If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of General Thoracic Surgery, Hospital Clinic, Fundació Clínic, Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), CIBER Enfermedades Respiratorias, Universitat de Barcelona, Barcelona, Spain
Department of General Thoracic Surgery, Hospital Clinic, Fundació Clínic, Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), CIBER Enfermedades Respiratorias, Universitat de Barcelona, Barcelona, Spain
Address for reprints: Paolo Macchiarini, MD, PhD, Department of General Thoracic Surgery, Hospital Clinico de Barcelona, University of Barcelona, c. Villarroel 170 E-08036 Barcelona, Spain.
Department of General Thoracic Surgery, Hospital Clinic, Fundació Clínic, Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), CIBER Enfermedades Respiratorias, Universitat de Barcelona, Barcelona, Spain
We sought to bioengineer a nonimmunogenic tracheal tubular matrix of 6 cm in length and test its structural, functional, and immunologic properties in vitro and in vivo.
Methods
Twelve-centimeter tracheal segments were harvested from Yorkshire boars. Half of each segment was subjected to a detergent–enzymatic method (containing sodium deoxycholate/DNase lavations) of decellularization for as many cycles as needed, and the other half was stored in phosphate-buffered saline at 4°C as a control. Bioengineered and control tracheas were then implanted in major histocompatibility complex–unmatched pigs (allograft) or mice (xenograft) heterotopically for 30 days. Structural and functional analysis and immunostaining were performed after each detergent–enzymatic method cycle and transplantation.
Results
Compared with control tracheas, bioengineered matrices displayed no major histocompatibility complex class I and II antigens after 17 detergent–enzymatic method cycles, without significant (P > .05) differences in their strain ability (rupture force, 56.1 ± 3.3 vs 55.5 ± 2.4 N; tissue deformation at 203% ± 13% vs 200% ± 8% or 12.2 ± 0.8 vs 12 ± 0.5 cm; and applied maximum force, 173.4 ± 3.2 vs 171.5 ± 4.6 N). Thirty days after implantation, significantly (P < .01) smaller inflammatory reactions (392 vs 15 macrophages/mm2 and 874 vs 167 T lymphocytes/mm2) and P-selectin expressions (1/6 vs 6/6) were observed in both the xenograft and allograft models with bioengineered matrices compared with those seen with control tracheas. There was no development of anti-pig leukocyte antigen antibodies or increase in both IgM and IgG content in mice implanted with bioengineered tracheas.
Conclusions
Bioengineered tracheal matrices displayed similar structural and mechanical characteristics to native tracheas and excite no immune response to 30 days when implanted as allografts or xenografts. This method holds great promise for the future of tissue-engineered airway replacement.
Unfortunately, the resectable length of the diseased trachea is usually restricted to approximately 30% of the total length in children and around 6 cm in adults, and any further increase in this resection rate depends on the development of a safe tracheal replacement.
This last is not yet clinically available because almost every attempt to provide an autologous or synthetic safe and reproducible tracheal graft has been disappointing thus far,
have been reported. However, the bioengineering process in these previous experiences was labor intensive and required too many steps to gain wide clinical acceptance. Thus, and encouraged by our recent in vitro generation of short but vital tracheal matrices,
we aimed to study bioengineered tubular tracheal matrices of at least 6 cm in length and evaluate their outcome without immunosuppression in allograft and xenograft animal heterotopic models.
Materials and Methods
Donor and recipient pigs were 25 outbred Yorkshire Duroc boars (Isoquimen S/L, Barcelona, Spain) weighing 42.4 ± 3.3 kg. Sixteen CD-1 mice (Charles River Laboratories Italia S.r.l., Calco, Italy) also acted as recipients. All animals received care in compliance with the “Principles of laboratory animal care” formulated by the National Society for Medical Research and the “Guide for the care and use of laboratory animals” prepared by the Institute of Laboratory Animal Resources, National Research Council, and published by the National Academy Press, revised 1996. The study was approved by the Animal Care and Use Committee and the Bioethics Committee of the University of Barcelona.
Study Design
Twelve-centimeter tracheal segments were harvested from donor pigs. Half of each was allocated to the bioengineering protocol previously described,
and half was stored as a control. Structural and morphologic studies were performed. When no major histocompatibility complex (MHC) could be detected, decellularization cycles ceased, and donor tracheal segments were implanted heterotopically into either MHC–unmatched pigs or mice (Figure 1). Biopsy specimens and blood samples were taken from recipient animals weekly until 30 days after implantation.
Figure 1Study design. The term native trachea represents natural untreated tracheas harvested from healthy animals. DEM, Detergent–enzymatic method; control trachea, native trachea.
Pigs (n = 13) were premedicated with intramuscular azaperone (4 mg/kg; Esteve S.A., Barcelona, Spain) and intravenous thiopental (10 mg/kg; B. Braun Medical S.A., Rubi, Barcelona, Spain) and relaxed with intravenous vecuronium (6 mg · kg−1 · h−1; Norcuron, Organon S.A., Barcelona, Spain). Orotracheal intubation was performed through a 7.5F endotracheal tube under bronchoscopic guidance. Anesthesia was maintained by means of continuous infusion of fentanyl (1 μg· kg−1 · h−1; B. Braun Medical S.A.) and propofol (3-5 mg· kg−1 · h−1; B. Braun Medical S.A.). Peripheral oxygen arterial saturation was permanently monitored with a pulse oximeter (BCI, Inc, Waukesha, Wis) placed at the pig's tail. A lower median cervicotomy permitted dissection of the entire trachea, as previously described.
Then death was induced by using an intravenous bolus of fentanyl, propofol, and potassium chloride (40 mEq; B. Braun, Melsungen, Germany). On extubation, the trachea from the first ring to the carina was harvested under sterile conditions. Pigs were then killed by using an approved method (overdose of intravenous anesthetic agent).
The harvested segment was then divided into two 6-cm halves. One half (control) was placed in a stock solution made of phosphate-buffered saline (PBS; Invitrogen S.A., Barcelona, Spain), containing 1% antibiotic and antimycotic solution (AA; Sigma Chemical Company, Barcelona, Spain). The other half was similarly stored for 24 hours but then submitted to the bioengineering protocol, as below. Additionally, small sections were stored at −80°C for immunohistology as a baseline native control.
Matrix Bioengineering
The overlying tissue of the harvested tracheas was stripped off, deprived of trachealis muscle, and rinsed 4 times (for 4 hours each) in PBS containing 1% antibiotic and antimycotic solution. Thereafter the protocol continued as previously reported.
Briefly, tissue was processed with multiple treatment cycles, including the following steps: tissue was stored in Aqua milliQ (Millipore, Madrid, Spain) for 48 hours at 4°C and then incubated in 4% sodium deoxycholate for 4 hours and 2000 kU of DNase-I in 1 mol/L NaCl (Sigma Chemical Company) for 4 hours. The samples were stored in PBS (containing 1% antibiotic and antimycotic) at 4°C up to a maximum of 2 months. The presence of cellular elements and MHC+ cells were verified histologically after each cycle.
Matrix Histology and Immunochemistry
Samples were washed thoroughly in saline before use. To quantify the remaining cells after each cycle of the detergent–enzymatic method (DEM) treatment, we analyzed 10 slides with an optical microscope. The samples were covered with Vectashield (Vector Laboratories, Inc, Burlingame, Calif) mounting medium for fluorescence with 4′-6-diamidino-2-phenylindole (Vector Laboratories, Inc), and the total number of nuclei was visualized at 250× magnification by using fluorescence microscopy. The cell density was expressed as the number of nuclei × 105/μm2. The mean ± 1 standard deviation (SD) was determined for each cycle analysis. Paraffin-embedded tissue sections measuring 5 μm were mounted on slides and stained with hematoxylin and eosin (Merck, Darmstadt, Germany) to evaluate morphologic changes. The presence of MHC markers was evaluated by means of immunostaining. After each DEM cycle, we fixed aliquots of the trachea with 10% neutral buffered formalin for 24 hours and embedded on paraffin. Vertical sections were incubated for 30 minutes at room temperature with PBS containing 10% fetal calf serum (Invitrogen S.A.). After that, samples were incubated at 37°C for 1 hour with monoclonal anti-MHC class I OX27 and anti-MHC class II OX4 antibodies (Abcam, Cambridge, United Kingdom) diluted in 1% fetal calf serum–PBS (1:400), and then labeled with the avidin/biotin–amplified immunoperoxidase method by using the Large Volume Dako LSAB Peroxidase Kit (Dako, Glostrup, Denmark). A biotinylated goat anti-mouse antibody (H+L; Vector Laboratories, Inc) served as a secondary antibody. Streptavidin–peroxidase conjugate was applied, and final staining was performed with diaminobenzidine slides and counterstained with hematoxylin. For fluorescence microscopy, a fluorescein isothiocyanate–conjugated secondary antibody was used.
Physical Strain Tests
Both native tracheas and bioengineered matrices were tested for mechanical properties after varying numbers of cycles of DEM. We used a tensile-test device (Zwick/Roell, version Z0.5TS, Barcelona, Spain) able to determine tensile stress under varying rates of strain or elongation with a precision of 0.0001 N and 0.001 mm in force and position, respectively. Each sample was subjected to increasing uniaxial tensile testing until rupture, which was confirmed by the loss of load and the appearance of tears in the tissue. After clamping the specimen into sample holders, a preload (preliminary force) of 2N was applied, and thereafter, the tensile trial started at a constant elongation rate of 1 mm/s at room temperature. The tensile tester recorded in real time the load and elongation to which the tissue was subjected.
Matrix Implantation as an Allograft in Pigs
The recipient animals (n = 12) were randomly assigned (by using a computer-generated code) to one of 2 different groups receiving the bioengineered matrices or native control tracheas. Animals were anesthetized as above, and the groin region was prepared. Tracheal samples were implanted subcutaneously, and their proximal and distal lumens were anchored to the surrounding skin to leave them open.
Postoperatively, antibiotics (cefazolin, 2 g administered intravenously; Lilly, Madrid, Spain) and analgesics (Enantyum, 40 mg/12 hours; Menarini, Barcelona, Spain) were administered intravenously. Animals were examined daily for 30 days for clinical signs of inflammation or rejection and for general health. Weekly general anesthetic endoscopies were performed to biopsy-implanted matrices and surrounding host tissue, as well as to take blood samples to check for the development of antibodies and increased inflammatory response. Thirty days after implantation, animals were killed as above, and the matrices were harvested with a cuff of donor tissue.
Matrix Implantation as a Xenograft in Mice
Mice (n = 16) were anesthetized with isoflurane (Halocarbon Laboratories, River Edge, NY). Matrix samples (1.5 × 1.5 cm) were implanted under the dorsal skin. Mice receiving the matrices were killed on days 7, 15, 23, and 30. Blood samples were taken, and tissue samples were analyzed both macroscopically and microscopically.
Analysis of Tissue From Pig Recipients
Samples were formalin fixed and paraffin embedded before staining with hematoxylin and eosin (Merck).
Anti–swine leukocyte antigen (anti-SLA) antibodies were tested at 7, 15, 23, and 30 days after matrix implantation by using a modification of the standard flow cytometric crossmatch on lymph node cells, as previously described.
Briefly, donor pig lymph node cells were incubated for 30 minutes with recipient serum and rinsed 3 times in PBS, and a fluorescein-marked anti-porcine–immunoglobulin and phycoerythrin-marked anti-porcine CD3 were added for 30 minutes. A FACS-Scan (FACSAria; Becton–Dickinson, Erembodegem, Belgium) was used to evaluate the double fluorescence. The coefficient between the mean channel in test serum and the negative control serum for T cells (defined by mentioned double fluorescence) was recorded. We considered a ratio of greater than 2.9 as positive based on previous samples obtained before implantation. To identify T lymphocytes, monocytes/macrophages, and polymorphonuclear granulocytes, immunochemistry was performed with a rabbit anti-CD3 antibody (Dako) and mouse anti-L1 antibody (Dako), respectively.
P-selectin (CD62P) expression was measured in animals by fixing matrix/native trachea connective blood vessels after 30 days with 2% paraformaldehyde for 3 hours (4°C) before immersing them in OCT medium and freezing in liquid nitrogen. Thereafter, sections were fixed in paraformaldehyde (4%), and the primary antibody (P-sel.KO.2.5) was incubated for 2 hours at room temperature. Negative controls were carried out similarly but omitting the primary antibody. A piece of harvested tissue was stimulated with lipopolysaccharides (Sigma Chemical; 30 μg/kg) and used as a positive control. The secondary antibody (Cy3 goat anti-mouse IgG; Jackson Immuno Research Laboratories, West Grove, Pa) was incubated for 1 hour at room temperature, and sections were mounted with Fluorescent Mounting Medium PCK (Dako). Finally, samples were analyzed by means of confocal microscopy (Leica TCS NT, Heidelberg, Germany), and the quantity of P-selectin was defined as low or high expression based on a computer-added system.
Analysis of Tissue From Murine Recipients
Tissue samples were immersed in 10% buffered formalin for 24 hours and embedded in paraffin. Then sections were stained with hematoxylin and eosin (Merck) and examined for the presence of inflammatory cells. Blood samples from mice were analyzed to identify levels of total circulating IgG and IgM (in milligrams per milliliter) with an enzyme-linked immunosorbent assay kit (ZeptoMetrix, Buffalo, NY).
Statistical Analysis
Continuous variables were compared by using the independent-samples t test. The odds ratio was calculated to perform between-group comparisons of categorical variables. Results are presented as means ± standard deviations of the mean. Analyses were made with the SPSS package (version 12.0; SPSS, Inc, Chicago, Ill).
Results
In Vitro Evaluation
The bioengineering process lasted 35 ± 1, days, corresponding to 17 DEM cycles, after which the bioengineered tracheal matrices showed few residual nuclei of chondrocytes but a complete removal of MHC class I and II antigens. By contrast, MHC class I and II expression was ubiquitous in control tracheas (Table 1 and Figure 2, Figure 3, Figure 4). The bioengineered tracheas showed no significant differences in their strain ability compared with control tracheas until the 17th cycle (Table 2 and Figure 4, B). However, a higher (≥18 cycles) number of DEM cycles led to a loss of the matrices' properties. The mechanical properties of the bioengineered tracheas were similar to those of the native tracheas, and there was no influence of airway diameter on the maximum breaking force. The breaking point for all tracheas was observed at the intercartilaginous septa as horizontal fissures with preservation of the physiologic tissue behavior. All matrices could be stored after the engineering process in PBS for at least 2 months without losing their characteristics.
Table 1In vitro morphologic and immunologic evaluation of native and bioengineered tracheas
The term native trachea represents natural untreated tracheas harvested from healthy animals. DEM, Detergent–enzymatic method; MHC, major histocompatibility complex; +++, very high expression; +, low expression; –, not expressed.
Figure 2Transverse sections of native trachea (A–D) and tracheal matrices (E–H) after 17 cycles of detergent–enzymatic treatment immunostained with 4′-6-diamino-2-phenylindole for nuclei (A, C, E, and G) or with monoclonal anti-MHC class I (B and F) and II (C and H). Neither MHC class I– nor MHC class II–presenting cells were detected in the bioengineered matrices. The term native trachea represents natural untreated tracheas harvested from healthy animals.
Figure 3Transverse sections of native trachea (A) and tracheal matrices (B) after 17 cycles of detergent–enzymatic treatment stained with hematoxylin and eosin. (Original magnification ×200.) Seventeen cycles were not able to completely decellularize tracheal specimens, and nuclei were still visible inside the cartilage ring, whereas cell walls disappeared by comparison with the native trachea. Arrows indicate intact chondrocytes in native trachea or nuclei with absent cell walls in bioengineered matrices. The term native trachea represents natural untreated tracheas harvested from healthy animals.
Figure 4A, Remaining nuclei of chondrocytes during the bioengineering procedure. ∗P < .01 versus native trachea. The line within the box corresponds with the median, and the lined border corresponds with the standard deviation. B, A box plot of the maximum force comparing native with bioengineered matrices. ∗P < .01 versus native trachea. NS, Nonsignificant differences compared with native trachea. The line within the box corresponds with the median, and the lined border corresponds with the standard deviation. NT, Native tracheas. The term native trachea represents natural untreated tracheas harvested from healthy animals.
The term native trachea represents natural untreated tracheas harvested from healthy animals. DEM, Detergent–enzymatic method; Maximum force, applied maximum force; N, Newton.
In both allotransplantation and xenotransplantation models, the bioengineered matrices displayed no signs of bioincompatibility in terms of hyperacute, acute, or chronic rejection; macroscopic inflammatory response; or health impairment during the study period. After 30 days of implantation, the bioengineered matrices showed a significantly (P < .01) lower inflammatory reaction in both models compared with that seen in control tracheas, which became partially necrotic. There was no detectable SLA antibody in recipient pigs or an increase in IgM or IgG levels in recipient mice. The expression of P-selectin was significantly greater in animals receiving control tracheas than in those receiving the bioengineered matrices (P < .01; Table 3, Table 4 and Figure 5).
Table 3In vivo evaluation of the bioengineered tracheal matrices implanted heterotopically in HLA-unmatched pigs
Figure 5A, High P-selectin expression on epithelial cells in small vessels of graft surrounding tissue in animals receiving native trachea. (Original magnification ×100.) B, Low P-selectin expression on epithelial cells in small vessels of graft surrounding tissue in animals receiving bioengineered matrices. (Original magnification ×100.) The term native trachea represents natural untreated tracheas harvested from healthy animals.
By contrast with other successful organ replacements, which take place in sterile mesenchymal environments (eg, the liver, kidney, and heart), the airway represents an interface between the mammal and the external environment. Unsurprisingly, its mucosa holds immunologically active cells playing a key role in airway transplantation,
a completely nonimmunogenic tracheal allograft with preserved functional and mechanical characteristics represents a necessary ideal for organ replacement.
The present results demonstrate that the DEM process is a simple and effective method to bioengineer not only in vitro
but also in vivo tracheal matrices lacking any MHC antigens while maintaining a structural integrity similar to that of native tracheas and, most importantly, of sufficient length to have potential clinical application. In contrast to other replacement methods, such as chemical fixation, cryopreservation, or lyophilization,
this study presents evidence that the time needed to constantly obtain a nonimmunogenic graft is around 35 days or 17 cycles and that this length of processing does not affect their mechanical properties.
Although all epithelial, interstitial, muscle, and gland cells disappeared after the 17 DEM cycles, some chondrocyte nuclei remained. The fact that they were unable to trigger any immunologic response might be explained by the fact that chondrocyte antigens are expressed only on their membranes and not on their nuclei. On the contrary, such nuclei might even have a positive effect on the biocompatibility of the matrices and the preservation of their excellent mechanical characteristics, which mimic those of natives trachea.
Interestingly, the immunologic and mechanical characteristics of the matrices remained unaffected by a 2-month storage in PBS (containing 1% antibiotic and antimycotic, at 4°C). Moreover, higher DEM cycles (>17) are to be avoided because they worsen the mechanical properties of the matrices to such an extent that implantation could be jeopardized. Undoubtedly, the DEM-generated matrices resemble normal trachea far more than those generated by using our previously reported protocols.
To confirm the nonimmunogenic properties of the DEM-generated matrices, we implanted them in HLA-unmatched pigs and mice, mimicking allotransplantation and xenotransplantation settings. Implants were left over a period of 30 days without immunosuppression. They displayed no histologic signs of local or graft rejection and significantly less inflammatory reaction and activation of adhesion molecules (eg, P-selectin) or expression of SLA antibodies when compared with control or untreated tracheas during the study period. One possible contributory factor to avoidance of hyperacute and acute rejection is the lack of an immediate dedicated blood supply, although vessels were seen to grow into the implanted tissue with time.
In conclusion, our findings suggest that 17 cycles or around 35 days of DEM are necessary to bioengineer a tracheal tubular matrix that is structurally and mechanically similar to native trachea. We show that higher cycles should be avoided to maintain mechanical force and strength. The immunologic ignorance resulting from this procedure avoids allorejection and xenorejection in the present animal models, and therefore there is no need for immunosuppression. We believe this protocol opens the door to the creation of clinically functional, fully tissue-engineered airway replacements in the near future.
We acknowledge the assistance of Laura Morte for her excellent organization and troubleshooting. Furthermore, we would like to express our gratitude to Pablo Engel and Jordi Sintes for kindly providing the P-Selectin antibody. Finally, we are indebted to Victor Peinado and others on his team for their superb assistance.
00772-7&id=gr1.jpg){kind=link}
00772-7&id=gr2.jpg){kind=link}
00772-7&id=gr3.jpg){kind=link}
00772-7&id=gr4.jpg){kind=link}
00772-7&id=gr5.jpg){kind=link}