Tankyrase-1 regulates RBP-mediated mRNA turnover to promote muscle fiber formation

Abstract

Poly(ADP-ribosylation) (PARylation) is a post-translational modification mediated by a subset of ADP-ribosyl transferases (ARTs). Although PARylation-inhibition based therapies are considered as an avenue to combat debilitating diseases such as cancer and myopathies, the role of this modification in physiological processes such as cell differentiation remains unclear. Here, we show that Tankyrase1 (TNKS1), a PARylating ART, plays a major role in myogenesis, a vital process known to drive muscle fiber formation and regeneration. Although all bona fide PARPs are expressed in muscle cells, experiments using siRNA-mediated knockdown or pharmacological inhibition show that TNKS1 is the enzyme responsible of catalyzing PARylation during myogenesis. Via this activity, TNKS1 controls the turnover of mRNAs encoding myogenic regulatory factors such as nucleophosmin (NPM) and myogenin. TNKS1 mediates these effects by targeting RNA-binding proteins such as Human Antigen R (HuR). HuR harbors a conserved TNKS-binding motif (TBM), the mutation of which not only prevents the association of HuR with TNKS1 and its PARylation, but also precludes HuR from regulating the turnover of NPM and myogenin mRNAs as well as from promoting myogenesis. Therefore, our data uncover a new role for TNKS1 as a key modulator of RBP-mediated post-transcriptional events required for vital processes such as myogenesis.

Graphical Abstract

Figure 1.

TNKS1 is required for muscle cell differentiation. (A) Total cell extracts were prepared from exponential and differentiating C2C12 myoblasts (Exp, Day 0, Day 1, Day 2, Day 3). (Top panel) Western blot experiments were performed using antibodies against pADPr or α-tubulin (loading control). (Bottom panel) Histogram representation of the quantification of the western blot in top panel. Values were quantified using ImageJ and normalized to tubulin. (B) Phase contrast images were taken of C2C12 cells treated with scrambled control (siCTRL), or specific siRNAs against PARP1, PARP2, PARP3, PARP4, TNKS1 or TNKS2 three days after induction of differentiation. Images of a single field are shown and represent three independent experiments (scale bars, 100 μm). (C) Immunofluorescence experiments demonstrating the differentiation of C2C12 myoblasts treated with the siRNAs described in (B). (Top panel) Cells were fixed three days post-induction of differentiation, and antibodies against known markers of muscle fibers (anti-MyHC, anti-myoglobin) were used for immunostaining. DAPI was used to stain nuclei. Images of a single field are shown and represent three independent experiments (scale bars, 20 μm). (Bottom panel) Quantification of the fusion index for the C2C12 cells described in top panel. The fusion index was calculated as the ratio of the nuclei number in myotubes versus the total number of nuclei. (D) (Top panel) Immunofluorescence experiments demonstrating the differentiation of primary myoblasts treated with siCTRL or siTNKS1. Cells were fixed 3 days post-induction of differentiation, and antibodies against known markers of muscle fibers (anti-MyHC, anti-myoglobin) were used for immunostaining. DAPI was used to stain nuclei. Images of a single field are shown and represent three independent experiments (scale bars, 20 μm). (Bottom panel) Quantification of the fusion index for the C2C12 cells described in top panel. The fusion index was calculated as the ratio of the nuclei number in myotubes versus the total number of nuclei. (E, F) Total extracts were prepared from C2C12 cells transfected with scrambled control (siCTRL) or siRNA against TNKS1 (siTNKS1) and collected from exponentially growing (EXP) and differentiating C2C12 myoblasts (Day 0 and Day 2). (E) The extracts were used in western blot analysis using antibodies against pADPr or α-tubulin (loading control). (F) Histogram representation of the quantification of the western blot in the left panel. Values were quantified using ImageJ, normalized to tubulin, and shown relative to the EXP siCTRL treated condition. (G, H) Total extracts were prepared from C2C12 cells transfected with scrambled control (siCTRL) or a combination of siRNA targeting PARP1 and PARP2 (siPARP1/2) and collected from exponentially growing (EXP) and differentiating C2C12 myoblasts (Day 0 and Day 2). (G) The extracts were subjected to western blot analysis using antibodies against pADPr or α-tubulin (loading control). (H) Histogram representation of the quantification of the western blot in the left panel. Values were quantified using ImageJ, normalized to tubulin, and shown relative to the EXP siCTRL treated condition. (I) (Top panel) Immunofluorescence images of C2C12 transfected with scrambled control (siCTRL) or a combination of siRNA against PARP1 and PARP2 (siPARP1/2) Cells were fixed three days post-induction of differentiation, and antibodies against known markers of muscle fibers (anti-MyHC, anti-myoglobin) were used for immunostaining. DAPI was used to stain nuclei. Images of a single field are shown and represent three independent experiments (scale bars, 20 μm). (Bottom panel) Quantification of the fusion index for the C2C12 cells in top panel. The fusion index was calculated as the ratio of the nuclei number in myotubes versus the total number of nuclei. Data shown in Figure 1 are presented ± the s.e.m. of three independent experiments with *P < 0.05, **P < 0.01, ***P < 0.001 by unpaired t-test.

Figure 2.

TNKS1 PARylates promyogenic RNA-Binding Proteins such as HuR. (A) Total cell extracts isolated from differentiating C2C12 myoblasts (Day 2) were used for immunoprecipitation experiments with pADPr or IgG antibodies and analyzed by mass spectrometry. The immunoprecipitation of pADPr was validated by western blot analysis using a pADPr antibody. The blot shown is representative of three independent experiments. (B) The identified proteins were classified according to their gene ontology molecular function. The percentage of proteins identified in each category is represented in a pie chart. (Lower panel) List of top 5 categories. (C) List of RNA-binding proteins previously associated with myogenesis identified in (B). (D) Immunoprecipitation experiments using pADPr or IgG antibodies were performed with extracts from confluent (D0) C2C12 cells. The association of RNA-binding proteins such as HuR and KSRP to pADPr was determined by western blot analysis. The blot shown is representative of three independent experiments. (E, F) Immunoprecipitation experiments using pADPr or IgG antibodies were performed with extracts from confluent (D0) C2C12 cell transfected with scrambled control (siCTRL) or siRNA against TNKS1 (siTNKS1). (E) The association of HuR to pADPr was determined by western blot analysis. The blot shown is representative of three independent experiments. (F) Histogram representation of the quantification of the western blot in the (E). (G) Schematic representation of the in vitro PARylation assay procedure. GST-HuR and biotinylated NAD+ were incubated with recombinant PARP1, PARP3, or TNKS1 enzymes. HRP-conjugated streptavidin was added to the reaction and the signal was measured as arbitrary units by chemiluminescence by detection with luminol. (H) Quantification of Ribosylation units demonstrating the PARylation of GST-HuR by PARP1 and TNKS1, but not PARP3, in vitro. Levels are normalized to GST control. Data shown in Figure 2 are presented ± the s.e.m. of three independent experiments with *P < 0.05, **P < 0.01, ***P < 0.001 by unpaired t-test

Figure 3.

TNKS1-mediates binding of HuR to myogenic mRNA targets during myogenesis. C2C12 muscle cells were transfected with scramble control (siCTRL) or siRNAs specific for TNKS1, or HuR. Total RNA and protein lysates were prepared from these cells 2 days post-induction of differentiation. (A) NPM (left panel) and Myogenin (right panel) mRNA levels were determined by RT-qPCR, standardized against GAPDH mRNA, and expressed relative to siCTRL conditions. (B, C) Total extracts were used for western blot analysis to determine NPM (B) and Myogenin (C) protein levels. (Bottom) Histogram representation of the quantification of the western blot. Values were quantified using ImageJ, normalized to tubulin, and shown relative to the siCTRL treated condition. (D, E) RNA-Immunoprecipitation coupled to RT-qPCR experiments was performed using anti-HuR and anti-IgG antibodies on total extracts from differentiating C2C12 cells treated with scrambled control (siCTRL) or siRNA against TNKS1 (siTNKS1). (D) Western blot assessing the immunoprecipitation of HuR. (E) NPM and Myogenin mRNA levels in the immunoprecipitates were normalized to the corresponding IgG sample and mRNA input. The levels of NPM and myogenin mRNA in siTNKS1 conditions were plotted relative to siCTRL conditions. (F, G) Actinomycin D (Act. D) pulse-chase assays were performed using C2C12 myoblasts transfected with scrambled control (siCTRL) or siRNA against TNKS1 (siTNKS1). 48 h post-induction of differentiation, the cells were treated for various periods of time with Actinomycin D to assess the stability of NPM (F) and Myogenin (G) mRNAs. Data shown in Figure 3 are presented ± the s.e.m. of three independent experiments with *P < 0.05, **P < 0.01, ***P < 0.001 by unpaired t-test.

Figure 4.

PARG inhibition ameliorates muscle fiber formation. (A) (Left panel) Total cell extracts were prepared from differentiating C2C12 cells (Day 2) treated with 5uM PARGi or DMSO as a control. These extracts were used for western blot analysis with antibodies against pADPr or α-tubulin (loading control) (Right panel). Histogram representation of the quantification of the western blot Levels of PARylated proteins were quantified using ImageJ and normalized to tubulin and shown relative to DMSO treated control condition. (B) Phase contrast images were taken of differentiating C2C12 cells treated with 5 uM PARGi or DMSO (Day 2). Images of a single field are shown and represent three independent experiments (scale bars, 50 μm). (C) (Left panel) Immunofluorescence experiments on differentiating C2C12 myoblasts treated with 5 uM PARGi or DMSO. Staining was performed with antibodies against known markers of muscle fiber formation, MyHC and myoglobin. DAPI was used to stain nuclei. Images of a single field are shown and represent three independent experiments (scale bars, 20 μm). (Right panel) Quantification of the fusion index for the C2C12 cells described in the left panel. The fusion index was calculated as the ratio of the nuclei number in myotubes with two or more nuclei versus the total number of nuclei. (D, E) (Top panels) Total cell extracts were prepared from differentiating C2C12 cells (Day 2) treated with 5 uM PARGi or DMSO as a control and used for western blot analysis with antibodies against NPM (D), Myogenin (E) or α-tubulin (loading control). (Bottom panels) Histogram representation of the quantification of the western blots on top panels. Values were quantified using ImageJ, normalized to tubulin, and shown relative to the DMSO-treated control condition. Data shown are presented ± the s.e.m. of three independent experiments with *P < 0.05, **P < 0.01, ***P < 0.001 by unpaired t-test.

Figure 5.

TNKS1-mediated PARylation of HuR regulates its cellular movement. (A) Immunofluorescence images showing the localization of HuR in exponentially growing (EXP) and differentiating (Day 2) muscle myoblasts transfected with scramble control (siCTRL) or siRNA against TNKS1 (siTNKS1). Immunofluorescence staining was performed with an antibody against HuR. DAPI was used to stain nuclei. Images of a single field are shown and are representative of three independent experiments (scale bars, 20 μm). (B) (Left panel) Immunoprecipitation experiments using TRN2 or IgG antibodies were performed with extracts from differentiating (D2) C2C12 cells transfected with scrambled control (siCTRL) or siRNA against TNKS1 (siTNKS1). The association of HuR to TRN2 was determined by western blot analysis. The blot shown is a representative of three independent experiments. (Right panel) Histogram representation of the quantification of the western blot in the left panel. (C) Total extracts from C2C12 myoblasts transfected at the exponential growth phase with scrambled control (siCTRL) or siRNA against TNKS1 (siTNKS1) and collected 2 days post-induction of differentiation, were used for western blot analysis (left panel) to determine HuR-CP1 protein levels. (Right panel) Histogram representation of the quantification of the western blot in left panel. Values were quantified using ImageJ and normalized to α-tubulin and shown relative to the siCTRL treated condition. Data shown in Figure 5 is presented ± the s.e.m. of three independent experiments with **P < 0.01, ***P < 0.001 by unpaired t-test.

Figure 6.

HuR contains a TNKS1 consensus binding motif. (A) Swiss-Prot Database (www.expasy.ch/prosite) blast of the putative TNKS1-binding motif identified in HuR. (B) Schematic of the GFP, GFP-HuR^WT, and GFP-HuR^G224D constructs. (C) (Left panel) Immunoprecipitation experiments using differentiating C2C12 cells expressing GFP, GFP- HuR^WT and GFP-HuR^G224D were performed using anti-IgG or anti-TNKS1 antibodies to assess the association of the exogenous proteins to TNKS1 by western blot analysis. The data are representative of three independent experiments. (Right panel) Histogram representation of the quantification of the western blot in the left panel. (D) In vitro PARylation assay: Quantification of Ribosylation units showing the PARylation of GST-HuR^WT and GST- HuR^G224D by recombinant TNKS1 in vitro. Levels are normalized to GST control. Data shown are presented ± the s.e.m. of three independent experiments with *P < 0.05, **P < 0.01 by one-way ANOVA, Tukey HSD test. (E) (Left panel) Immunoprecipitation experiments using differentiating C2C12 cells expressing GFP, GFP-HuR^WT or GFP- HuR^G224D, were performed with anti-IgG and anti-pADPr antibodies to assess the PARylation by western blot analysis. The blot is representative of three independent experiments. (Right panel) Histogram representation of the quantification of the western blot in the left panel. Data shown in Figure 6 is presented ± the s.e.m. of three independent experiments with **P < 0.01 by unpaired t-test.

Figure 7.

TNKS1-mediated PARylation of HuR is required for its pro-myogenic function. C2C12 cells were transfected with scrambled control (siCTRL) or siRNA against HuR (siHuR), and then constructs expressing GFP, GFP-HuR^WT or GFP-HuR^G224D. (A) Total RNA was isolated from these C2C12 and the level of NPM (left graph) and myogenin (right graph) mRNAs was assessed by RT-qPCR, standardized against GAPDH mRNA, and expressed relative to siCtrl + GFP conditions. Data are represented as mean is presented ± the s.e.m. of three independent experiments with *P < 0.05, **P < 0.01 by two-way ANOVA, Tukey HSD test. (B) (Left Panel) Immunofluorescence (staining with anti-MyHC and anti-GFP antibodies, as well as with DAPI to stain nuclei) images of cells described in (A) assessing rescue of the myogenic phenotype in HuR knockdown cells (scale bars 20 μm). (Right panel) A histogram representation of the fusion index of immunofluorescence shown in left panel. Data represented as mean ± the s.e.m. of three independent experiments with *P < 0.05, **P < 0.01, ***P < 0.001 by two-way ANOVA, Tukey HSD test. (C, D) RNA-Immunoprecipitation experiments using differentiating C2C12 cells expressing GFP, GFP-HuR^WT and GFP-HuR^G224D were performed using anti-GFP antibodies. (C) Western blot confirming the immunoprecipitation of GFP-conjugated HuR isoforms. The blot is representative of three independent experiments. (D) Association of GFP, GFP-HuR^WT and GFP-HuR^G224D to NPM (left panel) or myogenin (right panel) was determined by RT-qPCR analysis. Data in Figure 7D are presented as the mean ± SEM (n= 3) with **P < 0.01 by one-way ANOVA, Tukey HSD test.

Figure 8.

Model depicting the mechanism by which TNKS1-mediated PARylation of the promyogenic RBP HuR impact myogenesis. TNKS1-mediated PARylation promotes the cytoplasmic translocation of HuR. The observation reported in this study suggests that by retaining its PARylation status, HuR promotes myogenesis by differentially regulating the turnover and expression of its target mRNA NPM (decay) and myogenin (stability). Image was created with Biorender.com.