- Email: [email protected]
Affinity chromatography with immobilized DNA stationary phase for biological fingerprinting analysis of traditional Chinese medicines
Journal of Chromatography A, 1154 (2007) 132–137
Affinity chromatography with immobilized DNA stationary phase for biological fingerprinting analysis of traditional Chinese medicines Xingye Su, Lianghai Hu, Liang Kong, Xiaoyuan Lei, Hanfa Zou ∗ National Chromatographic R.&A. Centre, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China Received 6 February 2007; accepted 14 March 2007 Available online 19 March 2007
Abstract A stationary phase for high performance affinity chromatography with immobilization of DNA onto silica gel was prepared and characterized. The effect of the ionic strength, concentration of Mg2+ , EDTA and CH3 CN in the mobile phase on the retention of alkaloids were investigated. With this stationary phase, biological fingerprinting analysis of traditional Chinese medicines (TCMs) Coptis chinensis Franch and Rheum palmatum L. was performed with both one-dimensional (1-D) and two-dimensional (2-D) chromatography. The 1-D chromatography was performed with isocratic and gradient elution and 2-D chromatography was developed with immobilized DNA column combined with silica monolithic ODS column. It was found that 7 compounds in Coptis chinensis Franch including berberine, palmatine and jatrorrhizine, 14 compounds in Rheum palmatum L. including aloe-emodin, rhein, emodin, chrysophannol-8-O-glucophranoside and physionl-8-O-glucophranoside were active in binding to the immobilized DNA. © 2007 Elsevier B.V. All rights reserved. Keywords: Affinity chromatography; Biological fingerprinting analysis; Immobilized DNA stationary phase; Traditional Chinese medicines
1. Introduction Affinity chromatography, as a powerful method for probing small molecule-biomacromolecule interactions by immobilizing either of them on the solid support, has applied successfully in rapid selection and purification of compounds from complex mixtures [1]. Affinity chromatography with immobilized DNA as stationary phase was firstly prepared by Gilham [2] by attaching the oligo thymidylic acid (dT) and subsequently other single strands including oligo dA, dC, dG, and dU [3] on the cellulose. In 1990s, through covalent binding DNA oligomers onto the macroporous silica, the immobilized DNA stationary phase was employed in high performance affinity chromatography [4–7]. In further, enzymatic synthesized, PCR produced duplex DNA was immobilized onto silica support [8]. These immobilized DNA stationary phase has been used in the polynucleotide separations [9,10] and purification [5], and polynucleotide binding protein purification [11,12]. In recent years, aptamers immobilized stationary phases have been developed for numerous
∗
Corresponding author. Tel.: +86 411 84379610; fax: +86 411 84379620. E-mail address: [email protected] (H. Zou).
0021-9673/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2007.03.050
analytical applications [13–23]. Aptamers are highly structured single-stranded DNA (or RNA) molecules that are able to bind a ligand with high affinity and specificity [24–26]. This kind of stationary phases have been used in recognition, separation and purification of various molecules including proteins, peptides, drugs, and so on. TCMs, as an important kind of natural products, are gaining increasingly attentions for their long clinic experience of over 4000 years and integrated theory system for diagnosis and treatment [27]. TCMs are complex libraries usually containing up to hundreds even thousands of small molecules. However, only a few of them are responsible for the pharmaceutical and/or toxic effects [28], which make the screening and analysis of the bioactive compounds extremely difficult. Among the strategies developed for the bioactive components analysis in TCMs, the chromatographic methods have attracted much interest because of its advantage in high performance separation. However, there is little correlation between the chromatographic retention of the compounds and their bioactivities in conventional HPLC. To overcome this problem, biological fingerprinting analysis (BFA) based on the chromatographic and MS methods was proposed to apply in the multiple bioactive compounds screening from TCMs [29,30]. BFA is defined as the chromatograms and spectra
X. Su et al. / J. Chromatogr. A 1154 (2007) 132–137
of the complex small molecular system carrying with the information of their interaction with the biomolecules [31]. Affinity chromatography with immobilized biomolecules as stationary phase, as an important strategy in BFA, has been applied successfully in the bioactive compounds screening and analysis in TCMs with immobilized plasma proteins HSA [32–34], ␣1 -acid glycoprotein [28] and liposome [35] stationary phases in affinity chromatography. DNA is the molecular target of many antimicrobial, antiviral and antitumour active drugs [36]. The adduction of genotoxic carcinogens to DNA is believed to be the first step in chemically induced carcinogenesis [37,38]. It is essential to achieve a better understanding of their interactions in elucidating the molecular basis for the potent therapeutic or toxic activities of the compounds. In this paper, natural DNA was immobilized on silica gel as stationary phase for one-dimensional (1-D) and twodimensional (2-D) chromatography and BFA of TCMs Coptis chinensis Franch and Rheum palmatum L. was performed for screening and analysis of the DNA binding active compounds.
133
2. Experimental
chloroform extraction as previously described [40]. The standards of berberine chloride, palmatine chloride, jatrorrhizine chloride, rac-tetrahydropalmatine (THP), emodin, aloe-emodin and rhein were purchased from the National Institute for the Control of Pharmaceutical and Biological Products. (−)-THP was from Beijing Tianli Co., Ltd. (Beijing, China). (+)-THP was prepared from the racemic THP by HPLC on the chiral column packed with 5 m covalently bonded cellulose tris-(3,5dimethylphenyl carbamate) chiral stationary phase (CSP) under the nonaqueous mobile phase of hexane/isopropanol (60/40) at flow rate of 1.0 ml/min. The preparative chiral column (5 m, ˚ and the aminopropyl silica were prepared in house. 300 A) EDC (1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide) was purchased from Acros Organics (NJ, USA). Ultra pure grade Tris was purchased from Amresco (Solon, OH, USA). The HPLC gradient grade CH3 CN was from Merck (Darmstadt, Germany). Distilled water was further purified by a Milli-Q system (Millipore, Bedford, MA, USA); other chemicals were of analytical grade. The extracts of TCMs Coptis chinensis Franch (root) and Rheum palmatum L. were prepared as reported previously [29,30].
2.1. Instruments
2.4. Preparation of DNA immobilized stationary phase
The 2-D LC system was set up as previously reported [39]. It consists of two LC-10ATvp, two LC-10ADvp pumps (Shimadzu, Kyoto, Japan), an SPD-M 10Avp diode array detector (Shimadzu, Kyoto, Japan), and a two-position, ten-port valve with two loops (both 500 l). The 2-D LC system was controlled by a computer running a custom program written in-house with visual C++ 6.0 software (Microsoft Corp., Redmond, WA, USA) and the chromatographic data were collected with a data acquisition board (National Chromatographic R&A Center, Dalian, China). The 1-D HPLC system consists of two LC-10ATvp pumps (Shimadzu, Kyoto, Japan), a Rheodyne-type injector valve with a 10 L loop, a WatersTM 996 photodiode array detector (Waters, Milford, MA, USA) and a Millennium 32 workstation (Waters, Milford, MA, USA). LC-MS detection and analysis was performed on an APCIMS detector (Shimadzu, Kyoto, Japan).
The DNA immobilized silica was prepared with the procedure described by Rasmussen et al. [41] with modifications. Briefly, 25 mg ct-DNA in 20 ml 10 mM 1-methyl imidazole was sonicated for 30 min under an ice bath with a JY92-IIsonifier (Scientz Biotechnique Co., Ltd, Ningbo, China) to reduce the DNA length. To the DNA solution, 1.8 g aminopropyl silica ˚ was added. After mixed homogeneously, 2.7 ml (5 m, 300 A) fresh-made 200 mM EDC in 10 mM 1-methyl imidazole was added. Then the mixture was allowed to react at 50 ◦ C for 5 h. After washed by the mobile phase, the DNA immobilized silica was packed in the stainless steel column followed by rinsing with the mobile phase until the absorption peak at 260 nm on the UV spectrum disappeared. The column was allowed to equilibrate overnight prior to use. 3. Results and discussion 3.1. Characterization of the column
2.2. HPLC and MS conditions The mobile phase for the immobilized DNA column was 20 mM Tris–HCl (pH 7.4) containing NaCl, MgCl2 , EDTA and CH3 CN with various concentrations and that for silica monolithic ODS column was CH3 CN/H2 O. LC-MS was in negative ion detection mode. The APCI probe voltage was 1 800 V; the nebulizing gas flow was 2.5 l/min; the APCI, CDL and block temperature was 400, 250 and 200 ◦ C, respectively. The mass range [m/z] was from 100 to 800 and the scan rate was 2 s/scan. 2.3. Reagents and chemicals The calf thymus DNA (ct-DNA) was purchased from Sigma (St. Louis, MO, USA) and deproteinized with phenol
DNA has a maximal UV absorption at 260 nm. Therefore, DNA loading can be estimated via adsorption by subtraction of the recovered fraction of all wash steps from the amount used for reaction [7]. If done carefully, this procedure could be quite accurate [6]. Here, the DNA loading was determined as 11.25 ± 0.93 (n = 5) mg/g silica. Fig. 1 shows the chromatograms of berberine, a compound binding on DNA [29], on the DNA and the aminopropyl immobilized column, respectively. It can be seen that berberine has a great retention on the DNA immobilized column while little on aminopropyl immobilized column. Because the affinity interaction with DNA is responsible for the retention of berberine while the strong positive charge that berberine carries causes an electrostatic repulsion with the protonated amino group on the
134
X. Su et al. / J. Chromatogr. A 1154 (2007) 132–137
Fig. 1. Chromatograms of berberine on (a) aminopropyl and (b) the DNA immobilized stationary phase packed column. Chromatographic conditions: column, 50 mm × 4.6 mm I.D.; mobile phase, 20 mM Tris–HCl buffer (pH 7.4) containing 20 mM NaCl, 5 mM MgCl2 , and 2 mM EDTA; flow rate, 1.0 ml/min; detection wavelength, 345 nm.
stationary phase. Tetrahydropalmatine (THP) is a chiral alkaloid with binding activity to DNA [40]. As demonstrated in our previous work [40], THP enantiomers were able to be separated for their different affinity with DNA. All of those results demonstrated that the DNA molecules have been well immobilized on the silica gel and the separations were due to the interaction of the solute with DNA molecules. With intra-day RSD of 0.4% (n = 5) and inter-day RSD of 0.9% (n = 5) in five consecutive days, respectively, for the retention times, the DNA immobilized column showed good repeatability. With continuous pumping of the mobile phase, no peak or shift on baseline was observed at wavelength of 260 nm indicating that the immobilized DNA was stable under experimental conditions. However, due to the degradation of DNA caused by many factors such as the ubiquitous existence of DNAase in the environment, hydro-organic or high salt eluent conditions, the retention capability and the efficiency of the column decrease after a long time use or store. The column was stored at 4 ◦ C to slower the degradation.
Fig. 2. Effect of mobile phase composition on the retention factors of palmatine, berberine and jatrorrhizine. The mobile phases was 20 mM Tris–HCl buffer (pH 7.4) containing (a) 10 mM MgCl2 , 2 mM EDTA and NaCl varied from 20 to 200 mM; (b) 20 mM NaCl, 2 mM EDTA and MgCl2 varied from 2.5 to 25 mM; (c) 20 mM NaCl, 5 mM MgCl2 and EDTA varied from 2 to 9 mM; and (d) 20 mM NaCl, 5 mm MgCl2 , 2 mM EDTA and CH3 CN varied from 0 to 15%; other conditions are the same as in Fig. 1.
X. Su et al. / J. Chromatogr. A 1154 (2007) 132–137
135
3.2. Influence of mobile phase composition on separation The surface of the stationary phase was negative charged because of the phosphate groups in DNA molecules. Thus, the retention of the solutes is assumed to be partly from the ion exchange interaction. We investigate the influence of ionic strength of the mobile phase on retention factors of the solutes. As shown in Fig. 2a, the retention factors of the compounds berberine, palmatine and jatrorrhizine decreased slightly as the concentration of NaCl in the mobile phase increased from 20 to 200 mM. It indicates that the ion exchange mechanism contributes little on the retention. The stationary phase was stable with the elution of the mobile phase containing up to 200 mM NaCl. It was reported that Mg2+ bound on DNA and stabilized the duplex through binding on the phosphate moiety [13]. Therefore, the Mg2+ in the mobile phase may have great influence on the retention times of the solutes. With the addition of MgCl2 of various concentrations in the mobile phase, the influence of Mg2+ was determined and the results were shown in Fig. 2b. When Mg2+ concentration increased from 2.5 to 25 mM, the retention factors of the solutes decreased greatly. EDTA also affects the retention because of its complexation with Mg2+ . As shown in Fig. 2c, when the concentration of EDTA increases from 2 to 9 mM, the retention factors of the solutes also increase. CH3 CN has a great effluence on the retention. As seen in Fig. 2d, as CH3 CN concentration in the mobile phase increased from 0 to 15%, the retention factors of the solutes decreased greatly. The stationary phase showed to be stable and very little bleeding of DNA molecules with the elution of the mobile phase containing CH3 CN up to 15%. However, after a continuous elution with the mobile phase containing CH3 CN for a long time, the decrease in the column efficiency and the retention factors were also observed as compare to that without CH3 CN elution, which means that CH3 CN accelerated the denaturation and bleeding of DNA on the stationary phase. 3.3. Separation of TCM extracts with 1-D and 2-D chromatography Since the retention of the solutes on this immobilized DNA column was mainly contributed by their affinity with DNA, the retention factors of the solutes on the affinity column were an important parameter to indicate their interaction with DNA. Fig. 3a shows the separation of the extracts of Coptis chinensis Franch, which is a kind of TCMs containing DNAbinding compounds berberine, palmatine and jatrorrhizine [29]. There are seven peaks with retention factors greater than 20 in the chromatograms. Three of them were identified as palmatine, berberine and jatrorrhizine, respectively. Besides these three compounds, there were four unidentified compounds have greater retention on the column. With the comparison of their peak area percentage and UV-vis spectra, they showed to be corresponding with those unidentified compounds reported in our previous work [29]. It should be noted that the elution order of the three alkaloids berberine, palmatine and jatrorrhizine corresponds to their binding constants [29].
Fig. 3. Chromatograms of Coptis chinensis Franch extract on the immobilized DNA column with (a) isocratic and (b) gradient elution. Chromatographic conditions: column, 100 mm × 4.6 mm I.D.; mobile phase, 20 mM Tris–HCl buffer (pH 7.4) containing 20 mM NaCl, and 2 mM EDTA and (a) 5 mM MgCl2 , 3% CH3 CN and (b) MgCl2 from 5 to 50 mM and CH3 CN from 0 to 15% in 40 min; Peak identifications: (1) palmatine; (2) berberine and (3) jatrorrhizine; other conditions are the same as in Fig. 1.
Since the concentration of Mg2+ and CH3 CN in the mobile phase has great influence on the retention of the solutes, gradient elution with increasing Mg2+ and CH3 CN concentration was performed on the separation of the complex sample to improve the efficiency. As seen Fig. 3b, the extract of Coptis chinensis Franch was separated on the DNA column with Mg2+ and CH3 CN concentration in the mobile phase increasing from 5 to 50 mM and 0 to 15% in 40 min, respectively. Comparing with that obtained from isocratic elution, the chromatogram under gradient elution showed better separation efficiency, especially for the late-eluting peaks such as p6, which can hardly be discerned from the baseline with isocratic elution. It can be seen that through the gradient separation, the detection sensitivity of solutes was improved evidently. The extracts of many TCMs are extremely complex. Their separation on immobilized DNA column was not satisfactory. As an example, Rheum palmatum L. was separated on this column with its chromatogram shown in Fig. 4. It can be seen that the numerous peaks could not be separated efficiently even under the gradient elution. To improve the separation, comprehensive 2-D HPLC with immobilized DNA column as the first dimension and silica monolithic ODS column as the second dimension
136
X. Su et al. / J. Chromatogr. A 1154 (2007) 132–137
Fig. 4. Chromatogram of Rheum palmatum L. extract on the DNA immobilized column. Chromatographic conditions: column, 100 mm × 4.6 mm I.D.; mobile phase, 20 mM Tris–HCl buffer (pH 7.4) containing 20 mM NaCl, and 2 mM EDTA and MgCl2 with gradient elution (5 mM in 0–8 min, 5–50 mM in 8–30 min); flow rate, 1.0 ml/min; detection wavelength, 280 nm.
was performed. In this mode of separation, the complex sample was first separated based on their affinity interaction with DNA and automatically, the fractions collected in every 500 l were separated on the silica monolithic ODS column based on the reversed-phase mechanism. Since the separation mechanisms of the two dimensions are orthogonal, those compounds that co-eluted in the first dimension were expected to be separated more efficiently on the ODS column. The 2-D chromatograms for extract of Rheum palmatum L. were shown in Fig. 5. It can be seen that in the first ten cycles, the spots were thick dotted, which means that most components have weak interaction with the DNA stationary phase and thus were eluted first without separation. The late-eluting fractions on the DNA column were separated efficiently on the silica monolithic ODS column. Fig. 6 shows the extracted chromatograms on silica monolithic ODS column for cycles 8 and 43, it can be seen that the single fraction on the first dimension could be separated into more than ten peaks on the second dimension. Due to the relatively high concentration of inorganic salt in the mobile phase of the first dimension, the online MS detection of the active compounds was difficult to perform in the
Fig. 5. Two-dimensional chromatograms for Rheum palmatum L. extract with (a) 2-D and (b) 3-D view. Chromatographic conditions: first dimensional column, 100 × 4.6 mm I.D. immobilized DNA column; mobile phase, 20 mM Tris–HCl buffer (pH 7.4) containing 20 mM NaCl, and 2 mM EDTA and MgCl2 with gradient elution (5 mM in 0–80 min, 5–50 mM in 80–300 min); flow rate, 0.1 ml/min; second dimensional column, 50 mm × 4.6 mm I.D. silica monolithic ODS column; mobile phase, CH3 CN/H2 O; gradient elution, 0–3.5 min, 0–50% CH3 CN, 3.5–5.0 min, 0% CH3 CN; detection wavelength, 280 nm. Cycle time was 5 min.
Fig. 6. Chromatograms on the silica monolithic ODS column for the extracted fraction of the (a) 8th and (b) 43rd cycle eluted from the immobilized DNA column.
X. Su et al. / J. Chromatogr. A 1154 (2007) 132–137
second dimension. To identify these active compounds, the eluates from the first dimension were collected every 10 cycles into six fractions manually. After being desalted and concentrated, the fractions were analyzed by reversed-phase LC with online MS detection. Totally 14 compounds in the latter four fractions were identified by MS and PDA. Among these compounds, three were identified as aloe-emodin, rhein and emodin, respectively by comparing UV and MS spectra with standards. Two were primarily identified as chrysophannol8-O-glucophranoside and physionl-8-O-glucophranoside by comparing their UV and MS spectra with those reported previously [30,34,42]. Since these five compounds all had great retention on the first dimension, they are expected to be active in DNA binding, which was accordant with our previous report [30]. 4. Conclusions The DNA immobilized affinity HPLC column was prepared and applied to biological fingerprinting analysis of TCM extracts. It was demonstrated that the retention of the solutes was mainly based on their affinity with DNA. Analysis of TCMs with 1-D and 2-D chromatography was performed to probe the interaction of multiple compounds in TCM extracts with the immobilized DNA. It showed to be an effective alternative in screening and analyzing the multiple DNA binding active compounds in the complex samples such as natural products. Acknowledgement Financial supports from the National Natural Sciences Foundation of China (No.20075032), the China State Key Basic Research Program Grant (001CB510202), the China State HighTech Program Grant (2001AA233031-4) and the Knowledge Innovation program of DICP to Prof. Hanfa Zou are gratefully acknowledged. References [1] R.J. Tian, S.Y. Xu, X.Y. Lei, W.H. Jin, M.L. Ye, H.F. Zou, TrAC, Trends Anal. Chem. 24 (2005) 810. [2] P.T. Gilham, J. Am. Chem. Soc. 86 (1964) 4982. [3] P.T. Gilham, Methods Enzymol. 21 (1971) 191. [4] T.A. Goss, M. Bard, H.W. Jarrett, J. Chromatogr. 508 (1990) 279. [5] T.A. Goss, M. Bard, H.W. Jarrett, J. Chromatogr. 588 (1991) 157. [6] L.R. Massom, H.W. Jarrett, J. Chromatogr. 600 (1992) 221. [7] H.W. Jarrett, J. Chromatogr. 618 (1993) 315. [8] L.R. Solomon, L.R. Massom, H.W. Jarrett, Anal. Biochem. 203 (1992) 58.
137
[9] N. Tanner, Methods Enzymol. 180 (1989) 25. [10] J. Sambrook, T. Frich, T. Maniatis, Molecular Cloning-A Laboratory Manual, second ed., Cold Spring Harbor, NY, 1989, p. 7.1. [11] G. Vlatakis, V. Bouriotis, Anal. Biochem. 195 (1991) 352. [12] R. Blanks, L.W. McLaughlin, Nucleic Acids Res. 16 (1998) 10283. [13] Q. Deng, I. German, D. Buchanan, R.T. Kennedy, Anal. Chem. 73 (2001) 5415. [14] M. Michaud, E. Jourdan, A. Villet, A. Ravel, C. Grosset, E. Peyrin, J. Am. Chem. Soc. 865 (2003) 8672. [15] Q. Deng, C.J. Watson, R.T. Kennedy, J. Chromatogr. A 1005 (2003) 123. [16] M. Michaud, E. Jourdan, C. Rovelet, A. Villet, A. Ravel, C. Grosset, E. Peyrin, Anal. Chem. 76 (2004) 1015. [17] T.U. Vo, L.B. McGown, Electrophoresis 25 (2004) 1230. [18] C. Andre, A. Xicluna, Y. Guillaume, Electrophoresis 26 (2005) 3274. [19] A. Brumbt, C. Ravelet, C. Grosset, A. Ravel, A. Villet, E. Peyrin, Anal. Chem. 77 (2005) 1993. [20] C. Ravelet, R. Boulkedid, A. Ravel, C. Grosset, A. Villet, F. Jennifer, E. Peyrin, J. Chromatogr. A 1076 (2005) 62. [21] W.J. Chung, M.S. Kim, S. Cho, S.S. Park, J.H. Kim, Y.K. Kim, B.G. Kim, Y.S. Lee, Electrophoresis 26 (2005) 694. [22] A.C. Connor, L.B. McGown, J. Chromatogr. A 1111 (2006) 115. [23] N. Rupcich, R. Nutiu, Y.F. Li, Angew. Chem. Ed. 45 (2006) 3295. [24] D.S. Wilson, J.W. Szostak, Annu. Rev. Biochem. 68 (1999) 611. [25] M. Famulok, G. Mayer, M. Blind, Acc. Chem. Res. 33 (2000) 591. [26] S.E. Osborne, I. Matsumura, A.D. Ellington, Curr. Opin. Chem. Biol. 1 (1997) 5. [27] X.D. Huang, L. Kong, X. Li, X.G. Chen, M. Guo, H.F. Zou, J. Chromatogr. B 812 (2004) 71. [28] H.L. Wang, H.F. Zou, L. Kong, J.Y. Ni, J. Basic Clin. Physil. Pharmacol. 11 (2000) 155. [29] X.Y. Su, L. Kong, X. Li, X.G. Chen, M. Guo, H.F. Zou, J. Chromatogr. A 1076 (2005) 118. [30] X.Y. Su, L. Kong, X. Li, X.G. Chen, M. Guo, H.F. Zou, J. Comb. Chem. 8 (2006) 544. [31] X.Y. Su, L. Kong, X.Y. Lei, L.H. Hu, M.L. Ye, H.F. Zou, Mini-Rev. Med. Chem. 7 (2007) 87. [32] H.L. Wang, L. Kong, H.F. Zou, J.Y. Ni, Y.K. Zhang, Chromatographia 50 (1999) 439. [33] H.L. Wang, L. Kong, H.F. Zou, J.Y. Ni, Y.K. Zhang, Chin. J. Chromatogr. A 17 (1999) 123. [34] L.H. Hu, X. Li, S. Feng, L. Kong, X.Y. Su, X.G. Chen, F. Qin, M.L. Ye, H.F. Zou, J. Sep. Sci. 29 (2006) 881. [35] X.Q. Mao, L. Kong, Q. Luo, Z.X. Li, H.F. Zou, J. Chromatogr. B 779 (2002) 331. [36] U. Pindur, G. Fischer, Curr. Med. Chem. 3 (1996) 379. [37] E.C. Miller, Cancer Res. 38 (1978) 1479. [38] K. Hemminki, Carcinogenesis 14 (1993) 2007. [39] X.G. Chen, L. Kong, X.Y. Su, H.J. Fu, J.Y. Ni, R.H. Zhao, H.F. Zou, J. Chromatogr. A 1040 (2004) 169. [40] X.Y. Su, F. Qin, L. Kong, J.J. Ou, C.H. Xie, H.F. Zou, J. Chromatogr. B 845 (2007) 174. [41] S.R. Rasmussen, M.R. Larsen, S.E. Rasmussen, Anal. Biochem. 198 (1991) 138. [42] X. Yang, Z. Gu, C. Ma, H. Masao, N. Tsuneo, Chin. Herbal Med. 29 (1998) 5.
Affinity chromatography with immobilized DNA stationary phase for biological fingerprinting analysis of traditional Chinese medicines Xingye Su, Lianghai Hu, Liang Kong, Xiaoyuan Lei, Hanfa Zou ∗ National Chromatographic R.&A. Centre, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China Received 6 February 2007; accepted 14 March 2007 Available online 19 March 2007
Abstract A stationary phase for high performance affinity chromatography with immobilization of DNA onto silica gel was prepared and characterized. The effect of the ionic strength, concentration of Mg2+ , EDTA and CH3 CN in the mobile phase on the retention of alkaloids were investigated. With this stationary phase, biological fingerprinting analysis of traditional Chinese medicines (TCMs) Coptis chinensis Franch and Rheum palmatum L. was performed with both one-dimensional (1-D) and two-dimensional (2-D) chromatography. The 1-D chromatography was performed with isocratic and gradient elution and 2-D chromatography was developed with immobilized DNA column combined with silica monolithic ODS column. It was found that 7 compounds in Coptis chinensis Franch including berberine, palmatine and jatrorrhizine, 14 compounds in Rheum palmatum L. including aloe-emodin, rhein, emodin, chrysophannol-8-O-glucophranoside and physionl-8-O-glucophranoside were active in binding to the immobilized DNA. © 2007 Elsevier B.V. All rights reserved. Keywords: Affinity chromatography; Biological fingerprinting analysis; Immobilized DNA stationary phase; Traditional Chinese medicines
1. Introduction Affinity chromatography, as a powerful method for probing small molecule-biomacromolecule interactions by immobilizing either of them on the solid support, has applied successfully in rapid selection and purification of compounds from complex mixtures [1]. Affinity chromatography with immobilized DNA as stationary phase was firstly prepared by Gilham [2] by attaching the oligo thymidylic acid (dT) and subsequently other single strands including oligo dA, dC, dG, and dU [3] on the cellulose. In 1990s, through covalent binding DNA oligomers onto the macroporous silica, the immobilized DNA stationary phase was employed in high performance affinity chromatography [4–7]. In further, enzymatic synthesized, PCR produced duplex DNA was immobilized onto silica support [8]. These immobilized DNA stationary phase has been used in the polynucleotide separations [9,10] and purification [5], and polynucleotide binding protein purification [11,12]. In recent years, aptamers immobilized stationary phases have been developed for numerous
∗
Corresponding author. Tel.: +86 411 84379610; fax: +86 411 84379620. E-mail address: [email protected] (H. Zou).
0021-9673/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2007.03.050
analytical applications [13–23]. Aptamers are highly structured single-stranded DNA (or RNA) molecules that are able to bind a ligand with high affinity and specificity [24–26]. This kind of stationary phases have been used in recognition, separation and purification of various molecules including proteins, peptides, drugs, and so on. TCMs, as an important kind of natural products, are gaining increasingly attentions for their long clinic experience of over 4000 years and integrated theory system for diagnosis and treatment [27]. TCMs are complex libraries usually containing up to hundreds even thousands of small molecules. However, only a few of them are responsible for the pharmaceutical and/or toxic effects [28], which make the screening and analysis of the bioactive compounds extremely difficult. Among the strategies developed for the bioactive components analysis in TCMs, the chromatographic methods have attracted much interest because of its advantage in high performance separation. However, there is little correlation between the chromatographic retention of the compounds and their bioactivities in conventional HPLC. To overcome this problem, biological fingerprinting analysis (BFA) based on the chromatographic and MS methods was proposed to apply in the multiple bioactive compounds screening from TCMs [29,30]. BFA is defined as the chromatograms and spectra
X. Su et al. / J. Chromatogr. A 1154 (2007) 132–137
of the complex small molecular system carrying with the information of their interaction with the biomolecules [31]. Affinity chromatography with immobilized biomolecules as stationary phase, as an important strategy in BFA, has been applied successfully in the bioactive compounds screening and analysis in TCMs with immobilized plasma proteins HSA [32–34], ␣1 -acid glycoprotein [28] and liposome [35] stationary phases in affinity chromatography. DNA is the molecular target of many antimicrobial, antiviral and antitumour active drugs [36]. The adduction of genotoxic carcinogens to DNA is believed to be the first step in chemically induced carcinogenesis [37,38]. It is essential to achieve a better understanding of their interactions in elucidating the molecular basis for the potent therapeutic or toxic activities of the compounds. In this paper, natural DNA was immobilized on silica gel as stationary phase for one-dimensional (1-D) and twodimensional (2-D) chromatography and BFA of TCMs Coptis chinensis Franch and Rheum palmatum L. was performed for screening and analysis of the DNA binding active compounds.
133
2. Experimental
chloroform extraction as previously described [40]. The standards of berberine chloride, palmatine chloride, jatrorrhizine chloride, rac-tetrahydropalmatine (THP), emodin, aloe-emodin and rhein were purchased from the National Institute for the Control of Pharmaceutical and Biological Products. (−)-THP was from Beijing Tianli Co., Ltd. (Beijing, China). (+)-THP was prepared from the racemic THP by HPLC on the chiral column packed with 5 m covalently bonded cellulose tris-(3,5dimethylphenyl carbamate) chiral stationary phase (CSP) under the nonaqueous mobile phase of hexane/isopropanol (60/40) at flow rate of 1.0 ml/min. The preparative chiral column (5 m, ˚ and the aminopropyl silica were prepared in house. 300 A) EDC (1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide) was purchased from Acros Organics (NJ, USA). Ultra pure grade Tris was purchased from Amresco (Solon, OH, USA). The HPLC gradient grade CH3 CN was from Merck (Darmstadt, Germany). Distilled water was further purified by a Milli-Q system (Millipore, Bedford, MA, USA); other chemicals were of analytical grade. The extracts of TCMs Coptis chinensis Franch (root) and Rheum palmatum L. were prepared as reported previously [29,30].
2.1. Instruments
2.4. Preparation of DNA immobilized stationary phase
The 2-D LC system was set up as previously reported [39]. It consists of two LC-10ATvp, two LC-10ADvp pumps (Shimadzu, Kyoto, Japan), an SPD-M 10Avp diode array detector (Shimadzu, Kyoto, Japan), and a two-position, ten-port valve with two loops (both 500 l). The 2-D LC system was controlled by a computer running a custom program written in-house with visual C++ 6.0 software (Microsoft Corp., Redmond, WA, USA) and the chromatographic data were collected with a data acquisition board (National Chromatographic R&A Center, Dalian, China). The 1-D HPLC system consists of two LC-10ATvp pumps (Shimadzu, Kyoto, Japan), a Rheodyne-type injector valve with a 10 L loop, a WatersTM 996 photodiode array detector (Waters, Milford, MA, USA) and a Millennium 32 workstation (Waters, Milford, MA, USA). LC-MS detection and analysis was performed on an APCIMS detector (Shimadzu, Kyoto, Japan).
The DNA immobilized silica was prepared with the procedure described by Rasmussen et al. [41] with modifications. Briefly, 25 mg ct-DNA in 20 ml 10 mM 1-methyl imidazole was sonicated for 30 min under an ice bath with a JY92-IIsonifier (Scientz Biotechnique Co., Ltd, Ningbo, China) to reduce the DNA length. To the DNA solution, 1.8 g aminopropyl silica ˚ was added. After mixed homogeneously, 2.7 ml (5 m, 300 A) fresh-made 200 mM EDC in 10 mM 1-methyl imidazole was added. Then the mixture was allowed to react at 50 ◦ C for 5 h. After washed by the mobile phase, the DNA immobilized silica was packed in the stainless steel column followed by rinsing with the mobile phase until the absorption peak at 260 nm on the UV spectrum disappeared. The column was allowed to equilibrate overnight prior to use. 3. Results and discussion 3.1. Characterization of the column
2.2. HPLC and MS conditions The mobile phase for the immobilized DNA column was 20 mM Tris–HCl (pH 7.4) containing NaCl, MgCl2 , EDTA and CH3 CN with various concentrations and that for silica monolithic ODS column was CH3 CN/H2 O. LC-MS was in negative ion detection mode. The APCI probe voltage was 1 800 V; the nebulizing gas flow was 2.5 l/min; the APCI, CDL and block temperature was 400, 250 and 200 ◦ C, respectively. The mass range [m/z] was from 100 to 800 and the scan rate was 2 s/scan. 2.3. Reagents and chemicals The calf thymus DNA (ct-DNA) was purchased from Sigma (St. Louis, MO, USA) and deproteinized with phenol
DNA has a maximal UV absorption at 260 nm. Therefore, DNA loading can be estimated via adsorption by subtraction of the recovered fraction of all wash steps from the amount used for reaction [7]. If done carefully, this procedure could be quite accurate [6]. Here, the DNA loading was determined as 11.25 ± 0.93 (n = 5) mg/g silica. Fig. 1 shows the chromatograms of berberine, a compound binding on DNA [29], on the DNA and the aminopropyl immobilized column, respectively. It can be seen that berberine has a great retention on the DNA immobilized column while little on aminopropyl immobilized column. Because the affinity interaction with DNA is responsible for the retention of berberine while the strong positive charge that berberine carries causes an electrostatic repulsion with the protonated amino group on the
134
X. Su et al. / J. Chromatogr. A 1154 (2007) 132–137
Fig. 1. Chromatograms of berberine on (a) aminopropyl and (b) the DNA immobilized stationary phase packed column. Chromatographic conditions: column, 50 mm × 4.6 mm I.D.; mobile phase, 20 mM Tris–HCl buffer (pH 7.4) containing 20 mM NaCl, 5 mM MgCl2 , and 2 mM EDTA; flow rate, 1.0 ml/min; detection wavelength, 345 nm.
stationary phase. Tetrahydropalmatine (THP) is a chiral alkaloid with binding activity to DNA [40]. As demonstrated in our previous work [40], THP enantiomers were able to be separated for their different affinity with DNA. All of those results demonstrated that the DNA molecules have been well immobilized on the silica gel and the separations were due to the interaction of the solute with DNA molecules. With intra-day RSD of 0.4% (n = 5) and inter-day RSD of 0.9% (n = 5) in five consecutive days, respectively, for the retention times, the DNA immobilized column showed good repeatability. With continuous pumping of the mobile phase, no peak or shift on baseline was observed at wavelength of 260 nm indicating that the immobilized DNA was stable under experimental conditions. However, due to the degradation of DNA caused by many factors such as the ubiquitous existence of DNAase in the environment, hydro-organic or high salt eluent conditions, the retention capability and the efficiency of the column decrease after a long time use or store. The column was stored at 4 ◦ C to slower the degradation.
Fig. 2. Effect of mobile phase composition on the retention factors of palmatine, berberine and jatrorrhizine. The mobile phases was 20 mM Tris–HCl buffer (pH 7.4) containing (a) 10 mM MgCl2 , 2 mM EDTA and NaCl varied from 20 to 200 mM; (b) 20 mM NaCl, 2 mM EDTA and MgCl2 varied from 2.5 to 25 mM; (c) 20 mM NaCl, 5 mM MgCl2 and EDTA varied from 2 to 9 mM; and (d) 20 mM NaCl, 5 mm MgCl2 , 2 mM EDTA and CH3 CN varied from 0 to 15%; other conditions are the same as in Fig. 1.
X. Su et al. / J. Chromatogr. A 1154 (2007) 132–137
135
3.2. Influence of mobile phase composition on separation The surface of the stationary phase was negative charged because of the phosphate groups in DNA molecules. Thus, the retention of the solutes is assumed to be partly from the ion exchange interaction. We investigate the influence of ionic strength of the mobile phase on retention factors of the solutes. As shown in Fig. 2a, the retention factors of the compounds berberine, palmatine and jatrorrhizine decreased slightly as the concentration of NaCl in the mobile phase increased from 20 to 200 mM. It indicates that the ion exchange mechanism contributes little on the retention. The stationary phase was stable with the elution of the mobile phase containing up to 200 mM NaCl. It was reported that Mg2+ bound on DNA and stabilized the duplex through binding on the phosphate moiety [13]. Therefore, the Mg2+ in the mobile phase may have great influence on the retention times of the solutes. With the addition of MgCl2 of various concentrations in the mobile phase, the influence of Mg2+ was determined and the results were shown in Fig. 2b. When Mg2+ concentration increased from 2.5 to 25 mM, the retention factors of the solutes decreased greatly. EDTA also affects the retention because of its complexation with Mg2+ . As shown in Fig. 2c, when the concentration of EDTA increases from 2 to 9 mM, the retention factors of the solutes also increase. CH3 CN has a great effluence on the retention. As seen in Fig. 2d, as CH3 CN concentration in the mobile phase increased from 0 to 15%, the retention factors of the solutes decreased greatly. The stationary phase showed to be stable and very little bleeding of DNA molecules with the elution of the mobile phase containing CH3 CN up to 15%. However, after a continuous elution with the mobile phase containing CH3 CN for a long time, the decrease in the column efficiency and the retention factors were also observed as compare to that without CH3 CN elution, which means that CH3 CN accelerated the denaturation and bleeding of DNA on the stationary phase. 3.3. Separation of TCM extracts with 1-D and 2-D chromatography Since the retention of the solutes on this immobilized DNA column was mainly contributed by their affinity with DNA, the retention factors of the solutes on the affinity column were an important parameter to indicate their interaction with DNA. Fig. 3a shows the separation of the extracts of Coptis chinensis Franch, which is a kind of TCMs containing DNAbinding compounds berberine, palmatine and jatrorrhizine [29]. There are seven peaks with retention factors greater than 20 in the chromatograms. Three of them were identified as palmatine, berberine and jatrorrhizine, respectively. Besides these three compounds, there were four unidentified compounds have greater retention on the column. With the comparison of their peak area percentage and UV-vis spectra, they showed to be corresponding with those unidentified compounds reported in our previous work [29]. It should be noted that the elution order of the three alkaloids berberine, palmatine and jatrorrhizine corresponds to their binding constants [29].
Fig. 3. Chromatograms of Coptis chinensis Franch extract on the immobilized DNA column with (a) isocratic and (b) gradient elution. Chromatographic conditions: column, 100 mm × 4.6 mm I.D.; mobile phase, 20 mM Tris–HCl buffer (pH 7.4) containing 20 mM NaCl, and 2 mM EDTA and (a) 5 mM MgCl2 , 3% CH3 CN and (b) MgCl2 from 5 to 50 mM and CH3 CN from 0 to 15% in 40 min; Peak identifications: (1) palmatine; (2) berberine and (3) jatrorrhizine; other conditions are the same as in Fig. 1.
Since the concentration of Mg2+ and CH3 CN in the mobile phase has great influence on the retention of the solutes, gradient elution with increasing Mg2+ and CH3 CN concentration was performed on the separation of the complex sample to improve the efficiency. As seen Fig. 3b, the extract of Coptis chinensis Franch was separated on the DNA column with Mg2+ and CH3 CN concentration in the mobile phase increasing from 5 to 50 mM and 0 to 15% in 40 min, respectively. Comparing with that obtained from isocratic elution, the chromatogram under gradient elution showed better separation efficiency, especially for the late-eluting peaks such as p6, which can hardly be discerned from the baseline with isocratic elution. It can be seen that through the gradient separation, the detection sensitivity of solutes was improved evidently. The extracts of many TCMs are extremely complex. Their separation on immobilized DNA column was not satisfactory. As an example, Rheum palmatum L. was separated on this column with its chromatogram shown in Fig. 4. It can be seen that the numerous peaks could not be separated efficiently even under the gradient elution. To improve the separation, comprehensive 2-D HPLC with immobilized DNA column as the first dimension and silica monolithic ODS column as the second dimension
136
X. Su et al. / J. Chromatogr. A 1154 (2007) 132–137
Fig. 4. Chromatogram of Rheum palmatum L. extract on the DNA immobilized column. Chromatographic conditions: column, 100 mm × 4.6 mm I.D.; mobile phase, 20 mM Tris–HCl buffer (pH 7.4) containing 20 mM NaCl, and 2 mM EDTA and MgCl2 with gradient elution (5 mM in 0–8 min, 5–50 mM in 8–30 min); flow rate, 1.0 ml/min; detection wavelength, 280 nm.
was performed. In this mode of separation, the complex sample was first separated based on their affinity interaction with DNA and automatically, the fractions collected in every 500 l were separated on the silica monolithic ODS column based on the reversed-phase mechanism. Since the separation mechanisms of the two dimensions are orthogonal, those compounds that co-eluted in the first dimension were expected to be separated more efficiently on the ODS column. The 2-D chromatograms for extract of Rheum palmatum L. were shown in Fig. 5. It can be seen that in the first ten cycles, the spots were thick dotted, which means that most components have weak interaction with the DNA stationary phase and thus were eluted first without separation. The late-eluting fractions on the DNA column were separated efficiently on the silica monolithic ODS column. Fig. 6 shows the extracted chromatograms on silica monolithic ODS column for cycles 8 and 43, it can be seen that the single fraction on the first dimension could be separated into more than ten peaks on the second dimension. Due to the relatively high concentration of inorganic salt in the mobile phase of the first dimension, the online MS detection of the active compounds was difficult to perform in the
Fig. 5. Two-dimensional chromatograms for Rheum palmatum L. extract with (a) 2-D and (b) 3-D view. Chromatographic conditions: first dimensional column, 100 × 4.6 mm I.D. immobilized DNA column; mobile phase, 20 mM Tris–HCl buffer (pH 7.4) containing 20 mM NaCl, and 2 mM EDTA and MgCl2 with gradient elution (5 mM in 0–80 min, 5–50 mM in 80–300 min); flow rate, 0.1 ml/min; second dimensional column, 50 mm × 4.6 mm I.D. silica monolithic ODS column; mobile phase, CH3 CN/H2 O; gradient elution, 0–3.5 min, 0–50% CH3 CN, 3.5–5.0 min, 0% CH3 CN; detection wavelength, 280 nm. Cycle time was 5 min.
Fig. 6. Chromatograms on the silica monolithic ODS column for the extracted fraction of the (a) 8th and (b) 43rd cycle eluted from the immobilized DNA column.
X. Su et al. / J. Chromatogr. A 1154 (2007) 132–137
second dimension. To identify these active compounds, the eluates from the first dimension were collected every 10 cycles into six fractions manually. After being desalted and concentrated, the fractions were analyzed by reversed-phase LC with online MS detection. Totally 14 compounds in the latter four fractions were identified by MS and PDA. Among these compounds, three were identified as aloe-emodin, rhein and emodin, respectively by comparing UV and MS spectra with standards. Two were primarily identified as chrysophannol8-O-glucophranoside and physionl-8-O-glucophranoside by comparing their UV and MS spectra with those reported previously [30,34,42]. Since these five compounds all had great retention on the first dimension, they are expected to be active in DNA binding, which was accordant with our previous report [30]. 4. Conclusions The DNA immobilized affinity HPLC column was prepared and applied to biological fingerprinting analysis of TCM extracts. It was demonstrated that the retention of the solutes was mainly based on their affinity with DNA. Analysis of TCMs with 1-D and 2-D chromatography was performed to probe the interaction of multiple compounds in TCM extracts with the immobilized DNA. It showed to be an effective alternative in screening and analyzing the multiple DNA binding active compounds in the complex samples such as natural products. Acknowledgement Financial supports from the National Natural Sciences Foundation of China (No.20075032), the China State Key Basic Research Program Grant (001CB510202), the China State HighTech Program Grant (2001AA233031-4) and the Knowledge Innovation program of DICP to Prof. Hanfa Zou are gratefully acknowledged. References [1] R.J. Tian, S.Y. Xu, X.Y. Lei, W.H. Jin, M.L. Ye, H.F. Zou, TrAC, Trends Anal. Chem. 24 (2005) 810. [2] P.T. Gilham, J. Am. Chem. Soc. 86 (1964) 4982. [3] P.T. Gilham, Methods Enzymol. 21 (1971) 191. [4] T.A. Goss, M. Bard, H.W. Jarrett, J. Chromatogr. 508 (1990) 279. [5] T.A. Goss, M. Bard, H.W. Jarrett, J. Chromatogr. 588 (1991) 157. [6] L.R. Massom, H.W. Jarrett, J. Chromatogr. 600 (1992) 221. [7] H.W. Jarrett, J. Chromatogr. 618 (1993) 315. [8] L.R. Solomon, L.R. Massom, H.W. Jarrett, Anal. Biochem. 203 (1992) 58.
137
[9] N. Tanner, Methods Enzymol. 180 (1989) 25. [10] J. Sambrook, T. Frich, T. Maniatis, Molecular Cloning-A Laboratory Manual, second ed., Cold Spring Harbor, NY, 1989, p. 7.1. [11] G. Vlatakis, V. Bouriotis, Anal. Biochem. 195 (1991) 352. [12] R. Blanks, L.W. McLaughlin, Nucleic Acids Res. 16 (1998) 10283. [13] Q. Deng, I. German, D. Buchanan, R.T. Kennedy, Anal. Chem. 73 (2001) 5415. [14] M. Michaud, E. Jourdan, A. Villet, A. Ravel, C. Grosset, E. Peyrin, J. Am. Chem. Soc. 865 (2003) 8672. [15] Q. Deng, C.J. Watson, R.T. Kennedy, J. Chromatogr. A 1005 (2003) 123. [16] M. Michaud, E. Jourdan, C. Rovelet, A. Villet, A. Ravel, C. Grosset, E. Peyrin, Anal. Chem. 76 (2004) 1015. [17] T.U. Vo, L.B. McGown, Electrophoresis 25 (2004) 1230. [18] C. Andre, A. Xicluna, Y. Guillaume, Electrophoresis 26 (2005) 3274. [19] A. Brumbt, C. Ravelet, C. Grosset, A. Ravel, A. Villet, E. Peyrin, Anal. Chem. 77 (2005) 1993. [20] C. Ravelet, R. Boulkedid, A. Ravel, C. Grosset, A. Villet, F. Jennifer, E. Peyrin, J. Chromatogr. A 1076 (2005) 62. [21] W.J. Chung, M.S. Kim, S. Cho, S.S. Park, J.H. Kim, Y.K. Kim, B.G. Kim, Y.S. Lee, Electrophoresis 26 (2005) 694. [22] A.C. Connor, L.B. McGown, J. Chromatogr. A 1111 (2006) 115. [23] N. Rupcich, R. Nutiu, Y.F. Li, Angew. Chem. Ed. 45 (2006) 3295. [24] D.S. Wilson, J.W. Szostak, Annu. Rev. Biochem. 68 (1999) 611. [25] M. Famulok, G. Mayer, M. Blind, Acc. Chem. Res. 33 (2000) 591. [26] S.E. Osborne, I. Matsumura, A.D. Ellington, Curr. Opin. Chem. Biol. 1 (1997) 5. [27] X.D. Huang, L. Kong, X. Li, X.G. Chen, M. Guo, H.F. Zou, J. Chromatogr. B 812 (2004) 71. [28] H.L. Wang, H.F. Zou, L. Kong, J.Y. Ni, J. Basic Clin. Physil. Pharmacol. 11 (2000) 155. [29] X.Y. Su, L. Kong, X. Li, X.G. Chen, M. Guo, H.F. Zou, J. Chromatogr. A 1076 (2005) 118. [30] X.Y. Su, L. Kong, X. Li, X.G. Chen, M. Guo, H.F. Zou, J. Comb. Chem. 8 (2006) 544. [31] X.Y. Su, L. Kong, X.Y. Lei, L.H. Hu, M.L. Ye, H.F. Zou, Mini-Rev. Med. Chem. 7 (2007) 87. [32] H.L. Wang, L. Kong, H.F. Zou, J.Y. Ni, Y.K. Zhang, Chromatographia 50 (1999) 439. [33] H.L. Wang, L. Kong, H.F. Zou, J.Y. Ni, Y.K. Zhang, Chin. J. Chromatogr. A 17 (1999) 123. [34] L.H. Hu, X. Li, S. Feng, L. Kong, X.Y. Su, X.G. Chen, F. Qin, M.L. Ye, H.F. Zou, J. Sep. Sci. 29 (2006) 881. [35] X.Q. Mao, L. Kong, Q. Luo, Z.X. Li, H.F. Zou, J. Chromatogr. B 779 (2002) 331. [36] U. Pindur, G. Fischer, Curr. Med. Chem. 3 (1996) 379. [37] E.C. Miller, Cancer Res. 38 (1978) 1479. [38] K. Hemminki, Carcinogenesis 14 (1993) 2007. [39] X.G. Chen, L. Kong, X.Y. Su, H.J. Fu, J.Y. Ni, R.H. Zhao, H.F. Zou, J. Chromatogr. A 1040 (2004) 169. [40] X.Y. Su, F. Qin, L. Kong, J.J. Ou, C.H. Xie, H.F. Zou, J. Chromatogr. B 845 (2007) 174. [41] S.R. Rasmussen, M.R. Larsen, S.E. Rasmussen, Anal. Biochem. 198 (1991) 138. [42] X. Yang, Z. Gu, C. Ma, H. Masao, N. Tsuneo, Chin. Herbal Med. 29 (1998) 5.