Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2011 Aug;25(8):1311-25.
doi: 10.1210/me.2010-0420. Epub 2011 May 26.

Mutant Thyroid Hormone Receptors (TRs) Isolated From Distinct Cancer Types Display Distinct Target Gene Specificities: A Unique Regulatory Repertoire Associated With Two Renal Clear Cell Carcinomas

Affiliations
Free PMC article

Mutant Thyroid Hormone Receptors (TRs) Isolated From Distinct Cancer Types Display Distinct Target Gene Specificities: A Unique Regulatory Repertoire Associated With Two Renal Clear Cell Carcinomas

Meghan D Rosen et al. Mol Endocrinol. .
Free PMC article

Abstract

Thyroid hormone receptors (TRs) are hormone-regulated transcription factors that regulate a diverse array of biological activities, including metabolism, homeostasis, and development. TRs also serve as tumor suppressors, and aberrant TR function (via mutation, deletion, or altered expression) is associated with a spectrum of both neoplastic and endocrine diseases. A particularly high frequency of TR mutations has been reported in renal clear cell carcinoma (RCCC) and in hepatocellular carcinoma (HCC). We have shown that HCC-TR mutants regulate only a fraction of the genes targeted by wild-type TRs but have gained the ability to regulate other, unique, targets. We have suggested that this altered gene recognition may contribute to the neoplastic phenotype. Here, to determine the generality of this phenomenon, we examined a distinct set of TR mutants associated with RCCC. We report that two different TR mutants, isolated from independent RCCC tumors, possess greatly expanded target gene specificities that extensively overlap one another, but only minimally overlap that of the wild-type TRs, or those of two HCC-TR mutants. Many of the genes targeted by either or both RCCC-TR mutants have been previously implicated in RCCC and include a series of metallothioneins, solute carriers, and genes involved in glycolysis and energy metabolism. We propose as a hypothesis that TR mutations from RCCC and HCC may play tissue-specific roles in carcinogenesis, and that the divergent target gene recognition patterns of TR mutants isolated from the two different types of tumors may arise from different selective pressures during development of RCCC vs. HCC.

Figures

Fig. 1.
Fig. 1.
In the absence of hormone, RCCC-TR mutants repress a broad set of genes distinct from those repressed by WT- or HCC-TRs. A, Schematic of the WT and mutant TRs employed. The DBD and ligand-binding domains are shown, as are the locations of the relevant mutations (arrowheads). B, Venn diagrams of genes repressed by the different TR alleles; minus T3. RNA was isolated from stable transformants expressing the TR indicated and treated with carrier only. Gene expression levels were determined by microarray analysis as in Materials and Methods. Data were analyzed using a BH adjusted P value of <0.05 to identify genes repressed in each TR transformant compared with the vector-only control. C, Heat map clustering of representative repressed genes. Expression level of each gene is indicated by color (see key below panel). D, Venn diagrams of genes repressed by the different TRs, adjusted to different statistical cutoffs. Genes repressed in the absence of T3 by WT-TRα1 (green), WT-TRβ1 (pink), rc6-TRα1 (blue), rc15-TRβ1 (purple), hcI-TRα1 (yellow), or hcN-TRβ1 (orange) were identified as in panel B, but using BH-adjusted P values of <0.02 (darkest shades), 0.05 (intermediate shades), and 0.1 (lightest shades).
Fig. 2.
Fig. 2.
RCCC-TR mutants repress a shared set of target genes. A, Venn diagram of genes repressed by rc6-TRα1 compared with those repressed by rc15-TRβ1; minus T3. B, Microarray intensity signals of two representative target genes repressed in the absence of T3 by both rc6-TRα1 and rc15-TRβ1. Asterisks indicate significant repression (BH adjusted P value of <0.05) in the TR transformant vs. the empty vector control. The mean and sem values from three independent experiments are shown. NR, No receptor/empty vector.
Fig. 3.
Fig. 3.
Additional statistical analysis reveals an extensive overlap between target genes repressed by rc6-TRα1 and rc15-TRβ1. A, Microarray intensity signals of two target genes identified as repressed by rc6-TRα1, subsequently assayed for regulation by other TR alleles. Double asterisks indicate significant repression (BH-adjusted P value of <0.05) in the TR transformant vs. the empty vector control. Single asterisks indicate a potentially significant repression using a nonadjusted P value of <0.05 (see Materials and Methods). B, Microarray intensity signals of two target genes identified as repressed by rc15-TRβ1, subsequently assayed for regulation by other TR alleles as in panel A. C, Venn diagram comparing gene repressed in the absence of T3 by rc6-TRα1 vs. rc15-TRβ1. A two-step analysis was employed beginning with the genes identified in Fig. 2A and applying a second, nonadjusted P value of <0.05 (see Materials and Methods). D, Venn diagrams comparing genes repressed by WT and mutant TRα1 or WT and mutant TRβ1 in the absence of hormone. A two-step analysis was employed beginning with the genes identified in Fig. 1B and using a nonadjusted P value of <0.05 (see Materials and Methods). NA, Not applicable; NR, no receptor/empty vector.
Fig. 4.
Fig. 4.
Repression by RCCC-TR mutants is largely independent of T3. A, Microarray intensity signals of three target genes flagged as repressed by RCCC-TRs in the absence of T3 (Fig. 2A), subsequently compared with and without T3. Asterisks indicate significant repression (BH adjusted P value of <0.05) in the TR transformant vs. the empty vector control. B, Microarray intensity signals of additional RCCC-TR target genes flagged as repressed by RCCC-TRs in the presence of hormone (BH-adjusted P value of <0.05), subsequently compared with or without T3. The mean and sem values from three independent experiments are shown. NR, No receptor/empty vector.
Fig. 5.
Fig. 5.
RCCC-TR mutants activate an extended set of genes that are not targeted by the WT or HCC-TR mutants. A and B, Venn diagrams of genes activated by WT-TRα1 (green), WT-TRβ1 (pink), rc6-TRα1 (blue), RC15-TRβ1 (purple), hcI-TRα1 (yellow), or hcN-TRβ1 (orange) in the presence of T3 are shown, using BH-adjusted P values of <0.02 (darkest shades), 0.05 (intermediate shades), and 0.1 (lightest shades). C, Venn diagram of genes activated in the presence of T3 by rc6-TRα1 vs. rc15-TRβ1 (BH-adjusted P value of <0.05). D, Venn diagram comparing genes activated in the presence of T3 by rc6-TRα1 vs. rc15-TRβ1, using a two-step analysis as described in Fig. 3C. E and F, Venn diagrams comparing genes activated by the different TR alleles in the presence of T3, using a two-step analysis as described in Fig. 3D.
Fig. 6.
Fig. 6.
Activation by the RCCC-TR mutants is largely independent of T3. A, Microarray intensity signals of six target genes flagged as activated by RCCC-TRs in the presence of T3 (Fig. 5), subsequently compared with and without T3. Asterisks indicate a significant activation (BH-adjusted P value of <0.05) in the TR transformant vs. the empty vector control. B, Microarray intensity signals of additional target genes flagged as activated by RCCC-TRs in the absence of hormone (BH-adjusted P value of <0.05), subsequently compared with and without T3. The mean and sem values from three independent experiments are shown. NR, No receptor/empty vector.
Fig. 7.
Fig. 7.
An RTH-TRβ1 mutant displays impaired transcriptional properties and a unique gene expression signature. A, Reporter gene assays. Cells stably expressing WT-TRβ1, P453S-TRβ1, or an empty vector were transiently transfected with a DR4-TK-luciferase reporter and pCH110-lacZ as an internal control. The cells were treated with T3 as indicated 24 h after transfection, harvested 48 h after transfection, and analyzed for luciferase and β-galactosidase activity. Luciferase activity is shown relative to β-galactosidase activity. B, qRT-PCR assays on four genes known to be WT-TR targets. Cells transformed with WT-TRβ1, RTH-TRβ1, or an empty vector control, were treated with with and without 100 nm T3 for 6 h. RNA was isolated and analyzed by qRT-PCR using gene-specific primers. C, qRT-PCR assays on additional genes. Assay was as in panel B. Expression of each gene in the empty vector control in the absence of T3 is defined as = 1; the mean and sem values from three independent experiments are shown. NR, No receptor/empty vector.
Fig. 8.
Fig. 8.
Comprehensive illustration of all genes repressed by one or more TRα allele in the absence of T3. All genes repressed by rc6-TRα1, hcI-TRα1, and WT-TRα1 in the absence of hormone (as depicted in Fig. 1B) were evaluated for function and loosely grouped into categories based on their Entrez description. Genes within each category are listed alphabetically along the perimeter of the circle and marked by gray rectangles. Colored lines link a receptor (represented by the same colored rectangle) to the genes that it regulates. Genes repressed by WT-TRα1 are contacted by a green line, those repressed by rc6-TRα1 are contacted by a blue line, and those repressed by hcI-TRα1 are contacted by an orange line. Thin lines represent genes that met a BH-adjusted P value of at least <0.05; thick lines represent genes that met a BH-adjusted P value of <0.005. Genes contacted by multiple lines are repressed by multiple receptors. The diagram, exclusive of gene categorizations, was created using the open-source software package, Circos.

Similar articles

See all similar articles

Cited by 6 articles

See all "Cited by" articles

Publication types

MeSH terms

Feedback