Drug–polyionic block copolymer interactions for micelle formation: physicochemical characterisation

Journal of Controlled Release 75 (2001) 249–258 www.elsevier.com / locate / jconrel

Drug–polyionic block copolymer interactions for micelle formation: physicochemical characterisation a,b a, c c Thirumala Govender , Snjezana Stolnik *, Chengdong Xiong , Sheng Zhang , Lisbeth Illum a , Stanley S. Davis a a

School of Pharmacy and Pharmacology, University of Durban-Westville, Private Bag, X54001 Durban, South Africa b School of Pharmaceutical Sciences, University Park, University of Nottingham, Nottingham NG7 2 RD, UK c Chengdu Institute of Organic Chemistry, Chinese Academy of Sciences, Chengdu 610041, PR China Received 23 August 2000; accepted 4 April 2001

Abstract While covalent attachment of small drug molecules to AB copolymers for the formation of polymeric micelles for drug delivery has been investigated, few studies have focused on non-covalent interactions. The aim of this study was therefore to explore the potential of non-covalent interactions between an AB copolymer, Poly(aspartic acid)–poly(ethylene glycol) (Pasp–PEG), with anionic pendant groups and diminazene aceturate, a small molecular weight cationic drug. Micelles were prepared by mixing solutions of Pasp–PEG and diminazene in 25 mM Tris–HCl buffer. At all Pasp–PEG concentrations studied, the micelles appeared to be water soluble with a unimodal size distribution and ranged in size from approximately 22 to 60 nm. The polyionic micelles also displayed similar and small absolute zeta potential values at various drug:monomer molar ratios which confirmed stabilisation by the PEG corona. The scattering intensity was maximal and remained unchanged, while particle size increased slightly at pH range from 3.4 to 7.2. At this pH range both the polymer and drug would be ionised and ionic interactions possible to drive micellar formation. An increase in size and scattering intensity with addition of NaCl to the micelles was attributed to dehydration of the PEG corona which may have led to aggregation of the micelles. The absence of micellar dissociation upon addition of salt was attributed to the dominance of hydrogen bonding between Pasp and diminazene aceturate, as assessed by isothermal titration microcalorimetry. Morphological evaluation of these constructs showed them to be discrete and fairly uniform in size and shape. This study was therefore successful in confirming the potential of non-covalent interactions using an AB copolymer to form polyionic micelles for drug delivery.  2001 Elsevier Science B.V. All rights reserved. Keywords: Drug–polymer interactions; Polyionic micelles; Nanoparticles; Polyelectrolyte

1. Introduction

*Corresponding author. Tel.: 144-115-846-6074; fax: 144115-951-5102. E-mail address: [email protected] (S. Stolnik).

Recently, the use of AB copolymers to form polymeric micelles for drug delivery optimisation has received considerable attention [1–3]. In an aqueous solvent, the hydrophobic (A) blocks of the copolymer form the core whilst the hydrophilic (B)

0168-3659 / 01 / $ – see front matter  2001 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 01 )00353-4

250

T. Govender et al. / Journal of Controlled Release 75 (2001) 249 – 258

blocks form the shell or corona. The core-shell structure of these self-assembling systems presents many advantages. The hydrophobic core can be used for solubilisation of hydrophobic drug molecules [1,4] and also further to prevent drug molecules against possible in vivo degradation [5]. The hydrophilic (e.g., polyethylene glycol) shell should suppress opsonisation and in that way prolongs circulation times and influences the pharmacokinetics and biodistribution of the drug delivery system [6,7]. Since the polyethylene glycol (PEG) stabilising chains are chemically grafted to the core surface, the possibility of desorption or displacement by serum components is eliminated. The small size of copolymer micelles is also advantageous in that it can promote escape through the permeable vasculature at tumour sites or areas of inflammation [3,8]. Kataoka and co-workers have reported the selfassembly of block copolymers based on non-covalent interactions for proteins [5] and poly(amino acid) [9]. Recently, Thunemann et al. [10] prepared micelles from poly(lysine) and all-trans-retinoic acid via ion pairing. While hydrophobic drugs such as indomethacin [11,12]; adriamycin [13–15] and amphotericin B [4] have been mainly incorporated into AB block copolymer micelles by chemical conjugation and / or physical entrapment; few studies have focused on the self assembly of copolymers using non-covalent interactions such as ionic forces with less hydrophobic drugs. This approach is advantageous since the drug would be easily released from the micelles while with covalent attachment, cleavage of the drug from the polymer would be required. Also, since with non-covalent interactions mixtures of drug and polymer are simply mixed in a suitable solvent, the synthetic procedure of attaching the drug to the polymer is obviated. The aim of this study was to explore the potential of non-covalent drug–polymer interactions in forming micelles from an anionic polymer, poly(aspartic acid)–poly(ethylene glycol) and a cationic drug, diminazene aceturate. Physicochemical characterisation data based on size, surface charge and morphology of these micellar type systems are also identified. Diminazene aceturate (an antiprotozoal / antibacterial drug) was selected as a model drug since we have confirmed and thermodynamically characterised in previous calorimetric studies [16] its

interaction between poly(aspartic) acid. Also, diminazene is water soluble, cationic, inexpensive and readily available. Pasp–PEG was chosen as the functionalised copolymer of Pasp, which was chosen as the poly(amino acid) since it is water soluble, non-toxic and biodegradable [8,17] and also because of its confirmed interaction with diminazene aceturate [16]. The starting hypothesis was that micelle formation between Pasp–PEG and diminazene aceturate would occur as follows: When these two water soluble species are mixed in an aqueous solution, where both would be ionised, the cationic drug should interact ionically with the pendant carboxylic acid groups of Pasp making that portion of the copolymer hydrophobic. Since PEG is hydrophilic, this will lead to the drug / polymer complex self-assembling to form micellar type constructs. The interaction of the drug with pendant groups in the polymer may lead to an improved drug incorporation efficiency especially for relatively less hydrophobic, water soluble drugs. This is important since nanosystems generally suffer from poor drug incorporation efficiencies, particularly with water soluble drugs [18–19] due to their small size and loss of drug into the aqueous phase.

2. Materials and methods

2.1. Materials Poly(aspartic acid)–poly(ethylene glycol) (Pasp– PEG) was synthesised at the Chengdu Institute of Organic Chemistry (China). The molecular weight of Pasp was 6000 Da and that of the PEG block, 5000 Da. Diminazene aceturate was purchased from Sigma Chemical Co. (St Louis, MO, USA). Trizma  hydrochloride (Tris–HCl) and HEPES (as sodium salt) were also purchased from Sigma. All other chemicals used were of pharmaceutical grade.

2.2. Methods 2.2.1. Synthesis of Pasp–PEG polymer a-amino-v-methoxypoly(ethylene glycol) 5000: The tosyl ester of PEG was prepared as described previously [20]. 1 H-NMR of a-tosyl-v-methoxypoly(ethylene glycol) 5000: d 52.45 (CH 3 — of the

T. Govender et al. / Journal of Controlled Release 75 (2001) 249 – 258

tosyl ester), d 53.38 (CH 3 –O2), d 53.66 (CH 2 – CH 2 –O — of the tosyl ester), d 54.18 (CH 2 –CH 2 –OTs), d 57.32 and 7.80 (aromatic protons). The conversion of the tosylate into an amine was carried out by Gabriel synthesis according to reference method [21]. 1 H-NMR of a-amino-v-methoxypoly(ethylene glycol) 5000: d 52.78 (–CH 2 –CH 2 –NH 2 ), d 53.38 (CH 3 –O2), d 53.66 (CH 2 –CH 2 –O2). g-benzyl aspartic N-carboxyanhydride (BA-NCA): g-benzyl aspartic N-carboxyanhydrides were prepared using the classical method [22]. Poly(benzylaspartic acid)–poly(ethylene glycol) (PBAA–PEG): v-Amino PEG was used as an initiator for the polymerisation of BA-NCA. This macroinitiator and BA-NCA were added into a dried 100 ml glass reactor which was previously nitrogenpurged for several times. The solvent (1,4-dioxane / CHCl 3 53 / 2, v / v) was injected with a syringe. After 72 h reaction time at 258C, the copolymer was recovered by precipitation in methanol. IR (KBr): 1737 (vs; OC5O), 1665 (vs; NHC5O), 1550 (d ; CO–NH), 1115 (vs; C–O–C), 699 and 749 (v; C6H6 ) Deprotection of the benzyl group of poly(benzylaspartic acid)–poly(ethylene glycol): PBAA– PEG (10%w / v) was deprotected in CHCl 3 in the presence of excess NaOH. The reaction was kept for 5 h at room temperature. Following filtration to remove the NaOH, the copolymer was recovered by precipitation in diethyl ether. IR (KBr): 1665 (vs; NHC5O), 1550 (d ; CO–NH), 1115 (vs; C–O–C).

2.2.2. Micelle preparation using Pasp–PEG and diminazene All drug and polymeric solutions were prepared in 25 mM Tris–HCl buffer and adjusted to pH 5.3 before mixing. 2.2.3. Influence of Pasp–PEG concentration on micelle formation This study was performed with 5 mg, 12.5 mg and 25 mg Pasp–PEG and specified amounts of diminazene aceturate to obtain varying drug:monomer molar ratios. Pasp–PEG solution (4 ml) containing 5 mg, 12.5 mg or 25 mg Pasp–PEG was transferred to a scintillation vial. Specified volumes of diminazene aceturate solution were added to the polymeric

251

solution to obtain drug:monomer molar ratios ranging from (1 / 2) 0.12:1 to 3.6:1 and then magnetically stirred for 10 min. Prior to drug addition, specified volumes of the buffer solution were added to the polymer solution in the vials such that the final total volume would be 8 ml for all samples, thus obviating any dilution effects. The scattering intensity and size of these micellar type samples were determined by Photon Correlation Spectroscopy as described later. For each sample, the mean of five determinations was calculated for the scattering intensity measurements while for the micelle size analysis the mean6S.D. of six determinations was measured.

2.2.4. Effect of pH on micelle formation Pasp–PEG solution (4 ml) containing 5 mg Pasp– PEG was transferred to a scintillation vial. Diminazene aceturate solution (1.4 ml) containing 21.06 mg drug was added to the polymeric solution (to obtain a drug:monomer molar ratio of 0.84:1) and magnetically stirred for 10 min. Prior to drug addition, 2.6 ml of buffer was added to the polymeric solution to obtain a total volume of 8 ml for the final sample. The sample was then adjusted to various pH values ranging from 1.6 to 11.02 using either dilute HCl or NaOH. After each pH adjustment, the scattering intensity and size of an aliquot of the sample was determined by PCS. For each sample, the mean of five determinations was calculated for the scattering intensity measurements while for the micelle size analysis the mean6S.D. of six determinations was measured. A blank experiment was also performed by adjusting the pH of an 8 ml sample containing 5 mg Pasp–PEG only to similar pH values and then determining the scattering intensity of the aliquots as previously. 2.2.5. Effect of salt ( NaCl) concentration on micelle formation Pasp–PEG solution (4 ml) containing Pasp–PEG (5 mg) was transferred to a scintillation vial. Diminazene aceturate solution (1.4 ml) containing 21.06 mg drug was added to the polymeric solution (to obtain a mole drug:monomer ratio of 0.84:1) and magnetically stirred for 10 min. Specified volumes of a NaCl stock solution (14.39 g / 100 ml), also prepared in 25 mM Tris–HCl buffer, were added to the drug / polymeric samples to obtain salt concentrations

252

T. Govender et al. / Journal of Controlled Release 75 (2001) 249 – 258

ranging from 0.025 M to 0.8 M, stirred for 5 min and thereafter the scattering intensity and size analysis were determined by PCS. A separate sample for analysis was prepared for each salt concentration. For each sample, the mean of five determinations was calculated for the scattering intensity measurements, while for the micelle size analysis the mean6S.D. of six determinations were measured. Prior to addition of the NaCl solution, specified volumes of the buffer were added to the drug / polymeric mixture such that the total volume of all the final samples would be 8 ml thus obviating any dilution effects. A blank experiment was also performed by adding specified quantities of NaCl to a 8 ml solution containing 5 mg Pasp–PEG only to obtain similar salt concentrations as the test and then determining the scattering intensity of the aliquots by PCS as previously.

Anemometry (Malvern Zetasizer IV, Malvern Instruments Ltd, Malvern, UK). All analyses were performed on samples appropriately diluted with 1 mM HEPES buffer (adjusted to pH 7.4 with 1 M HCl) in order to maintain a constant ionic strength. For each sample the mean value6S.D. of four determinations were established.

2.2.7.3. Micelle morphology. Samples containing 5 mg Pasp–PEG and 21.06 mg of diminazene aceturate (mole drug:monomer ratio50.84:1) were prepared as described above and morphological evaluation performed using transmission electron microscopy (TEM) (Jeol JEM 1010 Electron Microscope, Japan) following positive staining with uranyl acetate.

3. Results and discussion

2.2.6. Effect of drug: monomer molar ratios on the zeta potential of micelles Pasp–PEG solution (4 ml) containing Pasp–PEG (5 mg) was transferred to a scintillation vial. Specified volumes of diminazene aceturate solution were added to the polymeric solution to obtain drug:monomer molar ratios ranging from 0.12:1 to 2.4:1 and then magnetically stirred for 10 min. Prior to drug addition, specified volumes of the buffer solution were added to the polymer solution in the vials such that the final total volume would be 8 ml for all samples; thus obviating any dilution effects. The zeta potential of the samples was determined by Laser Doppler Anemometry as described later. For each sample the mean6S.D. of four determinations was measured.

3.1. Influence of Pasp–PEG concentration on micelle formation The scattering intensity trends measured by PCS at various drug:monomer molar ratios for the various polymer concentrations studied are shown in Fig. 1. For each of the experiments with the various Pasp– PEG concentrations the scattering intensity was very low initially with no drug in the Pasp–PEG solution i.e., mole drug:monomer ratio50:1. This was due to the polymer being soluble in the aqueous phase

2.2.7. Physico-chemical characterisation 2.2.7.1. Micelle size. Micelle size was determined using Photon Correlation Spectroscopy (Malvern S4700 PCS System, Malvern Instruments Ltd, Malvern, UK). The analysis was performed at a laser wavelength of 488 nm, scattering angle of 908 and at a temperature of 258C. For each sample, the mean diameter6S.D. of six determinations were calculated applying multimodal analysis. 2.2.7.2. Zeta potential. The zeta potential of the micelles was determined by Laser Doppler

Fig. 1. Effect of Pasp–PEG concentration on the scattering intensity of micelles prepared from diminazene aceturate and Pasp–PEG in Tris–HCl buffer pH 5.3.

T. Govender et al. / Journal of Controlled Release 75 (2001) 249 – 258

therefore showing minimal scattering intensity values. However, with addition of diminazene aceturate the scattering intensity increased significantly. This may confirm the concept of the formation of micelles with addition of diminazene aceturate to the polymeric solution. Also, as the concentration of the polymer increased from 5 to 25 mg, the scattering intensity increased at equivalent charge ratios. This may have been due to an increasing number and / or possible aggregation of micelles with an increasing concentration of polymer. Although the scattering intensity increased at equivalent mole drug:monomer ratios for increasing polymer concentrations, the samples remained visually transparent. This was in contrast to the obvious visual appearance of complexes / precipitates formed from Pasp and diminazene aceturate. This apparent transparency of the diminazene aceturate / Pasp–PEG system may be due to their small size as well as the formation of micelles that are water soluble. The micelle size and polydispersity data at various mole drug:monomer ratios for each of the polymer concentrations studied are shown in Table 1. For all three polymer concentrations studied, the micelle size varied from approximately 22 nm to 60 nm. This size range is consistent with that reported in the literature for micelles formed from block copolymers using drug conjugation or physical entrapment [3,8,13]. For each of the polymeric concentrations, micelle

253

Fig. 2. Size distribution of micelles prepared from Pasp–PEG (5 mg) and diminazene aceturate (21.06 mg) in Tris–HCl buffer pH 5.3 (drug:monomer molar ratio50.84:1).

size increased with an increase in the drug:monomer molar ratio (Table 1). These results may be due to an increased number of Pasp portions of the polymer being neutralised by the cationic drug. This may lead to stronger hydrophobic forces and hence an increased aggregation number that may have contributed to the slight increase in micelle sizes observed. Fig. 2 further shows that although the polydispersity values are higher than desired, the micelles showed a unimodal size distribution. Also, at equivalent drug:monomer molar ratios, the micelles size of preparations with the 12 and 25 mg Pasp–PEG appear to be slightly larger than the preparation with 5 mg Pasp–PEG. At the higher polymeric concentrations, an increased number of unimers in the preparation may have led to an

Table 1 Influence of Pasp–PEG concentration on the size of micelles prepared from diminazene aceturate and Pasp–PEG Ratio* (1 / 2);

Particle size6S.D. (nm) (Polydispersity)

x:1

Pasp–PEG (5 mg)

Pasp–PEG (12.5 mg)

Pasp–PEG (25 mg)

0.06 0.12 0.15 0.18 0.36 0.60 0.84 1.08 1.20 2.40 3.60 4.80

25.060.7 22.060.4 23.460.3 26.760.6 31.760.3 35.261.2 35.360.7 33.961.2 34.861.0 36.861.2 43.461.1 49.162.3

36.761.2 (0.19060.020) 34.061.1 (0.16360.038) 36.562.6 (0.16360.008) 35.060.9 (0.13060.005) 41.563.6 (0.16260.009) 42.361.3 (0.12460.018) 41.360.6 (0.16860.009) 46.561.5 (0.18860.004) 47.962.6 (0.20960.014) 52.261.4 (0.23160.010) 53.860.7 (0.22660.017) 52.760.8 (0.25660.013)

27.460.4 (0.37260.010) 27.560.7 (0.34460.027) 27.560.8 (0.35960.020) 31.560.5 (0.28560.018) 41.260.3 (0.25860.011) 41.460.4 (0.27460.002) 41.960.4 (0.26460.009) 46.460.5 (0.28160.008) 46.560.7 (0.30060.034) 56.360.4 (0.28860.015) 58.560.4 (0.28960.010) 59.260.6 (0.31660.013)

*Drug:monomer molar ratio.

(0.42960.040) (0.38560.050) (0.32960.025) (0.28660.025) (0.26260.050) (0.24660.039) (0.19760.018) (0.21860.030) (0.21360.055) (0.31460.074) (0.34460.025) (0.39660.038)

254

T. Govender et al. / Journal of Controlled Release 75 (2001) 249 – 258

increased aggregation number and hence an increased micelle size. Future static light scattering studies are required to support this postulation.

3.2. Effect of pH on micelle formation The pKa of the carboxylic acid group is about 4 while diminazene aceturate is a weakly basic drug. Therefore, micelles were prepared in Tris–HCl buffer pH 5.3, since at this pH both species would be ionised and therefore non-covalent ionic interactions possible. Since pH will influence the ionisation of diminazene aceturate and Pasp, it was therefore necessary to determine its influence on the micelles. This was performed by altering the pH of the micelle solutions and then determining the micelle size and scattering intensity. As shown in Fig. 3, the scattering intensity was maximal and remained virtually unchanged in the pH range 3.4 to 7.2. These pH values correspond to those at which both Pasp and diminazene aceturate would be ionised, therefore providing the opportunity for ionic interactions to drive micelle formation. At pH values less than 3.4, the scattering intensity decreased since although diminazene aceturate was fully ionised, the degree of ionisation of Pasp decreased and therefore micelle formation may be decreased. At pH values greater than 7.2, the scattering intensity also decreased. Again, this may be due to the fact that although Pasp was fully ionised, in this instance the degree of ionisation of diminazene aceturate was decreased and therefore micelle formation was decreased. The

Fig. 3. Effect of pH on micelles prepared from Pasp–PEG (5 mg) and diminazene aceturate (21.06 mg) in Tris–HCl buffer pH 5.3 (drug:monomer molar ratio50.84:1).

blank experiment with the consistently low and unaffected scattering intensity values shown in Fig. 3 confirmed the absence of micelle formation when diminazene aceturate was not included in the polymer solution. No pronounced effect on micelle size was observed in the pH range 2.16 and 8.51 (Table 2). On either side of this pH range, micelle size could not be measured due to insufficient scattering counts.

3.3. Effect of salt ( NaCl) concentration on micelle formation It is known that the stability of polyionic complexes is strongly influenced by the ionic strength of the medium: being destabilised with an increase in ionic strength due to electrostatic shielding [9]. The influence of salt concentration on the micelles formed from diminazene aceturate and Pasp–PEG was there-

Table 2 Effect of pH on the size of micelles prepared from diminazene aceturate and Pasp–PEG pH

Particle size6S.D. (nm) (Polydispersity)

1.60 1.80 2.03 2.16 2.33 2.52 2.81 3.02 3.44 4.11 5.01 6.01 7.21 7.40 7.60 7.80 8.05 8.21 8.51 8.80 9.00 9.20 9.78 10.56 11.02

* * * 36.960.9 (0.13560.051) 37.460.8 (0.14060.049) 37.260.7 (0.16160.055) 36.660.7 (0.24160.027) 36.460.7 (0.28260.028) 37.861.7 (0.23360.062) 43.261.7 (0.23360.059) 41.260.5 (0.22560.190) 42.562.1 (0.25960.040) 47.761.3 (0.29060.046) 43.261.2 (0.24560.012) 40.060.8 (0.23960.048) 41.260.3 (0.19960.014) 43.862.9 (0.32060.090) 42.561.1 (0.27760.058) 42.761.3 (0.26560.056) * * * * * *

* Insufficient scattering intensity counts for size measurement.

T. Govender et al. / Journal of Controlled Release 75 (2001) 249 – 258

Fig. 4. Effect of salt (NaCl) concentration on micelles prepared from Pasp–PEG (5 mg) and diminazene aceturate (21.06 mg) in Tris–HCl buffer pH 5.3 (drug:monomer molar ratio50.84:1).

fore determined by measuring the scattering intensity and size of the micelle solutions after the addition of NaCl at various concentrations. The results in Fig. 4 and Table 3 clearly show that the scattering intensity and micelle size, respectively, were influenced by the salt concentration. As the NaCl concentration was increased, both micelle size and scattering intensity also increased. An increase in NaCl concentration may have induced dehydration of the PEG corona [9] thereby leading to a decrease in micelle steric repulsion. This may have therefore led to aggregation of the micelles thus explaining the increase in scattering intensity as shown in Fig. 4. The decreased steric repulsion may also be responsible for an increase in the micelle diameter due to the induced Table 3 Effect of salt (NaCl) concentration on the size of micelles prepared from diminazene aceturate and Pasp–PEG Salt (NaCl) concentration (M)

Particle size6S.D. (nm) (Polydispersity)

0.00 0.025 0.50 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80

31.760.3 42.261.9 46.561.3 48.260.8 56.060.2 61.260.2 72.660.9 96.461.9 107.161.8 142.761.9 146.161.1

(0.26260.050) (0.23360.016) (0.22760.009) (0.21260.003) (0.22060.006) (0.23060.012) (0.24660.012) (0.27360.020) (0.30960.064) (0.36560.027) (0.39560.013)

255

micelle aggregation and or rearrangement of the micelle structure. The blank experiment, with the consistently low and unaffected scattering intensity values shown in Fig. 4, confirmed the absence of micelle formation when diminazene aceturate was not included in the polymer solution. It should be noted that a decrease in scattering intensity with an increase in NaCl concentration, due to disassociation of the micelles caused by electrostatic shielding, was expected. Hence, the results in this study are contrary to what was expected. It seems that an increase in salt concentration does not lead to dissociation of the micelles. A possible reason for this binding may be based on our results reported previously [16]. Isothermal titration calorimetric analysis on the interaction of diminazene aceturate and Pasp showed that while ionisation of the interacting species and electrostatic interactions were essential for initiating complex formation, it was the non-electrostatic interactions particularly hydrogen bonding which was dominant in stabilising the complex. Hence, with the addition of salt to the micelle preparation, the non-electrostatic interactions continued to predominate in that way preventing the dissociation of micelles.

3.4. Zeta potential determination LDA studies were performed in order to determine the zeta potentials of the micelles formed. The results are shown in Table 4. The zeta potentials of the micelles with the various drug:monomer molar ratios had a small absolute value. These results are consistent with the formation of a small core-shell structure where the PEG corona prevents micelle aggregation through steric repulsion. Harada and Table 4 Zeta potential determination of micelles prepared with diminazene aceturate and Pasp–PEG copolymer Drug:monomer molar ratio (1 / 2; x:1)

Zeta potential6S.D. (mV)

0.12 0.48 0.84 1.08 1.20 2.40

23.660.4 20.660.1 2.660.1 1.860.3 1.560.9 2.960.3

256

T. Govender et al. / Journal of Controlled Release 75 (2001) 249 – 258

Kataoka [9] also reported small absolute zeta potential values for polyionic complex micelles formed from a poly( L-lysine) and poly(ethylene glycol)– poly(aspartic acid) block copolymer. The smallest absolute zeta potential value of 20.660.1 mV, which indicates that at this point the complex is almost neutral, is at a molar ratio of 0.48:1. This actually corresponds to a molar charge ratio of polymer to drug of approximately 2:1. These results correlate with previous calorimetric studies where the z-value, which according to Manning’s theory represents the number of electrostatic interactions of a drug molecule with a polymer, was calculated to be 2 for the interaction of Pasp with diminazene [16].

3.5. Morphology of micelles TEM using positive staining was employed to successfully confirm the morphology of this micellar type system. Fig. 5 showed them to be discrete units.

The absence of agglomerates suggested good stabilisation by PEG. They also appeared to be fairly similar in shape with a uniform size.

4. Conclusions The use of a PEG block copolymer and noncovalent interactions with a small cationic drug to drive micelle formation was investigated in the work described in this study. Pasp–PEG was chosen as the model block copolymer and diminazene aceturate as the model low molecular weight cationic drug. Micelles prepared at varying polymer concentrations appeared to be water soluble, showed a unimodal size distribution and ranged in size from approximately 22 to 60 nm for all polymer concentrations studied. The micelles displayed similar small absolute zeta potential values at various drug:monomer molar ratios which suggested stabilisation by the

Fig. 5. Morphology of micelles prepared from Pasp–PEG and diminazene aceturate in Tris–HCl buffer pH 5.3 (drug:monomer molar ratio50.84:1).

T. Govender et al. / Journal of Controlled Release 75 (2001) 249 – 258

PEG corona. No pronounced effect on the scattering intensity and micelle size was observed when the micelles where exposed to pH values ranging from 3.4 to 7.2. This corresponded to pH values where both species would be ionised and ionic interactions possible, in that way driving micelle formation. The addition of NaCl to the micelles led to an increase in size and scattering intensity. This was attributed to dehydration of the PEG corona which may have led to aggregation of the micelles. Also, the absence of micelle dissociation upon addition of salt was attributed to the dominance of non-electrostatic interactions between Pasp and diminazene aceturate which stabilise the system. It can therefore be concluded from the pH and salt studies in this paper as well as our previous results [14] that electrostatic interactions are essential to initiate interaction between drug and polymer since micelles were not observed at pH values where either of the interacting species were unionised. However, while electrostatic interactions are essential it is the non-electrostatic interactions which dominate and stabilise the complex. TEM studies confirmed the presence of these constructs as being discrete and fairly uniform in size and shape. This study has served to confirm the potential of non-covalent drug–polymer interactions in driving micellar formation for drug delivery and has also elucidated their preliminary physico-chemical characterisation. Clearly, a more detailed physicochemical characterisation of this drug delivery system is required for optimising drug incorporation efficiency and delivery.

Acknowledgements T. Govender is grateful to the Association of Commonwealth Universities for financial support. This work is also funded by AstraZeneca R&D, Charnwood. S. Stolnik is the AstraZeneca lecturer in Drug Delivery at the University of Nottingham. The authors would like to thank AstraZeneca R&D, Charnwood, for support, Dr Anne Brindley (AstraZeneca R&D, Charnwood) for her contribution to the work and Trevor Gray (Department of Histopathology, Queens Medical Centre) for assistance with TEM.

257

References [1] G.S. Kwon, M. Naito, K. Kataoka, M. Yokoyama, Y. Sakurai, T. Okano, Block copolymers micelles as vehicles for hydrophobic drugs, Colloids Surf. 2 (1994) 429–434. [2] G.S. Kwon, K. Kataoka, Block copolymer micelles as long circulating drug vehicles, Adv. Drug Deliv. Rev. 16 (1995) 295–309. [3] M. Yokoyama, Novel passive targetable drug delivery with polymeric micelles, in: T. Okano (Ed.), Biorelated Polymers and Gels: Controlled Release and Applications in Biomedical Engineering, Academic Press, Boston, 1998, pp. 193–229. [4] B.G. Yu, T. Okano, K. Kataoka, G.S. Kwon, Polymeric micelles for drug delivery: solubilisation and haemolytic activity of Amphotericin B, J. Controlled Release 53 (1998) 131–136. [5] A. Harada, K. Kataoka, Novel polyion complex micelles entrapping enzyme molecules in the core: preparation of narrowly distributed micelles from lysozyme and poly(ethylene glycol)–poly(aspartic acid) block copolymer in aqueous medium, Macromolecules 31 (1998) 288–294. [6] S.E. Dunn, A. Brindley, S.S. Davis, M.C. Davies, L. Illum, Polystyrene–poly(ethylene glycol)(PS–PEG 2000) particles as model systems for site specific drug delivery. 2. The effect of PEG surface density on the in vitro cell interaction and in vivo biodistribution, Pharm. Res. 11 (1994) 1016–1022. [7] S. Stolnik, L. Illum, S.S. Davis, Long circulating microparticulate drug carriers, Adv. Drug Deliv. Rev. 16 (1995) 195–214. [8] K. Kataoka, Design of nanoscopic vehicles for drug targeting based on micellisation of amphiphilic block copolymers, Pure Appl. Chem. A31 (1994) 1759–1769. [9] A. Harada, K. Kataoka, Formation of stable and monodispersive polyion complex micelles in aqueous medium from poly( L-lysine) and poly(ethylene glycol)–poly(aspartic acid) block copolymer, Pure Appl. Chem. A34 (1997) 2119–2133. [10] A. F Thunemann, J. Beyermann, H. Kukula, Poly(ethylene oxide)-b-poly( L-lysine) complexes with retinoic acid, Macromolecules 1 (2000) 5906–5911. [11] S.B. La, T. Okano, K. Kataoka, Preparation and characterisation of the micelle-forming polymeric drug indomethacinincorporated poly(ethylene oxide)–poly(b-benzyl L-aspartate) block copolymer micelles, J. Pharm. Sci. 85 (1996) 85–90. [12] I.G. Shin, S.Y. Kim, Y.M. Lee, C.S. Cho, Y.K. Sung, Methoxy poly(ethylene glycol) / e-caprolactone amphiphilic block copolymeric micelle containing indomethacin. I. Preparation and characterisation, J. Controlled Release 51 (1998) 1–11. [13] G.S. Kwon, M. Naito, M. Yokoyama, T. Okano, Y. Sakurai, K. Kataoka, Block copolymer micelles for drug delivery: loading and release of doxorubicin, J. Controlled Release 489 (1997) 195–201. [14] G.S. Kwon, M. Naito, M. Yokoyama, T. Okano, Y. Sakurai, K. Kataoka, Physical entrapment of adriamycin in AB block copolymer micelles, Pharm. Res. 12 (1995) 192–195. [15] M. Yokoyama, S. Fukushima, R. Uehara, K. Okamoto, K. Kataoka, Y. Sakurai, T. Okano, Characterisation of physical

258

T. Govender et al. / Journal of Controlled Release 75 (2001) 249 – 258

entrapment and chemical conjugation of adriamycin in polymeric micelles and their design for in vivo delivery to a solid tumour, J. Controlled Release 50 (1998) 79–92. [16] T. Ehtezazi, T. Govender, S. Stolnik, Hydrogen bonding and electrostatic interaction contributions to the interaction of a cationic drug with polyaspartic acid, Pharm. Res. 17 (2000) 871–878. [17] S. Roweton, S. J Huang, G. Swift, Poly(aspartic acid): synthesis, biodegradation and current applications, J. Environment. Polymer Degrad. 5 (1997) 175–181. [18] T. Niwa, H. Takeuchi, T. Hino, N. Kunou, Y. Kawashima, Preparations of biodegradable nanospheres of water-soluble and insoluble drugs with DL-lactide / glycolide copolymer by a novel spontaneous emulsification solvent diffusion method, and the drug release behaviour, J. Controlled Release 25 (1993) 89–98.

[19] T. Govender, S. Stolnik, M.C. Garnett, L. Illum, L.S.S. Davis, PLGA nanoparticles prepared by nanoprecipitation: drug loading and release studies of a water soluble drug, J. Controlled Release 57 (1999) 171–185. [20] K. Kugo, T. Uno, H. Yamano, J. Nishino, H. Masuda, Kobunshi Ronbunshu 42 (1985) 731. [21] K. Kugo, A. Ohji, T. Uno, J. Nishino, Synthesis and conformations of A–B–A tri-block copolymers with hydrophobic poly(benzyl L-glutamate) and hydrophilic poly(ethylene oxide), Polymer J. 19 (1987) 375–381. [22] P. Lloyd-Williams, F. Albericio and E. Giralt, Chemical approaches to the synthesis of peptides and proteins, CRC Press, Boca Raton, 1997.