Concomitant RAS,RET/PTC,or BRAF Mutations in Advanced Stage of Papillary Thyroid Carcinoma

Thyroid tumor specimens and cell lines

All tumor tissues were obtained at surgery with informed consent, and were immediately frozen in liquid nitrogen and stored at −70°C until processed. The clinical staging of thyroid cancer was based on the Tumor, Node, Metastasis (TNM) classification (11). Eighty-eight PTC diagnosed between 1987 and 2006 were selected randomly and included in the study: 82 CPTC, three tall-cell variants, and three FVPTC. The mean follow-up interval was 12 years. The following cell lines were studied to determine whether they have dual mutations in RET/PTC, RAS, or BRAF: K1, BCPAP, 8305C, 8505C, HTC-C3, BHT101, CAL62, TPC-1, and FRO. TPC-1 and FRO were kindly provided by Dr. James Fagin (Department of Medicine, Memorial Sloan-Kettering Cancer Center, New York, NY). Other cell lines were purchased from DSMZ (Braunschweig, Germany). ATC cell lines such as 8305C, 8505C, BHT101, FRO, and CAL62 were screened for concomitant mutations, since all of them carry either BRAF or RAS mutations, and it is not clear whether they have dual mutations in the MAPK pathway. The study was reviewed and approved by the Institutional Review Board.

Detection of BRAF and RAS mutations

Tumor tissue was obtained by standard sectioning, and DNA was extracted by standard proteinase-K treatment followed by phenol/chloroform extraction. BRAF exon 15 was amplified by PCR and directly sequenced as described previously (12). Mutations of KRAS, NRAS, and HRAS were analyzed by polymerase chain reaction (PCR) sequence analysis of exon 2 (for codon 12/13), exon 3 (for codon 61), and exon 4 in KRAS and NRAS; and exon 1 (codon 12/13), exon 2 (codon 61) and exon 3 in HRAS. Exon 4 was screened for non-hotspot mutations. PCR primers are listed in Supplementary Table S1 (Supplementary Data are available online at www.liebertpub.com/thy). The PCR conditions were 95°C for 5 min followed by 35 cycles of amplification (95°C for 40 s, 50–62°C for 40 s, and 72°C for 40 s) with final extension at 72°C for 5 min. The resulting PCR products were analyzed by gel electrophoresis. Each successfully amplified fragment was directly sequenced using the BigDye Terminator V3.1 Cycle Sequencing kit and ABI prism 3100 sequencer (Applied Biosystems, Foster City, CA). DNA from paraffin blocks of patients 27 and 78 was extracted as described previously (13).

Detection of RET/PTC rearrangements

Total RNA was extracted from tumor specimens and cell lines by the guanidine thiocyanate-phenol-chloroform method. The integrity of RNA was verified by denaturing gel electrophoresis. Two μg of total RNA were reverse-transcribed into cDNA using the Promega reverse transcription (RT) system (Promega, Madison, WI). Nested RT-PCR was used to amplify transcripts of RET/PTC rearrangements, using two sets of primers listed in Supplementary Table S1 and PCR conditions described above. The resulting PCR products were analyzed by gel electrophoresis and directly sequenced. GAPDH cDNA was used as an internal control for RNA quality. To rule out PCR contamination, positive samples were confirmed by repeating PCR using different batches of same specimen.

Screening for PAX8/PPARγ rearrangement in FVPTC

A fusion gene (PAX8/PPARγ) between the DNA binding domain of the thyroid-restricted transcription factor paired box gene 8 (PAX8) and the peroxisome proliferator-activated receptor γ gene (PPARγ) has been detected in about 10% of FVPTC. PAX8/PPARγ rearrangement was screened by RT-PCR in FVPTC as described previously (14).

Cloning and expression of mutant KRAS ^S65N and NRAS ^S65C

The full-length KRAS and NRAS cDNA clones were obtained from GenScript (Piscataway, NJ) and cloned into pcDNA3.1 under the control of the CMV promoter (Invitrogen, Carlsbad, CA). KRAS ^S65N and NRAS ^S65C mutants were obtained by site-directed mutagenesis and were verified by DNA sequencing. Wild-type KRAS and NRAS were used as controls. Equal amounts of the constructs were transfected into HEK293, BCPAP, and CAL62 cell lines using Lipofectamine (Invitrogen). Cells were collected 48 hours after transfection for MAPK and cAMP response element binding protein (CREB) activity.

Western blot analysis

Forty μg of protein from transfected cells was loaded onto a 12% SDS-polyacrylamide gel. Proteins were transferred to a polyvinylidene (PVDF) membrane and subject to Western blot analysis using anti-phospho-MEK1/2, anti-phospho-ERK 1/2 antibody, or anti-phospho-CREB (1:1000; Cell Signaling Technology, Inc., Danvers, MA).

Cell proliferation assay

Cell proliferation was measured by a non-radioactive MTS assay kit according to the manufacturer's procedure (Promega). Briefly, cells were plated in triplicate into 96-well plates (5×10³ cells/well) containing 1% or 10% serum respectively for 48 h. At the final 4 h of incubation, 20 μL of CellTiter 96^® AQ_ueous One Solution reagent was added into each well for measurement of cell viability.

Statistical analysis

Fisher's exact test (two-tailed) was used. A p-value of ≤0.05 was considered significant.

BRAF mutation

BRAF ^V600E mutation was detected in 39 of 82 classic PTC (48%) and in all three tall-cell variants (Fig. 1 and Table 1). No mutation was detected in FVPTC. The BRAF ^V600E mutation occurred more frequently in the advanced stages (stage III or IV) compared to the early stages (stage I or II) of CPTC: 17/26 (65%) versus 22/56 (39%) (Table 1; p<0.05, Fisher's exact test). The BRAF ^V600E mutation was also detected in 8305C, 8505C, HTC-C3, BHT101, BCPAP, K1, and FRO cells.

OPEN IN VIEWER

FIG. 1.

Detection of BRAF, RAS, and RET/PTC-1 mutations in papillary thyroid carcinoma (PTC). (A) RET/PTC-1 rearrangement was detected by reverse transcription polymerase chain reaction (RT-PCR) from patients. The PCR product was run on a 2% agarose gel. GAPDH cDNA was used as an internal control for RNA quality. (B) DNA sequence analysis of PCR products from three patients with concomitant mutations. The RAS and BRAF mutations are indicated by an arrow. The breakpoint of RET/PTC-1 rearrangement is indicated by a vertical line.

RAS mutation

KRAS mutation was detected in two CPTC: c.182A>G (p.Q61R) with stage III cancer, and c.194 G>A (p.S65N) with stage I cancer (Fig. 1). NRAS ^Q61R was detected in one CPTC case and two FVPTC cases. No mutation was found in HRAS. The overall rate of RAS mutation was 6% (5/88) in the cohort studied (Table 2). KRAS ^G12R was found in the CAL62 cell line. KRAS ^S65N has not been reported before in thyroid cancer. The significance of the mutation was further investigated first in matched normal tissue from paraffin blocks. No mutation was found in normal tissue, suggesting that it is not a germ-line polymorphism (data not shown). Multiple protein sequence alignments around codon 65 showed that serine⁶⁵ is highly conserved across different species (data not shown). Next, we used PolyPhen-2 to predict the possible effect of S65N substitution (http://genetics.bwh.harvard.edu/pph/). PolyPhen-2 predicted a possible damaging effect with a score of 0.919 of 1.0 (sensitivity: 0.81; specificity: 0.94). To confirm that KRAS ^S65N or NRAS ^S65C (reported previously in neuroendocrine lung cancer and melanoma but not characterized by functional studies) can activate the MAPK signaling pathway, we cloned and expressed KRAS ^S65N and NRAS ^S65C in HEK293 cells. As shown in Figure 2A, both KRAS ^S65N and NRAS ^S65C caused increased p-MEK and p-ERK expression in HEK293 cells. KRAS ^S65N can also increase MAPK activity and cell proliferation in both low- and high-serum conditions (Fig. 2B and C). We further investigated the effect of KRAS ^S65N on MAPK and CREB activity in the PTC cell line BCPAP and the ATC cell line CAL62. CREB is activated by phosphorylation at Ser133 by various signaling pathways, including the MAPK/ERK pathway. As shown in Figure 2B, the expression of both phospho-ERK and phospho-CREB was increased following KRAS ^S65N expression. The basal level of phospho-ERK was also higher in BCPAP and CAL62 as compared to HEK293, probably due to the fact that BCPAP and CAL62 contain BRAF ^V600E and KRAS ^G12R respectively.

OPEN IN VIEWER

FIG. 2.

Characterization of KRAS^S65N and NRAS^S65C on mitogen-associated protein kinase (MAPK) activity. (A) The mutant and wild-type constructs (KRAS ^S65N, NRAS ^S65C, and KRAS ^WT) were transfected into HEK293 cells. Cells were collected 48 h after transfection for Western blot analysis. Increased pMEK1/2 and pERK 1/2 expression was detected by an anti-pMEK1/2 and anti-pERK 1/2 antibody respectively in cells transfected with mutant RAS constructs. The membrane was reprobed with total ERK1/2 and GAPDH (37Kd) antibody to monitor protein loading. Quantification of the Western blot results is shown in Supplementary Figure S1. (B) The KRAS ^WT, KRAS ^S65N, and NRAS ^S65C were transfected into HEK293 cells. Cells were cultured in serum-free medium for an additional 2 days 24 h after transfection and collected for Western blot analysis. Increased pERK 1/2 and pMEK 1/2 expression was detected in both KRAS ^S65N and NRAS ^S65C transfected cells following 48 h serum starvation. (C) Cell proliferation of HEK293 cells transfected with KRAS ^S65N or KRAS ^WT. The cells were plated into 96-well plates (5×10³ cells/well) containing 1% or 10% serum respectively for 48 h. Data are expressed as mean±standard error of the mean (SEM) of triplicates and are representative of three different experiments (*p<0.01). (D) Increased MAPK and cAMP response element binding protein (CREB) activity in thyroid cancer cell lines following KRAS ^S65N transfection. Both phospho-CREB and-ERK1/2 expressions were increased in CAL62 and BCPAP cell lines. HRAS^G12V was transfected into CAL62 and BCPAP cell lines as a positive control. The basal level of pEKR1/2 was also higher in CAL62 and BCPAP as compared to HEK293.

RET/PTC rearrangements

A RET/PTC-1 rearrangement was detected in nine CPTC (9/82, 11%), one tall-cell variant, two FVPTC, and the TPC-1 cell line (Fig. 1 and Table 2). The overall rate of RET/PTC-1 was 14% (12/88). RET/PTC-2 and -3 were not detected in this series of samples. RET/PTC-1 occurred more frequently in the late stage (7/29, 24%) than in the early stage of PTC (5/59, 9%; p=0.0548).

PAX8/PPARγ rearrangement

No rearrangement was detected in the three FVPTC cases (data not shown).

Concomitant BRAF ^V600E , RAS, or RET/PTC-1 mutations

Two CPTC samples had both BRAF ^V600E and KRAS ^Q61R or KRAS ^S65N. Six CPTC and one tall-cell variant contained both BRAF ^V600E and RET/PTC-1. Two FVPTC samples had both NRAS ^Q61R and RET/PTC-1 (Table 1). None of the samples were found to have triple mutations. In total, concomitant mutations were found in 11/88 PTC samples (13%). Among them, three were found in early stages of PTC (3/59, 5%), and eight (8/29, 28%) in an advanced stage. The association of concomitant mutations with advanced stage was statistically significant (p<0.01). Eight cases with concomitant mutations were older than 45 years. Concomitant mutations were not found in nine cell lines studied. Co-existence of RET/PTC with RAS or BRAF has been reported in multiclonal PTC (9). We next examined different tumor sections of paraffin blocks for concomitant BRAF and RAS mutations found in our two cases. As shown in Table 3, the BRAF mutation was present in all tumor sections, whereas the RAS mutation was detected only in some sections. This may indicate intratumor heterogeneity. Histologic heterogeneity was not observed. Due to the limited sensitivity of Sanger sequencing to detect a low frequency of mutant alleles, we cloned the PCR fragments from tissue sections with negative RAS mutation into a TA vector, and screened 50 clones from each fragment for RAS mutation by direct Sanger sequencing. As shown in Table 3, a RAS mutation was detected in two sections that were negative by direct Sanger sequencing. We further evaluated the sensitivity of Sanger sequencing to detect KRAS mutations by mixing different amounts of wild-type and mutant KRAS DNA. As shown in Figure 3, about 20% mutant alleles are required in the sample to be reliably detected by Sanger sequencing (the detection limit is about 20%). In the lymph node metastases, the RAS mutation was significantly enriched and can be easily detected (Table 3).

OPEN IN VIEWER

FIG. 3.

Sensitivity of Sanger sequencing to detect KRAS mutations. Different amounts of wild-type and mutant KRAS DNA were mixed together in a total concentration of 100 ng DNA for PCR. PCR products were directly sequenced. At least 20% mutant alleles need to be present in the sample to be reliably detected by Sanger sequencing.

To determine whether concomitant mutations were associated with a poor clinical outcome, we analyzed disease-free survival of the patients who had concomitant mutations versus those without mutation or those who had only one mutation. As shown in Figure 4, the Kaplan–Meier curve shows a significant difference in the recurrence-free probability over a 10-year period among patients with no mutation, a single mutation, and double mutations (log-rank test χ²=18.2, p<0.0001), and between patients with single and double mutations (log-rank test χ²=4.2, p<0.041). These data indicate that patients with concomitant mutations had much poorer clinical outcomes than those with a single or no mutation.

OPEN IN VIEWER

FIG. 4.

Kaplan–Meier analysis for recurrence-free probability. The Kaplan–Meier curves shows a significant difference in the recurrence-free probability over a 10-year period after the diagnosis among patients with no mutation, a single mutation, and double mutations (log-rank test χ²=18.2, p<0.0001), and also between patients with single mutation and double mutations (log-rank test χ²=4.2, p<0.041).