Single-Step Process to Produce Surface-Functionalized Polymeric

Langmuir 2007, 23, 12275-12279

12275

Single-Step Process to Produce Surface-Functionalized Polymeric Nanoparticles Eric M. Sussman,†,| Michael B. Clarke, Jr.,‡,§ and V. Prasad Shastri*,‡,| UniVersity of PennsylVania, Departments of Bioengineering and Materials Science and Engineering, Philadelphia, PennsylVania 19104, Rohm and Haas Company, Spring House, PennsylVania 19477, and Vanderbilt UniVersity, Department of Biomedical Engineering, NashVille, Tennessee 37235 ReceiVed July 5, 2007. In Final Form: August 20, 2007 Nanoparticles (NPs) are a versatile medium for the localization of therapeutics to tumors and for cellular and tissue imaging. The ability to impart targeting capability or enhance cellular uptake is dependent in part on the presentation of relevant surface functionality, among other design parameters. Currently, the production of functionalized polymeric NPs requires the a priori synthesis of polymers bearing such functionality. Here we describe a process to produce functionalized polymeric NPs derived from nonfunctional polymers in a single step. This was achieved by tailoring the solvation of the polymer using a binary solvent system such that the addition of an aqueous phase rich in watersoluble polymer or polyelectrolytes results in the formation of NPs with the concomitant functionalization of NP surfaces with the polymeric moieties introduced into the aqueous phase. This strategy also allows for easy control over NP size independent of surface functionality. We have demonstrated that poly(lactic-co-glycolic acid) (PLGA) NPs bearing surface functionality as diverse as biological polysaccharides such as heparin, water-soluble ionic polymers, and poly(ethylene glycol) can be prepared under identical conditions in a single step, with surface coverage (mass %) ranging from 3 to >70%. We expect this novel process to enable complex surface engineering of NP chemistry that hitherto was impossible using existing approaches.

Introduction Nanoparticles (NPs) are a versatile medium for drug delivery and imaging.1-3 Efficacy in these applications is largely determined by their ability to avoid clearance by the Reticulo Endothelial System (RES) and localize in regions of maximum therapeutic value, both of which are determined by the surface chemistry of the NP.4,5 By enriching an NP surface with poly(ethylene glycol) (PEG), uptake by the RES can be significantly diminished.6,7 The surface chemistry of NPs also plays a role in transport across biological barriers such as the blood-brain barrier8,9 and in tumor targeting.10,11 Several studies have shown that polyions (e.g., poly(acrylic acid) and poly(lysine)) and polysaccharides (e.g., heparin) can play an important role in targeting moieties to tumors.12 However, the preparation of NPs with specific biomolecules such as a heparin-rich surface requires the synthesis of polymers possessing such functionality through * Corresponding author. E-mail: [email protected], [email protected]. † University of Pennsylvania, Department of Bioengineering. ‡ University of Pennsylvania, Department of Materials Science and Engineering. § Rohm and Haas Company. | Vanderbilt University. (1) Langer, R. Nature 1998, 392, 5-10. (2) Langer, R. Science 2001, 293, 58-59. (3) Moinard-Checot, D.; Chevalier, Y.; Briancon, S.; Fessi, H.; Guinebretiere, S. J. Nanosci. Nanotechnol. 2006, 6, 2664-2681. (4) Otsuka, H.; Nagasaki, Y.; Kataoka, K. AdV. Drug DeliVery ReV. 2003, 55, 403-419. (5) Otsuka, H.; Nagasaki, Y.; Kataoka, K. Langmuir 2004, 20, 11285-11287. (6) Avgoustakis, K. Curr. Drug DeliVery 2004, 1, 321-333. (7) Gref, R.; Minamitake, Y.; Peracchia, M. T.; Trubetskoy, V.; Torchilin, V.; Langer, R. Science 1994, 263, 1600-1603. (8) Brannon-Peppas, L.; Blanchette, J. O. AdV. Drug DeliVery ReV. 2004, 56, 1649-1659. (9) Fenart, L.; Casanova, A.; Dehouck, B.; Duhem, C.; Slupek, S.; Cecchelli, R.; Betbeder, D. J. Pharmacol. Exp. Ther. 1999, 291, 1017-1022. (10) Hattori, Y.; Maitani, Y. Cancer Gene Ther. 2005, 12, 796-809. (11) Pan, D.; Turner, J. L.; Wooley, K. L. Chem. Commun. (Cambridge) 2003, 2400-2401. (12) Mehvar, R. Curr. Pharm. Biotechnol. 2003, 4, 283-302.

complex synthetic routes7,13 or post-functionalization of the NP with the moiety of interest.14,15 NPs are typically produced by a solvent-evaporation process called spontaneous emulsification solvent diffusion (SED).3,16 NPs bearing surface functionality may also be prepared by a layer-by-layer (LBL) assembly of polyelectrolytes.17,18 This process is severely influenced by pH and the charge characteristics of the adsorbing species and can be time-consuming.19 The SED process is carried out in the presence of a nonionic surfactant such as poly(vinyl alcohol) (PVA) with the exception of when amphiphilic polymers such as PEG copolymers are employed. The role of PVA is to stabilize the NP during its formation and solidification through the formation of a shell around the NP. Achieving functionality on an NP surface in the SED process, however, is contingent upon the introduction of the functional moiety a priori into the polymer backbone6,7,11 and its subsequent segregation onto the NP surface during NP formation, which is precluded by the presence of PVA. In this article, we describe a process to produce surfacefunctionalized nanoparticles from degradable polymers in a single step, without the need to synthesize polymers with backbone functionality. It occurred to us that surface functionality could be introduced into a nanoparticle via entrapment of polyelectrolytes if a rapid phase inversion and solidification of the polymer phase could accompany such entrapment. We hypothesized that this may be achieved by the addition of an aqueous phase (13) Mazid, M. A.; Moase, E.; Scott, E.; Hanna, H. R.; Unger, F. M. J. Biomed. Mater. Res. 1991, 25, 1169-1181. (14) Stolnik, S.; Dunn, S. E.; Garnett, M. C.; Davies, M. C.; Coombes, A. G.; Taylor, D. C.; Irving, M. P.; Purkiss, S. C.; Tadros, T. F.; Davis, S. S.; Illum, L. Pharm. Res. 1994, 11, 1800-1808. (15) Vandorpe, J.; Schacht, E.; Dunn, S.; Hawley, A.; Stolnik, S.; Davis, S. S.; Garnett, M. C.; Davies, M. C.; Illum, L. Biomaterials 1997, 18, 1147-1152. (16) Murakami, H.; Kobayashi, M.; Takeuchi, H.; Kawashima, Y. Int. J. Pharm. 1999, 187, 143-152. (17) Caruso, F. AdV. Mater. 2001, 13, 11-22. (18) Decher, G. Science 1997, 277, 1232-1237. (19) Burke, S. E.; Barrett, C. J. Pure Appl. Chem. 2004, 76, 1387-1398.

10.1021/la701997x CCC: $37.00 © 2007 American Chemical Society Published on Web 10/27/2007

12276 Langmuir, Vol. 23, No. 24, 2007

Sussman et al.

(nonsolvent for the polymer) to a ternary system composed of two solvents with a combined polarity close to water and a polymer if the mixture resulted in the formation of an instantaneous stable microemulsion. Under such conditions, the thermodynamically stable microemulsion would serve as the matrix for the nucleation and growth of the polymeric nanoparticles, and any soluble species in the aqueous phase would then be trapped in the solidifying polymeric particulate phase, thus resulting in the functionalization of NPs in a single step. In addition, this proposed approach overcomes major limitations of current systems, including the need to use a surfactant to stabilize the NPs during the preparation and tedious synthesis of functionalized polymers.10 Experimental Procedures Materials. Poly(DL-lactide-co-glycolide) (PLGA, RG 503, MW ) 30 000) and poly(L-lactide) (PLA) MW ) 70 000, inherent viscosity 1.20 dL/g in CHCl3) were purchased from Birmingham Polymers (Birmingham, AL) and were purified by precipitation from methylene chloride in methanol prior to use. Tetrahydrofuran (THF), acetone (Ac), and 1-methyl-2-pyrolidone (NMP) were purchased from either Aldrich (Sigma-Aldrich, Milwaukee, WI) or Fisher (Fisher Scientific, Pittsburgh, PA) and used as received. All solvents were HPLC grade or the highest available purity. Poly(styrene sulfonate) (PSS, MW ) 70 000), poly(acrylic acid) (PAA, MW ) 2000), poly(L-lysine hydrochloride) (PLys, MW ) 22 100), poly(ethylene glycol) (PEG, MW ) 10 000), and porcine heparin were purchased from Sigma and used as received without further purification. Doubly distilled deionized (DI) water obtained from a Milli-Q water purification system (Millipore, Bedford, MA) was used throughout the study. Preparation of Nanoparticles. To prepare NPs, 1 mL of the aqueous phase was added to an equal volume of polymer dissolved in a binary solvent system. The typical polymer concentration was 10 or 20 mg/mL. The solvent pairs used in this study include THF/ acetone and NMP/acetone. The volumetric ratio of the solvent pair was optimized to yield NPs of various sizes. When surface functionalization was desired, the aqueous phase was supplemented with either a polyelectrolyte or a water-soluble polymer such as PEG at 0.05 w/v %. The resulting NP suspension had a blue tint (Tyndall effect), was purified by dialysis to remove organic components and any untrapped water-soluble species, and was concentrated further by dialysis to yield stable suspensions of ∼2.0 w/v %. Determination of NP Size and Zeta Potential (ζ). NP size and ζ were determined using a Malvern Zetasizer (3000HS, Malvern Instruments Ltd., Malvern, U.K.). All measurements were made in automatic mode, and the software supplied by the manufacturer was used to analyze the data. For size measurements, the NP suspension was diluted by a factor of 15 with DI water prior to analysis, and for ζ measurements, the pH of the NP suspension was adjusted to the desired pH using either HCl or NaOH prior to analysis. XPS Surface Analysis. For XPS analysis, 5 mL of the NP suspension in water was dialyzed against 500 mL of 50% ethanol, flash frozen in liquid nitrogen, and then lyophilized for 48 h to a powder. The NP powder was then placed on the sample stub, and a Kratos Axis-Ultra X-ray photoelectron spectrometer equipped with a monochromatic Al KR (1486 eV) X-ray source operating at 315 W (25 mA) was used to collect XPS data. High-resolution data was collected using a pass energy of 40 eV in 0.05 eV steps. The elemental composition was calculated, and curve-fitting routines were performed with CasaXPS software. Mass fraction of the functionalizing agents on the NP surface was determined by comparing the XPS spectra of functionalized NP with that of pure PLGA and functionalizing agent using CasaXPS software routine. Florescence Microscopy. DAPI, a water-soluble negatively charge fluorescent dye, was used to visualize the NPs. NPs containing DAPI were prepared by adding DAPI (Vecta Shield, Vector Laboratories, CA; solution in glycerol) to the aqueous phase prior to NP formation (3 drops in 5 mL of water). The NP suspension was dialyzed against deionized water for 48 h to remove free DAPI and then photographed using a Zeiss Axiophot fluorescence microscope at 400× magnification.

Figure 1. (a) Influence of acetone volume fraction in Sys I and Sys II on nanoparticle size (SD ( 50 nm). (b) Nanoparticle size as a function of increasing glycerol volume fraction (viscosity). Measurements reflect an average of at least three experiments.

Results Effect of Organic- and Aqueous-Phase Composition on NP Size. Drago’s solvent polarity index was utilized to select the binary solvent system that was capable of dissolving biodegradable polymer P(DL)LGA, one of the most commonly used polymers in injectable sustained release systems, at high concentrations (1-4 w/v %) while still exhibiting miscibility with water. Using this scale, we identified two solvent pairs that satisfied these requirements, namely, tetrahydrofuran/acetone (THF/Ac) (Sys I) and N-methylpyrrolidone/acetone (NMP/Ac) (Sys II) (polarity: water ≈ NMP > Ac > THF). The choice of these specific solvent pairs would also enable the verification of the role of water in NP formation, which is central to the hypothesis. In Sys I, increasing the volume fraction of acetone (υacetone) resulted in NPs of smaller mean diameter, whereas in Sys II, increasing υacetone yielded NPs of increasing mean diameter (Figure 1a). These results are consistent with what one would expect on the basis of the miscibility of the system with water because increased miscibility due to increased polarity (i.e., Sys I) should promote more rapid polymer-phase gelation (extendedcoil to collapsed-coil transition) and decreased miscibility due to lower polarity (i.e., Sys II) should slow down the kinetics of this gelation process (Figure 1a). All solvent systems yielding NPs will narrow the polydispersity index ranging from 0.05 to 0.09 (Supporting Information Table 1S). The effect of the aqueous-phase viscosity on NP size was also studied. The viscosity of the aqueous phase was modulated through the addition of glycerol, and its impact on NP size was studied (Figure 1b). A linear correlation between higher solution viscosity and increased NP was observed. More importantly, however, greater variability in NP size was observed upon increasing υglycerol. This is an expected outcome because increased aqueous-phase viscosity would impair water diffusion into the organic phase, making the sol-gel transition in the polymer phase less sharp.

Surface-Functionalized Polymeric Nanoparticles

Langmuir, Vol. 23, No. 24, 2007 12277

Figure 2. (a) Changes in the ζ potential of functionalized NPs as a function of pH. (b). Scanning electron micrographs of PLGA modified with PAA (A) and PSS (B) (scale bar 100 nm). The pIe of P-PEG could not be obtained because it coagulated at lower pH. Table 1. Nanoparticle (NP) Surface Characteristics: Correlation between the Isoelectric Point (pIe) of the NP Surface with the pKa of the Functional Groupa NP composition PLGA (P) P-PSS P-PAA P-PLys P-Hep P-PEG

pKa of surface group

pIe of the NP surface

∆ζ from PLGA

ζ at pH 7.4

∼2 ∼3.5 ∼10 2-4b

2.75 2.40 2.80 9.50 3.40 N/Ac

0 -0.35 +0.05 +6.75 +0.65

-26.7 -28.3 -26.2 16.9 -28.1 -28.3

a PSS, poly(styrene sulfonate); PAA, poly(acrylic acid); PLys, poly(Llysine); Hep, heparin; PEG, poly(ethyleneoxide). b Estimated. c NP coagulated before pIe could be determined.

Functionalization of the PLGA NP Surface. NP bearing various surface-bound functionalities such as PEG, heparin, poly(lysine), PSS, and PAA were prepared by incorporating the macromolecules bearing the functionality of choice (i.e., PEG, heparin, PSS, PAA, and PLys) at a low concentration of 0.05% during NP formation. The presence of the appropriate surface functionality was verified by measuring the isoelectric point (pIe) of the NP surface by mapping the ζ potential as a function of pH (Figure 2a, Table 1). As seen in Figure 2 and Table 1, pIe of the NP surface compared well with what would be expected on the basis of the ionizable moieties in the surface-bound functionality. A more quantitative, definitive verification was obtained by carrying out XPS analysis of the NP surface (Table 2, Figure 3). XPS analysis revealed not only information about the surface of the NP but also that very high surface coverage of functional moieties was attainable in some cases. For example, negatively charged high-molecular-weight species such as PSS yielded an excess of 50% surface coverage (Table 2, Supplementary Figure 1S). This suggests that the entrapment of functional groups on the NP surface during NP formation might be dictated by the molecular weight of the macromolecules bearing the functional groups. This is reasonable because an increase in the chain length of the macromolecule would improve entanglement and thus entrapment within the gelling polymer phase. However, before any broad conclusions can be drawn, further experimentation would be required and is currently ongoing. An important observation was that the morphology of the NP and the size were not significantly impacted by the introduction of a functionalization process (Figure 2b).

Discussion It was found that PLGA NPs could be prepared by the addition of water to both solvent-pair systems without the need for solvent evaporation or a hardening step. As described in the Experimental

Procedures section, in a typical process, rapid mixing of PLGA dissolved in an organic solvent pair with an equal volume of an aqueous phase resulted in the instantaneous formation of NPs that, when concentrated, yielded stable suspensions at even 2 w/v %. An interesting observation was that aging of the NP suspension did not result in an increase in NP size, suggesting that the solidification of the NP is rapid. The yields based on the initial polymer mass was >90%. We observed that with respect to NP size, increasing υacetone in Sys I decreased the NP size (R2 ) 0.996) whereas in Sys II it increased the NP size (R2 ) 0.918) (Figure 1a). This observation is consistent with a mechanism that involves the precipitation of the polymer (gelation) that is driven by the diffusion of water into the polymer solvation shell. In such a process, an increased rate of water diffusion should favor the faster transition of the polymer chains to a collapsed coil, yielding denser, smaller particles, whereas a diminution in water diffusion due to lower miscibility should slow down this process, resulting in larger particles. In fact, NPs ranging in size from 70-500 nm were obtained without the need for any steric stabilization agents. It is important to note that the ability to tune NP size is important for tumor targeting because NP size has been shown to be critical in the accumulation of NPs in tumor vasculature and within tumors through the passive “enhanced permeation retention” mechanism.6,13 The polydispersity index (PDI) of the NPs was determined and was found to be quite narrow, with values of less than 0.1 (Table 1S, Supporting Information). Further evidence to support a mechanism of NP formation through the diffusion of water was obtained by studying the effect of aqueous-phase viscosity on NP size. Upon increasing the viscosity of water by the addition of glycerol, which is capable of hydrogen bonding with water and hence is miscible with water, an increase in NP size was observed (Figure 1b). It was more pronounced at higher glycerol volume fractions and appears to be consistent with the slower diffusion of water with increased viscosity, resulting in a slower rate of nucleation and growth of NPs. We have carried out preliminary solution stability studies and have found that all of the functionalized PLGA-NP suspensions are stable over at least a 3 month period, as ascertained by visual inspection for aggregates and sediments and light scattering (data not shown). It has been our observation that once instability sets in rapid coagulation ensues and results in a translucent mass in the bottom of the test tube and a clear supernatant that does not have blue coloration. No such phase separation was observed in any of the samples over the 3 month period. On the basis of the above mechanistic insight, the formation of NPs using a water phase rich in synthetic polyions (0.05% w/v) was explored as a means of imparting functionality to the

12278 Langmuir, Vol. 23, No. 24, 2007

Sussman et al. Table 2. C 1s Composition of the NP Surface

COOR

COO-C-OR

PLGA std dev

38.2 0.0

36.7 0.0

nd

25.1 0.0

PLGA-PLys std dev

27.1 3.1

28.0 3.2

9.0 1.6

36.0 4.7

24 9

PLGA-PSS std dev

12.1 1.6

12.5 1.6

nd

75.5 3.2

66 4

PLGA-PAA std dev

36.9 0.7

35.5 0.7

nd

27.5 1.4

3 2

PLGA-heparin std dev

12.9 1.3

8.6 0.2

33.6 1.0

32.6 3.7

76 1

PLGA-PEG std dev

27.6 1.3

26.5 4.1

13.5 1.5

32.4 9.8

28 11

NP surface. It was observed that the introduction of polyions into the aqueous phase did not hinder the formation of NPs and had a minimal impact on the size and polydispersity of the NPs. Furthermore, we observed that NPs produced under these conditions possessed surface charge characteristics consistent with the chemical structure of the polyion in solution as determined by Zeta potential (ζ) measurements (Table 1). Specifically, the surface charge in these NPs exhibited a charge inversion close to the pKa of the ionizable group in the polyion (Figure 2a). This is consistent with our working hypothesis that surface functionality could be introduced into an NP via the entrapment of polyions from the water phase during NP formation. Using the approach described herein, NP bearing heparin, a naturally occurring anticoagulant, can be prepared under identical conditions (Figure 2a) without the a priori need to synthesize PLGA polymers bearing the heparin moiety. The functionalization of polymers with sugars is very challenging because of the complexity of sugar chemistry. The presence of heparin (Figure 1S, Supporting Information) and other functional polymers studied herein was verified using XPS. (XPS spectra of PLGA (without functionality), PEG, and heparin are shown, but XPS spectra of NP with PSS, PAA, and PLys are not shown.) The surface coverage of the functional polymers as determined by the contribution of the functional polymers to the mass of the NP surface ranged from 3 to >70% in the case of heparin (Table 2). To determine if the polyionic species was incorporated into the NP structure via physical entrapment or surface adsorption,

C-OR

C-Hx

percent of surface mass contributed by functional groups

we studied the changes in the ζ potential of PLGA-PLys as a function of increasing ionic strength. The choice of PLGAPLys for these studies was based on the rationale that because the unmodified PLGA surface has a negative ζ potential at pH 6 to 7 (pH range of the process) it is most likely to favor the electrostatic adsorption of polycations. Upon increasing the ionic strength of a PLGA-PLys NP suspension from 0.3 to 24 mM using potassium chloride, we observed that the surface charge characteristics were retained (ζbefore ) 48 mV; ζafter ) 34 mV), suggesting that the moieties contributing to the surface characteristics of the NP are not desorbed and hence are not electrostatically bound to the NP surface. However, the mechanism of surface entrapment needs to be investigated further, specifically with respect to the effect of the functional moieties’ molecular weight, lipophilicity, and molecular architecture on the efficiency of the NP-surface functionalization process. The ternary system described herein is not limited to the preparation of NPs with polyionic functionality but may be extended to include nonionic species such as poly(ethylene glycol) (PEG). NP bearing even low-molecular-weight PEG (MW ) 10 kDa) can be easily prepared by incorporating PEG into the water phase as a stable suspension without a significant impact on NP size. The presence of PEG functionality on the NP surface has been verified using XPS, as shown in Figure 3. Because the intended application of NPs is to deliver bioactive agents, we have verified that the process is amenable to the encapsulation of small water-soluble molecules using fluorescent dyes as models (Figure 2S, Supporting Information). In addition to P(DL)LGA, which is an amorphous polymer, functionalized NPs of poly(L-lactic acid), a highly crystalline biodegradable polymer with wide applications in drug delivery, have been prepared as well (data not shown).

Conclusions

Figure 3. XPS C 1s spectra of nanoparticles: (a) unmodified PLGA NP and (b) PEG-modified PLGA NP. The arrow in panel b indicates the new peak due to the -O-CH2 carbons from the PEG.

We have developed a process for producing surface-functionalized polymeric NP from polymers such as poly(lactic-coglycolide) and poly(L-lactic acid), whose backbones bear no functional groups. This was achieved by tailoring the solvation core of the polymer such that, upon the addition of a water-rich phase containing polyelectrolyte(s) or water-soluble polymer(s), NP formation proceeds concomitantly with NP surface modification. The method described herein could be useful in engineering NP delivery systems for tumor targeting and cellular imaging. However, to fully realize the pharmaceutical potential of this novel process in targeted therapy, future studies will focus on processing parameters to maximize the surface coverage of the functional moiety, developing conditions for the lyophilization

Surface-Functionalized Polymeric Nanoparticles

of NPs into freely flowing powders that are easily reconstituted, and long-term solution stability studies. Acknowledgment. This work was supported by the National Institutes of Health (R24-AI47739-03), the National Science Foundation REU Program (NSF DMR02-43676), and the Vanderbilt Institute for Integrative Biosystems Research and Education (VIIBRE). We thank Professor I-Wei Chen for access

Langmuir, Vol. 23, No. 24, 2007 12279

to the Zetasizer and the Rohm and Haas Company for the use of their XPS facility. Supporting Information Available: Nanoparticle size and polydispersity index for NPs produced using various solvent systems. XPS C 1s spectrum of the PLGA-heparing NP surface. Fluorescence image of NPs using the DAPI filter. This material is available free of charge via the Internet at http://pubs.acs.org. LA701997X