tRNA overexpression rescues peripheral neuropathy caused by mutations in tRNA synthetase

Abstract

Heterozygous mutations in six transfer RNA (tRNA) synthetase genes cause Charcot-Marie-Tooth (CMT) peripheral neuropathy. CMT mutant tRNA synthetases inhibit protein synthesis by an unknown mechanism. We found that CMT mutant glycyl-tRNA synthetases bound tRNA^Gly but failed to release it, resulting in tRNA^Gly sequestration. This sequestration potentially depleted the cellular tRNA^Gly pool, leading to insufficient glycyl-tRNA^Gly supply to the ribosome. Accordingly, we found ribosome stalling at glycine codons and activation of the integrated stress response (ISR) in affected motor neurons. Moreover, transgenic overexpression of tRNA^Gly rescued protein synthesis, peripheral neuropathy, and ISR activation in Drosophila and mouse CMT disease type 2D (CMT2D) models. Conversely, inactivation of the ribosome rescue factor GTPBP2 exacerbated peripheral neuropathy. Our findings suggest a molecular mechanism for CMT2D, and elevating tRNA^Gly levels may thus have therapeutic potential.

Fig. 1.. tRNA^Gly overexpression rescues inhibition of protein synthesis and peripheral neuropathy phenotypes in Drosophila CMT2D models.

(A) Schematic of the transgenes used for tRNA^Gly-GCC or tRNA^Gly-UCC overexpression. (B) Relative translation rate as determined by FUNCAT in motor neurons (OK371-GAL4) of larvae expressing E71G, G240R, or G526R GlyRS (2x: two transgene copies), in the presence or absence of the tRNA^Gly-GCC BAC transgene (10xtRNA^Gly-GCC). n=10–34 animals per genotype; ***p<0.001 by Kruskal-Wallis test. (C,E,G) Percentage of larvae with innervated muscle 24. GlyRS transgenes were expressed in motor neurons (OK371-GAL4), in the presence or absence of 10xtRNA^Gly-GCC BAC (C), 10xtRNA^Gly-GCC scramble (E), or 12xtRNA^Gly-UCC (G). n=19–26 (C), 8–22 (E), 12–27 (G) animals per genotype; *p<0.05; ***p<0.005 by Fisher’s exact test with Bonferroni correction. (D,F,H,I) Negative geotaxis climbing speed of 7-day-old female flies expressing GlyRS transgenes in motor neurons (OK371-GAL4), in the presence or absence of 10xtRNA^Gly-GCC BAC (D), 10xtRNA^Gly-GCC scramble (F), or 12xtRNA^Gly-UCC (H,I). n=11–13 (D), 8–52 (F), 7–19 (H,I) groups of 10 flies per genotype; **p<0.01; ***p<0.005 by two-way ANOVA (D) or Brown-Forsythe and Welch ANOVA (F,H,I). Controls in (B-I) are driver-only. Graphs represent mean ± SEM.

Fig. 2.. tRNA^Gly-GCC overexpression rescues peripheral neuropathy in CMT2D mouse models.

(A) Schematic of the genomic fragment used for generation of tRNA^Gly-GCC transgenic mice. (B,M) Hanging time in the inverted grid test of male Gars^C201R/+ x tRNA^Gly-high (B) or Gars^ΔETAQ/+ x tRNA^Gly-high (M) mice. n=8–9 (B), 11–13 (M) mice per genotype; ***p<0.0001 by one-sample t-test and two-tailed unpaired t-test with Bonferroni correction per time point. (C,I,N) 4-paw grip strength as measured by dynamometer. n=8–9 (C), 10–11 (I), 11–13 (N) mice per genotype; ***p<0.001 by two-way ANOVA with Tukey’s multiple comparisons test per time point (C,I) or Brown-Forsythe and Welch ANOVA (N). (D,E,J,K,O,P) Electromyography (EMG) at 12 (D,E,O,P) or 52 (J,K) weeks of age. (D,J,O) Latency time between sciatic nerve stimulation at sciatic notch level and detection of a compound muscle action potential (CMAP) in the gastrocnemius muscle. n=8–9 (D), 10–11 (J), 11–13 (O) mice per genotype; ***p<0.0001 by two-way ANOVA with Tukey’s multiple comparisons test (D) or Brown-Forsythe and Welch ANOVA (J,O). (E,K,P) CMAP amplitude in the gastrocnemius muscle. n=8–9 (E), 10–11 (K), 11–13 (P) mice per genotype; ***p<0.0005 by Brown-Forsythe and Welch ANOVA (E,P) or two-way ANOVA with Tukey’s multiple comparisons test (K). (F,L,Q) Ratio of muscle weight to body weight (MW:BW, shown as % of WT) of the gastrocnemius at 12 (F,Q) or 52 (L) weeks of age. n=8–9 (F), 10–11 (L), 11–13 (Q) mice per genotype; ***p<0.0001 by two-way ANOVA with Tukey’s multiple comparisons test. (G,H,R) Representative images (G) and quantification (H,R) of NMJ innervation status in plantaris muscle. In (G), neurofilament (NF) and SV2 label presynaptic nerve endings, while TRITC-conjugated bungarotoxin (BTX) labels postsynaptic acetylcholine receptors. n=5 mice per genotype; ***p<0.005 by Fisher’s Exact test with Bonferroni correction. Scale bar: 25μm. Graphs represent mean ± SEM.

Fig. 3.. tRNA^Gly sequestration by CMT-mutant GlyRS induces ribosome stalling.

(A) Size-exclusion chromatography of purified recombinant human GlyRS proteins. D:M= dimer:monomer ratio. (B) K_on and K_off values of tRNA^Gly-GCC binding and release to dimer and monomer forms of the indicated GlyRS variants. The (percentage) denotes the frequency of a measured value. (C) Quantification of tRNA^Gly bound to GlyRS in tRNA^Gly:GlyRS complexes immunoprecipitated from whole brains of Gars^C201R/+ and WT littermate control mice. tRNA^Gly/GlyRS ratio of WT is set as 100%; n=5 independent experiments; *p<0.05 by one-sample t-test. (D) Hanging time in the inverted grid test of male Gtpbp2^{+/? or −/−}; Gars^+/+ (control), Gtpbp2^+/?; Gars^C201R/+, and Gtpbp2^−/−; Gars^C201R/+ littermate mice at 4, 5, 6, 7 and 8 weeks of age. n=15–28 mice per genotype group; ***p<0.0005 by one-sample t-test and two-tailed unpaired t-test with Bonferroni correction per time point. (E) Nerve conduction velocity of the sciatic nerve at 8 weeks of age. n=13–20 mice per genotype group; ***p<0.0001 by Brown-Forsythe and Welch ANOVA. (F) Axon number in the motor branch of the femoral nerve at 8 weeks of age. n=8–13 per genotype group; ***p<0.0001 by one-way ANOVA with Tukey’s multiple comparisons test. Graphs represent mean ± SEM.

Fig. 4.. tRNA^Gly-GCC overexpression prevents ISR activation in CMT2D mouse models.

(A,B) Representative images (A) and quantification (B) of immunostaining intensity of phosphorylated eIF2α in motor neuron cell bodies in the spinal cord ventral horn of Gars^ΔETAQ/+ x tRNA^Gly-high mice. Scale bar: 100μm. n=4–5 mice per genotype; **p<0.01, ***p<0.0005 by two-way ANOVA with Tukey’s multiple comparisons test. (C-F) Representative images (C) and quantification of fluorescent in situ hybridization (FISH) for ATF4 target genes Gdf15 (D), Adm2 (E), and B4galnt2 (F). Scale bar: 50μm. n=5–6 mice per genotype; *p<0.05 by two-tailed Welch’s t-test with Bonferroni correction. (G-I) mRNA levels of ATF4 target genes Gdf15 (G), Fgf21 (H) and B4galnt2 (I) in spinal cord of Gars^C201R/+ x tRNA^Gly-high mice. n=3 per genotype; *p<0.05, **p<0.01 by Brown-Forsythe and Welch ANOVA. Graphs represent mean ± SEM.