Thyroid-Specific Inactivation of KIF3A Alters the TSH Signaling Pathway

ED’AMICO and others Kif3a and thyroid function 50:3 375–387 Research

Thyroid-speciﬁc inactivation of KIF3A alters the TSH signaling pathway and leads to hypothyroidism

Eva D’Amico1, Ste´phanie Gayral1, Claude Massart2, Jacqueline Van Sande2, Jeremy F Reiter3, Jacques E Dumont2, Bernard Robaye1 and Ste´phane Schurmans1,4,5,6

1Institut de Recherche Interdisciplinaire en Biologie Humaine et Mole´ culaire (IRIBHM), Institut de Biologie et de Me´ decine Mole´ culaires (IBMM), Universite´ Libre de Bruxelles (ULB), Rue des Professeurs Jeener et Brachet 12, 6041 Gosselies, Belgium 2IRIBHM, Universite´ Libre de Bruxelles (ULB), Campus Erasme, Route de Lennik 808, 1070 Brussels, Belgium 3Department of Biochemistry and Biophysics, University of California, Smith Cardiovascular Research Building, Mission Bay Boulevard South 555, San Francisco, California 94158-9001, USA 4Laboratoire de Ge´ ne´ tique Fonctionnelle, GIGA-Research Centre, Universite´ de Lie` ge (ULg), Rue de l’Hoˆ pital 1, Correspondence 4000 Lie` ge, Belgium should be addressed 5Welbio, Universite´ de Lie` ge, Lie` ge, Belgium to S Schurmans 6Secteur de Biochimie Me´ tabolique, De´ partement des Sciences Fonctionnelles, Universite´ de Lie` ge, Boulevard de Email Colonster 20, 4000 Lie` ge, Belgium [email protected]

Abstract

Kinesins, including the kinesin 2/KIF3 molecular motor, play an important role in intracellular Key Words traffic and can deliver vesicles to distal axon terminals, to cilia, to nonpolarized cell surfaces or to " thyroid epithelial cell basolateral membranes, thus taking part in the establishment of cellular polarity. " hypothyroidism We report here the consequences of kinesin 2 motor inactivation in the thyroid of 3-week-old " genetically-modified mouse C C Kif3aD/flox Pax8Cre/ mutant mice. Our results indicate first that 3-week-old Pax8Cre/ mice used " Kif3a Journal of Molecular Endocrinology in these experiments present minor thyroid functional defects resulting in a slight increase in " molecular motor circulating bioactive TSH and intracellular cAMP levels, sufficient to maintain blood thyroxine " kinesin 2 levels in the normal range. Second, Kif3a inactivation in thyrocytes markedly amplified the phenotype observed in Pax8Cre/C mice, resulting in altered TSH signaling upstream of the second messenger cAMP and mild hypothyroidism. Finally, our results in mouse embryonic fibroblasts indicate that Kif3a inactivation in the absence of any Pax8 gene alteration leads to

altered G protein-coupled receptor plasma membrane expression, as shown for the b2 adrenergic receptor, and we suggest that a similar mechanism may explain the altered TSH D/ﬂox Cre/C Journal of Molecular signaling and mild hypothyroidism detected in Kif3a Pax8 mutant mice. Endocrinology (2013) 50, 375–387

Introduction

Little is known about molecular motors responsible hormones, the TSH receptor and the sodium/iodide for intracellular trafﬁc in thyroid follicle cells. However, symporter (NIS) must reach the basolateral surface, like in other polarized cells, thyroid proteins must while thyroperoxidase and the dual oxidases must be speciﬁcally reach the apical or basolateral membrane transported to the thyrocyte apical membrane, and to be functional. In order to assure the synthesis of thyroglobulin (TG) exported through the apical mem-

tri-iodothyronine (T3)andthyroxine(T4)thyroid brane in the colloid.

http://jme.endocrinology-journals.org Ñ 2013 Society for Endocrinology Published by Bioscientiﬁca Ltd. DOI: 10.1530/JME-12-0219 Printed in Great Britain Downloaded from Bioscientifica.com at 09/28/2021 09:28:13PM via free access Research ED’AMICO and others Kif3a and thyroid function 50:3 376

Kinesins play an important role in intracellular traffic. Ethical Committee of the Medical and the Sciences School The superfamily of kinesins is constituted of motor of the Universite´ Libre de Bruxelles. proteins that use energy liberated from ATP hydrolysis to transport vesicles along microtubules. These proteins are Genotyping and analysis of Cre/lox recombination composed of a motor domain containing an ATP-binding specificity site and a microtubule-binding site, a domain that binds vesicles, and an oligomerization domain (Scholey 1996). PCR was performed on tail DNA with oligonucleo- Most of these motor proteins, including kinesin 2, mediate tide primers 1 (50-AGGGCAGACGGAAGGGTGG-30), vesicle transport over long distances to the plus end of 2(50-TCTGTGAGTTTGTGACCAGCC-30), and 3 (50-TGGC microtubules. This transport pathway is highly conserved AGGTCAATGGACGCAG-30). Using these three primers in through evolution. In view of microtubule organization, the same PCR, amplicons of 200, 360, and 490 bp were kinesins can deliver vesicles to distal axon terminals, obtained, corresponding to the delta, wild-type, and to cilia or flagella, to nonpolarized cell surfaces or to floxed Kif3a allele respectively. epithelial cell basolateral membranes, taking part in the establishment of cellular polarity (Hirokawa 2000). Cell culture

Kinesin 2 motor (or kinesin family member 3 (KIF3)) is C C K K Large T antigen-immortalized Kif3a / and Kif3a / composed of a heterotrimeric complex KIF3A–KIF3B– KAP3. KIF3A/KIFB are the motor proteins and KAP3 mouse embryonic fibroblasts (MEFs) were maintained in m regulates the binding between the cargo and the KIF3 DMEM supplemented with 100 U/ml penicillin, 100 g/ml heterodimer (Hirokawa 2000). This molecular motor is streptomycin, 15% fetal bovine serum, 1.1 mM sodium L m expressed in all tissues, including the thyroid. KIF3A or pyruvate, 3.97 mM -glutamine, 100 M nonessential m KIF3B inactivation in mice leads to abolition of molecular amino acids and 300 g/ml G418 (Sigma–Aldrich). Cells motor function and to an early embryonic lethality. were kept in a humidified incubator at 37 8C and 5% CO2. These knockout embryos exhibit loss of cilia ordinarily The medium was changed every 3 days. present on cells of the embryonic node, leading to a randomized establishment of left–right asymmetry and Western blotting numerous structural abnormalities (Nonaka et al. 1998, Thyroid and MEF proteins were extracted using RIPA buffer Marszalek et al. 1999). (Tris–HCl50mM,pH7.4,EDTA1mM,NaCl150mM,Triton In this study, we have explored the role of the kinesin 2 X-100 1%, Sodium deoxycholate 1%, SDS 0.1%). Thyroid Journal of Molecular Endocrinology molecular motor in mouse thyroid development, structure (25 mg) and MEF (60 mg) protein extracts were separated on a and function 3 weeks after birth. We used mice carrying a 10% SDS–PAGE and transferred onto PVDF membrane Kif3a floxed gene as well as knock-in mice expressing the (Amersham Biosciences) using a Mini Trans-Blot Transfer Cre recombinase under the control of the endogenous Cell system (Bio-Rad). Membrane was preincubated for 1 h et al et al Pax8 promoter (Marszalek . 1999, Bouchard . with 5% milk and 0.1% Tween in PBS, followed by overnight Cre/C 2004). Indeed, in these Pax8 mice, the expression of incubation at 4 8C with a Kif3a antibody (Sigma–Aldrich) the recombinase mimics the expression of the paired diluted 1:2000, or with a b-actin antibody (Sigma–Aldrich) domain transcription factor Pax8, including thyroid diluted 1:5000, or with a b2 adrenergic receptor (ADRB2) follicle cells, kidney, inner ear, and the mid–hindbrain antibody (Abcam, Cambridge, UK) diluted 1:1000 in the boundary region, all of them from the embryonic stage same buffer. The membrane was washed in PBS and then E8.5 (Plachov et al. 1990, Bouchard et al. 2002). incubated for 1 h at room temperature with peroxidase- conjugated protein A (Sigma–Aldrich) diluted 1:10 000. Proteins were visualized with Western Lightning plus-ECL Materials and methods (Perkin-Elmer) and exposed to X-ray film. Protein concen- Mice tration was determined with Bio-Rad protein assay reagents using BSA as standard. Kif3aD/flox mice, obtained from LS Goldstein (University of C California, San Diego, USA) and Pax8Cre/ mice, obtained Serum hormone assays from M Busslinger (Research Institute of Molecular Path-

ology, Vienna, Austria) were on a mixed Sv129!C56BL/6 Enzyme immunoassays for mouse T3 and T4 were

genetic background. All procedures were approved by the performed using Mouse/Rat T3 ELISA kit from Calbiotech

(Spring Valley, SK, Canada) and T4 ELISA kit from Immunohistochemistry DiaSource (Louvain-la-Neuve, Belgium) respectively. For TG analysis, sections were deparaffinized and treated TSH was measured in a bioassay using a line of with 0.3% hydrogen peroxide in methanol for 30 min. Chinese hamster ovary cells stably transfected with the They were incubated with blocking solution (10% goat human TSH receptor cDNA, as previously described (Perret serum in PBS) for 1 h and then with a human TG antibody et al. 1990, Moeller et al. 2003). Briefly, 50 000 cells (Dako, Heverlee, Belgium) diluted 1:2500 overnight at were seeded in individual test tubes and incubated for room temperature. 24 h in 100 ml Ham’s F-12 nutrient mixture supplemented For iodinated TG analysis, quenching of endogenous with 10% FCS, 100 IU/ml penicillin, 100 mg/ml strepto- peroxidases was performed, and then antigen retrieval was mycin, 1 mM sodium pyruvate and 2.5 mg/ml Fungizone. done by boiling the slides immersed in citrate buffer at pH 6. Cells were washed with 500 ml Krebs-Ringer HEPES They were incubated with blocking solution (10% BSA/2% buffer (pH 7.4), supplemented with 8 mM glucose and sheep serum in PBS) for 1 h and then with an antibody 0.5 g/l BSA and then preincubated for 30 min in 200 mlof directed against iodinated TG (a kind gift from Prof. thesamemedium.Themediumwasremovedand C Ris-Stalpers, Laboratory of Pediatric Endocrinology, 200 mloffreshbuffercontaining20mlofserumfor Amsterdam, The Netherlands) diluted 1:2000 overnight at 4 8C. TSH measurement and 25 mM Rolipram, a cAMP phos- For NIS analysis, sections were permeabilized for phodiesterase inhibitor, were added. The incubation was 20mininPBS–0.1%TritonX-100andtreatedwith continued for 1 h, at the termination of which the 0.3% hydrogen peroxide for 30 min. They were incubated medium was discarded and replaced with 0.1 M HCl. with blocking solution (1% BSA and 2% goat serum in PBS) cAMP was measured in the dried cell extract by RIA for 30 min and then with an NIS antibody (a kind gift according to the method of Brooker (Brooker et al. 1979). from Prof. Carrasco, Albert Einstein College of Medicine, Blanks were prepared as above but contained 20 mlof NY, USA) diluted 1:500 overnight at room temperature. human serum without TSH. Thus, this assay measures the stimulus to which the cells are exposed, which reflects both the concentration of the hormone and its Semiquantitative PCR specific activity, i.e. its glycosylation level. Total RNA was isolated from thyroid tissues using RNeasy kit (Qiagen) including a DNase treatment and was reverse Histology transcribed with the M-MLV Reverse Transcriptase

Journal of Molecular Endocrinology (Invitrogen) using 300 ng total RNA. The following genes Thyroid glands were removed from 3-week-old mice and were PCR amplified using sense and antisense oligonucleo- overnight fixed at room temperature in 3.7% formaldehyde. tide primers as follows: V-abl Abelson murine They were dehydrated through isopropanol series, cleared in leukemia viral oncogene homolog 1 (Abl (Abl1)) sense, Histoclear (National Diagnostics, Hessle, UK), embedded in 50-TCGGACGTGTGGGCATTT-30 and antisense, 50-CGCAT- paraffin and cut at 7 mm. For cryosectioning, thyroids were GAGCTCGTAGACCTTC-30; Kif3a sense, 50-ATGCCGAT- fixed overnight at 4 8C in 4% paraformaldehyde (PFA) in CAATAAGTCGGAGA-30,andantisense,50-GTTCCC HEPES-buffered saline (HBS), overnight treated at 4 8C in 30% CTCATTTCATCCACG-30;andNis sense, 50-CGCTA sucrose in HBS, embedded in Tissue-Tek R OCT TM CGGTCTCAAGTTTCTG-30 and antisense, 50-CGCAGTTC- compound (Klinipath, Duiven, The Netherlands) and cut at TAGGTACTGGTAGG-30. The reaction mixtures contained m 5 m. For histological analysis, slides were stained with 5 ml PCR buffer 10! (containing 15 mM MgCl2;Qiagen), hematoxylin/eosin (Klinipath). ImageJ Software (Thermo- 1 ml DNTP mix (10 mM each DNTP), 1 ml forward primer Scientific, Rockford, IL, USA) was used to quantify follicle area (10 mM), 1 ml reverse primer (10 mM), 2 U Taq DNA poly- as well as cell and follicle densities from a minimum of five merase(Qiagen), 15 ngtemplate cDNAand RNase-free water mice per genotype. For the follicle area, more than 20 follicles to a final volume of 50 ml. were measured in each thyroid gland. For each mouse, cell and follicle densities per mm2 were estimated by counting Quantitative PCR the number of cells and follicles in a total area of 0.1 mm2. Colloid and epithelial areas and volumes as well as the Total RNA from thyroid tissues was isolated using RNeasy Thyroid Activation Index were obtained as described kit (Qiagen) including a DNase treatment and was reverse (Kmiec et al. 1998). transcribed with the M-MLV Reverse Transcriptase

(Invitrogen) using 120 ng total RNA. Reactions for the and cells were incubated for 1 h at room temperature. quantification of Kif3a, Nis (or Slc5a5), Dio2, Tshr, Gnas, Cells were washed and incubated for 30 min on ice with an Adcy3, Adcy6 and Adcy9, Prkar1a and Prkar1b, Prkar2a and antibody directed against rabbit Ig coupled to Alexa fluor 488 Prkar2b, and Prkaca and Prkacb mRNAs were performed in (dilution 1:1000; Invitrogen). After washing, cells were a CFX96 Real-Time System (Bio-Rad), using SYBR Green as resuspended in FACS buffer and analyzed on a Cytomics detector dye. Porphobilinogen deaminase Hmbs and acidic FC500 cytometer (Beckman Coulter, Brea, CA, USA). ribosomal phosphoprotein P0 (36B4 (Rplp0)) were used as reference genes. The reaction mixtures contained 10 ml Statistical analysis SYBR Green PCR kit (Bio-Rad), 200 nM of each primer, 6 ng template cDNA and RNase-free water to a final All results are expressed as meanGS.E.M. and statistical analysis volume of 20 ml. The sequences of all oligonucleotide was done by unpaired t-test or Mann–Whitney U test, using primers are listed in Supplementary Table 1, see section on GraphPad Prism (GraphPad Software Inc., La Jolla, CA, USA). supplementary data given at the end of this article.

Results Ex vivo thyroid cAMP measurement Kif3a inactivation in mouse thyroid cells Thyroids were washed with cold PBS. Then, they were

incubated for 80 min (37 8C, 5% CO2) with 900 ml The Cre recombinase tissue specificity and efficiency at the C C Krebs-Ringer bicarbonate buffer (KRB, pH 7.4) containing Kif3a locus were investigated in Kif3aflox/ and Kif3aflox/ C 50 mM Rolipram, a cAMP phosphodiesterase inhibitor. Pax8Cre/ mice at the DNA level (Fig. 1A and B). Three Finally, 100 ml of 1 M HCl were added and the incubation oligonucleotide primers were designed to amplify the wild- was continued for 5 min. cAMP was measured by RIA type, flox, and exon 2-deleted (or D)Kif3aallelesinthe C (Brooker et al. 1979). same PCR. As expected, in Kif3aflox/ mice, amplicons of 360 and 490 bp, corresponding to the wild-type and the flox alleles respectively, were amplified out of the DNA In vitro MEF cAMP measurement C extracted from eight different tissues. In Kif3aflox/ C The day before the experiment, MEF cells (105) were seeded Pax8Cre/ mice, an additional 200 bp D signal was in 3.5-cm diameter dishes with 2 ml culture medium. specifically detected after amplification of DNA extracted Before the experiment, the culture medium was removed from the thyroid and the kidney (Fig. 1B). The presence of

Journal of Molecular Endocrinology and cells were rinsed with 1 ml KRB buffer, then pre- the D signal in these two tissues was associated with a incubated for 30 min in this buffer containing 50 mM decreased intensity of the ﬂox allele signal. These results

Rolipram at 37 8C–5% CO2. The buffer was then removed demonstrate that the floxed Kif3a locus is efficiently and replaced with fresh KRB buffer supplemented with recombined by the Cre enzyme. They also confirm previous 50 mM Rolipram and the different agonists: 10 mM iso- report indicating that the Cre recombinase is expressed in C proterenol (Sigma–Aldrich) or 10 mM forskolin (Sigma– thyrocytes and kidney cells of Pax8Cre/ mice. Aldrich). At the end of the 10 min incubation, the medium In order to investigate the role of the molecular motor C C was removed and 1 ml 0.1 M HCl was added to the dishes. component Kif3a in mouse thyrocytes, Kif3aD/ Pax8Cre/ C cAMP was measured by RIA (Brooker et al. 1979). mice were crossed with Kif3aflox/ mice to produce C C C C Kif3aD/flox Pax8Cre/ ‘mutant’ mice, Kif3a / and Kif3aflox/ C C C ‘control’ mice as well as Kif3a / Pax8Cre/ ‘Cre’ mice from Flow cytometry the same litter. RT-PCR, quantitative PCR (qPCR) and The day before the experiment, MEF cells (105) were western blot analyses demonstrated a severely reduced seeded in 3.5-cm diameter dishes with 2 ml culture Kif3a mRNA and protein in the thyroid of mutant mice, as medium. Cells were washed with PBS, incubated for compared with control and Cre mice (Fig. 1C, D and E). 20 min with PBS–EDTA 5 mM on ice and collected. After centrifugation, cells were resuspended in FACS buffer (PBS, Normal thyroid migration but hypothyroidism in 0.1% BSA and 0.1% sodium azide). For the permeabiliza- mutant mice tion, cells were incubated for 30 min in PBS–2% PFA and then for 3 min in PBS–0.1% NP40. A rabbit antibody Mice with all possible genotypes were recovered at directed to ADRB2 (dilution 1/50; Abcam) was added Mendelian frequency at birth, suggesting that Kif3a

A ATG 1 2 3

Wild-type Exon 1 Exon 2 Exon 3 ATG 1 2 3

Flox Exon 1 Exon 2 Exon 3

ATG 1 3 ∆

Exon 1 Exon 3

B Kif3aflox/+ Kif3aflox/+ Pax8cre/+

Tail ThyroidLung KidneySpleen Muscle Liver Testis Tail ThyroidLung KidneySpleen Muscle Liver Testis

506 bp Floc (490 bp) 396 bp Wild-type (360 bp) 344 bp 201 bp ∆ (200 bp)

MW MW MW

CD 1.5 * ** Ctrl Cre Mutant Kif3a 1.0 Journal of Molecular Endocrinology

Abl mRNA (AU) 0.5 Kif3a

0.0 Ctrl Cre Mutant E

Ctrl Cre Mutant

Kif3A (85 kDa)

β-Actin (42 kDa)

Figure 1 Strategy to inactivate Kif3a in thyroid follicle cells. (A) Mice with a wild- and the kidney. MW: molecular weight. Kif3a mRNA identification and type, a floxed, and/or a delta (D) Kif3a allele were used in the study. Black quantification by (C) RT-PCR and (D) qPCR performed on thyroid cDNA from boxes, exons of the Kif3a gene; triangles, loxP sites; arrowheads, primers 3-week-old mice. The Abl gene was used as control in the RT-PCR used to identify the mouse genotype in PCR. (B) Gel electrophoresis of a experiments. In qPCR, meanGS.E.M. of five independent experiments are PCR performed on DNA extracted from different tissues of Kif3aflox/C and represented. AU, arbitrary units. Statistics (t-test with Welch’s correction): C C Kif3aflox/ Pax8Cre/ mice, using the three primers shown in (A). Amplicons *P!0.05 and **P!0.01. (E) Western blot analysis of thyroid protein of 490, 360, and 200 bp are identified and correspond to the floxed, wild- extracts from 3-week-old normal, Cre, and mutant mice with a Kif3a type, and D alleles respectively. The D allele is solely present in the thyroid antibody. b-Actin served as loading control.

deletion in Pax8-expressing cells is not embryonically mutant mice, 787G33 mm2; nZ5 mice per group; PZ0.1770) lethal. Nevertheless, at weaning around 3 weeks of age, all and epithelial volume (Cre mice, 74G21 mm3;mutant mutant mice died. Dying mutant mice presented mild mice, 42G3mm3; nZ5 mice per group; PZ0.0952) was growth retardation and autopsy revealed a marked observed in mutant thyroids, but the differences did not increase in kidney weight and size, compared with control reach statistical significance when compared with Cre and Cre mice (Supplementary Figure 1A, B, C and D, see thyroids. A similar Thyroid Activation Index was found section on supplementary data given at the end of this in Cre and mutant mice (Cre mice, 1.891G0.2952; article). Histological analysis demonstrated the presence mutant mice, 2.090G0.3541; nZ5 mice per group; of numerous and large cysts in the kidneys of mutant PZ0.6772). As compared with control and Cre thyroids, mice, providing a probable cause for their reduced Tg mRNA and protein were normal in mutant thyroid, C survival (Supplementary Figure 1E). Indeed, in Pax8Cre/ but iodinated TG level in the colloid was significantly mice, the Cre recombinase is also expressed in kidney decreased in mutant mice, showing a very heterogeneous cells, and this was confirmed by the presence of severely expression in the different follicles (Fig. 3B and C, data decreased Kif3a mRNA and protein levels in mutant not shown). In order to define a potential mechanism of kidneys (Supplementary Figure 1F and G). No obvious this iodination defect, the expression of NIS, the NIS alteration was detected in other organs, including the responsible for the basal membrane iodide uptake, was brain (data not shown). In 3-week-old mutant mice, the investigated at the protein and mRNA levels (Fig. 3D, E thyroid was normally localized in the neck and was of and F). NIS protein expression and mRNA level were normal size and weight (Fig. 2A, data not shown). Mutant markedly decreased in mutant mice, as compared with

mice displayed hypothyroidism with low total T4 levels control and Cre mice, reaching about 20% of control C and increased TSH bioactivity, as compared with control level and 40% of Cre level. As expected from Pax8Cre/ and Cre mice (Fig. 2B and C). No difference was observed mice, which are heterozygous at the Pax8 locus, a slight

in total T3 levels between groups of mice (Fig. 2B). but significant decrease in both NIS protein and mRNA Interestingly, Cre mice, which are knock-in mice with a levels was observed in these mice. heterozygous insertion of the Cre cDNA into exon 3 of the Together, our results define a marked thyroid dysfunc- Pax8 gene, presented a mild but significant increase in TSH tion in mutant mice. This dysfunction is probably a bioactivity, as compared with control mice. However, in consequence of the nearly complete Kif3a inactivation these mice, TSH bioactivity was still markedly below the combined with the partial Pax8 gene inactivation, since level observed in mutant mice (Fig. 2C). Together, our mild thyroid alterations are already present in the C C Journal of Molecular Endocrinology results indicate that Kif3a inactivation using Pax8Cre/ Pax8Cre/ mice used to inactivate Kif3a in thyroid follicle mice results in early lethality, probably a consequence of cells. They also suggest a potential role of the TSH/cAMP kidney failure. Importantly, they indicate that Kif3a signaling pathway in the pathogenesis of the mutant inactivation in thyroid follicle cells leads to mild hypothyr- phenotype since Nis (Slc5a5) mRNA expression, which is oidism. They also suggest that, on the genetic background altered in mutant mice, is known to be regulated by this C studied, Pax8Cre/ mice have a mild thyroid dysfunction pathway in thyroid follicle cells. leading to a slight increase in bioactive TSH level.

Altered TSH signaling pathway in mutant mice Normal thyroid structure, but reduced TG iodination and Since Nis mRNA was found to be signiﬁcantly decreased in NIS expression in the thyroid of mutant mice mutant thyroid, mRNA levels of other genes known to be Histological examination of 3-week-old thyroids revealed involved in or induced by the TSH/cAMP signaling that the general thyroid architecture as well as the follicle pathway were investigated by qPCR in the thyroid of and cell densities was normal in Cre and mutant mice mutant, control, and Cre mice (Figs 4 and 5). Out of the (Figs 2D and 3A). A trend for decreased follicular area genes tested, only the mRNA levels of Adcy3 (the adenylate (Cre mice, 2049G428 mm2; mutant mice, 1394G135 mm2; cyclase 3 gene), Prkar2b (the PKA regulatory subunit 2b nZ5 mice per group; PZ0.2186), colloid area (Cre mice, gene), and Dio2 (the type 2 iodothyronine deiodinase 927G230 mm2; mutant mice, 607G111 mm2; nZ5 mice gene) were found signiﬁcantly decreased in mutant per group; PZ0.2469), colloid volume (Cre mice, thyroid, as compared with control and Cre thyroids. 46G17 mm3; mutant mice, 25G7mm3; nZ5 mice per Expression of the Dio1 gene was decreased both in Cre group; PZ0.2742), epithelial area (Cre mice, 1121G202 mm2; and mutant thyroids, suggesting that Dio1 mRNA may be

A Ctrl Cre Mutant

Tr Tr

B Total T4 Total T3 *** 8 2.5 *** 2.0 6 1.5 g/dl) 4 µ (ng/dl) ( 1.0 3 4 T T 2 0.5

0 0.0 Ctrl Cre Mutant Ctrl Cre Mutant

C TSH bioactivity *** 10 *** *** 8 6

TSH 4 2 (pmol cAMP/tube) 0 Ctrl Cre Mutant

Journal of Molecular Endocrinology D 400 15 000 2

300 2 10 000 200 5000

100 Cell per mm Follicles per mm Follicles 0 0 Ctrl Cre Mutant Ctrl Cre Mutant

Figure 2

Normal thyroid migration but hypothyroidism in mutant mice. (A) Sections mutant (nZ52) mice. MeansGS.E.M. are represented. Statistics through the neck of 3-week-old control (Ctrl), Cre, and mutant mice (Mann–Whitney U test), ***P!0.001. (D) Cell and follicle densities were showing the left and right lobes of the thyroid (arrows) and the trachea (Tr). determined in thyroids of 3-week-old control (nZ5–11), Cre (nZ5), and

(B) Total T4 in the blood of 3-week-old control (nZ71), Cre (nZ59), and mutant (nZ5–13) mice. MeansGS.E.M. are represented. No statistical

mutant (nZ46) mice and total T3, control (nZ29), Cre (nZ20), and mutant difference was observed between control, Cre, and mutant mice. Full colour (nZ17) mice and (C) TSH bioactivity: control (nZ75), Cre (nZ63) and version of this ﬁgure available via http://dx.doi.org/10.1530/JME-12-0219.

controlled directly or indirectly by PAX8 transcription among the genes tested, only Nis, Adcy3, Prkar2b, and factor. By contrast, mRNA levels for the TSH receptor, Gas, Dio2 mRNAs expression are known to be regulated by Adcy6, and Adcy9, the PKA a and b catalytic subunits as the second messenger cAMP, through CREB transcription well as for the Prkar1a, Prkar1b, and Prkar2a were similar in factor. Thus, we next investigated basal cAMP levels mutant, Cre, and control thyroids (Fig. 5). Interestingly, in mutant, Cre, and control thyroids (Fig. 6). In Cre

Hematoxylin/eosin Ctrl Cre Mutant

B TG

Ctrl Cre Mutant

Iodinated TG Ctrl Cre Mutant

D NIS

Ctrl Cre Mutant

E F Ctrl Cre Mutant 1.5 *** NIS ** 1.0 Abl

Journal of Molecular Endocrinology 0.5 mRNA (AU) Nis 0.0 Ctrl Cre Mutant

Figure 3 Normal histological structure, but decreased iodinated thyroglobulin and identification and quantification by (E) RT-PCR and (F) qPCR performed on NIS expression in mutant thyroid. (A) Hematoxylin/eosin-stained sections of thyroid cDNA from 3-week-old control, Cre, and mutant mice. The Abl gene thyroid from 3-week-old control (Ctrl), Cre, and mutant mice. Bars, 100 mm. was used as control in the RT-PCR experiments. In qPCR, meansGS.E.M.of Insets show follicles at higher magnification (!40). Immunodetection of five independent experiments are represented. AU: arbitrary units. (B) TG, (C) iodinated TG, and (D) NIS proteins on thyroid sections from 3- Statistics (unpaired t-test): *P!0.05 and ***P!0.001. Full colour version of week-old control, Cre, and mutant mice. Bars, 50 mm. Nis mRNA this figure available via http://dx.doi.org/10.1530/JME-12-0219.

thyroids, basal cAMP levels were signiﬁcantly increased, cAMP levels in mutant mice were found only slightly as compared with control thyroids. This is most probably increased when compared with control mice, and severely the consequence of the increased TSH bioactivity observed decreased when compared with Cre mice (Fig. 6). in these mice (Fig. 2C). Based on this result, basal cAMP Together, our results suggest that mutant mice present levels should be markedly increased in mutant thyroids, a defect in the proximal part of the TSH/TSH receptor/ since TSH bioactivity was about seven- and fourfold cAMP signaling pathway. They also conﬁrm that the C increased in these mice compared with control and Cre Pax8Cre/ mice used to induce recombination have a mice respectively. However, in agreement with a defect in thyroid dysfunction, which is compensated by increased the proximal part of the TSH/cAMP signaling pathway, TSH bioactivity.

** A 1.5 1.5 Ctrl * NS Cre 1.0 Mutant 1.0

0.5 mRNA (AU) mRNA (AU) 0.5 Dio1 0.0

0.0 Tshr Gnas1 Ctrl Cre Mutant B 2.0 ** NS 1.5 1.5 * *

1.0 * 1.0 mRNA (AU) 0.5 mRNA (AU) 0.5 0.0 Dio2 Adcy3 Adcy6 Adcy9

0.0 C 1.5 Ctrl Cre Mutant

1.0 Figure 4 Reduced Dio1 and Dio2 mRNA expression in thyroid of mutant mice. mRNA encoding Dio1 and Dio2 were quantified by qPCR performed on thyroid 0.5 cDNA from 3-week-old control, Cre, and mutant mice. MeansGS.E.M. of four mRNA (AU) or five independent experiments are represented. AU, arbitrary units. Statistics (unpaired t-test or Mann–Whitney U test): NS, nonsignificant; PO0.05, *P!0.05 and **P!0.01. Dio1, type 1 iodothyronine deiodinase; 0.0 Journal of Molecular Endocrinology Dio2, type 2 iodothyronine deiodinase. R1a R2a Ca

Abnormal plasma membrane G protein-coupled D 1.5 K/K * receptor transport in Kif3a MEFs * The faint TSH receptor expression on mouse thyroid 1.0 follicle cells and/or the absence of adequate antibody directed against this receptor did not allow us to clearly 0.5 localize or quantify the TSH receptor on the follicle cell mRNA (AU) basal membrane and to validate in vivo our hypothesis concerning the defect in the proximal part of the TSH/TSH 0.0 receptor/cAMP signaling pathway in mutant mice (data R1b R2b Cb not shown). We thus used a less complex cellular model K K to probe this hypothesis: Kif3a / MEF infected with Figure 5 RFP-tagged TSH receptor-expressing lentiviruses. Unfortu- Reduced Adcy3 and Prkar2b mRNA expression in thyroid of mutant mice. nately, infections demonstrated that both RFP-TSH mRNA encoding (A) Tshr and Gnas1, (B) Adcy3, 6, and 9 as well as (C) the a and (D) b subunits of Prkar1 (R1a and R1b), Prkar2 (R2a and R2b), and Prkac receptor mRNA and protein were very significantly less (Ca and Cb) were quantified by qPCR performed on thyroid cDNA from K/K C/C expressed in Kif3a than in Kif3a MEF, preventing 3-week-old control (Ctrl), Cre, and mutant mice. MeansGS.E.M. of four or the use of lentivirus to probe our hypothesis (data not five independent experiments are represented. AU, arbitrary units. Statistics (unpaired t-test or Mann–Whitney U test), *P!0.05. Tshr, TSH shown). As an experimental surrogate, we decided to study receptor; Gnas1, protein Gas; Adcy, adenylate cyclase; Prkar, protein kinase the subcellular localization and signaling downstream of A regulatory subunit; Prkac, protein kinase A catalytic subunit.

NS In man, at least ten different mutations have been 6 identiﬁed in the PAX8 gene, leading to thyroid dysgenesis ** and congenital hypothyroidism (Meeus et al. 2004,

g protein Montanelli & Tonacchera 2010). All affected people are µ 4 heterozygous for the PAX8 mutation and transmission is autosomal dominant (Meeus et al. 2004, Montanelli & Tonacchera 2010). Unexpectedly, no structural or 2 functional thyroid defect has been previously reported in C Pax8Cre/ mice, suggesting that, in this species, the cAMP (pmol)/100 presence of one functional Pax8 allele is sufficient for 0 normal organogenesis and function (Amendola et al. Ctrl Cre Mutant 2005). Unlike published data on these mice, we show C here that Pax8Cre/ mice present minor thyroid functional Figure 6 Altered basal cAMP level in thyroid of contrl (Ctrl), Cre, and mutant mice. defects, including a significant decrease in Nis mRNA Basal cAMP levels were analyzed in 3-week-old control (nZ7), Cre (nZ6), expression, resulting in a slight compensatory increase in and mutant (nZ6) thyroids. MeansGS.E.M. of cAMP levels per 100 mgof circulating TSH and intracellular cAMP levels, sufficient to thyroid protein extract are represented. Statistics (Mann–Whitney U test): NS, nonsignificant; PO0.05 and *P!0.05. maintain blood T4 levels in the normal range. The discrepancy between our results and those reported by Amendola (normal TSH and T levels) may be explained by another G protein-coupled receptor (GPCR) naturally 4 a difference in the mouse genetic background and/or in expressed by MEF: the ADRB2. Indeed, this receptor dietary levels of iodine. These findings certainly compli- belongs to the same GPCR subclass as the TSH receptor C C K K cate the interpretation of our results on the role of Kif3a in and is similarly expressed in Kif3a / and Kif3a / MEF the thyroid. (Fig. 7A). Despite the similar protein expression when It has been extensively demonstrated that the tested by western blotting on MEF protein extracts, a kinesin 2 molecular motor, including Kif3a, is essential significantly reduced ADRB2 expression was detected by K K for correct primary cilium assembly (Rosenbaum & Witman flow cytometry at the cell surface in Kif3a / cells 2002). Indeed, Kif3a participates in the intraflagellar (Fig. 7B). Therefore, in response to stimulation by transport in this antenna-like structure and Kif3a inacti- isoproterenol, an ADRB2 agonist, significantly lower K K vation in mice leads to the absence of primary cilium at the cAMP levels were observed in Kif3a / MEF, as compared Journal of Molecular Endocrinology C C cell surface and developmental alterations characteristic of with Kif3a / MEF (Fig. 7C). Altogether, these results ciliopathies (Marszalek et al. 1999, Lin et al.2003). suggest that the absence of Kif3a leads to alterations in Antenna-like structures related to primary cilium have b2 adrenergic GPCR transport to the plasma membrane been detected on thyroid cells during development and on and to defective downstream signaling. mature follicular cells, but major differences exist between species (Sobrinho-Simo˜es & Johannessen 1981). In adult mice, it has been reported that only 2% of thyroid cells Discussion present a cilium, and its precise role has not been addressed so far (Wetzel & Wollman 1969). We were Cre/C We report here for the first time that 3-week-old Pax8 unable to detect an acetylated a-tubulin-positive primary mice present minor thyroid functional alterations and cilium on thyroid cells in mouse at embryonic or adult that inactivation of Kif3a, a kinesin 2 molecular motor stages (data not shown). However, we show here that Kif3a subunit, with these Cre mice results in a mild hypo- inactivation from E8.5 in the thyroid bud had no obvious thyroidism at 3 weeks of age, probably secondary to defects effect on thyroid development and migration, suggesting in TSH receptor signaling. This mild hypothyroidism is that Kif3a and potential primary cilium have no essential characterized by a decreased circulating T4 and an increased role in these processes, at least in mice. TSH level, a probable consequence of a reduced production Correct cell-surface expression of receptors, including

of intracellular T3 from T4 in pituitary cells. Our results TSH and b2 adrenergic GPCR, is essential for binding of also indicate that a lack of Kif3a in cells leads to decreased water-soluble ligands and for initiating downstream cell surface expression of the endogenous ADRB2 and a intracellular signaling. In MEF in the absence of any decreased signaling in response to stimulation. Pax8 gene alteration, our results clearly indicate that Kif3a

ADRB2 expression A MEF Kif3a+/+ Kif3a–/– 1.5 HSP90 (90 kDa) 1.0

0.5 ADRB2

(65/55 kDa) Expression (AU) 0.0 Kif3a+/+ Kif3a–/–

B Non-permeabilized cells Permeabilized cells

–/– Kif3a+/+ Kif3a–/– Kif3a+/+ Kif3a Cell count Cell count Cell count Cell count

ADRB2ADRB2 ADRB2 ADRB2

ADRB2 surface expression ADRB2 total expression 150 150 *** NS 100 100

50 (% ctrl) 50 (% ctrl) Mean fluorescence

Mean fluorescence 0 0 Kif3a+/+ Kif3a–/– Kif3a+/+ Kif3a–/–

C 4000 *** Non-stimulated µ 3000 Isoproterenol 10 M Forskolin 10 µM g protein)

µ 2000 cAMP

1000 (pmol/100 0 Kif3a+/+ Kif3a–/–

Journal of Molecular Endocrinology Figure 7 C/C K/K Reduced surface expression and signaling response of b2 adrenergic stimulation were determined in Kif3a and Kif3a cells (nZ9). receptor in Kif3aK/K cells. Total cellular and cell surface expression of the Statistics (Mann–Whitney U test): NS, nonsigniﬁcant; PO0.05 and

b2 adrenergic receptor were respectively analyzed (A) by western blotting ***P!0.001. Adrb2, b2 adrenergic receptor. Full colour version of this (nZ3) and (B) by ﬂow cytometry (nZ8). Red, ADRB2; blue, negative ﬁgure available via http://dx.doi.org/10.1530/JME-12-0219. control. (C) cAMP levels in response to isoproterenol or forskolin

inactivation leads to significantly lower ADRB2 expression from the cytosol to the plasma membrane (Imamura et al. at the cell surface and, as a consequence, to markedly 2003, Wiesner et al. 2010, Hanania et al. 2012, Lin et al. decreased downstream signaling in response to agonist 2012). Unfortunately, we were unable to analyze the stimulation. The kinesin 2 molecular motor is known to intracellular traffic or the cell-surface expression of the participate in ‘non-intraflagellar’ transport in the cell, like TSH receptor in vivo or in vitro due to the lack of specific retrograde traffic between the Golgi and the endoplasmic antibody working on tissue sections (and probably also to reticulum (Stauber et al. 2006), endosome and lysosome the very low expression of the TSH receptor at the transports (Bananis et al. 2004, Brown et al. 2005), thyrocyte surface) and the unexpected consequence of endocytosis and recycling of cell-surface receptors like Kif3a inactivation on TSH receptor expression directed by transferrin, cubulin, and megalin receptors, or of other lentivirus infection. Given that the TSH and the ADRB2s C K proteins like Clc-5, the H /Cl exchange transporter belong to the same subclass of GPCR and have similar (Schonteich et al. 2008, Reed et al. 2010). The KIF3 posttranslational modifications (glycosylation, sialyla- molecular motor also plays an important role in the tion, and palmitoylation), we speculate here that they transport of vesicles containing GluR2 and GLUT4 use similar mechanisms for transport from the Golgi to the receptors or MT1-MMP and MMP9 metalloproteinases plasma membrane and/or for recycling at the cell surface.

Thus, based on our in vitro results on the ADRB2 in Kif3a- In conclusion, our results in 3-week-old Kif3aD/flox C C deficient MEF, we suggest that Kif3a inactivation in Pax8Cre/ mutant mice indicate first that the Pax8Cre/ thyrocytes may lead to altered transport and/or recycling mice used in these experiments present minor thyroid of the TSH receptor, decreased expression at the baso- functional defects resulting in a slight increase in lateral membrane and altered TSH signaling. TSH signaling circulating bioactive TSH and intracellular cAMP levels, D/flox Cre/C is clearly altered in Kif3a Pax8 mutant mice. First presumably allowing to maintain blood T4 levels in the the expression of specific genes controlled by cAMP normal range. Second, Kif3a invalidation in thyrocytes C through CREB transcription factors, like NIS, Adcy3, markedly amplifies the phenotype observed in Pax8Cre/ Prkar2b, and Dio2, is significantly decreased in these mice, resulting in an altered TSH signaling upstream of the mice. Secondly, as compared with thyroid of Cre mice, the second messenger cAMP and a mild hypothyroidism cAMP levels in mutant thyroid are markedly lower. 3 weeks after birth. Finally, our results in MEF indicate Thereby mutant mice are not able to compensate a thyroid that Kif3a inactivation in the absence of any Pax8 gene

dysfunction despite of increased TSH bioactivity, unlike alteration leads to altered b2 adrenergic GPCR plasma Cre mice. Our mutant phenotype is indeed reminiscent of membrane expression; we suggest that a similar the phenotype observed in TSH receptor-deficient mice. In mechanism may explain the altered TSH signaling and C the latter mice, normal thyroid development is also mild hypothyroidism observed in Kif3aD/flox Pax8Cre/ associated with TSH resistance, decreased Nis mRNA mutant mice. In future genetic studies on congenital expression and congenital hypothyroidism (Postiglione hypothyroidism or high TSH serum level, it will be et al. 2002). However, our mutant phenotype is less important to look for variants in the Kif3a gene. pronounced than the TSH receptor-deficient phenotype. K K As for the ADRB2 in Kif3a / MEF, it is probable that a significant fraction of the TSH receptor is still expressed at Supplementary data the thyrocyte cell surface and is functional. Partial This is linked to the online version of the paper at http://dx.doi.org/10.1530/ JME-12-0219. compensation of kinesin 2 deficiency by other kinesins, like Kinesin 1 which is expressed in thyrocytes (data not shown),tocarrycargototheirfinaldestinationis Declaration of interest abundantly described in the literature (Brown et al. 2005, The authors declare that there is no conflict of interest that could Hirokawa et al. 2009, Wiesner et al. 2010). It is noteworthy be perceived as prejudicing the impartiality of the research reported. here that mice were analyzed only at 3 weeks of age, given

Journal of Molecular Endocrinology the experimental difficulties to record similar data earlier C in mice. The thyroid phenotype of Kif3aD/flox Pax8Cre/ Funding mutant mice is thus unknown at earlier developmental This work was supported by the Fonds de la Recherche Scientifique-FNRS (FRS-FNRS to S S), the Fonds de la Recherche Scientifique Me´ dicale (FRSM to stages and it could eventually change with age, as reported S S), the Fondation Rose et Jean Hoguet and the Fonds David et Alice Van in type 3 deiodinase knockout mice (Hernandez et al. Buuren (to E D). 2006). Furthermore, the Kif3a mutation analyzed here is not strictly thyrocyte specific and the mice have a severe kidney phenotype, which probably explains the postnatal Author contribution statement E D is researched data, contributed to discussion and wrote manuscript. lethality. Severe chronic kidney disease in man results in S G and C M are researched data. J V S, J E D and B R are contributed decreased circulating T4,T3, and TSH levels, a consequence to discussion. J F R is supplied material. S S is designed project, of changes in peripheral hormone metabolism, thyroid contributed to discussion and wrote manuscript. hormone-binding proteins, and central defects (van Hoek & Daminet 2009). Although the consequences of chronic kidney disease have not been clearly reported in the Acknowledgements We thank D Christophe for discussion, S Costagliola for antibodies, literature for 3-week-old mice, the human phenotype is L S Goldstein for the Kif3aD/flox mice and M Busslinger for the D/flox Cre/C not identical to the Kif3a Pax8 mouse pheno- Pax8Cre/C mice.

type where only T4 level is decreased and where increased TSH level is associated with low thyroid cAMP level and with altered expression of speciﬁc genes controlled References by cAMP, which together sign the presence of TSH Amendola E, De Luca P, Macchia PE, Terracciano D, Rosica A, Chiappetta G, signaling defects. Kimura S, Mansouri A, Affuso A, Arra C et al. 2005 A mouse model

demonstrates a multigenic origin of congenital hypothyroidism. Moeller LC, Kimura S, Kusakabe T, Liao XH, Van Sande J & Refetoff S 2003 Endocrinology 146 5038–5047. (doi:10.1210/en.2005-0882) Hypothyroidism in thyroid transcription factor 1 haploinsufficiency Bananis E, Nath S, Gordon K, Satir P, Stockert RJ, Murray JW & Wolkoff AW is caused by reduced expression of the thyroid-stimulating hormone 2004 Microtubule-dependent movement of late endocytic vesicles receptor. Molecular Endocrinology 17 2295–2302. (doi:10.1210/me. in vitro: requirements for Dynein and Kinesin. Molecular Biology of the 2003-0175) Cell 15 3688–3697. (doi:10.1091/mbc.E04-04-0278) Montanelli L & Tonacchera M 2010 Genetics and phenomics of Bouchard M, Souabni A, Mandler M, Neubu¨ser A & Busslinger M 2002 hypothyroidism and thyroid dys- and agenesis due to PAX8 and TTF1 Nephric lineage specification by Pax2 and Pax8. Genes and Development mutations. Molecular and Cellular Endocrinology 322 64–71. (doi:10. 16 2958–2970. 1016/j.mce.2010.03.009) Bouchard M, Souabni A & Busslinger M 2004 Tissue-specific expression of Nonaka S, Tanaka Y, Okada Y, Takeda S, Harada A, Kanai Y, Kido M & cre recombinase from the Pax8 locus. Genesis 38 105–109. (doi:10. Hirokawa N 1998 Randomization of left-right asymmetry due to loss of 1002/gene.20008) nodal cilia generating leftward flow of extraembryonic fluid in mice Brooker G, Harper JF, Terasaki WL & Moylan RD 1979 Radioimmunoassay lacking KIF3B motor protein. Cell 95 829–837. (doi:10.1016/S0092- of cyclic AMP and cyclic GMP. Advances in Cyclic Nucleotide Research 10 8674(00)81705-5) 1–33. Perret J, Ludgate M, Libert F, Gerard C, Dumont JE, Vassart G & Brown CL, Maier KC, Stauber T, Ginkel LM, Wordeman L, Vernos I & Parmentier M 1990 Stable expression of the human TSH receptor in Schroer TA 2005 Kinesin-2 is a motor for late endosomes and CHO cells and characterization of differentially expressing clones. lysosomes. Traffic 6 1114–1124. Biochemical and Biophysical Research Communications 171 1044–1050. Hanania R, Song Sun H, Xu K, Pustylnik S, Jeganathan S & Harrison RE 2012 Plachov D, Chowdhury K, Walther C, Simon D, Guenet JL & Gruss P 1990 Classically activated macrophages use stable microtubules for matrix Pax8, a murine paired box gene expressed in the developing excretory metalloproteinase-9 (mmp-9) secretion. Journal of Biological Chemistry system and thyroid gland. Development 110 643–651. 287 8468–8483. (doi:10.1074/jbc.M111.290676) Postiglione MP, Parlato R, Rodriguez-Mallon A, Rosica A, Mithbaokar P, Hernandez A, Martinez ME, Fiering S, Galton VA & St Germain D 2006 Maresca M, Marians RC, Davies TF, Zannini MS, De Felice M et al. Type 3 deiodinase is critical for the maturation and function of the 2002 Role of the thyroid-stimulating hormone receptor signaling in thyroid axis. Journal of Clinical Investigation 116 476–484. (doi:10.1172/ development and differentiation of the thyroid gland. PNAS 99 JCI26240) 15462–15467. (doi:10.1073/pnas.242328999) Hirokawa N 2000 Stirring up development with the heterotrimeric kinesin Reed AA, Loh NY, Terryn S, Lippiat JD, Partridge C, Galvanovskis J, KIF3. Traffic 1 29–34. (doi:10.1034/j.1600-0854.2000.010105.x) Williams SE, Jouret F, Wu FT, Courtoy PJ et al. 2010 CLC-5 and KIF3B Hirokawa N, Noda Y, Tanaka Y & Niwa S 2009 Kinesin superfamily motor interact to facilitate CLC-5 plasma membrane expression, endocytosis, proteins and intracellular transport. Nature Reviews. Molecular Cell and microtubular transport: relevance to pathophysiology of Dent’s Biology 10 682–696. (doi:10.1038/nrm2774) disease. American Journal of Physiology. Renal Physiology 298 F365–F380. Imamura T, Huang J, Usui I, Satoh H, Bever J & Olefsky JM 2003 Insulin- (doi:10.1152/ajprenal.00038.2009) induced GLUT4 translocation involves protein kinase C-l-mediated Rosenbaum JL & Witman GB 2002 Intraflagellar transport. Nature Reviews. functional coupling between Rab4 and the motor protein kinesin. Molecular Cell Biology 3 813–825. (doi:10.1038/nrm952) Molecular and Cellular Biology 23 4892–4900. (doi:10.1128/MCB.23.14. Scholey JM 1996 Kinesin-II, a membrane traffic motor in axons, axonemes, 4892-4900.2003) and spindles. Journal of Cell Biology 133 1–4. (doi:10.1083/jcb.133.1.1) Kmiec Z, Kotlarz G, Smiechowska B & Mysliwski A 1998 The effect of fasting Schonteich E, Wilson GM, Burden J, Hopkins CR, Anderson K, and refeeding on thyroid follicule structure and thyroid hormone levels Goldenring JR & Prekeris R 2008 The Rip11/Rab11-FIP5 and kinesin in young and old rats. Archives of Gerontology and Geriatrics 26 161–175. II complex regulates endocytic protein recycling. Journal of Cell Journal of Molecular Endocrinology (doi:10.1016/S0167-4943(97)00040-X) Science 121 3824–3833. (doi:10.1242/jcs.032441) Lin F, Hiesberger T, Cordes K, Sinclair AM, Goldstein LS, Somlo S & Sobrinho-Simo˜es M & Johannessen JV 1981 Scanning electron microscopy Igarashi P 2003 Kidney-specific inactivation of the KIF3A subunit of of the normal human thyroid. Journal of Submicroscopic Cytology kinesin-II inhibits renal ciliogenesis and produces polycystic kidney 13 209–222. disease. PNAS 100 5286–5291. (doi:10.1073/pnas.0836980100) Stauber T, Simpson JC, Pepperkok R & Vernos I 2006 A role for Lin Y, Jones BW, Liu A, Vazque´z-Chona FR, Lauritzen JS, Ferrell WD & kinesin-2 in COPI-dependent recycling between the ER and the Marc RE 2012 Rapid glutamate receptor 2 trafficking during retinal Golgi complex. Current Biology 16 2245–2251. (doi:10.1016/j.cub. degeneration. Molecular Neurodegeneration 7 7. (doi:10.1186/1750-1326- 2006.09.060) 7-7) Van Hoek I & Daminet S 2009 Interactions between thyroid and kidney Marszalek JR, Ruiz-Lozano P, Roberts E, Chien KR & Goldstein LS 1999 Situs function in pathological conditions of these organ systems. General and inversus and embryonic ciliary morphogenesis defects in mouse Comparative Endocrinology 160 205–2015. (doi:10.1016/j.ygcen.2008. mutants lacking the KIF3A subunit of kinesin-II. PNAS 96 5043–5048. 12.008) (doi:10.1073/pnas.96.9.5043) Wetzel BK & Wollman SH 1969 Fine structure of a second kind of thyroid Meeus L, Gilbert B, Rydlewski C, Parma J, Roussie AL, Abramowicz M, follicle in the C3H mouse. Endocrinology 84 563–578. (doi:10.1210/ Vilain C, Christophe D, Costagliola S & Vassart G 2004 Character- endo-84-3-563) ization of a novel loss of function mutation of PAX8 in a familial case of Wiesner C, Faix J, Himmel M, Bentzien F & Linder S 2010 KIF5B and congenital hypothyroidism with in-place, normal-sized thyroid. KIF3A/KIF3B kinesins drive MT1-MMP surface exposure, CD44 Journal of Clinical Endocrinology and Metabolism 89 4285–4291. (doi:10. shedding, and extracellular matrix degradation in primary macro- 1210/jc.2004-0166) phages. Blood 116 1559–1569. (doi:10.1182/blood-2009-12-257089)

Received in ﬁnal form 8 March 2013 Accepted 19 March 2013 Accepted Preprint published online 19 March 2013