Fig 2: Proteasome activity in Pabpn1-DR cell culture and muscle tissue.A-C. Analyses of ß-subunits catalytic activity in muscle cell culture using activity gel. Specific binding of LWA300 to ß-subunits is demonstrated by pre-incubation with 0.5µM Epox for 1 hour (A). Proteasomal activity in Scram vs. shPab muscle cell cultures (B-C), activity gel shows a representative experiment (B) Equal loading was assessed by Coomassie staining (CS). Bar chart (C) shows LWA300 MFI in after normalization to CS. Averages are from triplicates. D. A representative western blot shows protein levels of proteasome-encoded genes in Scram and shPab cell cultures. E. Bar chart shows quantification of protein fold change after normalization to Scram cultures. Averages and standard variations are from three independent experiments. Statistical significance (p<0.05; *) was made with unpaired Student’s t-test. F. Images show subcellular localization of LWA300 in Scram or shPab cell cultures. Cultures were incubated with 62.5 nM LWA300, pre-incubation with 0.5µM Epox was carried out as control. Segmented nuclei are depicted with red circles. Scale bar is 5 µm. G. Plot shows cumulative distribution of nuclear LWA300 MFI in Scram or shPab C2C12 cells. Statistical significance (p<0.05) was assessed with the Kruskal-Wallis test. H. Bar chart shows flow cytometer analysis of LWA300 MFI in Scram or shPab cell cultures treated with or without Epox. Averages are from three experiments. Representative histogram of flow cytometer output is in S2 Fig. I-J. LWA300 signal in muscle cryosections. I. Images show GFP or LWA300 fluorescence in successive cryosections from Scram or shPab. Examples of the four same myofibers in GFP and LWA300 are depicted with numbers. Nuclei were counterstained with DAPI. Scale bar is 50 µm. J. Plot shows cumulative distribution of nuclear LWA300 MFI in cryosections from three mice. Over 400 myofibers are included per condition per mouse. Statistical significance (p<0.05) was assessed with Kruskal-Wallis test.

Fig 3: Autophagy is hyper activated in OPMD muscle cell modelExperiments were performed in stable muscle cell culture over-expressing wild type PABPN1 (A10) or expPABPN1 (A17). The expression of PABPN1 transgenes was induced by incubation with 2% HS. A. Western blot shows levels of transgenes A10 and A17 PABPN1-FLAG (55 kDa) and endogenous Pabpn1 (52 kDa). Autophagy activation is represented by LC3II and P62. Tubulin and Gapdh are used as loading controls. B. Images of representative immunofluorescence with anti-FLAG (green) and anti-LC3 (red) antibodies in cell cultures that were incubated with 2% HS for 3 hours or 48 hours. Scale bar is 10 µm. C. Chart bar shows mean LC3 puncta per nucleus in A10 or A17 culture. Mean and standard deviations are from 100 nuclei collected from three independent experiments. D. Bar chart shows mRNA levels of five ATG in A10 or A17 cultures. Fold change was obtained after normalisation to Hprt housekeeping gene and to parental culture. E. Bar chart shows the nuclear to cytosolic ratio of ATG transcripts from the distal primer set. F. Image shows a representative Western blot of nuclear and cytosolic fractions, marked by Emerin or Tubulin, respectively. G. Bar chart shows PABPN1 abundance in nuclear or cytosolic fractions in A10 or A17 cell cultures. PABPN1 abundance was calculated after normalization to loading control in each fraction. Averages and standard deviations are from 4 replicates. Averages and standard deviations are from three biological independent cultures. Statistical significance is assessed by the Student’s T-Test (p < 0.05 is denoted with *; p < 0.005 is denoted with **).

Fig 4: ATGs mRNA expression in the OPMD mouse model, A17.1A.-B. Bar charts show results from RT-qPCR experiments of eight ATG in EDL (A) or Soleus (B) muscles of A17.1 (grey bars) and FVB control mice. Expression fold change was calculated by normalisation to Hprt housekeeping gene and FVB control. C. Schematic presentation of PCR analysis of PAS utilization in the 3’UTR. Two primer sets were used: a distal set (D) amplifies long transcripts from the distal PAS, and a proximal set (P) that additionally amplifies short transcripts with proximal PAS. APA utilization is calculated from the ratio Distal to Proximal. D. Bar chart shows distal to proximal ratio for five ATG in EDL. Averages and standard deviations are from five mice. Student’s T-Test was used to assess statistical significance (0.05

Fig 5: A model for the effect of reduced PABPN1 levels on muscle atrophy.Reduced PABPN1 levels induced a alternative polyadenylation site (APA) utilization in the 3’-UTR of a subset of transcripts whose gene product regulates muscle waste, including Atrogin-1/Fbxo32 and proteasomal genes. In these muscles, Atrogin-1 is overexpressed, whilst proteasome activity is reduced. This leads to pathological hallmarks of muscle atrophy including reduction in myofiber cross-sectional area, increase in extracellular matrix (ECM) and a switch in myosin heavy chain (MyHC) expression pattern in myofibers.