A novel gene, thyroid cancer 1 (TC-1), was found recently to be overexpressed in thyroid cancer. TC-1 shows no homology to any of the known thyroid cancer-associated genes. We have produced stable transformants of normal thyroid cells that express the TC-1 gene, and these cells show increased proliferation rates and anchorage-independent growth in soft agar. Apoptosis rates also are decreased in the transformed cells. We also have expressed recombinant TC-1 protein and have undertaken a structural and functional characterization of the protein. The protein is monomeric and predominantly unstructured under conditions of physiologic salt and pH. This places it in the category of natively disordered proteins, a rapidly expanding group of proteins, many members of which play critical roles in cell regulation processes. We show that the protein can be phosphorylated by cyclic AMP-dependent protein kinase and protein kinase C, and the activity of both of these kinases is up-regulated when cells are stably transfected with TC-1. These results suggest that overexpression of TC-1 may be important in thyroid carcinogenesis.
Thyroid cancer is the most common endocrine malignancy, accounting for 90% of all of the endocrine cancers (1) . The incidence varies substantially around the world, with rates continuing to increase in many countries during the past several decades (2, 3, 4, 5) . Papillary thyroid carcinoma is the most common histologic type. In general, these are slow-growing malignancies with good prognoses; however, for a subset of patients, more aggressive metastatic disease develops. Treatment options vary from high-dose radioiodine treatment to surgical resection of metastatic foci, and although effective therapy reduces morbidity and mortality, some continue to have persistent aggressive disease (1 , 6) .
Our understanding of the molecular events that lead to the development and progression of thyroid tumors remains limited. There is a list of candidate genes with implicated roles in thyroid tumor formation, including ret/ptc, trk, and ras oncogenes, activating mutations of BRAF, the tumor suppressor gene p53, Pax8/peroxisome proliferator-activated receptor δ rearrangement, and candidate tumor suppressor genes on chromosomes 11q13, 3p, and 7q (7, 8, 9, 10, 11, 12, 13, 14, 15, 16) . However, there remain thyroid carcinomas for which none of the currently identified genes or mutations can be identified. This suggests that other novel genetic defects are involved in the initiation and/or progression of the disease.
We previously identified a novel gene, thyroid cancer 1 (TC-1; C8orf4), which is overexpressed in papillary thyroid cancer (17) . The TC-1 sequence shows no homology to any known gene, and no function has yet been assigned. It is expressed ubiquitously across a wide range of human tissues and also shows high homology at the protein sequence level to mouse, cattle, and chicken sequences. The sequence conservation and ubiquitous expression suggest TC-1 may have an important role in cell biology.
In the present study, our goals were to characterize the biophysical properties of the TC-1 protein and to explore the biological role of TC-1 in thyroid cells. The oligomeric state of the purified protein has been determined by analytical ultracentrifugation and its secondary structure probed by circular dichroism (CD) and nuclear magnetic resonance (NMR) spectroscopy. We also have produced stable transformants of normal thyroid cells that express the TC-1 gene, and these cells show increased proliferation rates and anchorage-independent growth in soft agar. We also demonstrate that apoptosis rates in these transformed cells are decreased. We show that the activities of cyclic AMP (cAMP)-dependent kinase (PKA) and protein kinase C (PKC) also are increased and that both of these kinases phosphorylate TC-1. Taken together, these data define the conformational properties of TC-1 and suggest that overexpression of TC-1 may be important in the pathogenesis and/or progression of thyroid cancer.
MATERIALS AND METHODS
Reagents for cloning and bacterial expression were from New England Biolabs (Beverly, MA), Novagen (Madison, WI), Invitrogen (Carlsbad, CA), MBI Fermentas (Hanover, MD), and Amyl Media Pty (Dandenong, Victoria, Australia). General laboratory biochemicals were from Asia Pacific Specialty Ltd. (Seven Hills, NSW, Australia) and Sigma-Aldrich (St. Louis, MO). Ni-NTA resin was obtained from Qiagen (Valencia, CA), and Ni-IDA was from Scientifix (Clayton, Victoria, Australia).
In Vitro Expression and TC-1 Purification
The gene for TC-1 was inserted into a pET-15b vector (Novagen) using NdeI and BamHI restriction sites. TC-1 protein was expressed as a His6-tagged protein in Escherichia coli BL-21 (DE3) cells. The protein formed inclusion bodies, and after separation from soluble cellular material, the inclusion bodies were solubilized in 0.1 m Tris HCl, 6 m GuHCl, and 1 mm EDTA (pH 8.0) by stirring at room temperature for 2 h. The pH of the solution then was adjusted to between pH 3 and 4, and insoluble material was removed by centrifugation. The solution was diluted fivefold with 5 mm imidazole, 20 mm Tris HCl, 0.5 m NaCl, and 0.75 m GuHCl (pH 7.9) before loading onto Ni-NTA agarose or Ni-IDA agarose charged with Ni2+. TC-1 was eluted with 1 m imidazole, 20 mm Tris HCl, 0.5 m NaCl, and 0.75 m GuHCl (pH 7.9). Fractions containing protein were pooled, and the pH of the sample was adjusted to <6.0 by addition of HCl. The sample then was dialyzed against 10 mm sodium acetate (pH 5.0). TC-1 also was purified by reverse-phase high-performance liquid chromatography on a C18 column (RCM semipreparative module, Waters, Milford, MA) using a gradient of acetonitrile in water with 0.1% trifluoroacetic acid.
Far-UV CD spectra were recorded on a J-720 spectropolarimeter (Jasco, Easton, MD) equipped with a Neslab RTE-111 temperature controller (Thermo Electron Corporation, Marietta, OH). CD data were collected using a 1-mm cuvette over the wavelength range of 190–250 nm and a speed of 20 nm/min, a resolution of 0.5 nm, a bandwidth of 1 nm, and a response time of 1 s. Final spectra were the average of three scans and were baseline corrected. The protein concentration was 0.21 mg/ml, estimated from A280 using E = 2560 m/cm, in 10 mm sodium acetate containing varying concentrations of NaF.
TC-1 was dissolved in 10 mm sodium acetate (pH 6.8) containing 150 mm NaCl at a concentration of 1.7 mg/ml (120 μm). NMR experiments were performed on a DRX-600 spectrometer (Bruker, Billerica, MA) equipped with a 5-mm triple-resonance gradient probe. Spectra were acquired at 293 K. All of the NMR spectra were recorded using a spectral width of 7200 Hz. Water suppression was achieved using pulsed field gradients and the water suppression by gradient-tailored excitation sequence (18) .
One-dimensional 1H spectra were acquired as 8192 data points. For the nuclear Overhauser effect spectroscopy (NOESY) experiment, 2048 points were acquired in F2, and 512 complex points were acquired in F1. The mixing time was 300 ms. Spectra were processed with Gaussian window functions and polynomial baseline correction functions using the Bruker XWINNMR software.
TC-1 protein was prepared for analytical ultracentrifugation by buffer exchange into 50 mm sodium acetate and 150 mm NaCl (pH 6.8) using a Centricon centrifugal filter device (Millipore, Billerica, MA) and then diluted to give concentrations of 1.25, 0.625, and 0.313 mg/ml. Equilibrium sedimentation experiments were performed using a Beckman model XL-A analytical ultracentrifuge (Beckman Coulter, Fullerton, CA) equipped with an An-60 Ti rotor. Data were collected in six-sector cells as absorbance (238 nm) versus radius scans (0.001-cm increments; 10 averages) at 24,000, 28,000, and 40,000 rpm. Scans were collected at 3-h intervals and compared to determine when the samples reached equilibrium. Analysis of the data was carried out using NONLIN software, and the best model and final parameters were determined by examination of the residuals derived from fits to several models (19) . The partial specific volume (0.736 ml/g) was determined from the sequence of TC-1 and adjusted for temperature using the program SEDNTERP (20) . The buffer density and viscosity were taken to be 1.00684 g/ml and 0.010295 Poise at 20°C, respectively.
PKA and PKC Phosphorylation Experiments
The catalytic subunits of PKA and PKC were purchased from Sigma-Aldrich and used without additional purification. TC-1 was prepared in 50 mm Tris and 2 mm MgCl2 (pH 7.5) at a concentration of 0.5 mg/ml for the PKA reactions or 1 mg/ml in 50 mm HEPES, 10 mm MgCl2, and 2 mm MnCl2 (pH 7.5) for the PKC reactions. For the PKA reactions, 55 nmol TC-1 were present per unit of PKA; for the PKC reactions, 60 μmol were present per unit of kinase. The reaction mixture was incubated overnight at 30°C and then quenched by addition of trifluoroacetic acid to 0.02%. Samples were prepared for mass spectrometry by treatment with a C18 ZipTip (Millipore) according to manufacturer’s instructions and analyzed by mass spectrometry using an LCQ mass spectrometer (Finnigan, San Jose, CA) to confirm the addition of a covalently bound phosphate group to the protein.
TC-1 that had been phosphorylated was digested with trypsin (Sigma-Aldrich) at a molar ratio of TC-1 to enzyme of 16:1 in 50 mm Tris (pH 8.5) containing 10% acetonitrile. The reaction was allowed to proceed for 18 h at room temperature. Peptide fragments were purified by reverse-phase high-performance liquid chromatography and identified by mass spectrometry as described previously.
The primary sequence of TC-1 was submitted for sequence alignment and secondary structure prediction at PredictProtein (Refs. 21 , 22 ; http://cubic.bioc.columbia.edu/predictprotein/) and for analysis of native disorder propensity to the Predictor of Natural Disordered Regions (PONDR; http://www.pondr.com; Refs. 23 , 24 ).
The mean hydrophobicity and the net charge of TC-1 were calculated as suggested by Uversky et al. (25) . The “boundary” mean hydrophobicity value (H) was calculated as follows: (H)boundary = (r + 1.151)/2.785 (26) . This equation, in which r corresponds to the mean net charge of the polypeptide, allows the calculation of a boundary hydrophobicity value for the protein below which it is predicted to have an unfolded native state.
The human thyroid cell line (Nthy), derived from normal thyrocytes transformed with SV40, was maintained in RPMI (Invitrogen) supplemented with 10% FCS (Invitrogen). The TC-1-expressing cell clones also were maintained in RPMI supplemented with 10% FCS and 250 μg/ml Zeocin (Invitrogen). The cell cultures were maintained routinely at 37°C in a humidified atmosphere of 95% air and 5% CO2, fed every 48–96 h, and subcultured 1:3 at 80–90% confluency.
RNA Extraction and Reverse Transcription-PCR
Total RNA was extracted with TRI reagent (Sigma-Aldrich). The quality of the extracted RNA was determined by absorbance (ratio A260 to A280), and quantity was measured by the SYBR Green II assay (Molecular Probes Inc., Eugene, OR). Two hundred fifty ng of total RNA were reverse transcribed into cDNA using random hexamers (50 ng) and Superscript II RNaseH− (Invitrogen). For real-time PCR, 10% of the reverse-transcription reaction was amplified using SYBR Green I core PCR reagents (Applied Biosystems, Foster City, CA) in a PE7700 machine (Applied Biosystems) using 20 pmol of each primer (TC-1 and β2-microglobulin; Refs. 17 , 27 ). Two-step cycling was performed at 94°C for 15 s and 68°C for 60 s for 45 cycles. Differences in expression were determined using the ΔΔCT method (28) .
Generation of Nthy Cells Overexpressing TC-1
Human TC-1 cDNA was subcloned into the pZEOSV2+ (Invitrogen) plasmid. Plasmid DNA was prepared by Qiagen MidiPrep, and 0.1 μg plasmid were used to transfect Nthy cells using Effectene (Qiagen). Twenty-four h after transfection, the cells were incubated in 250 μg/ml Zeocin (Invitrogen). This dose of Zeocin was determined by performing a Zeocin kill-curve with Nthy cells. This was the optimal dose that killed Nthy cells within 48 h. After a 6-week selection period, multiple clones of cells overexpressing TC-1 and vector without insert were obtained. Three TC-1-overexpressing clones (Clone 1, Clone 4, and Clone 6) were chosen randomly for study. Clones 1, 4, and 6 showed 20,000-, 4,000-, and 100,000-fold higher mRNA expression compared with parental Nthy cells. Two clones (Zeo-1 and Zeo-3) containing vectors without the TC-1 insert were chosen as transfection controls for the study. Parental (Nthy) cells also were used as controls.
Cells (1 × 104 cells/well) were plated in triplicate in 96-well plates containing RPMI supplemented with 10% FCS and Zeocin 250 μg/ml (except parental cells in which no Zeocin was added). After 24 h, cells were changed to RPMI medium supplemented with 0.1% FCS. Cell proliferation was assessed every 24 h for 96 h using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide assay (Cell Titer Proliferation Assay; Promega, Madison, WI). The dye taken up by cells during a 1-h incubation was measured by OD495 nm using an automated microplate reader.
Cells were seeded separately at 2 × 105 cells/well of a six-well dish in 2 ml of 0.3% agar in RPMI supplemented with Zeocin at 250 μg/ml (except parental cells in which no Zeocin was added). The suspension then was overlaid with 2 ml of 0.6% agar in complete medium. The cells were incubated in a humidified atmosphere of 95% air and 5% CO2 at 37°C. Cells were fed twice weekly with 1 ml of medium containing 0.3% agar without removing the basal medium. After 3 weeks, colonies were stained by 0.05% crystal violet, and colonies consisting of >20 cells then were counted with the aid of a microscope. Each assay was performed in triplicate.
TUNEL Apoptosis Assay
Cells were seeded on culture slides in RPMI supplemented with 10% FCS, and media then were changed to RPMI supplemented with 0.1% FCS for 48 h. The terminal deoxynucleotidyl transferase-mediated nick end labeling (TUNEL) method then was used to measure the apoptotic state of individual cells (DeadEnd; Promega). Slide images were captured digitally, and two independent investigators counted positive cells within a defined area (100,000 units = 80 cells).
COMET Apoptosis Assay
Cells were seeded in 25-cm2 culture dishes in RPMI supplemented with 10% FCS, and the media then were changed to RPMI supplemented with 0.1% FCS for 48 h. The COMET Assay (Trevigen, Inc., Gaithersburg, MD) was used to measure the amount of DNA fragmentation in individual cells. Slide images were captured digitally, and the tail lengths of at least 75 cells per slide were measured using Adobe Photoshop (San Jose, CA).
Caspase Activity Assay
Transfectants were seeded at 1 × 104 cells/96-well, and after overnight recovery, they were changed to 0.1% FCS. Forty-eight h later, caspase activity was measured using the CaspACE assay system (Promega). This system is based on caspase-3-dependent cleavage of the substrate, Ac-DEVD-pNA, which releases pNA and produces a yellow color. Color development is measured at OD405 nm. Specificity is determined by inhibiting caspase-3 activity with the addition of z-Val-Ala-Asp-fluoromethyl ketone to cultured cell media at the beginning of the experiment.
PKC, PKA, and PTK Assays
Cells were seeded at 5 × 106 cells in six-well culture dishes. After overnight recovery, media were changed to RPMI with 250 μg/ml Zeocin and 0.1% FCS for 48 h. Cells then were washed with PBS and homogenized in 1 ml of extraction buffer [25 mm Tris HCl (pH 7.4), 0.5 mm EDTA, 0.5 mm EGTA, 10 mm β-mercaptoethanol, 10 μl/ml protease inhibitor mixture (Sigma-Aldrich), and 0.5 mm phenylmethylsulfonyl fluoride]. Lysate was centrifuged at 14,000 × g for 5 min at 4°C in a microcentrifuge and supernatant from five wells, then passed through a 1-ml column of DEAE cellulose (pre-equilibrated with buffer). Column was washed once with extraction buffer, and the PKC-containing fraction then was eluted with 5 ml extraction buffer containing 200 mm NaCl. PKC activity was measured immediately using the PepTag PKC assay system (Promega) following the manufacturer’s instructions.
Cells were cultured as for the PKC assay. After washing, cells were homogenized in 1 ml extraction buffer (as for PKC) and centrifuged at 14,000 × g for 5 min at 4°C. Lysates were assayed immediately for PKA activity using the PepTag protein kinase A assay system (Promega).
Cells were cultured in 75-cm2 flasks. At 90% confluency, media were changed to RPMI with 250 μg/ml Zeocin and 0.1% FCS for 48 h. Cells then were washed twice with PBS. Cells were solubilized by adding 2 ml of extraction buffer [20 mm HEPES (pH 7.4), 1% Triton-X 100, 5 mm EDTA, 50 mm NaCl, plus 10 μl/ml protease inhibitor mixture (Sigma-Aldrich)] and incubating on ice for 10 min. Cells then were scraped, and the cell suspension was mixed by gentle pipetting before incubating for 15 min at 4°C, while rocking. The suspension then was centrifuged at 100,000 × g for 1 h at 4°C, and the supernatant was aliquoted to separate tubes and stored at −70°C. Protein tyrosine kinase (PTK) activity was measured using the SignaTECT PTK Assay System (Promega), following the manufacturer’s instructions.
The statistical significance of differences between the control and TC-1-expressing groups was estimated by the two-tailed t test.
Isolation of Stable TC-1 Protein.
TC-1 (sequence shown in Fig. 1 ⇓ ) was highly expressed, but most TC-1 protein was found in inclusion bodies. These were solubilized in 6 m GuHCl, and TC-1 then was purified by metal chelation affinity chromatography using buffers containing 0.75 m GuHCl. After purification, a single band corresponding to TC-1 was observed by SDS-PAGE. The molecular mass of the expressed protein is 14,384.0 ± 0.6 Da, in agreement with the mass predicted from the protein sequence (14,383.5 Da).
This affinity-purified protein was prone to precipitation at neutral pH in the absence of 0.75 m GuHCl. However, we observed that TC-1 could be resolubilized after precipitation by adjusting the pH of the solution to <pH 4–5. The protein also was considerably more soluble in acetate buffer than in phosphate-containing buffers. This, combined with the predicted isoelectric point, led us to consider the possibility that TC-1 might be binding and copurifying with nucleic acid. We used a set of oligonucleotide probes designed to include all of the possible 5-bp combinations (29) and performed electrophoretic mobility shift assays in the presence of purified TC-1 (data not shown). The protein formed large, soluble aggregates with the oligonucleotide probes, but we could not detect any specificity in the association. It appears that the protein binds considerable amounts of nucleic acid, probably through ionic interactions because the isoelectric point of the protein is >10 (measured by isoelectric focusing). After reverse-phase high-performance liquid chromatography, TC-1 was found to be much more soluble and stable.
Biophysical Characterization of TC-1.
Having isolated and purified TC-1, we next used analytical ultracentrifugation to determine the oligomeric state of the protein in solution (Fig. 1) ⇓ . The data fitted well to a monomeric model with an experimentally determined molecular mass of TC-1 of 15,600 Da.
Far-UV CD spectropolarimetry was used to study the secondary structure of TC-1 as a function of pH and salt concentration. These conditions were varied because of the observation that solubility of the protein was affected dramatically by the pH of the solution and the nature of the buffer salt. NaF, rather than NaCl, was used as the salt because Cl− ions in the latter interfere with the CD spectra. The position and intensity of the features in the far-UV CD spectrum are an indication of the secondary structure content of a protein. Fig. 2 ⇓ illustrates the difference in protein backbone conformation as a function of pH and in the presence of salt. At pH 3, the spectrum is dominated by the minimum at 202 nm, indicating that the protein has a mainly random coil conformation. The secondary structure of the protein undergoes a change with increasing pH; at pH 7, the spectrum shows a reduction in the size and shift in the position of the dominant minimum to 205 nm, an increase in ellipticity around 190 nm, and the development of a local minimum at 222 nm. These changes are consistent with a small increase in helical structure, but the protein conformation still is predominantly disordered. Addition of NaF to concentrations <500 mm appears to increase the helical content of the protein. It seems clear that the conformation of TC-1 is influenced strongly by the ionization state of charged groups.
One-dimensional and two-dimensional homonuclear NMR experiments also were used to characterize the structure of TC-1 (Fig. 3) ⇓ . The narrow peak widths and lack of chemical shift dispersion in the one-dimensional 1H NMR spectrum are consistent with the protein being monomeric and largely disordered in 10 mm sodium acetate (pH 6.8) and 150 mm NaCl (Fig. 3, A and B) ⇓ . The absence of a substantial number of cross-peaks in the two-dimensional NOESY spectrum indicates that the backbone of TC-1 is not folded into regular secondary structure and confirms that TC-1 does not have a stable globular three-dimensional structure (Fig. 3C) ⇓ .
Sequence Analysis of TC-1.
The PROF component of the PredictProtein sequence analysis suite classified TC-1 as an all-α protein. The α-helical content was predicted to be >55%, and the β structure was <5%, with the helical structure located in the COOH-terminal half of the protein between residues 45 and 95 (Fig. 4) ⇓ . This prediction is not consistent with our biophysical evidence that purified, monomeric TC-1 does not have a stable three-dimensional structure. Although the presence of the N-terminal His-tag may be expected to contribute to the random coil component observed in the far-UV CD spectrum, the absence of cross-peaks in the two-dimensional NOESY spectrum of TC-1 indicates that the protein does not have a folded structure. The limited amount of secondary structure under conditions of physiologic pH and salt concentration and the conformational plasticity displayed suggested to us that TC-1 could be a member of the increasingly large group of proteins known to be “natively unfolded.” These are proteins that have little or no ordered secondary structure under physiologic conditions of salt and pH, although some can undergo disorder-to-order transitions on binding to a target molecule (26 , 30 , 31) .
These results prompted us to analyze the sequence of TC-1 further using the series of neural network predictors that use amino acid sequence to predict disorder in a given region of a protein. The PONDR has been trained on a series of disordered regions using the premise that as sequence determines structure, sequence also should determine lack of structure (30) . Fig. 4 ⇓ illustrates the results of the PONDR and PROF predictions. PONDR predicts that TC-1 will be ∼60% disordered, with a long stretch of 52 residues in the COOH-terminal half of the protein (38–89) strongly predicted to be disordered. The main regions predicted by PROF to be helical fall within the stretch predicted by PONDR to be natively unfolded. The mean hydrophobicity and mean net charge of TC-1 were calculated by the method of Uversky et al. (25) to be 0.42 and 0.10, respectively. The boundary hydrophobicity for TC-1 was calculated to be 0.45. This indicates that the hydrophobicity of TC-1 (0.42) falls below the boundary value required to fold a protein of this net charge in a stable, globular form. Therefore, by this analysis TC-1 is predicted to be natively disordered.
Phosphorylation of TC-1.
Given the identification of possible sites of phosphorylation by the PROSITE motif search in PredictProtein and the results of a sequence analysis by NetPhos (32) , we tested whether purified TC-1 could be phosphorylated by the kinases PKA and PKC in vitro (33) . Incubation with PKA resulted in the covalent addition of a single phosphate group, as shown by the 80-Da mass increase (mass before incubation with PKA, 14,384.0 ± 0.9 Da; after incubation, 14,464.3 ± 0.7 Da). The site of phosphorylation was identified by tryptic digestion and mass spectrometric analysis as Ser102, five residues from the end of the polypeptide chain. The phosphorylated protein was purified and analyzed by far-UV CD, revealing that phosphorylation at this site does not affect the conformation of TC-1, which remained a random coil. Incubation with PKC gave rise to TC-1 with one, two, or three phosphate groups attached (masses after incubation with PKC, 14,464.1 ± 2.0 Da, 14,543.0 ± 0.1 Da, and 14,632.0 ± 3.3 Da; one, two, and three phosphate groups, respectively). These species could not be separated for analysis using far-UV CD. In the case of PKA, it seems that although phosphorylation may be the trigger for involvement of TC-1 in a cellular reaction cascade, it does not induce a conformational change on its own.
Growth Characteristics of TC-1-Transfected Nthy Cells.
Natively unfolded proteins often are involved in signaling, regulation, and control. To investigate such a role for TC-1 in thyroid malignancy, we established stable transformants of Nthy cells that constitutively express TC-1 as determined by real-time PCR. Clone 6 was found to demonstrate highest TC-1 expression (100,000-fold higher than Nthy cells). Clone 1 and clone 4 showed lower levels of expression (20,000- and 4,000-fold higher than Nthy cells, respectively). To evaluate the effect of overexpression of the TC-1 gene on the growth of the Nthy cell line, we examined the proliferative capacity of the various cell lines (parental, control vector transfected, and TC-1 transfected). Fig. 5 ⇓ shows that in 0.1% FCS-containing medium, the TC-1-transfected cell lines showed similar growth rates as the parental or control vector-transfected cell lines; however, the TC-1 clones showed greater survival rates by 48 h, which were sustained up to 96 h. These data suggest that overexpression of TC-1 modifies the survival characteristics of the Nthy cell line in vitro.
Effect of TC-1 on Anchorage-Independent Growth of Nthy Cells.
To evaluate the effects of TC-1 on anchorage-independent growth, we performed colony formation assays using Nthy parental cells, transfectants with empty vector, and transfectants with TC-1 DNA. On day 21 after seeding in soft agar, virtually no colonies per well were observed for control cells (parental or empty vector). In contrast, 19 ± 7, 47 ± 11, and 65 ± 14 colonies per well were observed for clone 1, clone 4, and clone 6, respectively (Fig. 6) ⇓ .
Effect of TC-1 on Cellular Apoptosis.
Parental cells, TC-1 clones, and empty vector controls were analyzed for apoptosis by examining DNA fragmentation (TUNEL and COMET assays, late apoptosis) and caspase-3 activity (early apoptosis). Signals for TUNEL were detected predominantly in the nuclei of control cells, whereas virtually no positive signals were detected in the TC-1 clones (Fig. 7) ⇓ . The COMET assay, which is another method to examine DNA fragmentation, showed that control cells, parental Nthy cells, zeo-1, and zeo-3 had increased levels of apoptosis than the TC-1-expressing clones (Fig. 8) ⇓ . By contrast, TC-1 clones showed higher activity of an apoptotic enzyme, caspase-3, relative to control cells (Fig. 9) ⇓ .
Effect of TC-1 on Cell Signaling.
We tested the effect of TC-1 expression on PKC, PKA, and PTK activities because these are the major signaling pathways in thyroid cells (34 , 35) . TC-1 clones were found to have higher PKA and PKC activities compared with Nthy parental cells or empty vector clones. There were no differences in PTK activity between TC-1 clones, empty vector clones, and parental Nthy cells (Fig. 10) ⇓ .
The TC-1 gene is highly expressed in papillary thyroid carcinoma compared with normal thyroid tissue. The TC-1 gene sequence shows no homology to other gene sequences but is highly conserved across several phyla. Furthermore, it is expressed across a wide range of tissues (17) .
To gain insight into the structure and function of this novel protein, we first expressed it recombinantly and purified it using standard chromatographic techniques. Sedimentation equilibrium experiments showed that TC-1 is strictly monomeric in solution, and CD measurements revealed that the protein contains little regular secondary structure under conditions of physiologic pH and NaCl concentration, although there is perhaps a small amount of nascent helical structure. The poor chemical shift dispersion and lack of cross-peaks in the two-dimensional NOESY spectrum confirmed the CD results, indicating that TC-1 is a natively unfolded protein (26 , 30 , 31 , 36) . Amino acid sequence analysis has demonstrated that natively disordered proteins are characterized by a combination of low mean hydrophobicity and relatively high net charge (25) . The primary sequence of TC-1 shares these features, with its average hydrophobicity of 0.42 falling below the boundary hydrophobicity of 0.45 calculated for a protein of its mean net charge (0.1).
Many natively unfolded proteins are known to perform key functions in the cell in the areas of cell cycle control and transcriptional and translational regulation. This rapidly expanding group of proteins includes the BclII antiapoptotic protein, calcineurin, cyclin-dependent kinase inhibitor p21, α-synuclein, high mobility group proteins, and transcription factors c-Fos and c-Jun (24 , 26 , 31) . Several natively unfolded proteins bind and regulate the assembly of cytoskeletal proteins such as tubulin and actin (e.g., MAP-2 and tau; Refs. 37 , 38 ); some bind calcium and/or calmodulin (e.g., chromogranin A and caldesmon; Refs. 39 , 40 ); and still others are phosphatase inhibitors (e.g., DARPP-32; Refs. 41 , 42 ). It seems that the flexible, extended structures of these proteins may be optimized for intermolecular interactions (43) . In some cases, the proteins undergo disorder-to-order transitions on binding to, for example, kinases, transcription factors, translation inhibitors, and nucleic acids (26) . These conformational switches may be triggered by covalent modifications, such as phosphorylation, or by binding to small molecules. The secondary structure that develops on complex formation often is not regularly repeating and relatively extended (30) . Intrinsic disorder may be useful in molecular recognition to achieve high specificity coupled to low affinity. It allows for one molecule to bind to multiple partners as a consequence of its structural plasticity and can create large interface regions from large target sites. Another suggestion is that native disorder allows for large protein interfaces while minimizing protein, genome, and cell size; for large interface regions to be stable on monomeric, globular proteins would require much larger proteins than if the interface were presented by a disordered protein (44) .
Analysis of the primary structure of TC-1 by PredictProtein suggested helical stretches from 46–52, 60–68, 75–84, and 92–97 in the chain. Given the evidence from CD that TC-1 was natively unfolded, we also evaluated the sequence using PONDR, and intriguingly, this predicts that most of the COOH-terminal half of TC-1 will be disordered (24) . Regions that “structure-based” algorithms predict to be ordered and PONDR predicts to be disordered often are the regions that assume structure in the functional complex. The secondary structure of TC-1 also is influenced by the ionic strength, and we have observed changes in the solubility of TC-1 that depend on the nature of the buffer ions. This suggests that subtle changes in cell pH, or in any of its components, can have an effect on the conformation and structural integrity of TC-1, and this could be a key mechanism in the control of intermolecular complexes that involve TC-1.
The characterization of TC-1 as a natively unfolded protein suggests possible functions of TC-1 by analogy with proteins with similar structural properties. To study the role of TC-1 in thyroid cancer, we stably transfected Nthy cells (SV40 transformed normal thyrocytes) with the cDNA of the TC-1 gene. Our results show that when transfected with TC-1, SV-immortalized thyroid cells, cultured in 0.1% FCS, showed increased cell number by 48 h, which continued until the experiment ended at 96 h. These data suggest that the TC-1 gene product may play a major role in cell survival.
Increased cell survival by TC-1-overexpressing Nthy cells correlated with decreased apoptosis as measured by TUNEL and COMET assay experiments. During the development and progression of cancers, cancer cells often acquire the capability of evading apoptosis; thus, this effect of TC-1 is in keeping with a tumorigenic phenotype by conferring an increased growth:apoptosis ratio (45 , 46) . Interestingly, TC-1 overexpression was found to increase activity of caspase-3, an early apoptosis marker. Caspase-3 is believed to be one of the main effectors in the apoptosis pathways (47) , and an increase in caspase-3 usually activates apoptosis (48) . Thus, this finding would appear to conflict with the TUNEL results. Therefore, we used the highly sensitive COMET assay to visualize the extent of DNA fragmentation in single cells. This assay confirmed the TUNEL results (i.e., TC-1-expressing clones showed less apoptosis induced by 48-h serum starvation, relative to control cells). Increased, rather than decreased, caspase-3 activation has been shown in breast cancers (49) and stomach cancers (50) , relative to their respective nonmalignant tissues. Apoptosis can be delayed or blocked via expression of apoptosis-inhibitory proteins. Such inhibitors of apoptosis can inhibit caspase-mediated pathways. Therefore, our findings suggest that increased caspase-3 activity induced by TC-1 overexpression does not lead to cell death, possibly because TC-1 inhibits the caspase-mediated pathway directly or via TC-1 stimulation of an inhibitor of apoptosis.
In keeping with enhanced cell survival, TC-1 overexpression was shown to increase PKC and PKA activity. These represent two of the three major cell-signaling pathways present in normal thyroid cells (33 , 35 , 51) . Activation of PKC and PKA has been implicated in the progression of thyroid cancer (51) . Thus, overexpression of TC-1 (as observed in thyroid cancer) leading to the activation of PKC and PKA could underlie the transformation of Nthy cells to a more tumorigenic phenotype. Our protein characterization studies also showed that PKA and PKC phosphorylate TC-1. This suggests that phosphorylation of TC-1 could be an important mode of regulation of its function and/or that TC-1 and PKC and/or PKA are involved together in a signal transduction pathway whereby they contribute to each other’s activation. More work is required to define the exact role of TC-1 and how it fits in the signal transduction pathway.
The TC-1-transfected cells also demonstrated significantly greater anchorage-independent growth as measured by colony formation in soft agar. Anchorage-independent growth is the transformed phenotype that most closely correlates with malignancy, suggesting that TC-1 appears to play a key role in the transformation of normal thyroid cells to a more malignant phenotype (41) .
In summary, these findings have demonstrated that TC-1 is a natively unfolded protein. The results presented have provided evidence that in solution TC-1 exists in a monomeric conformation that contains little secondary structure and no hydrophobic core. Many such proteins have been shown to have important biological roles, often involved in cell signaling through protein-protein interactions, and the disordered state appears to be critical for function in these cases. In keeping with this, our findings suggest that TC-1 may contribute to an aggressive tumorigenic phenotype. Because significant gaps still exist in our knowledge of the molecular events associated with thyroid carcinogenesis, characterization of TC-1 and its potential role in cellular signaling could be an important step toward understanding the molecular mechanisms underlying this neoplasm.
Grant support: Cure Cancer Foundation, Ramaciotti Foundation, and Sesqui Research and Development Scheme, University of Sydney, Australia.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: Alison Death, Heart Research Institute, 145 Missenden Road, Camperdown, NSW, 2050, Australia. Phone: 612-9550-3560; Fax: 612-9550-3302; E-mail:
- Received July 23, 2003.
- Revision received January 28, 2004.
- Accepted February 9, 2004.
- ©2004 American Association for Cancer Research.