Dexamethasone loaded bilayered 3D tubular scaffold reduces restenosis at the anastomotic site of tracheal replacement: in vitro and in vivo assessments

Sang Jin Lee§ ab, Ji Suk Choi c, Min Rye Eom c, Ha Hyeon Jo a, Il Keun Kwon b, Seong Keun Kwon *cd and Su A Park *a
aDepartment of Nature-Inspired Nanoconvergence Systems, Korea Institute of Machinery and Materials, 156 Gajeongbuk-ro, Yuseong-gu, Daejeon 34103, Republic of Korea. E-mail:; Tel: +82-42-868-7969
bDepartment of Dental Materials, School of Dentistry, Kyung Hee University, 26 Kyungheedae-ro, Dongdaemun-gu, Seoul 02447, Republic of Korea
cDepartment of Otorhinolaryngology-Head and Neck Surgery, Biomedical Research Institute, Seoul National University Hospital, Seoul, Republic of Korea. E-mail:; Tel: +82-2-2072-2286
dDepartment of Otorhinolaryngology-Head and Neck Surgery, Seoul National University College of Medicine, Seoul, Republic of Korea

Received 6th December 2019 , Accepted 17th January 2020

First published on 31st January 2020

Despite recent developments in the tracheal tissue engineering field, the creation of a patient specific substitute possessing both appropriate mechanical and biointerfacial properties remains challenging. Most tracheal replacement therapies fail due to restenosis at the anastomosis site. In this study, we designed a robust, biodegradable, 3D tubular scaffold by combining electrospinning (ELSP) and 3D (three-dimensional) printing techniques for use in transplantation therapy. After that, we loaded dexamethasone (DEX) onto the 3D tubular scaffold using mild surface modification reactions by using polydopamine (PDA), polyethyleneimine (PEI), and carboxymethyl–β-cyclodextrin (βCD). As a result, the fabricated 3D tubular scaffold had robust mechanical properties and the chemical modifications were confirmed to have proceeded successfully by physico-chemical analysis. The surface treatments allowed for a larger amount of DEX to be loaded onto the βCD modified scaffold as compared to the bare group. In vitro and in vivo studies demonstrated that the DEX loaded 3D tubular scaffold exhibited significantly enhanced anti-inflammation activity, enhanced tracheal mucosal regeneration, and formation of a patent airway. From our results, we believe that our system may represent an innovative paradigm in tracheal tissue engineering by providing proper mechanical properties and successful formation of tracheal tissue as a means of remodeling and healing tracheal defects for use in transplantation therapy.

1. Introduction

Tracheal tissue is prone to loss by unexpected trauma, industrial accidents, cancer, idiopathic diseases, congenital anomalies, and iatrogenic causes.1–3 Tracheal tissue damage happens regardless of gender or age. In these instances, rapid reestablishment of a patent airway is of paramount importance in the clinic.2 When tracheal stenosis occurs, surgeons remove the diseased, stenotic segment of the trachea and pull down the proximal and distal segments of the remaining trachea to secure together by anastomosis.4 However, this kind of surgery is impossible in cases where the diseased segment of the trachea is too long (6 cm or more), therefore a tissue engineering approach is needed.

Until now, to address these issues, many biomedical engineering researchers have developed implantable, decellularized trachea matrix (DTM) from human donor tissue.5,6 Although satisfactory results were reported in a very limited number of cases after transplantation, a great deal of cost and time is required for preparation of DTM. Often, post-transplantation interventions, such as multiple stent insertion and bronchoscopic toileting, are needed to avoid restenosis after implantation.7,8 Apart from the expense, the long preparation time may impede the clinical applications of the tracheal scaffold as patients need immediate replacement of the airway. Furthermore, DTM cannot provide an exact match in terms of size and mechanical properties for patient-specific airways.9

As an alternative solution, tissue-engineered scaffolds have gained much attention in recent years due to their unique properties of biodegradation, biocompatibility, and robust mechanical properties.2 Even though various scaffolds with stem cells have been developed, multiple steps, such as in vitro pre-culture and in vivo heterotopic-implantation and subsequent orthotopic transfer, make them clinically unattractive options.10,11 Even after long term preparation before orthotopic transplantation, regenerative treatments using tracheal scaffolds often fail due to restenosis. The trachea is a multi-layered complex structure which contains the pseudostratified ciliated columnar epithelium, submucosa, mature cartilaginous rings, and adventitia (from the inner layer to outer layer).12–14 Most of the previous studies have focused on regenerating the tracheal cartilage by mimicking its structure. However, without a strategy to achieve early regeneration of the inner respiratory epithelium, the tracheal tissue engineering will fail by restenosis. Such a transplant is unlikely to succeed and is expensive and impractical. Thus, a hybrid rapidly-manufacturable tracheal scaffold should be investigated. This can provide for both rapid transplantation into the patient and also promote regeneration of the tracheal tissue after transplantation.

In this study, we developed a dexamethasone (DEX) loaded, mussel-inspired, three-dimensional (3D) bilayered scaffold by combining electrospinning (ELSP) and 3D printing techniques (Fig. 1). In our previous reports, we designed a mechanically potent bio-tubular scaffold by combining an ELSP and a 3D printing system for artificial graft applications.15–17 Based on these findings, we expected that a PCL scaffold of the conduit type can be applied to tracheal replacement.17 PCL is a semi-crystalline, aliphatic polyester approved by the Food and Drug Administration (FDA), which has good biocompatibility and is biodegradable for use in medical devices.18,19 It also has been well utilized in investigations of tracheal reconstruction.8,17 Even though the developed 3D PCL scaffold provides for an attractive conduit structure for use in tracheal replacement, it does not have an optimal environment for regenerating tracheal tissue. In order to improve tissue regeneration, we loaded DEX directly onto the 3D tubular scaffold. DEX is a synthetic glucocorticoid used clinically as an anti-inflammatory drug.20 We hypothesized that restenosis would be alleviated by suppressing inflammation. Additionally, DEX is able to induce cartilage regeneration21 and airway epithelial repair.22,23 DEX has low aqueous solubility on account of it being a lipophilic substance, which is problematic for loading onto a scaffold.24,25 Therefore, a special mediator is required to place the DEX safely on the scaffold.

image file: c9nr10341d-f1.tif
Fig. 1 Schematic illustration of the preparation procedure to generate bilayered 3D tubular PCL scaffolds with surface modification for use in tracheal replacement. Steps: (i) 3D tubular conduit formation by combining ELSP and 3D printing system, (ii) surface modification through PDA, PEI, and βCD under aqueous conditions, (iii) DEX loading onto PCLDPβ scaffold.

In the present study, we designed a robust 3D tubular scaffold followed by mild surface treatments using biocompatible chemicals to load DEX for application in tracheal replacement therapy. Our developed 3D scaffolds were characterized by physico-chemical analysis. During in vitro evaluations, we carried out cell adhesion tests on scaffolds with human bronchial epithelial cells (hBEC) and anti-inflammatory response test with activated macrophages. Finally, a pre-clinical in vivo animal test was performed in a tracheal defected rabbit model for 4 weeks to confirm epithelium and cartilage regeneration.

2. Materials and methods

2.1 Materials

PCL (average MW = 45 kDa for 3D printing, 80 kDa for ELSP), Trizma® hydrochloride (reagent grade, ≥99.0% (titration), crystalline), 3-hydroxytyramine hydrochloride (dopamine hydrochloride), poly(ethyleneimine) solution (average Mn ∼ 1200, average Mw ∼ 1300 by LS, 50 wt% in H2O), carboxymethyl–β-cyclodextrin sodium salt, and dexamethasone (≥98% (HPLC), powder) were purchased from Sigma-Aldrich (St Louis, MO, USA). 1,1,1,3,3,3-Hexafluoro-2-propanol (HFIP, >99.0%), 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC), and N-hydroxysuccinimide (NHS) were purchased from TCI (Tokyo Chemical Industry CO., Ltd, Japan). Ethyl alcohol anhydrous (EtOH, 99.9%) was purchased from DaeJung (Chemical& Metals Co. Ltd, Korea). Deionized-distilled water (DDW) was produced by an ultrapure water system (Puris-Ro800; Bio Lab Tech., Korea). All other reagents and solvents were of analytical grade and used without further purification.

2.2 Fabrication of the 3D tubular PCL scaffold by combining ELSP and 3D printing systems

The fabrication of tubular PCL electrospun nanofibrous scaffold (ENs) was performed using ELSP as previously described with some modification.15–17 Briefly, PCL (MW = 80 kDa) was dissolved in HFIP solvent to produce a 10 wt% solution. For ELSP, the dissolved solution was loaded into a luer-lock syringe attached to a 20 gauge blunt-tip needle and electrospun on a rotating windmill mandrel of 7 mm diameter at 20 kV using a high-voltage DC power supply (Nano NC, Korea) with a 1 ml h−1 feed rate (KDS-200, KD Scientific Inc., USA) and a 15 cm needle tip-to-collector distance. Mandrel speed was 30 rpm. Finally, the resultant ENs were dried overnight under vacuum to remove any residual solvent. After that, tubular ENs were detached from the mandrel. Prior to depositing the PCL strand using the 3D printing system, tubular ENs were placed on the 3D printing mandrel. Then, 3D printing was performed by directly depositing onto tubular PCL ENs. This 3D printing process was performed as previously described.8,15–19,26 Briefly, 3D PCL strands were prepared using a 3D printing system (laboratory-made system at the Korea Institute of Machinery and Materials). The 3D printing instrument consisted of a 3D printing system equipped with a three-axis translation stage, software system, dispenser, nozzle, and compression/heat controller. PCL polymers (MW = 45 kDa) were melted at 100 °C in a heated dispenser. The nozzle size and distance between strands were 300 and 900 μm, respectively. After the PCL was melted, a continuous air pressure of 300 kPa was applied to the dispenser, and strands of molten PCL were applied layer-by-layer onto the ENs. The 7 mm drum holding the EN film was rotated between layers to produce a patterned 0°/45° degree porous structure (Fig. 2a).
image file: c9nr10341d-f2.tif
Fig. 2 Image of fabricated scaffold (a) and SEM images for characterization of surface morphology of 3D tubular scaffolds (“*” denotes electrospun nanofibrous layer and “**” denotes 3D printed layer) (b) Photograph of native rabbit trachea tissue after removal (c) characterization of mechanical property of fabricated scaffolds using compressive tests (d and e). Significant difference: **p < 0.01 and ***p < 0.001 as compared with the native tissue (e).

For in vitro tests, 2D membrane-type scaffolds were manufactured under the same conditions previously described. Briefly, an electrospun membrane was collected on a metal plate in order to fabricate a 2D electrospun membrane, and then the 3D printing was performed on the 2D electrospun membrane for manufacturing a 2D membrane-type scaffold. After cutting out the 2D membrane-type scaffold with a 10 mm diameter punch, the surface-modification reactions were performed under the same conditions as described below.

2.3 Measurement of mechanical properties using compression test

The compressive strength of the native rabbit trachea tissue and various types of scaffolds were measured using a compression test by a micro-load® system (R&D, Daejeon, Korea).15,16,27 The tubular samples evaluated using a 50 kgf load cell with a displacement rate of 1 mm min−1 at room temperature.

2.4 Surface modification of bare 3D PCL scaffold in order to load DEX

In order to load the DEX onto the 3D PCL scaffold, we carried out surface modification (Fig. 1). First, the 3D PCL scaffolds were immersed in dopamine hydrochloride solution (2 mg mL−1 in 10 mM of Tris buffer, pH 8.5) at room temperature. After 1 h, the PCLD scaffolds were washed with fresh DDW 3 times.16 Second, PCLD scaffolds were directly immersed in PEI solution (20 mg ml−1 in 10 mM of Tris buffer, pH 8.5) at room temperature to apply amine groups onto the scaffolds. After 2 h, the PEI-grafted PCLD scaffolds (PCLDP) were washed with deionized water. Third, the carboxymethyl–βCD solution was prepared for βCD grafting onto PCLDP scaffold. Carboxymethyl–βCD (10 mg ml−1) was dissolved in DDW containing 50 mM EDC and 25 mM NHS and vortex-stirred for 30 min. After that, PCLDP scaffolds were immersed in the EDC/NHS/carboxymethyl–βCD reaction solution overnight. These were then washed with fresh DDW 3 times. Fourth, PCLDPβ scaffolds were immersed in DEX/EtOH solution (1 mg ml−1) with shaking for two days as previously described.28,29 This was done with the bottles covered with aluminium-foil to protect the samples from light. Finally, the DEX loaded PCLDPβ scaffolds were washed with fresh DDW and dried at room temperature with protection from light.

2.5 Analytical equipment for characterization of scaffolds

The morphology of the scaffolds was observed using scanning electron microscopy (SEM, Hitachi S-4700, Japan) at an acceleration voltage of 15 kV. All samples were sputter-coated with platinum for 10 min. Additionally, the surface was observed using atomic force microscopy (AFM, Park-System XE-100, Korea) with a scan area of 10 × 10 μm2 at room temperature. Surface water contact angle measurements were performed with a goniometer (Phoenix 150, SEO, Korea) using 8 μl of distilled water at room temperature. X-ray photoelectron spectroscopy (XPS) was performed using K-Alpha 89 (Thermo Electron Manufacturing Ltd, UK) configured with a monochromatic Kα X-ray radiation. The survey scans were measured from 0 eV to 1300 eV.

2.6 Quantification of DEX loading amount on PCLDPβ

To quantify the loading of DEX on the PCLDPβ scaffold, the supernatant solution was collected after surface modification. As a control, bare PCL was also loaded with DEX under the same conditions and quantified. The supernatant solutions were diluted to be within the calibrated range and the ultraviolet-visible absorption was measured using an UV-1650PC spectrophotometer (Shimadzu, Japan). The actual loading amount was calculated from a standard curve of DEX of various concentrations (0, 10, 50, 100, 250, 500, and 750 μg ml−1). Quantification was based on the solution absorption at 260 nm.

2.7 Cell adhesion of human bronchial epithelial cells (hBEC) on scaffolds

In order to evaluate the efficacy of cell adhesion on scaffolds, the scaffolds were put into each well of a 48-well tissue culture plate (SPL Life Science, South Korea). To quantify non-adhesive cells, hBEC was labelled with a fluorescent membrane dye (PKH67 Green Fluorescent Cell Linker Kit, Sigma, St Louis, MO) according to the manufacturer's instructions. PKH67-labeled hBEC (Lonza, Walkersville, MD) was homogeneously suspended at a concentration of 2.7 × 106 cells per ml in a hBEC culture media based on a bronchial epithelial cell basal medium (Clonetics® BEGM, Lonza, Walkersville, MD) containing growth supplement (Clonetics® BEGM® SingleQuots®, Lonza, Walkersville, MD). PKH67-labeled hBEC solution (50 μl) was added into each well having a membrane for 10 minutes and then 450 μl of hBEC culture media was added. After 6 h, the media from each well was collected to measure the fluorescent intensity of non-adhesive PKH67-labeled hBEC. For this, 200 μl of collected media were transferred into each well of a 96-well plate, then the fluorescent intensity was measured using a multi-detection microplate reader (SpectraMax M5, Molecular Devices, Sunnyvale, CA) at an excitation wavelength of 485 nm and an emission wavelength of 538 nm. For the standard curve, the PKH67-labeled hBEC was used at a range of cell concentrations, 0.2 to 27.4 × 105 cells per ml in the hBEC culture media.

2.8 Anti-inflammatory response test of scaffolds through enzyme-linked immunosorbent assay (ELISA)

Each scaffold was placed into a well of a 48-well plate. The RAW 264.7 macrophages were seeded in each well (2 × 105 cells per well, n = 4 per group) and cultured for 24 h. Then, lipopolysaccharide (LPS, 1 μg ml−1) as a prototypical endotoxin added into each well to induce inflammatory response. After 24 h, the supernatant was collected from the each well to quantify pro-inflammatory cytokines, tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6), using ELISA (TNF alpha Mouse ELISA Kit and IL-6 Mouse ELISA Kit, Invitrogen, Carlsbad, CA, USA). The ELISA was performed according to the manufacturer's instructions.

2.9 3D tubular scaffold implantation at tracheal defect rabbit model

All animal procedures were approved by the Institutional Animal Care and Use Committee in Seoul National University Hospital (SNUH-IACUC) and animals were maintained in the facility accredited AAALAC International (#001169) in accordance with Guide for the Care and Use of Laboratory Animals 8th edition, NRC (2010). To evaluate the efficacy of the scaffolds to regenerate tracheal tissue in vivo, scaffolds were implanted into the tracheas of New Zealand rabbits (male, 3.0–3.5 kg, n = 3 per each group) and observed for 4 weeks (Fig. S1). The animal models were anesthetized using 0.83 ml kg−1 of xylanzine (23.32 mg ml−1, Rompun™, Bayer Korea) and 0.83 ml kg−1 of ketamine (57.68 mg ml−1, Yuhan Co. Ltd, South Korea) via intramuscular injection. Subsequently, the animals were placed in the supine position, shaved and disinfected at their necks in preparation for surgery. The surgical procedure was performed according to literature with some modifications.8,17,30,31 Briefly, a vertical incision was made at the midline of the neck until the trachea was exposed. A circumferential tracheal defect with a length of 8 mm was created. Subsequently, a 3D tubular scaffold with a diameter of 7 mm and a length of 10 mm was implanted into the tracheal defect using 4-0 PDS® Plus Antibacterial suture (Ethicon US, LLC., Somerville, NJ). After implantation, the strap muscles were sutured back into place using 4-0 Coated Vicryl® Plus Antibacterial suture (Ethicon US, LLC., Somerville, NJ) and the skin incision was closed using 3-0 Blue Nylon (AILEE Comp. Ltd, South Korea). Each animal was injected daily with 20 mg kg−1 of cephazolin (0.5 g ml−1, Yuhan Co. Ltd, South Korea) and 1 mg kg−1 of meloxicam (1.5 mg ml−1, Boehringer-Ingelheim Promeco SA de CV, Mexico) as an antibiotic and an analgesic, respectively, for a week. They were monitored daily for clinical signs of morbidity such as weight, cough, sputum, wheezing, and dyspnea. As a secondary endpoint, animals were euthanized when the animal showed 20% weight loss or symptoms of severe respiratory distress. Remaining animals were euthanized at 4 weeks after tracheal replacement.

2.10 In vivo evaluation of airway patency

Four weeks after tracheal replacement, each animal was anesthetized with 0.83 ml kg−1 of xylanzine and 0.83 ml kg−1 of ketamine via intramuscular injection, then the airway of the animal was observed using a 4.0 mm 30 degree rigid endoscope (Richard Wolf Medical Instruments Corp., Germany), attached to a digital camera (E4500, Nikon Corp. Japan). After euthanasia, the full trachea including the implanted scaffold was harvested for evaluation of airway patency using a micro-computed tomography (μCT) scanner (NFR Polaris-G90, NanoFocusRay, South Korea) at Seoul National University Hospital Biomedical Research Institute. The μCT images were taken at settings of 65 kV, 60 mA, and 500 ms with a resolution of 512 × 512 in Digital Imaging and Communication in Medicine (DICOM) format. Three-dimensional images of each replaced trachea by a scaffold were reconstructed using Lucion software (Infinitt Healthcare, South Korea).

2.11 Histological evaluation

After performing the μCT, the harvested trachea was fixed in 4% paraformaldehyde overnight, then the fixed trachea was dehydrated by serial ethanol washing and embedded in paraffin. A section of paraffin-embedded tissue (4 μm thick) was used for histological analyses. H&E staining, Masson's trichrome staining, and Alcian Blue staining were performed according to general procedures. Also, immunohistochemical staining (IHC) was performed to evaluate expression of cytokeratin 5 (CK5) with anti-CK5 (10 μg ml−1, monoclonal IgG, antibodies-online GmbH, Germany) and expression of β-tubulin with anti-β-tubulin (1 μg ml−1, monoclonal IgG, Thermo Fisher Scientific, Rockford, IL) according to manufacturer's instructions (VECTOR, Burlington, CA) with minor modification. Briefly, after deparaffinization and rehydration, tissues were exposed to citrate buffer (1×, pH 6.0) with heating. After washing with PBS (pH 7.4), the tissues were incubated with each primary antibody at room temperature for 1 h. After extensive washing with PBS (pH 7.4), the tissues were subsequently incubated with the pre-diluted pan-specific universal secondary antibody (R.T.U. VECTASTAIN®, VECTOR, Burlington, CA) at room temperature for 10 min. After washing three times with PBS for 5 min, the ready-to-use streptavidin and peroxidase complex was incubated on the tissues at room temperature for 5 min. After washing in the same manner as previously described, the tissues were incubated with peroxidase substrate solution until the desired intensity of the developed stain was achieved. Subsequently, the tissues were counter-stained with hematoxylin for 1 min and then mounted for imaging.

2.12 Statistical analysis

Statistical differences between the comparative groups were performed using one-way ANOVA (Tukey's post-test). The statistical analysis was processed using GraphPad Prism software.

3. Results

3.1 Scaffold characterization of surface structure and mechanical property (SEM and compressive strength test)

The surface morphology of manufactured 3D tubular scaffolds was observed via SEM analysis. In Fig. 2b, SEM images showed both PCL ENs layer (inner) and 3D printed PCL strand layer (outer). In the inner layer, randomly distributed ENs were observed. 3D printed PCL strands were also well deposited onto the ENs layer. Interestingly, both layers were well connected to each other without gaps. This phenomenon is due to the fact that we used the same PCL material during the manufacturing process of both ELSP and 3D printing.15,16 The 3D tubular scaffolds were still patent after deposition of PCL strand using 3D printing system and their structure were maintained well at room temperature (Fig. 2a).15,16 To certify whether the mechanical property is suitable for use as a scaffold for implantation in tracheal defect animal model, we performed a compressive test of the 3D tubular scaffolds and compared these to native rabbit tracheal tissue (Fig. 2e). Prior to analysis, we manufactured different types of 3D tubular scaffolds by modulating strand-to-strand distance (size of strand and distance between strands were 300 and 300/600/1200 μm. As shown in Fig. 2e, all the manufactured scaffolds represented stronger compressive strength than native rabbit tracheal tissue. An interesting result is that a narrower strand-to-strand spacing provided for a stronger compressive strength. In other words, the mechanical index of the scaffolds decreased gradually with increasing pore diameter.32 This result is due to the tight spacing making the strong scaffold. From these results, we determined that the manufacturing conditions of our 3D tubular scaffold provides for proper mechanical properties to transplant in vivo.

3.2 Characterization of surface morphology and chemical elements after surface modifications (AFM and XPS)

After surface treatments using PDA, PEI and βCD (Fig. 1), we carried out AFM and XPS analysis to confirm the physio-chemical changes of the scaffold's surface. The surface was found to become gradually rougher through the surface treatments (Fig. 3a).16,33 Roughness (Ra) measurements are depicted in the AFM images, indicating chemical molecules were well modified onto the scaffolds after each surface treatment step. We also carried out elemental XPS analysis to verify the surface chemical elements. As shown in Fig. 3b and Table. 1, the XPS spectra of the bare PCL tubular scaffold showed two major common peaks indicating C 1s and O 1s. However, the PDA coated scaffolds displayed characteristic N 1s peaks (amine functional group) and the peak notably increased after PEI treatment (Fig. 3b and c). This is due to PDA grafting through Michaels addition and Schiff-base reaction with PEI.34 Interestingly, the N 1s decreased and the O 1s group increased for the βCD treated scaffolds (Fig. 3b–d).35 This phenomenon is due to the grafting of βCD onto PCLDP scaffold, which has abundant hydroxyl groups.36 AFM and XPS analysis indicated that surface treatments were successful in modifying the surface. Thus, we anticipated that a large amount of DEX can be loaded onto the PCLDPβ scaffolds.37,38
image file: c9nr10341d-f3.tif
Fig. 3 Characterization of surface topologies of the 3D tubular scaffold by AFM. Rq denotes surface roughness (a). XPS spectrum of O, C, and N peaks (b), N 1s peak curve fitting (c), O 1s peak curve fitting (d) on bare of PCL, PCLD, PCLDP, and PCLDPβ, respectively.
Table 1 Chemical element content (%) of scaffolds by XPS analysis
Group abbreviations C N O
(a) Polycaprolactone (PCL) 77.68 0 22.32
(b) PCL + PDA (PCLD) 75.16 3.85 20.99
(c) PCL + PDA + PEI (PCLDP) 74.27 7.80 17.93
(d) PCL + PDA + PEI + βCD (PCLDPβ) 71.59 3.52 24.89

3.3 Characterization of loading amount of DEX on bare PCL and PCLDPβ

After grafting βCD on 3D tubular scaffolds, we loaded an excess amount of DEX on both bare PCL (control) and PCLDPβ scaffolds, respectively. The βCD grafted scaffold successfully loaded a larger amount of DEX (29.41 ± 6.06 μM) as compared to the bare PCL group (4.89 ± 1.98 μM) indicating a 5-fold increase in loading. This phenomenon is due to the formation of an inclusion complex between βCD and lipophilic soluble molecules such as DEX.20,39,40 From this result, we determined that grafting of βCD through surface modification can clearly improve the loading of DEX. Therefore, this outcome demonstrated that βCD was well grafted onto the 3D PCL scaffolds and allowed for subsequent DEX loading via facile surface treatment under mild aqueous conditions.

3.4. Cell adhesion on various surface-modified membranes

In order to evaluate the affinity of various surface-modified membranes for cell adhesion, the number of non-adherent hBEC was quantified after 6 h of cell seeding on each membrane (Fig. 4). All surface-modified membranes showed at least a 1.5-fold increase in cell adhesion as compared to bare PCL. The cell adhesion of PCLD and PCLDPβ + DEX were 1.8-fold and 2.0-fold higher, respectively, than that of bare PCL. Although additional surface-modifications were performed with PEI and βCD on PCLD, PCLDPβ + DEX showed similar cell adhesion to that of PCLD, suggesting PEI and βCD did not disrupt the cell adhesion.
image file: c9nr10341d-f4.tif
Fig. 4 Adhesion of human bronchial epithelial cells on various surface-modified membranes based on PCL fibrous membrane for 6 h. *p < 0.05.

3.5. Anti-inflammatory response of DEX loaded PCLDPβ scaffolds

In order to confirm the anti-inflammatory response of DEX on PCLDPβ, an amount of TNF-α and IL-6, well known inflammatory cytokines, secreted by the RAW 264.7 cells after treatment of LPS were quantified using ELISA. The cells on each membrane showed significantly lower expression of TNF-α and IL-6 under in an environment without LPS than that with the LPS treatment (Fig. 5a and b). The surface-modified materials on the PCL did not trigger pro-inflammatory responses due to the good biocompatibility of these materials. Consequently, the cells on each membrane expressed a large amount of TNF-α and IL-6 under treatment of LPS. However, the cells on PCLDPβ + DEX expressed significantly lower amount of TNF-α and IL-6, than those on PCL, while the expression of TNF-α and IL-6 by the cells on PCLD, CPLDP and PCLDPβ was higher or similar to that on PCL.
image file: c9nr10341d-f5.tif
Fig. 5 Anti-inflammatory response of various surface-modified membranes based on PCL fibrous membrane. The concentration of TNF-α (a) and IL-6 (b) secreted from the RAW 264.7 cells, with and without LPS treatment for 24 h were quantified by ELISA. **p < 0.01 and ***p < 0.001.

3.6. Maintenance of airway patency by DEX loaded PCLDPβ scaffolds

The airway patency of replaced trachea was evaluated four weeks after implantation using a rigid telescope and microCT. As observed in endoscopic images (Fig. 6a), the airway patency with PCL, PCLDP, and PCLDPβ was significantly decreased. Conversely, PCLD and PCLDPβ + DEX groups maintained their airway patency. Axial views of reconstructed microCT of each trachea (Fig. 6b) showed similar degrees of airway patency to those of endoscopic images. Moreover, when microCT images of the replaced tracheas were reconstructed with the longitudinal view (Fig. 6c), the stenosis was observed to be located at the anastomotic site of the replaced trachea with PCL and PCLDP. The narrow airway was observed along the entire length of the implanted PCLDPβ. Interestingly, we also observed cartilage-like signal represented with white-color on the area of implanted PCL, PCLDP and PCLDPβ + DEX, however, we did not confirmed any evidence regarding the cartilage-like signal through the histological evaluation (data not shown).
image file: c9nr10341d-f6.tif
Fig. 6 The airway patency of the replaced trachea with various surface-modified 3D printed tracheal scaffolds 4 weeks after tracheal replacement using (a) a bronchial endoscope and (b and c) microCT. The microCT images were reconstructed with (b) axial view and (c) longitudinal view. Cartilage rings and airway are represented with white and green color, respectively.

3.7. Histological evaluation of the replaced trachea with PCLDPβ-DEX

For the histological evaluation of replaced trachea, we used the middle part of the replaced trachea for histological staining. As shown in Fig. 7a, axial sections of the middle part of the implanted scaffolds showed similar degrees of airway patency to the images of microCT (Fig. 6). Also, the luminal surfaces of all implanted scaffolds, except PCL, were covered with tissue. While the luminal surface of the implanted PCLD was covered with squamous epithelium, the surface of PCLDPβ + DEX showed a wider area covered with pseudostratified epithelium which is the intrinsic respiratory epithelium. The deposition of collagen as a component of extracellular matrix (ECM) was evaluated by Masson's trichrome stain (Fig. 7b). Abundant and aligned collagen was observed at the migrated tissue of all implanted scaffolds, except PCL. Moreover, when expression of iNOS as a pro-inflammatory cytokine was confirmed by IHC, the tissue near the implanted PCLDPβ + DEX showed a significantly lower density of iNOS-positive brownish area than that of others (Fig. 7c). This aspect was analogous to the anti-inflammatory response with respect to the expression of TNF-α and IL-6 in vitro (Fig. 5).
image file: c9nr10341d-f7.tif
Fig. 7 Histological evaluation of the replaced trachea with various surface-modified 3D printed tracheal scaffolds 4 weeks after tracheal replacement by (a) H&E stain and (b) Masson's trichrome stain (blue; collagen, dark brown; nuclei). Red rectangle in each image of (a) indicates where each image of (b) locates. (c) Immunohistochemical (IHC) stain against anti-iNOS (iNOS; brown, deep blue-purple; nuclei). * and ** in each image of (c) indicate a 3D printed strand and a fibrous membrane, respectively.

Considering the airway epithelium is organized with basal cells, ciliated cells, and goblet cells, we confirmed CK5, β-tubulin, and mucin as a marker of basal cells, ciliated cells and goblet cells, respectively, for evaluation of the regeneration of the airway epithelium (Fig. 8). While a large number of CK5-positive cells were observed on all implanted scaffolds, except PCL, β-tubulin-positive cilia of ciliated cells were expressed only on the regenerated epithelium of the implanted PCLD and PCLDPβ + DEX. Interestingly, the regenerated epithelium of the implanted PCLDPβ + DEX showed stronger expression of β-tubulin as compared to that of the implanted PCLD. Moreover, significant expression of mucin was observed at the regenerated epithelium of the implanted PCLDPβ + DEX while other implanted scaffolds did not present mucin.

image file: c9nr10341d-f8.tif
Fig. 8 Regeneration of the airway epithelium with various surface-modified 3D printed tracheal scaffolds 4 weeks after tracheal replacement. IHC staining images of each epithelium against (a) anti-cytokeratin 5 (CK5) for basal cells (green; CK5, blue; nuclei) and (b) anti-β-tubulin for cilia of ciliated cells (green; β-tubulin, blue; nuclei). (c) Alcian blue staining images (blue; mucin and hyaluronic acid, red; nuclei) of each epithelium. Red arrowheads indicate mucin.

4. Discussion

In an otolaryngology-head and neck clinic, rapid availability of an artificial trachea scaffold to transplant into the defect area is necessary for the complete recovery of patients without further complications. Without this immediate replacement, life threatening events can occur. For this reason, otolaryngologists need a rapidly-available artificial scaffold possessing proper mechanical properties and capable of maintaining the patency of the airway.41 To resolve this problem, in trachea tissue engineering and regenerative medicine, bioengineering researchers have developed tracheal scaffolds by using DTM.5,6 Although a few satisfying achievements were obtained, there has been several problems. These include (i) time-consuming and expensive preparation, (ii) delayed mucosal regeneration resulting in the inflammation and restenosis of the trachea, and (iii) the lack of donor sources.42–44 As an alternative solution, based on these results, previous reports Tsao et al.10 and Luo et al.11 developed a biodegradable and biocompatible artificial scaffolds using biomimetic polymers. Nevertheless, these scaffolds require at least 2 weeks for preparation of a trachea that resembles the tissue. This creates a limitation in providing for immediate transplant to a defected area. Pre-made scaffolds are restricted in their patient application as the anatomical structure and physiological function of trachea differ depending on patient age and gender45

In order to overcome these issues in this study, we designed a hybrid rapidly-manufacturable tracheal scaffold by combining ELSP and 3D printing system with mild surface modifications to load a target drug (Fig. 1). There are several requirements for the tracheal scaffolds to be applicable to clinical translation: (i) rapid preparation with suitable mechanical property, (ii) ease of use, and (iii) improved tracheal regeneration. In the current study, we employed both ELSP and 3D printing systems to form a robust 3D tubular structure. In general, the ENs were fabricated via ELSP. Fabricated ENs can play an important role in tissue repair and reconstruction. It has several important benefits, including (i) high surface-to-volume ratio, (ii) high porosity with interconnected and tuneable pore structures, (iii) three-dimensional environment to support cell growth and proliferation.46 The ENs are similar to natural ECM, which enables them to provide both biophysical and biochemical properties required for functional cellular activity.47 Thus, we used the ENs as a base layer where the tracheal tissue directly touches the 3D tubular scaffolds. These fibrous webs, however, did not have sufficient mechanical properties for use as tracheal implant materials. For that reason, we used 3D printing to deposit the more rigid PCL polymer onto the fibrous web (Fig. 1, step I). In both ELSP and 3D printing procedures, we used the same material. This enabled the creation of a solid 3D tubular structure (Fig. 2a and b). Combinations of two systems are able to provide for rapid fabrication with suitable mechanical properties and custom preparation.15,16 From this approach, we obtained a robust artificial conduit scaffold (Fig. 2a), which exhibited solid properties comparative to native trachea tissue (Fig. 2e). After that, we performed surface treatments onto the scaffolds to load a large amount of DEX. As mediators between the 3D PCL scaffold and DEX, we grafted polydopamine (PDA), polyethylenimine (PEI), and carboxymethyl–β-cyclodextrin (βCD) on the 3D PCL scaffolds, sequentially. For grafting the carboxymethyl–βCD, we employed PDA and PEI chemistry in order to introduce an amine-rich surface onto the 3D PCL scaffold,26,34 which enabled grafting of carboxymethyl–βCD through amide bond formation.35 In our previous reports, we established PDA/PEI modification process on substrates as a means to load carboxylate bio-molecules.26,48 As is well known, PEI is a cationic polymer widely used in non-viral transfection of genes. This is due to its outstanding DNA entrapping characteristic and intrinsic endosomolytic activity.49 The biocompatibility of PEI depends on the molecular weight. Low molecular weight PEI has been shown to have less toxicity than high molecular weight PEI.49 Thus, many tissue-engineering researchers have used low molecular weight PEI as a mediator in production of scaffolds.50,51 Our lab has also previously used low molecular weight PEI in development of an antibiotic drug delivery system.48 In that study, the biocompatibility of low molecular PEI-treated titanium implants was measured using human adipose derived mesenchymal stem cells. Biocompatibility was determined to be good for this system. Based on these previous findings, we used low molecular weight PEI as it would not affect biocompatibility. The βCD is a cyclic oligosaccharide consisting of seven glucopyranose units with a hydrophilic outer surface and a lipophilic cavity in the center.37 In aqueous solutions, βCD is capable of forming inclusion complexes with various types of lipophilic drugs such as DEX.20,38 From these aspects, we speculated that the grafting of βCD on the scaffold surface may enable not only a large amount loading of DEX, but also regeneration of the tracheal tissue with anti-inflammation activity after transplantation in animal model. These materials are soluble in aqueous solution, therefore they do not collapse the polymer structure. In this way, we were able to treat the scaffold surface, directly. In our previous report, a 3D porous tubular scaffold was prepared by using PDA to induce a growth factor immobilization.16 The developed 3D tubular scaffold did not show any mechanical weakness. The properties can also be varied by modulating the design given by 3D printing (Fig. 2e). Consequently, we successfully carried out in vitro evaluation and in vivo transplantation (Fig. 4 and Fig. S1). According to previous reports, Hung et al. demonstrated that transplantation of decellularized tracheal scaffold in rabbit model died within 7–24 days because of collapse of the tracheal tubular structures.52 They described that the insufficient mechanical properties of the cartilage portions were unable to maintain mechanical hardness and sufficient integrity. However, our scaffold can sustain an artificial airway within the defect model after transplantation until sacrifice for analysis. In this case, we believe that our approach 3D tubular scaffold is better than current outcomes and is eligible for clinical use.

The poor cell adhesiveness of PCL is due to its strong hydrophobic property.53,54 This was improved by surface modification techniques (Fig. 4). From this, we hypothesized that the enhancement of the cell adhesion stimulates cell migration, and induces tissue regeneration. Although the surface modified scaffolds showed high cell affinity, the lack of a sufficient cue for selective adhesion of specific cells on the modified surface did not prevent the development of stenosis at the anastomotic site (Fig. 6). To provide for this, a scaffold must provide for selective affinity to epithelial cells which can lead to facilitated migration of all cells, including fibroblasts which are the main components of stenosis.55–58 The mortality of an animal after tracheal replacement occurs due to stenosis at the anastomotic site as this constricts the airflow. Unfortunately, a specific cue for enhancing adhesion of airway epithelium cells has not yet been revealed. In this study, we used an anti-inflammatory drug, dexamethasone, to suppress the development of stenosis. The environment of the airway is ceaselessly exposed to various antigens which induce inflammation during respiration.59 As a result, expressed inflammatory cytokines, especially TNF-α, accelerates expression of TGF-β1 which has been known to induce the migration of fibroblasts and ultimately stenosis.60 Accordingly, we investigated the anti-inflammatory effects of the surface-modified membranes through the expression of inflammatory cytokines by macrophages in vitro (Fig. 5). As the DEX loaded surface-modified membrane significantly suppressed the expression of inflammatory cytokines, we hypothesized that the anti-inflammatory effects of PCLDPβ + DEX is able to suppress the fibrosis and restenosis while enhancing the regeneration of the airway epithelium.

For these reasons, we used the PCLDPβ + DEX for tracheal replacement and evaluated its effect to both maintain airway patency and to regenerate the airway epithelium over the course of 4 weeks after tracheal replacement. The airway patency was maintained more thoroughly for PCLD and PCLDPβ + DEX scaffolds as compared to PCL, PCLDP and PCLDPβ scaffolds (Fig. 6). Animals implanted with PCL, PCLDP and PCLDPβ scaffolds died of restenosis at the anastomotic site. Although PCLD did not have the anti-inflammatory drug, dexamethasone, the airway patency of PCLD was similar to that of PCLDPβ + DEX at 4 weeks after tracheal replacement. We suspect that PDA acts as an ROS scavenger which acted to reduce the inflammatory response. Ultimately, this would lead to suppression of stenosis at the anastomotic site.61,62 This is supported by the expression of iNOS for which PCLD showed lower values as compared to PCL (Fig. 7). The higher expression of iNOS at PCL, PCLDP, and PCLDPβ induced death of animals by stimulating development of stenosis at the anastomotic site. Noticeably, the anti-inflammatory efficacy of PCLDPβ + DEX was clearly confirmed in vivo as well. Considering that inflammation induces the migration of fibroblasts,60 we supposed that the anti-inflammatory effects of DEX released from PCLDPβ + DEX contributed to prevent stenosis by decreasing cell migration. Basal cells have the potential of differentiating into epithelial cells, such as ciliated cells and goblet cells.63,64 The luminal surface of all scaffolds, except the PCL, showed CK5 positive basal cells. This was especially strong in PCLD and PCLDPβ + DEX. In particular, the epithelium of the implanted PCLDPβ + DEX showed abundant cilia formation as compared to that of the implanted PCLD. Moreover, mucin secreted by goblet cells was observed only at the epithelium of the implanted PCLDPβ + DEX. Considering that PDA has been widely employed to modify scaffold surfaces for enhancement of tissue regeneration, due to its superior cell adhesiveness,16,18,19,33,65 we concluded that the PDA acts to accelerate migration and adhesion of basal cells, which act as stem cells. Consequently, the airway epithelium can recover several intrinsic functions such as cilia formation and mucin secretion.

In this study, we aimed to create an artificial 3D tubular scaffold which could satisfy the requirements of mechanical property, anti-inflammation, convenient use, and prevention of restenosis. Animal model tested indicated success in maintaining the airway lumen patency and enhancing mucosal regeneration. Thus, we believe that our approach can be useful for the future of the tracheal tissue engineering field and may serve well for clinical applications.

5. Conclusion

In summary, we designed and prepared a robust 3D tubular scaffold followed by surface treatments to load DEX for use in trachea replacement therapy. By combining ELSP and 3D printing techniques, we could modulate the mechanical properties which was confirmed by compressive test. After surface treatments, we manufactured a PCLDPβ scaffold which enabled large amounts of DEX to be loaded onto the substrate. We investigated cellular behaviour regarding cell adhesion and anti-inflammation activity of these scaffolds. Interestingly, DEX loaded PCLDPβ scaffold exhibited higher cell adhesion and anti-inflammatory response as compared to the other groups. As an in vivo assessment, DEX loaded PCLDPβ scaffold significantly induced tracheal regeneration. As seen by these results, the developed 3D tubular scaffolds have good tracheal tissue regeneration potential and would be widely applicable for use in trachea transplantation therapy.

Conflicts of interest

There are no conflicts to declare.


This research was supported by the Bio & Medical Technology Development Program of the National Research Foundation (NRF) funded by the Korean government (MSIT) (No. 2019M3A9E2066348) and the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (HI18C1174).


  1. L. F. Jorge, J. C. Francisco, N. Bergonse, C. Baena, K. A. T. Carvalho, E. Abdelwahid, J. R. F. Neto, L. F. P. Moreira and L. C. Guarita–Souza, J. Tissue Eng. Regener. Med., 2018, 12, e1525–e1530 CrossRef CAS .
  2. M. Kalathur, S. Baiguera and P. Macchiarini, Cell. Mol. Life Sci., 2010, 67, 4185–4196 CrossRef CAS PubMed .
  3. P. Jungebluth, G. Moll, S. Baiguera and P. Macchiarini, Clin. Pharmacol. Ther., 2012, 91, 81–93 CrossRef CAS PubMed .
  4. S. Broor, G. S. Raju, P. Bose, D. Lahoti, G. Ramesh, A. Kumar and G. Sood, Gut, 1993, 34, 1498–1501 CrossRef CAS PubMed .
  5. P. Lange, K. Greco, L. Partington, C. Carvalho, S. Oliani, M. Birchall, P. Sibbons, M. Lowdell and T. Ansari, J. Tissue Eng. Regener. Med., 2017, 11, 800–811 CrossRef CAS PubMed .
  6. C. R. Butler, R. E. Hynds, C. Crowley, K. H. Gowers, L. Partington, N. J. Hamilton, C. Carvalho, M. Platé, E. R. Samuel and A. J. Burns, Biomaterials, 2017, 124, 95–105 CrossRef CAS PubMed .
  7. F. Sun, S. Pan, H. C. Shi, F. B. Zhang, W. D. Zhang, G. Ye, X. C. Liu, S. Q. Zhang, C. H. Zhong and X. L. Yuan, J. Biomed. Mater. Res., Part A, 2015, 103, 1509–1519 CrossRef PubMed .
  8. D. Y. Lee, S. A. Park, S. J. Lee, T. H. Kim, S. H. Oh, J. H. Lee and S. K. Kwon, Laryngoscope, 2016, 126, E304–E309 CrossRef CAS PubMed .
  9. R. J. Morrison, S. J. Hollister, M. F. Niedner, M. G. Mahani, A. H. Park, D. K. Mehta, R. G. Ohye and G. E. Green, Sci. Transl. Med., 2015, 7, 285ra264 Search PubMed .
  10. C.-K. Tsao, C.-Y. Ko, S.-R. Yang, C.-Y. Yang, E. M. Brey, S. Huang, I.-M. Chu and M.-H. Cheng, Biomaterials, 2014, 35, 1163–1175 CrossRef CAS PubMed .
  11. X. Luo, Y. Liu, Z. Zhang, R. Tao, Y. Liu, A. He, Z. Yin, D. Li, W. Zhang and W. Liu, Biomaterials, 2013, 34, 3336–3344 CrossRef CAS PubMed .
  12. E. M. Boazak and D. T. Auguste, ACS Biomater. Sci. Eng., 2018, 4, 1272–1284 CrossRef CAS .
  13. J. X. Law, L. L. Liau, B. S. Aminuddin and B. H. I. Ruszymah, Int. J. Pediatr. Otorhinolaryngol., 2016, 91, 55–63 CrossRef PubMed .
  14. C. Crowley, M. Birchall and A. M. Seifalian, J. Tissue Eng. Regener. Med., 2015, 9, 357–367 CrossRef PubMed .
  15. S. J. Lee, D. N. Heo, J. S. Park, S. K. Kwon, J. H. Lee, J. H. Lee, W. D. Kim, I. K. Kwon and S. A. Park, Phys. Chem. Chem. Phys., 2015, 17, 2996–2999 RSC .
  16. S. J. Lee, M. E. Kim, H. Nah, J. M. Seok, M. H. Jeong, K. Park, I. K. Kwon, J. S. Lee and S. A. Park, J. Colloid Interface Sci., 2019, 537, 333–344 CrossRef CAS PubMed .
  17. H.-J. Ahn, R. Khalmuratova, S. A. Park, E.-J. Chung, H.-W. Shin and S. K. Kwon, Tissue Eng. Regener. Med., 2017, 14, 631–639 CrossRef CAS PubMed .
  18. S. J. Lee, H.-J. Lee, S.-Y. Kim, J. M. Seok, J. H. Lee, W. D. Kim, I. K. Kwon, S.-Y. Park and S. A. Park, Nanoscale, 2018, 10, 15447–15453 RSC .
  19. S. J. Lee, D. Lee, T. R. Yoon, H. K. Kim, H. H. Jo, J. S. Park, J. H. Lee, W. D. Kim, I. K. Kwon and S. A. Park, Acta Biomater., 2016, 40, 182–191 CrossRef CAS .
  20. M. M. Doile, K. A. Fortunato, I. C. Schmücker, S. K. Schucko, M. A. Silva and P. O. Rodrigues, AAPS PharmSciTech, 2008, 9, 314–321 CrossRef CAS PubMed .
  21. S. E. Bae, D. H. Choi, D. K. Han and K. Park, J. Controlled Release, 2010, 143, 23–30 CrossRef CAS PubMed .
  22. A. Sekiyama, Y. Gon, M. Terakado, I. Takeshita, Y. Kozu, S. Maruoka, K. Matsumoto and S. Hashimoto, Int. Immunopharmacol., 2012, 12, 350–357 CrossRef CAS PubMed .
  23. Y. Kozu, Y. Gon, S. Maruoka, K. Kazumichi, A. Sekiyama, H. Kishi, Y. Nomura, M. Ikeda and S. Hashimoto, BMC Pulm. Med., 2015, 15, 80 CrossRef PubMed .
  24. T. Loftsson and E. Stefansson, J. Drug Delivery Sci. Technol., 2007, 17, 3–9 CrossRef CAS .
  25. P. O. Rodrigues, A. C. Becker, V. A. Dellagnelo, C. N. Benetti, E. M. Pereira, T. M. Wagner and M. A. S. Silva, Braz. Arch. Biol. Technol., 2014, 57, 887–894 CrossRef .
  26. S. J. Lee, H. H. Jo, K. S. Lim, D. Lim, S. Lee, J. H. Lee, W. D. Kim, M. H. Jeong, J. Y. Lim and I. K. Kwon, Chem. Eng. J., 2019, 378, 122116 CrossRef CAS .
  27. J. M. Seok, S. H. Oh, S. J. Lee, J. H. Lee, W. D. Kim, S.-H. Park, S. Y. Nam, H. Shin and S. A. Park, Mater. Today Commun., 2019, 19, 56–61 CrossRef CAS .
  28. V. T. Karathanos, I. Mourtzinos, K. Yannakopoulou and N. K. Andrikopoulos, Food Chem., 2007, 101, 652–658 CrossRef CAS .
  29. E. H. Santos, J. A. Kamimura, L. E. Hill and C. L. Gomes, LWT–Food Sci. Technol., 2015, 60, 583–592 CrossRef CAS .
  30. D. Y. Lee, J. H. Lee, H.-J. Ahn, S. H. Oh, T. H. Kim, H.-B. Kim, S.-W. Park and S. K. Kwon, Biomaterials, 2015, 44, 134–142 CrossRef CAS PubMed .
  31. S.-H. Kim, H. R. Lee, S. J. Yu, M.-E. Han, D. Y. Lee, S. Y. Kim, H.-J. Ahn, M.-J. Han, T.-I. Lee and T.-S. Kim, Proc. Natl. Acad. Sci. U. S. A., 2015, 112, 15426–15431 CrossRef CAS PubMed .
  32. S. Pan, Y. Zhong, Y. Shan, X. Liu, Y. Xiao and H. Shi, J. Biomed. Mater. Res., Part A, 2019, 107, 360–370 CrossRef CAS PubMed .
  33. S. J. Lee, J.-E. Won, C. Han, X. Y. Yin, H. K. Kim, H. Nah, I. K. Kwon, B.-H. Min, C.-H. Kim and Y. S. Shin, J. Colloid Interface Sci., 2019, 539, 468–480 CrossRef CAS PubMed .
  34. C. Zhao, F. Zuo, Z. Liao, Z. Qin, S. Du and Z. Zhao, Macromol. Rapid Commun., 2015, 36, 909–915 CrossRef CAS PubMed .
  35. J. B. Lee, J. E. Kim, D. A. Balikov, M. S. Bae, D. N. Heo, D. Lee, H. J. Rim, D. W. Lee, H. J. Sung and I. K. Kwon, Macromol. Biosci., 2016, 16, 1027–1038 CrossRef CAS .
  36. J. Zhang and P. X. Ma, Adv. Drug Delivery Rev., 2013, 65, 1215–1233 CrossRef CAS PubMed .
  37. R. F. Lopez, J. H. Collett and M. V. L. Bentley, Int. J. Pharm., 2000, 200, 127–132 CrossRef CAS PubMed .
  38. T. Loftsson, H. Friðriksdóttir, S. Thórisdóttir and E. Stefánsson, Int. J. Pharm., 1994, 104, 181–184 CrossRef CAS .
  39. C.-D. Radu, O. Parteni and L. Ochiuz, J. Controlled Release, 2016, 224, 146–157 CrossRef CAS PubMed .
  40. R. Mateen and T. Hoare, Int. J. Pharm., 2014, 472, 315–326 CrossRef CAS PubMed .
  41. J. S.-j. Lee, J. Park, D.-A. Shin, Y.-j. Ryu, H. C. Kim, J. C. Lee and S. K. Kwon, Auris, Nasus, Larynx, 2019, 46, 407–416 CrossRef .
  42. J. Basu and J. W. Ludlow, Trends Biotechnol., 2010, 28, 526–533 CrossRef CAS .
  43. P. M. Crapo, T. W. Gilbert and S. F. Badylak, Biomaterials, 2011, 32, 3233–3243 CrossRef CAS PubMed .
  44. G. Vogel, Science, 2013, 340, 266–268 CrossRef CAS PubMed .
  45. Y. Kang, C. Wang, Y. Qiao, J. Gu, H. Zhang, T. Peijs, J. Kong, G. Zhang and X. Shi, Biomacromolecules, 2019, 20, 1765–1776 CrossRef CAS PubMed .
  46. M. J. Mochane, T. S. Motsoeneng, E. R. Sadiku, T. C. Mokhena and J. S. Sefadi, Appl. Sci., 2019, 9, 2205 CrossRef .
  47. X. Yan, M. Yu, S. Ramakrishna, S. J. Russell and Y.-Z. Long, Nanoscale, 2019, 11, 19166–19178 RSC .
  48. J. S. Lee, S. J. Lee, S. B. Yang, D. Lee, H. Nah, D. N. Heo, H.-J. Moon, Y.-S. Hwang, R. L. Reis and J.-H. Moon, Appl. Surf. Sci., 2019, 496, 143675 CrossRef CAS .
  49. S. Taranejoo, J. Liu, P. Verma and K. Hourigan, J. Appl. Polym. Sci., 2015, 132, 42096 CrossRef .
  50. Y. Liu, R. Luo, F. Shen, L. Tang, J. Wang and N. Huang, Appl. Surf. Sci., 2015, 328, 163–169 CrossRef CAS .
  51. R. Luo, X. Wang, J. Deng, H. Zhang, M. F. Maitz, L. Yang, J. Wang, N. Huang and Y. Wang, Colloids Surf., B, 2016, 144, 90–98 CrossRef CAS PubMed .
  52. S. H. Hung, C. H. Su, S. E. Lin and H. Tseng, Laryngoscope, 2016, 126, 2520–2527 CrossRef CAS PubMed .
  53. J. S. Choi and H. S. Yoo, J. Bioact. Compat. Polym., 2007, 22, 508–524 CrossRef CAS .
  54. C. H. Kim, M. S. Khil, H. Y. Kim, H. U. Lee and K. Y. Jahng, J. Biomed. Mater. Res., Part B, 2006, 78, 283–290 CrossRef PubMed .
  55. J. S. Choi, P. B. Messersmith and H. S. Yoo, Macromol. Biosci., 2014, 14, 270–279 CrossRef CAS PubMed .
  56. S. H. Ku, J. S. Lee and C. B. Park, Langmuir, 2010, 26, 15104–15108 CrossRef CAS PubMed .
  57. C. Rodriguez-Emmenegger, C. M. Preuss, B. Yameen, O. Pop-Georgievski, M. Bachmann, J. O. Mueller, M. Bruns, A. S. Goldmann, M. Bastmeyer and C. Barner-Kowollik, Adv. Mater., 2013, 25, 6123–6127 CrossRef CAS PubMed .
  58. S. Hong, K. Y. Kim, H. J. Wook, S. Y. Park, K. D. Lee, D. Y. Lee and H. Lee, Nanomedicine, 2011, 6, 793–801 CrossRef CAS .
  59. M. Yao and G. Mainelis, J. Exposure Sci. Environ. Epidemiol., 2007, 17, 31–38 CrossRef CAS .
  60. L. A. Borthwick, T. A. Wynn and A. J. Fisher, Biochim. Biophys. Acta, Mol. Basis Dis., 2013, 1832, 1049–1060 CrossRef CAS PubMed .
  61. H. Zhao, Z. Zeng, L. Liu, J. Chen, H. Zhou, L. Huang, J. Huang, H. Xu, Y. Xu and Z. Chen, Nanoscale, 2018, 10, 6981–6991 RSC .
  62. L. Jin, F. Yuan, C. Chen, J. Wu, R. Gong, G. Yuan, H. Zeng, J. Pei and T. Chen, Inflammation, 2019, 42, 658–671 CrossRef CAS .
  63. J. R. Rock, S. H. Randell and B. L. Hogan, Dis. Models Mech., 2010, 3, 545–556 CrossRef CAS .
  64. R. Shaykhiev, Eur. Respir. J., 2015, 46, 894–897 CrossRef .
  65. S. J. Lee, H. H. Jo, S. K. Kwon, J. H. Lee, W. D. Kim, J. H. Lee, S. A. Park and I. K. Kwon, Macromol. Res., 2016, 24, 305–308 CrossRef CAS .


Electronic supplementary information (ESI) available. See DOI: 10.1039/c9nr10341d
These authors are considered as co-first authors and made equal contributions to the study.
§ Current address: Department of Bioengineering, University of Illinois at Chicago, Chicago, Illinois 60612, United States.
These corresponding authors equally contributed to this work.

This journal is © The Royal Society of Chemistry 2020