Biomaterials 32 (2011) 3822e3831

Contents lists available at ScienceDirect

Biomaterials
journal homepage: www.elsevier.com/locate/biomaterials

Effect of the dopant anion in polypyrrole on nerve growth and release of a neurotrophic protein
Brianna C. Thompson a, Simon E. Moulton a, Rachael T. Richardson b, Gordon G. Wallace a, *
a b

ARC Centre of Excellence for Electromaterials Science, Intelligent Polymer Research Institute, University of Wollongong, Wollongong, NSW 2522, Australia The Bionic Ear Institute, 384 Albert Street, East Melbourne, Victoria, Australia

a r t i c l e i n f o
Article history: Received 6 December 2010 Accepted 4 January 2011 Available online 25 February 2011 Keywords: Polypyrrole Neurotrophin-3 Dopant Electrical stimulation Neural

a b s t r a c t
The dopant anion in polypyrrole plays a critical role in determining the physical and chemical properties of these conducting polymers. Here we demonstrate an additional effect on the ability to incorporate and release a neurotrophic protein e neurotrophin-3. The multi-faceted role of the dopant is critical in ensuring optimal performance of polypyrroles in their use as platforms for nerve growth. In this paper, the effect of changing the co-dopant used in electrochemical polypyrrole synthesis on the compatibility with primary auditory nerve tissue is considered and compared to some of the physical properties of the lms. Signicant differences in the controlled-release properties of the lms were also observed. The ability of the polymers to enhance nerve growth and survival in vitro with neurotrophin-3 release was also studied, which is a function of both compatibility with the neural tissue and the ability of the polymer to release sufcient neurotrophic protein to affect cell growth. A small synthetic dopant, paratoluene sulphonate, was found to perform favourably in both aspects and ultimately proved to be the most suitable material for the application at hand, which is the delivery of neurotrophins for inner-ear therapies. 2011 Elsevier Ltd. All rights reserved.

1. Introduction Sensorineural hearing loss is the most common form of hearing loss and involves the loss of sensory hair cells. In severe cases, cochlear implantation is the only option available to restore hearing. However, the loss of hair cells leads to progressive degeneration of auditory neurons (spiral ganglion neurons) and eventual apoptosis of these cells [1,2]. The loss of these spiral ganglion neurons can have detrimental affects on the functioning of the cochlear implant, due to a decrease in the integrity of the nerveeelectrode interface [3]. Strategies to prevent the degeneration of the auditory nerve are therefore of interest to improve the function of the cochlear implant. The application of neurotrophins to the cochlea as a means of promoting survival of spiral ganglion neurons has been investigated over several years. Various neurotrophins, including neurotrophin-3 (NT-3), have been applied to cultured neurons and to the cochlea in vivo (for a review of potential clinical applications of neurotrophins in inner-ear therapies, see [4]). Both NT-3 and another neurotrophin, brain-derived neurotrophic factor (BDNF), have been shown to enhance nerve survival when applied either

* Corresponding author. Tel.: 61 2 4221 3127; fax: 61 2 4221 3114. E-mail address: gwallace@uow.edu.au (G.G. Wallace). 0142-9612/$ e see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2011.01.053

immediately post-deafening [5e7], or after partial neural degeneration has occurred post-deafening [8,9]. Two major challenges remain for therapeutic use of the neurotrophins in the deafened cochlea. Firstly, a safe method is required to deliver the therapeutic proteins into the cochlea that minimises the risk of infection and does not disrupt the function of the cochlear implant, and secondly, the neurotrophins must be delivered for an extended period of time for the rescued nerves to become established. Polypyrrole (PPy) has potential to act as a controlled-release material to provide neurotrophin delivery without disrupting the cochlear implant function, as it could coat the existing platinum electrodes without increasing the impedance, adding additional surgical steps, or introducing additional devices into the ear. The use of PPy as a matrix for storage and electrically-controlled delivery of drugs has been investigated for several molecules [10e18]. The authors have previously reported the controlled delivery of nerve growth factors, in particular NT-3 [19,20] and brain-derived neurotrophic factor [21,22]. An important consideration in use for the controlled release of molecules is the dopant used during the electrochemical synthesis of PPy lms. While often the dopant is the drug or molecule of interest, some systems use a co-dopant to reduce the amount of drug used, or to improve the electrical, mechanical or other physical properties of the polymer [10,11,14,19e24].

B.C. Thompson et al. / Biomaterials 32 (2011) 3822e3831

3823

The ability to incorporate and release NT-3 and BDNF is in fact peculiar in that the active molecule to be released should have an overall positive charge (according to its isoelectric point) at the pH values of the receiving media. Despite this, our previous studies have shown that these molecules can be effectively incorporated and released upon electrical stimulation [20e23]. As the bioactive molecule is not the primary dopant, the choice of anionic dopant becomes a critical factor. Given the known effects of the dopant on physical, chemical, electronic and electromechanical properties [25], we have explored the effect on the ability to incorporate and release NT-3. The effect of the dopant used is a complex interplay of the properties induced by the dopant. It is well documented that the dopant anion in polypyrrole plays a critical role in determining the physical and chemical properties of these conducting polymers. Therefore, in the work described in this paper we try to elucidate the effect of the dopant on the incorporation and release of the neurotrophin NT-3. Six different dopants are used in the synthesis of PPy, namely sodium salts of para-toluene sulfonate (pTS), dodecylbenzene sulfonate (DBS), poly (4-styrenesulfonate) (PSS), hyaluronic acid (HA) and chondroitin sulfate (CS) and the ammonium salt of poly(2-methoxyaniline-5sulfonic acid) (PMAS). The structures of each of these dopants are shown in Fig. 1. In addition, we investigated the behaviour of spiral

ganglion neurons (SGN) cultured on the resulting conducting polymer lms, with and without incorporated NT-3. The rationale behind the use of each of these anions varied. The rst three dopants, pTS, DBS and PSS, were chosen since they are commonly used as dopant molecules for PPy and have a similar dopant end group (-SO) tethered on small (pTS) to large (PSS) 3 backbones. Previous work using PPy/pTS has been published, both on the material properties [26], and on use of the material for biological applications [27]. Some work using PPy/DBS has been presented in the literature, both for neural applications [28] and for drug-delivery applications (using a biotin-modied pyrrole residue) [29]. PPy/PSS has been studied extensively for biological use, with both non-neural [30e32] and neural [28,33e39] applications investigated. The fourth dopant, PMAS, is a sulfonated and soluble form of the conducting polymer, polyaniline. PMAS has previously been used as a dopant molecule for PPy growth, and the resulting polymer exhibited interesting electrochemical properties [40]. The use of a conducting polymer dopant in the synthesis of another conducting polymer could have interesting implications for incorporation and release of biomolecules. The biocompatibility of PPy/PMAS has also been demonstrated [41], providing sound rationale for the use of this material in further neural cell growth studies.

Fig 1. Structures of dopant molecules used in PPy synthesis in this paper.

3824

B.C. Thompson et al. / Biomaterials 32 (2011) 3822e3831 2.3. Growth of polymers A two-electrode electrochemical cell was set up in a 1 cm 1 cm plastic cuvette, with stainless steel mesh as an auxiliary electrode, and gold-coated Mylar (masked to 1 cm2 area) as the working electrode. Differently-doped polymers were grown using slightly different conditions, as summarised in Table 1. The polymers were grown using a two layer approach as reported in our previous publication [19]. An initial layer (90 s for polymers grown at 2 mA/cm2 or 12 min for 0.25 mA/cm2 current density growth) was grown with no NT-3. In all cases, where NT-3 was incorporated during polymerisation, the NT-3 concentration was 2 ppm (2 mg/mL) and was grown into the second layer of the 2-layer growth for 60 min (2 mA/cm2) or 8 h (0.25 mA/cm2). Two polymerisation current densities were used in order to avoid the over oxidation of the polymers grown using PMAS, CS and HA dopants. These polymers were grown at the lower current densities. After growth, the polymers were rinsed for 30 s in MilliQ water. Upon completion of polymer growth and prior to release experiments, the polymers were stored in MilliQ water at 4  C. For cell culture, all polymerisation solutions were lter-sterilised and sterile techniques used for synthesis, rather than using post-growth sterilisation techniques. The time that current was applied for, varied between characterisation techniques used, but generally the charge for the polymerisation was kept constant for all lms. For example, for all NT-3 release and cell culture experiments PPy/pTS, PPy/DBS and PPy/PSS, which were synthesised at 2 mA/cm2, were grown for 1 h, while PPy/PMAS, PPy/HA and PPy/CS, which were synthesised at 0.25 mA/cm2 were grown for 8 h. This means that for all polymers the total charge passed was 7.2 C/cm2, so a similar mass of polymer should have been deposited for all PPy lms.

The nal two dopants, HA and CS, are both biomolecules and components of the extracellular matrix. HA is a non-sulfated glycosaminoglycan, which is found throughout many tissues of the body. It is a polymer based on a disaccharide, made up of D-glucuronic acid and D-N-acetylglucosamine linked together via alternating b-1,4 and b-1,3 glycosidic bonds (Fig. 1). HA polymers found in vivo range from 5 to 20,000 kDa and can be over 25,000 disaccharide units in length. HA interacts primarily with one cell receptor, CD44, and changes the proliferation, differentiation, movement and protein expression of cells which express this receptor [42]. Collier et al. have previously published work using HA as a dopant for polypyrrole along with a preliminary assessment of the suitability of the material as a substrate for tissue engineering [43]. CS is a sulfated glycosaminoglycan composed of a chain of the alternating sugars N-acetylgalactosamine and glucuronic acid (Fig. 1). A chondroitin chain can have over 100 individual sugars, each of which can be sulfated in variable positions and quantities. The chondroitin sulfate used in this work was chondroitin sulfate A, meaning the chondroitin chain is sulfated at carbon 4 of the N-acetylgalactosamine sugar. CS is often found attached to proteins in proteoglycation, and is also an important part of the extracellular matrix. CS has been ascribed roles in regulation of cell behaviour due to its interactions with cell membrane bound proteins as well as proteins in the extracellular matrix. In the nervous system, it has been suggested that CS may play a part in regulating the growth, repair and development of nervous tissue. Previously we have reported the use of CS as a dopant for polypyrrole, to act as a substrate for electrical stimulation studies on differentiating skeletal myoblasts [44].

2.4. Characterisation 2.4.1. SEM Scanning electron microscopy (JSM-7500F Field Emission Scanning Electron Microscopy) was used to observe the morphology of dried PPy lms. 2.4.2. Cyclic voltammetry Cyclic voltammetry (CV) of lms prepared with no NT-3 was performed to observe the electrochemical response of the lms grown for 1 h (lms grown at 2 mA/cm2) or 8 h (lms grown at 0.25 mA/cm2). CV was performed in 1 M NaNO3 in a three-electrode cell. Gold-coated Mylar coated with Ppy was used as a working electrode, with platinum mesh and silveresilver chloride electrodes as counter and reference, respectively. The scans were started at 0V, and scanned between 800 mV and 600 mV at 10 mV/s. 2.4.3. Contact angle measurement Sessile drop contact angle measurements were made with 1 mL water at 5 positions per lm on duplicates of each polymer. For this study, polymers were grown for 1 h (regardless of current density), so that Ppy/PMAS, Ppy/HA and Ppy/CS were grown with one-eight the total charge of the other Ppy lms. 1 mL drops of water were delivered onto the surfaces of lms and the left and right advancing angles of the water drop were averaged for each measurement. The 10 measurements per polymer were averaged and the standard error calculated.

2. Methods and materials 2.1. Materials Pyrrole was acquired from Merck, distilled before use and stored from light under nitrogen at 20  C. All chemicals used as dopants were purchased from SigmaeAldrich, except for PMAS which was synthesised and puried in-house. NT-3 was acquired from Chemicon (Australia), and iodination was performed by Prosearch International (Melbourne, Australia). Deionised MilliQ water (18 MU cm1) was used to prepare all solutions. Gold coated Mylar was purchased from CPFilms Inc (USA). ECIS plates were purchased from Applied Biophysics (USA) and used after 20% NaOH treatment to remove the non-conductive coating. Poly-ornithine was purchased from SigmaeAldrich and Natural Mouse laminin was purchased from Gibco. HEPES-buffered Eagles Medium (HEM) (with antibiotics) and high glucose Dulbeccos Modied Eagles Medium (DMEM) were also purchased from Gibco.

2.5. NT-3 release 2.2. Instrumentation Electrochemical synthesis and cyclic voltammetry characterisation was performed using EDAQ potentiostat/galvanostat controlled by Chart and EChem software, respectively. The morphology of the polymers was characterised using a Leica Stereoscan 440 SEM model scanning electron microscope (SEM). All radiation measurements were performed using a Cobra 5000 g-counter with a 125I detection efciency of approximately 75%. Contact Angle Measurements on polypyrrole lms were made using a DataPhysics OCA20 goniometer utilizing SCA21 software. A JOBIN Yvon Horiba Raman Spectrometer model HR800 was used to record Raman spectra. For all NT-3 release studies I-125 labelled NT-3 was used. Immediately following growth and rinsing, the polymers were placed into 1 mL 0.9% sodium chloride solutions. For stimulated release measurements, a stainless steel mesh counter electrode was also used in a two-electrode cell, and the electrodes were connected to a cochlea implant-mimicking stimulator, as described in electrical stimulation of cultured cells section. The procedure for determining the radiolabelled NT-3 release amounts is outlined in reference [19]. Briey, release samples were gamma counted for 5 min, from which an uncorrected counts per minute (CPM) value was obtained. This value was corrected for the decay rate of 125I, and the mass of NT-3 incorporated and released from the polymer was calculated using the specic activity of the labelled NT-3.

Table 1 Growth conditions for polypyrrole synthesised with different dopants for this study. Dopant pTS DBSA PSS PMAS HA CS
a

[dopant] 0.05 M 0.05 M 0.05 M repeating unit 2 mg/mL (0.2% w/v) 2 mg/mL (0.2% w/v) 2 mg/mL (0.2% w/v)

[pyrrole] (M) 0.2 0.2 0.2 0.2 0.2 0.2

Current density (mA/cm2) 2 2 2 0.25 0.25 0.25

Molecular Weight dopanta 194.19 Da 348.48 Da y70 kDa y10 kDa 10 kDae10 MDa 5 kDa-50 kDa

g dopant/25 mL solution 0.243 g 0.436 g 0.258 g 50 mg 50 mg 50 mg

Data provided by the supplier.

B.C. Thompson et al. / Biomaterials 32 (2011) 3822e3831 2.6. Cell culture 2.6.1. Polymer preparation Polymers were prepared at least one day before rat dissection to obtain auditory nerve tissue, and all work was carried out under sterile conditions. ECIS plates were prepared immediately before polymer growth, by treatment with 20% NaOH, followed by two MilliQ water rinses. Polymerisation solution was then placed into each well, and a stainless steel mesh counter electrode was placed into the solution. Polymers were grown for either 1 h (at 2 mA/cm2) or 8 h (0.25 mA/cm2), with PPy deposited on the gold working electrode of the ECIS plate. After growth, polymers were rinsed twice with MilliQ water, before y200 mL 0.5 mg/mL poly-ornithine in borate buffer was applied, left to coat overnight at room temperature.The poly-ornithine solution was removed, and the polymers rinsed with HEM, and 0.01 mg/mL laminin (in HEM) was applied. The plates were then placed in 37  C incubator until required (2e4 h). Immediately before use, the excess laminin was removed, plates were washed twice with MilliQ water, then lled with N2-supplemented DMEM media (DMEM/N2) at 37  C. 2.6.2. Animals Time-mated pregnant Albino-Wistar rats were purchased from the University of Adelaide. Rat pups were used at 4e6 days post-natal. National Institute of Health (NIH) guidelines for the care and use of laboratory animals (NIH Publication #85-23 Rev. 1985) were observed. The Animal Research Ethics Committee of the Royal Victorian Eye and Ear Hospital approved the care and use of the animals in this study (Ethics number #03/096A). 2.6.3. Dissection and culture of primary neurons Dissection of cochleae for cell culture was based on the method described by Aletsee et al. [45], with several differences. Day 4e6 post-natal Albino-Wistar rats

3825

were anaesthetised by induced hypothermia before decapitation. The remainder of the procedure was continued using aseptic conditions. The skull was opened midsagitally and the brain removed, before the two temporal bones were dissected and placed into ice-cold sterile HEM containing antibiotics. A dissecting microscope was then used to monitor the removal of the cochleae from the surrounding bony tissue, and the stria vascularis and organ of Corti were removed. The remaining tissue was then transferred to ice-cold DMEM/N2 media. The upper basal turn of the cochleae was isolated, and the tissue containing spiral ganglion neuron (SGN) cell bodies (referred to as SGN explants) was separated from the inner modiolus and dissected into 4e5 pieces. These pieces were then placed onto the prepared PPy or tissue culture slide, and cultured for 4 days at 37  C in 5% CO2, with or without 3 days of electrical stimulation, followed by xation of tissue and immunohistochemistry. 2.6.4. Electrical stimulation of cultured explants After 24 h culture, ECIS plates were attached to a pulse generator and oscilloscope, allowing charge-balanced, biphasic current pulses to be delivered to the materials and cells. The applied waveform applied to the SGN explants in culture were pulses of 1 mA at 100 ms with 20 ms interphase open circuit potential, and 3.78 ms short circuit. This stimulation was used to mimic the cochlear implant stimulation protocol. For more details on this stimulation setup, see [20]. 2.6.5. Immunohistochemistry and uorescent microscopy After removal of culture media, the wells of the culture slides were lled with ice-cold methanol for 15 min to x the SGN explants. The methanol was removed, and 2% foetal calf serum in PBS for 1 h at room temperature was used to block nonspecic binding. A 200 kDa rabbit anti-neurolament antibody (Chemicon) was then used diluted 1:1000 in the blocking buffer to label the SGNs. After 1 h incubation and three PBS washes, the primary antibody was detected with Alexauor

Fig 2. SEM images of polypyrrole synthesised with different dopants. a) PPy/pTS, b) PPy/DBS, c) PPy/PSS, d) PPy/PMAS, e) PPy/HA and f) PPy/CS. All polymers were grown using the same amount of charge (7.2 C/cm2). The scale bar represents 40 mm.

3826

B.C. Thompson et al. / Biomaterials 32 (2011) 3822e3831

488 goat anti-rabbit IgG (Molecular Probes) diluted 1:400 in blocking buffer. After a further 1 h incubation in the secondary antibody and three more PBS washes, Vectashield anti-fade mounting medium containing DAPI (Vector Laboratories) was applied to the SGN explants and they were cover slipped and viewed by uorescence microscopy. A Zeiss Axioplan uorescent microscope was used to image explants, and blind counts of the number of neurites exiting the explant at the explant edge were taken. Explants were discounted from statistical analysis in most cases if no non-neural DAPI-stained cell bed could be seen around the explant, as it was deemed that these had moved from the site of culture during the staining procedure. For validation of this explant assay, see [20].

3. Results and discussion PPy lms containing each of the dopants could be grown at potentials of z2.2V in the 2-electrode system described by galvanostatic deposition (7.2 C/cm2) using the conditions detailed in the Experimental Section. In all cases a uniform adherent coating was obtained on the gold-coated Mylar substrate. 3.1. Polypyrrole lm characterisation Scanning Electron Microscopy (SEM) images of PPy lms were taken to observe the effect of the dopant on morphology (Fig. 2). All polymers were grown to a total charge passed of 7.2 mC/cm2, with PPy/pTS, PPy/DBS and PPy/PSS grown at 2 mA/cm2 for 60 min, and PPy/PMAS, PPy/CS and PPy/HA grown at 0.25 mA/cm2 for 8 h. The different current densities were used in order to prevent the overoxidation of PPy, as the use of the higher molecular weight dopants increased the voltage of growth to potentially over-oxidising potentials when the higher current density was used. The charge passed during growth was the same for all lms in an attempt to keep the mass of polymer deposited similar. Changing the dopant and current density had a large effect on the morphology of lms. The lms grown at the higher current density of 2 mA/cm2 (PPy/pTS, PPy/DBS and PPy/ PSS) were rougher than the lms grown at the lower current density of 0.25 mA/cm2 (PPy/PMAS, PPy/CS and PPy/HA), implying that current density, rather than dopant is the major determinant. The PPy/pTS lm showed the roughest morphology, with cauliower-like nodules approximately 10e20 mm in diameter (Fig. 2a). DBS-doped PPy lms also showed many raised nodules of approximately the same size on the surface (Fig. 2b). PSS-doped lms were relatively smooth, with only small, blister-like nodules visible with SEM imaging (Fig. 2c). Images of PPy/PMAS revealed a very smooth lm with virtually no surface features (Fig. 2d). PPy/CS exhibited small micron/sub-micron sized nodules (Fig. 2e) whilst PPy/HA (Fig. 2f) was also observed to have small raised nodules 1e5 mm in diameter. The order of roughness appears to be pTS > DBS > PSS > CS > HA > PMAS using the conditions of growth described. Our previous publication [19] demonstrated that there is no morphological difference between PPy with and without NT-3 incorporated. Electrochemical characterisation of PPy lms grown with different dopants was performed using cyclic voltammetry. A comparison of the polypyrrole oxidation and reduction peaks for PPy lms grown with different dopants reveals that in all cases a conductive, electroactive material was formed, however, the electrochemistry of the polymer is heavily inuenced by the dopant (Fig. 3). PPy/DBS showed the most well-dened electrochemistry of all of the PPy lms, exhibiting well-dened oxidation (z0.2 V) and reduction (z-0.65 V) peaks. DBS-doped lms also showed the most negativeshifted PPy redox couple. PPy/PMAS lms exhibited clearly dened oxidation peaks at z -0.2 V, however the reduction peak was not as clearly dened (z-0.4 V). In addition to these PPy peaks, PMASdoped lms also showed small redox peaks at z 0.1 to 0.4 V and z -0.1 to 0.2 V, attributable to the redox-active PMAS [46]. PPy/pTS and PPy/PSS had poorly dened but distinguishable oxidation (z0.1 V for PPy/pTS and z -0.2 V for PPy/PSS) and reduction

Fig 3. Cyclic voltammograms of PPy lms grown with different dopants. Cyclic voltammetry performed in 1 M sodium nitrate at a scan rate of 10 mV/s using a AgjAgCl reference electrode. a) shows CV scans of PPy lms grown for 1 h at 2 mA/cm2 and b) shows CVs of lms grown for 8 h at 0.25 mA/cm2. Arrows indicate the direction of potential scan.

(z-0.2 V and z -0.35 V, respectively) peaks. The most poorly-dened peaks occurred with the biomolecules HA and CS as dopants. These were also the most positive-shifted of any of the redox couples. CSdoped PPy had oxidation and reduction peaks at 0.3 V and -0.2 V, respectively, and HA-doped PPy had peaks at z 0.2 V and z -0.1 V. The current magnitude of the PPy/CS and PPy/HA CVs are lower than the other polymers with the PPy.HA exhibiting the lowest currents. The poorly dened redox peaks and lower currents suggest reduced electroactivity of the PPy/CS and PPy/HA grown polymers. The contact angles of water droplets on the PPy lms were measured to determine the relative hydrophobicity of lms grown with various dopants. Most lms (PPy/CS, PPy/pTS, PPy/PMAS and PPy/PSS) had average contact angles of between 50 and 65 (Fig. 4). PPy/DBS was more hydrophobic than other PPy lms, at 81 3 presumably as a consequence of the long hydrophobic alkyl chains of the DBS dopant molecule. PPy/HA lms were extremely hydrophilic, with water drops spreading immediately over the polymer surface upon contact. Therefore, no accurate contact angle was determined and is thus reported as <10 . It is assumed that this phenomena was due to the highly hygroscopic nature of HA. 3.2. Biocompatibility of PPy lms grown with different dopants The compatibility of the PPy surfaces (polymerised without NT3) with auditory nerve tissues was assessed by growing SGN explants on the various PPy surfaces. In this case NT-3 was added exogenously to the culture media. The PPy lms were coated with poly-ornithine and laminin prior to cell culture, as previous work has shown that without cell adhesion molecules, auditory nerve

B.C. Thompson et al. / Biomaterials 32 (2011) 3822e3831

3827

Fig 4. Average water contact angle measurements on PPy lms grown with various dopants for 1 h. * indicates that PPy/HA was too hydrophilic to determine a true advancing contact angle measurement, so the contact angle is reported as <10 . Error bars show one standard error of the mean of 10 measurements across two samples.

explants do not remain attached during immunostaining and imaging of the cells [20]. Neurite extension from explants was used as a measure of survival and growth of auditory neurons. Since the concentration of exogenous NT-3 was constant the only factors that affected the neurite extension were the survival and adhesion of the auditory nerves on the PPy surfaces. Representative images of explants growing on PPy surfaces, and the quantication of the neurite outgrowth are shown in Fig. 5. These results showed that only two PPy lms facilitated neurite extension as well as, or better than, tissue culture plastic. PPy/pTS and PPy/DBS showed robust neurite extension from SGN explants, with signicantly more neurite extension quantied, as well as longer neurites being observed (data not shown). This indicates that these materials were highly compatible with the SGN tissues cultured. It should be noted that the cultured SGN explants contain not only auditory nerves, but also broblasts and Schwann cells, which provide support to the neurons. It was noted by Aletsee et al. [45] that the neurites often grow on an underlying layer of these support cells indicating that it may be these cell types, rather than nerve cells, which are compatible with the PPy lms. Fig. 5 shows some supporting evidence for this in that the SGN explants are surrounded by non-neural cells (which are not stained green with the specic antineurolament antibody) as evidenced by the DAPI-stained nuclei. These non-neural cell beds extend as far away from the SGN explants as the neurites extend; hence it is likely that it is these cells that contact the polymer surfaces, rather than the neural cells. Explants grown on PPy/PSS, PPy/CS and PPy/HA supported less neurite extension than those grown on tissue culture plastic, with shorter neurites and non-neural cell beds observed. While it was initially expected that PPy lms doped with HA and CS would be more biocompatible than lms doped with small, synthetic molecules, it appears that the biomolecule dopants did not enhance the biocompatibility of the polymers. This was particularly surprising considering that both HA and CS are components of the extracellular matrix upon which cells normally grow. PPy/PMAS supported virtually no neurite extension in any of the experiments performed, indicating that this material was not compatible with the SGN tissues. As shown in Fig. 5d, SGN explants grown adjacent to the autouorescent PPy/ PMAS lms on the tissue culture plastic extended a non-neural cell bed and neurites. This suggests that the observed lack of neurite extension is more likely due to a surface phenomenon rather than the PPy lm releasing toxic species into the media. Interestingly, the two surfaces which most supported neurite extension from SGN explants (PPy/pTS and PPy/DBS) also appeared to be the roughest as observed by SEM (Fig. 2), while the surface

Fig 5. Comparison of primary auditory neural explants growing on cell adhesion molecule-coated PPy lms grown with different dopants. a), b) and c) show explants growing on PPy/pTS, PPy/DBS and PPy/PSS, respectively, which were all synthesised for 1 h at 2 mA/cm2 with no NT-3. d), e) and f) show explants growing on PPy/PMAS, PPy/ HA and PPy/CS, respectively. g) shows the explants growing on poly-ornithine and laminin-coated tissue culture plastic as a comparison for the explant growth. Cells are stained with an anti-neurolament antibody (green) and the nuclear stain, DAPI (blue). For all images, NT-3 was included in the cell media at 40 ng/mL. h) shows the average numbers of neurites that were extended from nerve explants on the PPy lms. The error bars show the standard error of the mean of at least 20 explants analysed.

3828

B.C. Thompson et al. / Biomaterials 32 (2011) 3822e3831

which supported no cell growth was the most homogenous and smoothest (PPy/PMAS). This suggests that the heterogeneity and morphology of the lms inuenced SGN neurite extension on the surfaces more than the composition of the surfaces presented to the explant tissues. Additionally, the surfaces were coated with cell adhesion molecules (CAMs) prior to SGN culture, therefore the differences observed in SGN response to the materials may arise from differences in adhesion of the CAMs to the different PPy lms. QCM studies were undertaken in an attempt to quantify CAM binding to the various PPy surfaces, but no meaningful data was obtained, as the changes in mass of the underlying PPy lms (due to the uptake of ions from the supporting solution) masked any mass changes due to laminin and poly-ornithine deposition. Unfortunately, this meant that no conclusions could be made as to the origins of the differences between the PPy lms. However, based on the results described, it is clear that PPy/pTS and PPy/DBS lms best supported neurite extension from SGN explants. 3.3. Incorporation and release of NT-3 The amount of NT-3 that was incorporated into PPy lms grown with different dopants varied from 45 to 180 ng/cm2 (Fig. 6). PPy/ PMAS incorporated the most NT-3 at 180 30 ng/cm2, followed by PPy/DBS at 170 20 ng/cm2. PPy/pTS incorporated 130 20 ng/cm2, and PPy/PSS was very similar at 120 20 ng/cm2. Both PPy lms grown with biomolecule dopants, namely PPy/HA and PPy/CS, incorporated a lower amount of NT-3 at 47 8 ng/cm2 and 80 10 ng/ cm2, respectively. PPy/HA had the lowest incorporation of any of the PPy lms, and it may be worth noting that HA was the only dopant used which did not include a sulfate group in its structure. Interestingly, there was no trend in NT-3 incorporation with the size or chemical properties of the dopant molecules, or the hydrophobicity of the lms. It was thought that there may be some relationship between the properties of the dopant and the incorporation of NT-3 into PPy lms, which could potentially provide some insight into the mechanism of NT-3 interaction with PPy lms. However, no obvious relationship between the known physical properties of the dopants (size, log P values, charge or structural characteristics), or the hydrophobicity of the lms and NT-3 incorporation was observed. The only tentative correlation between the characteristics of the PPy lms and mass of NT-3 incorporated is an electrochemical one. Generally, the lms with better electrochemical properties, such as pronounced redox peaks and larger current (as determined by CV studies in Fig. 3) incorporated more NT-3, while the lms that had less well-dened CVs and lower doping levels and conjugation lengths incorporated less NT-3.

Fig 7. Average cumulative release of NT-3 from PPy lms grown with dopants as indicated. a) shows the cumulative mass of NT-3 released over 7 days with electrical stimulation, while b) shows the unstimulated (passive) release of NT-3 from the lms. PPy lms were grown for 1 h (PPy/pTS, PPy/DBS and PPy/PSS, grown at 2 mA/cm2) or 8 h (PPy/PMAS, PPy/HA and PPy/CS, grown at 0.25 mA/cm2). Each point represents the average of 3 measurements (error bars omitted for clarity). The average standard deviation for stimulated data was 5 ng e maximum was 16 ng, and the average standard deviation for unstimulated data was 1.4 ng e maximum was 3.1 ng).

The properties that are desirable for electrically-controlled release of NT-3 include a linear release prole over the longest possible duration, good control between stimulated and unstimulated release, and the ability to release NT-3 at a high rate (with a high rate of release meaning that rate can be controlled with lower stimulation for long term drug delivery applications). The electrically-stimulated NT-3 release proles of PPy lms are shown in Fig. 7a, and the unstimulated release proles are shown in Fig. 7b. The R2 values for the release proles (giving a measure of linearity of release) and ratio of stimulated:unstimulated release are summarised in Table 2. Using the above criteria, PPy/pTS and PPy/PMAS were the materials that performed the best for the release of NT-3. PPy/

Table 2 Summary of rates and R2 values for the release of NT-3 from PPy lms grown with different dopants. The ratio of stim:unstim is calculated from the ratio of the rates of release over the 7 day period of the study. PPy lm PPy/pTS PPy/DBS PPy/PSS PPy/PMAS PPy/HA PPy/CS Stimulated rate (ng/cm2/day) 8.0 0.63 3.1 3.6 1.1 0.87 R2 0.94 0.87 0.95 0.96 0.98 0.97 Unstimulated rate (ng/cm2/day) 0.77 0.67 0.55 0.32 0.29 0.21 R2 0.77 0.89 0.87 0.94 0.89 0.74 Ratio stim:unstim 10.3 0.9 5.7 11.2 3.8 4.1

Fig 6. Average mass of NT-3 incorporated by PPy lms grown for 1 h (PPy/pTS, PPy/ DBS and PPy/PSS, grown at 2 mA/cm2) or 8 h (PPy/PMAS. PPy/HA and PPy/CS, grown at 0.25 mA/cm2). Mass of NT-3 was determined by use of I-125-labelled NT-3. Error bars show the standard error of the mean of at least 8 samples.

B.C. Thompson et al. / Biomaterials 32 (2011) 3822e3831

3829

Fig 8. Response of auditory nerve tissue to electrically-controlled NT-3 release from PPy lms grown with different dopants. All polymers were grown with 2 mg/mL NT-3 in the polymerisation solution, and 7.2 C of charge were passed in the polymerisation of all lms. a)-g) show representative images of explants growing on various PPy lms after electricallystimulated release of NT-3 from the PPy. a) Explants on PPy/pTS, b) PPy/DBS, c) PPy/PSS, d) PPy/PMAS e) PPy/HA and f) PPy/CS. g) the average number of neurites extending from explants grown on electrically-stimulated and unstimulated lms over six experiments. The error bars show one standard error of the mean of 6e60 explants on each polymer.

3830

B.C. Thompson et al. / Biomaterials 32 (2011) 3822e3831

pTS had a linear release prole (R2 0.94) with electrical stimulation over the 7 days of release and released NT-3 at a reasonably high rate (8.0 ng NT-3/cm2/day). PPy/PMAS had a linear release prole (R2 0.96) with electrical stimulation over the 7 days of release and released NT-3 at a reasonably high rate (3.6 ng NT-3/cm2/day).The ratio of the rate of stimulated release to the rate of unstimulated release over the 7 days was 10.3 and 11.2 for PPy/pTS and PPy/PMAS respectively, indicating that the rate of release was increased more than 10-fold with electrical stimulation. Most PPy lms showed enhanced release of NT-3 with electrical stimulation. PPy/DBS was determined to be a less suitable material for controlled release of NT-3, due to less-linear release prole (R2 values of <0.9), but also due to the fact that electrical stimulation did not enhance NT-3 release, with the ratio of stimulated:unstimulated release rates being 0.9, due to slow NT-3 release with electrical stimulation (0.63 ng/cm2/day). All other PPy lms showed promise as controlled release materials with stimulated:unstimulated ratios ranging from 3.8 for PPy/HA to 11.2 for PPy/PMAS. However, the low stimulated release rates for PPy/HA and PPy/CS (1.1 ng/cm2/day and 0.87 ng/cm2/day, respectively) may mean that these materials cannot provide enough NT-3 to affect auditory nerve cell growth. The release of NT-3 from PPy lms onto cultured nerve cells supports this observation, as discussed below. 3.4. Effects of NT-3 release from PPy lms on cultured auditory nerves The average number of neurites per explant from SGN explants grown on various PPy lms with and without electrical stimulation is shown in Fig. 8. The only PPy material observed to signicantly enhance neurite outgrowth with electrical stimulation was PPy/pTS (p < 0.03, 1-tailed Students t-test). All other PPy lms either showed virtually no effect on neurite outgrowth between stimulated and unstimulated, or showed a slight (not statistically signicant) decrease in neurite outgrowth with stimulation. This was not an unexpected result in light of the NT-3 release kinetics of the lms (as summarised in Table 2), in which PPy/pTS and PPy/PMAS were found to have the best controlled-release characteristics, and the biocompatibility test, in which PPy/pTS and PPy/DBS were found to be the most compatible with the tissue (with PMAS-doped lms appearing to have a negative affect on the survival of auditory nerve tissue). PPy/ pTS was the only PPy lm that promoted a signicant amount of neurite extension at all, at an average of 40 neurites per explant with electrical stimulation, or 19 neurites per explant without stimulation to release NT-3. For other PPy lms no more than 10 neurites per explant were seen with or without electrical stimulation. This was not unexpected, based on the lower rates of NT-3 release from all other PPy lms (summarised in Table 2), however, it could also indicate that the PPy lm may leach some product of the polymerisation (excess dopant, or soluble oligomeric species), which may interfere with the metabolism of the cell. Toxicity data for pyrrole monomer indicates that the LD50 for subcutaneous exposure is 250 mg/mL in rabbits, but the metabolic effects on cells are unknown. Interestingly, the SGN response to PPy/pTS/NT-3 with electrical stimulation was greater than that observed in the biocompatibility experiments described above, where NT-3 was added exogenously at 40 ng/mL into the media (Fig. 5). While we have demonstrated that the NT-3 release from PPy/pTS lms over 3 days was up to 50 ng/cm2 with electrical stimulation (Fig. 7), these values were obtained with stimulation applied for 24 h a day, while the tissue culture experiments used only 8 h of stimulation per day. Additionally, the electrodes used in these SGN culture experiments were approximately 0.2 cm2 meaning that approximately 10 ng NT-3 would have been released into the 300 mL culture media giving a maximum theoretical concentration of w33 ng/mL. There are two

possible reasons for the increased number of neurites per explant observed with stimulated PPy. The proximity of the NT-3 release to the neurons may have had some effect, with release from PPy lms creating a higher local concentration around the explants. Alternatively, it has been reported in the literature that depolarisation of neurons can enhance the effects of neurotrophins [47], and it is likely that the dual effect of both electrical stimulation and neurotrophic effects have enhanced the actions of NT-3 [20]. 4. Conclusions We have shown that the dopant used during electrochemical PPy synthesis has a signicant effect on the drug-release capabilities of the resulting polymers, and a large effect on the compatibility of the materials with cultured cells. The role of the dopant in the synthesis of the polymer affects many properties of the PPy lm, and the measured physical properties, such as contact angles did not correlate in any straightforward way with the amount of NT-3 incorporated into the lms, or the kinetics of NT-3 release. Differences in controlled release from the lms may have been due to differences in surface area (caused by different roughness of lms), or differences in the electrochemical properties of the lms. A potential correlation with the electroactivity of the lms was found (with electroactivity of lms measured by cyclic voltammetry), which is unsurprising if the NT-3 release mechanism relies heavily on electrochemical processes. In terms of the compatibility of the PPy lms synthesised from various dopants, the smallest dopants (pTS and DBS) were found to produce PPy lms with the best biocompatibility with the neural tissue cultured. These lms were also found to be the roughest of the PPy lms studied. Surprisingly, biomolecule dopants did not enhance compatibility of the tissue with the PPy lms. Overall, one dopant performed best for the controlled release of NT-3 from PPy lms to enhance auditory nerve survival. The small, synthetic anion pTS was found to be the most suitable for delivery of the therapeutic protein, and most compatible with the auditory nerve. PPy/pTS was found to be the only polymer capable of signicantly enhancing auditory nerve survival through the electrically-controlled release of NT-3. Acknowledgments The authors wish to acknowledge the contributions of Rodney Millard in supply of the clinical stimulator used for the release and nerve stimulation work described in this paper. Additionally, the ongoing nancial support of the Australian Research Council is gratefully acknowledged. Biological work relating to improving cochlear implants was supported by the Royal National Institute for Deaf People. Appendix Figures with essential color discrimination. Figs. 5 and 8 in this article are difcult to interpret in black and white. The full color images can be found in the online version, at doi:10.1016/j.biomaterials.2011. 01.053. References
[1] Shepherd RK, Roberts LA, Paolini AG. Long-term sensorineural hearing loss induces functional changes in the rat auditory nerve. Europ J Neurosci 2004;20:3131e40. [2] Terayama Y, Kaneko Y, Kawamoto K, Sakai N. Ultrastructural changes of nerve elements following disruption of Organ of Corti. 1. Nerve elements in Organ of Corti. Acta Oto-Laryngologica 1977;83:291e302. [3] Shepherd RK, Hardie NA. Deafness-induced changes in the auditory pathway: implications for cochlear implants. Audiol Neuro-Otology 2001;6:305e18.

B.C. Thompson et al. / Biomaterials 32 (2011) 3822e3831 [4] Pettingill LN, Richardson RT, Wise AK, OLeary SJ, Shepherd RK. Neurotrophic factors and neural prostheses: potential clinical applications based upon ndings in the auditory system. IEEE Trans Biomed Eng 2007;54:1138e48. [5] Ernfors P, Duan ML, Elshamy WM, Canlon B. Protection of auditory neurons from aminoglycoside toxicity by neurotrophin-3. Nat Med 1996;2:463e7. [6] Miller J, Chi D, OKeeffe L, Kruszka P, Raphael Y, Altschuler RA. Neurotrophins can enhance spiral ganglion cell survival after inner hair cell loss. Int J Dev Neurosci 1997;15:631e43. [7] Staecker H, Kopke R, Malgrange B, Lefebvre P, Van de Water T. NT-3 and/or BDNF therapy prevents loss of auditory neurons following loss of hair cells. Neuroreport 1993;7:889e94. [8] Richardson RT, Wise AK, OLeary SJ, Hardman J. Tracing neurotrophin-3 diffusion and uptake in the guinea pig cochlea. Hearing Res 2004;198:25e35. [9] Wise AK, Richardson RT, Hardman J, Clark GM, OLeary SJ. Resprouting and survival of guinea pig cochlear neurons in response to the administration of the neurotrophins brain-derived neurotrophic factor and neurotrophin-3. J Comp Neurol 2005;487:147e65. [10] Zinger B, Miller LL. Timed release of chemicals from polypyrrole lms. J Am Chem Soc 1984;106:6861e3. [11] Li YJ, Dong SD. Electrochemically controlled release of Adenosine 5-triphosphate from polypyrrole lm. J Chem Soc Chem Comm 1992;1:827e8. [12] Pyo MH, Maeder G, Kennedy RT, Reynolds JR. Controlled-release of biological molecules from conducting polymer-modied electrodes - the potentialdependent release of adenosine 5-triphosphate from poly(pyrrole adenosine 5-triphosphate) lms. J Electroanal Chem 1994;368:329e32. [13] Pyo MH, Reynolds JR. Electrochemically stimulated adenosine 5-triphosphate (ATP) release through redox switching of conducting polypyrrole lms and bilayers. Chem Mater 1996;8:128e33. [14] Li YJ, Dong SJ. Electrochemical controlled-release of adenosine-triphosphate. Acta Chim Sinica 1995;53:163e7. [15] Kontturi K, Pentti P, Sundholm G. Polypyrrole as a model membrane for drug delivery. J Electroanal Chem 1998;453:231e8. [16] Massoumi B, Entezami A. Controlled release of sulfosalicylic acid during electrochemical switching of conducting polymer bilayers. Europ Poly J 2001;37:1015e20. [17] Massoumi B, Entezami A. Electrochemically controlled binding and release of dexamethasone from conducting polymer bilayer lms. J Bioactive Compatible Poly 2002;17:51e62. [18] Arbizzani C, Mastragostino M, Nevi L, Rambelli L. Polypyrrole: a drug-eluting membrane for coronary stents. Electrochimica Acta 2007;2:3274e9. [19] Thompson BC, Moulton SE, Ding J, Richardson RT, Cameron A, OLeary SJ, et al. Optimising the incorporation and release of a neurotrophic factor using conducting polypyrrole. J Control Release 2006;116:285e94. [20] Richardson RT, Thompson BC, Moulton SE, Newbold C, Lum MG, Cameron A, et al. The effect of polypyrrole with incorporated neurotrophin-3 on the promotion of neurite outgrowth from auditory neurons. Biomaterials 2007;8:513e23. [21] Evans AJ, Thompson BC, Wallace GG, Millard RJ, OLeary SJ, Clark GM, et al. J Biomed Mater Res 2009;91:241e50. [22] Thompson BC, Richardson RT, Moulton SE, Evans AJ, OLeary SJ, Clark GM, et al. Conducting polymers, dual neurotrophins and pulsed electrical stimulation e dramatic effects on neurite outgrowth. J Control Release 2010;141:161e7. [23] Miller LL, Zinger B, Zhou QX. Electrically controlled release of Fe(CN)6(4-) from polypyrrole. J Am Chem Soc 1987;109:2267e72. [24] Xiao YH, Che J, Li C, Sun CQ, Chua YT, Lee VS, et al. Preparation of nanotentacle polypyrrole with pseudo-molecular template for ATP incorporation. J Biomed Mater Res A 2007;80:925e31. [25] Diaz AF, Bargon J. Electrochemical synthesis of conducting polymers. In: Skotheim TA, editor. Handbook of conducting polymers. New York: Marcel Dekker; 1986. p. 81e115. [26] Masuda H, Asano DK. Preparation and properties of polypyrrole. Synth Met 2003;135:43e4.

3831

[27] Mattioli-Belmonte M, Gabbanelli F, Marcaccio M, Giantomassi F, Tarsi R, Natali D, et al. Bio-characterisation of tosylate-doped polypyrrole lms for biomedical applications. Mater Sci Eng C 2005;25:43e9. [28] George PM, Lyckman AW, LaVan DA, Hegde A, Leung Y, Avasare R, et al. Fabrication and biocompatibility of polypyrrole implants suitable for neural prosthetics. Biomaterials 2005;26:3511e9. [29] George PM, LaVan DA, Burdick JA, Chen CY, Liang E, Langer R. Electrically controlled drug delivery from biotin-doped conductive polypyrrole. Adv Mater 2006;18:577e81. [30] Wong JY, Langer R, Ingber DE. Novel applications for electrically conducting polymers: control of cell shape on polypyrrole thin lms. Proc Nat Acad Sci 1994;9:3201e4. [31] Gomez N, Lee JY, Nickels JD, Schmidt CE. Micropatterned polypyrrole: a combination of electrical and topographical characteristics for the stimulation of cells. Adv Funct Mater 2007;17:1645e53. [32] Aoki T, Tanino M, Sanui K, Ogata N, Kumakura K, Okano T, et al. Culture of mammalian cells on polypyrrole-coated ITO as a biomcompatible electrode. Synth. Met 1995;71:2229e30. [33] Schmidt CE, Shastri VR, Vacanti JP, Langer R. Stimulation of neurite outgrowth using an electrically conducting polymer. Proc Nat Acad Sci 1997;94: 8948e53. [34] Kim DH, Richardson-Burns SM, Hendricks JL, Sequera C, Martin DC. Effect of immobilized nerve growth factor on conductive polymers: electrical properties and cellular response. Adv Funct Mater 2007;17:79e86. [35] Kim DH, Abidian MR, Martin DC. Conducting polymers grown in hydrogel scaffolds coated on neural prosthetic devices. J Biomed Mater Res A 2004;71:577e85. [36] Gomez N, Schmidt CE. Nerve growth factor-immobilized polypyrrole: bioactive electrically conducting polymer for enhanced neurite extension. J Biomed Mater Res A 2007;81:135e49. [37] Kotwal A, Schmidt CE. Electrical stimulation alters protein adsorption and nerve cell interactions with electrically conducting biomaterials. Biomaterials 2001;22:1055e64. [38] Cui XT, Lee VA, Raphael Y, Wiler JA, Hetke JF, Anderson DJ, et al. Surface modication of neural recording electrodes with conducting polymer/ biomolecule blends. J Biomed Mater Res A 2001;56:261e72. [39] Cui XT, Wiler JA, Dzaman M, Altschuler RA, Martin DC. In vivo studies of polypyrrole/peptide coated neural probes. Biomaterials 2003;24:777e87. [40] Zhao HJ, Wallace GG. Polypyrrole poly(2-methoxyaniline-5-sulfonic acid) polymer composite. Polym Gels Netw 1998;6:233e45. [41] Liu XA, Gilmore KJ, Moulton SE, Wallace GG. Electrical stimulation promotes nerve cell differentiation on polypyrrole/poly (2-methoxy-5 aniline sulfonic acid) composites. J Neural Eng 2009;6:10e4. [42] Laurent TC, Fraser JRE. Hyaluronan. Faseb Journal 1992;6:2397e404. [43] Collier JH, Camp JP, Hudson TW, Schmidt CE. Synthesis and characterization of polypyrrole-hyaluronic acid composite biomaterials for tissue engineering applications. J Biomed Mater Res A 2000;50:574e84. [44] Gilmore KJ, Kita M, Han Y, Gelmi A, Higgins MJ, Moulton SE, et al. Skeletal muscle cell proliferation and differentiation on polypyrrole substrates doped with extracellular matrix components. Biomaterials 2009;30: 5292e304. [45] Aletsee C, Mullen L, Kim DH, Pak K, Brors D. The disintegrin kistrin inhibits neurite extension from spiral ganglion explants cultured on laminin. Audiol Neuro-otol 2001;6:57e65. [46] Masdarolomoor F, Innis PC, Ashraf S, Wallace GG. Purication and characterisation of poly(2-methoxyaniline-5-sulfonic acid). Synth Met 2005;153: 181e4. [47] Shepherd RK, Coco A, Epp SB, Crook JM. Chronic depolarization enhances the trophic effects of brain-derived neurotrophic factor in rescuing auditory neurons following a sensorineural hearing loss. J Comp Neurol 2005;486: 145e8.