Drug screening with human SMN2 reporter identifies SMN protein stabilizers to correct SMA pathology

This study has established a versatile human SMN2 reporter line for drug screening and identified a novel strategy of targeting cysteine proteases for rescuing SMA pathological phenotypes.

4.The authors should investigate the SMN2 protein levels in spinal cord of SMN model mice after treat ment wit h Z-FA-FMK and E64d to clarify the POC in vivo.
Minor point s 1)The qualit y of all images to show alt ered GFP signals is poor wit h a high background noise. The aut hors had bet ter use the Luc report er syst em.
2) The aut hors should add graphs to show the screening result s by plot ting alt ered rat ios compared wit h DMSO cont rol for each compound.
3)In Figure 2H, West ern blot ting band is not appropriat e. GAPDH band is also alt ered by the addit ion of compounds, including compound #8 up to 2-fold change. The aut hors should also add the calculat ed dat a to show the changes. Figure 4A, charact erizat ion of different iat ed mot or neurons is lacking. The aut hors should add the informat ion of cell-specific markers and their purit y (percent age per DAPI et c.) Figure 4B, the band of act in is also alt ered when adding Z-FA-FMK.

Referee #2 Review
Remarks for Aut hor: Wang et al generat ed human SMN2-GFP report er cell line, performed high-t hroughput screening, and ident ified candidat es that could increase SMN prot ein level. Especially, they found an irreversible inhibit or of cyst eine prot eases Z -FA-FMK as a most effect ive candidat e. They furt her test ed the effect of Z-FA-FMK in pat ient iPSCs and iPSCs-derived mot or neurons, and found that it is relat ed to protease/degradat ion pat hway inhibit ion. Furt hermore, aut hors demonst rat ed that 3 cyst eine prot eases CAPN1, CAPN7, and CTSL were import ant in mediat ing SMN prot ein degradat ion. Cellular phenot ypes of SMN, the abnormal mit ochondria behaviors and the neuron degenerat ion in mot or neurons, were also rescued by Z-FA-FMK treat ment . Finally, they demonst rat ed that Z-FA-FMK and anot her cyst eine prot ease inhibit or had prot ect ive effect s in SMA animal models. Overall, the st udy was well organized and clearly present ed, and the conclusions were st rongly support ed by their result s. Minor comment s include the below: 1. For most of the quantification data shown, the author indicated that mean {plus minus} s.e.m. was shown. The information about replicate/n number for those quantifications should be provided. Figure 4A, the time point for each stage should also be clearly stated either in the figure, or in the legend.

In
3. In Figure 7, the authors identified 3 cysteine proteases CAPN1, CAPN7, CTSL as important to regulate SMN protein stability. However, the effects of Z-FA-FMK on these specific proteases were not examined or discussed. Figure 8A, this figure is difficult to understand. The author should clearly explain the results. 5., In Figure 8 and 9, mitochondria transport and neuron degeneration were examined to confirm the functional relevance of SMN protein stabilization mediated by Z-FA-FMK. Ideally, neuronal activities should be examined to eventually confirm the functional relevance, or the author should discuss about this. 6. In Figure 9A and B, although there was a decrease in ISI1+ motor neurons in SMA+DMSO group, the proportion of DAPI+ nuclei looks similar (i.e. similar numbers of total cells were still growing in each condition). How do the authors address the effects of differentiation variations/motor neuron purity among different groups in their quantification? Similar issue is for Figure 10C and D. 9. In Figure 9C, the author may also enlarge part of the figure to better display the axon swelling and breakdown. 10, In Figure 10C and D, there is no WT control. Besides, if there is no DAPI staining for total nuclei, how did the aut hor quant ify the percent age of posit ive cells ( Figure 10D)?

Referee #3 Review
Remarks for Aut hor: This manuscript , "Drug screening wit h human SMN2 report er ident ifies SMN prot ein st abilizers t o correct SMA pat hology", describes const ruct ion of SMN2 report er cell line and it s applicat ion t o screen for compounds t hat correct SMN2 exon7 skipping, a common cause of SMA. A modest scale screen ident ified a compound Z -FA-FMK, a cyst eine prot ease inhibit or, which increases t he amount of SMN∆7. Z -FA-FMK had a posit ive effect on SMN's st abilit y and on several phenot ypic measures, including in SMA pat ient iPSC-derived mot or neurons, as well as a small increase in survival of SMA mice model. T he manuscript includes several int erest ing observat ions, however, it does not in my view add up t o a significant advance and quest ions remain unaddressed.
Several concerns: The suscept ibilit y of SMN∆7 t o prot eases has been previously elucidat ed and t he pot ent ial of inhibit ors of t his process for SMA have been previously proposed by st udies from Burnet t/ Fishbeck, Dreyfuss, and Rubin laborat ories.
Prot ease inhibit ors, such as Z -FA-FMK, are likely t o have pleiot ropic effect s rat her t han select ive effect on SMN∆7. T his needs t o be invest igat ed syst emat ically, for example, by mass spect romet ry. It is possible, and what ever effect s Z-FA-FMK has may be unrelat ed t o t he SMN increase and will be delet erious t o cells or an organism in longer t erm experiment s. Comment : Similar SMN2-based report er screens for SMA have been previously described, but the aut hors seem to have a nice set up which may be useful for larger scale screens. Thank you for transferring your manuscript entitled "Drug screening with human SMN2 reporter identifies SMN protein stabilizers to correct SMA pathology" to Life Science Alliance. The manuscript was assessed by expert reviewers at another journal before, and the editors transferred those reports to us with your permission. Furthermore, you had already provided an outline on how you could address the main concerns of the reviewers.
The reviewers thought that your work was well-executed, but noted that similar approaches have been used before and that insight into how SMN stabilization occurs is lacking. This is not a concern for publication here, and we would thus like to invite you to provide a revised version of your manuscript. Please provide a point-by-point response to all concerns raised and include the information and additional data that you already outlined upfront / provided upfront for editorial assessment. Please also make sure to adequately describe the HTS method (reviewer #1) and to add the requested quantifications. The potential pleiotropic effects of protease inhibitors should furthermore get mentioned in the manuscript text.
To upload the revised version of your manuscript, please log in to your account: https://lsa.msubmit.net/cgi-bin/main.plex You will be guided to complete the submission of your revised manuscript and to fill in all necessary information.
While you are revising your manuscript, please also attend to the below editorial points to help expedite the publication of your manuscript. Please direct any editorial questions to the journal office.
Thank you for this interesting contribution to Life Science Alliance. We are looking forward to receiving your revised manuscript.

A. THESE ITEMS ARE REQUIRED FOR REVISIONS
--A letter addressing the reviewers' comments point by point.
--An editable version of the final text (.DOC or .DOCX) is needed for copyediting (no PDFs).
--High-resolution figure, supplementary figure and video files uploaded as individual files: See our detailed guidelines for preparing your production-ready images, http://life-sciencealliance.org/authorguide --Summary blurb (enter in submission system): A short text summarizing in a single sentence the study (max. 200 characters including spaces). This text is used in conjunction with the titles of papers, hence should be informative and complementary to the title and running title. It should describe the context and significance of the findings for a general readership; it should be written in the present tense and refer to the work in the third person. Author names should not be mentioned.

B. MANUSCRIPT ORGANIZATION AND FORMATTING:
Full guidelines are available on our Instructions for Authors page, http://life-sciencealliance.org/authorguide We encourage our authors to provide original source data, particularly uncropped/-processed electrophoretic blots and spreadsheets for the main figures of the manuscript. If you would like to add source data, we would welcome one PDF/Excel-file per figure for this information. These files will be linked online as supplementary "Source Data" files. ***IMPORTANT: It is Life Science Alliance policy that if requested, original data images must be made available. Failure to provide original images upon request will result in unavoidable delays in publication. Please ensure that you have access to all original microscopy and blot data images before submitting your revision.*** Authors' Response to Reviewers: February 28, 2019 Major concerns from reviewers: 1. The susceptibility of SMN proteins to proteases has been previously elucidated.
Though previous studies have examined the stability of SMN proteins, they have been focused on ubiquitin proteasome and Calpain 1 (summarized in following Table 1). Instead, we have examined two major families of cysteine proteinases and studied their effects systematically from multiple angles. We first performed a screen for these proteases using RNAi (Fig. 7), then evaluated the candidates using co-Immunoprecipitation and Western Blot to exam their binding with SMN isoforms (Fig. 7), and finally examined the degradation of SMN proteins after overexpression of cysteine proteases using inducible Myc-SMN2a and SMN2d ( Fig. 7; Fig. S7). Our data showed that both non-lysosomal (i.e. Calpain 1/7) and lysosomal cysteine proteases (e.g. CTSL/CTSB) can degrade SMN-full length and SMN-Δ7. These data are significant because except Calpain 1, the role of Calpain2/CTSL/CTSB in degrading SMN proteins have not been reported. Moreover, cathepsins (CTSB/CTSL) belong to lysosomal cysteine protease and lysosome dysfunction has been implicated in many neurological diseases. Thus, we demonstrate that SMN proteins can be degraded by both non-lysosomal and lysosomal cysteine proteases (through either direct binding or other mechanisms including lysosomal-mediated pathway), providing novel targets for regulating SMN proteins.  (Cherry et al., 2013;Rietz et al., 2017)

ALLN
This work showed that SMN is a proteolytic target of Calpain 1. ALLN was used a tool to inhibit Calpain 1 but its effect was not tested in any SMA cell or animal models. Exogenous Calpain 1 was also used. (Fuentes et al., 2010)

Effects of Z-FA-FMK and E64d on SMN levels and motor neuron degeneration in vivo in SMA mice
were not clear We performed additional experiments to examine these and our data revealed a significant increase in the numbers of ventral horn motor neurons in lumbar segments after Z-FA-FMK and E64d treatment, which is coinciding with the significant increase of SMN proteins in spinal cord tissues after treatment (Fig. 10). This is a significant advantage over previous study about Calpain 1 and SMN (Fuentes et al., 2010), in which the effects of inhibiting cysteine proteases were not studied in any cell or animal SMA models. Our new data confirm the efficacy of both Z-FA-FMK and E64d in SMA mice in vivo, providing novel targets and approaches to treating SMA.
Since E64d can pass BBB and is easy to be administered to neonatal mice, this drug was injected into SMA mice daily (every day starting from day 1). This may be why E64d showed a stronger effect than Z-FA-FMK in long-term and significantly increased in the life span of SMA mice. Nevertheless, Z-FA-FMK which is only treated for 3 days, showed a trend in increasing the life span of SMA mice. Both Z-FA-FMK and E64d significantly increase the proportion of spinal motor neurons in spinal cord section at day 6 in SMA mice (Fig. 10). These data together demonstrate the effectiveness of candidate compounds in vivo and suggest a novel approach to treat SMA through targeting cysteine proteases. E64d is a potent inhibitor of thiol protease including CTSB (Hook et al., 2015;Inubushi et al., 1994;Romine et al., 2017;Trinchese et al., 2008;Tsubokawa et al., 2006). E64d has also proved to be safe to humans as shown in a clinical study for Alzheimer's disease and traumatic brain injury (this trial was stopped because lack of efficacy but the drug has proven to be safe to human). Given that E64d can significantly increase the proteins levels of SMN, the survival of spinal motor neurons, and the life span of SMA mice in vivo, our study not only identifies novel targets for regulating SMN proteins, but also provides a new candidate therapeutic agent for SMA. Table 1 summarized previous studies that are related to the susceptibility of SMN-∆7 proteins and the upregulation of SMN proteins for SMA. As we can see, though previous studies have shown that SMN-∆7 proteins are susceptible to degradation, they are focused on proteasome pathway, phosphorylation, and autophagy. The only study showed the role of cysteine protease in the SMN degeneration is from Strayer's group, in which only Calpain 1 was studied. Furthermore, the protective effects against motor neuron degeneration by cysteine protease inhibitors have not been examined before. During the review of our manuscript, a study reported and examined the role of Calpain 1 in SMA models (de la Fuente et al., 2018) , but did not examine any Cathepsin (lysosomal-dependent proteases) ; we have included a discussion of this study in the Discussion section of the revised manuscript.

Novelty and significance of our work versus previous studies
Our work is novel and significant in the following aspects: 1) first work to show that both non-lysosomal (e.g. CAPN1/7) and lysosomal cysteine proteinases (e.g. CTSL/CTSB) mediate the degradation of SMN proteins; 2) demonstrate the novel effects of Z-FA-FMK and E64d, two cysteine protease inhibitors, in increasing SMN proteins in SMA models both in vitro and in vivo; 3) reveal the efficacy of Z-FA-FMK and E64d in rescuing motor neuron loss in iPSCbased culture model and SMA mouse model; E64d also significantly increases the life span of SMA mice. Together, our work demonstrates the novel role of cysteine proteases (both non-lysosomal and lysosomal cysteine proteases) in regulating SMN protein stability, and reveal novel targets and candidate for rescuing motor neuron degeneration for the treatment of SMA.

Point-by-point response to reviewers' remaining questions:
Reviewer 1: 1. Evaluation of the HTS assay system is insufficient. Positive control is lacking and information of the HTS method is poor.
-Thank you for pointing this out. We have now included a detailed description about the drug screening method (in the Method section) and the examination of positive controls (Fig.  S1).
2. The authors described that Z-FA-FMK could elongate the life span of SMA model mice in the results section. However, Z-FA-FMK did not show any positive effect on mouse survival with statistical significance. Furthermore, it is unclear why E64d, which was not a hit compound, was selected in the next in vivo experiments, even though there were several cysteine protease inhibitors.
-We have now modified our description in the text to "Z-FA-FMK showed a trend in increasing the life span of SMA mice". Since Z-FA-FMK could not pass BBB, this compound was administrated to SMA mice only for a short period of time. E64d was selected because this drug can pass BBB and was proved to be safe from a previous clinical trial. Indeed, E64d significantly mitigated motor neuron degeneration and increased the life span of SMA mice (see our "Address to the major concern #2" for more details about the protective effects of E64d). We have added the related information in the revised manuscript.
3. Details of the method for in vivo experiments are lacking. The authors should describe how they injected Z-FA-FMK into lateral cerebral ventricles of postnatal day 1 -day 3 mice, and also how they decided the dose of compounds. It is unclear whether the amount of Z-FA-FMK 60 ng (155 microM) of 1 microL per day is appropriate for the treatment.
-These details are now added in the revised manuscript (Method section and Results section).
The concentration of Z-FA-FMK used in ICV injection is based on the amount in another paper (Hara et al., 1997), which is also cited in the revised paper.

4.
The authors should investigate the SMN2 protein levels in spinal cord of SMN model mice after treatment with Z-FA-FMK and E64d to clarify the POC in vivo.
-Thanks for this great suggestion. We have examined the proteins in SMA mice and found that both Z-FA-FMK and E64d can significantly increase the protein levels of SMN proteins (Fig.  10A, G).

(Minor points 1)
The quality of all images to show altered GFP signals is poor with a high background noise. The authors had better use the Luc reporter system.
-GFP signals represent SMN-Δ7 proteins which will degrade fast; therefore the GFP signals are weak in cells. When measuring the GFP fluorescence by ImageJ software, we always consider and calculate the background intensity and remove it from the cellular signals of each well. In the future, we will replace GFP with luciferase for further improvement of the system.

(Minor points 2)
The authors should add graphs to show the screening results by plotting altered ratios compared with DMSO control for each compound.
-We have now included a graph to show the screening results which identified 14 compounds that could brighten the GFP fluorescence by more than 0.5-fold compared to that of DMSOtreated cells (Fig. 2H). Figure 2H, Western blotting band is not appropriate. GAPDH band is also altered by the addition of compounds, including compound #8 up to 2-fold change. The authors should also add the calculated data to show the changes.

(Minor points 3) In
-GAPDH alteration is attributed to the bias of loading samples, rather than to the addition of compounds. We performed the Western blot again and quantified the data (n=4), as shown in Figure 3 (the relative levels of SMN proteins compared to GAPDH proteins were utilized for comparisons, which eliminate the variations caused by loading different amount of samples). From the quantification data, compound #8 is the most effective hit to increase endogenous SMN proteins in type I SMA fibroblast cells. Figure 4A, characterization of differentiated motor neurons is lacking.

(Minor points 4) In
-This protocol is based on our well-established protocol. We have previously examined and characterized the motor neuron differentiation from these SMA iPSCs (Xu et al., 2016). We have now described this work in more detail and added detailed information.
9. (Minor points 5) In Figure 4B, the band of actin is also altered when adding Z-FA-FMK.
-To accurately quantify the protein levels, actin was used as a loading control here. When comparing between different groups, the levels of SMN proteins relative to actin proteins were compared between different groups to eliminate the variations caused by the different loading amounts of samples. We also statistically analyzed the quantification data between different groups and observed a dose-dependent increase of SMN proteins by Z-FA-FMK.

Reviewer 2:
1. For most of the quantification data shown, the author indicated that mean ± s.e.m. was shown. The information about replicate/n number for those quantifications should be provided.
-Thank you for bringing this up. We have now included this information in the figure legends. Figure 4A, the time point for each stage should also be clearly stated either in the figure, or in the legend.

In
-We have now added this information in the figure legends.
3. In Figure 7, the authors identified 3 cysteine proteases CAPN1, CAPN7, CTSL as important to regulate SMN protein stability. However, the effects of Z-FA-FMK on these specific proteases were not examined or discussed.
-Thank you for this great suggestion. As we discussed in the above section (Address to the major concerns #1), we have examined the role of cysteine proteases from multiple angles and identified that both lysosomal and non-lysosomal proteases mediate the degradation of SMN proteins. We then examined the effects of Z-FA-FMK on these proteases and observed an inhibition of SMN protein degradation by Z-FA-FMK after overexpressing cysteine proteases (Fig.  S7). Figure 8A, this figure is difficult to understand. The author should clearly explain the results.

In
-In Figure 8A, these images represent the position (X-axis) versus timing (Y-axis) kymographs. X-axis in the kymograph represents the positions along neuronal axon, Y-axis represents the duration that we detected mitochondrial movement (i.e. 5 minutes). Each line in the image represents a mitochondrion; if this mitochondrion does not move during 5-minute imaging, it will not change its position along X-axis, which will lead to a vertical line. From these representative images, we could see less vertical lines in "SMA+Z-FA-FMK" group compared to "SMA+DMSO" group, suggesting an increase of mitochondrial transport by Z-FA-FMK. A detailed explanation has been added.

5.
In Figure 8 and 9, mitochondria transport and neuron degeneration were examined to confirm the functional relevance of SMN protein stabilization mediated by Z-FA-FMK. Ideally, neuronal activities should be examined to eventually confirm the functional relevance, or the author should discuss about this.
-We have now discussed the future examination of neuronal activities in the Discussion section.
6. In Figure 9A and B, although there was a decrease in ISI1+ motor neurons in SMA+DMSO group, the proportion of DAPI+ nuclei looks similar (i.e. similar numbers of total cells were still growing in each condition). How do the authors address the effects of differentiation variations/ motor neuron purity among different groups in their quantification?
-SMA is characterized by the specific degeneration of spinal motor neurons. As we reported before, the initial specification of spinal motor neuron lineage was not affected in SMA group. So the initial motor neuron proportion is similar between control and SMA groups (the efficient generation of Olig2+ motor neuron progenitors were observed in all groups). To ensure a rigorous comparison, multiple regions were selected blindly from triplicate samples as we described before. The degeneration of spinal motor neurons happened in long-term cultures (e.g. changes in morphology and accumulations of swellings), which is further supported by the increased apoptosis in the long-term cultures (Fig. 4E). We have now described these more in detail in the revised manuscript.
7. In Figure 9C, the author may also enlarge part of the figure to better display the axon swelling and breakdown.
-An enlarged part was included to show the axon swellings and breakdown.
8. In Figure 10C and D, there is no WT control. Besides, if there is no DAPI staining for total nuclei, how did the author quantify the percentage of positive cells ( Figure 10D)?
-Thank you for pointing this out. The WT group would be the same as shown for Z-FA-FMK treatment and here we focused on comparing vehicle-and E64d-treated groups. Yes, nuclei staining (Hoechst) was performed; it was not included before for showing other two staining more clear. We have now also included Hoechst in the images.
Reviewer 3: 1. The susceptibility of SMN∆7 to proteases has been previously elucidated and the potential of inhibitors of this process for SMA have been previously proposed by studies from Burnett/Fishbeck, Dreyfuss, and Rubin laboratories.
-Though the susceptibility of SMN∆7 to proteases has been previously elucidated, previous studies have been focused on ubiquitin proteasome and Calpain 1 (please see the summarized Table 1). In addition to Calpain 1, our data identified the novel role of both nonlysosomal (e.g. Calpain 7) and lysosomal cysteine proteases (e.g. CTSB/CTSL) in degrading SMN proteins ( Fig. 7; Fig. S7). This is of high importance because these members are different from Calpain 1 and have unique roles under pathological conditions. Please see the above section (Address to major concerns of the reviewers) for detailed information on the previous studies and our findings.

2.
Protease inhibitors, such as Z-FA-FMK, are likely to have pleiotropic effects rather than selective effect on SMN∆7. It is possible, and whatever effects Z-FA-FMK has may be on these specific proteases unrelated to the SMN increase and will be deleterious to cells or an organism in longer term experiments.
-Thanks for bringing up this great point. One of candidate compounds, E64d, has been used in a clinical trial for Alzheimer's disease and traumatic brain injury (from Medtrack). The drug has been approved to be safe to humans (this trial was stopped because lack of efficacy). In our study, E64d can significantly increase the proteins levels of SMN, the survival of spinal motor neurons, and the life span of SMA mice in vivo, suggesting that E64d serves a new candidate therapeutic agent for SMA. We have included a discussion on this issue in the revised manuscript.