cAMP-dependent posttranscriptional regulation of steroidogenic acute regulatory (STAR) protein by the zinc finger protein ZFP36L1/TIS11b

Abstract

Star is expressed in steroidogenic cells as 3.5- and 1.6-kb transcripts that differ only in their 3'-untranslated regions (3'-UTR). In mouse MA10 testis and Y-1 adrenal lines, Br-cAMP preferentially stimulates 3.5-kb mRNA. ACTH is similarly selective in primary bovine adrenocortical cells. The 3.5-kb form harbors AU-rich elements (AURE) in the extended 3'-UTR, which enhance turnover. After peak stimulation of 3.5-kb mRNA, degradation is seen. Star mRNA turnover is enhanced by the zinc finger protein ZFP36L1/TIS11b, which binds to UAUUUAUU repeats in the extended 3'-UTR. TIS11b is rapidly stimulated in each cell type in parallel with Star mRNA. Cotransfection of TIS11b selectively decreases cytomegalovirus-promoted Star mRNA and luciferase-Star 3'-UTR reporters harboring the extended 3'-UTR. Direct complex formation was demonstrated between TIS11b and the extended 3'-UTR of the 3.5-kb Star. AURE mutations revealed that TIS11b-mediated destabilization required the first two UAUUUAUU motifs. HuR, which also binds AURE, did not affect Star expression. Targeted small interfering RNA knockdown of TIS11b specifically enhanced stimulation of 3.5-kb Star mRNA in bovine adrenocortical cells, MA-10, and Y-1 cells but did not affect the reversals seen after peak stimulation. Direct transfection of Star mRNA demonstrated that Br-cAMP stimulated a selective turnover of 3.5-kb mRNA independent of AURE, which may correspond to these reversal processes. Steroidogenic acute regulatory (STAR) protein induction was halved by TIS11b knockdown, concomitant with decreased cholesterol metabolism. TIS11b suppression of 3.5-kb mRNA is therefore surprisingly coupled to enhanced Star translation leading to increased cholesterol metabolism.

Figure 1

Expression of Star mRNA with short and long 3′-UTR after stimulation of mouse MA-10 cells and primary bovine adrenocortical cells. A, Mouse and bovine AURE regions with TIS11b recognition sequences highlighted. B, MA-10 cells were treated with Br-cAMP (0.4 mm) for the indicated times. Cells were harvested and total RNA subjected to Northern blot analysis. Expression of 28S RNA was monitored and similar in each sample. C, BAC were treated for the indicated times with and without ACTH (10 nm). Northern blots for Star mRNA are compared along with quantification of total mRNA, using 18S rRNA as a loading control.

Figure 2

Induction of TIS11b mRNA and multiple TIS11b phosphoproteins in steroidogenic cells. A, TIS11b mRNA levels in MA-10 and Y-1 cells stimulated with Br-cAMP (0.4 mm); B, TIS11b mRNA in BAC stimulated by ACTH; C, immunoblots of TIS11b protein expression in MA-10 cells with and without stimulation by Br-cAMP in comparison with transfected human TIS11b vector (1 μg/24 h); D, immunoblot of TIS11b protein expression in Y-1 cells with and without dephosphorylation by λ-protein phosphatase. TIS11b mRNA in MA-10 and Y-1 cells was quantified by real-time RT-PCR and standardized to β-actin mRNA levels. In BAC, TIS11b mRNA was standardized to HPRT mRNA level. Data represent mean ± sd for triplicate samples.

Figure 3

Effects of transfected TIS11b on luciferase vectors modified by Star 3′-UTR. A, Diagram of luciferase-Star 3′-UTR chimeric constructs used in the cotransfection experiments. Deletion of the AURE (UTRdARE) represents removal of the terminal 250 bases that include three AURE octameric motifs. AURE chimeric constructs include only this 250-bp sequence. AU/CG mutations were made in the upstream pairs of octameric elements (see Fig. 1). B, Effects of the indicated amounts of cotransfected TIS11b on the steady-state expression of luciferase-Star vectors in MA-10 cells after 24 h. *, P < 0.05 compared with UTRS vector; §, P < 0.05 compared with UTRL. C, Similar cotransfections of TIS11b on the steady-state expression of luciferase-ARE vectors. *,P < 0.05 compared with pGL3P vector; §, P < 0.05 compared with ARE. D, Effects of the indicated amounts of cotransfected TIS11b on the steady-state expression of luciferase-Star vectors in BAC after 48 h. *, P < 0.05 compared with UTRS. E, Effect of ACTH stimulation for various times on the expression levels of transfected luciferase-Star vectors in BAC. *, P < 0.05 compared with UTRS. Vector expression was measured in each experiment from the luciferase activity and normalized to levels of cotransfected renilla luciferase activity. The different mean levels of luciferase expression for each reporter without the TIS11b vector (UTRS > UTRdARE > UTRL; see Ref. 28) or the expression without ACTH treatment are set to 1.0. All experiments show the mean ± sd from three separate transfections.

Figure 4

Effects of TIS11b in MA-10 cells on the expression of Star vectors that differ in the length of the 3′-UTR. A, Diagram of rat Star expression vectors used in the cotransfections. B, Effect of cotransfected TIS11b on StAR protein expression after 24 h. Upper panel shows STAR immunoblots, whereas lower panel shows quantification of STAR protein levels at these exposures. C, Effect of cotransfected TIS11b on the expression of total Star mRNA measured by RT-PCR using primers specific for rat Star mRNA. Data represent mean ± sd for three separate cultures. *, P < 0.05 compared with empty vector cotransfection.

Figure 5

Degradation of directly transfected 3.5-kb Star mRNA is selectively enhanced by Br-cAMP stimulation but is independent of AURE mutations. A, Diagram showing experimental design. Rat Star mRNA with different 3′-UTR was transcribed from Star cDNA, which each had 90-base poly-A tails. The mRNA was transfected into MA-10 cells, and the time course for entry into MA-10 cells was determined by RT-PCR. Steady-state expression was attained after 12 h. Cells were treated with or without Br-cAMP at the time of mRNA transfection and also after a wash removal of the liposome/mRNA mix. Cells were lysed at the indicated times. B, First-order decay kinetics (log Star mRNA/total RNA vs. time) are shown for representative experiments, with or without Br-cAMP. C, Half-lives for each mRNA species calculated by linear regression fit of the time points on semi-log plots. Included is a second experiment in which 3.5-kb mRNA is mutated at two AURE sites (StAR3.5k12m). §, P < 0.05 for basal half-life compared with Star no-UTR control; *, P < 0.05 for effects of Br-cAMP.

Figure 6

Effects of Tis11b and HuR suppression on STAR protein induction and steroidogenesis. A, Expression of TIS11b, STAR, COX-2, and β-actin proteins. At 48 h after transfection of MA-10 cells with TIS11b siRNA or scrambled siRNA, cells were then stimulated with Br-cAMP for the indicated times. β-Actin is used for standardization. B, Y-1 cells were similarly treated with TIS11b siRNA and Br-cAMP. C, Effects of Tis11b siRNA on the stimulation of cholesterol metabolism by Br-cAMP in MA-10 cells. Shown are the rates of pregnenolone synthesis in 5-min periods after addition of trilostane to inhibit the rapid conversion of pregnenolone to progesterone. Trilostane was added at the indicated times after Br-cAMP stimulation. Pregnenolone was determined by RIA and standardized to total cellular protein levels. Data represent mean of duplicate samples. Experiments were repeated two times with similar results. D, Effective knockdown of HuR by siRNA does not change the expression of STAR in MA-10 cells.

Figure 7

Effects of TIS11b knockdown on StAR mRNA expression in BAC, MA-10, and Y-1 cells. A, BAC were transfected with TIS11b siRNA or scrambled siRNA. After 48 h, cells were either lysed to evaluate TIS11b suppression or incubated with fresh medium containing 10 nm ACTH for the indicated period of time. At each time of stimulation by ACTH, total RNA was isolated, and RT-PCR analysis was performed to determine total StAR mRNA expression levels. HPRT mRNA levels were used for standardization. Inset shows effects on TIS11b mRNA. B, Primer pairs used for selective RT-PCR quantification of Star short (primer pair 2) and long transcripts (primer pair1) after reverse transcription using poly-dT primers. C, Br-cAMP stimulation of Star transcript levels in MA-10 cells treated with either nontarget siRNA or TIS11b siRNA for 48 h. Primer pairs 1 and 2 are compared with β-actin mRNA for standardization to measure 3.5- and 1.6-kb transcripts. D, Equivalent analyses for Y-1 cells. Data represent mean ± sd for triplicate cultures.

Figure 8

Effects of TIS11b suppression on StAR 3.5- and 1.6-kb mRNA and protein stimulated by Br-cAMP in MA-10 cells. Northern blot analysis was carried out on StAR mRNA after treatments with either nontarget siRNA or TIS11b siRNA for 48 h. Total RNA was used from the cultures represented in Fig. 7C. The 28S rRNA levels were used as loading control. Quantifications of phosphoimager analyses of Northern blots are shown below the blots as the ratios of StAR/28S signal intensities. Quantifications of StAR protein levels from Fig. 6A are compared.

Figure 9

TIS11b selectively binds to extended Star 3′-UTR in BAC. A, RNA-protein UV cross-linking assay. In vitro transcribed and ³²P-labeled Star UTRS and UTRL RNA probes were mixed with bacterial cell extracts containing GST alone or GST-TIS11B in the presence or in the absence of cold RNA competitor. The reaction mixtures were treated with UV irradiation and analyzed by SDS-PAGE. Nonrelevant RNA transcribed from pGEM plasmid was used as a negative control. B, RNP complex immunoprecipitation and analysis by RT-PCR. RNP complexes were immunoprecipitated after reversible cross-linking between RNA and protein. RNA was then isolated from immunoprecipitates, treated with DNase I, and reverse transcribed. A PCR amplification of Star transcripts (lanes 1, 2, 4, 5, and 6) was then carried out. The PCR products were analyzed by agarose gel electrophoresis. In lane 3, PCR was performed with GAPDH primers to evaluate the specificity of the interaction. NIS, Nonimmune serum. Anti-N-Ter and Anti-C-Ter refer to TIS11b antibodies directed against N- or C-terminal regions of TIS11b.