A lack of thyroid hormones rather than excess thyrotropin causes abnormal skeletal development in hypothyroidism

Abstract

By proposing TSH as a key negative regulator of bone turnover, recent studies in TSH receptor (TSHR) null mice challenged the established view that skeletal responses to disruption of the hypothalamic-pituitary-thyroid axis result from altered thyroid hormone (T(3)) action in bone. Importantly, this hypothesis does not explain the increased risk of osteoporosis in Graves' disease patients, in which circulating TSHR-stimulating antibodies are pathognomonic. To determine the relative importance of T(3) and TSH in bone, we compared the skeletal phenotypes of two mouse models of congenital hypothyroidism in which the normal reciprocal relationship between thyroid hormones and TSH was intact or disrupted. Pax8 null (Pax8(-/-)) mice have a 1900-fold increase in TSH and a normal TSHR, whereas hyt/hyt mice have a 2300-fold elevation of TSH but a nonfunctional TSHR. We reasoned these mice must display opposing skeletal phenotypes if TSH has a major role in bone, whereas they would be similar if thyroid hormone actions predominate. Pax8(-/-) and hyt/hyt mice both displayed delayed ossification, reduced cortical bone, a trabecular bone remodeling defect, and reduced bone mineralization, thus indicating that the skeletal abnormalities of congenital hypothyroidism are independent of TSH. Treatment of primary osteoblasts and osteoclasts with TSH or a TSHR-stimulating antibody failed to induce a cAMP response. Furthermore, TSH did not affect the differentiation or function of osteoblasts or osteoclasts in vitro. These data indicate the hypothalamic-pituitary-thyroid axis regulates skeletal development via the actions of T(3).

Figure 1

Growth and Skeletal Development of Pax8^−/− and Hyt/Hyt Mice A–C, Ulnas from P14 Pax8^−/− (A), P70 Pax8^−/− and T_3/4-treated (between P8 and P13) Pax8^−/− (B), and P49 hyt/hyt, HET, and WT littermates (C) stained with alcian blue and alizarin red. Arrows indicate delayed formation of secondary epiphyses and wider growth plates in Pax8^−/− and hyt/hyt mice. D, Mean lengths of tibia and ulna. *, P < 0.05; **, P < 0.01 mutant vs. WT; n =3–6, 2-tailed Student’s t test or ANOVA and Tukey’s multiple comparison post hoc test.

Figure 2

Histological Analysis of Ossification in Pax8^−/− and Hyt/Hyt Mice A and B, Proximal tibia sections stained with alcian blue/van Gieson from P70 Pax8^−/− and T_3/4-treated (between P8 and P13) Pax8^−/− mice (A) and P49 hyt/hyt, HET, and WT littermates (B). C, Widths of growth plate zones (RZ, PZ, and HZ) and total growth plate (RZ+PZ+HZ) from P70 WT, T_3/4-treated Pax8^−/−, and Pax8^−/− or P49 WT, HET, and hyt/hyt mice. D and E, Mid-diaphysis femurs are shown in P70 WT, T_3/4-treated Pax8^−/− and Pax8^−/− mice (D) or P49 WT, HET, and hyt/hyt mice (E). Graphs show cortical thickness, length/cortical thickness, and endosteal/periosteal diameter ratios in P70 WT, T_3/4-treated Pax8^−/−, and Pax8^−/− (D) and P49 WT, HET, and hyt/hyt mice (E). E, Epiphysis; GP, growth plate; IE, immature epiphysis. *, P < 0.05; **, P < 0.01; ***, P < 0.001 mutant vs. WT; n =3–7; ANOVA and Tukey’s post hoc test. Scale bars, 200 μm.

Figure 3

Expression of Fgfr1 in Osteoblasts in Pax8^−/−, T_3/4-Treated Pax8^−/−, and Hyt/Hyt Mice In situ hybridization of Fgfr1 expression in osteoblasts from P70 Pax8^−/− and T_3/4-treated Pax8^−/− mice (A) and P49 hyt/hyt, HET, and WT littermates (B). Arrows indicate layer of cortical bone lining osteoblasts. CB, Cortical bone. Scale bar, 50 μm.

Figure 4

Topographic Three-Dimensional BSE-SEM Analysis of Bone Structure in Pax8^−/− and Hyt/Hyt Mice BSE-SEM views of proximal humerus (A and B) from P70 WT, T_3/4-treated (between P8 and P13) Pax8^−/−, and Pax8^−/− mice and proximal tibia (C and D) from P49 WT, HET, and hyt/hyt mice. Scale bars, 200 μm.

Figure 5

qBSE-SEM Compositional Analysis of Mineralization Densities and BVF in Pax8^−/− and Hyt/Hyt Mice Mineralization densities in proximal humerus (A) from P70 WT, T_3/4-treated (between P8 and P13) Pax8^−/−, and Pax8^−/− mice and caudal vertebrae (D) from P49 WT, HET, and hyt/hyt mice. Gray-scale images were pseudocolored with low mineralization density in blue and high density in gray. Arrows indicate highly mineralized calcified cartilage. Scale bars, 200 μm. B and E, Relative and cumulative frequencies of mineralization densities. **, P < 0.01; ***, P < 0.001 mutant vs. WT; n = 2–4; Kolmogorov-Smirnov test. C and F, BMD, determined by dual-energy x-ray absorptiometry, and BVF, determined by BSE-SEM, in P70 WT, T_3/4-treated Pax8^−/−, and Pax8^−/− mice (C) and P49 WT, HET, and hyt/hyt mice (F). #, P < 0.05 for BMD or BVF in mutant vs. WT; n = 2–5 per group; ANOVA followed by Tukey’s multiple comparison post hoc test.

Figure 6

Primary Calvarial Osteoblast Cultures A, Southern blot RT-PCR of TSHα, TSHβ, and TSHR mRNAs on d 7, 14, 21, and 28 of culture with or without bTSH (10 U/liter) and expression of Runx2, Osterix, Twist2, Osteocalcin, and ColIa1 mRNAs on d 21. B, cAMP response (±sem) to 1) medium alone, 2) forskolin (10 μmol), 3) bTSH (10 U/liter), 4) bTSH (100 U/liter), 5) hTSH (10 U/liter), 6) hTSH (100 U/liter), 7) 3G4 nonstimulating TSHR antibody, and 8) M22 TSHR-stimulating antibody. C, Alkaline phosphatase activity (±sem). D, Representative Northern blot analysis of osteocalcin (OC) mRNA expression on d 7, 14, 21, and 28 of culture with or without bTSH (10 U/liter) and graph showing mean osteocalcin mRNA levels from three independent experiments normalized to expression of 18S rRNA. E, Alizarin red staining.

Figure 7

Osteoclast Cultures A, Southern blot RT-PCR of TSHα, TSHβ, and TSHR mRNAs on d 6, 9, and 12 of culture with or without bTSH (10 U/liter) and expression of Trap, Calcr, Ctsk, and Rank mRNAs on d 9. B, cAMP response (±sem) to 1) medium alone, 2) forskolin (10 μmol), 3) bTSH (10 U/liter), 4) bTSH (100 U/liter), 5) hTSH (10 U/liter), 6) hTSH (100 U/liter), 7) 3G4 nonstimulating TSHR antibody, and 8) M22 TSHR-stimulating antibody in osteoclasts and thyroid follicular FRTL-5 cells. C, Cell number (±sem), morphology, and TRAP activity (±sem) on d 9. D, Toluidine blue- and TRAP-stained osteoclasts cultured on dentine slices for 9 d. Graph shows the fraction of resorbed surface (±sem) determined by BSE-SEM. E, Resorption pits formed by bone marrow osteoclasts cultured for 12 h. Scale bars, 200 μm.

Figure 8

TSHR Protein Expression Western blot of TSHR expression in lysates from thyroid follicular FRTL-5 cells and primary osteoblasts (OB) and osteoclasts (OC).