Artificial trachea transplant: Tale of two stories.

South Korean-based team is first to successfully transplant 3D bioprinted artificial trachea. The success arises during scrutiny of artificial tracheal implants stemming from the denounced work of Dr. Paolo Macchiarini.

Changes in the Proportion of Each Cell Type After hiPSC-Derived Airway Epithelia Transplantation.

No radical treatment is available for the regeneration of dysfunction and defects in airway epithelia. Artificial tracheae made of polypropylene and collagen sponge were used in clinical studies to reconstitute tracheae after resection. For early epithelialization of the luminal surface of the artificial trachea, a model was established, that is, an artificial trachea covered with human-induced pluripotent stem cell-derived airway epithelial cells (hiPSC-AECs) was transplanted into a tracheal defect in an immunodeficient rat. Unlike the cell types of hiPSC-derived cells that are currently used in clinical studies, AECs maintain tissues by proliferation and differentiation of basal cells into various cell types that constitute AECs constantly. Therefore, post-transplantation, the proportion of each cell type, such as ciliated and goblet cells, may change; however, no studies have examined this possibility. In this study, using our hiPSC-AEC-transplanted rat model, we investigated changes in the proportion of each cell type in hiPSC-AECs pre-transplantation and post-transplantation. As a result, the proportion of each cell type changed post-transplantation. The proportion of ciliated, basal, and club cells increased, and the proportion of goblet cells decreased post-transplantation. In addition, the proportion of each cell type in engrafted hiPSC-AECs is more similar to the proportion of each cell type in normal proximal airway tissue than the proportion of each cell type pre-transplantation. The results of this study are useful for the development of therapeutic techniques using hiPSC-AEC transplantation.

Human tracheal transplantation: A systematic review of case reports.

BACKGROUND: Patients with long-segment airway stenosis not amenable to conventional surgery may benefit from tracheal transplantation. However, this procedure has been only anecdotally reported, and its indications, techniques, and outcomes have not been extensively reviewed. METHODS: We conducted a systematic Literature search to identify all original articles reporting attempts at tracheal transplantation in humans. RESULTS: Of 699 articles found by the initial search, 11 were included in the systematic review, describing 14 cases of tracheal transplantation. Patients underwent transplantation for benign stenosis in nine cases, and for malignancies in five cases. In 12 cases blood supply to the trachea was provided by wrapping the graft in a vascularized recipient's tissue, while in 2 cases the trachea was directly transplanted as a vascularized composite allograft. The transplantation procedure was aborted before orthotopic transplantation in two patients. Among the remaining 12 patients, there was 1 operative mortality, while 4 patients experienced complications. Immunosuppressants drugs were administered to the majority of patients postoperatively, and only one group of authors attempted their withdrawal, in five patients. At the end of follow-up, all 11 patients surviving the operation were alive, but 2 had a recurrent tracheal stenosis requiring an airway appliance for breathing. CONCLUSION: Human tracheal transplantation is still at an embryonic phase. Studies available in the Literature report different surgical techniques, and information on long-term outcomes is still limited. Future research is needed in order to understand the clinical value of this procedure.

De-Epithelialized Viable Tracheal Allotransplantation Without Immunosuppressants: 5-Year Follow-Up.

OBJECTIVE: Tracheal transplantation could be a better option for patients with long segmental laryngotracheal stenosis or defects, but the need for immunosuppressants limits its widespread use due to the antigenicity of the tracheal epithelium. Chemically treated or cryopreserved nonviable tracheal allografts have no immunogenicity but lead to necrosis and stenosis in long-term outcomes. The present report describes the 5-year outcomes of de-epithelialized viable tracheal allotransplantation without immunosuppressants in a patient with severe laryngotracheal stenosis. METHODS: The recipient was a 47-year-old female with relapsing polychondritis affecting the larynx and cervical trachea and producing a 5 cm long stenosis that could not be repaired using resection and anastomosis. A tracheal allograft was obtained from a 45-year-old male donor and treated with a combination of 3% sodium dodecyl sulfate (SDS) and organ preservation solution for 138 hours. The allograft was revascularized by heterotopical implantation in the infrahyoid muscles of the recipient for 3 months and then transplantation to the laryngotracheal defect with a split-thickness skin graft sutured to the lumen and a silicon T-tube. No immunosuppressants were used postoperatively. RESULTS: The allograft was de-epithelialized, and most of the cartilage rings remained viable after the treatment. The allograft was revascularized, viable, and mechanically stable after 3 months of heterotopic implantation. No apparent signs of rejection or destruction were observed. The T-tube was removed, and the internal lining of the allograft was repopulated 4 months after orthotopic transplantation, despite the skin graft necrotizing at 2 weeks. Endoscopy and computed tomography showed a patent airway 5 years after orthotopic transplantation. The patient was able to resume her usual quality of life. CONCLUSION: The present study demonstrates that transplantation of the de-epithelialized viable tracheal allograft without immunosuppressants is safe and promising for patients with long laryngotracheal stenosis or defects, especially for those with malignant tumor resections.

Murine Intrapulmonary Tracheal Transplantation: A Model for Investigating Obliterative Airway Disease After Lung Transplantation.

Murine intrapulmonary tracheal transplantation (IPTT) is used as a model of obliterative airway disease (OAD) following lung transplantation. Initially reported by our team, this model has gained use in the study of OAD due to its high technical reproducibility and suitability for investigating immunological behaviors and therapeutic interventions. In the IPTT model, a rodent tracheal graft is directly inserted into the recipient's lung through the pleura. This model is distinct from the heterotopic tracheal transplantation (HTT) model, wherein grafts are transplanted into subcutaneous or omental sites, and from the orthotopic tracheal transplantation (OTT) model in which the donor trachea replaces the recipient's trachea. Successful implementation of the IPTT model requires advanced anesthetic and surgical skills. Anesthetic skills include endotracheal intubation of the recipient, setting appropriate ventilatory parameters, and appropriately timed extubation after recovery from anesthesia. Surgical skills are essential for precise graft placement within the lung and for ensuring effective sealing of the visceral pleura to prevent air leakage and bleeding. In general, the learning process takes approximately 2 months. In contrast to the HTT and OTT models, in the IPTT model, the allograft airway develops airway obliteration in the relevant lung microenvironment. This allows investigators to study lung-specific immunological and angiogenic processes involved in airway obliteration after lung transplantation. Furthermore, this model is also unique in that it exhibits tertiary lymphoid organs (TLOs), which are also seen in human lung allografts. TLOs are comprised of T and B cell populations and characterized by the presence of high endothelial venules that direct immune cell recruitment; therefore, they are likely to play a crucial role in graft acceptance and rejection. We conclude that the IPTT model is a useful tool for studying intrapulmonary immune and profibrotic pathways involved in the development of airway obliteration in the lung transplant allograft.

[The first allogenic tracheal transplantation in humans 45 years ago : Procedure, course, and outcome]. / Die erste allogene Trachealtransplantation beim Menschen vor 45 Jahren : Vorgehen, Verlauf und Ergebnis.

In November/December 1978, the first successful tracheal transplantation in humans was performed at the University ENT Clinic in Cologne by the then senior physicians Kurt G. Rose (later chief physician in Dortmund) and Klaus Sesterhenn (later chief physician in Duisburg). Director of the clinic at that time was Prof. Dr. Dr. Fritz Wustrow [10]. The immunological foundations and preliminary work were laid by Sesterhenn in the context of a total of 338 tracheal transplants in Lewis rats in the 1970s (details in the text). The first successful tracheal transplantation was performed on 18 November 1978 in a, then 19-year-old patient who had previously had a motorcycle accident. The donor organ was explanted in the University Hospital Essen and transplanted about 160 min later in the Cologne University ENT Clinic, first into a pocket of the right sternocleidomastoid muscle. The definitive transplantation took place on 06 December 1978. In the article, the circumstances at that time and the perioperative course in the Cologne University ENT Clinic are described by an eyewitness. The former patient is still well and without complications after more than four decades.

Current status and future trends of reconstructing a vascularized tissue-engineered trachea.

Alternative treatment of long tracheal defects remains one of the challenges faced by thoracic surgeons. Tissue engineering has shown great potential in addressing this regenerative medicine conundrum and the technology to make tracheal grafts using this technique is rapidly maturing, leading to unique therapeutic approaches. However, the clinical application of tissue-engineered tracheal implants is limited by insufficient revascularization. Among them, realizing the vascularization of a tissue-engineered trachea is the most challenging problem to overcome. To achieve long-term survival after tracheal transplantation, an effective blood supply must be formed to support the metabolism of seeded cells and promote tissue healing and regeneration. Otherwise, repeated infection, tissue necrosis, lumen stenosis lack of effective epithelialization, need for repeated bronchoscopy after surgery, and other complications will be inevitable and lead to graft failure and a poor outcome. Here we review and analyze various tissue engineering studies promoting angiogenesis in recent years. The general situation of reconstructing a vascularized tissue-engineered trachea, including current problems and future development trends, is elaborated from the perspectives of seed cells, scaffold materials, growth factors and signaling pathways, surgical interventions in animal models and clinical applications. This review also provides ideas and methods for the further development of better biocompatible tracheal substitutes in the future.

Comparison of the biological properties between 3D-printed and decellularized tracheal grafts.

This study sought to characterize the differences between the 3D-printed and decellularized tracheal grafts, providing the basis for the synthesis of the more reasonable and effective tissue-engineered trachea. We compared the biomechanical properties and biocompatibility of the 3D-printed tracheal graft and decellularized tracheal graft in vitro and evaluated the biocompatibility, immune rejection and inflammation of the two materials through in vivo implantation experiments. Compared with the decellularized tracheal graft, the 3D-printed tracheal graft was associated with obviously higher biomechanical properties. The results demonstrated enhanced growth of BMSCs in the decellularized tracheal graft compared to the 3D-printed one when co-culture with two tracheal graft groups. Moreover, the CCK-8 assay demonstrated significant cell proliferation on the decellularized tracheal graft. Serum IgG and IgM measured in vivo by implantation testing indicated that the 3D-Printed tracheal graft exhibited the most significant inflammatory response. HE staining indicated that the inflammatory response in the 3D-printed tracheal graft consisted mainly of eosinophils, while little inflammatory cell infiltrates were observed in the decellularized tracheal graft. CD68 immunohistochemical analysis indicated that the infiltration of macrophages was not significant in both tracheal grafts. Our findings suggest that the biomechanical properties of the 3D-printed tracheal grafts are better than the decellularized tracheal grafts. Nonetheless, the decellularized tracheal graft exhibited better biocompatibility than the 3D-printed tracheal graft.

Application of aortic allograft in trachea transplantation.

BACKGROUND: The use of tracheal implants for tracheal reconstruction remains a challenge in thoracic medicine due to the complex structure of the trachea in mammalian organisms, including smooth muscles, cartilage, mucosa, blood vessels, cilia, and other tissues, and the difficulty in achieving tracheal regeneration using implants from either allografts or synthetic biomaterials. METHODS: This project used the Lee-Sung strain pig, a swine breed local to Taiwan, as the experimental subject. The aorta of the pig was harvested, decellularized to form the scaffold, and transplanted into the trachea of allogeneic pigs together with growth factors. Postoperative physiological function and tissue changes were observed. The postoperative physiological parameters of the LSP were monitored, and they were sacrificed after a certain period to observe the pathological changes in the tracheal epithelial cells and cartilages. RESULTS: Overall, six LSP tracheal transplantations were performed between March 4, 2020, and March 10, 2021. These included aortic patch anastomosis for pig 1 and aortic segmental anastomosis for pigs 2-6. The shortest and longest survival periods were 1 day and 147 days, respectively. Excluding the pig that survived for only 1 day due to a ruptured graft anastomosis, all other subjects survived for over 1 month on average. CONCLUSION: In this study, we grafted a decellularized porcine aorta into a recipient pig with a tracheal defect. We found cryopreservation of the allogeneic aorta transplantation was a feasible and safe method for the management of airway disease, and immunosuppressants were unnecessary during the treatment course.

Human Tracheal Transplantation.

Long-segment tracheal airway defects may be congenital or result from burns, trauma, iatrogenic intubation damage, or tumor invasion. Although airway defects <6 cm in length may be reconstructed using existing end-to-end reconstructive techniques, defects >6 cm continue to challenge surgeons worldwide. The reconstruction of long-segment tracheal defects has long been a reconstructive dilemma, and these defects are associated with significant morbidity and mortality. Many of these defects are not compatible with life or require a permanent extended-length tracheostomy that is fraught with complications including mucus plugging and tracheoesophageal fistula. Extensive circumferential tracheal defects require a reconstructive technique that provides a rigid structure able to withstand the inspiratory pressures, a structure that will biologically integrate, and contain functional ciliated epithelium to allow for normal mucociliary clearance. Tracheal transplantation has been considered the reconstructive "Holy Grail;" however, there has been a long-held scientific dogma that revascularization of the trachea was not possible. This dogma stifled research to achieve single-staged vascularized tracheal transplantation and prompted the introduction of many creative and inventive alternatives. Throughout history, alloplastic material, nonvascularized allografts, and homografts have been used to address this dilemma. However, these techniques have largely been unsuccessful. The recent introduction of a technique for single-staged vascularized tracheal transplantation may offer a solution to this dilemma and potentially a solution to management of the fatal tracheoesophageal fistula.

A Multimodal Approach to Quantify Chondrocyte Viability for Airway Tissue Engineering.

OBJECTIVES/HYPOTHESIS: Partially decellularized tracheal scaffolds have emerged as a potential solution for long-segment tracheal defects. These grafts have exhibited regenerative capacity and the preservation of native mechanical properties resulting from the elimination of all highly immunogenic cell types while sparing weakly immunogenic cartilage. With partial decellularization, new considerations must be made about the viability of preserved chondrocytes. In this study, we propose a multimodal approach for quantifying chondrocyte viability for airway tissue engineering. METHODS: Tracheal segments (5 mm) were harvested from C57BL/6 mice, and immediately stored in phosphate-buffered saline at -20°C (PBS-20) or biobanked via cryopreservation. Stored and control (fresh) tracheal grafts were implanted as syngeneic tracheal grafts (STG) for 3 months. STG was scanned with micro-computed tomography (µCT) in vivo. STG subjected to different conditions (fresh, PBS-20, or biobanked) were characterized with live/dead assay, terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL), and von Kossa staining. RESULTS: Live/dead assay detected higher chondrocyte viability in biobanked conditions compared to PBS-20. TUNEL staining indicated that storage conditions did not alter the proportion of apoptotic cells. Biobanking exhibited a lower calcification area than PBS-20 in 3-month post-implanted grafts. Higher radiographic density (Hounsfield units) measured by µCT correlated with more calcification within the tracheal cartilage. CONCLUSIONS: We propose a strategy to assess chondrocyte viability that integrates with vivo imaging and histologic techniques, leveraging their respective strengths and weaknesses. These techniques will support the rational design of partially decellularized tracheal scaffolds. LEVEL OF EVIDENCE: N/A Laryngoscope, 133:512-520, 2023.

Early reperfusion with hemoglobin vesicles into tracheal subepithelial capillaries in a mouse tracheal transplant model.

Hemoglobin vesicles (HbVs), liposomes containing concentrated hemoglobin extracted from outdated human red blood cells (RBC), are artificial oxygen carriers with a small particle size. To evaluate the reperfusion of capillaries with HbVs in a tracheal transplant model and compare it with that of RBC. Isogenic mice were used as donors and recipients in a parallel trachea transplant model. Both ends of the donor trachea were anastomosed end-laterally to the recipient trachea to form in parallel. After transplantation, 0.3 mL of HbV solution (Hb concentration, 10 g/dL) was administered via the tail vein. The recipients were euthanized 1, 4, 6, and 8 h after surgery (n = 5 in each group). The tracheas were harvested, and tracheal subepithelial capillaries (SEC) reperfusion was histologically evaluated. A significant number of particles defined as HbV by electron microscopy were observed in the SEC of the grafted tracheas 4 h after the transplant surgery and HbV administration when no RBC were found in the SECs. The number increased 6 and 8 h later. Our findings suggest that HbVs, which are smaller than RBC, can reperfuse the capillaries of grafts earlier than RBCs after transplantation and contribute to the oxygenation of transplanted tissues.

Nintedanib reduces alloimmune-induced chronic airway changes in murine tracheal allografts.

BACKGROUND: The major obstacle for long-term survival after successful lung transplantation is the development of bronchiolitis obliterans (BO) which is one phenotype of chronic lung allograft dysfunction (CLAD). Nintedanib has beneficial effects treating neoplastic diseases and idiopathic pulmonary fibrosis by blocking tyrosine kinase receptors. These receptors play an important role in alloimmune-mediated proliferative diseases. The aim of this study was to determine the effect of nintedanib on proliferative airway changes after orthotopic trachea transplantation in mice. METHODS: C57BL/6 mice (H-2b) donor tracheas were orthotopically transplanted into CBA/J mice (H-2k). After transplantation, recipients were daily treated with nintedanib (60 mg/kg; p.o.). Histological and immunofluorescence analysis were performed after 30 days and intragraft gene expression measurements after 14 days of treatment, respectively. RESULTS: Tracheal allografts from mice treated with nintedanib showed significantly less features of chronic rejection than untreated allografts reflected in a higher epithelium/lamina propria ratio (ELR) [ELR: 0.65 ± 0.13 nintedanib vs. 0.50 ± 0.07 untreated controls; p < 0.05] and a reduced submucosal smooth muscle actin (SMA) content [SMA: 1.26% ± 0.78% nintedanib vs. 2.18% ± 1.01% untreated controls; p < 0.01]. Furthermore, lower T cell, macrophage and dendritic cell infiltration was detected in the nintedanib treated grafts. The protein and intragraft mRNA expression of receptor subtypes was considerably decreased in grafts of nintedanib treated mice. The mRNA expression of relevant immune mediators was affected by nintedanib treatment. CONCLUSION: Receptor blocking by nintedanib reduced alloimmune-induced inflammation and chronic airway changes in mouse trachea allografts and might be a promising approach to diminish the development of BO in lung transplants.

CTLA4-Ig mediated immunosuppression favors immunotolerance and restores graft in mouse airway transplants.

CTLA4-Ig is a potent costimulatory blocker that inhibits T cell activation during alloimmune inflammation and increases graft survival and function. CTLA4-Ig-mediated immunosuppression has been demonstrated to support transplant function in various clinical trials and preclinical settings, but its effects on the balance between regulatory T cells (Tregs) and effector T cells (Teffs), as well as complement activation, are less well investigated. In the present study, we proposed to investigate the effects of CTLA4-Ig mediated immunosuppression on the phase of immunotolerance and the subsequent graft microvascular and epithelial repair during the progression of subepithelial fibrosis in a mouse model of orthotopic trachea transplantation. Briefly, CTLA4-Ig treated allografts (2 mg/kg, I.P.), untreated allografts, and syngrafts were serially monitored for peripheral FOXP3+ Tregs, antibody-mediated complement activation (C3d and C4d), tissue oxygenation, donor-recipient microvascular blood flow, and subsequent tissue remodeling following transplantation. Our data demonstrate that CTLA4-Ig mediated immunosuppression significantly results in late increases in both peripheral CD4+/CD8+ FOXP3+ Tregs and serum IL-10, but prevents the microvascular deposition of IgG, complement factor C3d, and epithelial C4d respectively, which proportionally improved blood flow and tissue oxygenation in the graft and, thus, promotes graft repair. Also, it restored the airway lumen, epithelium, and prevented the progression of subepithelial collagen deposition up to 90 days after transplantation. This study demonstrates that CTLA4-Ig-mediated immunosuppression potentially modulates both effector response and a late surge of regulatory activity to preserve graft microvasculature and rescue allograft from sustained hypoxia and ischemia and thereby limits subepithelial fibrosis.

Biomechanical properties of the ex vivo porcine trachea: A benchmark for three-dimensional bioprinted airway replacements.

PURPOSE: Combining tissue engineering and three-dimensional (3D) printing may allow for the introduction of a living functional tracheal replacement graft. However, defining the biomechanical properties of the native trachea is a key prerequisite to clinical translation. To achieve this, we set out to define the rotation, axial stretch capacity, and positive intraluminal pressure capabilities for ex vivo porcine tracheas. STUDY DESIGN: Animal study. MATERIALS AND METHODS: Six full-length ex vivo porcine tracheas were bisected into 5.5 cm segments. Maximal positive intraluminal pressure was measured by sealing segment ends with custom designed 3D printed caps through which a pressure transducer was introduced. Axial stretch capacity and rotation were evaluated by stretching and rotating the segments along their axis between two clamps, respectively. RESULTS: Six segments were tested for axial lengthening and the average post-stretch length percentage was 148.92% (range 136.81-163.48%, 95% CI 153-143%). The mean amount of length gain achieved per cartilaginous ring was 7.82% (range 4.71-10.95%, 95% CI 6.3-9.35%). Four tracheal segments were tested for maximal positive intraluminal pressure, which was over 400 mmHg. Degree of rotation testing found that the tracheal segments easily transformed 180° in anterior-posterior bending, lateral bending, and axial rotational twisting. CONCLUSIONS: We define several biomechanical properties of the ex vivo porcine trachea by reporting the rotation, axial stretch capacity, and positive intraluminal pressure capabilities. We hope that this will aid future work in the clinical translation of 3D bioprinted airway replacement grafts and ensure their compatibility with native tracheal properties.

Tracheal Macrophages During Regeneration and Repair of Long-Segment Airway Defects.

OBJECTIVES/HYPOTHESIS: Tissue-engineered tracheal grafts (TETGs) offer a potential solution for repair of long-segment airway defects. However, preclinical and clinical TETGs have been associated with chronic inflammation and macrophage infiltration. Macrophages express great phenotypic heterogeneity (generally characterized as classically activated [M1] vs. alternatively activated [M2]) and can influence tracheal repair and regeneration. We quantified and characterized infiltrating host macrophages using mouse microsurgical tracheal replacement models. STUDY DESIGN: Translational research, animal model. METHODS: We assessed macrophage infiltration and phenotype in animals implanted with syngeneic tracheal grafts, synthetic TETGs, or partially decellularized tracheal scaffolds (DTSs). RESULTS: Macrophage infiltration was observed following tracheal replacement with syngeneic trachea. Both M1 and M2 macrophages were present in native trachea and increased during early tracheal repair (P = .014), with an M1/M2 ratio of 0.48 ± 0.15. In contrast, orthotopic implantation of synthetic TETGs resulted in a shift to M1 predominant macrophage phenotype with an increased M1/M2 ratio of 1.35 ± 0.41 by 6 weeks following implant (P = .035). Modulation of the synthetic scaffold with the addition of polyglycolic acid (PGA) resulted in a reduction of M1/M2 ratio due to an increase in M2 macrophages (P = .006). Using systemic macrophage depletion, the M1/M2 ratio reverted to native values in synthetic TETG recipients and was associated with an increase in graft epithelialization. Macrophage ratios seen in DTSs were similar to native values. CONCLUSIONS: M1 and M2 macrophages are present during tracheal repair. Poor epithelialization with synthetic TETG is associated with an elevation of the M1/M2 ratio. Macrophage phenotype can be altered with scaffold composition and host-directed systemic therapies. DTSs exhibit M1/M2 ratios similar to those seen in native trachea and syngeneic tracheal replacement. LEVEL OF EVIDENCE: NA Laryngoscope, 132:737-746, 2022.

Bioinspired carbon monoxide delivery using artificial blood attenuates the progression of obliterative bronchiolitis via suppression of macrophage activation by IL-17A.

Carbon monoxide (CO) is expected to attenuate the progression of obliterative bronchiolitis (OB), which is a serious complication after lung transplantation. However, issues in terms of feasible exogenous CO supply, such as continuousness and safety, remain unsolved. Here, we applied nano red blood cells, namely hemoglobin vesicles (Hb-V), as a CO cargo based on the biomimetic concept and investigated the therapeutic potential of CO-loaded Hb-V on OB in orthotopic tracheal transplant model mice. The CO-loaded Hb-V was comprised of negatively charged liposomes encapsulating carbonylhemoglobin with a size of ca. 220 nm. The results of histological evaluation showed that allograft luminal occlusion and fibrosis were significantly ameliorated by treatment with CO-loaded Hb-V compared to treatment with saline, cyclosporine, and Hb-V. The therapeutic effects of CO-loaded Hb-V on OB were due to the suppression of M1 macrophage activation in tracheal allografts, resulting from decreased IL-17A production. Furthermore, the expression of TNF-α and TGF-ß in tracheal allografts was decreased by CO-loaded Hb-V treatment but not saline and Hb-V treatment, indicating that CO liberated from CO-loaded Hb-V inhibits epithelial-mesenchymal transition. These findings suggest that CO-loaded Hb-V exerts strong therapeutic efficacy against OB via the regulation of macrophage activation by IL-17A and TGF-ß-driven epithelial-mesenchymal transition.