Linear-g-hyperbranched and cyclodextrin-based amphiphilic block copolymer as a multifunctional nanocarrier

Summary In this paper, a novel, multifunctional polymer nanocarrier was designed to provide adequate volume for high drug loading, to afford a multiregion encapsulation ability, and to achieve controlled drug release. An amphiphilic, triblock polymer (ABC) with hyperbranched polycarbonsilane (HBPCSi) and β-cyclodextrin (β-CD) moieties were first synthesized by the combination of a two-step reversible addition-fragmentation transfer polymerization into a pseudo-one-step hydrosilylation and quaternization reaction. The ABC then self-assembled into stable micelles with a core–shell structure in aqueous solution. These resulting micelles are multifunctional nanocarriers which possess higher drug loading capability due to the introduction of HBPCSi segments and β-CD moieties, and exhibit controlled drug release based on the diffusion release mechanism. The novel multifunctional nanocarrier may be applicable to produce highly efficient and specialized delivery systems for drugs, genes, and diagnostic agents.


S3
The molecular structure parameters of the resulting polymers were determined on a DAWN EOS size exclusion chromatography/multiangle laser light scattering (SEC/MALLS) instrument equipped with a viscometer (Wyatt Technology, USA), and HPLC grade DMF containing LiCl (0.01 mol/L) (at 40 ºC) was used as the eluent at a flow rate of 0.5 mL/min.
The chromatographic system consisted of a Waters 515 pump, differential refractometer (Optilab rEX), and one column, MZ 10 3 Å 300 × 8.0 mm. A MALLS detector (DAWN EOS), a quasielastic light scattering (QELS), and a differential viscosity meter (ViscoStar) were placed between the SEC and the refractive index detector. The molecular weight (M w ) and molecular weight distribution (MWD) were determined by a SEC/DAWN EOS/OptilabrEX/QELS model. ASTRA software (Version 5.1.3.0) was utilized for acquisition and analysis of data.
The particle size and particle size distribution index (PDI) were measured by a Zetasizer Nano-ZS dynamic light scattering (DLS) (Malvern Instruments, UK) device. Samples were kept at equilibrium at a predetermined temperature for 5 min before data collection and were measured at 25 ºC.
Transmission electron microscopy (TEM) experiments were carried out on a JEOL JEM-3010 instrument (JEOL, Japan). The mMicelle solutions were dropped onto copper grids and dried at room temperature before measurement.
UV-vis spectrophotometry measurements were preformed on a Shimadzu UV-2550 model spectrophotometer (Shimadzu, Japan). All solutions were kept at 37 o C for 48 h prior to measurements.

Synthesis of macro chain transfer agent (PHEMA-macroCTA) (Similar as described in
Ref. S4) In a flask equipped with magnetic stirring, BDATTC (141 mg, 0.5 mmol) and AIBN (2.46 mg, 0.015 mmol) were first dissolved in tertbutyl alcohol (5 mL), and then HEMA (5.8 g, 0.05 mol) was added under vigorous stirring. The mixture was degassed under vacuum for 15 min, and then purged under dry nitrogen for 30 min. The procedure described above was repeated three S4 times. The flask was then sealed under vacuum. Polymerization was conducted at 80 o C for 10 h, and then the system was cooled down to ambient temperature. The final product was precipitated in cold methanol and dried under vacuum at 40 o C for 5 d.

Synthesis of triblock copolymer (PHEMA-b-PDMAEMA-b-PHEMA) (Similar as described in Ref. S4)
PHEMA-macroCTA (0.1 g) and AIBN (2.46 mg, 0.015 mmol) were first dissolved in DMF (4 mL), and then DMAEMA (7.9 g, 0.05 mol) was added under vigorous stirring. The mixture was degassed under vacuum for 15 min, and then purged under dry nitrogen for 30 min. The procedure described above was repeated three times. The flask was then sealed under vacuum.
Polymerization was conducted at 80 o C for 2 h, and then the system was cooled down to ambient temperature. The solvent was evaporated under reduced pressure at 80 o C/10 -4 bar, and the raw product obtained was again dissolved in 1,4-dioxane. The final product was precipitated in cold hexane and dried under vacuum at 40 o C for 5 d.

Synthesis of P1 (Similar as described in Ref. S3)
PHEMA-b-PDMAEMA-b-PHEMA (1.23 g), THF (50 mL), pyridine (10.81 g, 0.14 mol) and DMAP (10 mg) were mixed in a 500 mL three-neck round-bottomed flask with a thermometer in an ice water bath. Under vigorous stirring, AC (12.38 g, 0.14 mol) dissolved in THF (20 mL) was added dropwise to the mixture. The pyridinium hydrochloride could be observed as a white precipitate at an early stage, and then later changed to a yellowish solid. The reaction was then kept at ambient temperature for 24 h, and the precipitate was filtered out and the raw product was left in the filter. The solvent was evaporated under reduced pressure at 50 o C/10 -4 bar. The final product was precipitated in cold diethylether and dried under vacuum at ambient temperature for 5 d.

Synthesis of P2
P1 (0.97 g), and DMF (2 mL) solvents were first mixed in a 100 mL round-bottomed flask equipped with a funnel under nitrogen atmosphere. Then, the first generation AB 2 monmer (0.13 g) dissolved in DMF (1 mL) was added dropwise into the solution under vigorous S5 stirring. A H 2 PtCl 6 divinyltetramethyldisiloxane solution was used as the catalyst for the hydrosilylation addition reaction. The system was then moved to an oil bath at 50 o C under vigorous stirring until no Si-H groups were detected by FTIR (2136 cm -1 ). The second generation AB 2 monmer (0.26 g) (which was dissolved in DMF (1 mL)) was added again, and the same method was used to detect whether the reaction was finished. After the system was cooled down to ambient temperature, the mixture was precipitated in cold methanol, and then the precipitate as a cyclic product was filtered out, while the raw product was left in the filter.
The solvent was evaporated under reduced pressure at 40 o C/10 -4 bar. The final product was precipitated in cold diethylether and dried under vacuum at ambient temperature for 5 d.

Synthesis of P3 (Similar as described in Ref. S3)
P2 (0.3 g) and mono-6-I-β-CD (0.6 g) were first dissolved in dry DMF (3 mL), and then the system was kept under vigorous stirring at 90 o C in an oil bath for 72 h. After cooling down to ambient temperature, the mixture was dialyzed in a dialysis bag (molecular weight cut off: 3500) against distilled water for 7 d. It was refreshed at an interval of 5 h. The final product was lyophilized and kept in glassware under vacuum for further use.

Preparations of P1, P2 and P3 micelles (Similar as described in Ref. S5)
Preparations of P1, P2 and P3 micelles were carried out at room temperature using DMF and water as solvents. P1 (or P2, P3) (20 mg) was first dissolved in DMF (1mL), and then distilled water (20 mL) was added dropwise to the solution. The system was continuously stirred for 16 h. Later, the mixture was subjected to dialysis (molecular weight cut off: 3500) against water for 2 d. When the dialysis was complete, the micelle solution was subjected to centrifugation for the removal of larger aggregates. The volume of the obtained micelle solutions was about 23-25 mL, and they were designated as P1, P2 and P3 micelles, respectively.

Encapsulation behavior of P1, P2 and P3 micelles
The encapsulation behavior of P1, P2 and P3 micelles was measured by UV-vis spectrophotometry, using LND (5 × 10 -5 mol/L) as a guest molecule in a buffer solution with S6 ionic strength equal to 0.1 mol/L and pH = 7.4. Typically, the encapsulation procedures were similar to the preparation of micelles, and no further change were made except for using LND as a buffer solution instead of the distilled water. All solutions were maintained for more than 12 h to ensure binding equilibrium and then stirred prior to measurement.

Drug loading
P1 (or P2, P3) (20 mg) and LND (2 mg) were dissolved in DMF (1 mL) under vigorous stirring, and the subsequent procedures of preparing LND-loaded micelle solutions denoted as DLMP1 (or DLMP2, DLMP3) were the same as those for preparing polymer micelles. The encapsulation efficiency (EE) and loading content (LC) of LND were calculated by using Equations 1 and 2 as follows: EE = W drug in miclles /W feeding drug × 100% (1) where W drug in miclles was determined by UV-vis spectroscopy. Here, 1 mL of DLMP1 (or DLMP2, DLMP3) was diluted to 10 mL with DMF, and the LND concentration was determined by measuring the absorbance at 298.5 nm.  S7 where W 0 (mg) is the weight of the LND in DLMP1 (or DLMP2, DLMP3), C n (mg/ml) is the concentration of the LND in buffer solution (which was withdrawn n times), and C n-1 (mg/ml) is the concentration of the LND in buffer solution (which was withdrawn n-1 times).

Release kinetics (Similar as described in Ref. S5)
The release kinetics of DLMP1, DLMP2, and DLMP3 were determined by a simple semi-empirical equation (Equation 4) and a modified Equation 5 used to describe the release behavior of polymeric micelles [S6-S8].
where M t and M∞ are the cumulative amount of released LND at time t and infinite time, respectively; k is the release constant while k' is the constant proportional to k; the exponent n describes the kinetic and the release mechanism. For a diffusion-degradation controlled release system, n for spherical particles is usually between 0.43 and 0.85. When n is close to 0.43, diffusion is the major driving force, which is called the "Fickian diffusion." When n is close to 0.85, the release is mainly controlled by degradation [S7,S8]. In order to obtain a linear fit for the drug release data, Equation 5 Figure S1c) at δ 5.56 (-CH＝CH 2 ) and the 13 C NMR spectrum ( Figure S2c) at δ 126.8 (-CH 2 =CH 2 ), δ 129.5 (-CH 2 ＝CH 2 ) and δ 164.8 (-COO-) indicate that the acylation reaction was successfully carried out. Additional evidence comes from FTIR spectra ( Figure S3c) where the peak found at 1640 cm -1 , assigned to double bond absorption, appears after the grafting reaction, further confirming the acryloyl chloride was successfully grafted onto  and P1 (c) Si H P2 S14 Figure S5. SEC elution cure of P2 and 1.84 (-CH-) indicates that the α and β additions simultaneously took place during the hydrosilylation reaction according to our knowledge [S10]. The 13 C NMR spectrum in Figure S6b further confirmed this point. However, the 1 H NMR (δ 5.71) and 13 C NMR (δ 127.5, and δ130.2) data proved that the reservation of double bond groups resulting from the incomplete hydrosilylation reaction was probably due to steric hindrance. Frey et al. found the same result during the synthesis of linear-hyperbranched block copolymers consisting of polystyrene and HBPCSi blocks [S11].
Fortunately, the residual double bond groups have little effect on the subsequent micelle formation and drug loading.