*2.12. HA-AGMCs Cellular Attachment E*ffi*ciency*

To measure the precise amount of HA-AGMCs attached on chondrocytes, we examined the element content of Fe by inductively coupled plasma mass spectrometry (ICP-MS). To prepare the sample, we collected 1 mL medium in vials after chondrocytes mixing with various amounts of HA-AGMCs and seeded on hydrogel overnight. Hydrochloric acid (12 M), which can totally corrode the cell sample and redox iron in HA-AGMC, was mixed in all the samples. The content of Fe element was quantified by ICP-MS.

#### *2.13. RNA Isolation, cDNA Synthesis and Quantitative PCR Analysis*

The procedure of RNA isolation was bottomed on Molecular Research Center, Inc. In short, chondrocytes pellets were washed with PBS buffer and added 1mL of RNAzol® RT reagent to lyse cells. Subsequently, added DEPC-treated water (0.4 mL) to lysate and shook the mixture vigorously for 15 s. Centrifuged samples for 15 min at 13,500 rpm at 4 ◦C after storing for 15 min. The RNA remained soluble in the supernatant. Each of the supernatant (1 mL) was transferred to a new tube

and precipitated by mixing with 0.4 mL of 75% ethanol. The samples were centrifuged for 10 min at 12000 rpm at 4 ◦C after stored for 15 min. RNA precipitated and formed a white pellet at the bottom of the tube. Next, we washed the RNA twice with 75% ethanol to remove the ethanol and dissolve the RNA precipitation pellet with DEPC-treated water.

Complementary DNA was performed for reverse transcription using the GScript First-Strand Synthesis Kit (GeneDireX). Specific cDNA was amplified by PCR using Platinum® Blue PCR SuperMix. The protocol of experiment method followed Invitrogen manufacturer. cDNA samples were mixed up with Platinum® Blue PCR SuperMix and forward/reverse primer. The PCR amplification condition was illustrated as follows: heating up to 95 ◦C for 5 min, denaturing the DNA at 95◦C for 30 s, annealing with the primer at 55 ◦C for 30 s, extending the length of product at 72 ◦C for 30 s, and finally cycled for 35 times. PCR cycle ended at 72 ◦C for 3 min and cooled down to 4 ◦C. The samples were stored at −20 ◦C and electrophoresis experiment ran at voltage 80 V, 35 min. The experiment was also conducted in three times.

#### *2.14. Alcian Blue Staining*

Alcian blue solution was prepared beforehand by dissolving Alcian blue 8GX powder in 40% acetic acid and 60% ethanol mixture solvent in 1 wt % concentration, and the solution was stored at 4 ◦C. At the end of the culture period, cell pellets were washed with PBS buffer and fixed with 4% formaldehyde for 1 h. The fixed-cell samples were washed with PBS buffer and then incubated with Alcian blue solution overnight. The stained samples were washed with PBS buffer and mounted by mounting solution. The observation was using optical microscope. The experiment was run in triplicate and the results were statistically analyzed using two-way ANOVA Dunnett's multiple comparisons test comparing with the control group (\* *P* < 0.05, \*\* *P* < 0.01, \*\*\* *P* < 0.001, and \*\*\*\* *P* < 0.0001).

#### *2.15. Static Magnetic Field (MF) and Magnet-Derived Shear Stress (S) Stimulations*

A cylinder-shaped neodymium magnet with a magnetic field of 0.22 T was used to static magnetic field induction [24,28]. On the other hand, a magnetic stirrer of 50 rpm was applied for magnet-derived shear stress stimulation on the cell-pellets. For combined stimulation, cell-pellets were first placed on the top of the magnet and then transferred to the magnetic stirrer on pellets in consecutive order following static magnetic field and/or magnet-derived shear stress for 1 h one day for five consecutive days

#### **3. Results and Discussion**

#### *3.1. Synthesis and Characterization of AG-g-HA*

Figure 2A illustrated the reaction scheme for the synthesis of Hyaluronic acid-grafted amphiphilic gelatin (AG-*g*-HA). As a widely used biopolymer, gelatin is composed of a series of amino acids, serving as hydrophilic backbone and modified by hexanoic anhydride to gain amphiphilic characteristic. HA was conjugated on gelatin's Arginine (Arg) part by the well-known EDC/NHS method. The chemical signals of the gelatin (Figure S1 and Table S1) from peak 1 to peak 11 assigned to the protons on the primitive gelatin macromolecules [29]. According to the NMR spectra in Figure 2B, chemical signals of amphiphilic gelatin (AG) at peak 3.5 ppm and peak 1.2 ppm were the protons from the hexanoyl group. The disappearance of peak at 2.9 ppm which was referred to the primary amino group of the gelatin confirmed that amino group, arginine, was partially substituted to hexanoyl group forming AG. The α-carbonyl group (COCαH3) at 1.9 ppm and glucose group between 3.2 and 3.9 ppm gave the cue of glycosylation (Figure 2C). TNBS is a well-known reagent specific for primary amino groups, which can be quantified and measured at absorbance 335 nm [30]. The substitution rate of the hexanoyl group was measured to be 58.2%. The second HA conjugation rate was 4.0% (Figure S2). The results demonstrated that there were finally 37.8% free amino groups remaining in the arginine and lysine of the gelatin where hexanoyl group and HA substitution were at 58.2% and 4.0%, respectively.

*Polymers* **2019**, *11*, x FOR PEER REVIEW 7 of 16

**Figure 2.** (**A**) The reaction scheme for the synthesis of hyaluronic acid-grafted amphiphilic gelatin (AG-g-HA); 1H NMR spectrum of macromolecules in D2O for (**B**) amphiphilic gelatin (AG) and (**C**) AG-g-HA. As shown, the NMR spectra at 3.5, 1.2, 1.9 and around 3.5 ppm demonstrated the successful synthesis of AG-g-HA. **Figure 2.** (**A**) The reaction scheme for the synthesis of hyaluronic acid-grafted amphiphilic gelatin (AG-g-HA); <sup>1</sup>H NMR spectrum of macromolecules in D2O for (**B**) amphiphilic gelatin (AG) and (**C**) AG-g-HA. As shown, the NMR spectra at 3.5, 1.2, 1.9 and around 3.5 ppm demonstrated the successful synthesis of AG-g-HA.

#### *3.2. Characterization of HA-AGMCs and Cartilage Tissue-Mimetic Pellets*

*3.2. Characterization of HA-AGMCs and Cartilage Tissue-Mimetic Pellets*  Hyaluronic acid-graft-amphiphilic gelatin microcapsule (HA-AGMC) was synthesized via double emulsification using SPIOs and AG-g-HA. The diameter and zeta potential of HA-AGMCs were measured by dynamic light scattering (DLS) (Table S2). HA-AGMC exhibited an average size of 1.24 ± 0.1 μm in diameter with an excellent monodispersity (Figure S3). The surface charge was approximately −16 mV caused by the replacement of positively charged amino groups to hexanoyl group and HA. HA-AGMCs demonstrated the appearance of deflated balloons structure in scanning electron microscope (SEM) images (Figure 3A). The round hollow morphology can be further confirmed by transmission electron microscopy (TEM) image in Figure 3B where the little dark spots of SPIOs were clearly observed in the shell of microcapsules. The loading efficiency of SPIOs in HA-Hyaluronic acid-graft-amphiphilic gelatin microcapsule (HA-AGMC) was synthesized via double emulsification using SPIOs and AG-g-HA. The diameter and zeta potential of HA-AGMCs were measured by dynamic light scattering (DLS) (Table S2). HA-AGMC exhibited an average size of 1.24 ± 0.1 µm in diameter with an excellent monodispersity (Figure S3). The surface charge was approximately −16 mV caused by the replacement of positively charged amino groups to hexanoyl group and HA. HA-AGMCs demonstrated the appearance of deflated balloons structure in scanning electron microscope (SEM) images (Figure 3A). The round hollow morphology can be further confirmed by transmission electron microscopy (TEM) image in Figure 3B where the little dark spots of SPIOs were clearly observed in the shell of microcapsules. The loading efficiency of SPIOs in HA-AGMCs was 92.2% determined by ultraviolet-visible spectroscopy (UV-vis).

AGMCs was 92.2% determined by ultraviolet-visible spectroscopy (UV-vis). The biocompatibility of HA-AGMC with rabbit primary chondrocytes was investigated by MTS assay (Figure 3C). The cells were treated with different concentrations of HA-AGMC from 0.05 to 13.6 mg mL−1. HA-AGMC illustrated outstanding biocompatibility at high concentration in which The biocompatibility of HA-AGMC with rabbit primary chondrocytes was investigated by MTS assay (Figure 3C). The cells were treated with different concentrations of HA-AGMC from 0.05 to 13.6 mg mL−<sup>1</sup> . HA-AGMC illustrated outstanding biocompatibility at high concentration in which chondrocytes viability was higher than 84% after 24 h co-culture with HA-AGMC under 6.8 mg mL−<sup>1</sup> .

chondrocytes viability was higher than 84% after 24 h co-culture with HA-AGMC under 6.8 mg mL−1. In this study, hyaluronic acid (HA) on HA-AGMC surface was designed to help sticking chondrocytes and microcapsules together. The CD44 antigen was the main receptor for hyaluronic acid, which was responsible for cell proliferation, differentiation and migration [16,17,31]. Hyaluronic acid conjugated on the surface of microcapsule was capable of binding to the chondrocytes due to CD44/HA receptors and serving as a backbone to recruit proteoglycans and glycoproteins into extracellular matrix structures [32]. Attachment efficiency of HA-AGMCs to chondrocytes was used to evaluate the formation of cartilage tissue-mimetic pellets by co-culturing chondrocytes with HA-AGMCs on non-adhesive micro-molds (Figure 3D). The Fe amount was also quantified by using ICP-MS and increased with HA-AGMCs concentration. The attachment efficiency was over 90% in each HA-AGMCs group with the microcapsule concentration of 42.5 to 680 μg mL−1. The result In this study, hyaluronic acid (HA) on HA-AGMC surface was designed to help sticking chondrocytes and microcapsules together. The CD44 antigen was the main receptor for hyaluronic acid, which was responsible for cell proliferation, differentiation and migration [16,17,31]. Hyaluronic acid conjugated on the surface of microcapsule was capable of binding to the chondrocytes due to CD44/HA receptors and serving as a backbone to recruit proteoglycans and glycoproteins into extracellular matrix structures [32]. Attachment efficiency of HA-AGMCs to chondrocytes was used to evaluate the formation of cartilage tissue-mimetic pellets by co-culturing chondrocytes with HA-AGMCs on non-adhesive micro-molds (Figure 3D). The Fe amount was also quantified by using ICP-MS and increased with HA-AGMCs concentration. The attachment efficiency was over 90% in each HA-AGMCs group with the microcapsule concentration of 42.5 to 680 µg mL−<sup>1</sup> . The result demonstrated that mostly HA-AGMCs can easily attach to chondrocytes and form pellets together.

demonstrated that mostly HA-AGMCs can easily attach to chondrocytes and form pellets together.

**Figure 3.** Morphological appearance and characterization of the HA-AGMCs. (**A**) SEM and (**B**) TEM images. (**C**) Biocompatibility of HA-AGMCs to chondrocytes after 24 h incubation (n = 3). (**D**) Attachment efficiency of HA-AGMCs to chondrocytes in different concentrations of HA-AGMC. The **Figure 3.** Morphological appearance and characterization of the HA-AGMCs. (**A**) SEM and (**B**) TEM images. (**C**) Biocompatibility of HA-AGMCs to chondrocytes after 24 h incubation (n = 3). (**D**) Attachment efficiency of HA-AGMCs to chondrocytes in different concentrations of HA-AGMC. The results demonstrated HA-AGMCs have a spherical-shaped bilayers structure with average diameter of 1.2 µm and negligible cytotoxicity to chondrocytes.

results demonstrated HA-AGMCs have a spherical-shaped bilayers structure with average diameter of 1.2 μm and negligible cytotoxicity to chondrocytes. We fabricated cartilage tissue-mimetic pellets with different concentrations of microcapsule HA-AGMCs and also used macromolecule AG-g-HA as control group. The number of viable chondrocytes and the morphology of cartilage tissue-mimetic pellets were examined by MTS assay and optical microscope at day 7 and 14 respectively. HA-AGMCs and AG-g-HA showed little difference in optical density (O.D.) over the cultural period at different concentrations from 42.5 to 680 μg mL−1 (in AG-g-HA concentration) (Figure 4A,B). The results illustrated that the HA-AGMCs microcapsules containing SPIOs do not lower the number of chondrocytes. Of note, after 14 days of culture, HA-AGMCs at the concentration of 170 μg mL−1 had the relative highest cell number correlated to the optical microscope images (Figure 4A). The morphology of pellets in different HA-AGMCs concentrations after culturing 14 days was shown (Figure 4C). Pellets in each group showed We fabricated cartilage tissue-mimetic pellets with different concentrations of microcapsule HA-AGMCs and also used macromolecule AG-g-HA as control group. The number of viable chondrocytes and the morphology of cartilage tissue-mimetic pellets were examined by MTS assay and optical microscope at day 7 and 14 respectively. HA-AGMCs and AG-g-HA showed little difference in optical density (O.D.) over the cultural period at different concentrations from 42.5 to 680 µg mL−<sup>1</sup> (in AG-g-HA concentration) (Figure 4A,B). The results illustrated that the HA-AGMCs microcapsules containing SPIOs do not lower the number of chondrocytes. Of note, after 14 days of culture, HA-AGMCs at the concentration of 170 µg mL−<sup>1</sup> had the relative highest cell number correlated to the optical microscope images (Figure 4A). The morphology of pellets in different HA-AGMCs concentrations after culturing 14 days was shown (Figure 4C). Pellets in each group showed uniform size and rounded shape, and more microcapsules made the pellets looked darker responding to more SPIOs. During the culture period, a scatter of cell debris was found in the group with 680 mL−<sup>1</sup> HA-AGMCs, which resulted in a smaller size of pellet. In contrast, HA-AGMCs at the concentration of 170 µg mL−<sup>1</sup> had the biggest pellet size of 200 µm. Therefore, the HA-AGMCs at 170 µg mL−<sup>1</sup> was selected for the following all experiments.

μg mL−1 was selected for the following all experiments.

uniform size and rounded shape, and more microcapsules made the pellets looked darker responding to more SPIOs. During the culture period, a scatter of cell debris was found in the group with 680 mL−1 HA-AGMCs, which resulted in a smaller size of pellet. In contrast, HA-AGMCs at the concentration of 170 μg mL−1 had the biggest pellet size of 200 μm. Therefore, the HA-AGMCs at 170

**Figure 4.** Cartilage tissue-mimetic pellets proliferation and morphology. The relative cell proliferation after culturing chondrocytes with HA-AGMCs microcapsule or AG-g-HA at different concentrations (from 42.5 to 680 μg mL<sup>−</sup>1) at (**A**) 7 days and (**B**) 14 days (n = 4). (**C**) The cell pellet morphologies after culturing 14 days showed that HA-AGMCs at the concentration of 170 μg mL<sup>−</sup>1 had the biggest pellet **Figure 4.** Cartilage tissue-mimetic pellets proliferation and morphology. The relative cell proliferation after culturing chondrocytes with HA-AGMCs microcapsule or AG-g-HA at different concentrations (from 42.5 to 680 µg mL−<sup>1</sup> ) at (**A**) 7 days and (**B**) 14 days (n = 4). (**C**) The cell pellet morphologies after culturing 14 days showed that HA-AGMCs at the concentration of 170 µg mL−<sup>1</sup> had the biggest pellet size of 200 µm. Scale bar = 50 µm.

#### size of 200 μm. Scale bar = 50 μm. *3.3. Cartilage Tissue-Mimetic Pellets Live*/*Dead Assay*

viability in cartilage tissue-mimetic pellets.

*3.3. Cartilage Tissue-Mimetic Pellets Live/Dead Assay*  We tested long-term cell viability with Live/Dead assay to confirm cell condition, and to investigate whether HA-AGMCs had any influence on the cartilage tissue-mimetic pellets. In 3D culture, the supplement of oxygen was mainly controlled by diffusion, consequently resulting in oxygen gradient. Oxygen tension was often found in the center of the 3D structure and also dependent on the number of cells. The nutrient and oxygen supply in the central of pellets would be increasingly inadequate during the culture period [33,34]. In this study, fluorescence images showed after 14 days, the vast majority of cells remain viable in both pure cells control group and HA-AGMCs group (Figure 5A,B). The mean fluorescence intensity in each HA-AGMCs pellet showed more cells and higher viability compared to cell only group as shown in Figure 5C. In comparison to clear boundary between each cell in control group (Figure 5A), HA-AGMC group showed blurry edge, indicating cell pellets in HA-AGMC group had stronger connection to ECM (Figure 5B). This further confirmed that HA-AGMC are able to promote the connection between cells and construct chondrocytes into a compact pellet. To observe the localization of HA-AGMC, quantum dots were We tested long-term cell viability with Live/Dead assay to confirm cell condition, and to investigate whether HA-AGMCs had any influence on the cartilage tissue-mimetic pellets. In 3D culture, the supplement of oxygen was mainly controlled by diffusion, consequently resulting in oxygen gradient. Oxygen tension was often found in the center of the 3D structure and also dependent on the number of cells. The nutrient and oxygen supply in the central of pellets would be increasingly inadequate during the culture period [33,34]. In this study, fluorescence images showed after 14 days, the vast majority of cells remain viable in both pure cells control group and HA-AGMCs group (Figure 5A,B). The mean fluorescence intensity in each HA-AGMCs pellet showed more cells and higher viability compared to cell only group as shown in Figure 5C. In comparison to clear boundary between each cell in control group (Figure 5A), HA-AGMC group showed blurry edge, indicating cell pellets in HA-AGMC group had stronger connection to ECM (Figure 5B). This further confirmed that HA-AGMC are able to promote the connection between cells and construct chondrocytes into a compact pellet. To observe the localization of HA-AGMC, quantum dots were dissolved in chloroform with SPIOs and encapsulated in the shell of HA-AGMC. The confocal fluorescent images showed cooperative formation of HA-AGMC and chondrocytes at 14 days (Figure 5D), HA-AGMCs still remained in the pellets and scattered uniformly in the ball-shaped pellets. Our studies demonstrated that microcapsule HA-AGMCs can maintain great cell compatibility and viability in cartilage tissue-mimetic pellets.

dissolved in chloroform with SPIOs and encapsulated in the shell of HA-AGMC. The confocal fluorescent images showed cooperative formation of HA-AGMC and chondrocytes at 14 days (Figure 5D), HA-AGMCs still remained in the pellets and scattered uniformly in the ball-shaped pellets. Our studies demonstrated that microcapsule HA-AGMCs can maintain great cell compatibility and

*Polymers* **2019**, *11*, x FOR PEER REVIEW 10 of 16

**Figure 5.** Confocal fluorescent image of cartilage tissue-mimetic pellets after 14 days culture. Live/Dead fluorescent image of (**A**) pure cells control group and (**B**) cells with HA-AGMCs group. Calcein AM (green), Ethidium homodimer-1 (red). (**C**) The fluorescence intensity in each pellet was measured by ImageJ and further analyzed by unpaired *t*-test (*P* = 0.0123, n = 3). (**D**) Quantum dots were encapsulated in the shell of HA-AGMCs and used to visualize the localization of cartilage tissuemimetic pellets consisting of DAPI (blue), actin (green), HA-AGMC (red quantum dots). As shown, chondrocytes remained viable in HA-AGMCs group and had stronger connection to ECM. Scale bar **Figure 5.** Confocal fluorescent image of cartilage tissue-mimetic pellets after 14 days culture. Live/Dead fluorescent image of (**A**) pure cells control group and (**B**) cells with HA-AGMCs group. Calcein AM (green), Ethidium homodimer-1 (red). (**C**) The fluorescence intensity in each pellet was measured by ImageJ and further analyzed by unpaired *t*-test (*P* = 0.0123, n = 3). (**D**) Quantum dots were encapsulated in the shell of HA-AGMCs and used to visualize the localization of cartilage tissue-mimetic pellets consisting of DAPI (blue), actin (green), HA-AGMC (red quantum dots). As shown, chondrocytes remained viable in HA-AGMCs group and had stronger connection to ECM. Scale bar = 25 m.

#### = 25 m. *3.4. Gene expression of Cartilage Tissue-Mimetic Pellets under Physical Stimulations*

*3.4. Gene expression of Cartilage Tissue-Mimetic Pellets under Physical Stimulations*  HA-AGMC was designed as a multifunctional platform to guide the cartilage tissue-mimetic pellets and serve physical stimulations. Static magnetic field (MF) and magnet-derived shear stress (S) were applied on the HA-AGMCs one hour each day for five consecutive days. In addition to mimic cartilage ECM environment, joint movements mechanical forces also played a crucial part in growth and development of articular cartilage tissue. It was found that stem cell performed better properties when cells grew in scaffold under external mechanical stimulations [23,35,36]. More recently, magnetic nanoparticles captured more attention to bioreactor as they were capable of providing different mechanical forces and were more suitable for applying loading to improve cell condensation and scaffold seeding efficiency [37–39]. We performed biophysical stimulations via SPIOs and analyzed the gene expression of Aggrecan (Agg), collagen type I (Col I), collagen type II (Col II), and Sox9 with polymerase chain reaction (PCR) after 7 and 21 days of culture. The primer sequences were shown in Table S3 and housekeeping genes GAPDH were used in comparison with the samples. After 7 days culture period, the presence of HA-AGMCs significantly increased Col I, Col II and Sox9 gene expression compared to control group. HA-AGMC+MF+S and control group showed similar Agg gene expression (Figure 6A). The gene expression of Col II and Agg gave us an indication of chondrocytes' functionality as they attributed to the secretion of ECM to stabilize the structure of ECM [40]. It is worth noting that Sox9 gene regulated chondrogenesis and chondrocyte proliferation [41], and the expressions of Sox9 were dramatically upregulated 2-fold in all HA-AGMCs-added groups after 7 days. Taking up-regulation of Agg, Col II, and Sox9 contrasting to Col I gene expression altogether, HA-AGMC-treated group demonstrated a slightly better chondrocyte expansion. The HA-AGMC with physical stimulation-treated group showed significant Agg gene expression compared to control group after 21 days culture period (Figure 6B). Furthermore, gene expression of both Col I and Col II revealed significant difference between HA-AGMC+MF+S and control group. Taking the results together, the application of HA-AGMCs and biophysical stimulation did not generate dedifferentiation effect on chondrocytes due to the relatively lower expression of Col I than any other group during culture period. Furthermore, HA-AGMC+MF+S HA-AGMC was designed as a multifunctional platform to guide the cartilage tissue-mimetic pellets and serve physical stimulations. Static magnetic field (MF) and magnet-derived shear stress (S) were applied on the HA-AGMCs one hour each day for five consecutive days. In addition to mimic cartilage ECM environment, joint movements mechanical forces also played a crucial part in growth and development of articular cartilage tissue. It was found that stem cell performed better properties when cells grew in scaffold under external mechanical stimulations [23,35,36]. More recently, magnetic nanoparticles captured more attention to bioreactor as they were capable of providing different mechanical forces and were more suitable for applying loading to improve cell condensation and scaffold seeding efficiency [37–39]. We performed biophysical stimulations via SPIOs and analyzed the gene expression of Aggrecan (Agg), collagen type I (Col I), collagen type II (Col II), and Sox9 with polymerase chain reaction (PCR) after 7 and 21 days of culture. The primer sequences were shown in Table S3 and housekeeping genes GAPDH were used in comparison with the samples. After 7 days culture period, the presence of HA-AGMCs significantly increased Col I, Col II and Sox9 gene expression compared to control group. HA-AGMC+MF+S and control group showed similar Agg gene expression (Figure 6A). The gene expression of Col II and Agg gave us an indication of chondrocytes' functionality as they attributed to the secretion of ECM to stabilize the structure of ECM [40]. It is worth noting that Sox9 gene regulated chondrogenesis and chondrocyte proliferation [41], and the expressions of Sox9 were dramatically upregulated 2-fold in all HA-AGMCs-added groups after 7 days. Taking up-regulation of Agg, Col II, and Sox9 contrasting to Col I gene expression altogether, HA-AGMC-treated group demonstrated a slightly better chondrocyte expansion. The HA-AGMC with physical stimulation-treated group showed significant Agg gene expression compared to control group after 21 days culture period (Figure 6B). Furthermore, gene expression of both Col I and Col II revealed significant difference between HA-AGMC+MF+S and control group. Taking the results together, the application of HA-AGMCs and biophysical stimulation did not generate dedifferentiation effect on chondrocytes due to the relatively lower expression of Col I than any other group during culture period. Furthermore, HA-AGMC+MF+S group performed the highest Col II gene expression level

throughout the culture period, despite of HA-AGMC+S with sound Col II gene expression at the early stage. These data indicated that HA-AGMC can help chondrocytes present in cartilage tissue-specific gene at the beginning of pellets formation, and somewhat maintain functional gene expression better after applying both static magnetic field and magnet-derived shear stress. However, further studies are needed to optimize the mimic native cartilage environment in terms of biophysical parameters. group performed the highest Col II gene expression level throughout the culture period, despite of HA-AGMC+S with sound Col II gene expression at the early stage. These data indicated that HA-AGMC can help chondrocytes present in cartilage tissue-specific gene at the beginning of pellets formation, and somewhat maintain functional gene expression better after applying both static magnetic field and magnet-derived shear stress. However, further studies are needed to optimize the mimic native cartilage environment in terms of biophysical parameters.

**Figure 6.** Gene expressions of cartilage tissue-mimetic pellets under static magnetic field (MF) and magnet-derived shear stress (S) stimulations at (**A**) day 7 and (**B**) day 21. The results were statistically analyzed using two-way ANOVA Dunnett's multiple comparisons test comparing with the control group (\* *P* < 0.05, \*\* *P* < 0.01, \*\*\* *P* < 0.001, and \*\*\*\* *P* < 0.0001). HA-AGMC helped chondrocytes maintain cartilage tissue-specific gene at the beginning of pellets forming and applying both MF and S hold the gene expression better. **Figure 6.** Gene expressions of cartilage tissue-mimetic pellets under static magnetic field (MF) and magnet-derived shear stress (S) stimulations at (**A**) day 7 and (**B**) day 21. The results were statistically analyzed using two-way ANOVA Dunnett's multiple comparisons test comparing with the control group (\* *P* < 0.05, \*\* *P* < 0.01, \*\*\* *P* < 0.001, and \*\*\*\* *P* < 0.0001). HA-AGMC helped chondrocytes maintain cartilage tissue-specific gene at the beginning of pellets forming and applying both MF and S hold the gene expression better.

#### *3.5. Synthesis of Sulfated Glycosaminoglycan under Physical Stimulations 3.5. Synthesis of Sulfated Glycosaminoglycan under Physical Stimulations*

*3.6. Histological Analysis* 

The synthesis of sulfated glycosaminoglycan (sGAG) is one important index of chondrocyte biochemical function [41,42], and the diminishing presence of sGAG indicated a tendency to dedifferentiate into fibrochondrocytes [43]. The cartilage tissue-mimetic pellets productivity of sGAG under physical simulations was examined by Blyscan assay (Figure 7A). The sGAG production increased with incubation time. Physical stimulations further enhanced cartilage tissue-specific ECM production, and HA-AGMC+MF+S group exhibited the greatest sGAG secretion in medium after 21 days culture (Figure 7B). Alcian blue staining results further demonstrated that both and HA-AGMCs group had high content of sGAG excretion in comparison with pure cells control group in Figure 7C where, staining results were not further quantified because of the SPIOs interference. The synthesis of sulfated glycosaminoglycan (sGAG) is one important index of chondrocyte biochemical function [41,42], and the diminishing presence of sGAG indicated a tendency to de-differentiate into fibrochondrocytes [43]. The cartilage tissue-mimetic pellets productivity of sGAG under physical simulations was examined by Blyscan assay (Figure 7A). The sGAG production increased with incubation time. Physical stimulations further enhanced cartilage tissue-specific ECM production, and HA-AGMC+MF+S group exhibited the greatest sGAG secretion in medium after 21 days culture (Figure 7B). Alcian blue staining results further demonstrated that both and HA-AGMCs group had high content of sGAG excretion in comparison with pure cells control group in Figure 7C where, staining results were not further quantified because of the SPIOs interference. *Polymers* **2019**, *11*, x FOR PEER REVIEW 12 of 16

**Figure 7.** The cartilage tissue-mimetic pellets productivity of sGAG. (**A**) Blyscan assay detected sGAG content in culture medium under different stimulations at 7 and 21 day (n = 3). Alcian blue staining demonstrated the secretion of sGAG after 21 days culture of (**B**) control and (**C**) HA-AGMCs. scale bars = 200 μm. **Figure 7.** The cartilage tissue-mimetic pellets productivity of sGAG. (**A**) Blyscan assay detected sGAG content in culture medium under different stimulations at 7 and 21 day (n = 3). Alcian blue staining demonstrated the secretion of sGAG after 21 days culture of (**B**) control and (**C**) HA-AGMCs. scale bars = 200 µm.

implanted with a magnet as a source of magnetic force as shown in Figure 8A. In this in-vivo experiment, the HA-AGMCs are pre-cultured with chondrocytes before implantation into rabbit OA knee. Subsequently, after 6 weeks of surgery, the rabbits received implants of pellets containing either

After 4 weeks of implantation, the preliminary results revealed that control groups without HA-AGMCs displayed irregular layer of cartilage surface and degeneration of the cartilage tissue (as illustrated in Figure 8B). In addition, the abnormal growth and disorder of cartilage tissues were observed in pure cell without magnet and HA-AGMCs group shown in Figure 8C. More importantly, a newly formed tissue is evident in the group treated with a combination of chondrocytes with A-AGMCs and magnet in Figure 8D, which demonstrated that the cells treated with HA-AGMCs and magnetic force can improve the retention and biofunctionality of transplanted chondrocytes to form ordering arrangement in cartilage matrix, which is very important for cartilage repair. However, the comprehensive investigation about chondrogenic analysis and cartilage repair is in subsequent

only cells or HA-AGMCs with cells or nothing as control groups.

progress, which will be reported in the future.

The cell pellets with HA-AGMCs at the concentration of 170 μg mL−1 was used for animal
