Mutant dlx3b disturbs normal tooth mineralization and bone formation in zebrafish

Animals and ethics statement

All zebrafish studies were conducted under the guidance and approval of the Institutional Animal Care and Use Committee at Peking University (LSC-ZhangB-1). Zebrafish embryos were incubated at 37 °C until 6 days post-fertilization (dpf). The water was changed daily. Larvae were fed with paramecium from 6–10 dpf and commercial baby fish food from day 10 onward. The adult fishes were maintained at 28.5 °C with a 14-h light/10-h dark cycle in a circulating system that continuously filtered, UV treated, and aerated the circulating water.

Generation of the dlx3b^−∕− zebrafish line

TDO related DLX3 mutations are found to affect residues within, or adjacent to the homeodomain. The TALEN targeting site in the zebrafish dlx3b gene was designed at the 3-terminal end of the homeodomain, using TALE-NT (https://boglab.plp.iastate.edu/) and ZiFiT (http://zifit.partners.org/ZiFiT/). Then, the TALEN constructs were assembled using the unit assembly method (Huang et al., 2011). The TALEN mRNAs were synthesized using the SP6/T7 MESSAGE mMACHINE Kit, purified by LiCl precipitation, and injected into zebrafish embryos at the one-cell stage. The embryos were raised to adulthood and then outcrossed with wild-type fish to select founders carrying mutations. Mutations were further confirmed via HindIII digestion of PCR products and DNA sequencing. Two founders containing the targeted 4 bp deletion mutation were selected and outcrossed with wild type (WT) fish. The F1 embryos from the 2 selected founders were raised. Heterozygous F1 mutants were identified. Then, the F1 mutants were incrossed to breed enough F2 fish for subsequent experiments, including WT (dlx3b^+∕+), heterozygous mutants (dlx3b^+∕−), and homozygous mutants (dlx3b^−∕−). The WT (dlx3b^+∕+) were used as the control group. The number of specimens used in different experiments has been indicated in respective parts. All specimens used were identified using HindIII digestion of PCR products and DNA sequencing. The PCR primers used were: forward primer 5′-TTAGTGAGGCATTGTCGGCT-3′ and reverse primer 5′-CGCGGAGTATCTGTCAGCTT-3′. Zebrafish embryos at 6 dpf were allocated to three groups for further growth in separate dishes, starting with 100 embryos in each group. The dead fish were removed when replacing water. The number of surviving larval fish was counted and calculated from 6 to 10 dpf using Image J, and each experiment was repeated for three times.

Scanning electron microscopy

Adult F2 zebrafish were collected when they were a standard length of 25 mm, approximately coinciding with 90 dpf. The pharyngeal jaws were carefully dissected under a stereomicroscope, immersed in 0.1% KOH for 48 h to remove the soft tissues, rinsed overnight with sterile water, and then freeze-dried for 4 h. The samples were fixed to a specific platform using conductive adhesive and coated with a palladium-gold film before scanning. The surface morphology of each sample was recorded using a scanning electron microscope (S4800, n = 10 for each group). Energy dispersive spectrometer analysis was then performed on a certain area of tooth enameloid using EMAX350 (n = 5 for each group).

Micro-computed tomography analysis

To examine the three-dimensional structures of mineralized hard tissues, the adult zebrafishes were fixed in 4% paraformaldehyde for 48 h. The zebrafish samples were scanned using micro-CT (Inveon MM CT, SIEMENS, Munich, Germany) and then 3-D reconstructed using the NRecon (NRecon version 1.6.9.4, Skyscan, KONTICH, Belgium) software (n = 3 for each group). Eight-bit grayscale images, reconstructed using NRecon, were transferred to SkyScan^® CT analyser Version 1.13.5.1 (Kontich, Belgium). Three regions of interest (ROIs) were selected, and then the values for bone density from wide-type and homozygous mutants were calculated and analyzed.

Real-time RT-PCR

Eight larvae at 96 h post-fertilization (hpf) were pooled for total RNA isolation of each group. Total RNA was isolated with a TRIzol reagent (Invitrogen, Waltham, CA, USA) according to the manufacturer’s instruction. The cDNA synthesis was carried out in a 10-µl reaction mixture containing 2 µg total RNA, 400 mM reverse transcription primers, 4 U/µl M-MLV, 1 U/µl RNAsin, and 0.4 mM dNTP mix, using M-MLV reverse transcriptase (Promega, Madison, WI, USA). The amplification reaction was carried out in an ABI 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with SYBR Green Master Mix (Roche, Basel, Switzerland). The reaction mixture was incubated for 1 min at 95 °C. The amplification program consisted of 40 cycles of denaturation at 95 °C for 15 s and annealing and extension at 60 °C for 45 s. Oligonucleotide primer sequences and annealing temperatures are displayed in Table 1. Transcription levels were normalized against β-actin, and each value was the average of three independent extractions. Student’s t-test was employed to determine significant changes at 95% confidence level (P <0.05).

Alizarine Red S staining

Adult zebrafish with a standard length of 25 mm (coinciding with approximately 90 dpf) were collected. All fish samples were devitalized, fixed in 4% paraformaldehyde for 24 h, immersed in 0.1% KOH for 48 h, and then macerated with 1%–2% KOH for 8 h to thoroughly remove the residual soft tissue. The samples were then stained with 0.1% Alizarine Red S (Sigma, USA) in 0.5% KOH for 14–20 h, followed by rinsing with sterile water for 12 h; the water was replaced at intervals of one hour. The fifth ceratobranchial arch, which is also called the pharyngeal bone, was carefully dissected and photographed under a stereomicroscope (n = 8 for each group).

Statistical analysis

All data was presented as means ± standard deviation (SD). Comparisons between groups were conducted using Student’s t-test performed on Graphpad Prism 8 statistical software Graphpad Prism 8.0.2 (USA). Statistically significant differences were noted when P <0.05.

Generation of dlx3b mutant zebrafish lines

The TALEN targeting site in the zebrafish dlx3b gene model was designed at the 3′-terminal end of the homeodomain (Fig. 1A). By using the TALEN-mediated genome editing method, the predicted Hind III restriction enzyme digestion site (AAGCTT) in the dlx3b gene was disrupted, resulting in a four-bp deletion mutation from 544 bp to 547 bp (Fig. 1B). This dlx3b mutation introduced a frameshift and premature stop codon, shortening the protein sequence to 189 aa in total. Notably, in the mutant dlx3b protein, the 182nd lysine (K) was altered to phenylalanine (F) and followed by 7 novel amino acids (Fig. 1C). During fish development, the tails of 1-month-old zebrafish were collected and pooled after genotyping. Under Hind III restriction endonuclease digestion, the intact 513 bp PCR product of dlx3b^−∕− fish tail could not be cut into 286 bp and 226 bp fragments, which was consistent with the four-bp deletion mutation in the dlx3b gene (Fig. 2A). The PCR product was further confirmed the exact mutant sequence in the dlx3b^−∕− zebrafish model (Fig. 2B). During zebrafish breeding, we found that the survival rate of dlx3b^−∕− larvae decreased dramatically from 6 to 10 dpf, as compared with that of wild type (dlx3b^+∕+) and heterozygous (dlx3b^+∕−) siblings (Fig. 2C). Still, there were enough mutant fish available for subsequent investigations.

Curved spine and hypoplasia of vertebrates were observed in dlx3b^−∕− zebrafish

Starting in the early larval stages, a moderate curvature of the spine bone toward the dorsal side was found in dlx3b^−∕− fish, manifesting in 86 out of 100 fish in the dlx3b^−∕−group as compared to 3 out of 99 in the dlx3b^+∕+ group (Figs. 3A & 3B). At the adult stage (3–6 months), three of the thirty homozygotes exhibited prominent abnormal curvatures in the spine (Fig. 3E), while the wild-type siblings appeared normal (Fig. 3D). The micro-CT revealed that the abdominal and caudal spine bones of dlx3b^−∕− adult fish exhibited marked curvatures on both the dorsal–ventral (Fig. 3G&I) and medial–lateral (Figs. 3F & 3H) planes. A radiopaque area was shown in the caudal spine region of mutant fish under high magnification (Figs. 3J & 3K).

Abnormal morphology of tooth enameloid with reduction of mineral content in dlx3b^−∕−zebrafish

Unlike most mammalian teeth, which are supported by alveolar bones, the pharyngeal teeth of zebrafish directly attach to the pharyngeal bone without root. Alizarine Red S staining revealed that six of eight dlx3b^−∕− zebrafish mutants exhibited a bone-tooth joint deformity in the pharyngeal teeth, especially the ventral teeth (Figs. 4E–4H). None of the examined wild-type siblings exhibited these changes in the pharyngeal bone and dentition (Figs. 4A–4D). Enameloid was visible with the Alizarine Red S staining, but it did not display any visible abnormality at this level.

To elucidate the effects of mutant dlx3b on enameloid, we observed the dissected pharyngeal teeth with scanning electron microscopy (SEM). The fourth pharyngeal teeth on the ventral side were accurately identified (Figs. 5A–5I). Under high magnification, eighty percent of dlx3b^−∕− mutants showed pitting in the enameloid surface of the ventral side compared with the smooth enameloid surface of the wild-type and heterozygous siblings (Fig. 5I). Energy dispersive spectrometry (EDS) was performed on the pitted enameloid regions of dlx3b^−∕−fish to analyze the mineral contents (Figs. 5J–5M). We scanned the same areas in all groups. For each examined tooth, seven areas along the ventral side were chosen and measured. The O, Ca, P, and Au (or Pt) energy spectrum was acquired at each area (Figs. 5J & 5L). The results showed that the relative mass fraction of mineral elements was significantly decreased in dlx3b^−∕− enameloid, with the calcium content decreasing from 33.84% to 22.73% (Fig. 5N) and phosphorus content decreasing from 17. 52% to 11.71% (Fig. 5O). Specimens were examined by EDS and values were means for n = 5 individuals in each group. Nearly ninety percent of dlx3b^−∕− zebrafish showed a decrease in calcium and phosphorus content.

SCPP genes were downregulated in dlx3b mutants

The SCPP family is a group of tandemly duplicated genes involved in the mineralization of hard tissues, including bones and teeth. The principal enamel and dentin protein-related genes in mammals, such as Amelx, Enam, Odam and Dspp, all belong to the SCPP family. In dlx3b mutant larvae zebrafish, the expression of several SCPP genes, including odam, scpp9, spp1, scpp1, and scpp5, was significantly downregulated at 96 hpf compared with that in the wild-type siblings (Fig. 6, n = 3 independent extractions of 8 pooled individuals each). Among these SCPP genes, odam and scpp9 are mainly involved in the mineralization process of zebrafish enameloid, while spp1, scpp1, and scpp5 are closely related to dentin and bone development (Kawasaki, 2009).