Nicotinamide N-methyltransferase mediates lipofibroblast-myofibroblast transition and apoptosis resistance

Abstract

Metabolism controls cellular phenotype and fate. In this report, we demonstrate that nicotinamide N-methyltransferase (NNMT), a metabolic enzyme that regulates developmental stem cell transitions and tumor progression, is highly expressed in human idiopathic pulmonary fibrosis (IPF) lungs, and is induced by the pro-fibrotic cytokine, transforming growth factor-β1 (TGF-β1) in lung fibroblasts. NNMT silencing reduces the expression of extracellular matrix proteins, both constitutively and in response to TGF-β1. Furthermore, NNMT controls the phenotypic transition from homeostatic, pro-regenerative lipofibroblasts to pro-fibrotic myofibroblasts. This effect of NNMT is mediated, in part, by the downregulation of lipogenic transcription factors, TCF21 and PPARγ, and the induction of a less proliferative but more differentiated myofibroblast phenotype. NNMT confers an apoptosis-resistant phenotype to myofibroblasts that is associated with the downregulation of pro-apoptotic members of the Bcl-2 family, including Bim and PUMA. Together, these studies indicate a critical role for NNMT in the metabolic reprogramming of fibroblasts to a pro-fibrotic and apoptosis-resistant phenotype and support the concept that targeting this enzyme may promote regenerative responses in chronic fibrotic disorders such as IPF.

Keywords: apoptosis; fibrosis; lipofibroblasts; metabolism; myofibroblasts; nicotinamide N-methyltransferase.

Figure 1

NNMT expression is elevated in myofibroblasts of IPF lung tissues and TGF-β1 stimulated human lung fibroblasts.A–C, expression of NNMT transcript in myofibroblasts of control and IPF/ILD lungs in single-cell RNA-seq analysis obtained from different datasets (A) Kaminski/Rosas, (B) Banovich/Kropski, and (C) Lafyatis groups from IPF Cell Atlas. D, UMAP plot showing NNMT transcript expression in different mesenchymal cells of control and IPF lungs in single cell RNA-seq analysis obtained from Banovich-Kropski dataset of IPF Cell Atlas. E, violin plot showing NNMT mRNA in Gene Expression Affymetrix data (GSE150910) of control donor and IPF human lungs. Data presented as means ± s.e.m., n = 102 to 103, ∗∗∗∗p < 0.0001 (Student’s t test). F, mRNA expression of NNMT in Digital Spatial Profiling (DSP) representing six Regions of Interest (ROIs) from control donor and individual with IPF; n = 6, representing one individual. The data are presented as means ± s.e.m., ∗∗∗p = 0.0005 (Student’s t test). G, NNMT mRNA in Gene Expression Microarray of human fibroblasts treated with vehicle or TGF-β1 (2.5 ng/ml) for 24 h. Data presented as means ± s.e.m., n = 3, ∗p = 0.0171 (Student’s t test). H, expression of NNMT mRNA in human fibroblasts treated with or without TGF-β1 (2.5 ng/ml) for indicated time points up to 48 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA. I and J, Western blot and quantitative analysis of NNMT protein in human fibroblasts treated with TGF-β1 (0 or 2.5 ng/ml) for 48 h. Data presented as means ± s.e.m., n = 3, ∗∗∗p = 0.0006 (Student’s t test). K, Western blot demonstrating expression of NNMT and α-SMA at indicated time points in TGF-β1 (2.5 ng/ml) treatment of human fibroblasts (IMR-90). L, quantitative analysis of NNMT protein at indicated timepoints of TGF-β1 treatment from (K). Data presented as means ± s.e.m., n = 2, One-way ANOVA. M, flow cytometric analysis of control and IPF fibroblasts stained with APC-labelled NNMT and FITC-labelled α-SMA is shown. Cells were first gated for FSC and SSC followed by NNMT-APC and α-SMA-FITC. N, the percentage of cell populations gated in B2 quadrant positive for both NNMT-APC and α-SMA-FITC are shown. Results represent data from n = 3 representing fibroblasts of three donor and IPF subjects each. Data presented as means ± s.e.m., n = 3, ∗∗∗p = 0.0062 (Student’s t test).

Figure 2

NNMT regulates extracellular matrix production, cellular proliferation, and metabolic pathways in lung fibroblasts.A–H, Western blot and quantitative analysis of protein expression of NNMT, Col1a1, FN, SIRT3, ACLY, PCNA, and Cyclin D1 in IMR-90 fibroblasts transfected with non-targeting (NT) or NNMT siRNA (100 nM) for 72 h. Data presented as means ± s.e.m., n = 3, ∗p < 0.05 (Student’s t test). I–L, fold change in mRNAs of NNMT, ACTA2, Col1a1, FN1, and ACLY in RNA-seq analysis of human fibroblasts transfected with non-targeting or NNMT siRNA (100 nM) followed by treatment with TGF-β1 (0 or 2.5 ng/ml) for 48 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA.

Figure 3

NNMT modulates TGF-β1 induced extracellular matrix production and pro-fibrotic metabolic pathways in human lung fibroblasts.A–E, Western blot and quantitative analysis of protein expression of NNMT, α-SMA, SIRT3, and ACLY in human lung fibroblasts transfected with non-targeting or NNMT siRNA (100 nM) followed by treatment with TGF-β1 (0 or 2.5 ng/ml) for 48 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA.

Figure 4

NNMT regulates lipofibroblast-myofibroblast phenotypic transition.A–D, violin plots showing mRNA of PPARγ, TCF21, Acetyl-CoA-Carboxylase (ACAC), and Perilin2 in Gene Expression Affymetrix data (GSE150910) of lungs of control donor and IPF patients. Data presented as means ± s.e.m., n = 102 to 103, ∗p < 0.005 (Student’s t test). E-G, mRNA of PPARγ, TCF21, and Perilin2 in RNA-seq analysis of human fibroblasts transfected with non-targeting or NNMT siRNA (100 nM) followed by treatment with TGF-β1 (0 or 2.5 ng/ml) for 48 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA. H–K, Western blot and quantitative analysis of NNMT, PPARγ, and TCF21 in whole cell lysates of IMR-90 fibroblasts transfected with NT or NNMT siRNA (100 nM) for 72 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA. L, flow cytometric analysis of control and IPF fibroblasts stained with APC-labelled NNMT and FITC-labelled PPARγ. Cells were first gated for FSC and SSC followed by NNMT-APC and PPARγ -FITC. M, plots show the percentage of population positive for NNMT-APC (grouped quadrant B2) and PPARγ-FITC (grouped quadrant B4) from (L). Results represent data from n = 3 representing fibroblasts of three donor and IPF subjects each. Data presented as means ± s.e.m., n = 3, ∗∗∗P = <0.0002 (Student’s t test).

Figure 5

NNMT induces apoptosis resistance and abrogates anti-apoptotic effects of TGF-β1 in lung fibroblasts.A, Western blot showing NNMT protein in IMR-90 fibroblasts transfected with NT or NNMT siRNA (100 nM) for 72 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h. B, representative phase contrast images of IMR-90 fibroblasts transfected with NT or NNMT siRNA (100 nM) for 72 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h. Scale bar 50 μm. C–E, Western blot and quantitative analysis of cleaved caspase-3, and cleaved PARP in whole cell lysates of IMR-90 fibroblasts transfected with NT or NNMT siRNA (100 nM) for 72 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 48 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA. Please note that the β-actin loading control in Figure 5C is the same as shown for Figure 3A since the same gel was probed for different proteins (as indicated on the two figures). F, quantitation of the apoptotic cells stained with Annexin V and Propidium Iodide after transfection with NT or NNMT siRNA (100 nM) for 72 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA. G, quantitation of the apoptotic cells stained with Annexin V and Propidium Iodide after transfection with control or NNMT overexpression vector for 48 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA.

Figure 6

NNMT regulates the expression of Bcl-2 family member proteins in human lung fibroblasts.A–C, violin plots showing mRNA of Bim, BAD, and PUMA in Gene Expression Affymetrix data (GSE150910) of lungs of control donor and IPF patients. Data presented as means ± s.e.m., n = 102 to 103, ∗p < 0.05 (Student’s t test). D and E, mRNA of Bim and PUMA in RNA-seq analysis of human fibroblasts transfected with non-targeting or NNMT siRNA (100 nM) followed by treatment with TGF-β1 (0 or 2.5 ng/ml) for 48 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA. F–I, Western blot and quantitative analysis of expression of NNMT, Bim, and PUMA in whole cell lysates of human donor lung fibroblasts transfected with NT or NNMT siRNA (100 nM) for 72 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA. J, schematic diagram showing that NNMT regulates lipofibroblast–myofibroblast plasticity and susceptibility to apoptosis.

Supporting Figure S1

Extracellular matrix forming proteins are highly expressed in IPF.A–C, violin plots showing mRNA of α-SMA, Col1a1, and Fibronectin in Gene Expression Affymetrix data (GSE 150910) of lungs of control donor and IPF patients. Data presented as means ± s.e.m., n = 102 to 103, ∗p < 0.05 (Student’s t-test). D, immunofluorescence staining of NNMT in lung sections from control and IPF patients. NNMT -red; nuclei -blue. Scale bar 50 μm. E, fluorescence DSP images of control donor and IPF lung tissues showing α-SMA expression in red outlined Regions of Interest (ROI). Pan-keratin (green), CD31 (red), α-SMA (yellow), nucleus (blue). Scale bar 200 μm. F, heatmap showing comparative quantitation of a-SMA in ROIs of control and IPF lung tissues. G–I, mRNA of α-SMA, Col1a1, and Fibronectin (FN1) in Digital Spatial Profiling (DSP) representing six Regions of Interest (ROIs) from lung sections of control and individual with IPF; n = 6 representing one individual, each with analysis of six random Regions of Interest (ROIs). The data are presented as means ± s.e.m., ∗p < 0.05 (Student’s t-test).

Supporting Figure S2

NNMT mediates lipofibroblast-myofibroblast phenotype transition and metabolic reprogramming in human lung fibroblasts.A–E, Western blot and quantitative analysis of protein expression of NNMT, Col1a1, CTGF and ACLY in human IPF fibroblasts transfected with non-targeting or NNMT siRNA (100 nM) for 72 h. Data presented as means ± s.e.m., n = 3, ∗p < 0.05 (Student’s t-test). F, PPARγ mRNA expression in Digital Spatial Profiling (DSP) representing twelve Regions of Interest (ROIs) from control donor and individual with IPF; n = 12 in each group. The data represent means ± s.e.m., ∗∗∗p = 0.0086 (Student’s t-test).

Supporting Figure S3

NNMT mediates TGF-β1 induced anti-apoptotic phenotype in human lung fibroblasts.A, Western blot showing expression of NNMT in IMR-90 fibroblasts transfected with NT or NNMT siRNA (100 nM) for 72 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h. B, flow cytometric analysis of fibroblasts stained with Annexin V and Propidium iodide after transfection with NT or NNMT siRNA (100 nM) for 72 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h.

Supporting Figure S4

NNMT potentiates TGF-β1 induced anti-apoptotic phenotype in human lung fibroblasts.A, Western blot showing expression of NNMT in IMR-90 fibroblasts transfected with control or NNMT overexpressing vector for 48 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h. B, flow cytometric analysis of fibroblasts stained with Annexin V and Propidium iodide after transfection with control or NNMT overexpressing vector for 48 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h.

Supporting Figure S5

NNMT downregulates expression of pro-apoptotic Bcl-2 family proteins in human lung fibroblasts.A–D, Western blot and quantitative analysis showing expression of NNMT, Bim, and PUMA in whole cell lysates of IMR-90 fibroblasts transfected with NT or NNMT siRNA (100 nM) for 72 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h. Data presented as means ± s.e.m., n = 3, One-way ANOVA. E–H, Western blot and quantitative analysis showing expression of NNMT, Bim, and PUMA in whole cell lysates of IPF lung fibroblasts transfected with NT or NNMT siRNA (100 nM) for 72 h followed by treatment with vehicle or TGF-β1 (2.5 ng/ml) for 36 h. Data represent means ± s.e.m., n = 3, One-way ANOVA.