Cerebellar abnormalities in mice lacking type 3 deiodinase and partial reversal of phenotype by deletion of thyroid hormone receptor α1

Abstract

Thyroid hormone serves many functions throughout brain development, but the mechanisms that control the timing of its actions in specific brain regions are poorly understood. In the cerebellum, thyroid hormone controls formation of the transient external germinal layer, which contains proliferative granule cell precursors, subsequent granule cell migration, and cerebellar foliation. We report that the thyroid hormone-inactivating type 3 deiodinase (encoded by Dio3) is expressed in the mouse cerebellum at embryonic and neonatal stages, suggesting a need to protect cerebellar tissues from premature stimulation by thyroid hormone. Dio3(-/-) mice displayed reduced foliation, accelerated disappearance of the external germinal layer, and premature expansion of the molecular layer at juvenile ages. Furthermore, Dio3(-/-) mice exhibited locomotor behavioral abnormalities and impaired ability in descending a vertical pole. To ascertain that these phenotypes resulted from inappropriate exposure to thyroid hormone, thyroid hormone receptor α1 (TRα1) was removed from Dio3(-/-) mice, which substantially corrected the cerebellar and behavioral phenotypes. Deletion of TRα1 did not correct the previously reported small thyroid gland or deafness in Dio3(-/-) mice, indicating that Dio3 controls the activation of specific receptor isoforms in different tissues. These findings suggest that type 3 deiodinase constrains the timing of thyroid hormone action during cerebellar development.

Fig. 1.

Developmental expression of D3 and D2 in cerebellum. A, D3 activity peaked in mouse cerebellar homogenates at embryonic and early neonatal stages, whereas D2 activity peaked postnatally. B, D3 activity was undetectable in cerebellar or control eye samples from Dio3^−/− mice and was reduced by approximately 50% in both tissues in Dio3^+/− mice compared with +/+ mice at E18.5. C, In situ hybridization detected Dio3 mRNA in midcentral regions of the early cerebellar outgrowth (mc), Purkinje cell layer (PCL), and EGL (arrowhead) at E18. At P3, Dio3 mRNA signals decreased but were weakly detectable in the PCL and EGL (arrowhead). At P3, Thra mRNA signals strengthened in the inner region of the PCL. Scale bar, 0.1 mm.

Fig. 2.

Hematoxylin staining of midsagittal sections of cerebellum, revealing impaired foliation in adult Dio3^−/− mice. All foliae (numbered I–X) were present in rudimentary form in Dio3^−/− mice but were less elaborately expanded or folded than in +/+ mice. Scale bar, 1 mm.

Fig. 3.

Hematoxylin staining of midsagittal sections of cerebellum showing abnormally accelerated development in Dio3^−/−, mice and substantial reversal of this phenotype in Dio3^−/−;Thra1^−/− mice. A, Adult Dio3^−/−;Thra1^−/− mice displaying corrected foliation compared with Dio3^−/−, mice. Scale bar, 1 mm. B, At P14, Dio3^−/− mice display accelerated disappearance of the EGL compared with +/+ mice; the phenotype is largely reversed in Dio3^−/−;Thra1^−/− mice. Scale bar, 1 mm. C, Magnified view of boxed area in B. Scale bar, 0.1 mm. ML, Molecular layer. D, At P10, the EGL was prematurely thinner in Dio3^−/− mice compared with +/+ or Dio3^−/−;Thra1^−/− mice. Scale bar, 0.1 mm.

Fig. 4.

Abnormal cerebellar dimensions in Dio3^−/− mice and partial correction in Dio3^−/−;Thra1^−/− mice. A, Prematurely decreased thickness of the EGL in Dio3^−/− but not Dio3^−/−;Thra1^−/− mice at P10 and P14 (means ± sem). ***, P < 0.001 compared with +/+ mice. B, The IGL in adult Dio3^−/− mice was 41% smaller than in +/+ mice (mean size ± sem). *, P < 0.05 (P = 0.025) compared with +/+ mice. C, Cerebellar size measured on representative areas of histological sections was smaller in Dio3^−/− mice than +/+ mice, but this reduction was in proportion to the 30% lower body weight of Dio3^−/− mice at P10 (mean ± sem). ***, P < 0.001 compared with +/+ mice. P values shown were obtained using a t test or Mann-Whitney U test where appropriate.

Fig. 5.

Premature expansion of the molecular layer (ML) in Dio3^−/− mice but not in Dio3^−/−;Thra1^−/− mice. A, Calbindin staining of midsagittal sections of cerebellum of +/+, Dio3^−/−, Thra1^−/−, and Dio3^−/−;Thra1^−/− mice at P10 and at adult ages. PCL, Purkinje cell layer. B, Molecular layer thickness at P10 and adult ages and number of Purkinje cells at P10 (mean thickness ± sem). ***, P < 0.001 compared with +/+ mice. P values shown were obtained using a Student's t test or Mann-Whitney U test where appropriate. Scale bar, 0.05 mm.

Fig. 6.

Dio3^−/− mice have impaired locomotor activity, which is partially reversed in Dio3^−/−;Thra1^−/− mice. A, Time to invert and time to descend a vertical pole in male +/+, Dio3^−/−, Thra1^−/−, and Dio3^−/−;Thra1^−/− mice (mean ± sem). *, P < 0.05 (P = 0.02); **, P = 0.01; ***, P < 0.001 compared with wild-type. B, Time to invert on and time to descend a vertical pole for female +/+, Dio3^−/−, Thra1^−/−, and Dio3^−/−;Thra1^−/− mice (mean ± sem). * P < 0.05 (P = 0.02) compared with +/+ mice. C, Performance on a rotarod was not different between +/+ and Dio3^−/− mice, nor was there a difference in their learning capacity (mean ± sem). P values shown were obtained using a Student's t test or Mann-Whitney U test where appropriate.

Fig. 7.

Corrected body weight but not thyroid gland function by deletion of TRα1 in Dio3^−/−mice. A, Serum T₄ and T₃ levels were similar in Dio3^−/− and in Dio3^−/−;Thra1^−/− mice at P5 and P10 and in adults (mean ± sem). B, Hematoxylin and eosin staining of midregion thyroid sections demonstrating smaller thyroid glands with fewer colloid-filled follicles in Dio3^−/− and Dio3^−/−;Thra1^−/− mice than in +/+ or Thra1^−/− mice. C, Body weight in Dio3^−/−;Thra1^−/− mice was substantially recovered compared with Dio3^−/− mice (mean ± sem). ***, P < 0.001 compared with +/+ mice. D, Decreased thyroid size in Dio3^−/− mice was not proportionate to the decreased weight (mean ± sem). *, P < 0.05 (P = 0.02 for both) compared with +/+ mice. P values shown were obtained using a Student's t test or Mann-Whitney U test where appropriate. Scale bar, 0.1 mm.

Fig. 8.

Nonreversal of auditory defect in Dio3^−/− mice by deletion of TRα1. A, ABR thresholds for a click stimulus or pure tone stimuli of 8, 16, and 32 kHz were elevated in both Dio3^−/− and Dio3^−/−;Thra1^−/− genotypes. The magnitude and statistical significance of defects in Dio3^−/− mice compared with +/+ mice were consistent with previously reported data (means ± sem). *, P < 0.05 for Dio3^−/− and Dio3^−/−;Thra1^−/− groups, individually, compared with +/+ mice; **, P < 0.01 for Dio3^−/− and Dio3^−/−;Thra1^−/− groups, individually, compared with +/+ mice. B, Representative ABR waveforms for +/+, Dio3^−/−, and Dio3^−/−;Thra1^−/− mice with thresholds of 50, 75, and 85 dB sound pressure level, respectively, for a click stimulus. The normalized voltage scale is different for each genotype. P values were obtained using a Student's t test or Mann-Whitney U test where appropriate.