Orphan receptor small heterodimer partner suppresses tumorigenesis by modulating cyclin D1 expression and cellular proliferation

Abstract

The small heterodimer partner (SHP; NROB2), a member of the nuclear receptor superfamily, contributes to the biological regulation of several major functions of the liver. However, the role of SHP in cellular proliferation and tumorigenesis has not been investigated before. Here we report that SHP negatively regulates tumorigenesis both in vivo and in vitro. SHP-/- mice aged 12 to 15 months old developed spontaneous hepatocellular carcinoma, which was found to be strongly associated with enhanced hepatocyte proliferation and increased cyclin D1 expression. In contrast, overexpressing SHP in hepatocytes of SHP-transgenic mice reversed this effect. Embryonic fibroblasts lacking SHP showed enhanced proliferation and produced increased cyclin D1 messenger RNA and protein, and SHP was shown to be a direct negative regulator of cyclin D1 gene transcription. The immortal SHP-/- fibroblasts displayed characteristics of malignant transformed cells and formed tumors in nude mice.

Conclusion: These results provide first evidence that SHP plays tumor suppressor function by negatively regulating cellular growth.

Fig. 1

Spontaneous formation of HCC in mice lacking SHP. (A) SHP^−/− mice developed HCC at 15 months of age (panels a-c). Hema-toxylin-eosin staining of wild-type (WT) liver (panel d) and tumors b and c (panels b’ and c’). (B) Increased cyclin D1 expression in SHP^−/− liver compared with wild-type mice by northern blotting. (C) Massive hepatocyte proliferation in livers of SHP^−/− (SKO) mice, which was inhibited in SHP-transgenic (STG) mice. NC, nontransgene controls. Ten- to 12-month-old mice were used for the assay. Bar graph represents mean ± SD of PCNA-positive cells counted in three fields per slide (n = 3). (D) SHP inhibition of cyclin D1 and PCNA expression in mouse HCC cell line Hepa-1 cells as examined by semiquantitative polymerase chain reaction analysis.

Fig. 2

Enhanced proliferation of immortalized fibroblasts lacking SHP. (A) FACS analysis of SHP^+/+ and SHP^−/−MEFs synchronized at G₀/G₁ by hydroxyurea. (B) FACS analysis of SHP^+/+ and SHP^−/− MEFs synchronized at G₁/S by double thymidine block. (C) FACS analysis of SHP^+/+ and SHP^−/− MEFs arrested at G₂/M phase by nocodazole. (D) FACS analysis of cells synchronized at G₀ by serum starvation and released into the cell cycle. Left: SHP^+/+ and SHP^−/− cells in S phase; middle: re-expressing SHP in SHP^−/− cells by adenovirus; right: knocking down SHP in SHP^+/+ cells by RNA interference. GFP, control virus.

Fig. 3

Changes of cyclins and CDKs lacking SHP and inhibition of cyclin D1 mRNA by SHP. (A) Changes of cyclin D1/pRb-kinase activity and cyclin E/H1-kinase activity in SHP^+/+ and SHP^−/− fibroblasts. Cyclin D-associated or cyclin E-associated protein complexes were immunoprecipitated from cell lysates, and GST-Rb or histone H1 kinase activity was measured as described in Materials and Methods. (B) Expression of cell cycle regulatory proteins via western blotting in synchronized and restimulated SHP^+/+ and SHP^−/− fibroblasts. (C) Top: expression of cyclin D1 mRNA in synchronized and restimulated fibroblasts as determined by northern blot. Bottom: expression of cyclin D1 mRNA was increased in exponential growing SHP^−/− cells (left), decreased by overexpressing SHP (middle), and increased by knocking down SHP function (right), as determined via northern blotting. G, green fluorescent protein; S, SHP. (D) Cyclin D1 promoter activity. Both the SHP^+/+ and SHP^−/− cells were transfected with a mouse cyclin D1 promoter reporter with or without SHP expression vector, and promoter activity was measured via luciferase/β-gal assay. AdeS, adenovirus SHP.

Fig. 4

Transcriptional repression of cyclin D1 promoter by SHP. (A) Left: a − 187mCD1-Luc was transfected into CV-1 cells with LRH-1 (50, 100, 200 ng) in the absence (−) or presence (+) of SHP (100, 200, 300 ng). Right: reporter gene of −970mCD1-Luc was cotransfected with expression vectors for β-catenin (CA, 400 ng), LRH-1 (100 ng), and SHP (100, 200, 300 ng; fixed amount: 200 ng) as indicated. Luciferase activities were normalized against β-galactosidase and expressed as relative activity. (B) Mutagenesis study. The mCD1 gene contains two potential LRH-1 sites (L1, L2). Mismatches with the consensus sequence are indicated by lower case letters, and mutated sequences are underlined. The mutated constructs (mut1, mut2, and mut3) were transfected into CV-1 cells with or without LRH-1 (100 ng). (C) Gel shift assay. In vitro translated LRH-1 was incubated with the ³²P-labeled L1 and increasing amounts of unlabeled L1 (lanes 2–4), L2 (lanes 5–7), and mutated L1 (lanes 8–10) or HNF-4 binding sequences (ns: lanes 11–12). (D) Chromatin immunoprecipitation analysis. Chromatin preparations from MEFs were subjected to chromatin immunoprecipitation assay with anti-SHP antibodies. DNA in the immunoprecipitates was polymerase chain reaction-amplified with the primers shown in the illustration of the promoter region. DNA from equal amounts of chromatin was polymerase chain reaction amplified before immunoprecipitation (Input). Neg, immunoglobulin G.

Fig. 5

Transformation and tumorigenesis of immortalized fibroblasts lacking SHP. (A) Colony formation assay. Results from three independent SHP^+/+ and SHP^−/− cells were shown. (B) Tumorigenicity assay in nude mice. Four wild-type and five SHP mutant 3T3 clones were tested. Nude mice were still tumor-free 6 weeks after injection of wild-type MEF cells, whereas tumors (arrows) were observed 4 weeks after injection of SHP^−/− cells. Photographs were taken 6 weeks after injection. (C) Chromosome spreads of SHP^+/+ and SHP^−/− 3T3 cells to examine chromosome aberrations. Each metaphase spread was assessed for the frequency of chromosome abnormalities. Note that only parts of spreads are shown in order to highlight the short chromosomes (arrows) found in SHP^−/− cells. (D) Foci formation assay. SHP was overexpressed into SHP^−/− fibroblasts by adenovirus transduction. Statistical results represent the mean ± standard deviation of foci counts from two different fields of triplicate plates.