- Email: [email protected]
Equilibria and kinetics of the intercalation of Pt-proflavine and proflavine into calf thymus DNA
ABB Archives of Biochemistry and Biophysics 418 (2003) 63–70 www.elsevier.com/locate/yabbi
Equilibria and kinetics of the intercalation of Pt-proflavine and proflavine into calf thymus DNA Tarita Biver, Fernando Secco,* Maria Rosaria Tine, and Marcella Venturini Dipartimento di Chimica e Chimica Industriale, Universit a di Pisa, Via Risorgimento 35, 56126 Pisa, Italy Received 19 May 2003, and in revised form 22 July 2003
Abstract The interaction of the cis-platinum derivative of proflavine [{PtCl(tmen)2 }{HNC13 H7 (NHCH2 CH2 )2 }]þ (PRPt) with CTDNA is investigated by spectrophotometry and T-jump relaxation in 0.11 M NaCl, pH 7.0, and 25°C. The DNA–proflavine (PR) system is investigated under the same conditions. Static measurements indicate that base-dye interactions prevail and their analysis reveals that the site size for PRPt (n ¼ 2:6) is twice that found for PR (n ¼ 1:3). One relaxation effect is observed for the DNA/PR system and two effects for the DNA/PRPt system, the faster of them being similar to that of DNA/PR. The kinetics of the process are discussed in terms of the three-step sequence D þ S ¢ DSI ¢ DSII ¢ DSIII , where PR and the aromatic residues of PRPt intercalate into DNA by the same mechanism. The third step represents the penetration of platinum residues between base-pairs and is associated to remarkable enthalpy and entropy changes. Further mechanistic details are discussed. Ó 2003 Elsevier Inc. All rights reserved. Keywords: DNA; Metal acridines; Anticancer drugs; Intercalation; Kinetics; Equilibria
Acridines are known to interact with DNAs and double stranded RNAs mainly by intercalation [1–4]. The wide range of biological effects of intercalating drugs continues to stimulate investigations on this process, mainly in conjunction with the synthesis of new metal intercalators whose properties as drugs need to be tested. Actually, the activity of many anticancer, antimalarial, and antibacterial agents and that of aromatic carcinogens find its origin primary in intercalation [5–7]. Studies on this issue have shown that intercalation is more complex than it was supposed to be and the details of the process strongly depend on the intercalate structure. In this contest, particularly interesting appear to be bifunctional molecules bearing an aromatic residue and a metal containing residue [8–10]. The aromatic residue provides an anchorage for the molecule on the polymer chain by intercalation, whereas the metal containing residue may exert different functions. It can intercalate as well or it might remain outside of the cavity where the
* Corresponding author. Fax: +39-050-918260. E-mail address: [email protected] (F. Secco).
0003-9861/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/S0003-9861(03)00384-9
metal could interact with the polymer backbone. Studies in this sense are being performed by Barton and co-workers [11], who have synthesised special metal intercalators where the metal is intended to promote the cleavage of the phosphodiesteric bond at a selected point of the chain length. Other metal intercalators bear a platinum (II) complex appended to the aromatic residue. These substances could in principle be employed as anticancer drugs [8,12,13] and the knowledge of the mechanism of their binding to nucleic acids is of great importance, since this process does constitute a prerequisite for the subsequent slow attack of Pt(II) to the base nitrogen [14,15]. Recently, we have investigated the equilibria and kinetics of the intercalation of a cis-platinum derivative of proflavine [{PtCl(tmen)}2 {HNC13 H7 (NHCH2 CH2 )2 }]þ (PRPt)1 into double stranded poly(A) [15]. We report here on the intercalation of the same molecule into CTDNA. For comparison, the DNA–proflavine system has been investigated as well. 1
Abbreviations used: PRPt, Pt-proflavine; PR, proflavine.
64
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
Materials and methods Materials Calf thymus DNA was purchased from Pharmacia Biotech (Uppsala, Sweden), in the form of lyophilised sodium salt, dissolved in water, and sonicated as described below. Stock solutions of DNA were standardised spectrophotometrically using e ¼ 13,200 M1 cm1 at 260 nm, as obtained from the sample certificate. The DNA concentration is expressed in molarity of basepairs and will be indicated as CP . Proflavine hydrochloride (Fig. 1A) was from Sigma (St. Louis, MO) and its solutions were frequently prepared and standardised by absorbance measurements at 444 nm using e ¼ 4:2 104 M1 cm1 [16]. PRPt perchlorate (Fig. 1B) was synthesised as described elsewhere [17]. Stock solutions of PRPt were prepared by dissolving weighed amounts of the solid in doubly distilled water. The absorption spectrum and the extinction coefficient of PRPt in the visible region appear to be practically the same as those of proflavine. All the solutions were kept in the dark at 4°C and used within 2 days. Sodium nitrate was used to adjust the ionic strength, whereas sodium cacodylate (1 103 –1 102 M) was employed to keep the pH of the solutions at the value of 7.0. Doubly distilled water was used throughout. Methods Measurements of pH were made by a Radiometer Copenhagen (Copenhagen, Denmark) PHM84 pH meter equipped with a combined glass electrode. The DNA sonication was performed by an MSE-Sonyprep sonicator, by seven repeated cycles of 10 s sonication and 20 s pause, at an amplitude of 14 lm. The sonicator tip was directly introduced into the solution, this being kept in an ice bath to avoid any temperature increase due to the sonication. PAGE tests indicated that the polymer length was reduced to ca. 350 bp. Spectrophotometric binding measurements were carried out in the visible region where only the dyes absorb by using a Perkin–Elmer, € berlingen, Germany) Lambda 17 UV/Vis Gmbh (U
spectrophotometer. Fluorescence titrations were performed on a Jasco (Tokyo, Japan) FP-770 spectrofluorometer. The fluorescence was excited at kexc ¼ 444 nm and collected at kem ¼ 512 nm. The intensity of the emitted light was corrected by applying the equation Fcorr ¼ Fobs antilog ½ðAbsexc þ Absem Þ=2 [18], although under the conditions of the experiments the inner filter effect was largely reduced. Since both dyes are noticeably adsorbed on glass or quartz, all the optical experiments have been performed on polystyrene cuvettes. Blank tests ensured that the fluorescence from the plastic material was negligible under the conditions of the experiments. Both absorbance and fluorescence titrations were carried out by adding increasing amounts of DNA directly into the cell containing the dye solution. The concentration ranges are: 2 105 –5 104 M (DNA), 1 106 – 2 105 M (PR), and 5 106 –6 105 M (PRPt). The T-jump measurements were performed on a Messanlagen (G€ ottingen, Germany) instrument with Joule heating and absorbance detection, modified by inserting a polariser in front of the entrance cell window. The ability of the cell windows to depolarise the incident light was tested for several cells and found negligible for the cell used in this work [15]. The measurements were done under magic angle conditions (55°). When this work was in progress we completed the construction of a T-jump apparatus that was able to measure fast changes of fluorescence where suitable silica photodiodes (Hamamatsu, S1336, Japan) have been used in place of photomultipliers. Some additional experiments have then been performed in the fluorescence mode. The kinetic curves were collected by a Tektronix (Beaverton, OR) 2212 storage oscilloscope, transferred to a PC, and evaluated according to the method of Provencher [19]. The experiments were performed under pseudo-firstorder conditions and the concentration ranges of the reactants were the same as those used in the equilibrium measurements. The DNA/PRPt system displays two relaxation times differing by more than one-order of magnitude. Since our data storage device lacks the dual sampling scale, the experiments with DNA/PRPt have been recorded first at a high and then at a low sampling rate and the two effects were analysed separately. The temperature control was within 0.1 °C.
Results Equilibria The binding process can be represented by the apparent reaction D þ S ¢ DS Fig. 1. Molecular structures of (A) proflavine (PR) and (B) Pt-proflavine (PRPt). L–L ¼ N,N,N 0 ,N 0 –tetramethylethylenediamine.
ð1Þ
where D and DS denote the free and bound dyes, respectively, whereas S indicates the free sites on the
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
polymer. The equilibrium (1), in the absence of co-operativity, could be described by a binding constant and a site size. The data were first analysed in terms of Scatchard plots but they were found to display negative deviations from linearity at low values of r (r ¼ ½DS=CP ) which, at first sight, could suggest the occurrence of a co-operative behaviour. It should be noted, however, that the Scatchard variables, r and r=½D, are strongly correlated. This feature introduces remarkable deviations from linearity at the extremes of the r=½D vs. r plots and could lead to misinterpretation of the results. Hence, we have used a different method of analysis. First, the site size n, defined according to McGhee and von Hippel [20], has been evaluated by titrations at an ionic strength of 1 103 M, where the complex is quantitatively formed. The evaluation of n is shown in Fig. 2A for DNA/PR and in Fig. 2B for DNA/PRPt.
65
The titrating function, DW=CD , is plotted against CP =CD . Here, DW indicates a change of fluorescence, while in other cases it could indicate a change of absorbance; CD and CP denote the analytical concentrations of dye and DNA, respectively. The intersection of the two branches of the titration curve yields on the X axis the CP =CD value corresponding to n. The values of n are collected in Table 1. Second, the degree of binding, r, for each titration point was evaluated as r ¼ ½DS=CP ¼ DW=ðDU CP Þ, where DU ¼ eDS eD (absorbance) or DU ¼ uDS uD (fluorescence). A first estimate of this parameter is obtained from the amplitude of the titration curve. Third, the site concentration along the titration was determined as ½S ¼ CP f ðrÞ, being f ðrÞ ¼ n ð1nÞ ½1 nr ½1 ðn 1Þr [20,21]. Finally, the binding constant, K, of the apparent reaction (1) and DU were simultaneously obtained from a linear fit to the relationship
A
CD =DW ¼ 1=DU þ 1=ðKDUÞ 1=½S:
ð2Þ
The value of DU so obtained was then used to re-evaluate r and the procedure was repeated. Usually, four iterations were sufficient to reach convergence. Fits to Eq. (2) are shown in Fig. 3 for both the DNA/PR and DNA/PRPt systems. The values of K are collected in Table 1. The temperature dependence of K has been investi0 gated for DNA/PRPt and the value of DHapp , the apparent enthalpy of reaction (1), is reported in Table 1. The binding constant of the DNA/PRPt system depends on salt concentration as shown in Fig. 4. The trend of log K as a function of log½Naþ is linear, in agreement with the Manning and Record equation [22]. The intercept of the plot yields log K0 , which has been defined as the binding constant in the absence of electrostatic effects [22]. The slope, whose value is 1.01, corresponds to m0 w, where m0 is the number of phosphodiesteric residues occupied by one dye molecule and w is the extent of DNA charge shielded by counter-ions. Being w ¼ 0:88 [23], it follows that m0 ffi 1:2. The value of w is made by a shielding contribution (ws ) and by a contribution from ion condensation (wc ). Being for DNA w ffi wc , it turns out that m0 w represents the number of condensed sodium ions displaced by one dye molecule.
B
Fig. 2. Fluorescence titrations of the DNA/PR (A) and DNA/PRPt (B) systems at low ionic strength. I ¼ 1 103 M, pH 7.0, and T ¼ 25 °C. The two straight lines intersect to a point corresponding to the site size, n.
Table 1 Reaction parameters for the interaction of CT-DNA with proflavine (PR) and Pt-proflavine (PRPt); pH 7.0, I ¼ 0:11 M (NaCl), T ¼ 25 °C Dye
n
104 Ka (M1 )
104 Kb (M1 )
103 K1 (M1 )
K2 c
K3 c
103 k2 (s1 )
102 k2 (s1 )
k3 (s1 )
k3 (s1 )
0 DHapp (kcal mol1 )
PR PRPt
1.3 0.1 2.6 0.1
6.6 0.7 14 1
7.1 2.2 17 7
4.2 0.5 5.4 0.5
16 3 22 3
– 0.39 0.08
6.8 0.4 6.6 0.4
4.3 0.5 3.0 0.3
– 28 4
– 72 5
)5.9 0.3d )10 1e
a
From static measurements. K ¼ K1 ð1 þ K2 ð1 þ K3 ÞÞ. c From rate constants: K2 ¼ k2 =k2 , K3 ¼ k3 =k3 . d From amplitudes. e From static measurements at different temperatures. b
66
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
A
B
Fig. 3. Fluorescence titrations for the DNA/PR (A) and DNA/PRPt (B) systems. The fittings of the experimental points to Eq. (2) are shown as continuous lines. ½S ¼ CP f ðrÞ, I ¼ 0:11 M (NaCl), pH 7.0, and T ¼ 25 °C.
Fig. 4. Dependence of the DNA/PRPt equilibrium constant on salt concentration. pH 7.0, T ¼ 25 °C.
Kinetics DNA/PR The T-jump experiments with absorbance detection have been carried out at 430 nm with polarised light under magic angle conditions (55°) in order to avoid interference from polymer alignment caused by the capacitor discharge [24,25]. This precaution was maintained, although preliminary experiments revealed that at I ¼ 0:01 M (NaNO3 ) the time constant of the process of polymer deorientation is only 32 ls. This finding ensures that at the working ionic strength (0.11 M) the deorientation of the sonicated DNA is much faster and
Fig. 5. Relaxation curves for the DNA/PR system recorded by absorbance (A) and fluorescence (B) detection. I ¼ 0:11 M (NaCl), pH 7.0, and T ¼ 25 °C; (A) CP ¼ 4:6 105 M, CD ¼ 3:4 105 M; (B) CP ¼ 5:0 105 M, CD ¼ 2:5 106 M.
therefore it does not interfere with the dynamics of the binding process. This conclusion is demonstrated by the fact that relaxation curves recorded in the presence or in the absence of polariser are identical, except for a somewhat larger noise due to light absorption by the polariser. The measurements with fluorescence detection agree very well with those monitored by absorbance. Fig. 5 represents two relaxation curves recorded by absorbance detection (A) and fluorescence detection (B), respectively. In contrast with some of the previous studies on the DNA/PR system, where two chemical relaxation effects were observed [26–28], but in agreement with other investigations [29,30], here the kinetic curves are monoexponential, independent of the detection mode (absorbance or fluorescence). The dependence of the reciprocal relaxation time 1=s on F ðCÞ, a function of the reactant concentration, tends to a plateau (Fig. 6), thus revealing that the binding process could not be described by a single reaction. According to Jovin and Striker [21], F ðCÞ ¼ ½S f 0 ðrÞ½D. A two-step series model, leading to Eq. (3), appears to explain the kinetic behaviour 1=s ¼ K1 k2 F ðCÞ=ð1 þ K1 F ðCÞÞ þ k2 :
ð3Þ
The meaning of the reaction parameters, whose values are collected in Table 1, is made clear by inspecting scheme (5). DNA/PRPt This system displays two kinetic effects, as revealed by the biexponential relaxation curve shown in Fig. 7.
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
67
A
Fig. 6. Dependence of the reciprocal relaxation time, 1=s, on the reactant concentration for the DNA/PR system. The fitting of the experimental data to Eq. (3) is shown as a continuous line. F ðCÞ ¼ ½S f 0 ðrÞ½D, I ¼ 0:11 M (NaCl), pH 7.0, T ¼ 25 °C.
B
Fig. 8. Concentration dependencies of fast (A) and slow (B) relaxation times for the DNA/PRPt system. The fittings of the experimental data to Eqs. (3) and (4), respectively, are shown as continuous lines. (d), absorbance; (s), fluorescence; I ¼ 0:11 M (NaCl), pH 7.0, T ¼ 25 °C.
Fig. 7. Relaxation curve for the DNA/PRPt system (absorbance). I ¼ 0:11 M (NaCl), pH 7.0, T ¼ 25 °C.
The concentration dependence of the reciprocal fast relaxation time, 1=sf , tends to a plateau (Fig. 8A) in agreement with Eq. (3) (where now 1=s is replaced by 1=sf ). The reciprocal slow relaxation time, 1=ss , also tends to a plateau (Fig. 8B) in agreement with 1=ss ¼ ðKf k3 F ðCÞÞ=ð1 þ Kf F ðCÞÞ þ k3 ;
ð4Þ
where Kf ¼ K1 ð1 þ k2 =k2 Þ. The rate parameters, derived by least-square fits to Eqs. (3) and (4), are collected in Table 1. The behaviour of both the investigated systems could be rationalised on the basis of the general reaction scheme K1
k2
k3
k2
k3
D þ S ¢ DSI ¢ DSII ¢ DSIII
ð5Þ
where DSI , DSII , and DSIII represent different bound forms. For the DNA/PR system, the last step is lacking. According to scheme (5), the overall binding constant can be expressed as a combination of rate constants, as shown by K ¼ K1 f1 þ ðk2 =k2 Þ½1 þ ðk3 =k3 Þg: For the DNA/PR system, K ¼ K1 ½1 þ ðk2 =k2 Þ.
ð6Þ
Fig. 9. Salt concentration dependence of the kinetic effects for the DNA/PRPt system. (r), k ¼ K1 (M1 ); (j), k ¼ k2 (s1 ); and (d), k ¼ k3 (s1 ); pH 7.0, T ¼ 25 °C.
Table 1 shows that values of K from Eq. (6) agree with those evaluated from equilibria, thus proving the consistence of the results. Fig. 9 shows the salt concentration dependence of the kinetic effects. The largest variation is displayed by K1 . A plot of log K1 vs. log½Naþ is linear with a slope equal to unity, whereas plots of log k2 and log k3 vs. log½Naþ do not show any salt dependence. This finding reveals that the electrostatic effect is exerted on the first step of the binding process. Amplitude analysis DNA/PR The analysis of the amplitudes of the DNA/PR system has been performed according to [31]
68
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
ratio of the slow effect is too unfavourable to provide reliable results. Since the relaxation times of the two effects differ by more than one-order of magnitude, the fast relaxation has been associated with the same steps that are operative in the DNA/PR system, namely with the first of the two steps of scheme (5). Its amplitude, dA0I , is directly proportional to the amplitude factor CI as shown in Fig. 10B. In this case, however, the plot of dA0I vs. CI does not enable us to evaluate the reaction enthalpy, since DeI is unknown. However, if one makes the hypothesis that DeI is the same as that measured for DNA/PR, it turns out from the slope of the plot of Fig. 10B that DHI0 ¼ 5:0 kcal mol1 . This value is similar to that found for DNA/PR.
A
B
Discussion DNA/PR
Fig. 10. Amplitude analysis (A) for the DNA/PR system and (B) for the DNA/PRPt system. The fittings of the experimental data to Eq. (7) are shown as continuous lines; I ¼ 0:11 M (NaCl), pH 7.0, T ¼ 25 °C.
0 dA0 ¼ DeCDHapp dT =RT 2 ;
ð7Þ
where dA0 is the amplitude of the relaxation effect (expressed as a change of absorbance) following a jump of temperature of magnitude dT . The amplitude 1 factor [21] C ¼ ð1=½DS þ 1=½D f 0 ðrÞ=½SÞ has been evaluated, for each experiment, by using the already determined values of n and K. Fig. 10A shows that the plot of dA0 vs. C is linear, in accord with Eq. (7). Being De ¼ 2:06 104 M1 and dT ¼ 1:6 °C, the slope of the plot of Fig. 10A yields the enthalpy of the apparent reaction (1) whose value is reported in Table 1. DNA/PRPt The amplitude analysis of DNA/PRPt has been performed only on the fast effect since the signal-to-noise
This system has been widely investigated by previous authors with results apparently not converging. The principal discrepancy concerns the number of the observed kinetic effects. A number of investigators [26–28] have measured two relaxation times whereas others [29,30] report only a single relaxation time, which tends to level off at high reactant concentration. Nevertheless, the biphasic nature of the binding process has been established by all the studies far so performed. The results of the various kinetic investigations, including the present one, on the DNA/PR system are collected in Table 2. Concerning the site size, n, it can be noted that n is linked to the parameter B of the Scatchard equation by the relationship B ¼ 1=ð2n 1Þ [20]. In the present case, one obtains B ¼ 0:64 in NaCl 0.11 M, which agrees with the value that can be derived from the data of Marcandalli et al. [30] at the same salt concentration. DNA/PRPt A comparison between DNA/PRPt and DNA/PR shows that the platinum residues play a principal role in determining the differences of behaviour between the two systems. The site size for DNA/PRPt is twice that
Table 2 Reaction parameters for the binding of proflavine (PR) to DNA
CT-DNA CT-DNA CT-DNA Poly(dA-dT) CT-DNA CT-DNA
106 k1 (s1 )
103 k1 (s1 )
102 k2 (s1 )
101 k2 (s1 )
103 K1 (M1 )
K2
104 K (M1 )
0 DHapp (kcal mol1 )
14 17 28 – – –
13 0.12 11 – – –
69 33 18 36 33 68
42 20 11 8 26 43
1.1 3.3 2.5 4.6 6.7 4.2
17 16 16 46 13 16
2 – 4.3 – 9.2 6.6
)7.9 )12.9 – )0.38 – )5.9
(1) (2) (3) (2) (4) (5)
(1) I ¼ 0:2 M, T ¼ 25 °C, [26]; (2) I ¼ 0:11 M, T ¼ 17 °C, [27]; (3) I ¼ 0:24 M, T ¼ 10 °C, [28]; (4) I ¼ 0:2 M, T ¼ 25 °C, [29]; (5) I ¼ 0:11 M, T ¼ 25 °C, this work.
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
measured for DNA/PR. This result suggests that the cis-platinum moiety as well could accommodate into the DNA cavities and that this process is related to the kinetic effect with time resolution in the tenths of second. The time constant of the slow effect is far below any time constant measured for DNA/PR. Thus, this effect could not be associated to a monomolecular step where only the acridine residue is involved. It appears quite reasonable that the slowness of this effect should be ascribed to the work necessary to enlarge the polymer cavity in order to introduce the cis-platinum residues as well (step DSII ¢ DSIII ). The finding that for PRPt the value of n is twice that measured for PR could be alternatively rationalised, assuming that the platinumcontaining residue undergoes groove interaction, thus preventing the occupation of the nearest-neighbour site. The rate of interaction of groove dyes with DNA has been shown to be nearly diffusion controlled [32] and therefore is rather independent of the ligand nature. Taking for the rate constant of the binding step the value of 108 M1 s1 [33] it turns out that, in the range of DNA concentrations investigated (2 105 –5 104 M), the values of the relaxation times for step DSII ¢ DSIII would lie between 0.02 and 0.5 ms, i.e., 20–500 times faster than the experimental values. It should also be noticed that square-planar platinum complexes as [(bipy)2 Pt(en)]2þ are known to intercalate into DNA [34]. Although aware that, in the absence of any structural evidence, any conclusion about the nature of the slow binding is speculative, we tend to favour the first hypothesis on the basis of the above considerations. The enthalpy change associated with the step DSII ¢ DSIII for the DNA/PRPt system is remarkable. It can be evaluated noting that the overall equilibrium constant K is related to the individual-step equilibrium constants by Eq. (6). By differentiating this equation a relationship between the enthalpies can be obtained, 0 that is: DH30 ¼ ðDHapp DHI0 Þð1 þ K2 ð1 þ K3 ÞÞ=ðK2 K3 Þ, 0 where DHI , evaluated by the amplitude analysis, is the enthalpy of the fast normal reaction, i.e., a combination of DH10 and DH20 [35]. Thus, the enthalpy variation for the DSII ¢ DSIII step, DH30 , is found to be equal to )18 kcal M1 . The GibbÕs energy variation for the above step, DG03 ¼ RT ln K3 , is +0.56 kcal M1 ; hence, from the relationship DG03 ¼ DH30 T DS30 , it turns out that the entropy change, DS30 , is )62 cal M1 K1 . This finding would suggest that the interaction of the platinum-residues with the DNA bases is accompanied by considerable increase of rigidity. The data of Table 1 show that the affinity of PRPt for DNA is somewhat higher than that of PR. In particular, whereas the stability of the DSI complex is similar for the two systems, the value of K2 increases by 27% on changing PR to PRPt. The enhanced stability of DSII depends on k2 which decreases from 4.3 102 s1 (PR) to 3.0 102 s1
69
(PRPt). We ascribe this rate reduction to a stabilising effect of the platinum-residue positioned outside of the DNA cavity in DSII . By contrast, the presence of the platinum-residue inside of the cavity as in DSIII exerts a destabilising effect, as shown by the value of K3 which indicates that the platinum-residue stays inside of the cavity only for the 28% of the time.
Acknowledgment This work has been partially supported by MIUR with COFIN2002.
References [1] L.S. Lerman, J. Mol. Biol. 39 (1961) 461–477. [2] A. Aggarwal, S.A. Islam, R. Kuroda, S. Neidle, Biopolymers 23 (1984) 1025–1041. [3] A. Adams, Curr. Med. Chem. 9 (2002) 1667–1675. [4] L. Malinina, M.M. Soler-L opez, J. Aymamı, J.A. Subirana, Biochemistry 41 (2002) 9341–9348. [5] S. Neidle, Prog. Med. Chem. 16 (1979) 151–221. [6] H.M. Berman, P.R. Young, Annu. Rev. Biophys. Bioeng. 10 (1981) 87–114. [7] W.P. Wilson, R.L. Jones, Adv. Pharmacol. Chemother. 18 (1981) 177–222. [8] S. Lippard, Acc. Chem. Res. 11 (1978) 211–217. [9] C. Hiort, P. Lincoln, B. Norden, J. Am. Chem. Soc. 115 (1993) 3448–3454. [10] S. Arounaguiri, D. Easwaramoorthy, A. Ashokkumar, A. Dattagupta, G.M. Bhaskar, Proc. Indian Acad. Sci. (Chem. Sci.) 112 (2000) 1–17. [11] K.E. Errkila, D.T. Odom, J.K. Barton, Chem. Rev. 99 (1999) 2777–2795. [12] L.G. Marzilli, S. Ano, F.P. Intini, G. Natile, J. Am. Chem. Soc. 121 (1999) 9133–9142. [13] E.A. Boudreaux, Int. J. Quantum Chem. 83 (2001) 255–258. [14] B.E. Bowler, K.J. Ahmed, W.I. Sunquist, L.S. Hollis, E.E. Whang, S.J. Lippard, J. Am. Chem. Soc. 111 (1989) 1299– 1306. [15] C. Ciatto, M.L. DÕAmico, G. Natile, F. Secco, M. Venturini, Biophys. J. 77 (1999) 2717–2724. [16] G. Schwarz, S. Klose, W. Balthasar, Eur. J. Biochem. 12 (1970) 454–467. [17] E. Ceci, R. Cini, A. Karaulov, M.B. Hurthouse, L. Maresca, G. Natile, J. Chem. Soc. Dalton Trans. (1993) 2491–2497. [18] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Kluver Academic/Plenum Publishers, New York, 1999. [19] S.W. Provencher, J. Chem. Phys. 64 (1976) 2772–2777. [20] J.D. McGhee, P.H. von Hippel, J. Mol. Biol. 86 (1974) 469–489. [21] T.M. Jovin, G. Striker, Mol. Biol. Biochem. Biophys. 24 (1977) 245–281. [22] M.T. Record, C.F. Anderson Jr., T.M. Lohman, Q. Rev. Biophys. 11 (1978) 103–178. [23] T. Schelhorn, S. Kretz, H. Zimmermann, Cell. Mol. Biol. 38 (1992) 345–365. [24] M. Dourlent, J.F. Hogrel, C. Helene, J. Am. Chem. Soc. 96 (1974) 3398–3406. [25] F.J. Meyer-Almes, D. Porschke, Biochemistry 32 (1993) 4246– 4253. [26] H.J. Li, D.M. Crothers, J. Mol. Biol. 39 (1969) 461–477.
70
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
[27] J. Ramstein, M. Ehrenberg, R. Rigler, Biochemistry 19 (1980) 3938–3948. [28] M. Dourlent, J.F. Hogrel, Biochemistry 15 (1976) 430–436. [29] B. Marcandalli, C. Winzek, J.F. Holzwarth, Ber. Bunsenges. Phys. Chem. 88 (1984) 368–374. [30] B. Marcandalli, G. Stange, J.F. Holzwarth, J. Chem. Soc. Faraday Trans. 1 (84) (1988) 2807–2819. [31] M. Citi, F. Secco, M. Venturini, J. Phys. Chem. 22 (1988) 6399– 6404.
[32] S.Y. Breusegem, S.E. Sadat-Ebrahimi, K.T. Douglas, E.V. Bichenkova, R.M. Clegg, F.G. Loontiens, J. Med. Chem. 44 (2001) 2503–2506. [33] S.Y. Breusegem, R.M. Clegg, F.G. Loontiens, J. Mol. Biol. 315 (2002) 1049–1061. [34] S. Arnott, P.J. Bond, R. Chandrasekaran, Nature 287 (1980) 561– 563. [35] F. Secco, M. Venturini, J. Chem. Soc. Faraday Trans. 89 (1993) 719–725.
Equilibria and kinetics of the intercalation of Pt-proflavine and proflavine into calf thymus DNA Tarita Biver, Fernando Secco,* Maria Rosaria Tine, and Marcella Venturini Dipartimento di Chimica e Chimica Industriale, Universit a di Pisa, Via Risorgimento 35, 56126 Pisa, Italy Received 19 May 2003, and in revised form 22 July 2003
Abstract The interaction of the cis-platinum derivative of proflavine [{PtCl(tmen)2 }{HNC13 H7 (NHCH2 CH2 )2 }]þ (PRPt) with CTDNA is investigated by spectrophotometry and T-jump relaxation in 0.11 M NaCl, pH 7.0, and 25°C. The DNA–proflavine (PR) system is investigated under the same conditions. Static measurements indicate that base-dye interactions prevail and their analysis reveals that the site size for PRPt (n ¼ 2:6) is twice that found for PR (n ¼ 1:3). One relaxation effect is observed for the DNA/PR system and two effects for the DNA/PRPt system, the faster of them being similar to that of DNA/PR. The kinetics of the process are discussed in terms of the three-step sequence D þ S ¢ DSI ¢ DSII ¢ DSIII , where PR and the aromatic residues of PRPt intercalate into DNA by the same mechanism. The third step represents the penetration of platinum residues between base-pairs and is associated to remarkable enthalpy and entropy changes. Further mechanistic details are discussed. Ó 2003 Elsevier Inc. All rights reserved. Keywords: DNA; Metal acridines; Anticancer drugs; Intercalation; Kinetics; Equilibria
Acridines are known to interact with DNAs and double stranded RNAs mainly by intercalation [1–4]. The wide range of biological effects of intercalating drugs continues to stimulate investigations on this process, mainly in conjunction with the synthesis of new metal intercalators whose properties as drugs need to be tested. Actually, the activity of many anticancer, antimalarial, and antibacterial agents and that of aromatic carcinogens find its origin primary in intercalation [5–7]. Studies on this issue have shown that intercalation is more complex than it was supposed to be and the details of the process strongly depend on the intercalate structure. In this contest, particularly interesting appear to be bifunctional molecules bearing an aromatic residue and a metal containing residue [8–10]. The aromatic residue provides an anchorage for the molecule on the polymer chain by intercalation, whereas the metal containing residue may exert different functions. It can intercalate as well or it might remain outside of the cavity where the
* Corresponding author. Fax: +39-050-918260. E-mail address: [email protected] (F. Secco).
0003-9861/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/S0003-9861(03)00384-9
metal could interact with the polymer backbone. Studies in this sense are being performed by Barton and co-workers [11], who have synthesised special metal intercalators where the metal is intended to promote the cleavage of the phosphodiesteric bond at a selected point of the chain length. Other metal intercalators bear a platinum (II) complex appended to the aromatic residue. These substances could in principle be employed as anticancer drugs [8,12,13] and the knowledge of the mechanism of their binding to nucleic acids is of great importance, since this process does constitute a prerequisite for the subsequent slow attack of Pt(II) to the base nitrogen [14,15]. Recently, we have investigated the equilibria and kinetics of the intercalation of a cis-platinum derivative of proflavine [{PtCl(tmen)}2 {HNC13 H7 (NHCH2 CH2 )2 }]þ (PRPt)1 into double stranded poly(A) [15]. We report here on the intercalation of the same molecule into CTDNA. For comparison, the DNA–proflavine system has been investigated as well. 1
Abbreviations used: PRPt, Pt-proflavine; PR, proflavine.
64
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
Materials and methods Materials Calf thymus DNA was purchased from Pharmacia Biotech (Uppsala, Sweden), in the form of lyophilised sodium salt, dissolved in water, and sonicated as described below. Stock solutions of DNA were standardised spectrophotometrically using e ¼ 13,200 M1 cm1 at 260 nm, as obtained from the sample certificate. The DNA concentration is expressed in molarity of basepairs and will be indicated as CP . Proflavine hydrochloride (Fig. 1A) was from Sigma (St. Louis, MO) and its solutions were frequently prepared and standardised by absorbance measurements at 444 nm using e ¼ 4:2 104 M1 cm1 [16]. PRPt perchlorate (Fig. 1B) was synthesised as described elsewhere [17]. Stock solutions of PRPt were prepared by dissolving weighed amounts of the solid in doubly distilled water. The absorption spectrum and the extinction coefficient of PRPt in the visible region appear to be practically the same as those of proflavine. All the solutions were kept in the dark at 4°C and used within 2 days. Sodium nitrate was used to adjust the ionic strength, whereas sodium cacodylate (1 103 –1 102 M) was employed to keep the pH of the solutions at the value of 7.0. Doubly distilled water was used throughout. Methods Measurements of pH were made by a Radiometer Copenhagen (Copenhagen, Denmark) PHM84 pH meter equipped with a combined glass electrode. The DNA sonication was performed by an MSE-Sonyprep sonicator, by seven repeated cycles of 10 s sonication and 20 s pause, at an amplitude of 14 lm. The sonicator tip was directly introduced into the solution, this being kept in an ice bath to avoid any temperature increase due to the sonication. PAGE tests indicated that the polymer length was reduced to ca. 350 bp. Spectrophotometric binding measurements were carried out in the visible region where only the dyes absorb by using a Perkin–Elmer, € berlingen, Germany) Lambda 17 UV/Vis Gmbh (U
spectrophotometer. Fluorescence titrations were performed on a Jasco (Tokyo, Japan) FP-770 spectrofluorometer. The fluorescence was excited at kexc ¼ 444 nm and collected at kem ¼ 512 nm. The intensity of the emitted light was corrected by applying the equation Fcorr ¼ Fobs antilog ½ðAbsexc þ Absem Þ=2 [18], although under the conditions of the experiments the inner filter effect was largely reduced. Since both dyes are noticeably adsorbed on glass or quartz, all the optical experiments have been performed on polystyrene cuvettes. Blank tests ensured that the fluorescence from the plastic material was negligible under the conditions of the experiments. Both absorbance and fluorescence titrations were carried out by adding increasing amounts of DNA directly into the cell containing the dye solution. The concentration ranges are: 2 105 –5 104 M (DNA), 1 106 – 2 105 M (PR), and 5 106 –6 105 M (PRPt). The T-jump measurements were performed on a Messanlagen (G€ ottingen, Germany) instrument with Joule heating and absorbance detection, modified by inserting a polariser in front of the entrance cell window. The ability of the cell windows to depolarise the incident light was tested for several cells and found negligible for the cell used in this work [15]. The measurements were done under magic angle conditions (55°). When this work was in progress we completed the construction of a T-jump apparatus that was able to measure fast changes of fluorescence where suitable silica photodiodes (Hamamatsu, S1336, Japan) have been used in place of photomultipliers. Some additional experiments have then been performed in the fluorescence mode. The kinetic curves were collected by a Tektronix (Beaverton, OR) 2212 storage oscilloscope, transferred to a PC, and evaluated according to the method of Provencher [19]. The experiments were performed under pseudo-firstorder conditions and the concentration ranges of the reactants were the same as those used in the equilibrium measurements. The DNA/PRPt system displays two relaxation times differing by more than one-order of magnitude. Since our data storage device lacks the dual sampling scale, the experiments with DNA/PRPt have been recorded first at a high and then at a low sampling rate and the two effects were analysed separately. The temperature control was within 0.1 °C.
Results Equilibria The binding process can be represented by the apparent reaction D þ S ¢ DS Fig. 1. Molecular structures of (A) proflavine (PR) and (B) Pt-proflavine (PRPt). L–L ¼ N,N,N 0 ,N 0 –tetramethylethylenediamine.
ð1Þ
where D and DS denote the free and bound dyes, respectively, whereas S indicates the free sites on the
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
polymer. The equilibrium (1), in the absence of co-operativity, could be described by a binding constant and a site size. The data were first analysed in terms of Scatchard plots but they were found to display negative deviations from linearity at low values of r (r ¼ ½DS=CP ) which, at first sight, could suggest the occurrence of a co-operative behaviour. It should be noted, however, that the Scatchard variables, r and r=½D, are strongly correlated. This feature introduces remarkable deviations from linearity at the extremes of the r=½D vs. r plots and could lead to misinterpretation of the results. Hence, we have used a different method of analysis. First, the site size n, defined according to McGhee and von Hippel [20], has been evaluated by titrations at an ionic strength of 1 103 M, where the complex is quantitatively formed. The evaluation of n is shown in Fig. 2A for DNA/PR and in Fig. 2B for DNA/PRPt.
65
The titrating function, DW=CD , is plotted against CP =CD . Here, DW indicates a change of fluorescence, while in other cases it could indicate a change of absorbance; CD and CP denote the analytical concentrations of dye and DNA, respectively. The intersection of the two branches of the titration curve yields on the X axis the CP =CD value corresponding to n. The values of n are collected in Table 1. Second, the degree of binding, r, for each titration point was evaluated as r ¼ ½DS=CP ¼ DW=ðDU CP Þ, where DU ¼ eDS eD (absorbance) or DU ¼ uDS uD (fluorescence). A first estimate of this parameter is obtained from the amplitude of the titration curve. Third, the site concentration along the titration was determined as ½S ¼ CP f ðrÞ, being f ðrÞ ¼ n ð1nÞ ½1 nr ½1 ðn 1Þr [20,21]. Finally, the binding constant, K, of the apparent reaction (1) and DU were simultaneously obtained from a linear fit to the relationship
A
CD =DW ¼ 1=DU þ 1=ðKDUÞ 1=½S:
ð2Þ
The value of DU so obtained was then used to re-evaluate r and the procedure was repeated. Usually, four iterations were sufficient to reach convergence. Fits to Eq. (2) are shown in Fig. 3 for both the DNA/PR and DNA/PRPt systems. The values of K are collected in Table 1. The temperature dependence of K has been investi0 gated for DNA/PRPt and the value of DHapp , the apparent enthalpy of reaction (1), is reported in Table 1. The binding constant of the DNA/PRPt system depends on salt concentration as shown in Fig. 4. The trend of log K as a function of log½Naþ is linear, in agreement with the Manning and Record equation [22]. The intercept of the plot yields log K0 , which has been defined as the binding constant in the absence of electrostatic effects [22]. The slope, whose value is 1.01, corresponds to m0 w, where m0 is the number of phosphodiesteric residues occupied by one dye molecule and w is the extent of DNA charge shielded by counter-ions. Being w ¼ 0:88 [23], it follows that m0 ffi 1:2. The value of w is made by a shielding contribution (ws ) and by a contribution from ion condensation (wc ). Being for DNA w ffi wc , it turns out that m0 w represents the number of condensed sodium ions displaced by one dye molecule.
B
Fig. 2. Fluorescence titrations of the DNA/PR (A) and DNA/PRPt (B) systems at low ionic strength. I ¼ 1 103 M, pH 7.0, and T ¼ 25 °C. The two straight lines intersect to a point corresponding to the site size, n.
Table 1 Reaction parameters for the interaction of CT-DNA with proflavine (PR) and Pt-proflavine (PRPt); pH 7.0, I ¼ 0:11 M (NaCl), T ¼ 25 °C Dye
n
104 Ka (M1 )
104 Kb (M1 )
103 K1 (M1 )
K2 c
K3 c
103 k2 (s1 )
102 k2 (s1 )
k3 (s1 )
k3 (s1 )
0 DHapp (kcal mol1 )
PR PRPt
1.3 0.1 2.6 0.1
6.6 0.7 14 1
7.1 2.2 17 7
4.2 0.5 5.4 0.5
16 3 22 3
– 0.39 0.08
6.8 0.4 6.6 0.4
4.3 0.5 3.0 0.3
– 28 4
– 72 5
)5.9 0.3d )10 1e
a
From static measurements. K ¼ K1 ð1 þ K2 ð1 þ K3 ÞÞ. c From rate constants: K2 ¼ k2 =k2 , K3 ¼ k3 =k3 . d From amplitudes. e From static measurements at different temperatures. b
66
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
A
B
Fig. 3. Fluorescence titrations for the DNA/PR (A) and DNA/PRPt (B) systems. The fittings of the experimental points to Eq. (2) are shown as continuous lines. ½S ¼ CP f ðrÞ, I ¼ 0:11 M (NaCl), pH 7.0, and T ¼ 25 °C.
Fig. 4. Dependence of the DNA/PRPt equilibrium constant on salt concentration. pH 7.0, T ¼ 25 °C.
Kinetics DNA/PR The T-jump experiments with absorbance detection have been carried out at 430 nm with polarised light under magic angle conditions (55°) in order to avoid interference from polymer alignment caused by the capacitor discharge [24,25]. This precaution was maintained, although preliminary experiments revealed that at I ¼ 0:01 M (NaNO3 ) the time constant of the process of polymer deorientation is only 32 ls. This finding ensures that at the working ionic strength (0.11 M) the deorientation of the sonicated DNA is much faster and
Fig. 5. Relaxation curves for the DNA/PR system recorded by absorbance (A) and fluorescence (B) detection. I ¼ 0:11 M (NaCl), pH 7.0, and T ¼ 25 °C; (A) CP ¼ 4:6 105 M, CD ¼ 3:4 105 M; (B) CP ¼ 5:0 105 M, CD ¼ 2:5 106 M.
therefore it does not interfere with the dynamics of the binding process. This conclusion is demonstrated by the fact that relaxation curves recorded in the presence or in the absence of polariser are identical, except for a somewhat larger noise due to light absorption by the polariser. The measurements with fluorescence detection agree very well with those monitored by absorbance. Fig. 5 represents two relaxation curves recorded by absorbance detection (A) and fluorescence detection (B), respectively. In contrast with some of the previous studies on the DNA/PR system, where two chemical relaxation effects were observed [26–28], but in agreement with other investigations [29,30], here the kinetic curves are monoexponential, independent of the detection mode (absorbance or fluorescence). The dependence of the reciprocal relaxation time 1=s on F ðCÞ, a function of the reactant concentration, tends to a plateau (Fig. 6), thus revealing that the binding process could not be described by a single reaction. According to Jovin and Striker [21], F ðCÞ ¼ ½S f 0 ðrÞ½D. A two-step series model, leading to Eq. (3), appears to explain the kinetic behaviour 1=s ¼ K1 k2 F ðCÞ=ð1 þ K1 F ðCÞÞ þ k2 :
ð3Þ
The meaning of the reaction parameters, whose values are collected in Table 1, is made clear by inspecting scheme (5). DNA/PRPt This system displays two kinetic effects, as revealed by the biexponential relaxation curve shown in Fig. 7.
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
67
A
Fig. 6. Dependence of the reciprocal relaxation time, 1=s, on the reactant concentration for the DNA/PR system. The fitting of the experimental data to Eq. (3) is shown as a continuous line. F ðCÞ ¼ ½S f 0 ðrÞ½D, I ¼ 0:11 M (NaCl), pH 7.0, T ¼ 25 °C.
B
Fig. 8. Concentration dependencies of fast (A) and slow (B) relaxation times for the DNA/PRPt system. The fittings of the experimental data to Eqs. (3) and (4), respectively, are shown as continuous lines. (d), absorbance; (s), fluorescence; I ¼ 0:11 M (NaCl), pH 7.0, T ¼ 25 °C.
Fig. 7. Relaxation curve for the DNA/PRPt system (absorbance). I ¼ 0:11 M (NaCl), pH 7.0, T ¼ 25 °C.
The concentration dependence of the reciprocal fast relaxation time, 1=sf , tends to a plateau (Fig. 8A) in agreement with Eq. (3) (where now 1=s is replaced by 1=sf ). The reciprocal slow relaxation time, 1=ss , also tends to a plateau (Fig. 8B) in agreement with 1=ss ¼ ðKf k3 F ðCÞÞ=ð1 þ Kf F ðCÞÞ þ k3 ;
ð4Þ
where Kf ¼ K1 ð1 þ k2 =k2 Þ. The rate parameters, derived by least-square fits to Eqs. (3) and (4), are collected in Table 1. The behaviour of both the investigated systems could be rationalised on the basis of the general reaction scheme K1
k2
k3
k2
k3
D þ S ¢ DSI ¢ DSII ¢ DSIII
ð5Þ
where DSI , DSII , and DSIII represent different bound forms. For the DNA/PR system, the last step is lacking. According to scheme (5), the overall binding constant can be expressed as a combination of rate constants, as shown by K ¼ K1 f1 þ ðk2 =k2 Þ½1 þ ðk3 =k3 Þg: For the DNA/PR system, K ¼ K1 ½1 þ ðk2 =k2 Þ.
ð6Þ
Fig. 9. Salt concentration dependence of the kinetic effects for the DNA/PRPt system. (r), k ¼ K1 (M1 ); (j), k ¼ k2 (s1 ); and (d), k ¼ k3 (s1 ); pH 7.0, T ¼ 25 °C.
Table 1 shows that values of K from Eq. (6) agree with those evaluated from equilibria, thus proving the consistence of the results. Fig. 9 shows the salt concentration dependence of the kinetic effects. The largest variation is displayed by K1 . A plot of log K1 vs. log½Naþ is linear with a slope equal to unity, whereas plots of log k2 and log k3 vs. log½Naþ do not show any salt dependence. This finding reveals that the electrostatic effect is exerted on the first step of the binding process. Amplitude analysis DNA/PR The analysis of the amplitudes of the DNA/PR system has been performed according to [31]
68
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
ratio of the slow effect is too unfavourable to provide reliable results. Since the relaxation times of the two effects differ by more than one-order of magnitude, the fast relaxation has been associated with the same steps that are operative in the DNA/PR system, namely with the first of the two steps of scheme (5). Its amplitude, dA0I , is directly proportional to the amplitude factor CI as shown in Fig. 10B. In this case, however, the plot of dA0I vs. CI does not enable us to evaluate the reaction enthalpy, since DeI is unknown. However, if one makes the hypothesis that DeI is the same as that measured for DNA/PR, it turns out from the slope of the plot of Fig. 10B that DHI0 ¼ 5:0 kcal mol1 . This value is similar to that found for DNA/PR.
A
B
Discussion DNA/PR
Fig. 10. Amplitude analysis (A) for the DNA/PR system and (B) for the DNA/PRPt system. The fittings of the experimental data to Eq. (7) are shown as continuous lines; I ¼ 0:11 M (NaCl), pH 7.0, T ¼ 25 °C.
0 dA0 ¼ DeCDHapp dT =RT 2 ;
ð7Þ
where dA0 is the amplitude of the relaxation effect (expressed as a change of absorbance) following a jump of temperature of magnitude dT . The amplitude 1 factor [21] C ¼ ð1=½DS þ 1=½D f 0 ðrÞ=½SÞ has been evaluated, for each experiment, by using the already determined values of n and K. Fig. 10A shows that the plot of dA0 vs. C is linear, in accord with Eq. (7). Being De ¼ 2:06 104 M1 and dT ¼ 1:6 °C, the slope of the plot of Fig. 10A yields the enthalpy of the apparent reaction (1) whose value is reported in Table 1. DNA/PRPt The amplitude analysis of DNA/PRPt has been performed only on the fast effect since the signal-to-noise
This system has been widely investigated by previous authors with results apparently not converging. The principal discrepancy concerns the number of the observed kinetic effects. A number of investigators [26–28] have measured two relaxation times whereas others [29,30] report only a single relaxation time, which tends to level off at high reactant concentration. Nevertheless, the biphasic nature of the binding process has been established by all the studies far so performed. The results of the various kinetic investigations, including the present one, on the DNA/PR system are collected in Table 2. Concerning the site size, n, it can be noted that n is linked to the parameter B of the Scatchard equation by the relationship B ¼ 1=ð2n 1Þ [20]. In the present case, one obtains B ¼ 0:64 in NaCl 0.11 M, which agrees with the value that can be derived from the data of Marcandalli et al. [30] at the same salt concentration. DNA/PRPt A comparison between DNA/PRPt and DNA/PR shows that the platinum residues play a principal role in determining the differences of behaviour between the two systems. The site size for DNA/PRPt is twice that
Table 2 Reaction parameters for the binding of proflavine (PR) to DNA
CT-DNA CT-DNA CT-DNA Poly(dA-dT) CT-DNA CT-DNA
106 k1 (s1 )
103 k1 (s1 )
102 k2 (s1 )
101 k2 (s1 )
103 K1 (M1 )
K2
104 K (M1 )
0 DHapp (kcal mol1 )
14 17 28 – – –
13 0.12 11 – – –
69 33 18 36 33 68
42 20 11 8 26 43
1.1 3.3 2.5 4.6 6.7 4.2
17 16 16 46 13 16
2 – 4.3 – 9.2 6.6
)7.9 )12.9 – )0.38 – )5.9
(1) (2) (3) (2) (4) (5)
(1) I ¼ 0:2 M, T ¼ 25 °C, [26]; (2) I ¼ 0:11 M, T ¼ 17 °C, [27]; (3) I ¼ 0:24 M, T ¼ 10 °C, [28]; (4) I ¼ 0:2 M, T ¼ 25 °C, [29]; (5) I ¼ 0:11 M, T ¼ 25 °C, this work.
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
measured for DNA/PR. This result suggests that the cis-platinum moiety as well could accommodate into the DNA cavities and that this process is related to the kinetic effect with time resolution in the tenths of second. The time constant of the slow effect is far below any time constant measured for DNA/PR. Thus, this effect could not be associated to a monomolecular step where only the acridine residue is involved. It appears quite reasonable that the slowness of this effect should be ascribed to the work necessary to enlarge the polymer cavity in order to introduce the cis-platinum residues as well (step DSII ¢ DSIII ). The finding that for PRPt the value of n is twice that measured for PR could be alternatively rationalised, assuming that the platinumcontaining residue undergoes groove interaction, thus preventing the occupation of the nearest-neighbour site. The rate of interaction of groove dyes with DNA has been shown to be nearly diffusion controlled [32] and therefore is rather independent of the ligand nature. Taking for the rate constant of the binding step the value of 108 M1 s1 [33] it turns out that, in the range of DNA concentrations investigated (2 105 –5 104 M), the values of the relaxation times for step DSII ¢ DSIII would lie between 0.02 and 0.5 ms, i.e., 20–500 times faster than the experimental values. It should also be noticed that square-planar platinum complexes as [(bipy)2 Pt(en)]2þ are known to intercalate into DNA [34]. Although aware that, in the absence of any structural evidence, any conclusion about the nature of the slow binding is speculative, we tend to favour the first hypothesis on the basis of the above considerations. The enthalpy change associated with the step DSII ¢ DSIII for the DNA/PRPt system is remarkable. It can be evaluated noting that the overall equilibrium constant K is related to the individual-step equilibrium constants by Eq. (6). By differentiating this equation a relationship between the enthalpies can be obtained, 0 that is: DH30 ¼ ðDHapp DHI0 Þð1 þ K2 ð1 þ K3 ÞÞ=ðK2 K3 Þ, 0 where DHI , evaluated by the amplitude analysis, is the enthalpy of the fast normal reaction, i.e., a combination of DH10 and DH20 [35]. Thus, the enthalpy variation for the DSII ¢ DSIII step, DH30 , is found to be equal to )18 kcal M1 . The GibbÕs energy variation for the above step, DG03 ¼ RT ln K3 , is +0.56 kcal M1 ; hence, from the relationship DG03 ¼ DH30 T DS30 , it turns out that the entropy change, DS30 , is )62 cal M1 K1 . This finding would suggest that the interaction of the platinum-residues with the DNA bases is accompanied by considerable increase of rigidity. The data of Table 1 show that the affinity of PRPt for DNA is somewhat higher than that of PR. In particular, whereas the stability of the DSI complex is similar for the two systems, the value of K2 increases by 27% on changing PR to PRPt. The enhanced stability of DSII depends on k2 which decreases from 4.3 102 s1 (PR) to 3.0 102 s1
69
(PRPt). We ascribe this rate reduction to a stabilising effect of the platinum-residue positioned outside of the DNA cavity in DSII . By contrast, the presence of the platinum-residue inside of the cavity as in DSIII exerts a destabilising effect, as shown by the value of K3 which indicates that the platinum-residue stays inside of the cavity only for the 28% of the time.
Acknowledgment This work has been partially supported by MIUR with COFIN2002.
References [1] L.S. Lerman, J. Mol. Biol. 39 (1961) 461–477. [2] A. Aggarwal, S.A. Islam, R. Kuroda, S. Neidle, Biopolymers 23 (1984) 1025–1041. [3] A. Adams, Curr. Med. Chem. 9 (2002) 1667–1675. [4] L. Malinina, M.M. Soler-L opez, J. Aymamı, J.A. Subirana, Biochemistry 41 (2002) 9341–9348. [5] S. Neidle, Prog. Med. Chem. 16 (1979) 151–221. [6] H.M. Berman, P.R. Young, Annu. Rev. Biophys. Bioeng. 10 (1981) 87–114. [7] W.P. Wilson, R.L. Jones, Adv. Pharmacol. Chemother. 18 (1981) 177–222. [8] S. Lippard, Acc. Chem. Res. 11 (1978) 211–217. [9] C. Hiort, P. Lincoln, B. Norden, J. Am. Chem. Soc. 115 (1993) 3448–3454. [10] S. Arounaguiri, D. Easwaramoorthy, A. Ashokkumar, A. Dattagupta, G.M. Bhaskar, Proc. Indian Acad. Sci. (Chem. Sci.) 112 (2000) 1–17. [11] K.E. Errkila, D.T. Odom, J.K. Barton, Chem. Rev. 99 (1999) 2777–2795. [12] L.G. Marzilli, S. Ano, F.P. Intini, G. Natile, J. Am. Chem. Soc. 121 (1999) 9133–9142. [13] E.A. Boudreaux, Int. J. Quantum Chem. 83 (2001) 255–258. [14] B.E. Bowler, K.J. Ahmed, W.I. Sunquist, L.S. Hollis, E.E. Whang, S.J. Lippard, J. Am. Chem. Soc. 111 (1989) 1299– 1306. [15] C. Ciatto, M.L. DÕAmico, G. Natile, F. Secco, M. Venturini, Biophys. J. 77 (1999) 2717–2724. [16] G. Schwarz, S. Klose, W. Balthasar, Eur. J. Biochem. 12 (1970) 454–467. [17] E. Ceci, R. Cini, A. Karaulov, M.B. Hurthouse, L. Maresca, G. Natile, J. Chem. Soc. Dalton Trans. (1993) 2491–2497. [18] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Kluver Academic/Plenum Publishers, New York, 1999. [19] S.W. Provencher, J. Chem. Phys. 64 (1976) 2772–2777. [20] J.D. McGhee, P.H. von Hippel, J. Mol. Biol. 86 (1974) 469–489. [21] T.M. Jovin, G. Striker, Mol. Biol. Biochem. Biophys. 24 (1977) 245–281. [22] M.T. Record, C.F. Anderson Jr., T.M. Lohman, Q. Rev. Biophys. 11 (1978) 103–178. [23] T. Schelhorn, S. Kretz, H. Zimmermann, Cell. Mol. Biol. 38 (1992) 345–365. [24] M. Dourlent, J.F. Hogrel, C. Helene, J. Am. Chem. Soc. 96 (1974) 3398–3406. [25] F.J. Meyer-Almes, D. Porschke, Biochemistry 32 (1993) 4246– 4253. [26] H.J. Li, D.M. Crothers, J. Mol. Biol. 39 (1969) 461–477.
70
T. Biver et al. / Archives of Biochemistry and Biophysics 418 (2003) 63–70
[27] J. Ramstein, M. Ehrenberg, R. Rigler, Biochemistry 19 (1980) 3938–3948. [28] M. Dourlent, J.F. Hogrel, Biochemistry 15 (1976) 430–436. [29] B. Marcandalli, C. Winzek, J.F. Holzwarth, Ber. Bunsenges. Phys. Chem. 88 (1984) 368–374. [30] B. Marcandalli, G. Stange, J.F. Holzwarth, J. Chem. Soc. Faraday Trans. 1 (84) (1988) 2807–2819. [31] M. Citi, F. Secco, M. Venturini, J. Phys. Chem. 22 (1988) 6399– 6404.
[32] S.Y. Breusegem, S.E. Sadat-Ebrahimi, K.T. Douglas, E.V. Bichenkova, R.M. Clegg, F.G. Loontiens, J. Med. Chem. 44 (2001) 2503–2506. [33] S.Y. Breusegem, R.M. Clegg, F.G. Loontiens, J. Mol. Biol. 315 (2002) 1049–1061. [34] S. Arnott, P.J. Bond, R. Chandrasekaran, Nature 287 (1980) 561– 563. [35] F. Secco, M. Venturini, J. Chem. Soc. Faraday Trans. 89 (1993) 719–725.