**1. Introduction**

Articular cartilage disorder most commonly occurs at the conjunction between bones, and the condition is progressively worsened by constant and unavoidable mechanical degeneration. The loss of the normal cartilage tissue will lead to more serious joint diseases like osteoarthritis [1]. Cartilage reconstruction has been a clinical issue for decades because of poor intrinsic ability to repair defects and lacking of specific diagnostic biomarkers. Several types of clinical treatments and pharmacologic therapies are available and effective in reducing pain and increasing mobility of patients [2–4]. Nevertheless, those treatments are merely a temporary relief and unable to restore the damaged tissue into its original function [5]. The long-term clinical solutions for cartilage repair are still in demand, and diverse regenerative therapies are consequently being brought to the table.

The hydrogel-based scaffolds have received a big interest as providing a temporary three-dimensional structure [6–9]. High functional, strong biomechanical properties and long-term biocompatible scaffolds can be made by regulating different materials to promote the cartilage tissue therapies [10,11]. Scaffolds which allow physical supporting often combines with primary chondrocytes therapies to overcome low cell proliferation or dedifferentiation problems. However, these 3D networks still struggle with several challenges such as cell viability, growth factor burst release or low oxygen content. On the other hand, particles serve as a superior medicine approach. Particles have been widely developed for drug delivery systems in tissue engineering applications due to their wide variety and highly regulated potential [12]. In particular, microscale particles were popular among delivering anti-inflammatory drugs or growth factors and hold promise in performing as building block in scaffold for cartilage repair [13,14]. Cruz et al. [15] aggregated gelatin microparticles and chondrocytes into 3D pellets and found higher cell viability over long periods.

As one of the major components in chondrocytes extracellular matrix (ECM), hyaluronic acid (HA) had a linear polysaccharide structure which functioned as a structural element, providing backbones for the component distribution and aggrecan aggregation. HA was also well known to interact with specific receptors including CD44 to regulate signal transduction, cell migration and differentiation [16,17]. Gelatin was a denatured protein from collagen and widely used in drug delivery systems. Because of gelatin's biocompatibility, low antigenicity, chemical modification possibility and low-cost [18], many studies had encapsulated growth factor in modified gelatin-based particles to stimulate chondrogenesis [8,19,20]. Moreover, gelatin-based microparticles as building blocks for three-dimensional (3D) structures have been wildly used in cartilage engineering [13].

In addition to chemical stimulations and highly cartilage-like materials, mechanical forces acting as additional tools were also applied to improve cartilage reconstruction in recent studies to mimic in vivo environments. The proper biophysical stimulations were able to increase proteoglycan synthesis and cartilage-specific gene expression [21,22]. Magnetic nanoparticles are thought of as excellent candidates to apply remote magnetic induced physical stimulation, which also holds the capability of targeting a specific site. Nathalie et al. [23] labeled mesenchymal stem cells (MSC) with magnetic nanoparticles to enhance seeding density and condensation in scaffold. Together with a dynamic bioreactor, MSC differentiation performance was markedly improved. Son et al. [24] exposed bone-marrow-derived human MSC (BM-hMSC) to static magnetic field and magnet-derived shear stress via magnetic nanoparticle. Without hypertrophic effect, biophysical stimulation led BM-hMSC to higher chondrogenic differentiation efficiency.

We developed a brand-new platform, cartilage tissue-mimetic pellets, to combine biochemical and biophysical treatments to mimic native cartilage tissue. As illustrated in Figure 1, cartilage tissuemimetic pellets were composed by rabbit primary chondrocytes and hyaluronic acid-graft-amphiphilic gelatin microcapsules (HA-AGMCs). HA polymer chains on the microcapsules surface are expected to expand space between each microcapsule with its highly hydrophilic and polyanionic characteristics. In addition, HA can enhance chondrocytes attachment through CD44 receptors and act stabilize the pellets structure as ECM component at the beginning of formation. We encapsulated superparamagnetic iron oxide nanoparticles (SPIOs) in hydrophobic shells of HA-AGMCs to guide cells and serve physical stimulations by applying static magnetic field and magnet-derived shear stress. The inner hydrophilic space of microcapsules is capable of encapsulating growth factor or biomolecules for cell proliferation or repair. Such HA-AGMC approaches can be used to provide cartilage structure stability, rule biochemical and biophysical stimulations, and thereby promote faster and more complete cartilage reconstruction.

**Figure 1.** Synthesis of the hyaluronic acid-graft-amphiphilic gelatin microcapsules (HA-AGMCs) structure to fabricate cartilage tissue-mimetic pellets with combined biochemical and biophysical **Figure 1.** Synthesis of the hyaluronic acid-graft-amphiphilic gelatin microcapsules (HA-AGMCs) structure to fabricate cartilage tissue-mimetic pellets with combined biochemical and biophysical treatments.

#### treatments. **2. Materials and Methods**

#### **2. Materials and Methods**  *2.1. Materials*

*2.1. Materials*  Gelatin from porcine skin (type A 300 bloom), hexanoic anhydride, absolute ethanol (99.5%), sodium hydroxide, 2,4,6-Trinitrobenzene Sulfonic Acid (TNBS), 1,2-hexadecanediol (97%), oleic acid (90%), oleylamine (>70%), and iron(III) acetylacetonate (Fe(acac)3) were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA) Hyaluronic acid (Hyalo-Oligo) was purchased from Tannmer Enterprise Co., Ltd. (New Taipei City, Taiwan). N-hydroxysuccinimide (NHS) and 1-ethyl-3-(3 dimethylaminopropyl) carbodiimide hydrochloride (EDC) were purchased from Echo Chemical Co., Gelatin from porcine skin (type A 300 bloom), hexanoic anhydride, absolute ethanol (99.5%), sodium hydroxide, 2,4,6-Trinitrobenzene Sulfonic Acid (TNBS), 1,2-hexadecanediol (97%), oleic acid (90%), oleylamine (>70%), and iron(III) acetylacetonate (Fe(acac)3) were purchased from Sigma-Aldrich Co. (St. Louis, MO, USA) Hyaluronic acid (Hyalo-Oligo) was purchased from Tannmer Enterprise Co., Ltd. (New Taipei City, Taiwan). N-hydroxysuccinimide (NHS) and 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) were purchased from Echo Chemical Co., Ltd. (Miaoli, Taiwan). Platinum® PCR SuperMix and GScript First-Strand Synthesis Kit were purchased from Thermo Fisher Scientific (Waltham, MA, USA).

#### Ltd. (Miaoli, Taiwan). Platinum® PCR SuperMix and GScript First-Strand Synthesis Kit were purchased from Thermo Fisher Scientific (Waltham, MA, USA). *2.2. Synthesis of Hyaluronic Acid-Graft-Amphiphilic Gelatin (AG-g-HA)*

*2.2. Synthesis of Hyaluronic Acid-Graft-Amphiphilic Gelatin (AG-g-HA)*  Gelatin powder (3 g) was completely dissolved in deionized water (40 mL) at 70 °C with constant stirring for an hour. Eethanol (95%, 30 mL) and hexanoic anhydride (3 mL) were sequentially added dropwise in succession and stirred for 4 h. In this process, the amphiphilic gelatin (AG) form and was tuned to a pH value around 7. The final solution was dialyzed against a mixture of water and ethanol (3:4). AG was collected and dried at 60 °C and then ground into powder with grinding machine. Hyaluronic acid (HA) was grafted to the amino group of the gelatin via EDC/NHS method. HA (1 g) was first dissolved in phosphate buffer solution (40 mL) under stirring. 1-ethyl-3-(-3 dimethylaminopropyl) carbodiimide hydrochloride (EDC, 1 g) and N-hydroxysuccinimide (NHS, 1.9 g) were added to the solution and stirred for one hour at pH 5.5. After the pH was adjusted to 7 with NaOH (10 N), AG (1 g) was added directly and stirred for 2 h. After the reaction, AG-g-HA formed and dialyzed (Spectra/Por, MWCO = 20000) against water before dried by a freeze vacuum dryer. Gelatin powder (3 g) was completely dissolved in deionized water (40 mL) at 70 ◦C with constant stirring for an hour. Eethanol (95%, 30 mL) and hexanoic anhydride (3 mL) were sequentially added dropwise in succession and stirred for 4 h. In this process, the amphiphilic gelatin (AG) form and was tuned to a pH value around 7. The final solution was dialyzed against a mixture of water and ethanol (3:4). AG was collected and dried at 60 ◦C and then ground into powder with grinding machine. Hyaluronic acid (HA) was grafted to the amino group of the gelatin via EDC/NHS method. HA (1 g) was first dissolved in phosphate buffer solution (40 mL) under stirring. 1-ethyl-3-(-3-dimethylaminopropyl) carbodiimide hydrochloride (EDC, 1 g) and N-hydroxysuccinimide (NHS, 1.9 g) were added to the solution and stirred for one hour at pH 5.5. After the pH was adjusted to 7 with NaOH (10 N), AG (1 g) was added directly and stirred for 2 h. After the reaction, AG-g-HA formed and dialyzed (Spectra/Por, MWCO = 20000) against water before dried by a freeze vacuum dryer. AG-g-HA was analyzed by nuclear magnetic resonance spectroscopy (VARIAN, UNIYTINOVA 500 NMR) with 600-MHz 1H-NMR. Each copolymer (10 wt %) was dissolved in D2 O to obtain 1 H NMR and 13 C NMR spectra.

#### AG-g-HA was analyzed by nuclear magnetic resonance spectroscopy (VARIAN, UNIYTINOVA 500 NMR) with 600-MHz 1H-NMR. Each copolymer (10 wt %) was dissolved in D2 O to obtain 1 H NMR *2.3. Quantification of HA Grafting Rate*

and 13 C NMR spectra. *2.3. Quantification of HA Grafting Rate*  We used 2,4,6-Trinitrobenzene Sulfonic Acid (TNBS) reagent to determine the content of free We used 2,4,6-Trinitrobenzene Sulfonic Acid (TNBS) reagent to determine the content of free amino groups. TNBS (0.025% *w*/*v*) was dissolved in sodium hydrogen carbonate solution (4%) serving as reaction buffer. Gelatin, AG, and AG-g-HA were dissolved in dH2O (1 mL), and mixed with reaction buffer (0.5 mL) separately. Calibration curve was made by mixing Lysine (2–20 µg in 1 mL dH2O)

amino groups. TNBS (0.025% *w*/*v*) was dissolved in sodium hydrogen carbonate solution (4%) serving as reaction buffer. Gelatin, AG, and AG-g-HA were dissolved in dH2O (1 mL), and mixed with reaction buffer (0.5 mL). After incubating at 37 ◦C for two hours, sodium dodecyl sulfate (SDS, 10% *w*/*v*) and 1 N HCl were added to each sample. An absorbance peak at 336 nm was measured with UV-vis spectroscopy (UV-vis, Evolution 300, Thermo, Waltham, MA, USA). The number of free amines contained in each polymer was quantified by correlating with the calibration curve.
