Study on the interaction between rivanol and DNA and its application to DNA assay

SPECTROCHIMICA ACTA PART

A

Spectrochimica Acta Part A 53 (1997) 781-787

ELSEVIER

Study on the interaction between rivanol and DNA tind its application to DNA assay Wen-You Li, Jin-Gou Xu *, Xiang-Qun Guo, Qing-Zhi Zhu, Yi-Bing Zhao The Research

Luboratory

of SEDC Xiamerl

of Analytical Science jar Material and Lijk Chemistry, Uniuersitv, Xiamen 361005, People’s Republic of‘ China

Department

of Citendry.

Received 5 November 1996: accepted 3 January 1997

Abstract

Rivanol (RVN) binds to the double helical DNA with a high affinity, as deducedfrom the absorption and fluorescencespectral data. Extensive hypochromismand red shifts in the absorption spectra were observedwhen RVN binds to calf thymus DNA (CT DNA), which suggested the intercalation mechanismof RVN into DNA bases. Upon binding to DNA, the fluorescencefrom RVN wasefficiently quenchedby the DNA bases,with no shiftsin the emissionmaximum.The large increasesin the polarization upon binding to CT DNA supportedthe intercalation of RVN into the helix. Iodide quenchingstudiesshowedthat the magnitudeof KS,of the free RVN washigher than that of the bound RVN. The results of competitive binding studiesshowedthat RVN can be displacedby ethidium bromide. Thermal denaturation experimentsexhibited that the quenchingof the fluorescencefrom RVN by single strand (ssDNA) wassmallerthan that by doublestrand (dsDNA). The resultsof all above further studiesalsoproved the intercalation of RVN into DNA basestack. Quenchingof fluorescencefrom RVN by DNA can be employedfor sensitive detection Keywords:

of DNA.

The limit of detection

for CT DNA

was 16 ng ml-‘.

0 1997 Elsevier Science B.V.

Intercalation binding; Rivanol; DNA assay

1. Introduction

Several small molecules have been shown to interact with DNA at the molecular level by specific binding modes. These binding studies were driven partly by the need to understand the mechanism of anticancer drug action at the molecular level [1,2]. Several of these DNA binding studies were also directed at deciphering the -___ * Corresponding author. Fax: + 86 592 2188054: e-mail: [email protected]

chemical basis for the carcinogenicity of environmental pollutants and toxic chemicals [3,4]. A systematic investigation of the binding of antibiotics, heterocyclic cations and metal complexes with DNA has revealed several structural and electronic factors that control the DNA binding affinity and sequence specificity of small molecules [5,6]. For example, hydrophobicity, cationic charges and coordinatively active metal centers, planar heterocyclic cations and hydrogen bonding groups have been shown to affect the overall binding affinity and sequence specificity of

1386-1425,‘97/$17.00 0 1997 Elsevier Science B.V. All rights reserved. PIIS1425(97)00015-9

W.-Y. Li et al. 1 Spectrochimica Acta Part A 53 (1997) 781 -- 757

782

the interaction of small molecules with DNA [7]. The results of these various binding studies have been very useful in designing new and promising anticancer agents for clinical use [S-lo]. The DNA binding studies with drugs have also provided a molecular basis to understand the binding and DNA sequence recognition by proteins. Small molecules bind to the double helix by two dominant modes, referred to as intercalation and groove binding. Intercalation of small molecules into the helix involves the insertion of the small molecule-usually a planar aromatic cation-into the base stack of the helix [11,12]. Groove binding involves docking the thin ribbon-like molecules in the DNA minor groove, in close proximity to the sugar-phosphate backbone. Intercalated fluorescent molecules are well protected from the aqueous solvent and waterbound fluorescence quenchers [13- 151. In contrast, groove-bound fluorescent molecules may not be protected from water-bound anionic fluorescence quenchers [16]. Intercalative binding of molecules is favored by stacking interactions with the adjacent DNA bases [17], whereas electrostatic hydrogen bonding and hydrophobic interactions contribute to the stability of groove binding [18]. Hydrophobic and electrostatic interactions provide additional stabilization for the intercalated molecules. Rivanol is 6,9-diamino-2-ethoxyacridine lactate and is better known as ethacridine, its structure is given in Scheme 1. Rivanol has some pharmaceutical uses. such as mild bacteriostatic from pre-penicillin era, in abortion clinics, in veterinary practice and in various tests on antigens. Rivanol is one of acridine derivatives which are well-known intercalators. The kinetics

. CH,CHOHCOO-

Scheme

1. Structure

of rivanol.

0.40

0.30 6 0.20

0.10

i

330

360

390

mm> Fig. 1. Absorption spectra of RVN. (1) In the absence of CT DNA: (2)-(5) in the presence of CT DNA with the concentrations of 2.0, 5.0. 10.0 and 20.0 pg ml- ‘, respectively.

of the interaction between ethacridine and DNA has been studied [19]. The reaction is a two-step process, the details of which are not accurately known [19]. We chose rivanol (RVN) for the DNA binding studies to evaluate its mode of interaction with DNA and the effect of DNA binding on the photophysical properties of the bound RVN. Upon binding RVN to CT DNA, an extensive hypochromic effect and appreciable red shifts in the absorption spectra of RVN, and the efficient fluorescence quenching of RVN by the DNA bases, with no shifts in the emission maximum. were observed. The results of absorption spectra, KI quenching studies, fluorescence polarization measurements, competitive binding studies, and thermal denaturation experiments suggested that the interaction between RVN and DNA be intercalative.

W.-Y. Li et al. /Spectrochimica Acta Part A 53 (1997) 781-787

2. Experimental 2.1. Apparatus

All fluorescence measurements were made with a Hitachi 650-10s spectrofluorimeter equipped with a 125-W xenon lamp. Absorption spectra were recorded on a Shimadzu UV-240 ultraviolet-visible spectrophotometer. All pH measurements were made with a digital pH and temperature meter 631 (Extech, Boston, USA). 2.2. Reagents

Commercially prepared CT DNA, obtained from Sino-American Biotechnology Co., were directly dissolved in water at a final concentration of 100 ug ml - ’ and stored at 4°C. A RVN solution (2.0 x 10 -4”/o) was prepared by diluting 0.20 ml of RVN solution (O.lO%, Xiamen Pharmaceutical, China) into 100 ml of water and stored in the dark. All other chemicals were of analytical reagent grade. Deionized distilled water was used to prepare the solutions. A buffer solution of pH 6.0 was prepared by mixing 87.7 ml of 0.10 M NaH,PO, and 12.3 ml of 0.10 M Na,HPO+

783

the reagent blank (prepared in a similar manner without CT DNA) are allowed to incubate for 10 min. Record the fluorescence spectra of the mixed solution and the reagent blank on a Hitachi 65010s spectrofluorimeter. Measure the fluorescence intensities of the mixed solution (F) and the reagent blank (FO) with the following settings of the spectrofluorimeter: excitation wavelength (L,,), 361 nm; excitation slit (EX), 3 nm; emission wavelength (&,), 504 nm; emission slit (EM), 3 nm. Plot a calibration curve of Fe/F vs. the concentration of CT DNA. 2.3.3. Fluorescence quenching experiments

The fluorescence quenching experiments with potassium iodide were performed and the experimental data were plotted according to the SternVolmer equation

x .z

300

5 .-2

-3.3. Procedures 2.3.1. Experiments of absorption spectra

Transfer to a lo-ml standard flask 1.0 ml of buffer solution (pH 6.0) and 1.0 ml of RVN solution (0.10%). Add a known volume of CT DNA standard solution. Dilute to the volume with water and mix. Both the mixed solution and a reagent blank (prepared in a similar manner without RVN) are allowed to incubate for 10 min. Record the absorption spectra on a Shimadzu UV-240 ultraviolet-visible spectrophotometer. 2.3.2. Experiments of fluorescence spectra

Transfer to a lo-ml standard flask 1.0 ml of buffer solution (pH 6.0) and 1.0 ml of RVN solution (2.0 x 10e4%). Add a known volume of CT DNA standard solution, Dilute to the volume with water and mix. Both the mixed solution and

460

520

580

w4 Fig. 2. Fluorescence emissiom spectra of RVN. (1) in the absence of CT DNA; (2) in the presence of CT DNA (1.0 pg ml ‘). RVN at 2.0 x lo- “)/L

784

W.-Y. Li et al. /Specirockitnica

Acta Part A 53 (1997) 781-787

3. Results and discussion 3.1. Studies of absorption spectru 2.20

1.90 " . rz 1.60

1.30

1.oo 0

4

8

Concentration

12

16

of KI ( mh4 )

Fig. 3. Quenching of RVN fluorescence (2.0 x 10 W%) by KI in the absence of CT DNA (curve A) and in the presence of CT DNA with the concentrations of 0.10 pg ml-’ (curve B) and 3.0 pg ml -’ (curve C).

The absorption spectra of RVN in the presence of increasing amounts of CT DNA showed strong decreases in the peak intensities (hypochromic effect) (Fig. 1). The hypochromism was suggested to be due to a strong interaction between the electronic states of the intercalating chromophore and that of the DNA bases 120-221. Since the strength of this electronic interaction is expected to decrease as the cube of the distance of separation between the chromophore and the DNA bases [22], the observed hypochromism suggests a close proximity of the chromophore to the DNA bases. In addition to the decrease in intensity. a small red shift was also observed in the spectra. These spectral changes are consistent with the intercalation of RVN into the DNA base stack [20,21]. 3.2. Emission studies

Upon binding to DNA, the fluorescence from RVN was efficiently quenched by the DNA bases, but there is no shifts in the emission maximum (Fig. 2). where FO and F are the fluorescence intensities in the absence and in the presence of the quencher (Q). The Stern-Volmer quenching constant K,, was evaluated by linear least-squares analysis of the data in terms of the above equation. 23.4. Fluorescencepolarization experiments

The fluorescence polarization measurements were carried out on a Hitachi 650-10s spectrofluorimeter. The samples were excited at 361 nm and the fluorescence signals were monitored at 504 nm through a pair of polarizers. 23.5. Competitive binding experiments

The experiments of competitive binding between ethidium bromide and RVN for DNA were done by adding different quantities of ethidium bromide to the system with fixed concentrations of RVN and DNA, and monitoring the changes in the fluorescence intensity of RVN.

3.3. Iodide yzremhing studies

Since groove binding exposes the bound molecules to the solvent surrounding the helix much more than does the intercalation [23], the iodide quenching experiment was chosen to further establish the DNA binding affinity of RVN. According to Ref. [16], if RVN is intercalated into the helix stack, the magnitude of K,, of the free RVN should be higher than that of the bound RVN; in contrast, if RVN binds to DNA in the groove, the magnitude of K,, of the bound RVN should be higher than that of the free RVN. In aqueous solutions, iodide quenched the ffuorescence of RVN very efficiently. Addition of potassium iodide to a mixture of RVN and CT DNA resulted in decreased quenching of the fluorescence intensity (Fig. 3). K,, values for the free RVN and the bound RVN with CT DNA at 1

W.-Y.

Li et al. /Spectrochirnica

Acta Part

A 53 (1997)

785

781 -787

Table 1 Fluorescence polarization values of RVN in the presence of CT DNA” CT DNA (pg ml -‘) Polarization

0 0.0030

0.20 0.005 I

0.50 0.0058

1.0 0.011

2.0 0.019

3.0 0.022

j’ RVN concentration. 2 x 10-‘“/;1.

and 3 ug ml-’ were 81.1, 52.2 and 29.6 MP’, respectively. The results showed that the magnitude of KSy of the free RVN was higher than that of the bound RVN, which suggested the intercalation of RVN into the DNA bases.

enhanced fluorescence polarization [24]. The large increase of fluorescence polarization upon binding RVN to CT DNA supported the intercalation of RVN into the helix [24]. 3.5. Competitive binding studies

3.4. Fluorescencepolarization measurements

In the absence of DNA, the fluorescence of RVN was weakly polarized due to the rapid tumbling motion of the RVN molecule in aqueous media. However, if the RVN molecule intercalates into the helix, its rotational motion should be restricted and, hence, the fluorescence polarization of the bound chromophore should be increased (Table 1). Mere binding to the phosphate backbone or to the DNA grooves does not result in

Ethidium bromide (EB), a well known intercalator, binds to DNA with intercalation. The experiment of competitive binding between EB and RVN for DNA was performed by adding different quantities of EB to the mixture with fixed concentrations of RVN and DNA (RVN at 2.0 x 10 - ‘%, DNA at 1 ug ml - ‘). The results printed in Fig. 4 showed that RVN could be displaced by EB, which supported the above intercalation mechanism of RVN into DNA bases. 3.6. Comparison between the ejjkcts of dsDNA and ssDNA on the quenching of RVN fluorescence

260 250 x .z 2e, .-2 s 5 z r;’5 z $$ d

Double strand DNA (dsDNA) was converted into single strand DNA (ssDNA) with the open of its double helix by incubated at 100°C for 10 min and cooled in ice-water immediately. If RVN is intercalated into the helix stack, the quenching of

240

230

r

220

Table 2 Comparison between the effects of dsDNA and ssDNA on the quenching of RVN fluorescence”

210

200

Concentration of DNA (pg ml-‘)

Types of DNA

F,+F

1.0

dsDNA ssDNA dsDNA ssDNA dsDNA ssDNA

1.56 1.13 1.30 1.07 1.12 1.04

190 0

2

4

Concentration

6

0

10

12

of EB(pM)

Fig. 4. Competitive binding between RVN and ethidium bromide for DNA. RVN at 2.0 x IO-%, CT DNA at 1.0 pg ml ‘.

0.50 0.20 il RVN concentration. 2 x lo-%I.

W.-Y. Li et al. /Spectrochimica Acta Part A 53 (1997) 781-787

786

Table 3 Analytical parameters of this method

CT DNA

Linear range (pg ml-‘)

r*

LODb (ng ml-‘)

RSD’ (‘%I) (n = 7)

o-o.7 0.7-6.0

0.9970 0.9973

16

I .o

1.3

’ Y. correlation coefficient. h LOD, limit of detection. ’ RSD. relative standard deviation for seven measurements of 0.5 or 1.0 pg ml-’

DNA

the fluorescence from RVN by ssDNA would be smaller than that by dsDNA. The results of comparison experiments given in Table 2 exhibited that the quenching of the fluorescence from RVN by ssDNA was smaller than that by dsDNA, which also supported the intercalation of RVN into DNA helix.

RVN solution (2.0 x lo-‘%) was in the range of 0.5- 1.5 ml, and the F,/F values decreased outside this range of RVN concentration. Thus, 1.0 ml of RVN solution (2.0 x 10P4%) was chosen for use. The values of F,,/F showed a linear relationship with the concentrations of CT DNA. The linear concentration range, limit of detection, correlation coefficient and precision were given in Table

3.7. Application

3.

on DNA

assay

First, the effect of pH on F,/F was studied. The concentrations of RVN and DNA were maintained at 2.0 x 10 ~ 5% and 1 l.tg ml - ‘, respectively. The results indicated that the maximum F,,/F occurred in the pH range of 5.8-6.4 and the F,,/F values decreased for other pH values outside this range. In the subsequent studies, a pH of 6.0 was recommended for use. Next, the influence of RVN concentration on F,JF was investigated with constant concentration of DNA (1 pg ml - ‘) at pH 6.0. The results showed that the maximal and constant FJF was reached when the volume of

Table 4 The comparison of LOD of DNA Method Ethidium bromide 4’,6-diamidino-2-phenylindole Hoechst 33258 Eu’ +-tetracycline Tb’+-phenanthroline This method

LOD (ng ml-‘) 10 0.5

10 10 100 16

Reference P51

WI P71

P81 P91

The limit of detection (LOD) was given by the equation, LOD = K&/S, where K is a numerical factor chosen according to the confidence level desired, S, is the S.D. of the blank measurements (n = 9) and S is the sensitivity of the calibration graph. Here a value of 3 for K was used. The comparison of LOD for DNA among this method and some other commonly used methods was listed in Table 4. But Eu3 +-tetracycline is a specific regent for DNA 1281. Moreover, the optimum pH for the various methods was different. 4. Conclusion

RVN binds to the double helical DNA with a high affinity. Strong hypochromism and appreciable red shifts in the absorption spectra can be observed when RVN binds to CT DNA. Upon binding to DNA, the fluorescence of RVN was efficiently quenched by the DNA bases, with no shifts in the emission maximum. The results of absorption spectra, KI quenching studies, fluorescence polarization measurements, competitive binding studies, and thermal denaturation experiments suggested that the interaction between RVN and DNA is intercalative.

W.-Y.

Li et al. lSpectrochimica

Acknowledgements

The authors thank the China Nature Fund for financial assistance.

Science

References [l] M.J. Waring, in G.C.K. Roberts (ed.), Drug Action at the Molecular Level, Macmillan, London. 1977, pp. 167- 189. [2] M.J. Waring and K.R. Fox, in S. Neidle and M.J. Waring (eds.), Molecular Aspects of Anti-Cancer Drug Action, Macmillan, London, 1983, pp. 127-156. [3] C. Heidelberger, Ann. Rev. Biochem., 44 (1975) 79. [4] M.A. Warpehoski and L.H. Hurley, Chem. Res. Toxicol., 1 (1988) 313. [5] C.V. Kumar, A.L. Raphael and J.K. Barton, Biomolec. Struct. Dyn.. 3 (1986) 85. [6] J.K. Barton, J. Biomolec. Struct. Dyn., I (1983) 621. [7] D. Porschke. in W. Guschlbauer and W. Saenger (eds.), DNA-Ligand Interactions, Specificity and Dynamics of Protein-Nucleic Acid Interactions, Plenum, New York, 1986. p. 85. [8] M.P. Singh, T. Joseph, S. Kumar. Y. Bathini and J.W. Lown, Chem. Res. Toxicol.. 5 (1992) 597. [9] K.E. Rao and J.W. Lown, Chem. Res. Toxicol., 4 (1991) 661. [lOI J.W. Lown, Anticancer Drug Design, 3 (1988) 25. [ll] L.S. Lerman, J. Molec. Biol., 3 (1961) 18. [I?] H.M Berman and P.B. Young, Ann. Rev. Biophys. Bioeng., 10 (1981) 87.

Acta Parr

A 53 (1997)

781-787

‘81

[I 31 C.V. Kumar, J.K. Barton and N.J. Turro, J. Am. Chem. Sot., 107 (1985) 5518. [I41 J.K. Barton, C and E News, 66 (1988) 30. [I 51 C.V. Kumar and E.H. Asuncion. J. Chem. Sot., Chem. Commun., 6 (1992) 470. [16] C.V. Kumar, R.S. Turner and E.H. Asuncion. J. Photochem. Photobiol. A: Chem.. 74 (1993) 231. [17] A.M. Pyle, J.P. Rehmann, R. Meshoyrer. C.V. Kumar, N.J. Turro and J.K. Barton, J. Am. Chem. Sot.. I 11 (1989) 3051. [18] M.-K. Teng, C.A. Fredrick, N. Usmann and A.H.-J. Wang, Nucleic Acids Res., 16 (1988) 2671. [19] L.P.G. Wakelin and M.J. Waring, J. Mol. Biol., 144 (1980) 183. [20] E.C. Long and J.K. Barton, Act. Chem. Res.. 23 (lY90) 271. [21] H.M. Berman and P.R. Young, Annu. Rev. Biophys. Bioeng.. 10 (1981) 87. [22] C. Cantor and P.R. Schimmel, Biophysical Chemistry, Vol. 2, W.H. Freeman, San Francisco. 1980, p. 398. [23] A.H.-J. Wang and M.-K. Teng. in C.E. Bugg and S.E. Eahck (eds.), Crystallographic and Modeling Methods in Molecular Design. Springer, New York, 1990, p. 123. [24] C.V. Kumar and E.H. Asuncion. J. Am. Chem. Sot.. I 15 (1993) 8547. [25] J.B. LePecq and C. Paoletti, Anal. Biochem.. 17 (1966) 100. [26] J. Kapuscinski and B. Skoczylas. Anal. Biochem.. 83 (1977) 252. [27] C. Labarca and K. Paigen, Anal. Biochem., 102 (1980) 344. [28] Y.X. Ci,Y.Z. Li andX.J. Liu, Anal. Chem., 67 (1995) 1785. [29] Y.X. Ci. Y.Z. Li and W.B. Chang, Anal. Chim. Acta. 248 (1991) 589.