RESEARCH ARTICLE – Pharmaceutical Nanotechnology
Fabrication of Biocompatible and Tumor-Targeting Hyaluronan Nanospheres by a Modified Desolvation Method WENYI ZHENG, YONGQUAN LI, JINPING DU, ZONGNING YIN Key Laboratory of Drug Targeting and Drug Delivery Systems, West China School of Pharmacy, Sichuan University, Chengdu, Sichuan Province, China Received 25 November 2013; revised 23 January 2014; accepted 18 February 2014 Published online 6 March 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.23924 ABSTRACT: The aim of this work was to maximize the tumor targetability of biocompatible hyaluronan (HA) by construction of a novel nanocarrier, using HA as the single material. HA was prefunctionalized with active amino groups, desolvated by acetone, and crosslinked by glutaraldehyde. The process was further optimized with regard to yield, stability, and particle size. The cytotoxicity of HA nanospheres (HA-NPs) was evaluated by thiazolyl blue tetrazolium bromide and reactive oxygen species assays. A549 cells and H22bearing Kunming mice were employed to characterize the tumor targeting of fluorescein isothiocyanate-conjugated HA-NPs. Nanospheres prepared according to the optimal formulation were characterized by a maximum yield (90%), high stability over 7 days, regular spheres (97.42 nm in diameter), and negative charge (−32.7 mV in zeta potential). In vitro results revealed that HA-NPs had little cytotoxicity and efficiently accumulated into A549 cells in a HA-dependent manner. Following systemic administration in mice, HA-NPs selectively accumulated in the tumor as demonstrated by the frozen section examination and flow cytometry analysis. In conclusion, this work C 2014 Wiley successfully prepared HA-NPs and explored their potential applications for tumor targeting in terms of safety and efficacy. Periodicals, Inc. and the American Pharmacists Association J Pharm Sci 103:1529–1537, 2014 Keywords: hyaluronan; CD44; cancer; desolvation; drug targeting; nanospheres; polymeric drug delivery systems
INTRODUCTION Hyaluronan (HA) is a linear, negatively charged polysaccharide, which is composed of repetitive disaccharide units. It is involved in the regulation of several key biological processes by binding to its endogenous receptors including clusters of differentiation 44 (CD44), HA-mediated motility receptor, lymphatic vessel endothelial HA receptor-1, and HA receptor for endocytosis.1 Among these receptors, CD44 is reported to be expressed highly on the surface of tumor cells, especially the highly invasive tumor cells and stem cells, a discovery that is widely utilized in specific targeting therapy.2,3 In the field of pharmaceutics, delivery systems based on HA– CD44 interactions have been developed to facilitate drug/gene delivery. More specifically, HA-based polymeric prodrugs account for a large proportion of the existing delivery systems. Successful examples include doxorubicin, epirubicin, butyrate, paclitaxel, fluorouracil, and camptothecin.4–9 In spite of the enhanced therapeutic efficacy, this strategy exhibits some drawbacks, a main one of which is the rapid clearance rate of HA in the blood circulation.10 Various nanoparticulate systems for CD44 targeting are available at present. Thanks to the abundant carboxyl and hydroxyl groups, HA could be readily modified on lipid-based nanoparticle, FeO particles, or liposome.11–13 Alternative systems include micelles constructed by introducing hydrophobic segments (i.e., fatty amine, histidine, deoxycholic acid, cholanic acid, and ceramide)14–18 to the HA backbone, and nanocomplexs formed between positively charged chitosan, polyetherimide, or platinum and negatively charged HA.19–21 Correspondence to: Zongning Yin (Telephone: +86-28-85502917; Fax: +8628-85502917; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 103, 1529–1537 (2014)
C 2014 Wiley Periodicals, Inc. and the American Pharmacists Association
It is clear that HA, the targeting molecule, only accounts for a small part within these nanoparticulate systems. Qhattal and Liu13 concluded that the uptake efficiency of HA-grafted liposomes in CD44-overexpressing tumor cells rises with increasing grafting density. It is therefore reasonable to assume that nanospheres consisting entirely of HA would have the strongest targetability. In addition, because of the continuous distribution of HA from the interior to the exterior of the nanospheres, HA-NPs would remain its targetability during the course of degradation, providing an advantage over conventional carriers. Another advantage of HA-NPs is associated with their inherent hydrophilicity, which minimizes serum protein binding and subsequent reticuloendothelial uptake. To the best of our knowledge, there were few literatures reporting successful preparation of HA-NPs without using composite materials. In this study, we aimed to fill this gap by developing HA-NPs via a modified desolvation method. HA-NPs were further optimized in terms of yield, stability, and particle size, and evaluated for their cytotoxicity as well as targetability.
MATERIALS AND METHODS Materials Hyaluronan with molecular weights (MW) of 101 and 8 kDa were purchased from Dali (Liuzhou, Guangxi, China), whereas HA with MW of 1.44 mDa was from Freda (Jinan, Shandong, China). Thiazolyl blue tetrazolium bromide (MTT), trypan blue (TB), and sodium 3-nitrobenzenesulfonate (TNBS) were obtained from Sigma–Aldrich (St. Louis, Missouri, USA). Fluorescein isothiocyanate (FITC) and adipic dihydrazide (ADH) were purchased from Aladdin (Shanghai, China). Reactive oxygen species (ROS) assay kit was purchased from Beyotime Institute of Biotechnology (Haimen, Jiangsu, China).
Zheng et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:1529–1537, 2014
1529
1530
RESEARCH ARTICLE – Pharmaceutical Nanotechnology
Table 1.
Formulations Used in the Optimization Study of HA-NPs
Formulation A B C D E
Molecular Weight HA–ADH Acetone Glutaraldehyde of HA (kDa) (mg) (mL) (:mol) 101 101 101 101 /
5 5 5 / 5
/ 3.5 / 3.5 3.5
75 / 75 75 75
Formulation A and B were used to evaluate the effect of acetone on nondesolvated HA–ADH and the effect of glutaraldehyde on free amino groups. Formulations C, D, and E were used to assess the effect of acetone, HA–ADH concentration, and MW of HA on particle size, respectively. / means variable.
Hyclone (Logan, Utah, USA). Adenocarcinomic human alveolar basal epithelial cells (A549) and murine hepatocellular carcinoma (H22) cells were provided by the Key Laboratory of Drug Targeting and Drug Delivery Systems. Derivatization of HA Adipic dihydrazide (391.5 mg, 2.02 mmol) was added to a solution of HA (180 mg) in water (45 mL). After adjusting the pH of the reaction mixture to 4.75 with HCl solution, the reaction was initiated by adding 1-ethyl-3-[3-(dimethylamino)propyl] carbodiimide (86.4 mg, 0.44 mmol) under stirring at room temperature. The reaction was stopped 2 h later by adjusting the pH to 7.0. The mixture was exhaustively dialyzed against 0.1 M NaCl solution, 25% ethanol, and pure water, and then lyophilized to give HA–ADH.22 1 H-NMR (Unity Inova 400, Varian (Palo Alto, California, USA), 400 MHz, D2 O, *, ppm): 2.235–2.207 (s, 4H, –COCH2 ), 2.005–1.967 (t, 3H, –NCOCH3 ), 1.638 (s, 4H, –COCH2 CH2 ). The substitution degree of different conjugates was estimated to vary from 12.6% to 22.4% according to the 1 H-NMR spectrum. Preparation and Characterization of HA-NP
Figure 1. Percentage of nondesolvated HA–ADH in correlation with acetone volume added during the desolvation procedure.
4 ,6-Diamidino-2-phenylindole) (DAPI), collagenase type IV, and DNase I were purchased from Biosharp (Hefei, Anhui, China). Sephadex G50 was purchased from Pharmacia (Uppsala, Sweden). Roswell Park Memorial Institute (RPMI-1640) medium and fetal bovine serum (FBS) were purchased from
Hyaluronan–ADH was dissolved in 1 mL water and desolvated by adding acetone dropwise under stirring. After addition of glutaraldehyde, the mixture was subjected to reaction for 12 h. HA-NPs were obtained after removal of acetone. Further purification was performed on Sephadex G50 for cell experiment and in vivo study. The optimization was conducted according to different formulations as listed in Table 1. All experiments were performed in triplicate. The average particle size, polydispersity index (PDI), zeta potential, and count rate were measured using a dynamic light scattering analyzer (Zetasizer Nano ZS90, Malvern, UK). Before scanning electron microscope (SEM, JSM-6510LV, Jeol, Tokyo, Japan) and transmission electron microscopy (TEM, H6001V, Hitachi, Tokyo, Japan) observation, nanospheres were coated with gold and stained with phosphotungstic acid. In order to determine the cellular uptake by flow cytometry (FC500, Beckman Coulter, Brea, California, USA),
Figure 2. (a) Influence of the amount of glutaraldehyde on free amino groups of HA-NPs. (b) Diameter and size distribution variation of HA-NPs cross-linked with 75 :mol glutaraldehyde over 7 days after incubation with PBS at 4◦ C. Zheng et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:1529–1537, 2014
DOI 10.1002/jps.23924
RESEARCH ARTICLE – Pharmaceutical Nanotechnology
1531
mined by fluorescence spectrometry at Ex /Em wavelengths of 490/520 nm. Determination of Nondesolvated HA For the determination of the percentage of nondesolvated HA– ADH, the nanospheres were separated from the supernatant by centrifugation at 16 738 g for 90 min (Allgera X22, Beckman Coulter, California, USA). HA–ADH in the supernatant was quantified by the turbidimetric method, which is based on the formation of HA–hexadecyltrimmethylammonium bromide (CTAB) complex.24 Briefly, 50 :L standard solution (0.3– 1.2 mg/mL) or supernatant was added to a 96-well plate, and incubated with 50 :L acetate buffer (0.2 M, pH 6.0) with 0.15 M NaCl for 20 min at 37◦ C. Then, 100 :L of 10 mM preincubated CTAB solution in 0.5 M NaOH was added, and the absorbance was measured within 10 min at 570 nm using a microplate reader (Varioskan Flash, Thermo Scientific, Waltham, Massachusetts, USA). Quantification of Free Amino Groups after Cross-linking Free amino groups were determined by equimolar reaction with of TNBS.25 Briefly, 2 mL of 4 :mol/mL TNBS solution was added to 2 mL of 2.5 mg/mL HA-NPs in 4% sodium bicarbonate buffer (pH 8.5). The reaction mixture was shaken at 100 rpm for 1 h at 40◦ C and centrifuged (16 738 g, 90 min) to separate the nanospheres. To determine the amount of unreacted TNBS, 0.9 mL supernatant was withdrawn and incubated with 0.1 mL L-valine solution (40 :mol/L) in 1% trichloroacetic acid. After reaction for 1 h at 40◦ C, the absorbance at 425 nm was recorded. Therefore, the free amino group content of HA-NPs was calculated by subtracting the unreacted TNBS in the supernatant fraction from the amount of TNBS added initially. Cell Culture A549 cells were maintained in a 5% CO2 humidified incubator at 37◦ C and cultured in RPMI-1640 medium supplemented with 10% FBS and antibiotics (50 U/mL penicillin and 50 mg/mL streptomycin). Cytotoxicity Study
Figure 3. Optimization of the particle size of HA-NPs with regard to three influencing factors: (a) acetone volume, (b) HA–ADH concentration, and (c) HA molecular weight.
fluorescence microscopy (Axiovert 40CFL, Carl Zeiss, Oberkochen, Germany), and confocal laser scanning microscopy (CLSM, FV1000, Olympus, Tokyo, Japan), HA-NPs were labeled with FITC by the reaction between the isothiocyanate groups of FITC and the amino groups of HA–ADH.23 The content of FITC in the conjugate was about 9.4 mg/g as deterDOI 10.1002/jps.23924
A549 cells were seeded in 96-well plates at a density of 7 × 103 cells/well and incubated for 24 h. The cells were treated with HA–ADH or HA-NPs (0.01–1 mg/mL) for 24 h. Cells were then stained with 20 :L of 5 mg/mL MTT for 4 h. The formazan crystals formed were dissolved in 150 :L dimethyl sulfoxide, and the absorbance was measured at 570 nm.7 Untreated cells were used as the control. Reactive oxygen species assay was also used to evaluate the cytotoxicity of HA-NPs.26 2 × 105 cells per well were seeded in 12-well plates. After adherence, the cells were treated with HA–ADH or HA-NPs (1 mg/mL) for 24 h. A549 cells were then stained with 0.5 mL of 10 :M 2 ,7 -dichlorodihydrofluorescein diacetate (dissolved in serum-free RPMI-1640 medium) at 37◦ C for 20 min. The cells were harvested by trypsinization and washed with phosphate-buffered saline (PBS, pH 7.4). After centrifugation, the cell pellet was resuspended in 0.5 mL PBS before flow cytometry analysis. Untreated cells were used as the control. Zheng et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:1529–1537, 2014
1532
RESEARCH ARTICLE – Pharmaceutical Nanotechnology
Figure 4. Characteristics of the optimal HA-NPs. Morphology of HA-NPs measured using SEM (a) and TEM (b). Particle size distribution (c) and zeta potential (d) of HA-NPs in PBS determined by DLS.
Cell Uptake Analysis A549 cells were cultivated in 12-well plates at a density of 2 × 105 cells/well. FITC solution or FITC–HA-NPs (equivalent to 1 :g/mL of FITC) were added to the cells. After treatment for different time periods (from 5 min to 2 h), the cells were washed thrice with cold PBS and subjected to fluorescence microscopy analysis. For flow cytometry analysis, the cells were harvested, washed, and resuspended in PBS. In order to investigate the endocytosis mechanism of HANPs, 0.5 mL of 5 mg/mL TB was added to the wells immediately after the treatment with nanospheres. A549 cells were treated at 4◦ C and 37◦ C, respectively, to study the effect of temperature. To assess the effect of HA, competitive inhibition study was also conducted by addition of HA solution (1 or 10 mg/mL) 2 h before the treatment. In Vivo Study Kunming mice (18–22 g) were purchased from Laboratory Animal Center of Sichuan University. All in vivo procedures were approved by the Animal Ethical Experimentation Committee, in accordance with the requirements of the National Act on the use of experimental animals. Hepatoma model was established as follows.27 H22 cells (preserved in liquid nitrogen) were warmed to 37◦ C and diluted with sterile physiological saline. The mice were inoculated intraperitoneally with 0.2 mL of H22 cell suspension (1 × 106 cells/mL) to form ascites. After a week, the ascites were drawn, diluted to 1 × 107 cells/mL with physiological saline and then injected subcutaneously into the left armpit of Kunming mice. When the tumor diameter reached about 6 mm, FITC or FITC–HA-NPs (equivalent to 50 :g/mL of FITC) was injected through the tail vein. Tumor tissues were excised 6 h Zheng et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:1529–1537, 2014
later and frozen sectioned. Sections were stained with DAPI (2 :g/mL) and imaged by CLSM. For quantitative analysis, the excised tumor tissues were cut into small pieces, incubated with dissociation solution, which was prepared by dissolving collagenase type IV (1 mg/mL) and DNase I (30 :g/mL) in PBS. The obtained cell suspension was filtered through 70-:m mesh, washed with PBS for three times before flow cytometry analysis. Statistical Analysis Results were shown as mean ± SD. Comparison between data was carried out using t-test of independent samples. A p value 0.05, respectively.
glutaraldehyde, and subjected to stability test (suspended in PBS, stored at 4◦ C). As shown in Figure 2b, the particle size experienced minimal variation over 7 days, ranging from 97.94 ± 1.72 to 109.2 ± 1.25 nm, and the size distribution was comparable as well (PDI < 0.2) during this period, demonstrating that 75 :mol glutaraldehyde was enough to stabilize the HA-NPs. Particle size
Figure 7. Effect of TB (a) and incubation temperature (b) on cellular uptake of FITC–HA-NPs. * and # represent p < 0.05 and p > 0.05, respectively.
the precipitation of HA–ADH. The proportion of nondesolvated HA–ADH might dissolve in the mixed solvent. Stability As shown in Figure 2a, the amount of free amino groups on HA-NPs after being crosslinked with low amount of glutaraldehyde (1.25–5 :mol) was approximate to that of HA–ADH. It can be deduced that only a small proportion of HA-NPs had been successfully cross-linked and the nanospheres prepared were relatively unstable. This deduction was further evidenced by the redissolution of HA-NPs after removal of acetone.30 As the amount of glutaraldehyde increased from 12.5 to 75 :mol, the free amino groups decreased without reaching a nadir, indicating a progressive degree of cross-linking. With the increase of glutaraldehyde, HA-NPs were supposed to be more stable. Nevertheless, given the toxicity of glutaraldehyde, HA-NPs were cross-linked with no more than 75 :mol Zheng et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:1529–1537, 2014
Particle size is believed to play a critical role in targeting, therefore optimization of the particle size was conducted in consideration of three influencing factors (Fig. 3). The particle size gradually decreased from 149.2 ± 4.82 to 97.47 ± 1.19 nm with the addition of acetone (2–3 mL) as a result of enhanced hydrophobic interaction, and reached a plateau when acetone was more than threefold. During the latter stage, the count rate, which is proportional to both particle size and concentration,29 also reached the maximum (Fig. 3a). It is reasonable to deduce that the particle concentration was relatively stable. This deduction is consistent with the conclusion above that nondesolvated HA–ADH remained constant when more than threefold acetone was added. The maximal count rate was achieved when 2.5-fold acetone was added, and this phenomenon may be attributed to the synergistic effect of particle size and concentration. Figure 3b illustrated the change of particle sizes caused by the increase of HA–ADH concentrations. As the concentration increased (from 1 to 5 mg/mL), particle size dropped from 569.17 ± 51.25 to 98.3 ± 1.27 nm. This is inconsistent with the general belief that higher concentration favors larger particle.29 However, it might be reasonable considering that HA–ADH molecules in low concentration solutions are too rare to ensure proper distance between amino groups, which is a prerequisite for cross-linking. Low cross-linking level could lead to subsequent redissolution of HA-NPs, as evidenced by high PDI (PDI > 0.5) and low count rate (shown in Fig. 3b). As HA–ADH concentration rose to 10 mg/mL, larger nanospheres (144.4 ± 2.23 nm) were obtained as expected. Simultaneously, count rate surged with the increase of HA–ADH concentration, mainly benefiting from higher particle concentration. DOI 10.1002/jps.23924
RESEARCH ARTICLE – Pharmaceutical Nanotechnology
1535
Figure 9. Representative CLSM images of frozen sections (a) and flow cytometry images of cell suspension (b) of tumor tissues excised from H22-bearing mice 6 h after meditation.
Realizing that commercially available HA might differ in MW, we assessed the effect of MW on particle size (Fig. 3c). HA with MW of 8 and 101 kDa could produce uniform nanospheres with PDI of 0.13 ± 0.07 and 0.18 ± 0.03, respectively, whereas the nanospheres prepared with HA of low MW (361.33 ± 2.60 nm) were almost threefold larger than that prepared with HA of moderate MW (98.37 ± 1.27 nm). At the same mass concentration, more molecules are contained in HA–ADH solution of low MW, which could boost the chances for aggregation, thereby resulting in larger particle.29 HA of high MW (1.44 mDa) is not suitable for the fabrication of nanospheres because of its poor aqueous solubility (
Fabrication of Biocompatible and Tumor-Targeting Hyaluronan Nanospheres by a Modified Desolvation Method WENYI ZHENG, YONGQUAN LI, JINPING DU, ZONGNING YIN Key Laboratory of Drug Targeting and Drug Delivery Systems, West China School of Pharmacy, Sichuan University, Chengdu, Sichuan Province, China Received 25 November 2013; revised 23 January 2014; accepted 18 February 2014 Published online 6 March 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.23924 ABSTRACT: The aim of this work was to maximize the tumor targetability of biocompatible hyaluronan (HA) by construction of a novel nanocarrier, using HA as the single material. HA was prefunctionalized with active amino groups, desolvated by acetone, and crosslinked by glutaraldehyde. The process was further optimized with regard to yield, stability, and particle size. The cytotoxicity of HA nanospheres (HA-NPs) was evaluated by thiazolyl blue tetrazolium bromide and reactive oxygen species assays. A549 cells and H22bearing Kunming mice were employed to characterize the tumor targeting of fluorescein isothiocyanate-conjugated HA-NPs. Nanospheres prepared according to the optimal formulation were characterized by a maximum yield (90%), high stability over 7 days, regular spheres (97.42 nm in diameter), and negative charge (−32.7 mV in zeta potential). In vitro results revealed that HA-NPs had little cytotoxicity and efficiently accumulated into A549 cells in a HA-dependent manner. Following systemic administration in mice, HA-NPs selectively accumulated in the tumor as demonstrated by the frozen section examination and flow cytometry analysis. In conclusion, this work C 2014 Wiley successfully prepared HA-NPs and explored their potential applications for tumor targeting in terms of safety and efficacy. Periodicals, Inc. and the American Pharmacists Association J Pharm Sci 103:1529–1537, 2014 Keywords: hyaluronan; CD44; cancer; desolvation; drug targeting; nanospheres; polymeric drug delivery systems
INTRODUCTION Hyaluronan (HA) is a linear, negatively charged polysaccharide, which is composed of repetitive disaccharide units. It is involved in the regulation of several key biological processes by binding to its endogenous receptors including clusters of differentiation 44 (CD44), HA-mediated motility receptor, lymphatic vessel endothelial HA receptor-1, and HA receptor for endocytosis.1 Among these receptors, CD44 is reported to be expressed highly on the surface of tumor cells, especially the highly invasive tumor cells and stem cells, a discovery that is widely utilized in specific targeting therapy.2,3 In the field of pharmaceutics, delivery systems based on HA– CD44 interactions have been developed to facilitate drug/gene delivery. More specifically, HA-based polymeric prodrugs account for a large proportion of the existing delivery systems. Successful examples include doxorubicin, epirubicin, butyrate, paclitaxel, fluorouracil, and camptothecin.4–9 In spite of the enhanced therapeutic efficacy, this strategy exhibits some drawbacks, a main one of which is the rapid clearance rate of HA in the blood circulation.10 Various nanoparticulate systems for CD44 targeting are available at present. Thanks to the abundant carboxyl and hydroxyl groups, HA could be readily modified on lipid-based nanoparticle, FeO particles, or liposome.11–13 Alternative systems include micelles constructed by introducing hydrophobic segments (i.e., fatty amine, histidine, deoxycholic acid, cholanic acid, and ceramide)14–18 to the HA backbone, and nanocomplexs formed between positively charged chitosan, polyetherimide, or platinum and negatively charged HA.19–21 Correspondence to: Zongning Yin (Telephone: +86-28-85502917; Fax: +8628-85502917; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 103, 1529–1537 (2014)
C 2014 Wiley Periodicals, Inc. and the American Pharmacists Association
It is clear that HA, the targeting molecule, only accounts for a small part within these nanoparticulate systems. Qhattal and Liu13 concluded that the uptake efficiency of HA-grafted liposomes in CD44-overexpressing tumor cells rises with increasing grafting density. It is therefore reasonable to assume that nanospheres consisting entirely of HA would have the strongest targetability. In addition, because of the continuous distribution of HA from the interior to the exterior of the nanospheres, HA-NPs would remain its targetability during the course of degradation, providing an advantage over conventional carriers. Another advantage of HA-NPs is associated with their inherent hydrophilicity, which minimizes serum protein binding and subsequent reticuloendothelial uptake. To the best of our knowledge, there were few literatures reporting successful preparation of HA-NPs without using composite materials. In this study, we aimed to fill this gap by developing HA-NPs via a modified desolvation method. HA-NPs were further optimized in terms of yield, stability, and particle size, and evaluated for their cytotoxicity as well as targetability.
MATERIALS AND METHODS Materials Hyaluronan with molecular weights (MW) of 101 and 8 kDa were purchased from Dali (Liuzhou, Guangxi, China), whereas HA with MW of 1.44 mDa was from Freda (Jinan, Shandong, China). Thiazolyl blue tetrazolium bromide (MTT), trypan blue (TB), and sodium 3-nitrobenzenesulfonate (TNBS) were obtained from Sigma–Aldrich (St. Louis, Missouri, USA). Fluorescein isothiocyanate (FITC) and adipic dihydrazide (ADH) were purchased from Aladdin (Shanghai, China). Reactive oxygen species (ROS) assay kit was purchased from Beyotime Institute of Biotechnology (Haimen, Jiangsu, China).
Zheng et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:1529–1537, 2014
1529
1530
RESEARCH ARTICLE – Pharmaceutical Nanotechnology
Table 1.
Formulations Used in the Optimization Study of HA-NPs
Formulation A B C D E
Molecular Weight HA–ADH Acetone Glutaraldehyde of HA (kDa) (mg) (mL) (:mol) 101 101 101 101 /
5 5 5 / 5
/ 3.5 / 3.5 3.5
75 / 75 75 75
Formulation A and B were used to evaluate the effect of acetone on nondesolvated HA–ADH and the effect of glutaraldehyde on free amino groups. Formulations C, D, and E were used to assess the effect of acetone, HA–ADH concentration, and MW of HA on particle size, respectively. / means variable.
Hyclone (Logan, Utah, USA). Adenocarcinomic human alveolar basal epithelial cells (A549) and murine hepatocellular carcinoma (H22) cells were provided by the Key Laboratory of Drug Targeting and Drug Delivery Systems. Derivatization of HA Adipic dihydrazide (391.5 mg, 2.02 mmol) was added to a solution of HA (180 mg) in water (45 mL). After adjusting the pH of the reaction mixture to 4.75 with HCl solution, the reaction was initiated by adding 1-ethyl-3-[3-(dimethylamino)propyl] carbodiimide (86.4 mg, 0.44 mmol) under stirring at room temperature. The reaction was stopped 2 h later by adjusting the pH to 7.0. The mixture was exhaustively dialyzed against 0.1 M NaCl solution, 25% ethanol, and pure water, and then lyophilized to give HA–ADH.22 1 H-NMR (Unity Inova 400, Varian (Palo Alto, California, USA), 400 MHz, D2 O, *, ppm): 2.235–2.207 (s, 4H, –COCH2 ), 2.005–1.967 (t, 3H, –NCOCH3 ), 1.638 (s, 4H, –COCH2 CH2 ). The substitution degree of different conjugates was estimated to vary from 12.6% to 22.4% according to the 1 H-NMR spectrum. Preparation and Characterization of HA-NP
Figure 1. Percentage of nondesolvated HA–ADH in correlation with acetone volume added during the desolvation procedure.
4 ,6-Diamidino-2-phenylindole) (DAPI), collagenase type IV, and DNase I were purchased from Biosharp (Hefei, Anhui, China). Sephadex G50 was purchased from Pharmacia (Uppsala, Sweden). Roswell Park Memorial Institute (RPMI-1640) medium and fetal bovine serum (FBS) were purchased from
Hyaluronan–ADH was dissolved in 1 mL water and desolvated by adding acetone dropwise under stirring. After addition of glutaraldehyde, the mixture was subjected to reaction for 12 h. HA-NPs were obtained after removal of acetone. Further purification was performed on Sephadex G50 for cell experiment and in vivo study. The optimization was conducted according to different formulations as listed in Table 1. All experiments were performed in triplicate. The average particle size, polydispersity index (PDI), zeta potential, and count rate were measured using a dynamic light scattering analyzer (Zetasizer Nano ZS90, Malvern, UK). Before scanning electron microscope (SEM, JSM-6510LV, Jeol, Tokyo, Japan) and transmission electron microscopy (TEM, H6001V, Hitachi, Tokyo, Japan) observation, nanospheres were coated with gold and stained with phosphotungstic acid. In order to determine the cellular uptake by flow cytometry (FC500, Beckman Coulter, Brea, California, USA),
Figure 2. (a) Influence of the amount of glutaraldehyde on free amino groups of HA-NPs. (b) Diameter and size distribution variation of HA-NPs cross-linked with 75 :mol glutaraldehyde over 7 days after incubation with PBS at 4◦ C. Zheng et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:1529–1537, 2014
DOI 10.1002/jps.23924
RESEARCH ARTICLE – Pharmaceutical Nanotechnology
1531
mined by fluorescence spectrometry at Ex /Em wavelengths of 490/520 nm. Determination of Nondesolvated HA For the determination of the percentage of nondesolvated HA– ADH, the nanospheres were separated from the supernatant by centrifugation at 16 738 g for 90 min (Allgera X22, Beckman Coulter, California, USA). HA–ADH in the supernatant was quantified by the turbidimetric method, which is based on the formation of HA–hexadecyltrimmethylammonium bromide (CTAB) complex.24 Briefly, 50 :L standard solution (0.3– 1.2 mg/mL) or supernatant was added to a 96-well plate, and incubated with 50 :L acetate buffer (0.2 M, pH 6.0) with 0.15 M NaCl for 20 min at 37◦ C. Then, 100 :L of 10 mM preincubated CTAB solution in 0.5 M NaOH was added, and the absorbance was measured within 10 min at 570 nm using a microplate reader (Varioskan Flash, Thermo Scientific, Waltham, Massachusetts, USA). Quantification of Free Amino Groups after Cross-linking Free amino groups were determined by equimolar reaction with of TNBS.25 Briefly, 2 mL of 4 :mol/mL TNBS solution was added to 2 mL of 2.5 mg/mL HA-NPs in 4% sodium bicarbonate buffer (pH 8.5). The reaction mixture was shaken at 100 rpm for 1 h at 40◦ C and centrifuged (16 738 g, 90 min) to separate the nanospheres. To determine the amount of unreacted TNBS, 0.9 mL supernatant was withdrawn and incubated with 0.1 mL L-valine solution (40 :mol/L) in 1% trichloroacetic acid. After reaction for 1 h at 40◦ C, the absorbance at 425 nm was recorded. Therefore, the free amino group content of HA-NPs was calculated by subtracting the unreacted TNBS in the supernatant fraction from the amount of TNBS added initially. Cell Culture A549 cells were maintained in a 5% CO2 humidified incubator at 37◦ C and cultured in RPMI-1640 medium supplemented with 10% FBS and antibiotics (50 U/mL penicillin and 50 mg/mL streptomycin). Cytotoxicity Study
Figure 3. Optimization of the particle size of HA-NPs with regard to three influencing factors: (a) acetone volume, (b) HA–ADH concentration, and (c) HA molecular weight.
fluorescence microscopy (Axiovert 40CFL, Carl Zeiss, Oberkochen, Germany), and confocal laser scanning microscopy (CLSM, FV1000, Olympus, Tokyo, Japan), HA-NPs were labeled with FITC by the reaction between the isothiocyanate groups of FITC and the amino groups of HA–ADH.23 The content of FITC in the conjugate was about 9.4 mg/g as deterDOI 10.1002/jps.23924
A549 cells were seeded in 96-well plates at a density of 7 × 103 cells/well and incubated for 24 h. The cells were treated with HA–ADH or HA-NPs (0.01–1 mg/mL) for 24 h. Cells were then stained with 20 :L of 5 mg/mL MTT for 4 h. The formazan crystals formed were dissolved in 150 :L dimethyl sulfoxide, and the absorbance was measured at 570 nm.7 Untreated cells were used as the control. Reactive oxygen species assay was also used to evaluate the cytotoxicity of HA-NPs.26 2 × 105 cells per well were seeded in 12-well plates. After adherence, the cells were treated with HA–ADH or HA-NPs (1 mg/mL) for 24 h. A549 cells were then stained with 0.5 mL of 10 :M 2 ,7 -dichlorodihydrofluorescein diacetate (dissolved in serum-free RPMI-1640 medium) at 37◦ C for 20 min. The cells were harvested by trypsinization and washed with phosphate-buffered saline (PBS, pH 7.4). After centrifugation, the cell pellet was resuspended in 0.5 mL PBS before flow cytometry analysis. Untreated cells were used as the control. Zheng et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:1529–1537, 2014
1532
RESEARCH ARTICLE – Pharmaceutical Nanotechnology
Figure 4. Characteristics of the optimal HA-NPs. Morphology of HA-NPs measured using SEM (a) and TEM (b). Particle size distribution (c) and zeta potential (d) of HA-NPs in PBS determined by DLS.
Cell Uptake Analysis A549 cells were cultivated in 12-well plates at a density of 2 × 105 cells/well. FITC solution or FITC–HA-NPs (equivalent to 1 :g/mL of FITC) were added to the cells. After treatment for different time periods (from 5 min to 2 h), the cells were washed thrice with cold PBS and subjected to fluorescence microscopy analysis. For flow cytometry analysis, the cells were harvested, washed, and resuspended in PBS. In order to investigate the endocytosis mechanism of HANPs, 0.5 mL of 5 mg/mL TB was added to the wells immediately after the treatment with nanospheres. A549 cells were treated at 4◦ C and 37◦ C, respectively, to study the effect of temperature. To assess the effect of HA, competitive inhibition study was also conducted by addition of HA solution (1 or 10 mg/mL) 2 h before the treatment. In Vivo Study Kunming mice (18–22 g) were purchased from Laboratory Animal Center of Sichuan University. All in vivo procedures were approved by the Animal Ethical Experimentation Committee, in accordance with the requirements of the National Act on the use of experimental animals. Hepatoma model was established as follows.27 H22 cells (preserved in liquid nitrogen) were warmed to 37◦ C and diluted with sterile physiological saline. The mice were inoculated intraperitoneally with 0.2 mL of H22 cell suspension (1 × 106 cells/mL) to form ascites. After a week, the ascites were drawn, diluted to 1 × 107 cells/mL with physiological saline and then injected subcutaneously into the left armpit of Kunming mice. When the tumor diameter reached about 6 mm, FITC or FITC–HA-NPs (equivalent to 50 :g/mL of FITC) was injected through the tail vein. Tumor tissues were excised 6 h Zheng et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:1529–1537, 2014
later and frozen sectioned. Sections were stained with DAPI (2 :g/mL) and imaged by CLSM. For quantitative analysis, the excised tumor tissues were cut into small pieces, incubated with dissociation solution, which was prepared by dissolving collagenase type IV (1 mg/mL) and DNase I (30 :g/mL) in PBS. The obtained cell suspension was filtered through 70-:m mesh, washed with PBS for three times before flow cytometry analysis. Statistical Analysis Results were shown as mean ± SD. Comparison between data was carried out using t-test of independent samples. A p value 0.05, respectively.
glutaraldehyde, and subjected to stability test (suspended in PBS, stored at 4◦ C). As shown in Figure 2b, the particle size experienced minimal variation over 7 days, ranging from 97.94 ± 1.72 to 109.2 ± 1.25 nm, and the size distribution was comparable as well (PDI < 0.2) during this period, demonstrating that 75 :mol glutaraldehyde was enough to stabilize the HA-NPs. Particle size
Figure 7. Effect of TB (a) and incubation temperature (b) on cellular uptake of FITC–HA-NPs. * and # represent p < 0.05 and p > 0.05, respectively.
the precipitation of HA–ADH. The proportion of nondesolvated HA–ADH might dissolve in the mixed solvent. Stability As shown in Figure 2a, the amount of free amino groups on HA-NPs after being crosslinked with low amount of glutaraldehyde (1.25–5 :mol) was approximate to that of HA–ADH. It can be deduced that only a small proportion of HA-NPs had been successfully cross-linked and the nanospheres prepared were relatively unstable. This deduction was further evidenced by the redissolution of HA-NPs after removal of acetone.30 As the amount of glutaraldehyde increased from 12.5 to 75 :mol, the free amino groups decreased without reaching a nadir, indicating a progressive degree of cross-linking. With the increase of glutaraldehyde, HA-NPs were supposed to be more stable. Nevertheless, given the toxicity of glutaraldehyde, HA-NPs were cross-linked with no more than 75 :mol Zheng et al., JOURNAL OF PHARMACEUTICAL SCIENCES 103:1529–1537, 2014
Particle size is believed to play a critical role in targeting, therefore optimization of the particle size was conducted in consideration of three influencing factors (Fig. 3). The particle size gradually decreased from 149.2 ± 4.82 to 97.47 ± 1.19 nm with the addition of acetone (2–3 mL) as a result of enhanced hydrophobic interaction, and reached a plateau when acetone was more than threefold. During the latter stage, the count rate, which is proportional to both particle size and concentration,29 also reached the maximum (Fig. 3a). It is reasonable to deduce that the particle concentration was relatively stable. This deduction is consistent with the conclusion above that nondesolvated HA–ADH remained constant when more than threefold acetone was added. The maximal count rate was achieved when 2.5-fold acetone was added, and this phenomenon may be attributed to the synergistic effect of particle size and concentration. Figure 3b illustrated the change of particle sizes caused by the increase of HA–ADH concentrations. As the concentration increased (from 1 to 5 mg/mL), particle size dropped from 569.17 ± 51.25 to 98.3 ± 1.27 nm. This is inconsistent with the general belief that higher concentration favors larger particle.29 However, it might be reasonable considering that HA–ADH molecules in low concentration solutions are too rare to ensure proper distance between amino groups, which is a prerequisite for cross-linking. Low cross-linking level could lead to subsequent redissolution of HA-NPs, as evidenced by high PDI (PDI > 0.5) and low count rate (shown in Fig. 3b). As HA–ADH concentration rose to 10 mg/mL, larger nanospheres (144.4 ± 2.23 nm) were obtained as expected. Simultaneously, count rate surged with the increase of HA–ADH concentration, mainly benefiting from higher particle concentration. DOI 10.1002/jps.23924
RESEARCH ARTICLE – Pharmaceutical Nanotechnology
1535
Figure 9. Representative CLSM images of frozen sections (a) and flow cytometry images of cell suspension (b) of tumor tissues excised from H22-bearing mice 6 h after meditation.
Realizing that commercially available HA might differ in MW, we assessed the effect of MW on particle size (Fig. 3c). HA with MW of 8 and 101 kDa could produce uniform nanospheres with PDI of 0.13 ± 0.07 and 0.18 ± 0.03, respectively, whereas the nanospheres prepared with HA of low MW (361.33 ± 2.60 nm) were almost threefold larger than that prepared with HA of moderate MW (98.37 ± 1.27 nm). At the same mass concentration, more molecules are contained in HA–ADH solution of low MW, which could boost the chances for aggregation, thereby resulting in larger particle.29 HA of high MW (1.44 mDa) is not suitable for the fabrication of nanospheres because of its poor aqueous solubility (
