and did not performance liquid chromatography with a dynamically modified show any bias in a single direction.
amino column and evaporative light scattering detection. The HPLC–ELSD method was found to be rapid, relatively inex- J. Chromatogr. A 904: 113–17 (2000).
11. F.F. Camponovo, J.L. Wolfender, M.P. Maillard, O. Potterat, and pensive, straightforward, and reproducible.
K. Hostettmann. Evaporative light scattering and thermospray mass
spectrometry: two alternative liquid chromatographic determina-
tions of ginkgolides and bilobalide in Ginkgo biloba leaf extracts and
phytopharmaceuticals. Phytochem. Anal. 6: 141–48 (1995).
Acknowledgments
12. W.K. Li and J.F. Fitzloff. Determination of 24 (R)-pseudoginsenoside F11 in North American ginseng using high performance liquid chro- The authors are grateful to Dr. H.W. Leung and Mr. C.L. Chan matography with evaporative light scattering detection. J. Pharm.
Biomed. Anal. 25: 257–65 (2001).
of the Institute for the Advancement of Chinese Medicine, HongKong Baptist University for providing the reference standard.
Manuscript accepted on September 5, 2001.

Source: http://www.terraternal.it/Files/Ag4DetectionWithHPLCUsingELSD.pdf