PCCP View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
PAPER
Cite this: Phys. Chem. Chem. Phys., 2014, 16, 21657
View Journal | View Issue
Interaction of insulin with anionic phospholipid (DPPG) vesicles Bidisha Tah,a Prabir Pal,a Sabyashachi Mishrab and G. B. Talapatra*a The interaction between a protein/enzyme and a lipid is critical for pharmacological activity. Here, we study the interaction between insulin and the 1,2-dipalmitoyl-sn-glycero-3-phosphoglycerol (DPPG) lipid anionic vesicle by successfully entrapping the insulin molecule into DPPG vesicles, which are biocompatible liposomes. For the insulin–DPPG complex system, steady state emission spectroscopy at room temperature (300 K) shows a new broad and structured peak between 400 nm and 500 nm along with the tyrosine fluorescence peak at 303 nm. Temperature dependent and time resolved spectroscopy
Received 10th July 2014, Accepted 12th August 2014
reveal that the peak between 400 nm and 500 nm in the insulin–DPPG system arises due to the tyrosine
DOI: 10.1039/c4cp03028a
the liposome. A molecular dynamics study of the tyrosine–DPPG system shows that the rigidity of tyrosine
www.rsc.org/pccp
increases in the lipid layer. Dynamic light scattering (DLS), and zeta potential studies also establish the attachment of insulin with the anionic liposome.
phosphorescence phenomenon. This phosphorescence peak is the signature of insulin entrapment into
1. Introduction Insulin is a hormone, secreted from b cells of the pancreas as a proinsulin precursor and converted to insulin by enzymatic cleavage.1–3 An insulin molecule is composed of 51 amino acids, and arranged into two polypeptide chains, A and B. Two inter-chain disulphide bridges connect these chains. In the secondary structure, chain A consists of two anti-parallel a helices (A2 to A8 and A13 to A20) while chain B forms a single a helix (B9 to B19) followed by a turn and a b strand (B21 to B30).4,5 Insulin can associate into a monomer, dimers, tetramers, hexamers, and higher order aggregates.6 The monomer is the active form of insulin and these are stored as crystalline zinc bound hexamers in vesicles within pancreatic b cells from which secretion occurs in response to elevated blood glucose levels.7 Thus, insulin is mainly used for the treatment of diabetes. The use of insulin requires painful injections for diabetic patients. Oral delivery of insulin is a dream of patients and a big challenge for scientists. Many researchers are involved in developing non-toxic, stable, bioactive oral insulin delivery systems.8–10 Different approaches, such as chemical modification11 and co-administration with absorption enhancers and/or enzyme inhibitors have been tried. Insulin incorporation into carriers,12 such as liposomes,13,14 mixed micelles, lipid-based
a
Department of Spectroscopy, Indian Association for the Cultivation of Science, Jadavpur, Kolkata-700 032, India. E-mail: [email protected]; Fax: +91-33-24732805; Tel: +91-33-24734971 b Department of Chemistry, Indian Institute of Technology, Kharagpur-721 302, India
This journal is © the Owner Societies 2014
systems, microspheres, and nanoparticles15–18 has been explored to improve the bioavailability of insulin. Lipid vesicles, known as liposomes, are widely recognized as pharmaceutical carriers because of their low toxicity, biodegradability, and biocompatibility.19 Phospholipid vesicles immobilized on solid substrates have been considered as models for cell membranes.20,21 Vesicles have good longevity in the blood and can specifically target disease sites when various targeting ligands are attached to their membranes. In a recent publication, a sodium dodecyl sulfate (SDS)–cetyl trimethyl ammonium bromide (CTAB) catanionic vesicle was developed22 and insulin was immobilized in the vesicular membrane.23 The SDS–CTAB catanionic vesicle was used as a model system due to its easy availability, but it is not a biocompatible system and cannot be used as a drug delivery tool. Previous research23 showed that CTAB and insulin formed a charge transfer complex. Thus, for the SDS–CTAB system, if not all the CTAB is involved in catanionic vesicle formation, the free CTAB may denature insulin. 1,2-Dipalmitoyl-sn-glycero-3phosphoglycerol (DPPG) is chosen here due to its biocompatible nature. Moreover, it is suitable for drug delivery due to its smaller vesicular size. Several authors24–27 have studied protein/drug immobilization in phosphatidylcholine or lecithin liposomal systems, but many of these studies are in the context of pharmacological applications. The detailed mechanism by computer simulation and spectroscopic studies has not yet been investigated. However, the study of the insulin–DPPG complex is rather rare. In this paper, our main objective is to study the insulin– vesicle complex formation and the entrapment of insulin into the vesicular membrane. To confirm the self-assembled structure
Phys. Chem. Chem. Phys., 2014, 16, 21657--21663 | 21657
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
Paper
PCCP
formation of the insulin and vesicles, we have used emission and excitation spectroscopy, dynamic light scattering (DLS), and zeta potential measurements. To establish the entrapment of insulin into the vesicular membrane, we have also performed molecular dynamics (MD) simulations. The results establish the fact that, after the interaction of insulin with the vesicles, a new structured broad phosphorescence peak appears at room temperature aside from the tyrosine fluorescence. The appearance of the phosphorescence peak is the signature of insulin entrapment into the vesicular membrane.
2. Experimental section and computational method 2.1.
Materials
The anionic bi-tail lipid, 1,2-dipalmitoyl-sn-glycero-3-phosphoglycerol (DPPG), was purchased from Avanti Polar Lipids, Inc. (Alabaster, Alabama). The chloroform (UV Grade) and methanol were purchased from Spectrochem and Sisco Research Laboratory, (Mumbai, India), respectively. The human insulin (Actrapid) was purchased from Abbott India Limited, (Mumbai, India). 2.2.
Experimental procedures
2.2.(A) Preparation of vesicular solution and insulin–vesicle complex. The stock solution of DPPG, of concentration 0.2 mM, was prepared in a 1 : 1 chloroform and methanol solvent. The prepared lipid solution was heated in a water bath at 70 1C to evaporate the solvent. Millipore water was then added to form vesicles at 45 1C, which is above the phase transition temperature (41 1C) of DPPG.28,29 The aqueous solution of insulin, of concentration 0.4 mg ml1 (pH 7.0), was added during the vesicle formation and the system was sonicated for 10 min. 2.2.(B) Spectroscopic characterization. Fluorescence and excitation spectra were recorded for pure insulin and the insulin–DPPG system at 25 1C, using a cuvette with a path length of 1 cm on a fluorescence spectrometer (HITACHI F-4500). Using liquid nitrogen as a coolant, the temperature (from room temperature to liquid nitrogen temperature) dependent emission spectra were recorded. A phosphorescence decay profile was also recorded using this spectrometer. 2.2.(C) Dynamic light scattering. A Brookhaven BI-200SM goniometer (Brookhaven Instruments Corporation) with a 35 mW He–Ne laser (633 nm) was used to study the dynamic light scattering. A non-negativity constrained least-squares algorithm was used to determine the mean diameter of the molecules. 2.2.(D) Zeta-potential measurements. A Malvern Zetasizer Nano ZS (part no.: ZEN3600) instrument was used for zeta potential measurements at a scattering angle of 1751 and a detection angle of 12.81 with a He–Ne laser (power 4 mW, wavelength 632.8 nm and beam diameter 0.63 nm (1/e2)). The experiment was performed at 25 1C. On one hand, the machine has a size measurement range of 3.81 nm and on the other hand, there was no practical range for the zeta-potential measurement. Standard Malvern zeta-potential disposable capillary cells were used to hold the vesicular solutions and
21658 | Phys. Chem. Chem. Phys., 2014, 16, 21657--21663
insulin–vesicle system for the zeta-potential measurements. For verification, the experiment was repeated three times. 2.3.
Computational method
2.3.(A) Preparation of the insulin monomer. The initial structure of the insulin monomer was obtained by considering chains A and B of the crystal structure of the insulin hexamer (PDB ID 3AIY), shown in Fig. 1. Chains A and B consist of 21 and 30 amino acids, respectively. All Glu, Lys, and Arg side chains were modeled as charged, and all His were kept neutral. Three disulfide bonds between Cys residue pairs 6 (chain A)–11 (chain A), 7 (chain A)–7 (chain B) and 20 (chain A)–19 (chain B) were added. The N- and C-termini of both chains were capped with acetyl and methylamine groups, respectively. The missing hydrogen atoms were added and the overall charge (2) was neutralized by adding 2 potassium ions. 2.3.(B) Insulin in water. The protein was immersed in a cubic water box of side 60 Å, which ensured a 10 Å distance between the boundary of the water box and any solute atoms. A 150 mM salt concentration was achieved by adding 18 potassium and chloride ions. The simulation system contained 6419 water molecules and about 20 100 atoms in total, shown in Fig. 2A. The MD simulations were carried out at constant temperature and constant pressure (1 atm) using a Nose–Hoover Langevin thermostat and piston.30,31 Long range electrostatic interactions were treated by the particle mesh Ewald method with a 12 Å cutoff. Van der Waals interactions were truncated at a cutoff of 12 Å and a switch function was activated starting at 10 Å.32 The SHAKE algorithm33 was used to fix the length of the covalent bonds involving hydrogen atoms, which allowed an integration time step of 2 fs. Structures were saved every 1 ps for analysis. All simulations were performed employing the NAMD program34 with the CHARMM22 force field35 and TIP3 potential for water molecules.36 The system was first equilibrated for 2 ns at two different temperatures (100 K and 300 K), followed by equilibrium dynamics for 20 ns each at the two temperature values. 2.3.(C) Insulin in the lipid bilayer. When a lipid bilayer is rolled into an enclosed cell vesicles are formed. That is why we have modeled insulin in the lipid bilayer in the simulation. The initial structure of the monomer was oriented to align the vector between the Ca atoms of Cys7 (chain B) and Val18
Fig. 1 Crystal structures of the (A) insulin hexamer and (B) insulin monomer (PDB ID 3AIY).
This journal is © the Owner Societies 2014
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
PCCP
Fig. 2
Paper
(A) Insulin monomer solvated in a cubic water box. (B) Insulin monomer in the DPPG lipid bilayer. (C) Insulin monomer in the DPPG lipid bilayer and water.
(chain B) with the z-axis, and was then placed in a homogeneous lipid bilayer of DPPG molecules by following the so-called insertion method as implemented in CHARMM-GUI.37–39 The resulting system is shown in Fig. 2B. The number of DPPG molecules was determined by ensuring at least three lipid molecules were placed between two proteins in the primary and image systems.38 The insulin–lipid system was then solvated by adding a rectangular water box to maintain a water thickness of 15 Å on top of and below the protein. The system was then neutralized and a salt concentration of 150 mM was achieved by adding potassium and chloride ions. The final system contained the insulin monomer with 96 DPPG molecules, 3668 water molecules, 106 potassium ions, and 8 chloride ions, with a total of about 24 000 atoms, shown in Fig. 2C. The system was subjected to gradual equilibration for 375 ps as described in previous literature,38 followed by free equilibrium MD simulation for 20 ns, with the same cutoff parameters as described in the insulin in water system. Another MD simulation was carried out using the same procedure but at 100 K.
3. Results and discussion Fluorescence spectroscopy can be employed as a highly sensitive method to study protein structure, conformation, and kinetics. Intrinsic protein fluorescence generally arises from three
fluorophores, i.e., tryptophan (Trp), tyrosine (Tyr) and phenylalanine (Phe).40,41 Among these, the spectroscopic behavior of a protein is normally dominated by Trp, as the quantum yield of Trp is highest.42,43 However, in the case of insulin, tyrosine plays an important role for spectroscopic characterization, as tryptophan is absent. Fig. 3A and B show the peak normalized emission and excitation spectra of pure insulin and the insulin–DPPG system at room temperature (300 K), respectively. The emission spectrum of pure insulin (Fig. 3A, curve-a) shows a single peak around 305 nm, which is due to tyrosine fluorescence. In the absence of tryptophan in insulin, tyrosine is responsible for the emission peaks. Very interestingly, in the presence of vesicles, the emission spectrum of insulin drastically differs from that of only the protein. In curve-b of Fig. 3A, the peak around 303 nm is due to the fluorescence of the tyrosine residues. In comparison with pure insulin, this peak marginally shifts towards the blue end of the spectrum and shows a smaller full width at half maximum. This shift may presumably reflect the change in local environment upon movement of some tyrosine residues from an aqueous environment to the vesicular lipid membrane.44 Apart from this peak, another new structured broad band centered around 450 nm appears, possibly due to the phosphorescence of tyrosine. The discussion about this band is described below. Fig. 3B shows the excitation spectra of the pure insulin (curve-a) and insulin–DPPG complexes (curve-b). Both cases show
Fig. 3 (A) Emission spectra (monitored at lex = 275 nm) of pure insulin (curve-a) and the insulin–DPPG complex (curve-b) and (B) excitation spectra (monitored at lem = 305 nm) of pure insulin (curve-a) and the insulin–DPPG complex (curve-b).
This journal is © the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 21657--21663 | 21659
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
Paper
two excitation peaks. In pure insulin, these peaks, at 228 nm and 277 nm, originate from 1La and 1Lb transitions of the tyrosine phenol ring, respectively.45 In the case of the insulin–DPPG system, the higher energy band (228 nm) appears at the same wavelength as that of pure insulin, however, with reduced intensity, whereas the lower energy band (272 nm) is blue shifted. This shift is due to the interaction between the lipid and protein molecules. The change in relative intensity after vesicle inclusion may be due to the restriction of the 1Lb vibration. In many previous literature studies,46–48 a low temperature phosphorescence peak around 450 nm was observed for tyrosine. To confirm whether the emission peak around 450 nm in this case is due to phosphorescence, we have conducted a temperature dependent steady state and time resolved emission study of the long wavelength band of the insulin–DPPG complex. Fig. 4A shows the results. Between 350 nm and 500 nm, a broad peak appears at each temperature. With decreasing temperature, the intensity of the band increases drastically. The intensity is highest at liquid nitrogen temperature (100 K), whereas it is lowest at room temperature (300 K). Fig. 4B shows the lifetime of the insulin–DPPG complex (at lem = 430 nm) at different temperatures. It indicates that the lifetime decreases with the increase of temperature. The lifetime of the insulin–DPPG complex at room temperature is about 0.22 ms, whereas at liquid nitrogen temperature it increases to 2.3 ms. We have tried to fit the temperature dependent phosphorescence lifetime using a well known kinetic equation found in previous literature:49 1 1 DE þ constant (1) ¼ ln t t0 kT where t is the lifetime, t0 is the intrinsic lifetime, DE is the activation energy, k is the Boltzmann constant, and T is the temperature. Eqn (1) can be modified to eqn (2) by considering t0, DE, and k as constants: C (2) t¼1 A þ Be T where A, B, and C are the constants.
PCCP
Here we have fitted the data with eqn (2) (shown by the smooth curve in Fig. 4B). The R2 value is 0.97. From the curve, the millisecond order long lifetime and lifetime dependency of the temperature confirm that the structured broad emission peak between 400 nm and 500 nm is due to phosphorescence, originating due to the rigid geometry of the tyrosine molecule.50 It is to be noted that the emission spectrum at low temperature does not show a structured broad peak. At freezing conditions, the orientation of every insulin molecule in the insulin–DPPG complex is different. Due to this reason, in the frozen glassy state, the fine structures of the broad peak are not well resolved (Fig. 4A). Even at room temperature, when is insulin entrapped into the liposome membrane, the movement of tyrosine is restricted. In this restricted tyrosine geometry, a radiative transition between the lowest triplet state and ground state (intersystem crossing) becomes more efficient and we observe phosphorescence at room temperature. Fig. 5 shows particle size distribution curves resulting from DLS experiments on the DPPG–insulin system (curve-a) and DPPG vesicles (curve-b). The size of a DPPG vesicle is within 150–300 nm, which is much smaller than a SDS–CTAB catanionic vesicle23 and thus more penetrable and useful in drug delivery. In
Fig. 5 Size distribution of the insulin–DPPG complex (curve-a) and DPPG vesicles (curve-b). The symbols show the data and red lines show the fitted curves. The black and blue curves are the deconvoluted curves for the insulin–DPPG complex and DPPG vesicles, respectively.
Fig. 4 (A) Emission spectra (monitored at lex = 275 nm) of the insulin–DPPG complex at different temperatures. (B) Lifetime of the insulin–DPPG complex at different temperatures with a fitted curve.
21660 | Phys. Chem. Chem. Phys., 2014, 16, 21657--21663
This journal is © the Owner Societies 2014
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
PCCP
each case, two peaks are observed. These peaks indicate that there are two aggregation sizes in the system. By deconvulation of these curves using Microcal Origin 8, we get two peaks around 177 nm and 241 nm for the DPPG–insulin system (R2 = 0.998), whereas, similar peaks originate in a lower region (157 nm and 222 nm) for the pure DPPG system (R2 = 0.999). This observation clearly illustrates that the size enhancement occurs due to insulin entrapment into the lipid vesicles. The measured zeta-potentials of the insulin–DPPG system and pure DPPG vesicular system at pH 7.0 are 41 mV and 53 mV, respectively. The negative zeta potential is due to the anionic nature of the DPPG lipid.51 The zeta potential becomes less negative in the case of the DPPG–insulin system. We have added insulin during the preparation of the DPPG vesicles and, due to the attachment of insulin molecule with the vesicle membrane, the effective surface charge density at the outer surface may decrease. This result supports the previous results and confirms the entrapment of insulin into the DPPG vesicles. A molecular dynamics study can be a very useful technique to elucidate the origin of the experimental observations. In the case of insulin, the tyrosine residues play an important role in the spectroscopy of the molecule. A detailed study of tyrosine in insulin can reveal the real facts about the encapsulation of insulin in the lipid vesicles. Fig. 6A shows the root mean square fluctuation (RMSF) of the non-hydrogen atoms of the Tyr residues of the insulin monomer averaged over intervals of 2 ns molecular dynamic trajectories. The standard deviations of the RMSF values of different Tyr residues are shown as error bars. The RMSF values of the Tyr residues when insulin is solvated in water and in the DPPG lipid bilayer are shown as black diamonds (curve-a) and red diamonds (curve-b), respectively. Fig. 6A suggests a reduced fluctuation of all four Tyr residues when the protein is in the lipid bilayer compared to their respective RMSF values in water. In particular, the reduced movement of Tyr14 (chain A) in the lipid bilayer is significant, while the same effect is moderate for Tyr19 (chain A) and Tyr16 (chain B). Among these four tyrosine
Paper
residues, Tyr14 is most exposed to the lipid and therefore, it interacts strongly with the lipid, as reflected by the reduced fluctuation of Tyr14 in the lipid bilayer. Fig. 6B shows the positions of the tyrosine residues in the protein monomer during the interaction with the lipid. At 100 K the system becomes totally frozen and the fluctuations of all tyrosine residues in both water and lipid media are more or less the same and much reduced from the fluctuations seen at 300 K (Fig. 6A). From this observation, it can be concluded that tyrosine becomes rigid after the entrapment of insulin into the liposome membrane as well as at low temperature. This rigid geometry of tyrosine is responsible for the phosphorescence phenomenon at room temperature. For further information regarding the interaction between the lipid and protein, we have calculated the number of contacts between the DPPG atoms and insulin residues and the interaction energy between individual residues and lipid molecules. Fig. 7A shows the average number of contacts each residue makes with the lipid molecules which are within 4 Å of the corresponding residue. Fewer contacts are seen for the Ile2, Val3, Cys6, Cys7, Leu16, Glu17, Cys20 residues of chain A and His5, Leu15 and Cys19 of chain B. These residues form an a-helix segment in the protein, suggesting that the alpha helical segment has much less interaction with the lipid molecules. On the other hand, a larger number of contacts for His10 of chain A and Phe1 and Leu17 of chain B suggest that they play key roles in mediating the protein–lipid interaction. This calculation also shows that the average number of contacts of all four Tyr residues with the lipid molecules is higher than 25, suggesting the important role of Tyr in the protein–lipid interaction, which supports the experimental observations. To identify the residues contributing to the insulin–DPPG interaction, we have plotted the average interaction energy of each amino acid residue with the DPPG lipid molecules during the MD simulation in Fig. 7B. A strong favorable interaction energy of Glu4 (chain-A) and Glu13 (chain-B) with the DPPG molecules is seen. The origin of this favorable interaction
Fig. 6 (A) The RMSF values of the Tyr residues when insulin is solvated in water and in the DPPG lipid bilayer are shown as curve-a (300 K) and curve-c (100 K) and as curve-b (300 K) and curve-d (100 K), respectively. The standard deviations of the RMSF values of different Tyr residues are shown as error bars. (B) The insulin monomer in the DPPG layer after interactions. Four tyrosine residues are shown in ball–stick form and DPPG is shown as lines.
This journal is © the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 21657--21663 | 21661
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
Paper
PCCP
Fig. 7 (A) The number of contacts between DPPG atoms (within 4 Å) and residues of insulin for chain-A (curve-a) and chain-B (curve-b). The standard deviations of the contact values of different residues are shown as error bars. (B) The interaction energy between DPPG and the insulin residues of chain-A (curve-a) and chain-B (curve-b). The standard deviations of the energy values of different residues are shown as error bars.
energy can be traced back to the interaction of the carboxylate groups of Glu4 (chain A) and Glu13 (chain B) with the phosphorus atom of the DPPG head group, which carries a positive partial charge.52 On the other hand, the positively charged side chains of Arg22 (chain-B) show a stronger interaction with the negatively charged oxygen atoms of the lipid head group. Similar interactions are also responsible for the positive value of the interaction energy between Glu17 (chain A) and the lipid head group. In summary, the lipid head group is found to stabilize both acidic and basic residues depending on the relative orientation of the side chains of the amino acid residue and the lipid molecule.
4. Conclusion We have successfully entrapped insulin into the DPPG vesicle membrane. After entrapment, the protein remains in its native state. We have studied steady state and time resolved emission spectroscopy. A newly observed structured broad band, centered at 450 nm, is the signature of the entrapment of the insulin molecule into the vesicle membrane. From the lifetime measurements, we assigned the peak as the phosphorescence emission of tyrosine, which originates due to the limited movement of Tyr residues arising from their interaction with the lipid molecule. This is further supported by MD simulations of the insulin monomer in water and lipid bilayer media. The decrease in the RMSF of the Tyr residues of insulin in the lipid bilayer compared to in water establishes the fact that the Tyr residues become rigid after the entrapment of insulin into the vesicle membrane. Since DPPG–insulin is a biocompatible system, the present study is expected to be useful and effective in the field of drug delivery, biosensors, and bioengineering.
Acknowledgements We thank DST, Government of India (Project No. SR/S2/CMP0079/2010(G)) for partial financial support and the IACS for providing central instrumental facilities. SM thanks DST, India for the Ramanujan fellowship.
21662 | Phys. Chem. Chem. Phys., 2014, 16, 21657--21663
References 1 Y. Date, M. Nakazato, S. Hashiguchi, K. Dezaki, M. S. Mondal, H. Hosoda, M. Kojima, K. Kangawa, T. Arima, H. Matsuo and T. Yada, Diabetes, 2002, 51, 124–129. 2 Y. Itoh, Y. Kawamata, M. Harada, M. Kobayashi, R. Fujii, S. Fukusumi, K. Ogi, M. Hosoya, Y. Tanaka, H. Uejima, H. Tanaka, M. Maruyama, R. Satoh, S. Okubo, H. Kizawa, H. Komatsu, F. Matsumura, Y. Noguchi, T. Shinohara, S. Hinuma, Y. Fujisawa and M. Fujino, Nature, 2003, 422, 173–176. 3 M. N. Poy, L. Eliasson, J. Krutzfeldt, S. Kuwajima, X. Ma, P. E. MacDonald, S. Pfeffer, T. Tuschl, N. Rajewsky, P. Rorsman and M. Stoffel, Nature, 2004, 432, 226–230. 4 J. L. Whittingham, D. J. Scott, K. Chance, A. Wilson, J. Finch, J. Brange and G. G. Dodson, J. Mol. Biol., 2002, 318, 479–490. 5 G. D. Smith, D. C. Swenson, E. J. Dodson, G. G. Dodson and C. D. Reynolds, Proc. Natl. Acad. Sci. U. S. A., 1984, 81, 7093–7097. 6 B. Tah, P. Pal, S. Roy, D. Dutta, S. Mishra, M. Ghosh and G. B. Talapatra, Spectrochim. Acta, Part A, 2014, 129, 345–351. 7 R. K. Zalups and D. J. Koropatnick, Cellular and Molecular Biology of Metals, CRC Press, 2010. ´, P. Maincent and N. Ubrich, J. Controlled Release, 8 C. Damge 2007, 117, 163–170. 9 M. A. Kisel, L. N. Kulik, I. S. Tsybovsky, A. P. Vlasov, M. S. Vorob’yov, E. A. Kholodova and Z. V. Zabarovskaya, Int. J. Pharm., 2001, 216, 105–114. 10 K. Iwanaga, S. Ono, K. Narioka, K. Morimoto, M. Kakemi, S. Yamashita, M. Nango and N. Oku, Int. J. Pharm., 1997, 157, 73–80. 11 H. Ashada, T. Douen, Y. Mizokoshi, T. Fujita, M. Murakami, A. Yamamoto and S. Muranishi, J. Pharm. Sci., 1995, 84, 682–687. 12 E. Ziv, O. Lior and M. Kidron, Biochem. Pharmacol., 1987, 36, 1035–1039. 13 M. A. Kisel, L. N. Kulik, I. S. Tsybovsky, A. P. Vlasov, M. S. Vorobyov, E. A. Kholodova and Z. V. Zabarovskaya, Int. J. Pharm., 2001, 216, 105–114.
This journal is © the Owner Societies 2014
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
PCCP
14 J. H. Fendler and A. Romero, Life Sci., 1977, 20, 1109–1120. 15 J. Scott-Moncrieff, Z. Shao and K. Mitra, J. Pharm. Sci., 1994, 83, 1465–1469. ´ron, F. Puisieux and M. Seiller, Int. 16 A. Silva-Cunha, J. Che J. Pharm., 1997, 158, 79–89. 17 B. Y. Kim, J. Jeonga, K. Parkb and J. D. Kim, J. Controlled Release, 2005, 102, 525–538. 18 B. Sarmento, S. Martins, A. Rebeiro, F. Veiga, R. Neufeld and D. Ferreira, Int. J. Pept. Res. Ther., 2006, 12, 131–138. 19 A. Akbarzadeh, R. Rezaei-Sadabady, S. Davaran, S. W. Joo, N. Zarghami, Y. Hanifehpour, M. Samiei, M. Kouhi and K. Nejati-Koshki, Nanoscale Res. Lett., 2013, 8, 102. 20 B. Tah, P. Pal, M. Mahato and G. B. Talapatra, J. Phys. Chem. B, 2011, 115, 8493–8499. 21 T. Kamilya, P. Pal, M. Mahato and G. B. Talapatra, J. Nanosci. Nanotechnol., 2009, 9, 2956–2964. 22 T. Kamilya, P. Pal, M. Mahato and G. B. Talapatra, Mater. Sci. Eng., C, 2009, 29, 1480–1485. 23 B. Tah, P. Pal and G. B. Talapatra, J. Lumin., 2014, 145, 81–87. 24 K. Ariga, Y. M. Lvov, K. Kawakami, Q. Ji and J. P. Hill, Adv. Drug Delivery Rev., 2011, 63, 762–771. 25 M. A. Kisel, L. N. Kulik, I. S. Tsybovsky, A. P. Vlasov, M. S. Vorobyov, E. A. Kholodova and Z. V. Zabarovskaya, Int. J. Pharm., 2001, 216, 105–114. 26 X. Zhang, J. Qi, Y. Lu, X. Hu, W. He and W. Wu, Nanoscale Res. Lett., 2014, 9, 185. 27 R. Mo, T. Jiang, J. Di, W. Tai and Z. Gu, Chem. Soc. Rev., 2014, 43, 3595–3629. 28 R. Wohlgemuth, N. Waespe-Sarcevic and J. Seelig, Biophys. J., 1980, 19, 3315–3321. 29 T. Zaraiskaya and K. R. Jeffrey, Biophys. J., 2005, 88, 4017–4031. 30 G. Martyna, D. Tobias and M. Klein, J. Chem. Phys., 1994, 101, 4177–4189. 31 S. E. Feller, Y. Zhang, R. Pastor and B. R. Brooks, Chem. Phys., 1995, 103, 4613–4621. 32 P. J. Steinbach and B. R. Brooks, J. Comput. Chem., 1994, 15, 667–683. 33 W. Van Gunsteren and H. Berendsen, Mol. Phys., 1977, 34, 1311–1327.
This journal is © the Owner Societies 2014
Paper
34 J. C. Phillips, R. Braun, W. Wang, J. Gumbart, ´ E. Tajkhorshid, E. Villa, C. Chipot, R. D. Skeel, L. Kale and K. Schulten, J. Comput. Chem., 2005, 26, 1781–1802. 35 A. D. Mac Kerell, Jr., D. Bashford, M. Bellott, R. L. Dunbrack, Jr., J. D. Evan-seck, M. J. Field, S. Fischer, J. Gao, H. Guo, S. Ha, D. Joseph-McCarthy, L. Kuchnir, K. Kuczera, F. T. K. Lau, C. Mattos, S. Michnick, T. Ngo, D. T. Nguyen, B. Prodhom, W. E. Reiher, B. Roux, M. Schlenkrich, J. C. Smith, R. Stote, J. Straub, M. Watanabe, J. Wiorkiewicz-Kuczera, D. Yin and M. Karplus, J. Phys. Chem. B, 1998, 102, 3586–3616. 36 W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey and M. L. Klein, J. Chem. Phys., 1983, 79, 926–935. 37 S. Jo, J. B. Klauda and W. Im, J. Comput. Chem., 2008, 29, 1859–1865. 38 S. Jo, T. Kim and W. Im, PLoS One, 2007, 2(9), e880. 39 S. Jo, J. B. Klauda and W. Im, Biophys. J., 2009, 97, 50–58. 40 M. Mahato, P. Pal, T. Kamilya, R. Sarkar, A. Chawdhuri and G. B. Talapatra, J. Phys. Chem. B, 2010, 114, 7062–7070. 41 P. Pal, M. Mahato, T. Kamilya and G. B. Talapatra, Phys. Chem. Chem. Phys., 2011, 13, 9385–9396. 42 D. B. Wetlaufer, Adv. Protein Chem., 1962, 17, 303–390. 43 J. T. Vivian and P. R. Callis, Biophys. J., 2001, 80, 2093–2109. 44 H. S. Galdiero, A. Falanga, M. Vitiello, L. Raiola, L. Russo, C. Pedone, C. Isernia and M. Galdiero, J. Biol. Chem., 2010, 285, 17123–17136. 45 J. R. Lakowicz, Topics Fluorescence Spectroscopy, Kluwer academic publishers, 2002, vol. 3. 46 T. Truong, R. Bersohn, P. Brumer, C. K. Ltjk and T. Tao, J. Biol. Chem., 1967, 242, 2979–2985. 47 R. F. Steiner and R. Kolinski, Biochemistry, 1968, 7, 1014–1018. 48 K. W. Rousslang, P. J. Reid, D. M. Holloway, D. R. Haynes, J. Dragavon and J. B. A. Ross, J. Protein Chem., 2002, 21, 547–555. 49 I. G. Leubner, J. Phys. Chem., 1970, 74, 77–82. 50 G. B. Strambini and E. Gabellieri, Biophys. J., 1996, 70, 971–976. 51 H. Bensikaddour, K. Snoussi, L. Lins, F. V. Bambeke, P. M. Tulkens, R. Brasseur, E. Goormaghtigh and M. P. MingeotLeclercq, Biochim. Biophys. Acta, Biomembr., 2008, 1778, 2535–2543. 52 Y. N. Kaznessis, S. Kim and R. G. Larson, J. Mol. Biol., 2002, 322, 569–582.
Phys. Chem. Chem. Phys., 2014, 16, 21657--21663 | 21663
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
PAPER
Cite this: Phys. Chem. Chem. Phys., 2014, 16, 21657
View Journal | View Issue
Interaction of insulin with anionic phospholipid (DPPG) vesicles Bidisha Tah,a Prabir Pal,a Sabyashachi Mishrab and G. B. Talapatra*a The interaction between a protein/enzyme and a lipid is critical for pharmacological activity. Here, we study the interaction between insulin and the 1,2-dipalmitoyl-sn-glycero-3-phosphoglycerol (DPPG) lipid anionic vesicle by successfully entrapping the insulin molecule into DPPG vesicles, which are biocompatible liposomes. For the insulin–DPPG complex system, steady state emission spectroscopy at room temperature (300 K) shows a new broad and structured peak between 400 nm and 500 nm along with the tyrosine fluorescence peak at 303 nm. Temperature dependent and time resolved spectroscopy
Received 10th July 2014, Accepted 12th August 2014
reveal that the peak between 400 nm and 500 nm in the insulin–DPPG system arises due to the tyrosine
DOI: 10.1039/c4cp03028a
the liposome. A molecular dynamics study of the tyrosine–DPPG system shows that the rigidity of tyrosine
www.rsc.org/pccp
increases in the lipid layer. Dynamic light scattering (DLS), and zeta potential studies also establish the attachment of insulin with the anionic liposome.
phosphorescence phenomenon. This phosphorescence peak is the signature of insulin entrapment into
1. Introduction Insulin is a hormone, secreted from b cells of the pancreas as a proinsulin precursor and converted to insulin by enzymatic cleavage.1–3 An insulin molecule is composed of 51 amino acids, and arranged into two polypeptide chains, A and B. Two inter-chain disulphide bridges connect these chains. In the secondary structure, chain A consists of two anti-parallel a helices (A2 to A8 and A13 to A20) while chain B forms a single a helix (B9 to B19) followed by a turn and a b strand (B21 to B30).4,5 Insulin can associate into a monomer, dimers, tetramers, hexamers, and higher order aggregates.6 The monomer is the active form of insulin and these are stored as crystalline zinc bound hexamers in vesicles within pancreatic b cells from which secretion occurs in response to elevated blood glucose levels.7 Thus, insulin is mainly used for the treatment of diabetes. The use of insulin requires painful injections for diabetic patients. Oral delivery of insulin is a dream of patients and a big challenge for scientists. Many researchers are involved in developing non-toxic, stable, bioactive oral insulin delivery systems.8–10 Different approaches, such as chemical modification11 and co-administration with absorption enhancers and/or enzyme inhibitors have been tried. Insulin incorporation into carriers,12 such as liposomes,13,14 mixed micelles, lipid-based
a
Department of Spectroscopy, Indian Association for the Cultivation of Science, Jadavpur, Kolkata-700 032, India. E-mail: [email protected]; Fax: +91-33-24732805; Tel: +91-33-24734971 b Department of Chemistry, Indian Institute of Technology, Kharagpur-721 302, India
This journal is © the Owner Societies 2014
systems, microspheres, and nanoparticles15–18 has been explored to improve the bioavailability of insulin. Lipid vesicles, known as liposomes, are widely recognized as pharmaceutical carriers because of their low toxicity, biodegradability, and biocompatibility.19 Phospholipid vesicles immobilized on solid substrates have been considered as models for cell membranes.20,21 Vesicles have good longevity in the blood and can specifically target disease sites when various targeting ligands are attached to their membranes. In a recent publication, a sodium dodecyl sulfate (SDS)–cetyl trimethyl ammonium bromide (CTAB) catanionic vesicle was developed22 and insulin was immobilized in the vesicular membrane.23 The SDS–CTAB catanionic vesicle was used as a model system due to its easy availability, but it is not a biocompatible system and cannot be used as a drug delivery tool. Previous research23 showed that CTAB and insulin formed a charge transfer complex. Thus, for the SDS–CTAB system, if not all the CTAB is involved in catanionic vesicle formation, the free CTAB may denature insulin. 1,2-Dipalmitoyl-sn-glycero-3phosphoglycerol (DPPG) is chosen here due to its biocompatible nature. Moreover, it is suitable for drug delivery due to its smaller vesicular size. Several authors24–27 have studied protein/drug immobilization in phosphatidylcholine or lecithin liposomal systems, but many of these studies are in the context of pharmacological applications. The detailed mechanism by computer simulation and spectroscopic studies has not yet been investigated. However, the study of the insulin–DPPG complex is rather rare. In this paper, our main objective is to study the insulin– vesicle complex formation and the entrapment of insulin into the vesicular membrane. To confirm the self-assembled structure
Phys. Chem. Chem. Phys., 2014, 16, 21657--21663 | 21657
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
Paper
PCCP
formation of the insulin and vesicles, we have used emission and excitation spectroscopy, dynamic light scattering (DLS), and zeta potential measurements. To establish the entrapment of insulin into the vesicular membrane, we have also performed molecular dynamics (MD) simulations. The results establish the fact that, after the interaction of insulin with the vesicles, a new structured broad phosphorescence peak appears at room temperature aside from the tyrosine fluorescence. The appearance of the phosphorescence peak is the signature of insulin entrapment into the vesicular membrane.
2. Experimental section and computational method 2.1.
Materials
The anionic bi-tail lipid, 1,2-dipalmitoyl-sn-glycero-3-phosphoglycerol (DPPG), was purchased from Avanti Polar Lipids, Inc. (Alabaster, Alabama). The chloroform (UV Grade) and methanol were purchased from Spectrochem and Sisco Research Laboratory, (Mumbai, India), respectively. The human insulin (Actrapid) was purchased from Abbott India Limited, (Mumbai, India). 2.2.
Experimental procedures
2.2.(A) Preparation of vesicular solution and insulin–vesicle complex. The stock solution of DPPG, of concentration 0.2 mM, was prepared in a 1 : 1 chloroform and methanol solvent. The prepared lipid solution was heated in a water bath at 70 1C to evaporate the solvent. Millipore water was then added to form vesicles at 45 1C, which is above the phase transition temperature (41 1C) of DPPG.28,29 The aqueous solution of insulin, of concentration 0.4 mg ml1 (pH 7.0), was added during the vesicle formation and the system was sonicated for 10 min. 2.2.(B) Spectroscopic characterization. Fluorescence and excitation spectra were recorded for pure insulin and the insulin–DPPG system at 25 1C, using a cuvette with a path length of 1 cm on a fluorescence spectrometer (HITACHI F-4500). Using liquid nitrogen as a coolant, the temperature (from room temperature to liquid nitrogen temperature) dependent emission spectra were recorded. A phosphorescence decay profile was also recorded using this spectrometer. 2.2.(C) Dynamic light scattering. A Brookhaven BI-200SM goniometer (Brookhaven Instruments Corporation) with a 35 mW He–Ne laser (633 nm) was used to study the dynamic light scattering. A non-negativity constrained least-squares algorithm was used to determine the mean diameter of the molecules. 2.2.(D) Zeta-potential measurements. A Malvern Zetasizer Nano ZS (part no.: ZEN3600) instrument was used for zeta potential measurements at a scattering angle of 1751 and a detection angle of 12.81 with a He–Ne laser (power 4 mW, wavelength 632.8 nm and beam diameter 0.63 nm (1/e2)). The experiment was performed at 25 1C. On one hand, the machine has a size measurement range of 3.81 nm and on the other hand, there was no practical range for the zeta-potential measurement. Standard Malvern zeta-potential disposable capillary cells were used to hold the vesicular solutions and
21658 | Phys. Chem. Chem. Phys., 2014, 16, 21657--21663
insulin–vesicle system for the zeta-potential measurements. For verification, the experiment was repeated three times. 2.3.
Computational method
2.3.(A) Preparation of the insulin monomer. The initial structure of the insulin monomer was obtained by considering chains A and B of the crystal structure of the insulin hexamer (PDB ID 3AIY), shown in Fig. 1. Chains A and B consist of 21 and 30 amino acids, respectively. All Glu, Lys, and Arg side chains were modeled as charged, and all His were kept neutral. Three disulfide bonds between Cys residue pairs 6 (chain A)–11 (chain A), 7 (chain A)–7 (chain B) and 20 (chain A)–19 (chain B) were added. The N- and C-termini of both chains were capped with acetyl and methylamine groups, respectively. The missing hydrogen atoms were added and the overall charge (2) was neutralized by adding 2 potassium ions. 2.3.(B) Insulin in water. The protein was immersed in a cubic water box of side 60 Å, which ensured a 10 Å distance between the boundary of the water box and any solute atoms. A 150 mM salt concentration was achieved by adding 18 potassium and chloride ions. The simulation system contained 6419 water molecules and about 20 100 atoms in total, shown in Fig. 2A. The MD simulations were carried out at constant temperature and constant pressure (1 atm) using a Nose–Hoover Langevin thermostat and piston.30,31 Long range electrostatic interactions were treated by the particle mesh Ewald method with a 12 Å cutoff. Van der Waals interactions were truncated at a cutoff of 12 Å and a switch function was activated starting at 10 Å.32 The SHAKE algorithm33 was used to fix the length of the covalent bonds involving hydrogen atoms, which allowed an integration time step of 2 fs. Structures were saved every 1 ps for analysis. All simulations were performed employing the NAMD program34 with the CHARMM22 force field35 and TIP3 potential for water molecules.36 The system was first equilibrated for 2 ns at two different temperatures (100 K and 300 K), followed by equilibrium dynamics for 20 ns each at the two temperature values. 2.3.(C) Insulin in the lipid bilayer. When a lipid bilayer is rolled into an enclosed cell vesicles are formed. That is why we have modeled insulin in the lipid bilayer in the simulation. The initial structure of the monomer was oriented to align the vector between the Ca atoms of Cys7 (chain B) and Val18
Fig. 1 Crystal structures of the (A) insulin hexamer and (B) insulin monomer (PDB ID 3AIY).
This journal is © the Owner Societies 2014
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
PCCP
Fig. 2
Paper
(A) Insulin monomer solvated in a cubic water box. (B) Insulin monomer in the DPPG lipid bilayer. (C) Insulin monomer in the DPPG lipid bilayer and water.
(chain B) with the z-axis, and was then placed in a homogeneous lipid bilayer of DPPG molecules by following the so-called insertion method as implemented in CHARMM-GUI.37–39 The resulting system is shown in Fig. 2B. The number of DPPG molecules was determined by ensuring at least three lipid molecules were placed between two proteins in the primary and image systems.38 The insulin–lipid system was then solvated by adding a rectangular water box to maintain a water thickness of 15 Å on top of and below the protein. The system was then neutralized and a salt concentration of 150 mM was achieved by adding potassium and chloride ions. The final system contained the insulin monomer with 96 DPPG molecules, 3668 water molecules, 106 potassium ions, and 8 chloride ions, with a total of about 24 000 atoms, shown in Fig. 2C. The system was subjected to gradual equilibration for 375 ps as described in previous literature,38 followed by free equilibrium MD simulation for 20 ns, with the same cutoff parameters as described in the insulin in water system. Another MD simulation was carried out using the same procedure but at 100 K.
3. Results and discussion Fluorescence spectroscopy can be employed as a highly sensitive method to study protein structure, conformation, and kinetics. Intrinsic protein fluorescence generally arises from three
fluorophores, i.e., tryptophan (Trp), tyrosine (Tyr) and phenylalanine (Phe).40,41 Among these, the spectroscopic behavior of a protein is normally dominated by Trp, as the quantum yield of Trp is highest.42,43 However, in the case of insulin, tyrosine plays an important role for spectroscopic characterization, as tryptophan is absent. Fig. 3A and B show the peak normalized emission and excitation spectra of pure insulin and the insulin–DPPG system at room temperature (300 K), respectively. The emission spectrum of pure insulin (Fig. 3A, curve-a) shows a single peak around 305 nm, which is due to tyrosine fluorescence. In the absence of tryptophan in insulin, tyrosine is responsible for the emission peaks. Very interestingly, in the presence of vesicles, the emission spectrum of insulin drastically differs from that of only the protein. In curve-b of Fig. 3A, the peak around 303 nm is due to the fluorescence of the tyrosine residues. In comparison with pure insulin, this peak marginally shifts towards the blue end of the spectrum and shows a smaller full width at half maximum. This shift may presumably reflect the change in local environment upon movement of some tyrosine residues from an aqueous environment to the vesicular lipid membrane.44 Apart from this peak, another new structured broad band centered around 450 nm appears, possibly due to the phosphorescence of tyrosine. The discussion about this band is described below. Fig. 3B shows the excitation spectra of the pure insulin (curve-a) and insulin–DPPG complexes (curve-b). Both cases show
Fig. 3 (A) Emission spectra (monitored at lex = 275 nm) of pure insulin (curve-a) and the insulin–DPPG complex (curve-b) and (B) excitation spectra (monitored at lem = 305 nm) of pure insulin (curve-a) and the insulin–DPPG complex (curve-b).
This journal is © the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 21657--21663 | 21659
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
Paper
two excitation peaks. In pure insulin, these peaks, at 228 nm and 277 nm, originate from 1La and 1Lb transitions of the tyrosine phenol ring, respectively.45 In the case of the insulin–DPPG system, the higher energy band (228 nm) appears at the same wavelength as that of pure insulin, however, with reduced intensity, whereas the lower energy band (272 nm) is blue shifted. This shift is due to the interaction between the lipid and protein molecules. The change in relative intensity after vesicle inclusion may be due to the restriction of the 1Lb vibration. In many previous literature studies,46–48 a low temperature phosphorescence peak around 450 nm was observed for tyrosine. To confirm whether the emission peak around 450 nm in this case is due to phosphorescence, we have conducted a temperature dependent steady state and time resolved emission study of the long wavelength band of the insulin–DPPG complex. Fig. 4A shows the results. Between 350 nm and 500 nm, a broad peak appears at each temperature. With decreasing temperature, the intensity of the band increases drastically. The intensity is highest at liquid nitrogen temperature (100 K), whereas it is lowest at room temperature (300 K). Fig. 4B shows the lifetime of the insulin–DPPG complex (at lem = 430 nm) at different temperatures. It indicates that the lifetime decreases with the increase of temperature. The lifetime of the insulin–DPPG complex at room temperature is about 0.22 ms, whereas at liquid nitrogen temperature it increases to 2.3 ms. We have tried to fit the temperature dependent phosphorescence lifetime using a well known kinetic equation found in previous literature:49 1 1 DE þ constant (1) ¼ ln t t0 kT where t is the lifetime, t0 is the intrinsic lifetime, DE is the activation energy, k is the Boltzmann constant, and T is the temperature. Eqn (1) can be modified to eqn (2) by considering t0, DE, and k as constants: C (2) t¼1 A þ Be T where A, B, and C are the constants.
PCCP
Here we have fitted the data with eqn (2) (shown by the smooth curve in Fig. 4B). The R2 value is 0.97. From the curve, the millisecond order long lifetime and lifetime dependency of the temperature confirm that the structured broad emission peak between 400 nm and 500 nm is due to phosphorescence, originating due to the rigid geometry of the tyrosine molecule.50 It is to be noted that the emission spectrum at low temperature does not show a structured broad peak. At freezing conditions, the orientation of every insulin molecule in the insulin–DPPG complex is different. Due to this reason, in the frozen glassy state, the fine structures of the broad peak are not well resolved (Fig. 4A). Even at room temperature, when is insulin entrapped into the liposome membrane, the movement of tyrosine is restricted. In this restricted tyrosine geometry, a radiative transition between the lowest triplet state and ground state (intersystem crossing) becomes more efficient and we observe phosphorescence at room temperature. Fig. 5 shows particle size distribution curves resulting from DLS experiments on the DPPG–insulin system (curve-a) and DPPG vesicles (curve-b). The size of a DPPG vesicle is within 150–300 nm, which is much smaller than a SDS–CTAB catanionic vesicle23 and thus more penetrable and useful in drug delivery. In
Fig. 5 Size distribution of the insulin–DPPG complex (curve-a) and DPPG vesicles (curve-b). The symbols show the data and red lines show the fitted curves. The black and blue curves are the deconvoluted curves for the insulin–DPPG complex and DPPG vesicles, respectively.
Fig. 4 (A) Emission spectra (monitored at lex = 275 nm) of the insulin–DPPG complex at different temperatures. (B) Lifetime of the insulin–DPPG complex at different temperatures with a fitted curve.
21660 | Phys. Chem. Chem. Phys., 2014, 16, 21657--21663
This journal is © the Owner Societies 2014
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
PCCP
each case, two peaks are observed. These peaks indicate that there are two aggregation sizes in the system. By deconvulation of these curves using Microcal Origin 8, we get two peaks around 177 nm and 241 nm for the DPPG–insulin system (R2 = 0.998), whereas, similar peaks originate in a lower region (157 nm and 222 nm) for the pure DPPG system (R2 = 0.999). This observation clearly illustrates that the size enhancement occurs due to insulin entrapment into the lipid vesicles. The measured zeta-potentials of the insulin–DPPG system and pure DPPG vesicular system at pH 7.0 are 41 mV and 53 mV, respectively. The negative zeta potential is due to the anionic nature of the DPPG lipid.51 The zeta potential becomes less negative in the case of the DPPG–insulin system. We have added insulin during the preparation of the DPPG vesicles and, due to the attachment of insulin molecule with the vesicle membrane, the effective surface charge density at the outer surface may decrease. This result supports the previous results and confirms the entrapment of insulin into the DPPG vesicles. A molecular dynamics study can be a very useful technique to elucidate the origin of the experimental observations. In the case of insulin, the tyrosine residues play an important role in the spectroscopy of the molecule. A detailed study of tyrosine in insulin can reveal the real facts about the encapsulation of insulin in the lipid vesicles. Fig. 6A shows the root mean square fluctuation (RMSF) of the non-hydrogen atoms of the Tyr residues of the insulin monomer averaged over intervals of 2 ns molecular dynamic trajectories. The standard deviations of the RMSF values of different Tyr residues are shown as error bars. The RMSF values of the Tyr residues when insulin is solvated in water and in the DPPG lipid bilayer are shown as black diamonds (curve-a) and red diamonds (curve-b), respectively. Fig. 6A suggests a reduced fluctuation of all four Tyr residues when the protein is in the lipid bilayer compared to their respective RMSF values in water. In particular, the reduced movement of Tyr14 (chain A) in the lipid bilayer is significant, while the same effect is moderate for Tyr19 (chain A) and Tyr16 (chain B). Among these four tyrosine
Paper
residues, Tyr14 is most exposed to the lipid and therefore, it interacts strongly with the lipid, as reflected by the reduced fluctuation of Tyr14 in the lipid bilayer. Fig. 6B shows the positions of the tyrosine residues in the protein monomer during the interaction with the lipid. At 100 K the system becomes totally frozen and the fluctuations of all tyrosine residues in both water and lipid media are more or less the same and much reduced from the fluctuations seen at 300 K (Fig. 6A). From this observation, it can be concluded that tyrosine becomes rigid after the entrapment of insulin into the liposome membrane as well as at low temperature. This rigid geometry of tyrosine is responsible for the phosphorescence phenomenon at room temperature. For further information regarding the interaction between the lipid and protein, we have calculated the number of contacts between the DPPG atoms and insulin residues and the interaction energy between individual residues and lipid molecules. Fig. 7A shows the average number of contacts each residue makes with the lipid molecules which are within 4 Å of the corresponding residue. Fewer contacts are seen for the Ile2, Val3, Cys6, Cys7, Leu16, Glu17, Cys20 residues of chain A and His5, Leu15 and Cys19 of chain B. These residues form an a-helix segment in the protein, suggesting that the alpha helical segment has much less interaction with the lipid molecules. On the other hand, a larger number of contacts for His10 of chain A and Phe1 and Leu17 of chain B suggest that they play key roles in mediating the protein–lipid interaction. This calculation also shows that the average number of contacts of all four Tyr residues with the lipid molecules is higher than 25, suggesting the important role of Tyr in the protein–lipid interaction, which supports the experimental observations. To identify the residues contributing to the insulin–DPPG interaction, we have plotted the average interaction energy of each amino acid residue with the DPPG lipid molecules during the MD simulation in Fig. 7B. A strong favorable interaction energy of Glu4 (chain-A) and Glu13 (chain-B) with the DPPG molecules is seen. The origin of this favorable interaction
Fig. 6 (A) The RMSF values of the Tyr residues when insulin is solvated in water and in the DPPG lipid bilayer are shown as curve-a (300 K) and curve-c (100 K) and as curve-b (300 K) and curve-d (100 K), respectively. The standard deviations of the RMSF values of different Tyr residues are shown as error bars. (B) The insulin monomer in the DPPG layer after interactions. Four tyrosine residues are shown in ball–stick form and DPPG is shown as lines.
This journal is © the Owner Societies 2014
Phys. Chem. Chem. Phys., 2014, 16, 21657--21663 | 21661
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
Paper
PCCP
Fig. 7 (A) The number of contacts between DPPG atoms (within 4 Å) and residues of insulin for chain-A (curve-a) and chain-B (curve-b). The standard deviations of the contact values of different residues are shown as error bars. (B) The interaction energy between DPPG and the insulin residues of chain-A (curve-a) and chain-B (curve-b). The standard deviations of the energy values of different residues are shown as error bars.
energy can be traced back to the interaction of the carboxylate groups of Glu4 (chain A) and Glu13 (chain B) with the phosphorus atom of the DPPG head group, which carries a positive partial charge.52 On the other hand, the positively charged side chains of Arg22 (chain-B) show a stronger interaction with the negatively charged oxygen atoms of the lipid head group. Similar interactions are also responsible for the positive value of the interaction energy between Glu17 (chain A) and the lipid head group. In summary, the lipid head group is found to stabilize both acidic and basic residues depending on the relative orientation of the side chains of the amino acid residue and the lipid molecule.
4. Conclusion We have successfully entrapped insulin into the DPPG vesicle membrane. After entrapment, the protein remains in its native state. We have studied steady state and time resolved emission spectroscopy. A newly observed structured broad band, centered at 450 nm, is the signature of the entrapment of the insulin molecule into the vesicle membrane. From the lifetime measurements, we assigned the peak as the phosphorescence emission of tyrosine, which originates due to the limited movement of Tyr residues arising from their interaction with the lipid molecule. This is further supported by MD simulations of the insulin monomer in water and lipid bilayer media. The decrease in the RMSF of the Tyr residues of insulin in the lipid bilayer compared to in water establishes the fact that the Tyr residues become rigid after the entrapment of insulin into the vesicle membrane. Since DPPG–insulin is a biocompatible system, the present study is expected to be useful and effective in the field of drug delivery, biosensors, and bioengineering.
Acknowledgements We thank DST, Government of India (Project No. SR/S2/CMP0079/2010(G)) for partial financial support and the IACS for providing central instrumental facilities. SM thanks DST, India for the Ramanujan fellowship.
21662 | Phys. Chem. Chem. Phys., 2014, 16, 21657--21663
References 1 Y. Date, M. Nakazato, S. Hashiguchi, K. Dezaki, M. S. Mondal, H. Hosoda, M. Kojima, K. Kangawa, T. Arima, H. Matsuo and T. Yada, Diabetes, 2002, 51, 124–129. 2 Y. Itoh, Y. Kawamata, M. Harada, M. Kobayashi, R. Fujii, S. Fukusumi, K. Ogi, M. Hosoya, Y. Tanaka, H. Uejima, H. Tanaka, M. Maruyama, R. Satoh, S. Okubo, H. Kizawa, H. Komatsu, F. Matsumura, Y. Noguchi, T. Shinohara, S. Hinuma, Y. Fujisawa and M. Fujino, Nature, 2003, 422, 173–176. 3 M. N. Poy, L. Eliasson, J. Krutzfeldt, S. Kuwajima, X. Ma, P. E. MacDonald, S. Pfeffer, T. Tuschl, N. Rajewsky, P. Rorsman and M. Stoffel, Nature, 2004, 432, 226–230. 4 J. L. Whittingham, D. J. Scott, K. Chance, A. Wilson, J. Finch, J. Brange and G. G. Dodson, J. Mol. Biol., 2002, 318, 479–490. 5 G. D. Smith, D. C. Swenson, E. J. Dodson, G. G. Dodson and C. D. Reynolds, Proc. Natl. Acad. Sci. U. S. A., 1984, 81, 7093–7097. 6 B. Tah, P. Pal, S. Roy, D. Dutta, S. Mishra, M. Ghosh and G. B. Talapatra, Spectrochim. Acta, Part A, 2014, 129, 345–351. 7 R. K. Zalups and D. J. Koropatnick, Cellular and Molecular Biology of Metals, CRC Press, 2010. ´, P. Maincent and N. Ubrich, J. Controlled Release, 8 C. Damge 2007, 117, 163–170. 9 M. A. Kisel, L. N. Kulik, I. S. Tsybovsky, A. P. Vlasov, M. S. Vorob’yov, E. A. Kholodova and Z. V. Zabarovskaya, Int. J. Pharm., 2001, 216, 105–114. 10 K. Iwanaga, S. Ono, K. Narioka, K. Morimoto, M. Kakemi, S. Yamashita, M. Nango and N. Oku, Int. J. Pharm., 1997, 157, 73–80. 11 H. Ashada, T. Douen, Y. Mizokoshi, T. Fujita, M. Murakami, A. Yamamoto and S. Muranishi, J. Pharm. Sci., 1995, 84, 682–687. 12 E. Ziv, O. Lior and M. Kidron, Biochem. Pharmacol., 1987, 36, 1035–1039. 13 M. A. Kisel, L. N. Kulik, I. S. Tsybovsky, A. P. Vlasov, M. S. Vorobyov, E. A. Kholodova and Z. V. Zabarovskaya, Int. J. Pharm., 2001, 216, 105–114.
This journal is © the Owner Societies 2014
View Article Online
Published on 19 August 2014. Downloaded by University of Chicago on 26/10/2014 17:59:55.
PCCP
14 J. H. Fendler and A. Romero, Life Sci., 1977, 20, 1109–1120. 15 J. Scott-Moncrieff, Z. Shao and K. Mitra, J. Pharm. Sci., 1994, 83, 1465–1469. ´ron, F. Puisieux and M. Seiller, Int. 16 A. Silva-Cunha, J. Che J. Pharm., 1997, 158, 79–89. 17 B. Y. Kim, J. Jeonga, K. Parkb and J. D. Kim, J. Controlled Release, 2005, 102, 525–538. 18 B. Sarmento, S. Martins, A. Rebeiro, F. Veiga, R. Neufeld and D. Ferreira, Int. J. Pept. Res. Ther., 2006, 12, 131–138. 19 A. Akbarzadeh, R. Rezaei-Sadabady, S. Davaran, S. W. Joo, N. Zarghami, Y. Hanifehpour, M. Samiei, M. Kouhi and K. Nejati-Koshki, Nanoscale Res. Lett., 2013, 8, 102. 20 B. Tah, P. Pal, M. Mahato and G. B. Talapatra, J. Phys. Chem. B, 2011, 115, 8493–8499. 21 T. Kamilya, P. Pal, M. Mahato and G. B. Talapatra, J. Nanosci. Nanotechnol., 2009, 9, 2956–2964. 22 T. Kamilya, P. Pal, M. Mahato and G. B. Talapatra, Mater. Sci. Eng., C, 2009, 29, 1480–1485. 23 B. Tah, P. Pal and G. B. Talapatra, J. Lumin., 2014, 145, 81–87. 24 K. Ariga, Y. M. Lvov, K. Kawakami, Q. Ji and J. P. Hill, Adv. Drug Delivery Rev., 2011, 63, 762–771. 25 M. A. Kisel, L. N. Kulik, I. S. Tsybovsky, A. P. Vlasov, M. S. Vorobyov, E. A. Kholodova and Z. V. Zabarovskaya, Int. J. Pharm., 2001, 216, 105–114. 26 X. Zhang, J. Qi, Y. Lu, X. Hu, W. He and W. Wu, Nanoscale Res. Lett., 2014, 9, 185. 27 R. Mo, T. Jiang, J. Di, W. Tai and Z. Gu, Chem. Soc. Rev., 2014, 43, 3595–3629. 28 R. Wohlgemuth, N. Waespe-Sarcevic and J. Seelig, Biophys. J., 1980, 19, 3315–3321. 29 T. Zaraiskaya and K. R. Jeffrey, Biophys. J., 2005, 88, 4017–4031. 30 G. Martyna, D. Tobias and M. Klein, J. Chem. Phys., 1994, 101, 4177–4189. 31 S. E. Feller, Y. Zhang, R. Pastor and B. R. Brooks, Chem. Phys., 1995, 103, 4613–4621. 32 P. J. Steinbach and B. R. Brooks, J. Comput. Chem., 1994, 15, 667–683. 33 W. Van Gunsteren and H. Berendsen, Mol. Phys., 1977, 34, 1311–1327.
This journal is © the Owner Societies 2014
Paper
34 J. C. Phillips, R. Braun, W. Wang, J. Gumbart, ´ E. Tajkhorshid, E. Villa, C. Chipot, R. D. Skeel, L. Kale and K. Schulten, J. Comput. Chem., 2005, 26, 1781–1802. 35 A. D. Mac Kerell, Jr., D. Bashford, M. Bellott, R. L. Dunbrack, Jr., J. D. Evan-seck, M. J. Field, S. Fischer, J. Gao, H. Guo, S. Ha, D. Joseph-McCarthy, L. Kuchnir, K. Kuczera, F. T. K. Lau, C. Mattos, S. Michnick, T. Ngo, D. T. Nguyen, B. Prodhom, W. E. Reiher, B. Roux, M. Schlenkrich, J. C. Smith, R. Stote, J. Straub, M. Watanabe, J. Wiorkiewicz-Kuczera, D. Yin and M. Karplus, J. Phys. Chem. B, 1998, 102, 3586–3616. 36 W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey and M. L. Klein, J. Chem. Phys., 1983, 79, 926–935. 37 S. Jo, J. B. Klauda and W. Im, J. Comput. Chem., 2008, 29, 1859–1865. 38 S. Jo, T. Kim and W. Im, PLoS One, 2007, 2(9), e880. 39 S. Jo, J. B. Klauda and W. Im, Biophys. J., 2009, 97, 50–58. 40 M. Mahato, P. Pal, T. Kamilya, R. Sarkar, A. Chawdhuri and G. B. Talapatra, J. Phys. Chem. B, 2010, 114, 7062–7070. 41 P. Pal, M. Mahato, T. Kamilya and G. B. Talapatra, Phys. Chem. Chem. Phys., 2011, 13, 9385–9396. 42 D. B. Wetlaufer, Adv. Protein Chem., 1962, 17, 303–390. 43 J. T. Vivian and P. R. Callis, Biophys. J., 2001, 80, 2093–2109. 44 H. S. Galdiero, A. Falanga, M. Vitiello, L. Raiola, L. Russo, C. Pedone, C. Isernia and M. Galdiero, J. Biol. Chem., 2010, 285, 17123–17136. 45 J. R. Lakowicz, Topics Fluorescence Spectroscopy, Kluwer academic publishers, 2002, vol. 3. 46 T. Truong, R. Bersohn, P. Brumer, C. K. Ltjk and T. Tao, J. Biol. Chem., 1967, 242, 2979–2985. 47 R. F. Steiner and R. Kolinski, Biochemistry, 1968, 7, 1014–1018. 48 K. W. Rousslang, P. J. Reid, D. M. Holloway, D. R. Haynes, J. Dragavon and J. B. A. Ross, J. Protein Chem., 2002, 21, 547–555. 49 I. G. Leubner, J. Phys. Chem., 1970, 74, 77–82. 50 G. B. Strambini and E. Gabellieri, Biophys. J., 1996, 70, 971–976. 51 H. Bensikaddour, K. Snoussi, L. Lins, F. V. Bambeke, P. M. Tulkens, R. Brasseur, E. Goormaghtigh and M. P. MingeotLeclercq, Biochim. Biophys. Acta, Biomembr., 2008, 1778, 2535–2543. 52 Y. N. Kaznessis, S. Kim and R. G. Larson, J. Mol. Biol., 2002, 322, 569–582.
Phys. Chem. Chem. Phys., 2014, 16, 21657--21663 | 21663
