Molecular Sciences Synthesis, Characterization and in Vitro Evaluation of New Composite Bisphosphonate Delivery Systems

In this study, new composite bisphosphonate delivery systems were obtained from polyurethanes (PUs) and nanocrystalline hydroxyapatite (HA). The biodegradable PUs were first synthesized from poly(ε-caprolactone) diols (PCL diols), poly(ethylene adipate) diol, 1,6-hexamethylene diisocyanate, 1,4-butanediol and HA. Moreover, the PCL diols were synthesized by the ring-opening polymerization catalysed by the lipase from Candida antarctica. Next, composite drug delivery systems for clodronate were prepared. The mechanical properties of the obtained biomaterials were determined. The cytotoxicity of the synthesized polymers was tested. The preliminary results show that the obtained composites are perspective biomaterials and they can be potentially applied in the technology of implantation drug delivery systems.


Introduction
Bone metastasis is prevalent in many cancers, especially breast, prostate or lung cancer, the most common neoplasms in the world today [1,2]. Cancer patients with bone metastasis are exposed to numerous skeletal disorders, such as unexpected pathological fractures, serious hypercalcaemia or severe bone pain which is difficult to relieve [1]. Until now, the first-line treatment for bone metastasis has administered bisphosphonates (BPs) [1][2][3][4]. Their mechanism of action is now more clear [5][6][7]. It is commonly known that they inhibit bone resorption by suppressing osteoclastogenesis and osteoclast activity via the farnesyl pyrophosphate synthase enzyme (FPPS) in the mevalonic acid pathway [6,7]. Moreover, recent research shows that BPs may inhibit bone tumour growth and tumour cell invasion in the extracellular matrix. These studies suggest a preventive role played by BPs in tumour metastasis in bone tissues [8,9].
Among the most effective BPs in bone metastasis treatment are: clodronate (CLO), pamidronate, ibandronate and zoledronic acid [2,5,10]. They are administered to patients via two routes-oral or intravenous-though unfortunately they can cause some side effects, such as an acute systemic inflammatory reaction, ocular inflammation, nephrotic syndrome or electrolyte imbalance [11]. Moreover, when applied orally, the bioavailability of these drugs is very low and often insufficient. Thus, it seems reasonable to deliver BPs locally and as a consequence accelerate their local bioavailability.
Aliphatic or cycloaliphatic polyurethanes (PUs) demonstrate good biodegradability and biocompatibility in human tissues. These attributes make them advantageous and extremely useful for the technology of controlled DDS [14,24].
Until now, the studies into BPDDS have been carried out mostly using animal models, with only a few exceptions employing human clinical trials. As such, the preparation of novel BPDDS is especially interesting for the pharmaceutical industry and medicine in general.
Moreover, it is important to know that BPs exhibit a strong affinity to nanocrystalline hydroxyapatites [25,26]. It has been shown that they may strongly adsorb on the apatitic surface by an ion-exchange mechanism between phosphonate groups from BPs and phosphate ions from hydroxyapatite (HA) [27]. Several authors have reported that some BPs may also adsorb on HA through surface binding of phosphonate groups and Ca 2+ sites of HA [25,28]. Therefore, the use of HA nanoparticles as the delivery system of BPs has been widely studied [29,30].
The main aim of our study has been to prepare composite CLO DDS using biodegradable PUs hydroxyapatite (HA) as components. The structures and chemical compositions of the new biomaterials were investigated and discussed based on the results obtained.

Synthesis of Polyols and Polyurethanes
The aim of the first part of this study was to obtain poly(ε-caprolactone) diols (PCL diols) which could be applied as precursor of further polyurethanes (PUs) synthesis. The polymerization reactions of ε-caprolactone (CL) in the presence of diethylene glycol (DEG) and the lipase from Candida antarctica (CA) were conducted at 70 °C for 14 days. The molar ratio of CL/DEG was either 20:1 (PCL-1), 30:1 (PCL-2) or 40:1 (PCL-3). Reactions were carried out with one level of lipase concentration at the same scale of monomer (100 mg of CA).
The structure of the obtained PCL diols was confirmed by proton nuclear magnetic resonance ( 1 H NMR) or carbon-13 nuclear magnetic resonance ( 13 C NMR), Fourier transform infrared spectroscopy (FTIR) and matrix-assisted laser desorption/ionization mass spectrometry (MALDI-TOF MS) (Experimental Section).
In the MALDI-TOF MS spectra of the PCL diols, linear macromolecules were observed ( Figure 1). The first and most prominent series of peaks was assigned to polyols terminated with a hydroxyl group (residual mass: 15 Da, Na + adduct). This series of peaks was differing by 114 Da, which is equal to the mass of the repeating unit of PCL. In addition, low-intensity series of peaks corresponding to macromolecules (residual mass: 31 Da, K + adduct) was detected in the mass spectrum. The average molecular mass (Mn) values of PCL diols determined by the MALDI-TOF MS method ranged from 1500 to 2900 Da (polydispersity indexes (PD) 1.36-1.89). The weight method was used to determine the reaction yield. For the PCL-1, PCL-2 and PCL-3, the corresponding yield values were 84%, 73% and 69%, respectively.
The chemical structure of the PUs was confirmed by 1 H, 13 C NMR and FTIR (Figures 2 and 3). The data are shown in the Experimental Section. Table 1 shows the mechanical properties of the obtained PUs. The fail stress (FS), stress at 100% elongation (S100), Shore hardness (ShH) and stress at 100% elongation at break (ε) of the obtained PUs were determined. As is presented in Table 1, the ε was greater than 300%. The obtained materials showed FS within the range of 13.9-14.7 MPa and ShH within the range of 41-44 Shore A degrees.

The Polyurethane/Hydroxyapatite Composites' Fabrication
The polyurethane/hydroxyapatite composites (PU-HA composites) were obtained from previously synthesized PU-1, PU-2 and PU-3 ( Table 2). The ratio of PU to HA was 9:1 or 8:2 (w/w) ( Figure 4). The composites were formed by mixing the mixture of PU and nanocrystalline HA with NaCl. The size of the HA crystals used varied from 15 to 40 nm ( Figure 5). Table 2. Characterization of polyurethane/hydroxyapatite composites.
A typical scanning electron microscope (SEM) micrograph of the PU-HA composite shows the continuous structure of interconnected and somewhat regular pores ( Figure 6). The regular pores range from several microns up to a dozen or so microns, which is within the appropriate range for tissue engineering.

Cytotoxic Tests
Cytotoxic tests of the obtained PUs were carried using the luminescent bacteria V. fischeri and two ciliated protozoa S. ambiguum and T. termophila (Table 3). All the tested samples were not toxic to any of the tested bionts, whether bacteria or protozoa, due to the fact that a sample is considered toxic when the percentage of the toxicity effect (PE) is higher than 20. Table 3. Cytotoxicity of the obtained polyurethanes.

Drug Release from the Polyurethane/Hydroxyapatite Composites
CLO was incorporated into the PU-HA composites by immersing the material in an aqueous drug solution of a known concentration. The drug content in the PU-HA composites was 1% wt.
In vitro CLO release from the PU-HA composites was conducted in a phosphate buffer solution (PBS) buffer at 37 °C for 1-8 weeks. The kinetic rates of CLO released from the obtained biomaterials at pH 7.4 are shown in Figures 7 and 8.
Two factors could influence the release of CLO from the obtained PU-HA composites, namely the Mn of the PCL diols used in PU synthesis and the P of the composites.   It was found that the rate of CLO release increases with increasing the P and decreasing the Mn of PCL used in PU synthesis. The P of the PU-HA composites decreases with increasing HA content.
It was already known that the hydrolytic stability of PU-HA composites and CLO release from matrices depend on numerous factors, such as composition, kind of hard or soft segments, the crystallinity, and the size and form of the crystallite of the PU, etc. However, it seems that in our study the main influence on this property has a kind of polyols type soft segments.
The degradation tests of the obtained PUs or PU-HA composites and the kinetic rates of the CLO released were conducted in the same manner and under the same conditions.
The results directly comparing CLO release with the Mv of the PUs (Table 4) or the mass loss (WL) of the PU-HA matrices studies follow the same trend (Figures 9 and 10).
In vitro degradation of the synthesized PUs was controlled by the change of the mechanical properties and the Mv. The Mv of the obtained PUs were determined after four and eight weeks of degradation ( Table 4). The changes in the Mv for the obtained PUs were around 4.7%-6.0% after four weeks and 7.2%-11.4% after eight weeks. PU-1 degraded faster in comparison to PU-2 and PU-3. The above parameters are in a good agreement with the results of the kinetic rates of CLO release from the obtained PU-HA composites.    Furthermore, the kinetic rates of CLO release are in agreement with the change in the mechanical properties of the PUs and the in vitro degradation of the produced PU-HA composites. After eight weeks' degradation process, the PU-1 obtained from PCL-1 retained around 76% of the original value of FS, 84% of the value of ShH and 81% of the value of ε (Tables 1 and 5). The changes of these parameters were clearly smaller for the PUs obtained from PCL-2 and PCL-3, which confirms earlier reports of the higher hydrolytic stability of PUs containing longer polyester units. PU-3 retained around 88% of the original value of FS, 93% of the value of ShH and 91% of the value of ε (Table 5). In vitro degradation of the obtained PU-HA composites was controlled by the WL of the materials. The results are shown in Figures 9 and 10. The WL values of PU-HA-1 and PU-HA-2 were 22% and 12% after eight weeks of degradation, respectively. However, for PU-HA-5 and PU-HA-6 the WL was 9% and 3% (after eight weeks). The WL values of PU-HA increased slowly with increasing hydrolytic degradation time. These results correlate well with the change in the mechanical properties of the PUs.
The scanning electron microscopic images of the PU-HA composites, both in their original state and after eight weeks' degradation process, are shown in Figure 11. In comparison to the original composite (Figure 11a), the surface of the PU-HA composite after the degradation process shows severe cracking all over its surface, indicating significant oxidative/hydrolytic damage (Figure 11b).

Synthesis of Poly(ε-caprolactone) Diols
The CL, DEG and CA were weighed (under dry argon) into a cylindrical glass reactor. Before the reactions, the monomer, glycol and enzyme were dried in vacuo at room temperature for 1 h. The reaction vessel was placed into an oil bath. Polymerization of the CL (0.05 mol) in the presence of DEG (0.00125-0.0025 mol) and 100 mg CA was carried out in bulk (under dry argon at 70 °C for 14 days). After the polyreaction time was complete, the mixture was dissolved in CH2Cl2 and the insoluble enzyme was removed by filtration. Next, the obtained solution was washed with cold MeOH using vigorous stirring. The operation was repeated three times [31]. The final products (poly(ε-caprolactone) diols, PCL diols) were dried in vacuo at room temperature for 48 h.

Synthesis of Polyurethanes
The PUs were prepared following a two-step, pre-polymer synthesis method. The isocyanate index was about 1.05. First, all the reactants (HMDI, PEAD diol, PCL diol, BD) were dried in vacuo for 1 h at 60 °C. The reactor was vacuumed and then purged with argon. The polyols and catalyst (DABCO) were first mixed at 80-100 °C in a three-necked flask equipped with a stirrer and thermometer. Next, HMDI was added to the reaction mixture and all the components were mixed vigorously for about 5 min. The temperature of the reactor was reduced to 70-80 °C. A chain extender (BD) was then slowly added to the reaction mixture. The reaction was kept at 70-80 °C for 3 h. Next, the product was conditioned in vacuo at 50 °C for 24 h. The synthesized PUs were dissolved in DMSO and precipitated into distilled water. Next, precipitated PUs were dried in vacuo at 40-50 °C for one week.

Cytotoxicity Assays
The luminescent bacteria V. fischeri and two ciliated protozoans S. ambiguum and T. termophila were used to evaluate the cytotoxicity of the obtained PCL diols and PUs. The cytotoxicity tests were carried out according to procedures described in our earlier papers [32,33].

Composite Production
The previously synthesized PUs were first dissolved in DMSO at a concentration of 10%-20% (w/v). Next, the PUs solution were mixed with HA. Pores were created by mixing the mixture of PUs and HA with 0.5 g of NaCl crystals per 1.5 g of PU. The PU/salt mixtures were poured into a mould. Next, the mould were dried in vacuo at 40-50 °C for 24-48 h. The samples were washed for 24 h in distilled water to remove NaCl. The composite samples were later dried in vacuo at room temperature for about one week.

Clodronate Impregnation of the Polyurethane Composites
CLO was incorporated into the PU composites by immersing the material into an aqueous drug solution of known concentration. The solution was pulled into the pores of the biomaterials by repeated five-cycles of vacuum/argon. PU composites were dried in vacuo at room temperature until the weight of the impregnated materials remained unchanged. The gain in weight of the PU composites following impregnation was taken as the weight of the CLO incorporated into the biomaterials.

Clodronate Release from the Composites
The composite BPDDS were incubated in a phosphate buffer solution (PBS) (pH 7.4) at a ratio of 15 mg of composite to 1 mL of buffer at 37 °C. The mixture was stirred under constant agitation (50 cycles/min) and a sample was removed at selected intervals followed by fresh buffer replacement. The quantity of the released CLO was determined from the calibration curve previously obtained under the same conditions and analysed by means of the high performance liquid chromatography with charged aerosol detector (HPLC CAD) method.

Degradation Test
The hydrolytic degradation of the PUs and PU composites was measured by immersion for eight weeks in a PBS at 37 °C. After a certain period, the biomaterials were completely dried in a vacuum oven at 35 °C. Three individual experiments were performed in the degradation test, and then the average value was calculated. The degree of degradation was determined from the weight loss (WL) of the polymeric samples according to the equation: WL = [(W0 − Wd)/W0] × 100 (%), where W0 is the initial weight of the polymer sample and Wd is the weight of the dry polymer sample after degradation.

Measurements
1 H and 13 C NMR spectra of the PCL diols and PUs were recorded on a Varian 300 MHz spectrometer using CDCl3 or DMSO-d6 as a solvent. Tetramethylsilane was served as the internal standard. The FTIR spectra were measured from KBr pellets (PerkinElmer spectrometer, PerkinElmer, Warsaw, Poland).
The molar mass and molar mass distributions of the PCL diols were determined using a GPC instrument (GPC Max + TDA 305, Viscotek, Malvern, UK) equipped with Jordi DVB Mixed Bed columns (one guard and two analytical, Viscotek) at 30 °C in CH2Cl2 (HPLC grade, Sigma-Aldrich, Poznań, Poland) at a flow rate of 1 mL/min with RI detection and calibration based on narrow PS standards (ReadyCal Set, Fluka). The results were processed using the OmniSEC software (ver. 4.7) (Viscotek).
The MALDI-TOF MS spectra were performed in linear mode on an ultrafleXtreme™ (Bruker Daltonics, Poznań, Poland) mass spectrometer using a nitrogen gas laser and HABA as a matrix. The polymer samples were dissolved in tetrahydrofuran (THF) (5 mg/mL) and mixed with a solution of HABA.
The surface morphologies were studied by scanning electron microscope (SEM, LEO 435VP (Zeiss, Jena, Germany)) so as to compare them with the initial morphologies.
The density and porosity values of the PU composites were measured by the liquid displacement method [36]. Ethanol (EtOH) was used as the displacement liquid. A dry composite sample was placed in a cylinder filled with a predetermined volume of EtOH (V1). Next, the cylinder was placed in vacuo for 20 min. The total volume of EtOH containing the composite sample was recorded as V2 and the residual EtOH volume was recorded as V3. Three individual measurements were performed and then the average value was calculated.
The amount of open pores in the PU composites (P) was calculated according to the following equation: where (V2 − V3) denotes the total volume of the composite sample and (V1 − V3) denotes the volume of ethanol retained in the composite sample.
The density of the composite (d) was expressed as: where W refers to the weight of the sample. The quantity of the released CLO was analysed by means of HPLC CAD using the UHPLC Dionex Ultimate 3000 analytical system with a CAD detector. Chromatographic separations were carried out using the Luna C8 column (250 × 4.6 mm, 5 μm). The calibration curve was obtained by the analysis of different concentrations of CLO in PBS solutions (0.05-2.00 mg/mL). The analytical method was validated by the Pharmaceutical Research Institute in Poland.

Conclusions
New porous composite bisphosphonate delivery systems were prepared from biodegradable polyurethanes and nanocrystalline hydroxyapatite. The obtained polymer matrices were non-toxic. The rates of clodronate release were shown to be directly dependent upon the nature of the obtained polyurethanes and the porosity of the composites. The results demonstrate that the polyurethane/hydroxyapatite composites are promising materials for the controlled release of clodronate and they can find practical applications as effective medium-or long-term implantation drug delivery systems.