iMyoblasts for ex vivo and in vivo investigations of human myogenesis and disease modeling

The adult environment promotes the transcriptional maturation of human iPSC-derived muscle grafts

Pluripotent stem cell (PSC)-based cell therapy is an attractive option for the treatment of multiple human disorders, including muscular dystrophies. While in vitro differentiating PSCs can generate large numbers of human lineage-specific tissue, multiple studies evidenced that these cell populations mostly display embryonic/fetal features. We previously demonstrated that transplantation of PSC-derived myogenic progenitors provides long-term engraftment and functional improvement in several dystrophic mouse models, but it remained unknown whether donor-derived myofibers mature to match adult tissue. Here, we transplanted iPAX7 myogenic progenitors into muscles of non-dystrophic and dystrophic mice and compared the transcriptional landscape of human grafts with respective in vitro-differentiated iPAX7 myotubes as well as human skeletal muscle biospecimens. Pairing bulk RNA sequencing with computational deconvolution of human reads, we were able to pinpoint key myogenic changes that occur during the in vitro-to-in vivo transition, confirm developmental maturity, and consequently evaluate their applicability for cell-based therapies.

A human immune/muscle xenograft model of FSHD muscle pathology

Background

Facioscapulohumeral muscular dystrophy (FSHD) disease progression is associated with muscle inflammation, although its role in FSHD muscle pathology is unknown.

Methods

We have developed a novel humanized mouse strain, NSG-SGM3-W41, that supports the co- engraftment of human hematopoietic stem cells (HSCs) and muscle myoblasts as an experimental model to investigate the role of innate immunity in FSHD muscle pathology.

Results

The NSG-SGM3-W41 mouse supports the selective expansion of human innate immune cell lineages following engraftment of human HSCs and the co-engraftment and differentiation of patient-derived FSHD or control muscle myoblasts. Immunohistological and NanoString RNA expression assays establish that muscle xenografts from three FSHD subjects were immunogenic compared to those from unaffected first-degree relatives. FSHD muscle xenografts preferentially accumulated human macrophages and B cells and expressed early complement genes of the classical and alternative pathways including complement factor C3 protein, which is a mediator of early complement function through opsonization to mark damaged cells for macrophage engulfment. FSHD muscle xenografts also underwent immune donor dependent muscle turnover as assayed by human spectrin β1 immunostaining of muscle fibers and by NanoString RNA expression assays of muscle differentiation genes.

Conclusions

The NSG-SGM3-W41 mouse provides an experimental model to investigate the role of innate immunity and complement in FSHD muscle pathology and to develop FSHD therapeutics targeting DUX4 and the innate immunity inflammatory responses.

Flavones provide resistance to DUX4-induced toxicity via an mTor-independent mechanism

Facioscapulohumeral muscular dystrophy (FSHD) is among the most common of the muscular dystrophies, affecting nearly 1 in 8000 individuals, and is a cause of profound disability. Genetically, FSHD is linked to the contraction and/or epigenetic de-repression of the D4Z4 repeat array on chromosome 4, thereby allowing expression of the DUX4 gene in skeletal muscle. If the DUX4 transcript incorporates a stabilizing polyadenylation site the myotoxic DUX4 protein will be synthesized, resulting in muscle wasting. The mechanism of toxicity remains unclear, as many DUX4-induced cytopathologies have been described, however cell death does primarily occur through caspase 3/7-dependent apoptosis. To date, most FSHD therapeutic development has focused on molecular methods targeting DUX4 expression or the DUX4 transcript, while therapies targeting processes downstream of DUX4 activity have received less attention. Several studies have demonstrated that inhibition of multiple signal transduction pathways can ameliorate DUX4-induced toxicity, and thus compounds targeting these pathways have the potential to be developed into FSHD therapeutics. To this end, we have screened a group of small molecules curated based on their reported activity in relevant pathways and/or structural relationships with known toxicity-modulating molecules. We have identified a panel of five compounds that function downstream of DUX4 activity to inhibit DUX4-induced toxicity. Unexpectedly, this effect was mediated through an mTor-independent mechanism that preserved expression of ULK1 and correlated with an increase in a marker of active cellular autophagy. This identifies these flavones as compounds of interest for therapeutic development, and potentially identifies the autophagy pathway as a target for therapeutics.

Addressing the dNTP bottleneck restricting prime editing activity

Prime editing efficiency is modest in cells that are quiescent or slowly proliferating where intracellular dNTP levels are tightly regulated. MMLV-reverse transcriptase - the prime editor polymerase subunit - requires high intracellular dNTPs levels for efficient polymerization. We report that prime editing efficiency in primary cells and in vivo is increased by mutations that enhance the enzymatic properties of MMLV-reverse transcriptase and can be further complemented by targeting SAMHD1 for degradation.

Transfer learning enables predictions in network biology

Mapping gene networks requires large amounts of transcriptomic data to learn the connections between genes, which impedes discoveries in settings with limited data, including rare diseases and diseases affecting clinically inaccessible tissues. Recently, transfer learning has revolutionized fields such as natural language understanding1,2 and computer vision3 by leveraging deep learning models pretrained on large-scale general datasets that can then be fine-tuned towards a vast array of downstream tasks with limited task-specific data. Here, we developed a context-aware, attention-based deep learning model, Geneformer, pretrained on a large-scale corpus of about 30 million single-cell transcriptomes to enable context-specific predictions in settings with limited data in network biology. During pretraining, Geneformer gained a fundamental understanding of network dynamics, encoding network hierarchy in the attention weights of the model in a completely self-supervised manner. Fine-tuning towards a diverse panel of downstream tasks relevant to chromatin and network dynamics using limited task-specific data demonstrated that Geneformer consistently boosted predictive accuracy. Applied to disease modelling with limited patient data, Geneformer identified candidate therapeutic targets for cardiomyopathy. Overall, Geneformer represents a pretrained deep learning model from which fine-tuning towards a broad range of downstream applications can be pursued to accelerate discovery of key network regulators and candidate therapeutic targets.

Pathological modeling of glycogen storage disease type III with CRISPR/Cas9 edited human pluripotent stem cells

Introduction: Glycogen storage disease type III (GSDIII) is a rare genetic disease caused by mutations in the AGL gene encoding the glycogen debranching enzyme (GDE). The deficiency of this enzyme, involved in cytosolic glycogen degradation, leads to pathological glycogen accumulation in liver, skeletal muscles and heart. Although the disease manifests with hypoglycemia and liver metabolism impairment, the progressive myopathy is the major disease burden in adult GSDIII patients, without any curative treatment currently available. Methods: Here, we combined the self-renewal and differentiation capabilities of human induced pluripotent stem cells (hiPSCs) with cutting edge CRISPR/Cas9 gene editing technology to establish a stable AGL knockout cell line and to explore glycogen metabolism in GSDIII. Results: Following skeletal muscle cells differentiation of the edited and control hiPSC lines, our study reports that the insertion of a frameshift mutation in AGL gene results in the loss of GDE expression and persistent glycogen accumulation under glucose starvation conditions. Phenotypically, we demonstrated that the edited skeletal muscle cells faithfully recapitulate the phenotype of differentiated skeletal muscle cells of hiPSCs derived from a GSDIII patient. We also demonstrated that treatment with recombinant AAV vectors expressing the human GDE cleared the accumulated glycogen. Discussion: This study describes the first skeletal muscle cell model of GSDIII derived from hiPSCs and establishes a platform to study the mechanisms that contribute to muscle impairments in GSDIII and to assess the therapeutic potential of pharmacological inducers of glycogen degradation or gene therapy approaches.

Flavones provide resistance to DUX4-induced toxicity via an mTor-independent mechanism

Facioscapulohumeral muscular dystrophy (FSHD) is among the most common of the muscular dystrophies, affecting nearly 1 in 8000 individuals, and is a cause of profound disability. Genetically, FSHD is linked to the contraction and/or epigenetic de-repression of the D4Z4 repeat array on chromosome 4, thereby allowing expression of the DUX4 gene in skeletal muscle. If the DUX4 transcript incorporates a stabilizing polyadenylation site the myotoxic DUX4 protein will be synthesized, resulting in muscle wasting. The mechanism of toxicity remains unclear, as many DUX4-induced cytopathologies have been described, however cell death does primarily occur through caspase 3/7-dependent apoptosis. To date, most FSHD therapeutic development has focused on molecular methods targeting DUX4 expression or the DUX4 transcript, while therapies targeting processes downstream of DUX4 activity have received less attention. Several studies have demonstrated that inhibition of multiple signal transduction pathways can ameliorate DUX4-induced toxicity, and thus compounds targeting these pathways have the potential to be developed into FSHD therapeutics. To this end, we have screened a group of small molecules curated based on their reported activity in relevant pathways and/or structural relationships with known toxicity-modulating molecules. We have identified a panel of five compounds that function downstream of DUX4 activity to inhibit DUX4-induced toxicity. Unexpectedly, this effect was mediated through an mTor-independent mechanism that preserved expression of ULK1 and correlated with an increase in a marker of active cellular autophagy. This identifies these flavones as compounds of interest for therapeutic development, and potentially identifies the autophagy pathway as a target for therapeutics.

Induced Pluripotent Stem Cells for Modeling Physiological and Pathological Striated Muscle Complexity

Neuromuscular disorders (NMDs) are a large group of diseases associated with either alterations of skeletal muscle fibers, motor neurons or neuromuscular junctions. Most of these diseases is characterized with muscle weakness or wasting and greatly alter the life of patients. Animal models do not always recapitulate the phenotype of patients. The development of innovative and representative human preclinical models is thus strongly needed for modeling the wide diversity of NMDs, characterization of disease-associated variants, investigation of novel genes function, or the development of therapies. Over the last decade, the use of patient's derived induced pluripotent stem cells (hiPSC) has resulted in tremendous progress in biomedical research, including for NMDs. Skeletal muscle is a complex tissue with multinucleated muscle fibers supported by a dense extracellular matrix and multiple cell types including motor neurons required for the contractile activity. Major challenges need now to be tackled by the scientific community to increase maturation of muscle fibers in vitro, in particular for modeling adult-onset diseases affecting this tissue (neuromuscular disorders, cachexia, sarcopenia) and the evaluation of therapeutic strategies. In the near future, rapidly evolving bioengineering approaches applied to hiPSC will undoubtedly become highly instrumental for investigating muscle pathophysiology and the development of therapeutic strategies.

Generation of iMyoblasts from Human Induced Pluripotent Stem Cells

Skeletal muscle stem cells differentiated from human-induced pluripotent stem cells (hiPSCs) serve as a uniquely promising model system for investigating human myogenesis and disease pathogenesis, and for the development of gene editing and regenerative stem cell therapies. Here, we present an effective and reproducible transgene-free protocol for derivation of human skeletal muscle stem cells, iMyoblasts, from hiPSCs. Our two-step protocol consists of 1) small molecule-based differentiation of hiPSCs into myocytes, and 2) stimulation of differentiated myocytes with growth factor-rich medium to activate the proliferation of undifferentiated reserve cells, for expansion and cell line establishment. iMyoblasts are PAX3 ⁺ /MyoD1 ⁺ myogenic stem cells with dual potential to undergo muscle differentiation and to self-renew as a regenerative cell population for muscle regeneration both ex vivo and in vivo . The simplicity and robustness of iMyoblast generation and expansion have enabled their application to model the molecular pathogenesis of Facioscapulohumeral Muscular Dystrophy and Limb-Girdle Muscular Dystrophies, to both ex vivo and in vivo muscle xenografts, and to respond efficiently to gene editing, enabling the co-development of gene correction and stem cell regenerative therapeutic technologies for the treatment of muscular dystrophies and muscle injury. Graphical abstract.

Update on the Molecular Aspects and Methods Underlying the Complex Architecture of FSHD

Despite the knowledge of the main mechanisms involved in facioscapulohumeral muscular dystrophy (FSHD), the high heterogeneity and variable penetrance of the disease complicate the diagnosis, characterization and genotype–phenotype correlation of patients and families, raising the need for further research and data. Thus, the present review provides an update of the main molecular aspects underlying the complex architecture of FSHD, including the genetic factors (related to D4Z4 repeated units and FSHD-associated genes), epigenetic elements (D4Z4 methylation status, non-coding RNAs and high-order chromatin interactions) and gene expression profiles (FSHD transcriptome signatures both at bulk tissue and single-cell level). In addition, the review will also describe the methods currently available for investigating the above-mentioned features and how the resulting data may be combined with artificial-intelligence-based pipelines, with the purpose of developing a multifunctional tool tailored to enhancing the knowledge of disease pathophysiology and progression and fostering the research for novel treatment strategies, as well as clinically useful biomarkers. In conclusion, the present review highlights how FSHD should be regarded as a disease characterized by a molecular spectrum of genetic and epigenetic factors, whose alteration plays a differential role in DUX4 repression and, subsequently, contributes to determining the FSHD phenotype.

Generation of human myogenic progenitors from pluripotent stem cells for in vivo regeneration

Muscular dystrophy encompasses a large number of heterogeneous genetic disorders characterized by progressive and devastating muscle wasting. Cell-based replacement strategies aimed at promoting skeletal muscle regeneration represent a candidate therapeutic approach to treat muscular dystrophies. Due to the difficulties of obtaining large numbers of stem cells from a muscle biopsy as well as expanding these in vitro, pluripotent stem cells (PSCs) represent an attractive cell source for the generation of myogenic progenitors, given that PSCs can repeatedly produce large amounts of lineage-specific tissue, representing an unlimited source of cells for therapy. In this review, we focus on the progress to date on different methods for the generation of human PSC-derived myogenic progenitor cells, their regenerative capabilities upon transplantation, their potential for allogeneic and autologous transplantation, as well as the specific challenges to be considered for future therapeutic applications.

Generation of hiPSC-Derived Skeletal Muscle Cells: Exploiting the Potential of Skeletal Muscle-Derived hiPSCs

Cell therapies for muscle wasting disorders are on the verge of becoming a realistic clinical perspective. Muscle precursor cells derived from human induced pluripotent stem cells (hiPSCs) represent the key to unrestricted cell numbers indispensable for the treatment of generalized muscle wasting such as cachexia or intensive care unit (ICU)-acquired weakness. We asked how the cell of origin influences efficacy and molecular properties of hiPSC-derived muscle progenitor cells. We generated hiPSCs from primary muscle stem cells and from peripheral blood mononuclear cells (PBMCs) of the same donors (n = 4) and compared their molecular profiles, myogenic differentiation potential, and ability to generate new muscle fibers in vivo. We show that reprogramming into hiPSCs from primary muscle stem cells was faster and 35 times more efficient than from blood cells. Global transcriptome comparison revealed significant differences, but differentiation into induced myogenic cells using a directed transgene-free approach could be achieved with muscle- and PBMC-derived hiPSCs, and both cell types generated new muscle fibers in vivo. Differences in myogenic differentiation efficiency were identified with hiPSCs generated from individual donors. The generation of muscle-stem-cell-derived hiPSCs is a fast and economic method to obtain unrestricted cell numbers for cell-based therapies in muscle wasting disorders, and in this aspect are superior to blood-derived hiPSCs.

Anabolic Factors and Myokines Improve Differentiation of Human Embryonic Stem Cell Derived Skeletal Muscle Cells

Skeletal muscle weakness is linked to many adverse health outcomes. Current research to identify new drugs has often been inconclusive due to lack of adequate cellular models. We previously developed a scalable monolayer system to differentiate human embryonic stem cells (hESCs) into mature skeletal muscle cells (SkMCs) within 26 days without cell sorting or genetic manipulation. Here, building on our previous work, we show that differentiation and fusion of myotubes can be further enhanced using the anabolic factors testosterone (T) and follistatin (F) in combination with a cocktail of myokines (C). Importantly, combined TFC treatment significantly enhanced both the hESC-SkMC fusion index and the expression levels of various skeletal muscle markers, including the motor protein myosin heavy chain (MyHC). Transcriptomic and proteomic analysis revealed oxidative phosphorylation as the most up-regulated pathway, and a significantly higher level of ATP and increased mitochondrial mass were also observed in TFC-treated hESC-SkMCs, suggesting enhanced energy metabolism is coupled with improved muscle differentiation. This cellular model will be a powerful tool for studying in vitro myogenesis and for drug discovery pertaining to further enhancing muscle development or treating muscle diseases.