European Journal of Medicinal Chemistry 80 (2014) 285e294
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Localization and interaction of hydroxyflavones with lipid bilayer model membranes: A study using DSC and multinuclear NMR Ragini Sinha a, Akshada Joshi b, Urmila J. Joshi b, *, Sudha Srivastava a, Girjesh Govil a a b
National Facility for High Field NMR, Tata Institute of Fundamental Research, Homi Bhabha Road, Mumbai 400 005, India Principal K.M. Kundnani College of Pharmacy, Cuffe Parade, Mumbai 400 005, India
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 July 2013 Received in revised form 17 April 2014 Accepted 19 April 2014 Available online 22 April 2014
The localization and interaction of six naturally occurring flavones (FLV, 5HF, 6HF, 7HF, CHY and BLN) in DPPC bilayers were studied using DSC and multi-nuclear NMR. DSC results indicate that FLV and 6HF interact with alkyl chains. The 1H NMR shows interaction of flavones with the sn-glycero region. Ring current induced chemical shifts indicate that 6HF and BLN acquire parallel orientation in bilayers. 2D NOESY spectra indicate partitioning of the B-ring into the alkyl chain region. The DSC, NMR and binding studies indicate that 5HF and 7HF are located near head group region, while 6HF, CHY and BLN are located in the vicinity of sn-glycero region, and FLV is inserted deepest in the membrane. Ó 2014 Elsevier Masson SAS. All rights reserved.
Keywords: Flavones DPPC DSC NMR Chemical shifts CSA
1. Introduction Flavonoids have emerged as potential therapeutic drugs and are effective against several diseases. Prominent among them, are free radical mediated [1,2] and cell proliferative diseases [3,4]. Flavones and flavonols are found in nature and have shown promising antioxidant and anticancer activity [5,6]. The antioxidant activity of 7-hydroxyflavone in egg yolk phosphatidylcholine liposomes has been reported [7]. Tsuchiya has reported that chrysin and baicalein show significant inhibition of lipid peroxidation in 1, 2-diolieolylphosphatidylcholine liposomal membranes [8]. Greeff et al. have shown potent peroxyl radical scavenging activity of 6hydroxyflavone against fluoroscence decay of fluorescein [9]. Antioxidant activity of 5-hydroxyflavone, chrysin and baicalein has also been reported [10]. Chrysin (5, 7-dihydroxyflavone) possesses several biological properties [11,12], including antioxidant [13] and anticancer activity [14]. Arora et al. [6] have compared membrane interaction of nine flavonoids, and related them to their antioxidant effects. Flavones were not included in their studies, though they
* Corresponding author. Department of Pharmaceutical Chemistry, Prin. K.M. Kundnani College of Pharmacy, 23-Jote Joy Building, R.S. Marg, Cuffe Parade, Colaba, Mumbai, Maharashtra 05, India. E-mail addresses:
[email protected],
[email protected] (U.J. Joshi). http://dx.doi.org/10.1016/j.ejmech.2014.04.054 0223-5234/Ó 2014 Elsevier Masson SAS. All rights reserved.
have important bioactivity [15]. Membrane interaction of flavonoids has been compared and their effects on membrane leakage and lipid peroxidation have been evaluated [16,17]. There are very few studies on the interaction of structurally different flavones with membranes and their relation to bioactivity. The number, position and phenolic hydroxyl groups influence physiochemical properties such as dipole moment and hydrophobicity. In turn, these determine their partitioning into lipid membranes. Interaction of flavones with membranes plays a crucial role in their biological activity [18]. Understanding interaction of flavones with biological membranes may help to elucidate mechanisms of action of flavones as antioxidant and anti-proliferative agents. Model membranes mimic the behavior of biological membranes [19]. The phase transition temperature (Tm) of model membranes plays an important role in binding studies, which in turn depends on the lipid composition. DPPC lipid bilayers are ideally suited for drugemembrane interaction for physical studies due to its above room temperature Tm value 314 K [20,21]. In this paper we report membrane partitioning of six flavone molecules with varying number and position of hydroxyl groups (Fig. 1). DSC and multinuclear NMR techniques have been used to determine localization and interaction of the hydroxyflavones with model bilayer membranes formed by DPPC. Antioxidant and antiproliferative activities have also been studied.
286
R. Sinha et al. / European Journal of Medicinal Chemistry 80 (2014) 285e294
Fig. 1. Molecular structure of flavone and hydroxyflavones.
2. Results and discussion 2.1. Binding studies with MLV Binding studies were done using centrifugation and NMR methods. Fig. 2A shows a plot of fraction of flavones bound to multilamellar vesicles (MLVs) with concentration of lipid. In all cases, an increase in binding affinity is observed with increasing concentration of lipid. FLV shows nearly 80% binding even at relatively low (0.25 mM) concentration of lipid, and thereafter it nearly forms a plateau. Double inverse plot of fraction of drug bound and the lipid concentration (inset of Fig. 2) has been used to calculate binding constants. The apparent binding constants measured are as follows: FLV: 19569 M1, 5HF: 313 M1, 6HF: 800 M1, 7HF: 450 M1, CHY: 759 M1 and BLN: 252 M1. An alternative method using common titration by NMR has also been done to determine the binding constant. Fig. 2B shows a plot of change in chemical shift versus molar concentration (M) of the lipid. The data was analyzed using non-linear regression analysis. The difference in chemical shift of the lipideflavone complex and the flavone represented by Dmax were extracted. The binding constants were calculated using the equation given in Section 4.2 [22]. The apparent binding constants calculated are: FLV: 1817 M1, 5HF: 229 M1, 6HF: 572 M1, 7HF: 311 M1, CHY: 532 M1 and BLN: 215 M1. These values are in agreement with those calculated using centrifugation method. The discrepancy observed in case of FLV may be due to its low water solubility. The results indicate that these molecules bind to the MLV with variable degree of affinity in the order FLV > 6HF > CHY > 7HF > 5HF > BLN. FLV which is devoid of hydroxyl group shows highest binding with DPPC and BLN with highest number of hydroxyl group show least binding. This shows that hydrophobic interactions are the predominant binding forces. The position of the hydroxyl groups also alters binding, probably due to differential hydrogen bond formation with DPPC. Thus the binding of flavones to DPPC seems to be a result of combination of hydrogen bonding and hydrophobic interactions. Issues concerning the nature of this binding with regard to the thermotropic behavior and dynamics of lipids and also the position/number of hydroxyl groups affecting the intermolecular interactions have been addressed in the following sections. 2.2. DSC studies DSC is a sensitive technique for studying the effect of drugs on the packing order of the lipid bilayers. Thermotropic aspect of
drugelipid interactions can be studied by examining changes in the melting point and the shape of the DSC thermograms [23]. Fig. 3 shows thermogram of lipid bilayers incorporated with FLV, 5HF, 6HF, 7HF, CHY and BLN. In each case, the lowest graph (a) represents the thermogram for lipid bilayers alone. Here the pretransition (Tp) at 34.1 C indicates the mobility of the choline part of the polar head of DPPC. Mobility of the alkyl chain is seen in the main transition (Tm) at 41.9 C. All compounds under study cause a concentration dependent decrease in the Tm values and either reduced or abolished the Tp values. This suggests an interaction between these molecules and lipid bilayers [24]. The pre-transition peak is sensitive to the presence of impurities and is abolished even at small quantities of impurities [25]. This explains the abolition of pre-transition peak in case of FLV and 6HF. However, even at higher concentrations of 5HF, 7HF, CHY and BLN the pre-transition peak remains sharp. Moreover, a large decrease in Tp is observed for 5HF (3.9 C), 7HF (6.7 C), CHY (3.6 C) and BLN (3.9 C). This indicates interaction of these compounds with the head group region [26]. A decrease in the main transition temperature indicates that the compounds interact with the hydrophobic core of the lipid bilayer. The maximum decrease in Tm was seen for FLV (36.3 C) followed by 6HF (36.9 C) at drug/lipid molar ratio of 1:2 (Table 1) indicating a strong interaction with the hydrophobic core. In the case of CHY and BLN, an increase in drug concentration leads to a decrease in Tp and Tm values up to 1:10 drug/lipid molar ratio followed by an increase at higher concentration. The increase in Tm can be attributed to concentrations of these flavones in excess of the solubility in the lipid. As a result, further increase in concentration of the drug leads to aggregation [27]. The aggregation is facilitated by multiple hydroxyl groups in the structure, which enable intermolecular hydrogen bonding. The aggregates have a limited perturbing effect on the thermotropic behavior. Thus, it may be concluded that CHY and BLN affect the pre and the main transitions at lower concentrations, aggregate at higher concentrations and beyond a certain concentration no further binding results. Comparison of thermograms of FLV, 5HF, 6HF and 7HF shows broadening of main transition peaks of FLV and 6HF, as against narrowing of the main transition peaks of 5HF and 7HF at all concentrations. The broadening of peak indicates that alkyl chains have become ordered [28], while narrowing of the main transition peak suggests a disordered structure of the alkyl chains. Ordering of alkyl chains induced by FLV and 6HF is possibly due to their deeper
R. Sinha et al. / European Journal of Medicinal Chemistry 80 (2014) 285e294
Fig. 2. (A) Binding of drugs with DPPC multilamellar vesicles in H2O (pH 7.2) using centrifugation method. The fraction bound is determined as described in Section 4.2. Inset shows the double reciprocal plot for 6HF. (B) Binding curve using NMR chemical shift as described in Section 4.2. Inset shows fraction bound vs. D for 6HF. The curves are: FLV (C), 5HF (B), 6HF (;), 7HF (D), CHY (-) and BLN (,).
insertion into hydrophobic core. This is also indicated by maximum decrease in their main transition temperatures. 2.3. NMR experiments 1 H and 13C NMR spectra of flavones in DMSO-d6 have been assigned using 2D COSY (Supplementary Table 1). 1H NMR spectra of ULV of DPPC incorporated with flavones (1:5 drug/lipid molar ratios) at 323 K show significant shifts as compared to ULV of DPPC (Table 2). Fig. 4 shows the aromatic protons of all the flavones are shifted downfield on interaction with DPPC. Maximum downfield shift is seen for 6-H and 8-H of CHY which points to the involvement of both 5-OH and 7-OH in the hydrogen bonding with DPPC. A large downfield shift in the 30 , 50 -protons of the B-ring is also seen in FLV, 6HF and CHY, indicating their interaction with DPPC. Localization of compounds in lipid bilayers can be predicted from the changes in chemical shifts. Large induced chemical shifts indicate a strong lipideflavone interaction. Chemical shift changes of DPPC upon incorporation of different flavones (Fig. 5) indicate that all the compounds with the exception of 5HF interact with the
287
sn-glycero region of DPPC to varying degrees. 7HF show significant interaction with the polar head group region whereas FLV, 6HF, CHY and BLN interact with alkyl chains situated deeper. Previous reports have indicated that although benzene ring of the molecule is located preferentially in the hydrophobic core [29], addition of a directly bonded hydroxyl group shifts the location to the lipidewater interface [30]. Applying this to the flavones in the present study, it may be concluded that the A/C-rings of the flavones which contain phenolic hydroxyl groups orient towards the lipidewater interface and B-ring penetrates in the hydrophobic core. In such an orientation, the hydroxyl groups of the A-ring are able to form hydrogen bonds either with water molecules or phosphate and/or acyl oxygens of DPPC. Since all compounds interact significantly with the glycero region of DPPC, these molecules possibly form hydrogen bonds with acyl oxygens. The interaction at the head group region is likely to be a combination of hydrogen bonding with phosphate oxygens and cationep interaction, taking into consideration the structures of flavones and DPPC. This however cannot be extended to FLV, which in absence of any hydroxyl group may not show preferred orientation or an anchoring point. The only expected interaction of FLV with polar head group region may be cationep type of interaction. Location of the rings can be determined from the ring current effect on the lipid protons [31]. The chemical shift changes induced by ring current depend on the orientation of the ring with reference to the lipid protons in the proximity of the ring. The lipid resonances are shifted to high frequency upon interaction with ring edge and low frequency on interaction with ring center [32]. Thus, a high frequency shift indicates the ring plane to be perpendicular to the bilayer normal whereas a low frequency shift indicates a parallel orientation of the ring plane with reference to the bilayer normal. The flavones under consideration contain more than one aromatic ring. A uniform low frequency shift of the lipid protons in presence of 6HF and BLN indicates that both these rings i.e. A/C and B are orientated parallel to the bilayer normal pointing at a parallel orientation of the molecules. This may be due to the presence of a common 6-hydroxyl group. A parallel orientation has been reported for luteolin [33]. In CHY, a uniform high frequency shift indicates a perpendicular orientation of both the rings with reference to lipid protons. In absence of a uniform low frequency or high frequency shift for 5HF, 7HF and FLV, it is possible that one of the rings is parallel and the other perpendicular to the bilayer normal. To probe deeper into the nature of interaction, 1He1H NOESY [34] spectrum of ULV of DPPC and that of DPPC ULV incorporated with flavones have been recorded. The assignment of the drugs and lipid molecules are indicated along F2 and F1 dimensions, respectively. In the 2D NOESY spectra of 6HF and BLN, NOEs are seen between the 30 /50 protons of B-ring of these compounds and the e (CH2)n protons of the alkyl chain of lipid bilayer (Fig. 6). This suggests localization of the 6-hydroxy group of 6HF and BLN at the glycero group and its B-ring partitioning into the hydrophobic chain of lipid bilayer. In the 13C NMR spectrum of the DPPC bilayer and that of DPPC ULV incorporated with flavones (Supplementary Fig. 1), all signals arising from flavones are broadened compared to their pure form. The broadening of the signals arises due to an exchange at an intermediate NMR time scale between the bound and free form of the flavones [35]. Further, the >CO signal of CHY at 182.5 ppm shifts significantly to lower frequency by Dd w7 ppm and appears at 176.1 ppm in the ULV incorporated with chrysin. The >CO of the acyl group of the lipid (at 173.8 and 173.6 ppm) in DPPC do not shift. In other flavones the same peak is shifted to lower frequency by w3 ppm and is merged with the >CO peak of the acyl group of the lipid apart from broadening. A drastic shift in peak position of the >CO group in chrysin may be due to the involvement of adjacent 5-
288
R. Sinha et al. / European Journal of Medicinal Chemistry 80 (2014) 285e294
Fig. 3. DSC heating curves of hydrated MLVs of DPPC (50 mM) containing flavones and hydroxy-flavones, at drug/lipid molar ratios; a ¼ 0:100; b ¼ 1:20; c ¼ 1:10; d ¼ 1:5; e ¼ 1:2.
Table 1 DSC studies showing the pretransition (Tp) and main transition (Tm) temperatures values of DPPC (50 mM) with varying drug molar ratios. Drug:DPPC
0:100 1:20 1:10 1:5 1:2
FLV
5HF
6HF
7HF
CHY
BLN
Tp
Tm
Tp
Tm
Tp
Tm
Tp
Tm
Tp
Tm
Tp
Tm
34.1 e e e e
41.9 37.2 36.7 36.4 36.3
34.1 32.3 32.0 32.0 30.3
41.9 41.1 40.7 40.6 39.6
34.1 e e e e
41.9 37.4 37.3 37.1 36.9
34.1 31.9 31.8 e 27.4
41.9 40.7 40.5 40.4 39.8
34.1 31.0 30.4 30.5 31.6
41.9 40.6 39.8 39.9 40.1
34.1 32.1 30.2 32.9 33.8
41.9 40.4 39.7 40.6 40.9
Table 2 1 H NMR shifts (ppm) in D2O, for ULV of DPPC and ULV of DPPC incorporated with flavone/hydroxyflavones in 1:5 drug/lipid molar ratios at 323 K. a
Assignment protons
Alkyl chain 1 2 3 4 Polar head 10 9 8 sn-glycero 7 5b 5a 6 a
DPPC
FLV þ DPPC
5HF þ DPPC
6HF þ DPPC
7HF þ DPPC
CHY þ DPPC
BLN þ DPPC
1.13(s) 1.52(s) 1.85(s) 2.57(d)
1.18(s) 1.58(s) 1.84(s) 2.62(s)
1.12(s) 1.50(s) 1.81(s) 2.57(d)
1.08(s) 1.48(s) 1.76(s) 2.54(s)
1.13(s) 1.54(s) 1.83(s) 2.60(d)
1.23(s) 1.62(d) 1.92(s) 2.66(d)
1.10(s) 1.51(s) 1.76(s) 2.56(d)
3.50(s), 3.45(s) 3.93(d) 4.52(d)
3.55(s), 3.51(s) 3.93(d) 4.50(d)
3.49(s) 3.93(s) 4.54(d)
3.45(s), 3.43(s) 3.89(s) 4.50(s)
3.50(s) 3.92(s) 4.55(s)
3.58(s) 4.01(d) 4.65(s)
3.47(s), 3.42(s) 3.86(d) 4.51(s)
4.23(s) 4.40(s) 4.76(s) 5.52(d)
4.20(d) 4.42(s) 4.68(s) 5.54(s)
4.24(s) 4.40(s) 4.76(s) 5.52(d)
4.1(d), 4.23(s) 4.38(s) 4.67(s) 5.46(s)
4.24(s) 4.44(s) 4.45(s) 5.52(d)
4.33(s) 4.56(d) 4.77(s) 5.61(d)
4.11(d), 4.23(s) 4.38(s) 4.69(s) 5.47(s)
Assignment of protons of DPPC as shown in Fig. 4.
R. Sinha et al. / European Journal of Medicinal Chemistry 80 (2014) 285e294
289
Fig. 4. 500.13 MHz 1H NMR spectra of DPPC unilamellar vesicles and aromatic region of ULV of DPPC incorporated with FLV, 5HF, 6HF, 7HF, CHY and BLN in 1:5 drug:lipid molar ratio at 323 K in D2O.
hydroxy group in hydrogen bond formation. This is further supported by a distinct shift of 6H of CHY from 6.2 ppm to 6.8 ppm in the 1H NMR (Fig. 4). The absence of shift of carbonyl group in 5HF is in agreement with its minimum interaction with either phosphate or acyl oxygen atoms. 31 P NMR is sensitive to local motions and the orientation of the phosphate group in the membrane. It has been used for monitoring structural changes and detecting polymorphism in model membranes [36]. The line shape is determined by the CSA of the phosphate group and molecular motions near the head groups [37]. Lipid bilayers give a characteristic broad spectrum with a high field peak and low field shoulder. CSA can be measured from the low and high field shoulders of the spectrum (sk and st components). The effect of flavones on the 31P CSA has been measured as a function of concentration and temperature. Fig. 7 shows 31P NMR line shape of MLV of DPPC bilayer incorporated with different concentrations of flavones at temperature 323 K. The change in CSA with temperature and concentration is shown in Fig. 8. CSA shows an increase as compared to pure DPPC MLV in case of FLV, 5HF, 6HF and 7HF at all the three drugs to lipid molar ratios. In case of CHY, the CSA decreases with increasing concentration [Fig. 8 (A and B)] but increases at 1:2 drug to lipid molar ratio (Fig. 8C). As shown in Fig. 8, the highest concentration-dependent increase in CSA is shown by DPPC MLV incorporated with 5HF and 7HF. The bilayer features of the MLV of lipid are retained in these cases. However in case of FLV larger vesicles are formed at 1:2 drug/lipid molar ratios. In 6HF and BLN at the same concentration, the bilayer phase transforms into hexagonal phase. Increase in CSA indicates interaction of the drug with the head group region. It may also reflect a decrease in the local mobility of the phosphorus of the head group region [38]. High increase in CSA in the case of 5HF and 7HF therefore indicates their interaction with the phosphate region. The drug molecules probably form hydrogen bonds with the phosphate oxygens or water molecules present in the vicinity, thereby causing a decrease in the local mobility of
phosphorus. FLV, 6HF, CHY and BLN are inserted deeper than 5HF and 7HF show less increase in CSA. This may be due to their greater interaction with alkyl chains. The order of increase in CSA at 1:5 drugelipid molar ratio at 323 K is 5HF, 7HF, CHY > BLN > 6HF > FLV. The mobility of the head group is least affected by FLV and 6HF, thereby pointing to its localization near the hydrophobic core. Thus, NMR, DSC and binding studies together point towards that fact that in hydroxyflavones, the A-ring is probably at the lipide water interface, while ring B partitions into the alkyl chains. FLV which does not have any hydroxyl groups is located deep in the hydrophobic core of the lipid. The orientation and location of these molecules is schematically shown in Fig. 9A and B respectively. 5HF and 7HF with eOH group at position 5 and at position 7 respectively, are located near the head group region. The orientation of the rings in these molecules are not uniform and it seems that their eOH group position is such that it restricts their partitioning into the hydrophobic core. In case of 6HF the eOH group position seems to be favorable for a uniform orientation and better partitioning inside the hydrophobic core. Moreover, 6HF as well as BLN lie parallel to the bilayer normal, and probably the presence of eOH group at position 6 facilitates their partitioning inside the membrane. However, CHY with eOH groups at positions 5 and 7 lie perpendicular to the bilayer normal. CHY, 6HF and BLN to some extent interact with all the regions of the bilayer. In case of 5HF and 7HF, the interaction is weaker and is not uniform with all regions of the bilayer. This results in lower binding. The least binding of BLN may be due to the presence of three hydroxyl groups causing attenuation of membrane interaction as has been reported in case of myricetin having three-hydroxyl groups on B-ring [8]. 2.4. Antioxidant and antiproliferative activity DPPH assay has been used to estimate radical scavenging activity of the flavones. The radical scavenging activity is dependent
290
R. Sinha et al. / European Journal of Medicinal Chemistry 80 (2014) 285e294
Fig. 5. 500.13 MHz 1H NMR chemical shifts (ppm) in D2O, for ULV of DPPC and ULV of DPPC incorporated with flavones in 1:5 drug/lipid molar ratio at 323 K.
on the number and the position of the phenolic OH groups [39]. The compounds in the present study adhere to the above suggestion (Table 3). Therefore, increase in the number of phenolic hydroxyl group increases its antioxidant activity. FLV which does not have OH group shows negligible radical scavenging activity. Among the flavones with single OH group at different positions, it has been
reported that the bond dissociation energies follow the order of 7OH > 5-OH > 6-OH [40]. The radical scavenging activity in the present work is parallel to the bond dissociation energies of the Oe H bonds. The order of radical scavenging activity shown by these compounds is in the order BLN > CHY > 6HF > 7HFw5HF > FLV. The anti-proliferative activity (Table 4) of these compounds
Fig. 6. 2D NOESY spectrum of DPPC ULV incorporated with 6HF in 1:5 drug:lipid molar ratio at 323 K in D2O. Inset figure represents 2D NOESY spectrum of DPPC ULV incorporated with BLN under similar conditions.
R. Sinha et al. / European Journal of Medicinal Chemistry 80 (2014) 285e294
291
Fig. 7. 202.4 MHz 31P NMR spectra of DPPC (100 mM) multilamellar vesicles incorporated with FLV, 5HF, 6HF, 7HF, CHY and BLN. The additives:lipid molar ratios are (a) 0:100 (b) 1:10 (c) 1:5 and (d) 1:2. All experiments are at 323 K.
showed BLN, 6HF and FLV to be most effective against MCF-7, K562 and HepG2 cell lines respectively. No significant correlation was observed between localizationeorientation and inhibitory activity. 3. Conclusions The objective of the present study was to investigate localization and interaction of flavones having varying number of hydroxyl groups, when incorporated in DPPC bilayers. The results suggest a broad distribution of flavones in the membrane, which is dependent on their chemical structure. It is observed that the location of the compound along with the nature of the interaction affects its binding, orientation, antioxidant and anti-proliferative activity. The hydrophilic nature of flavonoid increases with the number of hydroxyl groups. As flavonoids become more hydrophilic, their membrane localization shifts towards the aqueous environment. Amongst the flavones, containing one eOH group, their position seems to affect their interaction with the membrane and position 6 is probably the most suited for ring orientation as well as better penetration inside the membrane. It may also be noted that higher concentration of CHY and BLN is not acceptable to the membrane structure as observed from DSC results. Thus, the hydroxyl group present at different positions on the A-ring of the flavone molecule influences its orientation, localization and thereby, interaction with the lipid membrane. 4. Materials and methods 4.1. Materials Flavone, 5-hydroxyflavone, 6-hydroxyflavone, 7-hydroxyflavone, chrysin, baicalein, 1, 2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and DPPH were purchased from Sigma chemicals Co. USA. 4.2. Determination of drugeMLV binding constants DrugeMLV binding constants were determined by the centrifugation method [41] as well as the NMR titration method [22]. The centrifugation method was carried out using MLV prepared by
varying lipid concentration from 0.25 mg/ml to 2 mg/ml and fixed drug concentration of 100 mM (giving rise to a drug to lipid ratios in the range 1:2.5 to 1:20). The resulting solutions were incubated for 2 h at 50 C and subsequently transferred into ultracentrifuge tubes. Separation of liposomes from the aqueous phase was achieved by centrifugation at 40,000 rpm for 2 h. The amount of compound bound to liposomes was determined from the difference in optical density measured for the pure compound and that of the supernatant. Optical density of 100 mM solution of the pure drug molecules has been measured at a wavelength range of 220e400l. The drugeliposome apparent binding constant (k) has been analyzed using the double reciprocal plot of 1/(fraction bound) versus 1/(lipid concentration) which yields a straight line with a slope 1/k. The NMR titration was done keeping the concentration of the drug constant (10 mM) and varying the lipid concentration (1 mMe 10 mM). Stock solutions of lipid and drugs were prepared in D2O and methanol respectively, and diluted suitably. To each sample Trimethylsilyl propanoic acid (TSP) was added for reference. The binding constant was calculated using equation:
D¼
Dmax K½L 1 þ K½L
where D is the difference in chemical shift for the pure compound and that on addition of lipid, Dmax is the difference in chemical shift of the lipideflavone complex and the flavone, [L] is the concentration of free lipid and K is the binding constant. The fraction bound is the ratio of D/Dmax. A plot of fraction bound (FB) vs. D should be a straight line. 4.3. NMR and DSC experiments NMR experiments have been recorded on a BRUKER AVANCE 500 MHz NMR spectrometer, 2D-NOESY spectra were recorded using standard pulse programs [42] at 323 K, with a mixing time of 400 ms. 31P and 13C NMR experiments have been carried out using a relaxation delay of 1 s using broadband proton decoupling. The data has been processed using Topspin 2.0. DSC measurements have been performed using differential scanning calorimeter VP-
292
R. Sinha et al. / European Journal of Medicinal Chemistry 80 (2014) 285e294
Fig. 8. Change in CSA (Hz) represented by Ds ¼ sk st with temperature at (A) 1:10 (B) 1:5 and (C) 1:2, drug:lipid molar ratios in D2O of DPPC MLV (C), FLV (B), 5HF (◄), 6HF (D), 7HF (-), CHY (,) and BLN (A).
Fig. 9. Schematic representation of (A) orientation and (B) localization of flavones into lipid bilayer.
DSC (Microcal, Northampton, MA, USA) using the standard procedure used earlier [43]. Repeated scans for the same samples were generally super imposable. Data were analyzed with the software ORIGIN provided by Microcal.
4.5. Determination of antioxidant activity by DPPH assay Different concentrations of the test compounds were added to methanolic solution of DPPH and 0.1 M TriseHCl buffer. After an incubation period of thirty minutes, the absorbance was measured at 517 nm. IC50 values were determined graphically [44].
4.4. Sample preparation for NMR and DSC experiments 4.6. Determination of anti-proliferative activity MLVs were prepared using standard procedure [43] wherein the desired quantity of DPPC was dissolved in chloroform. The solvent was then evaporated with a stream of nitrogen so as, to deposit a lipid film on the walls of the container. The last traces of the solvent were removed with vacuum for a period of 1 h. MLV samples thus prepared were hydrated with the required amount of D2O at pH 7.2, which was followed by incubation in water bath at 50 C with repeated vortexing. The lipid concentrations for NMR samples were maintained at 100 mM while the concentrations of the flavones were varied from 10 to 50 mM. For DSC experiments, samples were prepared by mixing the lipid and drug solutions to obtain drug/lipid ratios from 1:20 to 1:2 by maintaining the lipid concentration to 50 mM. Unilamellar vesicles (ULV) for NMR experiments were prepared by sonicating the lipid dispersions with a Branson Sonicator-450 at 50% duty cycles till optical clarity was obtained.
Antiproliferative activity was evaluated by the Sulforhodamine B assay method [45]. Three different cell lines, viz. MCF-7, K562 and Table 3 In-vitro antioxidant activity of different flavones by DPPH radical scavenging method. Compound
DPPH radical scavenging activity IC50 (mg/ml)a
FLV 5HF 6HF 7HF CHY BLN
>100 23.8 22.1 24.2 20.4 18.2
a The molar drug concentrations required to cause 50% inhibition (IC50) were determined from doseeresponse curves. Results represent means SE of at least three different experiments.
R. Sinha et al. / European Journal of Medicinal Chemistry 80 (2014) 285e294 Table 4 Effect of different flavones (at a concentration of 104 mg/ml) on cell growth of different human cancer cell lines. All values are expressed as mean SD of at least three independent experiments. Flavone
FLV 5HF 6HF 7HF CHY BLN
[14]
[15]
% Growth of different human cancer cell lines HepG2
MCF-7
K562
7.6 88.8 82.8 98.8 76.7 70.6
50.3 28.6 28.8 26.6 25.9 8.9
56.8 26.7 13.3 22.3 38.3 16.7
HepG2 were used. Cell lines were grown in RPMI 1640 medium containing 10% fetal bovine serum and 2 mM of L-glutamine. The test compounds were dissolved in dimethyl sulfoxide and diluted suitably before adding them to the culture medium. After incubation at standard conditions for 48 h, percent growth inhibition has been calculated.
[16]
[17]
[18]
[19] [20] [21]
[22]
Acknowledgements [23]
Ragini Sinha and Girjesh Govil thank the Indian National Science Academy and Urmila Joshi thanks Department of Biotechnology for funding the above project. The help and co-operation provided by Ms. Mamta Joshi and Mr. Devidas at the National Facility for High Field NMR located at TIFR are gratefully acknowledged.
[25]
Appendix A. Supplementary data
[26]
Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.ejmech.2014.04.054.
[27]
References
[28]
[1] P.I. Oteiza, A.G. Erlejman, S.V. Verstraeten, C.L. Keen, C.G. Fraga, Flavonoidmembrane interactions: a protective role of flavonoids at the membrane surface? Clinical and Developmental Immunology 12 (2005) 19e25. [2] B. Sengupta, J. Swenson, Properties of normal and glycated human hemoglobin in presence and absence of antioxidant, Biochemical and Biophysical Research Communications 334 (2005) 954e959. [3] F.V. So, N. Guntherie, A.F. Chambers, M. Mousca, K.K. Carroll, Inhibition of human breast cancer cell proliferation and delay of mammary tumorigenesis by flavonoids and citrus juices, Nutrition and Cancer 26 (1996) 167e181. [4] G. Di Carlo, N. Mascalo, A.A. Izzo, F. Capasso, Flavonoids : old and new aspects of a class of natural therapeutic drugs, Life Sciences 65 (1999) 337e353. [5] A.B. Hendrich, R. Malon, A. Pola, Y. Shirataki, N. Motohashi, K. Michalak, Differential interaction of Sophora isoflavonoids with lipid bilayers, European Journal of Pharmaceutical Sciences 16 (2002) 201e208. [6] A. Arora, T.M. Byrem, M.G. Nair, G.M. Strasburg, Modulation of liposomal membrane fluidity by flavonoids and isoflavonoids, Archives of Biochemistry and Biophysics 373 (2000) 102e109. [7] S. Chaudhari, B. Pahari, P.K. Sengupta, Ground and excited state proton transfer and antioxidant activity of 7-hydroxyflavone in model membranes: absorption and fluorescence spectroscopic studies, Biophysical Chemistry 139 (2009) 29e36. [8] H. Tsuchiya, Structure-dependent membrane interaction of flavonoids associated with their bioactivity, Food Chemistry 120 (2010) 1089e1096. [9] J. Greef, J. Joubert, S.F. Malon, S. VaneDyk, Antioxidant properties of 4-quinolones and structurally related flavones, Bioorganic & Medicinal Chemistry 20 (2012) 809e818. [10] N. Cotelle, J.L. Bernier, J.P. Catteau, J. Pommery, J.C. Wallet, E.M. Gaydou, Antioxidant properties of hydroxy-flavones, Free Radical Biology & Medicine 20 (1996) 35e43. [11] G. Pushpavalli, P. Kalaiarasi, C. Veeramani, K.V. Pugalendi, Effect of chrysin on hepatoprotective and antioxidant status in D-galactosamine-induced hepatitis in rats, European Journal of Pharmacology 631 (2010) 36e41. [12] G.K. Harris, Y. Qian, S.S. Leonard, D.C. Sbarra, X. Shi, Luteolin and chrysin differentially inhibit cyclooxygenase-2 expression and scavenge reactive oxygen species but similarly inhibit prostaglandin-E2 formation in RAW 264.7 cells, The Journal of Nutrition 136 (2006) 1517e1521. [13] J. Sathiavelu, G.J. Senapathy, R. Devaraj, N. Namasivayam, Hepatoprotective effect of chrysin on prooxidant-antioxidant status during ethanol-induced
[24]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36] [37]
[38]
[39]
293
toxicity in female albino rats, The Journal of Pharmacy and Pharmacology 61 (2009) 809e817. T. Zhang, X. Chen, L. Qu, J. Wu, R. Cui, Y. Zhao, Chrysin and its phosphate ester inhibit cell proliferation and induce apoptosis in Hela cells, Bioorganic & Medicinal Chemistry 12 (2004) 6097e6105. E. Middleton, C. Kandaswami, T.C. Theoharides, The effects of plant flavonoids on mammalian cells: Implications for inflammation, heart disease and cancer, Pharmacological Reviews 52 (2002) 103e108. F. Ollila, K. Halling, P. Vuorela, H. Vuorela, J.P. Slotte, Characterization of flavonoid biomembrane interactions, Archives of Biochemistry and Biophysics 399 (2002) 103e108. A.J. Erlejman, S.V. Verstracten, C.G. Fraga, P.I. Oteiza, The interaction of flavonoids with membranes: potential determinant of flavonoid antioxidant effects, Free Radical Research 38 (2004) 1311e1320. H. Tsuchiya, M. Nagayama, T. tanaka, M. Furusawa, M. Kashimata, H. Takeuchi, Membrane-rigidifying effects of anticancer dietary factors, Biofactors 16 (2002) 45e56. H. Hausen, G. Poupartm, in: P. Yeagle (Ed.), The Structure of Biological Membranes, vol. 3, CRC Press, London, 1992. F. Frezard, Liposomes :From biophysics to the design of peptide vaccines, Brazilian Journal of Medical and Biological Research 32 (1999) 181e189. C.H. Huang, S. Li, Calorimetric and molecular mechanics studies of the thermotropic phase behavior of membrane phospholipids, Biochimica et Biophysica Acta 1422 (1999) 273e307. T. Ramstad, C.E. Hadden, G.E. Martin, S.M. Speaker, D.L. teagarden, T.J. Thamann, Determination by NMR of the binding constant for the molecular complex between alprostadil and a-cyclodextrin implications for a freeze-dried formulation, International Journal of Pharmaceutics 296 (2005) 55e63. M.P. Lambros, E. Sheu, J.S. Lin, H.A. Pereira, Interaction of a synthetic peptide based on the neutrophil derived antimicrobial protein CAP 37 with dipalmitoyl-phosphatidylcholine membranes, Biochimica et Biophysica Acta 1329 (1997) 285e290. M.K. Jain, Introduction to Biological Membranes, Wiley, New York, 1988, pp. 122e165. R.N. McEihaney, Differential scanning calorimetric studies of lipid-protein interactions in model membrane systems, Biochimica et Biophysica Acta 864 (1986) 361e421. R.F. Epand, A. Ramamoorthy, R.M. Epand, Membrane lipid composition and the interaction of pardaxin: the role of cholesterol, Protein & Peptide Letters 13 (2006) 1e5. A. Saija, F. Bonina, D. Trombetta, A. Tomaino, L. Montenegro, P. Smeriglio, F. Castelli, Flavone-biomembrane interactions: a calorimetric study on dipalmitoylphosphatidylcholine vesicles, International Journal of Pharmaceutics 124 (1995) 1e8. L. Panicker, Interaction of propyl paraben with dipalmitoyl phosphatidylcholine bilayer: a differential scanning calorimetry and nuclear magnetic resonance study, Colloids and surfaces. B: Biointerfaces 61 (2008) 145e152. D. BassolinoeKlimas, H.E. Alper, T.R. Stouch, Mechanism of solute diffusion through lipid bilayer membranes by molecular dynamic simulation, Journal of the American Chemical Society 117 (1995) 4118e4129. K. Kachel, E. AsuncionePunzalan, E. London, Anchoring of Trp and Tyr analogs at the hydrocarbon-polar boundary in membranes: parallax analysis of fluorescence quenching induced by nitroxide-labeled phospholipids, Biochemistry 34 (1995) 15475e15479. T. Kimura, K. Cheng, K.C. Rice, K. Gawrisch, Location, structure and dynamics of synthetic cannabinoid ligand CP-55,940 in lipid bilayers, Biophysical Journal 96 (2009) 4916e4924. H. Stamm, H. Jackel, Relative ring-current effects based on a new model for aromatic-solvent-induced shift, Journal of the American Chemical Society 111 (1989) 6544e6550. H.A. Scheidt, A. Pampel, L. Nissler, R. Glebhardt, D. Huster, Investigation of the membrane localization and distribution of flavonoids by high-resolution magic angle spinning NMR spectroscopy, Biochimica et Biophysica Acta 1663 (2004) 97e107. J. Jeener, B.H. Meier, P. Bachmann, R.R. Ernst, Investigation of exchange processes by two-dimensional NMR spectroscopy, The Journal of Chemical Physics 71 (1979) 4546e4553. A. Waghmare, M. Kanyalkar, M. Joshi, S. Srivastava, In-vitro metabolic inhibition and antifertility effect facilitated by membrane alteration: search for novel antifertility agent using nifedipine analogues, European Journal of Medicinal Chemistry 46 (2011) 3581e3589. K.V.R. Chary, G. Govil, NMR in Biological Systems, from Molecules to Humans, Springer, The Netherlands, 2008. S. Srivastava, R.S. Phadke, G. Govil, Role of tryptophan in inducing polymorphic phase formation in lipid dispersions, Indian Journal of Biochemistry and Biophysics 25 (1988) 283e286. C. D’Souza, M. Kanyalkar, M. Joshi, E. Coutino, S. Srivastava, Search for novel neuraminidase inhibitors: design, synthesis and interaction of oseltamivir derivatives with model membrane using docking, NMR and DSC methods, Biochimica et Biophysica Acta 178 (2009) 1740e1751. S. Van Acker, M.J. deGroot, D.J. van den Berg, M. Tromp, G. den Kelder, W. Vijgh, A. Bast, A quantum chemical explanation of the antioxidant activity of flavonoids, Chemical Research in Toxicology 9 (1996) 1305e1312.
294
R. Sinha et al. / European Journal of Medicinal Chemistry 80 (2014) 285e294
[40] Hong-Yu Zhang, Structure-Activity relationships and rational design strategies for radical scavenging antioxidants, Current Computer-Aided Drug Design 1 (2005) 257e274. [41] H.Y. Cheng, C.S. Randall, W.W. Holl, P.P. Constantinides, T.L. Yue, G.Z. Feuerstein, Carvedilol-liposome interaction: evidence for strong association with the hydrophobic region of the lipid bilayers, Biochimica et Biophysica Acta 1284 (1996) 20e28. [42] W.P. Aue, E. Bartholdi, R.R. Ernst, Two-dimensional spectroscopy-application to nuclear magnetic resonance, The Journal of Chemical Physics 64 (1976) 2229e2246.
[43] R. Sinha, M.K. Gadhwal, U.J. Joshi, S. Srivastava, G. Govil, Interaction of quercetin with DPPC model membrane: molecular dynamic simulation, DSC and multinuclear NMR studies, Journal of the Indian Chemical Society 88 (2011) 1e8. [44] P. Molyneux, The use of the stable free radical diphenylpicrylhydrazyl (DPPH) for estimating antioxidant activity, Songklanakarin Journal of Science and Technology 26 (2004) 211e219. [45] V. Vichai, K. Kirtikara, Sulforhodamine B colorimetric assay for cytotoxicity screening, Nature Protocols 1 (2006) 1112e1116.