Early short-term PXT3003 combinational therapy delays disease onset in a transgenic rat model of Charcot-Marie-Tooth disease 1A (CMT1A)

Abstract

The most common type of Charcot-Marie-Tooth disease is caused by a duplication of PMP22 leading to dysmyelination, axonal loss and progressive muscle weakness (CMT1A). Currently, no approved therapy is available for CMT1A patients. A novel polytherapeutic proof-of-principle approach using PXT3003, a low-dose combination of baclofen, naltrexone and sorbitol, slowed disease progression after long-term dosing in adult Pmp22 transgenic rats, a known animal model of CMT1A. Here, we report an early postnatal, short-term treatment with PXT3003 in CMT1A rats that delays disease onset into adulthood. CMT1A rats were treated from postnatal day 6 to 18 with PXT3003. Behavioural, electrophysiological, histological and molecular analyses were performed until 12 weeks of age. Daily oral treatment for approximately 2 weeks ameliorated motor deficits of CMT1A rats reaching wildtype levels. Histologically, PXT3003 corrected the disturbed axon calibre distribution with a shift towards large motor axons. Despite dramatic clinical amelioration, only distal motor latencies were improved and correlated with phenotype performance. On the molecular level, PXT3003 reduced Pmp22 mRNA overexpression and improved the misbalanced downstream PI3K-AKT / MEK-ERK signalling pathway. The improved differentiation status of Schwann cells may have enabled better long-term axonal support function. We conclude that short-term treatment with PXT3003 during early development may partially prevent the clinical and molecular manifestations of CMT1A. Since PXT3003 has a strong safety profile and is currently undergoing a phase III trial in CMT1A patients, our results suggest that PXT3003 therapy may be a bona fide translatable therapy option for children and young adolescent patients suffering from CMT1A.

Conflict of interest statement

We have the following interests. This trial was financially supported by Pharnext, the employer of Serguei Nabirotchkin, Rodolphe Hajj and Daniel Cohen. DC, RH, MWS, TP and KAN submitted a patent based on this work: Full patent name: Early treatment of CMT disease Number: US2018/0000813. MWS, TP and KAN act as consultants to Pharnext. PXT3003 achieved a Phase 3 trial in CMT1A adult patients. There are no further patents related to this study, or marketed products to declare. This does not alter our adherence to all the PLOS ONE policies on sharing data and materials, as detailed online in the guide for authors.

Figures

Fig 1. Long-lasting restoration of phenotype deficits after early short-term PXT3003 therapy.

Early short-term therapy regimen followed by a long-term observation of motor and sensory, histological and molecular effects (A). Early postnatally PXT3003 application dose-dependently improved the grip strength impairment at the forelimbs after the treatment phase in 3 weeks aged CMT1A rats (B; WT controls 3.66±0.09, CMT1A controls 2.64±0.06, CMT PXT3003-3 2.93±0.10). First grip strength improvement was observed at the hindlimbs in 9 weeks aged CMT1A rats (C; WT controls 3.64±0.16, CMT1A controls 3.06±0.12, CMT PXT3003-3 3.47±0.12) lasting until 12 weeks (D; WT controls 5.38±0.23, CMT1A controls 4.16±0.22, CMT PXT3003-3 5.10±0.19). PXT3003 long-term effects increased over time reaching wildtype levels in the long-term observation (E). PXT3003 early short-term therapy improved motor deficits on the inclined plane in 3 (F; WT controls 0.10±0.07, CMT1A controls 1.00±0.18, CMT PXT3003-3 0.48±0.13) and 12 weeks aged CMT1A rats (G; WT controls 0.07±0.04, CMT1A controls 0.33±0.11, CMT PXT3003-3 0.07±0.04). Although some training effects were observed in the inclined plane behaviour in CMT1A controls, PXT3003 effects increased over time and reached wildtype levels at the study end (H). (ns = not significant, * = p

Fig 2. Improved motor latency correlating with the phenotype at study end.

Electrophysiological recordings including motor latency (red bar), nerve conduction velocity (NCV) and compound muscle potential (CMAP) analysis were performed as shown for representative measures (A). CMT1A controls showed prolonged motor latencies, which were dose-dependently improved by PXT3003 early short-term treatment (B; WT controls 1.66±0.10, CMT1A controls 3.35±0.22, CMT PXT3003-3 2.82±0.13). In long-term observation, the motor latency improvement correlated with the restored grip strength at the hindlimbs including all PXT3003 treated CMT1A rats (C). NCV and CMAP recordings were not affected by PXT3003 in long-term (D and F; WT controls 37.02±2.16, CMT1A controls 14.02±0,67, CMT PXT3003-3 14.86±0.53 and WT controls 5.31±0.47, CMT1A controls 1.49±0.25, CMT PXT3003-3 1.18±0.16, respectively), and did not correlate with the hindlimb grip strength of all PXT3003-treated CMT1A rats (E and G). (ns = not significant, * = p

Fig 3. Shift towards large-calibre axons in the long-term observation.

Histological analyses were performed on light microscopic level in the sciatic nerve as demonstrated in representative peripheral nerve cross sections. In contrast to WT controls, untreated CMT1A rats showed a loss of large-calibre axons (orange arrow) combined with a hypomyelination of predominantly large axons (green arrow) and a hypermyelination of predominantly small-calibre axons (blue arrow). A higher number of large-calibre axons appeared after PXT3003 treatment in CMT1A rats, without obvious differences between hypo- and hypermyelinated axons (A). Quantification of myelinated axons confirmed axonal loss in CMT1A controls not being affected by PXT3003 treatment (B; WT controls 9003±114, CMT1A controls 8654±74, CMT PXT3003-3 8727±68); however PXT3003 partially corrected the reduced mean axon diameter (C; WT controls 4.96±0.19, CMT1A controls 3.66±0.09, CMT PXT3003-2 3.97±0.08). More detailed analyses of axon calibre distribution confirmed an obvious loss of large-calibre axons in CMT1A rats, and a shift towards large-calibre axons after PXT3003 treatment (D). Mid- and large-calibre axons were corrected towards the wildtype situation as illustrated in more detail for 3–4μm and 5–6μm axons (E and F; WT controls 11.33±1.48, CMT1A controls 26.50±1.08, CMT PXT3003-3 22.44±1.43 and WT controls 21.83±1.54, CMT1A controls 11.11±1.12, CMT PXT3003-2 15.62±1.13, respectively). (ns = not significant, * = p

Fig 4. PXT3003 therapy acting on Pmp22 overexpression and downstream signalling in long-term at 12 weeks.

Early short-term applied PXT3003 was in long-term effective on the downregulation of disease-causing Pmp22 mRNA overexpression in CMT1A rats in the sciatic nerve (A; WT controls 1.00±0.04, CMT1A controls 1.48±0.07, CMT PXT3003-3 1.33±0.04; Pmp22 splice variant normalised to Cyclophilin A). In the brachial plexus, mRNA expression of the differentiation marker Sox2 (B; WT controls 1.00±0.09, CMT1A controls 3.68±0.55, CMT PXT3003-3 2.22±0.28; normalised to Cyclophilin A and Rplp0) and cJun (C; WT controls 1.00±0.09, CMT1A controls 2.18±0.34, CMT PXT3003-3 1.14±0.10; normalised to Cyclophilin A and Rplp0), as well as the biomarker Ctsa were influenced by PXT3003 in long-term observation (D; WT controls 1.00±0.08, CMT1A controls 1.28±0.16, CMT PXT3003-3 0.66±0.06; normalised to Cyclophilin A and Rplp0). Downregulated Hmgcr and Prx mRNA expression in CMT1A rats were not affected by PXT3003 (E; WT controls 1.00±0.08, CMT1A controls 0.40±0.03, CMT PXT3003-3 0.35±0.05 and F; WT controls 1.00±0.07, CMT1A controls 0.67±0.05, CMT PXT3003-3 0.63±0.08; normalised to Cyclophilin A and Rplp0). Pmp22 downstream signalling was investigated by Western Blot analysis using antibodies against p-AKT/AKT and p-MAPK/MAPK (G). At the end of long-term observation, PXT3003 early short-term treatment restored the decreased p-AKT/AKT signalling in CMT1A rats (H; WT controls 1.00±0.17, CMT1A controls 0.38±0.08, CMT PXT3003-3 0.93±0.21) and dose-dependently corrected the disrupted ratio of p-AKT/AKT to p-MAPK/MAPK signalling towards the wildtype situation (I; CMT1A controls 0.30±0.08, CMT PXT3003-1 0.56±0.15, CMT PXT3003-2 0.85±0.13, CMT PXT3003-3 1.36±0.52), which is characterized by a higher relative amount of p-AKT versus p-MAPK. (ns = not significant, * = p<0.05, **p<0.01 and *** = p<0.001).