SEARCH

SEARCH BY CITATION

Keywords:

  • Actin cytoskeleton;
  • assessment;
  • ex vivo assessment;
  • experimental transplantation;
  • graft function;
  • ischemia/reperfusion injury;
  • lung transplantation;
  • MAP kinase;
  • nitric oxide;
  • non-heart-beating donors;
  • nuclear factor-kappa B (NF-κB);
  • pulmonary vascular physiology;
  • rat;
  • TNF-alpha

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Lungs from non-heart-beating donors (NHBDs) would enhance the donor pool. Ex vivo perfusion and ventilation of NHBD lungs allows functional assessment and treatment. Ventilation of rat NHBD lungs with nitric oxide (NO) during ischemia, ex vivo perfusion and after transplant reduced ischemia-reperfusion injury (IRI) and improved lung function posttransplant. One hour after death, Sprague-Dawley rats were ventilated for another hour with either 60% O2 or 60% O2/40 ppm NO. Lungs were then flushed with 20-mL cold Perfadex, stored cold for 1 h, perfused in an ex vivo circuit with Steen solution and warmed to 37°C, ventilated 15 min, perfusion-cooled to 20°C, then flushed with cold Perfadex and stored cold. The left lung was transplanted and ventilated separately. Recipients were sacrificed after 1 h. NO-ventilation was associated with significantly reduced wet:dry weight ratio in the ex vivo circuit, better oxygenation, reduced pulmonary vascular resistance, increased lung tissue levels of cGMP, maintained endothelial NOS eNOS, and reduced increases in tumor necrosis factor alpha (TNF-α) and inducible nitric oxide synthase (iNOS). NO-ventilation had no effect on MAP kinases or NF-κB activation. NO administration to NHBDs before and after lung retrieval may improve function of lungs from NHBDs.


Abbreviations: 
cGMP

cyclic guanosine monophosphate

ERK

extracellular signal-regulated kinases

HMVECs

human pulmonary microvascular endothelial cells

HUVECs

human umbilical vein endothelial cells

IRI

ischemia-reperfusion injury

LTX

lung transplantation

MAPK

mitogen-activated protein kinase

NHBD

non-heart-beating donor

NF-κB

nuclear kappa factor B

NO

nitric oxide

NOS

nitric oxide synthase

eNOS

endothelial NOS

iNOS

inducible NOS

nNOS

neuronal NOS

p38

p38-MAPK

PA

pulmonary artery

PVR

pulmonary vascular resistance

SAPK/JNKs

stress-activated/c-Jun NH2-terminal kinases

TNF-α

tumor necrosis factor alpha

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Lung transplantation (LTX) palliates patients with end-stage lung diseases, but is constrained by an inadequate number of suitable lungs from conventional organ donors. We and others have demonstrated the feasibility of LTX from Maastricht category I non-heart-beating donors (NHBDs) (1) in animal models (2–6). In humans, variable warm ischemic times and other premorbid events may influence function of lungs retrieved from NHBDs after transplantation. Thus, there is a pressing need for a reliable method to predict function of lungs retrieved from NHBDs before LTX. Steen developed a method to perfuse and ventilate porcine lungs ex vivo to evaluate lung function before transplant (7), and used this to assess lungs retrieved from a NHBD prior to transplant (8). We and others developed similar ex vivo circuits to evaluate human lungs in (9, 10) and Steen has reported using his ex vivo circuit to ‘resuscitate’ lungs thought unsuitable for transplant (11).

Lungs from NHBDs experience a mandatory period of warm ischemia in the deceased donor before retrieval. This is a fundamental difference between ischemia-reperfusion injury (IRI) of lungs retrieved from conventional donors and lungs retrieved from NHBDs. IRI in the lung is manifest in two phases (12): an early phase, characterized by pulmonary edema, related to endothelial injury with alterations in cytoskeleton and permeability (13,14), confounded by altered alveolar fluid resorption; (15) and later generation of cellular adhesion molecules (CAMs) and cytokines probably due to upregulation of genes in the lung parenchyma and entrapped leukocytes (16,17). We studied the first phase of lung IRI and demonstrated reduced filtration coefficient (Kfc) and wet:dry weight ratio (W/D) in rat lungs retrieved from NHBDs reperfused with the nitric oxide (NO) donor nitroglycerine (NTG) (18) and in lungs retrieved from NHBDs treated with NO before retrieval in the NHBD, after retrieval in the circuit, or both (19).

We developed a model to rewarm, perfuse and ventilate rat lungs retrieved from NHBDs before LTX (20). This model resembles our human lung ex vivo perfusion/ventilation system. This study aimed to determine if ventilation of NHBD lungs with NO during warm ischemia, ex vivo perfusion, and perioperatively would reduce IRI and improve function posttransplant of lungs retrieved from NHBDs. We demonstrate a dramatic reduction of edema development in the ex vivo circuit and improved lung function following transplant of NO-ventilated lungs from NHBDs, perhaps related to NO-mediated inhibition of microvascular endothelial cell cytoskeletal abnormalities that occur during ischemia.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Donor lung retrieval and the ex vivo circuit

Specific details of donor lung retrieval and the ex vivo perfusion/ventilation circuit have been outlined previously (20). Male Sprague-Dawley rats (350–450 gm) were anesthetized with intraperitoneal pentobarbital sodium (7.5 mg/100 g; Abbott Laboratories, Chicago, IL). After tracheotomy with a 14-gauge catheter, laparotomy exposed the liver for intrahepatic injection of heparin (600 U; Elkins-Sinn, Cherry Hill, NJ). Donor rats were sacrificed with intrahepatic pentobarbital sodium (15 mg/100 gm). One hour after cardiac arrest, dead rats were ventilated for another hour with 60% O2 alone (O2-vent, n = 6) or 40 ppm NO in 60% O2 (NO-vent, n = 6) with a tidal volume of 0.75 mL/100 gm, rate 60/min, and positive PEEP 3 cm H2O. Through a median sternotomy, the right ventricle was cannulated with p60 tubing, and the left atrial appendage was incised to vent the pulmonary circulation. The lungs were flushed with 20 mL cold Perfadex (Vitrolife, Kungsbacka, Sweden) and stored at 4°C for 1 h. The donor heart-lung blocks were removed from cold storage, perfused in an ex vivo circuit with 75–90 mL Steen solution (Vitrolife) through the pulmonary artery (PA) using a Minipuls three peristaltic pump (Gilson Medical Electronics, Middleton, WI). Circuit perfusate temperature was manipulated by circulating water through a Plexiglas-jacketed venous reservoir with a pump, positioned in a cooler containing either warm (38°C) or ice cold water. PA and airway pressure (AP) were continuously monitored with transducers attached to National Instruments (Austin, TX) hardware connected to a computer with continuous real-time display. Perfusion flow rate was adjusted to maintain PA pressure below 20 mm Hg. When the circuit temperature reached 37°C, the lungs were mechanically ventilated (Harvard rodent respirator model 681, Harvard Apparatus Co, Millis, MA) with alveolar gas (5% CO2, 20% O2, 75% N2) for O2-vent and 40 ppm NO in alveolar gas for the NO-vent group at 60 breaths/min, tidal volume 0.75 mL/100 gm. Perfusate pH was continuously monitored with an Acumen pH probe (Fisher Scientific, Pittsburgh, PA) in the venous reservoir. The pH was maintained near 7.40 by adding dilute HCl or NaHCO3 as necessary. Ventilation and perfusion continued for 15 min at 37 °C to simulate ex vivo gas exchange assessment (as performed in the human circuit). The lungs were then perfusion-cooled to 20°C, flushed again with cold Perfadex and stored cold for 2.5 h before left LTX into a recipient. The unused donor right lung was partitioned, assessed for W/D ratio, flash frozen and stored at −80°C.

Rat lung transplantation technique

Left single LTX was performed using a modified cuff technique (20). The trachea was retrieved with the donor left lung, and the right mainstem bronchus was ligated, so the allograft was ventilated separately from the native right lung in the anesthetized recipient (21) with right/left tidal volume ratio of 2:1. Recipients were anesthetized, intubated, and ventilated with two Harvard rodent ventilators (Model 683) that delivered 60% O2 (O2-vent) alone or 40 ppm NO in 60% O2 (NO-vent) to both lungs with a total tidal volume of 0.75 mL/100 gm. rate 60/min, positive PEEP 3 cm H2O.

Lung function parameters

Transonic flow probes (Transonic Systems, Inc., Ithaca, NY) were placed around the recipient main PA and the left PA to measure total and left PA flow as a measure of relative vascular resistance of the graft. Left lung AP was measured and recorded to reflect dynamic compliance. Following 1 h of reperfusion after transplant, arterial blood gases were measured from a blood sample drawn from the left (transplanted lung) pulmonary vein (PV) and the aorta with an i-STAT device (Abbott Point of Care, East Windsor, NJ). The animals were sacrificed by cardiectomy, and portions of both recipient lungs were retrieved and partitioned. The apical portion was used for W/D, and remaining portions placed in 10% buffered formalin and flash frozen in liquid nitrogen and stored at −80°C.

W/D

At the completion of the experiment, the apical portion of non-transplanted donor right lung (after perfusion in the ex vivo circuit), the transplanted left lung and native right lung was excised and immediately weighed. It was dried in a 60°C oven for 48 h and reweighed, and the ratio was calculated. Fresh lungs were taken immediately after anesthesia from three rats, partitioned and flash frozen in liquid nitrogen and stored at −80°C as normal controls.

Cyclic guanosine monophosphate (cGMP) determination

cGMP was analyzed in lung tissue with the Biotrak enzyme-linked immunoassay (ELISA) system from Amersham Biosciences Corp (Code RPN 226, Buckinghamshire, UK) after extraction according to manufacturer's instructions. Data were corrected for water content by calculating W/D ratio; results were expressed as femtomoles of cGMP per milligram of lung dry weight.

Real-time reverse transcription-polymerase chain reaction (PCR)

Lung messenger RNA (mRNA) levels for IL-6, IL-1β, TNF-α, ICAM-1, eNOS, nNOS and iNOS were assessed by SYBR Green 2-step, real-time reverse transcription-polymerase chain reaction (RT-PCR). Total RNA and protein were extracted from lung samples with a TRIzol reagent (Sigma). One microgram of total RNA from each sample was used for reverse transcription with a High Capacity cDNA Reverse Transcription Kit (PE Applied Biosystems, Foster City, CA) to generate first-strand complementary DNA. The PCR mixture was prepared with Power SYBR Green PCR Master Mix (PE Applied Biosystems). Thermal cycling conditions were 10 min at 95°C to activate the Amplitaq Gold DNA polymerase, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 min on a 7500 Real-Time PCR System (PE Applied Biosystems). Expression of each gene was normalized to the mRNA content of β-actin and calculated relative to fresh samples. Specific oligonucleotide primers were purchased from Invitrogen according to coding regions of published sequences of rat genes. Primers are shown in Table 1.

Table 1.  Primers used in real-time RT-PCR
GenePrimer sequences 5′[RIGHTWARDS ARROW]3′Gene accession no.PCR product size
IL-6Forward: caagagacttccagccagtt Reverse: ccattaggagagcattggaagNM_012589510 bp
IL-1βForward: gtggcagctacctatgtctt Reverse: gccatctttaggaagacacgM98820596 bp
(TNF)-αForward: gagaagttcccaaatgggct Reverse: atctgctggtaccaccagttNM_012675203 bp
ICAM-1Forward: ggagatcacattcacggtgc Reverse: cacacttcacagttacttggtNM_012967500 bp
nNOSForward: agaaattggcagaggccgtc Reverse: tgaatcggaccttgtagctcNM_052799445 bp
eNOSForward: cactgctagaggtgctggaa Reverse: aggacttgtccaaacactccNM_021838492 bp
iNOSForward: cagaggttggaggccttgt Reverse: ctgagagtcatggagccgctNM_012611434 bp
β-actinForward: catggatgacgatatcgctg Reverse: ggatggctacgtacatggctNM_031144408 bp

Immunoblotting analyses

Proteins were separated by SDS/PAGE (4–12% gel) and 20 μg transferred to polyvinylidene difluoride membranes. Blots were incubated with anti-p44/42 MAP Kinase polyclonal (1/9000 dilution; Cell Signaling), anti-SAPK/JNK monoclonal (1/6000 dilution; Cell Signaling), anti-p38 mitogen-activated protein kinase (MAPK) monoclonal (1/9000 dilution; BD transduction) or anti-β actin monoclonal (1/500 000 dilution; Sigma) antibodies, followed by horseradish peroxidase-conjugated anti-mouse IgG or anti-rabbit IgG (Promega) and developed using SuperSignal West Pico Chemiluminescent Substrate (Pierce). Membranes were re blotted with anti-Phospho-p44/42 MAPK monoclonal (1/5000 dilution; Cell Signaling), anti-Phospho-SAPK/JNK (Thr183/Tyr185) polyclonal (1/1000 dilution; Cell Signaling), anti-Phospho-p38 MAPK monoclonal (1/3000 dilution; BD transduction), or anti-IκB-alpha polyclonal (1/1000 dilution; Cell Signaling) antibodies, after stripping from the same membranes.

p65 immunostaining

Five micron sections were deparaffinized and underwent staining for p65 antibody. Immunohistochemical staining was performed in the Anatomical Pathology Translational Core Lab at UNC Hospitals. Heat induced antigen retrieval was conducted at pH 6.0 for 30 min in DakoCytomation Target Retrieval Solution (S1699). Specimens were incubated with 1:200 rabbit polyclonal antibody against nuclear kappa factor B (NF-κB) p65 (sc-109, Santa Cruz Biotechnology, Inc., Santa Cruz, CA) at room temperature for 1 h. VECTASTAIN Elite ABC Kit (Rabbit) was used for detection according to the manufacturer's instructions. Sections were evaluated by a masked pathologist who graded nuclear localization as absent, mild (1+), or strong (2+). Sections were photographed and images analyzed by Aperio software to quantify nuclear localization.

Cell culture model of cold ischemia

We developed a model of simulated cold lung IRI with human umbilical vein endothelial cells (HUVECs) and showed that replacing cold media with Perfadex resulted in rapid alteration in actin cytoskeleton associated with gap formation in the endothelial monolayer (22). To determine the impact of exposure to 40 ppm NO on endothelial behavior, human pulmonary microvascular endothelial cells (HMVECs) were grown to confluence in nine P35 dishes. Six dishes were placed inside a sealed Plexiglas chamber and ventilated at 1 l/h with or without 40 ppm NO in 60% O2 and 5% CO2 (n = 3 each) for 1 h at room temperature to simulate ventilation of the NHBD. Then the media was replaced with cold Perfadex (Vitrolife) buffered with Tham (Abbott Pharmaceuticals, Chicago, IL) to pH 7.2. The chamber was placed in a cooler on ice and ventilated with or without 40 ppm NO in 60% O2/5% CO2 for another hour to mimic cold ischemia. Three dishes served as controls and were left in an incubator at 37°C. Cells were fixed with 4% paraformaldehyde, and stained with Alexa Fluor 568 phalloidin (Invitrogen, OR) to visualize F-actin cytoskeleton. For each dish, photographs were taken of 3 random adjacent fields at 20×. Monolayer gap surface area was quantified using Metamorph software (Universal Imaging Corp, Downingtown, PA).

Statistics

Calculation of means, standard errors, t-tests and ANOVA were performed with Excel 2003 (Microsoft, Redmond, WA). All values are reported as the mean ± standard error of the mean (SEM). Differences were considered significant when p < 0.05.

Animal studies

This study was approved by the Institutional Animal Care and Use Committee of the University of North Carolina at Chapel Hill. All animals received humane care in accordance with the Guide for the Care and Use of Laboratory Animals (National Academy Press, 1996).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Lung function

Parameters of lung function are depicted in Figure 1. After reperfusion in the circuit, NHBD lungs ventilated with NO developed significantly less edema than lungs ventilated with oxygen (Figure 1A). After LTX, graft lung W/D in the NO-vent group was lower than in the O2-vent group but the difference was not significant. Oxygenation in NO-ventilated lungs was significantly better (Figure 1B). In recipients of O2-vent NHBD lungs, pAO2/FiO2 in the graft was lower than in the aorta, suggesting that right lung oxygenation was better than in the transplanted lung. In NO-ventilated animals receiving lungs from NO-vent NHBDs, oxygenation in the transplanted lung was the same as the native right lung. We expressed flow to the graft lung as a ratio to total flow. Lungs treated with NO had much higher blood flow than controls (Figure 1C) although W/D was only slightly lower (Figure 1A). Despite more flow to the NO-ventilated graft, mean AP in the NO-vent group was lower than in the O2-vent group but the difference was not significant (p = 0.08) (Figure 1D).

image

Figure 1. Wet to dry weight ratios (W/D) and hemodynamic data. A: After perfusion in the circuit, donor lungs treated with NO (NO-vent) developed significantly less edema than O2-vent lungs (‡ p < 0.003). After LTX, graft lung W/D in the NO-vent group was lower than in the O2-vent group but the difference was not significant. B: Oxygenation in NO-ventilated lungs was significantly better († p < 0.001). In O2-ventilated lungs, graft PaO2/FiO2 was lower than aorta PaO2/FiO2 suggesting that oxygenation in the right lung was better than in the transplanted lung. In NO-ventilated animals, oxygenation in the transplanted lung was the same as the native right lung. C: Lungs treated with NO had much higher blood flow than O2-vent lungs (Φ p < 0.005). We expressed flow to the graft lung (L) as a ratio to total flow LPA = (LPA flow rate/Donor weight)/(PA flow rate/Recipient weight). D: The mean airway pressure (AP) in the NO-vent group was lower than that in the O2-vent group but the difference was not significant (Δ p = 0.08), mean ± SEM, n = 6 per group.

Download figure to PowerPoint

cGMP

cGMP levels in lung tissue were higher in lungs ventilated with NO either in the circuit or after transplantation (Figure 2). cGMP of NO-ventilated lungs in the circuit was much higher than NO-ventilated lungs after transplant because hemoglobin breaks down cGMP (23).

image

Figure 2. Lung tissue cyclic guanosine monophosphate (cGMP). cGMP levels increased when nitric oxide (NO) was administered either in the circuit (‡ p < 0.001) or after transplantation († p < 0.01). mean ± SEM, n = 6 per group.

Download figure to PowerPoint

RT-PCR

NO-ventilation of NHBD lungs dramatically reduced tumor necrosis factor-α (TNF-α) mRNA 1 h after LTX (Figure 3A). NO-ventilation attenuated increased inducible NOS (iNOS) mRNA, but the difference was not significant (Figure 3B). Compared with fresh lungs, endothelial NOS (eNOS) mRNA decreased both after reperfusion in the ex vivo circuit and after transplant. Decreased eNOS mRNA was significantly impeded by administration of NO after reperfusion in the circuit, and after transplant (Figure 3C). There were no differences between the two groups for nNOS, ICAM-1, IL-6 and IL-1β mRNA after transplantation (results not shown).

image

Figure 3. Changes in TNF-α, iNOS and eNOS mRNA after ex vivo reperfusion and transplantation. A: TNF-α. In the control group, TNF-α expression increased dramatically in the ex vivo circuit and after transplant. This upregulation was significantly attenuated by NO-ventilation in the circuit and after LTX (‡p = 0.01). B: iNOS. Ex vivo reperfusion had no effect on iNOS expression. NO-ventilation attenuated the increased iNOS expression after LTX, but the difference was not significant. C: eNOS. Compared with fresh lung, eNOS expression decreased both in the ex vivo circuit and after transplant. This decreased eNOS expression was impeded by the administration of NO (†p = 0.015 in the circuit, Φp = 0.02 after transplant).

Download figure to PowerPoint

MAPKs and IκBα

We assessed MAPK and NF-κB activation after LTX. Despite better lung function, reduced vascular resistance and dramatically reduced TNF-α mRNA, NO-ventilation had no impact on activation of MAPKs or NF-κB activation 1 h after LTX (Figures 4 and 5). Qualitative assessment of p65 localization by a masked pathologist agreed with the Aperio(tm) analysis depicted in Figure 5. All post-LTX specimens showed strong p65 nuclear localization 1 h after transplant, but this was absent or mild in fresh lung Controls.

image

Figure 4. Western blots of phospho-p44/42 MAPK, phospho-SAPK/JNK (Thr183/Tyr185) phospho-p38 MAPK, IκB-α and actin after LTX. MAPKs were activated and IκB-α degraded 1 h after LTX, irrespective of whether lungs were ventilated with NO.

Download figure to PowerPoint

image

Figure 5. Immunostaining for p65 shows nuclear localization after LTX. There were no differences between NO-vent and O2-vent lungs. Both a masked observer and Aperio software quantified nuclear localization and found no difference between NO-vent and O2-vent lungs 1 h after LTX (n = 3 lungs in each group) († p < 0.01, ANOVA).

Download figure to PowerPoint

Response of HMVECs to cold ‘Ischemia’

Replacement of warm cell culture media with cold Perfadex resulted in rearrangement of the actin cytoskeleton associated with formation of gaps in the confluent monolayer (Figure 6). The dishes ventilated with 40 ppm NO showed virtually no gap formation.

image

Figure 6. Exposing HMVECs to 1 h simulated cold ischemia was associated with significant gaps in the monolayer that were prevented by ventilation with 40 ppm NO. A. control HMVECs before cold ischemia. B. Representative field of HMVECs after 1 h cold ischemia (warm cell culture media replaced with cold Perfadex, Plexiglas boxes ventilated with 60% O2, 1 l/min). Actin cytoskeleton disassembles and gaps form in the previously confluent monolayer (white arrows). C. HMVECS ventilated with 40 ppm NO in 60% O2 show peripheral orientation of actin stress fibers but very little gap formation (white arrows). D. Gap surface area quantified by Metamorph software. Experiments performed in triplicate, three separate contiguous fields photographed at 40× for analysis (nine photos at each time point) (‡ p < 0.001, ANOVA).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Retrieving lungs from Maastricht Category 1 NHBDs might be an effective way to address the critical shortage of lungs for transplantation. Because of this shortage, strict listing criteria are espoused (24), denying thousands of end-stage lung disease patients an opportunity for transplant, and many patients die waiting for LTX. Lungs retrieved from NHBDs may need to be assessed before transplantation because of the variable warm ischemic times and other premorbid events (8, 25). Ex vivo perfusion provides an opportunity for both lung assessment and treatment. Keshavjee's group recently showed gene transfer during normothermic ex vivo perfusion (26).

In our earlier study, NO-ventilation of NHBD lungs in the donor or in the perfusion circuit reduced Kfc and W/D (19). NO was most effective when it was provided both to the NHBD and in the circuit with reperfusion. Thus, in this initial study, we used NO as a therapy prior to lung retrieval, in the ex vivo circuit, and after LTX to determine efficacy in the setting of lung retrieval and transplant from NHBDs. These interventions could easily be performed in the clinical setting.

Lung function was dramatically improved by NO-ventilation. In the ex vivo circuit, W/D of NO-ventilated NHBD lungs was almost normal despite 2 h of ischemia, while NHBD lungs ventilated after death with 60% oxygen became edematous. Following transplant, NO-ventilated lungs had almost normal vascular resistance (normal left:total PA flow in rats is 30% without vascular cuffs), and oxygenation in the graft was the same as in the aorta, i.e. the same as the right lung. Despite the increased blood flow to the graft (> twice that of untreated lungs), W/D and AP were lower in NO-vent lungs after LTX, although these differences were not significant.

Despite better lung function, equivalent MAPK and NF-κB activation occurred after LTX from both NO-treated and untreated lungs. Presumably, this would result in considerable gene transcription in the later phase of IRI. Although equivalent MAPK and NF-κB activation occurred, TNF-α upregulation was significantly blunted in NO-treated lungs both in the ex vivo circuit and after transplant, suggesting that these transcription factors are not responsible for the dramatic TNF-α upregulation. We did not investigate the source of TNF-α mRNA, but alveolar macrophages (AMs) are a well-known source, and have been implicated in lung IRI (27,28). Also, NO has been shown to reduce TNF-α production by AMs (29,30). We expected to see less NF-κB activation in NO-treated recipients because other studies have documented less myeloperoxidase (MPO) activity (a surrogate for PMN accumulation) in lungs when IRI was treated with NO (31). Perhaps increased MPO is not related to NF-κB activation. Similarly, we expected to see reduced ICAM-1 mRNA in NO-treated lungs, but ICAM-1 mRNA was unchanged, irrespective of NO-treatment. In unpublished studies from our lab, ICAM-1 gene expression was not increased in rat lungs 30 min after transplant, but was notably increased 3 h after transplant.

NO-treatment resulted in significantly increased concentration of lung cyclic GMP. NO stimulates soluble guanylate cyclase (32), which produces cGMP. Cyclic GMP regulates three classes of effector proteins: cGMP-dependent protein kinases, cGMP-gated ion channel protein kinases, and phosphodiesterases, mediating protein phosphorylation, cation influx, and cyclic nucleotide catabolism (33). NO is a major controller of cGMP in vascular smooth muscle (34). Endogenous NO production due to L-arginine conversion to citrulline, along with extrinsic organic nitrates, result in increased cGMP and relaxation in vascular smooth muscle (35).

The NO pathway is pivotal in regulation of microcirculation flow and vascular permeability associated with IRI (36,37). NO donors and stimulation of endogenous NO production reduce pulmonary injury due to IRI (38,39), reduce PMN influx after pulmonary IRI, assessed by myeloperoxidase (MPO) assay (31,40). NO inhibits neutrophil activation, aggregation and migration (37,41), platelet activation (42), and reduces cytokine-induced CAM expression (43). NO may inhibit expression of cytokines in lymphocytes, eosinophils, monocytes, and lung macrophages, including cytokines critical for the development of inflammatory processes like IL-1β, TNF-α, IL-6 or IFN- γ (29,44). NO increases heme oxygenase (HO-1) expression (45) and reduces IRI (46). We did not examine HO-1 mRNA levels; HO-1 is upregulated by a large number of cytokines and pathways (47,48) so it would be difficult to ascribe any increase to NO alone.

In tissues, NO is generated enzymatically by synthases (NOS). There are three NOS isoforms: NOS I (nNOS)—the neuronal form; NOS II—inducible nitric oxide synthase (iNOS), present in various cell types and increased by a variety of inflammatory stimuli; and NOSIII (or eNOS)—constitutive enzyme discovered in the endothelium (49). Low concentrations of NO produced by eNOS and nNOS inhibit adhesion molecule expression, cytokine and chemokine synthesis and leukocyte adhesion and transmigration (50), and may have antioxidative actions (51). Generation of endogenous endothelial cGMP or administration of membrane permeable cGMP analogs attenuates the endothelial barrier dysfunction caused by reactive oxygen species (52,53). However, large amounts of NO generated by iNOS can be toxic and pro-inflammatory (54), and may promote peroxynitrite radical formation (51). Induction of iNOS can be initiated by inflammatory cytokines IFN-γ, TNF-α or IL-1β (55).

In this study, we investigated mRNA expressions of the three NOS isoforms together with TNF-α and some other inflammatory cytokines. There were no differences in iNOS expression between two groups in the ex vivo circuit. iNOS expression increased after transplant but the difference was not significant; this increase was attenuated by NO administration. In another rat LTX study, iNOS mRNA increased and eNOS mRNA expression decreased after ischemia-reperfusion (56). In our O2-vent group, eNOS expression decreased both in the circuit and after LTX. NO-ventilation prevented reduced eNOS expression especially after transplant; eNOS mRNA levels of the NO-vent group after transplant were almost the same as in fresh lung samples.

Our cell culture studies shed light on why treatment of NHBD lungs with NO resulted in reduced edema development in the ex vivo circuit. NO is diminished in endothelial cells following ischemia and reperfusion (57,58). Exposing HMVECs to NO at 40 ppm prevented gap formation that occurred due to simulated cold ischemia. If the same phenotype occurred in vivo, maintaining integrity of the endothelial barrier would result in less early edema with reperfusion. To show this in vivo would require electron microscopy, but subtle differences in gaps of this small magnitude may be difficult to demonstrate. Our immunostained paraffin-embedded lung sections were also examined after H&E staining, but only evidence of scattered interstitial edema was apparent in both groups (data not shown). Recently, we showed that early lung edema due to IRI was mediated largely through toll-like receptor 4 (TLR4) on pulmonary parenchymal cells (not myeloid-derived AMs). In a model of simulated warm IRI, gap formation that occurred in HMVECs was prevented by a TLR4 inhibitor that also prevented edema in the warm lung in vivo IRI model (59). These data imply that TLR4-mediated changes in cytoskeleton associated with altered HMVEC permeability may be mediated through a NO mechanism. Kolluru et al showed that NO prevented actin cytoskeletal changes due to hypoxia in HUVECs (60). More recently, actin cytoskeletal changes in plant root cells were shown to be mediated by NO (61).

Our animal model differs from the clinical situation in some respects. We have some experience retrieving lungs from Maastricht Category 1 NHBDs. One hour of ‘hands off’ time is almost necessary before consent for ventilation and lung retrieval can be obtained, and transportation to a facility where ventilation is possible. Ventilation with NO is only available at large medical centers, but most Category 1 NHBDs will come from urban centers based on population demographics. For logistic reasons, we abbreviated cold storage time to 1 h in this study. Fortunately, Perfadex preservation has allowed for longer safe cold storage. In the clinical arena, longer storage would be necessary to perform serologies and identify a prospective recipient. Alternately, longer ex vivo perfusion may be an option (62). This study did not determine when NO-ventilation needs to be initiated. Takashima et al reported that NO-ventilation of canine recipients of NHBD lungs retrieved from nonventilated donors 3 h after death led to very acceptable gas exchange function, even when NO was administered for only 1 h after the onset of reperfusion (63).

Retrieving lungs from NHBDs might be an effective way to reduce the critical shortage of lungs for transplantation. Earlier we showed that ventilation of NHBDs with oxygen delayed loss of viability and ultrastructural integrity (64,65), and increased the safe ‘time window’ for transplantation of lungs retrieved after death to 4 h (66), although LTX was associated with edema and impaired oxygenation. The current study suggests that other simple interventions may help reduce IRI when lungs are retrieved from NHBDs. Administration of NO to NHBDs before and after lung retrieval reduced edema when lungs were perfused in the ex vivo circuit, and reduced PVR and improved oxygenation after transplant. NO administration was associated with dramatically reduced TNF-α mRNA, and normalized iNOS and eNOS mRNA levels due to IRI. Additional studies with longer-term follow-up are warranted to confirm the potential impact of this strategy in the clinical arena.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

The authors wish to acknowledge the technical assistance of Kimberlie Burns (UNC Cystic Fibrosis Center) for histology preparation, and Nana Nikolaishvili Feinberg and Dr. C. Ryan Miller (UNC Hospitals Anatomical Pathology Translational Core Lab) for assistance with immunostaining and analysis, and Dr. William Funkhouser (UNC Department of Pathology and Lab Medicine) for visual assessment, and the editorial assistance of Margaret Alford Cloud. Perfadex and Steen solution were provided by Vitrolife. This work was supported by grants from the Cystic Fibrosis Foundation (EGAN06P0) and the UNC Lung Transplant Research Fund.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References