Zebrafish as a Model for Monocarboxyl Transporter 8-Deficiency*

Background: Mutations in the thyroid hormone transporter MCT8 are associated with the psychomotor retardation Allan-Herndon-Dudley syndrome (AHDS). Results: In zebrafish, as in humans, mct8 is expressed primarily in the nervous system. Elimination of MCT8 causes severe neural impairment. Conclusion: MCT8 is a crucial regulator during zebrafish embryonic development. Significance: Establishment of the first vertebrate model for MCT8 deficiency, which exhibits a neurological phenotype. Allan-Herndon-Dudley syndrome (AHDS) is a severe psychomotor retardation characterized by neurological impairment and abnormal thyroid hormone (TH) levels. Mutations in the TH transporter, monocarboxylate transporter 8 (MCT8), are associated with AHDS. MCT8 knock-out mice exhibit impaired TH levels; however, they lack neurological defects. Here, the zebrafish mct8 gene and promoter were isolated, and mct8 promoter-driven transgenic lines were used to show that, similar to humans, mct8 is primarily expressed in the nervous and vascular systems. Morpholino-based knockdown and rescue experiments revealed that MCT8 is strictly required for neural development in the brain and spinal cord. This study shows that MCT8 is a crucial regulator during embryonic development and establishes the first vertebrate model for MCT8 deficiency that exhibits a neurological phenotype.

In all vertebrates, thyroid hormones (THs) 2 are essential regulators of development, neurogenesis, growth, and metabolism (1). TH actions are mediated via intracellular activation and inactivation of iodothyronine deiodinases and binding of triio-dothyronine (T 3 ) to nuclear TH receptors, which regulate gene transcription (1). Thus, to function, THs require efficient transport across the cell membrane. In the last two decades, several transmembrane transporters, which are required for the cellular uptake and efflux of THs, have been functionally described, including the T 3 -specific monocarboxylate transporter 8 (MCT8) (2), the T-type amino acid transporter MCT10 (3,4), and the organic anion-transporting polypeptide 1C1 (OATP1C1) (5), among others (6). Importantly, mutations in MCT8 were associated with the X-linked Allan-Herndon-Dudley syndrome (AHDS), which is characterized by elevated serum T 3 levels and severe psychomotor retardation (7,8). The mechanism underlying this disorder is thought to involve a defect in the MCT8-dependent neuronal entry of T 3 , leading to impaired neurological development. However, little is known about the role of MCT8 in regulating embryonic development and AHDS.
The abnormal serum TH levels observed in AHDS patients were closely replicated in MCT8-deficient mice; however, these mice did not display apparent neurological or behavioral phenotypes (9,10). Biochemical studies suggested that the transport of T 3 , but not its precursor, thyroxine (T 4 ), is impeded in MCT8-deficient mice, thus additional TH transporters, such as OATP1C1 and MCT10, might compensate for MCT8 deficiency in rodents (11,12). Clearly, the establishment of additional MCT8-deficient vertebrate models that mimic the pathophysiological condition of AHDS patients is required to complement the mouse model and to better understand the role of MCT8.
The zebrafish is a simple vertebrate model with conserved organization of the central nervous system (CNS), which is ideally suited to study genetics, transcriptional regulation, neuronal development, synaptogenesis, and behavior in live animals (13)(14)(15). The larval optical translucency provides the unique ability to visualize single neurons in live animals (13). Importantly, the zebrafish thyroid system, including the hypothalamus-pituitary-thyroid gland axis (16,17), and the main genes involved in TH signaling (18 -21), are largely conserved between zebrafish and mammals. Moreover, studies on cell lines have recently demonstrated that the zebrafish MCT8 is able to transport THs (22).
In this study, we isolated the complete mct8 gene and promoter and described the expression pattern of zebrafish mct8 during development. mct8 is mainly expressed in the nervous and vascular systems. Using knockdown experiments and in vivo imaging, we showed that the lack of MCT8 causes developmental and neurological impairment. These results establish the zebrafish as a model for studying the role of MCT8 and the mechanisms underlying AHDS.

EXPERIMENTAL PROCEDURES
Zebrafish Husbandry-Adult zebrafish were raised and maintained in fully automated zebrafish housing systems (Aquazone, Israel; temperature 28 Ϯ 0.5 ºC, pH 7.0, conductivity 300 S) under 14-h light/10-h dark cycling, and fed twice a day. Embryos were generated by natural spawning and raised in egg water in a 28 Ϯ 0.5 ºC, light-controlled incubator, as previously described (23). All animal protocols were reviewed and approved by the Bar-Ilan University Bioethics Committee.
Isolation of mct8 mRNA and Rapid Amplification of cDNA Ends (RACE)-The mouse MCT8 sequence (NM_009197.2) was used as a query in the BLAT algorithm of the University of California Santa Cruz zebrafish genome browser using the Zv9/ danRer7 version (genome.ucsc.edu). The partially predicted mct8 545-bp sequence that was found on chromosome 14 was used for designing gene-specific primers (GSPs). 5Ј and 3Ј sequences of the zebrafish mct8 mRNA were determined using the 5Ј and 3Ј systems for RACE, according to the manufacturer's protocol (Invitrogen). The specific primers 5ЈGSP1 (5Јggaatcatcatggacatcac-3Ј) 5ЈGSP2 (5Ј-tcttgactccaggtatgaggtctcc-3Ј) and 5ЈGSP3 (5Ј-agacgaagctccgatgcacaccagc-3Ј) were used for the 5Ј RACE analysis and the specific primers 3ЈGSP1 (5Ј-tgtgatgtttttcgtgcctc-3Ј) and 3ЈGSP2 (5Ј-ccgacgatcccagcatggac-3Ј) were used for 3Ј RACE analysis.
To isolate the mct8 promoter, a fragment containing 1,728 bp of genomic 5Ј flanking region and 272 bp 5Ј UTR of the mct8 gene was amplified (JQ966310) from zebrafish genomic DNA using the specific primers mct8Pro(2000)F, incorporating a SalI restriction site (5Ј-cgcctcgaggacaccaacacccccataatgggac-3Ј) and mct8UtrR, containing a BglII restriction site (5Ј-cgcagatctcctacagggagaggatgcagacgcg-3Ј). The PCR product was double-digested with SalI and BglII, and ligated into a BamHI/ SalI-digested pT2-ALR150G (25) upstream of the enhanced green fluorescent protein (EGFP) reporter gene to create the pT2-mct8:EGFP construct (Fig. 2B). This construct was later used for the preparation of the Tg(mct8:EGFP) transgenic line (see below). The GAL4-VP16 transcriptional activator was amplified and subcloned into a NcoI/BglII-digested pT2-mct8: EGFP, replacing the EGFP, to create the pT2-mct8:GAL4 construct. This construct was used to generate the Tg(mct8:GAL4) transgenic line (see below).
Whole Mount in Situ Hybridization and Immunohistochemistry Assays-In both whole mount ISH and immunofluorescence experiments, embryos and larvae were fixed in 4% paraformaldehyde overnight at 4 ºC, washed in PBS, and stored in 100% methanol. The location and level of mRNA expression were detected by whole mount ISH, as described (27,28). Digoxigenin-labeled full-length antisense riboprobes for mct8, mct10, oatp1c1, dio1, dio2, dio3, and myod were transcribed in vitro using the vector templates described above, and standard reagents followed the manufacturer's instructions (Roche Applied Science).
Establishment of Stable Transgenic Lines and Colocalization Experiments-To transiently express pT2-mct8:EGFP and pT2-mct8:GAL4/uas:EGFP in live fish, the constructs were diluted to a concentration of 50 ng/l and microinjected, using a micromanipulator and a PV830 Pneumatic Pico Pump (World Precision Instruments, Sarasota, FL), into one-cell stage eggs. The embryos were kept in Petri dishes, and the pattern of EGFP expression was monitored throughout their development. To generate Tg(mct8:EGFP) and Tg(mct8:GAL4) stable transgenic fish, the Tol2 system was used (25). Capped RNA encoding the Tol2 transposase and the pT2-mct8:EGFP or pT2-mct8:GAL4 constructs were co-injected independently (in a concentration of 25 ng/l each) into fertilized eggs at one-cell stage. The injected fish (generation F0) were raised to adulthood and screened for integration of the transgene into the germline. F0 fish, injected with pT2-mct8:EGFP, were crossed with wild-type fish, and F0 fish, injected with pT2-mct8:GAL4, were crossed with a Tg(uas:EGFP) transgenic line. Transgenic EGFP-positive lines (F1) were screened and isolated using a fluorescent stereomicroscope (M167FC, Leica, Wetzlar, Germany). Four Tg(mct8:EGFP) and three Tg(mct8:GAL4/uas: EGFP) transgenic lines were obtained. To obtain the Tg(mct8: GAL4) line, Tg(mct8:GAL4/uas:EGFP) fish were out-crossed with wild-type fish. All lines showed similar patterns of EGFP expression mainly in the central nervous system and along the spinal cord. The transgenic lines that showed the strongest EGFP expression were used in this study. Transgenic fish were established in the nacre Ϫ/Ϫ mutant (29) background to avoid pigmentation. Tg(mct8:EGFP/fli:DsRED) double transgenic larvae were produced by crossing adult Tg(mct8:EGFP) and Tg(fli:DsRED) (30).
Morpholino Design, Preparation, and Injection-Gene knockdown experiments were performed using the following morpholino-modified antisense oligonucleotides (MO, Gene Tools, Philomath, OR): Gene Tools standard control MO (5Јctcttacctcagttacaatttata-3Ј), mct8(E2I2)MO (5Ј-ataaaatcatgtatttacgtggcga-3Ј), and mct8(UTR)MO (5Ј-tcctacagggagaggatgcagacgc-3Ј). The mct8(UTR)MO was designed to block MCT8 translation. To validate the efficiency of the mct8(UTR)MO, it was injected into Tg(mct8:EGFP) embryos. Because the 5ЈUTR of mct8 is present in the mct8:EGFP transgene, EGFP expression was eliminated in all injected embryos (n ϭ 304, Fig. 5, F-H). The mct8(E2I2)MO was designed to interfere with the splicing of the second exon/intron and, thereby, to introduce a premature stop codon. Indeed, mct8 CDS and a 341-bp fragment incorporating parts of the second and third exons (mct8 E2E3) were PCR-amplified from the cDNA of control MO-injected embryos but not from mct8(E2I2)MO-injected embryos (Fig. 4A). In all experiments, Tg(mct8:EGFP), Tg(mct8:GAL4/ uas:EGFP), or wild-type embryos were injected with 0.3-1.6 pmol of MO. In rescue experiments, 80 pg of in vitro transcribed mct8 mRNA was co-injected in combination with 1 pmol of mct8(E2I2)MO or 0.3 pmol of mct8(UTR)MO. Injected embryos were monitored under a M167FC stereomicroscope (Leica, Wetzlar, Germany) and sorted into three groups: normal development, mildly altered development, and severely altered development. Following sorting, embryos were counted, and statistical significances between the different groups were determined by -square tests.
Real-time PCR Quantification Assays-The levels of mRNA expression of dio1, dio2, dio3, tshb, and ␤-actin were determined using quantitative real-time PCR assays. In three independent experiments, mRNA of 48 h post-fertilization (hpf) larvae injected with 0.5 pmol of mct8 (UTR) MO, mct8 (E2I2) MO, or standard control MO were extracted using the RNeasy Protect mini kit according to the manufacturer's instructions (Qiagen). A similar amount of mRNA (280 ng) was reverse transcribed, as described above. Transcript levels were determined by the 7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA) using the KAPA SYBR FAST qPCR Kit (Kapa Biosystems, Cambridge, MA) according to the manufacturer's instructions. Triplicate first-strand cDNA aliquots from each sample served as templates in real-time PCR. The relative quantification of dio1, dio2, dio3, and tsh␤ mRNA expression levels were normalized against ␤-actin mRNA expression levels and subjected to the ⌬⌬C T method (23). Statistical significance was determined using one-way analysis of variance.
Imaging-An epifluorescence stereomicroscope (Leica M167FC) was used to visualize larvae expressing fluorescent reporters and for imaging whole mount ISH-stained larvae. Pictures were taken using Leica Application Suite imaging software version 3.7 (Leica, Wetzlar, Germany). For confocal imaging, embryos and larvae were anesthetized with Tricaine (0.01%) and placed in low-melting point-agarose (0.5-1.0%) on a specially designed dish filled with embryo water. Similar mounting protocol was used to image fixed embryos subjected to immunohistochemistry. Confocal imaging was performed using either a Zeiss LSM710 or LSM780 upright confocal microscope (Zeiss, Oberkochen, Germany). All images were processed using ImageJ (National Institutes of Health, Bethesda, MD) and Adobe Photoshop (San Jose, CA) software.
Bioinformatical Analyses-The prediction of transcription factor-binding elements, within orthologous putative mct8 promoters, was performed using the MatInspector software tool (Genomatix, Munich, Germany). Sequences of 2000 bp of 5Ј flanking regions upstream to the putative translation start sites from the human, rat, mouse, and zebrafish genomes, were used. Predicted transcription factors found in all four putative promoter sequences, and with the lowest p value scores, were selected. The prediction of putative transmembrane domains (TMDs) within the TH transporters was achieved using the Simple Modular Architecture Research Tool (SMART) online software (Biobyte Solutions GmbH, Heidelberg, Germany), and the calculation of intra-and extracellular loops was performed manually. The prediction of PEST domains was performed using Mobyle@Pasteur version 1.0.4 online software (Pasteur Institute, Paris, France).

RESULTS
Isolation of the Full-length mct8 Transcript-We isolated a 1578-bp fragment of the mct8 coding sequence, a 272-bp fragment of the 5Ј UTR, and 557-bp 3Ј UTR fragments (Fig. 1A). BLAT analysis (UCSC genome browser) revealed that the complete mct8 mRNA (accession number JQ966311) consists of seven exons located on chromosome 14 in a region that was not fully sequenced in the latest version of the zebrafish genome (Sanger Institute, Zv9).
TH Transporter Proteins Are Well Conserved in Zebrafish and Mammals-Protein motif comparison of MCT8, MCT10, and OATP1C1 showed that the fish TH transporters share 12 similar TMDs, as well as similar intra-and extracellular loop spans, with their human orthologs (Fig. 1B). Confirming previous observations (22), the zebrafish protein sequence of MCT8 shares ϳ60% identity with its human homolog, including a PEST domain at the N-terminal end. In addition, an arginine residue within TMD8 and an aspartate residue within TMD10, which were previously identified as being involved in substrate interaction in mammalian MCT8 (31,32), are present in the zebrafish MCT8. Similarly, zebrafish MCT10 holds both the conserved arginine and aspartate residues as well as a PEST domain at the N-terminal end, and shares 71% identity with its human ortholog. Furthermore, OATP1C1 of zebrafish (19) and humans share 58% identity, and sensitive residues, such as two tryptophan amino acids in TMD6, a couple of glycine residues in TMD8, and an arginine within TMD11, are well conserved from fish to human (Fig. 1B). The conserved sequences and motifs of zebrafish and human MCT8, MCT10, and OATP1C1, suggest that the transport mechanism of TH is well conserved from fish to humans.
Expression Patterns of TH Transporters, mct8 Is Widely Expressed in the Nervous System-To characterize the spatial and temporal expression patterns of mct8, whole mount ISH was performed at several developmental stages. Ubiquitous mct8 expression was observed during the bud stage in 10-hpf embryos (Fig. 1C). Later, at 24 hpf, the expression pattern of mct8 was most abundant in the forebrain, midbrain, hindbrain, spinal cord, notochord, and eyes (Fig. 1, D and E). At 48 hpf, mct8 expression was mainly observed in the brain and along the spinal cord (Fig. 1, F and G). Unlike the broad expression of mct8 in the CNS, oatp1c1 expression was restricted to vascular structures within the brain (Fig. 1, H and I), and expression of mct10 was restricted to the liver and trigeminal ganglia (Fig. 1, J-M). These results suggest that MCT8 plays a general role during the development of the nervous system, and that other TH transporters are not likely able to fully compensate for MCT8 deficiency in zebrafish.
Isolation of a Functional mct8 Promoter and the Establishment of a Tg(mct8:EGFP) Transgenic Line-To visualize MCT8-positve cells in a live, developing animal, we sought to identify the zebrafish mct8 promoter. An in silico approach was applied to identify conserved DNA regulatory sequences within putative mct8 promoters. Comparison of our 5Ј RACE analysis with available genomic data revealed that the start codon is located within the first exon in zebrafish and mammals (Fig.  1A). Therefore, fragments of 2000 bp upstream to the ATG of zebrafish, mouse, rat, and human, were analyzed. Conserved transcription factor binding sites for the purine-rich singlestranded DNA-binding protein ␣ (PURA, p Ͻ 1.4E-04), the Huntington disease gene-regulatory region binding proteins (HDBP, p Ͻ 5.77E-04), and GATA (p Ͻ 0.05) were found in all four putative promoters ( Fig. 2A). To functionally test the zebrafish mct8 promoter in vivo, a 2000-bp genomic fragment (Fig. 2B, accession number JQ966310), located upstream of the start codon, was cloned upstream of an EGFP. The pT2-mct8: EGFP construct was microinjected into one-cell stage embryos. At 24 hpf, ubiquitous mosaic EGFP expression was observed, suggesting that the 2000-bp promoter can drive gene expression in vivo. To test whether this promoter can drive specific expression in all mct8-positive cells, a stable Tg(mct8:EGFP) transgenic line was established (Fig. 2, C-I). At 24 hpf, this Tg(mct8:EGFP) fish expressed EGFP mainly in the forebrain, midbrain, hindbrain, along the spinal cord and notochord, and in the eyes (Fig. 2, C-E). Expression in the notochord gradually disappeared during development. High resolution imaging revealed expression in various tissues, including cells along the spinal cord and notochord (Fig. 2F), epithelial cells above the otic vesicle (Fig. 2G), the choroid plexus (Fig. 2H), and olfactory bulbs (Fig. 2I). This expression pattern was consistent at least until 11 days post-fertilization (dpf).
mct8 Is Expressed in the Vascular System-As mct8 is a hormone transporter, we monitored its expression in the vascular system. At 3 dpf, EGFP expression was observed in vessel-like structures in Tg(mct8:EGFP) larvae (Fig. 2E, white dashed box). To examine this expression pattern, Tg(mct8:EGFP) and Tg(fli: DsRED) fish were crossed, and their progeny were imaged by confocal microscopy. In the Tg(fli:DsRED) line, blood and lymphatic vessels were marked with red fluorescent protein (30). In the Tg(mct8:EGFP/fli:DsRED) double transgenic larvae, EGFP and DsRED colocalized in trunk vessels, specifically, in the dorsal aorta, the posterior cardinal vein, and the intersegmental vessels (Fig. 3, A-C). In addition, colocalization was observed in the thoracic duct of the lymphatic system. To check whether mct8 is expressed in blood vessels in the brain, imaging was  performed in the head of Tg(mct8:EGFP/fli:DsRED) larvae (Fig.  3, D-F). Notably, colocalized expression was detected in the midbrain veins, the middle cerebral veins, the dorsal longitudinal vein, and the posterior cerebral veins (33). These results indicate that MCT8 is expressed in both blood and lymphatic vascular systems and is likely involved in TH transport into the CNS.
mct8 Is Expressed in Neurons and Neuron-supporting Cells-To specifically identify the cells that express mct8 in the CNS, immunofluorescence double labeling assays were conducted in Tg(mct8:GAL4/uas:EGFP) fish, which demonstrated more robust EGFP expression levels than the Tg(mct8:EGFP) line. Antibodies against EGFP and the HUC protein, a marker of developing neurons (34), or the glial fibrillary acidic protein (GFAP), a marker of astrocytes (35), were used. At 3 dpf, EGFP immunoreactive cell bodies were detected in the brain and along the spinal cord, where they colocalized with HUC-positive neurons (Fig. 3, G-L). In contrast, EGFP immunoreactive cell bodies were not colocalized with GFAP-positive astrocytes in the spinal cord (Fig. 3, M-O). These results indicate that mct8 is expressed in neurons but not in astrocytes. Furthermore, EGFP immunoreactive cells were observed in cells that were not stained by either HUC or GFAP antibodies (Fig. 3m G,  I, J, L, M, an O). The location and shape of these cells was similar to SOX10-positive oligodendrocyte; the myelin-forming cells (36,37). These results suggest that MCT8 transports THs from the blood vessels into oligodendrocytes and neurons, and regulates myelination and neuron development.
Establishment of MCT8-deficient fish, the Expression of THrelated Genes Is Not Altered-To establish an MCT8-deficient model in zebrafish, MCT8 was knocked down (KD) by injecting two different MOs into one-cell stage embryos. The mct8(UTR)MO was designed to block translation and the mct8(E2I2)MO was designed to interfere with the splicing of the second exon/intron. In vivo and in vitro experiments confirmed that both MOs are able to efficiently and specifically knockdown MCT8 (Figs. 4A and 5, respectively). To examine the effect of MCT8 knockdown on the TH endocrinological system, spatial and quantitative expression assays were performed on the three deiodinases (dio1, -2, and -3) (18,20) and the thyroid stimulating hormone ␤ (tsh␤) (38). The deiodinases selectively remove iodide from thyroxine and its derivatives, thus activating or inactivating THs. TSH is secreted by the pituitary and regulates the function of the thyroid gland. TSH consists of ␣and ␤-subunits (39,40). As the localization of the deiodinase genes had not been previously described in zebrafish, the spatial expression of the three deiodinases was monitored using whole mount ISH (Fig. 4, B-D). At 48 hpf, dio1 was strongly expressed in the liver and weakly expressed in the head. dio2 was specifically expressed in the thyroid gland, and dio3 was expressed primarily in the pronephros. Next, at 48 hpf, total mRNA was extracted from mct8(E2I2)MO-, mct8(UTR)MO-, or control MO-injected embryos, and the expression levels of dio1, dio2, dio3, and tsh␤ were quantified using quantitative RT-PCR. KD of MCT8 did not affect the expression levels of the deiodinases and tsh␤ (Fig. 4E). These results show that zebrafish MCT8 does not regulate deiodinase and tsh␤ gene expression in the whole larvae at early develop-mental stages. Nevertheless, MCT8 may affect these enzymes in specific tissues or at the protein and protein-activity levels.
Knockdown of mct8 Alters the Development of Zebrafish Embryos-To study the effect of MCT8 knockdown on the morphology and development of zebrafish embryos,  JANUARY 4, 2013 • VOLUME 288 • NUMBER 1 JOURNAL OF BIOLOGICAL CHEMISTRY 175 mct8(E2I2)MO, mct8(UTR)MO, or control MO were independently injected into one-cell stage Tg(mct8:EGFP) embryos. MO-injected embryos exhibited a normal survival rate (above 95% at 24 hpf). Notably, at 48 hpf, the injection of both mct8(E2I2)MO and mct8(UTR)MO caused an altered developmental phenotype, characterized by small eyes, decreased pigmentation, pericardial edema, and perturbed trunk and tail development (Fig. 5). A large portion of the MO-injected embryos exhibited a mild phenotype and their morphology was only slightly altered. The number of embryos displaying mild to severe altered development was significantly higher compared with the control MO (p Ͻ 1Xe Ϫ20 , 2 ϭ 10857.492, df ϭ 2 and p Ͻ 1Xe Ϫ20 , 2 ϭ 36118.169, df ϭ 2, respectively). Moreover, because MCT8-expressing cells were fluorescently labeled, we noticed that the brain, spinal cord, and notochord were severely deformed (Fig. 5, B, C, F,  and G). These results suggest that MCT8 is necessary for normal embryonic development.  D, E, and H), severe (B and F), and mild (C and G) phenotypes are shown. Injection of mct8 mRNA alone did not cause abnormal development and rescued the altered development phenotype observed in MO-injected embryos. I, embryos injected with MOs or MOsϩmct8 mRNA were sorted according to the morphological criteria shown in A-H. The percentage of embryos from each phenotype is presented. The numbers of mct8(E2I2)MO-and mct8(UTR)MO-injected embryos demonstrating altered development were significantly (p Ͻ 1 ϫ e Ϫ20 ) higher than the number of altered control MO-injected embryos. In rescue experiments, the injection of mct8 mRNA into MOs-injected embryos significantly (p Ͻ 1 ϫ e Ϫ20 ) reduced the number of embryos demonstrating altered morphology. Statistical significance was determined by square tests and by comparing the distribution of normal, mild, and severely altered development phenotypes.

MCT8 Regulates Neural Development in Zebrafish
To confirm that this altered development of the nervous system is specific to MCT8 deficiency, we performed rescue experiments. We initially injected mct8 mRNA into Tg(mct8:EGFP) one-cell stage embryos, and no apparent morphological changes were observed (Fig. 5E). Co-injection of mct8 mRNA with mct8(E2I2)MO or mct8(UTR)MO rescued the altered developmental phenotype (Fig. 5, D and H). Following co-injections of each MO with mct8 mRNA, a significantly lower percentage of embryos displayed an altered developmental phenotype (Fig. 5, I, p Ͻ 1Xe Ϫ20 , 2 ϭ 5392.65, df ϭ 2 for the mct8(E2I2)MO versus mct8(E2I2)MOϩmct8 mRNA; p Ͻ 1Xe Ϫ20 , 2 ϭ 16290.365, df ϭ 2 for the mct8(UTR)MO versus mct8(UTR)MOϩmct8 mRNA). These results indicate that mct8 mRNA can rescue the developmental defects caused by MCT8-KD, and that the MO-mediated phenotype is a specific result of MCT8 deficiency.
Knockdown of mct8 Specifically Alters Neural Development in Zebrafish Embryos-To investigate the function of MCT8 in specific tissues, whole mount ISH and immunohistochemistry experiments were performed in MO-injected embryos that demonstrated mild morphological phenotype. Because MCT8 deficiency in human patients affects muscle tone, the morphology and development of the muscles were studied. At 24 hpf, the expression pattern of myod mRNA, a marker for muscle development (24), was monitored in control MO-and mct8(E2I2)MO-injected embryos. Although the trunk was mildly deformed (as described above, Fig. 5), myod expression levels and muscle morphology were mostly similar in mct8(E2I2)MO-injected and wild-type embryos (Fig. 6, A and  B). Similarly, at 5 dpf, the morphology of the vascular system in mct8(E2I2)MO-injected Tg(fli:DSRED) larvae was mostly intact (Fig. 6, C and D). In contrast, in the CNS, the number of MCT8-positive cells was reduced and the cell organization was altered in the hindbrain, midbrain-hindbrain boundary, midbrain (Fig. 6, G and H), and spinal cord (Fig. 6, E and F) of mct8(E2I2)MO-injected 2 dpf larvae. These results strongly suggest that loss of MCT8 does not affect the development of muscles and vessel; however, it plays an essential role in the development of the CNS. Altogether, these experiments establish the zebrafish as a model for studying the role of MCT8 and the mechanisms underlying AHDS.

DISCUSSION
The effect of MCT8 on neural development and its role in the psychomotor retardation AHDS is not clear. To study the function of MCT8, we developed an MCT8-deficent zebrafish. This simple model provides genetic and imaging tools that are unique among all vertebrate models. To characterize the spatial and temporal expression of mct8, we cloned the complete mct8 mRNA and performed whole mount ISH on developing embryos. We found that mct8 was ubiquitously expressed at 10 hpf. As MCT8 transports THs, and maternal THs are important for early brain development (41,42), the observed mct8 expression implies that MCT8 may allow the accumulation of active THs, even at early developmental stages. At 1-2 dpf, mct8 was mainly expressed in the CNS, specifically in the forebrain, midbrain, hindbrain, and along the spinal cord and notochord. This expression pattern partially recapitulates the expression profile of mct8 in the mammalian nervous system (43,44). These results suggest that MCT8 plays a general role in early developmental stages and is possibly a key regulator during the development of the nervous system.
To understand a broader view of zebrafish TH transport, we also performed whole mount ISH using probes of other TH transporters. Although mct8 was widely expressed in the CNS, oatp1c1 expression was restricted to vasculature structures within the brain. Similarly, in rodents, oatp1c1 shows a strong expression in brain endothelial cells and in choroid plexus structures (6,12,45). MCT10, which is thought to act in the liver, intestine, kidneys, and growth plate chondrocytes in mammals (6,46,47), was observed in the liver and the trigeminal ganglia in zebrafish. Recently, OATP1C1 was implicated as compensating for the lack of MCT8 in MCT8-knock-out mice (12,48). The distinct expression patterns of oatp1c1 and mct10 suggest that, in contrast to the mouse model and as is the case in humans, they are not likely able to fully compensate for MCT8 deficiency in zebrafish.
To further study the expression pattern of MCT8 in live developing animals, a Tg(mct8:EGFP) stable transgenic line was generated. The EGFP pattern was similar to endogenous mct8 expression, thus, this line provides a platform for the imaging of MCT8-positive cells in live animals. High magnification imaging revealed EGFP expression in various cell types, including epithelial cells above the otic vesicle, in the olfactory bulb, and in the choroid plexus. This expression pattern is consistent with previous reports on chicken embryos (42,49) and mammals (43). The expression pattern in the choroid plexus suggests that MCT8 may play a role in the transport of THs into the cerebrospinal fluid. Furthermore, double immunofluorescence assays revealed that mct8 colocalized with HUC in neurons of the brain and spinal cord but not with GFAP in astrocytes, consistent with the mct8 expression pattern in mice (6,43,50). In addition, mct8 expression was observed in oval cell bodies with short projections that did not express both HUC and GFAP markers. Because the structure and the location of these cells are similar to sox10 expressing cells (36,37), they are likely to be oligodendrocytes. Similarly, in mammals, mct8 mRNA and protein were present in oligodendroglial cells derived from mice (50). These results suggest a role for MCT8 in the uptake of THs in neurons and oligodendrocytes. Thus, MCT8 regulates the maintenance and development of neurons. Because a key role of oligodendrocytes is to produce a myelin sheath, MCT8 may also be involved in the myelination of neurons. This hypothesis is consistent with recent reports of delayed myelination in AHDS patients (51). Future studies on live developing zebrafish are required to understand the role of MCT8 in neuron maintenance and myelination.
In live Tg(mct8:EGFP/fli:DsRED) double transgenic larvae, co-expression of EGFP and DsRED was observed in both the blood and lymphatic vascular systems. Thus, MCT8 is likely to play a key role in the transport of THs across the vessel membrane and into the nervous system. Colocalization of EGFP and DsRed was also observed in vessels surrounding the brain; however, not in the region where zebrafish blood-brain barrier was identified (52). This finding is consistent with previous reports on chickens (42). In contrast, in rodents, MCT8 mRNA and protein were observed in the blood-brain barrier (53), suggesting that MCT8 is not directly involved in TH transport across the blood-brain barrier in non-mammalian vertebrates. Our expression studies suggest that, as is the case in mammals (6,12,48), OATP1C1 can function in zebrafish blood-brain barrier, and MCT8 can regulate the transport of THs from the blood to the nervous system in distinct regions.
EGFP expression in the Tg(mct8:EGFP) closely mimics the pattern of expression of the endogenous mct8 mRNA, indicating that the mct8 promoter is functional in live animals. The effect of MCT8 on gene transcriptional regulation was previously investigated in fibroblasts derived from AHDS patients (54); however, very little is known about the transcriptional regulation of the mct8 gene. An in vitro study performed on mouse cells revealed the presence of a putative mct8 core promoter containing an SP1 binding site that might be important for mct8 transcriptional regulation (55). Here, we isolated, for the first time, a mct8 promoter that contains distinct regulatory regions and that is functional in vivo. Interestingly, we found that binding sites for the PURA, HDBP, and GATA transcription factors are conserved from fish to humans. Both the PURA and HDBP transcription factors have previously been associated with the development of the nervous system (56,57), and GATA is commonly present in regulatory elements of bloodrelated genes (58). The presence of these putative binding sites is consistent with the expression observed in the nervous and vascular systems, respectively. Promoter-bashing experiments will clarify the specific role of these putative elements in mct8 transcriptional regulation in animal models and humans. The zebrafish promoter and the transgenic lines will serve as powerful tools for functional promoter analysis in the future, thereby providing a platform for the identification of novel transcriptional mechanisms that regulate mct8 expression. Importantly, the promoter isolation in zebrafish could likely aid in the identification of a functional mct8 promoter in humans. This benefit will be a critical step toward applying tissue-specific gene therapy in AHDS patients.
Knockdown of MCT8 did not affect the expression levels of key TH genes: the deiodinases and tsh␤. These results suggest that although the zebrafish MCT8 is able to uptake TH in cell lines (22), it does not regulate deiodinase and tsh␤ gene expression at early developmental stages. In contrast, in MCT8knock-out mice, the mRNA levels of dio1 in the liver and dio2 in the cerebrum significantly increased (9). This discrepancy between zebrafish and mice could be explained by the relatively early developmental stage at which larvae were sampled. The establishment of a MCT8-mutant zebrafish will enable sampling of mRNA from specific tissues in adults rather than from whole larvae. Alternatively, it is possible that protein activity, rather than transcription, was affected by the KD of zebrafish MCT8. Nevertheless, because MCT8-KD had a major effect on embryonic development, it cannot be excluded that MCT8 play an elusive role that is independent of its function as a TH transporter.
The establishment of MCT8-deficient mouse models has significantly advanced our understanding of the endocrinological phenotype linked to AHDS, yet these murine models do not display obvious neurological and psychomotor deficiencies (9). Additional non-mammalian models, such as chicken and frog, were used to study the role of MCT8 in TH transport (2,42). However, although the role of MCT8 in regulating the TH endocrinological system is well characterized, the cause for the AHDS neurological symptoms and the mechanisms underlying MCT8 deficiency remain poorly understood. Here, we showed that KD of MCT8 results in a mild to severe altered developmental phenotype. The injection of both MCT8-MOs into embryos resulted in similar phenotypes. Moreover, the use of the Tg(mct8:EGPF) transgenic line revealed that the development of mct8-expressing cells was altered and caused deformed brain, spinal cord, and notochord. High magnification confocal imaging revealed that KD of MCT8 reduced the number and altered the organization of neural cells in the brain and spinal cord. This robust phenotype may result from insufficient levels of THs in the cerebrospinal fluid, neurons, and glial cells. On the other hand, the development of muscles and vessels was mostly intact and only a weak malformation was observed in these tissues. This neural-specific phenotype suggests that MCT8 plays a key role in the development of the nervous system. In turn, the altered nervous system may cause severe physiological and behavioral abnormalities.
The role of MCT8 was further studied using the overexpression of mRNA. The injection of mct8 mRNA alone, which results in homogenous and ubiquitous expression in the whole embryo, did not seem to cause any abnormal developmental defects. However, it was able to efficiently rescue the phenotype of MCT8-MO-injected embryos. These results strongly suggest that the described phenotype is specific to MCT8 deficiency and indicate that MCT8 does not have a toxic effect on non-MCT8 expressing cells, thus addressing important questions for gene therapy.
This study establishes the zebrafish as a promising model to study the role of MCT8 and the mechanisms underlying AHDS. Future studies using live time lapse imaging (13) in MCT8mutant embryos will enable testing fine changes in neuronal structures and circuit connectivity throughout the developmental stages. Furthermore, as the zebrafish has been increasingly used as a high-throughput genetic model organism for pharmacological screens (59,60), this study may be used as a platform for future screens of therapeutic reagents that may compensate for MCT8 deficiency and aid in the treatment of AHDS patients.