1. Introduction
Many pathologies affecting retinal tissue lead to irreversible vision impairments due to the loss of photoreceptors. Regenerative medicine represents a viable strategy for treating these conditions by replacing dead photoreceptors with healthy ones. Research advancements in cell therapies have shown promise in restoring retinal cells and retinal function [
1]. For instance, the use of induced pluripotent stem cells (iPSCs) to generate a functional retinal pigment epithelium has demonstrated potential in clinical trials [
2]. Yet, significant hurdles, including the long-term survival and integration of transplanted cells, remain [
3]. Moreover, photoreceptors are difficult to cultivate and maintain in vitro [
4]. To tackle these issues, cells can be transplanted using a scaffold, allowing for the delivery of an organized and functional cell layer that integrates better with the host tissue compared with cell injection. Yet, few studies on the fabrication of photoreceptor scaffolds have been reported in the literature. Most of these studies focused on creating scaffolds with an architecture that guides cell differentiation and orientation, which is crucial for fulfilling the light-sensing function [
5]. Steedman et al. proved that the presence of microtopography positively affected cell attachment and induced the differentiation of retinal progenitor cells [
6]. Similarly, Jung et al. showed that a 3D micropatterned scaffold can guide the attachment and the differentiation of pluripotent stem cells into photoreceptors [
7]. Besides topographical cues, scaffold stiffness has also been identified as a key parameter for driving cells to differentiate into photoreceptors and for maintaining lineage specification [
8,
9]. Therefore, selecting suitable substrate materials and fabrication techniques is crucial for recapitulating both the architecture and properties of the in vivo photoreceptor environment [
8]. Current substrates fail to mimic both the architecture and the stiffness of the native milieu, thus hindering progress in retinal regenerative approaches [
7,
10]. Hence, recreating the environment of photoreceptors in vitro still represents an engineering challenge.
Recently, 3D bioprinting has been suggested as a promising technique to fabricate scaffolds for photoreceptors, as it allows 3D porous architectures with precise control over pore size, geometry, and cell spatial distribution to be obtained without significantly affecting cell viability [
5,
11,
12,
13]. Different bioprinting technologies have been developed and employed according to the application [
14]. For instance, Masaeli et al. demonstrated the feasibility of fabricating a 3D in vitro retina model using an inkjet-based bioprinter [
12]. However, with this approach, cells had to be printed in culture medium to avoid clogging issues. So, the native stiffness could not be matched. Among the bioprinting approaches, pneumatic-driven extrusion-based bioprinting is one of the most common due to its versatility in printing a wide range of bioinks and its ease of use [
15]. Additionally, it enables the production of clinically relevant constructs, in terms of size, within a reasonable time. Bioinks employed in extrusion bioprinting must display shear-thinning behavior to enable successful extrusion and printing while preserving the structure given by the printing process [
16]. Shi et al. used an alginate–pluronic bioink that included a retinoblastoma-derived cell line to replicate the photoreceptor layer in a 3D in vitro model [
17]. They evaluated the impact of two different printing patterns on cell morphology; however, the stiffness of the bioink was not investigated. Hence, to date, a bioink capable of meeting the requirements for extrusion-based bioprinting and simultaneously accurately replicating retinal properties has yet to be developed [
11,
18].
Sodium alginate–gelatin (SA-G) blends have been successfully applied over the years in extrusion-based 3D bioprinting for engineering different tissues, such as liver and neural tissue [
19,
20,
21]. These blends have shown a good printability window, a rapid crosslinking rate, printing accuracy, and biocompatibility [
22,
23]. Moreover, they can be obtained via an easy and cost-effective process. We believe that an SA-G bioink holds great potential for bioprinting a functional photoreceptor layer, as gelatin provides cell attachment sites, while sodium alginate exhibits a suitable stiffness for the development of photoreceptors. Hunt et al. found that a sodium alginate-based hydrogel functionalized with arginine–glycine–aspartate groups promoted photoreceptor cell differentiation in comparison with hydrogels based on hyaluronic acid and pure gelatin due to its stiffness [
8].
The present research aims to introduce a novel platform that is potentially able to guide and sustain functional photoreceptors in vitro through both topographical and mechanical cues. To this end, we optimized a bioink composed of sodium alginate and gelatin to replicate retinal native properties, such as stiffness, while ensuring successful extrusion and high shape fidelity in printed constructs. To the best of our knowledge, no bioink with these features has been developed previously. To investigate the properties of the optimized SA-G bioink, we performed a comprehensive characterization based on well-established testing protocols as well as protocols developed specifically for this study.
2. Materials and Methods
A schematic representation of the research methodology is reported in
Figure 1.
2.1. Material Preparation
To prepare the SA-G bioink, low-viscosity sodium alginate (154725, Lot Number: SR01864) and gelatin (bovine skin type B, G9391, Lot Number: SLCF9893) powders were purchased from MP Biomedicals (Santa Ana, CA, USA) and Sigma-Aldrich (St. Louis, MO, USA), respectively.
The hydrogel precursor solution was prepared as follows. First, the powders were separately sterilized in 99.99% ethanol (Sigma-Aldrich, St. Louis, MO, USA) for 15 min. The sterilization process was repeated three times. After the complete evaporation of ethanol, 0.4 g of sodium alginate powders were dissolved in 10 mL of sterile phosphate-buffered saline (PBS, 1X, w/o Calcium, w/o Magnesium, pH = 7.4, Gibco, Thermo Fisher Scientific, Waltham, MA, USA) overnight under vigorous magnetic stirring. Similarly, 1 g of gelatin was dissolved in 10 mL of sterile PBS at 37 °C until total solubilization. The two resulting solutions were mixed together in a volume ratio of 1:1. Hence, the final alginate and gelatin concentrations in the precursor were 2% and 5% wt./v. These concentrations were selected after an optimization process based on preliminary extrudability and printability tests. The precursor solution was stored at 4 °C and used within one week.
The crosslinking solution was prepared by dissolving 2%
wt./
v of calcium chloride (CaCl
2, Sigma-Aldrich, St. Louis, MO, USA) into sterile distilled water and stored at 4 °C until use. The CaCl
2 concentration and the crosslinking duration were optimized based on previous studies [
24]. The use of CaCl
2 is widely reported in the literature to crosslink sodium alginate [
25]. The crosslinking mechanism is based on ionic bonds between calcium ions and carboxylate ions present in the alginate. Briefly, sodium alginate is a linear polysaccharide containing carboxylate ions, each bonded with a sodium ion. When alginate is immersed in a CaCl
2 solution, the dissociated calcium ions replace the sodium ions in the alginate, ionically bonding with the carboxylate ions. Since each calcium ion can bond with two carboxylate ions, the process results in the crosslinking of the polymer chains and, thus, in the formation of a hydrogel [
25].
2.2. Swelling Degree and Degradation Rate
To evaluate the hydrogel swelling and degradation properties, the SA-G precursor solution was transferred into a Petri dish and crosslinked at 4 °C for 10 min. Hydrogel cylindrical samples (n = 3) of 7 mm diameter and 3 mm thickness were obtained by using a biopsy punch. The samples were then crosslinked by submerging them in the CaCl2 crosslinking solution for 10 min at room temperature.
To evaluate the swelling degrees and degradation rates, the samples were placed in a 6-well plate, immersed in complete culture medium composed of high glucose Dulbecco’s Modified Eagle’s Medium (DMEM, Gibco
TM, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum albumin (FBS, American Type Culture Collection, Manassas, VA, USA) and 1% penicillin/streptomycin antibiotics (Gibco
TM, Thermo Fisher Scientific, Waltham, MA, USA), and incubated at 37 °C for the following time points: 15 min, 30 min, 1 h, 2 h, 3 h, 4 h, 5 h, and 6 h (same set of samples for each time point) [
26]. After specific incubation periods, the samples were weighed to determine the swelling degree. The swelling degree at each time point was computed according to Equation (1):
where W
t is the sample weight at each time point, and W
0 is the starting sample weight before adding the solution.
The same set of samples was employed to study the degradation rate. In more detail, after reaching the swelling equilibrium, samples were incubated at 37 °C for 1, 3, 5, 7, 10, 14, and 21 days. At each time point, the samples were weighed, and the degradation rate was calculated by quantifying the weight decrease as in Equation (2):
where W
t is the sample weight at each time point, and W
s is the sample weight reached at swelling equilibrium.
2.3. Mechanical Tests
Hydrogel mechanical properties were measured by performing uniaxial tensile tests through an ad hoc setup. This includes two poly(lactic acid) (PLA, K Kentstrapper, Florence, Italy) plates fabricated using fused-deposition modeling (Verve, Kentstrapper, Florence, Italy). Each plate consists of a base and a protrusion with dimensions of (30 × 30 × 10) mm and (10 × 10 × 20) mm, respectively (
Figure 2a). The protrusions were designed to be inserted and secured into the jaws of a testing machine (Synergie 200, MTS Systems, Eden Prairie, MN, USA) by tightening, thus preventing any slipping of the plates. Hydrogel samples (
n = 5) obtained as described in
Section 2.2 were tightly glued to the bases of the plates by using a biocompatible tissue adhesive (3M
TM Vetbond
TM, Thermo Fisher Scientific, Waltham, MA, USA). Subsequently, the plates along with the sample were mounted onto the testing machine, which was equipped with a 100N loading cell (
Figure 2b). Samples were pulled until failure at a displacement rate of 0.1 mm/s at room temperature (25 °C). The displacement rate was selected based on previous tensile tests on retinal samples [
27].
For the analyses, force (N)−elongation (mm) data were elaborated to obtain the stress [kPa]−strain [mm/mm] relationship for each sample. The stress and the strain were computed as indicated in Equations (3) and (4), respectively. The elastic modulus (E) was extracted from the stress–strain plots as the slope of the initial linear region of the stress–strain curve.
where σ and ε are the stress and the deformation, respectively. F is the force applied to the original cross section A, and ΔL is the change in length with respect to the original length L
0.
2.4. Rheological Characterization
Rheological analyses were carried out at 25 °C using a rotational rheometer (Discovery HR2, TA Instruments, New Castle, DE, USA) equipped with a cone-plate geometry (diameter: 20 mm). The testing temperature was selected based on the printing temperature. Prior to all tests, a conditioning step, consisting of a rotation of the tool at 10 s−1 for 30 s followed by a soak period of 30 s, was included.
To evaluate the rheological properties of the SA-G bioink, two specific tests were performed: a flow curve and a strain sweep. In the former, the viscosity and the shear stress exerted on the material in response to the applied shear rates were measured in a continuous flow experiment. Specifically, the shear rate was progressively increased in a logarithmic manner from 0.001 to 10,000 s
−1. Ten points per decade were registered, and the test time was 180 s., i.e., 30 s per decade. In the latter, material responses, namely the storage modulus (G′) and the loss modulus (G″), were measured by applying oscillatory strains in a logarithmic manner from 0.01% to 1000% at a frequency of 1 Hz. Ten points per decade were recorded. The testing parameters for the flow curve and strain sweep were determined by conducting preliminary tests. To assess the impact of cells on the bioink rheological properties, both tests were repeated on the SA-G bioink loaded with L929 cells at different cell densities (5 × 10
6 cells/mL and 10 × 10
6 cells/mL) [
16]. The cell encapsulation procedure will be illustrated in
Section 2.7.
To investigate the material recovery ability after shear application and removal, a strain recovery test and an elastic recovery test were carried out. In the strain recovery test, the viscosity was measured during the following consecutive steps that mimic the printing process: (i) pre-printing phase (shear rate of 0.05 s
−1 for 300 s); (ii) printing (sudden increase in shear rate to 811 s
−1 for 60 s); (iii) post-printing phase (shear rate of 0.05 s
−1 for 300 s). Thirty points per step were recorded. A 0.05 s
−1 shear rate was chosen to simulate material resting conditions, whereas 811 s
−1 was found to be the maximum shear rate value applied to the material during extrusion. It was computed by using Equation (5), which was derived from the Hagen–Poiseuille law with Rabinowitsch correction [
28].
where Q (mm
3s
−1) is the flow rate, R (mm) is the inner radius at the outlet of the nozzle used in the bioprinting process and (n − 1) is the slope of the viscosity versus shear rate graph on a log–log plot obtained from the flow curve. To assess the material elastic recovery after the printing process, oscillatory time sweeps made at a frequency of 1 Hz were performed with alternating high/low shear stresses, i.e., 7 × 10
−4 and 10
−6 MPa, respectively. The maximum shear stress value set in this test was the one corresponding to a shear rate of 811 s
−1. The material was allowed to recover for 2 min to simulate the non-printing phases. All rheological measurements were repeated three times.
2.5. In Vitro Cytotoxicity
Cylindrical hydrogel samples (
n = 3), prepared as described in
Section 2.2, were placed in 6-well plates and incubated in complete culture medium at 37 °C and 5% CO
2. The medium was collected at different incubation times (1, 5, and 7 days). Meanwhile, L929, a murine fibroblast cell line (Catalog No. CCL-1
TM, American Type Culture Collection, Manassas, VA, USA), was routinely cultured in complete culture medium at 37 °C and 5% CO
2, with medium replacement every 2 days. The cells were cultured for four sequential passages until they reached the desired density. L929 cells were then seeded in 12-well plates at a density of 1 × 10
5 cell/well and grown until they reached 70% confluency. Subsequently, these cells were cultured for 24 h with the medium that had been incubated with the hydrogel samples. As a control, cells were cultured with fresh complete culture medium. All experiments were performed in replicates of three (n = 3). After 24 h, cell viability was assessed using the alamarBlue
TM assay (Invitrogen
TM, Carlsbad, CA, USA) according to the manufacturer’s instruction. Alamar blue is an oxidation–reduction indicator that changes color upon reduction by living cells. To perform the assay, the alamar blue stock solution was diluted to 1:10 with complete culture medium. The resulting 10% alamar blue solution was incubated with the cells for 3 h at 37 °C. After incubation, a volume of 100 µL/well was transferred to a 96-well plate for absorbance reading. The absorbance at 570 nm and 600 nm was measured using a microplate reader (infinite 200Pro, Tecan Group Ltd, Männedorf, Switzerland), and the percentage viability was computed as the percent difference in reduction between treated and control cells.
2.6. Printability Assessment
Cellink INKREDIBLE+ (Cellink AB, Gothenburg, Sweden) was used to print the SA-G bioink. Such a 3D bioprinter is based on pneumatic extrusion technology and is equipped with dual printheads featuring a heating system and built-in photocuring modules. Moreover, it includes patented clean-chamber technology that efficiently filters air through an H13 HEPA filter, thus providing a clean printing environment. All structures introduced in this section were designed with Solidworks 2020 (Dassault Systems Solid WorksCorp, Waltham, MA, USA), sliced with PrusaSlicer 2.3.3. (Prusa Research, Prague, Czech Republic), and printed at 25 °C. Two conical nozzles with inner diameters of 22G (0.41 mm) and 25G (0.25 mm) were used. Prior to printing, the hydrogel precursor solution stored at 4 °C was incubated at 37 °C for 15 min, thus allowing the solution to be transferred to a printing cartridge. The printing cartridge was left at 25 °C for 40 min before printing to enable partial gelatin crosslinking. After printing, all structures were physically crosslinked at 4 °C for 5 min followed by chemical crosslinking in a 2% wt./v CaCl2 bath for 10 min.
Printability was assessed by evaluating the filament formation and printing accuracy in terms of filament merging and shape fidelity in multi-layered structures [
16]. For the first test mentioned, four lines were printed at different increasing velocities, i.e., 5, 10, 15, and 20 mm/s, by applying a pressure of 25 kPa. The test was repeated using a printing pressure of 30 kPa. The line widths were measured using ImageJ software (NIH, Stapleton, NY, USA, ver. 1.54i) after imaging the lines with an optical microscope (Eclipse Ti2, Nikon, Tokyo, Japan). Based on the results of this test, optimized printing parameters were chosen for further analyses. The printing accuracy on the x–y plane was determined through the filament fusion test [
29]. For this test, a precise pattern was designed that involved parallel lines spaced at different distances (from 0.5 mm to 3 mm), with a progressive 0.5 mm increase for each subsequent line. Ultimately, circular discs with a diameter of 30 mm and a thickness of 0.4 mm were designed and printed with four layers to evaluate the shape fidelity of multilayered constructs. Specifically, the dimensions of the printed structures were measured using ImageJ software and compared with the original dimensions set in the digital model of the structure. Moreover, optical images taken by the Eclipse Ti2 microscope were analyzed to investigate the pore geometry of the printed discs. Subsequently, the printability index (Pr) was computed using Equation (6) [
16]:
where L and A are the pore perimeter and area, respectively.
2.7. Bioprinting
Before bioprinting, L929 cells routinely cultured in complete culture medium were centrifuged and resuspended in fresh culture medium at a density of 5 × 10
6 cells/mL. The cell suspension was mixed with the SA-G precursor solution previously heated at 37 °C for 15 min by means of a luer-lock connector to homogeneously encapsulate cells within the solution. The SA-G bioink containing cells was transferred into a printing cartridge equipped with a 22G (0.41 mm) conical nozzle and left at 25 °C for 40 min before printing. Four-layered grids with dimensions of 10 mm × 10 mm × 1 mm were then printed using the printing parameters established with the printability tests (
Section 2.6). The printed structures were crosslinked as indicated in
Section 2.1 and cultured at 37 °C and 5% CO
2. Cell viability was evaluated with the LIVE/DEAD Cell Viability Assay (Invitrogen, Carlsbad, CA, USA) following manufacturer instructions. Fluorescent images were acquired with an optical microscope (Eclipse Ti2, Nikon, Tokyo, Japan) 1 and 5 days after printing.
2.8. Statistical Analyses
Statistical analyses were performed with IBM SPSS software (ver. 29). Results are presented as mean ± SD. To evaluate the effect of cell density on the SA-G viscosity, the Friedman test was used with a 95% confidence interval, whereas for the cytotoxicity analysis of degradation products, two-way ANOVA was employed with a 95% confidence interval. Significant differences were reported for p-values lower than 0.05, with * indicating a p-value < 0.05.
4. Conclusions
In this study, we aimed to optimize an SA-G bioink formulation to match the properties of the native photoreceptor layer and to be successfully extruded and printed with good resolution. To assess its suitability as a substrate for photoreceptors in in vitro culture, we conducted a comprehensive characterization of an SA-G bioink. Our results showed that the stiffness of the bioink closely mimics the in vivo photoreceptor stiffness. This parameter plays a crucial role in the development of a functional photoreceptor layer, as it influences the cell behavior guiding the cell phenotype and genotype. In addition, through rheological and printability tests, we demonstrated the bioink’s ability to fabricate multilayered structures characterized by high shape fidelity. This feature is crucial for obtaining a microstructure able to drive photoreceptor polarization, thus replicating their highly organized in vivo configuration. Ultimately, the viability of fibroblasts, both in contact with material extracts and embedded within the bioink, proved that the SA-G material is non-cytotoxic. Taken together, these findings suggest that the proposed SA-G bioink may be used as a tissue mimetic to produce a cellularized construct with the specific pore size and geometry capable of promoting photoreceptor development, alignment, and functional culture in vitro. Future studies should focus on incorporating stem cells into the SA-G bioink to form a functional photoreceptor layer in vitro. This step will require the selection of the appropriate culture medium and growth factors and the execution of several tests to determine the optimal cell density for mimicking the native high packing density of photoreceptors. Different printing patterns will then be explored to determine which microstructure best supports native-like spatial cell organization. The formation of a functional photoreceptor layer will be evaluated by conducting a thorough biological analysis, including of cell viability and proliferation and the expression of specific markers. Finally, in vivo studies need to be performed to assess the functional integration and stability of the engineered photoreceptor layer.