Cell replacement in human lung bioengineering

D1X XBrandon A. Guenthart, D2X XMD,a,b D3X XJohn D. O’Neill, D4X XPhD,b D5X XJinho Kim, D6X XPhD,b,c

D7X XKenmond Fung, D8X XCCP,d D9X XGordana Vunjak-Novakovic, D10X XPhD,b,e andD11X XMatthew Bacchetta, D12X XMDa

From the aDepartment of Surgery, Columbia University Medical Center, Columbia University, New York, New York, USA;bDepartment of Biomedical Engineering, Columbia University Medical Center, Columbia University, New York, NewYork, USA; cDepartment of Biomedical Engineering, Stevens Institute of Technology, Hoboken, New Jersey, USA;dDepartment of Clinical Perfusion, Columbia University Medical Center, Columbia University, New York, New York, USA;and the eDepartment of Medicine, Columbia University Medical Center, Columbia University, New York, New York, USA.

BACKGROUND: As the number of patients with end-stage lung disease continues to rise, there is a grow-ing need to increase the limited number of lungs available for transplantation. Unfortunately, attemptsat engineering functional lung de novo have been unsuccessful, and artificial mechanical devices havelimited utility as a bridge to transplant. This difficulty is largely due to the size and inherent complexityof the lung; however, recent advances in cell-based therapeutics offer a unique opportunity to enhancetraditional tissue-engineering approaches with targeted site- and cell-specific strategies.METHODS: Human lungs considered unsuitable for transplantation were procured and supported usingnovel cannulation techniques and modified ex-vivo lung perfusion. Targeted lung regions were treatedusing intratracheal delivery of decellularization solution. Labeled mesenchymal stem cells or airwayepithelial cells were then delivered into the lung and incubated for up to 6 hours.RESULTS: Tissue samples were collected at regular time intervals and detailed histologic and immuno-histochemical analyses were performed to evaluate the effectiveness of native cell removal and exoge-nous cell replacement. Regional decellularization resulted in the removal of airway epithelium withpreservation of vascular endothelium and extracellular matrix proteins. After incubation, deliveredcells were retained in the lung and showed homogeneous topographic distribution and flattened cellularmorphology.CONCLUSIONS: Our findings suggest that targeted cell replacement in extracorporeal organs is feasibleand may ultimately lead to chimeric organs suitable for transplantation or the development of in-situinterventions to treat or reverse disease, ultimately negating the need for transplantation.J Heart Lung Transplant 2019;38:215−224! 2018 International Society for Heart and Lung Transplantation. All rights reserved.

KEYWORDS:cell replacement;lung bioengineering;lung transplantation;ex-vivo lung perfusion(EVLP);stem cell;decellularization;organ regeneration

The growing need for transplantable organs continues todrive development and expand the capabilities of tissueengineering and regenerative medicine. The need for aviable alternative to traditional donor organ utilization isperhaps most evident in the field of lung transplantation.For thousands of patients with end-stage lung disease,

transplantation remains the only option. Despite efforts toexpand the donor organ pool,1−3 the number of lung trans-plantations performed annually has not risen significantly,and waitlist mortality continues to rise.4

Recapitulating the complexity of the lung and bioengi-neering functional tissue capable of gas exchange is challeng-ing. Historically, biomedical engineers have approached thisproblem from 2 directions: (i) A bottom-up (or modular)approach—involving the formation of a bio-inspired micro-architectural unit (e.g., the alveolus), which when scaled canform complex tissues capable of organ-specific function. Thelung, however, has over 40 distinct cell types,5 and multiple

Reprint requests: Matthew Bacchetta, MD, MBA, Department of Tho-racic Surgery, Vanderbilt University Medical Center, 609 Oxford House,

1313 21st Avenue South, Nashville, TN 37232. Telephone: 212-305-3408.

Fax: 615-936-7003.

E-mail address: matthew.bacchetta@vanderbilt.edu

1053-2498/$ - see front matter! 2018 International Society for Heart and Lung Transplantation. All rights reserved.

https://doi.org/10.1016/j.healun.2018.11.007

http://www.jhltonline.org

tissue types with region-specific composition and function. Itis estimated that 1 cm3 of lung parenchyma contains >170alveoli.6 Given the size and anatomic complexity of thehuman lung, this approach may be best suited for in-vitromicrosystems (e.g., organ-on-a-chip) and have little practicalapplication in whole-organ bioengineering. (ii) A top-downapproach—involving recellularization strategies using bio-degradable, artificial, or decellularized biomimetic scaffolds.In 2010, whole-organ perfusion-based decellularization wasreported in rodentmodels.7,8Although these studies, and sub-sequent scaled studies using porcine and human lungs,9,10

served as proofs-of-concept that decellularized scaffoldscould support the engraftment of various cell types, trans-planted lung constructs failed within hours due to severe pul-monary edema and microemboli within the pulmonaryvasculature. As progress with this approach has been limited,further study is needed to identify appropriate cell popula-tions and doses capable of regeneration to optimize deliverystrategies and develop more robust bioreactor systems forwhole-organ culture.

Currently, there is considerable interest in investigatingcell-based therapeutics to ameliorate lung injury or treatdisease. Mesenchymal stem cells (MSCs) have been shownto augment injury recovery through paracrine activity11−13

and mitochondrial transfer,14 and their use is currentlyunder investigation in a number of human clinical trials. Inaddition, investigators have explored the use of pluripotentcells, which have tremendous potential in treating or revers-ing the course of lung disease and in bioengineering newlung tissue. Since the landmark discovery of induced plu-ripotent stem cells (iPSCs) in 2006 by Yamanaka and col-leagues,15 there has been tremendous interest in autologouscell therapy. Somatic reprogramming and genetic manipu-lation offers the ability to produce an unlimited supply of

healthy, patient-specific progenitor cells. In the case of lungdisease, these cells have the potential for use in cell replace-ment therapy, a process involving the removal of diseasedor damaged cells and subsequent replacement with healthyprogenitors. As seen with hematopoietic cell‒based thera-pies, even low levels of chimerism can effectively reversethe disease phenotype.16,17 Similar evidence suggests that acorrection of only 5% to 10% of the airway epithelial cellsin patients with cystic fibrosis may restore normal chloridesecretion.18,19

Herein we assessed the potential for targeted cellremoval and cell replacement therapy using human lungsunacceptable for transplantation and supported withex-vivo lung perfusion (EVLP) (Figure 1A). In addition, toexplore the clinical potential of our approach, we furthercharacterized unacceptable human lungs and detailed ourEVLP set-up, cannulation techniques, and novel transpleu-ral imaging system for the in-situ visualization of deliveredcells in real time. Our data demonstrate that regional decel-lularization followed by recellularization using controlledmicro-volume delivery techniques20 in predetermined lungregions resulted in: (i) selective removal of pulmonaryepithelium; (ii) maintenance of native lung architecture andretention of essential extracellular matrix components;and (iii) efficient delivery, topographic distribution, andattachment of therapeutic cells into targeted lung segments.

Methods

Human lung procurement

To demonstrate the utility of our methods in a clinically relevantmodel, we procured human lungs considered unacceptable fortransplant. Lungs were harvested in standard clinical fashion21,22

A B

Figure 1 Cell replacement in ex-vivo lung bioengineering and envisioned in-situ application. (A) Ex-vivo lung bioengineering and cellreplacement strategy to recover lungs unsuitable for transplantation. (i) Schematic showing human lungs procured and placed on ex-vivosupport enabling regional cell replacement and creation of functionally improved chimeric lungs. (ii) Schematic showing the selectiveremoval of respiratory epithelium with preservation of native lung and extracellular matrix components followed by therapeutic cell deliv-ery, attachment, and differentiation. (iii) Additional diagnostic and therapeutic interventions enabled by ex-vivo lung perfusion and pro-longed extracorporeal support. (B) Envisioned in-situ cell replacement in patients with lung injury or disease. Collection of a patient’s owncells (1). Somatic reprogramming and genetic modification of patient-derived cells to obtain healthy patient-specific progenitor cells (iPSC)(2). Delivery of healthy progenitors into targeted lung regions in damaged lungs to treat or reverse the trajectory of disease (3).

216 The Journal of Heart and Lung Transplantation, Vol 38, No 2, February 2019

from donors after brain death. The use of human tissue forresearch received a waiver from the institutional review board(IRB) at Columbia University and approval from LiveOnNYorgan donor network.

Characterization of donors with lungs consideredunsuitable for transplantation

A comprehensive review was conducted to identify the literaturepublished worldwide describing donor characteristics in caseswhere lungs were considered unsuitable for transplantation. Anal-ysis only included lungs donated after brain death with no warmischemic time and excluded donors with cardiac death. Pooleddata are reported as mean § standard deviation or as number (%).

Lung cannulation

The aortic arch, pericardium, or another large donor vessel wasconnected to the left atrial cuff with a running 6-0 Prolene suture(Ethicon), and a custom 36F crenellated venous drainage cannulawas secured in place with a 2-0 Ti-Cron tie (Covidien). The useof a “bio-bridge” was adapted from previously described methodsin a porcine model.23 An 18F to 20F pulmonary artery cannulawas secured in place with a purse-string 5-0 Prolene suture, andthe trachea was cannulated with a 7.5-mm cuffed endotrachealtube (Sheridan).

Ex-vivo lung perfusion

A double-lined organ basin heated with a recirculating warmsaline circuit was placed in the preservation chamber (XVIVO),equipped with a humidification system and ambient temperaturecontrols. Lungs were placed in the prone position, and the cannu-las were secured. The EVLP technique was based on previouslydescribed methods.24 Thermal images of the lung were capturedusing an infrared camera (T430sc; FLIR Systems) EVLP wasconducted for up to 6 hours.

Regional decellularization

Decellularization of target lung regions was achieved by intratra-cheal delivery of decellularization solution. A mild detergentsolution, previously described,25 containing 8 mmol/liter3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate(CHAPS), 0.5 mol/liter NaCl, and 25 mmol/liter ethylene-diaminetetraacetic acid (EDTA), was introduced bronchoscopically intotargeted bronchopulmonary segments by a custom microcatheterdelivery system. Immediately after withdrawal of the broncho-scope, a 4F to 7F Fogarty occlusion catheter (Edwards Lifescien-ces) was inflated and left in place to prevent escape ofdecellularization fluid from the targeted region. The decellulariza-tion solution was allowed to dwell for 1 hour, and was thenreplaced with fresh solution for an additional 1 hour beforerepeated bronchoalveolar lavage with sterile normal saline.During airway decellularization, the vascular compartmentremained perfused and in the native condition (i.e., did notundergo decellularization). To assess the effectiveness of decellu-larization, lung wedge samples were collected after decellulariza-tion and before further intervention, then fixed, embedded,deparaffinized, and imaged, as previously described.

Therapeutic cell delivery

Cell delivery was performed using a 3.8-mm flexible bronchoscopewith a 1.2-mm lumen for therapeutic delivery and intervention.MSCs or airway epithelial cells labeled with carboxyfluorescein-succinimidyl ester (CFSE) were delivered locally into targetedlobes or segments. Subsequent live transpleural imaging of nearinfrared-labeled cells was performed using a custom imaging sys-tem developed in our lab, consisting of an LED excitation source(M780L3; Thorlabs), a camera (Zyla 4.2; Andor Technology), andan objective (Plan N 10 £; Olympus).23 To assess the distributionand evidence of early engraftment of delivered cells, lung wedgesamples were collected at the conclusion of the experiment, thenfixed, embedded, deparaffinized, and imaged, as previouslydescribed (refer to Supplementary Material available online atwww.jhltonline.org/).

Results

Human lung ex-vivo perfusion

Lungs considered unsuitable for transplantation were pro-cured in standard fashion from donors after brain death(n = 6). An EVLP platform with integrated circuit elementswas developed and used to provide ex-vivo lungs with6 hours of normothermic perfusion (Figure 2A and B). Theprocedure for cannulation and management of pulmonaryvenous drainage is outlined in Figure 3A. A donor vessel(e.g., aortic arch) or tissue (e.g., pericardium) was used as anatural bio-bridge to facilitate cannulation of the pulmonaryveins (PVs) (Figure 3B). Next, the pulmonary artery (PA)and trachea were cannulated (Figure 3A, iii). Target param-eters for extracorporeal lung perfusion and ventilationwere: PA pressure, <20 mm Hg; PV pressure, 3 to 5 mmHg; flow, 0.2 to 0.4 liter/min, temperature, 36˚ to 38˚C;respiratory rate, 6 to 8 beats/min; tidal volume, 6 ml/kg;and fraction of inspired oxygen (FIO2), 40%. Real-timemonitoring and feedback-regulated pressure-limited controlallowed for tight regulation of vascular flow (Figure 2C).Adjustments in the height differential between the lung andperfusate reservoir allowed for modulation of hydrostaticpressure and precise maintenance of the transpulmonarypressure gradient (Figure 2D).

Donor and lung characteristics

Representative imaging before procurement demonstratedlobar collapse and consolidation on computed tomography(CT) scan (Figure 3C, i) and opacities/atelectasis on chestX-ray (CXR) (Figure 3C, ii). Gross imaging of donor lungshighlighted lung regions that were not able to be recruited(i.e., ventilated) during the procurement process (Figure 3C,iii and iv). Bronchoscopy demonstrated inflammation andthick copious secretions in the airways (Figure 3C, v and vi),and periodic acid‒Schiff (PAS) staining of BAL specimensdemonstrated abundant cellular content (Figure 3E). Ther-mography showed lungs efficiently cooled and placed in thewarm organ basin after flush and cannulation (Figure 3D). Atthe start of EVLP, a fluorescent cell viability marker (CSFE)

Guenthart et al. Human Lung Bioengineering 217

was delivered to the distal lung, where uptake of CFSEconfirmed cell viability (Figure 3F).

To better understand and characterize lungs consid-ered unsuitable for transplantation, a comprehensiveliterature review was conducted. Data were combinedfrom previous reports of brain death donors whoselungs were unable to be used for transplantation(Table 1).26−43 Average (mean § SD) donor age was45.7 § 6.9 years, and 59.8% of the donors were male.Trauma was the leading cause of brain death (56.8%),followed by intracerebral hemorrhage (34.2%) and othercauses, such as cardiac arrest with return of spontaneouscirculation (ROSC) or ischemic brain injury (8.9%). Ofthose reported, 23.1% of lungs had a significant smokinghistory (>20 pack-years), and 6.9% had an underlyingpulmonary disease (e.g., asthma, pulmonary hyperten-sion, emphysema). Lungs, on average, had a partialpressure of oxygen/fraction of inspired oxygen (PaO2/FIO2) ratio of 259.18 § 66.92 mm Hg, with abnormalCXR or suspected pneumonia/infection present in 36.5%and 26.5% of cases, respectively.

Human lung decellularization

To demonstrate the ability to selectively remove airwayepithelium (Figure 4A), targeted lobes or bronchopulmonary

segments were decellularized by microcatheter delivery of adecellularization solution (CHAPS).25 Blue dye was addedto allow for bronchoscopic and gross visualization of thetreated area (Figure 4B). Histologic analysis by hematoxy-lin-and-eosin (H&E) staining demonstrated native lung withintact respiratory epithelium (Figure 4C, i‒iii) and treatedlung with loss of nuclear staining in distal alveoli andremoval of epithelium in large airways (Figure 4C, iv‒vi;see also Figure S1A in Supplementary Material online).Transmission electron microscopy (TEM) showed nativelung covered by type I pneumocyte cell membranes and ahealthy type II pneumocyte (arrow). After decellularization,TEM revealed the removal of type I pneumocytes andexposed basement membrane on either side of the blood‒gasbarrier. In Figure 4D (right), a preserved endothelial cell(star) is shown within a capillary adjacent to an injured typeII cell (arrow) detaching from the basement membrane. Tri-chrome staining of the lung demonstrated retention of extra-cellular matrix components, including collagen and elasticfibers (Figure 4E, i and ii), silver stain demonstrated preser-vation of reticular fibers (Figure S1B online), and Alcianblue stain showed retention of sub-mucosal glands and intactchondrocytes (see Figure S1C online) in decellularizedlung. Immunohistochemical staining similarly demon-strated the retention of basement membrane proteinscollagen IV and laminin in decellularized lung(Figure 4E, iii−vi). Methods to selectively target small

A B

C

D

Figure 2 Ex-vivo lung perfusion (EVLP) equipment and set-up. (A) Perfusion cart and organ preservation and housing perfusion, venti-lation, and imaging equipment for the extracorporeal lung. (B) EVLP circuit diagram with integrated circuit elements. A, airflow probe; H,heat exchanger; HC, humidified chamber; P, pressure sensor; PA, pulmonary artery; PV, pulmonary vein; Q, flow probe; T, temperatureprobe; WB, warm organ basin. Dhreservoir =§45 cm, Dhlung =§32 cm. (C) Flow (liters/min) within PA and PV cannulas of extracorporeallung (n = 6). (D) Transpulmonary gradient (TPG), which is the difference between PA and PV pressures (n = 6). All values indicate mean §standard deviation.

218 The Journal of Heart and Lung Transplantation, Vol 38, No 2, February 2019

airway segments are demonstrated on X-ray using a sin-gle customized dual-occlusion catheter (Figure 4F, i). Inthis approach, the catheter is inserted into the targetedairway, and the distal balloon is filled to occlusion.Then the decellularization solution is instilled, and theproximal balloon is expanded. In addition, the use ofindividually steerable micro-occlusion catheters alsoallowed for directed therapeutic delivery while isolatingproximal and/or distal areas during decellularizationtreatments (Figure 4F, ii), which resulted in the removalof pseudo-stratified epithelium in targeted airways(Figure 4F, iii).

Cell delivery and replacement in human lungs

Bronchoscopic cell delivery and the fundamental fluidmechanical forces acting on cells during deposition alongthe airway surface are shown schematically, where thehydrodynamic behaviors of cells delivered intratracheallyare influenced by the viscous force (FV), gravity (FG), andlift force (FL) (Figure 5A and B). A simplified force dia-gram of a spherical particle with diameter d can be appliedto cells given their spherical shape after trypsinization andresuspension. The buoyant force of the cell is negligible, ascell density is greater than that of the liquid. As the liquidplug travels through the airway, suspended cells carried bythe liquid will be deposited onto the airway wall surface.

Human MSCs cultured in vitro (Figure 5C, i) were labeledwith quantum dots (Figure 5C, ii) or CSFE (Figure 5C, iii),resuspended, and delivered bronchoscopically to targeted(non-decellularized) lung regions (Figure 5D). A compacttranspleural imaging system (Figure 5E, i) enabled real-time, non-invasive imaging of delivered cells without theneed for tissue biopsy. Labeled MSCs were visualizedtranspleurally immediately after delivery (Figure 5E, ii),and were found to be widely distributed throughout targetedregions of the distal lung. Post-procedural analysis of tissuesections confirmed homogeneous distribution and attach-ment of MSCs throughout the distal lung (Figure 5F) andwithin individual alveoli (Figure 5F, inset). Movement ofdelivered cells into the interstitium (see Figure S4A online)and coexpression with integrin-b1 (see Figure S4B online)suggested early engraftment.

After decellularization, heterogeneous populations ofpulmonary airway epithelial cells (AECs) were deliveredusing microvolume delivery techniques, as previouslydescribed.20,44 In larger proximal airways, a hydrogel com-prised of lung extracellular matrix enriched in lung-specificbasement membrane components was delivered before theAECs to facilitate attachment and counteract the mechani-cal forces of gravity and ventilation. Post-procedural analy-sis demonstrated AECs lining larger airways (Figure 5G, i)and distributed within alveoli in the distal lung (Figure 5G,ii). Their topographic distribution was confirmed on histo-logic analysis (see Figure S4C online), and costainingwith collagen IV (Figure 5H; see also Figure S4D online)demonstrated the retention of a key basement membranecomponent after decellularization as well as the importanceof extracellular matrix integrity for the attachment andintegration of newly delivered cells.

Discussion

Worldwide the number of patients with end-stage lungdisease continues to rise,45 and a viable alternative totraditional transplantation and donor lung utilization isdesperately needed. Cellular therapeutics may play a majorrole in solving this problem, and the use of patient-derivedcells circumvents many of the ethical and immunologicissues related to allogeneic cell‒based therapies.

Over the last decade, there have been significantadvancements in 2 lung-related fields: EVLP and lung bio-engineering approaches based on the use of decellularizedscaffolds. Many groups have used the EVLP techniquesand protocols pioneered by Steen and colleagues46 and theToronto group24 to support, assess, and recover marginalquality donor lungs for transplantation.31,47,48 Meanwhile,the tools and methods for tissue decellularization andscaffold characterization (e.g., analysis of extracellularmatrix components and biomechanics) have drasticallyimproved.49,50 However, the de-novo bioengineering offunctional human lungs suitable for transplantation remainschallenging despite advances in these technologies.

To address the challenges of whole-organ bioengi-neering, we developed an approach to selectively targetand intervene in specific lung regions. Cell replacement

Table 1 Characteristics of Donors With Lungs ConsideredUnsuitable for Transplantation (n = 756)

Age (years) (n = 432) 45.7 § 6.9Gender (n = 709)Male 424 (59.8%)Female 285 (40.2%)

Cause of brain death(n = 301)

Trauma 171 (56.8%)Intracerebral hemorrhage 103 (34.2%)Other 27 (8.9%)Presence of underlyinglung disease (n = 579)

40 (6.9%)

Smoking history of >20pack-years (n = 616)

142 (23.1%)

PaO2/FIO2 ratio (mm Hg)(n = 233)

259.18 § 66.92

PaO2 on 100% FIO2(mm Hg) (n = 49)

236.17 § 101.26

Abnormal CXR (edema orinfiltrates) (n = 411)

150 (36.5%)

Suspected PNA oraspiration (n = 491)

130 (26.5%)

Time on ventilator (hours) (n = 51) 108.7 § 47.1Time from brain death toprocurement (hours) (n = 192)

36.65 § 14.64

Data are expressed as mean § standard deviation or as number (%)and were taken from combined analyses from 18 peer-reviewed litera-ture reports. The total number of donors (n) is reported in the left col-umn for each category. CXR, chest X-ray; FIO2, fraction of inspiredoxygen; PaO2, partial pressure of oxygen; PNA, pneumonia.

Guenthart et al. Human Lung Bioengineering 219

strategies targeting the treatment of only 20% to 25%may be sufficient to improve lung function and make anorgan suitable for transplantation, or to reverse a spe-cific disease phenotype. This approach addresses manyof the challenges impeding current bioengineering

efforts, including: (i) Scale—the human lung is com-prised of roughly 2£ 1011 pneumocytes51 and 75£ 109

endothelial cells,52 with over 40 distinct cell pheno-types. Treating smaller lung volumes allows for the useof fewer cells and the selection of cell types most

Figure 3 Donor lung cannulation techniques and donor lung characteristics. (A) (i) Left atrial cuff after removal of the heart allowingvisualization of pulmonary veins (arrow heads). T, trachea; PA, pulmonary artery. (ii) Dissected human aortic arch used for cannulationand serving as a bio-bridge in the management of pulmomary venous drainage. DA, descending aorta; IA, innominate artery; LSC, left sub-clavian artery; LCC, left common carotid artery; RCC, right common carotid artery; RSC, right sub-clavian artery. Lung cannulation withthe aortic arch (AA) serving as a bio-bridge and a portion of descending aorta serving as a graft to facilitate the insertion of pulmonaryvenous (PV) and arterial (PA) cannulas. Endotracheal tube (ETT) secured within the trachea (iii). (B) Donor tissue options for use as a bio-bridge. (i) Aortic arch, (ii) abdominal aorta, (iii) pericardium, and (iv) cannulation of lungs using aortic arch or (v) pericardium. (C) Imagingexamples of lungs unsuitable for transplantation before procurement. (i) Computed tomography (CT) scan of chest demonstrating collapse/consolidation of the right lower lobe and left lower lobe base, and (ii) chest X-ray (CXR) with opacity/atelectasis of the right lower lobe.(iii, iv) Gross appearance of unsuitable lungs and (v, vi) bronchoscopic view of the airway demonstrating inflammation and removal of amucous plug, respectively. (D) Gross appearance (i) and thermography (ii) of human lungs after procurment, cold flush, and placementwithin a warm organ basin. (E) Periodic acid‒Schiff (PAS) stain of bronchoalveolar lavage fluid obtained immediately after human lungprocurement. (F) Delivery and uptake of cell viability marker (CFSE) by alveolar cells in the distal lung of unsuitable human donor lungs.

220 The Journal of Heart and Lung Transplantation, Vol 38, No 2, February 2019

appropriate for targeted delivery locations. In addition,this addresses the practical challenges related to themaintenance, expansion, and tumor-forming potential ofpluripotent cells during large-scale production.53 (ii)Blood gas barrier—most decellularization approachesare perfusion-based, and therefore necessitate recellulari-zation of both the vascular and airway compartments.The vascular networks of the lung are among the mostintricate in the human body. The integrity of this inter-face is vital to proper lung function and constitutes amajor engineering hurdle. Our method of intratrachealdecellularization preserves native lung vasculatureand ensures that delivered cells are supplied with

blood-based nutrients and factors necessary for engraft-ment and differentiation.54

Our platform enabled delivery, visualization, and attach-ment of MSCs. Evidence suggests MSCs exert a variety ofimmunomodulatory effects and that the use of allogeneicMSCs is safe. We recognize the limitations of MSCs for reca-pitulating the normal cellular distribution of cell types withinthe lung; however, they serve as an efficient and economicalprototype for other cellular strategies, such iPS or distal basalcells.55−57 Additional studies are needed to determine optimalcell doses and delivery route, and whether co-delivery withother cell types has benefit in the application of bioengineeredconstructs. Our regional decellularization technique resulted

Figure 4 Human lung decellularization, extracellular matrix preservation, and targeted therapeutic approach. (A) Schematic of lungdecellularization demonstrating removal of lung epithelium with preservation of the basement membrane (BM) and extracellular matrix(ECM). (B) Regional decellularization of the lung (blue dye added for visualization). (i) Bronchoscopic view and (ii) gross appearance afterdelivery into the lateral segment of the right middle lobe. (C) Histologic comparison of native and de-epithelialized lung. (i−iii) Hematoxy-lin-and-eosin (H&E) stain of native lung and (iv−vi) decellularized lung demonstrating removal of lung epithelium (arrows). (D) Transmis-sion electron microscopy (TEM) of native lung demonstrating intact blood‒gas barrier with type II cell (arrow) and decellularized lungwith removal of type I pneumocyte cell membranes, exposure of the basement membrane, detached type II cell (arrow), and intact endothe-lial cell (star). (E) Retention of ECM components. (i−ii) Trichrome stain showing retention of collagen (blue) and elastic fibers (purple) sur-rounding airways and vessels (stars) with removal of airway epithelium (arrows). (iii−vi) Collagen IV and laminin immunostaining innative and decellularized lung. (F) Targeted lung decellularizaiton methods. (i) X-ray imaging showing a double-balloon occlusion cathetersystem to treat segments of proximal conducting airways. Treatment area highlighted in blue and marked by arrow. (ii) Dual-catheter use toocclude 1 lung segment (arrow) while treating or also occluding more distal lung (star). (iii) Histologic analysis of large airway after treat-ment demonstrating removal of pseudo-stratified columnar epithelium (arrows).

Guenthart et al. Human Lung Bioengineering 221

in the selective removal of the airway epithelium whilepreserving the pulmonary endothelium and extracellularmatrix (Figure 4; see also Figure S1 online). This approachresulted in the creation of a lung niche that could supportthe attachment of airway epithelial cells in proximal anddistal airspaces (Figure 5; see also Figure S4 online).Targeted intervention into lobar and sub-segmental regionsas well as segments of proximal conducting airways wasfacilitated by the use of custom microcatheters, transpleuralimaging, lung extracellular matrix hydrogel, and balloonocclusion techniques—representing an integrated systemsapproach to tailored lung bioengineering. Aberrancy inairway epithelium has been linked to pathologic states ofdisease,58 and our understanding of disease pathogenesisand cellular derangement continues to improve. Furtherrefinements in these techniques and devices will allow formore localized targeting, and more sophisticated decellulari-zation agents (e.g., toxin-linked or antibody-mediated)may enable the highly selective removal and subsequentreplacement of specific cell types.

Our study was designed and limited to serve as an initialproof-of-concept to demonstrate the feasibility of targeteddecellularization and recellularization in extracorporeallungs. There are, however, several limitations to the currentstudy: (i) although techniques and cell interventions werebeing refined, investigations into lung function and gasexchange did not take place, as our EVLP system was basedon acellular perfusate and mechanics was not the focusof this study; and (ii) cell-level analysis, such as geneexpression and ability to observe differentiation, was notconducted, given the relatively short duration of time lungswere maintained on ex-vivo support. Currently, this is amajor limitation in the field and further cell delivery studiesare planned utilizing our cross-circulation platform forprolonged normothermic support to study cell replacementtherapy over the course of days.

Continued advancement in the field of human lungbioengineering will require improved physiologic andbiochemical assessment of donor lungs. Currently, there isno standardized reporting of donor characteristics, and

Figure 5 Cell replacement in human lungs. (A) Targeted intratracheal cell delivery and deposition onto airway surfaces. (B) Cell depo-sition by microvolume liquid plug delivery. Hydrodynamic motion of cells influenced by internal circulatory flow patterns and forces, FL(lift), FV (viscous), and FG (gravity). (C) Human MSCs (i) cultured in vitro and labeled with (ii) quantum dots (Qdot) or CSFE (iii). (D) X-ray image during bronchoscopic delivery (arrow marks distal tip). (E) (i) Transpleural imaging camera used to confirm the (ii) delivery anddistribution of delivered cells. (F) Histologic analysis of delivered MSCs. Global distribution and within a single alveolus (inset). (G) Cellreplacement by targeted delivery of airway epithelial cells (AECs) after decellularization. (H) Collagen IV immunostaining highlightingretention of a key extracellular matrix component critical for the attachment of cells to the basement membrane. Delivered cells (MSCs andAECs) demonstrate flattening and morphologic adaptations to surrounding alveoli.

222 The Journal of Heart and Lung Transplantation, Vol 38, No 2, February 2019

the reasons why lungs are not transplantable are oftenmultifactorial; thus, the true incidence of specific pathologyis difficult to determine. Biochemical markers of alveolarepithelial barrier dysfunction,59 donor exosome profiles,60

and detailed cytokine analyses61 of donor lungs mayimprove donor‒recipient matching, enable better predictionof outcomes, and help identify lungs most amenable to cellreplacement therapy in the future.

In this investigation we have demonstrated that theutilization of human lungs unsuitable for transplantationand supported by EVLP allowed for the development andrefinement of: (i) an integrated lung support platform withadvanced theranostics; (ii) selective removal of native lungepithelium; (iii) maintenance of extracellular matrix com-ponents essential for providing the biophysical andbiochemical cues to therapeutic replacement cells; and (iv)effective delivery and attachment of therapeutic cells(MSCs and AECs) into targeted lung regions. These resultssuggest that targeted cell replacement in extracorporealorgans is feasible and may ultimately lead to chimericorgans suitable for transplantation. Furthermore, as mecha-nisms for somatic reprogramming and genetic manipulationof patient-specific cell lines continue to advance, this workmay have profound translational application in patientsduring early in-situ interventions to halt or reverse diseasetrajectory (Figure 1B), and ultimately preclude the need fortransplantation altogether.

Disclosure statement

The authors have no conflicts of interest to disclose. Wethank S. Halligan and K. Cunningham for administrativeand logistical support; H. Wobma for providing mesenchy-mal cells; L. Cohen-Gould for TEM imaging; and T. Wu,D. Sun, and R. Chen (Herbert Irving Comprehensive Can-cer Center Molecular Pathology Shared Resources) for helpwith analytics. This study was supported by the NationalInstitutes of Health (HL134760, EB002520, HL007854),the Rick Bartlett Memorial Foundation, and the MikatiFoundation.

Supplementary data

Supplementary data are available in the online version ofthis article at https://doi.org/10.1016/j.healun.2018.11.007.

References

1. Cypel M, Keshavjee S. Extending the donor pool: rehabilitation of

poor organs. Thorac Surg Clin 2015;25:27-33.2. Elgharably H, Shafii AE, Mason DP. Expanding the donor pool: dona-

tion after cardiac death. Thorac Surg Clin 2015;25:35-46.

3. Botha P, Trivedi D, Weir CJ, et al. Extended donor criteria in lung

transplantation: impact on organ allocation. J Thorac Cardiovasc Surg2006;131:1154-60.

4. Valapour M, Skeans M, Heubner B, et al. OPTN/SRTR 2013 annual

data report: lung. Am J Transplant 2015;15:1-28.

5. Wagner DE, Bonvillain RW, Jensen T, et al. Can stem cells be used togenerate new lungs? Ex vivo lung bioengineering with decellularized

whole lung scaffolds. Respirology 2013;18:895-911.

6. Ochs M, Nyengaard JR, Jung A, et al. The number of alveoli in thehuman lung. Am J Respir Crit Care Med 2004;169:120-4.

7. Ott HC, Clippinger B, Conrad C, et al. Regeneration and orthotopic

transplantation of a bioartificial lung. Nat Med 2010;16:927-33.

8. Petersen TH, Calle EA, Zhao L, et al. Tissue-engineered lungs for invivo implantation. Science 2010;329:538-41.

9. Gilpin SE, Guyette JP, Gonzalez G, et al. Perfusion decellularization

of human and porcine lungs: bringing the matrix to clinical scale.

J Heart Lung Transplant 2014;33:298-308.10. Wagner DE, Bonenfant NR, Sokocevic D, et al. Three-dimensional

scaffolds of acellular human and porcine lungs for high throughput stud-

ies of lung disease and regeneration. Biomaterials 2014;35:2664-79.11. Aggarwal S, Pittenger MF. Human mesenchymal stem cells modulate

allogeneic immune cell responses. Blood 2005;105:1815-22.

12. Ortiz LA, DuTreil M, Fattman C, et al. Interleukin 1 receptor antago-

nist mediates the antiinflammatory and antifibrotic effect of mesenchy-mal stem cells during lung injury. Proc Natl Acad Sci 2007;104:

11002-7.

13. Wobma HM, Liu D, Vunjak-Novakovic G. Paracrine effects of mesen-

chymal stromal cells cultured in three-dimensional settings on tissuerepair. ACS Biomater Sci Eng 2018;4:1162-75.

14. Islam MN, Das SR, Emin MT, et al. Mitochondrial transfer from bone-

marrow-derived stromal cells to pulmonary alveoli protects againstacute lung injury. Nat Med 2012;18:759-65.

15. Takahashi K, Yamanaka S. Induction of pluripotent stem cells from

mouse embryonic and adult fibroblast cultures by defined factors. Cell

2006;126:663-76.16. Fitzhugh CD, Hsieh MM, Cordes S, et al. At least 20% donor myeloid

chimerism is necessary to reverse the sickle phenotype after allogeneic

hematopoietic stem cell transplantation. Am Soc Hematol 2017;130:

1946-8.17. Walters M, Patience M, Leisenring W, et al. Stable mixed hematopoi-

etic chimerism after bone marrow transplantation for sickle cell ane-

mia. Biol Blood Marrow Transplant 2001;7:665-73.18. Johnson LG, Olsen JC, Sarkadi B, et al. Efficiency of gene transfer for

restoration of normal airway epithelial function in cystic fibrosis. Nat

Genet 1992;2:21-5.

19. Ramalho AS, Beck S, Meyer M, et al. Five percent of normal cysticfibrosis transmembrane conductance regulator mRNA ameliorates the

severity of pulmonary disease in cystic fibrosis. Am J Respir Cell Mol

Biol 2002;27:619-27.

20. Kim J, O’Neill JD, Dorrello NV, et al. Targeted delivery of liquidmicrovolumes into the lung. Proc Natl Acad Sci 2015;112:11530-5.

21. Saxena P, Zimmet AD, Snell G, et al. Procurement of lungs for trans-

plantation following donation after circulatory death: the Alfred tech-

nique. J Surg Res 2014;192:642-6.22. Sundaresan S, Trachiotis GD, Aoe M, et al. Donor lung procure-

ment: assessment and operative technique. Ann Thorac Surg

1993;56:1409-13.23. O’Neill JD, Guenthart BA, Kim J, et al. Cross-circulation for extracor-

poreal support and recovery of the lung. Nat Biomed Eng 2017;1: 0037.

24. Cypel M, Yeung JC, Hirayama S, et al. Technique for prolonged nor-

mothermic ex vivo lung perfusion. J Heart Lung Transplant 2008;27:1319-25.

25. O’Neill JD, Anfang R, Anandappa A, et al. Decellularization of human

and porcine lung tissues for pulmonary tissue engineering. Ann Thorac

Surg 2013;96:1046-56.26. Cypel M, Liu M, Rubacha M, et al. Functional repair of human donor

lungs by IL-10 gene therapy. Sci Transl Med 2009;1:4ra9.

27. Wierup P, A!Haraldsson, Nilsson F, et al. Ex vivo evaluation of nonac-

ceptable donor lungs. Ann Thorac Surg 2006;81:460-6.

28. Egan TM, Haithcock JA, Nicotra WA, et al. Ex vivo evaluation of

human lungs for transplant suitability. Ann Thorac Surg 2006;81:

1205-13.29. Ingemansson R, Eyjolfsson A, Mared L, et al. Clinical transplantation

of initially rejected donor lungs after reconditioning ex vivo. Ann

Thorac Surg 2009;87:255-60.

30. Ware LB, Wang Y, Fang X, et al. Assessment of lungs rejected fortransplantation and implications for donor selection. Lancet 2002;360:

619-20.

Guenthart et al. Human Lung Bioengineering 223

31. Wallinder A, Ricksten S-E, Hansson C, et al. Transplantation ofinitially rejected donor lungs after ex vivo lung perfusion. J Thorac

Cardiovasc Surg 2012;144:1222-8.

32. Pego-Fernandes P, de Medeiros I, Mariani A, et al. Ex vivo lung perfu-

sion: early report of Brazilian experience. Transplant Proc 2010;42:440-3.

33. Boffini M, Ricci D, Barbero C, et al. Ex vivo lung perfusion increases

the pool of lung grafts: analysis of its potential and real impact on a

lung transplant program. Transplant Proc 2013;45:2624-6.34. Ware LB, Fang X, Wang Y, et al. High prevalence of pulmonary arte-

rial thrombi in donor lungs rejected for transplantation. J Heart Lung

Transplant 2005;24:1650-6.35. Medeiros IL, Pego-Fernandes PM, Mariani AW, et al. Histologic

and functional evaluation of lungs reconditioned by ex vivo lung

perfusion. J Heart Lung Transplant 2012;31:305-9.

36. Andreasson A, Karamanou DM, Perry JD, et al. The effect of ex vivolung perfusion on microbial load in human donor lungs. J Heart Lung

Transplant 2014;33:910-6.

37. Mariani AW, Medeiros IL, Pego-Fernandes PM, et al. Cold ischemia

or topical-ECMO for lung preservation: a randomized experimentalstudy. Sao Paulo Med J 2014;132:28-35.

38. Lee JW, Fang X, Gupta N, et al. Allogeneic human mesenchymal stem

cells for treatment of E. coli endotoxin-induced acute lung injuryin the ex vivo perfused human lung. Proc Natl Acad Sci 2009;106:

16357-62.

39. La Francesca S, Ting AE, Sakamoto J, et al. Multipotent adult progen-

itor cells decrease cold ischemic injury in ex vivo perfused humanlungs: an initial pilot and feasibility study. Transplant Res 2014;3:19.

40. Wallinder A, Ricksten S-E, Silverborn M, et al. Early results in trans-

plantation of initially rejected donor lungs after ex vivo lung perfu-

sion: a case−control study. Eur J Cardiothorac Surg 2014;45:40-5.41. McAuley DF, Curley GF, Hamid UI, et al. Clinical grade allogeneic

human mesenchymal stem cells restore alveolar fluid clearance in

human lungs rejected for transplantation. Am J Physiol 2014;306:L809-15.

42. McElhinney DB, Khan JH, Babcock WD, et al. Thoracic organ donor

characteristics associated with successful lung procurement. Clin

Transplant 2001;15:68-71.43. George TJ, Arnaoutakis GJ, Beaty CA, et al. A physiologic and bio-

chemical profile of clinically rejected lungs on a normothermic ex

vivo lung perfusion platform. J Surg Res 2013;183:75-83.

44. Kim J, Guenthart B, O’Neill JD, et al. Controlled delivery andminimally invasive imaging of stem cells in the lung. Sci Rep 2017;7.

45. Marciniuk D, Ferkol T, Nana A, et al. Respiratory diseases in the

world. Realities of today—opportunities for tomorrow. Afr J Respir

Med 2014;9.

46. Steen S, Sjoberg T, Pierre L, et al. Transplantation of lungs from anon-heart-beating donor. Lancet 2001;357:825-9.

47. Cypel M, Yeung JC, Liu M, et al. Normothermic ex vivo lung

perfusion in clinical lung transplantation. New Engl J Med

2011;364:1431-40.48. Zych B, Popov AF, Stavri G, et al. Early outcomes of bilateral sequen-

tial single lung transplantation after ex-vivo lung evaluation and

reconditioning. J Heart Lung Transplant 2012;31:274-81.

49. Wobma H, Vunjak-Novakovic G. Tissue engineering and regenerativemedicine 2015: a year in review. Tissue Eng B Rev 2016;22:

101-13.

50. Gilpin SE, Ott HC. Using nature’s platform to engineer bio-artificiallungs. Ann Am Thorac Soc 2015;12(suppl):S45-S9.

51. Mercer RR, Russell ML, Roggli VL, et al. Cell number and distribu-

tion in human and rat airways. Am J Respir Cell Mol Biol

1994;10:613-24.52. Stone KC, Mercer RR, Freeman BA, et al. Distribution of lung cell

numbers and volumes between alveolar and nonalveolar tissue. Am

Rev Respir Dis 1992;146:454-6.

53. Fox IJ, Daley GQ, Goldman SA, et al. Use of differentiated pluripotentstem cells in replacement therapy for treating disease. Science

2014;345:1247391.

54. Dorrello NV, Guenthart BA, O’Neill JD, et al. Functional vascularizedlung grafts for lung bioengineering. Sci Adv 2017;3:e1700521.

55. Huang SX, Green MD, de Carvalho AT, et al. The in vitro generation

of lung and airway progenitor cells from human pluripotent stem cells.

Nat Protoc 2015;10:413-25.56. Huang SX, Islam MN, O’Neill J, et al. Efficient generation of lung and

airway epithelial cells from human pluripotent stem cells. Nat Biotech-

nol 2014;32:84-91.

57. Chen Y-W, Huang, de Carvalho ALRT SX, et al. A three-dimensionalmodel of human lung development and disease from pluripotent stem

cells. Nat Cell Biol 2017;19:542.

58. Zemans RL, Matthay MA. Bench-to-bedside review: the role of thealveolar epithelium in the resolution of pulmonary edema in acute

lung injury. Crit Care 2004;8:469.

59. Frank JA, Briot R, Lee JW, et al. Physiological and biochemical

markers of alveolar epithelial barrier dysfunction in perfused humanlungs. Am J Physiol 2007;293:L52-9.

60. Gunasekaran M, Xu Z, Nayak D, et al. Donor-derived exosomes with

lung self-antigens in human lung allograft rejection. Am J Transplant

2017;17:474-84.61. Kaneda H, Waddell T, De Perrot M, et al. Pre-implantation multiple

cytokine mRNA expression analysis of donor lung grafts predicts

survival after lung transplantation in humans. Am J Transplant

2006;6: 544-51.

224 The Journal of Heart and Lung Transplantation, Vol 38, No 2, February 2019

Cell Replacement in Human Lung Bioengineering 1 2 Brandon A. Guenthart, John D. O’Neill, Jinho Kim, Kenmond Fung, Gordana Vunjak-Novakovic, 3 and Matthew Bacchetta 4 5 6 7 8

Online Data Supplement 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

Additional Methods: 52 53 54 Ex vivo lung perfusion (additional details) – The circuit was primed with preservation solution 55 (Perfadex) containing heparin 10,000U (Sagnet), cefazolin 500mg (WG Critical Care), albumin 56 100g, and methyleprednisolone 500mg (APP Pharmaceuticals), de-aired, and connected to the 57 extracorporeal lung. Initial flow rates were set to 5 – 10% of estimated cardiac output with a target 58 PA pressure of < 15 mmHg and PV pressure of 3 – 5 mmHg. The circuit and lungs were allowed 59 to acclimate to ambient temperature and slowly warmed to 37°C over ten minutes. Ventilation 60 was initiated within the first 10 minutes of perfusion with the following initial settings: volume 61 control mode, respiratory rate 6 – 8 bpm, tidal volume (TV) 6 – 8 mL/kg, positive end-expiratory 62 pressure (PEEP) 5cm H2O, and FiO2 40% (Oxylog 3000 plus; Dräger). Atelectatic lung regions 63 were recruited by increasing TV, PEEP (up to 10 cm H2O), and by performing inspiratory hold 64 maneuvers (up to 25 cm H2O). 65 66 Immunohistochemical staining – Sample collection was conducted using a surgical stapler with 67 medium/thick reloads. Specimens were immediately fixed in cold phosphate-buffered 4% 68 paraformaldehyde for 48–72h, embedded in paraffin, and sectioned at 3–5 µm thickness. Lung 69 sections were de-paraffinized, subjected to boiling citrate buffer (pH 6.0) for antigen retrieval, and 70 blocked with 10% normal goat serum in PBS for 2 hours at room temperature. Primary antibodies 71 were added and incubated for 12 hours at 4°C, or 4 hours at room temperature. Secondary 72 antibodies were diluted 1:200 and incubated for 1 hour at room temperature for all stains. Sections 73 were mounted in Vectashield Mounting Medium with DAPI (Vector Laboratories), coverslipped, 74 and imaged with an Olympus FSX100 microscope or Olympus BX61VS virtual slide microscope. 75 Additional sections were stained for silver reticulin, Alcian blue (pH 2), and trichrome by the 76 histology service Department of Molecular Pathology at Columbia University. 77 78 Transmission electron microscopy – Lung samples were fixed with 2.5% glutaraldehyde in 0.1M 79 Sorenson’s buffer (pH 7.2). Samples were then post-fixed with 1% OsO4 in Sorenson’s buffer for 80 1 hour. After dehydration, tissues were embedded in Lx-112 (Ladd Research Industries). Thin 81 sections were cut on the PT-XL ultramicrotome at 60 nm thickness. Sections were stained with 82 uranyl acetate and lead citrate and examined under a JEOL JEM-1200 EXII electron microscope. 83 Images were captured with an ORCA-HR digital camera (Hamamatsu) and recorded with AMT 84 Image Capture Engine. 85 86 Delivery of human mesenchymal stem cells – Isolation and culture of mesenchymal stem cells 87 followed protocols previously published by our group(1, 2). In brief, adipose derived MSCs were 88 obtained from human lipoaspirates (fully de-identified commercially available samples; LaCell). 89 MSCs were tested to confirm trilineage differentiation and found to express MSC surface markers, 90 according to established protocols(3). Extensive proteomic and metabolomic analyses and 91 description of cell behavior of the cell population used in this study has previously been 92 reported(1). Cells were cultured and expanded up to passage 3 before use. At the time of the 93 experiment, cells were trypsinized, counted with the Countess Automated Cell Counter 94 (Invitrogen), and re-suspended in PBS at a final concentration of 2.0 × 107 cells/mL for labeling 95 with CSFE (Abcam) or carboxyl Qdot solution (Invitrogen) according to the manufacturers’ 96 instructions. Cells were delivered at the start of EVLP to provide up to 6 hours incubation time 97 within the lung. Delivery was carried out in multiple 500 µm volumes, with periods of ventilation in 98 between(4), total volume of delivery varied from 1–10 mL (proportional to the size of targeted lung 99 tissue). 100 101

Delivery of human airway epithelial cells – Following decellularization, shake tests of BAL fluids 102 were conducted to confirm the removal of residual detergent before the introduction of cells. In 103 this study, commercially available cryopreserved human small airway epithelial cells (Lonza) were 104 utilized. These cells are obtained from distal lung regions (bronchioles up to 1mm) of normal 105 human lung. Methods for cell collection and isolation have been previously described(5, 6), as 106 well as analysis of cell composition and gene expression(7, 8). This cell population is capable of 107 mucin production, cilia formation, and the development of high transepithelial electrical resistance 108 (an indirect measure of formation of tight junctions)(9). Cultured cells were passaged by 109 trypsinization, labeled with CFSE according to the manufacturer’s instructions, and re-suspended 110 at a concentration of 5 – 10 × 106 cells/mL. Cells were delivered at the conclusion of 111 decellularization, providing 4 hours incubation time within the lung. Cells were delivered in 112 hydrogel (East River Bio.) – a lung-specific extracellular matrix hydrogel to facilitate cell 113 attachment and engraftment in proximal airways. Delivery was carried out in multiple 500 µm 114 volumes, with periods of ventilation in between(4), total volume of delivery varied from 1–10 mL 115 (proportional to the size of targeted lung tissue). To assess the distribution and attachment of 116 delivered cells, lung wedge samples were collected at the conclusion of the experiment, fixed, 117 embedded, de-paraffinized, and imaged, as previously described. 118 119 Thermography – Thermal images of donor lungs were captured following lung procurement, flush 120 and selected time points throughout the EVLP procedure using a T430sc infrared camera (FLIR) 121 and ResearchIR v. 4.40.6 software (FLIR). 122 123 124 125 Supplementary Text: 126 127 EVLP Strategy 128 129 Our group previously demonstrated prolonged extracorporeal support of lungs beyond 36 hours 130 utilizing a cross-circulation platform with flow rates of 5-10% of cardiac output. With this approach 131 we demonstrated the maintenance of lung function along with cell viability(10). It is our practice 132 and working hypothesis that protective ventilation in combination with a reduced flow rate is 133 protective during EVLP (and provides sufficient support to maintain cellular viability and function), 134 and that the gradual ramp-up to 40% of cardiac output can be safely implemented prior to 135 transplantation. Further work is needed to validate this approach and elucidate its potential 136 benefits and pitfalls. 137 138 Decellularization technique 139 140 We previously reported methods and results of single lung airway de-epithelialization (i.e., airway 141 decellularization) utilizing EVLP in a rat model. Decellularization of the airways resulted in the 142 reduction of DNA by 23.4%(11). Given the composition of rat lungs (43% endothelium, 32% 143 interstitial, 22% epithelial)(12), the treatment effect was largely confined to the epithelium. This 144 technique has also been used in a swine model with lungs on cross-circulation. De-145 epithelialization resulted in the removal of pulmonary epithelium, preservation of underlying 146 vasculature and ECM, and had no adverse effects on overall lung function or stability/safety of 147 the host(10). 148 149 Here we describe decellularization of human lungs on EVLP, targeting the epithelium within 150 specified and localized lung regions. Comprehensive histologic analysis of decellularized lung 151

regions demonstrated retention of lung structure, underlying pulmonary endothelium, and the 152 retention of key extracellular matrix components (Fig. 4, Sup. Fig. 1). 153 154 Cell delivery and attachment 155 156 In vitro MSCs display a fibroblast-like morphology (Sup. Fig. 2A,B). When preparing cells for 157 delivery they are subjected to trypsinization, after which cells are placed in suspension. Following 158 this process, cells display a distinctly round morphology (Sup. Fig. 2C). To observe the 159 appearance and behavior of non-viable cells delivered into the lungs a series of control 160 experiments were carried out. MSCs were treated with benzoanse (an endonuclease) and 161 delivered and allowed to incubate in the lung utilizing the same methodology and under the same 162 conditions as previously described. Following sample collection, fixation, sectioning, and staining, 163 very few cells were observed and retained within the lung. Cells that did remain were observed 164 as unattached single cells (Sup. Fig. 3A, i) or as cellular aggregates (Supplementary Figure 3a, 165 ii). These non-viable, unattached cells maintain a rounded morphology. 166 167 MSCs and AECs delivered into targeted regions of human lungs on EVLP demonstrated several 168 key morphologic changes that support the conclusion that cells attached and integrated with their 169 environment. MSCs and AECs are seen throughout targeted lung segments, and in high 170 concentrations. Cells demonstrate flattening and morphological adaptations (Fig 5; Sup. Fig. 4) 171 which were not observed in control experiments (Sup. Fig. 3). Col IV staining highlights the 172 retention of a key extracellular component, and its close approximation to cells attached. Cells 173 located within the interstitium (Supplementary Figure 4A), and Integrin- β1 staining 174 (Supplementary Figure 4B) provide further supporting evidence of cell viability, function, and 175 engraftment. 176 177 In an attempt to further characterize delivered cells, transmission electron microscopy (TEM) was 178 iteratively attempted on regions of recellularized epithelium, however, technical challenges limited 179 the success of these efforts. The extremely small sample size (~ 1 mm) required for TEM 180 sectioning and imaging constrained the areas of analysis and prohibited extensive assessment 181 of recellularized lung regions. Furthermore, identification via TEM of exogenous cells (e.g., 182 epithelial or mesenchymal stem cells delivered into the lungs) is immensely difficult given the 183 relatively small number of delivered cells compared to the vast number of endogenous cells. As 184 a result, our group developed a transpleural microscopic imaging system with fluorescence 185 capabilities to microscopically visualize and analyze delivered cells across much larger regions 186 of the lung. Lastly, TEM lacks the ability to easily label/stain and verify native versus exogenous 187 (i.e., delivered) cells. For these reasons, TEM following cell delivery was not included in the 188 analyses of this paper. The further refinement and development of advanced imaging modalities 189 will undoubtable contribute to our understanding of cell delivery into the lung. 190 191 192 Supplemental References 193 194 1. Wobma HM, Tamargo MA, Goeta S, Brown LM, Duran-Struuck R, Vunjak-Novakovic G: 195 The influence of hypoxia and IFN-γ on the proteome and metabolome of therapeutic 196 mesenchymal stem cells. Biomaterials 2018;167:226-34. 197 2. Wobma HM, Kanai M, Ma SP, et al.: Dual IFN-γ/hypoxia priming enhances 198 immunosuppression of mesenchymal stromal cells through regulatory proteins and metabolic 199 mechanisms. Journal of Immunology and Regenerative Medicine 2018;1:45-56. 200

3. Krampera M, Galipeau J, Shi Y, Tarte K, Sensebe L: Immunological characterization of 201 multipotent mesenchymal stromal cells—The International Society for Cellular Therapy (ISCT) 202 working proposal. Cytotherapy 2013;15:1054-61. 203 4. Kim J, Guenthart B, O’Neill JD, Dorrello NV, Bacchetta M, Vunjak-Novakovic G: Controlled 204 delivery and minimally invasive imaging of stem cells in the lung. Scientific Reports 2017;7. 205 5. Banerjee B, Kicic A, Musk M, Sutanto EN, Stick SM, Chambers DC: Successful 206 establishment of primary small airway cell cultures in human lung transplantation. Respiratory 207 research 2009;10:99. 208 6. Harvey B-G, Heguy A, Leopold PL, Carolan BJ, Ferris B, Crystal RG: Modification of gene 209 expression of the small airway epithelium in response to cigarette smoking. Journal of molecular 210 medicine 2007;85:39-53. 211 7. Hackett NR, Butler MW, Shaykhiev R, et al.: RNA-Seq quantification of the human small 212 airway epithelium transcriptome. BMC genomics 2012;13:82. 213 8. Buro-Auriemma LJ, Salit J, Hackett NR, et al.: Cigarette smoking induces small airway 214 epithelial epigenetic changes with corresponding modulation of gene expression. Human 215 molecular genetics 2013;22:4726-38. 216 9. Stewart CE, Torr EE, Jamili M, Nur H, Bosquillon C, Sayers I: Evaluation of differentiated 217 human bronchial epithelial cell culture systems for asthma research. Journal of allergy 218 2012;2012. 219 10. O’Neill JD, Guenthart BA, Kim J, et al.: Cross-circulation for extracorporeal support and 220 recovery of the lung. Nature Biomedical Engineering 2017;1:0037. 221 11. Dorrello NV, Guenthart BA, O’Neill JD, et al.: Functional vascularized lung grafts for lung 222 bioengineering. Science Advances 2017;3:e1700521. 223 12. Haies DM, Gil J, Weibel ER: Morphometric study of rat lung cells: I. Numerical and 224 dimensional characteristics of parenchymal cell population. American Review of Respiratory 225 Disease 1981;123:533-41. 226 227

Supplementary Figure 1 | Human lung decellularization. Histologic comparison of native and decellularized lung. (A)H&E of native and decellularized lung demonstrating removal of airway epithelium (arrows). (B) Silver stain of native anddecellularized conductive airways (i-ii) and native and decellularized distal alveoli (iii-iv) showing preservation of reticularfibers following decellularization and removal of airway epithelium (arrows). Blood vessel indicated by asterisk. (C) Alcianblue stain of native and decellularized lung demonstrating removal of airway epithelium (arrows) and retention ofsubmucosal glands and intact chondrocytes in airway cartilage (inset) following decellularization. (D) Transmissionelectron microscopy (TEM) of decellularized lung. Arrows indicate fragments of damaged epithelium. Capillary indicatedby asterisk. BM, basement membrane; EN, endothelium; INT, interstitium.

a

AN

ativ

eD

ecel

l

H&E

25 μm

25 μm

B

Nat

ive

Dec

ell

Silver Reticulin

50 μm 25 μm

50 μm 25 μm

*ii

i

C

Nat

ive

Dec

ell

Alcian Blue

100 μm

100 μm

D

BMBM

EN

Dec

ell

TEM

INT

*

500 nm

iii

iv

250 μm

C

Supplementary Figure 2 | MSCs in culture and suspension. (A) Brightfield image of MSCs in culture, demonstratingfibroblast-like morphology. (B) Fluorescence image of MSCs cultured on coverslips (Grace Bio-labs), labeled with Qdot,and stained with DAPI. (C) MSCs labeled with Qdot and stained with DAPI shown following trypsinization and suspension.Cells display a distinctly round morphology.

aA B

Qdot

2 μm

Qdot 2 μm 50 μm

MSC

ii

Supplementary Figure 3 | MSC delivery control. (A) Control MSCs treatedwith benzonase (an endonuclease), labeled with CSFE, and delivered into thelung. (i) Non-viable cells maintain a rounded morphology (similar to cells insuspension) and are observed unattached and floating in alveoli. Lack of DAPIstaining observed in cells treated with benzonase (arrows). (ii) Aggregatedcluster of cells observed in the lung. Cells fail to attach or migrate followingdelivery.

aA

50 μm 50 μm

i

500 μm

Supplementary Figure 4 | Cell attachment and integration following delivery. (A) Fluorescence imaging with brightfieldoverlay of delivered MSCs into human lungs; cells demonstrate morphologic changes adapting to alveolar geography withevidence of migration into the interstitium (arrows). (B) MSCs delivered into human lungs, co-localized with integrin β-1 . (C)AECs delivered into human lungs, location of delivery and relation to alveolar architecture. Alveoli marked by dotted white lines.(D) Collagen IV immunostaining highlighting retention of a key extracellular matrix and direct proximity to attached exogenouslydelivered AECs.

COL IV AECDAPI MERGE

A

B

i ii iii

25 μmMSC

MSC

Integrin β-1

25 μm

C

MSCs

AECs

D

xx μm

iii

AEC

AEC

50 μm

25 μm

MSC

Integrin β-1

100 μm

i

ii

25 μm