Colony-forming cells reduced the lung injury induced by cardiopulmonary bypass

Cardiopulmonary bypass (CPB) results in severe lung injury via inflammation and endothelial injury. The aim of this study was to evaluate the effect of endothelial colony-forming cells (ECFCs) on lung injury in rats subjected to CPB. Thirty-two rats were randomized into the sham, CPB, CPB/ECFC and CPB/ECFC/L-NIO groups. The rats in the sham group received anaesthesia, and the rats in the other groups received CPB. The rats also received PBS, ECFCs and L-NIO-pretreated ECFCs. After 24 hours of CPB, pulmonary capillary permeability, including the PaO 2 /FiO 2 ratio, protein levels in bronchoalveolar lavage fluid (BALF) and lung tissue wet/dry weight, was evaluated. The cell numbers and cytokines in BALF and peripheral blood were tested. Endothelial injury, lung histological injury and apoptosis were assessed. The oxidative stress response and apoptosis-related proteins were analysed.


3
During cardiopulmonary bypass (CPB) in cardiac surgery, the lung blood supply is significantly decreased and then recovers after CPB, which causes pulmonary dysfunction [1]. Postoperative lung injury after CPB is a rare but severe complication that prolongs the duration of mechanical ventilation and the hospital stay and even increases mortality [2]. The morbidity of post-CPB lung injury is 0.4-0.6%, but the mortality is approximately 15-41.5% [3]. CPB-induced lung injury is associated with systemic inflammation induced by the introduction of blood elements to artificial circuits [2] and lung ischaemia/reperfusion injury [4], and activated inflammatory cells ultimately contribute to alveolar inflammation [5,6]. Clinical and experimental studies have focused on a treatment for lung injury after CPB, but there is no ideal strategy for application in clinical work [7,8].
As the outgrowth of endothelial progenitor cells, endothelial colony forming cells (ECFCs) have high proliferative potency[9] and anti-inflammatory effects [10]. ECFCs reduced ischemic injury via the high proliferative ability, differentiation and promotion of revascularization [11] and ventilator-induced lung injury via anti-inflammatory effects [12].
ECFCs have been shown to protect against the renal reperfusion injury by secreting exosomes [13]. Considering the key role of inflammation in lung injury induced by CPB and the anti-inflammatory effects of ECFCs, we hypothesized that ECFCs can ameliorate lung injury after CPB. In this study, we established a rat CPB model to observe the effect of ECFCs on CPB-related lung injury. In Vitro Part

Isolation of ECFCs
We isolated and cultured ECFCs according to a previous study [12]. First, we collected peripheral blood and isolated mononuclear cells using density-gradient centrifugation with Ficoll-Plaque Plus (Amersham Pharmacia Biotech, Uppsala, Sweden). The mononuclear cells were cultured with endothelial growth medium-2 (containing 2% foetal bovine serum). The mononuclear cells were cultured in six-well plates, which were coated with human fibronectin at 37 °C for 21 days. After 21 days, the adherent cells were harvested for further characterization.

Characterization Of ECFCs
The cells were identified according to the results of our previous study [12]. Approximately 24 × 10 4 cells/well were incubated with fluorescein isothiocyanate (FITC)-conjugated Ulex europaeus agglutinin-1 (50 µg/ml) (UEA-1, Sigma-Aldrich, Saint Louis, USA) and DiI-acetyllow-density lipoprotein (LDL) (30 µg/ml) (Invitrogen, Carlsbad, USA). After incubation with UEA and LDL, the mononuclear cells were examined using fluorescence confocal microscopy. The mononuclear cells with dual-positive staining for UEA-1 and acetyl-LDL were defined as endothelial progenitor cells. The cells were also identified with staining for vascular endothelial growth factor receptor (VEGFR) 2 (Abcam, Cambridge, UK) and CD34 (Santa Cruz Biotechnology, Santa Cruz, USA) using a fluorescence microscope. The mononuclear cells with double-positive staining for VEGFR-2 and CD-34 were also identified as endothelial progenitor cells. Based on these results, the cells were further analysed with FITC-labelled CD14 and PE-labelled CD45 antibodies using flow cytometry.
The endothelial progenitor cells with double negative staining of CD14 and CD45 were 5 identified as ECFCs [14]. For analysis of the mechanism of ECFCs in lung injury, ECFCs were preincubated with N5-(1-iminoethyl)-l-ornithine (L-NIO, 10 µM, Santa Cruz Biotechnology) to observe the function of the ECFCs [15].

The Expression Of eNOS In ECFCs
The ECFCs treated with or without L-NIO were harvested, and the total protein of the ECFCs was extracted. The protein expression in ECFCs was detected by Western blots to investigate the effect of L-NIO on the expression of ECFCs.

In Vivo Part
Rat CPB model Thirty-two male Sprague-Dawley rats (400-450 g), obtained from the animal centre of the Second Affiliated Hospital Harbin Medical University, were randomized into 4 groups: the sham group, CPB group, EPC group and EPC/L group. The rats in the sham group only received anaesthesia and tracheal intubation. The cells were pre-treated with L-NIO.
Briefly, the rats were anaesthetized with 3% pentobarbital sodium (30 mg/kg) intraperitoneally. After anaesthesia, all rats were intubated and ventilated (Model 683, Harvard Apparatus, Boston, USA). The respiratory parameters were a tidal volume (Vt) of 10 ml/kg and a respiratory rate (RR) of 50 breaths/min. The fraction of inspired oxygen (FiO 2 ) and positive end-expiratory pressure (PEEP) were set at 50% and 2 cmH 2 O, respectively, and the inspiratory expiratory ratio was 1:1.
After heparinization (500 IU/kg heparin), under local analgesia, 18 G and 16 G catheters were inserted into the right carotid artery and right femoral vein, respectively, to inflow and outflow the blood. Moreover, a 22 G catheter was inserted into the right femoral artery to monitor and analyse the blood sample. The CPB circuit was constructed by a 20 ml venous reservoir, roller pump (Cole Parmer instrument company, Chicago, USA) and membrane oxygenator (MeicroPort, Dongguan, Guangdong, China). Before CPB, the circuit was primed with 0.2 ml heparin, 11 ml of hydroxyethyl starch solution and 0.5 ml 7% sodium bicarbonate solution [16]. During the experiment, the rectal temperature was monitored and maintained within 36-38 ℃ by a heat blanket. The flow rate was gradually adjusted to 100 ml/kg body weight/min and maintained for 60 min [17]. During CPB, the mean arterial pressure was maintained within the range of 60 to 80 mmHg using the continuous injection of adrenaline. The anesthesia was maintained with 3% pentobarbital sodium (10 mg/kg) and rocuronium (0.6 mg/kg) for a 1-hour interval. After 60 min of CPB, the outflow cannula was withdrawn, and the right femoral vein was ligated. The remaining priming solution was continuously infused when the haemodynamics were stable, and the inflow catheter was withdrawn. Immediately after withdrawal of the catheter, the rats in the sham and CPB groups were intravenously injected with 1 ml of PBS, and the rats in the ECFC and ECFC/L groups were intravenously injected with ECFCs or ECFCs pretreated with L-NIO (approximately 10 6 cells in 1 ml of PBS) [15]. To prevent infection, 2000 U/kg 7 penicillin was administered, and incisions were sutured. All rats were extubated when they recovered spontaneous breathing for 24 hours. All rats were sacrificed with an overdose of anaesthetics at 24 hours after ventilation [18]. In this study, we enrolled 10, 9 and 9 rats in the CPB, EPC and EPC/L groups, respectively, to achieve 8 rats in each group.

ECFCs And Alveolar-capillary Permeability
The arterial blood was analysed pre-CPB and at 24-hour after CPB using a Bayer Rapidlab 348 (Bayer Diagnostics, Germany). The PaO 2 /FiO 2 ratio was calculated to evaluate the effect of ECFCs on lung gas exchange function.
Moreover, part of the lung tissue from the right upper lung lobes was harvested. The lung tissues were weighed and dried at 60 °C for 48 hours and then weighed again. The wet/dry weight (W/D) was calculated to observe the effect of ECFCs on alveolar-capillary permeability. Moreover, the protein levels in BALF were also tested.

Histopathologic Injury Evaluation
The lung tissue from the right lower lobe was collected to estimate histological changes.
Lung tissue fixed with 4% paraformaldehyde was embedded in paraffin. The lung tissue was cut into 4-µm sections and stained with haematoxylin and eosin. Two independent pathologists were blinded and employed to evaluate lung histological injury with light microscopy.

ECFCs And Local And Systemic Inflammation
The right bronchi were blocked using an artery clamp. Sterile saline (15 ml/kg) at 4 °C was injected into the left lung via the left bronchi and was withdrawn 5 times. After 5 withdrawals, the bronchoalveolar lavage fluid (BALF) was collected and centrifuged at 4 °C and 1,000 g for 15 min, and then the supernatant was collected and stored at -80 °C. The peripheral blood was collected pre-CPB and at 24 hours after CPB. The blood was centrifuged at 4 °C and 1,500 g for 10 min, and the serum was collected and stored at -80 °C. The cytokines TNF-α, IL-1β, IL-6, and IL-10 were detected in the BALF and serum with the corresponding ELISA kits (Wuhan Boster Bio-Engineering Limited Company, Wuhan, China).
Moreover, the number of neutrophils and the levels of elastase in BALF were also detected.

Tracking Of ECFCs In Lung Tissue
To observe the distribution of ECFCs in lung tissue, approximately 1 × 10 6 ECFCs (with or without pre-treated L-NIO) labelled with acetyl-LDL were injected into rats of the ECFC and ECFC/L groups. After 24 hours of CPB, the lung tissue was harvested, and ECFC tracking was performed by fluorescence microscopy. A slice of lung tissue was prepared according to the histological analysis method. The pulmonary tissue slices were deparaffinized and 9 stained with 4,6-diamidino-2-phenylindole (DAPI) to stain the cell nuclei. The ECFCs in lung tissues were visualized by fluorescence confocal microscopy at a wavelength of 555 nm (acetyl-LDL).

Apoptosis Assay
Apoptosis in the lung tissue was investigated by TUNEL staining with an Apoptosis Assay kit (Roche, Mannheim, Germany). Briefly, the lung tissue slices were immersed in proteinase K at 37 °C for 30 min. The slices were washed twice with PBS. Then, the slides were incubated in the TUNEL reaction mixture (TdT and fluorochrome-conjugated dUTP) for 60 min in a dark chamber at 37 °C. After washing twice, the slides were further incubated with 1 µg/ml 4,6-diamidino-2-phenylindole for 30 min.
The slides were covered with 0.3% H 2 O 2 to inhibit endogenous peroxidase activity, incubated with extra-avidin peroxidase and then immersed in diaminobenzidine solution.
The nuclei that were stained brown were judged as apoptotic cells. In this study, apoptosis of the endothelium and epithelium was identified by two pathologists who analysed histological injury. The apoptosis index was calculated by the ratio of positive apoptotic cells to total cells in a random field from all slides.

Western Blot
First, the protein was extracted, and the protein levels were calculated with the Bradford assay. An equivalent protein volume of every sample was injected into the gel. After electrophoresis, the protein was transferred onto a polyvinylidene fluoride membrane. The membrane was blocked with 5% milk for 30 min and incubated with primary antibody

Statistical analysis
The primary outcome of this study is the PaO 2 /FiO 2 after 24 hours of CPB. In the preliminary study of 5 rats, the PaO 2 /FiO 2 at 24 hours post-CPB was 240 ± 33. The sample size was calculated using PASS 11. Eight rats were needed in each group to detect an increase of 30 in the PaO 2 /FiO 2 with a power of 0.9 and α of 0.05. All the data were normally distributed and are presented as the mean (SD). The data were analysed by oneway analysis of variance and an unpaired t test. All data were analysed using IBM SPSS Statistics 19.0 (SPSS, Chicago, IL, USA). A two-tailed p-value of < 0.05 was considered statistically significant.
To identify the sub-type of EPCs, the cells were analysed for the expression of CD14 and CD45 using flow cytometry (Fig. 1F). All the cells were CD14 − /CD45 − (Fig. 1G to H). These data indicated that the mononuclear cells were late outgrowth ECFCs [19][20][21].

ECFC Proliferation Ability
Compared with that of the normal ECFCs, the viability of the ECFCs that received the L-NIO treatment was significantly decreased (95.6 ± 7.9 vs 80.0 ± 5.43) (P < 0.001). We also found that L-NIO significantly reduced the expression of eNOS in the ECFCs (4.3 ± 0.8 vs 1.2 ± 0.3) (P < 0.001).

Detection Of ECFCs In The Lung Tissue 11
The rats in the ECFC and ECFC/L groups received an injection of ECFCs with or without L-NIO pre-treatment. No ECFCs were detected in the sham and CPB groups ( Fig. 2A and B).
The number of ECFCs in the lung tissue from rats in the ECFC and ECFC/L groups was calculated under a fluorescence microscope (Fig. 2C and D). The number of ECFCs in the ECFC/L group was significantly smaller than that in the ECFC group (7.3% ±2.1 vs. 13.7% ±3.5%, p < 0.05).

ECFC Reduced Histological Injury Induced By CPB
Compared to the sham group, we found typical pathological changes in the CPB group, including lung edema, bleeding, infiltration of inflammatory cells, and damaged alveoli.
Compared with that in the CPB group, the lung injury score was significantly reduced in the EPC group. However, the protective effect of EPCs on lung injury was reduced by the L-NIO (Fig. 3).
ECFCs Improved The Alveolar-capillary Permeability After CPD Compared with the sham group, PaO 2 /FiO 2 , the lung tissue W/D ratio and the concentration of protein in BALF were markedly deteriorated by CPB. After stimulation with CPB, PaO 2 /FiO 2 was increased, but the protein levels and W/D ratio were decreased by EPCs compared with those of the CPB group. Compared with those in the ECFC group, the improvements in PaO 2 /FiO 2 , protein levels and W/D weight ratios were significantly mitigated in the ECFC /L group (Fig. 4).

ECFCs Inhibited Local And Systemic Inflammation After CPB
Compared with those in the sham group, the cytokine levels and the number of cells were significantly increased in rats that received CPB. Compared to the CPB group, the ECFC group exhibited significantly reduced concentrations of TNF-α, IL-1β and IL-6 but elevated the levels of IL-10. The ECFCs also decreased the number of neutrophils and neutrophil elastase in the BALF (Fig. 5). Moreover, the expression of phosphorylated NF-kB and MLC was also inhibited by ECFCs (Fig. 6).
Second, pro-inflammatory factors in the serum were also reduced by ECFCs, but the antiinflammatory factor IL-10 was upregulated by the ECFCs. Compared with the ECFC group, the regulatory effect of ECFCs on inflammatory factors and proteins was partly reversed by L-NIO (Fig. 7).
ECFCs attenuated apoptosis of the endothelium and epithelium after CPB In the sham group, few apoptotic cells were detected. After CPB, many apoptotic endothelial and epithelial cells were observed in the lung tissue. Compared with that in the CPB group, the number of apoptotic cells was significantly reduced in the ECFC group.
However, the number of apoptotic cells in the ECFC/L group was significantly increased compared with that in the ECFC group (Fig. 8). We also found that Bax, Bcl-2 and Gelsolin levels were significantly increased in the rats that received CPB compared with those in the sham group. Compared to those in the CPB group, Bax, Gelsolin and cleaved caspase-3 levels were downregulated, but Bcl-2 was upregulated by the ECFCs. Compared with that in the ECFC group, the regulatory effect of ECFCs on apoptosis was reduced by L-NIO ( Fig. 6).

Discussion
In this study, the ECFCs ameliorated lung injury, improved alveolar-capillary permeability and gas exchange function, reduced local and systemic inflammation, and inhibited apoptosis induced by CPB.
During CPB, lung ischaemia and the introduction of an artificial circuit into the blood resulted in severe local and systemic inflammation[6], which led to prolonged mechanical ventilation, a prolonged stay in the ICU, and even respiratory failure and increased 13 mortality. ECFCs can reduce the ventilator-induced lung injury in rats with ARDS via antiinflammatory effects [12] and protect against renal reperfusion injury by secreting exosomes [13]. Therefore, in this study, we administrated intravenous ECFCs to observe the effect of EPCs on lung injury after CPB. Moreover, to avoid the effect of haemodilution and pressure of the roller pump on ECFCs, we injected the ECFCs after withdrawal of CPB.
In this study, we found that the ECFCs significantly improved gas exchange function and mitigated histological changes [22]. These results indicated that ECFCs can ameliorate lung injury and that ECFCs may be an alternative therapy for patients who undergo cardiac surgery combined with CPB. As previous studies indicated, inflammation, the oxidative response and apoptosis contributed to the lung injury induced by CPB. In this study, we investigated the mechanism of ECFCs in lung injury after CPB based on inflammation, the oxidative response and apoptosis.
Many studies suggested that the imbalance of inflammation plays a key role in the pathogenesis of lung injury after CPB [3]. During CPB, activation of NF-kB promoted the release of chemokines, such as MCP-1 and ICAM-1. Under the chemoattraction of MCP-1 and ICAM-1, neutrophils migrated into the lung and became activated [23,24]. The activated inflammatory cells released pro-inflammatory factors, including TNF-α, IL-1β, IL-8, elastase and MMP-9, which aggravated local inflammation [25]. These injure-related factors not only activated direct damage to the endothelium but also induced the migration of other inflammatory cells. In this study, the results indicated that EPCs mitigated local and systemic inflammation after CPB, which was consistent with previous studies [12,15]. This anti-inflammatory effect of ECFC may occur via the inhibitory effect of ECFCs on chemokines (MCP-1 and ICAM-1) [13] and regulation of the immune response [26]. The regulatory effect of ECFCs on inflammation induced by CPB was mainly attributed to the inhibition of NF-kB and MLC activation [12,15,27]. Moreover, the ability of ECFCs to induce anti-inflammatory IL-10 also played an important role. IL-10 opposed the injurious effect of TNF-α, IL-1β, and IL-6 and reduced the recruitment of inflammatory cells [28]. The anti-inflammatory effects of ECFC not only depended on the inhibitory effect of ECFCs on NF-kB but also the regulatory effect of ECFC on MLC [15,29]. During inflammation, MLC was activated and phosphorylated after endothelial injury [30].
Phosphorylated MLC can damage the contractile elements of the endothelium and lead to injury of the endothelium and lung edema [31,32]. In this study, ECFC treatment improved the pulmonary endothelial barrier and ameliorated pulmonary edema, and this result was consistent with our previous study [15]. The reduction of phosphorylated MLC may be another protective effect of ECFCs on the endothelium.
After CPB, apoptosis also played a pivotal role in lung injury [33]. Both inflammatory factors (TNF-α) and reactive oxygen species induced by lung ischaemia lead to cell apoptosis [34,35]. The apoptotic endothelium and epithelium deteriorated pulmonary function and increased capillary permeability. In this study, the ECFCs significantly reduced cell apoptosis after CPB. During apoptosis, Bax and Bcl-2 play a pivotal role. Bax is a pro-apoptosis protein that sends the apoptosis signal and promotes the activation of caspase-3 to produce cleaved caspase-3, which cuts the DNA and results in cell apoptosis.
In contrast to Bax, bcl-2 is an anti-apoptosis protein and inhibits the role of Bax. The ratio of bax to bcl-2 usually determines the survival or apoptosis of cells [36]. In this study, the ECFCs significantly reduced the expression of Bax and cleaved caspase-3 but increased the expression of Bcl-2. Moreover, the ability of ECFCs to reduce apoptosis was also attributed to the ability of ECFCs to decrease TNF-α expression. The inhibitory effect of ECFCs on apoptosis also contributed to the protective effect of ECFCs against lung injury [37]. eNOS has been suggested to play a key role in bioactivation of early endothelial 15 progenitor cells [38,39]. In our previous study, we also demonstrated that the eNOS inhibitor significantly decreased the infiltration of endothelial progenitor cells in transplanted lung tissue by interfering with the expression of eNOS [15]. In this study, we administrated the NOS inhibitor L-NIO to ECFCs to test whether eNOS plays a crucial role in the late endothelial progenitor cells. The results of this study indicated that treatment with L-NIO significantly decreased the number of ECFCs in the lung tissue and partly reduced the protective effect of ECFCs against lung injury after CPB. This result was consistent with previous studies.

Limitation
There are some limitations in this study. The first is that there is no uniform consensus regarding unique and specific markers of ECFCs[40], although some articles suggested that the method used in this study can identify ECFCs [19][20][21]. Specific markers of ECFCs are still needed. The second is that the ECFCs were harvested after 21 days of culture.
This duration is too long for applications in some patients receiving CPB. However, in clinical work, there are several patients with severe preoperative organ dysfunction who need long-term treatments. Therefore, ECFCs may benefit these patients. Moreover, the effects of ECFCs in rat studies may not be analogous to those in humans after CPB.

Conclusion
The results of this study suggested that ECFCs can attenuate lung injury, improve

Availability of data and materials
All data generated or analysed during this study are included in this published article