Supporting Information. for. Angew. Chem. Int. Ed Wiley-VCH 2004

Transcription

1 Supporting Information for Angew. Chem. Int. Ed Wiley-VCH Weinheim, Germany

2 Supporting Information crgd-functionalized Micelles for Targeted Doxorubicin Delivery orased asongkla, Xintao Shuai, Hua Ai, Brent D. Weinberg, John Pink, David A. Boothman, and Jinming Gao* Materials. ovasyn TGT alcohol and all of the amino acids were purchased from ovabiochem (CA). All solvents for crgd synthesis were peptide synthesis grade. Poly(ethylene glycol) monoethyl ether maleimide (H-PEG-MAL, MW = 3210 Da) was purchased from ektar Therapeutics (AL). Stannous (II) octoate (Sn(ct) 2, from Aldrich) was used as received. DX in aqueous solution (Doxorubicin-HCl, 2mg/mL) was purchased from the Bedford Laboratories (Bedford, H), and was deprotonated at ph 9.6 to obtain the hydrophobic DX. All organic solvents are of analytic grade. SLK cells were cultured in DMEM medium with 10% fetal bovine serum (FBS) at 37 C in 5% C 2 atmosphere. Synthesis of crgd using solid phase peptide synthesis chemistry. The synthetic scheme for crgd (Figure S1) is revised based on a reported procedure by Schatzlein et al. [see: C. F. Catherine, F. McCusker, P. J. Philip, J. Kocienski, T. Boyle, A. G. Andreas, G. Schätzlein, Bioorg. Med. Chem. Lett. 2002, 12, ]. ovasyn TGT alcohol (1.25 mmol) (ovabiochem, CA) was converted to its active chloride form by using acetyl chloride (1 ml/g ) (62.5 mmol) in toluene for 3 h at 60 0 C. Then the was washed with dry toluene and dichloromethane (DCM). The synthesis of linear peptide started with the attachment of aspartic acid, by mixing the chlorinated with a solution of Fmoc-Asp- All (2.5 equiv) and, -diisopropylethylamine (DIPEA) (10 equiv) in dry CH 2 Cl 2 at RT for h. The solution of CH 2 Cl 2, MeH and DIPEA was added to cap the unreacted sites of. After 30 min the was washed with dimethylformamide (DMF). The Fmoc protecting group was removed with a solution of piperidine DMF (1:4) at RT for 4 min, 2 times. The rest of amino acids were added 1

3 PEG H 2 C H Acetyl chloride in toluene PEG 3 h, 60 C H 2 C Cl Solid phase peptide synthesis, Fmoc chemistry PEG H 2 C Asp(All)-Gly-Arg(Pbf)-Lys(Dde)-Phe--Fmoc Deprotection by (Pd(PPh 3 ) 4 and piperidine Cyclization by HATU PEG H 2 C H H H 2 Deprotection Hydrazine monohydrate (Pbf) PEG H 2 C H H H -(Dde) (Pbf) Thiol addition -Succinimidyl (acetylthio)acetate PEG H 2 C H H H S CH 3 (Pbf) Deprotection Cleavage TFA/H 2 / 1, 2-ethanediol ( 90: 5: 5) H H H H 2 S CH 3 Figure S1. Schematic diagram of solid phase synthesis of crgdfk. consecutively (Fmoc-Gly-H, Fmoc-Arg(Pbf)-H, Fmoc-Arg(Pbf)-H, Fmoc-Lys(Dde)-H and Fmoc-D-Phe-H) using standard Fmoc strategy. The amino acid (2 equiv) was added first followed by,,','-tetramethyl--(7- azabenzotriazol-1-yl)uronium hexafluorophosphate (HATU) (2 equiv) and DIPEA (4 equiv) and allowed to react for 1.5 h at RT. The C-terminal allyl ester group of the aspartic acid was removed after addition of the last amino acid with Palladium-tetrakis(triphenylphosphine) (Pd(PPh 3 ) 4 ) (3 equiv) in a solution of CHCl 3, acetic acid and -methylmorpholine for 2 h at RT. The mixture was washed with DIPEA in DMF followed by 0.5%w/w diethyldithiocarbamic acid sodium salt in DMF. The head-to-tail cyclization was preformed by removal of the -terminal Fmoc group before addition of HATU (2 equiv) and DIPEA (4 equiv) in DMF at RT for 16 h. The protecting group of the amino side-chain of lysine (4, 4-dimethyl-2,6- dioxocyclohex-1-ylidene)ethyl, (Dde)) was removed by hydrazine monohydrate DMF (2:98) for 3 min at RT, 3 time. The thiol addition was preformed by swelling with DIPEA (15 equiv) in DMF, 2

4 followed by addition of S-Acetylthioglycolic acid -hydroxysuccinimide ester (SATA) solution (2 equiv) in DMF (15 mg/ml) for 2 h at RT. The was then washed with DMF followed by CH 2 Cl 2. ext, the protecting group of arginine (pentamethyl-dihydrobenzofuran-5-sulfonyl, (Pfb)) was removed using TFA CH 2 Cl 2 (1:1) for 2 h before washing with TFA CH 2 Cl 2 (1:9). The solutions were concentrated and precipitated with cold Et 2 several times. The precipitate was then redissolved in aqueous buffer, purified using reverse phase HPLC and lyophilized to give crgd peptide (90 mg, 10% yield based on H H H S CH 3 Figure S2. Proton assignments in S-acetyl crgdfk molecule initial loading of ). The purified peptide was characterized by 1 H MR and MALDI-TF mass spectroscopy: α H (600 MHz, D 2, Fig. S2) C18H2 = , C19H2 = 1.10, C17HaHb = , C17HaHb = , C14H2 = , C13HaHb = , C13HaHb = , C31H3 = 2.44, C11H2 = 2.76, C20H2 = , C21HaHb = 3.00, C21HaHb = 3.10, C15H2 = 3.22, C4HaHb = 3.51, C8H = , C4HaHb = 4.23, C6H = 4.38, C2H = , C10H = 4.64, C23H = 7.28, C27H = 7.28, C25H = 7.32, C26H = MALDI-TF-MS: m/z (M + 1) + calculated for crgdfk (C 31 H S), ; found Synthesis of diblock copolymer of maleimide-terminated poly(ethylene glycol) and poly(εcaprolactone) (MAL-PEG-PCL). MAL-PEG-PCL was synthesized by ring opening polymerization of ε-caprolactone at 68 C. Poly(ethylene glycol) monoethyl ether maleimide (H-PEG-MAL, Mn = 3,210 Da) was used as a macro-initiator. ε-caprolactone was added as a monomer and Stannous (II) octoate (Sn(ct) 2 ) was added as a catalyst. After reacting for 3 days, the mixture was allowed to cool down to room temperature. MAL-PEG-PCL was purified by redissolving in THF and precipitating in hexane 3 times. The degree of polymerization of the PCL was calculated by comparing integral intensity of 3

5 characteristic resonance of the PCL at 2.25 ppm (-C(=)-CH 2 -) and PEG resonance at 3.64 ppm (- CH 2 CH 2 -) in the 1 H MR spectrum. The amount of maleimide proton was calculated by comparing integral intensity of characteristic resonance at 6.74 ppm and PEG resonance at 3.64 ppm (-CH 2 CH 2 -). The molecular weight and polydispersity of MAL-PEG-PCL were also characterized by gel permeation chromatography (THF as eluent) and the results were found to be consistent with 1 H MR data. Two batches of MAL-PEG-PCL (Mn = 5.5 kd) were synthesized with 95% yield and used in this study. It is noteworthy that the commercially available MAL-PEG-H polymer has 76% of all PEG chains terminated with maleimide groups. This percentage is retained in the MAL-PEG-PCL copolymer, as demonstrated by 1 H MR. Preparation of crgd-dx-micelles. Doxorubicin (DX) containing polymeric micelles were prepared as following: 20 mg of MAL-PEG-PCL and 2 mg of doxorubicin were dissolved in 0.5 ml THF in a glass vial. ext, the mixture was slowly added into 10 ml of an aqueous solution of 0.05 M HEPES and 0.01 M EDTA under sonication (60 Sonic Dismembrator, Fisher Scientific). The mixture was vigorously stirred under argon for 3 h to remove THF. Then different amounts of c(rgdf(ε-sacetylthioacetyl)k and 0.05 M hydroxy amine in HEPES/EDTA aqueous solution were added into solutions of micelles with 5, 16, and 76% maleimide density. The conjugation was allowed to occur for 4 h followed by filtration through a Millipore centrifugal filter (pore size 0.45 µm) to remove DX aggregates in micelle solution. The crgd-micelles were dialyzed with Spectra/Por dialysis membrane (molecular weight cutoff = 50,000 Da) until free crgd was completely removed. Micelle solutions were lyophilized to obtain the powdery form to avoid DX release before cell culture experiments. Micelles were characterized by dynamic light scattering and atomic force microscopy. 1 H MR was used to confirm the formation of core-shell structure and conjugation of crgd to micelles. The strong resonance of methylene proton in PEG was detected where as all of caprolactone proton resonance were hardly observed demonstrating the core-shell structure of these micelles. The successful conjugation of crgd 4

6 onto the surface of micelles was verified by the appearance of phenyl protons of crgd at 7.4 ppm. These results are described in the manuscript (Figure 1B). DLC measurement. The DX-loading content (DLC, defined as the weight percentage of DX in the micelles) was quantified by UV-Vis analysis using a Perkin-Elmer Lambda 20 spectrophotometer. First, micelle solutions were frozen and lyophilized to yield the solid micelle samples. Then the dried samples were weighed and re-dissolved in a mixture of chloroform and DMS (1:1, v/v) before the UV-Vis analysis. The absorbance at 485 nm was measured to determine the DX content in the solution with a previously established calibration curve. Stability of micelles in solution. The critical micelle concentration of MPEG-PCL (Mn = 5.5 kd) was determined to be 60 µg/ml by a surface tension method with a manual digital tensiometer (Sigma 703, KSV Instruments Ltd, Finland, see reference 23 for detailed procedures). Dynamic light scattering measurement was carried out to examine the micelle stability during experiment and in storage. o obvious size change was detected for the micelle solutions up to 3 weeks at 4 C. Moreover, neither micelle aggregation nor dissociation leading to obvious micelle size changes was noted during experiments. Atomic force microscopy (AFM). DX-loaded micelles were prepared as mentioned above. Two group of micelle were used for AFM study. The first group was DX-micelles without crgd ligand. The second group was DX-micelles with 76% crgd density. The micelle suspension (2 µl) was placed on the mica surface, and allowed to dry at room temperature overnight. The size measurement was carried out on an atomic force microscope (Multimode, Digital Instruments, Santa Barbara, CA) operated in the tapping mode using a high resonance frequency AFM tip (type: CH, anoworld Innovative Technologies, Switzerland). The AFM tip has a pyramidal geometry with a height of 4 µm and a force constant of 42 /m. The nominal tip radius of curvature is 5-10 nm. The constant force mode was used with a scan frequency of 2 Hz. Both non-functionalized and 76% crgd containing micelles show discrete and round-shaped nanoparticles. These results were shown in Figure 2A and 2C. 5

7 Micelles with 76% crgd attachment (43.2 ± 3.9 nm, n=29) showed a mean size slightly larger than that of RGD-free micelles (37.5 ± 2.6 nm, n=29). Dynamic light scattering (DLS). DLS was performed on a 90 Plus Particle Size Analyzer (Brookhaven Instruments Corporation). Scattered light was detected at 90 o at room temperature and collected on an autocorrelator. The data for each sample was obtained in five measurements and the average number was used. The same two groups of micelles as in AFM studies were used for DLS characterization. The sizes of these micelles are 20.9 ± 1.7 and 24.4 ± 2.7 nm for crgd-free and 76% crgd micelles, respectively. The results were shown in Figure 2B and 2D. Flow cytometry analysis. SLK cells were seeded at 125,000 cells /well in 6-well plates in 2 ml DMEM with 10% FBS. After 24 h, 1 mg of micelles (from 3.3 mg/ml micelle suspension) for each micelle formulation with different crgd density (0, 5, 16, 76 %crgd) was added into each well and incubated at 37 C for 2 h. Then, cells were washed, trypsinized and neutralized. After centrifugation at 1200 rpm for 5 min, cells were re-suspended in 1 ml PBS, followed by filtration and analysis using flow cytometry. Cell uptake was found to increase up to 30-fold with 76% crgd-dx-micelles compared to those not attached with crgd (0% crgd). In the control experiment, SLK cells were first incubated with a free blocking peptide, Ala-Ala-Arg-Gly-Asp-Tyr (AARGDY), and then co-incubated with 76% crgd-functionalized micelles. Almost 100% inhibition by AARGDY at 9 mm concentration was observed as demonstrated by the flow cytometry histograms (Fig. S3). 6

8 A B C Figure S3. Flow cytometry histogram of micelle uptake in SLK tumor endothelial cells as a function of crgd density (A) 0% (B) 76% on the micelle surface. (C) Cell uptake of 76% crgd-micelles is inhibited by the presence of free RGD ligands (9 mm) in solution. Confocal laser scanning microscopy (CLSM). DX-micelles with 0 and 16% crgd density (0.5 mg/well) were incubated with SLK cells (6000 cells/well) culture wells for 2 h. Before the CLSM examination, cell nuclei were stained with Hoechst (Molecular Probes, Inc.). Cells were examined by a Zeiss LSM 510 microscope (Zurich, Switzerland, laser: Ar nm, Ar nm) with a confocal plane of 300 nm. Doxorubicin and Hoechst were excited at 485 and 352 nm, respectively. The emission wavelength of Doxorubicin and Hoechst are 595 and 455 nm, 7

9 respectively. A significantly increased amount of micelle uptake was observed in micelles with 16% crgd surface density (Fig. 3C) compared to those without crgd (Fig. 3B). This result is consistent with those from flow cytometry studies. 8