Live imaging of transgenic Tg(tg:nlsEGFP) zebrafish permits real-time analysis of thyroid morphogenesis. (A–E) Epifluorescence microscopy of live Tg(tg:nlsEGFP) zebrafish (ventral view, anterior is to the top) is sufficiently sensitive to monitor stage-dependent changes in thyroid morphology during normal thyroid development (see A–D) and to detect the goitrous thyroid enlargement caused by PTU treatment (E). For each embryo shown, three-fold magnified views of the thyroid region are displayed (GFP channel). (F–K) Confocal analyses of double transgenic Tg(tg:mCherry;tg:nlsEGFP) larvae (100 hpf) confirms thyroid-specific co-expression of membrane-tethered mCherry (F,G,I,J) and nuclear EGFP reporters (F,H,I,K). 3D reconstruction of confocal images (F–H) and individual confocal sections (I–K) are shown. Ventral view, anterior is to the left. (L) Live imaging of a developmental series of Tg(tg:nlsEGFP) fish by confocal microscopy highlights the progressive expansion of thyroid tissue along the anterio-posterior axis (ventral view, anterior is to the top), the increase in thyroid cell number and the re-organization of cord-like cell clusters (arrows) into definitive follicular structures (arrowheads) during normal development. (M) Quantification of thyroid follicular cell (TFC) number in untreated control and PTU-treated zebrafish. Results are shown as means ± S.E.M (N = 6–10). Asterisks denote significant differences between treatment means at a given developmental stage (*P < 0.05, ***P < 0.001, Student’s t-test). (N–P) Comparison of thyroid morphologies between untreated control (Ct) and PTU-treated fish at different time point during the course of PTU treatment. Note the increased size and hyperplasia of thyroids from PTU-treated fish at 6 and 11 dpf. 3D reconstruction of confocal images obtained in live fish are shown (ventral view, anterior is to the top). (Q–T) Confocal sections highlight the stage-dependent increases in follicle size and luminal diameter in untreated control (Ct) fish (see Q,R). In PTU-treated fish, live imaging (ventral view, anterior is to the left) permitted monitoring of goitrous responses including early onset of thyroid cell hypertrophy (S), collapse of follicular lumina (T) and thyroid cell hyperplasia (T). Scale bars: 100 µm (A–E) and 20 µm (F–L, N–T).

Live imaging of transgenic Tg(tg:nlsEGFP) fish reveals diminished reporter expression and thyroid hypoplasia in thyroxine-treated fish. (A) Model for thyroid hormone negative feedback loop along pituitary-thyroid axis with tissue-specific gene expression markers examined in this study. T4, thyroxine; TSH, thyroid-stimulating hormone. (B,C) Treatment of zebrafish embryos with 5.0 µg/L T4 affects pituitary and thyroid gene expression as demonstrated by WISH. Pituitary expression of tshb mRNA (see ventral views in B) is reduced in T4-treated fish compared to untreated controls (Ct). At the thyroid level, expression of the TSH-dependent gene slc5a5 is strongly decreased following T4 treatment (ventral views in C, upper panels). T4-treated fish also showed decreased tg mRNA expression and a reduced number of tg mRNA expressing cells (ventral views in C, lower panels), effects which were more marked at 100 hpf. (D) Epifluorescence microscopy of live Tg(tg:nlsEGFP) fish revealed a strongly reduced intensity of the fluorescent reporter signal in T4-treated fish relative to untreated controls (Ct). For each embryo shown (ventral view), three-fold magnified views of the thyroid region are displayed (GFP channel). (E) Comparative analysis of thyroid morphology in control (Ct) and T4-treated larvae by confocal live imaging of g(tg:nlsEGFP) fish. 3D reconstructions of confocal images (ventral views) are shown. At 6 dpf, thyroids of T4-treated fish appeared atrophic (smaller follicles and reduced reporter expression). (F) Quantification of thyroid follicular cell (TFC) number in untreated controls (Ct) and T4-treated fish. Results are shown as means ± S.E.M (N = X − Y). Asterisks denote significant differences between treatment means at the indicated developmental stages (*P < 0.05, **P < 0.01, Student’s t-test). Anterior is to the top in all images. Scale bars: 20 µm (B,E) and 100 µm (C,D).

Recovery of thyroid dysgenesis phenotypes in zebrafish pax2a crispants. (A) Epifluorescence live imaging of Tg(tg:nlsEGFP) zebrafish. Ventral views of the head region (anterior to the right, scale bar: 100 µm) and magnified views of the thyroid region (GFP channel only, scale bar: 50 µm) are shown for non-injected controls and pax2a crispants at 55 hpf, 80 hpf, and 6 dpf. Phenotypic pax2a crispants displayed athyreosis or severe hypoplasia. (B) Whole-mount immunofluorescence staining of Tg(tg:nlsEGFP) zebrafish (6 dpf) for EGFP (thyroid cells) and thyroxine (colloidal T4). Epifluorescence imaging of the thyroid region in 6 dpf larvae (ventral views, anterior to the right, scale bar: 100 µm) revealed variable but reduced T4 content in pax2a crispants with thyroid hypoplasia and confirmed the absence of detectable T4 synthesis in pax2a crispants lacking EGFP-tagged thyroid cells (athyreosis group). (C) Distribution of allelic variants as determined by Illumina HiSeq analysis of individual pax2a crispants confirmed high mutagenesis efficiency in larvae affected by thyroid dysgenesis. The percentage of WT alleles (no variant call), in-frame indels, or frameshift indels is shown for N = 4 larvae per phenotypic category (median values with interquartile range).

Recovery of dyshormonogenesis phenotypes in zebrafish duox crispants. (A) Epifluorescence live imaging of Tg(tg:nlsEGFP) zebrafish. Ventral views of the head region (anterior to the right, scale bar: 100 µm) and magnified views of the thyroid region (GFP channel only, scale bar: 50 µm) are shown for non-injected controls and duox crispants at 55 hpf, 80 hpf, and 6 dpf. No thyroid phenotypes were detectable at 55 hpf and 80 hpf but by 6 dpf, the majority of duox crispants displayed a goitrous thyroid enlargement with severe hyperplasia. (B) Whole-mount immunofluorescence staining of Tg(tg:nlsEGFP) zebrafish (6 dpf) for EGFP (thyroid cells) and thyroxine (colloidal T4). Epifluorescence imaging of the thyroid region in 6 dpf larvae (ventral views, anterior to the right, scale bar: 50 µm) revealed reduced T4 content in duox crispants as the most prevalent phenotypic alterations. Most of these hypothyroid larvae displayed concurrent thyroid hyperplasia. (C) Distribution of allelic variants as determined by Illumina HiSeq analysis of individual duox crispants confirmed high mutagenesis efficiency in larvae affected by thyroid dyshormonogenesis. The percentage of WT alleles (no variant call), in-frame indels, or frameshift indels is shown for N = 4 larvae per phenotypic category (median values with interquartile range).

Recovery of hypoplastic/atrophic thyroid phenotypes in zebrafish tshr crispants. (A) Epifluorescence live imaging of Tg(tg:nlsEGFP) zebrafish. Ventral views of the head region (anterior to the right, scale bar: 100 µm) and magnified views of the thyroid region (GFP channel only, scale bar: 50 µm) are shown for non-injected controls and tshr crispants (target: exon 4) at 55 hpf, 80 hpf, and 6 dpf. No thyroid phenotypes were detectable at 55 hpf and 80 hpf but by 6 dpf, tshr crispants presented a hypoplastic/atrophic thyroid phenotype. (B) Whole-mount immunofluorescence staining of Tg(tg:nlsEGFP) zebrafish (6 dpf) for EGFP (thyroid cells) and thyroxine (colloidal T4). Epifluorescence imaging of the thyroid region in 6 dpf larvae (ventral views, anterior to the right, scale bar: 50 µm) revealed that thyroid hypoplasia in tshr crispants was accompanied by a reduction in thyroidal T4 content. (C) Distribution of allelic variants as determined by Illumina HiSeq analysis of individual tshr crispants revealed high mutagenesis efficiency in hypothyroid larvae presenting a hypoplastic thyroid. The percentage of WT alleles (no variant call), in-frame indels, or frameshift indels is shown for N = 4–6 larvae per phenotypic category (median values with interquartile range).

tshr crispants are resistant to PTU-induced thyroid hyperplasia. (A) Confocal live imaging of Tg(tg:nlsEGFP) zebrafish at 6 dpf. Three-dimensional reconstructions of confocal images of the thyroid region (ventral views, anterior to the left, scale bars: 40 µm) are shown for tshr crispants and non-injected siblings raised in the presence or absence of 30 mg/L phenylthiourea (PTU). Non-injected control larvae developed severe thyroid hyperplasia in response to PTU treatment. In contrast, only about 50% of the PTU-treated tshr crispants displayed a similar thyroid enlargement (phenotype #2). The remaining tshr crispants showed an almost complete resistance to PTU-induced thyroid enlargement and instead presented with a hypoplastic/atrophic thyroid phenotype (phenotype #1) similar to that seen in tshr crispants raised in the absence of PTU. (B) Quantification of thyroid follicular cell number in the different phenotypic groups at 6 dpf. Means ± S.E.M. are shown (N = 12–16). Different letters indicate significant differences between groups (P < 0.05, Tukey’s multiple comparison test). (C) Confocal live imaging identified mosaicism in cellular responses to PTU treatment in several tshr crispants. Note the atrophic characteristics of the follicle in the middle (low reporter expression, flat epithelium) compared to the marked hypertrophy/hyperplasia of the two neighbouring follicles.

duoxa germline mutants develop goitrous thyroid phenotypes. (A) Zebrafish duoxa genomic locus on chromosome 25 with sequences for the wild-type (WT) allele and a mutant allele (duoxa Δ11) containing a 11 bp deletion in exon 2. The sgRNA target site is underlined in the WT sequence. (B) PCR analysis of genomic DNA allows for sensitive detection of WT and duoxa Δ11 mutant alleles in individual fish (F3 generation). Polyacrylamide gel electrophoresis of PCR amplicons of WT, heterozygous and homozygous duoxa Δ11 carriers. Full-length gel is shown in Supplementary Fig. 7 from which lanes 1, 2, 3, 4 and 6 are shown in the cropped gel image. (C) Thyroid phenotyping of duoxa Δ11 mutant fish maintained on a Tg(tg:nlsEGFP) background. Immunofluorescence staining (GFP and T4) of 6 dpf larvae (ventral view, anterior to the top, scale bars: 20 µm) showed goitrous thyroid enlargement and absence of detectable T4 staining in all homozygous duoxa Δ11 fish (N = 30). Larvae with a normal-looking thyroid morphology (N = 30) were genotyped as either WT or heterozygous carriers of the duoxa Δ11 allele. For each larvae shown, 3.5-fold magnified views of the thyroid region are displayed (merge of GFP/T4 and T4 only). (D) Proportion of larvae with goitrous thyroid phenotype as detected in the progeny of three independent inbreeding experiments with heterozygous duoxa Δ11 fish.

Phenotypic spectrum of pax2a and nkx2.4b crispants. (A) Epifluorescence live imaging of Tg(tg:nlsEGFP) zebrafish larvae injected with Cas9 protein and sgRNAs targeting pax2a or nkx2.4b (exon 1 and exon 2), respectively. Ventral views of the thyroid region (anterior to the right) are shown for non-injected control fish and four different crispants displaying hypoplasia of variable severity. Images were aquired at 55 hpf, 80 hpf, and 6 dpf. (B) Whole-mount immunofluorescence staining of Tg(tg:nlsEGFP) zebrafish for EGFP and colloidal T4 at 6 dpf. Ventral views of the thyroid region (anterior to the right) are shown for non-injected control fish and four different crispants displaying hypoplasia of variable severity. Scale bars: 50 μM.

Thyroid anlage specification is not perturbed in pax2a crispants. (A,B) Whole-mount in situ hybridization for the early thyroid marker nkx2.4b in non-injected control embryos and pax2a crispants. Thyroidal nkx2.4b expression (arrow) was not different between experimental groups at 28 hpf. Panel A shows dorsal views of stained specimen (anterior is to the top). Panel B shows lateral views (anterior is to the right). (C) Loss of thyroid marker expression in later stage pax2a crispants. Whole-mount in situ hybridization for the thyroid differentiation marker tg in non-injected control embryos and pax2a crispants at 55 hpf. In contrast to the strong tg staining in the compact and slightly ovoid thyroid primordium of control embryos, half of the pax2a crispants showed either a complete absence of detectable tg staining (data not shown) or presented thyroid primordia of greatly reduced size. Ventral views are shown (anterior is to the top). (D) pax2a crispants display strongly reduced immunoreactivity to a pax2a antibody. Whole-mount immunofluorescence staining of pax2a (magenta) was performed on non-injected control embryos and pax2a crispants at 24 hpf using a pax2a antibody directed against an epitope located C-terminal to the sgRNA target site. Approximately half of the pax2a crispants (53.7%; N=79/147) displayed a strongly reduced pax2a immunofluorescence signal. Lateral views are shown (anterior is to the right). Arrowheads: mid-hindbrain boundary; arrows: thyroid anlage; asterisks: otic vesicle. Scale bars: 100 μm.

Phenotypic spectrum of duox and duoxa crispants. (A) Epifluorescence live imaging of Tg(tg:nlsEGFP) zebrafish larvae injected with Cas9 protein and sgRNAs targeting duox or duoxa, respectively. Ventral views of the thyroid region (anterior to the right) are shown for non-injected control fish and four crispants. Images were aquired at 55 hpf, 80 hpf, and 6 dpf. No deviation from control thyroid morphology was evident in duox and duoxa crispants at 55 and 80 hpf. By 6 dpf, hyperplastic thyroid enlargement (goiter) was detectable in many duox and duoxa crispants. (B) Whole-mount immunofluorescence staining of Tg(tg:nlsEGFP) zebrafish for EGFP and colloidal T4 at 6 dpf. Ventral views of the thyroid region (anterior to the right) are shown for non-injected control fish and different crispants displaying hypothyroidism (decreased T4 staining) that was often but not always associated with a hyperplastic thyroid enlargement. Scale bars: 50 μm.

Phenotypic spectrum of tshr crispants. (A) Epifluorescence live imaging of Tg(tg:nlsEGFP) zebrafish larvae injected with Cas9 protein and sgRNAs targeting exon 4 or exon 10 of tshr, respectively. Ventral views of the thyroid region (anterior to the right) are shown for non-injected control fish and four crispants. Images were aquired at 55 hpf, 80 hpf, and 6 dpf. No deviation from control thyroid morphology was evident in tshr crispants at 55 and 80 hpf. By 6 dpf, thyroids of tshr crispants had an atrophic/hypoplastic appearance and displayed a greatly reduced GFP reporter signal. (B) Whole-mount immunofluorescence staining of Tg(tg:nlsEGFP) zebrafish for EGFP and colloidal T4 at 6 dpf. Ventral views of the thyroid region (anterior to the right) are shown for non-injected control fish and different tshr crispants displaying hypothyroidism (decreased T4 staining), variable thyroid hypoplasia and reduced GFP expression. Scale bars: 50 μm.

Characterization of adamtsl2 crispants. (A) adamtsl2 mRNA is expressed at high levels in the zebrafish thyroid primordium (arrow) as demonstrated by whole-mount in situ hybridization. Lateral and frontal views of 55 hpf embryos stained with an adamtsl2-specific riboprobe are shown. (B) Zebrafish adamtsl2 genomic locus on chromosome 5 with sequences for exon 1 and exon 13. Underlined sequences are target sites for sgRNA-ex1 and sgRNA-ex13, respectively. PAM sequences are highlighted in yellow and the boxed sequence in exon 1 is the BclI restriction enzyme recognition site. (C) Thyroid phenotyping of adamtsl2 crispants did not reveal alteration in thyroid morphogenesis or thyroid function. Results shown are from an injection experiment with sgRNA-ex13 and similar results were obtained with sgRNA-ex1. Epifluorescence live imaging of Tg(tg:nlsEGFP) zebrafish at 55 hpf, 80 hpf, and 6 dpf. Ventral views of the head region and 3-fold magnified views of the thyroid region (GFP channel only) are shown for non-injected controls and adamtsl2 crispants. Whole-mount immunofluorescence (IF) staining of 6 dpf Tg(tg:nlsEGFP) zebrafish for EGFP (thyroid cells) and thyroxine (colloidal T4). Ventral views of the head region (anterior is to the top) and 3-fold magnified views of the thyroid region (GFP, T4) are shown. Scale bars: 100 μm. (D) Polyacrylamide gel electrophoresis (PAGE) of PCR amplicons generated with primers spanning the target site of sgRNA-ex1. Upper panel shows a gel for PCR amplicons of WT larvae (single PCR product of 206 bp) and adamtsl2 crispants (multiple slow-migrating heteroduplex bands). Lower panel shows a gel for PCR amplicons digested with BclI. Note complete digestion of PCR amplicons from WT larvae resulting in two restriction fragments of 111bp and 95 bp, respectively. Images of fulllength gels are shown in Supplementary Figure 9A,B. (E) PAGE analysis of PCR amplicons generated with primers spanning the target site of sgRNA-ex13. Gel showing PCR amplicons of non-injected WT larvae (single PCR product of 151 bp) and adamtsl2 crispants (multiple slow-migrating heteroduplex bands in most samples). Full-length gel is shown in Supplementary Figure 9C.

Brightfield micrographs of crispants and germline mutant fish (A) Brightfield microscopy of control larvae and crispants displaying a thyroid phenotype after injection of gRNAs targetting pax2a, duox andtshr. Lateral views (anterior to the left) of 6 dpf old fish are shown. Arrows highlight defective brain development in pax2a crispants. (B) Brightfield microscopy of WT larvae and homozygous carriers of the duoxa Δ11 allele. Lateral views (anterior to the left) of 5 dpf old larvae (pigment-less casper fish) are shown.