*5.3. Characterization of the GNP*

## 5.3.1. Characterization of GNP by ESEM

Size, structure and overall morphology of the GNP samples (empty, FITC-protein A and BMP-2 loaded), were examined using ESEM (FEI quanta 250 FEG, FEI, Hilsboro, OR, USA). The impact of variations in the production method (temperature, pH, stirring speeds and times of addition of the crosslinker) on the formation of GNPs, size, shape and aggregation were investigated. For this purpose, double-sided adhesive polycarbonate-based conductive tabs were glued to pin sample plates made of aluminium. The freeze-dried samples in powder form could thus be fixed on the adhesive surface and were sputtered with gold before ESEM measurement (JFC-1200 fine coater, Jeol, Freising, Germany) to increase the conductivity and stability of the samples in high vacuum.

#### 5.3.2. Determination of the Inclusion Capacity

To assess how much protein was actually present in the nanoparticles, the supernatant obtained after centrifugation was measured with the fluorimeter (Ensight Multimode Plate Reader, PerkinElmer, Rodgau, Germany). The Kaleido software was used to obtain the data for quantitative analysis. A calibration curve was generated based on levels within the supernatant, which could then be used to approximate protein content within the solvent. Acetone and water in a ratio of 4:1 (*v*/*v*), plus a quarter of the amount of GTA were used. The greatest uncertainty was found for the amount of GTA, since the amount

of unreacted GTA after the crosslinking process was unknown. The difference between the amount of FITC-protein A used and the amount in the supernatant was defined as the amount included in the GNPs. In order to quantify how much BMP-2 was present in the produced GNPs, the supernatant obtained after the first centrifugation was analyzed using an ELISA. The evaluation of the color reaction was performed with the UV/Vis spectrometer (Spectrostar Nano, BMG Labtech, Ortenberg, Germany). A calibration curve was generated using the BMP-2 standard of the kit. The difference between the amount of BMP-2 used and the amount in the supernatant was defined as included BMP-2 in the GNPs.

$$\text{IC} = \frac{\text{m}(\text{FTC}\_{\text{total}}) - \text{m}\left(\text{FTC}\_{\text{superstart}}\right)}{\text{m}(\text{FTC}\_{\text{total}})} \tag{1}$$

IC: Inclusion capacity;

m(FITCsupernatant): Mass of FITC-protein A at supernatant; m(FITCtotal): Total mass of FITC-protein A.

#### *5.4. Biocompatibility*

In order to show that the GNPs were biocompatible with living cells and that after purification no cytotoxic residues such as acetone or GTA from the production steps were present, a cell culture experiment was carried out. MG-63 cells (ATCC CRL 1427) were used. For each sample 50,000 cells were placed in a 12-well plate on glass plates, which were previously coated with polylysine. The cells were cultivated overnight with Dulbecco's Modified Eagle Medium (DMEM F12 containing F12 nutrient and the additions of 1% penicillin/streptomycin (P/S, Sigma Aldrich (now Merck), Darmstadt, Germany) and 10% fetal bovine serum (FBS, Merck, Darmstadt, Germany) in a New Brunswick Galaxy 170 R incubator (Eppendorf, Hamburg, Germany) at 37 ◦C and a CO<sup>2</sup> saturation of 5%. The next day, a production batch (from originally 1.25 g gelatin) GNP were washed in 1 mL ethanol (60%) to prevent cell culture contamination. The GNP were dissolved in the culture medium and pipetted to the cells. Nunc™ Thermanox™ Coverslip (Thermo Fisher Scientific, Waltham, MA, USA) membranes were used as a control. For each time point at least three samples were used and all experiments were repeated at least three times.

#### 5.4.1. Live/Dead Assay

A total of 3 samples for each day and one blank sample (without GNP) for day 3, 7 and 10 were used for the live/dead assay. For the live/dead assay (L/D), the cells were stained with Calcein-AM and EthD-III (out of Live/Dead Cell Staining Kit II (PromoCell, Heidelberg, Germany)) on the respective days and observed under the fluorescence microscope (Olympus BX 51, Olympus, Hamburg, Germany), the evaluation was performed with the Stream Motion Software Version 1.7.1 from Olympus. Living cells exhibited a green fluorescence under blue light, and dead cells a red fluorescence. In a further experiment the influence of the released BMP-2 on the MG-63 cells was investigated. The L/D assay was repeated with GNPs loaded with BMP-2. This was followed by incubation in the incubator for the various experiments.

#### 5.4.2. Cell Proliferation Assay

Thermanox™ Coverslip membranes were used as a control. All samples and controls were equally covered with 50,000 cells in 200 µL. The cells were incubated for 4 h at 37 ◦C with a CO<sup>2</sup> saturation of 5% in the incubator for 3, 7 and 10 days. At the end of this period, 400 µL of the DMEM-F12 complete medium were added to each sample and incubated. After 24 h at day 0, 100 µL resuspended GNP + BMP2 were added. A medium change with the DMEM-F12 with the 10% FBS and 1% P/S additives was performed for days 7 and 10. For each WST evaluation the medium was aspirated, and the wells were washed three times with PBS. The samples and the Thermanox™ Coverslips were then transferred to a new well, and then 400 µL of the DMEM-F12 phenol red free (Art. No. 11039-021, Gibco, Grand Island, NE, USA) with the 1% P/S and 1% fetal bovine serum (FBS, Merck,

Darmstadt, Germany) additives were added to the wells with the sample. 400 µL of the medium were added to the previously used empty sample wells, Positive Control (C + R), Empty Control Well (C+), and Blank. The blank contained only the DMEM medium without phenol red and was measured to account for background absorption. 10% WST reagent (Art. No. 05015944001, Roche, Basel, Switzerland) was added to the corresponding volume of medium. Thus, 250 µL WST were added to the sample wells, and 40 µL were added to the old wells, the blank wells, and the positive control (C + R). This was incubated in an incubator at 37 ◦C for 2 h. After this time, the liquids were transferred into a 96 well plate. 100 µL of each solution were added to three separate wells. The absorption was then measured at 450 nm using a Spectrostar Nano microplate reader (BMG Labtech, Ortenberg, Germany). In a further experiment the influence of the released BMP-2 on the MG-63 cells was investigated. The WST-I was repeated with GNPs loaded with BMP-2.

#### 5.4.3. Lactate Dehydrogenase (LDH) Assay

Each experiment assessed three Thermanox™ cover slips, a negative control (cells only), a positive control (Triton X), and a blank to account for background absorbance in the ELISA reader. The experiments were repeated at least three times. A 200 µL cell solution containing 50,000 cells was seeded onto each scaffold, and a 100 µL cell solution containing 50,000 cells was seeded onto the Thermanox™ cover slips. One well was left empty for use as a blank. The well plate was placed in an incubator at 37 ◦C with 5% CO<sup>2</sup> for 4 h. Following incubation, 400 µL of DMEM-F12 phenol red free with the 1% P/S and 1% FBS additives were added into the sample-wells and control-wells. Since FBS itself contains LDH, a concentration of 10% in the medium might have triggered background absorption. Therefore, only a concentration of 1% FBS was added to the medium. For the positive controls, 1% Triton × 100 (Art. No. X100, Sigma Aldrich, Saint Louis, MO, USA) was added to the DMEM-F12 medium with 1% P/S and 1% FBS to kill the cells. The LDH experiments were carried out at 24, 48 and 72 h following seeding and the same procedure was repeated at each interval: Three 100 µL samples were taken from each well into a 96 well plate. An LDH reagent (100 µL) was added to each well in use, and the plate was incubated in darkness at room temperature for 30 min. Following incubation, the plate was placed in a Spectrostar Nano microplate reader (BMG Labtech, Ortenberg, Germany), and absorbance was measured at a λ of 490 nm with a reference λ of 600 nm. In a further experiment the influence of the released BMP-2 on the MG-63 cells was investigated. The LDH was repeated with GNPs loaded with BMP-2.

## *5.5. ADA-GEL Hydrogel*

#### 5.5.1. Preparation of Alginate-di-Aldehyde (ADA)

ADA was produced according to the method developed by Sarker et al. [45]. For this purpose, sodium alginate was dissolved in ethanol (99.8%) and sodium periodate dissolved in aqua bidest. was added. For the oxidation reaction, stirring was carried out for 6 h under exclusion of light (beaker was wrapped with aluminum foil). The reaction was stopped with ethylene glycol and stirring was continued for 30 min. The ADA was dialysed for 7 days against aqua bidest. to remove any remaining sodium periodate by using the dialysis system Spectra/Por (Repligen, Boston, MA, USA) with standard RC dialysis membranes (6–8 kD MWCO). The bidest. water was changed twice a day. After dialysis, the ADA was dried in a lyophilisator (FreeZone 2.5, Labconco, Kansas City, MO, USA) for another seven days.

#### 5.5.2. Crosslinking ADA and GEL

The cross-linking of the aldehyde groups produced by oxidation with the amino groups of the gelatin always took place shortly before the gel was used and is described in the following experiments.

#### *5.6. ADA-GEL Beads with Drugs*

#### 5.6.1. ADA-GEL Beads with Clindamycin and FITC-Protein A-Containing GNPs

A 5% *w*/*v* ADA solution was prepared by homogenizing freeze-dried ADA with aqua bidest. water for several hours on a magnetic stirrer in a beaker. Meanwhile, a 5% GEL solution was prepared from GEL A, 300 g Bloom (GELITA AG, Eberbach, Germany) with aqua bidest. water at 37 ◦C on the magnetic stirrer. Clindamycin was added to the gelatin after homogenization to achieve a final concentration in the beads of 50 mg/mL. Since clindamycin is light sensitive, the beaker was wrapped in aluminum foil. The amount of FITC-protein A GNPs from 2 production batches of 1.25 g each of initial GEL mass was added. GNP prepared with 2% FITC-protein A was used. The GNPs were added to the homogenized ADA. The clindamycin-containing gelatin and the ADA containing GNP were mixed together (1:1) and carefully dripped with the beaker into a 30 mM calcium chloride solution [19]. In addition to the active substance containing beads, hydrogel beads were prepared as blank samples without drugs.

#### 5.6.2. ADA-GEL Beads with CLI and BMP-2 Containing GNP

The beads were prepared with CLI and BMP-2 containing GNPs as described in 2.6.1. Before adding the gelatin, GNPs with BMP-2 from a batch of 1.25 g of gelatin (twice as much as described in 2.2) was added to the ADA. After the addition of the gelatin containing CLI, the mixture was homogenized under light for 30 s. The beads were then prepared by dropping the ADA-GEL into 30 mM CaCl<sup>2</sup> solution.

#### *5.7. Drug Release Experiments*

All drug release experiments were performed with a minimum of 10 samples and repeated at least three times.

#### 5.7.1. Drug Release from ADA-GEL Beads

1 g of the CLI loaded beads were stored, after being weighed with a precision balance (Secura 125-1CEU, Sartorius, Göttingen, Germany), with 3 mL bidest. water in 5 mL Eppendorf Tubes, at 37 ◦C for 28 days in an oven (Memmert IN 75, Schwabach, Germany). After 1, 2, 3, 6, 9, 14, 21 and 28 days, samples were taken and the liquid was completely removed and replaced by aqua bidest. The obtained liquid was frozen at −20 ◦C until analysis by HPLC.

#### 5.7.2. Dual Drug Release from ADA-GEL Beads

The dual release (CLI out of the ADA-GEL beads, BMP-2 out of GNPs within the ADA-GEL beads) was carried out here according to the same principle as described above. 1 g of loaded ADA-GEL beads was placed in a 5 mL Eppendorf tube with 3 mL bidest. water. The release was tested at 37 ◦C for 28 days. After 1, 2, 3, 6, 9, 14, 21 and 28 days, samples were taken, the complete liquid was removed once again and replaced by bidest. water. The samples were frozen at −20 ◦C in the same way as the release tests from the beads until they were analyzed by HPLC or fluorimeter (FITC-prot A) and ELISA kit (BMP-2).

#### 5.7.3. Quantitative Analysis

The released clindamycin from the ADA gelatin beads was quantitatively determined by HPLC. The HPLC System (Shimadzu, Kyoto, Japan) consisting of 2 Nexera XR LC-20AD pumps and a SIC-30AC autosampler, CTO 20 AC column oven, DGU-20A5R Degasser, SPD-M20A PDA detector, RF 20A fluorescence detector and a CBM-20A controller. A reversedphase column of butyl-modified silica gel was used as a separation column (NUCLEOSIL 300-5 C4, 5 µm, 250 × 4.6 mm, Macherey-Nagel, REF 761989.30, Düren, Germany). An acetonitrile/sodium hydrogen phosphate solvent with pH of 3.5 at a the ratio 29:71 was used according to Batzias et al. [46]. Flow rate was 0.66 mL/min at 25 ◦C. Clindamycin was eluted after 3.5 min and registered by a PDA detector at 193 nm. The released FITC-

protein A concentration was measured with a fluorimeter (Perkin Elmer EnSight, Waltham, MA, USA) at 490 nm excitation and 525 nm emission. The determination of the BMP-2 concentration was performed with an ELISA kit from Sino Biological according to their protocol. The 96 well plate already coated with a BMP-2 antibody by the manufacturer was first washed 3 times with 200 µL wash buffer. 100 µL of the samples were then pipetted into the wells and incubated at room temperature for two hours. The washing step was repeated 3 times before 100 µL of the detection antibody against BMP-2 were added conjugated with horseradish peroxidase. This was incubated for one hour at room temperature. Washing was repeated 3 times before the dye solution was added. After 20 min 50 µL of the Stop Solution were added to stop the color reaction. Using a UV/Vis spectrometer (SpectroStar Nano, BMG Labtech, Ortenberg, Germany), the color reaction was quantitatively evaluated using a calibration curve (2500 . . . 31.5 pg/mL) at λ = 450 nm.

#### 5.7.4. Kinetics Model

The release kinetics of CLI, FITC, and BMP-2 were fitted to a cumulative diagram according to the following Ritger [47] exponential relationship:

$$\frac{\mathbf{M}\_{\text{l}}}{\mathbf{M}\_{\infty}} = \mathbf{k} \mathbf{t}^{\text{n}} \tag{2}$$

where Mt/M<sup>∞</sup> = fractional solute release; t = release time; k = a constant; and n = diffusional exponent characteristic of the release mechanism. In the case of pure Fickian release, the exponent n has limited values 0.50, 0.45 and 0.43 for release from slabs, cylinders and spheres.

#### *5.8. Statistical Analysis*

Data was expressed as mean values ± standard deviation of the mean and analyzed by one-way analysis of variance (ANOVA). The level of statistical significance was set at *p* < 0.05. For statistical calculations, Origin 2020 Professional SR1 (OriginLab, Northampton, MA, USA) was used.

**Supplementary Materials:** The following supporting information can be downloaded at: https://www. mdpi.com/article/10.3390/gels8060365/s1. Figure S1: Image (live/dead staining) of agglomerated GNP-BMP2 with MG-63 cells; Figure S2: Cumulative releases, fit according to Ritger et al. [25]; (a) Clindamycin; (b) Conjugated FITC-protein A, with a diffusion coefficient n = 0.5 for Fickian' diffusion; (c) BMP-2 in relation to the BMP-2 amount used. According to previous published work [41], the fitting for the CLI release was split into a beginning part (nB) and final part (nF), both with anomalous diffusion (n<sup>B</sup> = 0.35 and n<sup>F</sup> = 0.01).

**Author Contributions:** M.S. and R.S. conceived and designed the experiments; M.S., L.R., P.S. and S.S., performed the experiments; L.R. sterilized and characterized the ADA-GEL Gels, M.S., L.R., P.S. and S.S. analyzed the data; M.S. and R.S. contributed reagents/materials/analysis tools; M.S. and S.S. wrote the paper. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by German Research Foundation (DFG) grant number 388988890. The article processing charge was funded by the Baden-Wuerttemberg Ministry of Science, Research and Art and the University of Freiburg in the funding programme Open Access Publishing.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author.

**Acknowledgments:** The authors would like to thank GELITA for providing the gelatin 300 bloom. We would also like to thank Logan Poehlman, Isabelle Caseley and Melanie Lynn Hart for proofreading the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.
