Author's response to reviews Title: Nuclear entrapment and extracellular depletion of PCOLCE is associated with muscle degeneration in oculopharyngeal muscular dystrophy Authors: Vered Raz (v.raz@lumc.nl) Ellen Sterrenburg (e.sterrenburg@prinsesbeatrixfonds.nl) Samantha Routledge (sam_routledge@hotmail.com) Seyed Yahya Anvar (s.y.anvar@lumc.nl) Andrea Venema (a.venema@lumc.nl) Barbara M van der Sluijs (b.van.der.sluijs@gelre.nl) Capucine Trollet (capucine.trollet@upmc.fr) George Dickson (G.Dickson@rhbnc.ac.uk) Baziel GM van Engelen (B.vanEngelen@neuro.umcn.nl) Silvère M van der Maarel (S.M.Maarel@lumc.nl) Michael N Antoniou (michael.antoniou@kcl.ac.uk) Version: 2 Date: 1 March 2013 Author's response to reviews: see over
Dear Editor Please find a resubmission of the MS entitled: Nuclear entrapment and extracellular matrix depletion of PCOLCE is associated with muscle weakness in oculopharyngeal muscular dystrophy. We appreciate the time and effort the reviewers have made to provide their comments on our manuscript. We believe that by responding to their invaluable comments our manuscript has significantly improved in its quality and consequently broadened the scope of its interest. In this resubmitted manuscript we have addressed all the reviewers comments. Our point-by-point response to the reviewers comments is highlighted in bold text below: Reviewer Satoru Noguchi Point 1. A major concern is co-localization of PCOLCE with aggregated PABPN1 shown in the merged image in Fig.7B. The green signals of PCOLCE were not completely co-localized with PABPN1 in red. - Figure 7B shows staining after KCl treatment of cells; this treatment allows identification of aggregated, insoluble nuclear proteins. However, treatment with KCl does not maintain nuclear morphology and exact co-localization cannot be determined. Exact co-localization must be determined with a high resolution image preferably with a confocal microscope, as shown in Figure 4Aii. Therefore, the exact co-localization statement is based on images in cell culture and the staining in Figure 7B shows that both PAPBN1 and PCOLCE aggregate in the same cell nucleus. Therefore, we conclude that PCOLCE entrapment within PABPN1 aggregates is relevant for OPMD. These points are now made clearer in the text on page 10 line 29 to page 11 line 2. On the authors theory, PCOLCE should be co-localized with PABPN1 for its nuclear entrapment. Why was PCOLCE localized in nucleus without interaction with PABPN1? In Figure 4 we show that PCOLCE is localized to the cell nucleus and binds to soluble PABPN1. Our model system suggests that PCOLCE interacts with nuclear PABPN1. These points have been made clearer in the text by reference to Figure 4 (page 9 line 30 to Page 10 line 4) and in the Discussion section (page 13 lines 12-14). The authors also need to explain the mechanism underlying recruitment of PCOLCE to aggregated PABPN1 in the nucleus, because PCOLCE protein is a secreted proteins translated with signal sequence, which would be synthesized in ER lumen and transported in Golgi and secretion pathway. This is a relevant point and we have included a statement addressing this issue in the Discussion section (page 13 lines 3-9). How and where do both proteins make a complex? Our data shows that PABPN1 and PCOLCE are bound together in soluble protein extracts
with co-localization found in the nuclear compartment. Since both PABPN1 and PCOLCE contain RNA-binding motifs, we suggest that these two proteins form a dynamic complex in the nucleus. This dynamic complex is disturbed in exppabpn1 expressing cells because exppabpn1 is prone to aggregation. We now better explain these points in the Discussion section (page 13 lines 3-14). Point 2. The present study did not fully prove that nuclear entrapment of PCOLCE and its ECM depletion contributes to muscle fibrosis in OPMD patients and murine models. Aberrant procollagen processing and deformed fibril morphology should be analyzed. We agree with the reviewer that this study does not directly show that depletion of PCOLCE at the ECM causes defects in procollagen processing. However, this was previously demonstrated in several in vitro and in vivo studies and also in muscular dystrophies. References to these studies, including additional references, are now clearly indicated in the Discussion section (page 12- lines 5-16). Point 3. The authors should provide explanation of a relationship between downregulation of PCOLCE gene and nuclear entrapment of PCOLCE protein. Several studies, including ours, have shown that expression levels of mrnas corresponding to proteins that are entrapped in nuclear aggregates are frequently dysregulated. The regulatory mechanism for this observation is not known, but it is possible that feedback regulatory loops coordinate protein accumulation and mrna expression. A statement discussing this point is now included in the Discussion section. Minor Essential Revisions 1. The authors used quadriceps muscle biopsies from OPMD patients (Fig. 2). The patient s clinical information; i.e, age at the first visit, kind of muscles in quadriceps, change in muscle imaging, should be supplied. We have now added Table 1 to the Methods section (under the heading Cell culture and collection of muscle biopsies, second paragraph), which summarizes the age of the patients, muscle strength and gender. The age of onset is not a clinical parameter, as it is highly influenced by family history. We have not as yet collected imaging information (MRI) to monitor change in muscle. The human muscle samples used in this study are Vastus lateralis and this is corrected throughout the manuscript. 2. The authors showed expression levels of PCOLCE in biopsied muscle from presymptomatic and symptomatic OPMD patients (Fig.3A). Are there any correlations between downregulation of PCOLCE gene and severity of muscle pathology? We show that PCOLCE levels are decreased between an age-matched control group and OPMD patients, and between symptomatic and pre-symptomatic OPMD patients. This indicates that a decrease in PCOLCE levels is associated with muscle symptoms. This point is now made clearer on page 7, last paragraph. We cannot further correlate PCOLCE levels with severity of muscle pathology as, (1) the cohort is too small for statistical correlation, and (2) we do not have quantitative measures of muscle pathology. Did error bars in Figs. 3B & C denote SD or SEM?
The error bars represent SD and this is now clearly indicated in the Figure legend. 3. Histogram in Fig.4B was missing. We apologize for this mistake. The text referring to the histogram in Figure 4B was accidently inserted and is removed from the revised version. 4. In Fig.5 two terms, PABPN1 and Pabpn1, are confusing. Image quality should be improved. According to the system of universal gene nomenclature, human genes are denoted in uppercase letters and murine genes with lowercase letters. Therefore, PABPN1 in this study always refers to the human gene, often with alanine-expansion (Ala17). This is now clearly indicated in the text and in figure legends. PCOLCE is written in either uppercase when the investigation is in human cells/tissues, or as Pcolce when the study is in a mouse system. In the Discussion it is always written as PCOLCE. As requested Figure 5 is now better presented: higher resolution is provided and background is corrected. 5. The authors stated that sections from the OPMD patient showed markedly less extracellular PCOLCE, which was accompanied by intense nuclear staining (Fig. 7A). Co-localization of PCOLCE with PABPN1 is difficult to be seen. An image for PCOLCE single staining should be supplied. An image for PCOLCE single staining is now included in this figure as requested. 6. Paper had many typing errors. These typing errors have now been corrected. 7. Quality of written English: Needs some language corrections before being Published The language of the paper has been revised to an appropriate higher standard. Reviewer: Corrado Angelini Major Compulsory Revisions: This is a paper that deals with PCOLCE mislocalisation in OPMD and correlates it with muscle weakness Title and abstract should be more focused. We agree with the reviewer and the title has been changed to reflect this principal point of our study. The specificity of the accumulation of PCOLCE is shown in fig 7, particularly in an histogram, however in myotonic dystrophy there are some similarities to OPMD, and authors do not take in account the central or peripheral location of nuclei, nor the possible participation of satellite cells nuclei. The reviewer s comment concerns some clinical similarities between OPMD and MD1: in both conditions muscle weakness initiates after mid-life, and in both aggregate formation is a histological hallmark of disease. Additional studies are required in order to conclude
whether these disorders share similar molecular mechanisms. Our study is one of the first attempts, which suggests some similarity concerning PCOLCE sub cellular localization. This point is discussed in the last paragraph of the Discussion. The scope of this study is not to compare histological changes in OPMD. Subsequent studies should investigate whether the change in PCOLCE sub cellular localization has an impact on location of nuclei in affected fibers and on satellite cells. There is in Methods no clinical detail about number and age of OPMD patients nor of DM1, FSHD and DMD cases analyzed, since the title mentions weakness it would be necessary to add an MRC scale of quadriceps strength. The information requested for OPMD patients is now included in Table 1. In cases of OPMD patients a single doctor determined the patients MRC score, and therefore scores can are comparable between patients. For DM1, FSHD and DMD patients clinical examination was performed by different doctors and thus scores cannot be directly compared. The aim of the experiments presented in Figure 8 was to show that the ECM defects found in other muscle disorders are not associated with PCOLCE subcellular localization. Since we have analyzed only one patient we cannot make a correlation with muscle weakness in these disorders. We appreciate the reviewers comments on this Figure and the text has changed to reflect these concerns that they have raised. We now suggest that a change in PCOLCE subcellular localization could be a general outcome in late-onset muscle disorders (page 8, last paragraph). In some Figures (Fig 3) data from OPMD patients and mice model data are mixed with different scales are used, it would be useful for the reader to separate muscle biopsies from mouse model data. The scales in this figure have now been altered to be similar in all cases. In addition, the figure legend has been altered to better explain its content. Were myotube cultures derived from biopsies or mice? The growth and differentiation of myotubes is different, in various species, this also should be stated in methods and figures more clearly. As described in the Methods section of the manuscript, the OPMD cell model system was generated from immortalized murine myoblast cells. This is now more clearly explained in the main text of the paper. Discretionary revisions: Fig 2 the use of trichrome stain is usually more useful for rimmed vacuoles and connective tissue or fat demonstration What do the authors mean by "signatures of muscle fibrosis", please explain. We appreciate the reviewer s suggestion for the histological staining and we will take it into consideration for future studies. To avoid confusion the statement signatures of muscle fibrosis is now removed from the text. Fig 2A have some abnormalities, but there is no description of central nuclei. Fig7B the experiment with 1 M KCL seems to have removed not only soluble but also some structural proteins, how long was 1 M KCL applied? Issues concerning the KCl treatment of the cells have been addressed under the response
to first Reviewer, in Point 1. As we did not provide quantification of central nuclei in this manuscript this statement in removed from the text. The use of dysferlin as sarcolemmal marker is not particularly clear, may be caveolin would be more clear. Many studies, and the images here show that dysferlin marks the periphery of the myofiber. We therefore do not believe that using caveolin as a sarcolemmal marker would add any greater clarity to the images from this immunofluorescence staining. Quality of written English: Needs some language corrections before being published We have extensively gone over the whole manuscript and have improved the quality of the English and corrected typing errors.