Impact of Dendrimer Surface Functional Groups on the Release of

Jan 27, 2014 - István Bányai,*. ,‡. Yun Zheng,. †. Mian Cao,. †. Xueyan Cao,. † and Xiangyang Shi*. ,†,§. †. College of Chemistry, Chem...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/JPCB

Impact of Dendrimer Surface Functional Groups on the Release of Doxorubicin from Dendrimer Carriers Mengen Zhang,† Rui Guo,† Mónika Kéri,‡ István Bányai,*,‡ Yun Zheng,† Mian Cao,† Xueyan Cao,† and Xiangyang Shi*,†,§ †

College of Chemistry, Chemical Engineering and Biotechnology, Donghua University, Shanghai 201620, People’s Republic of China Department of Colloid and Environmental Chemistry, Faculty of Science, University of Debrecen, H4032 Egyetem t.1, Debrecen, Hungary § CQM-Centro de Química da Madeira, Universidade da Madeira, Campus da Penteada, 9000-390 Funchal, Portugal ‡

S Supporting Information *

ABSTRACT: Generation 5 (G5) poly(amidoamine) dendrimers with acetyl (G5.NHAc), glycidol hydroxyl (G5.NGlyOH), and succinamic acid (G5.SAH) terminal groups were used to physically encapsulate an anticancer drug doxorubicin (DOX). Both UV−vis spectroscopy and multiple NMR techniques including one-dimensional NMR and two-dimensional NMR were applied to investigate the interactions between different dendrimers and DOX. The influence of the surface functional groups of G5 dendrimers on the DOX encapsulation, release kinetics, and cancer cell inhibition effect was investigated. We show that all three types of dendrimers are able to effectively encapsulate DOX and display therapeutic inhibition effect to cancer cells, which is solely associated with the loaded DOX. The relatively stronger interactions of G5.NHAc or G5.NGlyOH dendrimers with DOX than that of G5.SAH dendrimers with DOX demonstrated by NMR techniques correlate well with the slow release rate of DOX from G5.NHAc/DOX or G5.NGlyOH/DOX complexes. In contrast, the demonstrated weak interaction between G5.SAH and DOX causes a fast release of DOX, suggesting that the G5.SAH/DOX complex may not be a proper option for further in vivo research. Our findings suggest that the dendrimer surface functional groups are crucial for further design of multifunctional dendrimer-based drug delivery systems for various biomedical applications.



also lead to hemolysis even at low concentrations.32,33 To eliminate the terminal amine-induced cytotoxicity,34,35 the dendrimer surface primary amines have been partially or fully modified via PEGylation (PEG refers to poly(ethylene glycol)),11,36,37 acetylation,8,25,38−40 hydroxylation,39,41 or carboxylation.33,42 Both in vitro and in vivo experiments demonstrate that the biocompatibility of the modified dendrimers has been improved significantly. More importantly, the terminal groups of dendrimers also significantly impact the drug encapsulation efficiency, drug release property, and the biological activity of the drugs in vitro or in vivo.17,18,43,44 For example, Yang et al. compared the drug loading and release property of fully acetylated generation 5 (G5) PAMAM dendrimers with that of cationic G5 dendrimers using dexamethasone 21-phosphate (Dp21, an amphiphilic drug) as a model drug. They demonstrated that cationic dendrimers could encapsulate Dp21 under both acidic and neutral conditions, while acetylated dendrimers only formed inclusion complexes with Dp21 at acidic conditions with an increased aqueous solubility and better biocompatibility.45 In our previous work,44 we have shown that G5 PAMAM dendrimers

INTRODUCTION Poly(amidoamine) (PAMAM) dendrimers are a family of highly branched, monodispersed, synthetic macromolecules with well-defined structure and composition.1,2 Compared with traditional polymers, PAMAM dendrimers of larger generations possess a uniform spherical shape with a tunable nanometer size, which provides obvious advantages in various biomedical applications, including but not limited to gene delivery,3−6 molecular imaging,7−13 and drug delivery.14−22 Especially as a nanoscale drug delivery carrier, dendrimers are able to encapsulate hydrophobic anticancer drugs in their internal hydrophobic cavities and able to release the drug in a sustained manner.7,23,24 Meanwhile, drug molecules are also able to be covalently conjugated onto the dendrimer surface for delivery under certain circumstances.25−27 With the high density of functional groups on the dendrimer surface, dendrimers can be covalently conjugated with targeting ligands,28,29 imaging agents,26,30 or functional groups25,31 to form a multifunctional nanoplatform for various nanomedicinal applications. In general, before the design of a dendrimeric nanodevice, it is important to modify the dendrimer terminal amines with proper functional groups to eliminate the possible cytotoxicity caused by the amine termini. It has been known that cationic PAMAM dendrimers are not only able to induce cell membrane damage and result in a leakage of intracellular components but © 2014 American Chemical Society

Received: November 27, 2013 Revised: January 24, 2014 Published: January 27, 2014 1696

dx.doi.org/10.1021/jp411669k | J. Phys. Chem. B 2014, 118, 1696−1706

The Journal of Physical Chemistry B

Article

Scheme 1. Schematic Representation of the Synthesis of G5 Dendrimers with Different Terminal Functional Groups and the Molecular Structure of DOX·HCl

and co-workers deduced the existence of hydrogen bonds between drug molecules and dendrimers from the shifts in the UV−vis spectra of oligoethylene glycol-terminated PAMAM dendrimer/doxorubicin (DOX) complexes.49 Among all these methods, NMR techniques, including 1H and 13C NMR titrations, two-dimensional nuclear Overhauser enhancement spectroscopy (2D-NOESY), and diffusion-ordered spectroscopy (DOSY), have been demonstrated to be effective in studying the dendrimer host−drug guest interactions.50−52 For instance, Cheng and co-workers investigated the host−guest chemistry of dendrimer−drug complexes by means of NMR techniques using PAMAM dendrimers surface functionalized with positive or negative charges as hosts and four acidic drugs as guests. They concluded that cationic guests could only attach on the surface of anionic dendrimers via electrostatic interaction, and anionic guests were able to distribute both on the surface and in the interior cavities of the cationic dendrimers.48 DOX is an anticancer drug widely used in cancer chemotherapy. To improve the water solubility and bioavailability of DOX, dendrimers have been actively applied as a platform for targeted delivery of DOX.23,53−55 For instance,

with acetyl, carboxyl, and hydroxyl groups are able to encapsulate a potential anticancer drug 2-methoxyestradiol (2-ME) with approximately similar loading capacity. However, the 2-ME drug loaded within the carboxyl-terminated G5 dendrimers is unable to exert its therapeutic efficacy in vitro, which is believed to be due to the strong interaction between the dendrimer terminal succinamic acid groups and the dendrimer internal tertiary amines, limiting the effective release of the 2-ME drug. These studies highlight the importance to investigate the interactions between the drug guest and the dendrimer host. However, systematic studies of the interactions between dendrimers with different terminal groups and drug molecules and their influence on the property of drug delivery system have been seldom reported. To better understand the influence of dendrimer terminal groups on the properties of dendrimer-based drug delivery systems, the interaction between drug molecules and dendrimers has been investigated via various techniques, such as UV−vis spectroscopy,20,23 electron paramagnetic resonance,46 high performance liquid chromatography,47 and nuclear magnetic resonance (NMR).48 For example, Chandra 1697

dx.doi.org/10.1021/jp411669k | J. Phys. Chem. B 2014, 118, 1696−1706

The Journal of Physical Chemistry B

Article

hydrophobic DOX, followed by mixing with the 1.5 mL dendrimer aqueous solution. The mixture solution was vigorously stirred overnight to allow the evaporation of the methanol solvent. The dendrimer/DOX mixture solution was centrifuged (7000 rpm for 10 min) to remove the precipitate related to the noncomplexed free DOX, which is insoluble in water. The precipitate was collected and dissolved in 1 mL of methanol for UV−vis analysis. The supernatant was then lyophilized for 3 days to obtain the dendrimer/DOX complexes. Free DOX·HCl, dendrimers, and dendrimer/DOX complexes were characterized by a Lambda 25 UV−vis spectrophotometer (PerkinElmer). NMR Studies. For NMR measurements, all samples were dissolved in D2O. One- and two-dimensional (NOESY, DOSY) NMR experiments were carried out using Bruker DMX Avance III and DRX Avance I 400 MHz NMR spectrometers at Donghua University and at the University of Debrecen, respectively. In the case of G5.NHAc/DOX complex, 10 mg of G5.NHAc and 5 mg of DOX·HCl were codissolved within 1 mL of D2O, resulting in a sample of pH ≈ 3 (acidic sample). 0.5 μL of triethylamine (TEA) was added to 500 μL of acidic G5.NHAc/ DOX complex to get an alkaline G5.NHAc + DOX·HCl + TEA sample. 1H NMR and 2D NOESY and DOSY measurements of these systems were performed at Donghua University. 2DNOESY experiments were obtained with a mixing time of 500 ms and a relaxation delay of 2 s. Data were processed with cos2 window function in both f1 and f 2 dimensions according to the Bruker protocols. Each of the NOESY spectral data was processed with Mestre Nova software. DOSY experiments were performed with the following parameters: diffusion time (Δ) = 40 ms, gradient pulse (δ) = 8 ms, and relaxation delay = 3 s. DOSY spectral data were processed in 1D and 2D mode with Mestre Nova software. G5.NGlyOH and G5.SAH complexed with DOX were dissolved in D2O as well, in a concentration of approximately 10 mg/mL. 1H NMR and 2D NOESY and DOSY measurements were performed at the University of Debrecen. 2D NOESY measurements were carried out using the following parameters: mixing time = 150 and 300 ms for G5.NGlyOH/ DOX and G5.SAH/DOX, respectively; relaxation delay = 3 s. In DOSY experiments, we used a diffusion time (Δ) of 50 ms, gradient pulse (δ) of 10 ms, and relaxation delay of 4 s. Data processing was carried out in the same way as mentioned above. Kinetic Study of in Vitro Drug Release. Free DOX·HCl, G5.NHAc/DOX, G5.NGlyOH/DOX, or G5.SAH/DOX complex in water (1 mL) was sealed in a dialysis bag (MWCO of 10 000) and immersed in 20 mL of phosphate buffered saline (PBS) solution (pH 7.4, 0.2 M) or citrate buffer solution (pH 5.5, 0.2 M). The entire system was incubated in a vapor-bathing constant temperature vibrator at 37 °C. 3 mL of the buffer medium was taken out at each predetermined time interval and measured by UV−vis spectrophotometer. The volume of the outer aqueous phase was maintained constant by replenishing 3 mL of the corresponding buffer solution. Cell Biological Evaluation. KB cells were continuously grown in RPMI 1640 medium supplemented with 10% heatinactivated FBS, 100 U/mL penicillin, and 100 U/mL streptomycin at 37 °C and 5% CO2. Cells (1 × 104 cells per well) were seeded in 96-well culture plate one day before the experiment. DOX·HCl (5 μM) in PBS (10 μL) and three types of dendrimer/DOX complexes with DOX concentration of 5

Choi et al. used folate-modified carboxylic acid-terminated G5 PAMAM dendrimers to covalently conjugate photocaged DOX for targeted delivery of DOX to folate receptor-expressing cancer cells.55 Han et al. conjugated HAIYPRH peptide on the surface of G5 dendrimers through a PEG linker and demonstrated the use of this dendrimer-based carrier for effective DOX loading and targeted delivery of DOX to effectively inhibit tumor growth.54 To develop various multifunctional dendrimer-based carrier systems for targeted delivery of DOX, it is crucial to understand the influence of dendrimer surface functional groups on the drug encapsulation, release kinetics, and the therapeutic efficacy. In this study, we synthesized acetyl-terminated (G5.NHAc), glycidol hydroxyl-terminated (G5.NGlyOH), and carboxylterminated (G5.SAH) dendrimers for DOX loading and release. The host−guest interactions of the G5.NHAc/DOX, G5.NGlyOH/DOX, and G5.SAH/DOX complexes were carefully investigated by a combination of UV−vis spectroscopy, 1H NMR, 2D-NOESY, and DOSY techniques. The influence of dendrimer surface modification on the in vitro release kinetics was systemically studied at different pHs, and the therapeutic efficacy of the three kinds of dendrimer/DOX complexes against KB cells (a human epithelial carcinoma cell line) was explored by cell viability assay and cell morphology observation. To the best of our knowledge, this is the first study related to a systematical investigation of the influence of dendrimer surface functional groups on the interaction/encapsulation, release kinetics, and therapeutic effect of DOX complexed within the dendrimers.



EXPERIMENTAL SECTION Materials. Ethylenediamine core amine-terminated G5 PAMAM dendrimers (G5.NH2) were purchased from Dendritech (Midland, MI). The terminal amines of G5.NH2 dendrimers were reacted with acetic anhydride, glycidol, and succinic anhydride to form acetylated (G5.NHAc), hydroxylated (G5.NGlyOH), and carboxylated (G5.SAH) dendrimers, respectively (Scheme 1), according to procedures reported in our previous work.44,56 Doxorubicin hydrochloride (DOX·HCl, molecular structure shown in Scheme 1) was purchased from Beijing Huafeng Pharmaceutical Co., Ltd. (Beijing, China). KB cells were obtained from Institute of Biochemistry and Cell Biology, the Chinese Academy of Sciences (Shanghai, China). Roswell Park Memorial Institute-1640 (RPMI-1640), fetal bovine serum (FBS), penicillin, and streptomycin were purchased from Hangzhou Jinuo Biomedical Technology (Hangzhou, China). 3-[4,5-Dimethyl-2-thiazolyl]-2,5-diphenyl2H-tetrazolium bromide (MTT) and D2O were purchased from Sigma-Aldrich. The water used in all experiments was purified through Milli-Q Plus 185 water purification system (Millipore, Bedford, MA) with a resistivity higher than 18 mΩ cm. All other solvents and chemicals were of reagent grade and purchased from Sinopharm Chemical Reagent Co., Ltd. (China). Regenerated cellulose dialysis membranes (molecular weight cutoff, MWCO = 10 000) were acquired from Fisher. Preparation of Dendrimer/DOX Complexes. The complexation of dendrimers with DOX was carried out according to a method described in our previous work.23 Briefly, G5 dendrimers (10 mg) with different terminal groups (acetyl, hydroxyl, and carboxyl) were dissolved in water (1.5 mL). Doxorubicin hydrochloride (DOX·HCl) with 10 mol equiv of each dendrimer or derivative dissolved in 300 μL of methanol was neutralized with triethylamine (5 μL) to generate 1698

dx.doi.org/10.1021/jp411669k | J. Phys. Chem. B 2014, 118, 1696−1706

The Journal of Physical Chemistry B

Article

μM were added to cells and then incubated for 48 h at 37 °C before MTT assay. After 48 h incubation with free DOX·HCl or dendrimer/DOX complexes, MTT was added to each well, and the cells were incubated for an additional 4 h. Then, the medium was replaced with DMSO, and the plates were recorded at 570 nm using a microplate reader (MK3, Thermo Inc.). Mean and standard deviation for the triplicate wells for each sample were reported. After treatment with DOX·HCl or dendrimer/DOX complexes for 48 h, the cell morphology was also observed by phase-contrast microscopy (Leica DM IL LED inverted phase contrast microscope). The magnification was set at 200× for all samples. Statistical Analysis. One way ANOVA statistical analysis was performed to compare the cell viability upon treatment with dendrimers, dendrimer/DOX complexes, or free DOX· HCl. 0.05 was selected as the significance level, and the data were indicated with (∗) for p < 0.05, (∗∗) for p < 0.01, and (∗∗∗) for p < 0.001, respectively.



RESULTS AND DISCUSSION Preparation of Dendrimer/DOX Complexes. In our previous study, we complexed a potential anticancer agent 2Table 1. Complexation Capacity of DOX with G5 Dendrimers Having Different Terminal Groups

a

dendrimers

G5.NHAc

G5.NGlyOH

G5.SAH

Mn complexation capacitya

30 990 4.5 ± 1.1

38 382 5.2 ± 1.3

40 330 9.5 ± 0.7

Figure 2. 1H NMR spectra of free DOX·HCl (a), DOX·HCl in the presence of G5.NHAc (b), and DOX·HCl in the presence of G5.NHAc and TEA (c). All samples were dissolved in D2O. The DOX concentration in each group is 5 mg/mL.

Number of drug molecules per dendrimer.

complexes for cancer therapy applications. The host−guest interactions between dendrimers and DOX for the complexes were investigated by UV−vis spectroscopy and NMR techniques. During the preparation of dendrimer/DOX complexes, water-soluble DOX·HCl was first neutralized by TEA to form hydrophobic DOX, which can be efficiently encapsulated within the interiors of dendrimers via hydrophobic interaction or electrostatic interaction. Similar to our previous study,23 DOX was able to be successfully complexed with dendrimers by this method regardless of the type of the dendrimer terminal groups. The payloads of DOX within the dendrimers were evaluated by UV−vis spectroscopic analysis and are presented in Table 1. It can be seen that G5.NHAc and G5.NGlyOH have a similar drug loading capacity and are able to encapsulate approximately 5 DOX molecules per dendrimer. In contrast, G5.SAH dendrimer has the highest loading capacity and can load 9.5 DOX molecules per dendrimer. This result indicates that the terminal functional groups of dendrimers greatly impact the loading capacity of dendrimers, which may be due to the different interactions between the dendrimers and the drug molecules. The lyophilized powders of the formed dendrimer/ DOX complexes were able to be redispersed in water and were colloidally stable for at least 2 months at room temperature, similar to the corresponding dendrimers without DOX complexation (Figure S1, Supporting Information). UV−vis Spectroscopy. UV−vis spectroscopy has been demonstrated to be a useful and simple method to verify the loading of DOX and analyze the type of interactions between drug molecules and dendrimers.20,49 Therefore, in this study, the UV−vis spectra of dendrimers and dendrimer/DOX

Figure 1. UV−vis spectra of free DOX.HCl (0.16 mg/mL), G5.NHAc, G5.NGlyOH, and G5.SAH dendrimers and G5.NHAc/DOX, G5.NGlyOH/DOX, and G5.SAH/DOX complexes dissolved in water. The concentration of each dendrimer or dendrimer/DOX complex is 1 mg/mL.

ME with G5 dendrimers having different surface functional groups and demonstrated that the surface charges of dendrimers were crucial for the complexed 2-ME to exert its bioactivity.44 Molecular dynamics simulations of the dendrimer/2-ME complexes were applied to validate the experimental results. In this work, to evaluate the influence of the terminal groups of dendrimers on the loading capacity, release kinetics, and the bioactivity of the anticancer drug DOX, we synthesized G5.NHAc, G5.NGlyOH, and G5.SAH dendrimers (Scheme 1) and prepared different dendrimer/DOX 1699

dx.doi.org/10.1021/jp411669k | J. Phys. Chem. B 2014, 118, 1696−1706

The Journal of Physical Chemistry B

Article

Figure 3. 1H−1H NOESY spectra of G5.NHAc/DOX·HCl/TEA/D2O at a mixing time of 500 ms. The sample was prepared by dissolving 5 mg of G5.NHAc dendrimer in 0.5 mL of D2O, followed by mixing with 2.5 mg of DOX·HCl and 0.5 μL of TEA.

complexes were measured (Figure 1). Compared with the spectra of pure dendrimers, the spectra of all dendrimer/DOX complexes showed an absorption peak at 481 nm, which is associated with the typical absorption feature of DOX. This demonstrates that DOX has been loaded in the different kinds of dendrimers. It appears that the encapsulation of DOX within G5.NHAc or G5.NGlyOH dendrimers leads to slight blueshifts of the typical DOX absorption bands at 275 and 257 nm (n/p* absorption of the carbonyl groups of DOX), indicative of the existence of hydrogen bonds between the carbonyl groups of DOX (hydrogen acceptors) and −NH groups (hydrogen donors) of the dendrimers, in agreement with the results reported by Chandra and co-workers.49 Therefore, hydrogenbonding interaction between dendrimers and DOX may contribute to the encapsulation of DOX within these two types of dendrimers. As opposed to the G5.NHAc/DOX or G5.NGlyOH/DOX complexes, G5.SAH/DOX complex did not have apparent blue-shifts at the absorption bands of 275 and 257 nm when compared with free DOX·HCl. This may be because the electrostatic interaction between the surface carboxyl groups of G5.SAH dendrimers and the DOX amine groups contributes to the loading of the DOX drug. The larger drug loading capacity of G5.SAH than that of G5.NHAc or G5.NGlyOH dendrimers could be due to the numerous carboxyl groups on the periphery of the dendrimers, providing more interaction sites for DOX molecules. Likewise, the electrostatic interaction between DOX and G5.SAH may be

destroyed easily in ionic solution, and hence it is possible for G5.SAH/DOX complex to release DOX drug in a quick manner under physiological conditions. It should be noted that the slight absorption feature at 290 nm for both G5.NHAc and G5.NGlyOH dendrimers may be associated with a slight oxidation of dendrimers during the processing and storage, in agreement with the literature.57 1 H NMR Spectroscopy. NMR techniques have been proven to be a useful method to detect molecular interactions between the host and guest molecules.51 G5.NHAc/DOX complex was selected as a model. 1H NMR experiments were performed to detect the complex formation between the G5.NHAc dendrimer and DOX·HCl in the absence or in the presence of TEA in D2O (Figure 2). Compared with the 1H NMR spectrum of DOX·HCl (Figure 2a), the peaks of DOX protons show no obvious changes after mixing with G5.NHAc dendrimer in an acid circumstance (Figure 2b). However, after the addition of TEA, line broadening of the DOX proton signals is observed due to their decreased degree of freedom (Figure 2c), indicating that the neutralization of DOX·HCl by TEA leads to an enhanced interaction between the G5.NHAc and DOX. Moreover, the peaks of dendrimer protons shifted upfield in the spectrum of G5.NHAc + DOX·HCl + TEA (Figure 2c), indicative of the hydrophobic and hydrogenbonding interactions between the relative nonpolar pockets of dendrimers and the drug molecules, corroborating the UV−vis spectroscopic results. 1700

dx.doi.org/10.1021/jp411669k | J. Phys. Chem. B 2014, 118, 1696−1706

The Journal of Physical Chemistry B

Article

Figure 4. A representative pseudo-2D DOSY plot for (a) G5.NGlyOH/DOX complex (Ddendrimer = 4.9 × 10−11 m2/s, DDOX = 1.8 × 10−10 m2/s) and (b) G5.SAH/DOX complex (Ddendrimer = 5.62 × 10−11 m2/s, DDOX = 1.7 × 10−10 m2/s).

2D NMR Spectroscopy. In order to locate the specific structural units involved in the interaction between G5.NHAc and DOX, 2D-NOESY that can be used to detect protons spatially close to each other was performed. Figure 3 shows the NOESY spectrum of G5.NHAc/DOX complex in the presence of TEA. Different from a previous report dealing with dendrimers complexed with other drug molecules,48 there are no cross-peaks between DOX and dendrimer protons, and only cross-peaks of their own protons are shown. The phase of the cross-peaks between DOX protons is similar to that of the diagonal peaks, which means that DOX is incorporated into the G5.NHAc dendrimers. The identical phase shows that DOX behaves as a part of the macromolecule. Therefore, the most probable explanation is that DOX has a reversible interaction with G5.NHAc. There are both free and coordinated DOX in the solution. The exchange between them is fast at the 1H NMR (chemical shift) time scale; therefore, we cannot see separate peaks of them. However, the rotation correlation time of coordinated DOX is slow due to its behavior as a macromolecule after the incorporation into the G5.NHAc

dendrimer, and consequently it gives the same phase cross peaks as the dendrimer (and the diagonal). The exchange process transfers these in phase NOESY cross-peaks to the free DOX, and it compensates for the opposite phase NOE crosspeaks of the small molecule of free DOX. Similar to the result of G5.NHAc/DOX, we did not find cross-peaks between DOX and G5.NGlyOH or G5.SAH dendrimers (Figure S2, Supporting Information). Evidence for the interaction between G5.NHAc and DOX was further investigated by the DOSY technique. DOSY is able to give information about the average diffusion coefficients of the components of multicomponent systems separately, which reflects the effective size and shape of the specific molecule. In this study, the average diffusion coefficient (D) of DOX in G5.NHAc/DOX complex was measured to be 1.54 × 10−10 m2/s, which is much lower than that of free DOX·HCl (DDOX = 1.95 × 10−10 m2/s). This means that some of the DOX molecules move together with the G5.NHAc dendrimer, thus having limited diffusion rate. This D value was even lower than that of DOX in G5.NHAc + DOX·HCl sample (1.7 × 10−10 1701

dx.doi.org/10.1021/jp411669k | J. Phys. Chem. B 2014, 118, 1696−1706

The Journal of Physical Chemistry B

Article

Figure 6. Viability of KB cells after treatment with free DOX·HCl (5 μM) in 10 μL of PBS, G5.NHAc/DOX, G5.NGlyOH/DOX, and G5.SAH/DOX complexes with DOX concentration of 5 μM and G5.NHAc, G5.NGlyOH, and G5.SAH dendrimers with concentrations similar to those of the corresponding dendrimers used to complex DOX (5 μM). KB cells treated with 10 μL of PBS were used as control. The data are expressed as mean ± SD.

In Vitro Release Kinetics. We next investigated the in vitro release kinetics of the dendrimer/DOX complexes in PBS (pH 7.4) and citrate buffer (pH 5.5) at 37 °C. Figure 5 shows the release profiles of free DOX·HCl and DOX from G5.NHAc/ DOX, G5.NGlyOH/DOX, and G5.SAH/DOX complexes. It can be seen that nearly 80% of DOX was rapidly released within 2 h for the free DOX group. In contrast, less than 45% of DOX is released from all complexes in the same time period. The release of DOX from the three kinds of complexes follows a biphasic pattern, which is characterized by an initial fast release followed by release in a sustained manner. Therefore, the dendrimer/DOX complexes are able to release the DOX drug in a relative slower and sustained manner under both pH conditions, in agreement with our previous work.24 Three kinds of dendrimer/DOX complexes displayed different drug release rates under different pH conditions. Under pH 7.4, 5.5%, 18.5%, and 36.9% DOX is able to be released from the G5.NHAc/DOX, G5.NGlyOH/DOX, and G5.SAH/DOX complexes, respectively, at 48 h. In contrast, under pH 5.5, the DOX release percentage of 39.5%, 60.9%, and 68.6% from G5.NHAc/DOX, G5.NGlyOH/DOX, and G5.SAH/DOX complexes, respectively, can be achieved. The apparently faster release rate of each dendrimer/DOX complex at pH 5.5 than that at pH 7.4 may be due to the fact that the positively charged DOX molecules at acidic environment could be repelled by the protonated dendrimer branches with positive charges under pH 5.5, speeding up the release of DOX drug from the hydrophobic dendrimer interiors. For the case of G5.SAH/DOX complex, the major force driving the fast release of DOX could be due to the increased hydrophilicity of the DOX drug and the protonation of the dendrimer surface carboxyl groups. In addition, the release rate of DOX from dendrimer/drug complexes follows the order of G5.SAH/DOX > G5.NGlyOH/DOX > G5.NHAc/DOX. In agreement with the order of dendrimer/drug interaction at G5.NHAc/DOX > G5.NGlyOH/DOX > G5.SAH/DOX demonstrated by NMR studies, the slowest DOX release from G5.NHAc/DOX results

Figure 5. In vitro release profiles of DOX from dendrimer/DOX complexes in (a) PBS (pH 7.4) and (b) citrate buffer (pH 5.5) at 37 °C. In (a), free DOX·HCl dissolved in PBS (pH 7.4) was also dialyzed and used as control.

m2/s), indicating that the interaction between DOX and G5.NHAc is much weaker in acidic environment. Moreover, the average diffusion coefficients of DOX in G5.NGlyOH/ DOX (1.8 × 10−10 m2/s) and G5.SAH/DOX (1.7 × 10−10 m2/ s) complexes were smaller than that of free drug, which means that some drug molecules move together with the corresponding G5.NGlyOH or G5.SAH dendrimers, but are larger than that of DOX in G5.NHAc/DOX complex, indicative of the much weaker drug−dendrimer interaction in G5.NGlyOH/ DOX and G5.SAH/DOX complexes. This assumption was also validated by the pseudo-2D DOSY plot of G5.NGlyOH/DOX and G5.SAH/DOX complexes shown in Figure 4. It should be noticed that some of the representative points of DOX are in the same line with those of the dendrimers, demonstrating that DOX is complexed with the dendrimers and moves together with them (Figure 4a). More importantly, there are some of representative points of free DOX in the G5.SAH/DOX plot (Figure 4b). This could be ascribed to the released DOX from the complexes in the D2O solution within several minutes and also indicates the weak interaction between G5.SAH dendrimers and drug molecules. Therefore, according to the DOSY results, we can conclude that the interaction intensity between dendrimer and DOX follows the order of G5.NHAc/ DOX > G5.NGlyOH/DOX > G5.SAH/DOX. 1702

dx.doi.org/10.1021/jp411669k | J. Phys. Chem. B 2014, 118, 1696−1706

The Journal of Physical Chemistry B

Article

Figure 7. Phase contrast microscopy images of (a) control KB cells treated with 10 μL of PBS and KB cells treated with (b) G5.NHAc, (c) G5.NHAc/DOX complex (5 μM DOX), (d) free DOX·HCl (5 μM), (e) G5.NGlyOH dendrimers, (f) G5.NGlyOH/DOX complex (5 μM DOX), (g) G5.SAH dendrimers, and (h) G5.SAH/DOX complex (5 μM DOX) in 10 μL of PBS. The concentrations of dendrimers are similar to those of the dendrimers used to complex DOX (5 μM). The scale bar shown in each panel represents 20 μm.

of the corresponding complexes loaded with 5 μM DOX. Apparently, the viabilities of cells treated with G5.NHAc, G5.NGlyOH, and G5.SAH dendrimers are all over 95%, indicating that these dendrimers have no cytotoxicity (p > 0.05 for each), in agreement with our previous results.44 Our results suggest that the therapeutic activity of dendrimer/DOX complexes is solely related to the loaded DOX drug. The cytotoxic effect of dendrimer/DOX complexes was further confirmed by phase contrast microscopic visualization of the cell morphology. Figure 7 shows the morphology of the KB cells treated with PBS, free DOX·HCl, dendrimer carriers, and dendrimer/DOX complexes. It is clear that free DOX·HCl (Figure 7d), G5.NHAc/DOX (Figure 7c), G5.NGlyOH/DOX (Figure 7f), and G5.SAH/DOX (Figure 7h) complexes with similar DOX concentration (5 μM) induce similar cell morphology changes. A significant portion of the cells became detached from the bottom of the plate and were rounded in shape, indicating the cell death. In contrast, almost no detached and rounded cells can be observed in control cells treated with PBS (Figure 7a) and cells treated with the G5 dendrimers without DOX complexation, demonstrating that G5.NGlyOH (Figure 7b), G5.NHAc (Figure 7e), and G5.SAH (Figure 7g) vehicles are noncytotoxic at the given concentration. These results suggest that the bioactivity of dendrimer/DOX complexes is solely related to the loaded DOX. The qualitative cell morphology observation data corroborate the MTT assay results. In this study, dendrimer modified with different surface groups were applied to encapsulate DOX in order to improve the water solubility of the drug, release the drug in a sustained manner, and improve the bioavailability of the drug for potential in vivo applications. Although the used G5.SAH dendrimers have the highest drug loading capacity and exhibit

from the strongest interaction between DOX and G5.NHAc, and the fastest DOX release from G5.SAH/DOX is due to the weak electrostatic interaction between peripheral carboxyl groups of G5.SAH dendrimer and amine groups of DOX. Taken together, our NMR results correlate very well with the release kinetics of DOX from different dendrimer/DOX complexes. In Vitro Therapeutic Efficacy of Dendrimer/DOX Complexes. KB cells were chosen to assess the therapeutic efficacy of dendrimer/DOX complexes. We first tested the dose-dependent cytotoxicity of free DOX·HCl and dendrimer/ DOX complexes (Figure S3, Supporting Information). It can be seen that free drug shows an apparent cytotoxicity at the concentration as low as 0.5 μM, whereas the dendrimer/DOX complexes start to exert the therapeutic activity of DOX at around 1 μM. This should be due to the limited amount of DOX released from the complexes, resulting in a lower bioavailability of the drug. To ensure the therapeutic efficacy of the DOX drug, the concentration of DOX for all dendrimer/ DOX complexes was set at 5 μM in order to compare the cytotoxcity of the complexes and the dendrimers without drug. The MTT assay data are shown in Figure 6. It is clear that similar to the free DOX·HCl, G5.NHAc/DOX, G5.NGlyOH/ DOX, and G5.SAH/DOX complexes at the same DOX concentration of 5 μM cause a significant loss of cell viability when compared with control cells treated with 10 μL of PBS (p value