Genomic and Epigenetic Changes Drive Aberrant Skeletal Muscle Differentiation in Rhabdomyosarcoma

Abstract

Rhabdomyosarcoma (RMS), the most common soft-tissue sarcoma in children and adolescents, represents an aberrant form of skeletal muscle differentiation. Both skeletal muscle development, as well as regeneration of adult skeletal muscle are governed by members of the myogenic family of regulatory transcription factors (MRFs), which are deployed in a highly controlled, multi-step, bidirectional process. Many aspects of this complex process are deregulated in RMS and contribute to tumorigenesis. Interconnected loops of super-enhancers, called core regulatory circuitries (CRCs), define aberrant muscle differentiation in RMS cells. The transcriptional regulation of MRF expression/activity takes a central role in the CRCs active in skeletal muscle and RMS. In PAX3::FOXO1 fusion-positive (PF+) RMS, CRCs maintain expression of the disease-driving fusion oncogene. Recent single-cell studies have revealed hierarchically organized subsets of cells within the RMS cell pool, which recapitulate developmental myogenesis and appear to drive malignancy. There is a large interest in exploiting the causes of aberrant muscle development in RMS to allow for terminal differentiation as a therapeutic strategy, for example, by interrupting MEK/ERK signaling or by interfering with the epigenetic machinery controlling CRCs. In this review, we provide an overview of the genetic and epigenetic framework of abnormal muscle differentiation in RMS, as it provides insights into fundamental mechanisms of RMS malignancy, its remarkable phenotypic diversity and, ultimately, opportunities for therapeutic intervention.

Keywords: differentiation; epigenetics; genomics; rhabdomyosarcoma; skeletal muscle.

Figure 1

Normal skeletal muscle regeneration. Skeletal muscle is composed of bundles of terminally differentiated myofibers containing multiple post-mitotic nuclei. Muscle-resident stem cells, called satellite cells, are positioned beneath the basal lamina of mature myofibers. MYF5 is the only active MRF expressed alongside PAX7 in quiescent satellite cells in adult skeletal muscle. Upon muscle injury, MRFs are deployed to activate satellite cells, govern myoblast proliferation/fusion and, ultimately, withdrawal from the cell cycle and terminal differentiation. The transcriptional activity and protein availability of MRFs during myoblast proliferation/differentiation is tightly regulated. Regeneration cues rapidly induce MYOD1 expression. Activated MYOD1-/PAX7-/MYF5-expressing satellite cells (myogenic precursors or myoblasts) continue to proliferate (with MYOD1 and MYF5 expression depending on cell cycle phase) and down-regulate PAX7 while maintaining MYOD1, before they finally commit to myogenic differentiation via induction of MYF4. Other myoblasts maintain PAX7 but down-regulate MYOD1 and ultimately withdraw from the cell-cycle to regain quiescence and repopulate the satellite cell pool. This step-wise, hierarchical process is governed by complex epigenetic machinery. MRFs act as tissue-restricted transcription factors, which heterodimerize with bHLH proteins and bind to the regulatory regions of muscle-specific genes to activate transcription. MYOD1 binding also correlates with the opening of the chromatin structure at target genes through histone acetylation. Critical regulators of this process include p38α and a number of chromatin modifiers such as EZH2 and others.

Figure 2

RMS cell-to-cell heterogeneity recapitulating distinct states of muscle development. The PAX::FOXO1 fusion-positive (FP) and PAX::FOXO1 fusion-negative (FN) RMS cell pool harbors cells stalled at distinct developmental states: a transcriptionally immature stem cell-like/mesoderm/mesenchymal state (expressing PAX3, PAX7 CD44), a myoblast state (expressing MYF5) containing highly proliferative/cycling cells, and a more differentiated myocyte state (expressing MYF4/MYH3/MYH8) associated with better patient outcomes. Conventional chemotherapy depletes the myoblast state and enriches cells in the mesenchymal state, which characterizes a tumor-propagating subpopulation capable of re-creating original tumor heterogeneity. The tumor cell pool in alveolar RMS is more skewed towards later stages of myogenesis prominent compared to embryonal RMS. FP RMS also displays more prominent intrinsic plasticity, which may be the mechanism at the root of its aggressive behavior.

Figure 3

Core regulatory circuitries in RMS. Cell type-specific transcriptional programs are tightly controlled by master TFs (MTFs). CRCs are interconnected transcriptional loops in which individual MTFs control their own expression as well as the expression of the other involved TFs and exert feed-forward transcriptional control. These transcriptional networks control key transcriptional programs, which drive cell fate specification and differentiation. At the chromatin level, MTFs are accumulated at super-enhancers (SEs) and function under the control of SEs. CRCs have been identified for a number of different tissues, including normal skeletal muscle and RMS. Members of the MRF and MEF2 families are at the core of the normal muscle CRC, and MRFs appear to be the most common elements shared by CRCs in RMS. Generally speaking, the normal muscle CRC is active in RMS and complemented with additional modules: (A) The P3F+ RMS module relies on MYCN and P3F; the latter has pioneer factor activity and shapes the SE landscape in PF+ RMS cells. (B) The PF− RMS module involves PAX7 and AP1 family TFs.

Figure 4

Epigenetic enzymes involved in aberrant myogenic differentiation in RMS. In the nucleus, human chromosomal DNA is organized in chromatin fiber loops, which are coiled around nucleosomes. The latter contain histones whose tails are subjected to post-translational modifications, including acetylation (Ac) and methylation (Me). This schematic representation summarizes the epigenetic enzymes, which have been reported to be responsible for histone modifications identified in RMS (green arrows indicate stimulation; red signs indicate blockade).

Figure 5

Inhibition of myogenic differentiation by SNAI2. In proliferating myoblasts and PF− RMS cells, SNAI2 binding at super-enhancers dampens MYOD1 activity at myogenic genes, supporting proliferation at the expense of differentiation (left panel). Upon receipt of differentiation cues, SNAI2 is downregulated in differentiating myoblasts, thus permitting MYOD1 transcriptional activity at myogenic differentiation genes (MYF4, MEF2A, TNNT2, TNNI1) and inducing muscle differentiation (right panel). In PF− RMS, the myogenic differentiation process can be achieved through SNAI2 knockdown with short hairpin RNA (shRNA) alone or through treatment with the MEK inhibitor (MEKi) trametinib (right panel).