3D bioprinting of tyramine modified hydrogels under visible light for osteochondral interface

Recent advancements in tissue engineering have demonstrated a great potential for the fabrication of three-dimensional (3D) tissue structures such as cartilage and bone. However, achieving structural integrity between different tissues and fabricating tissue interfaces are still great challenges. In this study, an in situ crosslinked hybrid, multi-material 3D bioprinting approach was used for the fabrication of hydrogel structures based on an aspiration-extrusion microcapillary method. Different cell-laden hydrogels were aspirated in the same microcapillary glass and deposited in the desired geometrical and volumetric arrangement directly from a computer model. Alginate and carboxymethyl cellulose were modified with tyramine to enhance cell bioactivity and mechanical properties of human bone marrow mesenchymal stem cells-laden bioinks. Hydrogels were prepared for extrusion by gelling in microcapillary glass utilizing an in situ crosslink approach with ruthenium (Ru) and sodium persulfate photo-initiating mechanisms under visible light. The developed bioinks were then bioprinted in precise gradient composition for cartilage-bone tissue interface using microcapillary bioprinting technique. The biofabricated constructs were co-cultured in chondrogenic/osteogenic culture media for three weeks. After cell viability and morphology evaluations of the bioprinted structures, biochemical and histological analyses, and a gene expression analysis for the bioprinted structure were carried out. Analysis of cartilage and bone formation based on cell alignment and histological evaluation indicated that mechanical cues in conjunction with chemical cues successfully induced MSC differentiation into chondrogenic and osteogenic tissues with a controlled interface.


Introduction
Bioprinting is a rapidly evolving technique in tissue engineering allowing the fabrication of complex and multiphasic structures that can physiochemically mimic native tissues. Recent advancements in bioprinting exhibit great potential for manufacturing both soft and hard tissue structures, such as cartilage and bone, and for the precise deposition of various cells in concert with different biomaterials [1][2][3]. These advanced studies, which have targeted the regeneration of individual tissues, mostly include the use of encapsulated cells and bioactive molecules in hydrogels [4,5], scaffold-free approaches [6,7], and their combination to fabricate multi-layered constructs [8]. However, three-dimensional (3D) bioprinting of a hybrid structure for soft-hard tissue interfaces is still in its early stage.
Osteochondral (OC) interface is where viscoelastic soft cartilage meets hard bone tissue with different physiochemical and mechanical properties. The hierarchical and gradient nature of the interface facilitates its load-bearing applications and further assists structural functionality in articular joints and spine [9]. Osteoarthritis (OA) is one of the most debilitating OC-related degenerative diseases [9,10]. OA is a degenerative condition affecting the entire joint, not just a surface lesion. The breakdown of joint cartilage connection, synovial inflammation, subchondral bone abnormalities, as well as ligaments, muscles, and neurological tissues, all play a role in the disease's complex initiation and development [11]. However, current therapies for OA primarily seek to relieve pain without repairing damaged tissues [10]. In this context, the development of new medical treatments and the expansion of treatment options are urgently needed [12].
Bioprinting of hybrid soft-hard structures can mimic the intricate properties of tissue interface, providing a possible approach for the treatment of bone to cartilage tissue injuries [5,13]. Natural hydrogels such as gelatin, alginate (Alg) [14], and carboxymethyl cellulose (CMC) [15] in combination with their hybrids [16], have shown promising results in tissue engineering of OC defects, potentially stimulating the formation of extracellular matrix (ECM) [17]. An optimal hydrogel matrix for bone and cartilage engineering should encourage cell growth/ proliferation, preserve chondrocyte/osteoblast morphologies, and stimulate chondrogenic/osteogenic differentiation of stem cells for OC interface [1,18]. To enhance the mechanical properties of these hydrogels, several strategies including various crosslinking mechanisms [19][20][21][22], reinforcing with stiff materials [4,23,24], and increasing crosslinking density [25][26][27] have been developed. Various hydrogels, such as Alg [28], gelatin [29,30], and CMC [31], have been modified with tyramine for different tissue engineering applications. Through incorporation of phenolic hydroxyl groups (Ph), tyramine can enhance the hydrophobicity of the CMC hydrogel and hence, improve cell adhesion sites. Moreover, it was reported that by adjusting Ph content in CMC, hydrophobicity of the structures could be tuned resulting in controlling cell adhesion and proliferation [32,33]. Tyramine conjugations significantly enhanced biocompatibility and bioactivity, adhesion sites, and cell attachment, as well as improved mechanical properties.
Maintaining molecular communication between tissue layers, incorporating the chondro-osseous interface, and replicating each matrix composition are all extremely challenging and are major therapy barriers [34]. Biofabricated gradient hybrid structures can create a successful transition between cartilage and bone; furthermore, they can prevent interface uncertainty and better match the composition of OC interface [35]. By generating a smooth transition between cartilage and bone components, a continuous gradient scaffold can reduce the displacement at the tissue interface compared to distinct gradient scaffolds [36].
Crosslinking mechanism is another critical factor related to the mechanobiological properties of the bioprinted cell-laden hydrogel. Although common crosslinking methods including UV-photoinitiator, chemical reagents, enzymatic crosslinking are mostly used in the bioprinting of natural hydrogels with proper parameters, visible light-induced photocrosslinking has recently been in demand for the crosslinking of cell-laden gelatin-based hydrogels due to its non-damaging effect on the cell viabilities [37][38][39].
The visible-light crosslinking shows higher penetration depth and more potential for transdermal polymerization resulting in more uniform hydrogels [40,41]. In addition, it is cost-effective compared to enzymatic methods in the fabrication of hydrogel scaffolds [42]. Visible-light crosslinking using Ruthenium (Ru)/Sodium persulfate (SPS) systems have better cytocompatible results in comparison to UV crosslinking methods with cell viability of more than 85% [43]. The results were obtained at even higher Ru/SPS concentrations and light irradiation intensities and the structures were characterized for more than 21 d and showed less effect of oxygen inhibition compared to the other photo-crosslinking systems. The crosslinking process occurs in this way; Ru-metal complex is photolyzed with irradiation of visible light and photo-excited Ru 2+ oxidizes into Ru 3+ by donating an electron to SPS. At the same time, Ru 3+ reacts with aromatic residues including tyramine groups and tyrosine groups. Activated Ru 3+ complex extracts an electron from these aromatic groups, resulting in the formation of tyrosyl free radicals. The attack of these radicals to near tyramine and/or tyrosine moieties leads to covalent dityramine/tyrosine crosslinking bonds.
In situ crosslinking is one of the crosslinking strategies used in 3D bioprinting for photocrosslinkable non-viscous hydrogel inks, enabling the fabrication of complex and heterogeneous cellladen structures with high cell viability. Pre-and post-crosslinking methods, which means curing ink precursor before and after extrusion, can affect cell viability and printability by causing high and low extrusion forces, respectively. On the other hand, in situ crosslinking approach which includes light irradiation via photo-permeable capillary during extrusion prior to the deposition of the ink, enables stable hydrogel to be readily extruded on a variety of hydrogel types while lowering shear stresses that would otherwise lower cell viability [43,44].
In this study, we developed a novel 3D hybrid bioprinting method that uses visible-light + Ru/SPS crosslinking of tyramine-based hydrogel bioinks for mimicking OC interface. For this purpose, Alg and CMC were modified with tyramine and the crosslinking of mesenchymal stem cell (MSC)-laden bioink consisting of gelatin, Alg, and CMC occurred through gelatin's own tyrosine and tyramine groups on Alg  and CMC to enhance the bioactivity and mechanical properties of bioprinted structures for the cartilagebone tissue interface. To fabricate the narrow and fine interface region, a microcapillary extrusion-based bioprinting technique was developed via an in situ crosslinking approach. The biofabricated constructs were co-cultured in chondrogenic/osteogenic culture media for three weeks. Analyses of cartilage and bone formation were conducted based on cell alignment, histological assessment, and gene expression and the results showed that mechanical cues together with chemical cues lead to the differentiation of MSC into chondrogenic and osteogenic cells with a controlled OC interface. The mechanical properties of the two biomaterials were investigated with atomic force microscopy (AFM) and uniaxial compression tests. Our findings revealed that the printing structures had a high resolution in two distinct regions, indicating the potential of this approach for the biofabrication of OC tissue with a controlled interface. Figure 1 shows a schematic representation of the developed bioprinting method.

Modification of alginate and CMC with tyramine
The products of modification of alginate and CMC with tyramine were identified by Alg-Tyr and CMC-Tyr, respectively. Alg-Tyr was synthesized by a reaction of alginic acid with tyramine using EDC/NHS catalysis. Briefly, alginate (1% w/v) was dissolved in deionized (DI) water. After adding EDC (5 mM) and NHS (5 mM), and tyramine (2.5 mM) in that order, the solution was purged with nitrogen and stirred at room temperature overnight. The solutions were dialyzed against DI water for three days and then lyophilized [45]. CMC-Tyr was also produced by the same synthesis method as Alg-Tyr conjugate. After dissolving CMC (1% w/v) in DI water, EDC (5 mM), NHS (5 mM), and tyramine (6 mM) were added to the CMC solution, respectively. The solution was purged with nitrogen, and stirred overnight to occur the reaction. The CMC solution was dialyzed against 0.1 M NaCl solution, 25% ethanol, and DI water, in that order, for 2 d for each solution, and then lyophilized with freeze-dryer [15].

Preparation of Alg-Tyr-Gel and CMC-Tyr-Gel hydrogels
The final products were prepared by crosslinking of Alg-Tyr and CMC-Tyr together with gelatin, and denoted as Alg-Tyr-Gel, and CMC-Tyr-Gel, respectively. The hydrogel products with cells were casted for biochemical assays and mono-cultured histology test, while the hydrogels without cells were casted for uniaxial compression tests. The casted hydrogels were produced with Ru/SPS combination under visible light for mechanical properties evaluation. Briefly, gelatin was added at 8% w/v final concentration to the solutions of Alg-Tyr (0.5% w/v) and CMC-Tyr (0.05% w/v) in PBS. The solution was stirred until homogenization at 37 • C. Ru/SPS (0.1 mM/1 mM) coupling was added to the solution as a visible-light initiator system and the solution was transferred into a 2 ml syringe. The precursor solution in the syringe was exposed to visible light of around 425 nm at an intensity of 400 mW cm −2 for 10 min, and then hydrogel was removed from the syringe and sliced into 10 mm lengths.
As the casted hydrogels for uniaxial compression tests were prepared at a larger volume compared to the printed hydrogel structures (200 µl), we exposed them to light for a longer time to ensure the crosslinking was complete, although one minute was sufficient for crosslinking bioinks used in small amounts.

3D bioprinting of Alg-Tyr-Gel, CMC-Tyr-Gel hydrogels
Interface zone of the OC tissue was printed using customized NovoGen MMX bioprinter (Organovo, CA, USA) equipped with microcapillary cartridges. Briefly, the process was inspired by the method developed by Nadernezhad et al [46]. Glass microcapillary tube of 7.5 mm in length and 500 µm diameter was used to extrude the ink. Aspiration and dispensing of the ink were enabled by means of a plunger inside the tubes, which displaced upward and downward as the ink was extruded. Alg-Tyr-Gel and CMC-Tyr-Gel bioink solutions were used for the differentiation of cartilage and bone, respectively and were kept at 37 • C in the bioprinter's heater section before aspiration process. To achieve a transient interface, we extruded a specified length of previously aspirated bioink into the second bioink reservoir and then aspirated the second bioink reservoir as shown in figure 1. Using a computer model, the plunger moved upwards to aspirate the Alg-Tyr-Gel bioink (first ink) into the microcapillary glass. Microcapillary glass containing the first ink was then submerged in the CMC-Tyr-Gel ink (second ink) reservoir, and a desired amount of the first ink was released. Afterwards, the second bioink was aspirated into the microcapillary glass and subsequently formed a gradient bioink mixture inside the microcapillary tube. Following the aspiration of bioinks into the microcapillary glass, it was exposed to visible light for 15 s in order to make the liquid ink slightly viscous. Thus, the shape fidelity of the ink was achieved on the printing platform during the extrusion process. Then, the viscous bioink in the glass capillary was extruded on the platform at a 2 mm min −1 feed rate. Finally, the extruded bioink was completely crosslinked by exposing it to light for 1 min.

Printing parameters of the hydrogel structures
Different structures including circular, star, and grid shapes were fabricated to demonstrate the shape fidelity of the printed structures as follows: The inks were aspirated into the microcapillary glass using a plunger. The ink inside the microcapillary glass was then exposed to visible light for 15 s. The application of light exposure for 15 s is a part of the printing process for every hydrogel structure with and without cells to achieve in situ gelation. Aspiration and dispensing speeds were set to 2 mm min −1 . The circle structure of 6 mm diameter was printed as 2 concentric circles (figure 4(Ai)). The star shape with 5 mm arms was printed to indicate precise turns at the corners (figure 4(Aii)). For filament collapse assays, a single fiber was printed on discontinuous polydimethylsiloxane (PDMS) substrates with gaps of 2, 4, and 16 mm (figures 4(Bi) and (Bii)). Print fidelity was demonstrated by printing single fibers with various feature dimensions from 8 to 1 mm (figure 4(C)). The five-layer grid structure was printed at the size of 10 × 10 mm with a Z-axis difference of 400 µm (figure 4(D)).
To investigate structures with an interface, a mixture of two different fluorescent dyes (blue and red) were separately added to the two inks. The prepared inks were printed to form an interface of 0.5 mm. Moreover, for a sharp distinction between the two inks, a single fiber was fabricated without dispensing the first ink into the second one. Single (sharp and gradient interface) and multiline filaments (gradient) were aspirated with a 2 mm min −1 feed rate and their images were taken in a microcapillary glass, as a single fiber and multi-fiber structures crosslinked under visible light.

The morphological structures of the hydrogels
Morphological structures of the lyophilized gel samples were characterized using JEOL JSM 6010 scanning electron microscope (SEM). The samples were coated with Au-Pd under vacuum and scanned at an accelerating voltage of 3 kV and a working distance of less than 10 mm. SEM images were visualized with ImageJ software.

Fourier-transform infrared spectroscopy (FTIR) analysis
The infrared spectra of Alg-Tyr, and CMC-Tyr conjugate specimens were analyzed with Shimadzu FTIR spectrophotometer (Thermo-Nicolet iS10) equipped with an attenuated total reflectance accessory at room temperature. Their FTIR spectra were examined at a resolution of 4 cm −1 and in the wavenumber range of 600-4000 cm −1 .

Mechanical tests
Uniaxial compression measurements were performed on Alg-Tyr-Gel and CMC-Tyr-Gel hydrogels using a universal testing machine (Zwick Roell) equipped with a 200 N load cell. To show the effect of swelling on bioprinted hydrogels, the compression test was performed on both swollen and unswollen samples. The casted hydrogels were prepared in cylinder form with the dimensions of 9 mm in diameter and 10 mm in length for the compression tests. The measurements were conducted at a strain rate of 5 mm min −1 at 23 • C ± 2 • C up to 90% deformation. The linear region of the stress-strain curve between 4 and 6% strain was used to calculate Young's modulus, E of the specimens. In the strain-stress curves, nominal stress was plotted as a function of the deformation ratio.

AFM analysis
A gold-coated silicon AFM cantilever (PPP-NCLAuD) provided by Nanosensors (Neuchatel, Switzerland) was used for the stiffness measurements on the surface of bioprinted structures. The stiffness of the cantilever was measured at 38 ± 1 N m −1 using the thermal tuning method before attaching the glass bead. Firstly, the cantilever spring constant was measured by the thermal tuning method [47]. A glass bead with a diameter of 25 µm attached to the cantilever by a manual micropositioner and UV epoxy glue (Ossila E132). All the experiments were carried out in DI water and the measurements were taken at five nearby points. We captured force curves for each point and the Hertz model was employed to calculate Young's modulus, E from the force curves [48].

Cell culture and differentiation
Cryopreserved passage 2 hBMSCs were cultivated for two passages (P4) in the mesenchymal stem cell basal medium supplemented with mesenchymal stem cell supplement kit (Thermofisher Scientific), 10% FBS, and 1% antibiotics. The cells were incubated at 37 • C in a humid environment with 5% CO 2 . The culture medium was refreshed every three days and the cells were passaged using trypsin-EDTA when they were 80%-90% confluent. The amounts of 3 × 10 6 and 1 × 10 7 cells ml −1 were used for Alg-Tyr-Gel and CMC-Tyr-Gel hydrogels, respectively. Cellladen hydrogels were incubated for three weeks in different culture media. For osteogenic differentiation, they were treated with the osteocyte differentiation tool, and for chondrogenic differentiation they were treated with a medium composed of StemPro™ chondrogenesis/osteogenesis basal medium with a chondrogenesis differentiation kit (Thermofisher Scientific). Both differentiation media were supplemented with 10% FBS and 1% antibiotics. Bioprinted hydrogel structures were cultured for three weeks in a mixture of equal volumes of osteogenic and chondrogenic differentiation media and the mixture media were refreshed every three days.

Cell proliferation and morphology
The cell viability was performed on bioprinted CMC-Tyr-Gel hydrogel, while cellular morphology was made with both Alg-Tyr-Gel and CMC-Tyr-Gel hydrogel. The viability of cells was evaluated on days 1, 3, and 7 by staining with the Calcein-AM and PI to assess the cytotoxicity of the bioprinting processes. The bioprinted structures were dyed for 30 min at 37 • C with 1 M Calcein-AM (green fluorescence) and then for 5 min with 0.75 M PI (red fluorescence). To display the cellular morphology of the differentiated cells in the bioprinted structures, they were evaluated on the 7th, and 14th days of incubation by staining nuclei and F-actin. In brief, the bioprinted structures were fixated with 4% PFA for 60 min before being permeabilized with 0.1% Triton X-100 in PBS for 30 min. Alexa Fluor ® 546 Phalloidin was used to stain the Factin cytoskeleton for 60 min, and the nuclei of the cells were stained for 15 min in PBS with a (1:2000) ratio of DAPI.
Three-dimensional cellular images were taken by an inverted confocal microscope (Carl Zeiss LSM 710) to monitor cell viability and cellular morphology, with maximum excitation/emission wavelengths of 488/515 nm for viable cells, 561/625 nm for dead cells, 556/570 nm for f-actins, and 358/461 nm for nuclei staining.

Biochemical assays
To monitor differentiation of the structures, alkaline phosphatase (ALP) assay was used as an osteogenic marker, while sulfated glycosaminoglycan (sGAG) assay was utilized as a chondrogenic marker. ALP is an early osteoblastic differentiation marker [49]. GAGs are long disaccharides, which consist of negatively charged carboxylate/sulfate organizations that support chondrocytes and adhere to other matrix components. The DMMB assay is commonly used to determine the quantity of sGAG in MSCs during chondrogenic differentiation [50]. ALP activity was measured on 1, 7, 14, and 21 d after casted CMC-Tyr-Gel structures were lysate in M-PER ™ mammalian protein extraction reagent for 30 min at room temperature. The lysis solutions were centrifuged at 14 000 rpm for 15 min, and the supernatants were taken and incubated in a pNPP substrate at 37 • C for 60 min after incubation absorbance was measured at 405 nm. The total protein amount of each sample was determined using the BCA Protein test kit. BSA was used in a concentration range of 5-1000 g ml −1 to create protein standards. The standard concentrations were measured at 595 nm, and the results of the BCA test were used to normalize the ALP activity per unit of protein content.
Quantifying sGAG contents of casted Alg-Tyr-Gel hydrogels were examined on 1, 7, 14, and 21 d after using a DMMB assay with chondroitin-4-sulfate as a standard. Firstly, samples were lysed in 1 ml of 100 µg ml −1 protease K lysis buffer and the pH was adjusted to 7.5 using 10 mM Tris/HCl. Briefly, 16 mg of DMMB dissolved in 1 l DI water, followed by the addition of 95 ml of 0.1 M acetic acid, 1.6 g of 1 M NaCl, and 3.04 gr of Glycine, respectively. Casted samples were mixed with 1.25 ml DMMB dye for 30 min at room temperature and finally centrifuged at 10 000 rpm for 15 min at 16 • C. The 200 µl supernatant was withdrawn and the sample was measured at 595 nm. The standard curve was generated with chondroitin-4-sulfate serial dilution ranging from 0-800 µg ml −1 and after the addition of DMMB dye for each, concentration was measured at 525 nm.

Gene expression analysis
To observe the differentiation capacity of the hydrogel materials, we measured the expression of osteogenic and chondrogenic genes from samples of bone and cartilage. For osteogenic differentiation, we analyzed the expression of collagen type 1 (COL1), osteocalcin (OCN), and Runt-related transcription factor (RUNX2). For chondrogenic differentiation, we analyzed the expression of collagen type 2 (COL2), and aggrecan (ACAN). Gene expressions from the cultured cells were analyzed by semiquantitative reverse transcriptase polymerase chain reaction (PCR).

RNA isolation
Total RNA was extracted from bone and cartilage samples on Day 1 and Day 21 of differentiation. RNA was isolated with TRIzol™ Reagent according to the manufacturer's instructions. Briefly, 500 µl of TRIzol reagent was used for 50 µl of the OC interface hydrogel and pipetted until homogenization. After centrifugation of the lysate, the supernatant was incubated and 0.2 ml per 1 ml of TRIzol was added. After incubation and centrifugation, the aqueous phase was separated. Precipitation with isopropanol and ethanol yielded the RNA pellet, which was resuspended in 150 µl of molecular-grade water and stored at −80 • C until further analysis. The concentration of RNA was analyzed on an ND-1000 spectrophotometer (Nanodrop technologies).

Semiquantitative reverse transcriptase PCR
One microgram of RNA was used per PCR reaction. Semiquantitative reverse transcriptase PCR was performed with the Invitrogen EXPRESS One Step Superscript ® qRT-PCR mix for bone and cartilage. Non-fluorescent primers) were used (See table 1 for a list of primers). The one-step PCR reaction included 10 µl of EXPRESS SuperScript ® qPCR SuperMix Universal, 0.4 µl of 10 µM forward and reverse primers each, 2 µl of EXPRESS SuperScript Mix for One-Step qPCR, 1 µl of RNA template and water up to 20 µl. No-enzyme and no-template controls (RNA) were used. The PCR protocol had incubation at 50 • C for 15 min for initial cDNA reverse transcription followed by denaturation at 95 • C for 2 min and 40 cycles of 95 • C for 15 s, followed by annealing temperatures of 58 • C-65 • C depending on primers for 1 min, an extension at 72 • C for 40 s. The annealing temperatures varied for each primer pair, salt adjusted annealing temperature was 58 • C for glyceraldehyde 3-phosphate dehydrogenase (GAPDH), 60 • C for COL1, 62 • C for ACAN, RUNX2, and OCN, and 65 • C for COL2. The cycles were followed by cooling at 10 • C for 10 min and a final hold at 4 • C. For semi-quantitation of RNA expression cycles were run for 10, 20, 30, and 40 cycles for each gene. The samples for varying cycle runs were run on agarose gel electrophoresis for visualization and analysis using 90 volts for 30-45 min. To determine whether the bone samples exhibited cartilage markers and vice versa, the opposite markers were used to cross-compare the samples. Semi-quantitative PCR was conducted for 40 cycles, using identical controls. The samples were subsequently visualized using gel electrophoresis.
DNA ladder (GeneRuler 100 bp), 2% agarose gel in 1X TBE buffer, stained with GelRed Nucleic Acid Stain run at 70 V for 70 min was used for the electrophoretic separation to confirm the presence of the desired PCR products. Bio-Rad Gel Doc EZ Gel Documentation System (Bio-Rad, Hercules, CA, USA) was used for the gel visualization.

Histology staining
The bioprinted and casted cell-laden hydrogel structures were embedded in cryomedium for cryosectioning to collect histological sections after 21 d of incubation [51]. The specimens were fixed in a 4% PFA solution for 4 h. After cleaning the samples with PBS, they were embedded in cryomedium and stored at −80 • C overnight for frozen sectioning. Histological sections with a thickness of 5-10 µm were created by slicing the structure with a cryotome using standard protocols.
To show Ca 2+ deposition of the CMC-Tyr-Gel samples, casted cell-laden hydrogels were stained with %1 (w/v) Alizarin Red S (Sigma) with NH 4 OH adjusted pH 4.1 and samples were washed with 50% ethanol. Following sectioning, casted cell-laden Alg-Tyr-Gel hydrogels were stained with 1% w/v Alcian Blue in 3% v/v acetic acid to demonstrate chondrocyte proteoglycan formation [52]. After that, stained hydrogels were washed subsequently with a solution of 3% acetic acid, 25% ethanol, and 50% v/v ethanol. Gradient hybrid multiline bioprinted structures were stained with both prepared Alizarin Red and Alcian Blue staining to confirm the formation of the OC interface. The section of the bioprinted structures was first stained with Alcian Blue solution and cleaned with 50% ethanol to equalize ions charge and then stained with Alizarin Red S solution. Finally, inverted light microscopy was used to capture images of all sectioned hydrogels from both casted and bioprinted samples.

Statistical analysis
All the data for cell viability were presented as mean ± standard deviation and significant differences were analyzed by the one-way analysis of variance (ANOVA) method. P values less than 0.05 (P * < 0.05, P * * < 0.01) were considered as statistically significant.

Assessment of Alg-Tyr and CMC-Tyr with FTIR technique
Alg and CMC were conjugated with tyramine molecules using carbodiimide chemistry to acquire bioactivity. The carboxylic groups of Alg and CMC reacted with an amine group of tyramine in the presence of EDC/ NHS coupling.
The synthesis of Alg-Tyr and CMC-Tyr conjugate specimens was qualitatively analyzed by the FTIR technique. To confirm the tyramine conjugation of both alg and CMC, the infrared spectra of synthesized Alg-Tyr and CMC-Tyr were compared with that of Alg and CMC. Figures 2(Ai) and (Bi) show the FTIR spectra of Alg-Tyr and CMC-Tyr, respectively. FTIR spectrum of Alg displays typical absorption bands of carboxyl, ether, and hydroxyl groups. The peaks at the 3224, 2938, and 1600-1400 cm −1 were attributed to the vibration of OH bonds, the vibration of aliphatic CH bonds, and asymmetric and symmetric stretching vibrations of COO groups, respectively [53]. These bands also appeared in the FTIR spectrum of the Alg-Tyr precursor. A distinctive peak confirming the conjugation of tyramine was at 1516 cm −1 , which was related to the stretching vibration of aromatic rings of tyramine [54]. This peak was only in the spectrum of Alg-Tyr, while it did not appear in that of Alg. FTIR spectrum of CMC-Tyr precursor had similar typical absorption bands to those of Alg-Tyr, which corresponded to polysaccharides. The peak at 1516 cm −1 was also seen in the spectrum of CMC-Tyr, which is evidence of the conjugation of CMC with tyramine.

Morphological analysis
The pore morphology of cell-laden printed structures is crucial for the formation of the cellular habitat that encourages cell growth/ proliferation and migration. Therefore, the porous structures of Alg-Tyr-Gel and CMC-Tyr-Gel hydrogels were investigated with  the SEM technique. The molded hydrogels were subjected to SEM measurements both in unswollen and swollen states because of the change in their pore sizes after swelling. Figure 3 shows the SEM images of all the samples of Alg-Tyr-Gel and CMC-Tyr-Gel hydrogels at two different magnifications.
According to SEM results, pore diameters of Alg-Tyr-Gel and CMC-Tyr-Gel hydrogels increased from 8 ± 3 and 4 ± 2-410 ± 70 and 105 ± 20, respectively, after swelling because of their enlargement by entering water into the pores. When compared to CMC-Tyr-Gel hydrogels, the larger pore size of Alg-Tyr-Gel hydrogels before and after swelling tests may estimate further chondral differentiation specific collagen X content expression [55]. Besides, having a pore size of 200-700 µm after swelling can be related to higher levels of chondrocyte differentiation [56]. The heterogeneous distribution of CMC-Tyr-Gel hydrogel pore structures between 100-500 µm after swelling includes both the optimal pore size for bone growth (100-135 µm) and the vascularization pore size (>300 µm) required for long-term culturing [55]. The porous structures of all the samples have heterogeneous distributions. Before swelling, SEM analyses show that the pore difference between the Alg and CMC hydrogels is significant. Post-swelling hydrogel pore diameters, on the other hand, indicate that they have become suitable for cellular activities such as extracellular signal transmission or nutrient transportation.

Mechanical properties
For a successful scaffold treatment for OC tissue engineering, a physiochemical heterogeneous structure must be fabricated resembling the native transition of cartilage to bone tissue [57]. To this end, we investigated the mechanical properties of the two casted hydrogels of Alg-Tyr-Gel and CMC-Tyr-Gel by uniaxial compression tests at 23 • C ± 2 • C. To observe the effect of swelling for hydrogels, they were characterized in both swollen and unswollen conditions. Figure 3(E) and F present typical stress-strain curves of compression tests of the casted gels before and after swelling, respectively. Young's modulus, compressive strength, and fracture strain of the hydrogels are shown in a Table in figure 3(G). Young's moduli of both Alg-Tyr-Gel and CMC-Tyr-Gel hydrogels decreased from 4.71 and 8.41 kPa to 1.40 and 1.46 kPa, respectively, after equilibrium swelling because entangled polymer chains became expanded with penetration of water inside the gel network. It should be mentioned that the compressive modulus of the CMC-Tyr-Gel representing the bone part was 1.8-fold of the corresponding Alg-Tyr-Gel hydrogel for the cartilage. Their compressive strengths were also descended with the equilibrium swelling. On the other hand, Alg-Tyr-Gel hydrogel, which resisted up to 80% deformation in an unswollen state, was broken in 40% compressive strain after swelling, while CMC-Tyr-Gel hydrogel in swollen state sustained its resistance up to around 80% compressive strain. CMC-Tyr-Gel hydrogel exhibited mechanical stability by comparison to Alg-Tyr-Gel since the hydrogel including CMC had a lower swelling degree than those with Alg. In a similar study for OC interface regeneration that used agarose-hydroxyapatite, the compressive modulus of the scaffolds was measured for 2.9 ± 0.2 kPa [58]. Compressive strength of the CMC-Tyr-Gel was four-fold higher than that of Alg-Tyr-Gel. Considering that MSCs are sensitive to the mechanical properties of the substrate, our findings show that the stiffness of Alg-Tyr-Gel and CMC-Tyr-Gel could stimulate the differentiation to cartilage and bone, respectively.
Although stiffness does not solely determines cellular fate, a matrix with a stiffness of about 1 kPa stimulates both osteogenic and chondrogenic differentiation of human MSCs [59]. Moreover, mechanical stability is crucial especially for bone regeneration, since a mechanically unstable matrix can cause endochondral bone formation [60].

Printability
Printability is related to extrudability, shape fidelity, and filament integrity. In general, shape fidelity refers to a bioink's ability to retain its shape during and after deposition [61]. In this study, the dimensions of a single filament that can be printed in a single extrusion were limited by the length of the microcapillary glass tube which is 7.5 cm that was used for the in situ crosslink printing system.
Grid design, one of the most common designs, was used for this evaluation in our printed structures. Figures 4(Di) and (Dii) show the fabricated 10 × 10 mm grid structure with five layers and a 400 µm distance between the layers on the z-axis to allow the fusion and integration of layers. The extruded lines are rotated 90 • at each layer with a 2 mm gap between filaments.
A star shape was printed in addition to the grid structure to display fused segment length (fs) and filament thicknesses (ft) in printing, and a one-layered flowing pattern composed of parallel strands with decreasing gap distance was printed to investigate the filament fusion as shown in figures 4(Ai) and (Aii). The fusion between filaments was observed in the continuous printing in figure 4(C) according to a previously published study [23] starting from 8 mm spacing, in spite of that no differences were found in filament segment length (fs) as the filament distance decreased.
Filament deformation or rupturing during printing due to the hydrogel's weight and/or not fully crosslinked hydrogel cause printing challenges [38]. To analyze the printability of our in situ gelation bioprinting technique, filament collapse assay was performed at various gap distances. In this assay, a structure is printed on a supporting substrate having different gap lengths. The corresponding angles formed between the unsupported filament and supporting substrates and the hydrogel deformation on the gap of the supporting substrate indicate printability. A single printed hydrogel filament (figures 4(Bi) and (Bii)) showed almost zero collapse angle and no significant fractures or defects on the gap of the support structure up to 16 mm.
The aspiration-extrusion printing method was used to create a gradient hybrid interface. Sharp distinct and gradient interface areas were created with the different aspirating methods of the CMC-Tyr-Gel and Alg-Tyr-Gel inks materials which were marked with fluorescent dye prepared separately (figures 4(Ei) and (Fi)). As shown in figure 4(Fii), a gradient filament was fabricated successfully in microcapillary glass. Therefore, this design was selected as an appropriate environment for the differentiation of MSCs which can provide an improved perspective on the complex gradient of native OC units [62].

Cell viability and morphology
To show the effect of Ru/SPS crosslinking on bioprinted structures, cell viability of CMC-Tyr-Gel hydrogels were investigated. For this purpose, cell-laden CMC-Tyr-Gel hydrogels with 1 × 10 7 cell ml −1 were stained with Calcein-AM and PI to obtain cell viability after 1, 3, and 7 d of bioprinting. Green staining in 1, 3, and 7 d images indicates live cells, and red spots with red circles indicate dead cells. Cell viability was 85 ± 2.52% on day 1 (figure 5(Ai)) and then increased to 95.76 ± 0.72% on day 3 (figure 5(Aii)). No significant changes were observed on day 7 with 93.81 ± 8.02% cell viability ( figure 5(Aiii)). Therefore, the cell viability assay revealed no harmful effects and no cytotoxicity because of reactive by-products of the Ru/SPS catalyzed reaction during gelation. This study indicates that the Ru/SPS mediated crosslinking mechanism is appropriate for Biopolymers-Tyramine conjugates applications in a bioprinting approach. Previous studies in the literature proved that 0.1/1 mM and even higher concentration of Ru/SPS showed no harmful or cytotoxic effect on different cell types including MSCs [37,43,63].
Morphological and adherent properties of the cells such as osteoblasts and chondrocytes, strongly determine their functional behavior and fate during stem cell differentiation. Particularly, culturing chondrocytes in a 3D environment rather than a 2D surface affects rounded cell morphology [64]. CMC-Tyr-Gel and Alg-Tyr-Gel bioprinted hydrogels were cultured separately to assess cellular morphology and phenotype differences were evaluated using DAPI-Phalloidin staining. Cell nuclei were stained blue with DAPI staining, while cellular filamentous actions were stained red with Phalloidin staining at the 7th and 14th-day cultures of specimens.
According to the confocal images of both bioprinted hydrogels on the 7th day as seen in figure 5(C), the cells maintained their MSC-like morphology and did not show significant differences in terms of phenotypic differences. In the following 14th day culture, it was observed that the cells in the CMC-Tyr-Gel hydrogel cultured with bone exchange medium had aligned fibrillary and were similar to the natural osteonal structure. In contrast, rounded morphology cells in Alg-Tyr-Gel hydrogel images cultured with cartilage medium clearly exhibit more of their native chondrogenic morphology. Therefore, cell-laden bioprinted hydrogels cultured in special differentiation media induced cell differentiation into osteogenic and chondrogenic phenotypes by promoting various material properties, such as stiffness and swelling [65,66].

Biochemical and histological evaluation
As an early osteoblastic differentiation marker, ALP is an essential protein that is expressed during the calcification process. ALP enzyme activity was measured in casted CMC-Tyr-Gel hydrogels containing MSCs to show their differentiation at 7, 14, and 21 d. CMC-Tyr-Gel hydrogels cultured in mesenchymal stem cell medium without differentiation medium (Ctrl) were used as a control group. Compared to the control group, ALP activation in osteogenic differentiated was significantly increased by 1.2 and 1.85 10 5 U mg −1 until day 7 and day 14, respectively as shown in figure 6(A). On the 21st day, the ALP activation in CMC-Tyr-Gel decreased to 1.4 U mg −1 . By associating ALP enzyme activation with early osteoblasts, the decrease in the activation measured on the 21st day can be considered as an indication that osteoblasts have passed into the later phase. As a result of ALP enzyme activations, cellular response to CMC-Tyr-Gel hydrogel has been shown to be suitable for bone tissue differentiation.
As chondrocytes are surrounded by ECMs such as sGAGs, the presence of sGAG is crucial for regeneration of cartilage [67]. The amount of sGAG was determined on the 7th, 14th, and 21st days using a casted Alg-Tyr-Gel hydrogel in chondrogenic differentiation medium. Chondrogenic differentiation was compared with a control group cultured with a mesenchymal stem cell medium over the same period. As seen in figure 6(B), compared to day 7, the rate of sGAG formation increased almost 3-fold and 5fold on days 14 and 21, respectively. On the other hand, sGAG levels in control MSCs remained constant in all measurements until 21 d and were very poor in comparison to the differentiation medium. sGAG contents values have shown that the Alg-Tyr-Gel hydrogel can provide a suitable environment for chondrogenic differentiation. Moreover, it exhibited a certain potential increase depending on culture time.
The chondrogenic and osteogenic regions have different ECM components such as higher sGAGs and calcium nodules, respectively. To identify these components, histological analysis is usually preformed on the chondrogenic and osteogenic regions by Alcian Blue /Alizarin Red S staining [68,69]. Alizarin Red S, an anionic dye, turns to red color in the presence of calcium, thus allowing the observation of osteocytes in differentiation. After mono cultured osteogenic induction for 21 d, CMC-Tyr-Gel casted hydrogels were stained with Alizarin Red S to analyze ECM mineralization as shown in figure 6(Ci) and (Cii). According to Alizarin Red S staining results, the ECM of CMC-Tyr-Gel hydrogel exhibited osteoblastic characteristics of hMSC after 3 weeks of osteogenic incubation. Besides this, hMSC-laden hydrogels were stained with anionic Alcian Blue dye for chondrogenic identification. After 21 d of chondrogenic incubation, the Alg-Tyr-Gel hydrogels, which were stained by Alcian Blue dye, demonstrated high sGAG expression in figures 6(Di) and (Dii).

Gene expression of the bone and cartilage samples
Differentiation of the MSC-laden hydrogels was characterized using semi-quantitative PCR for the samples on days 1 and 21 of culture. Expression of the osteogenic and chondrogenic markers was evaluated using specified primers and final results were visualized using gel electrophoresis (Shown in figures 7(A) and (B)). For osteogenic genes, OCN and COL1 visible gene expression was seen after 40 cycles. Comparing day 1 and day 21 a thick band was observed on day 21 showing a higher gene expression as compared to day 1. This was consistent in OCN and RUNX2. RUNX2 is a critical transcription factor in osteoblast differentiation. RUNX2 showed an expression on day 1 and day 21 showing its constant expression before differentiation. However, the observation of bands at 30 cycles showed a higher expression compared to day 1. For COL 1 almost no expression was observed on day 1 and a visible band on day 21 showed expression after treatment. Our results show the differentiation of the cells towards osteogenesis.
For chondrogenic genes, ACAN and COL2 visible expression was observed on day 21 of treatment and none at day 1 showing differentiation towards cartilage formation. The results indicate an upregulation of ACAN and COL2 in differentiated samples at day 21 compared to control samples at day 1 of culture. Considering the expression rate of both cartilage and bone specific genes in three weeks, our hydrogels showed a promising potential for OC interface.
To fully demonstrate the behavior of the interface after differentiation of cells, the samples were subjected to cross-comparison, wherein the expression of RUNX2, OCN, and COL1 for cartilage samples, and ACAN, and COL2 for bone samples were investigated. A semi-quantitative PCR experiment with 40 cycles was conducted, with RNAand Enzserving as controls. The results were visualized using gel electrophoresis (figures 7(C) and (D)). As expected, the chondrogenic portion did not exhibit any expression of RUNX2. The expression of OCN was observed to be lower in the chondrogenic part compared to the osteogenic part, which is consistent with its known role in bone strength [70]. Interestingly, COL1 has been observed to be expressed in distinct regions of differentiated cartilage parts, suggesting that a subset of MSCs may undergo osteogenic differentiation, leading to the formation of bone tissue at the OC interface.
Gene expression analysis of OCN and ACAN in bone samples exhibited a similar pattern, with minimal expression observed at Day 21 of differentiation. These results support the aforementioned hypothesis of the presence of cartilage in the bone portion. Additionally, limited expression of COL2 expression was detected in bone sample, potentially indicating the presence of early stages of cartilage calcification [71]. These cross-comparisons indicate that the transition from bone to cartilage is not a distinct transition, but rather a continuous and gradual change.

Interface evaluation
The co-cultured bioprinted hydrogel constructs were stained with Alizarin Red S and Alcian Blue dyes, which signify osteogenic and chondrogenic differentiations, respectively, in order to further evaluate the capability of ECM accumulation and differentiate MSCs into two lineages. Figure 8(A) gives serial multiline printed continuous gradient interface structure of 2 cm. Increasing gradient concentration of Alizarin Red S dye through right side of bioprinted structures shows that the CMC-Tyr-Gel supports osteogenic differentiation ( figure 8(Bi)). At the same time, more intense staining of left side of the structure with Alcian Blue dye, demonstrates that Alg-Tyr-Gel promotes chondrogenic differentiation ( figure 8(Bii)).  As seen in the histology-stained images of cocultured bioprinted structures, the chondrogenic and osteogenic differentiations are seen in both Alginate and CMC parts of multiline bioprinted structure, which is natural because the whole multiline printed structure with stem cells is cultured in a mixture of equal volumes of osteogenic and chondrogenic differentiation basal medium. The important point here is that the Alg part of the multiline structure induced more chondrogenic differentiation compared to the CMC part, which stimulated osteogenic differentiation. It was clearly proved that CMC-Tyr-Gel and Alg-Tyr-Gel hydrogel bioinks promoted the MSCs to differentiate successfully into osteogenic and chondrogenic tissue in 21 d.
The micro-mechanical characteristic of the bioprinted OC interface, which was co-cultured for 21 d, were performed with AFM. To avoid the difficulty of locally measuring the mechanical differentiation of the bioprinted structures, AFM indentation was made using a glass bead attached cantilever, and the deviation was sensed by a photodiode. Glass bead diameter was selected 25 µm roughly as a cell diameter to better understand how a cell can affect the hydrogel surface locally. The compressive strength was measured from various points on the bioprinted structures. Young's modulus calculated from this measurement was plotted against the distance on gradient bioprinted structure in figure 8(C). As seen in the graph, Young's moduli gradually increased from 9.54 to 19.46 kPa as progressed from cartilage to bone region on the bioprinted structure. The mechanical gradient characteristic of the bioprinted structures was seen obviously after chondrogenic and osteogenic differentiation of MSCs on the structures cultured in 21 d.
As shown in figure 3(G), Young's moduli of swollen hydrogels with alginate and CMC are 1.40 and 1.46 kPa, respectively, while their Young's moduli increased to around 10 and 20 kPa after 21 d of chondrogenic and osteogenic differentiation ( figure 8(C)). In other words, the mechanical properties of our bioprinted structures enhance with increasing cell culture time of OC differentiation. A number of studies have already been reported showing that the mechanical strength of cell-laden structures increases during differentiation culture time [72,73]. In conclusion, we can say that our printed hydrogel is a suitable matrix for chondrogenic and osteogenic differentiation considering its mechanical properties and its gradient structure allowing the fabrication of the OC interface, which relies on culture time. Our study was a demonstration of how we could fine-tune and alter the mechanical properties of the interface and obtain a modular heterogeneous structure at a small scale.

Conclusion
Despite advancements in cell-laden hydrogels, the fabrication of hybrid structures for soft and hard bone tissue interfaces based on material selection is still extremely limited. Herein, for the first time, we developed an in situ crosslinked multi-material 3D bioprinting approach based on an aspirationextrusion microcapillary system for the fabrication of an OC interface with a controlled gradient structure. In situ crosslinking technique was used to create a gradient hybrid structure that can mimic the physiochemical characteristics of the OC interface. MSC-laden CMC-Tyr-Gel and Alg-Tyr-Gel hydrogels induced osteogenic and chondrogenic differentiation after three weeks of culture that demonstrate their suitability for osteogenesis and chondrogenesis. The gradient hybrid bioprinted structures, which were co-cultured in osteogenic/chondrogenic culture media, were analyzed histologically. The results showed that mechanical and biological cues associated with chemical indications lead to the differentiation of MSC into chondrogenic and osteogenic cells with a controlled gradient interface. The high stability of the printed structures with controlled and gradient regions demonstrated the potential of this approach for controlled OC interface application. The developed approach can easily be adapted for other tissue interfaces by changing the cells and hydrogels.

Data availability statement
All data that support the findings of this study are included within the article (and any supplementary files).