ASK1 Regulates Bleomycin-induced Pulmonary Fibrosis

Abstract

Pulmonary fibrosis (PF) is an abnormal remodeling of cellular composition and extracellular matrix that results in histological and functional alterations in the lungs. Apoptosis signal-regulating kinase-1 (ASK1) is a member of the mitogen-activated protein (MAP) kinase family that is activated by oxidative stress and promotes inflammation and apoptosis. Here we show that bleomycin-induced PF is reduced in Ask1 knockout mice (Ask1^−/−) compared with wild-type (WT) mice, with improved survival and histological and functional parameters restored to basal levels. In WT mice, bleomycin caused activation of ASK1, p38, and extracellular signal-regulated kinase 1/2 (ERK1/2) in lung tissue, as well as changes in redox indicators (thioredoxin and heme-oxygenase-1), collagen content, and epithelial–mesenchymal transition markers (EMTs). These changes were largely restored toward untreated WT control levels in bleomycin-treated Ask1^−/− mice. We further investigated whether treatment of WT mice with an ASK1 inhibitor, selonsertib (GS-4997), during the fibrotic phase would attenuate the development of PF. We found that pharmacological inhibition of ASK1 reduced activation of ASK1, p38, and ERK1/2 and promoted the restoration of redox and EMT indicators, as well as improvements in histological parameters. Our results suggest that ASK1 plays a central role in the development of bleomycin-induced PF in mice via p38 and ERK1/2 signaling. Together, these data indicate a possible therapeutic target for PF that involves an ASK1/p38/ERK1/2 axis.

Pulmonary interstitial diseases consist of a heterogeneous group of pathological processes involving varying degrees of inflammation and fibrosis (1). Pulmonary fibrosis (PF) is a distinct type of chronic, progressive interstitial disease with a usually fatal outcome (2). There are currently only two therapeutic drugs for PF, pirfenidone and nintedanib, which have been shown to extend life but do not reverse the disease (3). PF can be classified into two general groups: 1) fibrosis that results from known etiology such as genetic mutations, side effects from drugs, inhalation of inorganic/organic agents including particulates, or exposure to radiation; and 2) fibrosis caused by unknown etiology when there is no natural cause or apparent causative agent (idiopathic pulmonary fibrosis [IPF]) (4). The precise mechanisms of the development of PF have not been fully elucidated but include a complex set of pathways involving genetic and environmental factors, aging, and immune responses (5). One contributing mechanism in the development of PF is repetitive injury to the lung epithelium caused by the overproduction of reactive oxygen species (ROS) and other mediators by accumulated inflammatory cells in the lower respiratory tract (6). Repetitive injury to epithelial cells and accumulation of myofibroblasts is thought to remodel the lung parenchyma leading to thickening of the alveolar septa as well as narrowing of the airways (1, 5, 7).

The administration of bleomycin to mice is one of the most commonly used animal models to study pulmonary fibrosis (8). Bleomycin induces epithelial damage and alveolar inflammation in response to DNA damage and increased production of ROS (9), followed by fibroproliferative responses that have been demonstrated to include activation of MAP kinases including p38 and ERK1/2 (10). Epithelial cell injury and apoptosis are recognized as early features in bleomycin-induced PF in mice (11).

Apoptosis signal-regulating kinase-1 (ASK1), also known as mitogen-activated protein (MAP) kinase-kinase-kinase (MAP3K5), is ubiquitously expressed and functions as a redox-sensitive regulator of both c-Jun kinase (JNK)- and p38-mediated inflammation and apoptosis (12). ASK1 is activated in response to different stimuli, including oxidative stress, TNF-α, Fas ligand, and endoplasmic reticulum stress (13). Previous studies have shown that pharmacological inhibition of ASK1 reduced the development of liver fibrosis in rats (14) and that genetic deletion of Ask1 in mice ameliorated renal fibrosis (15). Since ROS production promotes ASK1 activation, we hypothesized that bleomycin-induced ROS would lead to ASK1 activation and that this would lead to MAPK-dependent development of fibrosis. We demonstrated here that activation of ASK1 promotes bleomycin-induced PF in mice and that treatment with an ASK1 inhibitor during the fibrotic phase of the disease significantly reduced PF.

Methods

Mouse Procedures

Ask1 knockout mice (Ask1^−/−) were kindly provided by H. Ichijo (University of Tokyo), and C57BL/6J mice were purchased from The Jackson Laboratory. All mouse lines were housed with food/water ad libitum, and weight was recorded daily. Male mice between 3 and 6 months of age were anesthetized by intraperitoneal injection of xylazine (AnaSed, Akorn) and ketamine (Ketathesia, Henry Schein) and received intratracheal instillation of a single dose of 0.3 U/kg of bleomycin sulfate from Streptomyces verticillus (Sigma-Aldrich) dissolved in 50 μl sterile saline. Control mice received saline intratracheally. In other experiments, mice were treated daily with 20 mg/kg selonsertib (ethanol:water; 50:50; Selleck Chemicals) by oral gavage starting 8 days after bleomycin instillation. Control mice were given vehicle (ethanol:water) by oral gavage on the same schedule. At the experimental endpoint (22 d), lung function measurements were made using the flexiVent system (SCIREQ), followed by the collection of BAL fluid (BALF) and lung tissue. The Institutional Animal Care and Use Committee for the University of Kentucky approved all animal procedures.

Lung Function Measurements

Mice were anesthetized by intraperitoneal injection of xylazine and ketamine. After a tracheotomy, the mice were connected to the flexiVent system. The computer-controlled small animal instrument ventilated the mice with a tidal volume of 10 ml/kg at a frequency of 150 breaths/minute. Following two total lung capacity maneuvers on the flexiVent, we performed a pressure-volume loop perturbation in each individual subject for three acceptable recorded measurements, of which an average was calculated for compliance, inspiratory capacity, and hysteresis.

Cell Recovery, Tissue Processing, and Assessment of Fibrosis

To perform BAL, the lungs were inflated two times with 0.7 ml PBS + EDTA via an intratracheal cannula, and the resulting BAL was centrifuged once at 3,000 revolutions per minute, and the supernatant was stored at −80°C. Cytospin sections from BAL cells were stained with Kwik-Diff (Thermo Fisher Scientific) for differential counts. The right lung was tied off, and the left lung was fixed with 10% formalin for histological analysis by hematoxylin and eosin staining and phospho-ASK1 immunostaining. The right lung was subsequently stored at −80°C and later homogenized in mammalian protein extraction reagent (Thermo Fisher Scientific) with Halt protease and phosphatase inhibitors (Thermo Fisher Scientific) for downstream analysis. Tissue and BAL total proteins were assayed by the BCA method. Tissue fraction and Ashcroft score (16) were used to assess the pulmonary fibrotic changes in lung tissue sections. Two observers counted blinded sections from at least four individual experimental mice in each group. Whole lung and BALF collagen levels were determined by Sircol assay (Biocolor). For further information, see supplemental methods.

Statistics

Data are expressed as mean ± SEM. Statistical significance was determined using one-way ANOVA for grouped or two-way ANOVA for multivariate analysis, as appropriate, with Neuman-Keuls post hoc comparisons. For all experiments, P < 0.05 was considered statistically significant. Graph generation and statistical analyses were performed using GraphPad Prism software (GraphPad Software Inc.).

Results

Ask1 Deletion Improved Survival and Lung Function in Response to Bleomycin

To determine whether ASK1 promotes PF, we instilled WT and Ask1^−/− mice with either vehicle (50 μl saline) or bleomycin (0.3 U/kg) and collected samples on Day 22 (Figure 1A). With this dose of bleomycin, 60% of WT mice and 80% of Ask1^−/− mice survived to Day 22 (Figure 1B). Weight loss was greater in WT mice compared with Ask1^−/− mice with the recovery of weight by Day 22 (data not shown). To determine whether Ask1 deletion caused changes in pulmonary function in response to bleomycin, we measured respiratory system compliance, inspiratory capacity, and hysteresivity before sacrifice. As expected, bleomycin caused a decrease in compliance (Figure 1C) and inspiratory capacity (Figure 1D) and an increase in hysteresivity (Figure 1E) in WT mice. While bleomycin also caused a reduction in compliance in Ask1^−/− mice, it was not decreased to the same degree in WT mice. Bleomycin also induced a decrease in inspiratory capacity in Ask1^−/− mice. The deletion of Ask1 did not cause a significant change in hysteresivity after bleomycin treatment.

Figure 1.

Ask1 (apoptosis signal-regulating kinase-1) deletion improved survival and lung function and reduced the development of lung fibrosis in response to bleomycin. (A) Timeline of bleomycin-induced pulmonary fibrosis (PF) in mice. A single dose of bleomycin (0.3 U/kg) was given on Day 0, and mice were killed on Day 22. The sham groups received 50 μl of vehicle (saline intratracheally). (B) Survival curve for wild-type (WT) (red line) and apoptosis signal-regulating kinase 1 knockout mice (Ask1^−/−) (orange line) that received bleomycin. Survival for sham groups (WT or Ask1^−/−) was 100% (not shown). n = 24 for WT bleomycin + vehicle (bleo) and n = 17 for Ask1^−/− bleo. (C–E) The flexiVent system was used to measure parameters of lung function, including compliance (C), inspiratory capacity (D), and hysteresis (E). Mean ± SEM; two-way ANOVA; n ⩾ 5 for each group. Tissue fraction (F) was evaluated from 10 random fields from hematoxylin and eosin (H&E) sections, and the Ashcroft score (G) was evaluated from two lower magnification fields. Evaluation of the Ashcroft score was performed by two investigators counting blinded sections. Mean ± SEM; two-way ANOVA; n ⩾ 5 for each group. (H and I) Collagen was measured in BAL fluid (BALF) (H) and tissue homogenates (I) using the Sircol method. (J) Representative images of H&E from lungs recovered from WT and Ask1^−/− mice receiving bleo or saline (sham). Scale bars, 500 μm. Whole lung images can be found in supplemental Figure E3. (K) Representative images of p-ASK1 immunostaining from lungs from WT and Ask1^−/− receiving bleo or saline. Scale bars, 100 μm. Mean ± SEM; two-way ANOVA; n ⩾ 6 for each group. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001. PV = pressure–volume.

Ask1 Deletion Reduced Lung Fibrosis in Response to Bleomycin

To assess the degree of PF, histological sections were quantified by tissue fraction analysis (fraction of lung tissue versus alveolar space) and Ashcroft scoring. As shown in Figure 1J, we observed that WT mice treated with bleomycin exhibited parenchyma rich in fibrous/scar tissue, while the Ask1^−/− mice appeared histologically more similar to the control group. Morphometric analysis showed that both WT and Ask1^−/− mice had an increase in tissue fraction with bleomycin treatment compared with their respective controls, but the increase in Ask1^−/− was significantly less than in WT mice (Figure 1F). The Ashcroft score was significantly increased in both WT and Ask1^−/− mice treated with bleomycin, but the score was significantly lower in Ask1^−/− mice (Figure 1G). To further confirm the development of PF, we measured collagen content in both BALF and lung tissue using the Sircol assay (Figures 1H and 1I, respectively). We observed an increase in collagen content in the lungs and BALF of WT and Ask1^−/− mice when compared with the respective control groups. However, the increase in collagen content was lower in the Ask1^−/− mice when compared with WT. To determine ASK1 activation, we performed immunostaining for phosphorylation of ASK1 at Thr845 (phospho-ASK1 [p-ASK1]), a classic indication of activation of the kinase. As shown in Figure 1K, there was relatively weak staining for p-ASK1 in epithelial cells in the lung parenchyma of control mice, but there was stronger staining in fibrotic tissue and epithelial cells in bleomycin-treated mice. As expected, there was no p-ASK1 staining in Ask1^−/− mice. These data confirm that p-ASK1 was increased in fibrotic tissue in bleomycin-treated mice and that deletion of Ask1 reduced the development of fibrotic tissue in response to bleomycin.

Ask1 Deletion Reduced Lymphocytes and Total Protein in BALF

To evaluate inflammatory responses and barrier function, we performed cell counts and measured protein in the BALF (Table 1). Bleomycin promoted a significant increase in the total number of cells and the number of macrophages in both WT and Ask1^−/− mice, and there were significantly more total cells in WT compared with Ask1^−/− treated with bleomycin. There was also a significant increase in lymphocytes in both WT and Ask1^−/− mice instilled with bleomycin, but significantly fewer lymphocytes were detected in the Ask1^−/− mice when compared with WT. These results suggest that lymphocytes may contribute to the development of PF and that ASK1 signaling impacts this response (17). Previous reports suggest a reduced role of neutrophils in PF (6, 18), and our results confirm this observation since very few neutrophils were found in BALF. BALF protein content was increased in both WT and Ask1^−/− in response to bleomycin, but Ask1^−/− mice had significantly less protein.

Table 1.

Inflammatory Cells and Total Protein in BAL Fluid

	WT Sham	Ask1−/− Sham	WT Bleo	Ask1−/− Bleo
Total cells	111 ± 12	139 ± 6	818 ± 64****	614 ± 41####,++
Macrophages	106 ± 12	138 ± 6	327 ± 50***	298 ± 24#
Lymphocytes	5 ± 1	3 ± 1	462 ± 63****	299 ± 33####,+
Neutrophils	0 ± 0	0 ± 0	27 ± 8**	15 ± 4
Total protein	136 ± 6	135 ± 9	1,818 ± 216****	823 ± 116#,++++

Definition of abbreviations: ASK1 = apoptosis signal-regulating kinase-1; bleo = bleomycin sulfate; WT = wild type.

Cell numbers ×10³/ml. Protein concentration is μg/ml. n = 8 (at least). Wild-type (WT) sham: WT control mice receiving 50 μl saline (intratracheally). Apoptosis signal-regulating kinase-1 (Ask1)^−/− sham: Ask1 knockout (KO) mice (Ask1^−/−) receiving 50 μl saline (intratracheally). WT bleomycin (bleo): WT mice receiving 50 μl bleo sulfate 0.3 U/kg (intratracheally). Ask1^−/− bleo: Ask1^−/− mice receiving 50 μl bleomycin sulfate 0.3 U/kg (intratracheally). Data are expressed as mean ± SEM. Statistical significance was determined using two-way ANOVA. **P < 0.01, ***P < 0.001, and ****P < 0.0001 when compared to the WT sham; ^#P < 0.01 and ^####P < 0.0001 when compared to the Ask1^−/− sham; ⁺P < 0.05, ⁺⁺P < 0.01, and ⁺⁺⁺⁺P < 0.0001 when compared to the WT bleo. See Figure E1 for cytospin images.

Bleomycin Caused Activation of ASK1 and p38 and Increased Expression of Trx1 and HO1

We next investigated ASK1 signaling in fibrotic tissue by examining expression in whole lung homogenates. As shown in Figure 2A, treatment of WT mice with bleomycin caused a significant increase in phosphorylation of ASK1 at Thr845 (p-ASK1) relative to total ASK1. As expected, there was no ASK1 or p-ASK1 in Ask1^−/− mice. ASK1 is known to activate both p38 and JNK MAP kinases in a tissue-specific manner (12, 19). Figure 2B demonstrates that p38 activation, as indicated by phosphorylation (p-p38), was significantly increased in WT mice treated with bleomycin relative to sham treatment. However, p38 phosphorylation was significantly reduced in Ask1^−/− mice in both sham and bleomycin-treated mice. There was no significant variation in total p38 expression in either strain. Phosphorylated JNK was significantly reduced in both WT and Ask1^−/− mice treated with bleomycin, and there did not appear to be an ASK1-dependent effect on JNK activation (Figure 2C).

Figure 2.

Ask1 deletion reduced mitogen-activated protein (MAP) kinase signaling and altered antioxidant responses, epithelial–mesenchymal transition (EMT) markers, and survival pathways following bleomycin treatment. Lung tissue was collected from sham and bleomycin-treated (0.3 U/kg; Day 22) WT and Ask1^−/− mice. Lung homogenates were then processed for western blotting for (A) phospho-ASK1 (p-ASK1) and total ASK1, (B) phospho-p38 and total p38, (C) c-Jun kinase (JNK) and total JNK, (D) thioredoxin 1 (Trx1), (E) heme-oxygenase-1 (HO1), (F) α-smooth muscle actin (SMA), (G) E-cadherin (E-cad), (H) collagen-1A1 (Col1A1), (I) pro-surfactant protein C (pro-SPC), (J) p-extracellular signal-related kinase (ERK)1/2 and total ERK1/2, (K) survivin, and (L) early growth response 1 (Egr1). Representative western blots are shown. Densitometry values for phosphorylated proteins were normalized to their respective total proteins, while other proteins were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and/or β-actin. These values were then further normalized to the sham WT value. In some cases, multiple proteins were evaluated from the same blot, and the loading control from that blot was then used for more than one protein (indicated by +, *). Mean ± SEM; two-way ANOVA; n ⩾ 3 for each group. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Bleomycin is known to promote the production of oxygen free radicals by stimulating mitochondrial production of ROS by increasing the expression of nicotinamide adenine dinucleotide phosphate oxidase-4 and by increasing inflammatory cells that also produce ROS (10, 18). An increase in ROS causes activation of ASK1 by dissociating it from thioredoxin (Trx1), which is normally bound to the N-terminal portion of ASK1 in the inactive state (12, 20). We observed that expression of Trx1 was significantly increased in WT mice treated with bleomycin (Figure 2D), possibly as compensation for the increase in ASK1 activity. Interestingly, Trx1 expression was significantly higher in untreated Ask1^−/− mice compared with WT, and there was a small reduction following bleomycin treatment. We speculate that Trx1 is upregulated in Ask1^−/− control mice due to the absence of its binding protein ASK1 (20). As a further indirect indication of oxidative stress, we measured the expression of heme-oxygenase-1 (HO1) in lung homogenates (Figure 2E). Although HO1 is a phase II enzyme of antioxidant response elements (similar to Trx1), we observed that its expression was increased in WT mice stimulated with bleomycin when compared with control mice. However, HO1 expression was unchanged in Ask1^−/− mice following bleomycin treatment. Since both Trx1 and HO1 were significantly increased in WT mice, these results suggest that the expression of these antioxidant response elements may be a compensatory mechanism following bleomycin stimulus and the development of PF.

Ask1 Deletion Reduced Epithelial–Mesenchymal Transition Markers, Collagen Content, and Pro-Surfactant Protein C in Response to Bleomycin

The development of PF typically involves increased expression of mesenchymal markers such as α-smooth muscle actin (α-SMA) and loss of epithelial markers such as E-cadherin (21, 22). Epithelial–mesenchymal transition (EMT) is a progressive alteration in epithelial cell morphology to a mesenchymal phenotype, and during this process, there is a reduction in cell junctions parallel with an increase in α-SMA that facilitates cell mobilization (23). As expected, following bleomycin treatment, we observed increased expression of α-SMA (Figure 2F) and decreased expression of E-cadherin (Figure 2G) in WT lung homogenates. Although we also observed a reduction in E-cadherin in bleomycin-treated Ask1^−/− mice, α-SMA expression in Ask1^−/− mice was similar to controls. When we compared collagen-1A1 content, the increase in WT mice was significantly greater than that in Ask1^−/− mice (Figure 2H). Since it has previously been shown that PF can cause an increase in alveolar type II (ATII) cells (22), we examined pro-surfactant protein C (pro-SPC) expression. As shown in Figure 2I, bleomycin caused an increase in pro-SPC expression in WT mice, but this response was significantly reduced in Ask1^−/− mice. Interestingly, the expression of pro-SPC was significantly higher in untreated Ask1^−/− mice compared with WT controls, suggesting a potential increase in ATII cells.

Ask1 Deletion Reduced ERK1/2 Activation and Influenced Survival Pathways

ERK1/2 is another member of the MAP kinase family that is involved in cell proliferation and prosurvival pathways, and previous studies have indicated a role in EMT and the development of fibrosis (10, 12, 24). While there is also evidence that this involves inhibition of apoptosis (25), there are no previous reports that link ASK1 to ERK1/2 signaling. When we measured ERK1/2 phosphorylation in response to bleomycin, we found that ERK1/2 activation was significantly increased in WT mice, but there was no increase in activation in Ask1^−/− mice (Figure 2J). These results suggest a potential regulation of ERK1/2 activity by ASK1 in response to bleomycin.

To further investigate prosurvival pathways, we investigated the expression of the survivin protein, which is a member of the apoptosis inhibiting protein family that inhibits caspase-3 (26). We found that while expression of survivin was increased in WT mice treated with bleomycin, this increase was prevented in Ask1^−/− mice (Figure 2K). These results suggest that apoptosis was reduced following bleomycin instillation. Survivin expression can be upregulated by activation of ERK1/2, and we observed a concomitant increase in phospho-ERK1/2 and survivin following bleomycin treatment in WT mice. There was no increase in phospho-ERK1/2 in Ask1^−/− mice, nor in survivin. Survivin is negatively regulated by early growth response protein 1 (Egr1), a transcription factor that regulates cell differentiation and mitogenesis (11). Figure 2L shows that Egr1 expression was significantly reduced in WT mice treated with bleomycin, consistent with the increased expression of survivin. Baseline expression of Egr1 was lower in untreated Ask1^−/− compared with WT, and there was no significant change following bleomycin. These results suggest that ASK1 may be partially regulating Egr1 expression.

ASK1 Inhibition during the Fibrotic Phase Increased Survival after Bleomycin

Instillation of bleomycin initially causes acute lung injury, which is then followed by the development of fibrosis (8, 27). The acute lung injury phase has been reported up to approximately Day 7, while the fibrotic phase has been reported to start around Day 8 and progress up to Day 21 (21, 28). Deletion of Ask1 may have resulted in reduced lung injury during the early inflammatory stages in response to bleomycin, as we have shown in a LPS-induced lung injury model (29), and this may have contributed to the reduced fibrosis we observed. To examine whether ASK1 contributes to the development of fibrosis independent of its role in the lung injury phase, we treated WT mice with an ASK1 inhibitor, selonsertib (GS-4997), daily by oral gavage beginning on Day 8 following bleomycin. This approach is outlined in Figure 3A. Figure 3B shows that treatment with selonsertib improved the survival of mice instilled with bleomycin compared with mice instilled with bleomycin and treated with vehicle (ethanol). When lung function measurements were obtained on Day 22, we observed that compliance was reduced in mice instilled with bleomycin regardless of whether or not they were treated with selonsertib (Figure 3C). There were no significant changes in either inspiratory capacity or hysteresivity (data not shown) in any group. These results demonstrate that selonsertib treatment improved survival despite the lack of an improvement in lung function.

Figure 3.

ASK1 inhibition improved survival and lung function and reduced the development of lung fibrosis in response to bleomycin. (A) Timeline of bleomycin-induced PF in mice. A single dose of bleomycin (0.3 U/kg in saline) was given on Day 0, and mice were killed on Day 22. The control group received 50 μl of saline (intratracheally). Selonsertib (sel) treatment (20 mg/kg in ethanol daily by oral gavage) was from Day 8 until Day 22 in bleomycin-treated mice (bleo + sel). Control and bleo groups received the vehicle, a mix of ethanol:water (50:50), daily by oral gavage (100 μl) using the same timeline as the selonsertib (bleo + sel)-treated group. (B) Survival for mice that received bleomycin and were treated with vehicle or selonsertib. Survival for the control group (saline + vehicle) was 100% (not shown). n = 15 for bleomycin (bleo) and n = 14 for bleo + sel. The flexiVent system was used to measure compliance (C). Tissue fraction (D) was evaluated from 10 random fields from H&E sections, and the Ashcroft score (E) was evaluated from two lower magnification fields. Evaluation of the Ashcroft score was performed by two investigators counting blinded sections. Mean ± SEM; two-way ANOVA; n ⩾ 4 for each group. Collagen was measured in BALF (F) and tissue homogenates (G) using the Sircol method. (H) Representative images of H&E from lungs recovered from mice receiving saline + vehicle (control), bleomycin + vehicle (bleo), or bleomycin + selonsertib (bleo + sel). Scale bars, 500 μm. (I) Representative images of p-ASK1 immunostaining from lungs recovered from mice receiving control, bleo, or bleo + sel. Scale bars, 100 μm. Whole lung images can be found in Figure E4. Mean ± SEM; one-way ANOVA n ⩾  for each group. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

ASK1 Inhibition Reduced PF

When we examined the lungs from mice treated with selonsertib during the fibrotic phase following bleomycin, there was a substantial improvement compared with mice treated with bleomycin + vehicle. Figure 3H shows that selonsertib treatment greatly attenuated fibrosis, as indicated by histological sections. There was some evidence of thickening of the alveolar septa in selonsertib-treated mice but not to the same degree as in bleomycin + vehicle mice. Measurements of lung tissue fraction and Ashcroft scoring confirmed the histological observations (Figures 3D and 3E). Bleomycin + vehicle mice had a significant increase in both tissue fraction and Ashcroft score compared with controls, but selonsertib treatment significantly reduced both parameters. Measurements of collagen in both BALF and lung tissue were also significantly reduced in mice instilled with bleomycin and treated with selonsertib compared with bleomycin + vehicle (Figures 3F and 3G, respectively). Treatment with selonsertib following bleomycin caused reduced expression of p-ASK1 compared with bleomycin alone (Figure 3I). The control group that received the vehicle had only a weak expression of p-ASK1.

ASK1 Inhibition Reduced the Number of Cells and Total Protein in BAL

Similar to the results in Table 1, bleomycin caused a significant increase in total cells in BALF in the vehicle-treated group compared with controls, including increases in macrophages and lymphocytes (Table 2). Total protein was also increased. Interestingly, treatment with selonsertib during the fibrotic phase after bleomycin was effective at reducing BAL cell numbers as well as total protein.

Table 2.

Inflammatory Cells and Total Protein in BAL Fluid

	Control Veh	Bleo Veh	Bleo Sel
Total cells	1.34 ± 7	925 ± 74****	474 ± 35**,####
Macrophages	127 ± 8	372 ± 39***	231 ± 28#
Lymphocytes	2 ±	500 ± 71****	225 ± 36##
Neutrophils	0 ± 0	10 ± 5	11 ± 4
Total protein	116 ± 8	1,037 ± 141***	590 ± 92*,#

Definition of abbreviations: sel = selonsertib; veh = vehicle.

Cell numbers ×10³/ml. Protein concentration is μg/ml. n = 10 (at least). Control vehicle: mice receiving 50 μl saline (intratracheally) at Day 0 and daily veh (ethanol:water; 50:50) orally from Day 8 until Day 21. Bleo veh: mice receiving 50 μl bleo sulfate 0.3 U/kg (intratracheally) at Day 0 and daily veh (ethanol:water; 50:50) orally from Day 8 until Day 21. Bleo sel: mice receiving 50 μl bleo sulfate 0.3 U/kg (intratracheally) at Day 0 and daily selonsertib (20 mg/kg) orally from Day 8 until Day 21. Data are expressed as mean ± SEM. Statistical significance was determined using one-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 when compared to the Control veh; ^#P < 0.05, ^##P < 0.01, and ^####P < 0.0001 when compared to the Bleo veh. See Figure E2 for cytospin images.

ASK1 Inhibition Decreased p38 Activation and Trx1 Expression

When WT mice instilled with bleomycin were treated with selonsertib during the development of fibrosis (beginning on Day 8), we confirmed that ASK1 activation was significantly decreased in lung tissue on Day 22 (Figure 4A). Similarly, p38 activation was also decreased in selonsertib-treated mice compared with vehicle treatment (Figure 4B). Expression of both Trx1 and HO1 were increased in bleomycin/vehicle-treated mice, but selonsertib treatment prevented the increased expression of Trx1 and HO1 (Figures 4C and 4D). These results suggest that inhibition of ASK1 during the development of fibrosis impacted Trx1 and HO1 responses.

Figure 4.

Inhibition of ASK1 with selonsertib reduced MAP kinase signaling and altered antioxidant responses, EMT markers, and survival pathways following bleomycin treatment. Lung tissue was collected from control, bleomycin-treated (0.3 U/kg; Day 22), and bleomycin + selonsertib-treated WT mice. Lung homogenates were then processed for western blotting for (A) p-ASK1 and total ASK1, (B) phospho-p38 and total p38, (C) Trx1, (D) HO1, (E) α-SMA, (F) E-cadherin, (G) Col1A1, (H) pro-SPC, (I) p-ERK1/2 and total ERK1/2, (J) survivin, and (K) Egr1. Representative western blots are shown. Densitometry values for phosphorylated proteins were normalized to their respective total proteins, while other proteins were normalized to GAPDH and/or β-actin. In some cases, multiple proteins were evaluated from the same blot, and the loading control from that blot was then used for more than one protein (indicated by *). Mean ± SEM; one-way ANOVA n ⩾ 3. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

ASK1 Inhibition Reduced EMT Markers, Collagen Content, and pro-SPC in Response to Bleomycin

We next investigated whether inhibition of ASK1 during the fibrotic stage would impact EMT markers and other changes associated with fibrosis. Bleomycin/vehicle-treated mice demonstrated an increase in α-SMA expression, but there was no increase in mice treated with selonsertib (Figure 4E). However, as shown in Figure 4F, there were no significant changes in E-cadherin expression in mice treated with bleomycin and then either vehicle or selonsertib. There was also a reduction in COL1A and pro-SPC expression in selonsertib-treated mice compared with vehicle-treatment (Figures 4G and 4H, respectively).

ASK1 Inhibition Altered Survival Pathways

Similar to the measurements with Ask1^−/− mice, inhibition of ASK1 during fibrosis prevented the increase in activation of ERK-1/2 observed in bleomycin-vehicle-treated mice (Figure 4I). In addition, the increased expression of survivin in response to bleomycin was reduced by selonsertib treatment (Figure 4J), while the decrease in expression of Egr1 was prevented by selonsertib (Figure 4K).

ASK1 (MAP3K5) Expression is Altered in Human Lung Fibrosis

To further investigate the role of ASK1 in pulmonary fibrosis, we mined the Idiopathic Pulmonary Fibrosis Cell Atlas database (https://p2med.shinyapps.io/IPFCellAtlas/) to evaluate the expression of MAP3K5 in different cell populations in lungs from IPF patients and chronic obstructive pulmonary disease patients compared with control lungs (30). As shown in Figure 5, expression of MAP3K5 was increased in several populations of epithelial cells (including ATII, basal, and club cells), immune cells (macrophages and T cells), and stromal cells (fibroblasts, pericytes). While gene expression was increased in some of these cell populations, it is unknown whether activation of MAP3K5 was increased.

Figure 5.

ASK1 (MAP3K5) is highly expressed in different cell populations from idiopathic pulmonary fibrosis (IPF) and chronic pulmonary obstructive disease (COPD) lungs. We analyzed a published database of lung cell distribution in patients with COPD, IPF, and control donors (https://p2med.shinyapps.io/IPFCellAtlas/). ASK1 (MAP3K5) gene expression (right column of images) was displayed from dark blue (zero expression) to yellow (high expression). There was an increased expression in IPF cells for epithelial (basal, goblet, ionocyte, club, and alveolar type II cells [AT2]), stromal (fibroblasts, pericytes, and smooth muscle), and immune cells (macrophages, B cells, alveolar macrophages, natural killer [NK] cells, T [CD] 4 and 8).

Discussion

Pulmonary fibrosis is a devastating disease for which there are limited treatment options. While many interventions have been proposed based upon extensive preclinical research, there has been little change in clinical outcomes (3, 10, 18, 31). In the current study, we demonstrated that ASK1 contributes to the development of PF in bleomycin-treated mice and that pharmacological inhibition of ASK1 with selonsertib partially abrogated fibrosis. While 40% of WT mice did not survive the treatment, the remaining mice exhibited the hallmark features of PF by Day 22 (Figures 1 and 3). There was a clear improvement in survival in Ask1^−/− mice instilled with bleomycin, and the functional and histological responses indicative of fibrosis were substantially reduced compared with WT mice. To confirm that ASK1 signaling was important during the development of PF, we treated bleomycin-injured WT mice with selonsertib only during the fibrotic stage. Inhibition of ASK1 also reduced the development of functional and histological changes that are normally seen in PF.

Bleomycin is known to increase the production of ROS in the lungs of mice (17, 21, 32). Since ROS causes dissociation of ASK1 from Trx1 and subsequent activation, we hypothesized that the persistent ROS production caused by bleomycin would promote ASK1 signaling, which was confirmed in lung homogenates (Figures 2A and 4A). Classically, activation of ASK1 has been demonstrated to cause subsequent activation of p38 (12, 13, 15, 19, 20, 33) and JNK in a tissue-specific manner (12, 14, 19, 20, 34). We demonstrated that both ASK1 and p38 activation were stimulated by bleomycin, but this was not observed in either Ask1^−/− mice or in selonsertib-treated WT mice (Figures 2A, 2B, 4A, and 4B). We did not observe ASK1-dependent activation of JNK (Figure 2C), but p-JNK was significantly reduced in WT mice following bleomycin.

One of the novel findings in this study is that both deletion and inhibition of ASK1 prevented bleomycin-induced activation of ERK-1/2 (Figures 2J and 4I). Previous studies involving cultured cells (such as RAW 264.7, human umbilical vein endothelial cells, and vascular smooth muscle cells) suggested that ASK1 did not influence ERK-1/2 activity (35–38). However, we demonstrated that ASK1 promoted ERK-1/2 activation since both genetic deletion and pharmacological inhibition with selonsertib prevented activation in response to bleomycin. Increased activation of ERK-1/2 signaling in PF has been previously reported (24, 39), but this is the first demonstration of the regulation of ERK-1/2 signaling by ASK1. There are known tissue- and cell–type-specific differences in the role of ERK-1/2 in survival versus apoptosis (40). ERK-1/2 activation is generally thought to promote prosurvival pathways through increased transcription of genes such as Nrf2 and activator protein-1 (AP1)-transcription factor (Jun), which inhibit apoptotic and necrotic pathways (41). Thus, it is possible that the reduced activation of ERK-1/2 with ASK1 deletion/inhibition caused a decrease in cell survival and reduced fibrosis (12, 37, 38). Our results demonstrating ASK1-dependent regulation of ERK1/2 are robust, but we have not identified a direct signaling pathway linking the two, and further studies are needed to identify the cell populations in which this signaling occurs. We hypothesize that fibroblast/myofibroblast populations may exhibit increased cell survival and proliferation, and studies examining ASK1-mediated effects on proliferation/survival in both cell culture and in vivo models will help to address this question. However, we cannot rule out that populations of ATII cells and macrophages may also have ASK1-mediated cell survival or possibly cell death. ATII cells are known to undergo apoptosis in the early stages following bleomycin-induced injury, and it is possible that this is mediated via ASK1. If this is the case, then inhibition or deletion of ASK1 could reduce the epithelial injury that occurs early.

We also investigated other survival pathway indicators. Survivin is a member of the apoptosis-inhibiting protein family, abundantly expressed during embryogenesis and in tumor tissues and virtually nonexistent in normal adult tissues, except in highly proliferative tissues (26). Survivin expression is downregulated during neoplasms by several molecular mechanisms, including increased proliferative activity, migration, and cell differentiation (25). We postulated that activation of MAP kinases during the development of fibrosis would promote the expression of survivin as part of the inhibition of the apoptotic response. From lung homogenates, we found that survivin was, in fact, increased in WT mice instilled with bleomycin (Figure 2K), but it was not increased in Ask1^−/− mice. Inhibition of ASK1 during the fibrotic phase reduced survivin expression following bleomycin (Figure 4J), corroborating the finding in Ask1^−/− mice. These results support the hypothesis that deletion or inhibition of ASK1 causes downregulation of apoptosis. Our results do not definitively demonstrate a causal relationship between MAP kinase signaling and survivin, but previous studies have suggested that ERK-1/2 positively regulates the expression of survivin during processes involving cell proliferation, migration, and differentiation (42, 43). Another regulator of survivin expression, the transcription factor Egr1 (11), was decreased in lung homogenates in response to bleomycin (Figure 2L). Interestingly, Egr1 expression was also significantly reduced in control Ask1^−/− mice compared with WT, and there was no change in expression following bleomycin treatment. Thus, for Ask1^−/− mice, there did not appear to be a direct correlation between the expression of survivin and the expression of Egr1, and this may suggest an alternative regulator for survivin in response to bleomycin. However, it should also be noted that both the deletion of Ask1 and treatment with selonsertib influenced the expression of survivin, which suggests a possible interaction between MAP kinases and survivin without the involvement of Egr1. A limitation of these studies is that our measurements from whole lung homogenate cannot identify the specific cell populations in which expression and changes in activity occurred. However, immunostaining for p-ASK1 in Figures 1K and 3I suggest increased expression in bronchial and alveolar epithelial cells as well as in fibrotic regions. Colocalization studies are needed to confirm the specific cell types. Figure 5 indicates that there are heterogeneous changes in MAP3K5 expression in different cell populations in IPF lungs compared with control lungs, but this does not address changes in phosphorylation/activation of ASK1.

pro-SPC functions in pulmonary homeostasis and plays a role in lung injury and repair (44, 45). Familial interstitial lung diseases have been associated with deficiencies in surfactant proteins and mutations in the genes for SFTPC and SFTPA2 (46). Because previous studies had shown an increase in ATII cells (22, 47), we measured pro-SPC in whole lung homogenates and found a significant increase in pro-SPC in WT mice treated with bleomycin (Figure 2I and 4H). This increase was partially reduced in Ask1^−/− mice and selonsertib-treated WT mice, potentially suggesting a reduction in ATII cell hyperplasia. Interestingly, pro-SPC expression was significantly increased in control Ask1^−/− mice compared with control WT. This could be due to increased ATII cells in Ask1^−/− mice, but we have not yet confirmed this. However, this suggests a potentially novel role for ASK1 in the regulation of ATII cell populations or expression of pro-SPC.

Because of the increased production of ROS in response to bleomycin, it is expected that there will be an increased production of antioxidant enzymes to maintain the redox balance between oxidants and antioxidants (6). However, if the production of ROS is exaggerated and persistent, this can lead to the depletion of redox sensors (12, 20, 38). We also expected a reduction in the expression of the antioxidant enzymes Trx1 and HO1, which are phase II enzymes that generally respond proportionally as part of the antioxidant response element family (41, 48). The deletion of Ask1 prevented the increase in Trx1 and HO1 expression (Figures 2D and 2E, respectively), and treatment with selonsertib prevented the increase in both as well (Figures 4C and 4D). These antioxidant enzymes may be regulated by other factors; for example, Trx1 expression may be linked to ASK1 given its regulatory role in ASK1 activity. This could, in part, explain the increase in Trx1 in mice instilled with bleomycin, but this response may also reflect changes in cellular composition. While the expression of HO1 and Trx1 are indicative of oxidative stress, they are not actual measures of oxidative stress, and further analysis is required to identify how these pathways are regulated by ASK1. In addition, changes in the expression of Trx1 may not be as important as its redox status in terms of regulating ASK1 activation (12, 20). However, we did not investigate the redox status of Trx1 in the current study.

The role of ASK1 in the development of fibrosis in other organs is unresolved. As mentioned above, reduction of ASK1 activity reduced liver fibrosis in rats (14) and renal fibrosis in mice (15). However, another study demonstrated that both genetic deletion and pharmacologic inhibition of ASK1 promoted the development of fibrosis in a mouse model of nonalcoholic steatohepatitis (NASH) through mechanisms involving reduced autophagy (49). These authors suggested that differences may be associated with different experimental models involving a high-fat diet, the specificity of ASK1 expression in the liver, or the use of different ASK1 inhibitors. Selonsertib was recently used in phase III clinical trials in an attempt to reduce fibrosis in patients with NASH and advanced liver scarring (50). While selonsertib was found to be safe for patients, and there was evidence for inhibition of ASK1 and some improvement in clinical measures, there was no efficacy observed in terms of disease progression or reduction of fibrosis. However, these studies are promising for the potential use of selonsertib for PF. It is interesting to note that we observed improved survival in selonsertib-treated mice, but lung function was not improved in these mice, unlike the Ask1^−/− mice. While morphological endpoints were improved by treatment with selonsertib, there was no improvement in inspiratory capacity or hysteresivity (data not shown). These differences are likely due to the fact that ASK1 was present and presumably active during the early inflammatory stage in WT mice treated with selonsertib, but there was no ASK1 present during this stage in the Ask1^−/− mice. The Ask1^−/− mice likely had reduced initial lung injury and inflammation. However, by treating with selonsertib beginning on Day 8, our selonsertib studies were designed to specifically investigate its impact on the fibrotic stage of the disease.

Our results reveal that ASK1 plays a central role in the development of bleomycin-induced PF in mice. Importantly, ASK1 appears to regulate both p38 and ERK-1/2 signaling during the development of PF. Treatment with selonsertib during the fibrotic phase of the disease ameliorated the development of PF similar to the reduction shown in Ask1^−/− mice. Together, these data indicate a possible therapeutic target for PF that involves ASK1/p38/ERK-1/2 signaling.