RNA binding protein IGF2BP2 expression is induced by stress in the heart and mediates dilated cardiomyopathy

The IGF2BP family of RNA binding proteins consists of three paralogs that regulate intracellular RNA localization, RNA stability, and translational control. Although IGF2BP1 and 3 are oncofetal proteins, IGF2BP2 expression is maintained in many tissues, including the heart, into adulthood. IGF2BP2 is upregulated in cardiomyocytes during cardiac stress and remodeling and returns to normal levels in recovering hearts. We wondered whether IGF2BP2 might play an adaptive role during cardiac stress and recovery. Enhanced expression of an IGF2BP2 transgene in a conditional, inducible mouse line leads to dilated cardiomyopathy (DCM) and death within 3-4 weeks in newborn or adult hearts. Downregulation of the transgene after 2 weeks, however, rescues these mice, with complete recovery by 12 weeks. Hearts overexpressing IGF2BP2 downregulate sarcomeric and mitochondrial proteins and have fragmented mitochondria and elongated, thinner sarcomeres. IGF2BP2 is also upregulated in DCM or myocardial infarction patients. These results suggest that IGF2BP2 may be an attractive target for therapeutic intervention in cardiomyopathies.

1.The authors performed a proteomic analysis to investigate the mechanism by which exogenous expression of human IGF2BP2 induces cardiomyopathy.However, given that IGF2BP2 functions as a reader for m6A mRNA modifications, the primary target of IGFPB2 should be the mRNA.It is likely that overexpression of hIGF2BP2 perturbs the transcriptome by altering the stability or localization of m6A-containing mRNAs that interact with IGF2BP2, ultimately leading to the translational defects.To elucidate the mechanism, the authors need to perform m6A RNAseq and/or conventional RNAseq in both control and hIGF2BP2-expressing hearts.Such experiments will provide valuable insights into the molecular mechanisms underlying the development of cardiomyopathy induced by IGF2BP2 overexpression.
2. Fig. 2F-Fig.4: The observation that induction of hIGF2BP2 expression beyond 16 days, but not 14 days, leads to mortality and cardiac dysfunction suggests that the duration of hIGF2BP2 protein induction is critical for myocardial fate.However, the temporal dynamics of hIGF2BP2 protein have not been studied in detail.To better understand this phenomenon, the authors should assess the protein levels of huIGF2BP2 at different time points.In particular, it is crucial to determine the time point at which hIGF2BP2 protein becomes detectable after transgene induction and the time point at which hIGF2BP2 disappears from the hearts after transgene elimination.
Minor comments: 3. Fig. 1C: It is not clear whether the authors generated the sFlt-expressing mice and obtained the expression data from their mice or reanalyzed the expression data from the previous study.If the former is the case, the authors need to add a description of the transgenic mice in the Methods section.If the latter is the case, Figure 1c should be moved to Supplementary Figure 1. 4. Fig. 1E-F: The experimental condition of Figure 1E-F is not clear.Because the expression level of IGF2BP2 seems to peak 24~48 h after TAC surgery and decrease to the basal level (Supplementary Figure 1A), authors need to specify the time point related to the data of Fig. 1E-F.In addition, authors need to describe the TAC procedure in detail in the Methods section.For Figure 1F, DAPI and IGF2BP2 images should be shown separately.The authors describe the role of IGF2BP2 in early-stage cardiac stress.This elegant study includes mouse models, primary cell cultures, and human patient material all confirming their findings.The study paves the path for novel studies into the potential therapeutic implications of IGF2BP2.
Please remove the double spaces throughout the manuscript (lines 218, 223, 227, 250, and perhaps more) Line 86-98 describes the expression of IGF2BP2 in cardiac stress models in literature and therefore seems to be introduction or discussion, more than something for the results section Line 122: please remove the remark "(see Materials and Methods)" Line 135: please explain what is meant by "sparse" Line 141: please refer to Figure 2G Line 192: minimal fibrosis is described in mouse V, however on the picture it seems to have increased fibrosis more on the epicardial side of the heart, which would be different from the other hearts.Please describe if this is the case.Line 231: MF20 was stained in the primary cardiomyocytes, however, it would be better to stain for any of the sarcomeric proteins like αActinin (line 246) which is down-regulated (antibody available according to the manuscript) and also one of the upregulated proteins Lines 259-270: This part rather seems to be a discussion instead of a result section.Line 276: the text indicates that multiple sections of human HCM patients have been used, however, it is not indicated how many samples have been used.Also, the qualitative scoring should be included in the figure.Line 415: dilution missing for αActinin Western blot Line 525: Unfortunately, just one gene (HPRT) is used as a reference gene for normalization.Why not perform proper normalization by measuring multiple refgenes and define the optimal number by genorm  Table 1: Please add that you are looking at protein expression in the legend Reviewer #3 (Remarks to the Author): The manuscript by Krumbein et all investigates the effect of IGFBP2 over-expression in mouse hearts using a conditional tg model.The authors show that IGFBP2 over-expression is detrimental and that this response is time-dependent.The authors further show that the several proteins are dysregulated in these animals, and that these proteins are mostly mitochondrial or sarcomeric.
The main issue with the paper is the conclusion that IGFBP2 has a direct impact on translation of these mRNAs.The observed changes can be the consequence of pathologic remodeling and not a direct consequence of IGFBP2 over-expression.Immunoprecipitation would be needed to make these statements.This reviewer suggests adding this to the limitations and to limit these conclusions in the text.
In Figure 1B, the levels of IGF2BP2 seem highly variable in P1.The quantification of IGF2BP1 is higher than BP2, but that is not a reflection of the blots.A better blot is needed.
The variability in the response after 16 days of transgene expression is surprising.What is the sex of those animals?
The authors state that several proteins are evaluated by Western blot, whereas only 4 are evaluated.Only one is a mitochondrial protein.Proteomics is sufficient for the protein analysis, but the authors should minimize the relative importance of the Western data in the text.
The authors can speculate on the effect of IGFBP2 on RNA stability in the Discussion, but not in the results, as they have no evidence that it is increasing RNA stability.
Quantification in Figure 7 needs to be shown as a bar graph.This reviewer agrees that timing of IGFBP2 regulation and suppression are likely important for a potential recovery.The authors need to show a time-course of IGFBP2 increased levels and the relationship to TAC and sFLT up-regulation and the association between its levels and cardiac function.
Line 328 -the authors state that they observed association of IGFBP2 with mitochondrial RNAs.The authors don't provide any association data.Do they mean association in the levels of IGFBP2 and RNAs?As written, this is misleading.In addition, I suspect these are not coded by the mitochondrial genome, correct?Please clarify if any of the identified proteins are coded by the mitochondria.

Dear Referees,
Thank you for your comments and recommendations.We have addressed all of the points mentioned in the reports.Each point is listed below, with our responses highlighted in yellow.Figures 1,2,5, and 7 and Extended Data Fig. 6 have been changed according to your requests, and new data has been added in Extended Data Figures 2A, B, F, and 5.We hope that you will agree with us that the paper has been significantly improved and is now appropriate for publication in Communications Biology.

Sincerely, Joel Yisraeli
Reviewer #1 (Remarks to the Author): Major Comments: 1.The authors performed a proteomic analysis to investigate the mechanism by which exogenous expression of human IGF2BP2 induces cardiomyopathy.However, given that IGF2BP2 functions as a reader for m6A mRNA modifications, the primary target of IGFPB2 should be the mRNA.It is likely that overexpression of hIGF2BP2 perturbs the transcriptome by altering the stability or localization of m6A-containing mRNAs that interact with IGF2BP2, ultimately leading to the translational defects.To elucidate the mechanism, the authors need to perform m6A RNAseq and/or conventional RNAseq in both control and hIGF2BP2-expressing hearts.Such experiments will provide valuable insights into the molecular mechanisms underlying the development of cardiomyopathy induced by IGF2BP2 overexpression.
We certainly agree that the question of the mechanism by which IGF2BP2 is working is very interesting.As the editors indicated, we have chosen to take a targeted approach to address the reviewer's comment.We have now performed rtPCR on 7 different genes that were downregulated in the proteomics data.The results, presented in Extended Data Fig. 5, show that some RNAs are downregulated, while others are not.We suspect this indicates that IGF2BP2 may be working through different regulatory pathways.The different possibilities are discussed in the Discussion.(Lines 231-238 and 331-345) 2. Fig. 2F-Fig.4: The observation that induction of hIGF2BP2 expression beyond 16 days, but not 14 days, leads to mortality and cardiac dysfunction suggests that the duration of hIGF2BP2 protein induction is critical for myocardial fate.However, the temporal dynamics of hIGF2BP2 protein have not been studied in detail.To better understand this phenomenon, the authors should assess the protein levels of huIGF2BP2 at different time points.In particular, it is crucial to determine the time point at which hIGF2BP2 protein becomes detectable after transgene induction and the time point at which hIGF2BP2 disappears from the hearts after transgene elimination.
The timing of the induction and silencing of IGF2BP2 expression in the transgenic mice indeed bears on the question of the rescue experiment.We have added a figure showing the kinetics of IGF2BP2 expression following withdrawal or addition of Tet to the drinking water.These data are shown in Extended Data Fig. 2A and  B.
Minor comments: 3. Fig. 1C: It is not clear whether the authors generated the sFlt-expressing mice and obtained the expression data from their mice or reanalyzed the expression data from the previous study.If the former is the case, the authors need to add a description of the transgenic mice in the Methods section.If the latter is the case, Figure 1c should be moved to Supplementary Figure 1.
We apologize for the lack of clarity.We did in fact reanalyze data from a previous study.The figure now appears in Extended Data Fig. 1D and E. 4. Fig. 1E-F: The experimental condition of Figure 1E-F is not clear.Because the expression level of IGF2BP2 seems to peak 24~48 h after TAC surgery and decrease to the basal level (Supplementary Figure 1A), authors need to specify the time point related to the data of Fig. 1E-F.In addition, authors need to describe the TAC procedure in detail in the Methods section.For Figure 1F, DAPI and IGF2BP2 images should be shown separately.
The details of the TAC surgery have been added to the Materials and Methods section (lines 401-403) along with the appropriate references.The RNA samples were taken 10 weeks after the surgery, and this is now explicitly mentioned in the legend and the M&M description.We apologize for any misunderstanding regarding the protocol.We have added a scale bar to the figure (which is now Extended Data Fig. 6).

line 184-185, related to
Reviewer #2 (Remarks to the Author): The authors describe the role of IGF2BP2 in early-stage cardiac stress.This elegant study includes mouse models, primary cell cultures, and human patient material all confirming their findings.The study paves the path for novel studies into the potential therapeutic implications of IGF2BP2.
Please remove the double spaces throughout the manuscript (lines 218, 223, 227, 250, and perhaps more) I have tried to remove them, but in some cases it was a result of the Word justification program.In addition, I was taught that after a period, one puts in a double space before beginning the next sentence.
Line 86-98 describes the expression of IGF2BP2 in cardiac stress models in literature and therefore seems to be introduction or discussion, more than something for the results section Because this discussion includes new analysis of previously published data (Extended Data Fig. 1A,B,and C, we felt it was more appropriate in the results section. Line 122: please remove the remark "(see Materials and Methods)" removed Line 135: please explain what is meant by "sparse" Changed to "less dense" (line 134) Line 141: please refer to Figure 2G The figure reference has been added (it is now Fig. 2E).
Line 192: minimal fibrosis is described in mouse V, however on the picture it seems to have increased fibrosis more on the epicardial side of the heart, which would be different from the other hearts.Please describe if this is the case.
The pathologist who examined the sections was not certain that the slight differences to which the reviewer refers were significant or perhaps just a result of fixation/sectioning/staining.Line 231: MF20 was stained in the primary cardiomyocytes, however, it would be better to stain for any of the sarcomeric proteins like αActinin (line 246) which is down-regulated (antibody available according to the manuscript) and also one of the upregulated proteins The reviewer raises an interesting question about whether we could see in primary cardiomyocytes the upregulation or downregulation of the proteins that we observed in the induced hearts.We were interested, however, in simply looking to see if the organization of the sarcomere was affected.The question of how well the primary cardiomyocytes reflect what is happening in the whole heart is a question we hope to address in the future.
Lines 259-270: This part rather seems to be a discussion instead of a result section.
The section has been slightly changed to reflect the addition of new data (OPA1 rtPCR).We felt that it was important to give a bit of context for the experiments.(lines 267-269) Line 276: the text indicates that multiple sections of human HCM patients have been used, however, it is not indicated how many samples have been used.Also, the qualitative scoring should be included in the figure .The number of samples is now indicated in the legend, and the scoring is now shown as a bar graph.
Line 415: dilution missing for αActinin Western blot In order to streamline the paper a bit, this western blot has been removed.
Line 525: Unfortunately, just one gene (HPRT) is used as a reference gene for normalization.Why not perform proper normalization by measuring multiple refgenes and define the optimal number by genorm We have done new rtPCRs using two reference genes and analyzed accordingly.These are shown in Extended Data Fig. 5 and primers are listed in Materials and Methods.The manuscript by Krumbein et all investigates the effect of IGFBP2 over-expression in mouse hearts using a conditional tg model.The authors show that IGFBP2 over-expression is detrimental and that this response is time-dependent.The authors further show that the several proteins are dysregulated in these animals, and that these proteins are mostly mitochondrial or sarcomeric.
The main issue with the paper is the conclusion that IGFBP2 has a direct impact on translation of these mRNAs.The observed changes can be the consequence of pathologic remodeling and not a direct consequence of IGFBP2 over-expression.Immunoprecipitation would be needed to make these statements.This reviewer suggests adding this to the limitations and to limit these conclusions in the text.
We agree completely with the reviewer about the limitations of conclusions about how IGF2BP2 is functioning.In addition to adding new rtPCR experiments on a number of RNAs encoding downregulated proteins (Extended Data Fig. 5), we have rewritten the sections related to possible modes of action for IGF2BP2, taken speculations out of the Results section, and put in the Discussion possibilities for how the protein may be functioning.In Figure 1B, the levels of IGF2BP2 seem highly variable in P1.The quantification of IGF2BP1 is higher than BP2, but that is not a reflection of the blots.A better blot is needed.
Because each antibody has its own affinity, it is not possible (with simple western blots) to compare the expression of the IGF2BP paralogs to each other but only relatively to themselves, between time points.Indeed, the way we presented the data was confusing.We have now separated the data, presenting on each graph the normalized, relative amount of one of the paralogs.
5. line 184-185, related to Fig. 4: To claim significance, the authors must perform a Kaplan-Meier test and report the P value.6. Supplemental fig 5: No scale bars in the images.Reviewer #2 (Remarks to the Author):

Figure 1 :
Figure 1: This figure is messy due to several reasons: 1) The bar graphs are depicted in different styles a. B is entirely different from the rest b.A and C have a title whereas E does not 2) The ChIP (please check the legend how it should be written) data is barely visible due to low quality 3) The IF staining should show a view on a higher magnification 4) The bar graph should include the dot plots as shown in 1N, 1O, 1P, 2, and 6

Figure 7 :
Figure 7: For comparison, this Figure would be more illustrative when a healthy heart would be included.
Fig. 4: To claim significance, the authors must perform a Kaplan-Meier test and report the P value.The text has been changed to "…the outcome in this group was variable" (line 184) 6. Supplemental fig 5: No scale bars in the images.

Figure 1 :
Figure 1: This figure is messy due to several reasons: 1) The bar graphs are depicted in different styles a. B is entirely different from the rest b.A and C have a title whereas E does not 2) The ChIP (please check the legend how it should be written) data is barely visible due to low quality 3) The IF staining should show a view on a higher magnification 4) The bar graph should include the dot plots as shown in 1N, 1O, 1P, 2, and 6The figure has been reformatted and reorganized along the lines that the reviewer indicated.

Figure 2 :
Figure 2: This figure includes lots of relevant information, but the layout is messy: Figure panels have different sizes (echocardiogram huge compared to the rest), font sizes differ, and therefore most of the text is difficult to read, lots of white spots in the figure.The fibrosis content in Masson`s Trichrome scan could be quantified.I don't see additive value in panels B and F. The legend indicates the precise p-values between brackets, but p-values are also indicated in asterisks.Please remove the precise numbers from the legend.The figure and legend have been reformatted and reorganized along the lines that the reviewer indicated.Quantification of fibrosis is shown in Extended Data Fig 2F.

Figure 7 :
Figure 7: For comparison, this Figure would be more illustrative when a healthy heart would be included.A panel showing a healthy heart has been added.

Fig. 1 -
Fig. 1 -The bar graphs are now dot plots in the same style.The ChIP panel has been moved, along with the associated bar graph, to Extended Data Fig. 1.Panel D has been separated to show the different channels.The graphs in panel B have been separated to show each paralog separately.

Fig. 2 (
Fig. 2 (above).The timeline panels in the previous version have been removed.The panels have been reformatted, in line with the reviewer's comments.

Fig. 7 -
Fig. 7 -A control heart section has been added (panel D).Quantification of the scoring has been added (panel E).

Table 1 :
Please add that you are looking at protein expression in the legend