Emma Sullivan HMB496Y Final Report, April 3, 2020 CRISPR-Cas9 Directed Removal of Duplications in Pelizaeus Merzbacher Disease

Emma Sullivan, Human Biology Program, University of Toronto and Eleonora Maino, Genetics and Genome Biology, SickKids

Pelizaeus-Merzbacher Disease (PMD) is a rare X-linked pediatric disorder affecting 1in 100,000 boys. The disease is caused by mutations of the PLP1 gene, encoding Proteolipid Protein 1, a ​ ​ key structural protein of the myelin that encapsulates axons for efficient cell-cell communication in the nervous system. Mutations in PLP1 result in abnormal myelination, resulting in severe ​ ​ motor and cognitive impairment, and limited life expectancy. Genomic duplications of various sizes, that contain the entire PLP1 gene, account for 70% of PMD cases. Currently, there is no ​ ​ cure for PMD, with treatment options limited to symptom management. In a previous study, we successfully tested a novel single guide CRISPR-Cas9 genome editing strategy that removed the PLP1 duplication in a PMD mouse model and led to an improvement in disease manifestations. ​ This project aims to utilize this approach to correct PLP1 duplications in three PMD patient cell ​ ​ lines. We characterized the nature of the duplications through Whole Genome Sequencing and identified a common duplicated region among 2 patient cell lines, then designed intergenic and intronic guides and assessed their efficacy in HEK293 cells. The most active guides were selected and will be transfected in wildtype fibroblasts before testing them in patient fibroblasts. Ultimately, a universal guide will be selected and injected in a PMD mouse model to check its efficacy in an in-vivo system. If the cells prove to be functional, they could be utilized for cerebral transplant in individuals affected by PMD. Successful completion of this project will lay the foundation for the development of novel therapies for the majority of PMD patients and potential avenues for the treatment of genetic disorders caused by genomic duplications.

Introduction

Leukodystrophies are a collection of genetic disorders resulting in the abnormal development or destruction of the white matter, or myelin, of the central nervous system (CNS) (1,2). Myelin, a lipid-rich substance, normally encapsulates the axons of neurons, acting as an insulator to promote rapid cell-cell communication. When myelin production or activity is impaired, the nervous system cannot function properly (1), as in Pelizaeus-Merzbacher disease (PMD), an X-linked leukodystrophy caused by a mutation in the proteolipid protein 1 (PLP1) gene (3, 4, 5). ​ ​ PMD is a rare condition, with an incidence rate of approximately 1:200,000 - 1:500,000 (2), primarily affecting males. PLP1, expressed exclusively in oligodendrocytes of the CNS (2), is a ​ ​ dosage sensitive gene that is responsible for the major protein component of myelin (1, 6). Genomic duplications containing the entire PLP1 gene are the most common cause of PMD, ​ ​ accounting for nearly 70% of all cases (1, 5, 7). These duplications most commonly occur as tandem head-to-tail repeats, and the duplicated region varies in size, ranging from approximately 56 kb to 11 Mb (5, 8). Although the mechanism is unknown, PLP1 overexpression is likely the ​ ​ cause of the hypomyelination seen in PMD patients (5). PMD patients exhibit phenotypic variability, although almost all cases are characterized by early neurological deficits and progressive degeneration (3, 9, 10). PMD was first decribed by Friedrich Pelizaeus in 1885, and later by Ludwig Merzbacher, in children displaying cognitive ​ impairments, spastic quadriplegia, nystagmus or rapid uncontrollable lateral eye movements, Emma Sullivan, HMB496Y Final Report, April 3, 2020 ataxia, hypotonia, and motor delays (3, 11). Based on how the disease manifests, PMD is classified into two primary types. In the most common form, patients with classic PMD exhibit nystagmus, hypotonia, and sometimes motor delays in childhood, and other symptoms including spasticity and ataxia appear in late adolescence. Most patients suffer from life-threatening and eventually fatal complications before they reach age 30 (10). A more severe and less common form, connatal PMD, presents earlier in infancy, and patients have a poor prognosis with a life expectancy that does not exceed early childhood (1, 8, 11). In suspected PMD patients with the aforementioned clinical manifestations, a diagnosis is typically made based on the results of ​ magnetic resonance imaging (MRI) showing abnormal white matter in the CNS (9, 11, 12), and a confirmed change in the PLP1 gene or copy number (4, 13), which distinguishes PMD from ​ ​ other leukodystrophies. Due to the rarity of the disease, patients with no familial history of X-linked inheritance are often mistakenly diagnosed with cerebral palsy or other neuropathies (14), further contributing to the already small population of identified individuals with PMD. This limited sample size has made general studies of PMD difficult, and few analyses have been done on PMD patients, so PLP1 mutations at the microscopic level have not been well characterized. ​ ​ Currently, therapeutic approaches are aimed at alleviating symptoms of PMD, as there is no cure for this disease (3, 7, 9), although there are several new approaches being pursued. As described ​ in a 2019 study by Gupta et al, neural stem cell transplantation has been explored as a treatment course for PMD, where hypomyelinating neural cells are replenished with CNS stem cells expanded in culture. Patients in this study all showed positive neural changes monitored via MRI, and all were healthy four years post transplant. However, some patients developed antibodies to the transplanted cells, demonstrating that this procedure can elicit an immune response with the potential for complications (2). In recent years, genome engineering methods ​ have been suggested as treatments for inherited disorders like PMD (15). Clustered regularly ​ interspaced short palindromic repeats (CRISPR) has emerged as an ideal candidate for this ​ approach. CRISPR-Cas systems are found in bacteria, and are used as immune systems to protect the organism from foreign DNA. CRISPR, remnants of harmful viruses, can guide a Cas9 enzyme to identical regions of the DNA where the Cas9 endonuclease acts as molecular scissors, creating double stranded breaks in DNA and effectively removing the intervening target sequence (16, 17). Previous research has shown that by pairing this system with a single guide RNA in vitro and more recently in vivo, it is possible to excise disease duplications in Duchenne muscular dystrophy patient cells, another X-linked neuromuscular disease (15, 18, 19). The customizable nature of this method allows for the targeted removal of any area of interest in the genome; in this case, the duplication region seen in most incidences of PMD. Our project focuses on the CRISPR-Cas9-mediated removal of genomic duplications in PMD patient cells. Due to ​ the heterogeneous nature of PLP1 duplications, the minimal common duplicated region will be ​ ​ identified through sequencing of the duplications in each patient cell line. An RNA guide will be designed to be used with CRISPR-Cas9 to excise the critical portion of the PLP1 duplication ​ ​ responsible for the disease phenotype. In this way, we hope to develop an approach that can be applied to all PMD patients with PLP1 duplications regardless of duplication size, and ​ ​ potentially also to other genetic disorders with similar duplication genotypes. To date, we have performed whole genome sequencing (WGS) on three fibroblast cell lines, successfully amplified the duplication junction in two cell lines, identified a common duplicated region Emma Sullivan, HMB496Y Final Report, April 3, 2020 among them, and designed and tested 20 guides in HEK293 cells. Our next steps are to test the most active guides in WT fibroblasts, and then in patient cells. We also hope to amplify the duplication junction in the third cell line, to see how that duplication lines up against the common duplicated region we have identified, and potentially further narrow down a conserved region responsible for the phenotypic manifestation of PMD.

Methods

Whole Genome Sequencing (WGS) and Analysis WGS was performed using the Illumina HiSeq X platform by The Centre for Applied Genomics in Toronto, producing 150 bp paired-end reads. Output FASTQ files were aligned to the hg19 genome and analyzed using the Integrative Genomics Viewer (IGV) Software. Duplicated regions were identified by an increase in reads mapping to the area. Exact genomic locations of the PLP1 duplications were determined by identifying the location of soft clipped reads, as ​ ​ previously described (20, 21) (Fig. 1a). 3’ forward and 5’ reverse primers were designed to around the duplication junction (patient 1: 3’F-TCTCACATGCCAAAAACAGG, ​ 5’R-CTCTCCTGCCCCAATCTCTA; patient 3: 3’F-CCCTGTTTGCCTCGGTATTA, 5’R-GGCAATGGCTTGAGAGTAGG) using the Primer 3 Plus Website, and PCR amplification ​ at the following cycling conditions (annealing temperature of 59°C, extension time 15s, and 35 ​ ​ cycles), followed by Sanger sequencing was used to confirm the duplication arrangement.

Cell Culture Fibroblast cells from three PMD patients with duplication mutations were received from the ​ biobank of Robert Debré Hospital, Paris, courtesy of Dr. Odile Boespflug-Tanguy and Dr. Anne Baron. Cells were maintained in DMEM with 10% fetal bovine serum (FBS), 1% penicillin- streptomycin, and 1% L-glutamine (Life Technologies). All cells were maintained at 37°C with ​ 5% CO2.

Transfection of HEK293 cells with guide plasmids was performed using Lipofectamine 3000, according to the manufacturer’s instructions (Invitrogen).

DNA Extraction DNA was extracted from each of the three cultured cell lines and from transfected HEK293 cells using the DNA Blood and Tissue Kit, according to the manufacturer’s protocol for cultured cells (Qiagen). sgRNA Design and Testing 20bp guides were designed in the Human GRCh37 (hg19) genome either between genes in the duplicated region (Fig. 1b) or within the introns of the PLP1 gene (Fig. 1c) using the online ​ ​ Benchling Tool. For each approach, ten guides with the highest off-target scores, indicating activity, were selected, as previously described (22). Emma Sullivan HMB496Y Final Report, April 3, 2020

Guide and Vector Preparation and Ligation The guide oligos were annealed using a 10x annealing buffer and amplified through PCR. Annealed oligos were then incubated with ATP, 10X PNK buffer, T4 PNK, and water at 37 ºC for 1 hour and then at 65 ºC for 20 minutes to phosphorylate the oligos. Meanwhile, the px601 plasmid was digested with the restriction enzyme BsaI, then purified. The purified vector was incubated with CutSmart enzyme and rSAP at 37 ºC for 30 minutes and then at 65 ºC for 5 minutes to dephosphorylate the vector. Finally, The Vector and annealed oligos were combined with 5X ligase buffer, ligase, ATP and water to form a guide plasmid.

Guide Transformation in Bacterial Cells and Plasmid Purification Guide Plasmids were transformed into STBL3 bacterial cells using a heat shock of 42°C for 30 seconds, followed by incubation on ice for 2-5 minutes, before incubating at 37 ºC overnight. Transformed cell colonies were then identified via colonyPCR (annealing temperature of 58°C, ​ ​ extension time 30s, and 35 cycles). In colonies that had been transformed, minipreparations were performed to purify the plasmids using the QIAcube, according to the manufacturer’s instructions.

RNA Extraction and Quantitative Reverse Transcription PCR RNA was extracted from cultured patient 1, 2, 3, and WT fibroblasts cells and cDNA synthesis was performed as previously described (23). qPCR was performed using SYBR green Master Mix (Qiagen) and analysis was carried out as previously described (24). PLP1 expression was ​ ​ analyzed using primers spanning the junctions of PLP1 exons 2 and 3, and endogenous GAPDH ​ ​ primers were used as a negative control. Emma Sullivan, HMB496Y Final Report, April 3, 2020 Results

WGS analysis identified the duplication locations in patient 1 and 3 cell lines In the patient 1 and patient 3 cell lines, duplications of 1.4 Megabases and 559 kilobases were identified respectively by determining their start and end locations (Fig. 2b), based on the whole ​ genome sequencing data viewed in IGV (Fig. 2a). Split reads were not identified in the sequencing data from patient 2 cells, so the precise location of the duplication in this cell line remains unknown. Emma Sullivan, HMB496Y Final Report, April 3, 2020 Validation of WGS results via PCR amplification Successful amplification of the duplication junction was achieved in both patient 1 (primers: patient 1 3’F and 5’R) and patient 3 (primers: patient 3 3’F and 5’R) cells using DreamTaq polymerase (Fig. 3c, 3d). The same result was obtained with different polymerases. The sequencing across the duplication junction obtained from gel extracted DNA is shown in Figure 3a and 3b, both of which confirm the WGS results. Similar experiments performed with patient 2 cells did not successfully amplify the predicted duplication junction.

Patients 1 and 3 share a common duplicated region of 532kb A common duplicated region of 532 kb spanning from chrX 102774001 - 103306000 was identified in the M1 and M3 cell lines, based on their genomic start and end sites (Fig. 2b) and depicted in figure 4 as it appears using the IGV software. Emma Sullivan, HMB496Y Final Report, April 3, 2020

Guide testing in HEK293 cells identifies the most active intergenic and intronic guides Of the 20 guides that were designed, 14 were successfully transfected into HEK293 cells, while 6 have yet to be transfected (intergenic guides 1, 4, 7, and intronic guides 2, 4, and 8). Following ICE analysis of the 14 transfected guides, the most active intergenic and intronic guides were identified (Fig. 5a, 5b). Intergenic guide 5 that cuts between the RAB40A and TCEAL4 genes in the duplicated region was the most active intergenic guide, with an indel rate of 10% (R² = 0.99) ​ ​ ​ (Fig. 5c). Intronic guide 3 that cuts between introns 2 and 3 of PLP1 was the most active intronic ​ ​ guide, with an indel rate of 18% (R² = 1) (Fig. 5d).

PLP1 expression is significantly increased in patient versus wildtype fibroblasts ​ When analyzed using quantitative reverse transcription PCR (qRT-PCR), it was determined that all three PMD patient fibroblast lines show a significant increase in PLP1 expression compared ​ ​ to healthy fibroblasts (Fig. 5e). Emma Sullivan, HMB496Y Final Report, April 3, 2020

Discussion

This project explores the use of a CRISPR-Cas9 sgRNA approach to removing duplication mutations in PMD patient cells. The data we obtained from WGS and later confirmed via PCR amplification and sequencing shows that we have two patients with duplications containing the PLP1 gene. This confirmation was essential not only to ensure the start and end locations of the ​ duplications were as predicted, but also that the arrangement of the duplication was a head to tail tandem duplication, which is essential for the sgRNA strategy to be effective in restoring correct copy number. Emma Sullivan, HMB496Y Final Report, April 3, 2020 Now that we have identified a common duplicated region, designed guides within this region, and tested them in HEK293 cells, our next steps are to select a narrowed down pool of the most active guides to test in WT fibroblasts. As they are the same cell type as the patient fibroblasts in which we will test the guides afterwards, we predict that they will be a better model for guide activity and more realistically demonstrate how the guides will behave in patient cells. Following guide testing in WT fibroblasts, we will select the most active guides and test them in our patient fibroblast cells. Since we showed that PLP1 expression is increased in PMD patient ​ ​ cells vs WT fibroblasts, we hope to show that following transfection with an active guide, PLP1 ​ expression in edited patient cells more closely reflects PLP1 expression in the WT fibroblasts. ​ ​ The increased PLP1 expression observed in PMD patient fibroblasts compared to wildtype ​ ​ ​ ​ fibroblasts also suggest that as previously predicted, overexpression of PLP1 is the likely cause ​ ​ of hypomyelination seen in the CNS of PMD patients (5), which helps us to better understand the mechanism of PMD, which remains unclear. It is also important to consider that of the 14 guides successfully transfected into HEK293 cells, the most active guide had an indel rate of only 18%. This may suggest that it could be worthwhile to design additional guides for initial testing in HEK cells to potentially find a more active guide, before moving into more difficult transfections in fibroblasts. As we pursue further guide testing in WT fibroblast and patient cells, another next step for this project would be to characterize the more complex duplication seen in patient 2. We were unable to determine the location in the same way as the other patients, as no split reads were identified in the WGS data. We did identify some reads that indicate a possible breakpoint, but sequencing across this location was unsuccessful, indicating that it is not likely to be the duplication junction. Once we are able to determine the location and arrangement of the duplication in this patient, we will align it with the duplication locations in M1 and M3, to see if we can further narrow down the common duplicated region, thus aiding in selecting a universal guide that is applicable to the highest number of duplications possible. The ultimate goal of this project, although far in the future, would be to show that genetically altered cells ameliorate disease in a PMD mouse model to show efficacy in vivo for external validity, and eventually we hope this approach could be applied to human cell therapy. This would involve the editing of human NSCs derived from the patient’s own skin cells with our guide, and the subsequent differentiation of these cells into oligodendrocytes to replace the affected ones in the patient. If this therapy is effective, we would expect to see re-myelination occuring in the CNS. While there is a very long road and there are considerable challenges that lie between this project and the application to patient therapy, it is an exciting potential outcome of this research. The importance of this project lies in the fact that there is no cure or even standard treatment option available for PMD, and research efforts have been limited by the poor characterization of PLP1 mutations in patient cells. We hope that by characterizing individual ​ ​ patient mutations we will gain a better understanding of the commonly duplicated region that causes the disease manifestations of PMD, and could thus be targeted in future therapeutic strategies. As outlined above, there are several potential avenues for further research to pursue to better our understanding of this rare disease and provide a viable treatment option for PMD patients. Further, the approach outlined in this project has the potential to be applied to other disorders featuring similar duplication mutations. Emma Sullivan, HMB496Y Final Report, April 3, 2020 Acknowledgements

I would like to express my immense gratitude to my primary supervisor, Eleonora Maino, who guided me throughout both this project and my first time in a working lab. Her encouragement, kindness, and advice was invaluable and made this an incredibly rewarding and positive experience. I would also like to thank Aiman Farheen, my partner in this project who carried out endless experiments, designed several figures included in the report, and helped me more times than I can count. Thank you to Steven Erwood, who assisted in navigating the software and websites used in the early stages of the project, helping us to overcome our first obstacle. A heartfelt thank you to each member of the Cohn Lab; Sonia Evagelou, Samar Rizvi, Reid Brewer, Kyle Lindsay, Matthew Rok, Ori Scott, Teija Bily, Monika Kustermann, Tatianna Wong, Antonio Mollica, and Shagana Visuvanathan; for welcoming me into their lab and their willingness to help set up a machine, explain a process, answer a question, or find a reagent, amongst many other things. Finally, my deepest thanks to Dr. Zhenya Ivakine and Dr. Ronald Cohn, for allowing me the opportunity to pursue my research project in this lab; it has been the most enriching and enjoyable time of my undergraduate career. Emma Sullivan, HMB496Y Final Report, April 3, 2020 References

(1) Nevin ZS, Factor DC, Karl RT, Douvaras P, Laukka J, Windrem MS, Goldman SA, Fossati V, Hobson GM, Tesar, PJ. Modeling the Mutational and Phenotypic Landscapes of Pelizaeus-Merzbacher Disease with Human iPSC-Derived Oligodendrocytes. Am J Hum Genet 2017; 100(4): 617-634.

(2) Gupta N, Henry RG, Kang S, Strober J, Lim DA, Ryan T, Perry R, Farrell J, Ulman M, ​ Rajalingam R, Gage A, Huhn SL, Barkovich AJ, Rowitch DH. Long-Term Safety, Immunologic Response, and Imaging Outcomes following Neural Stem Cell Transplantation for Pelizaeus-Merzbacher Disease. Stem Cell Reports 2019; 13(2): ​ 254-261.

(3) Osorio MJ, Goldman SA. Neurogenetics of Pelizaeus - Merzbacher disease. Handb Clin Neurol 2018; 148: 701-722.

(4) Kouga T, Koizume S, Aoki S, Jimbo E, Yamagata T, Inoue K, Osaka H. Drug screening for Pelizaeus-Merzbacher disease by quantifying the total levels and membrane localization of PLP1. Mol Genet Metab Rep 2019; 20: 100474. ​ ​

(5) Clark K, Sakowski L, Sperle K, Banser L, Landel CP, Bessert DA, Skoff RP, Hobson GM. ​ Gait Abnormalities and Progressive Myelin Degeneration in a New Murine Model of Pelizaeus-Merzbacher Disease with Tandem Genomic Duplication. J Neurosci 2013; 33(29): 11788-11799.

(6) Boison D, Bussow H, D'Urso D, Muller HW, and W Stoffel. Adhesive properties of proteolipid protein are responsible for the compaction of CNS myelin sheaths. J Neurosci 1995; 15(8): 5502-13.

(7) Li H, Okada H, Suzuki S, Sakai K, Izumi H, Matsushima Y, Ichinohe N, Goto Y, Okada T, Inoue K. Gene suppressing therapy for Pelizaeus-Merzbacher disease using artificial microRNA. JCI Insight 2019; 4(10).

(8) Woodward KJ. The molecular and cellular defects underlying Pelizaeus-Merzbacher disease. Expert Rev Mol Med 2008;10:e14.

(9) Gulati S, Jain P, Chakrabarty B, Kumar A, Gupta N, Kabra M. The spectrum of ​ ​ leukodystrophies in children: Experience at a tertiary care centre from North India. Ann Indian Acad Neurol 2016; 19(3): 332-338. Emma Sullivan, HMB496Y Final Report, April 3, (10)2020 Mądry J, Hoffman-Zacharska D, Królicki L, Jakuciński M, Friedman A. PLP1 gene ​ ​ duplication as a cause of the classic form of Pelizaeus-Merzbacher disease – case report. Neurol Neurochir Pol 2010; 44(5): 511-5.

(11) Velasco-Parra HM, Maradei-Anaya SJ, Acosta-Guio JC, Diaz A, Prieto-Rivera JC. Clinical and mutational spectrum of Colombian patients with Pelizaeus Merzbacher Disease. Colomb Med 2018;49(2): 182-187.

(12) Castro MAA, Fraiman PHA, Godeiro-Junior CO. Magnetic resonance imaging traits may help to differentiate Pelizaeus-Merzbacher and Pelizaeus-Merzbacher-like disease. Arq Neuropsiquiatr 2019; 77(8): 59.

(13)Wang X, He F, Yin F, Chen C, Wu L, Yang L, Penga J. The use of targeted genomic capture and massively parallel sequencing in diagnosis of Chinese Leukoencephalopathies. Sci Rep 2016; 6: 35936.

(14) Margraf RL, Durtschi J, Krock B, Newcomb TM, Bonkowsky JL, Voelkerding KV, Bayrak-Toydemir P, Lutz RE, Swoboda KJ. Novel PLP1 Mutations Identified With ​ ​ Next-Generation Sequencing Expand the Spectrum of PLP1-Associated Leukodystrophy ​ ​ Clinical Phenotypes. Child Neurol Open 2018; 5: 2329048X18789282.

(15) Wojtal D, Kemaladewi DU, Malam Z, Abdullah S, Wong TWY, Hyatt E, Baghestani Z, Pereira S, Stavropoulos J, Mouly V, Mamchaoui K, Muntoni F, Voit T, Gonorazky HD,1, Dowling JJ, Wilson MD, Mendoza-Londono R, Ivakine EA, Cohn RD. Spell Checking Nature: Versatility of CRISPR/Cas9 for Developing Treatments for Inherited Disorders. Am J Hum Genet 2016; 98(1): 90-101.

(16) Khanzadi MN, Khan AA. CRISPR/Cas9: Nature's gift to prokaryotes and an auspicious tool in genome editing. J Basic Microbiol 2019.

(17) Khadempar S, Familghadakchi S, Motlagh RA, Farahani N, Dashtiahangar M, Rezaei H, Gheibi Hayat SM. CRISPR-Cas9 in genome editing: Its function and medical applications. J Cell Physiol. 2019; 234(5): 5751-5761.

(18) Ousterout DG, Kabadi AM, Thakore PI, Majoros WH, Reddy TE, Gersbach CA. Multiplex CRISPR/Cas9-Based Genome Editing for Correction of Dystrophin Mutations that Cause Duchenne Muscular Dystrophy. Nat Commun 2015; 6: 6244.

(19) Wong TWY, Cohn RD. Therapeutic Applications of CRISPR/Cas for Duchenne Muscular Dystrophy. Curr Gene Ther 2017;17(4):301-308.

(20) Thorvaldsdóttir H, Robinson JT, Mesirov JP. Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinform 2013; 14: 178–192. Emma Sullivan, HMB496Y Final Report, April 3, 2020 (21) Robinson JT, Thorvaldsdóttir H, Winckler W, Guttman M, Lander ES, Getz G, Mesirov JP. Integrative Genomics Viewer. Nat Biotechnol; 29(1): 24–26. ​

(22) Doench J.G., Hartenian E., Graham D.B., Tothova Z., Hegde M., Smith I., Sullender M., Ebert B.L., Xavier R.J., Root D.E. Rational design of highly active sgRNAs for CRISPR-Cas9-mediated gene inactivation. Nat. Biotechnol. 2014; 32:1262–1267.

(23) Sunada, Y., Bernier, S. M., Utani, A., Yamada, Y. & Campbell, K. P. Identification of a ​ novel mutant transcript of laminin α2 chain gene responsible for muscular dystrophy and dysmyelination in dy 2J mice. Hum. Mol. Genet. 4, 1055–1061 (1995). ​ ​ ​ ​

(24) Kemaladewi, D.U., Bassi, P.S., Erwood, S., Al-Basha, D. Gawlik, K.I., Lindsay, K., Hyatt, E., Kember, R., Place, K.M., Marks, R.M., Durbeej, M., Prescott, S.A., Ivakine, E.A., Cohn, R.D. A mutation-independent approach for muscular dystrophy via upregulation of a modifier gene. Nature 2019; 572: 125–130.