Phosphorylation of tristetraprolin by MK2 impairs AU-rich element mRNA decay by preventing deadenylase recruitment

Abstract

mRNA turnover is a critical step in the control of gene expression. In mammalian cells, a subset of mRNAs regulated at the level of mRNA turnover contain destabilizing AU-rich elements (AREs) in their 3' untranslated regions. These transcripts are bound by a suite of ARE-binding proteins (AUBPs) that receive information from cell signaling events to modulate rates of ARE mRNA decay. Here we show that a key destabilizing AUBP, tristetraprolin (TTP), is repressed by the p38 mitogen-activated protein kinase (MAPK)-activated kinase MK2 due to the inability of phospho-TTP to recruit deadenylases to target mRNAs. TTP is tightly associated with cytoplasmic deadenylases and promotes rapid deadenylation of target mRNAs both in vitro and in cells. TTP can direct the deadenylation of substrate mRNAs when tethered to a heterologous mRNA, yet its ability to do so is inhibited upon phosphorylation by MK2. Phospho-TTP is not impaired in mRNA binding but does fail to recruit the major cytoplasmic deadenylases. These observations suggest that phosphorylation of TTP by MK2 primarily affects mRNA decay downstream of RNA binding by preventing recruitment of the deadenylation machinery. Thus, TTP may remain poised to rapidly reactivate deadenylation of bound transcripts to downregulate gene expression once the p38 MAPK pathway is deactivated.

FIG. 1.

Tethered TTP promotes mRNA deadenylation. (A) Northern blot assays of transcriptional pulse-chase mRNA decay assays in human HeLa (53) Tet-off cells of a reporter mRNA bearing six MS2 coat protein binding sites in the 3′ UTR (β-6bs) in the presence of exogenous MS2-TTP or MS2 protein alone, as indicated at the left. The time points at the top are minutes after transcriptional repression by tetracycline addition. β-6bs mRNA t_1/2s are indicated at the right and were calculated by normalizing to the constitutively expressed internal control, β-GAP mRNA. Arrows indicate the two distinct bands that reflect poly(A) (A_n) and deadenylated (A₀) mRNAs, as determined by comigration with deadenylated β-6bs mRNA generated by RNase H digestion in the presence of oligo(dT) (lane 7). (B) Northern blot assays as in panel A showing the decay of β-6bs in the presence of an exogenously expressed dominant negative form of the decapping enzyme hDcp2-E148Q together with MS2-TTP or MS2. (C) Northern blot assays showing the decay of β-6bs in the presence of an exogenously expressed dominant negative form of the deadenylase hCaf1b-D40,44A together with MS2-TTP or MS2.

FIG. 2.

TTP copurifies with deadenylase activity. (A) In vitro deadenylase assay with FLAG-tagged TTP purified from human cells. 5′-³²P-^m7G-labeled deadenylation substrate (ARE-A₆₀) (lanes 4 to 8) and/or nonadenylated control (Ctrl) RNA (lanes 4 to 12) were incubated for 1 h at 30°C with increasing amounts of FLAG-TTP immunopurified from HEK293T cells (lanes 5 to 12). The control (lane C), ARE-A₆₀ (A₆₀), and deadenylated ARE-A₆₀, [A₀; generated by RNase H-oligo(dT) cleavage] RNAs, are shown in lanes 1 to 3. (B) ARE-A₆₀ substrate incubated with increasing amounts of ARE binding-deficient TTP protein, TTP-F126N (lanes 4 to 7), and a mutated ARE substrate (ARE-MUT-A₆₀) incubated with increasing amounts of wild-type TTP (lanes 11 to 14).

FIG. 3.

TTP copurifies with multiple deadenylases. Shown are coimmunoprecipitation assays of Myc-tagged deadenylase proteins with FLAG-tagged TTP (lanes 3 and 4) coexpressed in HEK293T cells. RNase A-treated cell extracts were subjected to anti-FLAG immunopurification (IP), and precipitates were detected via immunoblotting with anti-Myc antibody. Negative immunoprecipitation controls (lanes 1 and 2) were derived from cells expressing no FLAG-tagged proteins. Lanes 1 and 3, 5% of total input fractions (T); lanes 2 and 4, pellet fractions (P). Myc-hnRNP A1, an RNA-binding protein, served as a negative coimmunoprecipitation control. (B) Deadenylase assay performed as described in the legend to Fig. 2 but with the addition of increasing amounts of FLAG-PABPC1 purified from RNase A-treated HEK293T cells. (C) Deadenylase assay using a 5′-³²P-monophosphorylated ARE-A₆₀ substrate instead of a cap-labeled substrate. A nonspecific 3′-truncated product derived from the ARE-A₆₀ substrate is indicated by the asterisk. wt, wild type.

FIG. 4.

Phosphorylation of TTP by MK2 inhibits TTP-mediated mRNA deadenylation in cells. (A) Northern blot assays showing decay rates of β-6bs in the presence of tethered wild-type TTP (WT) or TTP-AA, in which serines 52 and 178 have been mutated to alanines. Cells were transfected with TTP expression vectors in the presence of no exogenous kinase (−), a constitutively active mutant kinase, MK2EE, or a kinase-dead mutant kinase, MK2KR, as indicated on the left. Calculated β-6bs mRNA t_1/2s and average n-fold changes (average of three experiments) in the presence of phosphorylated TTP relative to nonphosphorylated TTP are given on the right. (B) Northern blot assays of mRNA decay assays as in panel A but without (lanes 1 to 7) or with (lanes 8 to 12) coexpression of the catalytically inactive form of the human decapping enzyme hDcp2-E148Q in the presence of MS2-TTP and MK2EE (top panel) or MK2KR (bottom panel).

FIG. 5.

Phosphorylation of TTP does not impair mRNA binding. (A) Coimmunoprecipitation assays performed with extracts derived from HEK293T cells coexpressing FLAG-tagged cytoplasmic poly(A)-binding protein (PABPC1) (lanes 2, 3, 5, and 6) together with Myc-tagged TTP, hnRNP A1, TTP_1-100 (lanes 1 to 6), and MK2EE (lanes 1 to 3) or MK2KR (lane 4 to 6). Immunoprecipitations (IP) were performed in the absence (lanes 1, 2, 4, and 5) or presence (lanes 3 and 6) of RNase A. Lanes 1 and 4 are negative controls from cells expressing no FLAG-tagged protein. Myc-tagged hnRNP A1 served as a positive RNA binding control, whereas the N-terminal domain of TTP (TTP_1-100) served as a negative control that does not bind RNA. Pellet and 5% of input fractions were loaded as indicated on the right and subjected to immunoblotting (Western blotting [WB]) with anti-myc or anti-FLAG antibodies as indicated on the left. (B) Coimmunoprecipitation assays as in panel A, except that cells were pretreated with the PP2A inhibitor okadaic acid (OA; lanes 1 to 3) or the vehicle (lanes 4 to 6) 2 h prior to lysate preparation. (C) Northern blot assays of RNA immunoprecipitated from cells coexpressing β-GAP and β-ARE reporter mRNAs together with FLAG-tagged wild-type (WT) TTP (lanes 1 and 2), or mutant versions of TTP with serines 52 and 178 mutated to alanines (AA, lanes 3 and 4) or bearing a point mutation that abolishes RNA binding (F126N, lanes 5 and 6). Prior to immunoprecipitation, cells were treated for 2 h with (+) or without (−) okadaic acid. TTP protein levels are shown in Western blot assays below each Northern blot assay. Upper panels, pellets; lower panels, 10% input. Lanes 7 and 8, immunoprecipitates from cells expressing no exogenous TTP protein.

FIG. 6.

Phosphorylation of TTP prevents deadenylase recruitment. (A) Anti-FLAG coimmunoprecipitation assays performed with extracts from HEK293T cells coexpressing FLAG-tagged deadenylase hPan2 or hCcr4b together with Myc-tagged TTP (WT) or the TTP-AA mutant protein as indicated. Cells were treated with the phosphatase inhibitor okadaic acid (OA) (lanes 1, 3, 7, and 10) or the vehicle (lanes 2, 4, 5, 6, 8, and 9). Immunoprecipitates (IP) from cells with no FLAG-tagged protein served as a negative control (lanes 5 and 8). (B) Western blot (WB) assays showing coimmunoprecipitation of myc-tagged TTP (WT or AA mutant) and FLAG-tagged 14-3-3ɛ (lanes 1 to 4). Cells were treated with (lanes 2, 4, 6, and 8) or without (lanes 1, 3, 5, and 7) okadaic acid. Myc-tagged hnRNP A1 (lanes 1 to 8) and immunoprecipitation from cells with no FLAG-tagged protein (lanes 5 to 8) served as negative controls. (C) Western blot assays of coimmunoprecipitation of transiently expressed FLAG-tagged 14-3-3ɛ and Myc-tagged TTP-WT, TTP-AA, or TTP-EE from HEK293T extracts. Pellet fractions (P) and 5% of the total (T) fractions are indicated above the panels. (D) Northern blot assays showing decay rates of β-6bs in the presence of tethered TTP-WT, TTP-AA, or TTP-EE with serines 52 and 178 mutated to negatively charged glutamate residues in an attempt to mimic serine phosphorylation. t_1/2s are indicated on the right.

FIG. 7.

Phosphorylation of TTP impairs target mRNA decay by preventing deadenylase recruitment. TTP activates deadenylation by binding mRNA and recruiting cytoplasmic deadenylases (shown as biting circles). Phosphorylation of TTP by MK2 inhibits deadenylation downstream of mRNA binding by preventing deadenylase recruitment in a manner inversely correlated with binding of the adaptor protein 14-3-3.