Kinetics of Protein−Protein Complex Coacervation and Biphasic ...

33 downloads 13620 Views 504KB Size Report
Dec 10, 2008 - molecules, gelatin-A and gelatin-B, having complementary charges that led to pH-induced liquid-liquid phase ... traverse some signature road maps. .... light scattering (DLS) technique using a Brookhaven-9000AT digital.
184

Biomacromolecules 2009, 10, 184–189

Kinetics of Protein-Protein Complex Coacervation and Biphasic Release of Salbutamol Sulfate from Coacervate Matrix Ananya Tiwari,† Sonal Bindal,‡ and H. B. Bohidar*,§ Department of Chemistry, St. Stephens College, Delhi, India, Department of Biological Science, Sri Venkateswara College, New Delhi, India, and Polymer and Biophysics Laboratory, School of Physical Sciences, Jawaharlal Nehru University, New Delhi, India Received October 15, 2008; Revised Manuscript Received November 4, 2008

Turbidimetric titration was used to initiate associative intermolecular interactions between a pair of protein molecules, gelatin-A and gelatin-B, having complementary charges that led to pH-induced liquid-liquid phase separation and the formation of complex coacervate. The stoichiometric binding ratio was found to be [gelatinA]/[gelatin-B] ) 3:2. The size of soluble intermolecular aggregates present in the supernatant exhibited interesting time-dependent coacervation because of residual electrostatic interactions. Dynamic light scattering and turbidity studies provided a systematic account of coacervation behavior. Rheology studies attributed the softening of the coacervate matrix to the presence of encapsulated salbutamol sulfate. The in vitro drug release kinetics was probed in simulated gastric fluid medium at physiological temperature (37 °C), which showed biphasic behavior. The initial release kinetics exhibited an exponential growth to saturation behavior, followed by a slower logarithmic release process.

Introduction Coacervation is usually defined as the spontaneous formation of a super dense liquid phase from a homogeneous macromolecular solution of poor solvent affinity.1 Intermolecular associative interaction between oppositely charged macromolecules, which is mostly driven by electrostatic forces, often leads to liquid-liquid phase separation (called coacervation transition). Here a homogeneous solution separates into a polymerrich phase (called coacervate) and its supernatant.2 In complex coacervation, the loss of solvation that lead to phase separation arises from the interaction of complementary macromolecular species. The formation of such macromolecule-rich fluids is well known in mixtures of complementary polyelectrolytes;3 it can also occur from mixtures of polyelectrolytes with colloidal particles, leading to condensed phases that are associated with interesting properties. Coacervation has been observed in a wide range of polyelectrolyte systems, such as poly(dimethyldiallylammonium chloride)-bovine serum albumin,4 gelatin-chitosan,5gelatin-agar,6 gelatin-gelatin,7 and so on. In addition, experiments that explored the formation of similar self-organized microscopic structures using nucleic acids were very successful.8 Properties of such self-assembled coacervates can be continuously varied from those of homogeneous liquids, to viscous gellike materials, and finally to amorphous solids,6 without protein denaturation.7 They offer the possibility of preparation of novel biocompatible and bioactive materials for a wide range of applications,9,10 extending well beyond the current utilization in encapsulation. These materials display very large shear viscosities but exhibit protein diffusivities that are only an order of magnitude below those in dilute protein solution.11,12 In the past, various methods were used to characterize the coacervates * Corresponding author. E-mail: [email protected]. Tel: +91 11 2760 4637. Fax: +91 11 2671 7562. † St. Stephens College. ‡ Sri Venkateswara College. § Jawaharlal Nehru University.

on many length and time scales, leading to a conclusive description of the microscopic structure and dynamics of these unique self-assembled materials.11 These results taken together revealed that the coacervate phase is a solutionlike state in which homogeneous fluidlike domains coexist with denser and more nearly charge-neutralized domains, which inhibit local diffusion.12 In simple coacervation, a polyampholyte molecule exhibits coacervation transition induced by a change in the solvent quality of the dispersion medium through the addition of salt13 or a nonsolvent. However, coacervation transitions universally traverse some signature road maps. These are as follows:14 (i) A homogeneous solution containing N1 molecules of solvent and N2 molecules of solute at temperature T and pressure P will remain stable as long as the free energy of the solute F2 in solution obeys the thermodynamic condition (∂2F2/∂N22)N1,T,P > 0. (ii) The liquid-liquid phase separation of the coacervate phase from the dilute supernatant is a dehydration process. (iii) Charge neutralization of polyion segments precedes phase separation. (iv) The polyions do not precipitate out of the solvent because of the entropy gain achieved by random mixing of polyions in the coacervate phase. In summary, the coacervation proceeds following two characteristic steps: first, the selective charge neutralization of polyions dictated by electrostatic interactions, and second, the gain in entropy through random mixing of polyions in the dense phase plus the gain in entropy due to the release of counterions to the solvent. Salbutamol is a hydrophilic drug used as a bronchodilator, prescribed for the relief of bronchospasm (constriction of airways) caused by asthma or chronic airways disease. It can also be used in the acute prevention of asthma in situations known to induce it, such as exercise. It quickly acts on the nerves that control the airway muscles and causes them to relax, dilating the airways.15 It also relieves swelling of the airways that can be caused by the allergic response and helps clear mucous that may contribute to asthmatic symptoms. The transdermal delivery of salbutamol sulfate (SS) has also been reported,16 and the

10.1021/bm801160s CCC: $40.75  2009 American Chemical Society Published on Web 12/10/2008

Kinetics of Protein-Protein Complex Coacervation

Figure 1. Binding curve indicating the stoichiometry of intermolecular of complexation between gelatin-A and gelatin-B. Optimum binding occurred at [gelatin-A]/[gelatin-B] ) 3:2.

effectiveness of this drug in controlling obesity and depressive behavior has been probed.17 Herein, we elucidate the protein-protein coacervation transition affected through pH-induced complexation of a low-chargedensity polyampholyte (gelatin-B) with the strong but moderatecharge-density polyelectrolyte (gelatin-A). This coacervate material was used for the encapsulation of a short-acting14 model drug, SS, and the in vitro release kinetics of this drug was probed in a simulated gastric fluid maintained at physiological temperature (37 °C). Whereas the phenomenology of coacervation might be considered to be well known, the nature of coacervates and their potential as drug carriers remains essentially unexplored because of the lack of systematic studies with modern techniques, which comprises the main focus of this work.

Materials and Methods Gelatin samples of both type-A (porcine skin extract, bloom ) 300, pI of 9) and type-B (bovine skin extract, bloom ) 75, pI of 5) obtained from Sigma Chemicals were used. SS (molecular formula: C13H21NO3 · 2H2SO4) was procured from Sigma Chemicals. All other chemicals used were of analytical grade and were bought from Thomas Baker, India. We prepared the gelatin-A and gelatin-B solutions (0.42 and 0.28% w/v) by dispersing gelatin in deionized water medium at 50 °C. We allowed the macromolecules to completely hydrate by stirring the solution; this took 1 to 1.5 h. Both gelatin solutions looked optically transparent. The pH (using 0.1 M HCl or 0.1 M NaOH) was first set as per the experimental requirement to pH 4. The gelation concentration of gelatin (both type-A and -B) in water is17 ∼2% (w/v); the concentration chosen in these experiments was deliberately kept lower than this to avoid the formation of gels.18 The actual concentration chosen was based on the optimum binding stoichiometry ascertained through an independent experiment (see Figure 1). We prepared complex coacervate samples by mixing these two solutions in equal volume. The mixture solution was titrated with 0.1 N NaOH to increase the pH, and the change in transmittance (% T) of the solution was continuously monitored using a turbidity meter (Brinkmann-910, Brinkmann Instruments) operating3 at 450 nm. This was continued until a maximum turbidity was registered. (See Figure 2.) The turbid samples were sealed and stored at room temperature (25 °C) for ∼5 days (to increase the yield) and were then subjected to centrifugation at 10 000 rpm for 30 min, which separated the turbid solution into two distinct liquid phases, namely, the dense coacervates at the bottom and the supernatant at the top. The polymer-rich phase at the bottom was collected after the supernatant was decanted. This was repeated at least three times, which yielded the coacervates. This is the normal procedure of extracting the coacervate from the reacted solutions.4-6

Biomacromolecules, Vol. 10, No. 1, 2009

185

The particle sizing measurements were done by the dynamic laser light scattering (DLS) technique using a Brookhaven-9000AT digital autocorrelator (Brookhaven Instruments). The scattering angle was fixed at 90°, and the data analysis was done using CONTIN software provided by Brookhaven Instruments. The instrument measures the translational diffusion coefficient of scattering moieties, which was used in Stoke-Einstein relation for determining the mean particle diameter (Dmean). The covariance of the diffusion coefficient yields the polydispersity in particle size. Further details of DLS technique and data analysis can be found elsewhere.19 The rheology experiments were performed using a AR-500 stress controlled rheometer (T.A. Instruments, UK). The dynamic rheology of coacervates was studied by using steel parallel plate geometry of a radius of 20 mm and with a truncation gap of 500 µm. The truncation gap used in the parallel plate geometry was intentionally kept this large to avoid the breaking of structures inside the coacervate samples. A metallic solvent trap was used to minimize evaporative losses. The zeta potential measurements were performed on an electrophoresis instrument (model: ZC-2000, Microtec, Japan). The sample solution was very diluted to isolate all individual particles from the aggregates to allow for the discovery of the surface charge of streaming particles. If one uses the zeta potential (ζ) as an approximation of the surface potential (φ) of a uniformly charged sphere, then the theory gives20

ζ = φ ) 4π(σ/εκ)

(1)

where σ is the surface charge density of the particle, and ε and κ are the dielectric constant and the Debye-Hu¨ckel parameter of the solution, respectively. It has been shown that the surface potential can be determined to a very good approximation from the potential existing at the hydrodynamic slip plane, which is called the zeta potential. The relationship between the mobility (µ) and the zeta potential (ζ) is ζ ) 4π(µη/ε). Then, µ can be written as µ ) σ/ηκ, where η is the viscosity of the solution. Because the polyelectrolytes are in random coil conformations, the quantitative application of eq 1 is not expected. SS is a water-soluble hydrophilic molecule that enables easy encapsulation. We loaded the drug in the coacervate matrix in two ways: (i) by adding the drug to the coacervating solution and (ii) by adding the drug to coacervate directly under sonication. These cases are referred to as “before” and “after” coacervation drug-loaded samples, respectively, in subsequent discussions. The drug concentrations were determined through UV absorbance data recorded at λ ) 276 nm. A dilute drug solution exhibits a nonambiguous absorption peak at this wavelength. The encapsulation percentage was determined to be 60 and 40% for before and after samples, respectively (obtained from graphical integration of release behavior). Typically, a 4 mg drug-loaded sample was put in a 20 mL (arbitrarily chosen) gastric fluid medium (0.5 N NaCl + 0.5 N HCl, pH 1.2) without pepsin, and its supernatant was examined by UV spectroscopy to determine the amount of drug released. The release experiments were performed inside an incubator maintained at 37 °C. It was visibly observed that the coacervate matrix slowly dissolved over a period of 72 h. An alternative method was to use a dissolution tester, which was not done.

Results and Discussion Coacervation and Coacervate Properties. Gelatin molecules in the solutions assume random coil conformation.21,22 Both static and DLS measurements were carried out on dilute gelatin solutions to evaluate their chain dimensions. Light scattering measurements assigned the following dimensions (radius of gyration, Rg, and hydrodynamic radius, Rh) to the chains: (i) for gelatin-A, Rg ) 55 ( 5 nm and Rh ) 58 ( 6 nm, and (ii) for gelatin-B, Rg ) 34 ( 3 nm and Rh ) 23 ( 3 nm. Thus, one can estimate the chain stiffness from the ratio23 Rh/Rg, which was 0.95 for gelatin-A and 0.67 for gelatin-B. This clearly attributes a fully flexible chain conformation to gelatin-B,

186

Biomacromolecules, Vol. 10, No. 1, 2009

Tiwari et al.

Figure 3. Frequency dependence of storage modulus measured at 25 °C. The data were fit to the power law given by eq 2, which yielded the exponents shown (χ2 > 0.93).

Figure 2. (A) Zeta potential profile indicating change in surface potential with pH variation. The intermolecular complexation leading to coacervation transition occurred at pH ≈ 6.5. This point corresponds to turbidity maximum. (See B.) (B) Titration profile indicating change in measured turbidity with pH variation. The intermolecular complexation ensues at pHc ≈ 5.5, turbidity was observed at pHΦ ≈ 6.3, and precipitation occurred at pHprep ≈ 6.5.

whereas gelatin-A chains appear to be more rigid. Gelatin is a polyampholyte molecule that makes the net charge on the molecule strongly dependent on pH. The electrophoretic mobility gives an estimation of the zeta potential. Figure 2A depicts the variation in zeta potential as a function of pH for both gelatin-A and -B molecules, and it is clearly seen that the isoelectric pH (pI) values were 9 and 5, respectively. Thus, associative electrostatic interaction between these molecules would prevail in the pH window defined by 5 < pH < 9. When gelatin-A and -B molecules were bound to form a complex, the complex assumed zeta potential values that were found to be intermediate between that of gelatin-A and -B values. This is clearly seen from Figure 2A data. The binding curve shown in Figure 1 yielded a stoichiometric binding ratio [gelatin-A]/[gelatin-B] ) 3:2. These observations set the optimum binding conditions. The simplicity and sensitivity of the turbidimetric titration method, as applied to protein-protein systems, is based on the fact that turbidity is proportional to both the molecular weight and the number density of particles present in dispersion. The change in turbidity mirrors the extent of interactions between the two biopolymers (gelatin and chitosan) prevailing at an instance. We observed the first occurrence of turbidity corresponding to the formation of soluble complexes that was measured (pHc). The titration process was continued until maximum turbidity (pHφ) was noticed (formation of insoluble complexes). The turbidimetric titration profile is shown in Figure 2B, which clearly establishes three distinguishable pHs: (i) pHC of 5.5, where associative intermolecular interaction ensues with the formation of soluble

Figure 4. Frequency dependence of loss modulus measured at 25 °C. The data were fit to the power law given by eq 3, which yielded the exponents shown (χ2 > 0.95). In the case of drug-loaded samples, no distinction could be made between before and after specimens.

aggregates and turbidity appears, (ii) pHΦ of 6.3, where turbidity reaches a maximum with the appearance of microscopic coacervate droplets, and (iii) pHprep of 6.5, where the solute precipitates out of the solvent. The zeta potential of the gelatinA-gelatin-B complex exhibited interesting features close to coacervation pHφ, where the zeta potential was ∼0. As gelatin-A and -B electrostatically bind to form a complex, these molecules become charge neutralized, which is shown in Figure 2A. There is close resemblance between turbidity and charge neutralization. (See Figure 2B). Consequently, close to the pHφ where coacervate droplets appear, the zeta potential of the complex becomes zero. This is discussed at length in refs 13 and 14. We collected the coacervate material close to pH 6.3 by centrifuging the coacervating solution for 30 min and decanting the supernatant. This procedure was repeated thrice, which produced the thick, light-yellow coacervate material located at the bottom of the centrifuge tube. This material was subjected to rheological studies to establish its viscoelastic signature. The frequency sweep data that define the frequency dependence of storage and loss moduli (G′ and G′′) are shown in Figures 3 and 4. The data could be adequately fit to the functions given by

G′ ∼ ωn′ and G′′ ∼ ωn′′

(2)

which gave n′ ) 1.7 ( 0.4 for coacervate and 2.0 ( 0.5 and 2.9 ( 0.8 for after- and before-loaded samples, respectively. The exponent n′′ ) 1.5 ( 0.2 and 1.6 ( 0.3 for the drug-encapsulated and for coacervate samples, respectively.

Kinetics of Protein-Protein Complex Coacervation

Figure 5. Derivative of storage modulus with temperature determined from isochronal temperature sweep measurements performed at 25 °C. The data clearly indicate melting temperatures, Tm ≈ 42, 37, and 35 °C for the coacervate matrix after and before drug-loaded samples, clearly implying that the drug was hosted inside the matrix.

Biomacromolecules, Vol. 10, No. 1, 2009

187

Figure 7. Temporal evolution of size polydispersity observed in the supernatant and measured at 25 °C. See the text for details.

Figure 6. Temporal evolution of supernatant turbidity and mean diameter of aggregates present in the supernatant, which was measured at 25 °C. See the text for details.

Thus, there was no discernible distinction between the before and after drug-loaded samples as far as G′′ dispersion behavior was concerned. Under low strain conditions, the viscoelastic theory predicts24 the elastic modulii scale with frequency to be G′ ∼ ω2 and G′′ ∼ ω1. The measured values for n′ for coacervate and one of the drug-loaded samples (after) were close to 2, whereas for the before sample, the exponent value was found to be high () 2.9). However, the exponent n′′ was found to be the same ()1.6 ( 0.1) for both the drug-loaded samples and the coacervate material. The storage modulus is the measure of elastic energy stored in the material, and thus a perfectly structured network will be associated with a high G′ value. The experimental results show that G′ consistently decreased as the drug was added to the coacervate matrix. This implies that the network was considerably disturbed and weakened because of the impregnation of the drug inside the matrix. This effect was profound in the case of before loaded drugs as compared with that of after loaded samples. If the drug molecules adhere to the outer surface of the matrix, then the dispersive behavior of G′ is not likely to change. However, the G′′ values for drug-loaded samples were higher than that of the coacervate matrix, indicating a gain in the fluidity of the material because of drug encapsulation. The isochronal temperature sweep experiments mapped the temperature profiles of G′, and Figure 5 depicts the temperature dependence behavior of (dG′/dT). This clearly establishes the existence of coacervate melting temperature, Tm ≈ 42 °C.

Figure 8. (A) The molecular structure of the drug salbutamol. Its sulfate variant has the molecular formula [C13H21NO3] · 2H2SO4.(B) Time-dependent release of drug salbutamol sulfate encapsulated inside coacervate matrix. The molecular structure of the drug is shown in A. A biphasic release pattern is clearly observed. See the text for details.

Gelatin-A and -B samples are associated with gelation melting temperatures that are typically ∼30 °C. Gelatin gels are formed because of extensive intermolecular hydrogen bonding, whereas the coacervate owes its origin to strong electrostatic interactions, which explains the higher melting temperature associated with this coacervate. The drug-loaded matrix was observed to melt at a temperature that was lower than Tm by 5 °C, indicating that the drug was hosted inside the matrix and that its presence altered the internal structure of the coacervate, which is supported by the frequency sweep results presented in Figure 3. Supernatant Behavior. We periodically examined the supernatant to find the presence of intermolecular soluble aggregates, and the results are presented in Figures 6 and 7. The mean aggregate diameter, Dmean, and the solution turbidity

188

Biomacromolecules, Vol. 10, No. 1, 2009

Figure 9. Fit of the initial portion of the drug release data (t < 10 h) to eq 3, which provides excellent χ2 values (>0.92). See the text for details (symbols used are same as in Figure 8).

Tiwari et al.

solution showed more encapsulation and better release percentage as compared with the drug that was added to the coacervate material. However, the release profiles were found to be identical. The release of the drug from the coacervate matrix is mostly driven by osmotic pressure. Several factors contribute to the release kinetics, which include the viscosity of the matrix, the concentration of the drug and its size, the solubility of the drug, the diffusion coefficient of the drug inside the matrix, and so on. It is nearly impossible to account for all of these contributions and tailor them into a model. Therefore, although many models for describing release kinetics have been proposed, none of these capture all of the observed experimental features. This encouraged us to describe our results through empirical fitting functions. The experimental data pertaining to the release occurring (Q(t) in %) in the first 8 h could be least-squares fitted (Figure 9) to the mathematical function (arbitrarily chosen)

Q(t) ) Q∞[1 - exp(-kt)]

Figure 10. Fit of the final portion of the drug release data (t > 10 h) to eq 4, which provides excellent χ2 values (>0.95). See the text for details (symbols used are same as in Figure 8).

exhibited an oscillatory time-dependent profile. The same was true with the polydispersity parameter. In the beginning, the aggregates had a mean diameter of ∼400 nm and a polydispersity of ∼27%, which monotonously grew to Dmean ≈ 900 nm and polydispersity ≈ 35% within a span of ∼10 h with a concomitant increase in turbidity from 18 to 45%. This, we believe, happens because of the simultaneous formation of larger aggregates from smaller units through residual interactions and Ostwald ripening of coacervate droplets. The 900 nm hydrated aggregates constitute coacervate droplets (turbidity appears at this stage) that immediately sediment to the bottom of the reaction beaker, thereby reducing the Dmean, turbidity, and polydispersity of the system. However, the residual interactions still present continue to drive the formation of larger aggregates, and in a period of another ∼10 h, the coacervate droplets again reach a size that ensues another round of sedimentation. As the concentration of available protein aggregates decrease with time, this process considerably slows down after ∼60 h, which is clearly evident from Figure 6. This observation reveals that the coacervation phenomenon persists in the supernatant for a long period of time even after the removal of coacervate via centrifugation. Release of Salbutamol Sulfate. The molecular structure of SS is shown in Figure 8A. The rheology data presented in Figures 3 and 4 clearly show substantial changes in the storage and loss modulii values for drug-containing coacervate samples. This implied that the drug was firmly encapsulated inside the coacervate matrix, which also contributed to the softening of the coacervate structure. The profile of release of drug to the gastric fluid medium is shown in Figure 8B, which indicates a biphasic release mechanism. The drug added to the coacervating

(3)

The exponential growth to saturation described by eq 3 is associated with the release time constant k, and at saturation, the released quantity is Q∞. The fitting yielded k ) ((2.0 ( 0.3) and (1.0 ( 0.2)) × 10-4 s-1 for the before and after coacervation drug-loaded samples, respectively. These data could not be fitted to any of the well-known model release functions,25 namely, zero-order (Q(t) ) Q∞kt), Higuchi (Q(t) ) Q∞kt1/2), Higson-Crowell (Q(t) ) Q∞(1 - kt)1/3), Korsmeyer (Q(t) ) Q∞ktn), or Peppas-Sahlin (Q(t) ) Q∞(k1tm + k2t2m)) equations. If one assumes that the amount of drug released from the coacervate matrix is directly proportional to that appearing in the gastric fluid, then the release kinetics can be written in a differential form as

[

dQ(t) Q(t) ) (Q∞k) 1 dt Q∞

]

(4)

Equations 3 and 4 are consistent. The second phase of release occurred in the time span of 23-55 h, which is shown in Figure 10. This data could be leastsquares fitted to a logarithmic growth function (arbitrarily chosen) defined by

Q(t) ) a + Q0 ln t

(5)

where a is a constant and Q0 defines the initial release. Equation 5 implies no saturation at t f ∞, which is a limitation of this representation. Thus, it is necessary to set up a cutoff for Q(t) at t f ∞ that is not trivial. Regardless, the release kinetics could be adequately described by this equation. A sufficient number of drug release reports of SS from any of the popular matrices is hard to find in the literature, which makes the possibility of comparison difficult. PEG-modified solid lipid nanoparticles were used to encapsulate SS, and the release kinetics was probed in phosphate buffer (pH 7.2). The release pattern was observed to be biphasic.26 The initial release occurred in the first 8 h, followed by a prolonged and sustained release that continued for 14 days. Though the exact timedependent release pattern from the coacervate matrix was different, nonetheless, there is a qualitative similarity between this result and our observations. It has also been reported that SS penetrates hydrophilic but not hydrophobic membranes,27 which increases its potential for transdermal delivery.16 Interestingly, the bioavailability of SS has been found to depend substantially on the delivery mode.28

Kinetics of Protein-Protein Complex Coacervation

Conclusions The kinetics of protein-protein coacervation transition leading to phase separation was studied under controlled conditions. The intermolecular complexation between gelatin-A and -B chains could be achieved through electrostatic interactions brought about by pH titration. The titration profile had distinguishable pHs where intermolecular interactions ensued (pHC), followed by the pH where turbidity appeared (pHΦ) and pHprep, where precipitation led to liquid-solid phase separation. The coacervate material was extracted, and the rheological analysis of the same attributed heterogeneous viscoelastic property to this substance. The model drug SS was found to be better encapsulated if it was added to the coacervating solution before it reached pHc. The in vitro release kinetics in simulated gastric fluid was found to be biphasic in nature. The drug release could not be described through any of the known release models. Several mechanisms are at play during the drug release process that need to be accounted for. These are the dissolution kinetics of the matrix, the permeation of solvent in the matrix, and the diffusion of drug in a crowded environment, which is thought to be non-Fickian. In fact, for the coacervate system, the polymer chains continue to interact through a residual electrostatic force that substantially increases the complexity of the problem. This encouraged us to describe the release kinetics through a set of empirical relations. Coacervate medium has been proposed to be an alternative encapsulation matrix for hydrophilic drugs with promising potential. Acknowledgment. We thank Dr. Javed Ali of Jamia Hamdard University, New Delhi, for material help and Dr. Anita Saxena for assistance in drug release work.

References and Notes (1) Bungenberg de Jong, H. G. Chapter III. In Colloid Science; Kruyt, H. R., Ed.; Elsevier: New York, 1949. (2) Tsuchida, E.; Abe, K. Interactions between Macromolecules in Solution and Intermacromolecular Complexes; Springer: Berlin, 1982.

Biomacromolecules, Vol. 10, No. 1, 2009

189

(3) Park, J. M.; Muhoberac, B. B.; Dubin, P. L.; Xia, J. Macromolecules 1992, 25, 290–295. (4) Kaibara, K.; Okazaki, T.; Bohidar, H. B.; Dubin, P. L. Biomacromolecules 2000, 1, 100–107. (5) Gupta, A. N.; Bohidar, H. B. J. Phys. Chem. B 2007, 111, 10137– 10145. (6) Singh, S. S.; Aswal, V. K.; Bohidar, H. B. Int. J. Biol. Macromol. 2007, 41, 301–307. (7) Burgess, D. J. J. Colloid Interface Sci. 1990, 140, 227–238. (8) Smith, A. E. Science 1967, 214, 1038–1040. (9) Zezin, A. B.; Izumrudov, V. A.; Kabanov, V. A. Makromol. Chem., Macromol. Symp. 1989, 26, 249. (10) Kokufuta, E. Prog. Polym. Sci. 1992, 17, 647–697. (11) Kayitmazer, A. B.; Bohidar, H. B.; Mattison, K. M.; Bose, A.; Sarkar, J.; Hashimoto, A.; Russo, P. S.; Jaeger, W.; Dubin, P. L. Soft Matter 2007, 3, 1064–1076. (12) Bohidar, H. B.; Dubin, P.; Majhi, P.; Tribet, C.; Jaeger, W. Biomacromolecues 2005, 6, 1573–1585. (13) Mohanty, B.; Bohidar, H. B. Biomacromolecues 2003, 4, 1080–1086. (14) Gupta, A.; Bohidar, H. B. J. Chem. Phys. 2006, 125, 054904-1– 054904-7. (15) Martindale: The Complete Drug Reference, 34th ed.; Sweetman, S. C., Ed.; Pharmaceutical Press: London, 2005. (16) Bendas, E. R.; Tadros, M. I. AAPS PharmSciTech 2007, 8, E1–E8. (17) Vardi, Y.; Regev, I.; Rosenbum, M.; Fletcher, S. J. Neurology 1983, 230, 43–55, See also: http://www.geocities.com/ask_lpumsun. (18) Veis, A. The Macromolecular Chemistry of Gelatin; Academic Press: New York, 1964. (19) Bohidar, H. B. In Handbook of Polyelectrolytes and Their Applications; Nalwa, H. S., Kumar, J., Tripathy, S. K., Eds.; American Scientific Publishers: Stevenson Ranch, CA, 2002; pp 117-144; Vol. II. (20) Ohshima, H. AdV. Colloid Interface Sci. 1995, 62, 189–235. (21) Jena, S. S.; Bohidar, H. B. J. Chem. Phys. 1994, 100, 6888–6895. (22) Pezron, I.; Djabourov, M.; Leblond, J. Polymer 1991, 32, 3201–3209. (23) Tanford, C. Physical Chemistry of Macromolecules; Wiley: New York, 1961. (24) Barnes, H. A. A Handbook of Elementary Rheology; University of Wales Institute of Non-Newtonian Fluid Mechanics: Aberystwyth, England, 2000. (25) Martineau, L.; Horan, M. A.; Rothwell, N. J.; Little, R. A. Clin. Sci. 1992, 83, 615–621. (26) Hong, Y.; Hu, F. Q.; Yuan, H. Pharmazie 2006, 61, 312–315. (27) Lin, S.-Y.; Lin, Y.-Y.; Chen, K. S. Pharm. Res. 1996, 13, 914–919. (28) Taha, E.; Zaghloul, A. A.; Samy, A. M.; Al-Saidin, S.; Kassema, A. A.; Khan, M. A. Int. J. Pharm. 2004, 279, 3–7.

BM801160S