Claw Amphiphiles with a Dendrimer Core - American Chemical Society

Mar 8, 2013 - Claw amphiphiles were prepared by attaching one end of comb- ... The linear chitosan amphiphile (GCPQA) forms the digits of the claw...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/Langmuir

Claw Amphiphiles with a Dendrimer Core: Nanoparticle Stability and Drug Encapsulation Are Directly Proportional to the Number of Digits Kar Wai Chooi,† Xue Liang Hou,† Xiaozhong Qu,† Ramesh Soundararajan,† and Ijeoma F. Uchegbu*,†,‡ †

School of Pharmacy, University College London, 29 − 39 Brunswick Square, London WC1N 1AX, U.K. Nanomerics, Approach Road, St. Albans AL1 1SR, U.K.

‡

S Supporting Information *

ABSTRACT: There are numerous pharmaceutical, food, and consumer product applications requiring the incorporation of hydrophobic solutes within aqueous media. Often amphiphiles and/or polymers are used to produce encapsulating nanostructures. Because the encapsulation efficiencies of these nanostructures directly impact on the process or product, it is often desirable to optimize this parameter. To produce these advanced functional materials, we hypothesized that an amphiphile with a claw shape would favor polymer aggregation into nanoparticles and hydrophobic compound encapsulation. Claw amphiphiles were prepared by attaching one end of combshaped chitosan amphiphile chains [N,N,N-trimethyl, N,N-dimethyl, N-monomethyl, N-palmitoyl, N-acetyl, 6-O-glycol chitosan (GCPQA)] to a central dendrimer core [generation 3 diaminobutane poly(propylenimine) dendrimer (DAB)] to give DABGCPQA. The linear chitosan amphiphile (GCPQA) forms the digits of the claw. These claw amphiphiles were very stable and had a high encapsulating efficiency. DAB-GCPQAs (Mn = 30 and 70 kDa) had extremely low critical micelle concentrations [CMCs = 0.43 μg mL−1 (13 nM) and 0.093 μg mL−1 (0.9 nM), respectively], and their CMCs were lower than that of linear GCPQA [Mn = 14 kDa, CMC = 0.77 μg mL−1 (38 nM)]. The claw amphiphile CMCs decreased linearly with the number of digits (r2 = 0.98), and drug encapsulation (hydrophobic drug propofol) in 4 mg mL−1 dispersions of the amphiphiles increased linearly (r2 = 0.94) with the number of digits. DAB-GCPQA70 (4 mg mL−1, 0.058 mM) encapsulated propofol (7.3 mg mL−1, 40 mM). Finally, despite their stability, claw amphiphile nanoparticles are able to release the encapsulated drug in vivo, as shown with the claw amphiphile−propofol formulations in a murine loss of righting reflex model.

■

region.1,9 We have also found that the molecular shape of dendrimer amphiphiles, in which low-molecular-weight dendrimers are covalently linked to fatty acid chains, affects their ability to form hydrophobic domains and in turn that the molecular shape of these dendrimer amphiphiles has a profound effect on their drug encapsulation efficiency.12 The best encapsulation efficiencies are obtained when the hydrophobic acyl group is attached to the dendrimer amphiphile via a hydrophilic spacer molecule because this creates flexibility in the molecule and enables it to form hydrophobic domains with the terminal acyl units.12 We have decided to explore this molecular flexibility observation further, and in the current work, we present a new type of amphiphile shapethe claw amphiphile in which linear comb-shaped polymers are covalently bound, at one end only, to a central dendrimer core (Figure 1). These amphiphiles form remarkably stable nanoparticle self-assemblies and are able to encapsulate hydrophobic solutes in a manner that is dependent on the

INTRODUCTION There are numerous pharmaceutical,1 food,2 and consumer product3 applications requiring the incorporation of hydrophobic solutes into aqueous media. This is usually accomplished by using self-assembling low-molecular-weight amphiphiles3−6 or self-assembling polymer amphiphiles.1,6,7 This dependence on surface-active agents has resulted in a thriving global surfactant industry, an industry that is predicted to generate $4.1 billion in revenue by 2018.8 Efficient amphiphiles are required for these industries (i.e. amphiphiles that create self-assemblies or multiphase systems that are stable on dilution and amphiphiles that form self-assemblies with a high encapsulation efficiency). We have produced linear polymeric amphiphiles, which are comb-shaped molecules comprising a hydrophilic polymer backbone and hydrophobic pendant groups,7,9,10 and found that these polymers self-assemble into micelles, bilayer vesicles, and dense nanoparticles, with the size and nature of the resulting nanoparticle finely controlled by the polymer’s chemistry.7,11 These comb-shaped amphiphiles by virtue of the high surface area occupied by the multiple acyl chains form extremely stable nanosized self-assemblies with critical micellar concentrations (CMCs) in the micromolar © 2013 American Chemical Society

Received: December 13, 2012 Revised: February 21, 2013 Published: March 8, 2013 4214

dx.doi.org/10.1021/la304909r | Langmuir 2013, 29, 4214−4224

Langmuir

Article

hydroxylamines, formic acid, or amides.14 Further confirmation of the aldehyde function is obtained via the hypsochromic shift in the wavelength of maximum absorbance of the purpald on addition to the aldehyde (Table 1). Nitrous acid-degraded GCPQA (d-GCPQA) was conjugated to DAB 16 via the formation of a Schiff’s base and the reduction of the resulting Schiff’s base (Scheme 1). Reduction was carried out with sodium cyanoborohydride (NaCNBH3). NaCNBH3 selectively reduces the imino bonds at a pH above 4.5.16 However, the pH of the reaction, although kept above 4.5, was kept below 6.5 to avoid polymer precipitation.16 The synthesis of DAB-GCPQA was confirmed by gel permeation chromatography with multiangle laser light scattering (GPC-MALLS, Figure 2, Table 2) and 1H NMR (Supporting Information Figure 1d). The number of digits per claw polymer (γ) was calculated from the DAB-GCPQA molecular weight (Mn) data using eq 1 Figure 1. Self-assembly of claw-shaped amphiphiles. The claw shape enables multiple hydrophobic contacts and the formation of larger hydrophobic domains.

M n = MDAB + γMdigit

(1)

where Mn is the number-averaged molecular weight of DABGCPQA, MDAB is the molecular weight of DAB 16 (Mw = 1686.8 Da), and Mdigit is the number-averaged molecular weight of d-GCPQA. Four claw amphiphiles were synthesized (Table 2) with 6, 8, 9, and 14 d-GCPQA (Mn = 4 to 5 kDa) digits. With the more hydrophilic d-GCPQA5, a greater number of digits could be conjugated to DAB 16 (up to 14 digits, Table 2) presumably because the lower level of palmitoyl substitution on dGCPQA5 (Table 2) favored the reactivity of these d-GCPQA molecules in the aqueous reaction media. For the linear GCPQA polymers, the number of palmitoyl units per molecule (N) was estimated using the formulas given in eqs 2 and 3.

number of digits in the claw. Furthermore, the amphiphile selfassemblies release the encapsulated solutes in vivo as demonstrated using a murine anesthesia model. This is the first report of claw-shaped amphiphiles.

■

RESULTS AND DISCUSSION Amphiphile Synthesis. The synthesis of claw amphiphile N-graf t-(N,N,N-trimethyl, N,N-dimethyl, N-methyl, N-acetyl, N-palmitoyl, 6-O-glycol chitosan)-diaminobutane poly(propylenimine) dendrimer (DAB-GCPQA) is shown in Scheme 1. This five-step synthesis protocol involves (a) the depolymerization of 6-O-glycol chitosan (GC),11 (b) the synthesis of N,N,N-trimethyl, N,N-dimethyl, N-methyl, Npalmitoyl, 6-O-glycol chitosan (GCPQ),1 (c) the acetylation of GCPQ to form N,N,N-trimethyl, N,N-dimethyl, N-methyl, Nacetyl, N-palmitoyl, 6-O-glycol chitosan (GCPQA), and finally (d) the conjugation of degraded GCPQA to the generation 3 diaminobutane poly(propylenimine) dendrimer (DAB 16). NAcetylation reduces the likelihood of the terminal GC aldehyde functions conjugating to GC amino groups to form cross-linked products, so exhaustive acetylation was carried out to prevent the reaction of any chitosan primary amino groups with the terminal aldehyde function. GCPQ consists predominantly of the following monomers: 6-O-glycol-chitosan, N-palmitoyl-6O-glycol chitosan, N,N,N-trimethyl-6-O-glycol chitosan, N,Ndimethyl-6-O-glycol chitosan, and N-monomethyl-6-O-glycol chitosan. Although the N,N,N-trimethyl and N-palmitoyl monomer contents may be estimated from 1H NMR data, the levels of dimethyl and monomethyl glycol chitosan monomers cannot be independently estimated from 1H NMR data because the signals for both types of methyl protons appear at δ = 2.5−3.0 ppm (Supporting Information Figure 1). The synthesis of GCPQA, a new compound was confirmed by 1H NMR and the terminal aldehyde function13 on degradation of GCPQA with nitrous acid was confirmed with the purpald reagent (Table 1). Purpald (4-amino-5-hydrazino1,2,4-triazole-3-thiol) in alkaline solution reacts with aldehydes to form unstable, oxygen-labile intermediates that rapidly oxidize (0.6, polydisperse 0.43 ± 0.01

263 ± 30 542 ± 10

182 ± 1

>0.6, polydisperse

n.d.

n.d.

polydispersity

LORR time (mins) 3.49 ± 033 4.89 ± 0.27

not diluted

0.55 ± 0.21 >0.6, polydisperse >0.6, polydisperse not diluted

n.d.

n.d.

3.24 ± 0.32

306 ± 14

5.07 ± 0.52 5.13 ± 0.59a

Significantly (p < 0.05) longer than the effect observed with Diprivan.

lated and the number of digits on the claw (Figure 3c), even when the initial polymer level was either 4 mg mL−1 (DABGCPQA30 and DAB-GCPQA70) or 5 mg mL−1 (DABGCPQA36). The relationship between the number of digits and level of drug encapsulated is expressed by eq 11 [propofol] = 2.88γ − 1.295 (r 2 = 0.94)

Previously we showed, with linear 15 kDa GCPQ polymers, that self-assembly and subsequent drug encapsulation increased with an increase in the number of palmitoyl groups.9 Additionally, a high palmitoylation level, even with a low molecular weight, also promoted aggregation.9 However, with the current data an increase in palmitoyl levels from 2.5 mol % (DAB-GCPQA70) to 18 mol % (DAB-GCPQA36) led to reduced polymer aggregation (an increase in the CMC, Tables 2 and 3) as the number of digits decreased on going from DABGCPQA70 to DAB-GCPQA36. Hence whereas with the linear GCPQ amphiphiles the level of palmitoylation is the main driver of aggregation, in the case of the claw amphiphiles the number of digits (or molecular weight) was the main driver of polymer aggregation. The intrinsic solubility of propofol is 1.1 mM, and hence the claw amphiphile increased the level of propofol incorporated within aqueous media by 60-fold (Figure 3c). Hydrophobic Compound Release in Vivo. To verify that the encapsulated molecules would be released when required, we tested the formulations in a murine loss of righting reflex (LORR) model using the anesthetic drug propofol.9,24 The formulations were prepared in glycerol; glycerol was added to make the formulation isotonic with blood. Nanoparticles were still observed with the glycerol formulations (Figure 4b,d). The formulations, with the exception of GCPQA4, were polydisperse (Figure 4b,d, Table 5) and showed a tendency toward increasing particle size on dilution (Table 5). The polymer amphiphile formulations, as has been reported previously for GCPQ amphiphiles,9 consisted of small 50 nm particles, presumably empty polymeric micelles, and a larger population of 200−500 nm particles, presumably drug-filled particles (Figure 4). No gross toxic effects were observed when the claw amphiphiles were administered intravenously at a concentration of 1 mg to mice (∼42 mg kg−1). All of the formulations resulted in a LORR time (Table 5), and there was no significant difference between the linear and claw GCPQ-based formulations although the unfiltered DAB-GCPQ36 formulation produced a significantly longer LORR time when compared to the commercial Diprivan formulation [diluted to 11 mM (2 mg mL−1) with 0.24 M glycerol]; this could be due to the fact that it contained a higher polymer−drug ratio. Filtration has been shown to reduce the polymer content by half as larger aggregates are filtered out.1 These experiments demonstrate that whereas the claw amphiphiles form stable self-assemblies with CMCs in the nanomolar range, when injected intra-

(11)

where [propofol] is the millimolar concentration of propofol encapsulated by a polymer concentration of 4 to 5 mg mL−1. As the molecular weight increases with the number of digits, there is an expected direct linear relationship between the amount of propofol encapsulated and the Mn of all of the linear (GCPQA14) and claw (DAB-GCPQA30, DAB-GCPQA36, and DAB-GCPQA70) amphiphiles (eq 12), [propofol] = 0.67M n − 5.39 (r 2 = 0.97)

(12)

suggesting that the aggregation and consequent drug encapsulation effects may be due simply to a change in molecular weight and not due to a change in polymer architecture. An examination of the molecular weight effects within the linear polymers reveals that all three linear polymers, GCPQA4, GCPQ14 and GCPQA20, encapsulate similar levels of propofol: 24.1 ± 2.8, 24.45 ± 0.017, and 23 ± 3.9 mM, respectively (Figure 3d, Table 4), at an initial polymer concentration of 10 mg mL−1 and an initial propofol concentration of 28 mM (GCPQA4 and GCPQA20) or 84 mM (GCPQA14). Hence the increase in polymer aggregation (eq 10) and consequent increase in drug encapsulation (eq 11), as the number of digits increase, is not due to molecular weight effects alone. The drug-loading data further demonstrate the advantage of the 3D claw architecture as with an increase in the number of digits the polymers were able to form more hydrophobic domains and encapsulate a higher quantity of drug. The fact that the level of palmitoylation varied over a wide range with the amphiphiles (2.5−18.0 mol %, Table 2) did not alter the drug loading versus claw digit relationship. These data strongly demonstrate that the number of digits and not the level of palmitoylation exerted greater control over the polymer’s aggregation into hydrophobic domains and its subsequent drug encapsulation capability and support our hypothesis that the claw amphiphiles, by virtue of their architecture, form hydrophobic domains more easily. 4220

dx.doi.org/10.1021/la304909r | Langmuir 2013, 29, 4214−4224

Langmuir

Article

(41 mL) and methanol (79 mL). To this solution was added dropwise a solution of acetic anhydride (42 mg) dissolved in methanol (19 mL). The reaction was stirred for 24 h and was stopped by adding ammonium hydroxide (25% w/v, 19 μL). The product was dialyzed exhaustively (MWCO 7 kDa) against water (5 L) over 24 h with six changes and lyophilized. The product was recovered as a white cottonlike solid. Once again, GCPQA20 was synthesized from GCPQ20, GCPQA14 was synthesized from GCPQ14, GCPQA8 was synthesized from GCPQ8, and GCPQA6 was synthesized from GCPQ6. GCPQA 1H NMR (CD3OD, D2O − 1:1): δ0.89 = CH3 (CH3− CH2−, palmitoyl), δ1.29 = CH2 (−CH2−CH2−CH2−, palmitoyl), δ1.58 = CH2 (−CH2−CH2−CO−, palmitoyl), δ1.90 = CH3 (CH3−CO− NH−, acetyl-glycol chitosan), δ2.0−2.1 = CH2(−CH2−CO, palmitoyl), δ2.35−2.65 = CH3 [N(CH3)2−CH−, -dimethylamino-glycol chitosan and NH(CH3)−CH−, −monomethylamino-glycol chitosan], δ3.38 = CH3 [N(CH 3 ) 3 −CH−, trimethylamino-glycol chitosan], δ 3.45−4.1 = [−CH(OH)− and −CH2−OH, glycol chitosan], δ3.3 = methanol and δ4.8 = water protons. GCPQA20 (n = 1): Yield = 0.192 g (59.8%), Mw = 26 kDa, Mn = 15.6 kDa, Mw/Mn = 1.30, mole % palmitoylation = 10, mole % quaternary ammonium groups = 4.3, mole % acetylation = 21.8. GCPQA14 (n = 1): Yield = 0.491 g (61.4%), Mw = 20.2 kDa, Mn = 14.3 kDa, Mw/Mn = 1.41, mole % palmitoylation = 6.7, mole % quaternary ammonium groups = 10.6, mole % acetylation = 30. GCPQA8 (n = 3): Yield = 0.225 ± 0.034 g (50 ± 12%), Mw = 16.6 ± 7.2 kDa, Mn = 8.6 ± 0.9 kDa, Mw/Mn = 2.0 ± 1.0, mole % palmitoylation = 20.2 ± 2.6, mole % quaternary ammonium groups =6.0 ± 2.3, mole % acetylation = 4.0 ± 0.3. GCPQA6 (n = 1): Yield = 0.281 g (71%), Mw = 8.2 kDa, Mn = 6.2 kDa, Mw/Mn = 1.33, mole % palmitoylation = 10.2%, mole % quaternary ammonium groups = 8.9%, mole % acetylation = 3.9%. Synthesis of Nitrous Acid Degraded GCPQA. Nitrous aciddegraded GCPQA was prepared by adapting a previously published method of GC degradation.13 GCPQA (0.263 g) was dissolved in acetic acid (2.5 v/v%, 26 mL). The solution was cooled on ice to 0−4 °C for 30 min. Dissolved oxygen was removed by bubbling N2 gas through the solution for 5 min while cooling the solution on ice. A freshly prepared solution of sodium nitrite, NaNO2 (0.89 g) in acetic acid (2.5 v/v%, 1 to 2 mL) was added immediately. The reaction was allowed to proceed in the dark without stirring for 72 h at 0−4 °C. The product was recovered by precipitation with acetone at 50 °C, followed by cooling to room temperature. The precipitate was separated by centrifugation (9000 rpm × 5 min, 10 °C, Hermle Z323K, VWR, Lutterworth, U.K.) and left overnight to air dry. The product, nitrous acid-degraded GCPQA (d-GCPQA), presented as a pale-yellow solid. Once again, d-GCPQA5 was synthesized from GCPQA14, d-GCPQA4 was synthesized from GCPQA7, and dGCPQA5A was synthesized from GCPQA6. d-GCPQA 1H NMR (CD3OD, D2O, 0.1 M HCl, 1:1:5 drops): δ0.89 = CH3 (CH3−CH2−, palmitoyl), δ1.28 = CH2 (−CH2−CH2−CH2−, palmitoyl), δ1.90 = CH3 (CH3−CO−NH−, acetyl-glycol chitosan), δ2.6−3.1 = CH3 [N(CH3)2−CH−, -dimethylamino-glycol chitosan and NH(CH3)−CH−, -monomethylamino-glycol chitosan], δ3.38 = CH3 [N(CH3)3−CH−, trimethylamino-glycol chitosan], δ3.49−4.18 = CH and CH2 [−CH(OH)− and −CH2−OH, glycol chitosan], δ3.3 = methanol protons and δ4.85 = water protons. d-GCPQA4 (n = 1): Yield = 0.111 g (48%), Mw = 7.5 kDa, Mn = 4.1 kDa, Mw/Mn = 1.8, mole % palmitoylation = 11.3, mole % quaternary ammonium groups = not detectable. d-GCPQA5 (n = 1): Yield = 0.337 g (52%), Mw = 6.9 kDa, Mn = 5.0 kDa, Mw/Mn = 1.38, mole % palmitoylation = 2.4, mole % quaternary ammonium groups = 7.8, mole % acetylation = 40. d-GCPQ5A (n = 1): Yield = 0.158 g (60%), Mw = 5.5 kDa, Mn = 5.1 kDa, Mw/Mn = 1.08, mole % palmitoylation = 9.7, mole % quaternary ammonium groups = not determined, mole % acetylation = not determined. Synthesis of Claw Amphiphile N-graft-(N,N,N-Trimethyl, NN-Dimethyl, N-Monomethyl, N-Acetyl, N-Palmitoyl, 6-O-Glycol Chitosan)-Diaminobutane Poly(propylenimine) Dendrimer

venously the encapsulated drug is still available to its brain target.

■

CONCLUSIONS A new polymer amphiphile architecture is presentedthe clawshaped amphiphilein which amphiphilic polymers bearing pendant hydrophobic groups extend from a generation 3 diaminobutane dendrimer core. These new claw amphiphiles self-assemble in aqueous media in the nanomolar to picomolar range, with the self-assembly being entropy-driven. The CMC of the amphiphiles decreases with the number of chitosan amphiphile digits, even when the chitosan amphiphile digits are substituted with fewer hydrophobic groups. Drug loading also increased with the number of chitosan amphiphile digits. We conclude from our data that the favorable self-assembly and drug-loading characteristics seen with the claw amphiphiles are a result of the claw architecture favoring more intramolecular and intermolecular hydrophobic contacts between polymer chains.

■

EXPERIMENTAL SECTION

Materials. Glycol chitosan, palmitic acid N-hydroxysuccinimide, sodium iodide, sodium bicarbonate, methyl iodide, N-methyl-2pyrrolidone, diaminobutane poly(propylenimine) generation 3 dendrimer (DAB 16), deuterated solvents, propofol, diethyl ether, and all other chemicals were supplied by Sigma-Aldrich Co., U.K., unless otherwise specified. Diprivan was supplied by AAH Pharmaceuticals Ltd., Romford, U.K. Organic solvents were supplied by the School of Pharmacy, University of London. Acetic anhydride was obtained from Fluka, U.K. All chemicals were used without further purification. Synthesis of Claw Amphiphiles N-graft-(N,N,N-Trimethyl, N,N-Dimethyl, N-Monomethyl, N-Acetyl, N-Palmitoyl, 6-OGlycol Chitosan)-Diaminobutane Poly(propylenimine) Dendrimer (DAB-GCPQA). The amphiphiles were synthesized as shown in Scheme 1. Synthesis of N,N,N,-Trimethyl, N,N-Dimethyl, N-Monomethyl, N-Palmitoyl, 6-O-Glycol Chitosan (GCPQ). GC was degraded in HCL (4 M) as previously described1,9,11 for 3, 15, 48, and 72 h to give GC3, GC15, GC48, and GC72. GCPQ was synthesized as previously described1,9 by reacting GC (GC3, GC48, and GC72, 0.969 g) with palmitic acid Nhydroxysuccinimide (PNS, 2.044 g), followed by alkylation of the resulting N-palmitoyl glycol chitosan (0.91 g) with methyl iodide (5.6 mL). The final products (GCPQ20 from GC3, GCPQ6 from GC48, and GCPQ8 from GC72) were isolated by precipitation with diethyl ether, washing with ethanol, and exhaustive dialysis [against sodium chloride solution (0.1 M, 5 L) with three changes and subsequently against water (5 L) over 24 h with six changes, molecular weight cut off = 7 kDa] and lyophilization. GCPQ20 (n = 1): Yield = 0.324 g (37%), Mw = 26.1 kDa, Mn = 19.5 kDa, Mw/Mn = 1.3, mole % palmitoylation = 5.4, mole % quaternary ammonium groups = 9.4, mole % acetylation = 9.2%. GCPQ6 (n = 1): Yield = 0.430g (47%), Mw = 8.08 kDa, Mn = 6.33 kDa, Mw/Mn = 1.27, mole % palmitoylation = 9.3%, mole % quaternary ammonium groups = 7.8%, mole % acetylation = not detectable. GCPQ8 (n = 3): Yield = 0.284 ± 0.084 g (38 ± 5%), Mw = 11.7 ± 0.5 kDa, Mn = 8.1 ± 0.4 kDa, Mw/Mn = 1.4, mole % palmitoylation = 16.5 ± 1.4, mole % quaternary ammonium groups = 9.2 ± 2.7, mole % acetylation = not detectable. GCPQ14 was synthesized from GC15 in a similar manner by reacting GC15 (0.1 g) with PNS (0.0396 g), followed by alkylation of the resulting N-palmitoyl glycol chitosan (0.15 g, pooled batches) with methyl iodide (0.93 mL). GCPQ14 (n = 1): Yield = 0.0478 g (47.8%). Synthesis of N,N,N,-Trimethyl, N,N-Dimethyl, N-Monomethyl, N-Acetyl, N-Palmitoyl, 6-O-Glycol Chitosan (GCPQA). GCPQ (394 mg) and sodium bicarbonate (276 mg) were dissolved in water 4221

dx.doi.org/10.1021/la304909r | Langmuir 2013, 29, 4214−4224

Langmuir

Article

(DAB-GCPQA). d-GCPQA4 was chemically conjugated to DAB 16 by reductive N-alkylation. GCPQA4 (0.144 g) was stirred in acetate buffer (0.1 M, pH 6.5, 12 mL) containing sodium chloride, NaCl (0.1 M), for 2h. To this solution was then added a solution of DAB 16 (4.5 mg) dissolved in acetate buffer containing NaCl. Sodium cyanoborohydride, NaCNBH3 (30 mg), in acetate buffer containing NaCl was then added to the solution after 2 and 24 h from the time that DAB16 was added. The reaction was allowed to proceed for 4 days with stirring. The product was purified by exhaustive dialysis (molecular weight cut off = 3.5 kDa) for 2 h against water (5 L), followed by chromatography of the dialyzate (3 mL each time) through a gel filtration column. The column (300 × 18 mm2) was packed with Sephadex G-100 and subsequently washed with acetate buffer (0.1 M, pH 4.3) before use. Six fractions (15 mL) were collected, and fractions 1−4 containing the product (DAB-GCPQA36) were further dialyzed exhaustively (molecular weight cut off = 3.5 kDa) against water (4 L) over 24 h with three changes and lyophilized. The synthesis of DABGCPQA42 from DAB 16 and GCPQA5A was accomplished using the same procedure as above. Nitrous acid-degraded GCPQA5 was chemically conjugated to DAB 16 by dissolving GCPQA5 (0.385 g) in acetate buffer (0.1 M, pH = 4.3, 2 mL) containing sodium chloride (0.1 M), and the solution was left to stir for 2 h prior to the addition of DAB 16 (0.012 g). The reaction was then left for 4 days, and then NaCNBH3 (100 mg) was added to the reaction 2 and 24 h after the DAB 16 had been added. At the end of the reaction, the reaction mixture was dialyzed against NaHCO3 (0.1 M, 1 × 5 L) and then water (6 × 5 L) all over a 24 h period and the dialyzate was freeze-dried. The freeze-dried product was dissolved in acetate buffer (0.1 M, pH = 4.3, 2 mL) and chromatographed over a Sephadex G 100 column (60 × 22 mm2). The column was washed with water and acetate buffer prior to use, and seven 15 mL eluates were collected and the molecular weight of each fraction was determined. Fraction 6 contained DABGCPQA30, and fraction 2 contained DAB-GCPQA70. The fractions were then further exhaustively dialyzed against water (5 × 5 L) over a 24 h period and freeze-dried. DAB-GCPQA 1H NMR (CD3OD, D2O, 0.1 M HCl, 1:1:5 drops): δ0.89 = CH3 (CH3−CH2−, palmitoyl), δ1.25 = CH2 (−CH2−CH2− CH2−, palmitoyl), δ1.90 = CH3 (CH3−CO−NH−, acetyl-glycol chitosan) and CH2 (N−CH2−CH2−CH2−CH2−, DAB 16), δ2.5−3.0 = CH2 (−CH2−N−, DAB 16), δ2.5−3.0 = CH3 [−N(CH3)2−CH−, -dimethylamino-glycol chitosan and NH(CH3)-CH−, -monomethylamino-glycol chitosan], δ3.37 = CH3 [−N(CH3)3−CH−, trimethylamino-glycol chitosan], δ3.45−4.05 = CH and CH2 [−CH(OH)− and −CH2−OH, glycol chitosan], δ3.27 = methanol protons and δ4.8 = water protons. DAB-GCPQA36 (n = 3): Yield = 0.037 ± 0.010 g (41 ± 14%, n = 3), Mw = 59.9 kDa, Mn = 36.3 kDa, Mw/Mn = 1.7, mole % palmitoylation relative to sugar monomers = 18.1, mole % quaternary ammonium groups relative to sugar monomers = 7.9, number of GCPQA4 digits = 8. DAB-GPQA30 (n = 1): Yield = 0.08 g (25%), Mw = 33 kDa, Mn = 30.2 kDa, Mw/Mn = 1.09, mole % palmitoylation = 4.1, mole % quaternary ammonium groups = 7.8, number of GCPQA5 digits = 6. DAB-GCPQA70: Yield = 0.046 g (15%), Mw = 103.3 kDa, Mn = 69.5 kDa, Mw/Mn = 1.49, mole % palmitoylation = 2.5, mole % quaternary ammonium groups = 7.8, number of dGCPQA5 digits = 14. DAB-GCPQA42: Yield = 0.055 g (38%), Mw = 49.2 kDa, Mn = 42.2 kDa, Mw/Mn = 1.16, mole % palmitoylation = 4.6, mole % quaternary ammonium groups = not detectable, number of dGCPQA5A digits = 8. Structural Characterization of Amphiphiles. 1H NMR and 1 H−1H COSY analyses were performed on a Bruker Avance 400 MHz spectrometer (Bruker, Coventry, U.K.). GC solutions were analyzed in D2O, GCPQ solutions were analyzed in CD3OD, GCPQA solutions were analyzed in CD3OD, D2O (1:1), and DAB-GCPQA solutions were analyzed in CD3OD, D2O, 0.1 M HCl (1:1:5 drops). The level of palmitoylation was calculated by comparing the ratio of palmitoyl methyl protons (δ = 0.95 ppm) to the sugar protons (δ = 3.5−4.2

ppm), the level of quaternization was calculated by comparing the ratio of quaternary ammonium (δ = 3.4 ppm) to sugar protons (δ = 3.5−4.2 ppm), and the level of acetylation (where possible) was calculated by comparing the ratio of acetyl protons (δ = 1.9 ppm) to sugar protons (δ = 3.5−4.2 ppm). The presence of the aldehyde functional group in d-GCPQA after degradation with sodium nitrite was verified using the purpald (4amino-5-hydrazino-1,2,4-triazole-3-thiol) test for aldehydes.15 Purpald (200 mg) was dissolved in sodium hydroxide solution (1 M, 2 mL). This alkaline purpald solution (1 mL) was then added to the polymer (15 mg mL−1, 1 mL), and the resulting solution was vortex mixed with aeration. The color of the vortex mixed solution was immediately recorded, and the ultraviolet absorption spectrum was recorded (Shimadzu UV 1650PC UV−vis spectrophotometer, Shimadzu Ltd., Milton Keynes, U.K.). A purple color or shift in the wavelength of maximum absorbance to ∼535 nm was indicative of the presence of an aldehyde function. The molecular weights of GCPQ, GCPQA, d-GCPQA, and DABGCPQA samples were determined by gel permeation chromatography−multiangle laser light scattering (GPC-MALLS) using a DAWN EOS 18 angle light scattering detector (λ = 690 nm), Optilab DSP interferometric refractometer (λ = 690 nm), and quasielastic light scattering detectors (Wyatt Technology Corporation, Santa Barbara, CA, USA) equipped with a Waters 515 pump and a Waters 717 Plus autosampler (Waters Ltd., Elstree, U.K.). Samples were eluted with a mobile phase of acetate buffer (0.3 M sodium acetate, 0.2 M acetic acid, pH 5.0), methanol (35:65). Filtered samples (0.2 μm, 200 μL) were chromatographed over a POLYSEP-GFC-P guard column (35 × 7.8 mm2, Phenomenex, Macclesfield, U.K.) and a POLYSEP-GFC-P 4000 column (300 × 7.8 mm2, exclusion limit for PEG = 200 kDa) at a loading concentration of 5−10 mg mL−1. Measurements were performed at room temperature and at a flow rate of 0.4 mL min−1. The data were processed using ASTRA for Windows version 4.90.08 software (Wyatt Technology Corporation, Santa Barbara, CA, USA). Toluene and bovine serum albumin were used for the DAWN EOS 18 angle light scattering 90° detector calibration and the normalization of all angles relative to the 90° detector, respectively. The specific refractive index increments (dn/dc) of GCPQ, GCPQA, d-GCPQA, and DAB-GCPQA were measured in acetate buffer (0.3 M sodium acetate/0.2 M acetic acid, pH 5.0), methanol (35:65) at 40 °C with an Optilab DSP interferometric refractometer (λ = 690 nm). Filtered samples (0.2 μm) of five different concentrations ranging from 0.2 to 1 mg mL−1 were manually injected (Rheodyne 7725, Gilson Scientific Ltd., Luton, U.K.) at a pump flow rate of 0.3 mL min−1. The data were processed using Wyatt DNDC for Windows software version 5.90.03.

■

SELF ASSEMBLY Amphiphile critical micellar concentrations (CMCs) were determined using the pyrene fluorescent probe 18 and isothermal calorimetry.12 With the pyrene probe, dispersions of the amphiphiles were added to the vials containing pyrene (added as a methanolic solution and evaporated to dryness) to achieve a final pyrene concentration of 0.5 μM. Pyrene− amphiphile samples were then equilibrated at 65 °C for 3 h and subsequently cooled to room temperature. The fluorescence emission spectra were recorded (λ = 350−550 nm) at an excitation wavelength of 339 nm. The intensity ratio of the first (375 nm) and third (383 nm) vibronic peaks was recorded. Isothermal calorimetry experiments involved a measurement of the enthalpy change of demicellization (ΔHdemic) and the CMC at 25 °C using a VP-ITC MicroCalorimeter (MicroCal, LLC, Northampton, MA, USA) using a previously published method.12 From these data, the thermodynamics parameters were calculated. The particle sizes (z-average mean hydrodynamic diameters) of the amphiphile aggregates were determined at 25 °C using 4222

dx.doi.org/10.1021/la304909r | Langmuir 2013, 29, 4214−4224

Langmuir photon correlation spectroscopy (Malvern Zetasizer 3000HSA, Malvern Instruments, Malvern, U.K.) before and after filtration (0.45 or 0.8 μm). Data were analyzed using the contin mode of analysis. Prior to experiments, the size of polystyrene standard dispersions (200 nm) was determined; the size recorded was in accordance with that stated by the manufacturer. Nanoparticles were imaged by transmission electron microscopy1 (TEM, Philips/FEI CM120 Bio Twin, Philips, Eindhoven, The Netherlands). Drug Loading. Drug loading was achieved by probe sonication of the amphiphiles (1−10 mg mL−1) in water for 5 min and in the presence of propofol (15 mg mL−1). Amphiphile−propofol dispersions were sonicated on ice and were filtered (0.45 μm). The filtrate was dissolved in methanol, water (80:20 v/v) and subsequently chromatographed (20 μL) over an Onyx Monolith C18 column (100 × 4.6 mm2, Phenomenex). The samples were eluted using an Agilent Technologies series 1200 high-performance liquid chromatography (HPLC) instrument and detected at a wavelength of 273 nm (Agilent Technologies, Wokingham, U.K.). The mobile phase was methanol, water (80:20 v/v) at a flow rate of 2 mL min−1. The limit of detection was 0.5 μg mL−1, and the calibration curve equation was y = 5.593x − 0.9122, r2 = 0.9998 over the range of 2.5−50 μg mL−1. Pharmacodynamic Activity. DAB-GCPQA36−propofol formulations were prepared by probe sonication of DABGCPQA36 (5 mg mL−1) and propofol (2 or 5 mg mL−1) in glycerol (0.24 M) on ice for 10 min. Samples were stored at room temperature for 24 h and filtered (0.8 μm) and analyzed by HPLC before use. GCPQA20−propofol formulations were prepared by probe sonication of GCPQA20 (10 mg mL−1) and propofol (5 mg mL−1) in glycerol (0.24 M) on ice for 10 min. Samples were stored at room temperature for 24 h and filtered (0.8 μm) and analyzed by HPLC before use. The 5 mg mL−1 propofol formulation was diluted to a final propofol concentration of 2 mg mL−1 and probe sonicated again for 10 min just prior to administration. All animal procedures were conducted under a Home Office license and approved by the School of Pharmacy local ethics committee. Male CD-1 mice (21−27 g) were injected intravenously via the tail vein with 400 μg of propofol (2 mg mL−1) administered in a 200 μL volume as a Diprivan, DABGCPQA36, or GCPQA20 formulation (Table 4). The loss of righting reflex (LORR) time was recorded. The LORR was confirmed in all dosed animals immediately after dosing by an animal failing to right itself when placed on its back. Statistics. Statistical analyses were performed via a one-way ANOVA test using Minitab 16 (Minitab Ltd., Coventry, U.K.) followed by Tukey’s post tests. Significant differences were indicated by a p value of less than 0.05.

■

ACKNOWLEDGMENTS

■

REFERENCES

This work was funded by grants from The Engineering and Physical Sciences Research Council of the United Kingdom.

(1) Siew, A.; Le, H.; Thiovolet, M.; Gellert, P.; Schatzlein, A.; Uchegbu, I. Enhanced oral absorption of hydrophobic and hydrophilic drugs using quaternary ammonium palmitoyl glycol chitosan nanoparticles. Mol. Pharm. 2012, 9, 14−28. (2) Muschiolik, G. Multiple emulsions for food use. Curr. Opin. Colloid Interface Sci. 2007, 12, 213−220. (3) Klier, J.; Tucker, C. J.; Kalantar, T. H.; Green, D. P. Properties and applications of microemulsions. Adv. Mater. 2000, 12, 1751−1757. (4) Gregoriadis, G. Liposome Technology; CRC Press: Boca Raton, FL, 2006; Volumes I−III. (5) Florence, A. T.; Attwood, D. Physicochemical Principles of Pharmacy, 4th ed.; McMillan Press: London, 2006; p 564. (6) Rowe, R. C.; Sheskey, P. J.; Owen, S. C. Handbook of Pharmaceutical Excipients; Pharmaceutical Press: London, 2006. (7) Lalatsa, A.; Schatzlein, A. G.; Mazza, M.; Le, T. B.; Uchegbu, I. F. Amphiphilic poly(l-amino acids) - new materials for drug delivery. J. Controlled Release 2012, 161, 523−36. (8) Market Study: Surfactants. Ceresana Research: Konstanz, Germany, 2012. (9) Qu, X. Z.; Khutoryanskiy, V. V.; Stewart, A.; Rahman, S.; Papahadjopoulos-Sternberg, B.; Dufes, C.; McCarthy, D.; Wilson, C. G.; Lyons, R.; Carter, K. C.; Schatzlein, A.; Uchegbu, I. F. Carbohydrate-based micelle clusters which enhance hydrophobic drug bioavailability by up to 1 order of magnitude. Biomacromolecules 2006, 7, 3452−3459. (10) Wang, W.; Qu, X. Z.; Gray, A. I.; Tetley, L.; Uchegbu, I. F. Selfassembly of cetyl linear polyethylenimine to give micelles, vesicles, and dense nanoparticles. Macromolecules 2004, 37, 9114−9122. (11) Wang, W.; McConaghy, A. M.; Tetley, L.; Uchegbu, I. F. Controls on polymer molecular weight may be used to control the size of palmitoyl glycol chitosan polymeric vesicles. Langmuir 2001, 17, 631−636. (12) Chooi, K. W.; Gray, A. I.; Tetley, L.; Fan, Y. L.; Uchegbu, I. F. The molecular shape of poly(propylenimine) dendrimers has a profound effect on their self assembly. Langmuir 2010, 26, 2301− 2316. (13) Tommeraas, K.; Varum, K. M.; Christensen, B. E.; Smidsrod, O. Preparation and characterisation of oligosaccharides produced by nitrous acid depolymerisation of chitosans. Carbohydr. Res. 2001, 333, 137−144. (14) Dickinson, R. G.; Jacobsen, N. W. A new sensitive and specific test for the detection of aldehydes: formation of 6-mercapto-3substituted-s-triazolo[4,3-b]-s-tetrazines. Chem. Commun. 1970, 1719−1720. (15) Dickinson, R. G.; Jacobsen, N. W. 6-Substituted s-triazolo[4,3b]-s-tetrazine-3-thiols: a sensitive and specific test for aldehydes. J. Chem. Soc., Perkin Trans. 1975, 975−979. (16) Tommeraas, K.; Koping-Hoggard, M.; Varum, K. M.; Christensen, B. E.; Artursson, P.; Smidsrod, O. Preparation and characterisation of chitosans with oligosaccharide branches. Carbohydr. Res. 2002, 337, 2455−2462. (17) Lamarque, G.; Chaussard, G.; Domard, A. Thermodynamic aspects of the heterogeneous deacetylation of β-chitin: reaction mechanisms. Biomacromolecules 2007, 8, 1942−1950. (18) Kalyanasundaram, K.; Thomas, J. K. Environmental effects on vibrionic band intensities in pyrene monomer fluorescence and their application in studies of micellar systems. J. Am. Chem. Soc. 1977, 99, 2039−2044. (19) Kujawa, P.; Goh, C. C. E.; Calvet, D.; Winnik, F. M. Do fluorocarbon, hydrocarbon, and polycyclic aromatic groups intermingle? Solution properties of pyrene-labeled bis(fluorocarbon/ hydrocarbon)-modified poly(N-isopropylacrylamide). Macromolecules 2001, 34, 6387−6395.

ASSOCIATED CONTENT

* Supporting Information S

Polymer characterization data. This material is available free of charge via the Internet at http://pubs.acs.org.

■

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: +44 207 753 5997. Fax: +44 207 753 5942. Notes

The authors declare no competing financial interest. 4223

dx.doi.org/10.1021/la304909r | Langmuir 2013, 29, 4214−4224

Langmuir

Article

(20) Sideratou, Z.; Tsiourvas, D.; Paleos, C. M. Solubilization and release properties of PEGylated diaminobutane poly(propylene imine) dendrimers. J. Colloid Interface Sci. 2001, 242, 272−276. (21) Eisenberg, H. Polyelectrolyte excluded volume and expansion compared to non-ionic polymers. Acta Polym. 1998, 49, 534−538. (22) Paula, S.; Sus, W.; Tuchtenhagen, J.; Blume, A. Thermodynamics of micelle formation as a function of temperature - a highsensitivity titration calorimetry study. J. Phys. Chem. 1995, 99, 11742− 11751. (23) Tanford, C. Interfacial free energy and the hydrophobic effect. Proc. Natl. Acad. Sci. U.S.A. 1979, 76, 4175−4176. (24) Irifune, M.; Takarada, T.; Shimizu, Y.; Endo, C.; Katayama, S.; Dohi, T.; Kawahara, M., Propofol-induced anesthesia in mice is mediated by gamma-aminobutyric acid-A and excitatory amino acid receptors. Anesth. Analg. 2003, 97, 424−429.

4224

dx.doi.org/10.1021/la304909r | Langmuir 2013, 29, 4214−4224