CD5L Deficiency Protects Mice Against Bleomycin-Induced Pulmonary Fibrosis

Background : Pulmonary fibrosis (PF), the most common clinical type of irreversible interstitial lung disease with one of the worse prognoses, has a largely unknown molecular mechanisms that underlies its progression. CD5 molecule-like (CD5L) functions in an indispensable role during inflammatory responses; however, whether CD5L functions in regulating bleomycin (BLM)-induced lung fibrosis is less clear. Methods : Herein, we describe the engineering of Cd5l knockout mice using CRISPR/Cas9 gene editing technology. The BLM-induced model of acute lung injury represents the most widely used experimental rodent model for PF. Results : Taking advantage of this model, we demonstrated that both CD5L mRNA and protein were enriched in the lungs of mice following BLM-induced pulmonary fibrosis. Inhibition of CD5L prevented mice from BLM-induced lung fibrosis and injury. In particular, a lack of CD5L significantly attenuated inflammatory response and promoted M2 polarization in the lung of this pulmonary fibrosis model as well as suppressing macrophage apoptosis. Conclusions : Collectively, our data support that CD5L deficiency can suppress the development of pulmonary fibrosis, and also provides new molecular targets for the use of immunotherapy to treat lung fibrosis.


Introduction
Pulmonary fibrosis (PF) has a poor prognosis and represents a severe health problem worldwide [1].PF is a chronically progressive lung disease characterized by the disruption of the pulmonary parenchyma and lung architecture, and accumulation of extracellular matrix (ECM) that leads to respiratory failure [2][3][4].Despite intensive studies, the complex pathogenesis and variable progression result in a median survival of only 3-5 years after diagnosis [5].The bleomycin (BLM)-induced model of acute lung injury represents the most widely used experimental rodent model for this disease, and simulates inflammatory and fibrotic events similar to pulmonary fibrosis symptoms seen clinically [6][7][8][9].BLM exposure results in release of inflammatory mediators and promotes initial inflow from neutrophils and, subsequently macrophages to sites of injury, with commensurate rises in ECM modulation and cytokine production including collagen, fibronectin, and hyaluronan [10].
During damage and fibrotic progression, abundant infiltration of macrophages occurs and portends poor prognosis for PF patients [11].In particular, macrophages play a crucial role in both inflammation and fibrotic phases of the disease where they adopt diverse phenotypes and differentiation in two distinct subsets.Classically, M1-like macrophages are activated by interferon-gamma (IFN-γ) and lipopolysaccharides (LPS), and promote proinflammatory cytokines [12].In contrast, macrophages may adopt an alternate M2-like phenotype and are stimulated by IL-4 and IL-13 to participate in wound healing and tissue remodeling [13].Notably, patients with idiopathic pulmonary fibrosis (IPF) were shown accumulate more M2-like macrophages during disease development and progression [14].Animal studies also indicate that M2-like polarization is considered a promising target for both prevention and therapy of pulmonary fibrosis [15][16][17].
CD5 molecule-like (CD5L), also known as apoptosis inhibitor of macrophage (AIM), has been reported to function in response to inflammation and microorganisms, as well as during fat metabolism.CD5L is a broad IgMbinding secreted protein [18,19].CD5L is principally expressed in macrophages and alveolar epithelial cells [20][21][22], and has been reported to protect against pro-apoptotic stimuli in macrophages, T cells, and NKT cells [19,23,24].CD5L has also been considered to be a pro-inflammatory factor in the pathogenesis of colitis [25], atherosclerosis [26], bacterial infection [27], and chronic obstructive pulmonary disease (COPD).Current evidence shows that higher circulating CD5L levels were associated with poor liver function in patients with liver cirrhosis [28,29].Moreover, CD5L is associated with lipid synthesis and function, for example, fatty acid synthesis, subcellular membranes [21,30], extracellular vesicles [31], LPSs and lipoteichoic acid [32].CD5L is potentially a pattern identification molecule [33]; however, whether it can be regarded as a potential treatment target for PF has not been explored.
In this study, we hypothesized that CD5L mediates macrophage apoptosis during lung fibrosis.We took advantage of the BLM-induced lung fibrosis mouse model to investigate: (1) Whether Cd5l expression would respond to the development of PF; (2) Whether Cd5l knockout (Cd5l −/− ) mice will exhibit physiologic differences compared to wild-type mice; (3) Whether deletion of Cd5l would reduce expression of fibrogenic genes and attenuate collagen deposition in early disease stages in a BLMinduced lung fibrosis mouse model; and (4) Whether deletion of Cd5l would regulate macrophage apoptosis during early stages of BLM-induced lung fibrosis.The results of our study provide empirical evidence that inhibition CD5L could be a beneficial preventative and therapeutic strategy for pulmonary fibrosis.

Laboratory Mice
C57BL/6 (wild-type-WT) mice were bred in specific pathogen-free animal conditions, and all procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Shanghai Laboratory Animal Research Center (Authorization numbers: 202202002).

Collection of Bronchoalveolar Lavage Fluid and Lung Sample
The lungs were prepared for lavage by cannulating the trachea with a catheter attached by a syringe, the bronchoalveolar lavage fluid (BALF) was collected by flashing the lung three times with 1 mL of sterile phosphate buffer saline (PBS) [35].Approximately 0.6 mL of BALF was routinely harvested from each mouse.After BALF recovery, the lungs were removed and stored at -80 °C until further evaluation.

Apoptosis Assessment
To determine fibroblast apoptosis following bleomycin, an Annexin V/propidium iodide (PI) apoptosis detection kit (Thermo Fisher Scientific, United States) was used to co-stain BALF [36] to discriminate between live cells and those in early and late apoptotic stages [37].Briefly, BALF cells were incubated with Annexin V and PI solutions for 15 min at room temperature in the dark.Fluorescence intensity was detected using a CytoFLEX Flow Cytometer (B5R3V5, Sadina, California, United States).For this evaluation by flow cytometry, a higher FITC-AnnexinV/lower PI value (termed Q3 region) defined early apoptotic cells whereas higher FITC-Annexin-V/higher PI (termed Q2 region) defined late apoptotic cells.Live cells were defined as lower FITC-Annexin-V/lower PI (termed Q4 region).The percentage of Annexin V-or PI-positive cells quantified the extent of apoptosis.

Macrophage Phenotyping
For classification of macrophages cells, the BALF was centrifuged for 10 min at 1500 rpm and immunostained.Lung macrophages cells were stained with antimouse CD16/CD32 (553140, Franklin Lakes, New Jer- Data analysis was performed using FlowJo software (v10, FlowJo LLC, Ashland, OR, USA).

Histological Analysis
The right lung tissue was excised and immersed in 4% neutral paraformaldehyde overnight at room temperature, and subsequently embedded in paraffin.Tissue was sectioned cut into 5 µm-thick sections for histological analysis using established protocols.After dewaxing, the alveolitis and fibrosis were investigated using hematoxylin and eosin (H&E) staining.Sirius red and Masson's trichrome staining method were conducted to reveal collagen deposition (blue staining).The severity of fibrosis changes was scored using the Ashcroft scoring system scoring from 0-8 [38,39] in a double-blind evaluation.Each object, such as aerated lung area (an indicator of lung consolidation extent) and sum or mean collagen volume area (an indicator of collagen extent), were determined using ImagePro-Plus version 6.0 (Media Cybernetics, Rockville, MD, USA).Mean values from all fields of view analyzed in for each mouse lung were calculated.

ELISA
Whole blood (approx. 1 mL) was collected into 1.5 mL tubes and allow to stand for 30 min at room temperature.The samples were subsequently centrifuged for 15 min at 350 g to obtain serum.Serum concentrations of TGFβ, IFNγ, and CCL2/MCP-1 in the serum were determined by enzyme-linked immunosorbent assay (ELISA) kits following the manufacturer's protocol (Biolegend, United States).

Western Blot Analysis
The protein in lung homogenate was extracted used by RIPA lysis buffer (Takara, Dalian, China).Western blot was analyzed using established procedures with indicated primary antibodies.Briefly, protein was separated on 10% polyacrylamide gel and electro-transferred to PVDF membrane.5% non-fat milk was used to block PVDF membranes for 1 h and then incubated overnight at 4 °C with primary antibody to αSMA (ab232784, Abcam, Cambridge, UK).After three washes with PBST (5 min each), Donkey anti-Rabbit secondary antibodies was applied from Odyssey (LI-COR, Lincoln, Nebraska, United States) and then visualized using LICOR (Odyssey, Shenzhen, China).

Generation of C57BL/6-Cd5l Knockout (Cd5l −/− ) Mice
To generate Cd5l −/− mice, CRISPR/Cas9-mediated genome editing was performed to generate Cd5l −/− mice (Fig. 1a).Two sgRNAs were designed to target the exons 2-4 according to the strategy indicated in Fig. 1a.Cas9 mRNA and sgRNAs were co-injected into the fertilized oocytes resulting in double-strand breaks and subsequent homologous recombination within the genomic targeted locus.PCR was performed to identify mice with Cd5l alteration.F0 mice with predict recombination were crossed with wild-type mice to generate 7 F1 mice, and the identification results are shown in Fig. 1b,c.As predicted, genome sequencing identified that the 7323 bp fragment encoded a knockout of the Cd5l gene with the mouse genome (Table 3).The Cd5l −/− mice were generated following inbreeding of F1 positive mice for at least four generations.Real-time PCR quantified the expression of Cd5l mRNA level.Cd5l was detected in tissues of wild-type mice, while this transcript was nearly undetectable in Cd5l −/− mice (Fig. 1c).Taken together, the Cd5l were knockout successfully through CRISPR/Cas9.

Pulmonary Fibrosis Manifests Altered Cd5l Expression
As Cd5l is associated with inflammation during the development of lung fibrosis [40].We first reasoned that Cd5l may be involved in the pathophysiology of pulmonary fibrosis and to test this, a 3 mg/kg dose of bleomycin was given once intratracheally to WT and Cd5l −/− mice to establish bleomycin-induced pulmonary fibrosis model (Fig. 2a).
We next sought to examine the expression of Cd5l in the lung of mice with BLM-induced pulmonary fibrosis.A 2.5-fold higher Cd5l RNA and 3-fold higher protein expression were detected in BLM-induced WT mice 21 days after BLM administration as compared with 0 day (Fig. 2b-d).
Collectively, these data indicate that CD5L is involved in the development of pulmonary fibrosis.

Loss of Cd5l Attenuates BLM-Induced Lung Injury and Fibrosis as well as Reducing Lung Profibrotic Gene Expression
Based on the above observations, the effect of Cd5l on inflammation and fibrotic in the lung of Cd5l −/− and WT mice treated by bleomycin were subsequently evaluated.We first sought to analyze weight loss of WT and Cd5l −/− mice with bleomycin-treated.Saline administration to WT and Cd5l −/− mice did not alter body weight (BW).Following day 3 of BLM induction, both Cd5l −/− and WT mice presented with a severe loss of weight, a commonly observed effect during pulmonary fibrosis [41,42], although a temporal recovery is commonly observed.Importantly, Cd5l −/− mice exhibited a mild weight loss throughout BLM induction when compared with WT mice (Fig. 3a).Moreover, the body weight of BLM-treated WT mice was lower than that measured in saline groups.Histopathological examination and relative lung weight confirmed no significant difference in lung morphology in untreated WT or Cd5l −/− mice; however, marked inflammatory cell infiltration with damage and fibrosis in the lungs were observed in BLM-treated animals at day 21 (Fig. 3b).Hematoxylin and eosin (H&E) stained lung sections of Cd5l −/− mice showed less inflammatory cell migration than WT mice, indicative of a weakened inflammatory response in Cd5l −/− mice.Reduced collagen deposition in Cd5l −/− mice was also noted using Masson and Sirius red staining compared to WT mice after BLM administration (Fig. 3ce, left panel).In particular, lower Ashcroft scores indicated that the severity of pulmonary fibrosis was substantially reduced (Fig. 3c, right panel).Collagen volume in bluestained areas (Fig. 3d) and aerated lung area was analyzed by Image-Pro Plus software and similar results were obtained (Fig. 3e, right panel).Based on both weight changes and histopathological staining, we conclude that Cd5l deficiency confers considerable protection against the development pulmonary fibrosis.

CD5L Reduced the Expression of Fibrosis Specific Indicators Induced by BLM
Next, the levels of fibrosis-specific markers were examined.Acta2 (αSMA), Col1a1 (collagen I), and Eln (Elastin) mRNA in lung were measured by qPCR (Fig. 4ac).As expected, the levels of αSMA, collagen I and Elastin were noticeably upregulated in the BLM groups compared with un-induced groups.
It has been have previously shown that the bleomycintreated respiratory distress is associated with an early inflammatory reaction [42,43], thus, we next examined lung inflammation.The inflammation-linked genes Tgfβ, Mcp1 and Ifnγ were measured on day 21 after BLM challenge.As expected, results showed significant increases in expression of Mcp1 and Tgfβ as a result of bleomycin-induced lung injury in WT mice, whereas expression of these genes was suppressed by deletion of Cd5l (Fig. 4d-f).Ifnγ expression was higher in BLM-induced lung of Cd5l −/− mice compared to saline Cd5l −/− mice.ELISA detection further proved similar levels of TGFβ, MCP-1 and IFNγ expression of lung between two BLM groups (Fig. 4g-i).Knockout of Cd5l did not alter either of the Tnfα or Il6 expression when compared with the BLM-WT group (data not shown).

Cd5l Deficiency Influence Macrophages Polarization
We characterized proportion of differential cells in the BALF obtained from various groups on day 21 after BLM-treatment.Flow cytometry showed that both WT and Cd5l −/− mice displayed enhanced CD45 + expression as a percentage of total live cells.Compared with controls, results showed a 2.8-fold (65%) increase in CD45 + cells in WT mice 21 days after BLM-induced damage.Cd5l −/− mice had a slightly inflammatory response to BLM (35%),  and a modest 30% upregulation in CD45 + cells compared with WT mice (Fig. 5a).CD11b + /F4/80 + macrophages in both groups showed no change after BLM-induction (Fig. 5b).We further analyzed the effect of Cd5l depletion on macrophage typing by examining the specific markers of M1 and M2 macrophages.M1 macrophage surface marker F4/80/MHCII confirmed BLM promoted a 1.5-fold induction of M1 macrophages in Cd5l −/− mice, but this effect was not observed in WT mice (Fig. 5c).In contrast, the accumulation of M2 macrophages, as measured by CD206 expression, increased 3-fold in BALF from BLM-WT mice.However, a 7-fold increase in the M2 phenotype was observed in Cd5l −/− mice after BLM challenge (Fig. 5d).Overall, these data suggest that CD5L may modulate the macrophage subtypes.

Cd5l Deficiency Reduced Macrophages Apoptosis
We next analyzed apoptosis in macrophages within BALF taken from BLM-treated mice by flow cytometry.An untreated control group were included in this analysis, so the death detected here may be conferred by treatment with TE buffer, and during assay preparation and centrifugation death may have occurred [44].Annexin V/PI staining showed that the proportions of early apoptotic cells in both models were found as significantly increased compared over controls, but after bleomycin treatment, apoptotic cells in BALF of Cd5l −/− mice displayed a remarkable reduction compared that in WT mice (Saline/WT: 0.664 ± 0.218; Saline/Cd5l −/− : 0.694 ± 0.199, BLM/WT: 11.6202 ± 1.493; BLM-Cd5l −/− : 2.77 ± 0.361, p = 0.0004).Flow cytometry revealed that Cd5l-deficiency repressed macrophage apoptosis after bleomycin stimulation (Fig. 6).

Discussion
The clinical role for CD5L in acute respiratory distress syndrome (ARDS), and trauma victims diagnosed with ARDS display a higher overall CD5L level, compared with healthy people.Further, serum CD5L was upregulated within 24 hours after trauma in patients, and severe trauma patients have higher CD5L values compared to more than mild trauma patients [45].The serum free state of CD5L had been shown to have increased response and expression during LPS-induced lung injury in a mouse model [46].
Based on previous reports, we first examined Cd5l expression in the BLM-induced fibrotic lung of WT mice, and found that Cd5l expression was gradually elevated in the lung of BLM-treated mice at 7 and 21 days after BLM-treatment compared to controls.Consistent with mRNA, Cd5l protein was enriched in the lung of treated mice (Fig. 2).A novel Cd5l knockout mouse model was generated and an in vivo model of moderate fibrosis was established.First, we examined a potential suppressive role for Cd5l played in pulmonary fibrosis has been verified by pathological assessment.Cd5l knockout results in tissue inflammation and collagen accumulation.Additionally, both the mRNA and protein levels of αSMA, collagen I, and elastin, the three most specific fibrotic indicators, were consistently and significantly reduced in Cd5l −/− mice following BLM-induced pulmonary fibrosis compared to BLM-treated WT mice (Fig. 4a-c).We also verified that Cd5l could effectively repress the progression of PF and provide more potential effective treatment for PF.Previously it had been demonstrated that CD5L blockade downregulated the expression of inflammatory cytokines and chemokines, and upregulated the production of anti-inflammatory factors, as well as the infiltration of leukocytes including macrophages, lymphocytes and neutrophils in peritoneal lavage fluid after CLP-treated sepsis [47].In this study, the level of inflammatory cytokines (Fig. 4d-i) and the infiltration of lymphocytes were also found to be suppressed by Cd5l depletion in pulmonary fibrosis (Fig. 5a).In the pathogenesis of inflammatory processes, CD5L is thought to control lipid metabolism by inhibiting macrophage apoptosis and also responses to TLR activation by promoting an anti-inflammatory cytokine profile [21].Initially, macrophages are divided into two functional subtypes.Classical active macrophages, the M1-like phenotype with representative markers iNOS or MHC II, are characterized by their association with several cytotoxic cytokines.M1 macrophages can be repolarized and transfer their phenotype toward M2 macrophages, which express the representative marker CD206, and which are defined by pro-regenerative and anti-inflammatory properties [48].There is emerging evidence that CD5L may function as a driver of M2 macrophage polarization [49].
One could hypothesize that the rise in CD5L during PF is an adaptive response to lung injury and fibrosis.To answer these questions, flow cytometry performed on BALF showed that M1 macrophages in Cd5l −/− mice are similar to WT mice at baseline.However, M1 macrophages from Cd5l −/− mice increased in the BALF after bleomycin exposure (Fig. 5c).According to previous studies [50], M2-like phenotype are crucial pathogenic factors during the course of BLM-induced pulmonary fibrosis, and the accumulated macrophage M2 program is generally related to collagen remodeling of internal tissues including the heart, kidneys, liver, and lungs.CD5L influences M1/M2 profiles and has positive feedback [49].We thus then carried out studies and provided evidence that M2 macrophage levels were higher in BLM-treated mice.In particular, we measured a 7-fold increase in M2 macrophages in BALF obtained from Cd5l −/− mice, which consistent with previous findings (Fig. 5d).
CD5L has previously been reported to have an antiapoptotic effect [19,24,51].We verified the effects of Cd5l on macrophage apoptosis in our mouse model.The proportion of apoptotic cells were significantly lower in Cd5l −/− mice than that measured in WT mice after BLM exposure (Fig. 6).AIM is associated with insulin-like growth factorbinding protein-4 [52], and it was reported that there is cross-talk between lipid mediators including SPMs [53], DHA [54], and PGE2 [55,56] that accelerate apoptosis relative to cell type and physiological conditions.Therefore, the pathogenic/inflammatory conditions promote apoptosis the regulatory function in apoptotic cell death may enable illumination of the mechanism during apoptosis.
We have shown that Cd5l protects mice from pulmonary fibrosis, but its regulatory signaling mechanisms remain unclear as does its function in inhibiting apoptosis.In future work, we plan to dissect the common signaling crosstalk as more details concerning the previously mentioned pathways and targeted factors need to be uncovered.

Conclusions
In conclusion, our findings demonstrate that CD5 molecule-like (CD5L)-deficiency protected mice from bleomycin-induced lung injury and fibrosis by attenuating the release of inflammatory mediators, regulating macrophages polarization and apoptosis, which could be a viable strategy for the prevention and treatment of pulmonary fibrosis in clinical settings.These findings may provide a sufficient theoretical basis for the clinical application of CD5L and development of novel therapeutics for treating pulmonary fibrosis.

Fig. 1 .
Fig. 1. Construction strategy of Cd5l knockout mice and Cd5l expression analysis.(a) Exons 2-4 were chosen as the region for gene knockout, and sgRNAs designed to target sites upstream and downstream of exons 2 and 4, CRISPR-Cas9 technology was used to produce a reading frame shifting mutation within the Cd5l.(b) PCR gel electrophoresis results of heterozygous and homozygous mice, HE-heterozygous mice, HO-homozygous mice, WT-wild type mice.(c) Statistics of Cd5l mRNA expression.

Fig. 2 .
Fig. 2. Expression of Cd5l in bleomycin-induced pulmonary fibrosis in mice.(a) A schematic diagram for BLM-induction is shown.(b) Cd5l mRNA expression in the lung of WT mice treated with saline or BLM.(c,d) Western blot analysis of Cd5l expression in the lung of WT mice following 21 days of BLM induction.Left panel: representative western blots are shown.Right panel: Immunoblot signals were quantified and a bar graph illustrating results is shown.Statistical testing used was a Student's t test.BLM, bleomycin.

Fig. 3 .
Fig. 3. Loss of Cd5l attenuates lung injury and fibrosis.(a) Weights of mice treated with BLM.Values are means ± SEM.(b) Lung tissue and relative weight in mice.(c-e) Histological examination of the extent of lung fibrosis in mice induced by BLM.Left panel: H&E (c), Masson (d), and Sirius red (e) staining.Collagen fibers are stained blue and muscle fibers are stained red.Black arrows indicate injury position of lung induced by BLM.Right panel: A bar graph depicting semiquantitative Ashcroft scores, collagen volume fraction (CVF), and aerated lung area fraction for the severity of fibrosis is given.Scale bar = 200 µm, error bars represent mean ± SEM (n = 5).*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Fig. 5 .
Fig. 5. CD5L predominantly induce M2 polarization in lung following BLM treatment.Left panel: Representative scatter diagram generated during FACS analysis.Right panel: A bar graph indicating the data obtained.(a) Flow cytometry analysis of CD45 + and macrophage differentiated cells following stimulation with or without BLM.(b) Percent of macrophages in the BALF obtained both BLM mice from both groups.The cells were first gated in zombie and CD45, and then to analyzed for CD11b and F4/80 expression.(c,d) BALF were stained with F4/80-BV421/MHCII-PE and F4/80-BV421/CD206FITC and cellular phenotype analyzed by FACS.The cells were initially gated using CD11b.Mean values from 10 replicates with mean ± SEM. *p < 0.05, **p < 0.01, ****p < 0.0001.