UvA-DARE (Digital Academic Repository)
Novel thyroid specific transcripts identified by SAGE: implication for congenital
hypothyroidism
Moreno Navarro, J.C.
Publication date
2003
Link to publication
Citation for published version (APA):
Moreno Navarro, J. C. (2003). Novel thyroid specific transcripts identified by SAGE:
implication for congenital hypothyroidism.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible.
CHAPTERR 5
Cloningg of NM41, a novel cystine-knot like protein
preferentiallyy expressed in the thyroid.
Joséé C. Moreno 1. Remco Keijser . Truus Veen boer , Jos B.G. Mulder Fiebo.. J.W. ten Kate 2. Jan J.M. de Vijlder 1 Carolyn Ris-Stalpers 1.
'' Laboratory of Pediatric Endocrinology. Emma Children's Hospital.
AcademicAcademic Medical Center. University of Amsterdam. The Netherlands. zz
Department of Pathology. Academic Medical Center. UniversityUniversity of Amsterdam. The Netherlands.
Cloningg of Ine NM41 gene
I n t r o d u c t i o n n
Congenitall hypothyroidism (CH) is the most common inborn endocrinopathy [1], If untreatedd or not optimally treated, it will result in severe and irreversible damage to thee development of the central nervous system. In 85% of cases, CH is caused by defectss at the thyroidal level (thyroidal CH) while in 15% of cases a defect at the pituitary/hypothalamicc level is causally related to the disease (central CH). There aree 2 main causes of thyroidal CH: dysgenesis, defects originated during the embryonicc development of the thyroid gland affecting its normal formation and/or migrationn from the primitive foregut and dyshormonogenesis, defects affecting the processs of thyroid hormone synthesis. Dysgenesis of the thyroid gland is responsiblee for most cases of thyroidal CH (80%). while dyshormonogenesis accountss for 20%.
Thee molecular basis of CH is largely unknown. Several proteins involved in thyroid hormonee synthesis have been identified and mutations have been found in the correspondingg genes of CH patients [1,2]. However, the cascade of molecular eventss required for the normal development of the thyroid gland is poorly understoodd at the molecular level. In general, "patterning" molecules responsible of developmentall processes of the embryo can be intrinsic, mainly transcription factors,, and extrinsic, represented by so called morphogens. Morphogens have beenn implicated not only in morphogenesis of embryonic structures but also in
morphostasis"" or maintenance of patterning events in postnatal life [3]. Both knock-outt mice and human models show that the development of the thyroid gland is partlyy under control of the transcription factors TTF1/NKX2.1, TTF2/FKHL15 and PAX88 [4-10]. Since inactivating mutations in these genes only account for a very smalll percentage of cases of thyroid dysgenesis, it is generally assumed that there mustt be more, currently unidentified factors involved in embryogenesis. migration andd follicle formation of the thyroid gland [11] Notably, no effector or mediator of developmentall processes in the thyroid has so far been characterized at the molecularr level.
Too identify novel thyroid specific genes we previously made a Serial Analysis of Genee Expression (SAGE) library of a normal thyroid gland [12]. Subsequent analysiss of the SAGE library revealed about 80 SAGE tags that could not be linked too any cDNA with an attributed function (no-match), but were expressed in thyroid at aa level similar to other well-known thyroid specific proteins as the TSH receptor and thyroperoxidase.. To expedite downstream analysis of these no-match SAGE tags wee developed a Tissue Preferential Expression (TPE) algorithm that allows in silico determinationn of SAGE tags with preferential thyroid expression [13]. This resulted inn the selection of no-match tag 41 (NM41) for further analysis. We report here the
cloningg and characterization of NM4J. a novel protein with preferential expression in thee thyroid.
Materiall and methods
CloningCloning of the NM41 gene.
AA SAGE library (TH4) was constructed from normal human thyroid tissue obtained fromm resection [12]. From the 4.26C no-match tags contained in the library (tags that doo not find a match in the Genbank database), the tag ccagctgcct was selected as preferentiallyy expressed in thyroid, using the TPE computational subtraction method [13].. The tag was linked to an EST sequence using the SAG E map program [14]. Tissuee preferential expression of NM41 was validated by RT-PCR using two primerss designed based on the AI375154 EST sequence (NM41frw1: 51 -ACAATTTCCAGATGGCTGCTCCTC-3'' and NM41rev1: 5'-TGCCTACTCAGGGCTTCCAAGAT-31).. In order to achieve the full-length cDNA of thee NM41 gene, a 3'-directional ZAP Express cDNA library was constructed from thee same human thyroid sample used for the construction of our TH4 SAGE library (Stratagene).. Approximately 40.000 plaques were screened with a 298 bp NM41 PCRR fragment after random-prime labeling with 'JjZP-</dATP. using standard hybridizationn conditions [15]. Aft er 2 rounds of re-screening and excision, positive cloness were sequenced. The chromosomal localization and organization of the genee were determined by comparison of the NM41 cDNA sequences with the humann genomic clone AC0007011 The wild type sequence was validated after directt seguencing of RT-PCR amplified NM41 from thyroid cDNA. using NM41 specificc primers NM41frw2: 5'- CCGCTGTGGTCCTCCTCACT-3" and NM41rev2: 5'-- CCTTAGATCCTGGAGGGC-3' . 5"-rapid amplification of cDNA ends (RACE) wass performed in order to map the transcription start site using the cDNA specific
r e v e r s ee primer 5'- CAGTAAGAGCAGGAGCAGCAGCG-3". located in exon 1.
inin silico determination of the NM41 expression pattern.
Tissuee expression profiles of the NM41gene were investigated by the presence of thee ccagctgcct tag in SAGE libraries of the CGAP/NCBI database. The EST libraries weree also screened for the abundance of EST sequences corresponding to the Hs. 117822 EST cluster. Abundance was determined by the number of times that the tag wass scored, or the number of EST-related cDNA clones present in SAGE or EST libraries,, respectively. Relative expression of NM41 per tissue was estimated by dividingg the abundance of NM4-SAGE tag / NM41-EST clones by the size of each SAGEE or EST library
Ciomrqq of the N.V141 gene
NorthernNorthern blot analysis.
Thee human multiple tissue northern blot "endocrine system" membrane (ClonTech). containingg 2 ug of mRNA per lane, was incubated at 68 C in hybridization solution forr 1 hour and hybridized (68 C. 4 h) against the 298 bp NM41 PCR product labeledd by random priming with 32P-f/.dATP (10 uCi) using the Klenow fragment of thee Taq polymerase. As probe for GAPDH detection, a labeled 300 bp PCR product amplifiedd from thyroid cDNA with the primers GAPDHfrw: 5'-ctgagaacgggaagcttgtc-3'' and GAPDHrev: 5'-gtgctaagcagttggtggt-3'. was used. Blots were washed followingg the instructions of manufacturers, exposed for 24 h in an Imager system andd visualized using AIDA software (Raytest).
IdentificationIdentification of functional motifs and homologies of NM41 protein.
Thee NM41 amino acid sequence was subjected to different computer programs [16] thatt can identify functional motifs, predict localization of the protein within the cell or assignn the sequence to known families of proteins. PROSITE and PSORT were usedd for general functional motif identification. SignalP to identify signal/anchor peptidee sequences, TOP PRED2 to localize hydrophobic stretches (putative transmembranee domains) and O-GLYCBASE to detect putative O-linked glycosylationn sites. General homology searches were performed by BLASTp and BLOCKS.. For specific homology with cystine-knot proteins the specific database at www.hormone.stanford.edu/cystine-knott was used.
InIn vitro expression of NM41 protein.
Thee NM41 protein was in vitro transcribed and translated from 200 ng of the fully sequencedd pBK-CMV-NM41 clone obtained after cDNA library screening, using a TnT77 T7 Coupled Reticulocyte Lysate System (Promega) according to instructions off the manufacturer with Tran35S-label (ICN Pharmaceutics), For experimental size determination,, samples were run on a 12.5% SDS-PAGE together with a protein markerr (Biorad). Radioactive signal of expressed in vitro translated proteins was detectedd using a Phosphoimager and Image Quant software (Molecular Dynamics).
CellCell culture and transfections.
HTorii 3 cells [17] were cultured in Dulbecco's modified Eagle's medium (Life Technologies;; Inc.) containing 4.5 g glucose/I supplemented with N a H C 03, L-Glutamine,, antibiotics and 10% fetal calf serum (Roche Molecular Biochemicals). Cellss were grown to 60-70% confluence at , 5% C 02 and 90% humidity and thenn were transfected with 2 pg of pCDNA3-NM41, using Fugene 6 (Boehringer
Mannheim)) After 48 hours, cells were harvested for protein isolation by standard procedures. .
DevelopmentDevelopment of antibodies against NM41.
Thee full-length cDNA sequence of NM41 subcloned in pCDNA3 vector was used to generatee polyclonal antibodies in mice (Eurogentec. Belgium). Four animals were injectedd with 200 pg of DNA of the plasmid purified in non-antigenic conditions (Endotoxinee Free Plasmid Mega kit. Promega). Antibody specificity was examined byy Western blotting using NM41 protein transiently expressed in HTon3 cells.
WesternWestern blot analysis.
100 pi of reticulocyte reaction, and molecular weight marker were subjected to SDS PAGEE (12.5%) according to Laerrmli's method. Geis were electro blotted to nitrocellulosee filters (0.1 pM: Schleicher & Schuell) and immunodetection was performedd at room temperature using PBS containing 0 . 1 % Tween 20. completed withh 5% low fat milk. Antibody complexes containing mouse-antiNM41 as first antibodyy (500 fold dilution) and GAMPO (DAKO) as second antibody were visualizedd using the ECL kit (Amersham International).
immunohistochemicalimmunohistochemical localization of NM41 protein.
Formalinn fixed paraffin embedded human biopsy specimens were obtained from the archivess of the pathology department of the Academical Medical Center. Sections of 44 urn were dewaxed and rehydrated in graded alcohols. Endogenous peroxidase activityy was quenched, antigen retrieval was performed by heating for 10 min. at
100 C in a 10/1 niM TRIS/EDTA pH 9,0 buffer, and non-specific staining was reducedd by a blocking step with normal goat serum. A 1000-fold dilution of NM41 antibodyy was applied in PBS containing 1 % bovine serum albumin and incubated forr 60 minutes at room temperature. A three step detection method was performed (ILL Immunologic), using the mouse primary antibody (60 min), rabbit anti-mouse antibodiess (15 mm) and goat anti-rabbit antibodies linked with horseradish peroxidasee (HRP) (30 min), following manufacturer's instructions. Peroxidase activityy was detected with diammobenzidine (DAB. Sigma), resulting in the formationn of a brown reaction product. Sections were briefly counterstained with hematoxilline.. dehydrated in graded alcohols and mounted Further controls consistedd of omitting the primary and secondary antibody and use of an appropriate control. .
CC c;r.ng o' the N U 4 " gene
R e s u l t s s
CloningCloning of NM41 from a thyroid SAGE library.
AA SAGE library was constructed from normal human thyroid tissue obtained by resectionn [12]. This library contains 10.994 tags of which 6.099 corresponded to uniquee tags. 1839 of these tags found a match in the GenBank database, correspondingg to known genes, while 4.260 (match tags) did not. These no-matchh tags correspond to putatively novel genes expressed in thyroid, in the search forr genes preferentially expressed in thyroid, the tag ccagctgcct was selected by the TPEE computational subtraction method [13] from 83 no-match tags with an expressionn level of 0.05% in thyroid. The NM41 TPE value of 8.36 (from a range 0-11)) was above the validated threshold of 7 and indicated thyroid specificity. The 10 bpp tag was linked to a 723 bp EST (AI375154) that contains a partial open reading frame.. This EST sequence was accepted based on 3 criteria: 1. it belongs to a 3'-orientedd cluster with a clear poly(A) signal: 2. it contains the SAGE tag in the proper orientation:: and 3. the tag is preceded by the most 3'-located CATG site within the ESTT sequence.
Thee EST sequence was used to screen a thyroid cDNA library and this resulted in thee isolation of 10 clones. Sequencing of these 10 clones identified a cDNA sequencee not present in the GenBank that extended 622 bp 5' of EST AI375154 Homologyy searches identified a match with the human genomic clone AC007011 in chromosomee 16p, from which the genomic organization of the gene was determinedd (Fig 1aj. The NM41 gene contains 8 exons and 7 introns and spans overr 20 kilo base pairs on human chromosome 16p. The transcription start site was mappedd by 5'RACE and fitted exactly the 5' end of the clones from the cDNA library. Thee 1345 bp mRNA transcript contains a 5' UTR of 22 base pairs, an open reading framee of 1083 base pairs and a 3' UTR of 240 bp The ATG translation initiation site iss preceeded by a Kozak consensus sequence [18] and the ORF encodes a protein off 361 amino acids (Fig. 1a. 1b).
TissueTissue distribution of NM41 mRNA.
Thee pattern of gene expression was assessed in silico by the determination of the presencee and relative abundance of NM41 SAGE tags in publicly available SAGE librariess and the presence and relative abundance of NM41 -related EST's in EST libraries.. NM41 expression is high in thyroid and present at lower levels in tissues likee stomach, lung and pancreas (Fig 2a).
Chapterr 5
--
Genomicc clone AC007011 AC007011II I I II
E x o n ss 1 2 3 4 5 6 G e n o m e e (Chr.. 16p) N M 4 11 c R N A 1.3451.345 bp ORF:ORF: 1.083 Qp N M 4 11 protein 3611 amino acids GGCC GGCC r : - : ? : . .Cloningg o*'tie N M ^ I gene
Too validate these results. NM41 expression was examined by Northern blot in a premadee membrane containing human mRNAs from thyroid and other tissues that showedd NM41 expression in silico. The mRNA signal observed at 1.35 kb fits the sizee of the full-length cloned NM41 cDNA. Expression of NM41 is high in thyroid, loww in pancreas and stomach (Fig 2b).
ProteinProtein analysis and prediction of functional motives.
Too elucidate the function of NM41 protein, computer programs were used to predict thee presence of functional motives in the NM41 amino acid sequence. A 19 amino acidd N-terminal cleavable signal peptide was identified (fig 1b). suggesting that NM411 is a secreted protein. No glycosylation sites were predicted. Prosite identified thee CXGXC motif, composed of 2 cysteines flanking 3 amino acids the intermediate off which is an obligatory glycine, corresponding to the glycoprotein hormone B1 signature.. Additionally, a partially conserved pattern of cysteine residues around thiss motif is present (Fig. 3). No homologies with other proteins can be detected by aminoo acid sequence comparison.
immunohistochemicalimmunohistochemical localization of NM41 in human tissues.
Expressionn of NM41 protein is examined in a range of tissues shown to contain NM411 mRNA by experimental or in silico means. The NM41 antibody can specificallyy detect NM41 protein with a molecular weight of approximately 40 kDa in HTon33 cells transiently transfected with the pCDNA3-NM41 construct (Fig. 4a). In thyroid.. NM41 is localized in the cytoplasm of thyroid cells, showing a coarse granularr staining. Interestingly, the follicular lumen is positively stained, fitting with thee hypothesis that NM41 is a scretable protein. Among thyrocytes. NM41 protein is unevenlyy distributed, some cells showing stronger staining and others faint or no NM411 signal (Fig. 4b,c).
Fig.. 1 The NM41 gene, transcript and protein, a, Genomic organization of the NM41
genegene present in genomic clone AC007011 from nucleotides 18.183 till 40.002. b,
ComplementaryComplementary nucleotide and deduced ammo-acid sequence of respectively human NM41 cDNAcDNA and protein. Amino-acid sequence is shown below the nucleotide sequence. The
signalsignal peptide, the cystine-knot related amino acids and the SAGE tag for the gene are shadedshaded in green, red and blue, respectively. The Kozak sequence is inderlined in italics and thethe polyadenylation signal in the 3' untranslated domain is underlined, (i p. 183)
Chapterr 5 o o Q. . 1/1 1/1 O O c c _o o u u H H CO O LU U !fl l CO O TO TO UJ J O O
< <
co o 'SAGE E -EST TII . . .
<// y y
z y ^ y y
re re CD D Ü Ü 0. . J5 5 3 3 a a "TO O c c < < Ü Ü ^ ^ H HTissues s
x x t t o o u u "TO O c c < < l/l l/l 'JJ j h--Ifl l F F > > 1--0) ) c c a a TO TO b b CO O o o TO TO (--O (--O CO O 2.44 kb 1.355 kb 1.355 kbNM41 1
GAPDH H
F i g .. 2 Tissue distribution of NM41 mRNA. a, Expression pattern of the NM41 gene
assessedassessed by relative expresssion levels in Serial Analysis of Gene Expression (SAGE) and ExpressedExpressed Sequence Tag (EST) libraries, b, Northern blot analysis of NM41mRNA expressionexpression in 8 human tissues. Upper panels are probed with NM41 mRNA and the lower withwith GAPDH mRNA.
Cloningg of the NM41 gene
Inn lung, an intense NM41 staining is present in epithelial cells of the bronchia and alveolii and is absent in mesoderm-derived muscle cells (Fig. 4d,e). In alveolar cells, aa coarse granular pattern is also detectable and, again, expression is restricted to groupss of cells, others showing negative staining (Fig. 4e). In pancreatic tissue, NM411 expression is restricted to cells of exocrine nature, showing an evenly distributedd pattern of coarse granules. Endocrine cells in the islets of Langerhans aree negative for NM41 staining.
s
C o n s e n s u ss . . . . < « < . C . < ( ( < . ( . . .
N M 4 11 R S D V ' C ' L V Q L L G T G I D S S t P C G - L S U l X RSLMTKPG» s G v ( ISHQl - FFLVVA -RMRG< l QQSQDÏ IM h< AVMMDI NRRA1 VI' 250
F M E N I M F C O M II i Kl :(SM l-AII s u n k ' . . . i i ( IEKLSKA1QYOQ i - C S S H M A A V A A L G G H Y I L A I YPPASR1 Pll .'52
PSTPPPPSSRR " ' '
Fig.. 3 Alignment of NM41 protein sequence with known cystine-knot proteins:
subfamilysubfamily of glycoprotein hormones. Amino acid residues involved in the cystine-knot structurestructure are shown in red. Disulfide bonds between conserved cysteine residues are indicatedindicated by dotted lines. The glycoprotein hormone beta-1 signature is depicted by a green box.box. Additional carboxiterminal cysteine residues in the NM41 sequence are shown in blue.
Chapterr 5 522 KDa 36.88 KDa
''
:
^£& ?/ .:'\.V
e e
V V
CC o'iir.g of the NM41 gene
D i s c u s s i o n n
Wee report the cloning, tissue distribution and initial characterization of a novel proteinn preferentially expressed in thyroid. NM41 contains a cleavable signal peptidee in its amino-termmal end. suggesting it is a secretable protein. In silico analysiss of functional motifs identifies a well conserved glycoprotein hormone Beta 11 signature (CXGXC). a five ammo acid motif present in the beta chains of TSH, FSH:: LH and CGH glycoprotein hormones and the novel thyrostimulin [19]. However,, the amino acid sequence of NM41 does not show significant additional homologiess to these hormones or to any known family of proteins.
Glycoproteinn hormones constitute a subfamily of proteins belonging to a broader andd growing family of molecules designated as cystine-knot. The name is derived fromm the spatial conformation of these proteins in a "knot", formed by 3 disulfide bridgess between a frame of six cysteines tightly distributed along the amino acid sequence.. Other members of the cystine-knot family of proteins are transforming growthh factor beta (TGF-(ï). the platelet-derived growth factor (PDGF) or the recentlyy incorporated subfamily of mucin-related proteins [20], Classification within thee family is based on the presence of a signal peptide; localization of the cysteine framee within the entire protein or the presence of additional motives.
NM411 conserves 5 out of the 7 cysteine residues typical for glycoprotein hormones, includingg the 2 cysteines of the B1 signature, with conserved spacing between residues.. The exceptions are the 2 C-terminal cysteines from the consensus domainn that are necessary for the formation of 2 disulfide bonds but are lacking in thee NM41 amino acid sequence. Theoretically, this would prevent the formation of thee knot However. 3 additional cysteines are present carboxyterminal to the consensuss motif in NM41, suggesting the formation of 2 slightly longer loops comparedd to the typical cystine-knot structure.
Fig,, 4 Immunohistochemical analysis of NM41 in human tissues, a, Specific
detectiondetection of NM41 protein transiently expressed in HTon3 celis with polyctonal NM41 antibodies.antibodies. Lane 1: J3S-labeled NM41 protein expressed in reticulocyte iysate system, as a controlcontrol for NM41 size: Lane 2: Protein extract from NM41-transfected HTon3 ceils and NM41 antibody:antibody: Lane 3: Protein extract from untransfected HTon3 celis and NM41 antibody b, c, NM41NM41 immunoreactivity in thyroid slices (100x magnification). The follicular lumen (Ft) and thethe positive granular staining (arrows) are indicated, d, e, immunostaining of NM41 in lung (d,
10x10x and e, lOOx magnification). Bronchia (Br) and Alveoli (Av) are indicated, f, g,
ImmunostainingImmunostaining of NM41 in the pancreas (f. 100x and g. 41 x magnification). Negative islets ofof Langerhans are indicated by dotted circles. 0 ' P- 185)
Basedd on the substantial conservator* of the cysteine frame in NM41. most similar too the glycoprotein subfamily, and in the presence of a cleavable signal peptide, characteristicc of other subfamilies of cysteme-knot proteins, we propose that the novell protein constitutes the model lor a new gender of proteins with a cystine-knot-likee sequence and tertiary conformation.
Too date, all known members of the family of cystine-knot proteins are extracellular proteins,, interacting with specific *"eceptors and-'or other extracellular proteins, mainlyy through the formation of h o m o or hetero-dimers [20.21]. These proteins play ann essential role during early embryonic development and. in adult life, have diverse physiologicall roles in cell cycle regulation, modification of the extracellular matrix andd regulation of other growth factors [20], Members of the TGF-(ï and u families havee been directly implicated in both the embryonic patterning and cell-fate of the primitivee gut. as well as in the maintenance of the ongoing patterning events ("morphostasis")) in postnatal life [22-24].
Thee gut tube is patterned in two major axes. Along the anterior-posterior axis, it developss regionally specific differentiation generating the foregut (.pharynx, esophaguss and stomachj. midgut (small intestines} and hindgut (colon). Along this axis,, the gut gives rise to several derivatives: thyroid, trachea, lungs, liver and pancreas.. NM41 is expressed at high levels in thyroid, but also present in most tissuess differentiated from the primitive gut (esophagus, stomach, colon) and derivativee structures (lungs and pancreas). The range of tissue expression of NM41 andd its similarity to cystine-knot proteins suggest a role in the development or maintenancee of tissue organization of the thyroid and other embryologicaiiy related structures. .
Furtherr characterization of the novel NM41 protein, including the determination of secretion,, self- or hetero- dimerizatton with partner molecules, or the computational modellingg of its tertiary structure are now open for research. Proof of the hypothesis thatt NM41 can be a morphogen involved in development of thyroid and other tissuess can lead to the identification of human molecular defects causing dysgenesiss of the thyroid gland or syndromic phenotypes associated with congenitall hypothyroidism.
CC o r n ci of t r fi \ ••!J 'j f; n e
References s
11 Vulsma. T. and de Vijlder J . J . M (2002) Thyroid disease in newborns, infants and children,, in Oxford Textbook of Endocrinology and Diabetes (Wass. J.A. and Shalet. S.M... eds.j. pp. 532-544. Oxford University Press.
22 Moreno. J.C. et al. (2002). inactivating mutations in the gene for thyroid oxidase 2 (THOX2)) and congenital hypothyroidism. N. Engl. J. Med. 347. 9 5 - 1 0 2 .
33 Potter, J.D. (2001) Morphostats: a missing concept in cancer biology. Cancer EpidemiolEpidemiol Biomarkers Prev 10: 161-170).
44 Kimura. S. et al. (1995) The T/ebp null mouse: thyroid-specific enhancer-binding proteinn is essential for the o r g a n o g e n e s i s of the thyroid, lung, ventral forebrain andd pituitary. Genes & Develop. 10. 60-69.
55 De Felice, M. et al. (1998) A mouse model for hereditary thyroid dysgenesis and cleftt palate. Nat. Genet. 19. 395-398.
66 Mansouri. A. et al. (1998) Follicular cells of the thyroid gland require Pax8 gene functionn Nat. Genet. 19. 87-90.
77 Krude, H. et al. (2002) Choreoathetosis, hypothyroidism and pulmonary alterationss due to h u m a n NKX2-1 haploinsufficiency. J. Clin. Invest. 109, 475-480. .
88 Pohlenz, J. et al. (2002) Partial deficiency of thyroid transcription factor 1 producess predominantly neurological defects in humans and mice. J. Clin. Invest. 109,, 469-473.
99 Clifton-Bligh. R.J. et al. (1998) Mutation of the gene encoding h u m a n TTF-2 associatedd with thyroid agenesis, cleft palate and choanal atresia. Nat. Genet. 19,
399-401. .
100 Macchia. P.E. et al. (1998) PAX8 mutations associated with congenital hypothyroidismm caused by thyroid dysgenesis. Nat. Genet. 19, 83-86.
111 Kimura, S. et al. (1999). Thyroid-specific enhancer-binding protein/thyroid transcriptionn factor 1 is not required for the initial specification of the thyroid and lungg primordia. Biochlmie 8 1 : 321-7.
122 Pauws. E. et al. (2000) Serial analysis of gene expression as a tool to assess the h u m a nn thyroid expression profile and to identify novel thyroidal genes. J. Clin. Endocrinol.Endocrinol. Metab. 8 5 : 1 9 2 3 - 1 9 2 7 .
O a ^ e rr 5
133 Moreno. J C. et ai. (2001 i C l o n ' i g of t ssue-spec fie genes ^ s ^ g s e r a i a n a l y s e of genee e x p r e s s i o n and a nove; : o m p u t a t i o o a l s u b s t r a c t o n aoproacn. Genomics 75:70-76 6
144 Lai. A. ef al (.2000) A publ c database for gene expression :n human cancers. CancerCancer Res. 59: 5 4 0 3 - 5 4 0 7
155 Mamatis. T. et al. (1982). Molecular cloning: A laboratory manual. Cold Sprang Harborr Laboratory Press. Cole Spring Harbor. New Yom.
166 B a x e v a m s . A . D . ( 2 0 0 0 ; The Molecular Biology Database collection: an online compilationn of relevant database resources Nucleic Acids Res. 28: 1-7
177 L e m o m e NR. et al (1989). Characterization of h u m a n tnyoic! ep:tneha; cells immortalisedd m vitro by simia" virus ^ 0 DNA t r a n s e c t i o n Br J. Cancer 60: 897-903. .
188 Kozak, M. (1989) The s c a n n r g moael for translation' an upcate. J.Cell. Biol. 1 08: 2 2 9 - 2 4 ' ' '
199 N a k a b a y a s h i . K. et al f'2002) T h y r o s t i m u l m . a heterodimer of two new h u m an glycoproteinn h o r m o n e s u b m i t s , activates the thyroid-stimulating hormone receptor.. J. Clin. Invest. 109: 1445-1452.
200 Vitt. U. A. et ai. (2002) Evci-tion and classification of cystine knot-containing h o r m o n e ss and related e x t r a o c u l a r signaling molecules. Mol. Endocrinol 1 5: 6 8 1 -694. .
211 Hsu. S.Y. et al. ("2002) Evc'ution of glycoprotein hormone subunit genes m bilaterall m e t a z o a : identification of two novei h u m a n glycoprotein hormone subumt familyy genes. GPA2 and GPE5 Mo!. Endocrinol 16: 1538-1551
222 Li. Q. et al. M 9 9 8 ) Stirnulatcn of activin receptor II signaling patnways inhibits differentiationn of multiple gas're epithelial lineages Mol. Endocrinol. 12: 181-192
233 Tagaki. H et al. ( 1 992) Hypertrophic gastropathy resembling Menetner's disease mm transgenic mice overexpressing transforming growth factor alpha J Clin Invest.Invest. 90:1161-1167
244 Dempsey. P.J e( al. ( i y 9 2 ) Possible roie of transforming y t o w t h factor alpna in thee pathogenesis of M e n e ; n e r s cisease. supportive evidence from human t transgenicc mice. Gastroenterology 103: 1950-1963.
