KLHL14 is a tumor suppressor downregulated in undifferentiated thyroid cancer

Abstract

KLHL14 is a substrate-binding subunit of Cullin-RING ligase 3 ubiquitin ligase complex, highly enriched in thyroid since early embryonic development, together with its antisense RNA KLHL14-AS. We have previously demonstrated that Klhl14-AS is a competing endogenous RNA regulating several differentiation and survival factors in thyroid cancer, acting as tumor suppressor. Recently, also KLHL14 has been shown to function as tumor suppressor in diffuse large B-cell lymphoma and in malignant mesothelioma. Here we show that KLHL14 expression is strongly reduced in anaplastic thyroid cancer, the less differentiated and most aggressive type of thyroid neoplasia. Such reduction is reproduced in different in vivo and in vitro models of thyroid cancer, being invariably associated with loss of differentiation. When Klhl14 expression is rescued in thyroid transformed cells, it reduces the cell proliferation rate and increase the number of apoptotic cells. On the other side, Klhl14 loss of function in normal thyroid cells affects the expression of several regulatory as well as functional thyroid markers. All these findings suggest that KLHL14 could be considered as a novel tumor suppressor in thyroid cancer, by also revealing its physiological role in the maintenance of a fully differentiated and functional thyroid phenotype.

Fig. 1. KLHL14 expression in human and experimental thyroid cancers.

Volcano Plots depicting differential gene expression in anaplastic thyroid carcinomas (ATC) from gene expression dataset GSE65144 (A), GSE33630 (B) and in papillary thyroid carcinomas (PTC) from GSE33630 (C), in which KLHL14 scattering is highlighted. The x-axis enlist the log2 fold change value representing the mean expression value of each gene, while the y-axis is the level of significance (adjusted p-value) reported in –log10; blue dots represent significantly down-regulated genes in the first group, red dots represent significantly up-regulated genes. D In situ hybridization micrograph showing Klhl14 RNA-probe signals in normal thyroid of Tg-rtTA; TetO/BRAF^V600E untreated mouse (NT) and in neoplastic thyroid of doxycycline treated mouse (+ Dox), at two different magnifications. Klhl14 mRNA levels analyzed through qRT-PCR in (E) thyroids from either untreated (NT) or treated (+Dox) Tg-rtTA; TetO/BRAF^V600E mice, (F) parental FRTL5 and transformed FRTL5-RAS cells, (G) FRTL5-ER^TM-RAS cells either treated with vehicle (DMSO) or 4-hydroxytamoxifen (4-OHT). H Luciferase activity induced by Klhl14 promoter (0.5 kb Klhl14) and empty reporter vector (pGL3Basic) in FRTL5-ER^TM-RAS cells either treated with vehicle (DMSO) or 4-hydroxytamoxifen (4-OHT). Relative luminescences were normalized on Renilla luciferase activity used as internal control. Statistical t-test significance is reported as: **, p-value < 0.01; *, p-value < 0.05; ns not statistically significant.

Fig. 2. Klhl14 expression affects proliferation and viability of transformed thyroid cells.

A Clonogenic assay of FRTL5-HRasG12V cells transfected with Klhl14 (Klhl14) or control plasmid (Ctrl), and their respective absolute and average colony numbers from two independent experiments. B Klhl14 expression measured by qRT-PCR in FRTL5, FRTL5-RAS, FRTL5-RAS control clone (EV) and in nine different FRTL5-RAS Klhl14-expressing clones. Clones were ordered based on ectopic Klhl14 expression levels, normalized on that of FRTL5-RAS cells. C Klhl14 expression measured by immunoblot in FRTL5, FRTL5-RAS, FRTL5-RAS control clone (EV) and in nine different FRTL5-RAS Klhl14-expressing clones, as well as in FRTL5-RAS control (Ctrl) and Klhl14 pools. Klhl14 was detected by both anti-FLAG antibody (upper panels) and anti-Klhl14 antibody (lower panels). Gapdh was used as loading control. Uncropped blots are in Supplementary Fig. 1. Cell viability of FRTL5-RAS Klhl14 and control pools assessed by trypan blue exclusion assay (D) and MTS assay (E). F growth curve of FRTL5-RAS Klhl14 and control pools. Cell numbers are reported as “cells x 10⁵”. Cell viability of control (EV) and two Klhl14-expressing FRTL5-RAS clones (K14-2 and K14-17) assessed by by trypan blue exclusion assay (G) and MTS assay (H). I Growth curve of control (EV) and two Klhl14-expressing FRTL5-RAS clones (K14-2 and K14-17). Cell numbers are reported as “cells x 10⁵”.

Fig. 3. Klhl14 expression reduces S-phase entry of transformed thyroid cells.

A Cell cycle analysis by flow cytometry using propidium iodide staining in two FRTL5-RAS Klhl14-expressing clones and one control clone (EV). The histograms report the percentage of cells in G1, S, and G2-M phases of the cell cycle. Fluorescence micrographs of BrdU incorporation assay in (B) FRTL5-RAS Klhl14-expressing and control (EV) clone and in (D) FRTL5-RAS Klhl14 expressing and control (Ctrl) mass populations, taken at 20x magnification. Cells from four acquired fields were counted for BrdU positivity and reported respectively in (C, E). BrdU assays shown are representative of two independent replicates. Scale bar (50 µm) is added for reference.

Fig. 4. Klhl14 expression increases cell death in transformed thyroid cells.

A Cytometric evaluation of cell viability using propidium iodide in non-permeabilized FRTL5-RAS Klhl14-expressing and control (EV) clones. Percentages of dead cells are reported. B Flow cytometry scatter plots of Klhl14 and EV clones. Size (forward scatter) and granularity (side scatter) On the x-axis are plotted the Forward Scattering intensity values (FSC-H), while on the y-axis there are the Side Scattering intensity values (SSC-H) relative to 10,000 events registered. Agarose gel electrophoresis of fragmented DNA from (C) FRTL5-RAS Klhl14-expressing (K14-2 and K14-17) and control (EV) clones and (D) FRTL5-RAS Klhl14-expressing (Klhl14) and control (Ctrl) mass populations, cultured either in the presence (+) or absence (−) of serum and hormones. Fluorescence micrographs of TUNEL assay in (E) two FRTL5-RAS Klhl14 expressing and onecontrol (EV) clones and (G) FRTL5-RAS Klhl14 expressing and control (Ctrl) mass populations, taken at 20x magnification. Cells from four acquired fields were counted for TUNEL positivity and reported, respectively, in (F, H). Presented results are representative of two independent replicates. Scale bar (50 µm) is added for reference.

Fig. 5. Thyroid differentiation markers expression in Klhl14 expressing thyroid transformed cells.

A Immunoblots showing protein expression of flagged-Klhl14 and thyroid markers in two FRTL5-RAS (K14-2 and K14-17) and control (EV) clones. Gapdh was assayed for each series of markers analyzed on the same blot. B Immunoblot signal quantifications of thyroid marker proteins normalized on the respective Gapdh. C Thyroid markers mRNA levels assessed by qRT-PCR on the same clones as in (A). Data are represented as mean ± SD of two independent replicates. Statistical t-test significance is reported as: *, p-value < 0.05; ns not statistically significant. Uncropped blots are in Supplementary Figs. 2–4.

Fig. 6. Effects of Klhl14 knockdown on normal thyroid differentiation.

Klhl14 was knocked down in FRTL5 cells by a specific LNA (Klhl14 LNA) or scrambled LNA (Ctrl LNA) and analyzed at 48H, 72H and 96H post-transfection. A Immunoblots showing protein expression of endogenous Klhl14 and thyroid markers. Gapdh, for each immunoblot, is shown as loading control. B Immunoblot signal quantifications of thyroid marker proteins normalized on the respective Gapdh. C Thyroid markers and endogenous Klhl14 mRNA levels assessed by qRT-PCR. Data are represented as mean ± SD of three independent experiments. Statistical t-test significance is reported as: *, p-value < 0.05; ns not statistically significant. Uncropped blots are in Supplementary Figs. 5−7.

Fig. 7. Klhl14 localization in normal and transformed thyroid cells.

A Immunoblot showing the expression of Klhl14 protein in nuclear and cytoplasmic fractions, in FRTL5 cells, FRTL5-RAS cells, FRTL5-RAS clones (K14-2 and K14-17) and control EV. Klhl14 was detected by anti-Klhl14 antibody. Gapdh and Histone H3 are shown as loading controls of cytoplasm and nucleus, respectively. Signals from either endogenous or ectopic Klhl14 were captured at different exposure times to avoid signal saturation in overexpressing clones. Uncropped blots are in Supplementary Fig. 8. B Immunofluorescence micrographs of Klhl14 in FRTL5 cells and FRTL5-RAS clones (K14-2 and K14-17) and control EV, taken at 63x magnification (oil immersion objective). Scale bar (20 µm) is added for reference. Images from either endogenous or ectopic Klhl14 were captured at different exposure times to avoid signal saturation in overexpressing clones. C Manders’ Overlap Coefficients of FRTL5 cells and FRTL5-RAS clones (K14-2 and K14-17) and control EV, calculated as overlapping signals of Klhl14 channel in Hoechst 33324 channel. Pearson Colocalization Coefficients: FRTL5 r = 0.518, EV r=not applicable, K14-2 r = 0.349, K14-17 r = 0.245. The p-value for this analysis was 100%, indicating a high probability of colocalization.