Deregulations of miR‐1 and its target Multiplexin promote dilated cardiomyopathy associated with myotonic dystrophy type 1

  1. Anissa Souidi
  2. Masayuki Nakamori
  3. Monika Zmojdzian
  4. Teresa Jagla
  5. Yoan Renaud
  6. Krzysztof Jagla

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Myotonic dystrophy type 1 (DM1) is the most common muscular dystrophy in adults. It is caused by the excessive expansion of noncoding CTG repeats, which when transcribed affects the functions of RNA‐binding factors with adverse effects on alternative splicing, processing, and stability of a large set of muscular and cardiac transcripts. Among these effects, inefficient processing and down‐regulation of muscle‐ and heart‐specific miRNA, miR‐1 , have been reported in DM1 patients, but the impact of reduced miR‐1 on DM1 pathogenesis has been unknown. Here, we use Drosophila DM1 models to explore the role of miR‐1 in cardiac dysfunction in DM1. We show that miR‐1 down‐regulation in the heart leads to dilated cardiomyopathy (DCM), a DM1‐associated phenotype. We combined in silico screening for miR‐1 targets with transcriptional profiling of DM1 cardiac cells to identify miR‐1 target genes with potential roles in DCM. We identify Multiplexin (Mp) as a new cardiac miR‐1 target involved in DM1. Mp encodes a collagen protein involved in cardiac tube formation in Drosophila . Mp and its human ortholog Col15A1 are both highly enriched in cardiac cells of DCM‐developing DM1 flies and in heart samples from DM1 patients with DCM, respectively. When overexpressed in the heart, Mp induces DCM, whereas its attenuation rescues the DCM phenotype of aged DM1 flies. Reduced levels of miR‐1 and consecutive up‐regulation of its target Mp/Col15A1 might be critical in DM1‐associated DCM.

Article activity feed

  1. Version published to 10.15252/embr.202256616
  2. Review Commons

    Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    2. Point-by-point description of the revision

    Reviewer #1 (Evidence, reproducibility and clarity (Required)):

    *The paper titled "Deregulations of miR-1 and its target Multiplexin promote dilated cardiomyopathy associated with myotonic dystrophy type 1" by the Jagla group studied the effect of down-regulation of miR-1 in myotonic dystrophy type 1 (DM1) using fly as the disease model. The study is based on previous findings that in DM1 MBNL1 is sequestered, CELF1 is stabilized, and miR-1 is down regulated. The authors further identified Multiplexin to be the target effector of miR-1 in the fly heart and studied its function with a series of gain- and loss-of-function and rescue experiments. The authors' findings represent a significant advance in understanding the genetic mechanisms that can explain the pathogenic causes of dilated cardiomyopathy associated with DM1. Overall, this paper is well written and organized, with well-designed experiments and a clear model. A few additional experiments are suggested to further strengthen the conclusion. *

    Answer: We are grateful to the Reviewer for appreciating quality and significance of our work.

    • Wild-type and Hand-Gal4 controls are missing for all experiments (only UAS lines outcrossed to w1118 are displayed). Also, Hand-Gal4 driver lines can cause mild dilation by itself, which would influence interpretation and statistics of data. * Answer: We agree and include the Hand-Gal4/+ control condition to all main and supplemental figures showing heart parameters. The differences in data statistics using Hand-Gal4/+ compared to UAS/+ control lines reinforce our data interpretation.

    They are listed below:

    • increase in diastolic diameter and reduction of fractional shortening become statistically significant in Hand>miR-1sponge hearts at 5 weeks (Fig. 1D, F);
    • reductions of fractional shortening become significant in Hand>Bru3 (Fig. 2C) and Hand>mblRNAi (Fig. 2F) contexts at 1 week;
    • increase in diastolic diameter of Hand>mblRNAi hearts at 5 weaks becomes statistically significant (Fig. 2D);
    • increase in diastolic diameter of Hand>Mp hearts at 5 weeks (Fig. 4E) becomes statistically significant ;
    • reduction of fractional shortening becomes statistically significant in Hand>Mp context at 1 week (Fig. 4G);
    • increase in diastolic diameter in Hand>960CTG context at 5 weeks becomes statistically significant (Fig. S2A). *Also, the authors should consider to confirm the miR-1 phenotype obtained with the sponge with a miR mutant, and also combine miR-1 het with miR sponge (worsening of phenotype?). Alternatively, knockdown efficiency of miR should be tested by qPCR or HCR/smFISH. *

    Answer: We are grateful for these comments. Below we refer to published data and to performed additional experiments that are in support of miR-1 sponge phenotypes:

    • UAS-miR-1 sponge line we used was generated and tested by Fulga et al., (Nat Comm, 2015). Fulga and colleagues apply UAS-miR1sponge line to attenuate miR-1 function in muscles and obtain miR-KO-like muscle phenotypes.
    • Here, we identify Mp as a new direct miR-1 target. To test whether miR-1 sponge attenuates miR-1 function we analyzed Mp protein levels in the hearts from wt and Hand>miR-1sponge flies. Mp expression is highly increased in Hand>miR-1sponge context indicating attenuation of miR-1 by the sponge transgene. These data are presented in new Fig. S5J;
    • We tested whether heterozygous dmiR-1 KO -/+ flies (homozygous dmiR-1 mutants are lethal) develop Hand>miR-1sponge-like heart phenotype. Indeed, at 5 weeks of age dmiR-1 KO -/+ flies show significantly increased diastolic and systolic heart diameters. Thus, in old flies loss of one copy of miR-1 mimics heart dilation observed in Hand>dmiR-1sponge context. Heart contractility remains unaffected in dmiR-1 KO -/+ flies, suggesting that loss of one copy of miR-1 has a weaker impact on heart function than heart-targeted miR-1sponge. These data are shown in a new supplemental figure (Fig. S4A-C).
    • It is surprising that one of the DM1 fly models, overexpression of 960 CTG repeats, did not show DCM, considering it is the primary cause of DM1 in humans due to excessive CTG repeats. It should be discussed why Hand>960 CTG does not lead to DCM, since the authors claim that this model with high number of CTG repeats shows a strong phenotype. Are Hand>bru3 and Hand>mbl stronger? *

    Answer: We thank Reviewer for pointing this out.

    Heart and muscle-specific DM1 models we established and tested (Hand> or Mef>960CTG, Hand> or Mef>mblRNAi and Hand> or Mef>Bru3) all develop the majority of DM1 phenotypes (Picchio et al., 2013 ; Picchio et al., 2018 ; Auxerre-Planté et al., 2019). However, some cardiac DM1 phenotypes such as conduction defects (Auxerre-Plantié et al., 2019) and described here DCM are only observed in Hand>Bru3 and Hand>mblRNAi contexts. We previously observed that the down-regulation of sarcomeric genes is more important in Mef>Bru3 than in Mef>960CTG context (Picchio et al., 2018). This could result from a milder effect of 960CTG repeats on Bru3 and Mbl levels when compared with Gal4-driven overexpression of Bru3 and RNAi-knockdown of mbl. We add a comment to Results section (page 5) to discuss this point: “The Hand>960CTG line shows cardiac dilation at 5 weeks of age characterized by significant increase in diastolic and systolic diameters but with normal cardiac contractility (Fig. S2A,B,C). We hypothesise that non-affected contractility in this DM1 line is due to a milder effect of 960CTG repeats on Bru3 and Mbl levels compared to GAL4-driven overexpression of Bru3 and RNAi-knockdown of mbl.”

    *Is miR-1 (and Mp) unaltered in these flies with 960 CTG repeats? *

    Answer: In Hand>960CTG context a reduced level of miR-1 and an increase in Mp are also observed (not shown). Hand>960CTG flies do not develop DCM but at 5 weeks of age show a significant increase in diastolic and systolic heart diameters. One possibility we favor is that deregulation of miR-1 and Mp in Hand>960CTG is under the level that induces DCM. By analogy, only DM1 patients with a high increase in Col15A1 develop DCM (Fig. 5).

    It would be interesting to overexpress 960 CTG in a miR-1 or mbl heterozygous mutant background, which may produce DCM.

    Answer: To test additional genetic context in which miR-1 is reduced we used miR-1 heterozygous KO flies. We found that in 5 weeks-old flies loss of one copy of miR-1 leads, like in Hand>miR-1sponge flies, to heart dilation (Fig. S4A-C), however mir-1KO +/- flies do not show affected contractility.

    We agree that combining Hand>960CTG with mir-1 heterozygous mutants would potentially result in an additional DCM producing context. As we are focusing on our DCM developing DM1 models (Hand>Bru3 and Hand>mblRNAi) and on conserved deregulation of miR-1 and its target Mp/Col15A1, we didn’t follow this suggestion.

    • In Figure 2H, the mean intensity is displayed as the readout of the smFISH quantification of miR-1 levels. If understood correctly, this is the wrong readout since smFISH detects single molecule fluorescence of transcripts, so the number of transcripts should be quantified. *

    Answer: The miR-1 quantification was done using FISH with miR-1-specific LNA probe (Qiagen miRCURY system). This is highly sensitive ISH but the resolution is not at a single molecule level. The Imaris-generated spots in Fig. 2G and 2G’ represent (for each of them) several miR-1 molecules. The mean intensity of the fluorescent signal for a given spot is proportional to the number of miR-1 molecules and the average of the mean intensities of all spots illustrates the level of miR-1 expression (Fig. 2H). We remove “sm” abbreviations from the text and Figure legend and provide a more detailed description of miR-1 quantification in the Method section.

    Furthermore, Hand-Gal4 is not expressed in the ventral longitudinal muscles (VLM). As a proof of principle, miR-1 levels should be quantified in VLM (no change in transcripts levels expected).

    Answer: When cross with UAS-GFP the Hand-Gal4 expression could be detected in VLMs even if VLM associated GFP signal is much lower than in cardioblasts (CB) and pericardial cells (PC). Representative views of Hand>GFP hearts labeled with anti-GFP are shown below.

    • Fig. 5: Since DCM- DM1 patients still show elevated COL15A1 levels but no DCM, it would be interesting to know if DCM phenotypes are COL15A1-dosage dependent. This could be easily tested in the fly model by testing UAS-Mp overexpression at different temperatures. *

    Answer: Heart parameters are to some extent temperature-sensitive (our observations) thus in our view increasing targeted Mp expression by elevating temperature is not appropriate for heart physiology experiments.

    Presented in the manuscript data on Mp overexpression at 25°C already provide some indication for Mp dose-dependent effect in the fly model. We observe that DCM is induced at both 1 and 5 weeks of age, but the cardiac tube dilation is less important in 1 than in 5 weeks-old flies (Fig. 4). Also, when analysing Hand>960CTG DM1 model we observed that young flies with a low Mp levels do not show cardiac dilation while aged Hand>960CTG flies display an increase in diastolic and systolic heart diameters concomitant with a higher Mp.

    • The authors elegantly show rescue of Hand>Bru3 flies by Mp RNAi. Their model would be further strengthened if a similar rescue can be shown with Hand>mblRNAi. *

    Answer: So far we were unsuccessful in generation of recombined mblRNAi ;MpRNAi line most probably because of incompatibility in chromosomal transgene locations. Thus, we were unable to perform this experiment.

    Because gene deregulations and DCM phenotypes we describe are highly similar in Hand>Bru3 and Hand>mblRNAi context we believe that rescue experiment we provide is representative for both DM1-associated DCM contexts.

    Minor points:

    *Fig. 1A,B: Ventral longitudinal muscles are covering the hearts on these images, so it's difficult to see the heart dimensions. This holds true for images throughout the manuscript. Where were the diameters measured (by the valves)? A better description and illustration would help the reader understand the situation. *

    Answer: In the lateral heart views as in Fig. 1A and B it is indeed difficult to appreciate heart dimensions. For this reason we always show transversal sections derived from 3D reconstruction (as in Fig. 1A’ and B’). In this context differences in the internal heart diameters could be appreciated (white lines). All diastolic and systolic heart diameter measures presented in the graphs are extracted from the SOHA registrations (see Methods).

    *Fig. 1 A',B': White line does not reflect the location where SOHA data are measured and should be horizontal for consistency. Where is ventral vs. dorsal? *

    Answer: We agree. We indicate where is ventral and dorsa. For consistency we remove white lines from panels 1A and 1B and maintain orientations of white lines in panels 1A’ and 1B’.

    Fig. 1D-F: Annotate 1 and 5 weeks in Figure, please. Also, why were 1 and 5 weeks tested? Is there an age-component in DM1 phenotype severity?

    Answer: We add 1 and 5 weeks indications to the figures and discuss in the text (Results section page 4) that 1 and 5 weeks analyses were applied because the severity of cardiac phenotypes increases with age.

    *Fig. 3A: Transcriptional analysis was done at which stage of development? *

    Answer: It was done at 5 weeks of age. We add information to the figure legend.

    *Fig. 3: It is not clear, in which set the authors looked for miR-1 bindings sites (144 genes or the whole set)? Not well annotated. What is meant by 'heart-targeted'? *

    Answer: In silico search was performed on the whole set of genes. We provide more precisions on in silico screen in Method section.

    *Fig. 4C,D: It looks like they are not shown in the same dorsal-ventral orientation. Also, it looks Mp is overexpressed in the VLM, but Hand-Gal4 only drives in the cardiomyocytes and pericardial cells? How was quantification done? *

    Answer: We are thanking for pointing this out. We revised heart orientations in panel 4C and 4D. As previously mentioned Hand-Gal4 is also expressed in the VLM. We present a more representative view in 4D with a lower Mp signal in VLMs. Quantification of Mp expression is not presented here but performed like in Fig. 3G.

    *Fig. 4I: Why are some myofibers indicated in red in the model? *

    Answer: In red are indicated additional actin filaments that form in the case of heart dilation. As we do not discuss this aspect we modify drawing in the model.

    Fig. 5 D-E: Genotypes need to be better indicated in the graphs.

    Answer: We provide now more complete genotypes.

    *Did the authors control for multiple UAS sites? Is UPRT a UAS control? *

    Answer: Yes, UPRT is the UAS line.

    *In the first paragraph of Result 3, the last sentence seems unfinished. "We identified a set of candidate genes, of which Multiplexin (Mp)" *

    Answer: We revise this sentence.

    *In Method, the in silico screening for miR-1 target should be explained in more detail. *

    Answer: We provide a more detailed in silico screening protocol in Method section.

    *Reviewer #1 (Significance (Required)): *

    The presented data is a significant advance our knowledge of our understanding of the molecular mechanisms involved in DM1. I expect that scientists in the muscular disease field and beyond will find this work of high interest. *

    Reviewer #2 (Evidence, reproducibility and clarity (Required)):

    *Summary: This well-written manuscript utilizes the Drosophila model system to demonstrate that reduction of micro-RNA miR-1 and the resulting increase in one of its down-regulated proteins (Multiplexin) contribute to a dilated cardiomyopathy (DCM) phenotype. This is of interest in that this particular micro-RNA is downregulated in myotonic dystrophy type 1 (DM1), and this correlates with the DCM phenotype observed in patients. Further, the authors show that the human ortholog of Multiplexin is enriched in human DM1-DCM hearts and that downregulation of this protein in Drosophila DM1 models improves the DCM phenotype. Hence, the work demonstrates a potential mechanism for disease development and its amelioration. *

    Answer: We are grateful to the Reviewer for appreciating our work and pointing out potential impact of our findings.

    Major comments:

    • As pericardial cells are probed and mentioned substantially in this paper, the authors should explain what these cells do in flies. While affiliated with the heart, they are not myocytes and are probably not particularly relevant to the human heart. In this regard, it is possible that the phenotypes observed in the heart are partially or completely the result of Hand-driven expression of transgenes in the pericardial cells. Although unlikely, this issue should be mentioned as well. * Answer: Hand-Gal4 driver is the most commonly used Drosophila cardiac driver. Regarding influence of Hand-Gal4 driven expression in pericardial cells on the heart phenotypes, we previously tested all our DM1 models using cardioblast-specific Tin-GAL4 driver. All cardiac phenotypes including DCM are also observed when using Tin-Gal4 driver (as shown for conduction defects in Fig.5B Auxerre-Plantié et al., Elife 2019) indicating that the phenotypes are mainly due to gene deregulations within the cardioblasts.

    *Does human miR-1 target Col15A1 transcripts based upon in silico analysis? This issue should be mentioned and discussed. *

    Answer: In silico analysis (new supplemental Fig.S4G) reveals that Col15A1 transcripts carry a perfect miR-1 seed site in 3’UTR region.

    Minor comments:

    • The abstract should explicitly state that Multiplexin is a form of collagen.* Answer: We mention this in the abstract.

    *More information on the identity between the Drosophila and human forms of miR-1 would be helpful to establish that they are conserved. What is the percent identity and are the sequences that target mRNAs homologous? *

    Answer: Mature Drosophila and human miR-1 are highly homologous. We provide their sequences in new supplemental Fig. S4F.

    • In Figure 1C, it appears that there is an increased heartbeat frequency and arrhythmicity. Are these mutant phenotypes as well? *

    Answer: We check it again and do not observe any significant change in heart period or in arrhythmia index in Hand>miR1 sponge context in both young and old flies. We show a new more representative view of M-modes in panel 1C.

    • Incomplete sentence (page 5): We identified a set of candidate genes, of which Multiplexin (Mp) *

    Answer: This sentence was revised.

    *What is the basis for studying Multiplexin function as opposed to other candidates that were identified? It would be useful to mention this in the Results, although it is mentioned in the Discussion ("We top-ranked Mp because of its known role in setting the size of the cardiac lumen"). *

    Answer: We add following sentences to Results section to clarify this point earlier in the manuscript.

    “Mp overexpression in the developing embryonic heart leads to an enlargement of heart lumen and is sufficient to promote an increase of the embryonic aorta diameter to that of the heart proper (Harpaz et al., 2013). We thus reasoned that Mp could be involved in DM1-associated DCM.”

    • "Mp was detected on the luminal and external surfaces of the cardiomyocytes ensuring cardiac contractions" Why does this ensure cardiac contractions? *

    Answer: We are grateful for pointing this out. Mp is not ensuring but could influence cardiac contractions. We revise this sentence by deleting its second part “ensuring cardiac contractions”.

    • Need to state in the text that the increased level of Col15A1 transcript expression in DM1 patients was not statistically significant. *

    Answer: We state this in the text.

    • Need a magnification bar for Figures 5F-H. *

    Answer: Scale bar is added.

    *Please speculate as to why the third DM1 model does not recapitulate the cardiac phenotypes. *

    Answer: Heart and muscle-specific DM1 models we established and tested (Hand> or Mef>960CTG, Hand> or Mef>mblRNAi and Hand> or Mef>Bru3) all develop the majority of DM1 phenotypes (Picchio et al., 2013 ; Picchio et al., 2018 ; Auxerre-Planté et al., 2019). However, some cardiac DM1 phenotypes such as conduction defects (Auxerre-Plantié et al., 2019) and described here DCM are only observed in Hand>Bru3 and Hand>mblRNAi contexts. We previously observed that downregulation of sarcomeric genes is higher in Mef>Bru3 than in Mef>960CTG contexts (Picchio et al., 2018). This could result from a milder effect of 960CTG repeats on Bru3 and Mbl levels when compared with Gal4-driven overexpression of Bru3 and RNAi-knockdown of mbl. We add a comment in Results section (page 5) to discuss this point: “The Hand>960CTG line shows cardiac dilation at 5 weeks of age characterized by significant increase in diastolic and systolic diameters but with normal cardiac contractility (Fig. S2A,B,C). We hypothesise that non-affected contractility in this DM1 line is due to a milder effect of 960CTG repeats on Bru3 and Mbl levels compared to GAL4-driven overexpression of Bru3 and RNAi-knockdown of mbl.”

    • Did the confocal studies indicate whether there was myofibrillar disarray in the heart tubes? *

    Answer: Thank you for this comment. Yes, we observe myofibrillar disarray. We show disarray phenotypes in all DCM developing contexts in a new supplementary figure (Fig.S1).

    • For the statistical comparisons in the figures, please indicate in the legends that statistically significant differences (p

    Answer: We provide this precision in Figure legends.

    *Please more thoroughly explain the UPRT control line. *

    Answer: We provide information about UPRT line in the Results section (p.8).

    *Figure S1 legend: "(red) and 5 (darck)"; the latter should read "(black)" *

    Answer: Revised.

    *Figure S2 panels J and K: it would be helpful to indicate what is being measured on the Y axis, e.g., Mean intensity of dmiR-1 levels. This is true for the various panels in other figures labeled CTCF on their Y axes. *

    Answer: Revised as suggested by the Reviewer.

    *CROSS-CONSULTATION COMMENTS All reviewers agree that this is a well-designed study and that the manuscript is well written. The missing Hand-Gal4 control mentioned by Reviewers 1 and 3 seems an important element that is missing. These reviewers also call into question the FISH quantification methodology. These two issues seem the most critical to resolve. The other additional experiments suggested deserve input from the authors as to whether they already have relevant data that can be cited, whether they are important to pursue or if they go beyond the scope of the current study. Reviewers 1 and 2 agree that further discussion of the fly model that does not show DCM should be provided. The question on fibrosis in the fly models is germane (Reviewer 3), although it might be indirectly addressed by the fact that a collagen molecule is upregulated here (a major player in fibrosis). All of the minor comments are reasonable and should be addressed by the authors. *

    Answer: We provide answer to all these comments.

    *Reviewer #2 (Significance (Required)): *

    This paper is significant in that it draws a more direct connection between the reduction in a microRNA that occurs in myotonic dystrophy and dilated cardiomyopathy that is affiliated with this disease. It shows that a form of collagen that is overexpressed in both Drosophila models and humans with DM1-caused DCM is causative/correlated with the increased heart diameters. Thus, the fly model provides important insights into the link between the mutant gene and the cardiac phenotype. This work will be of interest to those studying skeletal and cardiac muscle disease and scientists interested in developing potential therapeutics for treating DM1-caused DCM. Note that my expertise is in producing and studying skeletal muscle and cardiac disease models in the Drosophila system, which is relevant to evaluating this paper and defining its significance in the field. *

    Reviewer #3 (Evidence, reproducibility and clarity (Required)):

    *A current study demonstrates that miR-1 targets the newly identified heat-specific target Multiplexin, which, when upregulated, exhibits similar phenotypes observed for the Drosophila DM1 model. Furthermore, the authors additionally confirm some of their results using samples derived from DM1 patients and support the data obtained in flies. Overall, it is a good study with well-performed experiments. The data presented in the paper are convincing and most of the claims the authors provide are supported by their findings. Manuscript is clearly written and easy to read and understand. Statistical analysis and the description of the methods are appropriate. It is an interesting paper and I would highly support it to be accepted for publication; however, I few comments I would like authors to address. *

    Answer: We are grateful to the Reviewer for his enthusiastic and supportive comments on our work.

    Major comments:

    • The cardiac dilation/fractional shortening phenotype in Hand>dmiR-1-KD flies is only observed in young flies but not in old flies. However, heart-targeted Mp overexpression leads to DCM in aged flies. Could authors comment on this? * Answer: We speculate that attenuation of miR-1 in the heart could lead to more drastic pro-DCM alterations thus leading to earlier phenotypes than in the case of Mp overexpression.

    To better assess DCM phenotypes we included additional Hand-Gal4/+ control context and more heart samples from 5 weeks-old flies. These additional analyses presented in revised Fig. 1D-F reveal that old Hand>dmiR-1KD flies, like the young one, also develop DCM phenotype.

    • Since the HAND-Gal4 line was used to drive multiple transgenes, it would be important to have the cardiac dilation/fractional shortening phenotype measured in this line as a control. *

    Answer: We performed these control experiments and suggested by the reviewer they are now included to the graphs.

    • The use of a sponge line is highly appreciated as it allows for tissue-specific downregulation of miRNA. However, to corroborate the data, I would recommend including the knockdown mutant that is available in Bloomington as additional confirmation since no qPCR is provided for the efficacy of the sponge line. This line could also be used in combination with reporter lines to perform a targeting experiment. *

    Answer: We are grateful for these comments. Below we refer to performed additional experiments:

    • We tested whether heterozygous dmiR-1 KO -/+ flies (homozygous dmiR-1 mutants are lethal) develop Hand>miR-1sponge-like heart phenotype. Indeed, at 5 weeks of age dmiR-1 KO -/+ flies show significantly increased diastolic and systolic heart diameters. Thus, in old flies loss of one copy of miR-1 mimics heart dilation observed in Hand>dmiR-1sponge context. Heart contractility remains unaffected in dmiR-1 KO -/+ flies, suggesting that loss of one copy of miR-1 has a weaker impact on heart function than heart-targeted miR-1sponge. These data are shown in a new supplemental figure (Fig. S4A-C).
    • Here, we identify Mp as a new direct miR-1 target. To test whether miR-1 sponge attenuates miR-1 function we analyzed Mp protein levels in the hearts from wt and Hand>miR-1sponge flies. Mp expression is highly increased in Hand>miR-1sponge context indicating attenuation of miR-1 by the sponge transgene. These data are presented in new Fig. S5J;
    • The authors state that the reduced miR-1 levels have already been shown in DM1 patients. It would be a stronger argument if similar downregulation was shown in patient samples used in this manuscript (qPCR would be sufficient). *

    Answer: We performed suggested by the reviewer analyses of miR-1 in patient samples. We show that miR-1 is indeed down regulated. These new data supporting conserved pro-DCM deregulation of miR-1 and its target Mp/Col15A1 are shown in new Fig 5C.

    • Because fibrosis is a hallmark of myotonic dystrophy, do the authors have some makers or other methods to test whether observed phenotypes are due to fibrosis? *

    Answer: Fibrosis (replacement of muscle by fibrotic tissue) has not been reported in Drosophila and is not associated with degeneration of body wall or cardiac Drosophila muscles in so far described fly models of human muscular dystrophies. However, one could speculate that increase in Mp/Col15A1 levels within the ECM of diseased DM1 cardiac cells we observe, could have, a fibrotic-like, negative effect on cardiac function.

    • The explanation of the observation that pre-miR-1 levels are down-regulated only in young flies, whereas old flies show an opposite tendency, is missing. *

    Answer: Accumulation of pre-miR-1 in old flies is most probably due to the affected processing mediated by mbl. This is correlated with the reduction of the mature miR-1.

    The authors suggested that this is due to "impaired processing". To corroborate this interesting hypothesis, the authors performed only the smFISH intensity analyses, which are somewhat difficult to decipher. I would recommend, in addition to the pre-miRNA levels, to test and compare the mature miRNA expression using TaqMan qPCR.

    Answer: Impaired miR-1 processing is supported by the previous studies in human cells (Rau et al., 2011) and in Drosophila models (Fernandez-Costa et al., 2013). We believe that our method of quantification of miR-1 expression via highly sensitive miRCURY LNA FISH is a well-adapted method. It was performed with all necessary controls. In the method section we provide now more details for the LNA FISH based miR-1 quantification approach.

    In parallel, TaqManPCR-based miR-1 quantification was performed for human cardiac samples from DM1 patients.

    • The relationship between Bru3 and miR-1 shown in the schematic is not well-defined and would rather require a question mark or dotted line, as the authors provide no evidence that Bru3 can be directly involved in miR-1 processing. The authors suggest that CELF1 may bind UG-rich miRNAs and mediate their degradation by recruiting poly(A)-specific ribonuclease (PARN), but this is only a hypothesis and does not justify the placement of a direct line of repression on the schematic in the last figure. *

    Answer: We agree and modify scheme accordingly.

    • I also feel that the authors did not clearly explain the cardiac phenotypes in terms of systolic and diastolic diameter measurements. Which parameters clearly represent the DM1 model, specifically higher or lower diameters of systole and diastole? Results should be clearly indicated in figure legends. *

    Answer: We provide appropriate precisions in figure legends.

    Minor comments:

    • The full name of CELF1 on page 2: CUGBP Elav-like family member 1 should be added*. Answer: Revised
    • For better readability of the text and corresponding figures, consistent use of UAS-Mp or UAS-3HNC1 is recommended, but not a mixture of both. *

    Answer: We consistently use UAS-Mp in the revised version

    • Why is the Multiplexin overexpression line called UAS-3HNC1? *

    Answer: This is the name that resumes protein Mp domains: Collagen tripple helix and trimerization region (3H) and NC1 domain (C-terminal non-triple helical domain) comprising Endostatin domain. We provide this information in Methods section.

    • For all figures, it would be better if the genotypes were indicated in the panels and the graphs had the age of the flies instead of color coding. *

    Answer: We revised these points as suggested.

    • Figure 5. Were technical replicates performed for the western blot shown in 5B? *

    Answer: We didn’t perform technical replicates because of limited human sample amounts

    • Figure S1-S2. Why the data for qPCR of miR-1 is in figure S1 and not S2? *

    Answer: In the revised version all supplemental analyses on miR1 are included to the new Fig. S4

    • Figure S4. Misspelling in figure legend: Scale "barre" instead of scale "bar".*

    Answer:Revised

    Reviewer #3 (Significance (Required)):

    The manuscript "Deregulations of miR-1 and its target Multiplexin promote dilated cardiomyopathy associated with myotonic dystrophy type 1" by Souidi et al., reports a novel role of identified Multiplexin (Mp) as a new cardiac miR-1 target involved in myotonic dystrophy type 1 (DM1) using Drosophila as a model system. Myotonic dystrophy type 1 (MD1) is a severe disease that results in a multisystem disorder affecting the skeletal and smooth muscles as well as the eye, heart, endocrine system, and central nervous system. At the moment, no appropriate treatment has been identified to prevent it. Previous studies have also shown that heart-specific miR-1 levels are reduced in patients with DM1, but the role and targets of this miRNA in the heart have not been analyzed. Research presented in this paper is of a broad interest and provide new evidence that will help to better understating DM1 on molecular level. It will be interesting not only to scientists from the Drosophila field but will also contribute to medical research field.*

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    A current study demonstrates that miR-1 targets the newly identified heat-specific target Multiplexin, which, when upregulated, exhibits similar phenotypes observed for the Drosophila DM1 model. Furthermore, the authors additionally confirm some of their results using samples derived from DM1 patients and support the data obtained in flies. Overall, it is a good study with well-performed experiments. The data presented in the paper are convincing and most of the claims the authors provide are supported by their findings. Manuscript is clearly written and easy to read and understand. Statistical analysis and the description of the methods are appropriate. It is an interesting paper and I would highly support it to be accepted for publication; however, I few comments I would like authors to address.

    Major comments:

    1. The cardiac dilation/fractional shortening phenotype in Hand>dmiR-1-KD flies is only observed in young flies but not in old flies. However, heart-targeted Mp overexpression leads to DCM in aged flies. Could authors comment on this?

    2. Since the HAND-Gal4 line was used to drive multiple transgenes, it would be important to have the cardiac dilation/fractional shortening phenotype measured in this line as a control.

    3. The use of a sponge line is highly appreciated as it allows for tissue-specific downregulation of miRNA. However, to corroborate the data, I would recommend including the knockdown mutant that is available in Bloomington as additional confirmation since no qPCR is provided for the efficacy of the sponge line. This line could also be used in combination with reporter lines to perform a targeting experiment.

    4. The authors state that the reduced miR-1 levels have already been shown in DM1 patients. It would be a stronger argument if similar downregulation was shown in patient samples used in this manuscript (qPCR would be sufficient).

    5. Because fibrosis is a hallmark of myotonic dystrophy, do the authors have some makers or other methods to test whether observed phenotypes are due to fibrosis?

    6. The explanation of the observation that pre-miR-1 levels are down-regulated only in young flies, whereas old flies show an opposite tendency, is missing. The authors suggested that this is due to "impaired processing". To corroborate this interesting hypothesis, the authors performed only the smFISH intensity analyses, which are somewhat difficult to decipher. I would recommend, in addition to the pre-miRNA levels, to test and compare the mature miRNA expression using TaqMan qPCR.

    7. The relationship between Bru3 and miR-1 shown in the schematic is not well-defined and would rather require a question mark or dotted line, as the authors provide no evidence that Bru3 can be directly involved in miR-1 processing. The authors suggest that CELF1 may bind UG-rich miRNAs and mediate their degradation by recruiting poly(A)-specific ribonuclease (PARN), but this is only a hypothesis and does not justify the placement of a direct line of repression on the schematic in the last figure.

    8. I also feel that the authors did not clearly explain the cardiac phenotypes in terms of systolic and diastolic diameter measurements. Which parameters clearly represent the DM1 model, specifically higher or lower diameters of systole and diastole? Results should be clearly indicated in figure legends.

    Minor comments:

    1. The full name of CELF1 on page 2: CUGBP Elav-like family member 1 should be added.

    2. For better readability of the text and corresponding figures, consistent use of UAS-Mp or UAS-3HNC1 is recommended, but not a mixture of both.

    3. Why is the Multiplexin overexpression line called UAS-3HNC1?

    4. For all figures, it would be better if the genotypes were indicated in the panels and the graphs had the age of the flies instead of color coding.

    5. Figure 5. Were technical replicates performed for the western blot shown in 5B?

    6. Figure S1-S2. Why the data for qPCR of miR-1 is in figure S1 and not S2?

    7. Figure S4. Misspelling in figure legend: Scale "barre" instead of scale "bar".

    Significance

    The manuscript "Deregulations of miR-1 and its target Multiplexin promote dilated cardiomyopathy associated with myotonic dystrophy type 1" by Souidi et al., reports a novel role of identified Multiplexin (Mp) as a new cardiac miR-1 target involved in myotonic dystrophy type 1 (DM1) using Drosophila as a model system. Myotonic dystrophy type 1 (MD1) is a severe disease that results in a multisystem disorder affecting the skeletal and smooth muscles as well as the eye, heart, endocrine system, and central nervous system. At the moment, no appropriate treatment has been identified to prevent it. Previous studies have also shown that heart-specific miR-1 levels are reduced in patients with DM1, but the role and targets of this miRNA in the heart have not been analyzed. Research presented in this paper is of a broad interest and provide new evidence that will help to better understating DM1 on molecular level. It will be interesting not only to scientists from the Drosophila field but will also contribute to medical research field.

  4. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    Summary:

    This well-written manuscript utilizes the Drosophila model system to demonstrate that reduction of micro-RNA miR-1 and the resulting increase in one of its down-regulated proteins (Multiplexin) contribute to a dilated cardiomyopathy (DCM) phenotype. This is of interest in that this particular micro-RNA is downregulated in myotonic dystrophy type 1 (DM1), and this correlates with the DCM phenotype observed in patients. Further, the authors show that the human ortholog of Multiplexin is enriched in human DM1-DCM hearts and that downregulation of this protein in Drosophila DM1 models improves the DCM phenotype. Hence, the work demonstrates a potential mechanism for disease development and its amelioration.

    Major comments:

    1. As pericardial cells are probed and mentioned substantially in this paper, the authors should explain what these cells do in flies. While affiliated with the heart, they are not myocytes and are probably not particularly relevant to the human heart. In this regard, it is possible that the phenotypes observed in the heart are partially or completely the result of Hand-driven expression of transgenes in the pericardial cells. Although unlikely, this issue should be mentioned as well.

    2. Does human miR-1 target Col15A1 transcripts based upon in silico analysis? This issue should be mentioned and discussed.

    Minor comments:

    1. The abstract should explicitly state that Multiplexin is a form of collagen.

    2. More information on the identity between the Drosophila and human forms of miR-1 would be helpful to establish that they are conserved. What is the percent identity and are the sequences that target mRNAs homologous?

    3. In Figure 1C, it appears that there is an increased heartbeat frequency and arrhythmicity. Are these mutant phenotypes as well?

    4. Incomplete sentence (page 5): We identified a set of candidate genes, of which Multiplexin (Mp)

    5. What is the basis for studying Multiplexin function as opposed to other candidates that were identified? It would be useful to mention this in the Results, although it is mentioned in the Discussion ("We top-ranked Mp because of its known role in setting the size of the cardiac lumen").

    6. "Mp was detected on the luminal and external surfaces of the cardiomyocytes ensuring cardiac contractions" Why does this ensure cardiac contractions?

    7. Need to state in the text that the increased levels of Col15A1 transcript expression in DM1 patients was not statistically significant.

    8. Need a magnification bar for Figures 5F-H.

    9. Please speculate as to why the third DM1 model does not recapitulate the cardiac phenotypes.

    10. Did the confocal studies indicate whether there was myofibrillar disarray in the heart tubes?

    11. For the statistical comparisons in the figures, please indicate in the legends that statistically significant differences (p<0.05) are shown.

    12. Please more thoroughly explain the UPRT control line.

    13. Figure S1 legend: "(red) and 5 (darck)"; the latter should read "(black)"

    14. Figure S2 panels J and K: it would be helpful to indicate what is being measured on the Y axis, e.g., Mean intensity of dmiR-1 levels. This is true for the various panels in other figures labeled CTCF on their Y axes.

    CROSS-CONSULTATION COMMENTS

    All reviewers agree that this is a well-designed study and that the manuscript is well written. The missing Hand-Gal4 control mentioned by Reviewers 1 and 3 seems an important element that is missing. These reviewers also call into question the FISH quantification methodology. These two issues seem the most critical to resolve. The other additional experiments suggested deserve input from the authors as to whether they already have relevant data that can be cited, whether they are important to pursue or if they go beyond the scope of the current study. Reviewers 1 and 2 agree that further discussion of the fly model that does not show DCM should be provided. The question on fibrosis in the fly models is germane (Reviewer 3), although it might be indirectly addressed by the fact that a collagen molecule is upregulated here (a major player in fibrosis). All of the minor comments are reasonable and should be addressed by the authors.

    Significance

    This paper is significant in that it draws a more direct connection between the reduction in a microRNA that occurs in myotonic dystrophy and dilated cardiomyopathy that is affiliated with this disease. It shows that a form of collagen that is overexpressed in both Drosophila models and humans with DM1-caused DCM is causative/correlated with the increased heart diameters. Thus, the fly model provides important insights into the link between the mutant gene and the cardiac phenotype. This work will be of interest to those studying skeletal and cardiac muscle disease and scientists interested in developing potential therapeutics for treating DM1-caused DCM. Note that my expertise is in producing and studying skeletal muscle and cardiac disease models in the Drosophila system, which is relevant to evaluating this paper and defining its significance in the field.

  5. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    The paper titled "Deregulations of miR-1 and its target Multiplexin promote dilated cardiomyopathy associated with myotonic dystrophy type 1" by the Jagla group studied the effect of down-regulation of miR-1 in myotonic dystrophy type 1 (DM1) using fly as the disease model. The study is based on previous findings that in DM1 MBNL1 is sequestered, CELF1 is stabilized, and miR-1 is down regulated. The authors further identified Multiplexin to be the target effector of miR-1 in the fly heart and studied its function with a series of gain- and loss-of-function and rescue experiments. The authors' findings represent a significant advance in understanding the genetic mechanisms that can explain the pathogenic causes of dilated cardiomyopathy associated with DM1. Overall, this paper is well written and organized, with well-designed experiments and a clear model. A few additional experiments are suggested to further strengthen the conclusion.

    Major points:

    1. Wild-type and Hand-Gal4 controls are missing for all experiments (only UAS lines outcrossed to w1118 are displayed). Also, Hand-Gal4 driver lines can cause mild dilation by itself, which would influence interpretation and statistics of data. Also, the authors should consider to confirm the miR-1 phenotype obtained with the sponge with a miR mutant, and also combine miR-1 het with miR sponge (worsening of phenotype?). Alternatively, knockdown efficiency of miR should be tested by qPCR or HCR/smFISH.

    2. It is surprising that one of the DM1 fly models, overexpression of 960 CTG repeats, did not show DCM, considering it is the primary cause of DM1 in humans due to excessive CTG repeats. It should be discussed why Hand>960 CTG does not lead to DCM, since the authors claim that this model with high number of CTG repeats shows a strong phenotype. Are Hand>bru3 and Hand>mbl stronger? Is miR-1 (and Mp) unaltered in these flies with 960 CTG repeats? It would be interesting to overexpress 960 CTG in a miR-1 or mbl heterozygous mutant background, which may produce DCM.

    3. In Figure 2H, the mean intensity is displayed as the readout of the smFISH quantification of miR-1 levels. If understood correctly, this is the wrong readout since smFISH detects single molecule fluorescence of transcripts, so the number of transcripts should be quantified. Furthermore, Hand-Gal4 is not expressed in the ventral longitudinal muscles (VLM). As a proof of principle, miR-1 levels should be quantified in VLM (no change in transcripts levels expected).

    4. Fig. 5: Since DCM- DM1 patients still show elevated COL15A1 levels but no DCM, it would be interesting to know if DCM phenotypes are COL15A1-dosage dependent. This could be easily tested in the fly model by testing UAS-Mp overexpression at different temperatures.

    5. The authors elegantly show rescue of Hand>Bru3 flies by Mp RNAi. Their model would be further strengthened if a similar rescue can be shown with Hand>mblRNAi.

    Minor points:

    1. Fig. 1A,B: Ventral longitudinal muscles are covering the hearts on these images, so it's difficult to see the heart dimensions. This holds true for images throughout the manuscript. Where were the diameters measured (by the valves)? A better description and illustration would help the reader understand the situation.

    2. Fig. 1 A',B': White line does not reflect the location where SOHA data are measured and should be horizontal for consistency. Where is ventral vs. dorsal?

    3. Fig. 1D-F: Annotate 1 and 5 weeks in Figure, please. Also, why were 1 and 5 weeks tested? Is there an age-component in DM1 phenotype severity?

    4. Fig. 3A: Transcriptional analysis was done at which stage of development?

    5. Fig. 3: It is not clear, in which set the authors looked for miR-1 bindings sites (144 genes or the whole set)? Not well annotated. What is meant by 'heart-targeted'?

    6. Fig. 4C,D: It looks like they are not shown in the same dorsal-ventral orientation. Also, it looks Mp is overexpressed in the VLM, but Hand-Gal4 only drives in the cardiomyocytes and pericardial cells? How was quantification done?

    7. Fig. 4I: Why are some myofibers indicated in red in the model?

    8. Fig. 5 D-E: Genotypes need to be better indicated in the graphs. Did the authors control for multiple UAS sites? Is UPRT a UAS control?

    In the first paragraph of Result 3, the last sentence seems unfinished. "We identified a set of candidate genes, of which Multiplexin (Mp)"

    In Method, the in silico screening for miR-1 target should be explained in more detail.

    Significance

    The presented data is a significant advance our knowledge of our understanding of the molecular mechanisms involved in DM1. I expect that scientists in the muscular disease field and beyond will find this work of high interest.

  6. Version published to 10.1101/2022.09.06.506816 on bioRxiv