neurons. Compared with NEF-tTA/TDP-43M337V rats (Figure 2), ChAT–tTA-9/TDP-43M337V rats displayed a more severe loss of motor neurons, but displayed a similar progression of paralysis phenotype (Figure 6). We determined whether disease progression could be halted by stopping mutant TDP-43 expression after disease onset in ChAT–tTA-9/TDP-43M337V rats. When Dox was withdrawn from ChAT–tTA-9/TDP-43M337V rats at the age of 60 days (Figure 6A), the rats developed an early paralysis phenotype by 70 days of age and reached disease end stages by 78 days of age (Figure 6, C–F). Dox was thus withdrawn from rats that were 60 days old and was added back 12 days later, so that mutant TDP-43 was expressed in the rats from 63 to 77 days of age (Figure 6B and Figure 7B). We refer to restoring Dox as a treatment. Dox-untreated and Dox-treated rats all developed paralysis at comparable speeds before Dox produced an effect on disease gene expression (Figure 7C). Soft food was provided to rats with paralysis. While the disease rapidly progressed to end stages in Dox-untreated rats, the disease stopped progressing at paralysis stages in Dox-treated rats (Figure 7, C and D). Within a few days, Dox-treated rats began to gain body weight and to gain partial recovery of motor activity (Figure 7, C and D). Dox-untreated rats were terminated at disease end stages (by age of 80 days), and Dox-treated rats were terminated at 90 days of age.
In Dox-treated transgenic rats, H&E staining of gastrocnemius revealed remodeled motor units: some muscle fibers with centered nuclei, muscle fibers of varied sizes, and absence of atrophied muscle fibers (Figure 7, E–G, and T). Histochemistry for nonspecific esterase (Figure 7, H–J) and ATPase (Figure 7, K–M) showed that muscle cells of the same types, but with varied sizes, were grouped, suggesting a remodeling of motor units. During functional recovery, muscle fibers increased in size (Figure 7U). Toluidine blue staining revealed that most degenerated motor axons were cleared in the ventral roots (Figure 7, N–P). Interestingly, stereological cell counting revealed that the number of spinal motor neurons (L3–L5) between Dox-treated and Dox-untreated transgenic rats was similar (Figure 7V). These findings suggest that degenerating motor neurons cannot be rescued and that survived motor units can be remodeled.