Stress granules are dispensable for mRNA stabilization during cellular stress

Abstract

During cellular stress, protein synthesis is severely reduced and bulk mRNA is recruited to stress granules (SGs). Previously, we showed that the SG-recruited IGF2 mRNA-binding protein 1 (IGF2BP1) interferes with target mRNA degradation during cellular stress. Whether this requires the formation of SGs remained elusive. Here, we demonstrate that the sustained inhibition of visible SGs requires the concomitant knockdown of TIA1, TIAR and G3BP1. FRAP and photo-conversion studies, however, indicate that these proteins only transiently associate with SGs. This suggests that instead of forming a rigid scaffold for mRNP recruitment, TIA proteins and G3BP1 promote SG-formation by constantly replenishing mRNPs. In contrast, RNA-binding proteins like IGF2BP1 or HUR, which are dispensable for SG-assembly, are stably associated with SGs and the IGF2BP1/HUR-G3BP1 association is increased during stress. The depletion of IGF2BP1 enhances the degradation of target mRNAs irrespective of inhibiting SG-formation, whereas the turnover of bulk mRNA remains unaffected when SG-formation is impaired. Together these findings indicate that the stabilization of mRNAs during cellular stress is facilitated by the formation of stable mRNPs, which are recruited to SGs by TIA proteins and/or G3BP1. Importantly, however, the aggregation of mRNPs to visible SGs is dispensable for preventing mRNA degradation.

Figure 1.

The concomitant depletion of TIA proteins and G3BP1 impairs SG formation. (A) U2OS cells transfected with indicated siRNAs (siC, control; siSG, siTIA1, siTIAR and siG3BP1) for 72 h were stressed by arsenate (2.5 μM) for 2 h before immunostaining of indicated proteins. Enlargements of the boxed regions in the merged images are shown in the right panel. Bars, 25 μm. (B and C) The average number of SG containing cells (B) and the SG area fraction (C) was analyzed by immunostaining for IGF2BP1 and YB1 in U2OS cells transfected with indicated siRNAs. Both parameters were determined by an automated particle detection tool, adapted from (18), after indicated times of arsenate stress. Error bars indicate SD determined by analyzing at least 100 cells per condition in three independent experiments. (D) The phosphorylation of eIF2α in non-stressed (−) or arsenate (+, as in A) stressed U2OS cells transfected with control (siC) or siSG (as in A) siRNAs was determined by western blotting with indicated antibodies. VCL and TUB4A4 served as loading controls to determine knockdown efficiencies as indicated by numbers above each panel. Standard deviation was determined from three independent experiments. (E) The association of bulk (m)RNA with polysomes was monitored by linear (15–45% w/v) sucrose gradient centrifugation in stressed (+, arsenate) versus non-stressed (−, arsenate) U2OS cells transfected with control (siC) or siSG siRNAs, as in (A). The distribution of RNA was monitored by UV spectroscopy and is shown as the average absorbance determined for individual fractions isolated from siC- and siSG-transfected samples. Error bars indicate SD determined in three independent studies for siC- and siSG-transfected cells. (F) Protein synthesis in arsenate-stressed (+) versus non-stressed (−) U2OS cells transfected as in (A) was analyzed by metabolic labeling using S³⁵-methionine. The fraction of newly synthesized proteins was determined by western blotting using autoradiography (right panel). Equal loading was controlled by Ponceau staining (left panel).

Figure 2.

SG dynamics distinguish two classes of SG-associated RBPs. (A and B) The averaged recovery of GFP-fluorescence for the indicated proteins in SGs (red) or SG-free cytoplasm (black) in U2OS cells stably expressing and/or transiently transfected with the indicated proteins was determined by FRAP. Representative images of fluorescence intensities observed in SGs for GFP-G3BP1 (A) or GFP-ZBP1 (B) at indicated times of FRAP studies are shown in pseudo-colors (glow over/under) in the left panels. (C and D) The change in SG-localized photo-converted Dendra-fused G3BP1 (C) or ZBP1 (D) was determined by time-lapse microscopy. The averaged change of fluorescence intensities was determined over 30 s after photo-conversion. Representative images of fluorescence intensities observed at indicated time after photo-conversion are shown in left panels. Error bars indicate SD determined for the number of analyses summarized together with kinetic data in Supplementary Figure S8A. Bars, 5 μm.

Figure 3.

The association of G3BP1 with IGF2BP1 and HUR is enhanced during cellular stress. (A) The co-immunopurification of IGF2BP1 or HUR with G3BP1 from arsenate-stressed (+) or non-stressed (−) HEK293 cells was analyzed by western blotting. Where indicated, lysates were supplemented with RNaseA/T1. ProteinG dynabeads served as negative control. Copurification of IGF2BP1 or HUR with G3BP1 was quantified relative to immunopurified G3BP1 amounts by quantitative western blotting. The ratio of IGF2BP1/HUR copurified with G3BP1 was set to one, as indicated above by numbers above lanes. Standard deviation of copurification was determined from three independent experiments. (B) The co-immunopurification of endogenous G3BP1 with indicated proteins transiently expressed in HEK293 cells was analyzed by western blotting. Note that RNA binding of ZBP1-KH1-4 is substantially impaired by point mutation in all four KH domains (17). Flag-GFP served as negative control. (C) Schematic of G3BP1 domains, putative functions of the indicated domains and relative position of domains indicated by the numbering of residues. (D) Co-immunopurification of endogenous IGF2BP1 with indicated stably expressed GFP or GFP-fused G3BP mutant proteins. G3BP proteins analyzed: WT, wild type G3BP1; ΔRGG, G3BP1 lacking amino acids 425–466; FFVV, full-length G3BP1 with F-V conversion at residues 380 and 382; S149A, full-length G3BP1 with S-A conversion at residue 149. Western blotting for indicated proteins is shown for the input or co-immunopurified (α-GFP) protein fraction. HEK293 cells stably expressing GFP served as negative controls. * indicates degradation product.

Figure 4.

The forced expression of G3BP1 protein mutants modulates SG formation. (A) The formation of SGs was monitored in arsenate stressed (1 h) U2OS cells stably expressing GFP or indicated GFP-fused G3BP1 mutant proteins using immunostaining of SG-localized IGF2BP1 and TIA1. Enlargements of boxed regions depicted in the merged images are shown in the right panel. Bars, 25 μm. (B and C) The number of SG-positive cells (B) as well as the SG-area fraction (C) was determined as described in Figure 1B and C. Error bars indicate SD of at least three independent analyses including at least 30 cells per condition. Statistical significance was determined by Student's t-test: **P < 0.005.

Figure 5.

The control of mRNA turnover is independent of SG formation. (A) The turnover of indicated mRNAs was analyzed in U2OS cells transfected with indicated siRNAs (siC, control; siSG, siTIA1, siTIAR and siG3BP1) for 72 h. Cells were treated with arsenate and actinomycin D (ActD + arsenate) for indicated times. For RPLP0, PPIA and VCL mRNA abundance was determined relative to untreated samples by the ΔC_t method. For ACTB, MAPK4 and MYC, changes in mRNA levels relative to input (untreated) controls were determined by cross-normalization to RPLP0 using the ΔΔC_t method. (B) The turnover of indicated stress-induced mRNAs was analyzed in U2OS cells transfected as in A. Before monitoring mRNA turnover by ActD addition, cells were pre-stressed by arsenate for 1 h (gray). RNA levels observed after 1 h of arsenate stress were set to one. The decay of mRNAs was analyzed by qRT-PCR using the ΔΔC_t method and RPLP0 for internal cross-normalization as in A. Error bars indicate SD of at least three independent analyses. Statistical significance was determined by Student's t-test. (C–E) The abundance of RNAs in stressed (arsenate and ActD for 2 h) and non-stressed U2OS cells transfected as in A was analyzed by comparative microarray analyses. The concomitant knockdown of indicated proteins was confirmed by western blotting with VCL and ACTB serving as internal controls (C). The change of RNA abundance in response to the triple knockdown (siSG, as in A) was determined relative to siC-transfected controls (D). The siSG/siC ratio of transcripts is shown for the average of two independent analyses in stressed and non-stressed U2OS cells. Transcripts with an at least 2-fold distance to the origin in any direction are indicated in green. Transcripts selectively decreased at least 2-fold in the siSG-transfected populations during stress without significantly changed abundance under non-stressed conditions are indicated in red. Transcripts selectively upregulated during stress are indicated in blue. The number of transcripts in the depicted color-coded classes (D) or with an at least 2-fold distance to the origin in the four quadrants (D) was determined without any normalization of array data (E, w/o) or using indicated tools for normalization (E, Mas5 or RMA).

Figure 6.

IGF2BP1 stabilizes target mRNAs in a SG-independent manner. (A) The knockdown of indicated proteins in U2OS cells transfected with indicated siRNAs or siRNA mixtures (C, control; SG, siTIA1, siTIAR and siG3BP1) for 72 h was analyzed by western blotting. VCL served as loading control. (B) The abundance of indicated mRNAs in cells transfected as in (A) and stressed by arsenate for 2 h was determined by qRT-PCR relative to controls (siC) using the ΔΔC_t method and RPLP0 for internal cross-normalization. (C) The turnover of indicated mRNAs was analyzed in U2OS cells stably expressing GFP (black) or GFP-tagged ZBP1 (gray) treated with ActD and arsenate for indicated time. The abundance of indicated mRNAs was analyzed by qRT-PCR relative to untreated samples using the ΔΔC_t method and RPLP0 for internal cross-normalization. Error bars indicate SD of at least three independent analyses. Statistical significance was determined by Student's t-test. *P < 0.05; **P < 0.005.