Zebrafish Tric-b is required for skeletal development and bone cells differentiation

Introduction Trimeric intracellular potassium channels TRIC-A and -B are endoplasmic reticulum (ER) integral membrane proteins, involved in the regulation of calcium release mediated by ryanodine (RyRs) and inositol 1,4,5-trisphosphate (IP3Rs) receptors, respectively. While TRIC-A is mainly expressed in excitable cells, TRIC-B is ubiquitously distributed at moderate level. TRIC-B deficiency causes a dysregulation of calcium flux from the ER, which impacts on multiple collagen specific chaperones and modifying enzymatic activity, leading to a rare form of osteogenesis imperfecta (OI Type XIV). The relevance of TRIC-B on cell homeostasis and the molecular mechanism behind the disease are still unknown. Results In this study, we exploited zebrafish to elucidate the role of TRIC-B in skeletal tissue. We demonstrated, for the first time, that tmem38a and tmem38b genes encoding Tric-a and -b, respectively are expressed at early developmental stages in zebrafish, but only the latter has a maternal expression. Two zebrafish mutants for tmem38b were generated by CRISPR/Cas9, one carrying an out of frame mutation introducing a premature stop codon (tmem38b-/- ) and one with an in frame deletion that removes the highly conserved KEV domain (tmem38bΔ120-7/Δ120-7 ). In both models collagen type I is under-modified and partially intracellularly retained in the endoplasmic reticulum, as described in individuals affected by OI type XIV. Tmem38b-/- showed a mild skeletal phenotype at the late larval and juvenile stages of development whereas tmem38bΔ120-7/Δ120-7 bone outcome was limited to a reduced vertebral length at 21 dpf. A caudal fin regeneration study pointed towards impaired activity of osteoblasts and osteoclasts associated with mineralization impairment. Discussion Our data support the requirement of Tric-b during early development and for bone cell differentiation.


1
Supplementary Materials and Methods

Generation of zebrafish tmem38b knock out mutants using CRISPR/Cas9
A single guide RNA (gRNA) for the tmem38b gene (ENSDARG00000100549, Ensemble 94) was designed using the online freely available software CHOPCHOP (https://chopchop.rc.fas.harvard.edu). The target sequence was selected at the 5' end, in exon 7 (5'-GGTTCTCGTCACTTCCTTCATGG -3', 11391-11413 nt). The synthesis of target oligonucleotides (Eurofins Genomics), and the preparation of gRNA and Cas9 mRNA were carried out as described in (1). The genomic region surrounding the target sequence was PCR amplified using the following primers: forward 5'-TTACTGTCCGCTGGATGTGG-3' (11326-11345 nt) and reverse 5'-CAGAGCGTCGCTGTATTTGC-3' (11448-11467 nt), with annealing temperature of 56°C, generating a 142 bp amplicon. T7 endonuclease assay was performed to evaluate targeting efficiency as described in (1). The mutations in the heterozygous F1 zebrafish were determined by DNA extraction from tail clip of adult zebrafish followed by Sanger sequencing of the region surrounding the targeting.
To genotype tmem38b -/and tmem38b Δ120-7/120-7 mutants, amplicons were checked on 12% and 10% v/v electrophoresis acrylamide gel in TBE buffer (Tris HCl 0.1M, H3BO3 0.1M, EDTA 2mM, pH 8.2), respectively. The embryos generated from pairwise breeding were grown to 3 weeks post fertilization (wpf) and the genotyping data were used for Mendelian analysis of surviving homozygous knock out compared to WT and heterozygous zebrafish. Under the null hypothesis of no viability selection, progeny genotypes should conform to an expected Mendelian ratio of 1:2:1. Deviations from expected number of homozygous knock out were tested with goodness-of-fit Chi-square statistical analysis.

Bone Formation rate
Bone formation rate was performed on WT and tmem38b -/-. Briefly, WT (n=9) and tmem38b -/-(n= 10) were stained with 0.01% Alizarin Red S (Sigma-Aldrich) in KOH, pH 7.4, from 6 to 10 dpf for 15 min every day, according to Bensimon-Brito et al 2016 (4). The fish were then stained in 0.2% calcein (Sigma-Aldrich) in 0.9% NaCl at 1 mpf for 10 minutes and then washed over night. Images were acquired using a Leica M165 FC microscope connected to Leica DFC425 C digital camera. Bone formation rate was measured as the ratio between red operculum area on green operculum area, normalized to the standard length.

uCT
WT (n = 4) and tmem38b -/-(n = 4) zebrafish (9 mpf) were analyzed by µCT. Zebrafish were sacrificed, fixed overnight at 4 °C in 4% (w/v) PFA. Zebrafish were kept hydrated in parafilm and placed in a sample holder during µCT acquisitions (Skyscan 1272). For high-resolution scans and quantitative analysis of the 2 nd and 3 rd precaudal vertebrae, zebrafish were scanned at 40 kV and 230 mA with a voxel size of 3 µm. For all samples, ring artifact and beam hardening correction was kept constant and no smoothing was applied during reconstruction (NRecon, Bruker). After applying a constant global threshold to all samples, the morphological properties: vertebral body length (VBL, µm), bone volume (BV, mm 3 ), polar moment of inertia (MMIp, surrogate measure for resistance against torsion, mm 4 ), eccentricity, centrum thickness (C.Th, mm) and bone perimeter (B.Pm, mm) were analyzed according to previously established protocols (5, 6).

Nanoindentation
Elastic modulus E and hardness H of vertebral bone were determined using nanoindentation (iMicro, Nanomechanics, Inc., Oak Ridge, USA) based on previously established protocols (5). Whole fish samples (WT: n = 4; tmem38b -/and tmem38b Δ120-7/120-7 mutants n = 6, 2 mpf) were fixed in 4% PFA, embedded in polymethylmethacrylate, and ground plane parallel until the central sagittal plane was reached. To eliminate surface roughness, blocks were then polished with a 3 mm diamond suspension, followed by a 1 mm diamond suspension and final polishing with a 0.05 mm aluminum-oxide suspension. Finally, samples were ultrasonically cleaned in deionized water to remove surface debris. Using a Berkovich diamond tip and the depth-sensing continuous stiffness mode with a final depth of 1000 nm, indents were placed on the longitudinal plane of the vertebral bone cortex of 4 vertebrae per sample. Applying a Poisson's ratio of 0.3, E and H were extracted according to the Oliver-Pharr method (7) using in-house software (inView, Nanomechanics, Inc., Oak Ridge, USA).

Generation of tmem38b zebrafish models
A specific RNA guide (gRNA) targeting exon 7 of tmem38b was microinjected in 1-2 cell stage zebrafish fertilized embryos together with in vitro transcribed Cas9 mRNA. F0 mosaic zebrafish were screened for specific targeting at 1 day post fertilization (dpf) by T7 endonuclease I (T7EI) assay and by Sanger sequencing. The mutagenesis rate was 63%. To identify the F0 germ line zebrafish, mosaic fish were further outcrossed to AB WT. F1 progeny was initially screened by T7 endonuclease assay to discriminate the WT from the heterozygous mutant animals and the mutants finally confirmed by Sanger sequencing. The mutant F1 zebrafish carrying the c.524_530delTGAAGGA, predicted to insert a premature stop codon at amino acid 122 of Tric-b was chosen to obtain the F2 tmem38b knock out model (tmem38b -/-). The mutant F1 zebrafish carrying the c.517_540del24nt, predicted to introduce the in frame p.Ala120_Thr127 deletion was selected to generate the F2 tmem38b Δ120-7/Δ120-7 . To genotype both tmem38b -/and tmem38b Δ120-7/Δ120-7 different size of PCR products were evaluated; the expected amplicons were 142 bp for the WT and 135 bp and 118 bp for tmem38b -/and tmem38b Δ120-7/Δ120-7 , respectively (Supplementary Figure 1B-D).

Supplementary Figure 1. (A)
In situ hybridization performed on 48 and 72 hpf WT embryos using the sense oligonucleotide as negative control. Scale bar: 500 µm. Generation of tmem38b zebrafish models using CRISPR/Cas9: (B) F0 mosaic fish screening using T7 endonuclease. Amplicons were digested by T7 endonuclease in presence of heterozygosity. MW: molecular weight, nd: not digested.

Supplementary Tables
Supplementary Table 1