Nanoparticles for Intracellular Delivery - ACS Publications - American

Sep 30, 2010 - Confocal Raman Microscopy (CRM) is used to study the cell internalization of ... cell and does not provide information about localizati...
0 downloads 0 Views 4MB Size
Download PDF
Biomacromolecules 2010, 11, 2993–2999

2993

Surface Engineered Poly(lactide-co-glycolide) Nanoparticles for Intracellular Delivery: Uptake and CytotoxicitysA Confocal Raman Microscopic Study Gabriela Romero,† Irina Estrela-Lopis,‡ Jie Zhou,†,§ Elena Rojas,† Ana Franco,| Christian Sanchez Espinel,| Africa Gonza´lez Ferna´ndez,| Changyou Gao,§ Edwin Donath,‡ and Sergio E. Moya*,† CIC BiomaGUNE, Paseo Miramo´n 182 Ed. Emp. C, San Sebastia´n, Spain, Institute of Biophysics and Medical Physics, University of Leipzig, Leipzig, Germany, Department of Polymer Science and Engineering, Zhejiang University, Hangzhou 310027, China, and Laboratorio de Inmunologı´a, Edificio Ciencias Experimentales, Campus Lagoas Marcosende, Universidad de Vigo, CP 36310, Vigo, Pontevedra, Spain Received July 13, 2010; Revised Manuscript Received August 31, 2010

Confocal Raman Microscopy (CRM) is used to study the cell internalization of poly(lactide-co-glycolide) (PLGA) nanoparticles (NPs) fabricated by emulsion techniques with either poly(ethylene imine) (PEI) or bovine serum albumin (BSA) as surface stabilizers. HepG2 cells were exposed to PEI and BSA stabilized PLGA NPs. Spontaneous Confocal Raman Spectra taken in one and the same spot of exposed cells showed bands arising from the cellular environment as well as bands characteristic for PLGA, proving that the PLGA NPs have been internalized. It was found that PLGA NPs preferentially colocalize with lipid bodies. The results from Raman spectroscopy are compared with flow cytometry and confocal scanning laser microscopy (CLSM) data. The advantages of CRM as a label-free technique over flow cytometry and CLSM are discussed. Additionally, cell viability studies by means of quick cell counting solution and MTT tests in several cell lines show a generally low toxicity for both PEI and BSA stabilized PLGA NPs, with BSA stabilized PLGA NPs having an even lower toxicity than PEI stabilized.

Introduction Polymeric nanoparticles (NPs) of submicrometer size are appealing systems for drug delivery.1–7 They can be easily synthesized and in most of the cases the protocols for drug encapsulation are performed in parallel with fabrication. Furthermore, polymeric NPs are normally suitable to be functionalized for targeted delivery by multiple procedures. There is a wide variety of biocompatible and biodegradable polymers from natural sources like BSA, collagen, gelatin, chitosan, and alginate, which can be employed for the fabrication of devices for controlled drug delivery.8 During the last three decades, synthetic biodegradable polymers, for example, poly(amino acids), poly(alkyl-2-cyanoacrylates), polyesters, polyorthoesters, polyanhydrides, and polyurethanes, have been synthesized and employed in drug loading devices. Among them, the homopolymer of lactic acid (PLA)9,10 and its copolymer, poly(lactideco-glycolide) (PLGA),6,11–16 are commonly used to fabricate NPs. PLA and PLGA NPs are easy to produce and display excellent biocompatibility and biodegradability. Several techniques have been developed for the preparation of PLA and PLGA NPs such as simple or double emulsions-solvent evaporation,17 nanoprecipitation,18,19 spray drying,20 and so on. The simple or double emulsions-solvent evaporation method is most often employed. The size of the NPs can be easily controlled * To whom correspondence should be addressed. Fax: +34 943005311. E-mail: [email protected]. † CIC biomaGUNE. ‡ University of Leipzig. § Zhejiang University. | Universidad de Vigo.

with this method, and it can also be applied for the encapsulation of both hydrophobic and hydrophilic drugs.21–24 We have previously shown that PLGA NPs can be prepared employing PEI or BSA as surface stabilizers.25–27 The charge of PEI or BSA allows for the application of the layer by layer technique (LBL) for further noncovalent modification of the NPs. The LBL coating, which can be performed by the stepwise adsorption of the biocompatible chitosan and alginate,26 for example, has then been used for the attachment of folic acid to increase NP uptake in cancer cell lines.26,27 In our previous work, cellular uptake of the PLGA NPs has been quantified by means of the combination of flow cytometry and confocal laser scanning microscopy (CLSM). For labeling, either fluorescein or rhodamine was encapsulated in the PLGA NPs during preparation. It was shown that the uptake of NPs is influenced by the nature of the surface of the PLGA NPs. Large uptake was observed for PLGA NPs coated with PEI and BSA. A coating with chitosan/alginate or PEGylation27 reduced the uptake significantly. Subsequent modification with folic acid facilitated again uptake by specific targeting, because it is known that tumor cells overexpress on their membranes folate receptors (FRs).25–27 It is nevertheless difficult to differentiate whether the NPs have been internalized in the cells or remained attached to the cell membrane. Flow cytometry measures the fluorescence per cell and does not provide information about localization of labeled NPs, whether attached to or indeed taken up by the cells under study. Visualization of the NPs by CLSM is not a straightforward task because it is not always possible to distinguish between NPs that have been internalized in the cell and NPs attached to the cell membrane, which in many cases

10.1021/bm1007822  2010 American Chemical Society Published on Web 09/30/2010

2994

Biomacromolecules, Vol. 11, No. 11, 2010

needs to be labeled for cell visualization. Also, the encapsulated dye may be released and accumulated in the cell interior even though the PLGA NPs remained in the bulk phase. Even in the case that PLGA molecules are labeled covalently, PLGA could eventually degrade and be incorporated in the cell together with the marker. In these situations one could arrive to the false conclusion that the particles have been incorporated in the cells when it is not the case.28 In a recent paper, Peisheng Xu et al.28 demonstrated by means of coherent anti-Stokes Raman scattering (CARS) microscopy, that PLGA NPs stabilized with poly(vinyl alcohol) (PVA) were not taken up by cells, although CLSM and flow cytometry indicated that they were. Notwithstanding with this observation, Chernenko et al. showed by Confocal Raman Microscopy (CRM) that NPs stabilized with Pluronic F 108 are incorporated within cells and could be found in the Golgi associated vesicles of the late endosomes.29 The localization of PLGA NPs was also recently studied with immune fluorescence techniques in epithelial cells30 demonstrating an association with different intracellular compartments as a function of time. These results demonstrate that the chemical properties of the surface of the NPs play a decisive role regarding internalization and uptake of PLGA NPs. In this paper we show that CRM is a well-suited technique to study the uptake of NPs into culture cells. CRM does not require labeling and can therefore be used for studying a large variety of NPs without having the requisite of labeling them.31 This is an important advantage because authentic NP can be probed. The Raman spectra includes band characteristics of the PLGA NPs and of the cell environment in the vicinity of the NPs.29,31 Spectral decomposition can thus be used to explore the association of the NPs with particular intracellular environments. The information provided by CRM will be compared with CLSM and flow cytometry data. PLGA NPs with PEI or BSA as surface coatings are readily taken up, underlining the importance of a proper surface modification for delivery. In addition to the uptake studies, the cytotoxicity of these two types of particles was determined to evaluate the possible use of surface-modified PLGA NPs for intracellular delivery.

Experimental Section Materials. PLGA (D,L-lactide 85: glycolide 15, inherent viscosity within 0.55-0.75 dL/g) was purchased from LACTEL. Branched PEI, Mw ∼ 25 kDa, BSA, PIERCE BCATM Protein Assay Kit, Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS), penicillin-streptomycin and rhodamine 6G were purchased from SigmaAldrich. All chemicals were used as received. Hepatocarcinoma human cell line (HepG2) was obtained from American Type Culture Collection. NPs Preparation. PLGA NPs were prepared by the O/W emulsionsolvent evaporation method. First, 1 mL of 20 mg/mL PLGA dichloromethane solution (organic phase) was added to 4 mL of stabilizer (PEI or BSA) solution with different concentrations (1, 2, 3, 5% for PEI solution, 0.5, 1, 2, 3% for BSA solution), which was then emulsified by an ultrasonicator (SONICS1 VCX 500) for 20 s. This emulsion was immediately poured into 100 mL of distilled water and stirred for 3 h with a magnetic stirrer until the organic solvent totally evaporated. The PLGA NPs were collected by centrifugation at 10000 g for 5 min and were washed with Milli-Q water five times to remove the free stabilizer initially presented in the water phase. The samples for CLSM and flow cytometry were prepared by adding 0.5 mg of dye Rhodamine 6G (Rd6G) into the organic phase before sonication. Finally, the PLGA NPs were lyophilized with the lyophilizer Christ R 1-2 LD plus and stored at 4 °C before use. NPs Characterization. The size and shape of the PLGA NPs prepared with different concentrations of both stabilizers were char-

Romero et al. acterized by transmission electron microscopy (TEM) (JEOL JEM 2100F, Japan). The zeta potential and hydrodynamic diameters of PLGA NPs were measured with a Nanosizer (Malvern Nanosize, U.K.). All the zeta potential measurements were performed in 10 mM NaCl solution at pH 7.4 and at a cell drive voltage of 30 V, using a monomodal analysis model. Cellular Uptake by Flow Cytometry and CLSM. Hepatocarcinoma human cell line HepG2 was cultured in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum (FBS), 1000 U penicillin, and 10 mg/mL streptomycin at 37 °C and 5% CO2. When the cell confluence was around 70%, the cells were trypsinized and resuspended in culture medium. A total of 300000 cells were seeded into each well of a 24-well culture plate. PLGA NPs labeled with Rd6G were added into the culture medium 24 h later. The final NPs concentration was adjusted to 100 µg/mL. After cultured for 0.5, 1, 2, 4, 8, 12, and 24 h, the cells were washed twice with PBS, trypsinized, and studied by flow cytometry (BD FACSCanto II). The uptake ratio was calculated from the dotplot graph of forward scattering (FCS) versus fluorescence intensity (PE-A). Data were analyzed with the WinMDI program. First, a threshold of fluorescence was generated using a control sample, that is, the HepG2 cells without exposure to the NPs was set at a value of 340 arbitrary units and all runs were recorded using this value as threshold. All dots corresponding to the control sample are located at intensities below this threshold. Each experiment, was set to count 10000 events per run and each run was recorded always under the same parameters. The number of cells carrying fluorescently labeled NPs is obtained from the area corresponding to the dots located at higher intensities than the threshold. The cellular uptake ratio is equal to the (No. of events over the threshold/total No. of events) × 100%. CLSM (Carl-Zeiss LSM 10 META) was used to image HepG2 cells after incubation with NPs. A total of 100000 cells were seeded in a 3.5 cm culture plate containing a coverslip, which was placed into a 3.5 cm culture plate. After 24 h, the PLGA NPs containing Rd6G were added into the culture medium with a final concentration of 50 µg/mL. After incubation for another 12 h the culture medium was removed and the cells were rinsed three times with sterilized PBS. The cells were subsequently fixed with 3.7% formaldehyde solution in PBS for 30 min and rinsed again. Finally, the cells were incubated with DAPI (100 ng/mL) in 10 mM PBS for 20 min to stain the cell nucleus. After being washed five times with PBS, the cells on these coverslips were observed under CLSM employing a 63 × oil immersion objective with a pinhole aperture of 0.75 µm. Confocal Raman Microscopy. When cell confluence reached around 70%, the culture was trypsinized and 100000 cells were seeded in a glass bottom Petri dish for Raman experiments. After 24 h, the cells were incubated with UV sterilized NPs at a concentration of 150 µg/ mL of PLGA NPs for 12 h. Afterward, the plate was rinsed with PBS several times and the cells were fixed with 3.7% formaldehyde in PBS during 30 min and rinsed again. Micro-Raman analyses were performed using a Renishaw inVia Raman Microscope. Measurements were performed using the 532 nm laser excitation wavelength with a grating of 1800 mm-1. Spectra were taken using a 40× water immersion objective. The size of the focal spot was approximately 1 µm. Raman spectra were recorded in the region 300-3600 cm-1 with a resolution of approximately 7 cm-1. The system was calibrated to the spectral line of crystalline silicon at 520.7 cm-1. At least 8-15 accumulation scans at different spots in the various cell compartments, lipid bodies (LB), cytoplasm, and nucleus were taken to reduce the spectral noise. Control spectra of the different cell compartments were taken in cells not exposed to NPs. All spectra were corrected taking into account the PBS solution and glass coverslip baseline. Cell Viability. Four different human tumor cell lines were used for the cell viability study: Lymphoblastoid B (Hmy2), acute T lymphoblastic leukemia (Jurkat), myeloid-monocytic lymphoma (U937), and

PLGA Nanoparticles for Intracellular Delivery

Biomacromolecules, Vol. 11, No. 11, 2010

2995

Figure 1. (a) Size distribution by DLS and (b) TEM morphology of PLGA NPs stabilized with BSA.

prostate adenocarcinoma (PC3), obtained from American Type Culture Collection. Quick cell proliferation testing solution was used to evaluate the cell viability and proliferation. All cell lines were maintained in RPMI (Gibco, Life Technologies, Grand Island, Scotland) supplemented with 10% heat inactivated fetal bovine serum (FBS; PAA, Linz, Austria), penicillin (100 U/mL), streptomycin (100 µg/mL), and glutamine (2 mM; Gibco) at 37 °C in a humidified atmosphere containing 5% CO2. When culture confluence was around 70%, cells were seeded in 96-well plates in different amounts, depending on their proliferation rate as follows: PC3, 25.000 cells/well; Jurkat, 10.000 cells/ well; Hmy, 10.000 cells/well; and U937, 20.000 cells/well. After 24 h, NPs were added until a final concentration of 50 µg/mL was obtained, with the exception of the control, and incubated for 24 and 48 h. Then, the cell suspension from the plate was centrifuged 1 min at 1000 g and 100 µL of supernatant was removed. A total of 50 µL of Quick Cell Counting Solution was added to each well and incubated for 3 h. The reaction product was estimated by measuring its absorbance at 425 nm in a microplate reader (Multiskan EX, BioAnalysis Labsytems). Results are shown as percent of viability (V%) according to the following formula (blank discounted):

absorbance(cells + medium + NPs) V% )

absorbance(medium + NPs) × 100 absorbance(cells + medium)

In the case of HepG2, the cell viability was studied with the MTT assay. First, around 5000 HepG2 cells were planted into each well of 96-well plates, and after 24 h, PLGA NPs covered with PEI or BSA were added into culture media at different concentrations (0, 20, 50, 100, 200 µg/mL). After 3 days of coincubation, 20 mL of MTT solution (5 mg/mL in 10 mM PBS) was added into each well of the plates and incubated for 3 h. The absorbance was measured by means of a plate reader at 550 nm.

Results and Discussion NP Characterization. The size and morphology of PLGA NPs prepared with different concentrations of PEI or BSA were characterized by TEM and dynamic light scattering (DLS). The DLS distribution and TEM images show diameters of 450 and 270 nm, respectively, for BSA stabilized PLGA NPs, 3% BSA (Figure 1). TEM images of NPs stabilized with PEI are shown in Supporting Information (Figure S.1). DLS (Figure 1) revealed in general higher diameters for the NPs than those observed in TEM. This can be explained by the presence of hydrated stabilizers resulting in a larger hydrodynamic radius. For PEI-stabilized NPs, the diameter varied from ∼650 to ∼400 nm when the concentration of PEI increased from 1 to 5% (Figure S.2A). When BSA was used as the stabilizer, the diameter slightly decreased from 520 to 450 nm with increasing BSA concentration from 1 to 3% (Figure S.2B).

Figure 2. (A) Cellular uptake ratio and (B) normalized fluorescence intensity maximum for PLGA (PEI) NPs as function of the incubation time.

The variations between different batches were remarkably reduced with increasing BSA concentration. From the concentration dependence of the NP size, one may conclude that the BSA concentration affected the NP size to a lesser extent than the PEI concentration. This is probably related to the amphiphilic character of BSA, which is the reason for its frequent use as a surfactant in emulsions.32,33 The zeta-potential of the PEI-stabilized NPs showed values between +40 and +45 mV, being almost independent of the concentration of PEI used for stabilization. For BSA-stabilized NPs, the zeta-potential was negative and remained almost constant around -25 mV, independent of the amount of added BSA (Figure S.2). Cell Uptake and Nanoparticle Internalization. Flow cytometry data show that the uptake of BSA- and PEI-stabilized NPs labeled with rhodamine reaches a plateau during the first hours of incubation. For BSA the uptake was faster than for PEI-stabilized NPs reaching a plateau value corresponding to about 90% of the cell population within 6 h (Figure 2A). For PEI-stabilized NPs the uptake ratio increased more gradually and did not reach the values observed for BSA-stabilized NPs. The cell uptake ratio has been defined here as the percentage of cells, which are associated with NPs, having them either internalized or attached to the cell membrane. Alternatively, one may take the fluorescence intensity as a measure of uptake.

2996

Biomacromolecules, Vol. 11, No. 11, 2010

Romero et al.

Figure 3. CLSM images of HepG2 cells after being cocultured with (A) PLGA NPs stabilized with PEI and (B) PLGA NPs stabilized with BSA.

When comparing the net fluorescence of BSA- and PEIstabilized PLGA NPs it has to be taken into account that these two particle preparations have a priori different fluorescence intensities per NP. We measured the fluorescence distribution of the BSA- and PEI-stabilized PLGA NPs with the flow cytometer, employing the same settings as the experiments with cells (Figure S.3). The result was that the PEI-stabilized PLGA NPs have a fluorescence intensity almost three times the intensity of the BSA-stabilized PLGA NPs, implying that the latter had a lower degree of labeling than the PEI-stabilized NPs. In Figure 2B, consequently, we displayed the normalized intensity curves of PEI-stabilized PLGA NPs, taking the BSA-stabilized labeled particles as the reference. The data in Figure 2B thus provide a measure of the number of particles taken up by a single cell. In the case of BSA stabilized PLGA NPs the intensity reached a maximum within 3 h and then went down. The uptake of PEI stabilized NPs increased, on the contrary, more steadily reaching finally values comparable to BSA stabilized particles. These results are consistent with the cell uptake ratio, where one observes that the percentage of cells with NPs associated increases faster for BSA modified NPs than for PEI stabilized PLGA NPs. CLSM was used to have a direct visualization of the association of NPs with HepG2 cells. In Figure 3A the confocal images shows that PLGA NPs stabilized with PEI or BSA are associated with the cells. This follows from the red color indicating the rhodamine labeled NPs distributed around the blue stained nucleus. Similar images can be obtained for the PLGA NPs stabilized with BSA shown in Figure 3B. Because of the intrinsic difference in fluorescence between the PEI- and BSAstabilized particles the intensity of the red fluorescence in Figure 3B is weaker than in Figure 3A, however, the normalized plot in Figure 2B shows that the uptake of BSA-stabilized PLGA NPs per cell is not smaller than the uptake of PEI-stabilized PLGA NPs. Although flow cytometry proves the association of both BSAand PEI-modified PLGA NPs with cells, it is not clear if the NPs are really incorporated into the cell and have crossed the membrane barrier or if they remain only attached to the cell membrane. The CLSM images in Figure 3, where the nucleus and NPs have been labeled with DAPI (blue) and Rd6G (red), respectively, are representative of the behavior of the NPs in contact with Hep2G cells. The images clearly show that the NPs do not reach the nucleus, but an unambiguous conclusions about the environment in which the NPs are located within the cell or if they simply remained attached to the cell membrane cannot be drawn. On the other hand, the internalization of the NPs as a function of their surface modification is certainly a key issue regarding their potential use as delivery devices.

Figure 4. Raman spectrum recorded at different positions within a cell from the HepG2 line (ν indicates streching and δ deformation vibration modes; (l) denote vibrations of lipids and p of protein). The inset corresponds to the image of the cell under study. The numbers together with the color indicate the location in the cell at which the Raman spectra were recorded.

Therefore, alternative methods capable of proving internalization and intracellular location are highly desirable. Peisheng Xu et al.28 showed that PLGA NPs stabilized with poly(vinyl alcohol) (PVA) are not internalized by the cells. Employing CARS, they did not find the typical signal of PLGA at 2940 cm-1 within the cell cytoplasm, where the spectra of lipid bodies can be well detected. They concluded that although experiments with fluorescently labeled NPs hint the uptake of PLGA NPs, the latter remained attached to the cell membrane and did not enter the cytoplasm. Instead of using CARS and looking at defined Raman bands within the cell we applied spontaneous Raman confocal microscopy at different planes of the cell and at different spots in a plane. The advantage of spontaneous confocal Raman microscopy is that at each spot the whole spectrum is recorded. The disadvantage of this technique compared to CARS is that it takes much more time. The nucleus, cytoplasm, and lipid bodies can be identified by their chemical signature provided by the Raman spectrum. In the high frequency region the intensity ratio of the symmetric stretch bands of CH2 (2850 cm-1) to CH3 (2935 cm-1) is much higher in lipid bodies than in the cytoplasm. This can be explained by the lower density of CH2 groups in proteins compared with lipids. The nucleus region reveals the smallest intensity ratio of CH2 to CH3 bands and specific bands assigned as vibration of DNA bases of adenine (A) and guanine (G) (Figure 4). Figure 4 displays the changes in the Raman Spectra as we move to different positions within the cell. Then, cells incubated with PLGA NPs, which were either coated with PEI or BSA, respectively, were studied with

PLGA Nanoparticles for Intracellular Delivery

Figure 5. Spot Raman spectra (dark blue) in cells exposed to PLGA NPs covered with PEI (A) and BSA (B). In both cases, pink and green lines denote the component spectra of PLGA NPs and of the cells. The insets correspond to the image of the cell under study.

confocal Raman microscopy. The blue curves in Figure 5A, corresponding to PEI-stabilized PLGA NPs and in Figure 5B, corresponding to BSA-stabilized PLGA NPs, display features from the region of the lipid bodies as it is deduced from the relation of the bands between 2850 and 2900 cm-1, but they also show new bands in the methyl-methylene stretching region. These bands correspond to the characteristic CH2 and CH3 vibrations of PLGA NPs. For comparison, the spectra of the pure PLGA NPs and of the lipid bodies region have been included in Figure 5A,B (pink and green lines denote the spectra of PLGA NPs and the lipid body environment, respectively). It is clear that the dark blue curve displays the main bands of both the pure PLGA NPs and the lipid body region. From Figure 5 it can be concluded that it is possible to prove the colocalization of the PGLA NPs and lipid bodies within the dimensions of the confocal spot. Figure 6 displays Raman spectra of two spots taken at different Z-positions in cells which were exposed to PLGA NPs stabilized with either PEI or BSA. While moving away from the zero plane, the relative intensity of the signals of PLGA to that of the lipid bodies diminishes as we can observe for both the BSA- and the PEI-coated NPs. For the PEI-coated NPs, when moving downward, a region rich of lipid bodies is reached, where, however, NPs have not been detected as can be seen in Figure 6A. For the BSA-stabilized PLGA the cytoplasm is reached 5 µm below the zero plane and again NPs were not detected. The Z-scanning provides an ambiguous proof of internalization of the NPs because we move in distances of micrometers inside the cell, where the detection of PLGA attached to the cell membrane from the outside is very unlikely. The analysis of the Raman spectra revealed a colocalization of both PEI- and BSA-stabilized NPs with lipid bodies in the sense that the PLGA NPs were always found together with lipid

Biomacromolecules, Vol. 11, No. 11, 2010

2997

bodies. However, the reverse conclusion cannot be drawn. Furthermore, a colocalization of the PLGA NPs and the nucleus region was never observed. The reason, why BSA- and PEI-stabilized NPs were internalized while the PVA-stabilized NPs were not, is most likely related to the particular surface coating employed. It is known that polyelectrolytes carrying cationic amines like PEI can interact with lipid membranes crossing the membrane barrier. PEI is, for example, a widely used vector for DNA delivery in cell culture.34,35 Most likely, the PEI molecules around the PLGA NPs form a spherical brush, with chains protruding out of the surface of the NPs.25,27 These chains will have some conformational freedom, which can facilitate their interaction with the plasma membrane. It is also well-known that the positive charge of the NPs will attract proteins of the media, mostly BSA but also others like the apolipoprotein E, which can bind to a specific receptor in liver cells and this could facilitate uptake.36,37 For BSA stabilized NPs protein adsorption has to be less. Their negative surface charge will diminish the binding to the generally negatively charged cell surface. Therefore, a priori a smaller interaction with the cell surface as compared with PEI can be expected. Nevertheless, we observe that the number of cells, which have incorporated NPs is higher for BSA-modified NPs than for the PEI-modified ones. It is difficult to give a proper explanation to this observation at this point but it is probably related to a mechanism of uptake linked to a receptor.38,39 Cell Viability. Cell viability of PLGA NPs stabilized with PEI and BSA was first studied in HepG2 cells by means of the MTT assay after 3 days of coincubation (Figure S.4). For both PEI- and BSA-stabilized PLGA NPs the cell viability decreased with higher NPs concentrations. At 100 µg/mL, the highest concentration of NPs, cell viability was around a 50% for PEIstabilized PLGA NPs and almost 60% for BSA-stabilized PLGA NPs, indicating that PEI-stabilized PLGA NPs have a higher cytotoxicity than BSA-stabilized PLGA NPs. Cell toxicity is a critical aspect regarding the use of NPs in drug delivery. Therefore, we have extended cell viability experiments to several lines, applying other tests than MTT.40 The cell viability for NPs stabilized with PEI and BSA for four different cell lines are compared in Figure 7. Cell viability was measured after 1 day of exposure to the NPs (Figure 7A) and after 2 days of exposure (Figure 7B). It is interesting to notice how the viability changes for the different cell types. For example, the cell viability of the Hmy line was not affected by the NPs while the cell viability of PC3 was around 90% for BSA stabilized NPs (Figure 7A). Also, we see that for PEIstabilized NPs the viability of the Jurkat line is close to 80%, while in Hmy is less than 40%. For each cell line the viability in presence of NPs stabilized with PEI is always lower than for NPs stabilized with BSA. For the Hmy and U937 lines this difference is most pronounced. After a day of exposure to NPs we observe viabilities of less than 40% for PEI-coated NPs and over 100 for BSA-coated NPs. The smallest difference in viability within a cell line is observed for the PC3 line, where the viability for the PEI-stabilized NPs is close to 90% and the viability of the BSA-stabilized NPs is close to 100%. Cell viability behavior after two days of exposure of the cells to NPs differed in some cases from what is observed after a day of exposure, but no particular correlation can be established between one and two days exposure of the cells to NPs. To resume cell viability of PEI-stabilized NPs was always lower than that of BSA-stabilized NPs for all cell lines, which means that the cytotoxicity of these NPs was higher than the toxicity

2998

Biomacromolecules, Vol. 11, No. 11, 2010

Romero et al.

Figure 6. Z scans of HepG2 cells treated with PLGA NPs covered with PEI (A) and with BSA (B). The height level, Z ) 0, refers to the plane where the intensity at 2950 cm-1 referring to the PLGA NPs is maximal. The transmission images in the inset correspond to the scanned cells; the arrow points out the position where the measurement has been performed in the cell.

of the NPs stabilized with BSA. PEI molecules are cytotoxic even condensed with DNA,41,42 therefore, it is understandable that the PEI coating on the surface of PLGA presents certain toxicity within the cell lines tested our work. On the other hand, albumin, as the most abundant protein in serum, has excellent cytocompatibility for most of cell types. This would explain that when the NPs stabilized with BSA were added into the cell culture medium, almost no cytotoxicity could be found. BSA-coated NPs seem to have a positive effect on cellular viability, which reaches values over the 100% for several cell lines. From their moderate to low toxicity, the PLGA NPs fabricated here, especially those stabilized with BSA, are excellent candidates for intracellular drug delivery, as we have proved by CRM, that can be internalized within cells.

Conclusions By means of Confocal Raman microscopy we have shown that PLGA NPs either stabilized with PEI or with BSA can be internalized in HepG2 cells. This behavior is different from other stabilized NPs, which are only retained at the surface of the cells. The Raman data unambiguously prove the interpretation given by flow cytometry and confocal laser scanning microscopy that the NPs were internalized. PEI-stabilized NPs were taken over a longer period of time compared with BSA-stabilized PLGA NPs. Confocal Raman microscopy proves that both kinds of PLGA NPs are found in the region of the lipid bodies. Cell viability studies show that PEI-stabilized NPs are moderately toxic, while BSA-stabilized NPs observe a very good biocompatibility, resulting in cell viability percentages comparable or

Figure 7. Viability of different cell lines incubated with PLGA nanoparticles prepared by PEI or BSA as stabilizer after (A) 1 day and (B) 2 days.

PLGA Nanoparticles for Intracellular Delivery

better than the control. These results envisage the use of PEI/ BSA-coated PLGA NPs for intracellular delivery. Acknowledgment. This work was supported by the European Commission in the framework of FP7 Theme 4 - NMP Nanosciences, Nanotechnologies, Materials and New Production Technologies, Proposal No. CP-FP 228825-2 HINAMOX, as well as by Grant MAT 2007-60458 from the Spanish Ministry of Science and Innovation. S.E.M. is a Ramon y Cajal Fellow and he thanks this program of the Spanish Ministry of Science and Innovation for support. Supporting Information Available. TEM images of PEIstabilized PLGA NPs, data on the size and zeta potential dependence of the PLGA NPs with the percentage of stabilizator, the fluorescence distribution of rhodamine-labeled PEIand BSA-stabilized PLGA measured by flow cytometry and MTT data. This material is available free of charge via the Internet at http://pubs.acs.org.

References and Notes (1) Hans, M. L.; Lowman, A. M. Curr. Opin. Solid State Mater. Sci. 2002, 6, 319–327. (2) Langer, R. Science 1990, 249, 1527–1533. (3) Uhrich, K. E.; Cannizzaro, S. M.; Langer, R. S.; Shakesheff, K. M. Chem. ReV. 1999, 99, 3181–3198. (4) Ishihara, T.; Mizushima, T. Expert Opin. Drug DeliVery 2010, 7, 565– 575. (5) Patil, Y. B.; Swaminathan, S. K.; Sadhukha, T.; Ma, L. A.; Panyam, J. Biomaterials 2010, 31, 358–365. (6) Yuan, X. D.; Li, L.; Rathinavelu, A.; Hao, J. S.; Narasimhan, M.; He, M.; Heitlage, V.; Tam, L.; Viqar, S.; Salehi, M. J. Nanosci. Nanotechnol. 2006, 6, 2821–2828. (7) Cohen, H.; Levy, R. J.; Gao, J.; Fishbein, I.; Kousaev, V.; Sosnowski, S.; Slomkowski, S.; Golomb, G. Gene Ther. 2000, 7, 1896–1905. (8) Jalil, R.; Nixon, J. R. J. Microencapsulation 1990, 7, 297–325. (9) Allemann, E.; Leroux, J. C.; Gurny, R.; Doelker, E. Pharm. Res. 1993, 10, 1732–1737. (10) Zambaux, M. F.; Bonneaux, F.; Gref, R.; Maincent, P.; Dellacherie, E.; Alonso, M. J.; Labrude, P.; Vigneron, C. J. Controlled Release 1998, 50, 31–40. (11) Crotts, G.; Sah, H.; Park, T. G. J. Controlled Release 1997, 47, 101– 111. (12) Panyam, J.; Labhasetwar, V. AdV. Drug DeliVery ReV. 2003, 55, 329– 347. (13) Scholes, P. D.; Coombes, A. G. A.; Illum, L.; Davis, S. S.; Vert, M.; Davies, M. C. J. Controlled Release 1993, 25, 145–153. (14) Weiss, B.; Schneider, M.; Muys, L.; Taetz, S.; Neumann, D.; Schaefer, U. F.; Lehr, C. M. Bioconjugate Chem. 2007, 18, 1087–1094. (15) Mundargi, R. C.; Babu, V. R.; Rangaswamy, V.; Patel, P.; Aminabhavi, T. M. J. Controlled Release 2008, 125, 193–209. (16) Bala, I.; Hariharan, S.; Kumar, M. N. V. R. Crit. ReV. Ther. Drug Carrier Syst. 2004, 21, 387–422.

Biomacromolecules, Vol. 11, No. 11, 2010

2999

(17) Yang, Y. Y.; Chia, H. H.; Chung, T. S. J. Controlled Release 2000, 69, 81–96. (18) Hornig, S.; Heinze, T.; Becer, C. R.; Schubert, U. S. J. Mater. Chem. 2009, 19, 3838–3840. (19) Bilati, U.; Allemann, E.; Doelker, E. Eur. J. Pharm. Sci. 2005, 24, 67–75. (20) Jensen, D. M. K.; Cun, D.; Maltesen, M. J.; Frokjaer, S.; Nielsen, H. M.; Foged, C. J. Controlled Release 2010, 142, 138–145. (21) Fessi, H.; Puisieux, F.; Devissaguet, J. P.; Ammoury, N.; Benita, S. Int. J. Pharm. 1989, 55, R1–R4. (22) O’Donnell, P. B.; McGinity, J. W. AdV. Drug DeliVery ReV. 1997, 28, 25–42. (23) Zambaux, M. F.; Bonneaux, F.; Gref, R.; Dellacherie, E.; Vigneron, C. J. Controlled Release 1999, 60, 179–188. (24) Doiron, A. L.; Homan, K. A.; Emelianov, S.; Brannon-Peppas, L. Pharm. Res. 2009, 26, 674–682. (25) Zhou, J.; Moya, S.; Ma, L.; Gao, C. Y.; Shen, J. C. Macromol. Biosci. 2009, 9, 326–335. (26) Zhou, J.; Romero, G.; Rojas, E.; Ma, L.; Moya, S.; Gao, C. Y. J. Colloid Interface Sci. 2010, 345, 241–247. (27) Zhou, J.; Romero, G.; Rojas, E.; Moya, S.; Ma, L.; Gao, C. Y. Macromol. Chem. Phys. 2010, 211, 404–411. (28) Xu, P. S.; Gullotti, E.; Tong, L.; Highley, C. B.; Errabelli, D. R.; Hasan, T.; Cheng, J. X.; Kohane, D. S.; Yeo, Y. Mol. Pharm. 2009, 6, 190–201. (29) Chernenko, T.; Matthaus, C.; Milane, L.; Quintero, L.; Amiji, M.; Diem, M. ACS Nano 2009, 3, 3552–3559. (30) Cartiera, M. S.; Johnson, K. M.; Rajendran, V.; Caplan, M. J.; Saltzman, W. M. Biomaterials 2009, 30, 2790–2798. (31) Estrela-Lopis I., Romero G. Rojas E., Llarena I., Donath E., Moya S. E. , 2010, submitted for publication. (32) Boury, F.; Ivanova, T.; Panaiotov, I.; Proust, J. E. Langmuir 1995, 11, 599–606. (33) He, C. X.; Liu, J. H.; Ye, X. D.; Xie, L. Y.; Zhang, Q. L.; Ren, X. Z.; Zhang, G. Z.; Wu, C. Langmuir 2008, 24, 10717–10722. (34) Jiang, H. L.; Kwon, J. T.; Kim, Y. K.; Kim, E. M.; Arote, R.; Jeong, H. J.; Nah, J. W.; Choi, Y. J.; Akaike, T.; Cho, M. H.; Cho, C. S. Gene Ther. 2007, 14, 1389–1398. (35) Jiang, H. L.; Kim, Y. K.; Arote, R.; Nah, J. W.; Cho, M. H.; Choi, Y. J.; Akaike, T.; Cho, C. S. J. Controlled Release 2007, 117, 273– 280. (36) Mokuno, H.; Yamada, N.; Shimano, H.; Ishibashi, S.; Mori, N.; Takahashi, K.; Oka, T.; Yoon, T. H.; Takaku, F. Biochim. Biophys. Acta 1991, 1082, 63–70. (37) Allemann, E.; Gravel, P.; Leroux, J. C.; Balant, L.; Gurny, R. J. Biomed. Mater. Res. 1997, 37, 229–234. (38) Cu, Y.; LeMoellic, C.; Caplan, M. J.; Saltzman, W. M. Nanomed. Nanotechnol. Biol. Med. 2010, 6, 334–343. (39) Wright, T. L.; Lysenko, N.; Ockner, R. K.; Weisiger, R. A. Hepatology 1987, 7, 294–301. (40) Diaz, B.; Sanchez-Espinel, C.; Arruebo, M.; Faro, J.; de Miguel, E.; Magadan, S.; Yague, C.; Fernandez-Pacheco, R.; Ibarra, M. R.; Santamaria, J.; Gonzalez-Fernandez, A. Small 2008, 4, 2025–2034. (41) Beyerle, A.; Merkel, O.; Stoeger, T.; Kissel, T. Toxicol. Appl. Pharmacol. 2010, 242, 146–154. (42) Mok, H.; Park, T. G. Macromol. Biosci. 2009, 9, 731–743.

BM1007822