Biomimetic in situ tracheal microvascularization for segmental tracheal reconstruction in one‐step

Abstract Formation of functional and perfusable vascular network is critical to ensure the long‐term survival and functionality of the engineered tissue tracheae after transplantation. However, the greatest challenge in tracheal‐replacement therapy is the promotion of tissue regeneration by rapid graft vascularization. Traditional prevascularization methods for tracheal grafts typically utilize omentum or muscle flap wrapping, which requires a second operation; vascularized segment tracheal orthotopic transplantation in one step remains difficult. This study proposes a method to construct a tissue‐engineered tracheal graft, which directly forms the microvascular network after orthotopic transplantation in vivo. The focus of this study was the preparation of a hybrid tracheal graft that is non‐immunogenic, has good biomechanical properties, supports cell proliferation, and quickly vascularizes. The results showed that vacuum‐assisted decellularized trachea‐polycaprolactone hybrid scaffold could match most of the above requirements as closely as possible. Furthermore, endothelial progenitor cells (EPCs) were extracted and used as vascularized seed cells and seeded on the surfaces of hybrid grafts before and during the tracheal orthotopic transplantation. The results showed that the microvascularized tracheal grafts formed maintained the survival of the recipient, showing a satisfactory therapeutic outcome. This is the first study to utilize EPCs for microvascular construction of long‐segment trachea in one‐step; the approach represents a promising method for microvascular tracheal reconstruction.


| INTRODUCTION
Tracheal transplantation remains an unresolved clinical challenge worldwide. This is primarily owing to the lack of ideal tracheal substitutes and the obstruction of graft vascularization, which is caused by special anatomical structures. 1,2 Longer segment trachea (more than 50% in adults or 30% in children) lesions caused by tracheal tumors, stenosis, trauma, and softening cause added complications in surgical operations. In such situations, end-to-end anastomosis fails, and serious complications, such as anastomotic leakage and tracheal rupture, are caused owing to high anastomotic tension. 3 Tracheal transplantation with a substitute is then an effective treatment method to achieve healthy airway repair. 4 Clinical tracheal transplantation using bioprosthesis, 5 allograft, [6][7][8] or autologous tissue reconstruction [9][10][11] has been performed; however, but the effects have been highly unsatisfactory. The main shortcomings are as follows: (1) owing to poor biocompatibility and rigid structure, the bioprosthesis can easily cause hyperplasia of granulation tissue and tracheal stenosis; (2) long-term use of immunosuppressive agents is required by patients for the allograft to be immunogenic, and a lack of effective blood supply to the graft leads to necrosis; (3) reconstruction of the structure of the bionic native trachea through autologous tissue is challenging and may lead to its abnormal functioning, increasing the possibility of surgical trauma. 12 Tissue-engineering technology combines living cells with scaffolds and has the potential to construct highly bionic functionalized tracheal substitutes. 13 However, this technique still has not achieved the desired effect in clinical practice owing to postoperative complications such as softening, collapse, lumen stenosis, and blood circulation disorders of grafts. [14][15][16] A key scientific issue with regenerative materials is the need to provide tissue-specific biochemical and biomechanical properties, while facilitating cell proliferation and migration for functional graft formation and the repair of normal tissue. In addition, the development of a suitable biomaterial that contains angiogenesisinducing substances, that can promote angiogenic factor secretion and endothelial differentiation of seeded cells, is a meaningful way to promote the vascularization in tissue engineering. 17 From the perspective of tissue composition and structure, the decellularized extracellular matrix (dECM) from natural tissue is an ideal biomaterial that preserves complex molecular components and complete tissue structure. 18,19 Our recent study showed that vacuum-assisted decellularized trachea (VADT) exhibited extremely low immunogenicity and did not induce an inflammatory response when transplanted in vivo. 20 However, insufficient biomechanical properties are a major limitation of these natural materials. Thus, constructing tissue structures with the desired biological functions and biomechanical properties from existing biomaterials remains a challenge. One approach involves a hybrid structure of synthetic and natural materials: synthetic materials provide suitable mechanical properties and a complete physical structure for tissue remodeling, and natural materials provide a suitable biochemical microenvironment for cell adhesion, proliferation, and differentiation. 21,22 Polycaprolactone (PCL) has been widely demonstrated to have excellent biocompatibility, suitable mechanical properties, and printability for 3D printing a tracheal scaffold. [23][24][25] However, PCL is unfavorable for cell affinity, angiogenesis, and tissue regeneration when used alone, owing to its hydrophobicity. 26 Based on these properties, a VADT-PCL hybrid scaffold is expected to provide both biomechanical and stable cell affinity for tracheal grafts.
Foremost among the main challenges of all approaches in regenerative medicine is the ability to construct the vasculature of engineered tissues and translate them into a clinical model. 27 The goals of vascularized construction include ensuring that the entire graft is perfused, and adequate nutrients and oxygen are received for long-term survival. In addition, a well-developed vascular network, such as capillaries and microvessels, is necessary for neonatal tissue. 28 The blood supply to the trachea depends on small blood vessels that infiltrate between the cartilage rings and the mucosal layer of the inner wall to provide a segmental blood supply. Vascularization of the tissue-engineered trachea is crucial for the survival of tracheal epithelial cells. More importantly, the vascularized trachea also plays an important role in resisting infection and reducing the necrosis and stenosis of the graft. 29,30 Neovascularization can occur via angiogenesis or vasculogenesis. 31 Angiogenesis is the formation of new blood vessels by sprouting or by splitting based on pre-existing vessels. 32 By contrast, vasculogenesis refers to the process in which endothelial precursor cells aggregate, differentiate, and reorganize during the embryonic period, involving the formation of new blood vessels from bone marrow (BM) derived endothelial progenitor cells (EPCs). 33 Owing to their high potential for in vitro expansion and acquisition of the mature endothelial cell phenotype, EPCs have gained considerable interest as an alternative source of endothelial cells (ECs), as well as seed cells that effectively promote neovascularization when transplanted in vivo. 34 However, the role of BM-EPCs in promoting vascularization in tissueengineered tracheas in vivo remains unknown. Therefore, this study proposes the development of a hybrid tracheal graft that meets the following clinical needs: (1) low immunogenicity, good biocompatibility, and provision of a good microenvironment for cell growth; (2) good biomechanical properties to keep the lumen unobstructed; and (3) ability to form a microvascular network in a short time to promote the formation of functional grafts and long-term survival of the recipient. In this study, we prepared VADT scaffolds that exhibited low immunogenicity and supported cell growth and angiogenesis. Subsequently, PCL macroporous mesh stents that matched the diameter of the VADT were prepared using 3D-printing technology with good compression and resilience, and they were used to construct hybrid grafts in combination with the VADT that matched the biomechanical properties of the native trachea. Next, the EPCs were separated from BM and seeded on the surface of the hybrid grafts in vitro. Finally, Matrigel loaded with EPCs and vascular endothelial growth factor (VEGF) was used as a coating of the hybrid grafts during surgery for the orthotopic transplantation of segment tracheae, to promote the formation of vascularization and the functionalization of the grafts in vivo.

| Preparations and characterization of VADT
The VADT was prepared in a vacuum (À0.96 MPa) created by a microcomputer negative vacuum pump (Fujiwara Tools Co., Ltd., Taizhou, China) and processed in a shaking incubator at 80 rpm according to a previously described technique. 20 The native tracheal tissues (approximately 5-cm long segments) that were obtained from adult New Zealand white rabbits were incubated in sterile distilled water at 4 C for 24 h, and then were incubated in a detergent solution containing 0.25% Triton X-100 (Biofroxx, Einhausen, Germany) and 0.25% sodium deoxycholate (Sigma, CA, USA) at 37 C for 24 h.
The segments were then washed in sterile distilled water three times for 30 min, and the scaffolds were subjected to enzymatic digestion with 1 kU/mL DNAse (Sigma, CA, USA) and 2 U/mL RNAse (Biofroxx, Einhausen, Germany) in 1 M NaCl at 37 C for another 24 h. Finally, the decellularized tracheal segments were stored in phosphate-buffered saline (PBS) containing 1% antibiotic and antimycotic solution at 4 C.
To determine the decellularization efficacy, hematoxylin and

| Extraction and identification of EPCs
The EPCs were isolated and cultured using whole marrow differential adherence. Briefly, BM was isolated from the femurs of 3-week-old female rabbits and cultured in DMEM-F12 medium containing 10% fetal bovine serum (FBS, Gibco, NY, USA) in an incubator (HERAcell 150i, Thermo Fisher, Waltham, USA) at 37 C in 5% CO 2 for 20 h.

| Biocompatibility of the hybrid scaffolds
The biocompatibility of the hybrid scaffolds was evaluated as follows.
First, a cell-counting kit-8 (CCK-8, Biosharp, Hefei, China) assay was performed to examine the proliferation of EPCs on the scaffolds after 1, 3, and 5 days of culture. Next, live/dead staining (KeyGEN, Nanjing, China) was performed to examine the viability of the EPCs on the scaffolds after 24 h of culture. Finally, the microstructure of the cellscaffold composite was observed using SEM (GeminiSEM 300, Hitachi, Japan).

| Chicken embryo chorioallantoic membrane assays
Chorioallantoic membrane (CAM) assays were used as in vivo models to evaluate the angiogenic properties of the bioengineered tracheae.
As previously reported, 35

| In vivo neovascularization assay
To verify the in vivo angiogenic properties of the Matrigel loaded with VEGF and EPCs, it was embedded in Wistar rats. Matrigel was placed on ice overnight, and then added to 48-well plates (100 μL/well) in three groups (three sub-wells per group): VEGF, EPCs, and VEGF + EPCs. VEGF group was inoculated with VEGF 1 μg, EPCs group with 1 Â 10 6 EPCs, and VEGF + EPCs group with VEGF 1 μg and 1 Â 10 6 EPCs. Next, the Matrigel was incubated at 37 C for 1 h to form the gel and implanted subcutaneously in the abdomen of Wistar rats (n = 9). The grafts were obtained 7 days postoperatively for gross visualization and histological analysis. SOG staining showed that GAG was abundantly expressed in the cartilage of the native trachea. After decellularization, the expression of GAG was still evident in the VADT group (Figure 1a, E-F). GAG quantitative analysis showed that the GAG content in VADT scaffolds (9.64 ± 0.22 μg/mg) was significantly lower than in the native group (10.8 ± 0.23 μg/mg) ( p < 0.01, Figure 1d). However, 89% of the GAG content was retained in the VADT group compared with that in the native group. IF staining showed that Col-II was highly expressed in the cartilage area of both the native trachea and VADT scaffolds ( Figure 1b, G-H). There was no difference in the total collagen content between the VADT group (8.49 ± 0.31 μg/mg) and the native group (8.76 ± 0.48 μg/mg) ( p > 0.05, Figure 1e).

| Fabrication and characterization of 3D hybrid scaffolds
Although the VADT scaffold retained an intact extracellular matrix structure while becoming non-immunogenic, the longitudinal compression performance was significantly reduced. PCL is widely used in regenerative medicine, owing to its good mechanical properties and biocompatibility. 39 To extract the respective advantages of natural and synthetic materials and further improve the biomechanical properties of VADT, 3D printed PCL stents were prepared, and hybrid tracheal grafts were constructed. The morphology of the 3D printed stents and VADT/PCL hybrid grafts are shown in Figure S1.  (Table S1).

| Extraction and identification of EPCs
In this study, the whole BM differential adherence method combined with an EGM-2MV medium was used to isolate and culture the EPCs.
Flow cytometry analysis showed that CD31 (positive rate 98.4%) and CD34 (positive rate 99.6%) were expressed at significantly high levels, whereas CD44 (positive rate 6.78%) and CD105 (positive rate 3.27%) were expressed at significantly low levels in the group of EPCs ( Figure 3a). The IF staining showed that the CD31, CD34, and VEGFR2 were expressed significantly in these cells (Figure 3b). These experimental results comprehensively indicated that the extracted cells were EPCs.  and EPCs for 1 week showed that the in vivo microvascular formation effect of Matrigel mixed with VEGF alone or EPCs alone was less than that when they were combined, indicating that the rational use of VEGF can further enhance the in vivo microvascular formation of EPCs (Figure 6b).

| Orthotopic transplantation studies
We subsequently performed a tracheal transplantation in a rabbit model system to further evaluate the preclinical utility of biomimetic in situ tracheal microvascularization for segment tracheal reconstruction in one-step. SEM showed significant cell adhesion on the luminal and external surfaces of the grafts in Groups E and F after preoperative cell seeding in vitro (Figure 6c). Without ectopic embedding for microvascularization, the grafts were used to repair a segmental tracheal defect using end-to-end anastomosis directly (Figure 7a). None of the rabbits in Groups A, B, and C survived for 4 weeks, owing to anastomotic stenosis, tracheomalacia, or asphyxia due to phlegm blockage. In Groups D and E, the 4-week survival rate was only 20%, and the main causes of death were anastomotic stenosis, graft necrosis, and phlegm blockage asphyxiation. In contrast, 80% of the rabbits survived for 4 weeks with no evidence of respiratory distress in Group F ( Table 1)

| DISCUSSION
Although tissue-engineering technology has provided considerable development opportunities for tracheal reconstruction surgery, it also faces many challenges, such as the preparation of low-immunogenic tracheal grafts, maintaining good biomechanical properties of the grafts, and rapid microvascularization of the grafts. These are the key challenges that determine the success of tracheal-replacement therapy and the long-term survival of the recipients. In this study, we developed a novel hybrid scaffold with extremely low immunogenicity, suitable biomechanical properties, and good biocompatibility for tissue-engineered tracheal transplantation. Furthermore, we developed a new method for the microvascularization of tracheal grafts in one step, which directly promoted the microvascular network formation of the graft in situ. Rapid vascularization is crucial for cell proliferation, differentiation, and migration that effectively promote graft epithelialization, functionalization, and the long-term survival of transplant recipients. 30 In the preparation of a full biomimetic tracheal graft, it is important to select a suitable tracheal scaffold that has extremely low immunogenicity, a complete extracellular matrix structure, no toxic side effects, and is suitable for cell adhesion and proliferation. It has been confirmed that a natural decellularized tracheal extracellular matrix (dtECM) that meets the above requirements can be effectively obtained using decellularization. 20,35,36 It has been demonstrated that dtECM supports neo-epithelialization, endothelialization, and chondrocyte viability, 40 and it may serve as a promising biomaterial for tracheal reconstitution. 41 Although natural dtECM meets the requirements of F I G U R E 7 (a) Tracheal transplantation in a rabbit model system to evaluate the preclinical utility of biomimetic in situ tracheal microvascularization for long-segment tracheal reconstruction in one step. (b) Bronchoscopy and macroscopic observations. Group D, VADT + PCL + VEGF grafts transplantation; Group E, VADT + PCL + EPC grafts transplantation; and Group F, VADT + PCL + VEGF + EPC grafts transplantation.
most idealized tracheal grafts, the critical limitation is that the biomechanical properties of the scaffold are considerably reduced, resulting in serious complications, such as the softening and stenosis of the grafts in the early stage after orthotopic transplantation. 14,42 Pure synthetic tracheal grafts have favorable biomechanical properties; however, they exhibit numerous graft-related complications, including obstructive granulation tissue and anastomotic leaks, lack of vascularization, epithelial lining, or fusion into surrounding tissues. 43 Therefore, PCL stents that were matched with the native trachea were prepared using 3D printing technology. Hybrid tracheal grafts were constructed using VADT and PCL stent; the advantages of both natural and synthetic materials were utilized to further optimize the performance of the bionic tracheal grafts. Sun and colleagues 44 demonstrated that PCL remained structurally intact in rats during 2 years of in vivo implantation and broke into low molecular weight pieces at the end of 30 months.
Radiation tracer tests confirm that the material does not accumulate in tissues and can be completely excreted from the body. Dias and colleagues 45 demonstrated that PCL degrades more slowly in vivo than in vitro and does not cause rejection in vivo. Therefore, the graft degradation in vivo was not performed in this research. In the following studies, tritium-labeled hybrid scaffolds will be implanted into rabbits and tested for radioactive tracers in plasma, feces, and urine to study their absorption and excretion. Notably, the decellularized implants lack an epithelial layer; hence, they may not be suited to resolve complications associated with the elimination/clearance of mucus, airborne particulates, and bacteria. Therefore, promoting rapid epithelial cell migration from the host trachea growth on the graft luminal of the primary trachea is extremely important.
A critical challenge for the orthotopic transplantation of a tissueengineered trachea is the construction of the vascularization of the graft to provide nourishment to seed cells and promote the functionalization of neonatal tissues. Without effective microvascularization, cells in a biological microenvironment cannot acquire sufficient nutrients and oxygen or remove metabolic waste, which eventually leads to cell death and graft necrosis. 46 Because the tracheal blood supply is primarily sourced from the inferior thyroid artery, superior bronchial artery, esophagotracheal microvascular network, and a lack of arteriovenous vessels for anastomosis, microvascularization of the tracheal graft using vascularized seed cells is particularly important. EPCs are superior to vascular-derived ECs in forming vascular networks, and the vascular structures formed in vitro and in vivo possess permeability values similar to those of vascular-derived ECs. 47 Moreover, EPCs can be non-invasively isolated from peripheral blood, BM, and umbilical cord blood, as well as from human induced pluripotent stem cells, avoiding trauma and immunogenicity issues. [48][49][50][51][52] In this study, we isolated and cultured EPCs as microvascular seed cells using a differential adherence method. Flow cytometry analysis showed that CD31 (positive rate 98.4%) and CD34 (positive rate 99.6%) were significantly overexpressed, while CD44 (positive rate 6.78%) and CD105 (positive rate 3.27%) were underexpressed (Figure 3a). Furthermore, IF staining showed that the CD31, CD34, and VEGFR2 were significantly expressed in these cells (Figure 3b). Therefore, we identified the extracted cells as EPCs. In vivo experiments using a CAM confirmed clear microvascular formation both on the surface and around the matrix in the VADT + EPC group ( Figure 5). This shows that EPCs can induce VADT angiogenesis in vivo without rejection.
Matrigel is the most frequently used material in determining angiogenesis by mimicking the physiological cell substances in vitro and in vivo. 53 When VEGF is directly applied to scaffolds for angiogenesis, proper angiogenesis cannot occur due to the rapid diffusion of VEGF. Therefore, Matrigel could be used as a gel coating to maintain the VEGF concentration gradient and to control its release. 54    A long-term (more than 6 months) survival outcome after tubular tracheal transplantation needs to be explored in the future.
The microvascular density of the grafts in this study was analyzed via analysis of thin section staining (e.g., CD31, ɑ-SMA, and vWF).

| CONCLUSIONS
We developed a 3D printing macro-PCL/VADT hybrid tracheal graft based on BM-EPC seeding that supports vascularization in vivo. The hybrid tracheal graft has extremely low immunogenicity, good biocompatibility, angiogenesis-inducing properties, and biomechanical properties that match those of the native trachea. Furthermore, EPCs were isolated and purified as vascularized seed cells using the differential adhesion method, which further promoted microvascularization in VADT in vivo. Subsequently, the segment hybrid tracheal graft was subjected to one-step in situ replacement therapy in vivo.
Additionally, EPCs were implanted preoperatively and intraoperatively combined with VEGF management, which rapidly promoted graft microvascularization. Therefore, this study demonstrated that the 3D Hongcan Shi: Funding acquisition (lead); project administration (lead); supervision (lead).

ACKNOWLEDGMENTS
This study was financially supported by the National Natural Science Foundation of China (No. 82070020) and Outstanding Doctoral Dissertation Fund of Yangzhou University.