Making Human Skeletal Muscle Tissue in Mice

We have developed a method to xenograft immortalized human muscle precursor cells from patients with FSHD and first-degree relative controls into the tibialis anterior muscle compartment of immuno-deficient mice, generating human muscle xenografts. We report that FSHD cells mature into organized and innervated human muscle fibers with minimal contamination of murine myonuclei. They also reconstitute the satellite cell niche within the xenografts. FSHD xenografts express DUX4 and DUX4 downstream targets, retain the 4q35 epigenetic signature of their original donors, and express two novel protein biomarkers of FSHD, TPM3 and SLC34A2. Ours is the first scalable, mature in vivohuman model of FSHD. It should be useful for studies of the pathogenic mechanism of the disease as well as for testing therapeutic strategies targeting DUX4 expression. 

Our Approach^*

We use immunodeficient Nod-RAG (NRG; NOD.Cg-Rag1^tm1Moml2rg^tm-Wjl/SzJ) mice.  We first X-irradiate their hindlimbs, to prevent the regeneration of the hindlimb muscles in later steps. We then inject a myotoxin along the length of the tibialis anterior (TA) muscles of each limb. (We have tested a number of different toxins, the best being cardiotoxin and BaCl₂.  As cardiotoxin is no longer commercially available, we typically use BaCl₂.)  The following day, we inject a concentrated suspension of human myogenic precursor cells (hMPCs) along the muscle.  One week later, we electrically stimulate the lower hindlimb muscles, including the TA, via the peroneal nerve for 30 min a day, 3 times a week.  This intermittent neuromuscular electrical stimulataion (iNMES) is continued for 4 weeks.  At that point, the xenograft has developed hundreds of human muscle fibers that are mature and that contain very few, if any, murine myonuclei (<2%). Mice may then be euthanized or used for further studies. 

An overview of the timeline is shown above. Details of the procedures are reported in references 1 and 2.

*These methods are patented in the US (reference 3)

---

---

Characteristics of the Engrafted Human Muscle

The fibers in graft are human, with minimal contamination by murine myonuclei.  They are also mature, by several criteria.

Identifying the Human Myofibers

Human myofibers in the grafts are readily identified by immunofluorescence methods, following incubation with antibodies that recognize human but not murine muscle proteins.  We typically use monoclonal antibodies to human β-spectrin (h-β-spectrin)  and to human lamin A/C (h-lamin A/C; see figure).  Human fibers in the xenograft are labeled green with antibodies to h-β-spectrin; fibers around the periphery of the graft are not, and thus are murine in origin.  Similarly, myonuclei in fibers labeled for h-β-spectrin are also positive for h-lamin A/C (purple). The occasional murine myonucleus (blue) is not.

State of Differentiation

We examined myosin in the contractile apparatus in the human myofibers generated following our standard protocol, with iNMES, and without electrical stimulation.  Fibers stimulated electrically eliminate nearly all their embryonic myosin, whereas unstimulated human fibers do not (see figure).  This suggests that our protocol generates mature human skeletal muscle tissue.

• 

• 

The presence of fully differentiated fibers in the xenografts is also indicated by the fact that desmin is organized in striations, when fibers are observed in longitudinal section, and in a reticulum surrounding the myofibrils, when fibers are observed in cross section (desmin in red; see Figure).

• 

• 

Innervation

The human fibers are innervated, as indicated by the presence of presynaptic termini labeled for synaptophysin (Syp, green) and postsynaptic membranes enriched in acetylcholine receptors (AchR, labeled with fluorescent α-bungarotoxin, blue) on the fiber surfaces (labeled for h-β-spectrin, red; see Figure).

• 

• 

• 

Satellite Cells

The human fibers in the grafts are associated with human satellite cells, that have repopulated their niche between the plasma membrane (labeled for h-β-spectrin and h-lamin A/C; red) and the basal lamina (labeled for laminin; purple).   The satellite cell nucleus is labeled with DAPI (blue) and antibodies to Pax7 (green).  Thus, the grafts have the capacity to grow and regenerate.

• 

• 

• 

• 

• 

Pathological Markers

The disease we have been focusing on is Facioscapulohumeral Muscular Dystrophy (FSHD) which occurs in individuals with chromosomal contractions at position 4q135 that result in a reduction in the number of D4Z4 satellite repeats at that locus. This contraction leads to a decrease in methylation of the terminal D4Z4 repeat and an increase in its transcription. When a polyadenylation sequence is included in the transcript, an mRNA encoding the transcription factor, DUX4, is produced. DUX4’s activity as a transcription factor activates dozens of downstream genes.  Our results confirm that the xenografts prepared with FSHD cells replicate these biochemical features of FSHD.  They have additional, unexpected features, that we are documenting:

Genetic

DUX4 is expressed in grafts developed from FSHD but not control cells.

All the gene products downstream of DUX4 that we have tested in FSHD vs control grafts are also upregulated at the mRNA level, including TRIM43, MBD3L2, ZSCAN4, LEUTX, and SLC34A2.  Data for the first three are shown here.  Control gene products (not shown) do not differ between FSHD and control grafts.

Epigenetic

Methylation of the terminal D4Z4 region of chromosome 4q35 is reduced in grafts prepared from FSHD but not control cells. (Methylated sites are indicated by red boxes, unmethylated, by blue boxes)

Morphology

Overall fiber diameter and spacing is not significantly different between FSHD and control grafts, but necrotic fibers, identified with an antibody to a potential biomarker of FSHD, SLC34A2, are present in significantly higher numbers in the FSHD grafts (arrow, Figure)

---

Work in Progress with the Xenografts

Collaborations with a number of laboratories and biopharma companies have been established to test the abilities of different approaches to treating FSHD with small molecules (inhibitors of p38 kinase, gapmers), oligonucleotide-antibody conjugates, and adeno-associated viral vectors carrying CRISPR/Cas9 or DUX4 pseudosubstrates.

We are also using the grafts in experiments to identify biomarkers of FSHD that have the potential to be used to monitor the efficacy of different therapies currently being developed for the clinic.  Our primary focus is currently on SLC43A2, a protein that is upregulated by DUX4 expression.

Experiments to explore variations of our xenografting protocols are also in progress, with3  goals: producing grafts with still higher numbers of human skeletal myofibers, reducing variability among grafts, and minimizing the effort required to generate the grafts.

Uses of the Xenografting Methods

Disease Models

Many studies of neuromuscular diseases, such as muscular dystrophies and myopathies, have been performed in animal models, especially in mice, in which the mutations that cause disease in man can be generated.  Although the murine models can sometimes only approximate the human disease, they have been very useful in identifying some of the underlying pathogenic mechanisms and in initial testing of therapies.  But murine models of human diseases that are not the result of simple genetic mutations are much harder to generate.  Our ability to prepare mice carrying mature human skeletal muscle tissue for individuals with such diseases offers the best murine model available for mechanistic and therapeutic investigations.  

Testing the Efficacy and Specificity of Therapies

Therapies designed to treat muscular dystrophies typically use either small molecules as drugs or genetic approaches to replace the protein missing in the disease or to suppress the expression of a protein that causes the disease. Testing the efficacy and specificity of these treatments on patients is a lengthy and expensive process.  Having mature human muscle tissue to test in mice provides a quicker and cheaper alternative.  Off-target and other non-specific effects as well as the efficacy of different treatments can be readily assessed in our xenografts. An example of our use of the xenografts to test a small molecule therapy for FSHD is the work we did with Fulcrum Therapeutics to examine the effects of p38 kinase inhibitors on the expression of DUX4 and its downstream products. Data we obtained with Fulcrum testing the Lilly drug, LY 2228820, is shown here.  The results show that the drug reduced expression of mRNA encoding 3 of DUX4’s downstream products (black, controls; blue, + LY compound)

Licensing

Any parties interested in licensing this technology for research and/or commercial purposes should contact the Office of Technology Transfer at the University of Maryland, Baltimore here.

References

1. Sakellariou, P., A. O’Neill, A.L. Mueller, G. Stadler, W.E. Wright, J.A. Roche and R.J. Bloch (2016) Neuromuscular Electrical Stimulation Promotes Development in Mice  of Mature Human Muscle from Immortalized Human Myoblasts. Skeletal Muscle 6, 4   [PMID:26925213; PMC4769538].
2. Mueller, A.L., A. O’Neill, T.I. Jones, A. Llach-Martinez, L.A. Rojas, P. Sakellariou, G. Stadler,  W.E. Wright, D. Eyerman, P.L. Jones, R.J. Bloch (2019) Muscle Xenografts Reproduce Key Molecular Features of Facioscapulohumeral Muscular Dystrophy. Exp. Neurol 320:113011 [PMID:31306642; PMC6730665]
3. Bloch, R.J., P. Sakellariou, A. O’Neill, and J.A. Roche (2020) “Methods of Generating Mature Human Muscle Fibers” US Patent Number 10,632,305 issued 4/28/2020. Valid thru 12/8/2036.