rhCygb reverses bleomycin-induced established pulmonary �brosis in rats

Purpose Recombinant human cytoglobin (rhCygb) has been demonstrated to anti-inammation, anti-oxidation, anti-brosis in liver and kidney in animal disease models. However, the effect of rhCygb on the progression of pulmonary �brosis is still unclear. The aim of this study was to investigate the therapeutic effect of rhCygb in the bleomycin (BLM)-induced pulmonary �brosis rats. Methods We tested whether rhCygb would reverse lung �brosis (at day 28 and 56) in Sprague-Dawley rats treated with bleomycin. Bleomycin (5mg/kg) resulted in �brosis by CT and serum measurement at day 7. Effects of rhCygb treatment on morphological and CT imaging of the lung, as well as serial serum levels of biomarkers with progressive lung �brosis were tested at day 28 and 56.


Background
Idiopathic pulmonary brosis (IPF) is rare but fatal lung disease, which is characterized by chronic alveolar wall and interstitial brosis of lung tissue, causing progressive decline in lung function and respiratory failure. [1]Most patients with IPF may combine with pulmonary hypertension, emphysema or obstructive sleep apnea; the estimated median survival time is 2-5 years following diagnosis. [2]Typical high-resolution computed tomography (HRCT) of IPF displays brotic changes in both of the basal and peripheral lungs, with reticular opacities, honeycomb changes and ground-glass opacities. [3]An O cial American Thoracic Society Workshop Report con rmed that there was a consensus opinion of the intratracheal BLM model as "the best-characterized animal model available for preclinical testing".
Although the molecular pathogenesis remains incompletely elucidated, an updated pathological mechanism is advocated that epithelial injuries and subsequent aberrant repair play key roles in its development and progression.As a result of alveolar epithelial cells damage, they secrete several mediators to stimulate the migration, proliferation of broblasts and myo broblasts, and the deposition of excessive extracellular matrix (ECM), leading to remodeling of a so-called "honeycomb" lungs. [4]tracellular matrix is a highly dynamic macromolecular structure that provides both three-dimensional tissues and function support to organ. [5]In the lung, the ECM is generally composed of brillar proteins, glycoproteins, glycosaminoglycans, and the basement membrane laminin.There are still some variants of ECM constituents which have not been characterized. [6] Yilang et al showed signi cantly higher serum levels of ECM molecules in IPF patients, involving laminin (LN), type IV collagen (IVC), procollagen III N-terminal peptide(PIIINP), and hyaluronic acid (HA), compared with healthy individuals; they pointed out that serum levels of LN, IVC, PIIINP and HA might indicate IPF progression and might be indicators for the severity of IPF. [9]toglobin (Cygb) is a member of the globin family, which was discovered in hepatic stellate cells in 2001. [10]The biological and pathophysiological importance has increased progressively in the past decade.[18] Recombinant human cytoglobin (rhCygb) is a biologically synthesized protein.It has been reported to be an active oxygen scavenger as well as a functional enzyme like NADH oxidase and catalase. [19][22][23] Therefore, in this study, we intended to investigate the e cacy of rhCygb in a lung brosis rat model induced with BLM.

Methods
Male rats (Sprague-Dawley, SD) with an average weight of 180-220 g were purchased from the Guangdong Medical Laboratory Animal Center, China.Care of the animals in this investigation was done according to the guidelines approved by the Chinese Association of Laboratory Animal Care.Rats were housed in groups of 5 per cage under constant temperature (20 ± 2℃) and humidity (70%) with a 12 h light-dark cycle and had free access to food and water.The animals were acclimatized for 1 week prior to the experiments.All animal protocols were approved by the Ethics Committee for Animal Research of Southern Medical University.
To induce pulmonary brosis, rats were anesthetized with 2% pentobarbital (50 mg/kg body weight) and placed on an intubation stand facing upward at an angle of approximately 30-35 by using an elastic string carefully positioned under the animal's front incisors.The tongue was gently pulled out with forceps and the trachea was intubated.BLM-treated rats were intratracheally instilled with 5 mg/kg bleomycin sulfate (Nippon Kayaku, Tokyo, Japan ) solution in 0.9% saline as previously described [24] , and slowly instilled through the catheter into the trachea.The control rats simultaneously received the same volume of 0.9% saline solution without bleomycin.
SD rats were randomly allocated to the following groups: SD rats receiving a single intratracheal instillation of saline solution and sacri ced after 56 days (n = 11); SD rats receiving a single intratracheal instillation of BLM (5 mg/kg) and sacri ced after 56 days (n = 11); SD rats receiving a single intratracheal instillation of BLM (5 mg/kg), and 7 days later, treated with rhCygb (4 mg/kg, daily, by subcutaneous) for 49 days (n = 12); SD rats receiving a single intratracheal instillation of BLM (5 mg/kg), and 7 days later, treated with DXMS (3 mg/kg, daily, by subcutaneous) for 49 days (n = 11); A general schedule of the treatment is shown in Figure 1.The rats were scanned with multi-slice spiral CT at day 7 and 56.Finally, the rats were sacri ced and specimens of lung were harvested and processed for subsequent analyses.Furthermore, the rats were subjected to measurement of biomarkers at day 7, 28 and 56.Finally, rats were sacri ced with a sealed euthanasia device, before that, carbon dioxide was put into the device, so that the rats could enter anesthesia faster with reduced fear and pain.Then specimens of lung were harvested and processed for subsequent analyses.Furthermore, the rats were subjected to measurement of biomarkers at day 7, 28 and 56.Drug rhCygb was produced according to the method described previously. [20]It was dissolved in phosphate buffer saline and administered at 4 mg/kg subcutaneously.Selection of this dose of rhCygb was based on our previous literature. [20]mputed tomography analysis Seven days after BLM treatment, CT scan was performed for assessment of BLM-induced changes in lung density.Rats were scanned at 7 and 56 days postoperatively.The rats were placed in the prone position, under 2% vol.iso urane anesthesia, and then underwent plain CT scan with multi-slice spiral CT (SOMATOM Emotion 16; Siemens, Erlangen, Germany).Plain CT scan conditions were as follows: 80 kV with 500µA, a pitch of 0.7 mm, slice thickness of 0.7 mm and acquisition time of 32 s, 512 × 512 matrix, FOV 9.6 cm, and 1000 projections per scan.Images were reconstructed and analyzed.Mean lung density (MLD) for the entire lung volumes for each rat was calculated using the Somaris/5VB10B software.The MLD was assessed with quantitative determination of the overall lung Houns eld units (Hu). [25]Before scanning, the MLD for air was calibrated as −1000, while the water was 0.

Histological assessment of lung injury
At day 56, rats were sacri ced after the CT scan were performed and the lungs were carefully isolated to histological analysis.The lungs were xed with neutral buffered formalin (10%), dehydrated and embedded in para n, and then sectioned at 5 µm thickness.The slices were stained with hematoxylin and eosin (H&E) and Masson trichrome staining (Sigma-Aldrich, Saint Louis, Unite States), and observed under a microscope (Seepack TX510, Shenzhen, China) for assessing lung injury.Masson trichrome staining was performed to detect left lung brosis.Four elds of each sample were randomly selected and collagen volume fraction (CVF) was assessed by Image-Pro Plus 6.0 to represent the degree of brosis in the left lung.CVF (%) refers to the percentage of the area stained positive for collagen relative to the total area in a eld of view. [26]droxyproline content To obtain a quantitative measure of lung collage at the end of 56 days, hydroxyproline as the major component of collagen was measured in the lung tissues using the Hydroxyproline Assay Kit (Nanjing jiancheng bioengineering institute, Nanjing, China).After euthanizing the rat and harvesting the lung tissue, 30-60 mg (wet weight) of tissue was taken and put into the glass vial.One ml hydrolysate was added to the vial, capped tightly and hydrolyzed at 60℃ for 20 min.The hydrolysate was then centrifuged at 3500 rpm for 10 min according to the instructions of HYP assay kit.Absorbance was measured at 560 nm by a spectrophotometer, and HYP content of lung tissue was calculated according to the manufacture's protocol.
Measurement of serum levels of HA, LN, PIIINP, IVC On day 7, 28 and 56, blood samples were collected from the veil and serum was separated by centrifugation at 4℃ and stored at -20℃ for future use.The levels of HA, LN, PIIINP and IVC in serum were determined using a chemiluminescence immunoassay kit (Shenzhen new industry biomedical engineering Ltd., Shenzhen, China) according to the protocol by automatic chemiluminescence analyzer (Shenzhen new industry biomedical engineering Ltd., Shenzhen, China).

Statistical Analysis
Statistical analysis was performed with the statistical analysis software Graph Pad Prism 7 (GraphPad Software, San Diego, CA ).The experimental results were given as mean ± standard errors of mean.Group comparisons were performed by one-way or two-way ANOVA followed by the Bonferroni post hoc multiple comparison test.Results were considered statistically signi cant at P < 0.05.

Establishment of BLM-induced pulmonary brosis model in rats
CT is a routine, rapid and non-invasive method for diagnosis and characterisation of IPF. [27]To evaluate the BLM-induced rat model at day 7, CT was used to analysis the marked change in lung.Compared to control rats, the lungs of BLM-treated rats indicated parenchymal opacity, with a main features of a peripheral, predominantly basal pattern of coarse reticulation (Figure 2).Moreover, the CT MLD for BLMtreated rats was -(704.78±19.76)Hu (n=11), and for the control rats the MLD was -(401.71±22.29)Hu (n=34) (Table 1).[30] The serum levels of LN, IVC, PIIINP and HA in BLM-treated rats signi cantly increased compared with control rats (Table 1).Together with the CT results, it showed that the IPF rat model was well established.
rhCygb reversed BLM-induced histopathological and brotic changes in the lung After the establishment of IPF rat model, the rats were separated into BLM, rhCygb+ BLM (4 mg/kg) and DXMS+ BLM group (3 mg/kg).In gure 3A, CT scan images at day 56 are shown the differences in brotic and non-brotic lung in BLM-treated and control groups, while rhCygb had improved lung architecture with observable reduced collage deposition and loss of air spaces.Rats received DXMS also had effect on the lungs as compared to that of BLM-treated rats.
Morphologically, compared with the smooth surface of control lung, BLM-treated rats showed the rough, ugly surface of the lung with gray brous nodules.On the other hand, rhCygb obviously improved pulmonary morphology in rats with no brous nodules (Figure 3B).
H&E staining showed normal lung tissue structure and intact alveolar cavity in the control group.In the BLM-treated group, there was signi cant alveolar cavity collapse.The alveolar septum, deposition with hemoglobin, was thickened.The alveolar septum was less thickened in the rhCygb+BLM and DXMS+BLM groups (Figure 3C).
Masson trichrome staining (Figure 3D) revealed the lung tissue of the control group had no obvious broblast foci.In the BLM-treated group, BLM enhanced pulmonary alveolus in ammation and broadened alveolar septa.Moreover, there was a lot of collagen deposition in lung tissue and the pulmonary brosis was severe.In the rhCygb and DXMS treatment groups, the alveolar wall was less thickened and in ammatory cell in ltration decreased.Additionally, there were decreased broblasts in lung tissues.However, microthrombus were found in the bronchus of the DXMS treatment group, as compared to control.The Masson CVF results (Figure 3E) showed that the average percentage of lung brosis in the control group was 1.86 ± 1.51 %.In contrast, rats treated with BLM resulted in a higher degree of lung brosis, which was 40.33 ± 2.08 %.In the rhCygb and DXMS treatment groups, the CVF was 20.81 ± 1.65 % and 30.76 ± 1.57 %, respectively.The comparison between each group was statistically signi cantly different (Control vs. BLM group, P < 0.001; BLM group vs. rhCygb+BLM group, P < 0.001; BLM group vs. DXMS+BLM group, P < 0.001).
Hydroxyproline levels in lung tissue of BLM-treated group were signi cantly higher than that in control group (P < 0.05).In the rhCygb and DXMS treatment, HYP levels were obviously lower than that of BLM group (P < 0.05).Moreover, there was signi cant difference in HYP level between the rhCygb and DXMS groups (P < 0.05) (Figure 3F).This result was consistent with the Masson staining results.
The CT MLD acquired by experiments was shown in Figure 3G.Since most part of the lung is composed by air, the MLD of control lung is tendency towards to -800 Hu.The CT MLD in BLM-treated group was signi cantly higher than that in control group (P < 0.001).In the rhCygb and DXMS treatment, the CT MLD were obviously lower than that of BLM group (P < 0.001).

Effect of rhCygb on HA, LN, P NP and C serum levels
The levels of HA, LN, P NP and C of rats at day 7, 28 and 56 were measured in blood corresponding to the CT analysis in Figure 4.With a longer process of brosis, the HA, LN, P NP and C levels increased from 7 days to 56 days.Serum HA, LN, P NP and C levels in BLM-treated group were approximately 2.14, 1.61, 1.69, and 1.48 folds of those in the controls, respectively.Both rhCygb and DXMS signi cantly reduced HA, LN, P NP and C levels in blood of pulmonary brosis rats (P < 0.001).And at day 56, there was no signi cant difference in HA, LN, P NP levels between rhCygb and control groups except C levels (P > 0.1).

Discussion
The establishment of BLM-induced lung brosis model The model of BLM-induced lung brosis has played a pivotal role in the search for anti brotic agents for treating pulmonary brosis. [31]An O cial American Thoracic Society Workshop Report con rmed that there was a consensus opinion of the intratracheal BLM model as "the best-characterized animal model available for preclinical testing". [32]Nevertheless, sex and age differences in mice might affect the experiment results.Young mice, aged 8-12 weeks, has been reported to undergo spontaneous resolution of BLM-induced pulmonary brosis, while this phenomenon is not obtained in aged mice. [33]Thus, the aged male mice are recommended for experiment research, which may provide a more clinically relevant model of IPF.In our study, adult male SD rats were used in the BLM-induced pulmonary brosis model.
Recently, some researchers point out that therapies are usually administered within < 7 days following BLM exposure, which may represent a stage of in ammation or early brosis; so the therapeutic effects may through prevention of the in ammatory cascade rather than reversal of brosis, and this is in contrast to the clinical situation when brosis has already been established. [34]Thus, more recent studies have begun to explore administration of drugs after 7 days. [35,36]However, those papers were based on the hypothesis that lung brosis was induced 3 or 4 weeks after bleomycin instillation without diagnosing the establishment of lung brosis model.Surprisingly, Choi reported that more than half of bleomycin-instilled mice did not show lung brosis at 3 weeks or even 6 weeks by micro-CT. [37]So it is very essential to develop assessment methods or tools for establishment of IPF animal models.CT is used to evaluate the IPF as a non-invasive technique in humans, with the feature of honeycombing or traction bronchiectasis and a reticular abnormality consistent with brosis present in a basal and peripheral predominance. [38]Until now, with increasing technical possibilities, micro-CT provides highresolution anatomical images of small animals, and it allows repeated measurements, which avoid animal euthanasia. [37,39]Interestingly, in our study, we con rmed that a brosis phase in the BLM-induced brosis model has been well established at 7 days by spiral CT, showing alveolar septal thickening, interstitial brosis and honeycombing.This may be due to the animal breed, gender, dose, and administration affecting the results.Peng reported that bleomycin induced a dose-dependent increase in lung brosis, and signi cant brosis was observed in the groups of mice with higher dose. [40]These observations would support our hypothesis that testing of candidate (rhCygb or DXMS) during the stage of late brosis is likely to translate into a therapeutic bene t in the clinic, evidence that comes from the clinical situation in which treatment is initiated after onset of symptoms and when brosis has already been established.Furthermore, the combination of clinical parameters and biological markers has been studied in order to achieve more accurate results regarding the prognosis of IPF.Finding biomarkers for IPF has been a central challenge for a long time and would aid the ful llment of the need for noninvasion diagnostic of patients with IPF.Yiliang et al investigated serum levels of HA, LN, PIIINP, IVC in 323 patients and 160 healthy controls, and found that serum HA, LN, PIIINP, IVC levels were all signi cantly higher in the patients with IPF than in the control groups and had a signi cant positive association with HRCT score in patients with IPF. [9]However, whether serum levels of HA, LN, PIIINP and IVC could re ect the development and progression of IPF, a serial measurement of the serum levels of HA, LN, PIIINP and IVC at different disease stage is required.In the present study, we tested elevation of HA, LN, PIIINP, IVC serum levels at 7, 28 and 56 days of the same rats in BLM-treated group when the lung brosis obviously occurred.Surprisingly, compared with the progressive serum levels of HA, LN, PIIINP and IVC, the CT MLD of BLM-treated rats at 56 days almost equal to that at 7 days.Hence, we suggest that it is necessary to assess the progress of brosis in the BLM-induced brosis model using CT combined with biomarker measurement.

Reversed effect of rhCygb on BLM-induced lung brosis
Although we have previously reported that rhCygb displays antioxidative, anti-in ammatory, and anti brotic properties in vitro and in animal models of liver brosis, [20][21][22][23] its e cacy against the pulmonary brosis in BLM-induced rats has never been studied.Herein, we test of the candidate (rhCygb and DXMS) in the well-established pulmonary brosis rat model evident by the CT diagnosis combined with biomarker measurement including HA, LN, PIIINP, and IVC.
In the current study, we found that rats treated with rhCygb had improved lung architecture than BLMtreated group, with observable reduced collage deposition and loss of air spaces in CT imaging, and the CT MLD were obviously lower than that of BLM group.It is known that CT MLD provide prognostic estimation of disease severity in patients with IPF and other ILDs and high MLD values may indicate severe IPF. [25]Consistent with these bene cial effects of rhCygb (4 mg/kg) treatment in our lung brosis rats, we found a marked attenuation of the HA, LN, P NP and C levels in serum and HYP concentration in rat lungs compared to BLM-treated rats.
Recent advances targeting ECM production and repair have provided novel approaches that could be used to treat chronic lung diseases.The ECM basement membrane matrix is composed of non brillar collagens (e.g., collagens IV and V), LN, and proteoglycans, and the lung's interstitial connective tissue is composed of complex networks of brillar collagen (e.g., collagens I and ) as well as HA and proteoglycans.Excessive accumulation of ECM molecules within the interstitial matrix is thought to underlie the pathogenesis of IPF. [41,42]In clinic research, serum LN, IVC, PIIINP, and HA signi cantly increased in the patients (323 patients) with IPF or CTD-ILD compared with the healthy controls.After treatment, the survived patients had signi cantly lower serum LN, IVC, PIIINP, and HA than the dead patients. [9]In the current study, we found there was no signi cant difference in HA, LN, P NP levels between rhCygb+BLM group and control group at 56 days, so we inferred that the reversing effect of rhCygb may be involved in the regulation of ECM turnover.
This study also has several limitations.The molecular mechanism underlying the reversing effect of rhCygb in lung brosis remains unknown; and the anti-brotic effects of rhCygb was investigated only against the rat model, future works will be done to validate the effects of rhCygb using other animal IPF models.However, we believe that the experiment costs, time and labor could be lessened due to the noninvasive CT radiologic and biomarkers measurement.We will further verify the reversing lung brosis effects of rhCygb in other models in our future study.

Conclusions
In conclusion, our results indicate that rhCygb plays a signi cant reversing role in bleomycin-induced lung brosis rats.This bene cial effect is mediated by decreasing the EMC molecules of HA, LN, PIIINP and IVC levels near to the control levels.Our data also indicate that rhCygb lower the CT MLD, HYP and Masson CVF, which re ect the progress of lung brosis.CT MLD positively correlates with serum levels of HA, LN, PIIINP, IVC, and HYP, and especially with CVF.Our ndings encourage CT combined with biomarkers should be performed to evaluate whether or not lung brosis model is established.