Preclinical testing of the glycogen synthase kinase-3β inhibitor tideglusib for rhabdomyosarcoma

Abstract

Rhabdomyosarcoma (RMS) is the most common childhood soft tissue sarcoma. RMS often arise from myogenic precursors and displays a poorly differentiated skeletal muscle phenotype most closely resembling regenerating muscle. GSK3β is a ubiquitously expressed serine-threonine kinase capable of repressing the terminal myogenic differentiation program in cardiac and skeletal muscle. Recent unbiased chemical screening efforts have prioritized GSK3β inhibitors as inducers of myodifferentiation in RMS, suggesting efficacy as single agents in suppressing growth and promoting self-renewal in zebrafish transgenic embryonal RMS (eRMS) models in vivo. In this study, we tested the irreversible GSK3β-inhibitor, tideglusib for in vivo efficacy in patient-derived xenograft models of both alveolar rhabdomyosarcoma (aRMS) and eRMS. Tideglusib had effective on-target pharmacodynamic efficacy, but as a single agent had no effect on tumor progression or myodifferentiation. These results suggest that as monotherapy, GSK3β inhibitors may not be a viable treatment for aRMS or eRMS.

Keywords: GSK3β; myodifferentiation; patient-derived xenograft; preclinical testing; rhabdomyosarcoma.

Figure 1. GSK3α and GSK3β expression in RMS cell lines, patient samples, and normal muscle

(A) Schematic representation of full length GSK3α1/α2/α3 and GSK3β1/β2 showing their catalytic domain (kinase), sites of serine (S) and tyrosine (Y) phosphorylation. (B) RNA sequencing was performed on 31 RMS cell lines, 105 RMS patient samples, and 19 normal muscle tissue samples and the resulting Log2-scaled RPKM values for 4 isoforms of GSK3α and GSK3β are shown. Different sample types (RMS cell line, RMS patient sample, normal muscle) are indicated by the color-coded bars at the top of the figure. The heat scale is given on the side, ranging from green (low expression; RPPKM = −3), to black (RPKM = 0), to red (high expression; RPKM=3).(C & D) Table showing the different spliced variant of GSK3α and GSK3β with their respective ensemble ID, gene symbol, protein length (a.a) and their expression across, aRMS, eRMS patient samples and cell lines (color code matching heat map above). (E) Western blotting showing pattern of GSK3 α/β expression across aRMS, eRMS, or primary tissue versus cell line samples. GSK3α and GSK3β have molecular weight of 51 and 47 KDa.

Figure 2. Effects of tideglusib on tumor growth, myodifferentiation in vivo

(A & B) Western blotting of vehicle and tideglusib treated human PDX derived primary culture (PCB82 and PCB380) for detection of GSK-3β mediated phosphorylation of β-catenin which showed reduction and an increase in total β-catenin. (C) Kaplan-Meier curve showing eRMS (PCB82) and aRMS mice (PCB380) treated with 200 mg/kg tideglusib via oral gavage daily experienced no effect on survival. (D & E) Western blotting of vehicle and tideglusib treated human PDX tumors (PCB82 and PCB380) for detection of GSK3β mediated phosphorylation of β-catenin which showed reduction and an increase in total β-catenin. (F) Densitometric analysis shows the reduction in phos-β-catenin upon tideglusib treatment in eRMS (PCB82) (upper panel) and aRMS (PCB380) (lower panel) to be statistically significant (*p>0.05; **p>0.001). Error bars represent mean ± S.D. (G & H) Western blotting of vehicle and tideglusib treated human PDX derived primary culture (PCB82 and PCB380) for detection of myogenin and myosin heavy chain (MHC). Differentiated HSMM used as a positive control.