Research ArticleTissue Engineering

Production and transplantation of bioengineered lung into a large-animal model

See allHide authors and affiliations

Science Translational Medicine  01 Aug 2018:
Vol. 10, Issue 452, eaao3926
DOI: 10.1126/scitranslmed.aao3926

New life for lungs

Lungs are complex organs to engineer: They contain multiple specialized cell types in extracellular matrix with a unique architecture that must maintain compliance during respiration. Nichols et al. tackled the challenges of vascular perfusion, recellularization, and engraftment of tissue-engineered lungs in a clinically relevant pig model. Nanoparticle and hydrogel delivery of growth factors promoted cell adhesion to whole decellularized pig lung scaffolds. Autologous cell–seeded bioengineered lungs showed vascular perfusion via collateral circulation within 2 weeks after transplantation. The transplanted bioengineered lungs became aerated and developed native lung-like microbiomes. One pig had no respiratory symptoms when euthanized a full 2 months after transplant. This work represents a considerable advance in the lung tissue engineering field and brings tissue-engineered lungs closer to the realm of clinical possibility.


The inability to produce perfusable microvasculature networks capable of supporting tissue survival and of withstanding physiological pressures without leakage is a fundamental problem facing the field of tissue engineering. Microvasculature is critically important for production of bioengineered lung (BEL), which requires systemic circulation to support tissue survival and coordination of circulatory and respiratory systems to ensure proper gas exchange. To advance our understanding of vascularization after bioengineered organ transplantation, we produced and transplanted BEL without creation of a pulmonary artery anastomosis in a porcine model. A single pneumonectomy, performed 1 month before BEL implantation, provided the source of autologous cells used to bioengineer the organ on an acellular lung scaffold. During 30 days of bioreactor culture, we facilitated systemic vessel development using growth factor–loaded microparticles. We evaluated recipient survival, autograft (BEL) vascular and parenchymal tissue development, graft rejection, and microbiome reestablishment in autografted animals 10 hours, 2 weeks, 1 month, and 2 months after transplant. BEL became well vascularized as early as 2 weeks after transplant, and formation of alveolar tissue was observed in all animals (n = 4). There was no indication of transplant rejection. BEL continued to develop after transplant and did not require addition of exogenous growth factors to drive cell proliferation or lung and vascular tissue development. The sterile BEL was seeded and colonized by the bacterial community of the native lung.


Whole bioengineered lungs (BELs) produced on acellular lung scaffolds have been transplanted in small animal models, but lungs failed due to intravascular coagulation and defects in endothelial barrier function leading to pulmonary edema (1, 2). No approach has allowed for long-term survival of BELs after transplantation.

A work examining passive diffusion of gas into the lung suggests that nonvascularized lung can survive for periods of time without vascular support (3), such as ligation of the pulmonary artery (4). Here, we focused on development of the bronchial systemic circulation in nonimmunosuppressed pigs to support BEL growth and survival after transplantation. We performed a pilot study to establish feasibility of BEL transplantation, with an airway anastomosis but without a vascular (pulmonary) anastomosis. We relied on the development of collateral systemic circulation to support tissue survival (5, 6). BELs were created using autologous cells isolated from a left lung pneumonectomy for n = 6 pigs. Four pigs received implanted BELs 30 days after pneumonectomy, whereas two animals were euthanized before receiving a BEL. This approach allowed the opportunity to enhance our understanding of pulmonary vascular development, initiate examination of the BEL transcriptome, evaluate BEL tissue development after transplant, examine BEL immune response, evaluate acute and chronic rejection, and examine reestablishment of the microbiome within the BEL.



Porcine acellular lung scaffolds were produced, as described previously (7, 8), with one modification. A dextrose pretreatment step was added before decellularization of whole lungs. This was done to enhance protein stability (9), reducing collagen loss during decellularization. Established multiphoton microscopy (MPM) and second harmonic generation methods (7, 8) demonstrated that collagen fibers were less damaged (fig. S1A compared to fig. S1B), and significantly more collagen (P < 0.002) was retained in scaffolds using dextrose-SDS decellularization (fig. S1C). Bronchoscopy was performed on all scaffolds before recellularization (movie S1). Table S1 lists all of the abbreviations used in the manuscript.

Supplementation of scaffold

In past studies, acellular lung scaffolds were supplemented with platelet-rich plasma (PRP)–loaded pluronic F-127 hydrogel (BASF) (7, 8) before installation of cells. The ability of hydrogels and nanoparticles to target delivery in support of vascular tissue development has been previously demonstrated (10, 11). We combined microparticle (MP) delivery of vascular endothelial growth factor (VEGF) with hydrogel delivery of PRP, fibroblast growth factor 2 (FGF2), and keratinocyte growth factor (KGF). Discoidal porous silicon MPs (12) with 30- or 60-nm pores delivered VEGF to vascular portions of acellular scaffolds. Images of 1-μm VEGF-MPs show MP shape and structure (fig. S1, D and E). Suitability of growth factors was determined by measuring attachment of primary lung–derived vascular cells to 3 × 3 × 0.5–cm pieces of acellular blood vessel scaffold pretreated with media, VEGF-MP, a mixture of VEGF-MP and FGF2 hydrogel, or FGF2 hydrogel (fig. S1, F to I). Use of VEGF-MP, FGF2-loaded hydrogel, or VEGF-MP mixed with FGF2 hydrogel enhanced cell attachment (fig. S1J) beyond media or FGF2 hydrogel alone, and VEGF-MP and FGF2 hydrogel provided best cell attachment. Use of MPs with different pore sizes allowed for staged release of VEGF (fig. S1K). Hydrogels loaded with FGF2 or KGF released at a steady rate over time (fig. S1L). In whole acellular lung scaffolds, VEGF-MP (fig. S1, M and N) and FGF2 hydrogel (fig. S1O) were deposited within the small vessels and capillaries of the scaffold. Tracheal delivery of KGF hydrogel was also used to support cell attachment (fig. S1P).

Mesenchymal stem cells (MSCs) support angiogenesis, produce immunomodulatory factors, promote lung repair (13), regulate macrophage function (14), and combined with M2 macrophages contribute to tissue regeneration (15). To study the effects of MSC and M2 cells on lung tissue development, we added porcine MSCs, unpolarized macrophages, M1 or M2 macrophage subsets, mononuclear leukocytes (MNLs), or lipopolysaccharide (LPS)–stimulated MNLs or culture supernatants from these cell types to primary lung cells seeded onto 3 × 3 × 0.5–cm pieces of acellular lung scaffold. Increased cell attachment (fig. S2A) and proliferation, measured by Ki67 staining, occurred when primary lung cells were cultured on KGF hydrogel–pretreated scaffolds (fig. S2B) or with addition of MSC supernatant, MSC, M2 cell supernatant, or M2 cells to primary lung cultures, justifying use of these supplements in the production of whole BEL.

Production of BEL for transplantation

Procedures for recellularization of whole acellular pediatric scaffolds with adult lung–derived cells (7) were modified for use in this study. Changes included installation of VEGF-MP and FGF2 hydrogel into the pulmonary artery of whole acellular scaffolds 2 hours before primary vascular cell installation and addition of KGF hydrogel 2 hours before primary lung cell installation.

The primary lung cell preparation included aquaporin-5–positive (AQP5+) alveolar epithelial type I (AEC I) cells, prosurfactant protein C–positive (P-SPC+) AEC II, smooth muscle actin–positive (SMA+) cells, and fibroblast-specific protein 1–positive (FSP-1+) fibroblasts (fig. S3, A to I). Cells in the primary lung–derived vascular cell preparation contained CD31+ and vascular endothelial cadherin–positive (VE-CAD+) cells with SMA+ and FSP-1+ cells (fig. S3, J to R). Primary tracheal-bronchial cells were pan-cytokeratin–positive (Ck+), Ck-18+, epithelial cell adhesion molecule–positive (Ep-CAM+) cells, with few Clara cell protein-10+ (CC10) or FSP-1+ cells included (fig. S3S to CC). Cell installation information (table S2) and numbers of autologous cells installed into lung scaffold (table S3) are provided. Primary lung–derived vascular cells were installed into the pulmonary artery and primary lung, and primary tracheal-bronchial cells were installed into the trachea; MSC supernatant, MSCs, M2 macrophage supernatant, and M2 cells were added during BEL culture (tables S2 and S3). Oxygen concentrations were uniform for media alone or media and scaffold cultures over 30 days. There was a slow decrease in oxygen concentration over the 30-day BEL culture period, as oxygen was consumed by the cells of the BEL (fig. S4). In a subset of scaffolds, carboxyfluorescein succinimidyl ester (CFSE)–labeled primary lung cells were installed into scaffolds, and a Spectrum in vivo imaging system (IVIS) was used to examine cell dispersal on pieces of acellular scaffolds (fig. S5, A and B) or whole-lung scaffolds (fig. S5, C to F).

An overview of BEL production is shown in Fig. 1 (A to H). The left lungs removed from donor pigs (Fig. 1A) were used to produce left lung scaffolds for this study (Fig. 1B). The pulmonary artery, pulmonary vein, and trachea of the scaffold were cannulated as described (Fig. 1, B to D) (7).

Fig. 1 Study overview.

(A) Left lung scaffolds were produced from whole acellular pig lungs. (B) Catheters were placed into the trachea (TR), pulmonary artery (Pa), and pulmonary vein (Pv) and (C and D) were positioned in the chamber to permit visualization of catheters. (E) Diagram of the fluidic system shows the microfluidic and pumping system. OS, oxygen sensor. (F) Photograph of the system outlined in (E) (arrow points to oxygenator). (G and H) BEL on culture day 30. The BEL in (G) and (H) was produced using the scaffold in images (A) to (D). (H) Photograph of BEL being prepared for transplantation. Scale bar, 20 cm. (I) BEL in surgical suite and (J) after the trachea-to-trachea anastomosis. n = 6 BELs were created.

Transplantation and outcomes

Six nonimmunosuppressed pigs were slated to receive a BEL transplant, with two pigs euthanized before BEL transplantation because of surgical complications related to the left lung pneumonectomy. Four animals received autologous BELs 30 days after a left pneumonectomy and were euthanized at 10 hours (pig 2), 2 weeks (pig 1), 1 month (pig 4), and 2 months (pig 5) after transplantation (fig. S6A). After surgery, pig 5 developed a partial airway occlusion that reduced lung expansion. Pulse oximetry remained at 100% throughout the testing period. All pulmonary function measurements showed a trend toward return to baseline values, suggesting that transplanted lungs had normal pressures and volumes (fig. S6, B to D). Bronchoscopy of BELs was performed before (Fig. 2A and movie S2) and after transplantation (Fig. 2B, fig. S2C, and movie S3). Small blood vessels were seen mid-trachea and at the anastomosis site (Fig. 2, B and C) in animals that survived for longer than 10 hours. Computed tomography (CT) angiograms of the thorax of pig 1 (survived 2 weeks) comparing BEL and native lung (Fig. 2, D and E) depict the development of collateral blood circulation in BEL by 2 weeks after transplant. Figure 2F is a gross image of this BEL. Micro-CTs of native lung and BEL demonstrated that both BEL and native lung contained open airways and comparable tissue density (Fig. 2, G and H).

Fig. 2 Gross assessment of BEL and native lung.

(A to C) Bronchoscopy images of (A) BEL before and after transplant and of (B) the area above the anastomosis site showing left main stem bronchus, and (C) BEL trachea-to-trachea anastomosis (black arrow). (D and E) CT angiograms of the thorax of pig 1, 2 weeks after transplant. Native lung (NL) and BEL. (D) Collateral circulation in BEL is highlighted (white arrows), and aerated regions appear black in this colorized image. (E) Collateral vessels formed in BEL after transplantation (black arrows). (F) Gross image of BEL [left lung (LL)] from pig 1 after transplant. Black arrow indicates anastomosis site. (G and H) Micro-CTs of open airway in nonventilated (G) native lung and (H) BEL of pig 1. (I to N) BEL of pig 4, 1 month after transplant. (I) CT angiogram of the thorax in the arterial phase, axial image showing BEL in the left thoracic cavity (red dots denote edges of BEL). The pulmonary artery (Pa), aorta (Ao), right ventricle (RV), and left ventricle (LV) are shown, as well as the left side (L) of the animal in this coronal image. (J) Coronal x-ray image showing BEL in the left hemithorax (red dots denote edges of BEL). (K) Axial CT image of both native lung and BEL. The heart (Ht) and left side of the animal are noted. (L) Coronal and (M) axial images of MRI angiography showing peripheral enhancement outlining the left BEL, indicating capillary vascularization. (M) MRI image showing full expansion of both right native lung and left BEL. “R” denotes the right side of the animal and “L” the left side. (L and M) White arrows point to large collateral vessels in BEL. (N) Gross image of the BEL after necropsy showing native lung and the smaller BEL.

In pig 4 (survived 1 month), a CT angiogram of the thorax in the arterial phase, axial image, shows aerated portions of the BEL (Fig. 2I). The aorta, pulmonary artery, right ventricle, and left ventricle are noted. A coronal image of this animal shows the BEL in the left hemithorax (Fig. 2J). Hyperinflation of the right lung resulted in herniation of the native lung into the inferior left hemithorax. This contributed to the restricted expansion of the BEL (Fig. 2J), although the transplanted left lung became aerated during breathing (Fig. 2K). Coronal and axial images of magnetic resonance imaging (MRI) angiography display the peripheral enhancement outlining the left BEL due to capillary revascularization (Fig. 2, L and M). A large intercostal vessel arising from the aorta with branches extending toward the BEL is noted on the axial image (Fig. 2M, arrow). The gross image of the BEL from pig 4, after necropsy, shows the smaller size of the left BEL compared to the right native lung (Fig. 2N).

BEL transcriptome profile

We initiated the examination of BEL gene expression (GE) profiles at 1 month after transplant (pig 4) to determine whether angiogenesis or tissue development was still in progress and to identify key time points for examination of the BEL transcriptome in later studies. We tested 4128 genes isolated from BEL or native lung samples isolated from the same animal, setting the GE of native lung as reference and calculating fold changes (FCs) of GE for BEL. Here, FC was defined asEmbedded Image

Although there were variations in levels of GE in BEL compared to native lung, the types of genes expressed were similar (Fig. 3A). Compared with native lung, an average of 11.79% of the genes were down-regulated (0 < FC ≤ 0.5), and 15.93% were up-regulated (FC ≥ 2) in BEL (Fig. 3B). The majority of genes (72.28%, 0.5 < FC < 2) remained at the same expression level as found in native lung (Fig. 3B). We performed a paired Student’s t test between BEL and native lung GE with log2 transformation. Genes with FC ≥ 2 or FC ≤ 0.5 (P < 0.05) were defined as potential differentially expressed genes (Fig. 3, A to C, and tables S4 to S5).

Fig. 3 Genes expressed in BEL.

RNA sequence analysis of GE in BEL. FC > 1 indicates that GE value of BEL was greater than value of native lung, FC = 1 indicates that GE value of engineered lung was equal to value of native lung, and FC < 1 indicates that GE value of engineered lung was less than the value of native lung. (A) Heat map of the top 1000 genes (ranked by P values) from samples removed from three different regions of the BEL and native lung. (B) Table summarizing the number of genes in BEL exhibiting FC between 0 and 0.5, between 0.5 and 2, or >2 FCs in expression compared to native lung for tissue sets 1 to 3. (C) Table of BEL genes of interest related to angiogenesis with FC > 1 as compared to native lung.

Angiogenesis-related genes that were up-regulated in the BEL at 1 month after transplant included MAPK14 (FC = 5.00), TGFB2 (FC = 5.00), PDGFC (FC = 3.00), VCAM1 (FC = 3.00), VEGFD (FC = 3.00), HEY 1 (FC = 2.50), SRY-Box-9 (SOX-9; FC = 3.00), PDGFRA (FC = 2.50), SHH (FC = 2.25), SRY-Box-15 (SOX-15; FC = 2.00), FGFR1 (FC = 2.00), SELP (FC = 2.00), Wnt10B (FC = 2.00), ETV2 (FC = 2.00), and ICAM1 (FC = 2.00; Fig. 3C). Other up-regulated genes included KDR/VEGF2R (FC = 1.55), CXCL12/SDF-1 (FC = 1.33), NRP1 (FC = 1.28), SRY-Box-4 (SOX-4; FC = 1.25), and CXCR4 (FC = 1.17; Fig. 3C), as well as ITG2AB/CD41, ETS1, TGFB1, HEY2, PROX1, VEGFC, PECAM1, NOS2, NOS1, and SELE (table S4). In normal vascular development, one of the major signaling pathways is Notch (16). Although there was increased expression of downstream Notch target ligands in the BEL, Hey1 (FC = 3.00) and HeyL (FC = 2.50), this expression was not as robust as would have been expected if production of the BEL was purely a developmental process (17). Genes expressed in the BEL also included lung lineage gene NKX2-1 (FC = 1.40) and AEC I cell–associated genes AQP5 (FC = 2.00), SCNN1G (FC = 2.00), CAV-1, and RAGE/Ager or AEC II–associated genes SFTPC, SFTPB, SFTPD, and SFTPA1 (table S5). Other lung epithelial cell–associated genes expressed in the BEL included KRT19, MUC20, MUC13, MUC15, TP63, MUC1, and KRT5 (table S4). Genes normally expressed by neuroendocrine cells (CHGA, ENO2, and FOXF2), Clara cells (SCGβ3A2), or muscle cells (ACTG2, ACTA1, ACTB, and ACTA2; table S5) were also expressed.

The gene profile of the BEL was similar to that of the native lung, although the BEL exhibited distinct expression profiles. Despite this study’s limitations due to the small sample size, the information generated provides an important GE data set to build from in the future.

BEL vascular and lymphatic development

Pretransplant capillaries in BEL contained no red blood cells (Fig. 4A). Posttransplantation collateral circulation developed in pig 1 within 2 weeks (Fig. 4B) and developed in all animals that survived 2 weeks or longer (Fig. 2E). BEL microvasculature appeared normal (Fig. 4, C and D), and, on the basis of CFSE labeling of primary vascular cells, vessels in the BELs were formed from installed cells (Fig. 4, E to G), although few CFSE-labeled cells were found at the junction point between the BEL circulation and the animals’ normal vasculature (Fig. 4I, with Fig. 4H as the control). Pigs were spontaneously breathing 21% oxygen before BEL harvest. The average partial pressure of oxygen (pO2) in the BEL pulmonary artery of these pigs was 123 ± 10 mmHg, indicating it was receiving oxygenated blood and not venous blood from the collateral circulation. The lack of an oxygen gradient at the alveolus capillary junction prevented gas exchange, as has been documented in past studies (4, 6).

Fig. 4 Vascular tissue development in BELs.

(A) Transmission electron microscopy (TEM) of BEL on day 30 of bioreactor culture demonstrating capillaries (ca; black arrows) without red blood cells. (B) Hematoxylin and eosin (H&E) image of BEL at 2 weeks after transplant and (C) TEM of red blood cell–filled collateral capillaries (black arrows). (D) Cross-sectional H&E image of collateral blood vessels in BEL 2 weeks after transplant. (E to W) BEL harvested 1 month after transplant. (E) Cross section of CFSE-labeled (green) vessel in BEL. (F and G) Blood vessels within BEL formed from CFSE-labeled (green) primary lung–derived vascular cells. (G) Overlay of CD31+ (red) staining with CFSE+ (green). (H) 4′,6-Diamidino-2-phenylindole (DAPI; blue, nuclei) staining control lung for (I). (I) Junction of collateral vessel outside of the BEL. VE-cadherin+ (VE-Cad; red) endothelial cells and CFSE-labeled primary lung–derived vascular cells (white arrow) were found where collateral vessels joined with the BEL vasculature. (J to S) Cross sections of BEL blood vessels. (J, L, N, P, and R) DAPI staining controls and sections stained for (K and M) CD31+ (red), (O) ERG+ (red), (Q) eNOS+ (red), and (S) ACE+ (red) cells, all of which are indicators of endothelial cell function in the BEL. (T and V) DAPI control and representative image showing LYVE-1+ (green) lymphatic cells at (U) 2 weeks and (W) 1 month after transplant.

Vessels in BELs expressed CD31 (Fig. 4, K and M; J and L are controls) and angiogenesis markers including transcription factor early growth response protein 1 (ERG1) (Fig. 4O, with Fig. 4N as the control), endothelium nitric oxide synthase (eNOS; Fig. 4Q, with Fig. 4P as the control), and angiotensin-converting enzyme (ACE; Fig. 4S, with Fig. 4R as the control), which contributes to vascular muscle tone and blood flow (18). Lymphatic vessel endothelial receptor 1–positive (LYVE-1+) areas were seen at 2 weeks (Fig. 4U, with Fig. 4T as the control), and by 1 month after transplant, lymphatic vessels were found throughout the BEL (Fig. 4W, with Fig. 4V as the control), suggesting reestablishment of pulmonary lymphatics.

BEL tissue development

Acellular distal lung scaffold lacks structure (Fig. 5A). After recellularization, on day 30 of bioreactor culture, the BEL contained well-developed alveolar areas (Fig. 5, B and C), although nonaerated regions (Fig. 5, B and C, arrows) were evident. PSP-C+ AEC II (Fig. 5D) was the predominant cell type before transplant. After transplantation, normal breathing enhanced aeration of the BEL, although occasional nonaerated areas were present (Fig. 5E). PSP-C+ AEC II (Fig. 5, F and I, with Fig. 5H as the control) and AQP5+ AEC I were present in all animals (Fig. 5, G and K, with Fig. 5J as the control). Compared to the animal that survived for 10 hours (Fig. 5L), the total number of cells increased in BEL (Fig. 5M) of animals that survived for 1 month (Fig. 5N) and 2 months (Fig. 5O). The total number of AEC I also increased in pigs that survived for 2 weeks and 1 month but not in pig 5, in which a partial occlusion (Fig. 5O) reduced aeration and stretch of tissues.

Fig. 5 Lung tissue development in BELs.

(A) Scanning electron microscopy (SEM) of acellular scaffold. (B) Methylene blue–stained thin section of BEL before transplantation highlighting nonaerated (black arrows) and aerated spaces. (C) SEM of alveoli of BELs before transplant. White arrows indicate nonaerated areas. (D) TEM image of AEC II (inset, lamellar body) before transplant. (E to K) Evaluation of pig 4 BEL at 1 month after transplant. (E) SEM of BEL after transplant demonstrating increased aerated regions due to normal breathing. A small compressed area remains (white arrow). (F) TEM image of AEC II (inset, lamellar body) after transplant. (G) TEM of BEL containing AEC I pneumocytes (arrow). (H) DAPI-stained control and (I) P-SPC+ (red) AEC II. (J) DAPI-stained control and (K) AQP5+ (green) AEC I cells. (L to O) Mean counts of total number of cells and number of AEC I cells in native lungs or BELs for pigs that survived for (L) 10 hours, (M) 2 weeks, (N) 1 month, or (O) 2 months. Student’s t test was used to compare total number of cells and total numbers of AEC I in native lung and BEL. Analysis of variance (ANOVA) was used to assess statistical significance in the comparison of (L) to (M), (N), and (O) (*P < 0.001) and (M) and (N) to (O) (***P < 0.0001).

Acellular trachea scaffold (fig. S7A) was supplemented with FGF2 hydrogel to support tracheal cell attachment (fig. S7, B and C). FGF2 hydrogel provided for better cell attachment than media or hydrogel alone (fig. S7D). Two weeks after transplant, cells were dispersed (fig. S7, E and F) and lacked cell-to-cell contacts. By 2 months, cells in the BEL trachea had reestablished cell-cell contacts and developed intercellular junctions (fig. S7, G to J) in most areas.

Ki67+, proliferating cells were found in bronchioles (fig. S8B, with fig. S8A as the control) and lungs (fig. S8D, with fig. S8C as the control) of all BELs. Higher numbers of Ki67+ cells were found in animals that survived for 2 weeks and 1 month after transplantation (fig. S8E). Key changes after transplantation included continued angiogenesis and development of the epithelial lining of the trachea, bronchi, and bronchioles in BEL. Although alveoli and bronchioles were not well developed at 10 hours (fig. S8, F to H) or 2 weeks (fig. S8, I to K), continued development occurred in animals that survived for 1 month (fig. S8, L to N) or 2 months (fig. S8, O to Q). BEL alveoli and bronchi were indistinguishable from native lung except within nonaerated regions in pig 5.

Ck-18+ cells (fig. S9B, with fig. S9A as the control) and CC10+ Clara cells were present in developing bronchioles (fig. S9D, with fig. S9C as the control) as were mucin-producing cells (fig. S9E) including mucin 5a (MUC5a), a protein marker of developing airway epithelium (fig. S9G, with fig. S9F as the control), and MUC1 (fig. S9I, with fig. S9H as the control). Low numbers of lung progenitor cell phenotypes, such as CK5+/P63+ cells (19), were found (fig. S9K, with fig. S9J as the control) in BELs.

Pig 5 developed an occlusion of the first branch of the main stem bronchus of the BEL after transplantation. Both passageways at the point of the carina were open, as shown by bronchoscopy, 2 months after transplantation (Fig. 6A), but the left bifurcation of the lung was occluded (Fig. 6, B to D). Chest x-ray was performed due to breath sounds in the left chest cavity. The left lung appeared small, dense, and partially aerated (Fig. 6, D and E), although CT images indicated the presence of multiple intercostal vessels (Fig. 6F). P-SPC+ AEC II (Fig. 6H, with Fig. 6G as the control) were found in compressed, nonaerated areas, and many of these cells were undergoing apoptosis [terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick end labeling–positive (TUNEL+)] (Fig. 6, J, K, and O, with Fig. 6I as the control). More FSP-1+ fibroblasts were found in nonaerated versus aerated regions of the lung or compared to native lung (Fig. 6, M to O, with Fig. 6L as the control). As expected, native lung contained more cells and more P-SPC+ cells than BEL (Fig. 6O).

Fig. 6 Pig 5 tissue development.

Pig 5 developed a partial airway occlusion after transplantation. Bronchoscopy images of (A) carina, (B) bronchial occlusion (arrow) and open airway, and (C) image of bronchial occlusion alone (black arrow) in BEL. (D) Anastomosis site of BEL in recipient’s trachea. (E) Chest x-ray of the nonaerated BEL, seen as the dense homogeneous opacity, projecting over the mediastinum. Extensive compensatory hyperinflation of the native lung occurred. (F) CT image of chest in venous phase through the left hemithorax showing collapsed BEL containing multiple small intercostal vessels (arrows). (G, I, and L) DAPI-stained controls and BEL in (G and H) aerated and (J to N) nonaerated regions containing P-SPC+ (red) and TUNEL+ (green) AEC II. (H) P-SPC+ (red) cells with inset of enlarged image. (J and K) BEL in nonaerated region cells stained for P-SPC+ (red) and TUNEL+ (green), a marker indicative of cells undergoing apoptosis. (L) DAPI-stained control and (M and N) FSP-1+ (red) fibroblasts in nonaerated regions. Five randomly selected areas from 10 different sections of tissue immunostained were examined for TUNEL+ or FSP-1+ cells in native lung and in aerated or nonaerated sections of BEL. (O) Averaged number of cells, number of P-SPC+ AEC II, TUNEL+ AEC II, and FSP-1+ cells ± SD are shown for native lung and aerated and nonaerated BEL. Data were analyzed using ANOVA. *P < 0.05, **P < 0.005, ***P < 0.0005. NS, not significant.

Immune response of BEL

Contamination of long-term bioengineered tissue cultures is a common problem. There is also increased susceptibility to infection of pulmonary grafts after transplantation, due to contact with microbial contaminants during breathing (20, 21). As a preventative antimicrobial strategy, the immune systems of BELs were reconstituted. Autologous MNLs were added on day 11 of bioreactor culture and autologous serum, alveolar macrophages, and MNLs on day 30 before transplantation.

Cytokine analysis was performed on bonchioalveolar lavage (BAL) fluid isolated from native lung and BEL. Before transplant, there were low concentrations of proinflammatory cytokines in native lungs and BELs (Fig. 7A). Pig 2, euthanized 10 hours after transplant, had a measurable proinflammatory response due to an undiagnosed infection at the time of euthanasia (Fig. 7B). Cytokine concentrations decreased as survival time of animals increased (Fig. 7B). There was no indication of a T cell response after BEL transplantation as demonstrated by low concentrations of interleukin-2 (IL-2) or IL-12p70 and lack of increase in CD8+ cell numbers in BAL (Fig. 7C). Numbers of CD4+, CD8+, perforin-containing cells, and CD20+- or immunoglobulin G–expressing B lymphocytes were not significantly different in BEL compared to native lung (Fig. 7D). There was also no difference in location of CD8+ lymphocytes in airways of native lung (Fig. 7F, with Fig. 7E as the control) and BELs (Fig. 7H, with Fig. 7G as the control) or number of CD8+ cells in tissues (Fig. 7I). These data indicate that the autologous BELs were well tolerated, with no infiltration of leukocytes into tissues or up-regulation of T cell responses indicative of graft dysfunction or rejection.

Fig. 7 BEL immune response.

BAL was performed on BEL and native lungs of all animals. (A and B) Examination of IL-8, 1L-1β, IL-6, IL-10, IL-12p70, IL-2, IL-4, and interferon-γ (IFN-γ) concentrations in BAL of (A) native lungs and BELs before transplant. (B) Polar plot showing cytokine concentrations for native lungs and overlay of data from BALs evaluated at 10 hours, 2 weeks, 1 month, or 2 months after transplantation. The polar plot highlights each animal’s immune response. (A to D) BALs performed on native lung at the time of the pneumonectomy and on BEL after euthanasia. (C) Number of CD8+ cells isolated from BALs. (D) Percentage of CD4+, CD8+ T lymphocytes, perforin-positive cells, and CD20+ B lymphocytes are shown. (E and G) DAPI-stained controls and (F and H) representative images of CD8+ T lymphocytes (green) in tissues of (F) native lung or (H) BEL. (I) Averaged CD8+ lymphocyte counts for tissue sections from pigs 1, 2, 4, and 5 native lungs and BELs. Analyses to compare native lung to BEL per pig were done using Student’s t test.

BEL microbiome development

Transplantation of the sterile BELs provided an opportunity to observe the establishment of the pulmonary microbiome communities in the respiratory tree (21). Native lung contains a well-developed microbiome (Fig. 8A), but BELs are sterile before transplant (Fig. 8B), and no organisms were found in BEL until after transplantation (Fig. 8C). Evaluation of the established microbiome over a time course helped us address the BEL from the perspective of the bacterial community. We completed initial next-generation sequencing of samples from four pigs housed in our facility to identify the core microbiome of the respiratory tree. The resulting data were consistent with the limited published microbiome data for pig lung (22) and identified the most common genera or species present in these laboratory animals. Optimized quantitative polymerase chain reaction (qPCR) targets and assays were then established to quantify the common bacteria and selected minor species associated with pathogenic infections to evaluate the seeding of BEL (table S8). One bacterial target was based on identification of 16S sequences that did not align with sequences in the SILVA rRNA database. Specifically, IOLA (infectious organism lurking in airways) (22, 23) 16S was seen in one of the transplanted and two of the control animals studied. This result was confirmed through cloning and sequencing of additional genomic fragments using published PCR primers (23).

Fig. 8 Analysis of BEL microbiome.

(A) Representative SEM image of the native lung of pig 1 demonstrating the normal microbiome. (B) SEM image of sterile BELs before transplantation. (C) SEM of BEL 2 weeks after transplantation showing reduced microbial colonization of BEL. The composition of each microbiome was evaluated for (D) tracheal and (E) lung samples from native lung and BELs of pigs that survived for 10 hours, 2 weeks, or 1 month and are shown as proportional bar charts (average of at least two independent evaluations per sample). Data for native lung and BELs are labeled at the top of each bar.

The composition of each microbiome was evaluated for tracheal and lung samples from each of three pigs and is shown as proportional bar charts (average of at least two independent evaluations per sample) in Fig. 8. Tracheal and lung colonization occurred within 10 hours of the transplant; however, the profile of these communities appeared to be less stabilized, with more bacterial targets detected in the trachea of the transplant relative to the normal lung. Respiratory problems forced early euthanasia of this animal. The qPCR detected extremely high amounts of Mycoplasma flocculare in the bioengineered trachea and lung communities, suggesting that this organism may have contributed to the signs of clinical disease that warranted early euthanasia. The 2-week BEL tissue showed slight but not significant differences in proportions of Moraxella species and Staphylococcus species in the trachea (P > 0.05). The paired native lung and BEL samples for the 2- and 1-month transplants also showed similar bacterial communities with nearly identical representation and proportions. There were some notable differences in the 1-month tissues, including the M. flocculare (24) observed in the 10-hour samples.


To date, regenerative laboratories have attempted to engineer only a few whole organs. This endeavor requires engineering not only the organ but also vascular tissues to maintain a healthy organ with full functionality. We concentrated our initial efforts on developing the microvasculature and systemic support in the BEL and found that collateral systemic circulation developed in all animals that survived 2 weeks or longer. Because BEL was supplied with oxygenated rather than deoxygenated blood, we were unable to assess gas exchange due to a lack of an oxygen gradient at the alveolar capillary junction.

GE related to angiogenesis and lung tissue development indicated that tissue development was still in progress 1 month after transplantation. Histological examination of tissues indicated that collateral circulation developed in all animals as early as 2 weeks after transplant. Histological evaluation showed progression in lung and airway epithelial cell development with an associated increase in overall cell numbers and AEC I in animals that survived from 10 hours to 2 months. Cells associated with lung-specific lineages were found in all animals at all time points examined, although there were few Clara cells in the developing airways of animals due to the lack of primary Clara cells in the primary tracheal bronchial cell preparation.

One obvious limitation of this study is the small sample size related to genome analysis. However, this finding was supported by other methods of analysis and indicated that genes related to angiogenesis and lung cell lineages remained elevated in BEL 2 months after transplantation. Another limitation is the need to continue survival of animals beyond 2 months with subsequent evaluation of the ability of these animals to survive relying only on oxygen provided by their BEL alone.

Acute lung rejection, characterized by perivascular and subendothelial mononuclear infiltrates or by lymphocytic bronchitis and bronchiolitis, was not seen in BELs. We did not see a significant increase in the presence of proinflammatory cytokines in tissues isolated from BELs, except in pig 2 (10-hour survival). This animal was later shown to have high numbers of M. flocculare, a swine pathogen. We saw no indication of primary graft rejection in animals that survived for 10 hours, 2 weeks, 1 month, or 2 months based on BAL evaluations or histopathology. No marked structural abnormalities were found in BEL tissues in pigs 1, 2, or 4. Pig 5, however, developed a partial airway occlusion after surgery and showed some underdeveloped lung areas. Representative images indicated that aerated regions of the lung displayed normal lung architecture.

Our study provided the opportunity to examine the reestablishment of the microbiome in a sterile BEL after transplantation. Recent reports highlight a role for lung microbiota in control of lung injury and remodeling after transplantation (25) and development of bronchiolitis obliterans syndrome, which impacts long-term survival (26). The sterile tissues appear to have been seeded via the trachea, as noted from the results of the animal that survived for 10 hours; however, more work will be required to confirm this route of colonization. The distinct bacterial communities we observed were consistent with other reports for the swine lung (21) and were consistently reproduced in BEL. These evaluations also led to the observation of an infectious organism present in airways, suggesting that this organism may be of pathogenic concern in swine; moreover, IOLA had previously been reported only in the human respiratory tract in association with clinical disease (23).

In conclusion, we have shown that the nanoparticle and growth factor hydrogel modification of acellular scaffolds was essential to the success of this study and that continued vascular development occurs in animals after transplantation of BELs. These results also support the utility of the platform used to produce and transplant BEL for the general study of BEL development, including the transcriptome, vascular tissue development, immune response related to rejection, and microbiome formation. This platform would also allow examination of the influence of the microbiome on BEL survival and function in future studies. Together, these findings represent a significant advance in our understanding of the production of bioengineered tissues for transplantation. Future studies should concentrate on procedures to allow continued maturation of the BEL in vivo and establishment of vascular flow via the pulmonary artery and pulmonary vein.


Study design

The objective of this study was to explore development of the systemic circulation after transplanting BELs into a large-animal (pig) model with tracheal anastomosis but without reattachment of the pulmonary circulation. We used a three-dimensional model of porcine lung tissue to select methods of growth factor delivery and scaffold supplements that enhanced vascular and lung tissue development. BELs were created from autologous primary lung, and vascular cells were isolated from a left lung pneumonectomy performed 30 days before BEL transplantation. Porcine lungs for acellular scaffold production were obtained as discarded surgical materials at the University of Texas Medical Branch (UTMB) or were obtained following the Institutional Animal Care and Use Committee (IACUC)–approved protocols at the Texas Methodist Hospital Research Institute. Animal handling and surgical procedures for obtaining porcine peripheral blood or BEL transplantation were performed according to protocols approved by the IACUC of UTMB at Galveston and were compliant with guidelines of the American Association for the Accreditation of Laboratory Animal Care. Animals were not immunosuppressed in this study. Replicate numbers of each experiment are included in the figure captions. Tissues from n = 6 BELs before transplant and n = 4 BELs after successful transplantation were randomized before examination. Histology analysis and cell counts were performed by trained individuals who were blinded to the study. One animal (pig 2) was euthanized early due to respiratory complications at 10 hours. Pig 5 developed an airway occlusion after surgery, limiting BEL development, and samples from this animal were not used for microbiome analysis. Animals that survived for 2 weeks, 1 month, and 2 months demonstrated development of collateral systemic circulation, BEL survival, and tissue development after transplantation. Antibodies used in this study are listed in tables S6 and S7.

Statistical analysis

All viability, genomic, histology, imaging, and microbiome analyses data for each pig compared each animal’s BEL to its native lung. For cell phenotype analysis, 10,000 cells were collected for each flow cytometry sample examined. For specified data comparisons, a paired samples Student’s t test was used to compare means. For other data sets, ANOVA was used as noted. Statistical analyses for these data were performed using GraphPad Prism v7.0.04. Mean values and SDs are reported. Mean differences in the values were considered significant when P < 0.05. For microbiome analysis, mathematical analyses were performed using Excel (Microsoft Corp.). Graphing was competed using Excel or GraphPad InSTAT software (version 2003).


Materials and Methods

Fig. S1. Scaffold production and modification.

Fig. S2. BEL culture supplements.

Fig. S3. Cell phenotypes installed in BEL.

Fig. S4. Bioreactor culture BEL pO2 measurements.

Fig. S5. IVIS imaging to estimate cell dispersal.

Fig. S6. Information regarding study animals.

Fig. S7. Evaluation of tracheal development.

Fig. S8. BEL tissue development.

Fig. S9. Cell phenotypes in BEL.

Table S1. Abbreviations used in the manuscript.

Table S2. Cell installation information.

Table S3. Number of cells Installed in scaffolds.

Table S4. RNA sequence data: Angiogenesis.

Table S5. RNA sequence data: Cell lineage.

Table S6. Antibodies used for histochemical cell phenotype analysis.

Table S7. Antibodies used for flow cytometry analysis.

Table S8. Microbiome primers used in this study.

Movie S1. Bronchoscopy of acellular pig lung scaffold.

Movie S2. Bronchoscopy of BEL before transplant.

Movie S3. Bronchoscopy of BEL after transplant.

References (2744)


Acknowledgments: We thank the UTMB Animal Resource Center and our veterinary staff C. Klages, D. Brining, and D. Deyo for help with support of our animals. We thank A. Duarte, D. Christiani, and J. Leduc for reading and suggesting edits for the manuscript. We thank J. Barral for his help regarding use of osmolytes to protect proteins from denaturation. We thank M. Riddle, E. Suarez, and C. Bryant for their help with this study. We thank M. Susman and C. Holubar for editorial assistance. Funding: This work was supported in part by NIH U18 Grant (grant no. U18TR000560-01). J. Leduc provided funds to support BEL production and transplantation. Additional funding for production of nanoparticles was provided from startup funds provided to J.S. from Houston Methodist Research Institute of Houston. Author contributions: J.E.N. managed the BEL production team and data analysis and prepared the manuscript. S.L.F. performed surgeries including transplantation of the BEL and was involved in the preparation of the manuscript. S.P.V., J.E.N., J.A.N., and L.B.A. were involved in the production of BEL, data preparation, data analysis, and preparation of the manuscript. L.F. did the MRI and CT analyses of animals. J.S., X.L., and J.R. provided MP or produced porcine lung scaffolds. G.H. managed TEM and SEM. G.V. and R.P. did MPM. D.C.C., R.Z., B.E.H., and S.L. did the genomic analysis. R.B.P. and A.M. performed the microbiome analysis. L.W. performed bronchoscopic evaluations. F.B. developed the dextrose decellularization protocol. A.W. produced anesthesia protocol and provided anesthesia to animals. E.U., M.G., and D.W. did the histological and histopathological analyses. I.P. performed micro-CTs and IVIS. R.M. examined the respiratory function testing. J.C. managed the clinical transplant team and contributed to data analysis and preparation of the manuscript. Competing interests: J.S., J.E.N., and J.A.N. are inventors listed on Patent Cooperation Treaty (PCT) International Application No. PCT/US2016/057977 for use of dextrose in production of whole-lung scaffolds for BEL production. J.E.N., J.C., S.L.F., J.A.N., and J.S. are inventors on a U.S. provisional patent application #62/659,321 that is being submitted by UTMB at Galveston that covers the process of producing scaffolds for use in the production of BELs, use of nanoparticles to facilitate this process, and the method of transplantation of BEL into a large-animal model. S.L.F. worked for the Methodist Research Institute at the beginning of this project and now works for Biostage. J.S. worked for the Methodist Research Institute at the beginning of this project and now works for NanoMedical Systems Inc. Data and materials availability: All data associated with this study are present in the paper or the Supplementary Materials.

Stay Connected to Science Translational Medicine

Navigate This Article