Modulation of Conductivity of Alginate Hydrogels Containing Reduced Graphene Oxide through the Addition of Proteins

Modifying hydrogels in order to enhance their conductivity is an exciting field with applications in cardio and neuro-regenerative medicine. Therefore, we have designed hybrid alginate hydrogels containing uncoated and protein-coated reduced graphene oxide (rGO). We specifically studied the adsorption of three different proteins, BSA, elastin, and collagen, and the outcomes when these protein-coated rGO nanocomposites are embedded within the hydrogels. Our results demonstrate that BSA, elastin, and collagen are adsorbed onto the rGO surface, through a non-spontaneous phenomenon that fits Langmuir and pseudo-second-order adsorption models. Protein-coated rGOs are able to preclude further adsorption of erythropoietin, but not insulin. Collagen showed better adsorption capacity than BSA and elastin due to its hydrophobic nature, although requiring more energy. Moreover, collagen-coated rGO hybrid alginate hydrogels showed an enhancement in conductivity, showing that it could be a promising conductive scaffold for regenerative medicine.


Introduction
Hydrogels are three-dimensional scaffolds made up of highly hydrophilic polymers. Because they absorb so much water, these hydrogels swell, representing a high degree of flexibility, closer to that of natural tissue [1]. Hydrogels also represent high porosity, excellent biocompatibility, and controllable degradability [2], triggering their application in biomedicine including, applications of soft contact lenses in the correction of vision [3], developing a tissue engineering process [2,4,5], diagnostics [6], and embolizing cells [7]. Depending on the type of bonding, these hydrogels can be classified as either physical or chemical. Physical bonding, such as hydrogen bonding, and hydrophobicity result in physical gels, which are often reversible and affected by environmental factors [1]. tion [48]. Hydrothermal processing of alginate and graphene oxide in an aqueous solution yields hybrid alginate-rGO hydrogels with high porosity. In the hydrothermal process, graphene nanosheets and alginate form a porous structure as a result of auto-assembly; afterwards, the hybrid hydrogel is produced by ionically linking polymer networks of alginate [49]. Thus, rGO has become more applicable in tissue engineering, particularly for neuronal regeneration [48,50] or cardiomyocytes regeneration [51]. Here, we are using reduced graphene-based materials, as it is one of the best redox species that could be studied on an electrode. The redox peaks will give a clear indication of changes to the double layer on the electrode's surface. When rGO-protein-alginate is incorporated over an electrode the double layer changes, this can affect the double-layer capacitance and the electron transfer resistance. Therefore, by monitoring the charge transfer resistance, we could understand the charge transfer properties of the double layer. Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) are interrelated electroanalytical methods to examine the electrochemical double layer on the electrodes. In this work, we aimed to create conductive protein-rGO-alginate hydrogels using different proteins in order to study their adsorption capacity and electrochemical characteristics to identify the best composition.

Protein Adsorption
rGO powder in a 3/1 mixture of water and DMSO (v/v) to get a suspension of 4 mg/mL rGO, homogenizing by sonication for 60 min in a bath. The resulting rGO dispersion was diluted to 2.5 mg/mL with 18 MΩ cm resistivity deionize water (DI). Then, 90 µL of 200 µg/mL BSA, 200 µg/mL elastin, or 500 µg/mL collagen were mixed with 10 µL of 250 µg/mL rGO suspension for 120 min at 37 • C under agitation at 400 rpm. The samples were spun down at 15,000 rpm for 15 min to collect supernatants. The lack of adsorbed protein was determined with the BCA kit (Thermo Fisher, Massachusetts, MA, USA) in a M 200 TECAN microplate reader (TECAN Trading AG, Männedorf, Switzerland) at 562 nm. At least three samples were quantified to ensure accuracy and repeatability. The % of protein sorption was estimated by Equation (1), and the adsorption capacity qe (µg/µg) was estimated by Equation (2), where C 0 (µg/mL) and Ce (µg/mL) are the original protein concentration and the protein concentration at steadiness, respectively, S is the sample volume (mL), and m is the mass of rGO (µg). Adsorbability (%) = (C 0 − Ce) × 100/C 0 (1) Adsorption capacity qe = (C 0 − Ce) × S/m (2) Freundlich and Langmuir's adsorption isotherm models were implemented to estimate the adsorption isotherm. The Langmuir model is displayed in Equations (3) and (4), where the concentration of the adsorbed protein at steady-state is Ce (µg/mL), qe is the adsorption capacity (µg/ug), qmax (µg/µg) is the maximum quantity of protein sorbed per unit mass of rGO, K L (mL/µg) is the Langmuir factor associated with the surface affinity for the protein, C 0 is the initial protein concentration, and R L is the separation factor that specifies the Langmuir isotherm's fundamental aspects [52].
Freundlich model is described as follows (Equation (5)), where K F and m are the Freundlich constant and intensity adsorption, respectively: Log qe = log K F + 1/m·log Ce (5)

Thermodynamics of Protein Adsorption
250 µg/mL rGO was suspended either in 50 µg/mL BSA, 200 µg/mL collagen, or 50 µg/mL elastin solution. Mixtures with each protein were incubated and agitated at the following temperatures: 5, 10, 15, 25, 37, and 39 • C. After spinning down at 15,000 rpm for ten minutes, supernatants were collected to quantify the amount of non-adsorbed protein with BCA kit (Thermo Fisher) in a M 200 TECAN microplate reader (TECAN Trading AG, Männedorf, Switzerland) at 562 nm. At least three samples were quantified for each condition. The fundamental thermodynamic factors, such as entropy change (∆S • ), enthalpy change (∆H • ), and Gibbs free energy change (∆G • ), were calculated using Equations (10)-(13) [53][54][55], where R is the gas constant (8.314 J/mol K), T the absolute temperature (K), K d the equilibrium constant, qe (µg/µg) the quantity of protein adsorbed per mass unit of rGO at equilibrium, and Ce (µg/mL) the equilibrium concentration of each protein: Kd = qe/Ce (10)

Characterization of the rGO-Protein Binding
The binding of each protein to rGO surface was studied by Raman and Fourier transform infrared spectroscopy. First, 250 µg/mL rGO was suspended either in 50 µg/mL BSA, 200 µg/mL collagen, or 50 µg/mL elastin solution, and agitated at 37 • C for 2 h. The supernatants were collected after spinning down at rpm for 15 min and lyophilized in a Telstar Lyobeta 15 lyophilizer. rGO without protein incubation and proteins without rGO were also studied. At least three samples were studied for each condition.
Raman spectrum was obtained by confocal Raman imaging (Alpha 300 M, Company WITec, Ulm, Germany) with a 532 nm laser (5% laser power, a contact time of the 50 s, and four accumulations). FT-IR, using an attenuated total reflectance (ATR) technique, was performed in a FT-IR Bruker IFS 66/S Spectrometer, with 32 scans at a resolution of 4 cm −1 between the wavelength ranges of 4000-400 cm −1 . Air background was applied as a blank. At least three samples were analyzed for each condition.

EPO and Insulin Adsorption Blocking Study
100 µg/mL rGO was suspended either in DI water, 50 µg/mL BSA, 200 µg/mL collagen, or elastin and incubated for two hours at 37 • C. After being spun down at 15,000 rpm for 15 min, supernatants were removed, and protein-coated rGO was incubated with 200 µL of 200 mIU/mL recombinant EPO or 150 mIU/mL recombinant insulin for 24 h at 37 • C. Uncoated rGO was used as a reference. Next, samples were spun down by centrifuging for 15 min at 15,000 rpm and supernatants were collected. Non adsorbed EPO and insulin were quantified with Quantikine IVD EPO (R&D Systems) and Insulin Elisa (Mercodia), respectively, following manufacturer recommendation. At least three samples were analyzed for each condition.

Preparation of Alginate Hydrogels Containing rGO and Protein-Coated rGO
At room temperature, 1.87 g of high pure sodium alginate was dissolved in 1% mannitol by magnetic string at 200 rpm for 2 h; then, it was mixed and homogenized with either rGO or protein-coated rGO suspension, obtaining a final concentration of 1.5% alginate and 50 µg/mL rGO. To prepare hybrid alginate hydrogels, alginate solutions were mixed with 60 µL of 1.22 M calcium sulfate through a connector (Braun) between two Luer Lock syringes (BS Syringe). The mixed solutions were dispensed between two glass slides with 2 mm spacing, leaving them for 30 min to form hydrogel disks, 14 mm in diameter.

Conductivity of Alginate Hydrogels Containing Protein-Coated rGO
Electrochemical impedance spectra (EIS) were measured using a potentiostat (Princeton Applied research, Oak Ridge, TN, USA), with a screen-printed electrode (Dropsens, Oviedo, Spain) based on carbon, and a silver electrode as reference. Samples were immersed in 0.1 M PBS buffer at room temperature, applying frequency series from 10 −1 to 105 Hz. Cyclic Voltammetry (CV) measurements were performed in 0.1 M PBS buffer between the potential range from −0.2 to 0.5 V and at different scan speeds (100 mVs −1 ). Specific capacitance was estimated from CV curves by Equation (14) [16,56], where C (F·g −1 ) is the specific capacitance, Q the mean charge throughout the charging and discharging procedure, the potential range V (Volt), and the mass m (g) of the hydrogel disk:

Statistical Analysis
Statistical analysis was performed with GraphPad Prism 9.0 (GraphPad Inc., San Diego, CA, USA) and SPSS (version 27.00, IBM, New York, NY, USA) software. Results were presented as mean ± standard deviation. A normality test was performed, considering p < 0.05 as statistically significant values after ANOVA and Tukey's post hoc test bivariate correlation testing.

Adsorption of Proteins on rGO Surface
We began studying the adsorption capacity of rGO to several proteins usually located in FBS. Thus, we observed that rGO adsorption capacity (qe) was enhanced at low protein dose values, indicating the presence of available active groups on the rGO surface. However, no significant modifications in qe values at high initial protein doses (Co) were quantified, suggesting that no further protein loading on the rGO surface was allowed. Among the studied proteins, collagen showed the highest adsorption capacity (qe = 0.022 µg/µg), while other hydrophilic proteins, such as BSA and elastin showed a low affinity for rGO surface (qe between 0.0049-0.0067 µg/µg) (Figure 1a).

Adsorption of Proteins on rGO Surface
We began studying the adsorption capacity of rGO to several proteins usually located in FBS. Thus, we observed that rGO adsorption capacity (qe) was enhanced at low protein dose values, indicating the presence of available active groups on the rGO surface. However, no significant modifications in qe values at high initial protein doses (Co) were quantified, suggesting that no further protein loading on the rGO surface was allowed. Among the studied proteins, collagen showed the highest adsorption capacity (qe = 0.022 µg/µg), while other hydrophilic proteins, such as BSA and elastin showed a low affinity for rGO surface (qe between 0.0049-0.0067 µg/µg) (Figure 1a). In order to understand the adsorption phenomena involved on rGO surface, Langmuir and Freundlich's models were applied to the qe values recorded at a constant temperature, calculating the required parameters for the aforementioned models (Table 1 and Figure 1b). Protein adsorption phenomena on the rGO surface was better specified by the Langmuir than the Freundlich model, with R2 values between 0.968 and 0.996 for the Langmuir model and convergence between calculated qmax values and experimental qe results. Therefore, we suggest that adsorption phenomena occur on a homogeneous surface of rGO, with a specific number of adsorption sites on rGO surface binding to protein active sites and forming a monolayer [56]. In fact, Langmuir variable values (RL) < 1 would suggest advantageous adsorption onto the rGO surface for the studied proteins, with irreversibile adsorption for collagen and elastin (RL ≈ 0) [57]. Table 1. Parameters required for Langmuir and Freundlich adsorption isotherms. qe (µg/µg) is the Adsorption capacity at equilibrium, KL (mL/µg) is the Langmuir factor, R2 is the coefficient of determination for the Langmuir model, and qmax (µg/µg) is the maximum adsorption capacity of the proteins by rGO. For the Freundlich model, where KF is the Freundlich constant and m is the intensity adsorption, R2 is the coefficient of determination for the Freundlich isotherm.  In order to understand the adsorption phenomena involved on rGO surface, Langmuir and Freundlich's models were applied to the qe values recorded at a constant temperature, calculating the required parameters for the aforementioned models (Table 1 and Figure 1b). Protein adsorption phenomena on the rGO surface was better specified by the Langmuir than the Freundlich model, with R 2 values between 0.968 and 0.996 for the Langmuir model and convergence between calculated qmax values and experimental qe results. Therefore, we suggest that adsorption phenomena occur on a homogeneous surface of rGO, with a specific number of adsorption sites on rGO surface binding to protein active sites and forming a monolayer [56]. In fact, Langmuir variable values (R L ) < 1 would suggest advantageous adsorption onto the rGO surface for the studied proteins, with irreversibile adsorption for collagen and elastin (R L ≈ 0) [57]. Table 1. Parameters required for Langmuir and Freundlich adsorption isotherms. qe (µg/µg) is the Adsorption capacity at equilibrium, K L (mL/µg) is the Langmuir factor, R 2 is the coefficient of determination for the Langmuir model, and qmax (µg/µg) is the maximum adsorption capacity of the proteins by rGO. For the Freundlich model, where K f is the Freundlich constant and m is the intensity adsorption, R 2 is the coefficient of determination for the Freundlich isotherm.

Langmuir Model
Freundlich Model q e µg/µg Studying the adsorption capacity (qt) of rGO over time, we observed a quick adsorption process, completed after 20 min (Figure 2a). The intraparticle-diffusion model was implemented to attain a suitable mechanism that fits the protein adsorption on rGO surface ( Figure 2b). However, since qt vs. t 1/2 plotting showed linearity without going across zero, we think that intraparticle-diffusion is not the only process controlling protein adsorption on the rGO surface; film diffusion also contributes to the protein adsorption [40]. Calcu-lated intraparticle-diffusion parameters, such as the diffusion rate constant (KP) and the impediment layer wideness (C) ( Table 2), showed collagen with the highest intra-particle diffusion rate Kp (0.0001 µg·µg −1 min −1/2 ) and boundary layer thickness (0.0132), correlating to a strong hydrophobic attraction between collagen and the hydrophobic rGO surface. This result suggests that π-π bonding between collagen and rGO might also be responsible for this bonding [57]. However, the hydrophilicity of BSA and elastin would result in the decrease in interference forces and therefore the affinity with rGO.  Table 2. Calculated parameters from intra-particle diffusion pseudo-first-order and pseudo-second-order model. Calculated parameters from intra-particle diffusion are the intra-particle diffusion rate constant (K p ), constant for the intra-particle diffusion model (C) and coefficient of determination (R 2 ). The presented parameters for the pseudo-first-order are the adsorption capacity at equilibrium (qe), rate constant (K 1 ), and the coefficient of determination (R 2 ). For the pseudo-secondorder model, the parameters are the rate constant (K 2 ) and the adsorption capacity at different times (qt). Among adsorption kinetic mathematical models, the low R 2 values calculated in the pseudo-first-order model (0.35-0.82), with vast difference between estimated qe and experimental qt (Table 2), discarded this model for describing adsorption phenomena on the rGO surface. However, the pseudo-second-order model showed high R 2 values (0.93-0.99) and convergence between the estimated qe and the experimental qt values (Table 2); therefore, this model can be considered the best kinetic model to define the studied protein adsorption phenomena on the rGO surface [58,59]. Interestingly, pseudo-secondorder constant K 2 reduced while hydrophilicity increased with the lowest K 2 for elastin (2.16·µg −1 ·in −1 ) and with the highest for collagen (27.27 µg·µg −1 ·min −1 ). Representing qt values versus time, while a linear pseudo-second-order plot was more suitable for collagen and BSA (Figure 2c), a nonlinear pseudo-second-order plot was fit for elastin adsorption (Figure 2d) [60]. Collagen showed a ten-fold higher adsorption rate constant (K 2 ) than BSA and elastin, indicating the highest affinity for rGO than elastin or BSA and due to the strong hydrophobic-hydrophobic interactions between rGO surface and collagen [61].

Thermodynamics of Protein Adsorption onto rGO Surface
The study of adsorption capacity (q T ) with BSA and elastin on rGO surface increasing temperature revealed the exothermic character (∆H • < 0) of the adsorption (Figure 3a, Table 3). However, collagen showed endothermic adsorption when the temperature was increased (∆H • = 2.44 kJ/mol). We consider that the protein adsorption could start with an endothermic hydration step, followed by exothermic adsorption, but in collagen, there would be a hydration step caused by its hydrophobic nature that requires more energy than the other studied proteins [62]. This hypothesis would explain why an increase in the temperature would enhance the adsorption capacity of collagen, while it would increase the kinetic energy of BSA and elastin causing their elution from rGO surface. Moreover, the low ∆G • values (Table 3) indicated the physio-sorption nature of the adsorption [58], a nonspontaneous phenomenon that is a feature of positive ∆G • values. It could be attributed to the presence of an energy barrier in the migration of the studied proteins towards the rGO surface, with water forming a hydration shell around the proteins that would hinder their adsorption on rGO. Finally, the remarkable reduction in entropy during the adsorption of the studied proteins [58] suggests that molecular motion at the solid-liquid interface is more organized [63]. Among adsorption kinetic mathematical models, the low R 2 values calculated in the pseudo-first-order model (0.35-0.82), with vast difference between estimated qe and experimental qt (Table 2), discarded this model for describing adsorption phenomena on the rGO surface. However, the pseudo-second-order model showed high R 2 values (0.93-0.99) and convergence between the estimated qe and the experimental qt values (Table 2); therefore, this model can be considered the best kinetic model to define the studied protein adsorption phenomena on the rGO surface [58,59]. Interestingly, pseudo-second-order constant K2 reduced while hydrophilicity increased with the lowest K2 for elastin (2.16·µg −1 ·in −1 ) and with the highest for collagen (27.27 µg·µg −1 ·min −1 ). Representing qt values versus time, while a linear pseudo-second-order plot was more suitable for collagen and BSA (Figure 2c), a nonlinear pseudo-second-order plot was fit for elastin adsorption (Figure 2d) [60]. Collagen showed a ten-fold higher adsorption rate constant (K2) than BSA and elastin, indicating the highest affinity for rGO than elastin or BSA and due to the strong hydrophobic-hydrophobic interactions between rGO surface and collagen [61].

Thermodynamics of Protein Adsorption onto rGO Surface
The study of adsorption capacity (qT) with BSA and elastin on rGO surface increasing temperature revealed the exothermic character (∆H° < 0) of the adsorption (Figure 3a, Table 3). However, collagen showed endothermic adsorption when the temperature was increased (∆H° = 2.44 kJ/mol). We consider that the protein adsorption could start with an endothermic hydration step, followed by exothermic adsorption, but in collagen, there would be a hydration step caused by its hydrophobic nature that requires more energy than the other studied proteins [62]. This hypothesis would explain why an increase in the temperature would enhance the adsorption capacity of collagen, while it would increase the kinetic energy of BSA and elastin causing their elution from rGO surface. Moreover, the low ΔG° values (Table 3) indicated the physio-sorption nature of the adsorption [58], a non-spontaneous phenomenon that is a feature of positive ΔG° values. It could be attributed to the presence of an energy barrier in the migration of the studied proteins towards the rGO surface, with water forming a hydration shell around the proteins that would hinder their adsorption on rGO. Finally, the remarkable reduction in entropy during the adsorption of the studied proteins [58] suggests that molecular motion at the solidliquid interface is more organized [63].

Surface Chemistry of Protein Adsorbed rGO
In order to confirm the adsorption of the studied proteins onto the rGO surface, we studied rGO and protein-adsorbed rGO by Raman spectroscopy and FT-IR. In Raman spectra, G and D bands at~1595 cm −1 and~342 cm −1 in rGO Raman spectrum indicated the occurrence of defects due to the reduction process of GO [52] (Figure 4a), also detected in protein adsorbed-rGO samples (Figure 4b-d). After proteins were adsorbed on the rGO surface, the G and D bands shifted to~1588-1601 cm −1 and~1342-1351 cm −1 , respectively. The intensity ratio between those bands (ID/IG) suggested an sp 2 electron distribution in all the samples [63], being higher than those ID/IG ratios previously described in graphene [64], and slightly increased when proteins were adsorbed on rGO (Table 4). Finally, the 2D band position and their intensity ratio with G band (2D/G) increase would indicate more structural defects, most likely attributed to protein adsorption on rGO [65], confirming the adsorption of the studied proteins on the rGO surface [66].

Surface Chemistry of Protein Adsorbed rGO
In order to confirm the adsorption of the studied proteins onto the rGO surface, we studied rGO and protein-adsorbed rGO by Raman spectroscopy and FT-IR. In Raman spectra, G and D bands at ~1595 cm −1 and ~342 cm −1 in rGO Raman spectrum indicated the occurrence of defects due to the reduction process of GO [52] (Figure 4a), also detected in protein adsorbed-rGO samples (Figure 4b-d). After proteins were adsorbed on the rGO surface, the G and D bands shifted to ~1588-1601 cm −1 and ~1342-1351 cm −1 , respectively. The intensity ratio between those bands (ID/IG) suggested an sp 2 electron distribution in all the samples [63], being higher than those ID/IG ratios previously described in graphene [64], and slightly increased when proteins were adsorbed on rGO (Table 4). Finally, the 2D band position and their intensity ratio with G band (2D/G) increase would indicate more structural defects, most likely attributed to protein adsorption on rGO [65], confirming the adsorption of the studied proteins on the rGO surface [66].   Table 4. Parameters from Raman spectra: the D band is related to scattering from local defects or disorders present in carbon, and the G bands arise from the in-plane diverging stretching of the C-C bonds in the graphitic structure, and (2D) is related to the number of graphene layers. ID/IG is the intensity ratio of the D and G bands, which refers to the amount of defects existing in the graphene matter. The intensity ratio of 2D/G is related to the number of graphene layers in the matrix. Then, we confirmed the adsorption of the studied proteins on the rGO surface by FT-IR using an attenuated total reflectance (ATR) technique. Thus, we detected an absorption peak in rGO sample at 1603 cm −1 related to C=C stretching vibration, and at 1736 cm −1 and 1266 cm −1 , corresponding to carboxyl C=O and carbonyl C-O stretching vibrations, respectively [56] (Figure 5). When rGO incubated with the studied proteins was analyzed by FT-IR, an obvious shift of the C=C stretching band from 1603 cm −1 to 1590-1600 cm −1 and a shift of the C-O stretching band from 1266 cm −1 to 1227-1250 cm −1 were detected, which were ascribed to the π-π bonding between the benzene ring from proteins and the rGO surface (Table 5) [67]. Although due to the incomplete reduction in rGO, a hydroxyl signal was detected in FT-IR spectra, forming H-bonding when incubated with proteins (3692-3701 cm −1 ), these slight changes are not strong enough for stable H-bond formation. Therefore, we consider that the studied protein adsorption on the rGO surface is attributed mainly to π-π interactions between the rGO surface and the proteins [52,68].

Sample
Pharmaceutics 2021, 13, x FOR PEER REVIEW 10 of 18 Table 4. Parameters from Raman spectra: the D band is related to scattering from local defects or disorders present in carbon, and the G bands arise from the in-plane diverging stretching of the C-C bonds in the graphitic structure, and (2D) is related to the number of graphene layers. ID/IG is the intensity ratio of the D and G bands, which refers to the amount of defects existing in the graphene matter. The intensity ratio of 2D/G is related to the number of graphene layers in the matrix. Then, we confirmed the adsorption of the studied proteins on the rGO surface by FT-IR using an attenuated total reflectance (ATR) technique. Thus, we detected an absorption peak in rGO sample at 1603 cm −1 related to C=C stretching vibration, and at 1736 cm −1 and 1266 cm −1 , corresponding to carboxyl C=O and carbonyl C-O stretching vibrations, respectively [56] (Figure 5). When rGO incubated with the studied proteins was analyzed by FT-IR, an obvious shift of the C=C stretching band from 1603 cm −1 to 1590-1600 cm −1 and a shift of the C-O stretching band from 1266 cm −1 to 1227-1250 cm −1 were detected, which were ascribed to the π-π bonding between the benzene ring from proteins and the rGO surface (Table 5) [67]. Although due to the incomplete reduction in rGO, a hydroxyl signal was detected in FT-IR spectra, forming H-bonding when incubated with proteins (3692-3701 cm −1 ), these slight changes are not strong enough for stable H-bond formation. Therefore, we consider that the studied protein adsorption on the rGO surface is attributed mainly to π-π interactions between the rGO surface and the proteins [52,68].

Therapeutic Protein Adsorption on Protein Coated rGO
In order to determine if blocking the rGO surface with the studied proteins precluded further adsorption of other proteins, we studied the adsorption of two different therapeutic proteins as erythropoietin and insulin ( Figure 6) for 24 h at 37 • C. No differences were detected among the studied coating proteins when adsorption studies were performed with either erythropoietin (EPO) or insulin. However, this behavior with the therapeutic proteins was slightly different, since EPO trapping was lower than insulin, indicating a higher blocking for EPO protein by the protein coating.

Therapeutic Protein Adsorption on Protein Coated rGO
In order to determine if blocking the rGO surface with the studied proteins precluded further adsorption of other proteins, we studied the adsorption of two different therapeutic proteins as erythropoietin and insulin ( Figure 6) for 24 h at 37 °C. No differences were detected among the studied coating proteins when adsorption studies were performed with either erythropoietin (EPO) or insulin. However, this behavior with the therapeutic proteins was slightly different, since EPO trapping was lower than insulin, indicating a higher blocking for EPO protein by the protein coating. The mechanism of insulin adsorption by graphene and GO was shown to be different. The high adsorption of insulin hormone by graphene can be explained by a strong ππ interaction between the phenyl rings of insulin and the graphene surfaces [69]. On the other hand, insulin adsorption on the GO surface was attributed to electrostatic interactions and hydrogen bonds with the oxygenated functional groups. This interaction would be enough to overcome the blocking exerted by the coated studied proteins onto rGO, being unable to block the adsorption of further proteins with low molecular weight (6 kDa) and high hydrophilicity, such as insulin [16].

Conductivity Studies of Alginate Hydrogels Containing Protein Coated rGO
rGO has properties close to graphene and shows excellent electrical properties, including high electrical conductivity and high mobility [43]. During the reduction process, structural defects are formed, such as physical holes after removing the oxygen functional groups or alkyl groups. These holes can act as charge carriers and are responsible for the The mechanism of insulin adsorption by graphene and GO was shown to be different. The high adsorption of insulin hormone by graphene can be explained by a strong π-π interaction between the phenyl rings of insulin and the graphene surfaces [69]. On the other hand, insulin adsorption on the GO surface was attributed to electrostatic interactions and hydrogen bonds with the oxygenated functional groups. This interaction would be enough to overcome the blocking exerted by the coated studied proteins onto rGO, being unable to block the adsorption of further proteins with low molecular weight (6 kDa) and high hydrophilicity, such as insulin [16].

Conductivity Studies of Alginate Hydrogels Containing Protein Coated rGO
rGO has properties close to graphene and shows excellent electrical properties, including high electrical conductivity and high mobility [43]. During the reduction process, structural defects are formed, such as physical holes after removing the oxygen functional groups or alkyl groups. These holes can act as charge carriers and are responsible for the high conductivity of rGO [56]. In fact, the incorporation of rGO within scaffold matrices has improved their conductivity, such as rGO-gelatin methacryloyl (GelMA) hybrid hydrogels displaying improved electrical conductivity and mechanical properties [63]. Hybrid scaffolds with carbon nanofibers (CNFs) incorporated in alginate and gelatin hydrogels have demonstrated high electroconduction. Furthermore, this preparation method permits the elaboration of homogeneously dispersed hydrogels with integrated CNFs. The hybrid composite hydrogels including rGO were reported to display excellent electrical conductivity in the range of 4.1 × 10 −4 ± 2 × 10 −5 S/cm [70]. Therefore, we decided to study the electrochemical activity of protein-coated-rGO embedded within alginate hydrogels through the impedance spectroscopy (EIS) and cyclic voltammetry (CV) measurements. We quantified the phase angle (deg) and impedance modulus |Z| (ohms) within a range of frequencies between 10 −1 and 10 6 Hz in EIS, followed by analyzing the experimental data by Bode model (Figure 7) to determine the insulating or conducting behavior for the studied hydrogels. We determined that the addition of rGO or protein-coated-rGO to alginate matrix induced a slight decrease in hydrogel impedance at low frequencies.
Moreover, we observed that the phase angle stayed close to 90 • at low frequencies, decreasing towards zero at ultra-high frequencies in all the hydrogels studied (Figure 7a), quantifying the highest phase angle in collagen-coated rGO-alginate hydrogels. The magnitude of the impedance is inversely proportional to capacitance: ideal capacitors have lower impedance [71]. With rising frequency, the impedance of any given capacitance decreases. The frequency response of impedance is depicted in two portions on the Bode plot ( Figure 7a). The first is associated with the hydrogel and is below 10 1 Hz, whereas the second is related to the charge transfer resistance between the hybrid hydrogel and the electrolyte [72]. high conductivity of rGO [56]. In fact, the incorporation of rGO within scaffold matrices has improved their conductivity, such as rGO-gelatin methacryloyl (GelMA) hybrid hydrogels displaying improved electrical conductivity and mechanical properties [63]. Hybrid scaffolds with carbon nanofibers (CNFs) incorporated in alginate and gelatin hydrogels have demonstrated high electroconduction. Furthermore, this preparation method permits the elaboration of homogeneously dispersed hydrogels with integrated CNFs. The hybrid composite hydrogels including rGO were reported to display excellent electrical conductivity in the range of 4.1 × 10 −4 ± 2 × 10 −5 S/cm [70]. Therefore, we decided to study the electrochemical activity of protein-coated-rGO embedded within alginate hydrogels through the impedance spectroscopy (EIS) and cyclic voltammetry (CV) measurements. We quantified the phase angle (deg) and impedance modulus |Z| (ohms) within a range of frequencies between 10 −1 and 10 6 Hz in EIS, followed by analyzing the experimental data by Bode model (Figure 7) to determine the insulating or conducting behavior for the studied hydrogels. We determined that the addition of rGO or protein-coated-rGO to alginate matrix induced a slight decrease in hydrogel impedance at low frequencies.
Moreover, we observed that the phase angle stayed close to 90° at low frequencies, decreasing towards zero at ultra-high frequencies in all the hydrogels studied (Figure 7a), quantifying the highest phase angle in collagen-coated rGO-alginate hydrogels. The magnitude of the impedance is inversely proportional to capacitance: ideal capacitors have lower impedance [71]. With rising frequency, the impedance of any given capacitance decreases. The frequency response of impedance is depicted in two portions on the Bode plot ( Figure 7a). The first is associated with the hydrogel and is below 10 1 Hz, whereas the second is related to the charge transfer resistance between the hybrid hydrogel and the electrolyte [72]. The charge transfer resistance, as seen in Figure 7b, has the largest variance, where alginate hydrogels showed the highest impedance value, decreasing when rGO or protein-rGO was introduced into the hydrogels, especially with collagen that showed the lowest impedance value. At low frequencies, the charge transfer resistance can be obtained by interpolating the semi-circle to the real x-axis [73]. The current density of the rGOproteins alginate hydrogels rises as the hydrogel resistance and charge transfer resistance fall [73]. Figure 8 depicts the charge transfer resistance (Rct) derived by analyzing the EIS spectra. Because the charge transfer resistance characterizes the electron flow at the counter electrode to a great extent, the lower the resistance, the faster the electron flow rate in the hydrogel [73]. and rGO-Elastin). Bode plots representing (a) phase angle Z vs. frequency, and (b) the impedance modulus |Z| (ohms) vs. frequency. Measurements were performed in 0.1 M PBS buffer at room temperature.
The charge transfer resistance, as seen in Figure 7b, has the largest variance, where alginate hydrogels showed the highest impedance value, decreasing when rGO or protein-rGO was introduced into the hydrogels, especially with collagen that showed the lowest impedance value. At low frequencies, the charge transfer resistance can be obtained by interpolating the semi-circle to the real x-axis [73]. The current density of the rGO-proteins alginate hydrogels rises as the hydrogel resistance and charge transfer resistance fall [73]. Figure 8 depicts the charge transfer resistance (Rct) derived by analyzing the EIS spectra. Because the charge transfer resistance characterizes the electron flow at the counter electrode to a great extent, the lower the resistance, the faster the electron flow rate in the hydrogel [73]. The impedance of rGO-proteins alginate hydrogels reduced greater and that was attributed to the π-π bonding between rGO and the adsorbed proteins [74], the electrons of the π-π bonding have higher mobility than that of the holes in rGO, which is the main charge carrier in rGO [75,76].
Cyclic voltammetry (CV) is a technique used to assess redox properties, stability, and surface area of electrodes for biosensing, using materials such as graphene [77]. In this work, we applied the CV to study the electrochemical activity of the alginate hydrogel and hybrid alginate hydrogel with rGO and rGO-proteins. Through the CV the specific capacitance of each hydrogel can be estimated. Voltammetry (CV) measurements showed excellent capacitance for all the tested hydrogels, maintaining the box-like shape even at different scan rates (Figure 9). The impedance of rGO-proteins alginate hydrogels reduced greater and that was attributed to the π-π bonding between rGO and the adsorbed proteins [74], the electrons of the π-π bonding have higher mobility than that of the holes in rGO, which is the main charge carrier in rGO [75,76].
Cyclic voltammetry (CV) is a technique used to assess redox properties, stability, and surface area of electrodes for biosensing, using materials such as graphene [77]. In this work, we applied the CV to study the electrochemical activity of the alginate hydrogel and hybrid alginate hydrogel with rGO and rGO-proteins. Through the CV the specific capacitance of each hydrogel can be estimated. Voltammetry (CV) measurements showed excellent capacitance for all the tested hydrogels, maintaining the box-like shape even at different scan rates (Figure 9).
When rGO-proteins were added to the alginate matrix, the specific capacitance of the hybrid alginate hydrogel was modified ( Figure 10). Collagen-coated rGO maximized the capacitance reaching the highest value (3.17 × 10 −5 F/g), while BSA-rGOalginate hydrogel and elastin-rGO-alginate hydrogel minimized it (to 1.88 × 10 −5 F/g and 1.20 × 10 −5 F/g, respectively) ( Table 6). Therefore, CV results confirmed the measurement quantified in EIS, indicating that collagen-coated rGO hydrogel enhances the conducting behavior of rGO hydrogels [74], modifying the electrical properties of alginate hydrogel. Pharmaceutics 2021, 13, x FOR PEER REVIEW 14 of 18 Figure 9. Cyclic voltammograms (CV) for hybrid rGO-alginate hydrogel at various scan rates. The potential (V) was plotted against the current (A). All specimens were dipped in 0.1M PBS and CV measurements were done at the potential range from −0.2 to 0.5 V at different scan rates (100 mVs −1 ).
When rGO-proteins were added to the alginate matrix, the specific capacitance of the hybrid alginate hydrogel was modified ( Figure 10). Collagen-coated rGO maximized the capacitance reaching the highest value (3.17 × 10 −5 F/g), while BSA-rGO-alginate hydrogel and elastin-rGO-alginate hydrogel minimized it (to 1.88 × 10 −5 F/g and 1.20 × 10 −5 F/g, respectively) ( Table 6). Therefore, CV results confirmed the measurement quantified in EIS, indicating that collagen-coated rGO hydrogel enhances the conducting behavior of rGO hydrogels [74], modifying the electrical properties of alginate hydrogel. Figure 10. Cyclic voltammetry (CV) potential (V) vs. current curves for different hybrid alginate hydrogels containing uncoated and protein-coated rGOs (rGO-BSA, rGO-Collagen, and rGO-Elastin). All specimens were dipped in 0.1M PBS and CV measurements have been done at the potential range of −0.2 to 0.5 V, though the rate is specified as 100 mV/s. . Cyclic voltammograms (CV) for hybrid rGO-alginate hydrogel at various scan rates. The potential (V) was plotted against the current (A). All specimens were dipped in 0.1M PBS and CV measurements were done at the potential range from −0.2 to 0.5 V at different scan rates (100 mVs −1 ). Figure 9. Cyclic voltammograms (CV) for hybrid rGO-alginate hydrogel at various scan rates. The potential (V) was plotted against the current (A). All specimens were dipped in 0.1M PBS and CV measurements were done at the potential range from −0.2 to 0.5 V at different scan rates (100 mVs −1 ).
When rGO-proteins were added to the alginate matrix, the specific capacitance of the hybrid alginate hydrogel was modified ( Figure 10). Collagen-coated rGO maximized the capacitance reaching the highest value (3.17 × 10 −5 F/g), while BSA-rGO-alginate hydrogel and elastin-rGO-alginate hydrogel minimized it (to 1.88 × 10 −5 F/g and 1.20 × 10 −5 F/g, respectively) ( Table 6). Therefore, CV results confirmed the measurement quantified in EIS, indicating that collagen-coated rGO hydrogel enhances the conducting behavior of rGO hydrogels [74], modifying the electrical properties of alginate hydrogel. Figure 10. Cyclic voltammetry (CV) potential (V) vs. current curves for different hybrid alginate hydrogels containing uncoated and protein-coated rGOs (rGO-BSA, rGO-Collagen, and rGO-Elastin). All specimens were dipped in 0.1M PBS and CV measurements have been done at the potential range of −0.2 to 0.5 V, though the rate is specified as 100 mV/s. Figure 10. Cyclic voltammetry (CV) potential (V) vs. current curves for different hybrid alginate hydrogels containing uncoated and protein-coated rGOs (rGO-BSA, rGO-Collagen, and rGO-Elastin). All specimens were dipped in 0.1M PBS and CV measurements have been done at the potential range of −0.2 to 0.5 V, though the rate is specified as 100 mV/s.

Conclusions
Aiming to improve alginate hydrogels, we incorporated reduced graphene oxide (rGO) coated with proteins-BSA, collagen, or elastin-into the alginate hydrogel matrix. Our finding demonstrated that the adsorption of these three proteins onto the rGO surface occurs through the π-π interactions. Moreover, the hydrophobic nature of the rGO surface has increased its affinity for hydrophobic protein and decreased its affinity for hydrophilic proteins. Among the studied proteins, the rGO surface showed the highest adsorption capacity to collagen (qe = 0.0220 µg/µg), while BSA and elastin represented three to five times lower qe values (0.0070 µg/µg and 0.0049 µg/µg, respectively). The thermodynamic study showed that the adsorption of these proteins onto rGO is a nonspontaneous phenomenon. Moreover, the adsorption of collagen by rGO is an endothermic process, while the adsorption of BSA and elastin is exothermic. When rGO-protein matrices were incorporated into the alginate hydrogels, the protein coat on the rGO surface was able to preclude further adsorption of erythropoietin. This collagen-coated rGO addition to alginate hydrogel enhanced alginate's conductivity, leading to the lowest impedance modulus and the highest specific capacitance, 3.17 × 10 −5 (F/g). This enhancement would be extremely helpful for its application in tissue engineering for neuronal or cardiomyocyte regeneration, cell-based therapy, and tissue engineering.