A role for transportin in deposition of TTP to cytoplasmic RNA granules and mRNA decay

Abstract

Importin-beta family members, which shuttle between the nucleus and the cytoplasm, are essential for nucleocytoplasmic transport of macromolecules. We attempted to explore whether importin-beta family proteins change their cellular localization in response to environmental change. In this report, we show that transportin (TRN) was minimally detected in cytoplasmic processing bodies (P-bodies) under normal cell conditions but largely translocated to stress granules (SGs) in stressed cells. Fluorescence recovery after photobleaching analysis indicated that TRN moves rapidly in and out of cytoplasmic granules. Depletion of TRN greatly enhanced P-body formation but did not affect the number or size of SGs, suggesting that TRN or its cargo(es) participates in cellular function of P-bodies. Accordingly, TRN associated with tristetraprolin (TTP) and its AU-rich element (ARE)-containing mRNA substrates. Depletion of TRN increased the number of P-bodies and stabilized ARE-containing mRNAs, as observed with knockdown of the 5'-3' exonuclease Xrn1. Moreover, depletion of TRN retained TTP in P-bodies and meanwhile reduced the fraction of mobile TTP to SGs. Therefore, our data together suggest that TRN plays a role in trafficking of TTP between the cytoplasmic granules and whereby modulates the stability of ARE-containing mRNAs.

Figure 1.

Subcellular localization of importin-β family transporters in arsenite-treated cells. Double immunofluorescence was performed using a monoclonal antibody against TRN (A) or importin-β1 (B) as well as polyclonal anti-eIF4A in mock-treated and arsenite-treated (ARS) HeLa cells. For each of (C–F), an analogous immunofluorescence experiment was performed in HeLa cells that transiently expressed the indicated epitope-tagged importin-β family member using a polyclonal antibody against HA, FLAG or eIF4A and monoclonal anti-PABP or anti-HA. In A and B, arrows indicate colocalization of importin-β proteins with endogenous SG marker proteins.

Figure 2.

Subcellular colocalization of TRN with SG and P-body markers. (A) HeLa cells that transiently expressed GFP-TIA1, or HA-tagged HuR or PABP1 were treated with arsenite (ARS) and subjected to immunofluorescence using anti-TRN. GFP-TIA1 fluorescence was directly detected under a microscope whereas HA-tagged HuR and PABP1 were detected by using anti-HA. Merged signals indicated by arrows represent SGs. (B) Double immunofluorescence was performed to detect endogenous TRN and Dcp1a or EDC4 under normal and arsenite-stressed conditions using monoclonal anti-TRN and polyclonal anti-Dcp1a or anti-Edc4. Insets represent enlargements of indicated regions. (C) The myc-tagged Dcp1a was transiently overexpressed in HeLa cells. Double immunofluorescence was performed to detect endogenous TRN or importin-β1 and myc-Dcp1a. (D) Immunofluorescence was used to detect endogenous importin-β1 or the indicated transiently expressed nuclear transporter as well as endogenous Dcp1a in arsenite-stressed HeLa cells. Epitope-specific antibodies were used to detect tagged proteins in C and D. Merged signals indicated by arrowheads represent P-bodies in B–D.

Figure 3.

FRAP analysis of granule-localized proteins in arsenite-treated cells. FRAP was performed in HeLa cells that transiently expressed GFP-TRN, GFP-TIA1 or GFP-CPEB1. FRAP was performed following arsenite treatment. Fluorescence signal of selected granules was irradiated with the 488-nm laser, and its recovery was recorded for 200 s. (A) Selected images of GFP-fusion protein-expressing cells before bleaching and at the indicated intervals post-bleaching (from 0 to 200 s) are shown. Arrows indicate the targeted SGs. (B) The fluorescence intensity of GFP-fusion proteins after photobleaching was normalized to that before bleaching as described in ‘Materials and Methods’ section; relative intensity of each protein was averaged from at least three independent experiments (n = 5 for each experiment). The t_1/2 value is also indicated.

Figure 4.

Effect of TRN knockdown on cytoplasmic granules. (A) HeLa cells were mock-transfected or transfected with siRNA against TRAP150, lamin A/C, TRN or Xrn1. Representative images of double immunostaining using anti-TRN and anti-Dcp1a. Bar graph: percentage of P-bodies in mock-transfected or siRNA-transfected cells (n, the number of cells used for quantification). Proteins in total cell lysates were subjected to immunoblotting using antibodies against the indicated proteins. (B) As in A, TRN was depleted by siRNA in HeLa cells. Cells were mock-treated or treated with arsenite followed by double immunostaining with anti-Dcp1a and anti-PABP. Merge indicates superimposed images. Immunoblotting of TRN and β-actin was performed.

Figure 5.

The interaction between TRN and TTP. HEK293 cells were transiently transfected with the FLAG-TTP expression vector or empty vector. (A) Cells were treated with arsenite or mock-treated for 1 h before harvest. The cell lysates were subjected to immunoprecipitation with anti-FLAG in the presence of RNase A. The precipitates were analyzed by immunoblotting with anti-TRN and anti-FLAG. (B) The in vitro translation reaction containing ³⁵S-labeled TTP was incubated with GST or GST-TRN followed by chromatography on glutathione-Sepharose. Bound proteins were fractionated on SDS–PAGE and detected by autoradiography. (C) HeLa cells that transiently expressed FLAG-TTP were double immunostained with anti-FLAG and anti-TRN. (D) Upper panel: ³⁵S-labeled HuR and TTP proteins were each subjected to the pull-down assay as in B. Recombinant RanQ69L was loaded with GTP and subsequently added into the pull-down reaction (lanes 4 and 8). Bottom panel: HeLa cells that transiently expressed FLAG-TTP and HA-Ran or HA-RanQ69L were double immunostained with anti-FLAG and anti-HA.

Figure 6.

Association of TTP with ARE-containing mRNAs. (A) The TNFα ARE-containing β-globin reporter (TNF-UTR) or the control reporter (TNF-ΔARE), as shown in the top schematics, was transfected into HeLa cells. The cell lysates were subjected to immunoprecipitation using anti-TRN. Precipitated RNAs were analyzed by RT–PCR using primers specific to the reporters. (B) HeLa cells were mock-transfected or transiently transfected with siRNA against TRN. Depletion of TRN was examined by immunoblotting using anti-TRN. For mRNA stability analysis, the reporters used were the same as those in A, and meanwhile a GFP expression vector was cotransfected as a reference. Cells were harvested at the indicated time points after doxycycline addition. RNase protection was performed using ³²P-labeled probes specific to the reporter or the GFP mRNA. The graph shows the level of reporter mRNA remaining at each time point relative to that of the GFP reference. Arbitrary mRNA intensity was determined by phosphoimaging. Average and standard deviation were obtained from three independent experiments. The representative gel shows an experiment using the TNF-UTR reporter.

Figure 7.

Localization of TTP in cytoplasmic granules in TRN-depleted cells. (A) HeLa cells were transiently transfected without or with TRN siRNA and the FLAG-TTP expression vector. Double immunostaining was performed using anti-FLAG and an antibody specific to Dcp1a or PABP1. Arrowheads and arrows indicate merged immunofluorescence of FLAG-TTP with that of Dcp1a and PABP1, respectively. Bar graph: percentage of transfected cells that contained FLAG-TTP-costained SGs or P-bodies. The average was obtained from three independent experiments (∼100 transfected cells were counted in each experiment). (B) FRAP analysis of GFP-TTP in mock- and TRN-depleted cells. Following photobleaching, the fluorescence recovery of GFP-TTP in granules was recorded for ∼200 s. Relative fluorescence intensity of GFP-TTP was measured and calculated as in Figure 3.