- Email: [email protected]
pH- and Thermo-sensitive Hydrogel Nanoparticles
JOURNAL OF COLLOID AND INTERFACE SCIENCE ARTICLE NO.
206, 361–368 (1998)
CS985692
pH- and Thermo-sensitive Hydrogel Nanoparticles Sanjeeb Kumar Sahoo,* Tapas K. De,* P. K. Ghosh,†,1 and Amarnath Maitra*,1 *Department of Chemistry, University of Delhi, Delhi 110 007, India; and †Department of Biotechnology, Government of India, New Delhi 110 003, India Received January 21, 1998; accepted May 26, 1998
pH- and temperature-sensitive hydrogel nanoparticles of copolymers of vinylpyrrolidone (VP) and acrylic acid (AA) cross-linked with NN* methylene bis acrylamide (MBA) of sizes up to 50 nm diameter loaded with marker compound FITC-dextran (mol wt. 19.3 kD) were prepared in the aqueous core of reverse micellar droplets and were dispersed in aqueous buffer. These particles have high entrapment efficiency, and the lyophilized powder can be redissolved in buffer without any significant agglomeration. The release of FITC-dextran from these particles was found to be pH- and temperature-dependent. The release was slow in acid solution, but it increased considerably as the pH of the medium was increased. The release rate was also increased with the increase of temperature. © 1998 Academic Press Key Words: PVP; hydrogel nanoparticles; pH-controlled release; bioadhesive polymers.
INTRODUCTION
Oral delivery of drugs can be significantly improved by using nanoparticles as carriers (1–3). The extent and the pathway of uptake of the nanoparticle has been found to be different in different parts of the intestine (4). The phenomenon and the mechanism of orally administered particles into the bloodstream are indeed complex. Investigation of the fate of particles of less than 60 nm in diameter orally administered for delivery into the lymphatic system or blood is expected to come into sharp focus in future studies (5). Since the major pathway of uptake of these particles appears to be via the M-cells and payer’s patches in the gut (6), the uptake would increase with increasing hydrophobicity (7) and decreasing particle size (3, 8). Anticipating this, we got interested in the preparation of hydrophilic polymeric nanoparticles of 10 to 100 nm diameter with narrow size distribution (9 –11). We used aqueous core of reverse micellar droplets as host nanoreactors to regulate the size of these particles. In addition to drug delivery via uptake of intact particles, enhanced delivery was also observed through a direct interaction of the nanoparticles with membranes (12). Oral drug delivery with nanoparticles, therefore, may be further enhanced by addition of mucoadhesive substances to the nanoparticles (3). 1
To whom correspondence should be addressed. E-mail: maitra@ giasdl01.vsnl.net.in
Among the controlled oral drug delivery systems, hydrogels have been extensively exploited for biomedical applications due to their high water content and excellent biocompatibility (13, 14). The pH-sensitive hydrogels containing pendant acidic or basic groups such as carboxylic acids, sulphonic acids, primary amines, or ammonium salts which change ionization in response to change in the pH have become the subject matter of major interest for use as carriers in oral drug delivery research (15, 16). The extent of interaction, adhesion, and uptake of nanoparticles of broad spectrum sizes after oral administration have been reported to be highest for the smallest particles (11). In this paper we report the preparation of nanoparticles of up to 50 nm diameter which are co-polymers of biocompatible materials made from vinylpyrrolidone and acrylic acid monomers crosslinked with NN9methylene bis acrylamide and which were prepared in reverse micelles for precisely controlling the particle size. FITC-dextran was used as a marker compound which was entrapped in these nanoparticles. We observed that these smart hydrogel polymers were immensely sensitive to pH and temperature effects on the release of the entrapped marker compound, as has been discussed here. EXPERIMENTAL
Materials AOT (Sodium bis 2-ethylhexylsulphosuccinate), N,N,N9,N9 tetramethyl-ethylene diamine (TMED), N,N9methylene bis acrylamide (MBA), and fluorescein isothiocyanate dextran (FITC-Dx) were products of Sigma, USA, and were used directly without further purification. n-Hexane (99%), sodium monohydrogen phosphate and dihydrogen phosphate, and Ferrous ammonium sulfate (FAS) were procured from SRL (India). Acrylic acid and vinylpyrrolidone were purchased from Fluka and were used freshly distilled before polymerization. Doubly distilled water was used. Preparation of Nanoparticles The nanoparticles of these copolymers were prepared following the methods described in our recent patent and communication (11), the outline of which is described as follows. The surfactant, sodium bis-2-ethylhexylsulfosuccinate, or
361
0021-9797/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.
362
SAHOO ET AL.
Aerosol TO (i.e., AOT), has been dissolved in n-hexane (usually 0.03 to 0.1 M of AOT solution). Water soluble monomers, vinylpyrrolidone, and acrylic acid were used in different molar ratios (90:10, 80:20, and 60:40), and the polymer was crosslinked with N,N9 methylene bis acrylamide, MBA) for nanoparticle preparation. Aqueous solutions of monomer, crosslinking agent, initiator, and FITC dextran were added to AOT solution in hexane and the polymerization was carried out following the standard procedure. An additional amount of buffer may be added in reverse micelles in order to get the host micellar droplets of the desired size. In a typical experiment for the preparation of nanoparticles of copolymer of vinylpyrrolidone with acrylic acid, copoly[VP-AA], containing FITC-Dx as an encapsulated marker material (mol wt 19.3 kD), we have taken 40 ml of 0.03 M AOT solution in hexane in which 280 mliter freshly distilled vinylpyrrolidone and 36 mliter freshly distilled acrylic acid, 100 mliter MBA (0.049 g/ml) as crosslinking agent, 20 mliter 1% FAS, 20 mliter 11.2% aqueous solution of TMED, 30 mliter 20% ammonium persulphate as initiator, and 50 mliter of FITC-Dx (160 mg/ml) were added. The solution was homogeneous and optically transparent. Polymerization was done in N2 atmosphere at 35°C for 8 h in a thermostatic bath with continuous stirring. The above method produced co-poly[VP-AA] nanoparticles cross-linked with MBA and containing FITC-Dx as encapsulated material. The organic solvent was evaporated off in a rotary evaporator, and the dry mass was resuspended in 5 ml of water by sonication. A calculated amount of 30% CaCl2 solution was added drop by drop with continuous stirring to precipitate the surfactant as calcium salt of bis(2-ethylhexyl)sulphosuccinate, (Ca(DEHSS)2). The centrifuged (10,000 rpm for 10 min) aqueous solution contains nanoparticles which are homogeneous and transparent. The cake of Ca(DEHSS)2 after centrifugation contains some nanoparticles absorbed in it. It was dissolved in 10 ml n-hexane, and the hexane solution was washed 2–3 times each time with 1 ml of water. The phase-separated aqueous layer was drained out and added to the original centrifugate. The total aqueous dispersion of nanoparticles was then dialyzed for about 2–3 h using a spectrapore membrane dialysis bag (12 kD cut off). The dialyzed solution was lyophilized immediately to dry powder for subsequent use. Lyophilized nanoparticles are easily redisperseable in aqueous buffer. The sizes of these particles in buffer were found to be the same before and after lyophilization. It established that aggregation does not take place by drying these particles. The flow diagram for the preparation of the nanoparticles through reverse micelles is shown in Fig. 1. Size Determination (i) QELS measurements. Dynamic laser light scattering measurements for determining the size of the nanoparticles were performed using a Brookhaven 9000 instrument with a BI200SM goniometer. An air-cooled argon ion laser was op-
erated at 488 nm as the light source. The time dependence of the intensity autocorrelation function of the scattered intensity was derived by using a 128-channel digital correlator. The size of the nanoparticles was determined from diffusion of the particles using the Stokes–Einstein equation and the representative size distribution spectra are shown in Fig. 2a. (ii) TEM pictures. Two hundred microliters of the dialyzed solution was diluted to 50 ml in buffer to have a clear solution, and the samples for TEM were prepared with this. The TEM picture was taken in a JEOL JEM2000 Ex 200 model electron microscope. The pictures are shown in Fig. 2b. Entrapment Efficiency The entrapment efficiency (E%) of copoly[VP-AA] nanoparticles containing FITC-Dx was determined after filtering the particles from free FITC-Dx molecules using a millipore UFP2THK24 (100kD cut off) membrane filter and measuring the quantity of free FITC-Dx in the solution. The original amount of FITC-Dx (entrapped 1 free) is already known. In Vitro Release Kinetic Studies A known amount of lyophilized copoly[VP-AA] nanoparticles encapsulating FITC-Dx was suspended in 10 ml buffer of particular pH in which release kinetics are to be studied. The solutions were distributed at 500 mliter each in 20 microtubes kept at constant temperature. At a predetermined interval of time the solution was filtered through the millipore filter, as indicated above, to separate free FITC-Dx from the nanoparticles. The concentration of free FITC-Dx was determined spectrophotometrically at pH 8.6 at 494 nm wavelength. The release kinetics were studied at pH 3.0, 7.0, and 10.0 at 22°C. The release kinetics at 4, 22, and 37°C were also studied in a similar way at pH 7.0. Polymer Swelling Studies Solid hydrogel of copoly[VP-AA20] was prepared in an aqueous medium as follows. A mixture of 5.6 ml VP, 760 mliter acrylic acid, 40 mliter TMED, 40 mliter FAS, and 2 ml MBA (same concentration as used earlier) was freed from oxygen by passing nitrogen gas for about 15 min, and then 300 mliter ammonium persulfate was added while the nitrogen bubbling was continued for 1 h. A solid transparent gel was formed which was dried at 50°C in a vacuum oven for 24 h. The weighed dry gel in the form of a tablet was placed in a beaker and an aqueous buffer was added into it for swelling. The increase in weight of the polymer material was noted at different times. RESULTS
Size and Polydispersity of the Hydrogel Nanoparticles The sizes of the copoly[VP-AA] nanoparticles entrapping FITC-Dx measured by QELS and by TEM are shown in Fig. 2.
FIG. 1.
Flow diagram for the preparation of nanoparticles through reverse micelles.
HYDROGEL NANOPARTICLES
363
364
SAHOO ET AL.
FIG. 2a.
QELS spectra of the copoly[VP-AA] nanoparticles (i) [VP-AA0], (ii) [VP-AA10], and (iii) [VP-AA20].
Four samples containing different molar ratios of VP and AA were prepared. The resulting polymers were named as copoly[VP-AA00], copoly[VP-AA10], copoly[VP-AA20], and copoly[VP-AA40] which contain 0, 10, 20, and 40% AA, respectively, in moles per 100 mol of total monomers. The results clearly indicate that with the increased concentration of AA in the mixture the size of the nanoparticles is also increased from 35 nm diameter to about 60 nm diameter when the AA concentration was increased from 0 to 40%. With further increased concentration of AA in the mixture, the particles’
size becomes too large for them to dissolve in aqueous buffer. The self-adhesive behavior of the AA-derived polymer is imminent from such results. Entrapment Efficiency The E% of these nanoparticles of material, copoly[VPAA20], entrapping FITC-Dx was found to be about 40% in the present 2.5% w/w loading of FITC-Dx in the polymeric material. The E% was, however, found to be more or less constant
HYDROGEL NANOPARTICLES
FIG. 2b.
TEM pictures of copoly[VA-AA] nanoparticles (i) [VP-AA0], (ii) [VP-AA10], and (iii) [VP-AA20].
365
366
SAHOO ET AL.
tion as a function of swelling time. Swelling is increased quickly with the time of treatment, faster in the initial steps, reaching a constant value after a time depending on the pH of the medium. The time to reach equilibrium in acidic solution is much longer than that required in neutral and alkaline solutions.
FIG. 3. Time-dependent release of FITC-dextran from copoly[VP-AA20] hydrogels in various pH solutions at 22°C.
even at higher loading up to 3.5% w/w, above which the E% falls rapidly. The entrapment efficiency remains within 40 – 50% for copolymers of other compositions also. In Vitro Release Kinetic Studies Effect of pH. The in vitro release profiles of the entrapped FITC-Dx from copoly[VP-AA20] nanoparticles at pH 3.0, 7.0, and 10.0 were determined at 22°C, and the results are shown in Fig. 3. It is evident from the figure that the release rate was considerably slow in acidic pH compared to that in higher pH, and the release rate was found to increase exponentially with a rise in the pH of the media. As shown in Fig. 3, the in vitro release profile of FITC-Dx from the copoly[VP-AA20] nanoparticles exhibits an initial burst of about 60% release for the first day in pH 10 solution followed by a gradually reduced release for the next two days, giving rise to about 85% release of total dye beyond which the release rate is practically reduced to zero. At pH 7 only 40% of the entrapped material is released on the first day, and it takes about a week to release about 80% of the material. The release is extremely slow in acidic pH. Extrapolated release data indicate that at pH 1 (which is the pH of gastric fluid), only 2–3% of the material is released on the first day. One of the most interesting characteristics of these copolymers is their ability to swell at different pH values. The swelling behavior of the solid copolymer gels was studied by measuring the increase in weight of the polymer due to swelling. The percentage of hydration of gel was obtained according to the equation H% 5
Ww 2 Wo 3 100, Wo
where Wo and Ww are the weights of the polymer before and after swelling, respectively. Figure 4 shows the dynamic swelling expressed as the variation of the percent of hydra-
Effect of temperature. The temperature-dependent release profile of copoly[VP-AA20] nanoparticles for FITC-Dx from its matrix dispersed at pH 7.0 also shows (Fig. 5) that the release is highly temperature dependent. The percentage of release of FITC-Dx in pH 7 solution is about 18% at 4°C while it is about 30 and 50% at 22 and 37°C, respectively. The release rate is more or less steady in the first 24 h, after which it is slowed down considerably. Increase in release rate with temperature perhaps is direct evidence of a diffusion-controlled process of FITC-Dx release from the nanoparticles. DISCUSSION
The results of these experiments have enabled us to establish the experimental conditions needed to prepare copoly[VP-AA] hydrogel nanoparticles of a size of up to 50 nm diameter using aqueous core of reverse micellar droplets as nanoreactors to control the particle size. The size of the particle can be reduced down to 35 nm diameter by decreasing the quantity of acrylic acid fraction in the monomer mixture. Thus, the amount of acrylic acid in the polymer was found to influence strongly the size and the size distribution profile of the nanoparticles formed. VP and AA in the mole percent ratio 50:50 and below leads to the formation of particles that are too big (.100 nm) to dissolve completely in aqueous buffer. The concentration of acrylic acid in the polymerization medium demonstrates the strong influence on the size and distribution profiles of the nanoparticles. Pure polyacrylic acid is a tacky polymer. Therefore, as the concentration of AA is increased in the VP 1 AA mixture the more agglomeration of particles due to increased self-adhesion gives rise to the formation of bigger particles. This leads us to conclude, therefore, that the acrylic acid concentration in the monomer mixture has to be adjusted so that well-characterized nanoparticles of the desired size may be obtained without the particles sticking together. We have selected copoly[VP-AA20] as an optimum composition which produces particles of an average size of about 50 nm with controlled self-adhesiveness. FITC-Dx is a large molecule (mol wt 19.3 kD). Therefore simple diffusional release of the dye molecules from the gel matrix of the nanoparticles is difficult unless the polymer is swelled and/or eroded in an aqueous medium. The structure of hydrogels based on vinylpyrrolidone and acrylic acid crosslinked with NN9 methylene bis acrylamide can be schematically represented as follows.
HYDROGEL NANOPARTICLES
FIG. 4.
Time-dependent swelling of copoly[VP-AA20] solid polymer in different buffers.
©CH¤©CH©CH¤©CH©CH¤©CH©CH¤©CH© 1 2 x
N
O
COOH
CO
N
O
NH CH¤ NH O
N
COOH
367
CO
O
N
©CH¤©CH©CH¤©CH©CH¤©CH©CH¤©CH© 1 2 x
Polymeric erosion through chain cleavage may take place through amide bond hydrolysis with consequent generation of loose polymeric networks. However, Torchilin et al. (17) indicated that the rate of crosslink cleavage of amide bonds in MBA crosslinked polymers was extremely slow, particularly when the networks are prepared with more than 1% w/w MBA as in the present polymeric system. The rate of release of the large mole-
cules can, however, be controlled by polymeric erosion rather than diffusion by readily manipulating the simple structural variation of the hydrogels. Thus, when an easily hydrolyzable group such as ester is present in the network, the polymer becomes more erodible and the release of entrapped molecules is preferentially controlled by the erosion of the polymer rather than diffusion of the former (18). The stability of the amide linkage has been experimentally confirmed to be more than that of the ester linkage (19). Therefore, it is certain that the release of FITC-Dx from hydrogel nanoparticles precedes polymer degradation and it is a diffusioncontrolled process from highly swelled gels. From Fig. 4 it is apparent that these hydrogels, which are water insoluble hydrophilic polymers, swell in water enormously depending on the pH of the medium (20, 21). The swelling is much faster than the release of the dye, and the release is relatively much faster than the polymer hydrolysis (17). Polymer erosion starts long after the sufficient amount of dye is released to reach an equilibrium. In acidic solutions, acrylic acid based polymers are more lypophilic and, therefore, cannot swell sufficiently so as to allow the diffusion of FITC-Dx out of the nanoparticles to a significant extent
368
SAHOO ET AL.
REFERENCES
FIG. 5. Time-dependent release of FITC-dextran from copoly[VP-AA20] hydrogels in pH 7.0 solution and at different temperatures.
(22, 23). The swelling behavior of poly(ethoxy triethyleneglycol monomethylacrylate) was studied previously (24) and was ascribed to a transition from compact coiled globular morphology in acidic medium to an extended polymeric chain in nonacidic medium. Copoly[VP-AA] hydrogels are hydrophobic in acidic medium and hydrophilic in neutral and alkaline media due to different degrees of dissociation of the carboxylic groups in the chain. It is, therefore, expected that hydrophobic residues would favor the formation of compact coil arrangement at low pH values while an extended polymeric chain in extensively swelled polymer in higher pH is responsible for an increase in porosity of the matrix and thus would favor the slow diffusion of the entrapped dye. CONCLUDING REMARKS
By judicially adjusting the proportion of VP and AA monomers, the copolymer prepared by using MBA as cross-linking agent could be imparted with manipulative release properties of the substances entrapped in the nanoparticle matrices. The sticky properties of the pure polyacrylic acid material could be substantially modulated, and the particle size and size distribution can be well regulated through the present preparative method of making polymeric nanoparticles. These smart materials could be of enormous value in delivering various substances into the body through in vivo applications.
1. Maincent, P., LeVerge, R., Sado, P., Couvreur, P., and Devissaguet, J. P., J. Pharm. Sci. 75, 955 (1986). 2. Beck, P. H., Kreuter, J., Muller, W. E. G., and Schatton, W., Eur. J. Pharm. Biopharm. 40, 134 (1994). 3. Kreuter, J., in “Colloidal Drug Delivery Systems” (J. Kreuter, Ed.), pp. 219 –342. Dekker, New York, 1994. 4. Michel, C., Aprahamian, M., Dafontaine, L., Couvreur, P., and Damge, C., J. Pharm. Pharmacol. 43, 1 (1991). 5. Florence, A. T., Pharm. Res. 14, 259 (1997). 6. Kreuter, J., Adv. Drug Delivery Rev. 7, 71– 86 (1991). 7. Eldridge, J. H., Hammond, C. J., Meulbroek, J. A., Staas, J. K., Gilley, R. M., and Tice, T. R., J. Controlled Release 11, 205 (1990). 8. Jani, P., Halbert, G. W., Langridge, J., and Florence, A. T., J. Pharm. Pharmacol. 41, 809 (1989); 42, 821 (1990). 9. Munshi, N., Chakravarty, K., De, T. K., and Maitra, A. N., Colloid Polym. Sci. 273, 464 (1995). 10. Munshi, N., De, T. K., and Maitra, A. N., J. Colloid Interface Sci. 180, 387 (1997). 11. Maitra, A. N., Ghosh, P. K., De, T. K., and Sahoo, S., A Method for the Preparation of Drug Loaded Hydrophilic Polymeric Nanoparticles. U.S. Patent Application No. 08/779,904, dated 7.1.97; Sahoo, S., Gaur, U., De, T. K., Ghosh, P. C., Ghosh, P. K., and Maitra, A., RES Evading and Long Circulating Hydrophilic Nanoparticles, communicated, 1998. 12. Kreuter, J., in “Topics in Pharmaceutical Sciences” (D. D. Breimer and P. Speiser, Eds.), pp. 359 –370. Elsevier, Amsterdam, 1983. 13. Kost, J., and Langer, R., in “Hydrogels in Medicine and Pharmacy” (N. A. Peppas, Ed.), Vol. 3, pp. 95–108. CRC, Boca Raton, FL, 1986. 14. Pitt, C. G., Int. J. Pharm. 59, 173 (1990). 15. Heller, J., Chang, A. C., Rodd, G., and Grodsky, G. M., J. Controlled Release 13, 295 (1990). 16. Siegal, R. A., and Firstone, B. N. A., Macromolecules 21, 3254 (1988). 17. Torchilin, V. P., Tschenko, E. G., Sminov, V. N., and Chazov, E. D., J. Biomed. Mater. Res. 11, 223 (1997). 18. Heller, J., Helwing, R. F., Baker, R. W., and Tuttle, M. E., Biomaterials 4, 262 (1983). 19. Lindhardt, R. J., in “Controlled Release of Drugs: Polymers and Aggregated Systems” (M. Rosoff, Ed.), p. 53. VCH, Weinheim/New York, 1988. 20. Hoffman, A. S., Chen, G. H., Kaang, S. Y., Ding, Z. L., Randeri, K., and Kabra, B., Adv. Biomater. Biomed. Engng. Drug Delivery Syst. [Iketani Conf. Biomed. Polym., 5th 1995], 62, (1996). 21. Ravichandran, P., Shantha, K. L., and Panduranga Rao, K., Int. J. Pharm. 154, 89 (1997). 22. Leong, K. W., Brott, B. C., and Langer, R., J. Biomed. Mater. Res. 2, 167 (1985). 23. Ron, E., Turek, T., and Mathiowitz, E., Proc. Natl. Acad. Sci. USA 90, 4176 (1993). 24. Vazquez, B., Gurruchaga, M., Goni, I., Narvarte, E., and San Roman, J., Polymer 36, 3327 (1995).
206, 361–368 (1998)
CS985692
pH- and Thermo-sensitive Hydrogel Nanoparticles Sanjeeb Kumar Sahoo,* Tapas K. De,* P. K. Ghosh,†,1 and Amarnath Maitra*,1 *Department of Chemistry, University of Delhi, Delhi 110 007, India; and †Department of Biotechnology, Government of India, New Delhi 110 003, India Received January 21, 1998; accepted May 26, 1998
pH- and temperature-sensitive hydrogel nanoparticles of copolymers of vinylpyrrolidone (VP) and acrylic acid (AA) cross-linked with NN* methylene bis acrylamide (MBA) of sizes up to 50 nm diameter loaded with marker compound FITC-dextran (mol wt. 19.3 kD) were prepared in the aqueous core of reverse micellar droplets and were dispersed in aqueous buffer. These particles have high entrapment efficiency, and the lyophilized powder can be redissolved in buffer without any significant agglomeration. The release of FITC-dextran from these particles was found to be pH- and temperature-dependent. The release was slow in acid solution, but it increased considerably as the pH of the medium was increased. The release rate was also increased with the increase of temperature. © 1998 Academic Press Key Words: PVP; hydrogel nanoparticles; pH-controlled release; bioadhesive polymers.
INTRODUCTION
Oral delivery of drugs can be significantly improved by using nanoparticles as carriers (1–3). The extent and the pathway of uptake of the nanoparticle has been found to be different in different parts of the intestine (4). The phenomenon and the mechanism of orally administered particles into the bloodstream are indeed complex. Investigation of the fate of particles of less than 60 nm in diameter orally administered for delivery into the lymphatic system or blood is expected to come into sharp focus in future studies (5). Since the major pathway of uptake of these particles appears to be via the M-cells and payer’s patches in the gut (6), the uptake would increase with increasing hydrophobicity (7) and decreasing particle size (3, 8). Anticipating this, we got interested in the preparation of hydrophilic polymeric nanoparticles of 10 to 100 nm diameter with narrow size distribution (9 –11). We used aqueous core of reverse micellar droplets as host nanoreactors to regulate the size of these particles. In addition to drug delivery via uptake of intact particles, enhanced delivery was also observed through a direct interaction of the nanoparticles with membranes (12). Oral drug delivery with nanoparticles, therefore, may be further enhanced by addition of mucoadhesive substances to the nanoparticles (3). 1
To whom correspondence should be addressed. E-mail: maitra@ giasdl01.vsnl.net.in
Among the controlled oral drug delivery systems, hydrogels have been extensively exploited for biomedical applications due to their high water content and excellent biocompatibility (13, 14). The pH-sensitive hydrogels containing pendant acidic or basic groups such as carboxylic acids, sulphonic acids, primary amines, or ammonium salts which change ionization in response to change in the pH have become the subject matter of major interest for use as carriers in oral drug delivery research (15, 16). The extent of interaction, adhesion, and uptake of nanoparticles of broad spectrum sizes after oral administration have been reported to be highest for the smallest particles (11). In this paper we report the preparation of nanoparticles of up to 50 nm diameter which are co-polymers of biocompatible materials made from vinylpyrrolidone and acrylic acid monomers crosslinked with NN9methylene bis acrylamide and which were prepared in reverse micelles for precisely controlling the particle size. FITC-dextran was used as a marker compound which was entrapped in these nanoparticles. We observed that these smart hydrogel polymers were immensely sensitive to pH and temperature effects on the release of the entrapped marker compound, as has been discussed here. EXPERIMENTAL
Materials AOT (Sodium bis 2-ethylhexylsulphosuccinate), N,N,N9,N9 tetramethyl-ethylene diamine (TMED), N,N9methylene bis acrylamide (MBA), and fluorescein isothiocyanate dextran (FITC-Dx) were products of Sigma, USA, and were used directly without further purification. n-Hexane (99%), sodium monohydrogen phosphate and dihydrogen phosphate, and Ferrous ammonium sulfate (FAS) were procured from SRL (India). Acrylic acid and vinylpyrrolidone were purchased from Fluka and were used freshly distilled before polymerization. Doubly distilled water was used. Preparation of Nanoparticles The nanoparticles of these copolymers were prepared following the methods described in our recent patent and communication (11), the outline of which is described as follows. The surfactant, sodium bis-2-ethylhexylsulfosuccinate, or
361
0021-9797/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.
362
SAHOO ET AL.
Aerosol TO (i.e., AOT), has been dissolved in n-hexane (usually 0.03 to 0.1 M of AOT solution). Water soluble monomers, vinylpyrrolidone, and acrylic acid were used in different molar ratios (90:10, 80:20, and 60:40), and the polymer was crosslinked with N,N9 methylene bis acrylamide, MBA) for nanoparticle preparation. Aqueous solutions of monomer, crosslinking agent, initiator, and FITC dextran were added to AOT solution in hexane and the polymerization was carried out following the standard procedure. An additional amount of buffer may be added in reverse micelles in order to get the host micellar droplets of the desired size. In a typical experiment for the preparation of nanoparticles of copolymer of vinylpyrrolidone with acrylic acid, copoly[VP-AA], containing FITC-Dx as an encapsulated marker material (mol wt 19.3 kD), we have taken 40 ml of 0.03 M AOT solution in hexane in which 280 mliter freshly distilled vinylpyrrolidone and 36 mliter freshly distilled acrylic acid, 100 mliter MBA (0.049 g/ml) as crosslinking agent, 20 mliter 1% FAS, 20 mliter 11.2% aqueous solution of TMED, 30 mliter 20% ammonium persulphate as initiator, and 50 mliter of FITC-Dx (160 mg/ml) were added. The solution was homogeneous and optically transparent. Polymerization was done in N2 atmosphere at 35°C for 8 h in a thermostatic bath with continuous stirring. The above method produced co-poly[VP-AA] nanoparticles cross-linked with MBA and containing FITC-Dx as encapsulated material. The organic solvent was evaporated off in a rotary evaporator, and the dry mass was resuspended in 5 ml of water by sonication. A calculated amount of 30% CaCl2 solution was added drop by drop with continuous stirring to precipitate the surfactant as calcium salt of bis(2-ethylhexyl)sulphosuccinate, (Ca(DEHSS)2). The centrifuged (10,000 rpm for 10 min) aqueous solution contains nanoparticles which are homogeneous and transparent. The cake of Ca(DEHSS)2 after centrifugation contains some nanoparticles absorbed in it. It was dissolved in 10 ml n-hexane, and the hexane solution was washed 2–3 times each time with 1 ml of water. The phase-separated aqueous layer was drained out and added to the original centrifugate. The total aqueous dispersion of nanoparticles was then dialyzed for about 2–3 h using a spectrapore membrane dialysis bag (12 kD cut off). The dialyzed solution was lyophilized immediately to dry powder for subsequent use. Lyophilized nanoparticles are easily redisperseable in aqueous buffer. The sizes of these particles in buffer were found to be the same before and after lyophilization. It established that aggregation does not take place by drying these particles. The flow diagram for the preparation of the nanoparticles through reverse micelles is shown in Fig. 1. Size Determination (i) QELS measurements. Dynamic laser light scattering measurements for determining the size of the nanoparticles were performed using a Brookhaven 9000 instrument with a BI200SM goniometer. An air-cooled argon ion laser was op-
erated at 488 nm as the light source. The time dependence of the intensity autocorrelation function of the scattered intensity was derived by using a 128-channel digital correlator. The size of the nanoparticles was determined from diffusion of the particles using the Stokes–Einstein equation and the representative size distribution spectra are shown in Fig. 2a. (ii) TEM pictures. Two hundred microliters of the dialyzed solution was diluted to 50 ml in buffer to have a clear solution, and the samples for TEM were prepared with this. The TEM picture was taken in a JEOL JEM2000 Ex 200 model electron microscope. The pictures are shown in Fig. 2b. Entrapment Efficiency The entrapment efficiency (E%) of copoly[VP-AA] nanoparticles containing FITC-Dx was determined after filtering the particles from free FITC-Dx molecules using a millipore UFP2THK24 (100kD cut off) membrane filter and measuring the quantity of free FITC-Dx in the solution. The original amount of FITC-Dx (entrapped 1 free) is already known. In Vitro Release Kinetic Studies A known amount of lyophilized copoly[VP-AA] nanoparticles encapsulating FITC-Dx was suspended in 10 ml buffer of particular pH in which release kinetics are to be studied. The solutions were distributed at 500 mliter each in 20 microtubes kept at constant temperature. At a predetermined interval of time the solution was filtered through the millipore filter, as indicated above, to separate free FITC-Dx from the nanoparticles. The concentration of free FITC-Dx was determined spectrophotometrically at pH 8.6 at 494 nm wavelength. The release kinetics were studied at pH 3.0, 7.0, and 10.0 at 22°C. The release kinetics at 4, 22, and 37°C were also studied in a similar way at pH 7.0. Polymer Swelling Studies Solid hydrogel of copoly[VP-AA20] was prepared in an aqueous medium as follows. A mixture of 5.6 ml VP, 760 mliter acrylic acid, 40 mliter TMED, 40 mliter FAS, and 2 ml MBA (same concentration as used earlier) was freed from oxygen by passing nitrogen gas for about 15 min, and then 300 mliter ammonium persulfate was added while the nitrogen bubbling was continued for 1 h. A solid transparent gel was formed which was dried at 50°C in a vacuum oven for 24 h. The weighed dry gel in the form of a tablet was placed in a beaker and an aqueous buffer was added into it for swelling. The increase in weight of the polymer material was noted at different times. RESULTS
Size and Polydispersity of the Hydrogel Nanoparticles The sizes of the copoly[VP-AA] nanoparticles entrapping FITC-Dx measured by QELS and by TEM are shown in Fig. 2.
FIG. 1.
Flow diagram for the preparation of nanoparticles through reverse micelles.
HYDROGEL NANOPARTICLES
363
364
SAHOO ET AL.
FIG. 2a.
QELS spectra of the copoly[VP-AA] nanoparticles (i) [VP-AA0], (ii) [VP-AA10], and (iii) [VP-AA20].
Four samples containing different molar ratios of VP and AA were prepared. The resulting polymers were named as copoly[VP-AA00], copoly[VP-AA10], copoly[VP-AA20], and copoly[VP-AA40] which contain 0, 10, 20, and 40% AA, respectively, in moles per 100 mol of total monomers. The results clearly indicate that with the increased concentration of AA in the mixture the size of the nanoparticles is also increased from 35 nm diameter to about 60 nm diameter when the AA concentration was increased from 0 to 40%. With further increased concentration of AA in the mixture, the particles’
size becomes too large for them to dissolve in aqueous buffer. The self-adhesive behavior of the AA-derived polymer is imminent from such results. Entrapment Efficiency The E% of these nanoparticles of material, copoly[VPAA20], entrapping FITC-Dx was found to be about 40% in the present 2.5% w/w loading of FITC-Dx in the polymeric material. The E% was, however, found to be more or less constant
HYDROGEL NANOPARTICLES
FIG. 2b.
TEM pictures of copoly[VA-AA] nanoparticles (i) [VP-AA0], (ii) [VP-AA10], and (iii) [VP-AA20].
365
366
SAHOO ET AL.
tion as a function of swelling time. Swelling is increased quickly with the time of treatment, faster in the initial steps, reaching a constant value after a time depending on the pH of the medium. The time to reach equilibrium in acidic solution is much longer than that required in neutral and alkaline solutions.
FIG. 3. Time-dependent release of FITC-dextran from copoly[VP-AA20] hydrogels in various pH solutions at 22°C.
even at higher loading up to 3.5% w/w, above which the E% falls rapidly. The entrapment efficiency remains within 40 – 50% for copolymers of other compositions also. In Vitro Release Kinetic Studies Effect of pH. The in vitro release profiles of the entrapped FITC-Dx from copoly[VP-AA20] nanoparticles at pH 3.0, 7.0, and 10.0 were determined at 22°C, and the results are shown in Fig. 3. It is evident from the figure that the release rate was considerably slow in acidic pH compared to that in higher pH, and the release rate was found to increase exponentially with a rise in the pH of the media. As shown in Fig. 3, the in vitro release profile of FITC-Dx from the copoly[VP-AA20] nanoparticles exhibits an initial burst of about 60% release for the first day in pH 10 solution followed by a gradually reduced release for the next two days, giving rise to about 85% release of total dye beyond which the release rate is practically reduced to zero. At pH 7 only 40% of the entrapped material is released on the first day, and it takes about a week to release about 80% of the material. The release is extremely slow in acidic pH. Extrapolated release data indicate that at pH 1 (which is the pH of gastric fluid), only 2–3% of the material is released on the first day. One of the most interesting characteristics of these copolymers is their ability to swell at different pH values. The swelling behavior of the solid copolymer gels was studied by measuring the increase in weight of the polymer due to swelling. The percentage of hydration of gel was obtained according to the equation H% 5
Ww 2 Wo 3 100, Wo
where Wo and Ww are the weights of the polymer before and after swelling, respectively. Figure 4 shows the dynamic swelling expressed as the variation of the percent of hydra-
Effect of temperature. The temperature-dependent release profile of copoly[VP-AA20] nanoparticles for FITC-Dx from its matrix dispersed at pH 7.0 also shows (Fig. 5) that the release is highly temperature dependent. The percentage of release of FITC-Dx in pH 7 solution is about 18% at 4°C while it is about 30 and 50% at 22 and 37°C, respectively. The release rate is more or less steady in the first 24 h, after which it is slowed down considerably. Increase in release rate with temperature perhaps is direct evidence of a diffusion-controlled process of FITC-Dx release from the nanoparticles. DISCUSSION
The results of these experiments have enabled us to establish the experimental conditions needed to prepare copoly[VP-AA] hydrogel nanoparticles of a size of up to 50 nm diameter using aqueous core of reverse micellar droplets as nanoreactors to control the particle size. The size of the particle can be reduced down to 35 nm diameter by decreasing the quantity of acrylic acid fraction in the monomer mixture. Thus, the amount of acrylic acid in the polymer was found to influence strongly the size and the size distribution profile of the nanoparticles formed. VP and AA in the mole percent ratio 50:50 and below leads to the formation of particles that are too big (.100 nm) to dissolve completely in aqueous buffer. The concentration of acrylic acid in the polymerization medium demonstrates the strong influence on the size and distribution profiles of the nanoparticles. Pure polyacrylic acid is a tacky polymer. Therefore, as the concentration of AA is increased in the VP 1 AA mixture the more agglomeration of particles due to increased self-adhesion gives rise to the formation of bigger particles. This leads us to conclude, therefore, that the acrylic acid concentration in the monomer mixture has to be adjusted so that well-characterized nanoparticles of the desired size may be obtained without the particles sticking together. We have selected copoly[VP-AA20] as an optimum composition which produces particles of an average size of about 50 nm with controlled self-adhesiveness. FITC-Dx is a large molecule (mol wt 19.3 kD). Therefore simple diffusional release of the dye molecules from the gel matrix of the nanoparticles is difficult unless the polymer is swelled and/or eroded in an aqueous medium. The structure of hydrogels based on vinylpyrrolidone and acrylic acid crosslinked with NN9 methylene bis acrylamide can be schematically represented as follows.
HYDROGEL NANOPARTICLES
FIG. 4.
Time-dependent swelling of copoly[VP-AA20] solid polymer in different buffers.
©CH¤©CH©CH¤©CH©CH¤©CH©CH¤©CH© 1 2 x
N
O
COOH
CO
N
O
NH CH¤ NH O
N
COOH
367
CO
O
N
©CH¤©CH©CH¤©CH©CH¤©CH©CH¤©CH© 1 2 x
Polymeric erosion through chain cleavage may take place through amide bond hydrolysis with consequent generation of loose polymeric networks. However, Torchilin et al. (17) indicated that the rate of crosslink cleavage of amide bonds in MBA crosslinked polymers was extremely slow, particularly when the networks are prepared with more than 1% w/w MBA as in the present polymeric system. The rate of release of the large mole-
cules can, however, be controlled by polymeric erosion rather than diffusion by readily manipulating the simple structural variation of the hydrogels. Thus, when an easily hydrolyzable group such as ester is present in the network, the polymer becomes more erodible and the release of entrapped molecules is preferentially controlled by the erosion of the polymer rather than diffusion of the former (18). The stability of the amide linkage has been experimentally confirmed to be more than that of the ester linkage (19). Therefore, it is certain that the release of FITC-Dx from hydrogel nanoparticles precedes polymer degradation and it is a diffusioncontrolled process from highly swelled gels. From Fig. 4 it is apparent that these hydrogels, which are water insoluble hydrophilic polymers, swell in water enormously depending on the pH of the medium (20, 21). The swelling is much faster than the release of the dye, and the release is relatively much faster than the polymer hydrolysis (17). Polymer erosion starts long after the sufficient amount of dye is released to reach an equilibrium. In acidic solutions, acrylic acid based polymers are more lypophilic and, therefore, cannot swell sufficiently so as to allow the diffusion of FITC-Dx out of the nanoparticles to a significant extent
368
SAHOO ET AL.
REFERENCES
FIG. 5. Time-dependent release of FITC-dextran from copoly[VP-AA20] hydrogels in pH 7.0 solution and at different temperatures.
(22, 23). The swelling behavior of poly(ethoxy triethyleneglycol monomethylacrylate) was studied previously (24) and was ascribed to a transition from compact coiled globular morphology in acidic medium to an extended polymeric chain in nonacidic medium. Copoly[VP-AA] hydrogels are hydrophobic in acidic medium and hydrophilic in neutral and alkaline media due to different degrees of dissociation of the carboxylic groups in the chain. It is, therefore, expected that hydrophobic residues would favor the formation of compact coil arrangement at low pH values while an extended polymeric chain in extensively swelled polymer in higher pH is responsible for an increase in porosity of the matrix and thus would favor the slow diffusion of the entrapped dye. CONCLUDING REMARKS
By judicially adjusting the proportion of VP and AA monomers, the copolymer prepared by using MBA as cross-linking agent could be imparted with manipulative release properties of the substances entrapped in the nanoparticle matrices. The sticky properties of the pure polyacrylic acid material could be substantially modulated, and the particle size and size distribution can be well regulated through the present preparative method of making polymeric nanoparticles. These smart materials could be of enormous value in delivering various substances into the body through in vivo applications.
1. Maincent, P., LeVerge, R., Sado, P., Couvreur, P., and Devissaguet, J. P., J. Pharm. Sci. 75, 955 (1986). 2. Beck, P. H., Kreuter, J., Muller, W. E. G., and Schatton, W., Eur. J. Pharm. Biopharm. 40, 134 (1994). 3. Kreuter, J., in “Colloidal Drug Delivery Systems” (J. Kreuter, Ed.), pp. 219 –342. Dekker, New York, 1994. 4. Michel, C., Aprahamian, M., Dafontaine, L., Couvreur, P., and Damge, C., J. Pharm. Pharmacol. 43, 1 (1991). 5. Florence, A. T., Pharm. Res. 14, 259 (1997). 6. Kreuter, J., Adv. Drug Delivery Rev. 7, 71– 86 (1991). 7. Eldridge, J. H., Hammond, C. J., Meulbroek, J. A., Staas, J. K., Gilley, R. M., and Tice, T. R., J. Controlled Release 11, 205 (1990). 8. Jani, P., Halbert, G. W., Langridge, J., and Florence, A. T., J. Pharm. Pharmacol. 41, 809 (1989); 42, 821 (1990). 9. Munshi, N., Chakravarty, K., De, T. K., and Maitra, A. N., Colloid Polym. Sci. 273, 464 (1995). 10. Munshi, N., De, T. K., and Maitra, A. N., J. Colloid Interface Sci. 180, 387 (1997). 11. Maitra, A. N., Ghosh, P. K., De, T. K., and Sahoo, S., A Method for the Preparation of Drug Loaded Hydrophilic Polymeric Nanoparticles. U.S. Patent Application No. 08/779,904, dated 7.1.97; Sahoo, S., Gaur, U., De, T. K., Ghosh, P. C., Ghosh, P. K., and Maitra, A., RES Evading and Long Circulating Hydrophilic Nanoparticles, communicated, 1998. 12. Kreuter, J., in “Topics in Pharmaceutical Sciences” (D. D. Breimer and P. Speiser, Eds.), pp. 359 –370. Elsevier, Amsterdam, 1983. 13. Kost, J., and Langer, R., in “Hydrogels in Medicine and Pharmacy” (N. A. Peppas, Ed.), Vol. 3, pp. 95–108. CRC, Boca Raton, FL, 1986. 14. Pitt, C. G., Int. J. Pharm. 59, 173 (1990). 15. Heller, J., Chang, A. C., Rodd, G., and Grodsky, G. M., J. Controlled Release 13, 295 (1990). 16. Siegal, R. A., and Firstone, B. N. A., Macromolecules 21, 3254 (1988). 17. Torchilin, V. P., Tschenko, E. G., Sminov, V. N., and Chazov, E. D., J. Biomed. Mater. Res. 11, 223 (1997). 18. Heller, J., Helwing, R. F., Baker, R. W., and Tuttle, M. E., Biomaterials 4, 262 (1983). 19. Lindhardt, R. J., in “Controlled Release of Drugs: Polymers and Aggregated Systems” (M. Rosoff, Ed.), p. 53. VCH, Weinheim/New York, 1988. 20. Hoffman, A. S., Chen, G. H., Kaang, S. Y., Ding, Z. L., Randeri, K., and Kabra, B., Adv. Biomater. Biomed. Engng. Drug Delivery Syst. [Iketani Conf. Biomed. Polym., 5th 1995], 62, (1996). 21. Ravichandran, P., Shantha, K. L., and Panduranga Rao, K., Int. J. Pharm. 154, 89 (1997). 22. Leong, K. W., Brott, B. C., and Langer, R., J. Biomed. Mater. Res. 2, 167 (1985). 23. Ron, E., Turek, T., and Mathiowitz, E., Proc. Natl. Acad. Sci. USA 90, 4176 (1993). 24. Vazquez, B., Gurruchaga, M., Goni, I., Narvarte, E., and San Roman, J., Polymer 36, 3327 (1995).