Journal of Controlled Release 194 (2014) 168–177

Contents lists available at ScienceDirect

Journal of Controlled Release journal homepage: www.elsevier.com/locate/jconrel

A novel soluble supramolecular system for sustained rh-GH delivery Stefano Salmaso, Sara Bersani, Anna Scomparin, Anna Balasso, Chiara Brazzale, Michela Barattin, Paolo Caliceti ⁎ Department of Pharmaceutical and Pharmacological Sciences, University of Padova, Via F. Marzolo 5, 35131 Padova, Italy

a r t i c l e

i n f o

Article history: Received 28 July 2014 Accepted 23 August 2014 Available online 2 September 2014 Keywords: Growth hormone rh-GH Sustained protein delivery Amphiphilic polymers Protein/polymer association Poly(ethylene glycol)

a b s t r a c t Methoxy-poly(ethylene glycol)s bearing a terminal cholanic moiety (mPEG5kDa–cholane, mPEG10kDa–cholane and mPEG20kDa–cholane) were physically combined with recombinant human growth hormone (rh-GH) to obtain supramolecular assemblies for sustained hormone delivery. The association constants (Ka) calculated by Scatchard analysis of size exclusion chromatography (SEC) data were in the order of 105 M−1. The complete rh-GH association with mPEG5kDa–cholane, mPEG10kDa–cholane and mPEG20kDa–cholane was achieved with 7.5 ± 1.1, 3.9 ± 0.4 and 2.6 ± 0.4 w/w% rh-GH/mPEG–cholane, respectively. Isothermal titration calorimetry (ITC) yielded association constants similar to that calculated by SEC and showed that rh-GH has 21–25 binding sites for mPEG–cholane, regardless the polymer molecular weight. Dialysis studies showed that the mPEG– cholane association strongly delays the protein release; 80–90% of the associated rh-GH was released in 200 h. However, during the first 8 h the protein formulations obtained with mPEG10kDa–cholane and mPEG20kDa– cholane showed a burst release of 8 and 28%, respectively. Circular dichroism (CD) analyses showed that the mPEG5kDa–cholane association does not alter the secondary structure of the protein. Furthermore, mPEG5kDa– cholane was found to enhance both the enzymatic and physical stability of rh-GH. In vivo pharmacokinetic and pharmacodynamic studies were performed by subcutaneous administration of rh-GH and rh-GH/mPEG5kDa– cholane to normal and hypophysectomised rats. The study showed that mPEG5kDa–cholane decreases the maximal concentration in the blood but prolongs the body exposure of the protein, which resulted in 55% bioavailability increase. Finally, rh-GH formulated with mPEG5kDa–cholane yielded prolonged weight increase of hypophysectomised rats as compared to rh-GH in buffer or formulated with mPEG5kDa–OH. After the second administration the weight of the animals treated with rh-GH formulated with mPEG5kDa–cholane was about 2 times higher than that obtained with equal dose of non-formulated rh-GH. © 2014 Elsevier B.V. All rights reserved.

1. Introduction In the last decades, a growing number of proteins and peptides has been emerging as novel potential therapeutics. However, despite these biotherapeutics have been revolutionising the pharmaceutical scenario by prospecting novel therapeutic opportunities, the poor biopharmaceutical properties and inconvenient pharmacokinetic profiles have been often representing the bottleneck for their pharmaceutical development, which dramatically thwarts their widespread application [1]. Therefore, innovative delivery systems that enhance their stability, promote the biological barrier diffusion, prolong their permanence in the circulation and favour their localisation in the disease sites have been actively investigated [2,3]. Human growth hormone and its recombinant forms are widespread used for the treatment of childhood growth disorders and adult GH deficiency [4]. Optimal growth hormone treatments require continuous long-term adherence that, at present, is achieved by daily subcutaneous injection because its half-life after subcutaneous administration is ⁎ Corresponding author. Tel.: +39 0498275695; fax: +39 0498275366. E-mail address: [email protected] (P. Caliceti).

http://dx.doi.org/10.1016/j.jconrel.2014.08.024 0168-3659/© 2014 Elsevier B.V. All rights reserved.

limited to 5.3 h [5]. In order to overcome the issues associated with patient compliance, several user-friendly devices as well as controlled and sustained delivery systems have been developed. Sustained rh-GH delivery formulations include fusion protein derivatives, polymer bioconjugates, microspheres and hydrogels [6–12]. However, formulation of rh-GH drug delivery systems suffers from the general issues of protein formulation. Indeed, rh-GH is prone to denaturation and degradation during processing and storage making its formulation rather complicated. Furthermore, issues associated to the high cost of manufacturing of the rh-GH delivery systems or their inefficient therapeutic profiles and side effects have so far strongly limited their development and commercialisation [13]. Soluble polymers offer a versatile solution for the unmet needs for protein formulation. Hydrophilic supramolecular colloidal systems with tailored biopharmaceutical, pharmacokinetic and biological profiles have in fact been obtained by chemical conjugation of proteins with polysaccharides, polyacrylates, polyvinyls and polyoxyalkanes [11,14–17]. PEGylation, in particular, has successfully yielded a novel class of biotherapeutics, a few of which have reached the pharmaceutical market and rapidly become blockbusters. Nonetheless, the chemical attachment of soluble polymers to proteins often results in dramatic

S. Salmaso et al. / Journal of Controlled Release 194 (2014) 168–177

reduction of bioactivity deriving from reduced binding affinity of the protein to the biological target, which reduces the chance of application of this technique to a wide range of proteins. Delivery systems relying on physical protein/polymer association represent an attractive alternative to chemical polymer bioconjugation [18,19]. This approach does not generate new molecular entities, which is beneficial for the approval process by the regulatory agencies, and may prevent biological inactivation due to irreversible alterations of the protein structure. On the other side, soluble polymer association is a useful tool to enhance the protein stability and body exposure, which in turn results in the amelioration of the pharmaceutical performance. So far, a variety of amphiphilic polymers have been explored to produce non covalent protein/polymer supramolecular associations [18]. Aside from polymers designed for specific interactions with protein domains through selective biorecognition, neutral and charged macromolecules may strongly interact with proteins through unspecific coulombic or hydrophobic interactions. Polyoxyethylene, polyvinyl and polyacrylate derivatives as well as hydrophobised polysaccharides and comb-like hydrophobically modified polyaminoacids have been found to efficiently associate with a few proteins, including hormones and cytokines. Polymer association was found to enhance the biopharmaceutical and pharmacokinetic properties of protein drugs, modify their immunological profile and promote the cell up-take [20–23]. Amphiphilic macromolecules obtained by derivatisation of hydrophilic polymers with natural steroidal molecules, namely cholesterol and other polycyclic surfactants secreted in mammalian bile, can strongly associate with proteins. Indeed, due to their peculiar structural features, namely flexible side chains and asymmetric polycyclic faces, natural steroidal molecules can tightly interact with hydrophobic domains on the protein surface [24–27]. Accordingly, dendritic PEGs end functionalised with cholesterol have been used for protein brain targeting [28]. Cholesterol derivatised polyakylamines have been investigated for insulin delivery [29]. Cholesterol derivatised pullulan was shown to strongly associate with high molecular weight proteins, namely bovine serum albumin (65 kDa) and β-galactosidase (540 kDa) and promote their cell up-take [30]. Recent studies have shown that PEGs end-derivatised with a cholanic moiety can be properly exploited for protein delivery [31]. Indeed, the polycyclic structure of β-cholanic acid can dock into hydrophobic pockets of proteins or interact with surface hydrophobic spots thus yielding protein/polymer assemblies. According to the promising results obtained with the protein model rh-G-CSF, we investigated the possibility to expand the “mPEG–cholane association technology” to other proteins of pharmaceutical interest that have not found successful formulation yet. In particular, we investigated the mPEG–cholane association with the recombinant human growth hormone (rh-GH), which has an unmet need of adequate formulation for sustained release. Actually, rh-GH has been found to associate with polymeric surfactants and multifunctional amphiphilic polymers obtained by derivatisation of polymeric backbones with aliphatic chains [23], suggesting that it is a suitable candidate for association with mPEG–cholane derivatives, similarly to what was observed with rh-G-CSF. On the other hand, since rh-GH and rh-G-CSF have different biophysical properties, namely hydrophobic surface area and melting point, the polymer interaction can occur with different behaviour and mechanism. In order to investigate whether the mPEG– cholane association technology can be a useful approach to produce effective rh-GH sustained delivery systems and can be applied to proteins with different biophysical properties, an rh-GH/mPEG–cholane association study was undertaken. Therefore, this study is also aimed at obtaining information about the effect of the biophysical properties of the protein on the protein/polymer association. In order to select a suitable hormone/polymer formulation, a comparative association study was carried out using polymers with different size (mPEG5kDa–cholane, mPEG10kDa–cholane, mPEG20kDa– cholane). The association properties were examined by size exclusion

169

chromatography and isothermal calorimetry. In vitro studies were carried out to evaluate the biopharmaceutical properties, namely protein release and stability, whilst in vivo studies were performed to assess the pharmacokinetic and therapeutic profile of the new formulation. 2. Materials and methods Recombinant human growth hormone (rh-GH) was a kind gift of Bio-Ker (Pula, Italy). Linear 5, 10 and 20 kDa monoaminomonomethoxy-poly(ethylene glycol)s (mPEG5kDa–NH2, mPEG10kDa– NH2 and mPEG20kDa–NH2, respectively) were obtained from Nektar (Huntsville, AL, USA). 5β-cholanic acid was purchased from Fluka Chemika (Buchs, Switzerland). Triethylamine (TEA) and trinitrobenzenesulfonic acid (TNBS) were purchased from Aldrich (Milwaukee, WI, USA). In vivo studies were carried out using male Sprague–Dawley rats weighing 200–220 g and female hypophysectomised Sprague Dawley rats weighing 75–85 g. The animals were fed ad libitum. The care and handling of animals used for the pharmacokinetic studies were in accordance with the provisions of EU Council Directive 86/209 (recognised and adopted by the Italian Government with approval decree D.M. No. 230/95-B) and NIH publication No. 85–23, revised in 1985. The in vivo study protocols were approved by the Ethical Committee of the University of Padua. 2.1. Synthesis of mPEG–cholane The mPEG–cholane derivatives were prepared according to the procedure reported elsewhere [31]. Briefly, cholanyl chloride was obtained by the reaction of 5β-cholanic acid with thionyl chloride by 3 h refluxing under nitrogen atmosphere. The cholanyl chloride recovered by distillation and desiccated under vacuum was reacted overnight with mPEG5kDa–NH2, mPEG10kDa–NH2 or mPEG20kDa–NH2 (2:1 cholanic chloride/mPEG–NH2 molar ratio) in methylene chloride added of TEA. The product (mPEG–cholane) was precipitated in diethyl ether, dissolved in water, and centrifuged and the aqueous solution was lyophilised. A 5 mg/mL aqueous solution of product was analysed by the TNBS assay [32] to determine the concentration of free amino groups of unreacted mPEG–NH2 and by iodine test [33] to determine the PEG concentration. The iodine assay was performed by mixing 1–10 μL of polymer solution with 250 μL of 50 mM I2 and 2%KI in water, 250 μL 5% BaCl2 in 1 N HCl and 1 mL water. After 15 min, the optical density was determined at 535 nm and referred to a titration curve obtained by using mPEG–OH solutions at known polymer concentrations. The results obtained by iodine and TNBS assay were elaborated to derive the PEG/cholane conjugation yield. An aliquot of product dissolved in CDCl3 was analysed by 1 H NMR spectrometry using a Spectrospin 300 spectrometer (Bruker, Fallanden, Switzerland). The integration of signals corresponding to protons of PEG and the cholanyl moiety at δ 0.64 [s, 3H, CH3 (C19) of cholane], δ 0.91–0.93 [s + d, 6H, CH3 (C18) + CH3 (C21) of cholane], δ 3.38 [s, 3H, CH3O–PEG], δ 3.64 [s, 4nH, −(CH2CH2O)n– of PEG] were elaborated to obtain the degree of PEG/cholane molar ratio. The spectrometric results showed that 99.1–99.5% of mPEG–NH2 was derivatised with the cholanyl moiety. 2.2. rh-GH/mPEG–cholane associations rh-GH/mPEG–cholane samples were prepared by adding 40 μL of 2 mg/mL rh-GH solution in 20 mM phosphate buffer, 154 mM NaCl (PBS), pH 7.4, to 0, 5, 10, 20, 30, 40, 60, or 80 μL of a 15 mg/mL mPEG– cholane solution in the same buffer. The samples were added of volumes of PBS, pH 7.4, to reach a final volume of 120 μL. The protein/polymer solutions were maintained overnight at room temperature under mild top-down mixing. The samples were centrifuged at 5000 rpm for 3 min and analysed by size exclusion chromatography (SEC) using a Bio-Gel SEC 40XL (300 × 7.8 mm) column (Bio Rad, Hercules, CA,

170

S. Salmaso et al. / Journal of Controlled Release 194 (2014) 168–177

USA) operated on a HPLC system. The column was isocratically eluted with 97% of 63 mM phosphate buffer and 3% isopropanol, pH 7.4, and the UV detector was set at 280 nm. The amount of associated protein was indirectly determined by elaborating the peak area of the nonassociated protein using a standard titration curve obtained by eluting 2–80 μg/mL rh-GH solutions in PBS, pH 7.4 (y = 10,859x; R2 = 0.9916). Each experiment was repeated four times. The size exclusion chromatography data (bound protein and free protein) were elaborated according to the Scatchard analysis to calculate the protein/polymer affinity constants (Ka) [34].

2.3. Microcalorimetric analysis Isothermal titration calorimetry (ITC) was carried out using a MSCITC equipment (Microcal Inc. Northampton, MA). mPEG–cholane and rh-GH solutions in PBS, pH 7.4, were filtered with 0.2 μm cut-off filters, sonicated and thermostated at 25 °C. The experimental conditions were set up according to the molar peptide/polymer association calculated by SEC. In the first set of experiments, every 4 min, 2 or 5 μL of 2.236 mM mPEG5kDa–cholane, mPEG10kDa–cholane and mPEG20kDa–cholane solutions was injected into the calorimetric cell containing 1.439 mL of PBS, pH 7.4, with an automated syringe. The protein solution in the cell was maintained under stirring at 400 rpm, and the analyses were performed at 25 °C. In the second set of experiments, 10 or 5 μL of 2 mM mPEG5kDa–cholane, mPEG10kDa–cholane and mPEG20kDa–cholane solutions was injected into the calorimetric cell containing 1.439 mL of 7 μM rh-GH solutions in PBS, pH 7.4. All measurements were replicated three times, and data processing was performed with the Microcal Origin 7 software.

2.4. Dialysis studies

2.6. Proteolytic studies Three samples were prepared by mixing 500 μL of a 0.4 mg/mL rhGH solution in 20 mN HCl with: A. 500 μL of 20 mN HCl; B. 500 μL of 5.4 mg/mL mPEG5kDa–cholane in 20 mN HCl; C. 500 μL of 5.4 mg/mL or mPEG5kDa–OH in 20 mN HCl. The samples were added of 500 μL of 25 μg/mL pepsin in 20 mN HCl and incubated at room temperature. At scheduled times, 50 μL volumes were withdrawn and added of 50 μL of 0.1 M borate, pH 8.0 and analysed by RP-HPLC as reported above to evaluate the undegraded rh-GH. Each experiment was repeated five times.

2.7. Shear stress studies Samples were prepared by mixing 500 μL of a 0.4 mg/mL rh-GH solution in PBS, pH 7.4, with 500 μL of PBS, pH 7.4, or 500 μL of a 5.4 mg/mL of mPEG5kDa–cholane or mPEG5kDa–OH in PBS, pH 7.4. The solutions were maintained under stirring at room temperature in sealed vials and 20 μL samples were withdrawn at scheduled times, diluted with 80 μL of PBS, pH 7.4, centrifuged for 3 min at 5000 rpm and analysed by RP-HPLC as reported above to evaluate the undenaturated rh-GH. Each experiment was repeated five times. 2.8. Pharmacokinetic studies Twenty-four male Sprague–Dawley rats were randomly divided into 4 groups of 6 animals each. The animals were subcutaneously administered on the flank with: Group 1 (Control), 200 μL of a mPEG5kDa– cholane solution in PBS, pH 7.4, in order to inject 33.3 mg/kg of polymer; Group 2, 200 μL of rh-GH in PBS, pH 7.4, in order to inject 1 mg/kg hormone; Group 3, 200 μL of 7.5 w/w% rh-GH/mPEG5kDa–OH in PBS, pH 7.4, in order to inject 1 mg/kg of rh-GH; Group 4, 200 μL of 7.5 w/w% rh-GH/mPEG5kDa–cholane in PBS, pH 7.4, in order to inject 2.5 mg/kg of rh-GH. Blood samples were withdrawn from the caudal vein at 0, 1, 2, 4, 8, 24, 32, 48 and 72 h after administration and analysed by using a h-GH ELISA kit (Abnova GmbH, Heidelberg, Germany). The ELISA titration was preliminarily validated for the quantitative determination of rh-GH formulated with mPEG5kDa–cholane in blood. The mean concentration values of rh-GH at each time point were elaborated to derive the main pharmacokinetic parameters: maximal concentration (Cmax), maximal concentration time (tmax), half-life (t1/2) and area under the curve (AUC0-t).

rh-GH/mPEG–cholane mixtures were prepared by adding 200 μL of a 1 mg/mL rh-GH solution in PBS, pH 7.4, to 300 μL of 8.9, 6.7, and 4.5 mg/mL mPEG5kDa–cholane or 17.1 mg/mL mPEG10kDa–cholane or 25.6 mg/mL of mPEG20kDa–cholane solutions in PBS, pH 7.4. Similarly, rh-GH/mPEG–OH mixtures were prepared using mPEG5kDa–OH, mPEG10kDa–OH and mPEG20kDa–OH in place of the mPEG–cholane counterparts. Samples (500 μL) with same mPEG–cholane or mPEG–OH concentration reported above but without rh-GH and samples (500 μL) of 0.4 mg/mL of rh-GH in PBS, pH 7.4, were used as reference. The solutions were placed into 50 kDa cut-off dialysis Float-A-Lyzer tubes (Spectrum Laboratories, Los Angeles, CA, USA) and dialyzed against 200 mL of PBS pH 7.4 at room temperature. At scheduled times, 10 μL of the samples in the donor compartment were withdrawn, added of 50 μL buffer and analyzed by the iodine colorimetric assay for the determination of the polymer concentration and by RP-HPLC for the determination of the rh-GH content. The RP-HPLC analyses were performed using an analytical C-18 Luna column, 250 × 4.6 mm, (Phenomenex, Torrance, CA, USA) eluted with H2O/0.05% TFA (eluent A) and acetonitrile/0.05% TFA (eluent B) in a gradient mode: 0–3 min 45% eluent B, 3–23 min from 45% to 70% eluent B. The UV detector was set at 210 nm. The protein amount was determined on the basis of a protein titration curve obtained using dilutions of a protein solution in PBS pH 7.4 (y = 15,995x; R2 = 0.9965). The experiment was replicated four times.

Sixteen hypophysectomised female Sprague Dawley rats weighing 75–85 g were randomly divided into four groups of 4 animals each, which were subcutaneously treated on days 0 and 3 with 300 μL of PBS, pH 7.4 containing: Group A, 1.6 mg of mPEG5kDa–cholane; Group B, 120 μg of rh-GH in; Group C, 120 μg of rh-GH equivalent dose of 7.5 w/w% rh-GH/mPEG5kDa–cholane; Group D, 120 μg of rh-GH equivalent dose of 7.5 w/w% rh-GH/mPEG5kDa–OH. The animals were daily weighed and the activity of the hormone was estimated by changes in the body weight as reported in the literature [12].

2.5. Spectroscopic analysis

3. Results

Circular dichroism (CD) analyses were performed using a J-810 spectrodichrograph (Jasco, Tokyo, Japan). The spectra were recorded in the far UV (200–250 nm) using 0.05 mg/mL rh-GH solutions or equimolar rh-GH concentrations of a 7.5 w/w% rh-GH/mPEG5kDa–cholane or rh-GH/mPEG–OH5kDa mixtures in PBS, pH 7.4.

The mPEG–cholane derivatives represented in Scheme 1 (mPEG5kDa– cholane, mPEG10kDa–cholane and mPEG20kDa–cholane) were synthesised according to a simple and reproducible protocol, which allows for the extensive end-functionalisation of mPEG–NH2 with one cholanyl moiety [31].

2.9. Pharmacodynamic studies

S. Salmaso et al. / Journal of Controlled Release 194 (2014) 168–177

171

Scheme 1. Representation of the mPEG–cholane structure.

3.2. Isothermal titration calorimetry Calorimetric studies were undertaken to elucidate the nature of the rh-GH/mPEG–cholane association and the structure of the supramolecular assemblies. Isothermal titration calorimetry (ITC) was carried out by injecting volumes of buffer into rh-GH solution in the cell (blank1)

Associated rh-GH (mg/mL)

The physical association of rh-GH with the three mPEG–cholane conjugates was evaluated by size exclusion chromatography (SEC) of rh-GH/mPEG–cholane mixtures at different hormone/peptide molar ratios. Typical SEC profiles obtained with rh-GH with and without mPEG–cholane are reported in Fig. 1. As the rh-GH/mPEG–cholane molar ratio decreased, the peak area corresponding to non-associated rh-GH decreased whilst peaks at lower elution times corresponding to the associated rh-GH/mPEG– cholane were observed. Since the SEC peaks corresponding to the hormone/polymer assemblies were broad with shoulders, the extent of rh-GH/mPEG–cholane association in each sample was indirectly derived from the peak area of the non-associated rh-GH, which eluted as a well-defined narrow peak. The use of a titration curve allowed for the quantitative determination of the non-associated protein. The association profiles (w/w and mol/mol rh-GH/mPEG–cholane association) are reported in Fig. 2 and the association parameters obtained by the elaboration of the experimental data are summarised in Table 1. The w/w association profiles depicted in Fig. 2A and the w/w% polymer loading capacities reported in Table 1 show that the mPEG–cholane association capacity decreases as the polymer molecular weight increases. However, the mol/mol association profiles showed in Fig. 2B and the mol/mol% values reported in the table demonstrate that, according to the different molecular weight of the polymers, similar molar association ratios were obtained with the three mPEG–cholane derivatives. The polymer loading capacity was calculated as the rh-GH/mPEG– cholane w/w% and mol/mol% corresponding to complete rh-GH association. The association constants were obtained by the elaboration of the experimental data by Scatchard analysis.

The association constants (Ka) reported in Table 1 show that the mPEG5kDa–cholane and mPEG10kDa–cholane have similar affinity for rh-GH whilst mPEG20kDa–cholane displays significantly lower affinity as compared to the other polymers.

A 0.6

0.4

0.2

0 0

2

4

6

8

10

12

[mPEG-cholane] (mg/mL) 0.03

Associated rh-GH (mM)

3.1. Size exclusion chromatographic studies

B

0.02

0.01

0 0

0.5

1

1.5

2

[mPEG-cholane] (mM)

Fig. 1. Size exclusion chromatography (SEC) rh-GH profiles: rh-GH without mPEG–cholane (\); 13.3 w/w% rh-GH/mPEG5kDa–cholane (•••••); 8.8 w/w% rh-GH/mPEG5kDa–cholane (- - -).

Fig. 2. Association profiles of rh-GH with mPEG5kDa–cholane (●), mPEG10kDa–cholane ( ) and mPEG20kDa–cholane (○) obtained by SEC. A. rh-GH/mPEG–cholane w/w association profiles. Linear fitting: mPEG5kDa–cholane, y = 0.0745x R2 0.98; mPEG10kDa–cholane, y = 0.0411x R2 0.97; mPEG20kDa–cholane, y = 0.0262x R2 0.93. B. rh-GH/mPEG–cholane mol/mol association profiles. Linear fitting: mPEG5kDa–cholane, y = 0.018 R2 0.98; mPEG10kDa–cholane, y = 0.019 R2 0.97; mPEG20kDa–cholane y = 0.024x R2 0.93.

172

S. Salmaso et al. / Journal of Controlled Release 194 (2014) 168–177

Table 1 rh-GH/mPEG–cholane association parameters calculated by SEC analysis: association constant (Ka), polymer loading capacity expressed as the rh-GH/mPEG–cholane w/w% and mol/mol% inducing the complete rh-GH association. Association constant (Ka) by Scatchard

mPEG5 kDa–cholane mPEG10 kDa–cholane mPEG20 kDa–cholane

rh-GH/mPEG–cholane loading by SEC

M−1

w/w%

mol/mol%

6.6 × 105 7.7 × 105 0.4 × 105

7.5 ± 1.1 3.9 ± 0.4 2.6 ± 0.4

1.8 ± 0.3 1.8 ± 0.2 2.3 ± 0.4

and polymer solutions into the cell containing buffer (blank 2) or into rh-GH solutions (samples). The buffer injection into rh-GH solution (blank 1) did not elicit any significant thermal effect. Instead, the injection of aliquots of polymer solutions into the cell containing buffer (blank 2) produced bimodal exothermic ITC profiles (data not shown). The amount of heat released by single mPEG–cholane injections (siΔΗ) increased up to a maximal value. The polymer concentrations in the cell corresponding to the maximal siΔH were in the range of 10–60 μM depending on the mPEG–cholane size, which are in good agreement with the CMC previously determined by spectrometric analyses [31]. After the maximal siΔH value was achieved, constant siΔH values were maintained for further polymer additions. It should be noted that the heat released by single polymer solution injection (siΔH) in the plateau region was found to increase as the polymer molecular weight increased. Fig. 3 shows the ITC profiles obtained by injecting volumes of polymer solutions into the cell containing rh-GH solutions. The profiles show that the rh-GH association with the mPEG–cholane conjugates occurs according to a multimodal behaviour. Similarly to what observed with “blank 2”, the upper panels of Fig. 3 show that in the first step of the association the siΔH increases to maximal siΔH values, which correspond to mPEG5kDa–cholane, mPEG10kDa– cholane and mPEG20kDa–cholane concentrations in the cell of 26, 42 and 65 μM, respectively. After that, the siΔH values constantly decrease to achieve, in the last part of the thermogram, a plateau (steady state) with constant siΔH values. At the steady state, the siΔH values measured for the rh-GH/mPEG–cholane association were similar to the

Table 2 Association parameters of rh-GH with mPEG5kDa–cholane, mPEG10kDa–cholane, mPEG20kDa–cholane: number of association sites of rh-GH for mPEG–cholane (N), association constant (Ka), association enthalpy (ΔH), association entropy (ΔS). N mPEG5kDa–cholane mPEG10kDa–cholane mPEG20kDa–cholane

21.5 ± 0.25 22.2 ± 0.43 23.47 ± 1.07

Ka M−1 5

1.24 ± 0.13 10 1.05 ± 0.15 105 0.38 ± 0.85105

ΔH

ΔS

−641.0 ± 10.8 −769.6 ± 23.1 −1020.6 ± 67.1

25.3 25.4 24.3

values obtained at the steady state of “blank 2” (polymer solution injected in buffer without rh-GH). The calorimetric raw data obtained by mPEG–cholane addition to the rh-GH solution, subtracted of the values obtained with the corresponding blanks, were found to fit a one set binding site elaboration model. The association parameters derived from the analysis are summarised in Table 2. The ITC profiles depicted in Fig. 3 and the data reported in Table 2 show that rh-GH has a similar number of binding sites (N) for the three polymers whilst the association constants (Ka) were found to decrease as the mPEG–cholane molecular weight increases. The overall rh-GH/mPEG–cholane association process was found to occur by enthalpy (ΔH) decrease and entropy (ΔS) increase. The overall association absolute ΔH value was found to increase as the polymer molecular weight increases whilst the ΔS was similar for all the mPEG– cholane derivatives.

3.3. Dialysis studies Dialysis studies were carried out using rh-GH and rh-GH/mPEG– cholane mixtures at a protein/polymer w/w% corresponding to the complete rh-GH association with the polymer as calculated by SEC: 7.5 w/w% rh-GH/mPEG5kDa–cholane, 3.9 w/w% rh-GH/mPEG10kDa– cholane and 2.6 w/w% rh-GH/mPEG20kDa–cholane. For comparison, rh-GH/mPEG–OH samples prepared using mPEG5kDa–OH, mPEG10kDa– OH and mPEG20kDa–OH with the same hormone/polymer w/w% composition used in the case of the rh-GH/mPEG–cholane samples were investigated.

Fig. 3. ITC profiles obtained by rh-GH titration with mPEG–cholane: A. mPEG5kDa–cholane; B. mPEG10kDa–cholane; and C. mPEG20kDa–cholane.

S. Salmaso et al. / Journal of Controlled Release 194 (2014) 168–177

A

Released rh-GH (%)

100

Released rh-GH (%)

80 60 40 20 0 0

50

100

150

200

150

200

Time (hours)

B

100 80 60 40 20 0 0

50

100

Time (hours) Fig. 5. rh-GH and mPEG–cholane dialysis profiles from rh-GH and rh-GH/mPEG–cholane solutions. A. rh-GH release: rh-GH (■); 7.5 w/w% rh-GH/mPEG5kDa–cholane (●); 3.9 w/w% rh-GH/mPEG10kDa–cholane ( ); 2.6 w/w% rh-GH/mPEG20kDa–cholane (○). B. mPEG–cholane release; 7.5 w/w% rh-GH/mPEG5kDa–cholane (●); 3.9 w/w% rh-GH/mPEG10kDa–cholane ( ); 2.6 w/w% rh-GH/mPEG20kDa–cholane (○).

above, the product containing rh-GH at 7.5 w/w% rh-GH/mPEG5kDa– cholane did not show burst release and about 80% of protein release was achieved in 200 h.

80 60

3.4. Spectroscopic studies

40 20 0 0

20

40

60

80

Time (hours)

Spectroscopic studies were carried out to evaluate the effect of the polymer association on the secondary conformation of rh-GH. Fig. 7 shows that similar circular dichroism spectra were obtained with rh-GH in buffer alone and in association with mPEG5kDa–OH or mPEG5kDa–cholane indicating that the polymer association does not alter the secondary structure of the hormone. The circular dichroic

B

100

100

80

Released rh-GH (%)

Released mPEG-OH (%)

A

100

Released mPEG-cholane (%)

Another set of studies was carried out using rh-GH/mPEG5kDa– cholane with same rh-GH concentration and different hormone/ polymer w/w%: 15, 10 and 7.5 w/w%. Fig. 4 reports the rh-GH and mPEG–OH dialysis profiles obtained with rh-GH and rh-GH/mPEG–OH mixtures. Fig. 4A shows that rh-GH alone and combined with mPEG–OH have similar dialysis rate indicating that mPEG–OH does not affect the hormone release. Over 70% of the protein was released in 24 h and over 90% in 48 h. Fig. 4B shows that mPEG–OH is rapidly released by dialysis from the rh-GH/mPEG–OH solutions, regardless the molecular weight. Fig. 5 reports the rh-GH and mPEG–cholane dialysis profiles obtained with rh-GH alone or rh-GH combined with mPEG–cholane. The dialysis profiles depicted in Fig. 5A show that the three polymers significantly decrease the release rate of rh-GH. rh-GH/mPEG20kDa– cholane showed a rh-GH burst release of about 28% in 8 h followed by a constant release to achieve about 77% released hormone in 200 h. rh-GH/mPEG10kDa–cholane showed a rh-GH burst release of about 8% in 8 h and also in this case after the burst the hormone was linearly released to yield about 85% drug release in 200 h. Finally, rh-GH/ mPEG5kDa–cholane did not show burst release and 80% of rh-GH was linearly released in 200 h. Fig. 5B shows that mPEG–cholane was also slowly dialysed. In all cases, the polymer release was significantly lower as compared to the mPEG–OH counterparts (Fig. 4B). Less than 50% of mPEG–cholane was released in 200 h. Fig. 6 shows the rh-GH release profiles obtained with 15, 10 and 7.5 w/w% rh-GH/mPEG5kDa–cholane. The figure shows that the protein release rate increases as the rh-GH/mPEG5kDa–cholane ratio increases. rh-GH was rapidly released in 65 h from the 15 w/w% rh-GH/ mPEG 5kDa –cholane sample. The rh-GH release form the 10 w/w% rh-GH/mPEG5kDa–cholane sample occurred with a 20% burst during the first 8 h and then it achieved about 90% release in 150 h. As reported

173

60 40 20 0 0

20

40

60

80 60 40 20

80

Time (hours) Fig. 4. rh-GH and mPEG–OH dialysis profiles from rh-GH and rh-GH/mPEG–OH solutions. A. rh-GH release: rh-GH (■); 7.5 w/w% rh-GH/mPEG5kDa–OH (●); 3.9 w/w% rh-GH/mPEG10kDa–OH ( ); 2.6 w/w% rh-GH/mPEG20kDa–OH (○). B. mPEG–OH release: 7.5 w/w% rh-GH/mPEG5kDa–OH (●); 3.9 w/w% rh-GH/mPEG10kDa–OH ( ); 2.6 w/w% rh-GH/mPEG20kDa–OH w/w% (○).

0 0

50

100

150

200

Time (hours) Fig. 6. rh-GH dialysis profiles from rh-GH and rh-GH/mPEG5kDa–cholane mixtures: rh-GH (■); 7.5 w/w% rh-GH/mPEG5kDa–cholane (●); 10 w/w% rh-GH/mPEG5kDa–cholane ( ); 15 w/w% rh-GH/mPEG5kDa–cholane (○).

174

S. Salmaso et al. / Journal of Controlled Release 194 (2014) 168–177

rh-GH in plasma (ng/mL)

1000.0

10.0

0.1 0

20

40

Time (hours) Fig. 7. Circular dichroism profiles of rh-GH (\), rh-GH/mPEG5kDa–cholane (- - - -) and rh-GH/mPEG5kDa–OH (•••••).

spectra analysed by the Varselec fit programme showed that rh-GH either in polymer-free solution or in the polymer assembly has same alpha helix content. 3.5. Stability studies The effect of the mPEG5kDa–cholane association on enzymatic and physical stability of rh-GH was comparatively examined by using rhGH, and the 7.5 w/w% rh-GH/mPEG5kDa–cholane or rh-GH/mPEG5kDa– OH solutions. The enzymatic stability of rh-GH was evaluated using pepsin, a proteolytic enzyme that was previously found to rapidly degrade the hormone [11]. Fig. 8A shows that the rh-GH association with mPEG5kDa–cholane remarkably prevents the hormone proteolysis whilst mPEG5kDa–OH does not possess a substantial protecting effect on the rh-GH stability. Indeed, the addition of mPEG5kDa–OH was

Undegraded rh-GH (%)

120

A

100 80 60 40 20 0 0

1

2

3

4

Time (hours) 120

B

Folded rh-GH (%)

100 80 60 40

Fig. 9. Pharmacokinetic profiles of rh-GH after subcutaneous administration to rats of 1 mg/kg dose of rh-GH in buffer (○) and 2.5 mg/kg rh-GH dose of 7.5 w/w% rh-GH/mPEG5kDa–cholane (●).

found to slightly increase the half-life degradation time of rh-GH from 1 to 1.25 h. The half-life of rh-GH associated with mPEG5kDa–cholane was 3.75 h and the protein was completely degraded in 10 h. Fig. 8B shows the shear stress stability of rh-GH and rh-GH combined with mPEG5kDa–cholane or mPEG5kDa–OH. Both mPEG5kDa–cholane and mPEG5kDa–OH were found to delay the rh-GH denaturation under stressing conditions, even though the former was much more efficient than the latter. The hormone denaturation half-life time increased from 0.8 to 3.4 h in the presence of mPEG5kDa–OH and from 0.8 to 36.2 h in the presence of mPEG5kDa–cholane. After 72 h the rh-GH formulated with mPEG– cholane was completely denaturated. 3.6. Pharmacokinetic studies Pharmacokinetic studies were carried out by subcutaneous injections of rh-GH (1 mg/kg dose), 7.5 w/w% rh-GH/mPEG5kDa–OH (1 mg/kg rhGH dose) and 7.5 w/w% rh-GH/mPEG5kDa–cholane (2.5 mg/kg rh-GH dose) to rats. Plain buffer and polymer solutions were administered as controls. A preliminary pharmacokinetic study was carried out to set up the experimental conditions. Accordingly, hormone dose and experimental time points were selected to obtain detectable hormone concentrations in the bloodstream and perform accurate PK analyses. The protein concentrations in the blood at the experimental time points were assessed by ELISA according to a protocol that was preliminary validated. No differences were found in the rh-GH pharmacokinetic profiles obtained with rh-GH in the absence of polymer or combined with mPEG5kDa–OH (profile not shown). The pharmacokinetic profiles reported in Fig. 9 and the pharmacokinetic parameters summarised in Table 3 show that the mPEG5kDa– cholane association delays the rh-GH absorption by several hours (tmax) and prolongs the half-life (t1/2). Despite the polymer association lowers the maximal protein concentration in the blood (Cmax), the normalised systemic bioavailability (AUC/dose) obtained with rh-GH/mPEG5kDa–cholane was 55% higher than that obtained with the protein in the absence of the polymer.

20 0 0

2

4

6

8

10

Table 3 Pharmacokinetic parameters: peak concentration time (tmax); maximal concentration (cmax); area under the curve (AUC); half-life time (t1/2).

Time (hours) Fig. 8. rh-GH stability profiles. A. rh-GH degradation profile by pepsin proteolysis: rh-GH (○); 7.5 w/w% rh-GH/mPEG 5kDa –OH ( ); 7.5 w/w% rh-GH/mPEG5kDa–cholane (●). B. rh-GH denaturation by stirring: rh-GH (○); 7.5 w/w% rh-GH/mPEG5kDa–OH ( ); 7.5 w/w% rh-GH/mPEG5kDa–cholane (●).

tmax (h) cmax (ng/mL) AUC (ng mL−1 h) t1/2 (h)

rh-GH

rh-GH/mPEG5 kDa–cholane

0.5 374 854 2

8 187 3324 24

S. Salmaso et al. / Journal of Controlled Release 194 (2014) 168–177

16

Weight increase (g)

14 12 10 8 6 4 2 0 -2

0

5

10

Time (days) Fig. 10. Pharmacodynamic profiles obtained by subcutaneous administration on days 0 and 3 (arrows) of: 1.6 mg of mPEG5kDa–cholane (x); 120 μg rh-GH (●); 120 μg rh-GH equivalent dose of 7.5 w/w% rh-GH/mPEG5kDa–cholane (○); 120 μg rh-GH equivalent dose of 7.5 w/w% rh-GH/mPEG5kDa–OH ( ).

3.7. Pharmacodynamic studies In vivo pharmacological studies were performed by twice a week subcutaneous administrations of 120 μg of rh-GH or equivalent rh-GH doses of 7.5 w/w% rh-GH/mPEG5kDa–cholane or rh-GH/mPEG5kDa–OH to hypophysectomised female rats. Animals were also treated with mPEG5kDa–cholane alone as control. The control animals, treated with mPEG5kDa–cholane without rh-GH, did not show any weight increase during the experimental time. On the contrary, Fig. 10 shows that all groups treated with rh-GH, alone or in combination with mPEG5kDa–OH or mPEG5kDa–cholane elicited weight increase after hormone treatment. The administration of rh-GH and rh-GH/mPEG5kDa–OH stimulated similar weight increase profile over time. During the first 48 h from the first administration, the three rhGH formulations elicited similar weight increase. At the third day, the weight of the animals treated with rh-GH and rh-GH/mPEG5kDa–OH decreased whilst the weight of the animals treated with rh-GH/ mPEG5kDa–cholane still kept increasing. After the second injection, the weight of the animals treated with rh-GH and rh-GH/mPEG5kDa–OH showed a significant increase. However, since the day after the second injection the weight of these animals was practically constant. On the contrary, the weight of the rats treated with rh-GH/mPEG5kDa–cholane continuously increased throughout the days. After 10 days from the first administration the weight increase of the animals treated with rh-GH/ mPEG5kDa–cholane was about 2 times higher as compared to the animals treated with rh-GH and rh-GH/mPEG5kDa–OH. 4. Discussion Size exclusion chromatography (SEC) and isothermal titration calorimetry (ITC) studies showed that the amphiphilic mPEG–cholane derivatives efficiently associate with rh-GH to yield non-covalent supramolecular protein/polymer assemblies. The quantitative analysis of rh-GH/mPEG–cholane association by SEC was rather complex. In all cases, the addition of mPEG–cholane to the rh-GH solution resulted in both increase of the size of the hormone/polymer assemblies and decrease of the relative abundance of the non-associated rh-GH. This behaviour indicates that mPEG– cholane was either engaged with the protein to form nanocomplexes or with the protein/polymer nanocomplex to increase the apparent size of the supramolecular assemblies. Therefore, equilibrium occurs amongst non-associated rh-GH and rh-GH/mPEG–cholane assemblies with different protein/polymer composition. The rh-GH/mPEG–cholane association constants calculated by Schatchard elaboration of the SEC data are in fair agreement the values calculated by ITC and are consistent with hydrophobic interactions between the polymer and the hormone [35,36]. Amphiphilic polymers,

175

namely polymeric surfactants and self-assembling polymers, can in fact interact with proteins through hydrophobic bonding, electrostatic interactions and hydrogen bonds. The mPEG–cholane derivatives do not present charges, whilst hydrogen bonding usually occurs only under strict conformational conditions since requires not only the presence of hydrogen-bond donors and acceptors, but also the spatial alignment of the donor–acceptor pairs [36]. On the other side, rh-GH is a small protein (192 aminoacids, 22.26 kDa) characterised by large hydrophobic surfaces, which are involved in the interactions with the two human growth hormone binding protein of the extracellular domain of the membrane receptor, hGHbpI and hGHbpII, respectively [37]. In the native state, about 58% of total rh-GH exposed area is nonpolar; 1230 and 500 Å2 of the protein surface are engaged for the hGHbpI and hGHbpII binding, respectively. These non-polar surfaces are potential interaction sites with hydrophobic moieties [38]. Although studies carried out with hydrophobic fluorescent probes have demonstrated that the oxyethylene units of PEG can directly interact with protein surfaces through nonspecific hydrophobic interactions, the protein–polymer attraction in the short range is usually weak [39,40]. Therefore, it is reasonable to explain the Ka values derived by the SEC and ITC analysis as descriptors of the association efficiency between the cholane moiety of the polymer bioconjugates and hydrophobic spots or pockets on the protein surface. mPEG20kDa–cholane was found to associate with rh-GH with a lower Ka with respect to the other polymers. This result seems to indicate that high molecular weight PEG chains, 20 kDa, hamper the cholane interaction with the hydrophobic spots of the protein. This effect is ascribable to the high free energy of the bulky PEG chains, which increases with the polymer molecular weight, thus reducing the energetic gain deriving from the enthalpy of binding with the protein. The ITC studies showed the typical profile occurring in the case of proteins titrated with micelle forming polymers, namely surfactants [41]. The rh-GH titration with mPEG–cholane revealed three thermal events associated to polymer demicellisation, polymer dilution and protein/polymer association. The mPEG–cholane demicellisation occurs according to an endothermic process whilst both polymer dilution, corresponding to the bulk hydration effect, and protein/polymer association occur by exothermic processes. During the first additions of mPEG–cholane to the rh-GH solution, the endothermic process due to demicellisation is compensated by the heat release due to the polymer dilution and protein/polymer association. As a result, the absolute value of heat released by single polymer addition (siΔH) is initially low and increases until the CMC in the cell is achieved when the maximal siΔH values were registered. After that, only exothermic events due to polymer dilution and protein/polymer association take place. The polymer dilution occurs with constant heat release at each single injection (constant siΔH) whilst the siΔH derived from the protein/polymer interaction decreases throughout the rh-GH surface saturation with the polymer. Therefore, the resulting absolute siΔH values due to the combination of polymer dilution and protein/polymer association decrease throughout the titration process up to a plateauing of the siΔH values. At plateau, constant siΔH values correspond to polymer dilution only when all rh-GH has been completely associated with the polymer. It should be noted that the siΔH values in the plateau region increase as the polymer molecular weight increases. This is in agreement with the fact that the exothermal event depends on the molar concentration of the oxyethylene units during dilution; the higher the molecular weight of the mPEG–cholane, the higher the oxyethylene unit molar concentration. The ITC data fitted a single set binding site model and showed that rh-GH has 20–25 association sites for the mPEG–cholane. Since the dichroic profile showed that the mPEG–cholane association does not alter the secondary structure of rh-GH, we speculated that the cholane interaction takes place on the protein surface or with hydrophobic pockets without generation of newly exposed hydrophobic areas, which would imply the appearance of secondary binding sites [42].

176

S. Salmaso et al. / Journal of Controlled Release 194 (2014) 168–177

This hypothesis seems to be supported by previous studies showing that the mPEG–cholane association with rh-G-CSF occurs according to a bimodal behaviour with two sets of binding sites. In such a case, in fact, the association behaviour could be ascribed to binding sites available on the protein surface and into pockets formed by the polymer association that was found to induce slight reversible structural protein alteration. Furthermore, it should be noted that the two proteins possess different melting points, which is reported to be 76 °C and 60 °C in the case of rh-GH and rh-G-CSF, respectively [43,44]. Since this biophysical property is related to the physical stability of the protein, it seems reasonable to expect that the less stable rh-G-CSF undergoes partial unfolding by the amphiphilic polymer interaction more easily than the more stable rh-GH. The overall ITC results indicate that the rh-GH/mPEG–cholane association is enthalpically and entropically favoured. Therefore, it is conceivable that the hormone/polymer association is stronger than the polymer/polymer association in the micelles and the polymer chains on the protein surface are not as ordered as in the micelle structure. The dialysis studies showed that the association of rh-GH with mPEG–cholane allows for the slow release of the hormone, which confirms the hypothesis of this study. The parallel studies performed with mPEG–OH showed that delayed protein dialysis does neither depend on the interactions of rh-GH with the PEG chains nor on the viscosity due to the polymer. The rh-GH release rate was found to depend on the mPEG–cholane molecular weight and rh-GH/mPEG–cholane ratio. According to its lower association constant, the rh-GH/mPEG20kDa–cholane assembly yielded initial burst release. On the contrary, the rh-GH association with mPEG5kDa–cholane, which has the highest affinity for the protein, did not produce any burst effect. After the first 8 h, the protein release from all the formulations occurred with zero order-like kinetic for about 200 h, without significant differences amongst the three polymers, indicating that the protein release is not dictated by diffusion only. Indeed, the polymer dialysis occurring during the protein dialysis partially compensates the decrease of the rh-GH/mPEG–cholane ratio in the donor compartment thus affecting the hormone release rate. The dialysis results obtained with different rh-GH/mPEG5kDa–cholane ratios show that the protein release rate decreases as the rh-GH/ mPEG–cholane molar ratio decreases. Therefore, proper rh-GH/mPEG– cholane molar ratio can be selected to modulate the hormone release in vivo. Based on the results obtained by SEC, ITC and dialysis, mPEG5kDa– cholane was selected for the development of supramolecular systems for rh-GH delivery. Indeed, by virtue of its low molecular weight and high association constant, a smaller amount of polymer is required to associate rh-GH and yield slow protein release without burst effect. The capacity of mPEG–cholane to enhance the physical and enzymatic stability of proteins represents a main advantage for the pharmaceutical exploitation of rh-GH/mPEG–cholane formulations. Actually, rh-GH is a fragile protein susceptible to denaturation and degradation during manipulation, formulation and delivery, which limit the development of adequate delivery systems. In the native state, the growth hormone exposes a large number of aromatic amino acids, which enhance its tendency to misfold and aggregate. [45]. The physical stress promotes the exposure of uncompensated hydrophobic surfaces that lead to the peptide tendency to irreversibly unfold and aggregate forming insoluble products. Amphiphilic polymers, namely PEG and polysorbates, have been shown to enhance the rh-GH stability to shear stress and other physical and chemical harsh conditions [38,46, 47]. In this study mPEG5kDa–cholane was found to possess a higher rhGH stabilising effect than mPEG5kDa–OH. This result once again demonstrates that the cholane moiety allows for tight polymer anchoring to the protein hydrophobic surface through stable bonding and possibly promotes the stabilisation of transiently unfolded peptide conformations. The resulting polymer corona on the rh-GH surface protects the protein from exposure of hydrophobic regions to protein–protein and

protein–surface interactions that initiate the irreversible unfolding process. Similarly, the polymer corona protects rh-GH from enzymatic degradation, which represents one of the main reasons of its in vivo inactivation. It is worth to note that the protective effect of mPEG5kDa– cholane on the proteolytic stability of rh-GH is comparable to that previously reported after chemical conjugation of PEG and PEGMA [11]. The in vivo studies performed by subcutaneous administration of rhGH and rh-GH/mPEG5kDa–cholane to healthy and hypophysectomised rats showed that the formulation of rh-GH by association with mPEG5kDa–cholane enhances the rh-GH bioavailability and its pharmacological performance. Despite the polymer association reduces the maximal rh-GH concentration in the blood, the bioavailablity, calculated by normalising the AUC by the dose, was found to increase as a consequence of the sustained protein absorption that compensates the rapid clearance of the hormone from the bloodstream. Indeed, in the absence of mPEG5kDa–cholane the rh-GH maximal concentration drops to 1% in a few hours whilst mPEG5kDa–cholane maintains high hormone concentrations for several hours. Therefore, the mPEG5kDa–cholane association to rh-GH can mitigate the blood peak concentration of the hormone and yield formulations with prolonged effect thus avoiding frequent injections and resulting in ameliorated pharmacological performance. According to the PK data, the results obtained with hypophysectomised rats showed that the slow protein release from the injection site produces a prolonged animal body weight growth as compared to animal treated with rh-GH in buffer or in combination with mPEG5kDa–OH. Therefore, the pharmacokinetic profile of rh-GH alone is much less effective than that obtained by the hormone associated with mPEG5kDa–cholane. Indeed, the spike of rh-GH concentration in the blood does not produce a prolonged pharmacological effect and its activity lasts for a very short time whilst in the case of the rh-GH/ mPEG5kDa–cholane the activity is maintained for a long period although the maximal concentration is remarkably lower than that obtained with the rh-GH in buffer. These results demonstrate that mPEG5kDa–cholane is a good candidate to obtain sustained release formulations resulting in enhanced therapeutic performance.

5. Conclusions The results reported in this study show that mPEG–cholane can be successfully exploited to produce rh-GH formulations for sustained hormone release. The polymer association can in fact enhance the rh-GH physicochemical and enzymatic stability without chemical and structural alterations of the protein that may yield hormone inactivation during formulation, storage and delivery. Interestingly, the mPEG–cholane association platform consists of a straightforward physical protein/ polymer assembly that avoids chemical modification of the protein and hence the generation of “new drug entities” with benefit for the approval by the safety agencies for human use. The supramolecular assembly can yield favourable pharmacokinetic profiles that allow for avoiding frequent drug injections, which is currently a main issue to the rh-GH therapy. Therefore, the mPEG–cholane supramolecular formulation may result in ameliorated patient compliance and improved therapeutic performance. Furthermore, the possibility to modulate the protein release by proper adjusting the protein/polymer composition makes this technology flexible and finely adaptable to the patient therapeutic requirements. According to the results generated so far with this platform, we believe that it is a versatile and exploitable pharmaceutical platform for a variety of applications. However, the present study shows that the biophysical properties of the protein are paramount in the polymer association. The extent and conformation of hydrophobic surfaces as well as the presence of hydrophobic pockets determine the association parameters, namely number of associated polymer chains, association constants, and in turn the biopharmaceutical properties, namely the protein release profile. Furthermore, polymer association can induce denaturation of structurally unstable proteins. Therefore, suitable

S. Salmaso et al. / Journal of Controlled Release 194 (2014) 168–177

protein candidates for PEG–cholane association technology could be selected according to their biophysical properties. Acknowledgements We acknowledge the University of Padova for financial support through the “Ex-60 %” (grant code CUP C98C13001480005) and “Progetto di Ricerca di Ateneo” (grant code CPDA 121714–CUP C94H12000020005) funding schemes. References [1] B. Leader, Q.J. Baca, D.E. Golan, Protein therapeutics: a summary and pharmacological classification, Nat. Rev. Drug Discov. 7 (2008) 21–39. [2] A. Jain, A. Jain, A. Gulbake, S. Shilpi, P. Hurkat, S.K. Jain, Peptide and protein delivery using new drug delivery systems, Crit. Rev. Ther. Drug Carrier Syst. 30 (2013) 293–329. [3] S. Salmaso, S. Bersani, A. Semenzato, P. Caliceti, Nanotechnologies in protein delivery, J. Nanosci. Nanotechnol. 6 (2006) 2736–2753. [4] D. Clemmons, Growth hormone in health and disease: long-term GH therapy— benefits and unanswered questions, Nat. Rev. Endocrinol. 9 (2013) 317–318. [5] T. Laursen, J.O.L. Jørgensen, S. Susgaard, J. Møller, J.S. Christiansen, Subcutaneous absorption kinetics of two highly concentrated preparations of recombinant human growth hormone, Ann. Pharmacother. 27 (1993) 411–415. [6] E. Lee, Y. Kim, H. Lee, S. Park, H. Jung, J. Lee, Y.-H. Ahn, J. Kim, Stabilizing peptide fusion for solving the stability and solubility problems of therapeutic proteins, Pharm. Res. 22 (2005) 1735–1746. [7] H.J. Chung, Y. Lee, T.G. Park, Thermo-sensitive and biodegradable hydrogels based on stereocomplexed Pluronic multi-block copolymers for controlled protein delivery, J. Control. Release 127 (2008) 22–30. [8] O. Johnson, W. Jaworowicz, J. Cleland, L. Bailey, M. Charnis, E. Duenas, C. Wu, D. Shepard, S. Magil, T. Last, A.S. Jones, S. Putney, The stabilization and encapsulation of human growth hormone into biodegradable microspheres, Pharm. Res. 14 (1997) 730–735. [9] M.-R. Park, C. Chun, S.-W. Ahn, M.-H. Ki, C.-S. Cho, S.-C. Song, Sustained delivery of human growth hormone using a polyelectrolyte complex-loaded thermosensitive polyphosphazene hydrogel, J. Control. Release 147 (2010) 359–367. [10] M.-R. Park, B.-B. Seo, S.-C. Song, Dual ionic interaction system based on polyelectrolyte complex and ionic, injectable, and thermosensitive hydrogel for sustained release of human growth hormone, Biomaterials 34 (2013) 1327–1336. [11] J.P. Magnusson, S. Bersani, S. Salmaso, C. Alexander, P. Caliceti, In situ growth of sidechain PEG polymers from functionalized human growth hormone—a new technique for preparation of enhanced protein− polymer conjugates, Bioconjug. Chem. 21 (2010) 671–678. [12] R. Clark, K. Olson, G. Fuh, M. Marian, D. Mortensen, G. Teshima, S. Chang, H. Chu, V. Mukku, E. Canova-Davis, T. Somers, M. Cronin, M. Winkler, J.A. Wells, Long-acting growth hormones produced by conjugation with polyethylene glycol, J. Biol. Chem. 271 (1996) 21969–21977. [13] K.C.J. Yuen, R. Amin, Developments in administration of growth hormone treatment: focus on Norditropin® Flexpro®, Patient Prefer Adherence 5 (2011) 117–124. [14] S. Salmaso, P. Caliceti, Chapter 11—peptide and protein bioconjugation: a useful tool to improve the biological performance of biotech drugs, in: C.V.D. Walle (Ed.), Peptide and Protein Delivery, Academic Press, Boston, 2011, pp. 247–290. [15] M. Liu, P. Tirino, M. Radivojevic, D.J. Phillips, M.I. Gibson, J.-C. Leroux, M.A. Gauthier, Molecular sieving on the surface of a protein provides protection without loss of activity, Adv. Funct. Mater. 23 (2013) 2007–2015. [16] D.C. González-Toro, S. Thayumanavan, Advances in polymer and polymeric nanostructures for protein conjugation, Eur. Polym. J. 49 (2013) 2906–2918. [17] E.J. Oh, K. Park, K.S. Kim, J. Kim, J.-A. Yang, J.-H. Kong, M.Y. Lee, A.S. Hoffman, S.K. Hahn, Target specific and long-acting delivery of protein, peptide, and nucleotide therapeutics using hyaluronic acid derivatives, J. Control. Release 141 (2010) 2–12. [18] S. Salmaso, P. Caliceti, Self assembling nanocomposites for protein delivery: supramolecular interactions of soluble polymers with protein drugs, Int. J. Pharm. 440 (2013) 111–123. [19] P. van Rijn, Polymer directed protein assemblies, Polymers 5 (2013) 576–599. [20] A. Jintapattanakit, V.B. Junyaprasert, T. Kissel, The role of mucoadhesion of trimethyl chitosan and PEGylated trimethyl chitosan nanocomplexes in insulin uptake, J. Pharm. Sci. 98 (2009) 4818–4830. [21] M. Simon, M. Wittmar, U. Bakowsky, T. Kissel, Self-assembling nanocomplexes from insulin and water-soluble branched polyesters, poly[(vinyl-3-(diethylamino)propylcarbamate-co-(vinyl acetate)-co-(vinyl alcohol)]-graft- poly(l-lactic acid): a

[22]

[23]

[24] [25]

[26]

[27] [28]

[29]

[30] [31]

[32] [33] [34] [35]

[36] [37]

[38] [39]

[40]

[41]

[42] [43]

[44]

[45]

[46] [47]

177

novel carrier for transmucosal delivery of peptides, Bioconjug. Chem. 15 (2004) 841–849. Y.-P. Chan, R. Meyrueix, R. Kravtzoff, O. Soula, G. Soula, Basulin, a long-acting formulation of human insulin based on medusa nanoparticles, Nanobiotechnology 1 (2005) 317–318. S. Salmaso, R. Schrepfer, G. Cavallaro, S. Bersani, F. Caboi, G. Giammona, G. Tonon, P. Caliceti, Supramolecular association of recombinant human growth hormone with hydrophobized polyhydroxyethylaspartamides, Eur. J. Pharm. Biopharm. 68 (2008) 656–666. U. Westphal, Steroid-protein interaction: from past to present, J. Steroid Biochem. 19 (1983) 1–15. X. Tao, Q. Zhang, W. Yang, Q. Zhang, The interaction between human serum albumin and cholesterol-modified pullulan nanoparticle, Curr. Nanosci. 8 (2012) 830–837. J. Fantini, F.J. Barrantes, How cholesterol interacts with membrane proteins: an exploration of cholesterol-binding sites including CRAC, CARC and tilted domains, Front. Physiol. 4 (2013). U. Westphal, S.D. Stroupe, S.L. Cheng, G.B. Harding, Mechanism of steroid binding to serum proteins, J. Toxicol. Environ. Health 4 (1978) 229–247. L. Liu, K. Guo, J. Lu, S.S. Venkatraman, D. Luo, K.C. Ng, E.-A. Ling, S. Moochhala, Y.-Y. Yang, Biologically active core/shell nanoparticles self-assembled from cholesterolterminated PEG–TAT for drug delivery across the blood–brain barrier, Biomaterials 29 (2008) 1509–1517. C.J. Thompson, L. Tetley, I.F. Uchegbu, W.P. Cheng, The complexation between novel comb shaped amphiphilic polyallylamine and insulin—towards oral insulin delivery, Int. J. Pharm. 376 (2009) 46–55. K. Akiyoshi, J. Sunamoto, Supramolecular assembly of hydrophobized polysaccharides, Supramol. Sci. 3 (1996) 157–163. S. Salmaso, S. Bersani, F. Mastrotto, G. Tonon, R. Schrepfer, S. Genovese, P. Caliceti, Self‐assembling nanocomposites for protein delivery: supramolecular interactions between PEG‐cholane and rh‐G‐CSF, J. Control. Release 162 (2012) 176–184. A.F.S.A. Habeeb, Determination of free amino groups in proteins by trinitrobenzenesulfonic acid, Anal. Biochem. 14 (1966) 328–336. G.E.C. Sims, T.J. Snape, A method for the estimation of polyethylene glycol in plasma protein fractions, Anal. Biochem. 107 (1980) 60–63. G. Scatchard, The attractions of proteins for small molecules and ions, Ann. N. Y. Acad. Sci. 51 (1949) 660–672. H.P. Jennissen, Protein binding to two-dimensional hydrophobic binding-site lattices: sorption kinetics of phosphorylase b on immobilized butyl residues, J. Colloid Interface Sci. 111 (1986) 570–586. J.Y. Gao, P.L. Dubin, Binding of proteins to copolymers of varying hydrophobicity, Biopolymers 49 (1999) 185–193. A. de Vos, M. Ultsch, A. Kossiakoff, Human growth hormone and extracellular domain of its receptor: crystal structure of the complex, Science 255 (1992) 306–312. M.R. Kasimova, S.J. Milstein, E. Freire, The conformational equilibrium of human growth hormone, J. Mol. Biol. 277 (1998) 409–418. E.L. Furness, A. Ross, T.P. Davis, G.C. King, A hydrophobic interaction site for lysozyme binding to polyethylene glycol and model contact lens polymers, Biomaterials 19 (1998) 1361–1369. J. Wu, C. Zhao, R. Hu, W. Lin, Q. Wang, J. Zhao, S.M. Bilinovich, T.C. Leeper, L. Li, H.M. Cheung, S. Chen, J. Zheng, Probing the weak interaction of proteins with neutral and zwitterionic antifouling polymers, Acta Biomater. 10 (2014) 751–760. N.B. Bam, J.L. Cleland, T.W. Randolph, Molten globule intermediate of recombinant human growth hormone: stabilization with surfactants, Biotechnol. Prog. 12 (1996) 801–809. M. Sonenberg, S. Beychok, Circular dichroism studies of biologically active growth hormone preparations, Biochim. Biophys. Acta, Protein Struct. 229 (1971) 88–101. P. Luo, R.J. Hayes, C. Chan, D.M. Stark, M.Y. Hwang, J.M. Jacinto, P. Juvvadi, H.S. Chung, A. Kundu, M.L. Ary, B.I. Dahiyat, Development of a cytokine analog with enhanced stability using computational ultrahigh throughput screening, Protein Sci. 11 (2002) 1218–1226. I. Gomez-Orellana, B. Varinano, J. Miura-Fraboni, S. Milstein, D.R. Paton, Thermodynamic characterization of an intermediate state of human growth hormone, Protein Sci. 7 (1998) 1352–1358. D.E. Otzen, B.R. Knudsen, F. Aachmann, K.L. Larsen, R. Wimmer, Structural basis for cyclodextrins' suppression of human growth hormone aggregation, Protein Sci. 11 (2002) 1779–1787. M. Katakam, L.N. Bell, A.K. Banga, Effect of surfactants on the physical stability of recombinant human growth hormone, J. Pharm. Sci. 84 (1995) 713–716. N.B. Bam, J.L. Cleland, J. Yang, M.C. Manning, J.F. Carpenter, R.F. Kelley, T.W. Randolph, Tween protects recombinant human growth hormone against agitation-induced damage via hydrophobic interactions, J. Pharm. Sci. 87 (1998) 1554–1559.