Cardiac glycosides restore autophagy flux in an iPSC-derived neuronal model of WDR45 deficiency

Reviewer #1 (Public Review):

In the current study, Papandreou et al. developed an iPSC-based midbrain dopaminergic neuronal cell model of Beta-Propeller Protein-Associated Neurodegeneration (BPAN), which is caused by mutations in the WDR45 gene and is known to impair autophagy. They also noted defective autophagy and abnormal BPAN-related gene expression signatures. Further, they performed a drug screening and identified five cardiac glycosides. Treatment with these drugs effectively in improved autophagy defects and restored gene expression. Seeing the autophagy defects and impaired expression of BPAN-related genes adds strength to this study. Importantly, this work shows the value of iPSC-based modeling in studying disease and finding therapeutic strategies for genetic disorders, including BPAN.

Reviewer #2 (Public Review):

Summary:

In this manuscript, the authors aim to demonstrate that cardiac glycosides restore autophagy flux in an iPSC-derived mDA neuronal model of WDR45 deficiency. They established a patient-derived induced pluripotent stem cell (iPSC)-based midbrain dopaminergic (mDA) neuronal model and performed a medium-throughput drug screen using high-content imaging-based IF analysis. Several compounds were identified that ameliorate disease-specific phenotypes in vitro.

Strengths:

This manuscript engaged in an important topic and yielded some interesting data.

Author response:

The following is the authors’ response to the original reviews.

eLife assessment

This valuable manuscript reports alterations in autophagy present in dopaminergic neurons differentiated from iPSCs in patients with WDR45 mutations. The authors identified compounds that improved the defects present in mutant cells by generating isogenic iPSC without the mutation and performing an automated drug screening. The methodological approaches are solid, but the claims still need to be completed: showing the effects of the identified compounds on iron-related alterations is crucial. The effects of these drugs in vivo would be a great addition to the study.

Thank you for this assessment. We agree that further hit validation would be a great addition to the study. At present, we provide this through RNAseq data but not at the protein level. Further validation using in vivo models would also be warranted but is beyond the scope of the current work.

Public Reviews:

Reviewer #1 (Public Review):

Summary:

In the current study, Papandreou et al. developed an iPSC-based midbrain dopaminergic neuronal cell model of Beta-Propeller Protein-Associated Neurodegeneration (BPAN), which is caused by mutations in the WDR45 gene and is known to impair autophagy. They also noted defective autophagy and abnormal BPAN-related gene expression signatures. Further, they performed a drug screening and identified five cardiac glycosides. Treatment with these drugs effectively in improved autophagy defects and restored gene expression.

Strengths:

Seeing the autophagy defects and impaired expression of BPAN-related genes adds strength to this study. Importantly, this work shows the value of iPSC-based modeling in studying disease and finding therapeutic strategies for genetic disorders, including BPAN.

Weaknesses:

It is unclear whether these cells show iron metabolism defects and whether treatment with these drugs can ameliorate the iron metabolism phenotypes.

We are pleased to ascertain that the reviewer feels the work is an important step in the field for BPAN. We also absolutely agree that secondary hit validation assays showing cardiac glycoside efficacy in restoring patient-related in vitro phenotypes would be very valuable.

We set up assays to investigate iron metabolism phenotypes, including western blotting for Ferritin Heavy Chain 1, Transferrin and Ferroportin 1 (SLC40A1) at day 65 of differentiation, but found no significant difference when comparing patient lines to controls (data not shown).

We also performed cell viability studies using the Alamar Blue assay on Day 11 ventral midbrain progenitors after 24 hour exposure to a) glucose starvation, b) media with no antioxidants (L-ascorbic acid and B-27 supplement), c) oxidative stressors MPP+ 1mM and FeCl3 100 uM (MPP+ and FeCl3 as suggested by Seibler et al (Brain 2018 PMID: 30169597). We found no difference in cell viability between patients, age-matched controls and CRISPR lines (data not shown). Additionally, we examined lysosomal function in BPAN Day 11 progenitors (2 age-matched controls, 3 patient lines, 2 isogenic controls); again, using the autophagy flux treatments mentioned above) via LAMP1 high content imaging immunofluorescence. We have seen no difference in LAMP1 puncta production between patient lines and controls and, therefore, have not included this data in our revision.

Overall, we agree with the reviewer that more validation of the compound hits’ ability to restore robust BPAN-related in vitro and in vivo phenotypes (including studies of iron metabolism/ homeostasis) will be needed in the future – this could be undertaken in more mature 2D culture systems, 3D organoid models and disease-relevant animal models.

Reviewer #2 (Public Review):

Summary:

In this manuscript, the authors aim to demonstrate that cardiac glycosides restore autophagy flux in an iPSC-derived mDA neuronal model of WDR45 deficiency. They established a patientderived induced pluripotent stem cell (iPSC)-based midbrain dopaminergic (mDA) neuronal model and performed a medium-throughput drug screen using high-content imaging-based IF analysis. Several compounds were identified to ameliorate disease-specific phenotypes in vitro.

Strengths:

This manuscript engaged in an important topic and yielded some interesting data.

Weaknesses:

This manuscript failed to provide solid evidence to support the conclusion.

We are pleased that the reviewer assesses the work as conceptually important and interesting. We also agree that more work to understand the pathophysiology underpinning BPAN, and the mechanisms through which cardiac glycosides help restore affected intracellular pathways are warranted. More validation of the compound hits’ ability to restore broader disease-specific in vitro and in vivo phenotypes is also needed in future studies.

Recommendations for the authors:

Reviewer #1 (Recommendations For The Authors):

Overall, this is a nicely executed study. Here are my suggestions:

(1) Showing the iron phenotypes in these cells and testing if treatment with these drugs rescues iron-related phenotypes will add significant value to this work.

We absolutely agree that secondary hit validation assays showing glycoside efficacy in restoring disease-related in vitro phenotypes is warranted. The main issue here is identifying how WDR45 deficiency leads to cellular dysfunction or dyshomeostasis and early death. Unfortunately, the mechanism by which this happens is not yet delineated, and more relevant future work is needed.

In our lab, we set up such assays. Regarding iron metabolism-related phenotypes, we performed western blotting for Ferritin Heavy Chain 1, Transferrin and Ferroportin 1 (SLC40A1) but found no significant difference when comparing patient lines to controls (data not shown). We also performed cell viability studies using the Alamar Blue assay on Day 11 ventral midbrain progenitors after 24 hour exposure to a) glucose starvation, b) media with no antioxidants (L-ascorbic acid and B-27 supplement), c) oxidative stressors MPP+ 1mM and FeCl3 100 uM (MPP+ and FeCl3, as suggested by the Seibler et al paper, Brain 2018 PMID: 30169597). We found no difference in cell viability between patients, age-matched controls and CRISPR lines (data not shown). Additionally, we examined lysosomal function in BPAN Day 11 progenitors (2 age-matched controls, 3 patient lines, 2 isogenic controls; again, using the autophagy flux treatments mentioned above) via LAMP1 high content imaging immunofluorescence. We have seen no difference in LAMP1 puncta production between patient lines and controls and, therefore, have not included this data in our revision.

(2) Assessing the effects of these drugs in an in vivo model will strengthen this study.

This is a valid point, and we agree that further validation using in vivo models such as the reported BPAN mouse models, would be warranted in the future.

Reviewer #2 (Recommendations For The Authors):

While this manuscript engaged in an important topic and yielded exciting data, there are still some concerns for the authors to address.

(1) The biggest concern is that the characterization of autophagic flux solely with LC3 is not convincing enough. Although ATG2A and ATG2B are required for phagophore formation during autophagy, their interaction with WDR45 seems dispensable for phagophore formation for a mild autophagy defect observed in WDR45 knockout cell models and mouse models. All wdr45/- mice are born normally and survive the postnatal starvation period, unlike mice lacking essential ATG proteins, like ATG5, ATG7, and VMP1. The functional relevance of WDR45 and autophagy remains to be fully established. Overall, this manuscript failed to provide solid evidence to support the conclusion.

This is a valid point. We have looked at autophagy flux in fibroblasts and Day 11 ventral midbrain stage. For fibroblasts, 1 control line and three patient lines were used; for Day 11 progenitors, 2 control lines, 2 patient lines and one isogenic control were used. Cells from different lines were cultured on the same 96-well plates, at the same plating density, and treated concurrently to minimise fluctuations in flux due to unaccounted factors, e.g., confluence, incubator temperature etc. Treatments consisted of a) DMSO (basal condition), b) Bafilomycin A1 (flux inhibition via autophagosome/ lysosome fusion blockage), c) Torin A1 (mTOR inhibitor, flux inducer) and d) combination of Bafilomycin A1 and Torin 1, for a total of 3 hours. In all these conditions, LC3 puncta production in BPAN lines was reduced when compared to controls. We believe that these results indicate defective autophagy flux in BPAN in different cell types.

Moreover, we have demonstrated defects in autophagy-related gene (ATG) expression through RNA sequencing, that is restored after CRISPR/Cas9-mediated correction of the disease-causing mutation in a patient derived line, but also after treatments with torin 1 and digoxin. These results suggest a dysregulated ATG network in WDR45 deficiency.

(2) WDR45 is linked to BPAN. Do the authors detect any iron accumulation in DA progenitors or mDA neurons?

Regarding iron metabolism-related phenotypes, we performed western blotting for Ferritin Heavy Chain 1, Transferrin and Ferroportin 1 (SLC40A1) but found no significant difference when comparing patient lines to controls (data not shown). We agree that more studies into the links between WDR45 deficiency, iron metabolism and neurodegeneration are needed.

(3) It is necessary to detect LC3 protein levels by western blot to distinguish LC3I and LC3II and gain a more accurate understanding for the process of LC3 - marked autophagosome.

Thank you for this valid point.

Due to the very dynamic nature of autophagy, and many factors influencing flux , we have not been able to meaningfully examine autophagy-related markers in an iPSC-derived system that is also inherently prone to variability. Therefore, LC3 and p62 values exhibited high variability, and hence we are unable to adequately interpret them (data not shown). Instead, in this manuscript we have focused on high-content assays with cells cultured and treated simultaneously at Day 11 of differentiation, which have shown autophagy flux defects.

We have looked at autophagy flux in fibroblasts and at Day 11 ventral midbrain stage. For fibroblasts, 1 control line and three patient lines were used; for Day 11 progenitors, 2 control lines, 2 patient lines and one isogenic control were used. Cells from different lines were cultured on the same 96-well plates, at the same plating density, and treated concurrently to minimise fluctuations in flux due to unaccounted factors, e.g., confluence, incubator temperature etc. Treatments consisted of a) DMSO (basal condition), b) Bafilomycin A1 (flux inhibition via autophagosome/ lysosome fusion blockage), c) Torin A1 (mTOR inhibitor, flux inducer) and d) combination of Bafilomycin A1 and Torin 1, for a total of 3 hours. In all these conditions, LC3 puncta production in BPAN lines was reduced when compared to controls. We believe that these results indicate defective autophagy flux in BPAN in different cell types.

(4) Some methodological details need to be included - detailed descriptions of various quantifications for IF staining should be provided. For example, it is unclear how "% cells+ ve for marker combination" (Fig.1B) was quantified, and there are many unconventional units such as "% cells+ ve for marker combination "; please check and correct them.

Thank you for pointing this out. We have changed the legends in Figure 1B and Supplementary Figure 2C to ‘percentage of cells positive for marker combination’. Moreover, in our Methods section (Immunocytochemistry sub-section), we have updated the text as follows, to give more clarification on the process of marker quantification (Page 25, Paragraph 2): ‘For quantification, 4 random fields were imaged from each independent experiment. Subsequently, 1200 to 1800 randomly selected nuclei were quantified using ImageJ (National Institutes of Health). Manual counting for nuclear (DAPI) staining and co-staining with the marker of interest was performed, and percentages of cells expressing combinations of markers were calculated as needed.’

(5) In Figure 3 and Figure 4, the quantifications for IF images were inconsistent with the shown IF image, for example, the representative IF image for detection of LC3 with Tor1 treatment.

Due to space restrictions, we have not included representative images from all patient lines, and every treatment condition depicted in the graphs. In Figure 3 (describing the set-up of the LC3 screening assay), only one control line and one patient line is shown in basal (DMSO-treated) conditions. In Supplementary Figure 4D, only one patient line and the corresponding isogenic control line are depicted after Torin 1 treatments.

Quantification of the LC3 puncta in this assay (20 fields per well, each condition in a technical duplicate, n=8 biological replicates) was automated, using ImageJ and R Studio, with subsequent statistical significance calculation on GraphPad Prism. Hence, the immunofluorescence figures depict a reduction in LC3 puncta per nuclei numbers in patient-derived lines versus controls, but not the exact difference after automated image analysis. We have detailed this in the Methods section (High content imaging-based immunofluorescence subsection) of our manuscript (Page 26, Paragraph 2): ‘For all high content imaging-based experiments, the PerkinElmer Opera Phenix microscope was used for imaging. 20 fields were imaged per well, at 40 x magnification, Numerical Aperture 1.1, Binning 1. Image analysis was performed using ImageJ and R Studio.60 For the drug screen, puncta values were normalised according to positive and negative controls from each plate and Z-scores for each compound screened were generated. Statistical significances were calculated on GraphPad Prism V.

8.1.2. software (GraphPad Software, Inc.; https://www.graphpad.com/scientific-software/prism/).’

(6) In Figure 4C, LC3 should be co-stain with the DA progenitor maker to indicate that the intercellular LC3 level within the projectors.

Thank you for raising this point. The images from Figure 4C were obtained during the medium throughput drug screen, where the FOXA2 co-stain was not used. The FOXA2 stain was only used during the initial set-up of the LC3 screening assay, to confirm that the Day 11 cells had ventral midbrain identities. Indeed, most of the Day 11 cells used in the high content imaging-related experiments were FOXA2-positive, as shown in Figure 3 and Supplementary Figure 4.

(7) Examining P62, one of the most important indicators for autophagic flux, should be parallel with LC3 detection. In Figure 5A, P62 accumulation seems not significant in patient 02 Day 11 ventral midbrain projectors; how about that in Day 65?

The reviewer is raising a valid point. We have not examined p62 and LC3 staining in parallel in high content imaging-based experiments but agree that this would be good to examine in future studies.

Some other minor points

(8) It needs to give a more detailed description of the tested compounds you mentioned in the text.

Thank you for this point. We have elaborated on the contents of the Prestwick library used for the screening, as below (Page 9, Paragraph 3): ‘We then utilised this high-content imaging LC3 assay to identify novel compounds of potential therapeutic interest for BPAN by screening the Prestwick Chemical Library containing 1,280 compounds, of which more than 95% FDA/ EMA approved.’

In the Methods Section, Page 25, Paragraph 5, we also detail the library as follows: ‘For drug screening, the Prestwick Chemical Library (1,280 compounds, 95% FDA/ EMA approved, 10 mM in DMSO, https://www.prestwickchemical.com/screening-libraries/prestwick-chemical-library/) was used; cells were treated with compounds for 24 hours at 10 μM final concentration.’

(9) Please pay attention to the abbreviation; many gene names only have abbreviations without full names when they first appear in the context.

Thank you for this point. We have corrected this in various places throughout the manuscript and especially in the introduction section.

(10) Almost all figures have the problem of insufficient image resolution, or the font of the indicated words needs to be bigger to be distinguished clearly, like in Fig.1B, 1C, 1E.

Thank you for this point, we have ensured that all figures have adequate image resolution as specified by the journal requirements.

(11) The sample size or biological repeated times should be given in figure legends.

Thank you for this point. We have now indicated numbers of biological replicates where appropriate.