Tmem263 deletion disrupts the GH/IGF-1 axis and causes dwarfism and impairs skeletal acquisition

Abstract

Genome-wide association studies (GWAS) have identified a large number of candidate genes believed to affect longitudinal bone growth and bone mass. One of these candidate genes, TMEM263, encodes a poorly characterized plasma membrane protein. Single nucleotide polymorphisms in TMEM263 are associated with bone mineral density in humans and mutations are associated with dwarfism in chicken and severe skeletal dysplasia in at least one human fetus. Whether this genotype-phenotype relationship is causal, however, remains unclear. Here, we determine whether and how TMEM263 is required for postnatal growth. Deletion of the Tmem263 gene in mice causes severe postnatal growth failure, proportional dwarfism, and impaired skeletal acquisition. Mice lacking Tmem263 show no differences in body weight within the first two weeks of postnatal life. However, by P21 there is a dramatic growth deficit due to a disrupted GH/IGF-1 axis, which is critical for longitudinal bone growth. Tmem263-null mice have low circulating IGF-1 levels and pronounced reductions in bone mass and growth plate length. The low serum IGF-1 in Tmem263-null mice is associated with reduced hepatic GH receptor (GHR) expression and GH-induced JAK2/STAT5 signaling. A deficit in GH signaling dramatically alters GH-regulated genes and feminizes the liver transcriptome of Tmem263-null male mice, with their expression profile resembling a wild-type female, hypophysectomized male, and Stat5b-null male mice. Collectively, our data validates the causal role for Tmem263 in regulating postnatal growth and raises the possibility that rare mutations or variants of TMEM263 may potentially cause GH insensitivity and impair linear growth.

Keywords: GHR; IGF-1; Laron dwarfism; growth hormone; growth hormone insensitivity; skeletal dysplasia.

Figure 1.. TMEM263 is a novel and highly conserved plasma membrane protein.

(A) Sequence alignment of full-length human (NP_689474), mouse (NP_001013046), chicken (NP_001006244), xenopus frog (NP_989399), and zebrafish (NP_998306) transmembrane protein 263 (TMEM263) using Clustal-omega (82). Identical amino acids are shaded black and similar amino acids are shaded grey. Gaps are indicated by dash lines. The two predicted transmembrane domains are indicated in red. (B) TMEM263 expression across normal human tissues based on the consensus Human Protein Atlas (HPA) and Gene-Tissue Expression (GTeX) datasets. The data can be accessed via the HPA database (www.proteinatlas.org). nTPM denotes normalized protein-coding transcripts per million and it corresponds to the mean values of the different individual samples from each tissue. Bars are color-coded based on tissue groups with functional features in common. (C) Tmem263 expression across mouse tissues (n = 11). Relative expression across tissues were first normalized to β-actin, then normalize to the tissue (pancreas) with the lowest expression. (C) Immunoblot analysis of cell lysate from HEK293 cells transfected with a control pCDNA3 empty plasmid or plasmid encoding human TMEM263 tagged with a C-terminal Myc-DDK epitope. Immunoblots were probed with an anti-FLAG (DDK) antibody (left panel) or an anti-TMEM263 antibody (right panel). (D) TMEM263 is localized to the plasma membrane. Surface biotinylation was carried out on transfected HEK293 cells. Biotinylated plasma membrane proteins were captured with Avidin-agarose beads, eluted, and immunoblotted for TMEM263 with an anti-FLAG antibody.

Figure 2.. Deletion of Tmem263 gene causes dwarfism.

(A) Generation of Tmem263 knockout (KO) mice. The exon 3 that encodes ~81% of the full-length protein was deleted using CRISPR/Cas9 method and confirmed with DNA sequencing. The location and sequence of the two guide RNAs (gRNA) used to generate the deletion were underlined. Filled-in black boxes indicate part of the exon that codes for Tmem263 protein, and white boxes indicate part of the exon that codes for 5’ and 3’ UTR of the transcript. (B) Wild-type (WT) and KO alleles were confirmed by PCR genotyping. (C) The complete loss of Tmem263 transcript in KO mice was confirmed by qPCR in male and female mouse liver and hypothalamus (WT, n = 6-8; KO, n = 6-8). (D) The expected Mendelian versus observed genotype distributions in P1 pups (n = 82). (E) Representative images of WT and Tmem263-KO pups at postnatal day 1 (P1). Milk spots are indicated by a red arrow. (F) Representative Alcian blue and Alizarin red staining of axial skeletal and cartilage in WT and KO P1 pups. (G) Body weights of WT and Tmem263-KO pups at P1 (WT = 17; Het = 42; KO = 23), P7 (WT = 15; het = 28; KO = 5), P14 (WT = 16; het = 34; KO = 4), and P21 (WT = 33; het = 41; KO = 13). For panel G, we combined the data of male and female pups from P1 to P21. (H) Representative images of adult WT and Tmem263-KO mice at 9 weeks of age. (I-J) Body weights and body length of WT (+/+), heterozygous (+/−), and KO (−/−) male and female mice at 9 weeks of age. Sample size for males (WT = 45; het = 73; KO = 9) and females (WT = 30; het = 59; KO = 8). (K) The growth curve trajectory based on the combined data in G and I. All data are presented as mean ± S.E.M. **** P < 0.0001 (One-way ANOVA with Tukey’s multiple comparisons test).

Figure 3.. Mice lacking TMEM263 exhibit pronounced skeletal dysplasia.

(A) Representative microCT images of bone (femur) showing a dramatic reduction in size in Tmem263 KO (−/−) mice relative to WT (+/+) and heterozygous (+/−) controls at 8 weeks of age. (B) Femur length of WT (+/+), heterozygous (+/−), and KO (−/−) male and female mice at 8 weeks of age. (C) Quantification of trabecular bone volume per tissue volume (BV/TV) in the distal femur of WT (+/+), heterozygous (+/−), and KO (−/−) male and female mice. (D) Quantification of trabecular number (Tb. N) in the distal femur. (E) Trabecular bone thickness (Tb. Th). (F) Cortical tissue area (Tt. Ar). (G) Cortical area per tissue area. (H) Cortical thickness. (I) Tissue area per femur length. (J) Cortical thickness per femur length in male and female mice. (K) Representative images of tibial growth plate histology in WT and Tmem263-KO male mice. (L-N) Quantification of growth plate length (L), proliferative zone length (M), and hypertrophic zone length (N) in WT (n = 10) and KO (n = 10) male mice. All data were collected on 8-week-old mice. Sample size for panel B-J: males (WT = 6; het = 6; KO = 11) and females (WT = 6; het = 6; KO = 9). All data are mean ± S.E. ** P < 0.01; *** P < 0.001; **** P < 0.0001 (One-way ANOVA with Tukey’s multiple comparisons test).

Figure 4.. TMEM263 deficiency results in marked reduction in circulating IGF-1, IGFBP3, and IGFALS levels.

Serum levels of growth hormone (GH; A), IGF-1 (B), IGFBP3 (C), IGFALS (D), insulin (E), glucose (F), calcium (G), and phosphate (H) in WT (+/+), heterozygous (+/−) and Tmem263-KO (−/−) male and female mice at 8 weeks old. Sample size for panel A (GH): males (WT = 9; het = 9; KO = 5) and females (WT = 10; het = 7; KO = 8). Panel B (IGF-1): males (WT = 9; het = 9; KO = 9) and females (WT = 10; het = 7; KO = 7). Panel C (IGFBP3): males (WT = 15; het = 16; KO = 11) and females (WT = 12; het = 18; KO = 10). Panel D (IGFALS): males (WT = 10; het = 10; KO = 8) and females (WT = 10; het = 10; KO = 8). Panel E (insulin): males (WT = 12; het = 17; KO = 8) and females (WT = 10; het = 16; KO = 8). Panel F (glucose): males (WT = 9; het = 14; KO = 9) and females (WT = 6; het = 15; KO = 8). Panel G and H (calcium and phosphate): males (WT = 10; het = 10; KO = 8) and females (WT = 10; het = 10; KO = 8). (I) Ratio of calcium-to-phosphate in WT, heterozygous, and KO male and female mice. Sample size for males (WT = 10; het = 10; KO = 8) and females (WT = 10; het = 10; KO = 8). All data are presented as mean ± S.E.M. * P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001 (One-way ANOVA with Tukey’s multiple comparisons test).

Figure 5.. Reduced hepatic growth hormone receptor (GH-R) protein level and signaling in TMEM263 KO mice.

(A) The GH/IGF-1 axis required for postnatal skeletal growth. At the onset of growth spurt, growth hormone releasing hormone (GHRH) from the hypothalamus causes the release of growth hormone (GH) from the anterior pituitary. Circulating GH binds to its receptor (GHR) in liver and other peripheral tissues to induce the synthesis and secretion of IGF-1, which then acts in an endocrine, paracrine, and/or autocrine manner to induce skeletal growth. (B) Expression levels of Ghr (growth hormone receptor), Igf-1, Igfals (IGF binding protein acid labile subunit), and Igfbp3 (IGF binding protein 3) transcripts in the liver of WT and KO mice. Sample size of male mice (WT, n = 8; KO, n = 8) and Female mice (WT, n = 8; KO, n = 8). (C) Immunoblot analysis of growth hormone receptor (GH-R) protein levels in the liver of WT (n = 7) and KO (n = 7) mice. Molecular weight markers are indicated on the left. (D) Quantification of the immunoblot results as shown in C (n = 7 per genotype). (E-F) Reduced hepatic GH-induced signaling in KO (−/−; n = 5) mice relative to WT (+/+; n =4) controls. Immunoblot analysis of phospho-JAK2 (Tyr1008), total JAK2, phospho-STAT5 (Y694), and total STAT5 in liver lysates from control male mice not injected with GH (E) and male mice injected with recombinant GH (F). Molecular weight markers are indicated on the left of the gel. (G) Quantification of the immunoblot results as shown in F (WT, n = 4; KO, n =5). All data are presented as mean ± S.E.M. ** P < 0.01; *** P < 0.001; **** P < 0.0001 (One-way ANOVA with Tukey’s multiple comparisons test for data in B and two-tailed student’s t-test for data in D and F).

Figure 6.. Loss of TMEM263 disrupts GH-regulated gene expression in the male mouse liver.

(A) Volcano plot of male mouse liver transcriptome. (B) Enrichment analysis of differentially expressed genes (DEGs) by Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Reactome (http://www.reactome.org) databases. Plots of -Log(adj. p) vs gene ratio to show complete spread of all enrichment results for each analysis. The most effected categories are highlighted and labeled with the category name and number of up- and down-regulated genes within each. (C) Volcano plot showing the expression of all transcription factor (TF) genes detected in the male mouse liver transcriptome. TFs that are significantly up- or down-regulated in KO male mouse liver are highlighted. (D) Heat map of DEGs involved in growth and metabolism. (E) Heat map of all protein-coding DEGs from the cytochrome P450 (Cyp) gene family. (F) Heat map of all protein-coding DEGs from the Major urinary protein (Mup) gene family. (G-I) Tmem263-KO male liver DEG overlap comparison to three separate public data sets of mouse liver gene expression: (G) WT male vs. WT female mice (45), (H) Hypophysectomized vs. sham control male mice (48), and (I) Stat5b-KO vs WT male mice (35). (J) Summary of key findings underpinning the dwarfism and skeletal dysplasia phenotypes of Tmem263-null mice. All heat map data is shown on a column z-score scale. Only significantly different genes (adjusted p-value < 0.05 and Log₂(FC) <−1 or >1) are shown for all heat maps. PCG = protein coding gene, NPCG = non-protein coding gene. Sample size for male WT (n = 8) and KO (n = 8) mice were 9 weeks old for all data shown.