- Email: [email protected]
solvent evaporation
Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
Surfactive water-soluble copolymers for the preparation of controlled surface nanoparticles by double emulsion/solvent evaporation David Chognot 1 , Mich`ele L´eonard, Jean-Luc Six, Edith Dellacherie ∗ Laboratoire de Chimie Physique Macromol´eculaire, UMR CNRS-INPL 7568, ENSIC, BP 20451, 54001 Nancy Cedex, France Received 18 November 2005; received in revised form 4 April 2006; accepted 4 April 2006 Available online 2 May 2006
Abstract We have already shown that polylactide (PLA) nanoparticles covered with a hydrophilic polymeric layer can be prepared by simple emulsion/solvent evaporation by using amphiphilic copolymers as surfactants during the procedure. The external layer is then constituted by the hydrophilic part of the macromolecular surfactant. This kind of nanospheres is useful for the encapsulation of lipohilic molecules. The use of amphiphilic copolymers as surfactants in the preparation of PLA nanospheres with controlled surface properties, was then applied to the double emulsion/solvent evaporation procedure. The aim was to allow the encapsulation of water-soluble bioactive molecules in PLA particles with controlled surface properties. In this paper, we describe the results obtained with three different water-soluble monomethoxypolyethylene oxide (MPEO)-b–PLA diblock copolymers used as surfactants in the preparation of nanoparticles by double emulsion/solvent evaporation. After organic solvent evaporation, the obtained nanospheres were proved to be really covered by a MPEO layer whose characteristics were determined. It was firstly shown that the MPEO-covered particles did not flocculate at 25 ◦ C, even in 4 M NaCl while suspensions of bare nanospheres were destabilized for a NaCl concentration as low as 0.04 M. On the other hand, the suspensions of MPEO-covered nanoparticles in 0.3 M Na2 SO4 were found to be very sensitive to temperature as they flocculated at a temperature lying between 45 and 55 ◦ C depending on the MPEO-b–PLA composition. This property was attributed to the fact that MPEO is a polymer with a low critical solution temperature. The concentration of MPEO at the nanoparticle surface was then calculated for the three kinds of particles, from the initial flocculation temperature, and was found to be comparable to the value determined directly. © 2006 Elsevier B.V. All rights reserved. Keywords: Polymeric surfactants; Double emulsion/solvent evaporation; MPEO-covered PLA nanoparticles; Colloidal stability
1. Introduction Polymer-based colloids or nanoparticles have received considerable attention as drug delivery systems, particularly by the intravenous route. To overcome the main drawback of these systems which is related to their rapid sequestration by the mononuclear phagocyte system, and thus to improve their in vivo life time, surface-modified nanoparticles have been developed. In this aim, various hydrophilic polymers have been used for particle coating. Among them, polyethylene oxide (PEO) has been widely studied and it has been proved to be an efficient
∗ 1
Corresponding author. Tel.: +33 383175221; fax: +33 383379977. E-mail address: [email protected] (E. Dellacherie). Present address: Flamel Technologies, 69693 V´enissieux Cedex, France.
0927-7765/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2006.04.003
polymer to limit adsorption of plasma proteins and cells [1–4]. Other water-soluble polymers such as polysaccharides were also used as particle external layers: thus coating with dextran was found to decrease protein adsorption [5,6] and to modify the interactions with proteins [7], and heparin increases the blood circulation time [8]. Copolymers of dextran and monomethoxypolyethylene oxide (MPEO) also limit protein adsorption on polystyrene nanospheres [6]. The main procedures which can lead to surface-modified nanoparticles, are firstly the adsorption of water-soluble amphiphilic copolymers on preformed hydrophobic particles [5,6,9]. Another method based on emulsion/solvent evaporation or nanoprecipitation procedures, consists in provoking the self-assembly of non water-soluble amphiphilic copolymers [10–15] or, as we recently showed it, in using water-soluble amphiphilic copolymers [5,16–18]. In particular, we proved that
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
PLA nanospheres covered with MPEO could be obtained by a simple emulsion, with PLA in the organic phase and biocompatible water-soluble MPEO-b–PLA diblock copolymers as surfactants in the water phase [18]. These copolymers were constituted by low molar mass PLA chains (between 1100 and 1700 g/mol) and relatively high MPEO chains (between 5000 and 20,000 g/mol). However, when prepared by simple emulsion/solvent evaporation, the nanoparticles can only be loaded with lipophilic molecules [19,20]. To encapsulate water-soluble molecules, one can proceed by w/o/w double emulsion as already described [21]. By using this procedure and in order to control the surface properties of particles, we have replaced the classical stabilizer of the secondary emulsion, polyvinylalcohol (PVA), by biocompatible/biodegradable water-soluble amphiphilic copolymers. We thus expected that the external layer would be constituted of the hydrophilic part of the polymeric surfactant instead of PVA which is known to be non degradable in vivo [22]. This present paper deals with the preparation of PLA nanospheres by the double w/o/w emulsion using biocompatible/biodegradable water-soluble MPEO-b–PLA diblock copolymers as surfactants in the secondary emulsion. By this way, it was expected that the MPEO block would be at the particle surface after solvent evaporation. These nanoparticles were characterized in terms of size, zeta potential and MPEO layer thickness and concentration. 2. Materials and methods 2.1. Materials ¯ n = 20, 000 g/mol = 20 K, as given by the MPEO20K (M ¯ n = 10, 000 g/mol = 10 K), MPEO5K supplier), MPEO10K (M ¯ n = 5000 g/mol = 5 K), PLA45K constituting the particle (M ¯ n = 45, 000 g/mol), d,l-lactide and all the reagents and core (M solvents used for the synthesis and the emulsions, were obtained from the suppliers indicated by Chognot et al. [18]. Bovine serum albumin (BSA) was purchased from Sigma (St. Quentin Fallavier, France). 2.2. Polymer synthesis MPEO-b–PLA samples were synthesized by pseudo-anionic ring-opening polymerization of d,l-lactide from the hydroxyl end group of MPEO after activation with stannous octoate, according to a procedure already described [18,23]. ¯ n and M ¯ w of the copolymers The average molar masses M were determined using 1 H NMR and multiangle laser light scattering (MALLS) coupled to size exclusion chromatography (SEC) according to Chognot et al. [18]. Surface tension measurements were carried out at 25 ◦ C with a K8 surface tensiometer (Kr¨uss, Germany) using the Wilhelmy technique. All samples were equilibrated for a sufficient time to reach constant readings (5 min to 1 h depending on the copolymer). All values were expressed as means of three experiments.
87
2.3. Preparation of nanospheres by the double emulsion/solvent evaporation procedure Typically, a primary w/o emulsion was prepared by mixing an organic phase (4 mL ethylacetate) containing PLA (25 g/L) with an aqueous internal phase (0.2 mL) containing BSA (50 g/L). The mixture was then sonicated (10 W, continuous mode, 10 s, in an ice bath) using a Vibracell model 600 W (Sonics & Materials, Danbury, USA). This primary emulsion was poured into a second aqueous phase (8 mL) (external aqueous phase) containing a water-soluble MPEOb–PLA copolymer (0.5–5 g/L). The w/o/w emulsion was then obtained by sonication (10 W, pulsed mode, 20 s, in an ice bath). This double emulsion was then transferred into an aqueous dispersing phase (40 mL) and stirred for 5 min. The organic solvent was evaporated under vacuum for 30 min and the collected solid nanospheres were resuspended in water, then centrifuged again in order to remove the excess MPEO-b–PLA. This purification procedure was repeated two times. 2.4. Physico-chemical characterization of nanospheres The size distribution of nanospheres (mean diameter and polydispersity index) was determined by photon correlation spectroscopy (PCS) using a Malvern spectrometer 4600 (Malvern Instruments, UK) in 10−3 M NaCl at 30 ◦ C. Their electrophoretic mobility was investigated using a Malvern Zetasizer 4, in NaCl, as a function of ionic strength. The zeta potential (ζ) was calculated from the modified Booth equation [24] and the electrokinetic MPEO layer thickness (d ) was determined from the Eversole and Boardman equation [25] as already described by Chognot et al. [18]. The specific surface area (Sp ) of the nanosphere in m2 /g was calculated according to: Sp =
3 rρ
where r is the nanosphere mean radius determined by PCS and ρ the nanosphere density (here ρPLA = 1.26 × 106 g/m3 as measured by a helium pycnometer). The amount of MPEO-b–PLA at the particle surface was directly evaluated by UV spectrophotometry at 500 nm using the I2 /KI assay (1% I2 , 2% KI, w/v) [26] after PLA hydrolytic degradation of nanoparticles (5 mg) in 1 M NaOH (1 mL) for 5 h. The MPEO-b–PLA surface coverage (Γ MPEO ) of the nanospheres was calculated using: mMPEO ΓMPEO = mNS Sp where mMPEO is the MPEO amount at the particle surface and mNS the mass of degraded particles. PLA is not water-soluble unlike MPEO and the two polymers are not compatible as proved by a differential scanning calorimetry analysis. One can thus assume that during the secondary emulsion, the copolymer MPEO blocks are repelled to
88
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
the water phase while the PLA blocks are trapped into the PLA organic phase. After solvent evaporation, it is thus quite probable that all the copolymer MPEO chains contribute to form the particle external layer whereas the PLA blocks are anchored inside the PLA matrix. With this hypothesis, the average distance between two MPEO chains (d) and the surface area occupied by one MPEO chain at the particle surface (A) could be calculated according to the approach described by Chognot et al. [18] and using the following relationships: 1/2 M d= ; A = d2 ΓMPEO NA where M is the molecular weight of MPEO and NA the Avogadro number. The Flory radius of a MPEO chain at the particle surface, was calculated by: Rf = aN 3/5 where N is the MPEO polymerization degree and a the size of the EO monomer unit (here 0.4 nm [27]). The colloidal stability was studied by turbidimetry in NaCl solutions of various concentrations, using the method described by Long et al. [28]. This simple technique is based on the measurement of the solution turbidity in the 400–700 nm range. A general relationship was proposed by Heller and McCarthy [29] for the specific turbidity of a sol as a function of the wavelength (λ) of the light used: τ = kλ−n c c→0 where τ is the turbidity at concentration c, k a constant which is a function in turn of particle size and relative refractive index particle-medium. For particles of diameter much smaller than the used wavelength (<λ/10), n = 4 (Rayleigh scattering). For larger particles, n gradually decreases passing through zero and then oscillates positive and negative. This parameter may therefore be used as an indication of particle size and will be sensitive to any flocculation occurring in the system. From the Heller equation and considering that k is constant over the wavelength range considered, a plot of log τ versus log λ yields an approximate value of n. Experimentally, particles were added to NaCl solutions at a concentration of 0.005% weight at 25 ◦ C. The optical density (OD) was then measured in the 400–700 nm range, every 50 nm, and n was calculated as the slope of the log (OD)/log λ straight line. The critical flocculation concentration (CFC) was determined at the slope breaking of the n versus [NaCl] plots. The influence of temperature on flocculation was studied by similar experiments but in 0.3 M Na2 SO4 instead of NaCl, in the 10–66 ◦ C range. The initial flocculation temperature (IFT) was determined at the slope breaking of the n versus temperature plots. 3. Results and discussion To prepare PLA nanospheres by double emulsion/solvent evaporation procedure, one generally uses polyvinylalcohol
(PVA) to stabilize the secondary emulsion. However, PVA adsorbs at the materials surface [22] and cannot be easily removed. On the other hand, the presence of PVA at the surface is a favourable parameter as it hampers the particle aggregation in aqueous suspensions (sterical repulsion between nanospheres) as well as the adsorption of proteins onto the PLA matrix. However, PVA is toxic in vivo and it is not accepted for intravenous administration. To change the surface composition, nanospheres were prepared by double emulsion using an organic phase containing mixtures of PLA and non water-soluble MPEO-b–PLA copolymers. PVA was needed to stabilize the secondary emulsion. The obtained particles were thus constituted of a PLA core and a MPEO shell [11]. We have developed a general method for controlling the composition of the nanosphere external layer by using surfactive water-soluble copolymers designed to lead to the expected shell and specially synthesized for it. This method has been already applied to the preparation of PLA nanoparticles by simple emulsion and we employed biocompatible and biodegradable polymers such as water-soluble MPEO-b–PLA or watersoluble amphiphilic polysaccharides to stabilize the emulsion. We thus showed that nanospheres could be easily formed and that after solvent evaporation, the particles possessed an external hydrophilic layer constituted of MPEO [18] or of polysaccharides [5,14,17]. This paper describes the application of this procedure to the preparation of nanoparticles by double emulsion. 3.1. Properties of the MPEO-b–PLA copolymers used as stabilizers of the secondary emulsion Three water-soluble MPEO-b–PLA copolymers with different MPEO molar masses (5 kg/mol, 10 kg/mol and 20 kg/mol) and a high MPEO weight content, were prepared as described by Chognot et al. [18]. One of the objectives was to examine the influence of the MPEO length on the surface properties of the nanoparticles. The characteristics of the three copolymers are seen in Table 1. To determine if these copolymers possessed surface active properties, their behaviour at the air/water interface was studied at steady state as a function of the copolymer concentration. Fig. 1 shows that the three copolymers exhibit a behaviour similar to that already described for polymeric surfactants [30]. Thus, water surface tension slowly decreases when polymer concentration increases, from a value of about 72 mN/m down to 47–50 mN/m, and then levels off. Various water-soluble amphiphilic copolymers constituted of a PEO block and of a hydrophobic one such as polypeptides [31], poly(ε-caprolactone) [32] or PLA [33] have been already described for applications in drug delivery systems as well as other ones constituted for example of a block of PLA and a block poly(N-vinyl-2-pyrrolidone) [34]. But at our knowledge, these copolymers have been always employed under the form of micelles directly usable as delivery vehicles and not for the stabilization of emulsions aimed at the nanosphere preparation.
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
89
Table 1 Characteristics of the water-soluble MPEO-b–PLA copolymers used in the double emulsion/solvent evaporation experiments* Copolymer used as surfactant
¯ n MPEO a (g/mol) M
MPEO contentb (wt%)
¯ n PLA c (g/mol) M
¯ n MPEO−PLA d (g/mol) M
¯ w MPEO−PLA d (g/mol) M
MPEO20K MPEO20K MPEO20K
19350 11000 5000
92 91 77
1700 1100 1500
20750 12250 8150
21300 12550 8650
K: kg/mol. a Determined by SEC-MALLS in 0.1 M NaNO . 3 b Determined by 1 NMR according to Chognot et al. [18]. c Determined from 1 NMR data and M ¯ n MPEO values. d Determined by SEC-MALLS in tetrahydrofuran as described by Chognot et al. [18]. * By permission of D. Chognot, J.-L. Six, M. L´ eonard, F. Bonneaux, C. Vigneron, and E. Dellacherie.
cally with PLA, should minimize the adsorption of a protein of interest to be encapsulated, at the water/organic solvent interface and then on the hydrophobic PLA surface, which should limit its denaturation [36]. The effect of BSA on the emulsion stability was investigated and it was found that as expected, the emulsion was only stable when prepared in the presence of both BSA in the aqueous phase and PLA in the organic phase. In this case the droplet average diameter was found between 100 and 300 nm, depending on BSA concentration.
Fig. 1. Plots of surface tension (σ) of aqueous polymer solutions for increasing concentrations of MPEO20K –PLA1.7K (♦), MPEO10K –PLA1.1K () and MPEO5K –PLA1.5K ().
3.2. Nanosphere preparation and characterization 3.2.1. Stabilization of the primary w/o emulsion The primary emulsion is formed between an aqueous phase containing the bioactive molecule and an organic phase (ethylacetate) containing PLA. To stabilize this emulsion, rather than using an emulsifier dissolved in the organic phase together with PLA, we preferred to add bovine serum albumin (BSA) to the aqueous phase. In fact it was proved that, in the presence of PLA in the organic phase, BSA rapidly formed a strong film at the w/o interface thus stabilizing the emulsion in a more efficient manner than F68 Pluronic did when dissolved in the organic phase [35]. Another interest to use BSA, is that the formation of such an interfacial film in which BSA interacts hydrophobi-
3.2.2. Nanosphere characterization Nanospheres were obtained after the evaporation of ethylacetate of the double emulsion stabilized by one of the three copolymers described above. Their main characteristics are gathered in Table 2. The copolymer concentration in the aqueous phase, needed to reach the maximal value of the MPEO coverage, was determined for each copolymer, according to the results presented in Fig. 2. It should be noted that, under the conditions described in the experimental part, nanospheres with an average diameter close to 200 nm and a low dispersity index could be obtained. Various other parameters were determined following the approach already described for nanoparticles prepared by simple emulsion/solvent evaporation [18]. The values calculated in the case of double emulsion were similar to those determined for the simple emulsion, except that they were obtained with higher MPEO-b–PLA concentrations in the aqueous phase (3–5 g/L for the double emulsion against 2 g/L for the simple one). It can be seen that here also the ζ potential is close to zero even at low ionic strength (10−2 M NaCl), whereas that of bare nanospheres is close to −45 mV under similar conditions. This means that in particles prepared with MPEO-b–PLA surfactants, the –COO−
Table 2 Characteristics of nanospheres prepared by the double emulsion/solvent evaporation procedure in the presence of the various copolymers Copolymer used as surfactant
dH (nm)
Ip
ζ (mV)
CMPEO–PLA (g/L)
Γ MPEO (mg/m2 )
d (nm)
d (nm)
A (nm2 /MPEO chain)
Rf (nm)
MPEO20K –PLA1.7K MPEO10K –PLA1.1K MPEO5K –PLA1.5K
229 ± 2 195 ± 0.5 199 ± 2
0.09 0.11 0.14
−1.8 −2.1 −2.4
5 5 3
2.6 1.6 2.2
30 16 13
3.6 3.2 2
13 10 4
13.1 10.4 6.9
dH : hydrodynamic diameter (measured by PCS); Ip : polydispersity index; ζ: zeta potential in 10−2 M NaCl; CMPEO–PLA : minimal MPEO-b–PLA concentration in the external aqueous phase allowing the maximal MPEO surface coverage (determined after optimisation studies); Γ MPEO : maximal MPEO coverage; dζ : electrokinetic thickness of the MPEO layer; d: average distance between two MPEO chains at the particle surface; A: surface area occupied by one MPEO chain at the particle surface; Rf : Flory radius of a MPEO chain.
90
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
Fig. 2. Influence of the MPEO-b–PLA concentration in the external aqueous phase, on the nanoparticle coverage by MPEO (Γ ). (): MPEO20K –PLA1.7K ; (): MPEO10K –PLA1.1K ; (䊉): MPEO5K –PLA1.5K . Values expressed as means of three experiments.
groups of the PLA core are masked by the MPEO chains present at the surface. Finally, the spacing between two MPEO chains (d) is lower than the theoretical Flory radius (Rf ) of MPEO in the three cases. This means that the MPEO chains overlap and probably adopt a brushlike conformation. Moreover, it can also be seen in Table 2, that, as expected, a decrease in the MPEO length results in a decrease of the MPEO layer electrokinetic thickness, of the surface area of one MPEO chain at the surface and of the average distance between two MPEO chains. On the other hand, as ζ potential is very low for the three MPEO lengths, no significant influence of this parameter on the surface charge can be pointed out. 3.3. Nanosphere stability 3.3.1. Stability during freeze-drying The storage of nanoparticles implies that in fine they could be obtained as a stabilized dried powder. In fact, the storage of aqueous suspensions of particles containing water-soluble active molecules, cannot be envisaged because of the huge risks of diffusion in the supernatant. Freeze-drying has thus been considered and we studied the influence of this treatment on the characteristics of nanospheres. When the nanoparticle suspensions were freeze-dried as such, just after the last washing and centrifugation, aggregation phenomena were observed leading to a particle increased diameter, probably because of the crystallization of superficial MPEO (Fig. 3). To avoid this aggregation, sucrose, which is known for its protection effect during protein freeze-drying [37], was then added to the particle suspensions. The influence of the sucrose/nanosphere weight ratio on the dHf /dHi ratio – where dHf and dHi are the particle hydrodynamic diameters after and before freeze-drying, respectively, – is shown in Fig. 3. Indeed sucrose allows to avoid aggregation during freeze-drying probably because this disaccharide inhibits crystallization of MPEO in frozen solutions as already reported by Izutsu et al. [38]. The
Fig. 3. Influence of the sucrose/nanosphere weight ratio used during freezedrying, on the nanoparticle size. dHi = average hydrodynamic diameter before freeze-drying. dHf average hydrodynamic diameter after freeze-drying. ( ): MPEO20K –PLA1.7K ; ( ): MPEO5K –PLA1.5K ; ( ): MPEO10K –PLA1.1K . Values expressed as means of two experiments.
optimal sucrose/nanosphere weight ratio is between 5 and 16, depending on the copolymer used as surfactant. However, no satisfactory explanation was found for the strong aggregation observed at a sucrose/nanosphere weight ratio of 21. 3.3.2. Colloidal stability in NaCl solutions at 25 ◦ C The variation of the n parameter as a function of NaCl concentration for the particles obtained in the presence of the three polymeric surfactants, is described in Fig. 4. The results obtained with bare PLA nanoparticles (prepared with PVA as the stabilizer of the secondary emulsion) are also given for comparison. For bare nanoparticles, a flocculation phenomenon appears at a NaCl concentration close to 5 × 10−3 M and it is ended for 10−2 M
Fig. 4. Variation of n at 25 ◦ C as a function of NaCl concentration for bare PLA nanoparticles () and particles prepared in the presence of MPEO20K –PLA1.7K (䊉), MPEO10K –PLA1.1K () and MPEO5K –PLA1.5K (♦). CFC: critical flocculation concentration.
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
91
NaCl. At low ionic strength, the nanoparticles repeal each other because of the presence of PLA carboxylate groups at their surface. When the ionic strength is increased, the negative charges are progressively screened and the Van der Waals forces become preponderant, which leads to particle aggregation. On the other hand, with MPEO chains at the particle surface, no flocculation is observed when NaCl concentration is increased from 10−4 to 4 M, which proves that the suspension is sterically stabilized, whatever the characteristics of coverage (Γ MPEO from 1.6 to 2.6 mg/m2 and d from 13 to 30 nm, see Table 2). 3.3.3. Colloidal stability in aqueous solutions as a function of temperature PEO is a low critical solution temperature (LCST) polymer and LCST depends on the polymer molecular weight and on the presence of electrolytes in the water solutions. It was thus expected that the MPEO-covered nanoparticle suspensions could flocculate with increasing temperature because of the decreasing of the MPEO hydration sphere around the particles. For the study of the temperature effect, NaCl was replaced by Na2 SO4 because Na2 SO4 lowers the initial flocculation temperature (IFT) which can thus be in a range of measurable values. First experiments were then carried out to determine the salt concentration favouring the phase separation of the various MPEO (5 K, 10 K and 20 K) in aqueous Na2 SO4 solutions in a reasonable temperature range. A range of 20–200 g/L MPEO concentrations was chosen as it corresponds to the theoretical concentrations at the nanoparticle surface. Two Na2 SO4 concentrations, were studied, 0.5 M and 0.3 M. However, at 0.5 M, the cloud temperatures were relatively low (<0 ◦ C for a MPEO concentration higher than 100 g/L). The results of Fig. 5 concern therefore the 0.3 M concentration. It is noted that, as expected, the cloud temperature decreases when the MPEO molecular weight decreases or when the MPEO concentration increases. This 0.3 M Na2 SO4 concentration was thus chosen for the study on nanospheres.
Fig. 6. Stability in 0.3 M Na2 SO4 , of nanoparticles prepared with the different copolymers, as a function of temperature. MPEO20K –PLA1.7K (䊉), MPEO10K –PLA1.1K (♦), MPEO5K –PLA1.5K (). IFT: initial flocculation temperature.
Fig. 6 shows the influence of temperature on the stability of MPEO-covered nanoparticle suspensions in 0.3 M Na2 SO4 . The various suspensions do not flocculate when temperature is lower than about 40 ◦ C. However, above this temperature, n rapidly decreases, which corresponds to the particle flocculation. This means that, when temperature becomes higher than 40 ◦ C, 0.3 M Na2 SO4 is no more a good solvent for the superficial MPEO chains, which results in a decrease in layer solvency. Therefore, the chains shrink and the sterical stabilisation is no more efficient, which provokes the suspension flocculation. The initial flocculation temperature (IFT) is defined as shown in Fig. 6 and the IFT values obtained for the three types of particles are listed in Table 3. Assuming that at the IFT, the MPEO chains present at the nanoparticle surface, are at a concentration corresponding to the MPEO cloud point temperature, it is thus possible to use the results shown in Fig. 5 to evaluate the concentration of MPEO at the nanoparticle surface. The values of these concentrations for the three kinds of particles are given in Table 3 (method A). On the other hand, we directly calculated these concentrations by quantifying particle MPEO and using the data of Table 2 (in particular, the values of MPEO coverage and of electrokinetic MPEO layer thickness). The results are also shown in Table 3 (method B). It can be noted that the values calculated by the two methods are comparable as already found for some polystyrene nanoparticles coated with dextran–MPEO copolymers [6]. The fact that the MPEO concentrations calculated by method A are Table 3 MPEO concentration (cMPEO ) at the nanoparticle surface, calculated by two methods Copolymer used as surfactant
MPEO20K –PLA1.7K MPEO10K –PLA1.1K MPEO5K –PLA1.5K Fig. 5. Variation of the cloud temperatures of MPEOs with various molecular weights as a function of their concentration in 0.3 M Na2 SO4 . MPEO20K (); MPEO10K (); MPEO5K ().
IFT (◦ C)
53 52 46
cMPEO (g/L) A
B
110 170 210
90 100 170
(A) cMPEO was calculated from the values of the initial flocculation temperature (IFT) of the nanoparticle suspension. (B) The nanospheres were degraded by 1 M NaOH and MPEO was quantified by a I2 /KI assay. cMPEO was then calculated from data of Table 2 (Γ MPEO and d ).
92
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
higher than those obtained by method B could mean that the MPEO chains are really at the particle surface and have not be trapped inside the PLA matrix. 4. Conclusion We have shown that it is possible to prepare surface-modified PLA nanoparticles by using a double emulsion/solvent evaporation procedure in which the surfactant of the secondary emulsion is a water-soluble diblock MPEO-b–PLA copolymer. After evaporation of the organic solvent, the obtained nanospheres present at their surface MPEO chains probably in the form of a brush. These MPEO chains are strongly anchored inside the PLA matrix thanks to the PLA blocks of the copolymers. They confer to the aqueous particle suspensions, a good stability in the presence of high NaCl concentrations at 25 ◦ C. However, as MPEO is a polymer with a low critical solution temperature, the replacement of NaCl by Na2 SO4 , a salt lowering the cloud temperature of MPEO aqueous solutions, leads to the suspension flocculation at a temperature between 45 and 55 ◦ C depending on the copolymer composition. Concerning the influence of the copolymer MPEO length on the particle properties, we showed that the hydrophilic layer electrokinetic thickness, the surface MPEO concentration and the average distance between two MPEO chains at the surface, depend on the MPEO length. As it has been demonstrated that all these parameters influence the plasma protein adsorption [39], they will have to be considered for the design of long circulating nanoparticles based on this new way of preparation. The main advantage of the procedure here described, is that it can allow the encapsulation of water-soluble bioactive macromolecules such as for example proteins or polysaccharides, in PLA nanoparticles covered with any kind of hydrophilic polymer, provided that an appropriate amphiphilic copolymer is used as surfactant in the secondary emulsion. There is thus no need for other surfactants whose removal from the particle surface is often a problem. References [1] R. Gref, Y. Minamitake, M.T. Peracchia, V. Trubetskoy, V. Torchilin, R. Langer, Science 263 (1994) 1600. [2] R. Gref, G. Miralles, E. Dellacherie, Polym. Int. 48 (1999) 251. [3] M.-F. Zambaux, F. Bonneaux, R. Gref, E. Dellacherie, C. Vigneron, J. Biomed. Mater. Res. 44 (1999) 109. [4] S. Stolnik, C.R. Heald, J. Neal, M.C. Garnett, S.S. Davis, L. Illum, S.C. Purkiss, R.J. Barlow, P.R. Gellert, J. Drug Target 9 (2001) 361. [5] C. Rouzes, M. L´eonard, A. Durand, E. Dellacherie, Coll. Surf. B: Biointerf. 32 (2003) 125. [6] A. De Souza Delgado, M. L´eonard, E. Dellacherie, Langmuir 17 (2001) 4386.
[7] C. Lemarchand, R. Gref, C. Passirani, E. Garcion, B. Petri, R. M¨uller, D. Costantini, P. Couvreur, Biomaterials 27 (2006) 108. [8] C. Passirani, G. Baratt, J.-P. Devissaguet, D. Labarre, Pharm. Res. 15 (1998) 1046. [9] R.H. M¨uller, K.H. Wallis, Int. J. Pharm. 89 (1993) 25. [10] K. Akiyoshi, S. Kobayashi, S. Schichibe, D. Mix, M. Baudys, S.W. Kim, J. Sunamoto, J. Control. Release 54 (1998) 313. [11] P. Quellec, R. Gref, L. Perrin, E. Dellacherie, F. Sommer, J.-M. Verbavatz, M.J. Alonso, J. Biomed. Mater. Res. 42 (1998) 45. [12] M.-F. Zambaux, F. Bonneaux, R. Gref, E. Dellacherie, C. Vigneron, Int. J. Pharm. 212 (2001) 1. [13] C. Lemarchand, P. Couvreur, M. Besnard, D. Costantini, R. Gref, Pharm. Res. 20 (2003) 1284. [14] C. Nouvel, C. Frochot, V. Sadtler, P. Dubois, E. Dellacherie, J.-L. Six, Macromolecules 37 (2004) 4981. [15] I. Ydens, P. Degee, C. Nouvel, E. Dellacherie, J.-L. Six, P. Dubois, E-polymers 046 (2005) http://www.e-polymers.org. ISSN 1618 7229. [16] E. Rotureau, M. L´eonard, E. Dellacherie, A. Durand, Phys. Chem. Chem. Phys. 6 (2004) 1430. [17] A. Durand, E. Marie, E. Rotureau, M. L´eonard, E. Dellacherie, Langmuir 20 (2004) 6956. [18] D. Chognot, J.-L. Six, M. L´eonard, F. Bonneaux, C. Vigneron, E. Dellacherie, J. Coll. Interf. Sci. 268 (2003) 441. [19] A. Sanchez, J.L. Vila-Jato, M.J. Alonso, Int. J. Pharm. 99 (1993) 263. [20] T. G¨orner, R. Gref, D. Michenot, F. Sommer, T.M. Duc, E. Dellacherie, J. Control. Release 57 (1999) 259. [21] D. Blanco, M.J. Alonso, J. Eur. Pharm. Biopharm. 43 (1997) 287. [22] M.-F. Zambaux, F. Bonneaux, R. Gref, E. Dellacherie, C. Vigneron, J. Control. Release 60 (1999) 179. [23] P. Bouillot, A. Petit, E. Dellacherie, J. Appl. Polym. Sci. 68 (1998) 1695. [24] S.R. Deshiikan, K.D. Papadopoulos, Coll. Polym. Sci. 276 (1998) 117. [25] W.G. Eversole, W.W.J. Boardman, Chem. Phys. 9 (1941) 798. [26] B. Baleux, G.C.R. Champetier, Acad. Sci. Paris, S´er. C 274 (1972) 1617. [27] P. Molyneux, Water-Soluble Synthetic Polymers: Properties and Behaviour, CRC Press, Boca-Raton, FL, 1983. [28] J.A. Long, D.W.J. Osmond, B.J. Vincent, Coll. Interf. Sci. 42 (1972) 545. [29] W. Heller, H.J. McCarthy, Chem. Phys. 29 (1958) 78. [30] I. Piirma, Polymeric Surfactants, Surfactant Sciences Series, vol. 42, Marcel Dekker, Inc, New York, 1992, Chapter 5. [31] G. Kwon, M. Naito, M. Yokoyama, Y. Sakurai, R. Kataoka, J. Control. Release 48 (1997) 195. [32] C. Allen, Y.S. Yu, D. Maysinger, A. Eisenberg, Bioconjug. Chem. 9 (1998) 564. [33] S.A. Hagan, A.G.A. Coombes, M.C. Garnett, S.E. Dunn, M.C. Davies, L. Illum, S.S. Davis, S.E. Harding, S. Purkiss, P.R. Gellert, Langmuir 12 (1996) 2153. [34] A. Benahmed, M. Ranger, J.C. Leroux, Pharm. Res. 18 (2001) 323. [35] Ch. Schugens, N. Laruelle, N. Nihant, Ch. Grandfils, R. J´erome, P. Teyssi´e, J. Control. Release 32 (1994) 161. [36] H. Sah, J. Control. Release 58 (1999) 143. [37] J.L. Remmeb, C. Stushnoff, J.F. Carpenter, Pharm. Res. 14 (1997) 1548. [38] K. Izutsu, S. Yoshioka, S. Kojima, T.W. Randolph, J.F. Carpenter, Pharm. Res. 13 (1996) 1393. [39] R. Gref, M. L¨uck, P. Quellec, M. Marchand, E. Dellacherie, S. Harnisch, T. Blunk, R.H. M¨uller, Coll. Surf. B: Biointerf. 18 (2000) 301.
Surfactive water-soluble copolymers for the preparation of controlled surface nanoparticles by double emulsion/solvent evaporation David Chognot 1 , Mich`ele L´eonard, Jean-Luc Six, Edith Dellacherie ∗ Laboratoire de Chimie Physique Macromol´eculaire, UMR CNRS-INPL 7568, ENSIC, BP 20451, 54001 Nancy Cedex, France Received 18 November 2005; received in revised form 4 April 2006; accepted 4 April 2006 Available online 2 May 2006
Abstract We have already shown that polylactide (PLA) nanoparticles covered with a hydrophilic polymeric layer can be prepared by simple emulsion/solvent evaporation by using amphiphilic copolymers as surfactants during the procedure. The external layer is then constituted by the hydrophilic part of the macromolecular surfactant. This kind of nanospheres is useful for the encapsulation of lipohilic molecules. The use of amphiphilic copolymers as surfactants in the preparation of PLA nanospheres with controlled surface properties, was then applied to the double emulsion/solvent evaporation procedure. The aim was to allow the encapsulation of water-soluble bioactive molecules in PLA particles with controlled surface properties. In this paper, we describe the results obtained with three different water-soluble monomethoxypolyethylene oxide (MPEO)-b–PLA diblock copolymers used as surfactants in the preparation of nanoparticles by double emulsion/solvent evaporation. After organic solvent evaporation, the obtained nanospheres were proved to be really covered by a MPEO layer whose characteristics were determined. It was firstly shown that the MPEO-covered particles did not flocculate at 25 ◦ C, even in 4 M NaCl while suspensions of bare nanospheres were destabilized for a NaCl concentration as low as 0.04 M. On the other hand, the suspensions of MPEO-covered nanoparticles in 0.3 M Na2 SO4 were found to be very sensitive to temperature as they flocculated at a temperature lying between 45 and 55 ◦ C depending on the MPEO-b–PLA composition. This property was attributed to the fact that MPEO is a polymer with a low critical solution temperature. The concentration of MPEO at the nanoparticle surface was then calculated for the three kinds of particles, from the initial flocculation temperature, and was found to be comparable to the value determined directly. © 2006 Elsevier B.V. All rights reserved. Keywords: Polymeric surfactants; Double emulsion/solvent evaporation; MPEO-covered PLA nanoparticles; Colloidal stability
1. Introduction Polymer-based colloids or nanoparticles have received considerable attention as drug delivery systems, particularly by the intravenous route. To overcome the main drawback of these systems which is related to their rapid sequestration by the mononuclear phagocyte system, and thus to improve their in vivo life time, surface-modified nanoparticles have been developed. In this aim, various hydrophilic polymers have been used for particle coating. Among them, polyethylene oxide (PEO) has been widely studied and it has been proved to be an efficient
∗ 1
Corresponding author. Tel.: +33 383175221; fax: +33 383379977. E-mail address: [email protected] (E. Dellacherie). Present address: Flamel Technologies, 69693 V´enissieux Cedex, France.
0927-7765/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2006.04.003
polymer to limit adsorption of plasma proteins and cells [1–4]. Other water-soluble polymers such as polysaccharides were also used as particle external layers: thus coating with dextran was found to decrease protein adsorption [5,6] and to modify the interactions with proteins [7], and heparin increases the blood circulation time [8]. Copolymers of dextran and monomethoxypolyethylene oxide (MPEO) also limit protein adsorption on polystyrene nanospheres [6]. The main procedures which can lead to surface-modified nanoparticles, are firstly the adsorption of water-soluble amphiphilic copolymers on preformed hydrophobic particles [5,6,9]. Another method based on emulsion/solvent evaporation or nanoprecipitation procedures, consists in provoking the self-assembly of non water-soluble amphiphilic copolymers [10–15] or, as we recently showed it, in using water-soluble amphiphilic copolymers [5,16–18]. In particular, we proved that
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
PLA nanospheres covered with MPEO could be obtained by a simple emulsion, with PLA in the organic phase and biocompatible water-soluble MPEO-b–PLA diblock copolymers as surfactants in the water phase [18]. These copolymers were constituted by low molar mass PLA chains (between 1100 and 1700 g/mol) and relatively high MPEO chains (between 5000 and 20,000 g/mol). However, when prepared by simple emulsion/solvent evaporation, the nanoparticles can only be loaded with lipophilic molecules [19,20]. To encapsulate water-soluble molecules, one can proceed by w/o/w double emulsion as already described [21]. By using this procedure and in order to control the surface properties of particles, we have replaced the classical stabilizer of the secondary emulsion, polyvinylalcohol (PVA), by biocompatible/biodegradable water-soluble amphiphilic copolymers. We thus expected that the external layer would be constituted of the hydrophilic part of the polymeric surfactant instead of PVA which is known to be non degradable in vivo [22]. This present paper deals with the preparation of PLA nanospheres by the double w/o/w emulsion using biocompatible/biodegradable water-soluble MPEO-b–PLA diblock copolymers as surfactants in the secondary emulsion. By this way, it was expected that the MPEO block would be at the particle surface after solvent evaporation. These nanoparticles were characterized in terms of size, zeta potential and MPEO layer thickness and concentration. 2. Materials and methods 2.1. Materials ¯ n = 20, 000 g/mol = 20 K, as given by the MPEO20K (M ¯ n = 10, 000 g/mol = 10 K), MPEO5K supplier), MPEO10K (M ¯ n = 5000 g/mol = 5 K), PLA45K constituting the particle (M ¯ n = 45, 000 g/mol), d,l-lactide and all the reagents and core (M solvents used for the synthesis and the emulsions, were obtained from the suppliers indicated by Chognot et al. [18]. Bovine serum albumin (BSA) was purchased from Sigma (St. Quentin Fallavier, France). 2.2. Polymer synthesis MPEO-b–PLA samples were synthesized by pseudo-anionic ring-opening polymerization of d,l-lactide from the hydroxyl end group of MPEO after activation with stannous octoate, according to a procedure already described [18,23]. ¯ n and M ¯ w of the copolymers The average molar masses M were determined using 1 H NMR and multiangle laser light scattering (MALLS) coupled to size exclusion chromatography (SEC) according to Chognot et al. [18]. Surface tension measurements were carried out at 25 ◦ C with a K8 surface tensiometer (Kr¨uss, Germany) using the Wilhelmy technique. All samples were equilibrated for a sufficient time to reach constant readings (5 min to 1 h depending on the copolymer). All values were expressed as means of three experiments.
87
2.3. Preparation of nanospheres by the double emulsion/solvent evaporation procedure Typically, a primary w/o emulsion was prepared by mixing an organic phase (4 mL ethylacetate) containing PLA (25 g/L) with an aqueous internal phase (0.2 mL) containing BSA (50 g/L). The mixture was then sonicated (10 W, continuous mode, 10 s, in an ice bath) using a Vibracell model 600 W (Sonics & Materials, Danbury, USA). This primary emulsion was poured into a second aqueous phase (8 mL) (external aqueous phase) containing a water-soluble MPEOb–PLA copolymer (0.5–5 g/L). The w/o/w emulsion was then obtained by sonication (10 W, pulsed mode, 20 s, in an ice bath). This double emulsion was then transferred into an aqueous dispersing phase (40 mL) and stirred for 5 min. The organic solvent was evaporated under vacuum for 30 min and the collected solid nanospheres were resuspended in water, then centrifuged again in order to remove the excess MPEO-b–PLA. This purification procedure was repeated two times. 2.4. Physico-chemical characterization of nanospheres The size distribution of nanospheres (mean diameter and polydispersity index) was determined by photon correlation spectroscopy (PCS) using a Malvern spectrometer 4600 (Malvern Instruments, UK) in 10−3 M NaCl at 30 ◦ C. Their electrophoretic mobility was investigated using a Malvern Zetasizer 4, in NaCl, as a function of ionic strength. The zeta potential (ζ) was calculated from the modified Booth equation [24] and the electrokinetic MPEO layer thickness (d ) was determined from the Eversole and Boardman equation [25] as already described by Chognot et al. [18]. The specific surface area (Sp ) of the nanosphere in m2 /g was calculated according to: Sp =
3 rρ
where r is the nanosphere mean radius determined by PCS and ρ the nanosphere density (here ρPLA = 1.26 × 106 g/m3 as measured by a helium pycnometer). The amount of MPEO-b–PLA at the particle surface was directly evaluated by UV spectrophotometry at 500 nm using the I2 /KI assay (1% I2 , 2% KI, w/v) [26] after PLA hydrolytic degradation of nanoparticles (5 mg) in 1 M NaOH (1 mL) for 5 h. The MPEO-b–PLA surface coverage (Γ MPEO ) of the nanospheres was calculated using: mMPEO ΓMPEO = mNS Sp where mMPEO is the MPEO amount at the particle surface and mNS the mass of degraded particles. PLA is not water-soluble unlike MPEO and the two polymers are not compatible as proved by a differential scanning calorimetry analysis. One can thus assume that during the secondary emulsion, the copolymer MPEO blocks are repelled to
88
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
the water phase while the PLA blocks are trapped into the PLA organic phase. After solvent evaporation, it is thus quite probable that all the copolymer MPEO chains contribute to form the particle external layer whereas the PLA blocks are anchored inside the PLA matrix. With this hypothesis, the average distance between two MPEO chains (d) and the surface area occupied by one MPEO chain at the particle surface (A) could be calculated according to the approach described by Chognot et al. [18] and using the following relationships: 1/2 M d= ; A = d2 ΓMPEO NA where M is the molecular weight of MPEO and NA the Avogadro number. The Flory radius of a MPEO chain at the particle surface, was calculated by: Rf = aN 3/5 where N is the MPEO polymerization degree and a the size of the EO monomer unit (here 0.4 nm [27]). The colloidal stability was studied by turbidimetry in NaCl solutions of various concentrations, using the method described by Long et al. [28]. This simple technique is based on the measurement of the solution turbidity in the 400–700 nm range. A general relationship was proposed by Heller and McCarthy [29] for the specific turbidity of a sol as a function of the wavelength (λ) of the light used: τ = kλ−n c c→0 where τ is the turbidity at concentration c, k a constant which is a function in turn of particle size and relative refractive index particle-medium. For particles of diameter much smaller than the used wavelength (<λ/10), n = 4 (Rayleigh scattering). For larger particles, n gradually decreases passing through zero and then oscillates positive and negative. This parameter may therefore be used as an indication of particle size and will be sensitive to any flocculation occurring in the system. From the Heller equation and considering that k is constant over the wavelength range considered, a plot of log τ versus log λ yields an approximate value of n. Experimentally, particles were added to NaCl solutions at a concentration of 0.005% weight at 25 ◦ C. The optical density (OD) was then measured in the 400–700 nm range, every 50 nm, and n was calculated as the slope of the log (OD)/log λ straight line. The critical flocculation concentration (CFC) was determined at the slope breaking of the n versus [NaCl] plots. The influence of temperature on flocculation was studied by similar experiments but in 0.3 M Na2 SO4 instead of NaCl, in the 10–66 ◦ C range. The initial flocculation temperature (IFT) was determined at the slope breaking of the n versus temperature plots. 3. Results and discussion To prepare PLA nanospheres by double emulsion/solvent evaporation procedure, one generally uses polyvinylalcohol
(PVA) to stabilize the secondary emulsion. However, PVA adsorbs at the materials surface [22] and cannot be easily removed. On the other hand, the presence of PVA at the surface is a favourable parameter as it hampers the particle aggregation in aqueous suspensions (sterical repulsion between nanospheres) as well as the adsorption of proteins onto the PLA matrix. However, PVA is toxic in vivo and it is not accepted for intravenous administration. To change the surface composition, nanospheres were prepared by double emulsion using an organic phase containing mixtures of PLA and non water-soluble MPEO-b–PLA copolymers. PVA was needed to stabilize the secondary emulsion. The obtained particles were thus constituted of a PLA core and a MPEO shell [11]. We have developed a general method for controlling the composition of the nanosphere external layer by using surfactive water-soluble copolymers designed to lead to the expected shell and specially synthesized for it. This method has been already applied to the preparation of PLA nanoparticles by simple emulsion and we employed biocompatible and biodegradable polymers such as water-soluble MPEO-b–PLA or watersoluble amphiphilic polysaccharides to stabilize the emulsion. We thus showed that nanospheres could be easily formed and that after solvent evaporation, the particles possessed an external hydrophilic layer constituted of MPEO [18] or of polysaccharides [5,14,17]. This paper describes the application of this procedure to the preparation of nanoparticles by double emulsion. 3.1. Properties of the MPEO-b–PLA copolymers used as stabilizers of the secondary emulsion Three water-soluble MPEO-b–PLA copolymers with different MPEO molar masses (5 kg/mol, 10 kg/mol and 20 kg/mol) and a high MPEO weight content, were prepared as described by Chognot et al. [18]. One of the objectives was to examine the influence of the MPEO length on the surface properties of the nanoparticles. The characteristics of the three copolymers are seen in Table 1. To determine if these copolymers possessed surface active properties, their behaviour at the air/water interface was studied at steady state as a function of the copolymer concentration. Fig. 1 shows that the three copolymers exhibit a behaviour similar to that already described for polymeric surfactants [30]. Thus, water surface tension slowly decreases when polymer concentration increases, from a value of about 72 mN/m down to 47–50 mN/m, and then levels off. Various water-soluble amphiphilic copolymers constituted of a PEO block and of a hydrophobic one such as polypeptides [31], poly(ε-caprolactone) [32] or PLA [33] have been already described for applications in drug delivery systems as well as other ones constituted for example of a block of PLA and a block poly(N-vinyl-2-pyrrolidone) [34]. But at our knowledge, these copolymers have been always employed under the form of micelles directly usable as delivery vehicles and not for the stabilization of emulsions aimed at the nanosphere preparation.
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
89
Table 1 Characteristics of the water-soluble MPEO-b–PLA copolymers used in the double emulsion/solvent evaporation experiments* Copolymer used as surfactant
¯ n MPEO a (g/mol) M
MPEO contentb (wt%)
¯ n PLA c (g/mol) M
¯ n MPEO−PLA d (g/mol) M
¯ w MPEO−PLA d (g/mol) M
MPEO20K MPEO20K MPEO20K
19350 11000 5000
92 91 77
1700 1100 1500
20750 12250 8150
21300 12550 8650
K: kg/mol. a Determined by SEC-MALLS in 0.1 M NaNO . 3 b Determined by 1 NMR according to Chognot et al. [18]. c Determined from 1 NMR data and M ¯ n MPEO values. d Determined by SEC-MALLS in tetrahydrofuran as described by Chognot et al. [18]. * By permission of D. Chognot, J.-L. Six, M. L´ eonard, F. Bonneaux, C. Vigneron, and E. Dellacherie.
cally with PLA, should minimize the adsorption of a protein of interest to be encapsulated, at the water/organic solvent interface and then on the hydrophobic PLA surface, which should limit its denaturation [36]. The effect of BSA on the emulsion stability was investigated and it was found that as expected, the emulsion was only stable when prepared in the presence of both BSA in the aqueous phase and PLA in the organic phase. In this case the droplet average diameter was found between 100 and 300 nm, depending on BSA concentration.
Fig. 1. Plots of surface tension (σ) of aqueous polymer solutions for increasing concentrations of MPEO20K –PLA1.7K (♦), MPEO10K –PLA1.1K () and MPEO5K –PLA1.5K ().
3.2. Nanosphere preparation and characterization 3.2.1. Stabilization of the primary w/o emulsion The primary emulsion is formed between an aqueous phase containing the bioactive molecule and an organic phase (ethylacetate) containing PLA. To stabilize this emulsion, rather than using an emulsifier dissolved in the organic phase together with PLA, we preferred to add bovine serum albumin (BSA) to the aqueous phase. In fact it was proved that, in the presence of PLA in the organic phase, BSA rapidly formed a strong film at the w/o interface thus stabilizing the emulsion in a more efficient manner than F68 Pluronic did when dissolved in the organic phase [35]. Another interest to use BSA, is that the formation of such an interfacial film in which BSA interacts hydrophobi-
3.2.2. Nanosphere characterization Nanospheres were obtained after the evaporation of ethylacetate of the double emulsion stabilized by one of the three copolymers described above. Their main characteristics are gathered in Table 2. The copolymer concentration in the aqueous phase, needed to reach the maximal value of the MPEO coverage, was determined for each copolymer, according to the results presented in Fig. 2. It should be noted that, under the conditions described in the experimental part, nanospheres with an average diameter close to 200 nm and a low dispersity index could be obtained. Various other parameters were determined following the approach already described for nanoparticles prepared by simple emulsion/solvent evaporation [18]. The values calculated in the case of double emulsion were similar to those determined for the simple emulsion, except that they were obtained with higher MPEO-b–PLA concentrations in the aqueous phase (3–5 g/L for the double emulsion against 2 g/L for the simple one). It can be seen that here also the ζ potential is close to zero even at low ionic strength (10−2 M NaCl), whereas that of bare nanospheres is close to −45 mV under similar conditions. This means that in particles prepared with MPEO-b–PLA surfactants, the –COO−
Table 2 Characteristics of nanospheres prepared by the double emulsion/solvent evaporation procedure in the presence of the various copolymers Copolymer used as surfactant
dH (nm)
Ip
ζ (mV)
CMPEO–PLA (g/L)
Γ MPEO (mg/m2 )
d (nm)
d (nm)
A (nm2 /MPEO chain)
Rf (nm)
MPEO20K –PLA1.7K MPEO10K –PLA1.1K MPEO5K –PLA1.5K
229 ± 2 195 ± 0.5 199 ± 2
0.09 0.11 0.14
−1.8 −2.1 −2.4
5 5 3
2.6 1.6 2.2
30 16 13
3.6 3.2 2
13 10 4
13.1 10.4 6.9
dH : hydrodynamic diameter (measured by PCS); Ip : polydispersity index; ζ: zeta potential in 10−2 M NaCl; CMPEO–PLA : minimal MPEO-b–PLA concentration in the external aqueous phase allowing the maximal MPEO surface coverage (determined after optimisation studies); Γ MPEO : maximal MPEO coverage; dζ : electrokinetic thickness of the MPEO layer; d: average distance between two MPEO chains at the particle surface; A: surface area occupied by one MPEO chain at the particle surface; Rf : Flory radius of a MPEO chain.
90
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
Fig. 2. Influence of the MPEO-b–PLA concentration in the external aqueous phase, on the nanoparticle coverage by MPEO (Γ ). (): MPEO20K –PLA1.7K ; (): MPEO10K –PLA1.1K ; (䊉): MPEO5K –PLA1.5K . Values expressed as means of three experiments.
groups of the PLA core are masked by the MPEO chains present at the surface. Finally, the spacing between two MPEO chains (d) is lower than the theoretical Flory radius (Rf ) of MPEO in the three cases. This means that the MPEO chains overlap and probably adopt a brushlike conformation. Moreover, it can also be seen in Table 2, that, as expected, a decrease in the MPEO length results in a decrease of the MPEO layer electrokinetic thickness, of the surface area of one MPEO chain at the surface and of the average distance between two MPEO chains. On the other hand, as ζ potential is very low for the three MPEO lengths, no significant influence of this parameter on the surface charge can be pointed out. 3.3. Nanosphere stability 3.3.1. Stability during freeze-drying The storage of nanoparticles implies that in fine they could be obtained as a stabilized dried powder. In fact, the storage of aqueous suspensions of particles containing water-soluble active molecules, cannot be envisaged because of the huge risks of diffusion in the supernatant. Freeze-drying has thus been considered and we studied the influence of this treatment on the characteristics of nanospheres. When the nanoparticle suspensions were freeze-dried as such, just after the last washing and centrifugation, aggregation phenomena were observed leading to a particle increased diameter, probably because of the crystallization of superficial MPEO (Fig. 3). To avoid this aggregation, sucrose, which is known for its protection effect during protein freeze-drying [37], was then added to the particle suspensions. The influence of the sucrose/nanosphere weight ratio on the dHf /dHi ratio – where dHf and dHi are the particle hydrodynamic diameters after and before freeze-drying, respectively, – is shown in Fig. 3. Indeed sucrose allows to avoid aggregation during freeze-drying probably because this disaccharide inhibits crystallization of MPEO in frozen solutions as already reported by Izutsu et al. [38]. The
Fig. 3. Influence of the sucrose/nanosphere weight ratio used during freezedrying, on the nanoparticle size. dHi = average hydrodynamic diameter before freeze-drying. dHf average hydrodynamic diameter after freeze-drying. ( ): MPEO20K –PLA1.7K ; ( ): MPEO5K –PLA1.5K ; ( ): MPEO10K –PLA1.1K . Values expressed as means of two experiments.
optimal sucrose/nanosphere weight ratio is between 5 and 16, depending on the copolymer used as surfactant. However, no satisfactory explanation was found for the strong aggregation observed at a sucrose/nanosphere weight ratio of 21. 3.3.2. Colloidal stability in NaCl solutions at 25 ◦ C The variation of the n parameter as a function of NaCl concentration for the particles obtained in the presence of the three polymeric surfactants, is described in Fig. 4. The results obtained with bare PLA nanoparticles (prepared with PVA as the stabilizer of the secondary emulsion) are also given for comparison. For bare nanoparticles, a flocculation phenomenon appears at a NaCl concentration close to 5 × 10−3 M and it is ended for 10−2 M
Fig. 4. Variation of n at 25 ◦ C as a function of NaCl concentration for bare PLA nanoparticles () and particles prepared in the presence of MPEO20K –PLA1.7K (䊉), MPEO10K –PLA1.1K () and MPEO5K –PLA1.5K (♦). CFC: critical flocculation concentration.
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
91
NaCl. At low ionic strength, the nanoparticles repeal each other because of the presence of PLA carboxylate groups at their surface. When the ionic strength is increased, the negative charges are progressively screened and the Van der Waals forces become preponderant, which leads to particle aggregation. On the other hand, with MPEO chains at the particle surface, no flocculation is observed when NaCl concentration is increased from 10−4 to 4 M, which proves that the suspension is sterically stabilized, whatever the characteristics of coverage (Γ MPEO from 1.6 to 2.6 mg/m2 and d from 13 to 30 nm, see Table 2). 3.3.3. Colloidal stability in aqueous solutions as a function of temperature PEO is a low critical solution temperature (LCST) polymer and LCST depends on the polymer molecular weight and on the presence of electrolytes in the water solutions. It was thus expected that the MPEO-covered nanoparticle suspensions could flocculate with increasing temperature because of the decreasing of the MPEO hydration sphere around the particles. For the study of the temperature effect, NaCl was replaced by Na2 SO4 because Na2 SO4 lowers the initial flocculation temperature (IFT) which can thus be in a range of measurable values. First experiments were then carried out to determine the salt concentration favouring the phase separation of the various MPEO (5 K, 10 K and 20 K) in aqueous Na2 SO4 solutions in a reasonable temperature range. A range of 20–200 g/L MPEO concentrations was chosen as it corresponds to the theoretical concentrations at the nanoparticle surface. Two Na2 SO4 concentrations, were studied, 0.5 M and 0.3 M. However, at 0.5 M, the cloud temperatures were relatively low (<0 ◦ C for a MPEO concentration higher than 100 g/L). The results of Fig. 5 concern therefore the 0.3 M concentration. It is noted that, as expected, the cloud temperature decreases when the MPEO molecular weight decreases or when the MPEO concentration increases. This 0.3 M Na2 SO4 concentration was thus chosen for the study on nanospheres.
Fig. 6. Stability in 0.3 M Na2 SO4 , of nanoparticles prepared with the different copolymers, as a function of temperature. MPEO20K –PLA1.7K (䊉), MPEO10K –PLA1.1K (♦), MPEO5K –PLA1.5K (). IFT: initial flocculation temperature.
Fig. 6 shows the influence of temperature on the stability of MPEO-covered nanoparticle suspensions in 0.3 M Na2 SO4 . The various suspensions do not flocculate when temperature is lower than about 40 ◦ C. However, above this temperature, n rapidly decreases, which corresponds to the particle flocculation. This means that, when temperature becomes higher than 40 ◦ C, 0.3 M Na2 SO4 is no more a good solvent for the superficial MPEO chains, which results in a decrease in layer solvency. Therefore, the chains shrink and the sterical stabilisation is no more efficient, which provokes the suspension flocculation. The initial flocculation temperature (IFT) is defined as shown in Fig. 6 and the IFT values obtained for the three types of particles are listed in Table 3. Assuming that at the IFT, the MPEO chains present at the nanoparticle surface, are at a concentration corresponding to the MPEO cloud point temperature, it is thus possible to use the results shown in Fig. 5 to evaluate the concentration of MPEO at the nanoparticle surface. The values of these concentrations for the three kinds of particles are given in Table 3 (method A). On the other hand, we directly calculated these concentrations by quantifying particle MPEO and using the data of Table 2 (in particular, the values of MPEO coverage and of electrokinetic MPEO layer thickness). The results are also shown in Table 3 (method B). It can be noted that the values calculated by the two methods are comparable as already found for some polystyrene nanoparticles coated with dextran–MPEO copolymers [6]. The fact that the MPEO concentrations calculated by method A are Table 3 MPEO concentration (cMPEO ) at the nanoparticle surface, calculated by two methods Copolymer used as surfactant
MPEO20K –PLA1.7K MPEO10K –PLA1.1K MPEO5K –PLA1.5K Fig. 5. Variation of the cloud temperatures of MPEOs with various molecular weights as a function of their concentration in 0.3 M Na2 SO4 . MPEO20K (); MPEO10K (); MPEO5K ().
IFT (◦ C)
53 52 46
cMPEO (g/L) A
B
110 170 210
90 100 170
(A) cMPEO was calculated from the values of the initial flocculation temperature (IFT) of the nanoparticle suspension. (B) The nanospheres were degraded by 1 M NaOH and MPEO was quantified by a I2 /KI assay. cMPEO was then calculated from data of Table 2 (Γ MPEO and d ).
92
D. Chognot et al. / Colloids and Surfaces B: Biointerfaces 51 (2006) 86–92
higher than those obtained by method B could mean that the MPEO chains are really at the particle surface and have not be trapped inside the PLA matrix. 4. Conclusion We have shown that it is possible to prepare surface-modified PLA nanoparticles by using a double emulsion/solvent evaporation procedure in which the surfactant of the secondary emulsion is a water-soluble diblock MPEO-b–PLA copolymer. After evaporation of the organic solvent, the obtained nanospheres present at their surface MPEO chains probably in the form of a brush. These MPEO chains are strongly anchored inside the PLA matrix thanks to the PLA blocks of the copolymers. They confer to the aqueous particle suspensions, a good stability in the presence of high NaCl concentrations at 25 ◦ C. However, as MPEO is a polymer with a low critical solution temperature, the replacement of NaCl by Na2 SO4 , a salt lowering the cloud temperature of MPEO aqueous solutions, leads to the suspension flocculation at a temperature between 45 and 55 ◦ C depending on the copolymer composition. Concerning the influence of the copolymer MPEO length on the particle properties, we showed that the hydrophilic layer electrokinetic thickness, the surface MPEO concentration and the average distance between two MPEO chains at the surface, depend on the MPEO length. As it has been demonstrated that all these parameters influence the plasma protein adsorption [39], they will have to be considered for the design of long circulating nanoparticles based on this new way of preparation. The main advantage of the procedure here described, is that it can allow the encapsulation of water-soluble bioactive macromolecules such as for example proteins or polysaccharides, in PLA nanoparticles covered with any kind of hydrophilic polymer, provided that an appropriate amphiphilic copolymer is used as surfactant in the secondary emulsion. There is thus no need for other surfactants whose removal from the particle surface is often a problem. References [1] R. Gref, Y. Minamitake, M.T. Peracchia, V. Trubetskoy, V. Torchilin, R. Langer, Science 263 (1994) 1600. [2] R. Gref, G. Miralles, E. Dellacherie, Polym. Int. 48 (1999) 251. [3] M.-F. Zambaux, F. Bonneaux, R. Gref, E. Dellacherie, C. Vigneron, J. Biomed. Mater. Res. 44 (1999) 109. [4] S. Stolnik, C.R. Heald, J. Neal, M.C. Garnett, S.S. Davis, L. Illum, S.C. Purkiss, R.J. Barlow, P.R. Gellert, J. Drug Target 9 (2001) 361. [5] C. Rouzes, M. L´eonard, A. Durand, E. Dellacherie, Coll. Surf. B: Biointerf. 32 (2003) 125. [6] A. De Souza Delgado, M. L´eonard, E. Dellacherie, Langmuir 17 (2001) 4386.
[7] C. Lemarchand, R. Gref, C. Passirani, E. Garcion, B. Petri, R. M¨uller, D. Costantini, P. Couvreur, Biomaterials 27 (2006) 108. [8] C. Passirani, G. Baratt, J.-P. Devissaguet, D. Labarre, Pharm. Res. 15 (1998) 1046. [9] R.H. M¨uller, K.H. Wallis, Int. J. Pharm. 89 (1993) 25. [10] K. Akiyoshi, S. Kobayashi, S. Schichibe, D. Mix, M. Baudys, S.W. Kim, J. Sunamoto, J. Control. Release 54 (1998) 313. [11] P. Quellec, R. Gref, L. Perrin, E. Dellacherie, F. Sommer, J.-M. Verbavatz, M.J. Alonso, J. Biomed. Mater. Res. 42 (1998) 45. [12] M.-F. Zambaux, F. Bonneaux, R. Gref, E. Dellacherie, C. Vigneron, Int. J. Pharm. 212 (2001) 1. [13] C. Lemarchand, P. Couvreur, M. Besnard, D. Costantini, R. Gref, Pharm. Res. 20 (2003) 1284. [14] C. Nouvel, C. Frochot, V. Sadtler, P. Dubois, E. Dellacherie, J.-L. Six, Macromolecules 37 (2004) 4981. [15] I. Ydens, P. Degee, C. Nouvel, E. Dellacherie, J.-L. Six, P. Dubois, E-polymers 046 (2005) http://www.e-polymers.org. ISSN 1618 7229. [16] E. Rotureau, M. L´eonard, E. Dellacherie, A. Durand, Phys. Chem. Chem. Phys. 6 (2004) 1430. [17] A. Durand, E. Marie, E. Rotureau, M. L´eonard, E. Dellacherie, Langmuir 20 (2004) 6956. [18] D. Chognot, J.-L. Six, M. L´eonard, F. Bonneaux, C. Vigneron, E. Dellacherie, J. Coll. Interf. Sci. 268 (2003) 441. [19] A. Sanchez, J.L. Vila-Jato, M.J. Alonso, Int. J. Pharm. 99 (1993) 263. [20] T. G¨orner, R. Gref, D. Michenot, F. Sommer, T.M. Duc, E. Dellacherie, J. Control. Release 57 (1999) 259. [21] D. Blanco, M.J. Alonso, J. Eur. Pharm. Biopharm. 43 (1997) 287. [22] M.-F. Zambaux, F. Bonneaux, R. Gref, E. Dellacherie, C. Vigneron, J. Control. Release 60 (1999) 179. [23] P. Bouillot, A. Petit, E. Dellacherie, J. Appl. Polym. Sci. 68 (1998) 1695. [24] S.R. Deshiikan, K.D. Papadopoulos, Coll. Polym. Sci. 276 (1998) 117. [25] W.G. Eversole, W.W.J. Boardman, Chem. Phys. 9 (1941) 798. [26] B. Baleux, G.C.R. Champetier, Acad. Sci. Paris, S´er. C 274 (1972) 1617. [27] P. Molyneux, Water-Soluble Synthetic Polymers: Properties and Behaviour, CRC Press, Boca-Raton, FL, 1983. [28] J.A. Long, D.W.J. Osmond, B.J. Vincent, Coll. Interf. Sci. 42 (1972) 545. [29] W. Heller, H.J. McCarthy, Chem. Phys. 29 (1958) 78. [30] I. Piirma, Polymeric Surfactants, Surfactant Sciences Series, vol. 42, Marcel Dekker, Inc, New York, 1992, Chapter 5. [31] G. Kwon, M. Naito, M. Yokoyama, Y. Sakurai, R. Kataoka, J. Control. Release 48 (1997) 195. [32] C. Allen, Y.S. Yu, D. Maysinger, A. Eisenberg, Bioconjug. Chem. 9 (1998) 564. [33] S.A. Hagan, A.G.A. Coombes, M.C. Garnett, S.E. Dunn, M.C. Davies, L. Illum, S.S. Davis, S.E. Harding, S. Purkiss, P.R. Gellert, Langmuir 12 (1996) 2153. [34] A. Benahmed, M. Ranger, J.C. Leroux, Pharm. Res. 18 (2001) 323. [35] Ch. Schugens, N. Laruelle, N. Nihant, Ch. Grandfils, R. J´erome, P. Teyssi´e, J. Control. Release 32 (1994) 161. [36] H. Sah, J. Control. Release 58 (1999) 143. [37] J.L. Remmeb, C. Stushnoff, J.F. Carpenter, Pharm. Res. 14 (1997) 1548. [38] K. Izutsu, S. Yoshioka, S. Kojima, T.W. Randolph, J.F. Carpenter, Pharm. Res. 13 (1996) 1393. [39] R. Gref, M. L¨uck, P. Quellec, M. Marchand, E. Dellacherie, S. Harnisch, T. Blunk, R.H. M¨uller, Coll. Surf. B: Biointerf. 18 (2000) 301.