Comparative evaluation of PEGylated bovine hemoglobin as a perfusate in rat ex vivo lung perfusion

Abstract

Background

Ex vivo lung perfusion (EVLP) is pivotal for assessing and preserving marginal donor lungs. While red-blood-cell (RBC)-based perfusates enhance oxygen delivery, they increase pulmonary vascular resistance, thereby potentially exacerbating perfusion-induced injury. This study compared a novel acellular oxygen carrier, PEGylated bovine hemoglobin (PEG-bHb), against RBC-based and acellular perfusates in a rat EVLP-transplantation model.

Methods

A total of 49 rats were included in this study. Lungs underwent 4 h EVLP with acellular, RBC or PEG-bHb solutions, followed by left-lung transplantation and 2 h reperfusion; unperfused lungs served as controls. Respiratory mechanics, gas exchange, inflammatory cytokines, oxidative-stress markers, tight junction integrity (Occludin), and histology were evaluated. Bulk RNA-sequencing, qRT-PCR, and immunofluorescence for endoplasmic reticulum (ER) stress markers (ATF-6, ATF-4, CHOP and ubiquitin) were employed to explore underlying mechanisms.

Results

PEG-bHb demonstrated lower pulmonary vascular resistance and higher dynamic compliance than the RBC group. Inflammatory cytokine (IL-6, IL-1β, TNF-α) and Oxidative markers (MDA, HIF-1α) in lung tissue were reduced compared to acellular and RBC groups. Occludin expression indicated superior alveolar–capillary integrity versus control and acellular groups and parity with RBC group. Post-transplantation, the PEG-bHb group exhibited the least histological injury and the highest IL-10 expression, indicating attenuated early graft damage. Transcriptomic analysis suggested concurrent regulation of endoplasmic-reticulum stress and cytokine pathways; immunofluorescence demonstrated reduced ATF-6, increased ATF-4, and stable CHOP expression.

Conclusion

PEG-bHb reduced early post-transplant injury by preserving lung compliance, sustaining barrier integrity, and mitigating inflammatory stress during EVLP. It represents a promising candidate for clinical EVLP application.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12931-026-03710-6.

Introduction

Ex vivo lung perfusion (EVLP) has become an integral technique in the clinical management of marginal donor lungs, enabling preservation, evaluation, and functional recovery prior to transplantation [1–3]. Compared to static cold storage, EVLP has demonstrated superior efficacy in routine preservation of standard donor lungs, significantly reducing the incidence of severe primary graft dysfunction [4]. Central to the EVLP process is the utilization of specific perfusion solutions to sustain graft viability. Currently, perfusion solutions are broadly categorized into red blood cell (RBC)-based solutions and acellular solutions.

RBC-based perfusates provide enhanced oxygen delivery but are associated with logistical constraints and potential risks, including hemolysis-related microcirculatory injury during ex vivo perfusion [5]. To address these limitations, artificial hemoglobin has been explored as alternative perfusion agents due to their stability and scalability.

Artificial hemoglobin serves as an efficient oxygen carrier, its advantages include scalability for large-scale production, extended storage stability, and ease of transportation. Extensive studies have already explored its application in perfusion of other solid organs, including the liver and kidney [6, 7]. Among these, PEGylated bovine hemoglobin (PEG-bHb) has garnered particular attention. As a conjugate of bovine hemoglobin and polyethylene glycol, PEG-bHb demonstrates efficient oxygen transport, reduced immunogenicity, and limited nitric oxide scavenging, which may help maintain vascular tone during perfusion [8, 9]. This component can now be mass-produced and stored for one year under frozen conditions at -28 °C.

This study was designed to systematically evaluate PEG-bHb as a perfusate in a rat EVLP-transplantation model, in comparison with standard acellular and RBC-based solutions. Lung grafts were assessed at the end of EVLP for physiological function, inflammatory responses, and barrier integrity. Following transplantation, early reperfusion injury was further evaluated.

Methods

Animals and study design

Male inbred Sprague-Dawley rats, aged 7–8 weeks and weighing 250–350 g, were procured from GemPharmatech Co. (Jiangsu, China). To minimize variability, the weight difference between donor and recipient rats was maintained within 25 g. Rats were acclimated for two days and fasted for 24 h prior to surgery with free access to water.

A total of 49 rats were included in this study (Fig. 1). Unperfused native lungs served as baseline controls (n = 4). An additional cohort underwent direct left lung transplantation without EVLP, followed by 2 h of reperfusion, serving as transplant controls (n = 4). The remaining allocated lungs underwent 4 h of EVLP sequentially utilizing acellular (n = 13), RBC-based (n = 14), or PEG-bHb (n = 14) solutions. Within each EVLP cohort, a subset of lungs (n = 6 per group) was exclusively perfused without subsequent transplantation to evaluate end-ischemic states. The remaining lungs in each group (acellular, n = 7; RBC, n = 8; PEG-bHb, n = 8) underwent 4 h EVLP followed by left lung transplantation and a 2 h reperfusion phase to assess early graft injury. Sub-allocations for downstream histological, immunohistochemical, transcriptomic, and immunofluorescence analyses were predetermined as detailed in Fig. 1. All initial samples were successfully analyzed without experimental exclusions.

Fig. 1.

Experimental design and sample allocation. A total of 49 rats were used. Control group (n = 8): lungs were procured and either sampled immediately (Control-Baseline, n = 4) or transplanted directly without EVLP (Control-Transplant, n = 4) followed by 2 h reperfusion. Acellular EVLP, n = 13 (of which n = 6 perfused only and n = 7 transplanted & reperfused 2 h); RBC EVLP, n = 14 (n = 6 perfused only, n = 8 transplanted & reperfused 2 h); PEG-bHb EVLP, n = 14 (n = 6 perfused only, n = 8 transplanted & reperfused 2 h). RNA-seq: n = 3 per transplanted group; Biochemical assays (post-EVLP): n = 4 per group; Histology (2 h reperfusion): n = 4 per transplanted group. No animals were excluded after allocation. RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin, SPC: surfactant protein C, IL -10: interleukin − 10, ER: Endoplasmic Reticulum

All procedures, including surgery and EVLP, were performed at the First Affiliated Hospital of Guangzhou Medical University, and experiments were performed under a project license (No. 2021035) granted by the Experimental Animal Ethics Committee of Guangzhou Medical University, in compliance with Chinese guidelines for the care and use of animals.

Lung procurement

Anesthesia was induced using ≤ 1.0 mL of inhaled isoflurane in an induction chamber. Once respiratory rate decreased to approximately 1 breath per second and complete muscle relaxation was achieved, rats were placed on an operating table. Anesthesia was maintained by intraperitoneal administration of 1% ketamine (10 mL/kg). Tracheotomy and intubation were performed, followed by mechanical ventilation (100% oxygen, 15 cmH₂O inspiratory pressure, 70 breaths/min).

A midline incision was made from the xiphoid process to the mandible after sterilizing the thoracoabdominal region. The inferior vena cava (IVC) was exposed, and 500 IU of heparin sodium was administered via the IVC. Once systemic heparinization was confirmed, the IVC was transected, and blood was withdrawn until no further flow was observed. The rat was subjected to 2 min of warm ischemia to mimic donation after circulatory death organ procurement. Median sternotomy was performed to expose the heart and lungs. Both atrial appendages were excised, the vena cava ligated, and a 2-mm incision was made in the right ventricular outflow tract. A cannula was inserted into the pulmonary artery and secured with silk sutures. The lungs were flushed with 15 mL of low-potassium dextran (LPD) solution under 20 cmH₂O perfusion pressure. The trachea was transected, and the heart-lung block was carefully excised.

Perfusion solutions

Perfusion solutions were prepared as described in Table 1 [10]. RBC-based perfusion solutions were created by adding red blood cells to the acellular base solution to achieve a hematocrit of 15%. The PEG-bHb solution was formulated by adding PEGylated bovine hemoglobin (Zhongbaomu Co., China) to the acellular solution to achieve a hemoglobin concentration of 30 g/L. Methylprednisolone (0.25 mg/mL) and cefazolin (0.3 mg/mL) were added to the perfusion solutions immediately before EVLP. All solutions were filtered to eliminate particulates and potential bacterial contamination and stored for a maximum of 48 h to prevent bacterial growth.

Table 1.

 Compositions of the reagents

	LPD solution	Acellular	RBC	PEG-bHb
Dextran 40 (g/L)	50	5	5	5
Modified gelatin (g/L)	-	40	40	40
Human albumin (g/L)	-	-
Bovine serum albumin (g/L)	-	20	20	20
Na+ (mmol/L)	138	148	148	148
K+ (mmol/L)	6	4	4	4
Cl− (mmol/L)	142	103	103	103
Mg2+ (mmol/L)	0.8	1	1	1
Ca2+ (mmol/L)	-	1	1	1
Glucose (mmol/L)	5.5	11	11	11
Buffer (mmol/L)	Tham 1	HCO3− 24	HCO3− 24	HCO3− 24
	HPO42− 0.13
	H2PO4− 0.46
Added RBC			Hct 15%
PEG-bHb				HGB 30 g/L

LPD low-potassium dextran, RBC red blood cell, PEG-bHb PEGylated bovine hemoglobin

EVLP protocol

The excised heart-lung block was connected to the perfusion system. EVLP was initiated at a controlled temperature of 21 ± 1 °C and maintained for 4 h. Perfusion flow was gradually increased over the first 1.5 h to the target flow, corresponding to 20% of the estimated cardiac output of the rat. Calculated cardiac output was defined as 75 mL/min per 250 g body weight [10]. Mechanical ventilation was initiated immediately with the following settings: inspiratory pressure, 15 cmH₂O; frequency, 40 breaths/min; positive end-expiratory pressure (PEEP), 3 cmH₂O. A gas mixture (8% CO₂, 6% O₂, 86% N₂) was supplied to the oxygenator at a flow rate of 25 mL/min for deoxygenation. Every 30 min, recruitment maneuvers were performed with 100% oxygen at an inspiratory pressure of 20 cmH₂O for 5 min. Continuous monitoring included pulmonary artery pressure, peak airway pressure, tidal volume; dynamic lung compliance, and pulmonary vascular resistance (PVR) were calculated as described [11].

Perfusate samples were collected hourly for blood gas analysis, including measurements of PCO₂, PO₂, and lactate levels. After 4 h of EVLP, the donor lungs were flushed with 10 mL of LPD solution.

Left lung transplantation

Following EVLP, the left lung was isolated from the heart-lung block. Custom cuffs (16G, 18G, and 14G catheters) were prepared for the pulmonary vein (PV), pulmonary artery (PA), and main bronchus, respectively, and secured with 9 − 0 nylon sutures. Recipient rats were anesthetized using inhaled isoflurane for induction and maintained with intraperitoneal injection of 1% ketamine (7.5 mL/kg).

The left lung was transplanted into the recipient using a nonsuturing external cuff technique as described [12]. A tracheotomy and left thoracotomy were performed to expose the left pulmonary hilum. Microvascular clamps were applied to occlude the proximal hilum. Small incisions were made in the distal PA, PV, and main bronchus to insert and secure the cuffs. The clamps were then removed to restore perfusion and ventilation.

Inflammatory and oxidative stress marker analysis

Left lung tissue samples collected after 4 h of EVLP were stored at − 80 °C. For analysis, frozen tissues were weighed and homogenized in ice-cold lysis buffer at a standardized tissue-to-buffer ratio. Homogenates were centrifuged, and the supernatants were collected for subsequent assays. Cytokines (Interleukin-6 (IL-6), Interleukin-6 (IL-1β), Tumor Necrosis Factor-Alpha (TNF-α)) and oxidative stress markers (8-Hydroxy-2’-Deoxyguanosine (8-OHdG), Hypoxia-Inducible Factor-1 Alpha (HIF-1α), Malondialdehyde (MDA)) were quantified using multiplex immunoassay kits (Jianglai Bio, China) according to the manufacturer’s instructions. Total protein concentration of each homogenate was measured by BCA assay and all marker concentrations were normalized to total protein and expressed as pg (or ng) per mg protein.

Immunofluorescence staining of lung tissue before reperfusion

Donor lungs fixed in 4% paraformaldehyde and paraffin-embedded after 4 h of EVLP were analyzed for occludin expression and Terminal Deoxynucleotidyl Transferase dUTP Nick End Labeling (TUNEL)-positive apoptotic cells using a double immunofluorescence staining protocol [13]. The TUNEL apoptosis assay kit (Servicebio, China) was used as per the manufacturer’s instructions. Fluorescence images were captured with a CSU-W1 confocal microscope (OLYMPUS, Japan) and analyzed using ImageJ software.

Histological assessment

After 2 h of reperfusion, donor lungs were fixed with 4% paraformaldehyde and paraffin-embedded. Tissue sections were stained with hematoxylin and eosin (H&E) and evaluated using a Leica DM 6000 B microscope (Wetzlar, Germany). All sections were evaluated in a double-blind manner by two independent pathologists following anonymization. Lung injury was scored based on the evaluation of four randomly selected lung fields per mouse. The assessment included the presence of neutrophils in alveolar and interstitial spaces, hyaline membranes, proteinaceous debris filling the airspaces, and alveolar septal thickening. The scoring criteria were based on the guidelines established by the official American Thoracic Society workshop report on features and measurements of experimental acute lung injury in animals [14].

Immunohistochemistry and immunofluorescence of lung tissue after 2-hour of reperfusion

Paraffin-embedded lung sections were deparaffinized, subjected to heat-induced antigen retrieval, and blocked to suppress nonspecific binding. The tissues were incubated overnight at 4 °C with primary antibodies against Activating Transcription Factor 4 (ATF4), C/EBP homologous protein (CHOP), Activating Transcription Factor 6 (ATF6), ubiquitin, surfactant protein C (SPC), and Interleukin-10 (IL-10). SPC and IL-10 expression was evaluated by bright-field microscopy, whereas all other fluorescent signals were captured with a CSU-W1 spinning-disk confocal microscope (Olympus, Japan) and analyzed using ImageJ software.

RNA extraction and library preparation

The left lung lobe was excised 2 h after reperfusion, immediately snap-frozen in liquid nitrogen, and stored at − 80 °C. Total RNA was isolated with TRIzol Reagent (Invitrogen, USA) and its integrity assessed on an Agilent 2100 Bioanalyzer (Agilent Technologies Inc., USA). Poly(A) + RNA was enriched using Oligo(dT) magnetic beads, randomly fragmented in fragmentation buffer containing divalent cations, and subjected to first- and second-strand cDNA synthesis, end repair, and adaptor ligation. cDNA fragments of approximately 370–420 bp were purified with AMPure XP beads (Beckman Coulter, Inc., USA), PCR-amplified, and re-purified with AMPure XP to generate the final libraries. Libraries passing quality control were pooled according to effective molarity and desired sequencing depth and sequenced on an Illumina system.

Bioinformatic analysis

Libraries were preliminarily quantified with a Qubit 2.0 Fluorometer (Invitrogen, USA), diluted to 1.5 ng/µL, and the insert size verified on an Agilent 2100 Bioanalyzer; libraries meeting the expected size were accurately quantified by qPCR, and only those with an effective concentration ≥ 1.5 nM were loaded for sequencing. Clean reads were obtained after adaptor and quality trimming, mapped to the reference genome, and gene-level counts were generated with featureCounts. Differentially expressed genes were identified with DESeq2 (|log₂ fold-change| ≥ 0.585, adjusted p < 0.05), and Gene Ontology enrichment analysis was performed using clusterProfiler v3.8.1.

Expression validation using quantitative polymerase chain reaction

Total RNA was extracted from 30 mg of rat lung tissue using TRIzol, treated with DNase, and 1 µg was reverse-transcribed. Quantitative PCR (7500 Fast, Applied Biosystems) was performed with Universal Blue SYBR Green qPCR Master Mix (Servicebio, G3326) and exon-spanning primers for Lipocalin 2 (Lcn2), Colony Stimulating Factor 3 (Csf3), C-X-C Motif Chemokine Ligand 11 (Cxcl11), Cxcl10, Hspa5, Derl3 with Log2FC > 1 from RNA-seq analysis and reference β-actin. Using Actin expression as a normalization control, expression was analysed by 2^-ΔΔCt and reported as mean ± SEM. The primer sequences are provided in Table S1.

Statistical analysis

Normality was assessed utilizing the Shapiro-Wilk test. Data are presented as mean ± SEM (normal distribution) or median (25th, 75th percentiles) (non-normal distribution). Intra-group longitudinal changes relative to the initial baseline were analyzed using the Friedman test followed by Dunn’s post hoc test. Concurrently, inter-group comparisons at specific individual time points were evaluated using the Kruskal-Wallis test with Dunn’s post hoc test. For single-endpoint comparisons between two independent groups, the Student’s t-test or Mann-Whitney U test was employed as appropriate. A two-sided P < 0.05 was considered statistically significant. Analyses were performed using SPSS 25.0, with graphs generated via GraphPad Prism 9.4.1.

Results

Physiological parameters during EVLP

All donor lungs underwent 4 h of EVLP successfully. Throughout the perfusion course, lung compliance in the PEG-bHb group improved significantly from its initial baseline to the 4-hour endpoint (intra-group progression: 0.165 [0.153, 0.190] vs. 0.225 [0.195, 0.263] mL/cmH₂O, P < 0.01), a temporal trend of elevation not observed within the other groups. At the end of perfusion (4 h), the PEG-bHb group exhibited significantly superior lung compliance compared to the RBC group (inter-group analysis: 0.225 [0.195, 0.263] vs. 0.150 [0.143, 0.200] mL/cmH₂O, P < 0.05; Fig. 2A). In terms of the overall magnitude of change, the PEG-bHb group demonstrated a profoundly greater increase (+ 0.060 [0.035, 0.080] mL/cmH₂O) than the net decrease observed in the RBC group (− 0.025 [− 0.043, − 0.015] mL/cmH₂O, P < 0.01; Fig. 2E).

Fig. 2.

Physiological parameters during ex vivo lung perfusion. Compliance (A), lactate levels (B), pulmonary vascular resistance (C), and oxygenation capacity (D) in the RBC, PEG-bHb, and Acellular groups at indicated time points. Changes (Δ) in dynamic compliance (E), lactate levels (F), and pulmonary vascular resistance (G) from baseline to post-EVLP. RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin. ^##P < 0.01 indicates a significant change from baseline within the PEG-bHb group. ^####P < 0.0001 indicates a significant change in lactate levels over time within all groups. **P < 0.05, *P < 0.01 indicate significant differences between groups at the end of EVLP. RBC: red blood cell; PEG-bHb: PEGylated bovine hemoglobin, EVLP: Ex Vivo Lung Perfusion

Metabolically, lactate concentrations exhibited a progressive, statistically significant intra-group elevation across all study groups during EVLP (P < 0.0001 for all). However, at each evaluated time point, levels differed significantly among the three groups (all P < 0.001), with the RBC group consistently showing the highest median values. By the 4-hour mark, the lactate level in the RBC group reached 3.70 [3.43, 3.90] mmol/L, an elevation substantially more pronounced than those in the acellular (1.10 [0.78, 1.55] mmol/L, P < 0.01) and PEG-bHb groups (1.30 [1.13, 2.20] mmol/L, P < 0.05) (Fig. 2B and F).

No significant temporal changes in PVR were observed within any group during EVLP (P > 0.05). In contrast, inter-group comparisons at each time point revealed significant differences (P < 0.001 for all), with the RBC group showing consistently higher PVR than the PEG-bHb and acellular groups throughout perfusion (Fig. 2C and G). Specifically, at 4 h, PVR in the RBC group peaked at 78.16 [72.50, 89.82] mmHg/min/mL, whereas the PEG-bHb and acellular groups maintained significantly lower resistances of 30.64 [29.89, 52.51] and 39.36 [19.64, 47.31] mmHg/min/mL, respectively (P < 0.05 for both). Finally, no statistically significant inter-group differences in oxygenation capacity were detected at any time point during the procedure (P > 0.05; Fig. 2D).

Inflammatory Cytokines and Oxidative Stress Marker levels in Lung Tissue after 4 h of EVLP

Inflammatory cytokine levels (IL-6, IL-1β, TNF-α) were significantly elevated in the acellular and RBC groups compared to the PEG-bHb group (P < 0.0001, P < 0.05, P < 0.0001, respectively; Fig. 3A-C). The acellular group also showed a higher IL-6 level relative to the RBC group (P < 0.001; Fig. 3A). In terms of oxidative stress, the acellular group had markedly higher MDA and HIF-1α levels than the PEG-bHb group (P < 0.01, P < 0.0001, respectively; Fig. 3E, F). Similarly, the RBC group exhibited increased 8-OHdG and HIF-1α levels compared to the PEG-bHb group (P < 0.01, P < 0.05, respectively; Fig. 3D, F). Furthermore, HIF-1α was significantly higher in the acellular group than in the RBC group (P < 0.01; Fig. 3F).

Fig. 3.

Cytokine and oxidative stress marker levels in lung tissue. Concentrations of IL-6 (A), IL-1β (B), TNF-α (C), 8-OHdG (D), MDA (E), and HIF-1α (F) in lung tissue homogenates. Data are presented as protein-normalized concentrations. IL: interleukin, TNF-α: tumor necrosis factor α, 8-OHdG: 8Hydroxy2′deoxyguanosine, MDA: Malondialdehyde, HIF-1α: Hypoxia Inducible Factor1 Alpha, RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001

Occludin expression and apoptosis post-EVLP

TUNEL staining showed no significant differences in the levels of apoptosis among the three EVLP groups. However, apoptosis levels in the control group were significantly lower than those observed in the EVLP groups (P < 0.01; Fig. 4). Immunofluorescence analysis of Occludin, a tight junction protein, revealed robust expression in the PEG-bHb group, comparable to that in the RBC group, but significantly higher than in the acellular solution and control groups (P < 0.01; Fig. 4).

Fig. 4.

Representative images and quantification of occludin and TUNEL. The objective magnifcation was 20×, Gamma: 0.8, scale bar: 50 μm. RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin, MFI: mean fuorescence intensity. **P < 0.01, ***P < 0.001, ****P < 0.0001

Lung assessment post-reperfusion

Two hours after reperfusion, the PEG-bHb group exhibited uniform tissue texture and less atelectasis upon macroscopic evaluation, while varying degrees of structural damage were observed in the other groups (Fig. 5). H&E staining revealed that lung tissue in the PEG-bHb group was structurally intact, with minimal interstitial edema, reduced hemorrhage, and lower inflammatory cell infiltration compared to the other groups (Fig. 6, S1). Pathological scoring confirmed that the PEG-bHb group exhibited the least tissue damage among all groups (P < 0.0001; Fig. 7A). Immunohistochemical analysis showed that SPC were less expressed in the acellular group, comparable to those in the RBC and PEG-bHb groups (P < 0.0001, P < 0.05, respectively; Figs. 6 and 7B), and IL-10 expression was significantly higher in the PEG-bHb group than that in the acellular and RBC groups (P < 0.001, P < 0.01, respectively; Figs. 6 and 7C). These findings aligned with the macroscopic observations of reduced atelectasis and preserved tissue integrity.

Fig. 5.

Appearance of donor lungs before and after reperfusion in the study groups. Two hours after reperfusion, the PEG-bHb group exhibited superior lung compliance and uniform tissue texture upon macroscopic evaluation, while varying degrees of atelectasis and structural damage were observed in the other three groups. RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin

Fig. 6.

Representative photomicrographs of lung sections from the study groups 2 h after reperfusion. Histological changes are shown by H&E staining; lung compliance and anti-inflammatory response were assessed by immunostaining for SPC and IL-10 (brown), respectively. Sections are shown at x20 magnification, scale bar:50 μm. RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin, SPC: surfactant protein C, IL -10: interleukin − 10

Fig. 7.

Histological Assessment. A Graft injury scores blindly graded after hematoxylin–eosin staining; (B) Optical-density values for surfactant protein C; (C) Optical-density values for interleukin − 10. RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin, SPC: Surfactant protein C, IL-10: Interleukin-10. *P < 0.05, **P < 0.01, ****P < 0.0001

Identification of Differentially Expressed Genes (DEGs) and Gene -Ontology Enrichment Post-Reperfusion

To explore potential mechanisms underlying the observed physiological differences, we performed exploratory bulk RNA-seq on graft tissue collected after reperfusion (n = 3 per group). The differential transcriptome profiles obtained from the three perfusate groups are summarised in the volcano plots (Fig. 8A-C) and heat-maps (Fig. 8). Sixty DEGs were identified between the PEG-bHb and RBC groups (26 up- and 34 down-regulated; Fig. 8A, Table S2), 73 DEGs between the PEG-bHb and Acellular groups (38 up- and 35 down-regulated; Fig. 8B, Table S3), and 23 DEGs between the RBC and Acellular groups (8 up- and 15 down-regulated; Fig. 8C, Table S4). Gene Ontology (GO) biological-process analysis was performed for every comparison (Fig. 9) and identified distinct functional patterns associated with each perfusate.

Fig. 8.

Heatmaps illustrate hierarchical clustering of differentially expressed RNAs (|log₂FC| ≥0.585, adjusted p < 0.05). Volcano plots visualize RNA expression differences between pairwise experimental groups: sixty DEGs were identified between the PEG-bHb and RBC groups (26 up- and 34 down-regulated; A), 73 DEGs between the PEG-bHb and Acellular groups (38 up- and 35 down-regulated; B), and 23 DEGs between the RBC and Acellular groups (8 up- and 15 down-regulated; C). (A) PEG-bHb vs. RBC, (B) PEG-bHb vs. Acellular, and (C) RBC vs. Acellular. RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin

Fig. 9.

Gene Ontology (GO) enrichment analysis of differentially expressed RNAs. A Significantly enriched GO terms for up-regulated RNAs in PEG-bHb vs. RBC (|log₂FC| ≥0.585, adjusted p < 0.05). B Enriched GO terms for up-regulated RNAs in PEG-bHb vs. Acellular. C Enriched GO terms for down-regulated RNAs in PEG-bHb vs. Acellular. D Enriched GO terms for down-regulated RNAs in RBC vs. Acellular. RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin

PEG-bHb vs. RBC

Up-regulated genes were enriched for processes that attenuate endoplasmic-reticulum stress, including “response to unfolded protein”, “negative regulation of Unfolded Protein Response (UPR)”, “endoplasmic reticulum (ER)-associated degradation” and “retrograde protein transport, ER to cytosol” (Fig. 9A). These categories included genes such as Hspa5, Ficd, Derl3, Herpud1, Hsp90b1 and Hyou1. Conversely, down-regulated genes clustered in “response to lipopolysaccharide”, “cytokine-mediated signalling pathway” and “cellular calcium ion homeostasis” (Fig. 9B), and comprised Lcn2, Cxcl10, Csf3, Fgf23, Adora1 and Ptgir.

PEG-bHb vs. Acellular

Enrichment analysis highlighted oxygen-transport and haematopoietic terms such as “erythrocyte development”, “gas transport” and “negative regulation of inflammatory response to antigenic stimulus” (Fig. 9C); leading genes included Hbb-b1, Hba-a2, Alas2, Slc4a1, Cxcr4, Npy and Il1b. The term “negative regulation of inflammatory response to antigenic stimulus” included IL-10 and Npy among its contributing genes.

RBC vs. Acellular

Up-regulated transcripts were enriched for adaptive-immune processes—including “antigen-receptor-mediated signalling”, “B-cell activation” and “adaptive immune response” (Fig. 9D). Genes associated with these categories included Cd79a, Lax1, and Pax5. In contrast, down-regulated genes mapped to “complement activation” and “protein activation cascade”, with Masp1 and Vsig4 as the principal contributors to these complement-associated terms.

Endoplasmic-reticulum stress profile

To determine the influence of each perfusate on endoplasmic-reticulum stress (ERS) signalling, paraffin-embedded lung sections collected 2 h after transplantation were subjected to immunofluorescence staining for ATF-6, ATF-4, CHOP and total protein ubiquitination. Quantitative analysis demonstrated pronounced inter-group differences (n = 4 per group; Fig. 10). The mean fluorescence intensity of ATF-6 was significantly higher in the Acellular and RBC groups than in the PEG-bHb group (P < 0.0001). In contrast, ATF-4 staining was greatest in the PEG-bHb group and differed significantly from both the Acellular and RBC groups (P < 0.0001). No significant differences were detected in CHOP or ubiquitination signals among the three groups (P > 0.05).

Fig. 10.

Immunofluorescent evaluation of endoplasmic-reticulum stress markers in rat lung grafts 2 h after reperfusion. Representative micrographs of ATF-4, CHOP, ATF-6 and total ubiquitin staining, respectively, in lungs perfused with the Acellular, RBC and PEG-bHb solutions. The objective magnifcation was 20×, Gamma: 1, scale bar: 50 μm. RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin, ATF: activating transcription factor, CHOP: C/EBP homologous protein

Validation of mRNA expression using quantitative polymerase chain reaction

We performed quantitative polymerase chain reaction (qRT-PCR) validation on six inflammation- and endoplasmic-reticulum–related genes that were enriched in the GO analysis and exhibited LogFC > 1 between the RBC and PEG-bHb groups. As shown in Fig. 11, Lcn2, Csf3, Cxcl10, and Cxcl11 were significantly downregulated in the RBC group compared to the PEG-bHb group (P < 0.05), whereas Hspa5 and Derl3 were significantly upregulated in the PEG-bHb group (P < 0.05). These qRT-PCR results exhibited expression trends entirely consistent with the RNA-seq findings, providing validation of the transcriptomic data.

Fig. 11.

qRT-PCR validation of inflammation- and ER stress–related gene expression between RBC and PEG-bHb groups. RBC: red blood cell, PEG-bHb: PEGylated bovine hemoglobin, Lcn2: Lipocalin 2, Csf3: Colony Stimulating Factor 3, Cxcl11: C-X-C Motif Chemokine Ligand 11, Cxcl10: C-X-C Motif Chemokine Ligand 10. *P < 0.05

Discussion

This study systematically evaluated the effects of different perfusion solutions on donor lung preservation during EVLP and the mitigation of early reperfusion injury following transplantation. The findings highlight the potential of PEG-bHb as a superior perfusion solution, indicating improved alveolar-capillary barrier integrity and surfactant protein expression compared to acellular solutions. Moreover, PEG-bHb exhibited comparable performance to RBC-based solutions while avoiding their associated drawbacks, such as elevated vascular resistance, inflammation and oxidative stress, leading to better lung compliance during EVLP and reduced reperfusion injury after transplantation.

Comparison with RBC-Based perfusion solutions

RBC-based perfusion solutions enhance oxygen delivery by incorporating red blood cells, thereby supporting tissue metabolism and promoting cellular repair. Additionally, the higher viscosity of RBC solutions likely contributed to increased endothelial shear stress, stimulating the production of vasodilators such as NO and prostacyclin, which enhance perfusion. In this study, the RBC group exhibited favorable expression of alveolar-capillary barrier-associated proteins, comparable to the PEG-bHb group and significantly superior to the acellular and control groups. These results align with previous studies where RBC-based STEEN solutions outperformed acellular solutions in reducing the wet-to-dry weight ratio and improving perfusion outcomes [5].

However, RBC-based solutions have inherent limitations. During ex vivo perfusion, RBCs are susceptible to mechanical damage and hemolysis, releasing heme and free hemoglobin that act as potent inducers of oxidative stress and endothelial dysfunction [15]. Moreover, the increased viscosity of RBC-based perfusates contributes to higher vascular resistance [16], which can cause uneven tissue perfusion. In this study, the significantly higher levels of inflammatory cytokines (IL-6, IL-1β, TNF-α) and oxidative stress markers (8-OHdG, HIF-1α) in the RBC group compared to the PEG-bHb group are consistent with more severe tissue damage. These findings highlight the limitations of RBC-based perfusates in maintaining optimal physiological conditions during perfusion.

Advantages of PEG-bHb perfusion

PEG-bHb appears to offer several distinct advantages over RBC-based solutions. Its small molecular size may facilitate effective oxygen delivery to microcirculatory regions, potentially ensuring adequate oxygenation in poorly perfused areas. Although oxygen partial pressures were similar across the groups, the total oxygen content was markedly higher in the PEG-bHb group compared to the acellular group, addressing the limitations of acellular solutions that lack oxygen carriers [17]. By providing sustained oxygen delivery, PEG-bHb may effectively prevent hypoxia-induced damage to alveolar epithelial cells and maintains the integrity of the alveolar-capillary barrier.

The unique properties of PEG modification further enhance the functionality of PEG-bHb. The PEG chains shield the heme group, reducing NO scavenging and preventing vasoconstriction and excessive vascular resistance. The increased molecular size minimizes extravasation into the vascular wall, enhancing safety during perfusion. PEG modification is thought to reduce the immunogenicity of hemoglobin, which may in turn lower the risk of complement activation and immune rejection post-transplantation. Studies on PEGylated erythrocruorin from Lumbricus terrestris have demonstrated improvements in microcirculation, tissue perfusion, and organ function recovery [18]. Additionally, the use of bovine hemoglobin as the base material for PEG-bHb provides advantages such as high availability, low cost, and structural homology (> 85%) with human hemoglobin, ensuring effective oxygen transport without the need for 2,3-DPG regulation [19, 20].

Beyond its physiological advantages, PEG-bHb also presents a substantial economic benefit. Unlike RBC-based perfusates, which incur costs associated with blood collection, cross-matching, storage, and screening, PEG-bHb can be produced at scale with lower logistical and regulatory demands. This cost-effectiveness may enhance accessibility and feasibility of EVLP, especially in resource-constrained settings.

Early reperfusion injury after transplantation

At two hours post-reperfusion, both the PEG-bHb and RBC groups exhibited better-preserved alveolar architecture and higher SPC expression compared to the acellular group, consistent with the macroscopic observations of reduced atelectasis. Histological analysis, however, revealed a distinct injury gradient: tissue damage was most severe in the acellular group, intermediate in the control and RBC groups (Figure S1), and mildest in the PEG-bHb group, indicating a potential reparative benefit of PEG-bHb.

We hypothesize that the reduced injury in the PEG-bHb group relative to the RBC group is attributable to its ability to lower PVR during EVLP—a parameter recognized as an early predictor of transplantation outcomes [5]. In the same study, it was observed that porcine lungs perfused with RBC-based solution exhibited higher pulmonary artery pressure and pulmonary vascular resistance compared to the acellular solution group. However, due to its superior oxygen-carrying capacity, the graft lung function post-transplantation in the RBC group was better. In the present study, compared with the PEG-bHb group, the RBC group no longer retained this advantage. Instead, its higher pulmonary vascular resistance likely contributed to uneven tissue perfusion, resulting in significantly elevated levels of oxidative stress markers compared to the PEG-bHb group. This, in turn, exacerbated ischemia-reperfusion injury following transplantation.

Transcriptomic features associated with attenuated injury in the PEG-bHb group

To elucidate the mechanisms underlying these inter-group differences, we performed exploratory bulk-lung transcriptomic sequencing. The PEG-bHb group exhibited a transcriptomic profile suggestive of attenuated lung injury, characterized by coordinated modulation of ER stress–related and inflammatory pathways. Compared with the RBC group, genes enriched in the PEG-bHb group participated in multiple layers of negative regulation of ER stress: Up-regulation of Hspa5 (BiP) may dampen unfolded protein response sensitivity, as UPR transducer activity is inversely correlated with available BiP [21]; Derl3 supports ER-associated degradation, preventing sustained UPR activation.

Moreover, as ATF-6 is an early sensor of the canonical UPR, reduced ATF6 expression may suggest diminished activation of the canonical UPR branch, although this interpretation warrants further functional validation. Concomitantly, PEG-bHb lungs displayed increased ATF-4 without a corresponding increase in CHOP, which may indicate a bias toward adaptive stress signaling rather than pro-apoptotic transcriptional programs [22]. ATF-4 has been reported to drive macrophage polarisation toward the M2 phenotype and to promote IL-10 transcription, thereby exerting anti-inflammatory effects [23]. Consistent with this mechanism, we observed elevated IL-10 levels in the PEG-bHb group, paralleling the ATF-4 induction reported here; however, functional blockade studies targeting ATF-4 or IL-10 are still needed to establish causality.

Conversely, transcripts linked to acute inflammatory amplification were down-regulated in the PEG-bHb arm. Lcn2, which transports iron and drives oxidative stress [24], and Cxcl10, a chemokine implicated in immune cell recruitment via CXCR3 signaling, were reduced in the PEG-bHb group, consistent with diminished histological evidence of inflammatory infiltration observed in this group [25]. Relative to the acellular group, IL-10 displayed robust over-expression in the PEG-bHb group, which may contribute to restraint of excessive inflammatory signaling [26].

Given the limited sample size and the exploratory nature of bulk RNA-seq analysis, these transcriptomic findings should be interpreted cautiously. The differential expression results are hypothesis-generating and require validation in larger cohorts and through functional experiments to establish mechanistic causality.

Limitations

This study has several limitations. First, although the findings support PEG-bHb as a promising EVLP perfusate, the underlying mechanisms remain to be fully elucidated; the transcriptomic and immunofluorescence findings are exploratory in nature and require validation in future studies. Second, this study was conducted in a rat model, and the translational relevance to human lungs requires validation in large animal or clinical settings. Third, to minimize confounding variables, assessments were limited to 4 h of EVLP and 2 h post-transplantation [27, 28]; whether PEG-bHb confers protection during the critical 24–72 h window when primary graft dysfunction peaks remains unknown. Fourth, post-transplant physiological parameters were not assessed. Future investigations in large animal or long-term survival models should incorporate such functional measurements to fully validate the translational potential of PEG-bHb. Finally, only a single PEG-bHb concentration (30 g/L) was tested; dose-optimization studies are needed to determine the concentration that best preserves graft viability and outcomes for clinical translation [29].

Conclusion

In summary, PEG-bHb shows promise as an EVLP perfusate, potentially offering advantages in attenuating inflammatory injury, improving pulmonary compliance, and preserving the alveolar–capillary barrier. Lungs in the PEG-bHb group exhibited reduced hypoxic and inflammatory damage, better compliance during EVLP, and milder overall injury during the early post-transplant period. However, further studies are needed to optimize its concentration, validate its efficacy in large-animal and long-term survival transplant models to fully elucidate its therapeutic potential. With continued research and refinement, PEG-bHb may emerge as a valuable tool in EVLP and contribute to improved outcomes in organ transplantation.

Abbreviations

8-OHdG

8-Hydroxy-2’-Deoxyguanosine

ATF4

Activating Transcription Factor 4

ATF6

Activating Transcription Factor 6

CHOP

C/EBP Homologous Protein

Csf3

Colony Stimulating Factor 3

Cxcl10

C-X-C Motif Chemokine Ligand 10

DEGs

Differentially Expressed Genes

ER

Endoplasmic Reticulum

ERS

Endoplasmic-Reticulum Stress

EVLP

Ex Vivo Lung Perfusion

GO

Gene Ontology

H&E

Hematoxylin and Eosin

HIF-1α

Hypoxia-Inducible Factor-1 Alpha

IL-6

Interleukin-6

IL-10

Interleukin-10

IVC

Inferior Vena Cava

Lcn2

Lipocalin 2

LPD

Low-Potassium Dextran

MDA

Malondialdehyde

NO

Nitric Oxide

PEG-bHb

PEGylated Bovine Hemoglobin

PVR

Pulmonary Vascular Resistance

qRT-PCR

Quantitative Reverse Transcription Polymerase Chain Reaction

RBC

Red Blood Cell

SPC

Surfactant Protein C

TNF-α

Tumor Necrosis Factor-Alpha

TUNEL

Terminal Deoxynucleotidyl Transferase dUTP Nick End Labeling

UPR

Unfolded Protein Response

Authors’ contributions

(I) Conception and design: J Zhang, X Zhang, H Yuan; (II) Administrative support: J He; (III) Provision of study materials or patients: J He; (IV) Collection and assembly of data: J Zhang, X Zhang, Jingfeng Ou, Guanyi Wu; (V) Data analysis and interpretation: J Zhang, X Zhang, H Yuan, Jingfeng Ou, Guanyi Wu; (VI) Manuscript writing: All authors; (VII) Final approval of manuscript: All authors.

Funding

None.

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Declarations

Ethics approval and consent to participate

Experiments were performed under a project license (No. 2021035) granted by the Experimental Animal Ethics Committee of Guangzhou Medical University, in compliance with Chinese guidelines for the care and use of animals.

Competing interests

The authors declare no competing interests.