Deficiency of the zinc finger protein ZFP106 causes motor and sensory neurodegeneration

Abstract

Zinc finger motifs are distributed amongst many eukaryotic protein families, directing nucleic acid-protein and protein-protein interactions. Zinc finger protein 106 (ZFP106) has previously been associated with roles in immune response, muscle differentiation, testes development and DNA damage, although little is known about its specific function. To further investigate the function of ZFP106, we performed an in-depth characterization of Zfp106 deficient mice (Zfp106(-/-)), and we report a novel role for ZFP106 in motor and sensory neuronal maintenance and survival. Zfp106(-/-) mice develop severe motor abnormalities, major deficits in muscle strength and histopathological changes in muscle. Intriguingly, despite being highly expressed throughout the central nervous system, Zfp106(-/-) mice undergo selective motor and sensory neuronal and axonal degeneration specific to the spinal cord and peripheral nervous system. Neurodegeneration does not occur during development of Zfp106(-/-) mice, suggesting that ZFP106 is likely required for the maintenance of mature peripheral motor and sensory neurons. Analysis of embryonic Zfp106(-/-) motor neurons revealed deficits in mitochondrial function, with an inhibition of Complex I within the mitochondrial electron transport chain. Our results highlight a vital role for ZFP106 in sensory and motor neuron maintenance and reveal a novel player in mitochondrial dysfunction and neurodegeneration.

Figure 1.

Behavioural analyses of Zfp106^−/− mice reveal severe motor abnormalities. Black bars: Zfp106^−/−, grey bars: Zfp106^+/−, white bars: Zfp106^−/−. (A) Photograph of mutant Zfp106^−/− mouse at 15-weeks of age displaying pronounced kyphosis and unable to coordinate hind limbs on wire grate. (B) qPCR analysis of Zfp106 expression levels in brain and spinal cord of 6-week-old male WT and Zfp106^−/− littermates (n = 3 per genotype); expression in Zfp106^−/− mice is normalized to Zfp106 expression in respective WT tissues, taken as a value of 1 (Log scale). (C) Female weights recorded weekly from ages 3 to 15 weeks; at least five mice were assessed per genotype per time point. Weight is diminished in Zfp106^−/− female mice from 3-weeks of age (P = 0.04) and continues to significantly decrease, compared with WT and Zfp106^+/− animals, to 15-weeks of age (P < 0.001). (D) Female grip strength and (E) accelerated Rotarod performance are reduced in Zfp106^−/− mice at 6- and 13-weeks of age compared with WT and Zfp106^+/− littermates (n≥ 6 per genotype). (F and G) Open field assessment of (F) distance moved, and (G) velocity for female mice at 7- and 14-weeks of age (see ‘Materials and Methods’). A reduction in distance moved (F), and velocity (G), was seen in 14-week-old female Zfp106^−/− mice when compared with 7-week female Zfp106^−/− mice; P values are indicated (n ≥ 5 per genotype and time point) and 14-week old WT littermates. (H) Female mice assessed for defects in coordination of front and rear legs at 7- and 14-weeks of age using the Locotronic^® system (see ‘Materials and Methods’). Zfp106^−/− mice made significantly more front and rear leg placement errors compared with WT littermates. Rear leg errors also significantly increased for 14-week-old Zfp106^−/− mice compared with 7-week-old Zfp106^−/− mice (n ≥ 5 per time point and genotype; P value indicated). Numbers and percentages shown represent the mean ± SEM. *P < 0.05; **P < 0.001. (I) Survival of male and female Zfp106^−/− mice as determined by their humane endpoint (see ‘Materials and Methods’).

Figure 2.

Assessment of hind limb muscle force, motor unit survival and muscle fatigue on female mice. (A and B) Summary of the maximum tetanic muscle force generated from the TA muscle. (A) TA tetanic muscle force generated by Zfp106^−/− mice at 7-weeks of age (47 g ± 4 g) is reduced compared with WT littermates (133 g ± 4 g). TA tetanic tension is further reduced in 15-week-old Zfp106^−/− mice (32 g ± 4 g) compared with 7-week-old Zfp106^−/− mice. (B) Representative traces of TA tetanic tension from WT and Zfp106^−/− mice at 7- and 15-weeks of age. (C and D) Tetanic muscle force generated from the EDL at 7- and 15-weeks of age (see ‘Materials and Methods’). (C) EDL tetanic muscle force generated by Zfp106^−/− mice at 7-weeks (29 g ± 1 g) is reduced compared with WT littermates (37 g ± 2 g) and becomes further reduced in 15-week-old Zfp106^−/− mice (21 g ± 3 g). (D) Representative traces of EDL tetanic tension from WT, 7- and 15-week old Zfp106^−/− mice. (E and F) The number of surviving motor units in the EDL muscle of female mice at 7- and 15-weeks of age; motor units are reduced in 7-week-old Zfp106^−/− mice (26 ± 0.9) compared with WT littermates (31 ± 0.4), and are further reduced in 15-week-old Zfp106^−/− mice (23 ± 1). (F) Representative motor unit traces from EDL muscles of WT, 7- and 15-week old Zfp106^−/− mice. Each twitch trace recording is a single motor unit. (G and H) FI (see ‘Materials and Methods’ section) of EDL muscle for female mice at 7 and 15-weeks of age. FI of 7-week old Zfp106^−/− mice (0.48 ± 0.05) is increased compared with WT littermates (0.32 ± 0.04) but not significantly (P = 0.07). FI increases further in 15-week-old Zfp106^−/− mice (0.69 ± 0.05) compared with 7-week old Zfp106^−/− mice. (H) Representative fatigue traces from WT, 7- and 15-week Zfp106^−/− mice, produced by repetitive stimulation of the EDL muscle for 180 s. Each line in the trace represents a single tetanic tension. Results are presented as the mean ± SEM. At least 10 legs were assessed per genotype per time point. *P ≤ 0.001, or indicated within each chart when comparing Zfp106^−/− mice from 7- and 15-weeks of age.

Figure 3.

Motor neuron and axonal degeneration and gliosis in Zfp106^−/− mouse spinal cord. (A and B) Representative toluidine blue stained sections of lumbar L4 and L5 ventral roots from 15-week old WT (A) and Zfp106^−/− (B) mice. Zfp106^−/− mice have fewer motor axons than WT littermates (spaces are marked with asterisks), and have degenerating axons, indicated with arrow head. Scale bar is 20 μm. (C) Representative images of the ventral horn of lumbar spinal cord sections stained for Nissl from WT, 7- and 15-week-old Zfp106^−/− mice. The number of motor neurons in the sciatic pool (inset) was counted for each genotype and age to calculate motor neuron survival (D). Scale bars: main images 200 μm inset 100 μm. (D) The survival of motor neurons in the sciatic pool in female mice at 7- and 15-weeks of age. Motor neuron survival in 7-week-old Zfp106^−/− mice (294 ± 11) is reduced compared with WT littermates (481 ± 9) and further reduced in 15-week-old Zfp106^−/− mice (217 ± 9). Numbers represent the mean ± SEM. At least five animals were assessed per genotype per time point (*P < 0.001). (E) Representative toluidine blue stained section of 15-week-old Zfp106^−/− lumbar spinal cord showing an example of a vacuolated (arrow) motor neuron. (F and G) Lumbar spinal cord sections of 15-week-old WT (F) and Zfp106^−/− (G) mice stained for IBA-1 (green), GFAP (red) and Nissl (blue). Immunoreactivity for micro- and astrogliosis is dramatically increased in the lumbar of 15-week-old Zfp106^−/− mice compared with WT littermates. Scale bar is 20 μm.

Figure 4.

Axonal degeneration of sensory neurons. (A, B) Representative toluidine blue stained sections of lumbar L4 and L5 dorsal roots from WT (A) and Zfp106^−/− (B) mice. Zfp106^−/− mice have fewer sensory axons, and degenerating axons, indicated with arrow head. Scale bar is 20 μm. (C) Electron micrograph of Zfp106^−/− dorsal root showing degenerating myelinated axons (arrow head) and evidence of loss of unmyelinated axons with denervated Schwann cells (arrow). Scale bar is 5 μm. (D and E) Representative toluidine blue stained section of lumbar DRG from WT (D) and Zfp106^−/− (E) mice. Numerous DRG neurons from Zfp106^−/− mice (E) are shrunken and irregular and some show signs of chromatolysis, with flattened nuclei (indicated with arrows). Scale bar is 20 μm. (F and G) Representative toluidine blue stained sections of the dorsal fasciculi from the lumbar spinal cord from WT (F) and Zfp106^−/− (G) mice showing fewer axons and axonal degeneration (arrow heads). Scale bar is 20 μm. All images are of 15-week-old WT or Zfp106^−/− mice, images representative of three animals per genotype.

Figure 5.

Neuronal survival and axonal morphology at P9. (A and B) Representative images of the ventral horn of lumbar spinal cord sections stained for Nissl from P9 WT and Zfp106^−/− mice. The number of motor neurons in the sciatic pool (inset) was counted for each genotype. Scale bars: main images 200 μm, inset 100 μm. (C) Survival of motor neurons in the sciatic pool in female mice at P9. Motor neuron survival is unaffected in P9 Zfp106^−/− mice (530 ± 15) compared with WT littermates (508 ± 23). Numbers represent the mean ± SEM. At least five animals were assessed per genotype per time point. (D and E) Representative toluidine blue stained sections of lumbar L4 and L5 ventral roots from P9 WT (D) and Zfp106^−/− (E) mice. (F) Distribution graph relating g-ratio to axon diameter of all myelinated axons analysed from ventral roots show no differences between WT and Zfp106^−/−. Scale bar is 20 μm. (G and H) Representative toluidine blue stained sections of lumbar L4 and L5 dorsal roots from P9 WT (G) and Zfp106^−/− (H) mice. Scale bar is 20 μm. (I) Distribution graph relating g-ratio to axon diameter from dorsal roots. Zfp106^−/− mice have a significant shift in the frequency distribution towards a thinner axon diameter and an increased g-ratio compared with WT (P < 0.001). (J and K) Representative toluidine blue stained sections of the sciatic nerve from P9 WT (J) and Zfp106^−/− (K) mice. Scale bar is 10 μm. (L) Distribution graph relating g-ratio to axon diameter for the sciatic nerve. Zfp106^−/− mice show a significant shift in the frequency distribution towards a smaller axon diameter compared with WT animals (P < 0.001) but no change in g-ratio.

Figure 6.

Muscle histology of WT and Zfp106^−/− male mice. (A) Muscle weights for the TA (WT, 62 mg ± 1 mg; Zfp106^−/−, 34 mg ± 1 mg), EDL (WT, 11 mg ± 0.4 mg; Zfp106^−/−, 11 mg ± 0.4 mg), gastrocnemius (WT, 155 mg ± 1 mg; Zfp106^−/−, 29 mg ± 23 mg) and soleus (WT, 11 mg ± 0.2 mg; Zfp106^−/−, 5 mg ± 0.3 mg) from male 14-week old WT and Zfp106^−/− littermates (n = 3). TA, gastrocnemius and soleus weights are significantly different from WTs (*P < 0.01). (B–I) Representative images of muscle histopathology from 14-week-old WT and Zfp106^−/− mice: (B and C) H&E staining of the gastrocnemius shows significant atrophy, reduced fibre size and centralized nuclei (yellow arrows) in Zfp106^−/− muscle compared with WT littermates; (D and E) Masson's trichome staining of gastrocnemius muscle reveals fibrosis in Zfp106^−/− muscle (E), evident with increased blue staining; (F and G) NADH-TR staining of the gastrocnemius shows greatly increased dark blue staining in Zfp106^−/− muscle (G) compared with WT littermates, indicative of a higher proportion of slower twitch, type I muscle fibres; (H and I) ATPase staining of Zfp106^−/− soleus muscle (I) showing an increase in darkly stained fibres, indicating an increase in the proportion of type I fibres overall. Scale bars: main images 20 μm, inset 10 μm. (J) Type I (WT, 253 ± 6 fibres; Zfp106^−/−, 147 ± 10 fibres), type IIa (WT, 579 ± 95 fibres; Zfp106^−/−, 141 ± 40 fibres) and total fibre number (WT, 833 ± 92 fibres; Zfp106^−/−, 261 ± 38 fibres) were counted for the soleus muscle from 14-week-old WT and Zfp106^−/− littermates using ATPase staining. The soleus from Zfp106^−/− mice has significantly fewer muscle fibres than that of WT mice (P < 0.01).

Figure 7.

Mitochondrial abnormalities in Zfp106^−/− embryonic motor neurons. (A) Resting Δψ_m from embryonic WT, Zfp106^+/− and Zfp106^−/− motor neurons, estimated by live cell imaging using TMRM. Data are normalized to resting Δψ_m for WT (100%) and represent the mean ± SEM. (B) Representative traces of Δψ_m (TMRM, arbitrary units) in response to oligomycin (2 μg/ml) (inhibitor of the F₁F_O-ATP synthase), rotenone (5 μm) (inhibitor of complex I) and the minimum TMRM fluorescence achieved after addition of FCCP 1 μm (uncoupling agent) for WT and Zfp106^−/− embryonic motor neurons. Following oligomycin addition WT motor neuron Δψ_m hyperpolarises, whilst Zfp106^−/− depolarises. (C) NADH autofluorescence monitored in WT and Zfp106^−/− motor neurons, the addition of FCCP maximizes mitochondrial respiration, thus minimizes mitochondrial NADH. NaCN was added to block mitochondrial respiration and therefore maximize mitochondrial NADH. Redox index (the initial redox level expressed as a percentage of the range) and mitochondrial NADH level (the difference in absolute values between the minimum and maximum NADH autofluorescence) are described graphically. (D) NADH redox index for WT, Zfp106^+/− and Zfp106^−/− motor neurons calculated from (C). NADH redox index is increased in Zfp106^−/− embryonic motor neurons (58% ± 4%) compared with WT littermates (39% ± 3%). (E) Mitochondrial pool of NADH represented as % of WT. NADH levels are increased in Zfp106^−/− (138% ± 9.1%) motor neurons when compared with WT. (F) FAD autofluorescence monitored in WT and Zfp106^−/− motor neurons, the addition of FCCP maximizes respiration, increasing FAD autofluorescence to maximal levels, whilst NaCN inhibits respiration, reducing FAD autofluorescence to minimal levels. (G) FAD redox index for WT, Zfp106^+/− and Zfp106^−/− motor neurons, calculated by normalizing the FCCP response to 100% and the NaCN response to 0%. FADH redox index is increased in Zfp106^−/− embryonic motor neurons (89% ± 6%) compared with WT littermates (49% ± 3%). (H) Mitochondrial pool of FAD represented as % of WT. FAD levels are decreased in Zfp106^−/− motor neurons (45.6% ± 5.2%) when compared with WT. Data from (A–H) are represented as mean ± SEM; n ≥ 34 motor neurons per genotype, per condition. *P < 0.05; **P < 0.01.