Materials with Fungi-Bioinspired Surface for Efficient Binding and

Apr 8, 2014 - Tom Coenye,. ⊥. Angel Concheiro,. § and Carmen Alvarez-Lorenzo*. ,§. †. Departamento de Química de Radiaciones y Radioquímica, ...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/Biomac

Materials with Fungi-Bioinspired Surface for Efficient Binding and Fungi-Sensitive Release of Antifungal Agents Tania Segura,†,‡ Ana M. Puga,‡,§ Guillermina Burillo,*,† José Llovo,∥ Gilles Brackman,⊥ Tom Coenye,⊥ Angel Concheiro,§ and Carmen Alvarez-Lorenzo*,§ †

Departamento de Química de Radiaciones y Radioquímica, Instituto de Ciencias Nucleares, Universidad Nacional Autónoma de México, Av. Universidad 3000, Ciudad Universitaria, 04510 México, D.F. México § Departamento de Farmacia y Tecnología Farmacéutica, Facultad de Farmacia, Universidad de Santiago de Compostela, 15782-Santiago de Compostela, Spain ∥ Servicio de Microbiología y Parasitología, Complejo Hospitalario Universitario de Santiago de Compostela, 15782-Santiago de Compostela, Spain ⊥ Laboratory of Pharmaceutical Microbiology, Faculty of Pharmaceutical Sciences, Ghent University, Harelbekestraat 72, 9000 Ghent, Belgium S Supporting Information *

ABSTRACT: Materials with fungi-bioinspired surface have been designed to host ergosterol-binding polyene antibiotics and to release them via a competitive mechanism only when fungi are present in the medium. Silicone rubber (SR) surfaces were endowed with selective loading and fungi-triggered release of polyene antifungal agents by means of a two-step functionalization that involved the grafting of glycidyl methacrylate (GMA) via a γ-ray preirradiation method (9− 21.3% wt grafting) and the subsequent immobilization of ergosterol (3.9−116.8 mg/g) to the epoxy groups of polyGMA. The functionalized materials were characterized using FTIR and Raman spectroscopy, thermogravimetric analysis (TGA), and fluorescence, scanning electron microscopy (SEM), and atomic force microscopy (AFM) image analyses. Specific interactions between natamycin or nystatin and ergosterol endowed SR with ability to take up these polyene drugs, while immobilization of ergosterol did not modify the loading of antifungal drugs that did not interact in vivo with ergosterol (e.g., miconazole). In a buffer medium, polyene-loaded ergosterol-immobilized slabs efficiently retained the drug (99.5% titration), 1,2-dioleyl-sn-glycerol3-phosphocoline (DOPC), cholesterol, dioxane, and boron trifluoride etherate (BF3-Et2O) were from Sigma-Aldrich (St. Louis MO, U.S.A.). Natamycin USP was from AK Scientific (Union City, CA, U.S.A.), nystatin from Alfa Aesar (Karlsruhe, Germany), miconazole nitrate from Fagron Iberica (Terrasa, Spain), and sodium chloride from Panreac (Barcelona, Spain). GMA was vacuum-distilled before use. Analytical grade isopropyl alcohol, methanol, dichloromethane, toluene, and tetrahydrofuran (THF, dried with sodium metallic) were from Baker (Mexico). Other reagents were analytical grade. Silicone Rubber Modification. GMA Grafting. SR slabs (1 × 5 cm) were irradiated with 60Co γ-rays (GammaBeam 651 PT from Nordion Co., Ottawa, Canada) in air, with a dose of 20 kGy, at a dose rate of 13.5 kGy/h. The preirradiated SR slabs were placed into glass ampules containing different concentrations of GMA in methanol solution (8 mL), deoxygenated in a vacuum line by repeated freeze− thaw cycles for 1 h, and then sealed and placed in a water bath at 60 °C for different time periods. The grafted slabs were washed with dichloromethane (8 h) at room temperature and then heated to reflux for 24 h in order to remove ungrafted homopolymer and residual monomers. Finally, the grafted slabs (SR-g-GMA) were dried under vacuum and then placed in an oven at 60 °C until constant weight. The grafting degree was calculated according to the following equation:

%grafting =

W − Wo × 100 Wo

(1)

where Wo and W represent the weight of SR slab before and after grafting, respectively. Immobilization of Ergosterol. SR-g-GMA slabs (10−20% grafting degree) were swollen in dioxane or dry THF for 2 h in a flask equipped with a reflux condenser and a magnetic stirrer. After that, an amount equimolar of ergosterol (relative to the content of GMA grafted in each slab) was added. The reaction flask was evacuated with argon and then BF3-Et2O (5% mol) was added via a syringe. The reaction system was stirred under reflux for about 24h. Immediately afterward, the slabs were removed from the reaction medium, washed with toluene several times under continuous shaking, and then heated to reflux with toluene for 4 h to remove unreacted ergosterol. The slabs were dried under vacuum and finally in an oven at 60 °C until constant weight. The amount of ergosterol covalently immobilized to the SR-g-GMA slab was calculated as

YE =

We − Wg WeMe

× 106

Mt = k 0t M∞

(3)

Mt = kH t M∞ and Peppas equation (eq 5): Mt = k × tn M∞

(2)

(4) 33

(5)

To prepare the liposomes, DOPC (63 mg) and ergosterol (7.9 mg) or cholesterol (7.9 mg) were dissolved in 3 mL of chloroform in 20 mL vials.34 Solvent was evaporated under nitrogen flow and dried overnight under vacuum. The lipid film was hydrated with previously heated (50 °C) 10 mL of HEPES 10 mM−NaCl 100 mM pH 7 and vortexed until all lipids were removed from the vials walls. Then, liposomes dispersions (8 mM DOPC and 2 mM ergosterol or cholesterol) were frozen using liquid nitrogen and defrosted at 50 °C; the freeze−thaw cycles were repeated 8 times. Finally, the liposomes were extruded 12 times through a polycarbonate membrane filter with a pore size of 0.2 μm (Avanti Polar Lipids Inc., Alabaster AL, U.S.A.).

where Wg and We represent the mass of SR-g-GMA slab before and after immobilization of ergosterol, respectively, and Me is the molecular weight of ergosterol. Physical Characterization. FTIR spectra of SR slabs before and after GMA grafting and ergosterol immobilization were recorded using a PerkinElmer Spectrum 100 spectrometer (PerkinElmer Cetus Instruments, Norwalk, CT) with 16 scans, within the range of 4000 to 600 cm−1. Raman spectra were taken at 20, 200, and 500 μm of depth using an InVia Reflex Raman spectrometer (Renishaw, U.K.) employing a 250 mW laser diode at 785 nm with 100% of laser power, 1862

dx.doi.org/10.1021/bm500257s | Biomacromolecules 2014, 15, 1860−1870

Biomacromolecules

Article

Liposomes colloidal dispersions were kept at 4 °C until use. Liposomes size was measured at 25 °C by means of an ALV-5000F (ALV-GmbH, Germany) dynamic light scattering (DLS) instrument equipped with vertically polarized incident light of wavelength 488 nm supplied by a CW diode-pumped Nd:YAG solid-state laser (Coherent Inc., Santa Clara, CA) operated at 2 W. The intensity scale was calibrated against scattering from toluene. Measurements were made at the scattering angle of 90° to the incident beam. Microbiological Tests on Agar Plate and in Suspension. Natamycin- and nystatin-loaded slabs were removed from the loading solutions, rinsed with water, dried and cut as discs of 5 mm in diameter. Nonloaded pristine SR and SR-g-GMA with and without ergosterol were used as negative controls. C. albicans (ATCC 90028) and A. fumigatus (ATCC KM 8001; Culti-Loop, Thermo Scientific) were inoculated onto Mueller Hinton agar (BBL, Beckton, Dickinson and Company, Le Pont de Claix, France). The discs were placed in direct contact with fungi and incubated for 48 h at 35 °C. Digital pictures of discs were taken using a Leica CLS-150 stereomicroscope connected to a Nikon digital camera. In parallel, discs were placed in cation-adjusted Mueller Hinton broth in which C. albicans was inoculated and incubated for 48 h at 35 °C. Then, discs were removed and directly stained using 1:1 mixture of a 0.1% aqueous solution of Tinopal UNPA-G (Calcofluor white M2R, Sigma-Aldrich, Madrid, Spain) and an aqueous solution of KOH (10%) and glycerin (10%). Slab pieces were also examined directly using an epi-fluorescence microscope at 200× and 400× magnification (Eclipse 1000, Nikon Izasa S.A., Madrid, Spain; 100 W mercury vapor lamp) using indistinctly a 340−380 nm excitation filter and a LP-420 barrier filter from the Nikon UV-2A filter set or a 400−440 nm exciting filter and a LP-470 barrier filter from Nikon BV-2A filter set. Biofilm Assay. Candida albicans SC5314 strain was grown overnight in Sabouraud dextrose broth (SDB). The cells were harvested and washed three times with 5 mL of 0.9% (w/v) NaCl (PS) and the pellet was resuspended in 5 mL of 0.9% (w/v) NaCl. 200 μL of this suspension was added to 50 mL yeast nitrogen base (YNB). A total of 1 mL of this suspension was added to the wells of a 24-well microtiter plate containing the 1 cm2 slabs and this plate was placed for 1 h at 37 °C. After this 1 h adhesion step, the inoculum was removed and the slabs were washed three times with 1 mL PS. After this washing step, the slabs were placed in the wells of a sterile 24-well microtiter plate containing 1 mL of a 0.2× YNB medium. Biofilms were allowed to develop on the slabs for 24 h at 37 °C, after which the medium was removed and slabs were washed with PS to remove nonadhered cells. To enumerate culturable cells in the biofilms, each slab was transferred to 10 mL of PS and sessile cells were removed from the slabs by three cycles of 30 s sonication (Branson 3510, 42 kHz, 100 W, Branson Ultrasonics Corp., Danbury, CT, U.S.A.) and 30 s vortex mixing. This procedure ensured that all cells were removed while maintained the levels of cultivability. Serial 10-fold dilutions of the resulting cell suspensions were plated on SDA and plates were incubated for 24 h at 37 °C, after which colonies were counted. The biofilm experiments were performed on at least three slabs of each material. The normal distribution of the data was checked using the Shapiro−Wilk test. Statistical Analysis. Linear regression and statistical analysis were carried out using SPSS software, version 19.0 (SPSS, Chicago, IL, U.S.A.) and Statgraphics Centurion XVI (Statpoint Technologies Inc., Warranton, VA, U.S.A.). Data from drug release studies and biofilm inhibition tests were analyzed using a one-way ANOVA Dunnet statistical analysis.

peroxides decomposition and trigger the grafting of polyGMA brushes (Figure 1). Once SR-g-GMA was obtained, the epoxy groups of polyGMA reacted with ergosterol by means of a nucleophilic substitution reaction. The reaction was catalyzed with BF3-Et2O, which coordinates to the oxygen of the epoxy group and allows the epoxy ring-opening by ergosterol. In this reaction, the solvent plays an important role swelling the grafting slab and promoting the diffusion of ergosterol within the grafted layer. Two solvents were used, namely dioxane and THF. Important differences were observed in terms of ergosterol immobilization efficiency, being THF the solvent that rendered higher immobilization yield and enabled the obtaining of a direct correlation between the GMA groups available and the amount of ergosterol immobilized (Figure 2;

Figure 2. Dependence of the amount of ergosterol immobilized onto SR-g-GMA on the GMA grafting percentage, when dioxane (solid symbols) or THF (open symbols) were used as solvents.

slope 6.02 ± 0.91; α < 0.001). Roughly, the GMA/ergosterol molar ratio was in the 4−6 range, which indicates that not all GMA groups reacted with ergosterol probably because steric hindrance due to the larger size of ergosterol compared to the distance between two adjacent GMA monomers. The contents in GMA and ergosterol of functionalized slabs are shown in Table 1. The experimental setup for preirradiation of the slabs limited their size to 1 × 5 cm pieces. More than 40 slabs were irradiated; some slabs were subjected to the same processing conditions to have replicates, and some other slabs were exposed to different processing conditions in order to cover a wide range of GMA grafting and consequently to elucidate the minimum and maximum levels of functionalization with ergosterol that can be attained. Even in the case of the replicates, each slab has its own code (as reported in Table 1) for a precise description of the GMA grafting percentage and the ergosterol content. Grafting of GMA was confirmed by FT-IR analysis (Figure 3). The typical bands of SR appeared at 3000 cm−1 related to Si−OH, at 1259 cm−1 due to deformation vibration of -Si-CH3 bond, and at 994 cm−1 attributed to the stretching vibration of Si−O bonds in the Si−O−Si structure. The SR-g-GMA spectrum showed a signal at 1727 cm−1 assigned to CO stretching vibration, and a signal at 1123 cm−1 due to the −C− O−O stretching asymmetric vibration, which are both characteristic of the ester group. The signal at 907 cm−1 corresponded to the epoxy ring. After immobilization of ergosterol in the graft copolymer, new signals appeared at 2961



RESULTS AND DISCUSSION Silicone Rubber Modification and Characterization. The grafting of GMA onto SR was carried out by a γ-ray peroxidation method that involved two key steps: first SR was irradiated in air to generate alkylperoxides and hydroxyperoxides within the polymer matrix.32 Then, the preirradiated SR was immersed in the GMA monomer solution in the absence of oxygen, and then the system was heated at 60 °C to induce the 1863

dx.doi.org/10.1021/bm500257s | Biomacromolecules 2014, 15, 1860−1870

Biomacromolecules

Article

Table 1. GMA and Ergosterol Content of the SR Slabs code TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS TAS

113 125 127 131 135 143 146 152 156 158 159 160 177 178 187 188 193 195 196 197 198 201 202 203 206 208 209 210 213 217 219 251 252 253 272 276 277 285 287 288

GMA grafting (%)

ergosterol content (mg/g)

18.4 15.9 16.8 16.8 15.0 14.3 20.7 21.3 20.9 15.9 14.9 16.7 17.3 19.0 17.6 10.2 13.0 11.2 10.1 10.5 10.3 16.0 9.3 14.9 14.0 15.1 13.5 17.8 13.4 16.4 13.6 9.7 11.5 16.5 10.8 13.7 15.2 9.5 16.7 14.2

15.2 79.4 77.9 82.9 92.2 13.5 3.9 11.6 0 116.8 110.4 104.3 85.4 9.5 20.4 45.2 1.6 5.4 3.1 44.4 14.0 61.0 11.2 74.7 73.5 107.4 70.3 86.3 67.5 100.5 69.5 53.4 15.75 100.8 0 0 0 0 0 0

Figure 3. IR spectra of (a) SR, (b) SR-g-GMA, (c) SR-g-GMA with ergosterol immobilized, and (d) ergosterol.

ergosterol, polyGMA, and SR matrix degradation, respectively. Regarding the texture and appearance of the slabs, grafting of GMA at low proportions led to white nuclei homogeneously dispersed along the SR (Figure 4A). The nuclei were more abundant but smaller as the GMA grafting percentage increased. Immobilization of increasing amounts of ergosterol was visualized as a progressive rise in the fluorescence of the slabs observed under 366 nm light (Figure 4B) due to the presence of dehydroergosterol that absorbs in the UV range.24 The homogeneous fluorescence of the SR surfaces indicates the even distribution of ergosterol. SEM images of SR-g-GMA with or without immobilized ergosterol confirmed the quite homogeneous coating of the whole surface (Figure S4 in Supporting Information). Topography and roughness were measured by means of AFM at each step of the functionalization process in order to get an insight into resulting surface changes at the nanometric scale. In general, both pristine SR and SR-g-GMA before and after immobilization of ergosterol were found to present highly heterogeneous features in terms of topography with protrusions and recessions (pits) all over their surfaces (Figure 5). Pristine SR displayed smaller, worm-like structures in the Z-axis (from 158 to 647 nm), as compared to their modified counterparts (from 740 to 1000 nm for SR-g-GMA with 15% GMA grafting, and from 487 to 1600 nm for SR-g-GMA with 92.2 mg ergosterol/g); reflecting a random distribution of GMA chains all over the surface of the functionalized slabs (giving rise to higher protrusions randomly distributed). This fact was more clearly observed from the 2D profiles (Figure 6), extracted from representative zones of the 3D images: pristine SR showed surface protrusions not higher than 50 nm, while its modified counterparts had features as high as about 350 nm. Importantly, the surface topography was not significantly changed by the

and 2864 cm−1 due to symmetric and asymmetric vibration of the −C−H bond of the CH3 group. These two signals confirmed the successful covalent immobilization of ergosterol on the copolymer. Moreover, Raman spectra recorded at various depths allowed to elucidate to what extent the grafting occurred at the surface of the slab or inside the matrix (Figure S2 in Supporting Information). Compared to pristine SR, notable changes were observed at 20 μm depth for SR-g-GMA before and after binding of ergosterol. The changes attenuated at 200 μm depth and disappeared at 500 μm depth, which confirmed that ergosterol mainly remains at the surface of the slabs. The TGA profile of pristine SR slab showed one step due to thermal degradation of the matrix above 500 °C (Figure S3 in Supporting Information). The thermogram of the SR-g-GMA slab evidenced two degradation steps at 321 and 546 °C due to thermal decomposition of grafted polyGMA and SR, respectively. The grafted slabs modified with ergosterol showed three decomposition steps at 277, 389, and 570 °C because of 1864

dx.doi.org/10.1021/bm500257s | Biomacromolecules 2014, 15, 1860−1870

Biomacromolecules

Article

Figure 4. Digital images (from left to right) of unmodified silicone rubber, TAS197 (44.4 mg/g ergosterol), TAS209 (70.3 mg/g ergosterol), and TAS158 (116.8 mg/g ergosterol) taken under white (A) and UV light (366 nm, B). Codes as in Table 1.

Figure 5. 3D images of SR and SR-g-GMA slabs without (TAS277) and with immobilized ergosterol (TAS135) (∼15% GMA grafting in both cases) as characterized at three macroscopically separated regions of each. Beneath each image is the RMS value obtained.

subsequent immobilization of ergosterol, as demonstrated by the equivalent profiles of the SR-g-GMA curves of slabs with and without ergosterol. The root-square-roughness (RMS), which denotes the standard deviation of Z values along the studied area, was 53.4 (s.d. 34.7), 170.6 (s.d. 26.6), and 184.0 (s.d. 66.2) for pristine SR and for SR-g-GMA before and after immobilization of ergosterol, respectively. RMS values confirmed that (i) all slabs have highly heterogeneous patterns with varying surface features, that is, RMS levels, yielding mean values with broad standard deviations depending on the studied zone, and (ii) pristine SR presents a much lower RMS outcome (∼50 nm) than its modified counterparts (∼180−200 nm), which in turn appear not to display statistical differences compared one to

each other (confirming the functionalization with ergosterol as a not altering process). Natamycin, Nystatin, and Miconazole Loading. Slab pieces were loaded by soaking in aqueous solutions of the antifungal agents. Pristine and GMA-grafted SR slabs without ergosterol showed very limited ability to uptake natamycin (Figure 7). Immobilization of ergosterol onto SR-g-GMA notably improved the loading of natamycin, showing an ergosterol-content-dependent uptake. Two other antifungal agents, namely, nystatin and miconazole, were also tested in order to elucidate if differences in their pharmacological mechanisms of action affect to the loading capacity of the ergosterol-immobilized silicone slabs. Nystatin behavior was similar to that of natamycin, although the total amount loaded was higher. Thus, the greater the content in ergosterol, the 1865

dx.doi.org/10.1021/bm500257s | Biomacromolecules 2014, 15, 1860−1870

Biomacromolecules

Article

Figure 6. Representative 2D profiles of SR and SR-g-GMA slabs without and with immobilized ergosterol (∼15% GMA grafting in both cases) obtained all along the X-axis for a given Y-coordinate for a selected region.

higher the polyene amount loaded. Simple grafting of GMA (i.e., SR-g-GMA pieces) did not improve the performance of the uptake. Polyenes (i.e., natamycin, nystatin, or amphotericin B) are known to block fungal growth by binding specifically to ergosterol, which results in a broad spectrum of activity and the absence of development of resistance.22,34 This latter advantage is a distinctive feature of polyenes compared to other antifungal agents. Both natamycin (665.73 Da) and nystatin (926.1 Da) have a ring structure in which several conjugated double bonds are located opposite to a number of hydroxyl groups, and at one end there is a mycosamine group close to a carboxyl moiety, which renders the molecule amphoteric (Figure S1 in Supporting Information). The results of our loading experiments are in agreement with previous findings on that polyenes do not show affinity for hydrophobic surfaces containing no sterols.34 The double bonds in the B ring and the hydroxyl group in ergosterol seem to be the target of the hydrophobic and the hydrophilic regions of polyenes, respectively.22 The binding constant between natamycin or nystatin and ergosterol has been estimated to be in the 2.5−5.7 × 104 M−1 range with a stoichiometry of 1:1 mol ratio.34 Such relatively high affinity explains the role of ergosterol in the loading of the functionalized SR slabs. Differently from polyenes, the interaction of azoles with ergosterol is only driven by unspecific hydrophobic interactions in aqueous medium. The primary target of azoles, such as miconazole or ketoconazole, is the heme protein which cocatalyzes the demethylation of lanosterol during the synthesis of ergosterol.4 As expected, loading of miconazole was relatively low and occurred rapidly and independently on the amount of ergosterol immobilized onto the slabs (Figure 7). Natamycin, Nystatin, and Miconazole Release. Drug release was tested in HEPES buffer pH 7.0 solely or with liposomes mimicking fungal membrane (DOPC 0.5 mM and ergosterol 0.125 mM) in order to simulate an environment contaminated by fungi.24 Ergosterol liposomes represent a wellestablished model to study the interactions of antibiotics with the fungal membrane; ergosterol is inserted inside the lipidic bilayer of the liposomes in a similar way to that occurs in the yeast plasma membrane. 24,34 The ergosterol liposome population was formed by an unimodal population of vesicles of 160.3 nm hydrodynamic radius (Figure S5 in Supporting

Figure 7. Natamycin, nystatin, and miconazole loaded by pristine SR (open circle) and SR-g-GMA slabs before (open symbols) and after ergosterol immobilization (solid symbols) when soaked in natamycin (25 mg/L), nystatin (20 mg/L), and miconazole (7 mg/L) solutions, respectively. Codes as in Table 1.

Information). Release medium was replaced daily in order to avoid changes in the kinetics of the release and interferences in the UV quantification of the drug released due to the destabilization/aggregation of liposomes after interaction of the drug released with the ergosterol present in the membrane of the vesicles. It is important to note that ergosterol plays an important role in the stabilization of liposomes and that partitioning of the antifungal drug into the lipid bilayer may notably alter the permeability.24 Once placed in the HEPES buffer solely, polyene-loaded SR and SR-g-GMA slabs slowly released the unspecifically adsorbed drug (ca. 30% released at day 7), while the slabs with the immobilized ergosterol efficiently retained the drug ( SR > ergosterolimmoblized SR-g-GMA slabs (α < 0.001); no significant differences in release rates were observed among the different SR-g-GMA slabs with ergosterol immobilized. The faster release of nystatin from SR-g-GMA can be related to the fact that GMA enhances the hydrophilic character of SR surface and, as a consequence, retains worse the drug. The slowest release from ergosterol-immobilized slabs is directly related to the strong affinity of nystatin for ergosterol. By contrast, when the slabs were tested in HEPES buffer with ergosterol-based liposomes, comparison of the release rate constants (either KH or K0) did not reveal statistically significant differences among the slabs tested, which means that even the ergosterol-immobilized slabs can release nystatin at a rate similar to that recorded for nonfunctionalized SR slabs. For a given slab, the presence of ergosterol-based liposomes in the release medium markedly enhanced the release rate of nystatin (α < 0.001), but the magnitude of the release rate increase was notably larger in the case of SR-g-GMA slabs with ergosterol immobilized. The fungi-responsive release was found to be consistent for slabs prepared with a wide range of GMA graft percentage and amounts of ergosterol immobilized, probing the robustness of the approach (Table 2).

Figure 8. Natamycin and nystatin release profiles in HEPES buffer solely (A) and in the presence of liposomes of DOPC 0.5 mM and ergosterol 0.125 mM (B) or of liposomes of DOPC 0.5 mM and cholesterol 0.125 mM (C). Unfilled symbols refer to slabs without ergosterol. Codes as in Table 1.

Interestingly, in the case of miconazole (an antifungal agent that does not interact with ergosterol), similar release rates were found for all slabs in plain HEPES buffer (Table S2 in 1867

dx.doi.org/10.1021/bm500257s | Biomacromolecules 2014, 15, 1860−1870

Biomacromolecules

Article

Table 2. Nystatin Release Rate Parameters from SR, SR-g-GMA, and Ergosterol-Immobilized SR-g-GMA Slabs Obtained by Fitting to Square Root Kinetics (KH; %·h−0.5) and to Zero-Order Kinetics (K0; %·h−1)a HEPES buffer without liposomes

a

HEPES buffer with ergosterolbased liposomes

slab

KH (%·h−0.5); R2

K0 (%·h−1); R2

KH (%·h−0.5); R2

K0 (%·h−1); R2

SR TAS277 TAS198 TAS213 TAS210 TAS160

12.38; 0.948 19.89; 0.964 2.24; 0.964 2.21; 0.970 2.49; 0.961 1.39; 0.985

2.61; 4.14; 0.45; 0.45; 0.50; 0.28;

34.82; 31.56; 30.05; 32.31; 29.77; 25.18;

7.25; 6.36; 6.20; 6.65; 6.18; 5.73;

0.988 0.977 0.923 0.928 0.901 0.952

0.971 0.979 0.991 0.994 0.984 0.991

0.986 0.931 0.987 0.985 0.993 0.957

slab SR TAS276 TAS143 TAS127 TAS131 TAS208

HEPES buffer with cholesterolbased liposomes

HEPES buffer with ergosterolbased liposomes

KH (%·h−0.5); R2

K0 (%·h−1); R2

KH (%·h−0.5); R2

K0 (%·h−1); R2

18.43; 15.54; 18.03; 20.30; 17.35; 17.61;

3.76; 3.24; 3.81; 4.20; 3.63; 3.62;

31.19; 33.40; 30.62; 31.19; 33.17; 33.82;

6.52; 6.85; 6.43; 6.53; 6.99; 7.06;

0.992 0.967 0.947 0.977 0.963 0.993

0.966 0.988 0.988 0.981 0.989 0.983

0.972 0.997 0.965 0.972 0.956 0.975

0.997 0.983 0.995 0.997 0.996 0.996

The goodness of the fitting (R2) is also given. Standard deviations were in all cases below 10% (n = 3).

adequate levels of nystatin loading and a not too large affinity that could prevent the transfer of nystatin to the fungi. Nystatin-loaded slabs were also evaluated in Mueller-Hinton broth inoculated with C. albicans. Slabs without drug were colonized by a large number of fungi in the hyphal form (Figure S8 in Supporting Information). Nystatin-loaded ergosterolimmobilized slabs showed the presence of less fungi and in the form of yeasts. It is well-known that the ability of C. albicans to convert from yeast to hyphal forms is a key feature for its virulence.37,38 Thus, the functionalized slabs loaded with nystatin were able to inhibit the formation of hyphae and pseudohyphae of Candida sp. over the discs and thus to attenuate the virulence of the yeasts. To gain further insight into the capability of the nystatinloaded slabs to transfer the antifungal agent to the plasma membrane-inserted ergosterol of fungi, they were challenged with C. albicans in suspension and evaluated regarding the capability to inhibit biofilm formation (Figure 9). Although C. albicans was capable of forming a biofilm on the slabs, significant differences were observed between pristine SR and the ergosterol-immobilized SR-g-GMA slabs. Nonloaded slabs

Supporting Information), indicating that the fungi-responsiveness of ergosterol-immobilized slabs specifically affects to the antifungal agents able to strongly bind to ergosterol. To gain an insight into the role of ergosterol in the liposomes, similar release tests were carried out in HEPES containing cholesterol-based liposomes. It is well-known that liposomes by themselves can uptake a variety of drugs. Contribution of specific interactions of the antifungal agents (e.g., nystatin) with ergosterol can be demonstrated by comparison of the release profiles in ergosterol-based liposomes (fungi mimickers) medium with those recorded in cholesterolbased liposomes (mammalian cells mimickers) medium. The double bonds in the B-ring and the sp2 hybridization of C-7 in ergosterol results in a 1,3-diplanar chair conformation, particularly suitable for the interaction with the rigid structure of polyenes. By contrast, cholesterol with only one double bond adopts a half-chair conformation less suitable for polyene binding.24,34 Therefore, the competitive mechanism that drives the release from the ergosterol-immobilized slabs should be more efficient in media with ergosterol than with cholesterol. As can be observed in Figure 8, the presence of cholesterolbased liposomes did trigger much slower release (Table 2; α < 0.001) than that observed in ergosterol-based liposomes medium, confirming the fungi-responsiveness of nystatin release from ergosterol-immobilized slabs. Antifungal Tests. Several in vitro studies were carried out in order to verify the capability of the drug-loaded slabs to inhibit the growth of C. albicans and A. fumigatus, responsible for most of the invasive fungal infections. Nonloaded SR, SR-gGMA, or ergosterol-immobilized SR-g-GMA did not show antifungal activity. In fact, colonization of the discs by the fungi could be clearly observed (Figures S6 and S7 in Supporting Information). Interestingly, discs containing natamycin did not cause growth inhibition either, while those loaded with nystatin led to remarkable inhibition zones in both fungi seeded plates, which can be related to the lower MIC of nystatin (3 vs 5 mg/ L).36 SR and TAS287 were the materials with the lowest amount of nystatin loaded but they led to certain growth inhibition (especially in case of A. fumigatus), likely because the lack of specificity of the junctions between nystatin and the slabs prompted a rapid release of the whole loaded drug. Discs containing ergosterol showed different behaviors: TAS251 and TAS217 were more efficient than TAS113 and TAS203 to inhibit the growth of fungi (see Figure 7 for the amounts loaded). The greatest inhibition zone could be observed for TAS251 against both C. albicans and A. fumigatus. This slab contained an intermediate level of ergosterol immobilized (lower than for TAS203 or TAS217) and thus combined

Figure 9. Biofilm formation of C. albicans SC5314 on unmodified silicone rubber (SR nonloaded or loaded with 0.30 mg nystatin/g), SR-g-GMA (TAS 82 and TAS 287 having 11.2% and 16.7% GMA, respectively; TAS 287 contained 0.33 mg nystatin/g) and SR-g-GMA with ergosterol immobilized (TAS 178, TAS 201, TAS 135, TAS 193, TAS 196, and TAS 195 containing 0.21, 0.21, 0.26, 0.47, 0.45, and 0.46 mg nystatin/g). Significant differences *p < 0.005; **p < 0.01 compared to SR discs without nystatin (SR-nonloaded). 1868

dx.doi.org/10.1021/bm500257s | Biomacromolecules 2014, 15, 1860−1870

Biomacromolecules

Article

(CN2012/045) Spain, FEDER, and CYTED (Red iberoamericana de nuevos materiales para el diseño de sistemas avanzados de liberación de fármacos en enfermedades de alto impacto socioeconómico, RIMADEL). A.M.P. is grateful to the Spanish Ministerio de Economia y Competitividad for FPI (BES-2009024735) Grant. M. Alatorre-Meda is acknowledged for the help with AFM experiments and analysis.

did not prevent the biofilm formation. As mentioned above, SR took up a small quantity of nystatin and caused a minor decrease in the CFU/cm2. In contrast, ergosterol-immobilized SR-g-GMA slabs that were loaded with nystatin led to almost two-orders of magnitude decrease in the CFU/cm2 (Figure 9), which confirms the interest of the approach.





CONCLUSION Functionalization of materials following a bioinspired strategy that mimics the composition of fungal membrane has resulted in polyene antibiotic delivery systems with improved loading and fungi-responsive release. The reactions that lead to ergosterol immobilization on silicone rubber can be easily implemented to functionalize other polymeric substrates or matrices, resulting in a quite versatile approach to prepare delivery systems, drug-eluting medical devices or active packaging materials that release the drug in response to fungi infection or contamination. This bioinspired strategy can be considered safe regarding both the biocompatibility of ergosterol and the absence of risk of antimicrobial resistance. Optimization of GMA grafting and ergosterol immobilization may result in devices that release on demand polyene agents, which may notably improve the performance of the pharmacological treatments with polyene drugs and effectively prevent fungi biofilm formation on polymeric substrates.



(1) Cuenca-Estrella, M.; Bernal-Martinez, L.; Buitrago, M. J.; Castelli, M. V.; Gomez-Lopez, A.; Zaragoza, O.; Rodriguez-Tudela, J. L. Int. J. Antimicrob. Agents 2008, 32, S143−147. (2) Pfaller, M. A. J. Hosp. Infect. 1995, 30, 329−338. (3) Pfaller, M. A.; Diekema, D. J. Clin. Microbiol. Rev. 2007, 20, 133− 163. (4) Gannoum, M. A.; Rice, L. B. Clin. Microbiol. Rev. 1999, 12, 501− 517. (5) Soll, D. R. Acta Trop. 2002, 81, 101−110. (6) Sims, C. R.; Ostrosky-Zeichne, L.; Rex, J. H. Arch. Med. Res. 2005, 36, 660−671. (7) Douglas, L. J. Trends Microbiol. 2003, 11, 30−36. (8) Jan-Harriot, M. M.; Noverr, M. C. Trends Microbiol. 2011, 19, 557−563. (9) Mukherjee, P. K.; Chandra, J. Drug Resist. Update 2004, 7, 301− 309. (10) Fleet, G. H. Curr. Opin. Biotechnol. 2007, 18, 170−175. (11) Pawlowska, A. M.; Zannini, E.; Coffey, A.; Arendt, E. K. Adv. Food Nutr. Res. 2012, 66, 217−238. (12) Paulo, C. S. O.; Vidal, M.; Ferreira, L. S. Biomacromolecules 2010, 11, 2810. (13) Leslie, A. D. MMG 445 Basic Biotechnol. 2011, 7, 6−11. (14) Emam, W.; El-Seideek, L.; Ibrahim, N.; Eissa, H. A. J. Appl. Sci. Res. 2012, 8, 2821−2831. (15) Phillips, C. A.; Laird, K.; Allen, S. C. Food Res. Int. 2012, 47, 310−314. (16) Junqueira-Gonçalves, M. P.; Alarcón, E.; Niranjan, K. Food Control 2013, 33, 455−460. (17) Nava-Ortiz, C. A.; Burillo, G.; Concheiro, A.; Bucio, E.; Matthijs, N.; Nelis, H.; Coenye, T.; Alvarez-Lorenzo, C. Acta Biomater. 2010, 6, 1398−1404. (18) Braceras, I.; Oyarbide, J.; Azpiroz, P.; Briz, N.; Ipiñazar, E.; Alvarez, N.; Atorrasagasti, G.; Fratila, R. M.; Aizpurua, J. M. Plasma Process. Polym. 2013, 10, 328−335. (19) Alvarez-Lorenzo, C.; Bucio, E.; Burillo, G.; Concheiro, A. Expert Opin. Drug Delivery 2010, 7, 173−185. (20) Swiss, G. F.; Schwabe, S.; Moriarty, R. M. PCT Int. Appl., WO 2013036771 A1 20130314, 2013. (21) Cado, G.; Aslam, R.; Seon, L.; Garnier, T.; Fabre, R.; Parat, A.; Chassepot, A.; Voegel, J. C.; Senger, B.; Schneider, F.; Frere, Y.; Jierry, L.; Schaaf, P.; Kerdjoud, H.; Metz-Boutigue, M. H.; Boulmedais, F. Adv. Funct. Mater. 2013, 23, 4801−4809. (22) Gray, K. C.; Palacios, D. S.; Dailey, I.; Endo, M. M.; Uno, B. E.; Wilcock, B. C.; Burke, M. D. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 2234−2239. (23) Kathiravan, M. K.; Salake, A. B.; Chothe, A. S.; Dudhe, P. B.; Watode, R. P.; Mukta, M. S.; Gadhwe, S. Bioorg. Med. Chem. 2012, 20, 5678−5698. (24) Silva, L.; Coutinho, A.; Fedorov, A.; Prieto, M. Biophys. J. 2006, 90, 3625−3631. (25) Ranney, D. F. Biochem. Pharmacol. 2000, 59, 105−114. (26) Bayer, C. L.; Peppas, N. A. J. Controlled Release 2008, 132, 216− 221. (27) Siegel, R. A.; Gu, Y.; Lei, M.; Baldi, A.; Nuxoll, E. E.; Ziaie, B. J. Controlled Release 2010, 141, 303−313. (28) Yoo, J. W.; Irvine, D. J.; Discher, D. E.; Mitragotri, S. Nat. Rev. Drug Discovery 2011, 10, 521−535.

ASSOCIATED CONTENT

S Supporting Information *

Structure of the antifungal agents and of ergosterol and cholesterol; Raman spectra of pristine SR, SR-g-GMA and ergosterol-immobilized SR-g-GMA slabs; TGA of SR, polyGMA (pGMA), SR-g-GMA, and ergosterol-immobilized SR-gGMA slabs; SEM images of unmodified silicone rubber, SR-gGMA (TAS 276), and SR-g-GMA with ergosterol immobilized (TAS 206) at two different magnifications; DLS of DOPC (0.5 mM): ergosterol (0.125 mM) liposomes dispersed in HEPES buffer; zones of inhibition observed on agar plates seeded with C. albicans or A. fumigatus; epifluorescence images of C. albicans over a silicone rubber and over TAS251 disc loaded with nystatin; and natamycin and miconazole release rate parameters from SR, SR-g-GMA, and ergosterol-immobilized SR-g-GMA slabs obtained by fitting to square root kinetics (KH; %·h−0.5) and to zero-order kinetics (K0; %·h−1). This material is available free of charge via the Internet at http://pubs.acs.org.



REFERENCES

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Author Contributions ‡

These authors contributed equally (T.S. and A.M.P.).

Funding

The work described in this paper is the subject of patent application P201330893 filed by the University of Santiago de Compostela and the Universidad Nacional Autónoma de México. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS Work supported by MICINN (SAF2011-22771; INNPACTO IPT-060000-2010-14, MIPFOOD project), Xunta de Galicia 1869

dx.doi.org/10.1021/bm500257s | Biomacromolecules 2014, 15, 1860−1870

Biomacromolecules

Article

(29) Ribeiro, A.; Veiga, F.; Santos, D.; Torres-Labandeira, J. J.; Concheiro, A.; Alvarez-Lorenzo, C. Biomacromolecules 2011, 12, 701− 709. (30) Alvarez-Lorenzo, C.; Concheiro, A. Curr. Opin. Biotechnol. 2013, 24, 1167−1173. (31) Curtis, J.; Colas, A. In Biomaterials Science; Ratner, B. D., Hoffman, A. S., Schoen, F. J., Lemons, J. E., Eds.; Elsevier Academic Press: San Diego, CA, 2004; pp 697−707. (32) Segura, T.; Burillo, G. Radiat. Phys. Chem. 2013, 91, 101−107. (33) Siepmann, J.; Peppas, N. A. Int. J. Pharm. 2011, 418, 6−12. (34) te Welscher, Y. M.; ten Napel, H. H.; Balagué, M. M.; Souza, C. M.; Riezman, H.; de Kruijff, B.; Breukink, E. J. Biol. Chem. 2008, 293, 6393−6401. (35) Jacquier, N.; Schneiter, R. J. Steroid Biochem. Mol. Biol. 2012, 129, 70−78. (36) Moore, D.; Robson, G. D.; Trinci, A. P. J. 21st Century Guidebook to Fungi; Cambridge University Press: Cambridge, U.K., 2011; Chapter 18, p 517. (37) Lo, H. J.; Köhler, J. R.; Didomenico, B.; Loebenberg, D.; Cacciapuoti, A.; Fink, G. R. Cell 1997, 90, 939−949. (38) Corner, B. E.; Magee, P. T. Curr. Biol. 1997, 7, R691−R694.

1870

dx.doi.org/10.1021/bm500257s | Biomacromolecules 2014, 15, 1860−1870