- Email: [email protected]
Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres
European Journal of Pharmaceutics and Biopharmaceutics xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
European Journal of Pharmaceutics and Biopharmaceutics journal homepage: www.elsevier.com/locate/ejpb
Note
Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres Stefanie Wöhl-Bruhn a, Andreas Bertz b, Judith Kuntsche c,1, Henning Menzel b, Heike Bunjes a,⇑ a
Institute of Pharmaceutical Technology, Technische Universität Braunschweig, Braunschweig, Germany Institute of Technical Chemistry, Technische Universität Braunschweig, Braunschweig, Germany c Department of Pharmaceutical Technology and Biopharmaceutics, Martin Luther University Halle-Wittenberg, Halle, Saale, Germany b
a r t i c l e
i n f o
Article history: Received 8 November 2012 Accepted in revised form 28 February 2013 Available online xxxx Keywords: Hydrogel microspheres Polyethylene glycol Hydroxyethyl starch (HES) HES-HEMA HES-MA HES-P(EG)6MA Aqueous two-phase system (ATPS) Molecular weight Particle size distribution Encapsulation efficiency
a b s t r a c t Hydrogel microspheres, e.g. for the use as protein carriers, can be prepared without the use of organic solvents via an emulsified aqueous two-phase system (ATPS) that is based on two immiscible polymer solutions. The type and concentration of the polymers can affect the ATPS and finally the distribution of incorporated drugs between the aqueous phases. For the preparation of hydrogel microspheres based on hydroxyethyl starch–hydroxyethyl methacrylate (HES–HEMA), hydroxyethyl starch–methacrylate (HES–MA), and hydroxyethyl starch–polyethylene glycol methacrylate (HES–P(EG)6MA), polyethylene glycol 12,000 (PEG 12,000) was used as second polymer. The particle size distribution and encapsulation efficiency of the microspheres depended dramatically on the type of PEG 12,000 that was used in the second phase of the ATPS. Analysis of different PEG 12,000 brands by various methods revealed differences in the salt composition and molecular weight distribution of the polymers which can explain the effects on the production process. The results illustrate that the range of product specifications may not always be tight enough to avoid variability in pharmaceutical processes like the preparation of hydrogel microspheres by an aqueous two-phase preparation process. Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction Hydrogel microspheres are promising drug carrier systems, for instance, for the encapsulation of therapeutic proteins [1]. They can be prepared without the use of organic solvents via an emulsified aqueous two-phase system (ATPS) based on two immiscible polymer solutions. The formation of an ATPS is a thermodynamic phenomenon which depends on the type and concentration of the polymers. The chemical nature of the polymers affects the partition coefficient of incorporated drug substances [2]. We previously published studies concerning the production of hydrogel microspheres via such a system using hydroxyethyl starch–hydroxyethyl methacrylate (HES–HEMA), hydroxyethyl starch–methacrylate (HES–MA), and hydroxyethyl starch polyethylene glycol methacrylate (HES– P(EG)6MA) as microsphere forming polymers [3,4]. Polyethylene glycol 12,000 (PEG 12,000) was used as second polymer within the water-in-water emulsion process. In this study, we present data regarding the influence of different PEG 12,000 brands on the parti⇑ Corresponding author. Technische Universität Braunschweig, Institute of Pharmaceutical Technology, Mendelssohnstr. 1, 38106 Braunschweig, Germany. Tel.: +49 531 391 5657; fax: +49 531 391 8108. E-mail address: [email protected] (H. Bunjes). 1 Current address: University of Southern Denmark, Department of Physics, Chemistry and Pharmacy, Campusvej 55, 5230 Odense, Denmark.
cle size distribution and encapsulation efficiency of microspheres obtained by this procedure. In order to elucidate differences observed when using different PEG 12,000 brands, analysis of their Na+, K+, and Ca2+ content by atomic absorption spectrometry (AAS) and their molecular weight distribution by size exclusion chromatography (SEC) and by asymmetrical flow field-flow fractionation measurements (AF4) was performed. 2. Materials and methods 2.1. HES-based polymers All polymers were based on hydroxyethyl starch-(HES 130/0.4; Fresenius Kabi Deutschland GmbH, Bad Homburg, Germany). HES– P(EG)6MA (hydroxyethyl starch polyethylene glycol methacrylate) was synthesized and characterized as described previously [4]. HES to polyethylene glycol methacrylate–imidazole carbamate (P(EG)6MACI) to 4-dimethylaminopyridine (DMAP) was used in a molar ratio of 5:1:1. After a 24 h reaction time, a polymer with a degree of substitution (DS) of 0.07 was obtained. Hydroxyethyl starchmethacrylate (HES–MA, DS 0.12) was synthesized within 20 h using HES: GMA: DMAP in a molar ratio of 4:2:1 [4]. HES–HEMA (DS 0.04) was synthesized as described by Harling et al. [5] using a molar ratio of HES/HEMACI/DMAP of 4:1:1 and a reaction time of 24.5 h.
0939-6411/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ejpb.2013.02.018
Please cite this article in press as: S. Wöhl-Bruhn et al., Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres, Eur. J. Pharm. Biopharm. (2013), http://dx.doi.org/10.1016/j.ejpb.2013.02.018
2
S. Wöhl-Bruhn et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2013) xxx–xxx
2.2. Further materials
2.4. Characterization of polyethylene glycol raw materials
Fluorescein isothiocyanate-labeled dextran with a molecular weight of 70 kDa (FD70) was purchased from Sigma Aldrich (Germany). Polyethylene glycol 12,000 (PEG 12,000) was purchased from Fluka (Steinheim, Germany; PEG 12,000, Mw = 12,860, named as ‘‘brand A’’) and from Sigma–Aldrich (Steinheim, Germany PEG 12,000, Mw = 16,955, named as ‘‘brand C’’). PluriolÒ E 12,000 (named as ‘‘brand B’’) and PluriolÒ E 8000 were a gift from BASF (Ludwigshafen, Germany; Mw unknown). IrgacureÒ2959 (I2959) was purchased from Ciba Specialty Chemicals (Basel, Switzerland).
Atomic absorption spectrometry (AAS) was used for the quantitative element analysis of Na+, K+, and Ca2+ in differently concentrated PEG samples (1, 0.5, 0.25 mg/l) by a PerkinElmer 305 spectrometer according to the PerkinElmer manual (Waltham, Massachusetts, USA). A gas mixture of air/acetylene at 2300 °C, a specific Merck standard (Na: 170238, K: 19505, Ca: 119778), and the appropriate hollow cathode lamp (Na–K 12 mA, Ca–Mg 25 mA) were used. Size exclusion chromatography (SEC) measurements were carried out with a system consisting of a Merck HITACHI L6000A pump (Tokyo, Japan), PSS Suprema columns (10 lm pre-column, 100 Å 10 lm, 2 3000 Å 10 lm, Mainz, Germany), a Wyatt DAWN DSP light scattering detector (Santa Barbara, USA), and a high-sensitive RI (RI-101, Shodex) detector (Tokyo, Japan). Millipore water with 0.05 wt% sodium azide was applied as eluent and as solution medium for sample preparation. For the measurements 100 ll of the samples with a concentration of 0.5 mg/ml were analyzed in triplicate at a flow rate of 1 ml/min and 40 °C. Astra 3.6 was used for data processing. By asymmetrical flow field-flow fractionation, molar masses were determined with an Eclipse F separation system (Wyatt) connected with a multi-angle laser light scattering (MALLS, DAWN EOS, K5 flow cell, Wyatt) and a high-sensitive RI (RI101, Shodex) detector (Tokyo, Japan). The separation channel was equipped with a trapezoidal spacer (for channel dimensions see [6]) and a polyethersulfone membrane with a molecular weight cut-off of 5000 g/mol (Nadir) served as accumulation wall. 50 mM sodium chloride solution, filtered (0.1 lm pore size, VVLP, Millipore) and preserved with sodium azide 0.02% (w/v), was used as solvent and eluent. 100 ll polymer solution (2.5 mg/ml) was injected into the channel, focused with a focus flow of 2 ml/min, and eluted with a channel flow of 1 ml/min and a constant crossflow of 3 ml/min. Samples were measured in triplicate, and molar mass calculations were done with the Astra software version 4.90.08 (Wyatt) using the Debye fitting method. The dn/dc of the polymers (0.139) was determined by direct injection of polymer solutions with six different concentrations into the RI-detector.
2.3. Microsphere preparation and characterization Hydrogel microspheres from all polymers (HES–HEMA, HES– P(EG)6MA, or HES–MA) were prepared according to a general procedure described earlier [3,4]. In the standard procedure (Fig. 1) used in this study, an aqueous two-phase system was used containing 9.0 g of a 2 wt% HES-based polymer solution and 0.9 g of a 0.1 wt% IrgacureÒ2959 solution (as photoinitiator) in one aqueous phase. 0.45 mg of the model drug FITC-dextran 70 kDa was added to this phase as aqueous solution (45 ll). The second aqueous phase consisted of 6.0 g of a 30 wt% PEG 12,000 solution. This composition corresponds to a mass ratio of gel forming polymer to PEG 12,000 of 1:10. All substances were dissolved in 20 mM sodium phosphate buffer pH 7.0 and added into a 20 ml glass vial in the mentioned order. After cooling to 0 °C (10 min), the system was vigorously mixed for 1 min with a vortex-Genie mixer (Bender & Hobein AG, Zurich, Switzerland). The resulting water-in-water emulsion was exposed to UV light (Nu-8 KL, Benda, Wiesloh, Germany) for 30 min at 366 nm for crosslinking. The microsphere suspension was centrifuged and washed five times with demineralized water. Finally, the microspheres were lyophilized at 65 °C and 0.068 mbar for storage and further experiments (alpha 1-4 LO plus, Christ, Osterode, Germany). To investigate the influence of the second polymer, either 30 wt% PEG 12,000 brand A, B or C, or a mixture of 30 wt% brand B and PluriolÒ E 8000 (50 wt%: 50 wt%; named as ‘‘brand B mix’’) were used as second aqueous phase (all in 20 mM sodium phosphate buffer, pH 7.0). The microspheres were characterized for the particle size distribution by laser diffractometry and for their encapsulation efficiencies [4]. The volume size distributions were calculated, and the results are given as box plots with d10, d50, and d90 values (e.g. d90 means that 90% of the particles are smaller than the given size) as an average of three samples prepared from three different microsphere batches. Results concerning mean particle sizes (d50 values) and encapsulation efficiencies were analyzed statistically [4]. Significance (p < 0.05, ) and non-significance (°) are marked in the figures.
3. Results and discussion The influence of the PEG 12,000 brand A or B on the particle size distribution of resulting microspheres was investigated for the starch-based polymers (HES-MA, HES-P(EG)6MA, and HES-P(EG)10MA). In Fig. 2A, the d10, d50, and d90 values of the corresponding particle size distributions are presented in a box-plot diagram. For the three polymers, a trend to slightly higher d50 values and wider particle size distributions was observed for particles produced with PEG brand B as second polymer. Regarding the amount of FD70
Vortexing
+
UV light (366 nm, 30 min) Polymerization Washing 2µm
Polymer FD70 Irgacure® 2959 Buffer
PEG 12,000 Buffer
ATPS
W/W-Emulsion
Microspheres
Fig. 1. Production process of microspheres incorporating FITC-dextran 70 kDa (FD70) via an emulsified aqueous two-phase system (ATPS).
Please cite this article in press as: S. Wöhl-Bruhn et al., Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres, Eur. J. Pharm. Biopharm. (2013), http://dx.doi.org/10.1016/j.ejpb.2013.02.018
3
S. Wöhl-Bruhn et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2013) xxx–xxx
encapsulated in hydrogel microspheres significantly higher encapsulation efficiencies were obtained for particles prepared with brand A (70–80%) compared to those made with PEG brand B (30–40%) (Fig. 2B). As such a tremendous influence of the type of PEG 12,000 is highly relevant for the microsphere production process, the potential causes of this phenomenon were investigated in more detail. Parameters affecting the distribution behavior of proteins in aqueous two-phase systems have been studied widely. The properties of the polymer solutions and of the size and structure of incorporated substances are known to have strong influence [7]. They include, e.g. the size and charge of the incorporated model substance, temperature, pH value, salt concentration as well as the type, molecular weight and concentration of the polymers forming the phase separating aqueous solutions [7]. Since in this study just the second polymer was replaced and all other factors were kept constant, it was hypothesized that either a highly different salt concentration or differences in the molecular weight distribution of the PEG may cause the formation of larger particles and the decrease in encapsulation efficiencies. Regarding the influence of salt concentration on the aqueous two-phase system, ions contained in the two raw materials A
d90 d50
°
brand A
brand B
° brand A
° brand B
60
°
d10
brand A
Particle size (µm)
80
brand B mix
100
brand B
A
40 20 0
HES-HEMA
*
*
HES-P(EG) 6 MA
brand B mix
° brand B
brand A
* brand B
60
brand A
80
brand B
100
brand A
Encapsulation efficiency (%)
B
HES-MA
40 20
and B were analyzed by AAS. The data revealed a significant difference in ion content of the two brands: while brand A contained a high concentration of Na+ and no K+, brand B contained a low concentration of Na+ but a very high K+ concentration. Almost no difference in Ca2+ concentration was detected (Table 1). Due to the fact that the replacement of 20 mM sodium phosphate buffer against phosphate buffered saline (containing 8 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4, 0.2 g KH2PO4 and MilliQ water up to 1000 ml) as solvent for all substances did not lead to any negative effects on the particle size distribution and encapsulation efficiency, the ion concentration was considered not to be the main factor for the influences detected. Subsequently, a potential influence of the molecular weight distribution of these two PEG 12,000 brands was studied by size exclusion chromatography (SEC) and asymmetrical flow field-flow fractionation (AF4) measurements: Investigations by SEC with refractive index and light scattering detection revealed slight differences of the elution curves of the tested substances (Fig. 3A). In comparison with the signal of brand B, the elution curve of brand A was obtained at earlier time points. As higher molecular weight substances are eluted prior to those with lower molecular weight in SEC, brand A was assumed to contain a polymer with slightly higher molecular weight. To confirm this assumption, AF4 measurements were performed. In good agreement with the SEC data, the signal of brand A indicated slightly higher molecular weights (in AF4 molecules with higher molar mass elute later. Fig. 3B). Molar mass calculations from the AF4 data revealed an average molecular weight (n = 3) of brand A of 13,960 ± 400 g/ mol (Mw/Mn = 1.04 ± 0.03) and of brand B of 11,830 ± 100 g/mol (Mw/Mn = 1.06 ± 0.01). To gain further information about the variability of different PEG 12,000 polymers, ion concentration and molecular weight distribution studies were also performed for a third brand (brand C). The salt composition of brand C polymer showed differences to brand A as well as B, but this was assumed to have a negligible influence on the microsphere characteristics as discussed above (Table 1). Also SEC and AF4 measurements were performed with brand C polymer. While the SEC data showed a good comparability of the elution curves of brand A and C, the elution curve of brand C from AF4 measurements was situated in between curve A and B but slightly shifted to the curve of polymer A (Fig. 3). Considering the molecular weight distribution of brand C (13,720 ± 310 g/mol, Mw/Mn 1.02 ± 0.00; calculated from RI and MALLS detection data, not based on the retention times), a higher similarity to brand A polymer was obvious. To study this difference on the microsphere characteristics, encapsulation efficiency studies were performed for microspheres prepared with brand C PEG. The encapsulation of FD70 in HES–P(EG)6MA hydrogel microspheres resulted in similar encapsulation efficiencies as it was shown for brand A PEG (data obtained with HES–P(EG)6MA polymer of different degree of substitution (DS 0.04) and thus not shown here). To test the assumption that differences in molecular weight distribution can have major effects on the microsphere characteristics, HES–P(EG)6MA hydrogel microspheres were produced using PEG with lower average molecular weight. For this purpose, brand
0
HES-MA
HES-HEMA
HES-P(EG) 6MA
Fig. 2. Particle size distributions (d10, d50, d90 (A); statistical significance was tested between the d50 values) and encapsulation efficiencies of FD70-containing hydrogel microspheres (B) made from various polymers (HES–MA, HES–HEMA, HES–P(EG)6MA). Microspheres were prepared via an aqueous two-phase system containing polyethylene glycol 12,000 of two different brands (brand A and brand B) or a 1:1 mixture of polyethylene glycol 12,000 and 8000 (brand B mix) as second aqueous phase. Statistically significant differences are marked with (p < 0.05), those without statistical significance with °.
Table 1 Concentration of Na+, K+, and Ca2+ ions in PEG 12,000 brands as determined by AAS measurements (n P 3). Polyethylene glycol 12,000
Na+ (mmol/kg)
K+ (mmol/kg)
Ca2+ (mmol/kg)
Brand A Brand B Brand C
6.08 1.55 0
0 10.16 6.24
0.14 0.21 0.16
Please cite this article in press as: S. Wöhl-Bruhn et al., Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres, Eur. J. Pharm. Biopharm. (2013), http://dx.doi.org/10.1016/j.ejpb.2013.02.018
4
S. Wöhl-Bruhn et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2013) xxx–xxx
A
detector signal
brand A brand A brand B brand B brand C brand C
17
18
19
20
21
22
23
time (min)
B
detector signal
brand A brand A brand B brand B brand C brand C
was accompanied by a decrease in viscosity of the continuous phase. The detection of larger particles was attributed to the influence of lower viscosity on the shear forces during vortexing of the aqueous two-phase systems [8]. The data obtained with the PEG mixture supported the assumption that the tested PEG 12,000 brands differ in their average molecular weight resulting in essential influences on the waterin-water emulsion process. The molecular weight of commercially available substances can differ in spite of apparently similar product specifications. Due to the wide range in the specification of the molecular weight (e.g. for PEG 12,000 a range of the molecular mass between 11,000 and 15,000 g/mol was stated from the supplier of brand C) all substances met the product specification of the supplier. However, these results may point out that the range of product specification is sometimes not tight enough to avoid variability in pharmaceutical processes. In case of insufficient information on such parameters a simple exchange of substances (new supplier or even new batch) can lead to extensive negative effects on the production process and – as shown here – on the characteristics of HES-based microspheres. Acknowledgements The authors thank the German Research Foundation (DFG) for funding their project within the collaborative research centre 578 (SFB 578, Subproject D1). We also like to thank Birgit Niehoegen for the AAS measurements (Institute of Technical Chemistry, TU Braunschweig). Fresenius Kabi Deutschland and Ciba Specialty Chemicals and BASF are acknowledged for material support. References
6
8
10
12
14
16
time (min) Fig. 3. Representative elution curves of three measurements from SEC (A) and AF4 (B) measurements of different PEG brands (brands A–C) (N/D/ : RI-detector, d/s/ I: light scattering detector).
B was mixed with a lower weight PEG from the same brand and used as second phase polymer (named as ‘‘brand B mix’’). Purposely decreasing the average molecular weight of the second phase polymer led to a decrease in encapsulation efficiency (Fig. 2B). Furthermore, an increase in the mean particle size and broadening of the particle size distribution of the microspheres (Fig. 2A) was observed. Influences of the PEG molecular weight on the size of hydrogel microparticles prepared via aqueous twophase systems based on dextran and PEG were previously discussed by Stenekes et al. Decreasing the molecular weight of PEG
[1] T.R. Hoare, D.S. Kohane, Hydrogels in drug delivery: progress and challenges, Polymer 49 (2008) 1993–2007. [2] P.Å. Albertsson, Partition of Cell Particles and Macromolecules, Almquist & Wiksells, Uppsala, 1960. [3] A.D.A. Schwoerer, S. Harling, K. Scheibe, H. Menzel, R. Daniels, Influence of degree of substitution of HES–HEMA on the release of incorporated drug models from corresponding hydrogels, European Journal of Pharmaceutics and Biopharmaceutics 73 (2009) 351–356. [4] S. Wöhl-Bruhn, A. Bertz, S. Harling, H. Menzel, H. Bunjes, Hydroxyethylstarchbased polymers for the controlled release of biomacromolecules from hydrogel microspheres, European Journal of Pharmaceutics and Biopharmaceutics 81 (2012) 573–581. [5] S. Harling, A. Schwoerer, K. Scheibe, R. Daniels, H. Menzel, A new hydrogel drug delivery system based on hydroxyethylstarch derivatives, Journal of Microencapsulation 27 (2010) 400–408. [6] J. Kuntsche, K. Klaus, F. Steiniger, Size determinations of colloidal fat emulsions: a comparative study, Journal of Biomedical Nanotechnology 5 (2009) 384–395. [7] J.A. Asenjo, A.S. Schmidt, F. Hachem, B.A. Andrews, Model for predicting the partition behaviour of proteins in aqueous two-phase systems, Journal of Chromatography 668 (1994) 47–54. [8] R.J.H. Stenekes, O. Franssen, E.M.G. van Bommel, D.J.A. Crommelin, W.E. Hennink, The preparation of dextran microspheres in an all-aqueous system: effect of the formulation parameters on particle characteristics, Pharmaceutical Research 15 (1998) 557–561.
Please cite this article in press as: S. Wöhl-Bruhn et al., Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres, Eur. J. Pharm. Biopharm. (2013), http://dx.doi.org/10.1016/j.ejpb.2013.02.018
Contents lists available at SciVerse ScienceDirect
European Journal of Pharmaceutics and Biopharmaceutics journal homepage: www.elsevier.com/locate/ejpb
Note
Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres Stefanie Wöhl-Bruhn a, Andreas Bertz b, Judith Kuntsche c,1, Henning Menzel b, Heike Bunjes a,⇑ a
Institute of Pharmaceutical Technology, Technische Universität Braunschweig, Braunschweig, Germany Institute of Technical Chemistry, Technische Universität Braunschweig, Braunschweig, Germany c Department of Pharmaceutical Technology and Biopharmaceutics, Martin Luther University Halle-Wittenberg, Halle, Saale, Germany b
a r t i c l e
i n f o
Article history: Received 8 November 2012 Accepted in revised form 28 February 2013 Available online xxxx Keywords: Hydrogel microspheres Polyethylene glycol Hydroxyethyl starch (HES) HES-HEMA HES-MA HES-P(EG)6MA Aqueous two-phase system (ATPS) Molecular weight Particle size distribution Encapsulation efficiency
a b s t r a c t Hydrogel microspheres, e.g. for the use as protein carriers, can be prepared without the use of organic solvents via an emulsified aqueous two-phase system (ATPS) that is based on two immiscible polymer solutions. The type and concentration of the polymers can affect the ATPS and finally the distribution of incorporated drugs between the aqueous phases. For the preparation of hydrogel microspheres based on hydroxyethyl starch–hydroxyethyl methacrylate (HES–HEMA), hydroxyethyl starch–methacrylate (HES–MA), and hydroxyethyl starch–polyethylene glycol methacrylate (HES–P(EG)6MA), polyethylene glycol 12,000 (PEG 12,000) was used as second polymer. The particle size distribution and encapsulation efficiency of the microspheres depended dramatically on the type of PEG 12,000 that was used in the second phase of the ATPS. Analysis of different PEG 12,000 brands by various methods revealed differences in the salt composition and molecular weight distribution of the polymers which can explain the effects on the production process. The results illustrate that the range of product specifications may not always be tight enough to avoid variability in pharmaceutical processes like the preparation of hydrogel microspheres by an aqueous two-phase preparation process. Ó 2013 Elsevier B.V. All rights reserved.
1. Introduction Hydrogel microspheres are promising drug carrier systems, for instance, for the encapsulation of therapeutic proteins [1]. They can be prepared without the use of organic solvents via an emulsified aqueous two-phase system (ATPS) based on two immiscible polymer solutions. The formation of an ATPS is a thermodynamic phenomenon which depends on the type and concentration of the polymers. The chemical nature of the polymers affects the partition coefficient of incorporated drug substances [2]. We previously published studies concerning the production of hydrogel microspheres via such a system using hydroxyethyl starch–hydroxyethyl methacrylate (HES–HEMA), hydroxyethyl starch–methacrylate (HES–MA), and hydroxyethyl starch polyethylene glycol methacrylate (HES– P(EG)6MA) as microsphere forming polymers [3,4]. Polyethylene glycol 12,000 (PEG 12,000) was used as second polymer within the water-in-water emulsion process. In this study, we present data regarding the influence of different PEG 12,000 brands on the parti⇑ Corresponding author. Technische Universität Braunschweig, Institute of Pharmaceutical Technology, Mendelssohnstr. 1, 38106 Braunschweig, Germany. Tel.: +49 531 391 5657; fax: +49 531 391 8108. E-mail address: [email protected] (H. Bunjes). 1 Current address: University of Southern Denmark, Department of Physics, Chemistry and Pharmacy, Campusvej 55, 5230 Odense, Denmark.
cle size distribution and encapsulation efficiency of microspheres obtained by this procedure. In order to elucidate differences observed when using different PEG 12,000 brands, analysis of their Na+, K+, and Ca2+ content by atomic absorption spectrometry (AAS) and their molecular weight distribution by size exclusion chromatography (SEC) and by asymmetrical flow field-flow fractionation measurements (AF4) was performed. 2. Materials and methods 2.1. HES-based polymers All polymers were based on hydroxyethyl starch-(HES 130/0.4; Fresenius Kabi Deutschland GmbH, Bad Homburg, Germany). HES– P(EG)6MA (hydroxyethyl starch polyethylene glycol methacrylate) was synthesized and characterized as described previously [4]. HES to polyethylene glycol methacrylate–imidazole carbamate (P(EG)6MACI) to 4-dimethylaminopyridine (DMAP) was used in a molar ratio of 5:1:1. After a 24 h reaction time, a polymer with a degree of substitution (DS) of 0.07 was obtained. Hydroxyethyl starchmethacrylate (HES–MA, DS 0.12) was synthesized within 20 h using HES: GMA: DMAP in a molar ratio of 4:2:1 [4]. HES–HEMA (DS 0.04) was synthesized as described by Harling et al. [5] using a molar ratio of HES/HEMACI/DMAP of 4:1:1 and a reaction time of 24.5 h.
0939-6411/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ejpb.2013.02.018
Please cite this article in press as: S. Wöhl-Bruhn et al., Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres, Eur. J. Pharm. Biopharm. (2013), http://dx.doi.org/10.1016/j.ejpb.2013.02.018
2
S. Wöhl-Bruhn et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2013) xxx–xxx
2.2. Further materials
2.4. Characterization of polyethylene glycol raw materials
Fluorescein isothiocyanate-labeled dextran with a molecular weight of 70 kDa (FD70) was purchased from Sigma Aldrich (Germany). Polyethylene glycol 12,000 (PEG 12,000) was purchased from Fluka (Steinheim, Germany; PEG 12,000, Mw = 12,860, named as ‘‘brand A’’) and from Sigma–Aldrich (Steinheim, Germany PEG 12,000, Mw = 16,955, named as ‘‘brand C’’). PluriolÒ E 12,000 (named as ‘‘brand B’’) and PluriolÒ E 8000 were a gift from BASF (Ludwigshafen, Germany; Mw unknown). IrgacureÒ2959 (I2959) was purchased from Ciba Specialty Chemicals (Basel, Switzerland).
Atomic absorption spectrometry (AAS) was used for the quantitative element analysis of Na+, K+, and Ca2+ in differently concentrated PEG samples (1, 0.5, 0.25 mg/l) by a PerkinElmer 305 spectrometer according to the PerkinElmer manual (Waltham, Massachusetts, USA). A gas mixture of air/acetylene at 2300 °C, a specific Merck standard (Na: 170238, K: 19505, Ca: 119778), and the appropriate hollow cathode lamp (Na–K 12 mA, Ca–Mg 25 mA) were used. Size exclusion chromatography (SEC) measurements were carried out with a system consisting of a Merck HITACHI L6000A pump (Tokyo, Japan), PSS Suprema columns (10 lm pre-column, 100 Å 10 lm, 2 3000 Å 10 lm, Mainz, Germany), a Wyatt DAWN DSP light scattering detector (Santa Barbara, USA), and a high-sensitive RI (RI-101, Shodex) detector (Tokyo, Japan). Millipore water with 0.05 wt% sodium azide was applied as eluent and as solution medium for sample preparation. For the measurements 100 ll of the samples with a concentration of 0.5 mg/ml were analyzed in triplicate at a flow rate of 1 ml/min and 40 °C. Astra 3.6 was used for data processing. By asymmetrical flow field-flow fractionation, molar masses were determined with an Eclipse F separation system (Wyatt) connected with a multi-angle laser light scattering (MALLS, DAWN EOS, K5 flow cell, Wyatt) and a high-sensitive RI (RI101, Shodex) detector (Tokyo, Japan). The separation channel was equipped with a trapezoidal spacer (for channel dimensions see [6]) and a polyethersulfone membrane with a molecular weight cut-off of 5000 g/mol (Nadir) served as accumulation wall. 50 mM sodium chloride solution, filtered (0.1 lm pore size, VVLP, Millipore) and preserved with sodium azide 0.02% (w/v), was used as solvent and eluent. 100 ll polymer solution (2.5 mg/ml) was injected into the channel, focused with a focus flow of 2 ml/min, and eluted with a channel flow of 1 ml/min and a constant crossflow of 3 ml/min. Samples were measured in triplicate, and molar mass calculations were done with the Astra software version 4.90.08 (Wyatt) using the Debye fitting method. The dn/dc of the polymers (0.139) was determined by direct injection of polymer solutions with six different concentrations into the RI-detector.
2.3. Microsphere preparation and characterization Hydrogel microspheres from all polymers (HES–HEMA, HES– P(EG)6MA, or HES–MA) were prepared according to a general procedure described earlier [3,4]. In the standard procedure (Fig. 1) used in this study, an aqueous two-phase system was used containing 9.0 g of a 2 wt% HES-based polymer solution and 0.9 g of a 0.1 wt% IrgacureÒ2959 solution (as photoinitiator) in one aqueous phase. 0.45 mg of the model drug FITC-dextran 70 kDa was added to this phase as aqueous solution (45 ll). The second aqueous phase consisted of 6.0 g of a 30 wt% PEG 12,000 solution. This composition corresponds to a mass ratio of gel forming polymer to PEG 12,000 of 1:10. All substances were dissolved in 20 mM sodium phosphate buffer pH 7.0 and added into a 20 ml glass vial in the mentioned order. After cooling to 0 °C (10 min), the system was vigorously mixed for 1 min with a vortex-Genie mixer (Bender & Hobein AG, Zurich, Switzerland). The resulting water-in-water emulsion was exposed to UV light (Nu-8 KL, Benda, Wiesloh, Germany) for 30 min at 366 nm for crosslinking. The microsphere suspension was centrifuged and washed five times with demineralized water. Finally, the microspheres were lyophilized at 65 °C and 0.068 mbar for storage and further experiments (alpha 1-4 LO plus, Christ, Osterode, Germany). To investigate the influence of the second polymer, either 30 wt% PEG 12,000 brand A, B or C, or a mixture of 30 wt% brand B and PluriolÒ E 8000 (50 wt%: 50 wt%; named as ‘‘brand B mix’’) were used as second aqueous phase (all in 20 mM sodium phosphate buffer, pH 7.0). The microspheres were characterized for the particle size distribution by laser diffractometry and for their encapsulation efficiencies [4]. The volume size distributions were calculated, and the results are given as box plots with d10, d50, and d90 values (e.g. d90 means that 90% of the particles are smaller than the given size) as an average of three samples prepared from three different microsphere batches. Results concerning mean particle sizes (d50 values) and encapsulation efficiencies were analyzed statistically [4]. Significance (p < 0.05, ) and non-significance (°) are marked in the figures.
3. Results and discussion The influence of the PEG 12,000 brand A or B on the particle size distribution of resulting microspheres was investigated for the starch-based polymers (HES-MA, HES-P(EG)6MA, and HES-P(EG)10MA). In Fig. 2A, the d10, d50, and d90 values of the corresponding particle size distributions are presented in a box-plot diagram. For the three polymers, a trend to slightly higher d50 values and wider particle size distributions was observed for particles produced with PEG brand B as second polymer. Regarding the amount of FD70
Vortexing
+
UV light (366 nm, 30 min) Polymerization Washing 2µm
Polymer FD70 Irgacure® 2959 Buffer
PEG 12,000 Buffer
ATPS
W/W-Emulsion
Microspheres
Fig. 1. Production process of microspheres incorporating FITC-dextran 70 kDa (FD70) via an emulsified aqueous two-phase system (ATPS).
Please cite this article in press as: S. Wöhl-Bruhn et al., Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres, Eur. J. Pharm. Biopharm. (2013), http://dx.doi.org/10.1016/j.ejpb.2013.02.018
3
S. Wöhl-Bruhn et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2013) xxx–xxx
encapsulated in hydrogel microspheres significantly higher encapsulation efficiencies were obtained for particles prepared with brand A (70–80%) compared to those made with PEG brand B (30–40%) (Fig. 2B). As such a tremendous influence of the type of PEG 12,000 is highly relevant for the microsphere production process, the potential causes of this phenomenon were investigated in more detail. Parameters affecting the distribution behavior of proteins in aqueous two-phase systems have been studied widely. The properties of the polymer solutions and of the size and structure of incorporated substances are known to have strong influence [7]. They include, e.g. the size and charge of the incorporated model substance, temperature, pH value, salt concentration as well as the type, molecular weight and concentration of the polymers forming the phase separating aqueous solutions [7]. Since in this study just the second polymer was replaced and all other factors were kept constant, it was hypothesized that either a highly different salt concentration or differences in the molecular weight distribution of the PEG may cause the formation of larger particles and the decrease in encapsulation efficiencies. Regarding the influence of salt concentration on the aqueous two-phase system, ions contained in the two raw materials A
d90 d50
°
brand A
brand B
° brand A
° brand B
60
°
d10
brand A
Particle size (µm)
80
brand B mix
100
brand B
A
40 20 0
HES-HEMA
*
*
HES-P(EG) 6 MA
brand B mix
° brand B
brand A
* brand B
60
brand A
80
brand B
100
brand A
Encapsulation efficiency (%)
B
HES-MA
40 20
and B were analyzed by AAS. The data revealed a significant difference in ion content of the two brands: while brand A contained a high concentration of Na+ and no K+, brand B contained a low concentration of Na+ but a very high K+ concentration. Almost no difference in Ca2+ concentration was detected (Table 1). Due to the fact that the replacement of 20 mM sodium phosphate buffer against phosphate buffered saline (containing 8 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4, 0.2 g KH2PO4 and MilliQ water up to 1000 ml) as solvent for all substances did not lead to any negative effects on the particle size distribution and encapsulation efficiency, the ion concentration was considered not to be the main factor for the influences detected. Subsequently, a potential influence of the molecular weight distribution of these two PEG 12,000 brands was studied by size exclusion chromatography (SEC) and asymmetrical flow field-flow fractionation (AF4) measurements: Investigations by SEC with refractive index and light scattering detection revealed slight differences of the elution curves of the tested substances (Fig. 3A). In comparison with the signal of brand B, the elution curve of brand A was obtained at earlier time points. As higher molecular weight substances are eluted prior to those with lower molecular weight in SEC, brand A was assumed to contain a polymer with slightly higher molecular weight. To confirm this assumption, AF4 measurements were performed. In good agreement with the SEC data, the signal of brand A indicated slightly higher molecular weights (in AF4 molecules with higher molar mass elute later. Fig. 3B). Molar mass calculations from the AF4 data revealed an average molecular weight (n = 3) of brand A of 13,960 ± 400 g/ mol (Mw/Mn = 1.04 ± 0.03) and of brand B of 11,830 ± 100 g/mol (Mw/Mn = 1.06 ± 0.01). To gain further information about the variability of different PEG 12,000 polymers, ion concentration and molecular weight distribution studies were also performed for a third brand (brand C). The salt composition of brand C polymer showed differences to brand A as well as B, but this was assumed to have a negligible influence on the microsphere characteristics as discussed above (Table 1). Also SEC and AF4 measurements were performed with brand C polymer. While the SEC data showed a good comparability of the elution curves of brand A and C, the elution curve of brand C from AF4 measurements was situated in between curve A and B but slightly shifted to the curve of polymer A (Fig. 3). Considering the molecular weight distribution of brand C (13,720 ± 310 g/mol, Mw/Mn 1.02 ± 0.00; calculated from RI and MALLS detection data, not based on the retention times), a higher similarity to brand A polymer was obvious. To study this difference on the microsphere characteristics, encapsulation efficiency studies were performed for microspheres prepared with brand C PEG. The encapsulation of FD70 in HES–P(EG)6MA hydrogel microspheres resulted in similar encapsulation efficiencies as it was shown for brand A PEG (data obtained with HES–P(EG)6MA polymer of different degree of substitution (DS 0.04) and thus not shown here). To test the assumption that differences in molecular weight distribution can have major effects on the microsphere characteristics, HES–P(EG)6MA hydrogel microspheres were produced using PEG with lower average molecular weight. For this purpose, brand
0
HES-MA
HES-HEMA
HES-P(EG) 6MA
Fig. 2. Particle size distributions (d10, d50, d90 (A); statistical significance was tested between the d50 values) and encapsulation efficiencies of FD70-containing hydrogel microspheres (B) made from various polymers (HES–MA, HES–HEMA, HES–P(EG)6MA). Microspheres were prepared via an aqueous two-phase system containing polyethylene glycol 12,000 of two different brands (brand A and brand B) or a 1:1 mixture of polyethylene glycol 12,000 and 8000 (brand B mix) as second aqueous phase. Statistically significant differences are marked with (p < 0.05), those without statistical significance with °.
Table 1 Concentration of Na+, K+, and Ca2+ ions in PEG 12,000 brands as determined by AAS measurements (n P 3). Polyethylene glycol 12,000
Na+ (mmol/kg)
K+ (mmol/kg)
Ca2+ (mmol/kg)
Brand A Brand B Brand C
6.08 1.55 0
0 10.16 6.24
0.14 0.21 0.16
Please cite this article in press as: S. Wöhl-Bruhn et al., Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres, Eur. J. Pharm. Biopharm. (2013), http://dx.doi.org/10.1016/j.ejpb.2013.02.018
4
S. Wöhl-Bruhn et al. / European Journal of Pharmaceutics and Biopharmaceutics xxx (2013) xxx–xxx
A
detector signal
brand A brand A brand B brand B brand C brand C
17
18
19
20
21
22
23
time (min)
B
detector signal
brand A brand A brand B brand B brand C brand C
was accompanied by a decrease in viscosity of the continuous phase. The detection of larger particles was attributed to the influence of lower viscosity on the shear forces during vortexing of the aqueous two-phase systems [8]. The data obtained with the PEG mixture supported the assumption that the tested PEG 12,000 brands differ in their average molecular weight resulting in essential influences on the waterin-water emulsion process. The molecular weight of commercially available substances can differ in spite of apparently similar product specifications. Due to the wide range in the specification of the molecular weight (e.g. for PEG 12,000 a range of the molecular mass between 11,000 and 15,000 g/mol was stated from the supplier of brand C) all substances met the product specification of the supplier. However, these results may point out that the range of product specification is sometimes not tight enough to avoid variability in pharmaceutical processes. In case of insufficient information on such parameters a simple exchange of substances (new supplier or even new batch) can lead to extensive negative effects on the production process and – as shown here – on the characteristics of HES-based microspheres. Acknowledgements The authors thank the German Research Foundation (DFG) for funding their project within the collaborative research centre 578 (SFB 578, Subproject D1). We also like to thank Birgit Niehoegen for the AAS measurements (Institute of Technical Chemistry, TU Braunschweig). Fresenius Kabi Deutschland and Ciba Specialty Chemicals and BASF are acknowledged for material support. References
6
8
10
12
14
16
time (min) Fig. 3. Representative elution curves of three measurements from SEC (A) and AF4 (B) measurements of different PEG brands (brands A–C) (N/D/ : RI-detector, d/s/ I: light scattering detector).
B was mixed with a lower weight PEG from the same brand and used as second phase polymer (named as ‘‘brand B mix’’). Purposely decreasing the average molecular weight of the second phase polymer led to a decrease in encapsulation efficiency (Fig. 2B). Furthermore, an increase in the mean particle size and broadening of the particle size distribution of the microspheres (Fig. 2A) was observed. Influences of the PEG molecular weight on the size of hydrogel microparticles prepared via aqueous twophase systems based on dextran and PEG were previously discussed by Stenekes et al. Decreasing the molecular weight of PEG
[1] T.R. Hoare, D.S. Kohane, Hydrogels in drug delivery: progress and challenges, Polymer 49 (2008) 1993–2007. [2] P.Å. Albertsson, Partition of Cell Particles and Macromolecules, Almquist & Wiksells, Uppsala, 1960. [3] A.D.A. Schwoerer, S. Harling, K. Scheibe, H. Menzel, R. Daniels, Influence of degree of substitution of HES–HEMA on the release of incorporated drug models from corresponding hydrogels, European Journal of Pharmaceutics and Biopharmaceutics 73 (2009) 351–356. [4] S. Wöhl-Bruhn, A. Bertz, S. Harling, H. Menzel, H. Bunjes, Hydroxyethylstarchbased polymers for the controlled release of biomacromolecules from hydrogel microspheres, European Journal of Pharmaceutics and Biopharmaceutics 81 (2012) 573–581. [5] S. Harling, A. Schwoerer, K. Scheibe, R. Daniels, H. Menzel, A new hydrogel drug delivery system based on hydroxyethylstarch derivatives, Journal of Microencapsulation 27 (2010) 400–408. [6] J. Kuntsche, K. Klaus, F. Steiniger, Size determinations of colloidal fat emulsions: a comparative study, Journal of Biomedical Nanotechnology 5 (2009) 384–395. [7] J.A. Asenjo, A.S. Schmidt, F. Hachem, B.A. Andrews, Model for predicting the partition behaviour of proteins in aqueous two-phase systems, Journal of Chromatography 668 (1994) 47–54. [8] R.J.H. Stenekes, O. Franssen, E.M.G. van Bommel, D.J.A. Crommelin, W.E. Hennink, The preparation of dextran microspheres in an all-aqueous system: effect of the formulation parameters on particle characteristics, Pharmaceutical Research 15 (1998) 557–561.
Please cite this article in press as: S. Wöhl-Bruhn et al., Variations in polyethylene glycol brands and their influence on the preparation process of hydrogel microspheres, Eur. J. Pharm. Biopharm. (2013), http://dx.doi.org/10.1016/j.ejpb.2013.02.018