Constitutive AKT Activity Predisposes Lung Fibrosis by Regulating Macrophage, Myofibroblast and Fibrocyte Recruitment and Changes in Autophagy

The etiology and pathogenesis of pulmonary fibrosis is poorly understood. We and others reported that M-CSF/CSF-1, M-CSF-R and downstream AKT activation plays an important role in lung fibrosis in mice models and in IPF patients. To understand potential molecular pathways used by M-CSF-R activation to direct lung fibrosis, we used a novel transgenic mouse model that expresses a constitutively-active form of AKT, myristoylated AKT (Myr-Akt), driven by the c-fms (M-CSF-R) promoter. We were particularly interested in the basal immune state of the lungs of these Myr-Akt mice to assess M-CSF-R-related priming for lung fibrosis. In support of a priming effect, macrophages isolated from the lungs of unchallenged Myr-Akt mice displayed an M2-tropism, enhanced co-expression of M-CSF-R and α-SMA, reduced autophagy reflected by reduced expression of the key autophagy genes Beclin-1, MAP1-Lc3a(Lc3a), and MAP1-Lc3b(Lc3b), and increased p62/STSQM1 expression compared with littermate WT mice. Furthermore, Myr-Akt mice had more basal circulating fibrocytes than WT mice. Lastly, upon bleomycin challenge, Myr-Akt mice showed enhanced collagen deposition, increased F4/80+ and CD45+ cells, reduced autophagy genes Beclin-1, Lc3a, and Lc3b expression, and a shorter life-span than WT littermates. These data provide support that M-CSF-R/AKT activation may have a priming effect which can predispose lung tissue to pulmonary fibrosis.


Introduction
Idiopathic pulmonary fibrosis (IPF) is a progressive respiratory disease with a 5-year mortality rate of 50% -70% [1]. IPF is characterized pathologically by destruction of lung architecture, fibrotic remodeling, accumulation of myofibroblasts, and deposition of extracellular matrix (ECM) [2]. To date, the etiology and pathogenesis of IPF remains unclear.
The serine/threonine protein kinase, AKT, is downstream of many receptor tyrosine kinases and plays an important role in cell survival, differentiation, and cellular activation [3]. In macrophages, activation of the M-CSF receptor (M-CSF-R) induces stimulation of phosphoinositide 3-kinase (PI3K) leading to AKT recruitment to the cell membrane and subsequent Thr308 and Ser437 phosphorylation to activate AKT [4]. In fibroblasts, activated AKT regulates collagen I and III production [5] and contributes to human hepatic fibrosis [6] and bleomycin-induced lung fibrosis in mice [7] [8]. Importantly, PI3K/AKT activity is enhanced in IPF fibroblasts due to reduced levels of PTEN phosphatase function [9]. Moreover, PI3K/AKT inhibitors such as PX-866 are potential therapeutic agents for IPF [7] [10]. Thus, understanding the role of activated AKT is important in IPF.
In addition to M2 macrophages and myofibroblasts, circulating fibrocytes are also associated with pulmonary fibrosis [24]. Fibrocytes are bone marrow-derived mesenchymal cells and express leukocyte, hematopoietic cell and fibroblast markers [25]. Fibrocytes express chemokine receptors and adhesion molecules and can traffic to fibrotic regions [24]. Furthermore, fibrocytes can deposit the ECM proteins collagen I and III [26] and can also differentiate into fibroblasts and myofibroblasts, which further contribute to ECM deposition [27]. Therefore, circulating fibrocytes can contribute to the pathogenesis of pulmonary fibrosis [24] [28].
Reduced cellular homeostatic processes like autophagy are also found in the lungs of patients with IPF [29] [30] [31]. Autophagy is an evolutionarily-conserved lysosome-dependent cellular pathway responsible for the processing of unhealthy organelles and cellular waste to maintain cellular homeostasis during cellular stress [32]. Successful autophagy results in the delivery of proteins and/or organelles into double-membrane autophagosomes for lysosomal degradation [32] [33] and cellular homeostasis. Interestingly, autophagy activity correlates with organismal longevity [34] and interruption of autophagy is associated with cellular aging. Several studies report reduced autophagic vesicle formation (autophagosomes) and activity in human IPF lung tissue as measured by reduced LC3A/B-II (microtubule associated protein 1 light chain 3) and increased p62 (chaperone protein to autophagosome) [29] [30] [31].
BECLIN-1, a major regulator of autophagy, is decreased in IPF lung fibroblasts.
Furthermore, BECLIN-1 also binds to BCL-2 and decreases apoptosis in these fibroblasts [31] [35]. As an autophagy regulator, mammalian target of rapamycin, or mTOR, kinase activity is elevated in IPF lung fibroblasts [36] via enhanced AKT activation and PTEN suppression [37]. Thus, pathways regulating autophagy in IPF likely contribute to lung fibrosis. Previous data generated from our laboratory revealed that CSF1 contributes to pulmonary fibrosis in mice as CSF1-/-mice showed less fibrosis in response to bleomycin challenge [38]. However, the specific mechanism(s) of M-CSF-R-mediated signaling and regulatory pathways connecting CSF1 and M-CSF-R signaling to specific target cell populations is unclear. Current knowledge in the field is that AKT activation, M2 macrophage polarity, increased myofibroblast and fibrocytes numbers, and reduced autophagy are observed in the lungs of patients with IPF. Here, we extend our previous study by using a murine model, c-fms-directed Myr-Akt expression, to examine the role of constitutive AKT activation driven by the M-CSF-R promoter and identify potential priming of the lungs to fibrosis. With this murine model, we study a potential causal relationship between M-CSF-R-directed AKT activation and cellular phenotype to drive fibrosis. We observed that lungs from Myr-Akt mice had increased expression of genes representative of M2 macrophages, increased α-SMA positive cells, increased number of fibrocytes and reduced autophagy-related transcripts compared to WT mice. These data suggest M-CSF-R/AKT activation may predispose Advances in Bioscience and Biotechnology a pulmonary fibrotic phenotype.

Ethics Statement
This study was carried out in strict accordance with the recommendations in the
Briefly, a c-fms-Myr-Akt plasmid was used to generate the mouse expressing a membrane-bound form of Akt1 in mononuclear phagocytes. The myristoylation tag on Akt facilitates membrane binding resulting in constitutive activation (Supplemental Figure 1(a)). 6 -8 week old male and female mice were sacrificed by cardiac puncture after isoflurane inhalation and lungs perfused with 25 -35 mL DPBS. Tissues were fixed in 10% formalin for histologic and immunohistochemical analysis, snap-frozen in liquid nitrogen for RNA analysis, or digested immediately to obtain a single cell suspension for flow cytometry analysis.

Immunohistochemistry (IHC)
Immunohistochemical processing and staining with hematoxylin and eosin (H&E), Masson's Trichrome, F4/80 or α-SMA staining were done at the Histology Core at OSU Veterinarian School as described [41]. For LC3B staining, mouse lungs were fixed in freshly made 4% paraformaldehyde for 24 hours and processed and stained at OSU Wexner Medical Center Pathology Core. Ten pictures per section were captured using the Olympus IX50 inverted microscope-equipped Nikon camera (Olympus). The average number of pixels corresponding to specific stains was quantified using Adobe Photoshop CS and histogram analysis (Adobe) as described [41]. Data are presented as percent positive cells per high-power field (HPF) and are expressed as the mean ± SEM.

Single Cell Suspension and Flow Cytometry
Mouse tissue were harvested and immediately homogenized with the McIlwain Tissue Chopper (Ted Pella, Inc.) to obtain a single-cell suspension as described previously [41].

Fibrocytes Isolation
Primary murine fibrocytes were isolated from freshly collected peripheral blood as described [42]. Briefly, blood was collected from cardiac puncture and PBMCs were isolated using Lympholyte-Mammal (Cedarlene). PBMCs were cultured in FibroLife basal media (Lifeline Cell Technology) supplemented with 20 mM HEPES, 2× NEAA, 2 mM sodium pyruvate, 4 mM glutamine, 100 U/mL penicillin, 100 μg/ml streptomycin, and 2× ITS-3 for 5 days to generate mouse fibrocytes. Cells were counted and an equal number of cells were stained with antibodies specific for CD45 and CXCR4, permeabilized/fixed, and stained for Collagen I. The cells were analyzed by flow cytometry.

Analysis of Lung Inflammation by Pathological Assessment
A board-certified veterinary pathologist subjectively analyzed in a blinded manner and assessed the lungs for inflammation, the number of macrophages, and the amount of fibrous connective tissue present within the H&E-and Trichrome-stained slides, and scored them accordingly as previously described [38].

Western Blot
Single cell suspensions were obtained and lysed in Cell Lysis Buffer with protease inhibitors (Cell Signaling Technology). Protein concentration was determined by the Dc protein assay (Bio-Rad) and equal amounts of protein were separated on 10% or 4% -12% NuPAGE Novex bis-tris gel (Thermofisher Scientific), transferred to a nitrocellulose membrane, probed with the indicated antibodies and detected by ECL (GE Healthcare Bio-Sciences Corp.). ECL signal and band density were quantified using Quantity One program (Bio-Rad). Protein expression was normalized to total protein or β-ACTIN.

Statistical Analyses
Data were analyzed using analysis of variance (one-way ANOVA) and multiplicities were adjusted by Tukey's method to control the family-wise error rate at 0.05 using JMP 9 software (SAS Institute, Inc; Cary, NC). P ≤ 0.05 was defined as statistically significant in these studies. Hypothesis testing of synergistic effect were tested by interaction contrasts. Differences in survival between groups were compared by log-rank test. SAS 9.3 software (SAS Institute, Inc; Cary, NC) were used for data analysis.

M-CSF-R Driven Myr-Akt Induces an M2 Macrophage Tropism in Lung
Since bone marrow-derived macrophages (BMM) collected from mice expressing Myr-Akt driven by the M-CSF-R promoter had increased survival [40], we first checked if Myr-Akt mice produced more BMM than WT mice. Penetrance galy. For this reason, classification as Myr-Akt genotype in this study was confirmed using both murine genotyping (Supplemental Figure 1(a)) and spleen size (Supplemental Figure 1(b)). Immunohistochemical staining for the macrophage marker F4/80 revealed a significant increase in macrophage numbers in the spleens of Myr-Akt mice compared to WT littermates (p < 0.001) ( Figure   1(a), top and Figure 1(b)), however, we did not observe any differences in F4/80 expression in the lungs between these Myr-Akt and WT mice (Figure 1(a), bottom and Figure 1(b)). We next derived macrophages from bone marrow and measured F4/80 expression with flow cytometry. Our data suggest that Myr-Akt mice have increased numbers of F4/80+ BMMs compared to WT mice at day 1 (p = 0.02) and day 2 (p = 0.02) (Figure 1(c)). However, BMM numbers were similar after day 3 in culture (data not shown).
Since we did not observe more macrophages in the lungs of the Myr-Akt mice, we next asked if tissue macrophages from the lungs or spleens from Myr-Akt mice expressed a phenotype representing M2 phenotype [43] [44]. We isolated total RNA from whole lung or spleen tissue and performed qRT-PCR for the expression of M1 and M2 genes. There was no significant difference be-

α-SMA+ and α-SMA+/M-CSF-R+ Cell Influx in the Lungs of Unchallenged Myr-Akt Mice
Since α-SMA positive myofibroblasts mediate pulmonary fibrosis in human IPF [21] [45]. We next wanted to determine basal differences in α-SMA expression in lung cells [46] from Myr-Akt and WT mice. Using immunostaining with the myofibroblast marker α-SMA, we found more α-SMA expression in lung tissue of Myr-Akt mice than WT mice (p = 0.001) that were mostly confined in the ep-

Myr-Akt Mice have Altered Expression of Key Autophagy Genes in the Lung
Autophagy is a process that maintains cell and tissue homeostasis by degrading and recycling cellular proteins and organelles during times of cellular stress.
Autophagy activity is associated with organismal longevity and is reduced in  (Figure 3(a) and Figure 3(b)). Interestingly, we observed a significant decrease in LC3B in lungs from Myr-Akt mice compared to lungs from WT mice (p = 0.016), implying autophagy is impaired in the lung of Myr-Akt mice.

Myr-Akt Mice Produce Increased Numbers of Circulating Fibrocytes in Vivo
Bone-marrow derived circulating fibrocytes can traffic to the lung during the pathogenesis of pulmonary fibrosis [24] [28]. We inquired whether Myr-Akt mice could generate more circulating fibrocytes, in vivo. We generated fibrocytes from mouse PBMCs cultured in fibrocyte-deriving medium and observed more spindle-shaped fibroblast-like cell morphology from Myr-Akt mice compared with WT mice (Figure 4(a)). Since CD45+/COL1+/CXCR4+ fibrocytes were expanded in the bone marrow and have been reported to contribute to bleomycin-induced pulmonary fibrosis [24], we next stained the PBMC-differentiated cells with antibodies specific for CD45, Collagen I, and CXCR4 (Figure 4(b)). Our data suggest Myr-Akt mice generate significantly more circulating CD45+/COL1+/CXCR4+ fibrocytes than WT mice (p = 0.05), basally.

Bleomycin Challenge Enhances Collagen Deposition in the Lungs and Reduces Survival in Myr-Akt Mice
Because Myr-Akt macrophages express an M2 tropism, increased α-SMA+ myofibroblasts, increased circulating CD45+/COL1+/CXCR4+ fibrocytes, and reduced autophagy-related gene expression, we hypothesized that Myr-Akt mice would develop more fibrosis in response to repeated administration of bleomycin. Kaplan-Meier survival analysis ( Figure 5(a)) illustrates decreased survival in Myr-Akt mice challenged with bleomycin compared to bleomycin-treated WT mice (p = 0.0284). While there was no mortality in the Myr-Akt or WT mice groups treated with PBS (vehicle), the Myr-Akt mice started to die at day 7 post-bleomycin treatment while all WT mice survived until day 28 on the bleomycin protocol. The mice receiving bleomycin and surviving to day 33 showed typical fibrosis in the lung compared to vehicle-treated mice, including blue stained pixels (Trichrome) as an indication of collagen deposition (Figure 5(b) and Figure 5(c)) and significant inflammation ( Figure 5(d)). Bleomycin    (Figure 5(e)). Our results suggest that Myr-Akt enhances the bleomycin effect leading to the increased lung fibrosis and decreased overall survival.

cin treatment significantly increased CD45+ cells in the lungs of WT and
Myr-Akt mice (both p < 0.01). Additionally, more CD45+ cells were present in the lungs of Myr-Akt mice compared to the lungs of WT mice after bleomycin challenge (p = 0.001) (Figure 6(a) and Figure 6(b)). Importantly, unlike CD45+ cells, and as seen in Figure 1(a) and Figure 1

Bleomycin Suppresses Autophagy Gene Expression in the Lungs
Reduced levels of autophagy in IPF induce epithelial cell senescence and accelerate lung fibroblast differentiation into myofibroblasts [29]. Our data suggest that Myr-Akt mice both basally and post-bleomycin contained more α-SMA+ lung myofibroblasts and fewer autophagy-related genes expressed in lung tissue than WT mice. Specifically, we initially observed (Figure 3) that basal Myr-Akt expression decreased the key autophagy genes Beclin-1, LC3a and LC3b compared to WT. This data was confirmed as vehicle-treated Myr-Akt mice had significantly reduced Beclin-1, LC3a and LC3b mRNAs in whole lung tissue compared to WT mice (p < 0.001 for Beclin-1, LC3a, and LC3b). Predictably, bleomycin  (Figure 7). Interestingly, a synergistic decrease of autophagy genes was observed after bleomycin treatment in Myr-Akt mice (p < 0.001 for beclin-1, LC3a and LC3b) (Figure 7).

Discussion
We previously reported that CSF1 plays an important role in pulmonary fibrosis in both mice and IPF patients [38]. However, that report lacked critical mechanistic data about how CSF1 and M-CSF-R-mediated signaling contributes to  stromal formation, and clearance of cell debris in the ECM [16].
In addition to our observations about the role of M-CSF-R activation in IPF [38], another study demonstrated elevated M2 genes CCL18, CCL22, and CD206 from alveolar macrophages of IPF patients. In this study, the authors proposed L-4 and IL-10 are responsible for the M2 phenotype shift [55]. Most importantly, our findings suggest a link between M-CSF-R-regulated AKT activation and an M2 macrophage polarity as whole lung tissue from Myr-Akt mice had increased IL-10 mRNA and an M2 macrophage polarization in both lungs and spleens. Admittedly, alveolar macrophages are not the only producers of IL-10, and increases in IL-10 mRNA from whole lung tissue could be from other inflammatory cells as well as type I alveolar epithelial cells in this model. Studies are underway to confirm the sources of lung IL-10 in the mouse model. Indeed, the negative regulator of PI3K/AKT, src homology 2-containing inositol phosphatase (SHIP), interferes with lung M2 macrophage skewing as peritoneal and alveolar macrophages from SHIP1-/-mice express a strong M2 gene signature in terms of arginase-1 and Ym-1 mRNAs along with M2-mediated lung pathology [44]. Bone marrow-derived macrophages from SHIP1-/-mice display an M2 phenotype only upon TGF-β or IL-3 stimulation [14] [43]. Furthermore, the PI3K inhibitor LY294002 decreases the expression of M2 genes from alveolar macrophages [2] [14]. It is interesting that we did not observe any significant morphological changes or increase in macrophage numbers in the lungs from the Myr-Akt compared to WT mice (Figures 1(a), bottom and Figures 1(b)).
Therefore, our data of macrophage M2 phenotype tropism, increased myofibroblast and circulating fibrocytes population, and reduced autophagy genes Interestingly, phosphatase and tensin homolog (PTEN), a negative regulator of AKT activity, is down-regulated in myofibroblasts [56] isolated from IPF patients [9]. We observed elevated numbers of basal α-SMA+ myofibroblasts in untreated Myr-Akt lungs. Moreover, myofibroblasts have been shown to be more resistant to apoptosis than normal fibroblasts, which may interrupt epithelial cell repair [19] [30] [57]. As constitutively-active AKT was observed in Increasing evidence indicates detection of fibrocytes in fibrotic lungs. Fibrocytes are important sources of fibroblasts and myofibroblasts during tissue repair and remodeling processes [24]. Indeed, a key marker for fibrocytes, CXCR4 expression, is regulated by the PI3K/AKT/mTOR pathway and hypoxia as inhibiting PI3K/AKT/mTOR activity reduced PDGF and hypoxia-induced CXCR4 [27]. These data support increases in circulating fibrocytes in our Myr-Akt mice. Interestingly, a recent report suggests that fibrocytes do not contribute to significant type I collagen deposition in fibrotic lungs and that they may play an indirect role in fibrosis by orchestrating recruitment of other cells [58]. Studies to investigate the role of AKT in the differentiation of macrophages into fibroblasts and their paracrine effects on fibroblasts are ongoing.
In addition to the predisposed M2 polarity, and myofibroblast and fibrocyte presence, we also found cells from lungs of Myr-Akt mice had altered expression of key autophagy-regulatory genes, indicating reduced autophagy. Since autophagy is a critical pathway for cellular longevity and homeostasis, and is reduced in lungs of IPF patients, understanding the molecular regulation of this pathway may contribute to effective pharmacological intervention. Based on existing literature, we chose to measure LC3-II (lipidated/autophagosome membrane associated form of LC3), Beclin-1 and p62 (ubiquitin binding autophagy substrate receptor) proteins as markers for autophagic. Use of three markers to assess autophagic helps overcome the limitations of any one single marker in assessing such a complex system, facilitating more accurate interpretation of results. Our results indicate that lung tissue from Myr-Akt mice had reduced autophagy. Consistent with our observation in Myr-Akt mice, others demonstrate that inhi-biting autophagy by knocking down Lc3b and Atg5 mRNAs enhance expression of α-SMA and type I collagen in lung fibroblasts [29]. Similarly, inducing autophagy by inhibiting mTOR activity in lung fibroblasts with rapamycin reduces cellular α-SMA and fibronectin expression [30] linking autophagy and α-SMA expression with lung fibrosis in Myr-Akt mice. Of note, in addition to qRT-PCR for expression of lung collagen mRNAs, we elected to use Trichrome staining in lieu of Sircol or hydroxyproline assay on whole lung tissue to determine potential difference in location of collagen deposition. Further, we failed to assess autophagic flux with lysosome inhibitors, in vivo. Autophagic flux is the complete process of autophagy from phagophore formation to substrate degradation and release of breakdown products. Measuring autophagy flux will be included in our future autophagy analysis.
Interestingly, the regulation of autophagy may be cell-specific. Araya et al. reported that activating autophagy in human bronchial epithelial cells suppresses endoplasmic reticulum (ER) stress and cell senescence [29]. In contrast, inhibiting autophagy also promotes myofibroblast differentiation in the lungs of IPF patients [29]. Our study showed a reduction in key autophagy genes, Beclin-1, Lc3a and Lc3b in Myr-Akt mouse lungs. Several lines of evidence indicate that autophagy, apoptosis, and senescence are tightly regulated to maintain cellular and organ homeostasis. For example, reduced apoptosis in lung myofibroblasts from human IPF samples results in dysregulated lung repair [59] correlating with the inhibition of autophagy in IPF lung tissue [30] [31] [47]. Importantly, AKT-induced mTOR activation is critical in reducing autophagy as rapamycin has been shown to induce autophagy [47]. We observed elevated mTOR activity in lungs of Myr-Akt mice suggesting decreased autophagy via the AKT/mTOR pathway. Furthermore, in our Myr-Akt murine model, constitutive AKT activation correlated with reduced autophagy gene expression, which was further reduced by bleomycin challenge. It is plausible that the combination of constitutively-active AKT and bleomycin creates a scenario where clearance of unwanted proteins and organelles is hindered, leading to reduced survival times in Myr-Akt mice. We are currently investigating if AKT activation in macrophages, and the observed changes in lung expression of autophagy markers, are directly attributable to changes in the macrophages themselves or if this AKT activity translates into alterations of other cell phenotypes.
This report provides links between M-CSF-R-induced AKT activity and M2 macrophage bias, myofibroblast marker expression, and decreased autophagy gene expression. We used a transgenic murine model to demonstrate a pro-fibrotic role for CSF1 and AKT in the lungs, and believe that these observations may be applicable to clinical lung disease, as we showed overexpressing AKT predisposes lungs to fibrosis in the presence of bleomycin. This prediction is based on our previous data that showed CSF1 is important in IPF, and we surmise that this report clarifies a potential mechanism for this effect as a priming induced by CSF1 to predispose lung fibrosis in humans.