Prevention and treatment of bleomycin-induced pulmonary fibrosis with the lactate dehydrogenase inhibitor gossypol

Abstract

Pulmonary fibrosis is a chronic and irreversible scarring disease in the lung with poor prognosis. Few therapies are available; therefore it is critical to identify new therapeutic targets. Our lab has previously identified the enzyme lactate dehydrogenase-A (LDHA) as a potential therapeutic target in pulmonary fibrosis. We found increases in LDHA protein and its metabolic product, lactate, in patients with idiopathic pulmonary fibrosis (IPF). Importantly, we described lactate as a novel pro-fibrotic mediator by acidifying the extracellular space, and activating latent transforming growth factor beta (TGF-β1) in a pH-dependent manner. We propose a pro-fibrotic feed-forward loop by which LDHA produces lactate, lactate decreases pH in the extracellular space and activates TGF-β1 which can further perpetuate fibrotic signaling. Our previous work also demonstrates that the LDHA inhibitor gossypol inhibits TGF-β1-induced myofibroblast differentiation and collagen production in vitro. Here, we employed a mouse model of bleomycin-induced pulmonary fibrosis to test whether gossypol inhibits pulmonary fibrosis in vivo. We found that gossypol dose-dependently inhibits bleomycin-induced collagen accumulation and TGF-β1 activation in mouse lungs when treatment is started on the same day as bleomycin administration. Importantly, gossypol was also effective at treating collagen accumulation when delayed 7 days following bleomycin. Our results demonstrate that inhibition of LDHA with the inhibitor gossypol is effective at both preventing and treating bleomycin-induced pulmonary fibrosis, and suggests that LDHA may be a potential therapeutic target for pulmonary fibrosis.

Fig 1. LDHA is increased in bleomycin-induced pulmonary fibrosis.

Mice were exposed to saline or bleomycin via oropharyngeal aspiration (OA) and euthanized at indicated time points. (A) Total RNA was isolated from the right lung lobes and qRT-PCR was performed to measure LDHA mRNA. Data are displayed as fold change from saline controls normalized to 18S. *p≤0.05 compared to saline controls by t-test. n = 4–5 per group. (B) LDH activity was measured on lung tissue homogenates in mice administered either saline or bleomycin. ***p<0.001. (C-D) Lung tissue from mice at day 21 post bleomycin was stained for LDHA by immuno-histochemistry in red. Images were taken at 20X magnification, scale bars represent 50 μm. (E-F) Serial section immunohistochemistry was performed for LDH shown in panel E and αSMA shown in panel F. Images were taken at 40X magnification. LDHA expression localized to both macrophages (bold arrows) and areas co-staining for αSMA (open arrows), a marker of myofibroblast differentiation.

Fig 2. Gossypol prevents bleomycin-induced pulmonary fibrosis.

(A) Mice were exposed to 1.5U/kg bleomycin via OA and treated daily with sub-cutaneous injections of the LDHA inhibitor gossypol at 5, 10, or 20 mg/kg until euthanasia at day 21. (B) Collagen levels in the right lung lobes were measured by hydroxyproline assay. Data are displayed as mean ± SEM. *p≤0.05 by ANOVA. n = 9–10 mice per group. (C) Fibrosis scores were performed on trichrome stained sections. Data are displayed as mean ± SEM. *p≤0.05 by ANOVA (D-H) Lung tissue sections were Trichrome stained for collagen fibers in blue. (D) saline + vehicle, (E) saline + gossypol 20mg/kg, (F) bleomycin + vehicle, (G) bleomycin + gossypol 20mg/kg, (H) bleomycin + gossypol 10mg/kg, (I) bleomycin + gossypol 5mg/kg. Images were taken at 20X magnification, scale bar represents 50 μm.

Fig 3. Gossypol prevents collagen gene expression.

Mice were exposed to bleomycin and treated with gossypol as indicated in Fig 2 and sacrificed on day 7. Collagen 1α1 and collagen 3α1 gene expression were measured by qRT-PCR. Data are displayed as fold changes from saline and vehicle treated controls normalized to 18S. Collagen 1α1 mRNA was measured on day 7 (A) and day 21 (B). Collagen 3α1 mRNA was measured on day 7(C) and day 21 (D). Data are displayed as mean ± SEM. *p≤0.05 by ANOVA. n = 3–9 mice per group.

Fig 4. Gossypol prevents TGF-β1 induced fibronectin expression.

Mice were exposed to bleomycin and treated with gossypol as indicated in Fig 2 and sacrificed on day 21. Fibronectin expression was measured in whole lung homogenates by Western blot (A) which was quantified using densitometry (B). Data are displayed as mean ± SEM. **p<0.01 by ANOVA. n = 8 mice per treatment group. (C-F) Lung tissue sections were stained for fibronectin in red and counterstained with hematoxylin. (C) Saline + vehicle, (D) saline + gossypol 20mg/kg, (E) Bleomycin + vehicle at (F), bleomycin + gossypol 20mg/kg. All images were taken at 20X magnification (G). Scale bars represent 50 μm. The inset in (C) is an isotype negative control stain.

Fig 5. Gossypol inhibits LDH activity and prevents TGF-β1 expression and activation.

Mice were exposed to bleomycin and treated with gossypol as indicated in Fig 2. (A) LDH Activity in whole lung homogenates was measured using a standardized protocol. Data are displayed as mean ± SEM. ***p<0.001 by ANOVA. n = 8 mice per treatment. (B) Lactate was measured in whole lung homogenates by mass spectroscopy. Data are displayed as mean ± SEM. ***p<0.001 and *p<0.05 by ANOVA. n = 8 mice per treatment. (C-D) Total TGF-β1 was measured in whole lung tissue homogenates at 7 days (C) and at 21 days (C). *p<0.5 by ANOVA. n = 8 mice per treatment. (E-I) Lung tissue sections were stained for active TGF-β1 by IHC in red and counterstained with hematoxylin. (E) saline + vehicle, (F) saline + gossypol 20mg/kg, (G) bleomycin + vehicle, (H) bleomycin + gossypol 20mg/kg. Antibody neutralization was performed by incubating the primary antibody with human recombinant TGF-β1 for 30 minutes prior to staining to confirm specificity for active TGF-β1 (I). Images were taken at 20X objective, scale bar represents 50 μm.

Fig 6. Gossypol does not alter inflammatory cell profiles in the lung.

Mice were exposed to bleomycin and treated with gossypol as indicated in Fig 2. Mice were sacrificed on day 7 and the left lung was lavaged with PBS and total cells in lavage fluid were counted (A). Cytospins were prepared, stained with Diff-Quik, and differential counts were performed for macrophages, lymphocytes, and polymorphonuclear (PMN) cells (B). Differential counts are expressed as percent of total cells. Data are displayed as mean ± SEM. *p≤0.05 by ANOVA. n = 5–10 mice.

Fig 7. Gossypol is an effective anti-fibrotic treatment.

Mice were exposed to 1.5 U/kg bleomycin by OA. (A) Starting at day 7 post-bleomycin, mice were dosed with 20 mg/kg gossypol or vehicle until day of sacrifice at 21 days. (B) Collagen levels in the right lung lobes were measured by hydroxyproline assay. Data are displayed as mean ± SEM. *p≤0.05 by ANOVA. n = 7–10 mice per treatment. (C) Fibrosis scoring was performed on trichrome stained lung tissue sections. (D-F) Collagen fibers were visualized by Trichrome stain in blue. (D) Saline + vehicle, (E) bleomycin + vehicle, (F) saline + gossypol 20mg/kg, (G) bleomycin + gossypol 20mg/kg. Images were taken at 20X objective, scale bar represents 100 μm.

Fig 8. Gossypol treats bleomycin-induced fibronectin expression.

Mice were exposed to bleomycin and treated with gossypol as indicated in Fig 7. Lung homogenates were probed for fibronectin protein by western blot and quantified using densitometry (A). One representative blot is shown, each lane represents one mouse (B). Data are displayed as mean ± SEM. *p≤0.05 by ANOVA. n = 3–6 mice per group. (C-H) Lung tissue sections were stained for fibronectin in red and counterstained with hematoxylin. (C) Saline + vehicle, (D) saline + gossypol 20mg/kg, Bleomycin + vehicle at 10X magnification (E) and 20X magnification (F), bleomycin + gossypol 20mg/kg at 10X magnification (G) and 20X magnification (H). Scale bars represent 50 μm. Inset in (C) is an isotype negative control stain.