Light modulation ameliorates expression of circadian genes and disease progression in spinal muscular atrophy mice

Abstract

Physiology and behaviour are critically dependent on circadian regulation via a core set of clock genes, dysregulation of which leads to metabolic and sleep disturbances. Metabolic and sleep perturbations occur in spinal muscular atrophy (SMA), a neuromuscular disorder caused by loss of the survival motor neuron (SMN) protein and characterized by motor neuron loss and muscle atrophy. We therefore investigated the expression of circadian rhythm genes in various metabolic tissues and spinal cord of the Taiwanese Smn-/-;SMN2 SMA animal model. We demonstrate a dysregulated expression of the core clock genes (clock, ARNTL/Bmal1, Cry1/2, Per1/2) and clock output genes (Nr1d1 and Dbp) in SMA tissues during disease progression. We also uncover an age- and tissue-dependent diurnal expression of the Smn gene. Importantly, we observe molecular and phenotypic corrections in SMA mice following direct light modulation. Our study identifies a key relationship between an SMA pathology and peripheral core clock gene dysregulation, highlights the influence of SMN on peripheral circadian regulation and metabolism and has significant implications for the development of peripheral therapeutic approaches and clinical care management of SMA patients.

Figure 1

Dysregulation of diurnal expression of core clock genes in several tissues of pre-symptomatic SMA mice. Diurnal expression of core clock genes (Clock, Bmal1, Per1, Per2, Cry1 and Cry2) in TA, WAT, BAT, liver, heart and SC of postnatal day (P) 2 Smn^−/−;SMN2 mice compared to healthy littermates. Data are mean ± SEM, n = 3–5 mice per ZT, ^*P-value < 0.05, ^**P-value < 0.01, ^***P-value < 0.001, ^****P-value < 0.0001 (two-way ANOVA), # indicates circadian rhythmicity. ZT1 data are duplicated.

Figure 2

Dysregulation of diurnal expression of core clock genes in several tissues of symptomatic SMA mice. Diurnal expression of core clock genes (Clock, Bmal1, Per1, Per2, Cry1 and Cry2) in TA, WAT, BAT, liver, heart and SC of postnatal day (P) 7 Smn^−/−;SMN2 mice compared to healthy controls. Data are mean ± SEM, n = 3–4 mice per ZT, ^*P-value < 0.05, ^**P-value < 0.01, ^***P-value < 0.001, ^****P-value < 0.0001 (two-way ANOVA), # indicates circadian rhythmicity. ZT1 data are duplicated.

Figure 3

Dysregulation of diurnal expression of clock output genes in several tissues of pre-symptomatic and symptomatic SMA mice. Diurnal expression of clock output genes (Nr1d1, Dbp,) in TA, WAT, BAT, liver, heart and SC of postnatal day (P) 2 (A) and P7 (B) Smn^−/−;SMN2 mice compared to healthy littermates. P2 data are mean ± SEM, n = 3–5 mice per ZT, P7 data are mean ± SEM, n = 3–4 per ZT, ^*P-value < 0.05, ^**P-value < 0.01, ^***P-value < 0.001 (two-way ANOVA), # indicates circadian rhythmicity. ZT1 data are duplicated.

Figure 4

Summary of dysregulations of diurnal expression of clock and clock output genes in SMA tissues during disease progression. An X indicates either a change in phase, change in amplitude or differential expression of core clock and clock output genes in postnatal (P) 2 (A) and P7 (B) SMA mice compared to healthy littermates at one or more time points during a 24-h period.

Figure 5

Diurnal expression of MyoD and myogenin is not significantly affected in skeletal muscle of SMA mice during disease progression. Diurnal expression of MyoD and myogenin in TA of postnatal day (P) 2 (A) and P7 (B) Smn^−/−;SMN2 mice compared to healthy littermates. P2 data are mean ± SEM, n = 3–6 mice per ZT, P7 data are mean ± SEM, n = 3–5 mice per ZT, ^*P-value < 0.05, ^**P-value < 0.01 (two-way ANOVA), # indicates circadian rhythmicity. ZT1 data are duplicated.

Figure 6

The Smn gene displays an age-, tissue- and light-dependent diurnal expression. Diurnal expression of Smn in TA, SC, liver, heart, WAT and BAT of postnatal day (P) 2 (A) and P7 (B) Smn^+/−;SMN2 mice healthy control mice. P2 data are mean ± SEM, n = 3–4 mice per ZT, P7 data are mean ± SEM, n = 3–4 mice per ZT, ^*P-value < 0.05, ^**P-value < 0.01, ^***P-value < 0.001 (One-way ANOVA), # indicates circadian rhythmicity. ZT1 data is duplicated. (C) Diurnal expression of Smn in WAT and BAT from P7 WT and Per1/2 mutants (KO) exposed to constant darkness. Data are mean ± SEM, n = 3–7 mice per ZT, ^*P-value < 0.05 (two-way ANOVA), # indicates circadian rhythmicity. ZT1 data are duplicated. (D) Diurnal expression of Smn in epididymal eWAT and BAT from 3- to 4-month-old male WT and Per1/2 mutants (KO). Data are mean ± SEM, n = 3 mice per ZTs. ZT1 data ar duplicated.

Figure 7

Light modulation impact molecular and phenotypic parameters in SMA mice. (A) Lifespan and weight development of Smn^−/−;SMN2 mice in CL (n = 20) and RL (n = 14) conditions. P-value = 0.0003 for Kaplan–Meier log-rank (Mantel-Cox); Data are mean ± SEM, ^*P-value < 0.05, ^****P-value < 0.0001 (two-way ANOVA). (B) Weight curves of healthy littermates in CL (n = 19) versus RL (n = 14). Data are mean ± SEM, ^**P-value < 0.01, ^***P-value < 0.001, ^****P-value < 0.0001 (two-way ANOVA). (C) Corticosterone levels in serum of female breeders (n = 2 for RL and CL), healthy littermates (n = 7 for RL and CL) and Smn^−/−;SMN2 mice (n = 11 for RL and 9 for CL). Data are mean ± SEM, ^***P-value < 0.001, ns = not significant (two-way ANOVA). (D) Circadian rhythm genes dysregulated in ZT9 BAT from P7 Smn^−/−;SMN2 mice compared to healthy controls in CL conditions. (E) Heat map of circadian rhythm genes in ZT9 BAT of P7 Smn^−/−;SMN2 mice and healthy littermates from regular (RL) and CL shows genes dysregulated in Smn^−/−;SMN2 mice in RL and restored in Smn^−/−;SMN2 mice in CL. n = 4 for all experimental groups. F. Heat map of circadian rhythm genes in ZT17 BAT of P7 Smn^−/−;SMN2 mice and healthy littermates from RL and CL shows genes dysregulated in Smn^−/−;SMN2 mice in RL and restored in Smn^−/−;SMN2 mice in CL. n = 4 for all experimental groups. (G) Lifespan and weight development of Smn^−/−;SMN2 mice in CL (n = 19) versus light pulses (n = 12). ns = not significant for Kaplan–Meier log-rank (Mantel-Cox); data are mean ± SEM, ^*P-value < 0.05, ^**P-value < 0.01, ^***P-value < 0.001, ^****P-value < 0.0001 (two-way ANOVA). (H) Weight curves of healthy controls in CL (n = 16) and light pulses (n = 11). Data are mean ± SEM, ^*P-value < 0.05 (two-way ANOVA).

Figure 8

Light modulation impacts certain canonical pathological markers of SC and muscle pathology of SMA mice. TA and SC were harvested from postnatal (P) 7 Smn^−/−;SMN2 mice (n = 7 for RL and 3 for CL) and healthy littermates (n = 5 for RL and 6 for CL) in CL and RL conditions at ZT1 (9 am). Expression of MuRF-1 (A) and atrogin-1 (B) in TA of P7 Smn^−/−;SMN2 mice and healthy littermates from RL and CL. Expression of Fas (C) and Pmaip1 (D) in SC of P7 Smn^−/−;SMN2 mice and healthy littermates from RL and CL. Data are mean ± SEM, ^*P-value < 0.05, ^**P-value < 0.01, ns = not significant (one-way ANOVA).