| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLE |
Institut National de la Santé et de la Recherche Médicale, Unité 369, Institut Fédératif de Recherche Laennec, 69372 Lyon, Cedex 08, France
Address all correspondence and requests for reprints to: Professor Bernard Rousset, Institut National de la Santé et de la Recherche Médicale Unité 369, Institut Fédératif de Recherche Laennec, Rue Guillaume Paradin, 69372 Lyon Cedex 08, France. E-mail: u369{at}laennec univ-lyon1.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
J were isolated. pNIS-D encodes a protein of 643 amino acids highly homologous to the human, rat, and mouse NIS. pNIS-F and J differ from each other and from pNIS-D in their C-terminal part. pNIS-J lacks a six-amino-acid segment within the putative transmembrane domain 10. Transiently expressed in Cos-7 cells, the four pNIS-cDNAs led to the synthesis of proteins targeted at the plasma membrane and conferred perchlorate-sensitive iodide uptake activities to Cos-7 cells, except pNIS-J, which was devoid of activity. PNIS-D probably derives from the 3.5-kb transcript and pNIS-F, J, and J from the 3-kb transcript. The relative abundance of pNIS-D, F, and J transcripts in porcine thyrocytes was about 60%, 35%, and 5%, respectively; the J transcript was not present in detectable amount. By comparing porcine NIS genomic and cDNA sequences, splice donor and acceptor sites accounting for the generation of pNIS-F, J, and J transcripts were identified. None of the combinations of alternative splice sites found in the pig was present in the human, rat or mouse NIS gene. Our data show that porcine NIS gene, contrary to the NIS gene from other species, gives rise to splice variants leading to three active and one inactive NIS proteins. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
One of the privileged in vitro models to analyze different aspects of thyroid cell biology, especially iodine metabolism and thyroid hormone biosynthesis, is the porcine thyroid cell primary culture system. Indeed, freshly isolated pig thyrocytes are capable of reconstituting functional thyroid follicles in short-term culture (12, 13, 14), allowing one to study the expression of thyroid differentiation in vitro, under conditions rather close to the in vivo physiological context. With the aim of analyzing the regulation of NIS expression in this model, we tried to identify porcine NIS (pNIS) by Western blot, using antipeptide antibodies recognizing the C-terminal domain of either rNIS (15) or hNIS (Trouttet-Masson S., S. Selmi-Ruby, F. Bernier-Valentin, and B. Rousset, unpublished data). None of the antibodies did react with any porcine thyroid molecular species. Similarly, cDNA probes used to detect the rNIS transcript by Northern blot were ineffective for the identification of NIS transcripts in pig. Using a cDNA fragment generated by PCR with primers corresponding to conserved sequences in rNIS and hNIS, we found that pig thyrocytes contained two NIS transcripts of about 3.5 and 3 kb. This preliminary information suggesting substantial species differences in NIS primary structure and expressed forms prompted us to generate a porcine thyroid cDNA library to clone pNIS. Here, we report that the pNIS gene gives rise to different transcripts that are produced by the use of alternative splice sites not present in the NIS gene from other species (human, rat, and mouse). Among pNIS transcripts, three encode a functional protein and the fourth one, exhibiting a deletion in a putative transmembrane domain number 10, encodes a protein targeted to the plasma membrane but without activity.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RNA extraction
Total RNA was extracted from thyrocytes or Cos-7 transfected cells according to the phenol-chloroform method of Chomczynski and Sacchi (18). Purification of mRNA was carried out using a kit from Promega Corp. (Madison, WI). The RNA concentration was determined by absorbance measurements at 260 nm and the RNA purity and integrity were assessed by determination of the A260/A280 ratio and ethidium bromide staining of ribosomal 28S and 18S bands after electrophoretic separation on 1% agarose gels.
Molecular cloning of pNIS from a pig thyroid cDNA library
A cDNA library was synthesized from poly A+ RNA prepared from porcine thyroid cells cultured for 4 d in the presence of TSH, using a cDNA synthesis kit and the XhoI-oligo-deoxythymidine primer (OriGene Technologies, Inc., Rockville, MD). An EcoRI adapter was ligated at the 5' end of cDNAs that were then cloned into the EcoRI-SalI sites of the pCMV6-XL3 vector owing to the fact that XhoI and SalI have compatible cohesive ends. The pig cDNA library containing about 5.4 x 106 independent clones was screened by hybridization according to standard procedures (19) using a 0.8-kb porcine NIS cDNA fragment generated by RT-PCR (probe A). The primers used to generate probe A derived from the rat NIS (rNIS) sequence; they corresponded to nucleotides 559–584: 5'CGC GCC TGC GCT CAT CCT GAA CCA AG3' and to nucleotides 1309–1333: 5'CAG CAG TGA GGA CAG AGC CAC AG3' of the rNIS cDNA coding region (4). The amplified cDNA was cloned into the pGEM-T vector and certified by sequencing (Genome Express, Grenoble, France) using templates from the pGEM-T vector. Finally, the probe generated by digestion with Sph1 and Pst1 was randomly labeled with [
32P]-deoxy-CTP. After overnight hybridization at 42 C with the labeled probe, nylon membranes were washed at high stringency at 65 C in 0.2x SSC, 0.5% sodium dodecyl sulfate (SDS) (1x SSC corresponds to 150 mM NaCl and 15 mM sodium citrate at pH 7.0); positives clones were selected and amplified.
Preparation of pNIS cDNA constructs
Constructs corresponding to each pNIS cDNA clone isolated from the porcine thyroid cDNA library were generated by PCR. To optimize expression, the 5' and 3' untranslated regions of pNIS cDNAs were removed. PCR amplification was performed with 200 ng of cDNA template, 50 pmol of each primer in 5 µl 10x reaction buffer, and 2.6 U Expand High Fidelity Taq polymerase (Roche Diagnostics, Meylan, France) in a final volume of 50 µl. As the 5' end of the open reading frame was the same in all clones, we used the same forward primer to generate the different constructs. The primer sequence corresponding to nucleotides (1–30), numbered in the 5' to 3' orientation and beginning with the first base of the start codon, was: 5'ATG GCG ACC GTC GAG GGA GGC GCG CGG GCC3'. The reverse primer represented the complementary sequence of the C terminus of each pNIS cDNA clone followed by a BamH1 site (GAGGGATCC); the sequence of the primers was: 5'CGC TTA GAG GTC CGT CTC ACG CAG GTC3' for pNIS-D (1908–1934); 5'TCA GCA TCC ACC TTT GTC ATG TTC ACT GC3' for pNIS-F (1884–1912) and 5'GAG ATG CAT AAA GTG TCT AGA CGC TGA AG3' for pNIS-J (2036–2064) and pNIS J (2019–2047). Other constructs containing pNIS cDNA sequences fused in 5' to the sequence encoding the Flag epitope were generated by PCR using an oligonucleotide with a start codon followed by the 21 nucleotides corresponding to the Flag sequence and the first 24 bases of the 5' end of pNIS coding region. The 3' primers were those described above. The amplification reaction included a presoak step at 94 C for 3 min followed by 35 cycles for 1 min at 94 C, 1 min at 65 C, and 1 min at 72 C followed by a final expansion period of 10 min at 72 C. PCR products were cloned into the pTarget vector (Promega Corp.). The sequence of each construct, either untagged or N-Flag tagged, was verified by automated sequencing.
Transient transfection of Cos-7 cells
Cos-7 cells were maintained and propagated in DMEM containing 10% (vol/vol) fetal calf serum, at 37 C under air/CO2 (95%/5%) atmosphere. One day before transfection, cells were trypsinized and plated on 10-cm culture dishes. Cells at about 80% of confluency were transfected with pNIS cDNA vectors using Fugene transfection reagent (Roche Diagnostics; 0.7 µl of Fugene 6/µg DNA) according to the manufacturer’s instructions. Forty-eight hours after transfection, cells were assayed for NIS expression by Northern blot, Western blot, and indirect immunofluorescence using anti-Flag antibodies and for iodide uptake activity.
Western blot analysis
Membrane fractions from NIS expressing Cos-7 cells were prepared as described in Ref. 20 . In brief, transfected cells were washed with PBS, harvested, and resuspended in PBS containing aprotinin, leupeptin, and pepstatin (each at a concentration of 1 µg/ml), and 1 mM phenyl methylsulfonyl fluoride. Cells were then lysed by a freezing/thawing cycle and sonicated for 40 sec at 25 W using the vibra-cell apparatus from Bioblock Scientific (Illkirch, France). Crude membrane fractions were obtained by centrifugation at 100,000 x g for 20 min at 4 C. Protein concentration was assayed by the Lowry method after solubilization in 0.1% sodium desoxicholate using BSA as standard. Membrane protein samples (40 µg) were fractionated by electrophoresis on 9% polyacrylamide gel in presence of SDS and electrotransferred onto Immobilon-P membranes (Millipore Corp., Molsheim, France). Immobilon-P membranes were preincubated with 5% (wt/vol) nonfat dry milk and 0.2% (vol/vol) Tween-20 in PBS for 1 h at room temperature and incubated with a mouse anti-Flag M2 monoclonal antibody (Sigma-Aldrich, St. Quentin Fallavier, France) at a 1:2000 final dilution, overnight at 4 C. After three washes in PBS-0.2% Tween solution, Immobilon-P membranes were incubated with an antimouse IgG conjugated to horseradish peroxidase (Bio-Rad Laboratories, Inc., Richmond, CA; 1:5000 final dilution) for 1 h at room temperature. Detection of immune complexes was performed using an enhanced chemiluminescence kit from Amersham (Orsay, France) followed by an exposure for 1–3 min to Kodak (Rochester, NY) X-OMAT-AR film.
Indirect immunofluorescence
Cos-7 cells were fixed with 4% paraformaldehyde for 30 min at room temperature, then washed three times with PBS containing 10 mM glycine. Cells were permeabilized or not with 1% Triton X-100 in the same buffer for 10 min and washed three times with PBS containing 10 mM glycine. Cells were then preincubated in PBS containing 4% BSA for 1 h and incubated with the mouse anti-Flag M2 monoclonal antibody at a 1:2000 dilution in PBS containing 4% BSA for 1 h. Cells were washed and further incubated with a fluorescein-labeled goat antimouse secondary antibody (1:100 dilution; DAKO Corp., Trappes, France) for 1 h in the dark. Nuclei were stained with 1 µg/ml Hoechst-33342 reagent (Molecular Probes, Eugene, OR) for 10 min at room temperature. Fluorescence images, obtained with a Carl Zeiss (Jena, Germany) Axiophot fluorescence microscope, were monitored using a cooled charge-coupled device camera (LHESA Electronique, Cergy Pontoise, France). Images were collected in a computer equipped with a Cyclope imaging card (Digital Vision, Chatillon, France) and analyzed using graphic software.
Iodide uptake activity (IUA) measurements
Transfected cells were trypsinized, distributed into 3-cm wells, and cultured for 24 h. Cells were washed twice with Earle’s balanced salt solution, pH 6.8, and incubated in the same medium containing 0.5–1.0 µCi Na125I for 40 min at 37 C. Incubations were performed in the absence and in the presence of 0.1 mM sodium perchlorate, a competitive inhibitor of the sodium/iodide symporter. At the end of the incubation period, the medium was removed, and cells were quickly washed twice with ice-cold Earle’s medium. The procedure was achieved within 40 sec. Cells were lyzed in 200 µl cell culture lysis buffer (Promega Corp., Madison, WI) for 5 min and the cell lysate was counted in a 
counter from Packard Instruments Co., Inc. (Groningen, The Netherlands). Incubations were made in triplicate.
Analyses of the relative expression of each pNIS splice variant by thyroid cells in primary culture
Total RNA (25 µg) from cultured thyroid cells was fractionated by electrophoresis on 1% agarose gel containing formaldehyde, transferred to Hybond-N+ membrane (Amersham) and hybridized with two different probes (probe A and B) using standard Northern blot procedures (19). Probe A was the 0.8 kb pNIS cDNA fragment described above and probe B was a 476-bp pNIS cDNA fragment generated by PCR using pNIS-D as template and the following primers: 5'-primer (1908–1932) 5'ACC TGC GTG AGA CGG ACC TCT AA3'; 3'-primer (2361–2383) 5'GTA TAG GGG TTG GGC CTC AGG AC3'. Probes were labeled using the random priming kit from Roche Diagnostics. Hybridization was carried out in a 50% (vol/vol) formamide solution containing the [32P]-labeled probe (1–2 x 106 cpm/ml), overnight at 42 C. Filters were washed under stringent conditions (0.2x SSC, 0.5% SDS at 60 C) and exposed to Kodak XAR-5 films (Eastman Kodak Co.) at -80 C using intensifying screens.
The identification of transcripts corresponding to the different splice variants among pig thyroid cell RNA was pursued using RT-PCR. Two sets of primers, 1/2 and ß1/ß2, flanking the alternatively spliced regions were used; their position on the pNIS-D cDNA is shown (see Fig. 5
). Their sequence was: 1(1805–1828) 5'GAC TTC CTG TCC ACT AAT GAG GAC3'; 2 (2571–2591) 5'AGT GTC TAG ACG CTG AAG ATG3'; ß1 (933–962) 5'CCT CTC CTG GCA GGG CAT ATC TCT GCC CCA3' and ß2 (1294–1323) 5'CGT ATT GCA GGA GGG GAG GAA CAT TCC CAG3'. One microgram of total RNA isolated from pig thyrocytes was reverse-transcribed with the specific antisense primer and 200 U of RT-SuperScript II (Invitrogen SARL, Cergy Pontoise, France) and used as template for PCR amplification. The protocol included a presoak step at 94 C for 3 min followed by 35 cycles of 1 min at 94 C, 1 min at 60 C and 1 min at 72 C followed by a final expansion period of 10 min at 72 C. The resulting PCR products were separated by electrophoresis on 2% agarose gels stained with 0.5 µg/ml ethidium bromide. The size of the expected fragments generated from 1/2 primers was 295 and 249 bp for pNIS-F and pNIS-J/J, respectively. The size of the expected amplicons generated from ß1/ß2 primers was 393 bp and 377 bp for pNIS-J and pNIS-J, respectively. Given the small difference in size, the amplified products were submitted to a Bpu1 digestion to discriminate between pNIS-J and pNIS-J.
|
Set A, 5'primer (1275–1301)-exon 8 5'TGA AGA CCT GCC TGG AGT CCC TGG GCT3';
3'primer (1523–1548)-exon 10 5'CTG ATG ACT CCC ATG ACG GTG AAT GA3'.
Set B, 5'primer (2014–2037)-exon 14 5'GCT GCC CTG GAT GAC AGC CTG ATG3';
3'primer (2044–2067)-exon 15 5'TGC CAA AGG CAA TTC CTC AGC ACC3'.
Set C, 5'primer (1275–1301)-exon 15 5'GGT GCT GAG GAA TTG CCT TTG3';
3'primer (2852–2880)-exon 15 5'GAG ATG CAT AAA GTG TCT AGA CGC TGA AG3'.
PCR amplification products were purified using agarose gel electrophoresis and the DNA extraction kit from QIAGEN S.A. (Courtaboeuf, France); they were cloned into the pGEM-T vector and subjected to automated sequencing.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The two transcripts detected in freshly isolated thyroid cells disappeared within 48 h of culture of thyrocytes in the absence of TSH. When thyrocytes were cultured in the presence of TSH, the level of the two transcripts first declined (at 24 h) then increased up to 5- to 10-fold after 48 or 72 h. Under these various circumstances, the 3.5-kb transcript represented the major form and the relative level of the two transcripts remained rather constant. The identification of two NIS transcripts in pig thyroid gland was in marked contrast with what was known in other species; only one transcript of 3.7 kb had been detected in humans and 2.9 kb in mice and rats, respectively. To determine whether the different pNIS mRNAs encode active Na+/I- symporter proteins, we decided to clone pNIS cDNA sequences.
|
J have been isolated. The size of these cDNAs as well as the size of their open reading frame and corresponding polypeptide chain are reported in Table 1. The four cDNA sequences have been submitted and are available in GenBank/EMBL/DDBJ under the accession number mentioned in Table 1. The 3'-untranslated region (variable in length) of the four pNIS cDNAs presented two typical AATAAA polyadenylation signal (22) with a poly (A) tract.
|
). The human NIS and the pNIS-D proteins both composed of 643 amino acids differ from rat and mouse NIS proteins (formed of 618 amino acids) by two insertions of 5 residues in the putative last intracellular loop and 20 residues in the C-terminal intracellular domain. Polypeptide sequences of pNIS-F and pNIS-J, respectively, share the first 635 and 620 amino acids with pNIS-D but present distinct C-terminal sequences (Fig. 2B). The pNIS-J sequence differs from that of pNIS-J by a six-amino-acid deletion (amino acid 391–396) (Fig. 2C). Secondary structure prediction analysis suggested that pNIS-D (as well as other isoforms) is an intrinsic membrane protein with 13 transmembrane domains with the amino terminal domain located on the extracellular side of the plasma membrane and the carboxy terminal end inside the cytoplasm in accordance with the revised topological analysis of rNIS by Levy et al. (11). The three N-glycosylation sites (Asn-X-Ser/Thr) conserved in hNIS, rNIS, and mNIS are present in all pNIS isoforms (positions 225, 489, and 501). The potential phosphorylation site by cAMP-dependent protein kinase at position 554–557 on hNIS, which is conserved in rNIS and mNIS is absent in pNIS isoforms. Two consensus sequences for phosphorylation by protein kinase A have been found on pNIS at positions 52–55 and 240–243; they are located at the start of predicted helices forming the putative transmembrane domain nos. 2 and 7.
|
, the anti-Flag M2 monoclonal antibody recognized a single protein band, with an electrophoretic mobility of approximately 62 kDa, in cells transfected with either pNIS-D, pNIS-F, pNIS-J, or pNIS-J constructs.
|
. Immunofluorescence images of nonpermeabilized cells revealed the presence of the recombinant proteins at the plasma membrane. A similar labeling was obtained for the four pNIS isoforms. In permeabilized cells (+ Triton X-100), the immunostaining was not restricted to the plasma membrane but was also localized in intracellular compartments.
|
. Cells transfected with control vectors did not show any IUA. By contrast, a specific perchlorate sensitive iodide uptake was observed in cells expressing pNIS-D, pNIS-F, and pNIS-J. Cos-7 cells expressing the untagged (dark bars) or the N-Flag-tagged pNIS (gray bars) isoforms exhibited the same IUA. Despite a protein expression level and localization indistinguishable from that obtained with the other cDNAs, pNIS-J did not confer IUA to Cos-7 cells. To demonstrate that the deletion of the 6 residues in the putative transmembrane domain 10 (the difference between pNIS-J and J) was responsible for the lack of IUA, we generated a pNIS-D construct (pNIS-18) carrying the deletion of 18 nucleotides. No specific iodide transport was observed in cells expressing pNIS-18 (data not shown). Cotransfection of Cos-7 cells with equal amounts of pNIS-D and pNIS-18 or pNIS-J and pNIS-J resulted in a normal IUA (data not shown), showing that the inactive form of pNIS had no effect on the functional forms.
As shown in Fig. 3C, the three functional pNIS isoforms (pNIS-D, pNIS-F, and pNIS-J) transport iodide with the same Km value, about 20 µM.
Relationship between isolated pNIS cDNAs and thyroid pNIS transcripts
Complementary approaches including Northern blot, RT-PCR, and restriction analyses were used to establish a correspondence between isolated cDNA clones and porcine NIS transcripts expressed by pig thyrocytes in primary culture. As shown in Fig. 1, probe A (used for the isolation of the four cDNA clones) hybridized with transcripts of 3.5 and 3 kb. The relative abundance of the two bands did not vary with the culture conditions; they were 1) absent in cells cultured without TSH for 4 d, 2) present in high amounts when cells were cultured with TSH from the time of cell seeding, and 3) present at intermediate levels when cells were first cultured without TSH for 20 h and then activated by TSH for 3 d (Fig. 5A). Estimation of the ratio of these two bands in different experiments indicates that the amount of the 3.5-kb species was on average 1.7-fold higher than that of the 3-kb species. To demonstrate that pNIS-D clone actually corresponded to the 3.5-kb transcript (as suggested by their respective size), total RNA from cultured thyrocytes was subjected to Northern blot analysis using probe B, which specifically hybridized with pNIS-D cDNA. As expected, only the 3.5-kb transcript was detected. Consequently, the 3-kb transcript might possibly correspond to pNIS-F, J, and J clones lacking 492 bp, 538 bp, and 538 + 18 bp, respectively (as depicted at the top of Fig. 5). To identify the presence of mRNA corresponding to pNIS-F, J, and J, within the 3-kb molecular species, we designed two sets of primers (1/2 and ß1/ß2) flanking the deleted regions, as shown in Fig. 5B. The 1/2 primers should allow one to distinguish the transcript corresponding to pNIS-F from those corresponding to pNIS-J and pNIS-J cDNAs. Accordingly, two fragments of 295 and 249 bp were generated by RT-PCR from pig thyrocyte RNA. As controls, amplicons of the same sizes were generated by PCR from pNIS-F and pNIS-J cDNA vectors, respectively (Fig. 5B). The identity of the products amplified by RT-PCR and PCR was ascertained by restriction analyses (data not shown). Thus, the 3-kb band probably contained mRNA species corresponding to clones pNIS-F and pNIS-J and/or pNIS-J. It was estimated (from several experiments) that pNIS-F mRNA was six to eight times more abundant than pNIS-J plus pNIS-J mRNAs. To discriminate between mRNAs deriving from pNIS-J and pNIS-J which differ by only 18 nucleotides, we used the primers ß1/ß2 for PCR amplification and a digestion at a Bpu1 site only present in pNIS-J. Products amplified by PCR from vectors containing pNIS-J and pNIS-J cDNA were slightly different in size (expected size: 393 and 377 bp) and the amplicon generated from pNIS-J was cleaved by Bpu1 treatment, whereas that from pNIS-J remained intact. RT-PCR from total RNA from cultured thyrocytes yielded a band migrating as the 393-bp fragment, which was almost completely digested by Bpu1. This result suggests that, if present, pNIS-J mRNA was of very low abundance. It must be noticed that the amplification of pNIS-J mRNA species could be understated, as the most abundant transcripts, pNIS-D and pNIS-F were also amplified using ß1/ß2 primers. From the semiquantitative measurements of the relative abundance of the 3.5-kb and 3-kb transcripts on Northern blots and the relative abundance of pNIS-F, pNIS-J, and pNIS-J-related 3-kb transcripts by RT-PCR, one can estimate that the transcripts corresponding to clone D, F, J, and J could account for about 60, 35, 5, and less than 1% of total pNIS mRNA.
Origin of the different pNIS isoforms
To identify the mechanism whereby several transcripts and consequently several NIS proteins differing in their C-terminus are generated, we analyzed the porcine genomic DNA corresponding to the 3' part of the NIS gene. Primers were designed on the basis of the known structure of the hNIS gene consisting of 15 exons interrupted by 14 introns (5). Porcine genomic DNA regions spanning exon 8 to exon 10 and exon 14 to exon 15 (numbering of exons refers to hNIS gene) were amplified, cloned, and sequenced. The exon/ intron boundaries of the pNIS gene were established on the basis of consensus spliced signal sequences according to the gt/ag rule for RNA splicing (23), and the sequence homologies between the four pNIS cDNA clones schematically illustrated in Fig. 6. Two alternative splice donor sites (GT1 and GT2) and one alternative splice acceptor site (AG1) located within the last exon (exon 15 by reference to hNIS gene) were unequivocally identified. The use of GT2/AG1 splice sites allows the skipping of the stop codon operating in the primary transcript and generates pNIS-F that differs from pNIS-D by a 492-nucleotide deletion in the 3' translated and untranslated region (segment 2). The deletion does not alter the reading frame and the use of a downstream stop codon results in a protein shorter than pNIS-D by 6 residues (637 instead of 643 amino acids) (Fig. 6A, Table 1). The use of GT1/AG1 splice sites leads to a deletion of the same region plus an additional upstream sequence of 46 nucleotides (segment 1) and results in a frame shift that produces pNIS-J with an extended (665 amino acids instead of 643 or 637 for forms D and F) and markedly different C terminus. A second alternative splice acceptor site (AG2) was found in exon 10. The use of the regular donor splice site of exon 9/intron 9 boundary and the acceptor site AG2 (Fig. 6B) leads to a deletion of 18 nucleotides within exon 10 (segment 3), which does not alter the reading frame and results in a six-amino-acid shorter pNIS-J protein (659 instead of 665 amino acids). All the alternative splicing events identified in the porcine species would specifically occur in that species. Indeed, the donor splice sites GT1 and GT2 and the acceptor splice sites AG1 and AG2 are either absent or present in inappropriate combination (to generate splice variants) in human, rat, or mouse NIS genes (Fig. 7).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
J clone could not be clearly identified from porcine thyrocyte RNA.
Using the splice site prediction program obtained by neural network (www.fruitly.org), we tried to evaluate the frequency of use of the consensus splice sites found in the pNIS gene. The program gives a score for the usage frequency that is based on the analysis of nucleotides at each position of the splice complex. A splice complex containing the most frequent bases at each position would have a consensus value of 1, whereas that containing the least frequent bases would yield a consensus value of 0. Thus, by comparing the consensus value of two splice sequences, one can identify which one is most likely prone to be used. The consensus values obtained for the splice donor sites named GT2 and GT1 are 0.54 and 0.08, respectively. The splice acceptor site named AG1 within the last exon of pNIS had a score of 0.98, suggesting that it could be used with a high frequency. Thus, the relative score obtained for the two donor sites GT2 and GT1 i.e. 0.54/0.08 = 6.75 may give the relative frequency of the use of GT2/AG1 vs. GT1/AG1 and the relative proportion of pNIS-J and pNIS-F transcripts. Interestingly, the predicted value of 6.75 is very close to the estimated value of about 7 found for the pNIS-F/pNIS-J transcript ratio. Concerning the splice acceptor site AG2 present in the putative exon 10 expected to generate pNIS-J, we compared its consensus value with that of the conserved acceptor site of intron 9 (normally used in the generation of all pNIS transcripts). Values obtained (0.79 for the conserved acceptor site and only 0.04 for AG2) indicate that the acceptor site AG2 is probably used at a very low frequency and that the pNIS transcript lacking the 18-bp segment in the putative exon 10 is of very low abundance. The impossibility to firmly identify this transcript by RT-PCR amplification from thyroid cell RNA is in keeping with this hypothesis. Finally, information drawn from computer analyses appears in full agreement with the experimental data.
The alternative splicing events occurring on pNIS transcripts are probably not subjected to regulation. Indeed, the relative proportion of the 3.5- and 3-kb molecular species as well as the relative proportion of the pNIS-F and J transcripts (within the 3-kb band) did not markedly vary in thyroid cells upon generation of transcripts following TSH activation of transcription or upon disappearance of transcripts following TSH withdrawal. Despite the presence of TSH, the pNIS transcript level markedly decreased within the first 24 h of culture of freshly dispersed porcine thyrocytes. This decline does not seem to be related to a period of TSH refractoriness that could result from TSH receptor alterations caused by the proteolytic enzyme treatment used for cell isolation. Indeed, within the same period of time, thyrocytes readily respond to TSH in forming cell aggregates from which thyrocytes reconstitute follicle structures. The decline in NIS transcript level that was similar in the presence and in the absence of TSH, would rather reflect a transient and probably selective decrease of NIS gene transcription or increase of mRNA degradation; we observed that the level of transcripts of other thyroid-specific genes, thyroglobulin, Pax-8 did not significantly change during the same period of time.
Recombinant proteins expressed in Cos-7 cells from pNIS-D, F, J, and J clones had a molecular mass (
62 kDa) lower than that expected from available data on rNIS, mNIS, and hNIS. Indeed, in these species, both thyroid NIS and recombinant NIS proteins (expressed in Cos-7 or CHO cells) are glycoproteins with an apparent molecular mass ranging from 80–90 kDa (7, 11, 15, 24). Using anti-pNIS antibodies recently generated in our laboratory, we found that thyroid pNIS had the same molecular mass as murine NIS or hNIS and that its deglycosylation using N-glycosidase F led to 60-to 65-kDa polypeptides (Trouttet-Masson, S., F. Bernier-Valentin, S. Selmi-Ruby, and B. Rousset, unpublished data). Thus, the recombinant pNIS proteins produced from Cos-7 cells in this study correspond to nonglycosylated proteins. This is very likely due to a limitation of Cos-7 cells that we used; indeed, we observed a similar lack of glycosylation for another recombinant protein expressed by the same cells. Although performed on nonglycosylated proteins, our functional data should be fully reliable because Carrasco and her colleagues (11) have convincingly demonstrated that glycosylated, partially glycosylated, and nonglycosylated rNIS had the same activity and iodide transport kinetics.
Recombinant proteins expressed from pNIS-D, F, and J (for which the natural mRNAs have been firmly identified) differ in their C terminus and are all endowed with a function of active iodide transporter eliciting very close, if not identical, kinetics. These data indicate that the sequence and/or the length of the C-terminal intracellular domain of NIS (at least the last 23 amino acids) are not determinant for the activity of the protein. A similar conclusion could be drawn from the sequence comparison between hNIS and murine NIS. Indeed, compared with hNIS, the functional rNIS and mNIS lack 18 of the last 24 amino acids. The immunofluorescence detection of the flag epitope positioned at the N-terminal end of the recombinant pNIS proteins, without membrane permeabilization, confirms that the NIS protein has its N terminus outside the cells.
The characterization of a nonfunctional NIS protein (pNIS-J), deleted of 6 amino acids (391–396) in the putative transmembrane domain 10, suggests that at least one of these amino acids is important for the ability of the protein to transport iodide. This observation is in keeping with data obtained in patients with congenital hypothyroidism in which a missense mutation G395R on hNIS causes an iodide transport defect (25). Like the pNIS-J protein, the G395R hNIS mutant protein was properly targeted to the plasma membrane of Cos-7 cells. In terms of structure-activity relationship, it is worth noticing that the residues with a potential importance for the function and/or the stability of the protein (26, 27) corresponding to identified mutations of hNIS in position: G93, Q267, C272, T354, that cause a thyroid iodide transport defect and congenital hypothyroidism, are conserved in the three functional pNIS proteins. The nonfunctional pNIS-J protein did not exert any dominant negative effect in cotransfection experiments with one or the other active isoforms.
At the moment, we do not know whether the existence of several active NIS protein isoforms in the porcine species instead of only one in the human, rat, and mouse species could have physiological implications for the thyroid iodide economy. We already know that protein isoforms do not differ in their affinity for iodide. Production of antibodies capable of discriminating the three pNIS protein isoforms would allow one to analyze whether they have distinct cellular properties.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received September 16, 2002.
Accepted for publication December 2, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  R. Opitz, A. Trubiroha, C. Lorenz, I. Lutz, S. Hartmann, T. Blank, T. Braunbeck, and W. Kloas Expression of sodium-iodide symporter mRNA in the thyroid gland of Xenopus laevis tadpoles: developmental expression, effects of antithyroidal compounds, and regulation by TSH. J. Endocrinol., July 1, 2006; 190(1): 157 - 170. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Bernier-Valentin, S. Trouttet-Masson, R. Rabilloud, S. Selmi-Ruby, and B. Rousset Three-Dimensional Organization of Thyroid Cells into Follicle Structures Is a Pivotal Factor in the Control of Sodium/Iodide Symporter Expression Endocrinology, April 1, 2006; 147(4): 2035 - 2042. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Zhang, Y.-Y. Liu, and S. M. Jhiang Cell Surface Targeting Accounts for the Difference in Iodide Uptake Activity between Human Na+/I- Symporter and Rat Na+/I- Symporter J. Clin. Endocrinol. Metab., November 1, 2005; 90(11): 6131 - 6140. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Porra, C. Ferraro-Peyret, C. Durand, S. Selmi-Ruby, H. Giroud, N. Berger-Dutrieux, M. Decaussin, J.-L. Peix, C. Bournaud, J. Orgiazzi, et al. Silencing of the Tumor Suppressor Gene SLC5A8 Is Associated with BRAF Mutations in Classical Papillary Thyroid Carcinomas J. Clin. Endocrinol. Metab., May 1, 2005; 90(5): 3028 - 3035. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Frohlich, A. Witke, B. Czarnocka, and R. Wahl Retinol has specific effects on binding of thyrotrophin to cultured porcine thyrocytes J. Endocrinol., December 1, 2004; 183(3): 617 - 626. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Faivre, J. Clerc, R. Gerolami, J. Herve, M. Longuet, B. Liu, J. Roux, F. Moal, M. Perricaudet, and C. Brechot Long-Term Radioiodine Retention and Regression of Liver Cancer after Sodium Iodide Symporter Gene Transfer in Wistar Rats Cancer Res., November 1, 2004; 64(21): 8045 - 8051. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Trouttet-Masson, S. Selmi-Ruby, F. Bernier-Valentin, V. Porra, N. Berger-Dutrieux, M. Decaussin, J.-L. Peix, A. Perrin, C. Bournaud, J. Orgiazzi, et al. Evidence for Transcriptional and Posttranscriptional Alterations of the Sodium/Iodide Symporter Expression in Hypofunctioning Benign and Malignant Thyroid Tumors Am. J. Pathol., July 1, 2004; 165(1): 25 - 34. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
