Proteomic analysis of cap-dependent translation identifies LARP1 as a key regulator of 5'TOP mRNA translation

Abstract

The mammalian target of rapamycin (mTOR) promotes cell growth and proliferation by promoting mRNA translation and increasing the protein synthetic capacity of the cell. Although mTOR globally promotes translation by regulating the mRNA 5' cap-binding protein eIF4E (eukaryotic initiation factor 4E), it also preferentially regulates the translation of certain classes of mRNA via unclear mechanisms. To help fill this gap in knowledge, we performed a quantitative proteomic screen to identify proteins that associate with the mRNA 5' cap in an mTOR-dependent manner. Using this approach, we identified many potential regulatory factors, including the putative RNA-binding protein LARP1 (La-related protein 1). Our results indicate that LARP1 associates with actively translating ribosomes via PABP and that LARP1 stimulates the translation of mRNAs containing a 5' terminal oligopyrimidine (TOP) motif, encoding for components of the translational machinery. We found that LARP1 associates with the mTOR complex 1 (mTORC1) and is required for global protein synthesis as well as cell growth and proliferation. Together, these data reveal important molecular mechanisms involved in TOP mRNA translation and implicate LARP1 as an important regulator of cell growth and proliferation.

Keywords: 5′TOP; LARP1; mRNA; mTOR; proteomics; translation.

Figure 1.

Global quantitative assessment of proteins that associate with the mRNA 5′ cap. (A) Enrichment of eIF4E (immunoblot in the bottom panel) and associated factors after specific elution with m⁷GTP (Coomassie gel; fifth lane). (B) Schematic diagram of the multiplex work flow developed for identification and quantitation of the 5′ cap-binding complex by combining m⁷GTP pull-down and TMT⁶ labeling in the presence or absence of RNase/Benzonase for 30 or 60 min. (C) Distribution of ∼160 proteins found to be associated to the 5′ cap-binding complex. The RNase dependence ratios (treated/untreated) are plotted on a log₂ scale, normalized with respect to eIF4E abundance. Proteins with a log₂ ratio ≤1.3 (PABPC1 threshold), such as DDX1, NONO, and IGF2BP3, were considered RNA-dependent. (D) List of the most abundant associated proteins ranked according to the numbers of peptides identified. Gray shading indicates proteins previously shown to interact directly with eIF4E (purple). Accession numbers are from the International Protein Index (IPI). (E) Classification of RNA-insensitive candidates according to cellular and molecular functions or canonical signaling pathways using the DAVID bioinformatics database (http://david.abcc.ncifcrf.gov) or the Ingenuity Pathway Analysis platform (IPA; http://www.ingenuity.com) according to adjusted P-value. The gray line indicates minimum threshold (P = 0.05).

Figure 2.

mTOR-dependent regulation of the 5′ cap-binding complex. (A) Schematic diagram of the multiplex work flow strategy for the identification and quantitation of mTOR-regulated components. Cell treatments were performed in triplicate, and all m⁷GTP pull-down assays were processed and TMT⁶ labeled individually prior to being pooled for LC-MS/MS analysis and quantification. These experiments were also performed in the presence of RNase. (B) Effects of mTOR activity on eIF4F complex assembly using mTOR agonists and antagonists. Treatment of serum-growing cells with the dual mTOR/PI3K inhibitor PI-103 (1 μM) caused dissociation of eIF4G and eIF4A with an increased association of 4E-BP1 (second lane, top left panels), while insulin treatment (100 nM) for 30 min had an opposite effect (fourth lane, top right panels). The bottom panels show loading controls and treatment effects (4E-BPs phosphorylation shifts). (C) Bar graphs of TMT⁶-based quantifications for the same conditions. (***) P < 0.0001 using two-way ANOVAs. (ns) Nonsignificant. (D) Similar effects were seen in wild-type MEFs, but complete loss of regulation was observed in 4E-BP1/2 DKO cells. (E) Bar graphs of TMT values from DKO cells also showing a loss of regulation.

Figure 3.

Identification of novel components of the 5′ cap-binding complex. (A) Plot highlighting the 5′ cap association ratios in response to PI-103 (PI-103/DMSO; red squares) and insulin treatment (insulin/untreated; blue squares). Distributions obtained with 4E-BP1/2 DKO cells show loss of mTOR regulation (red shading) and were used to set minimum threshold for wild-type cells (gray shading). Proteins with a log₂ PI-103/DMSO ratio ≥0.7 and insulin/untreated ratio ≤0.7, such as FAM98A, hnRNPU, DDX6, LARP1, ILF2, eIF4G1, eIF4A, eIF3L, and PABPC1, were considered mTOR-regulated. Only 4E-BP1/2 showed opposite mTOR-dependent regulation. (B) Scattered plot distribution showing that a majority of insulin-regulated proteins were also inversely regulated by PI-103 (R² = 0.6186). (C,D) Bar graphs of TMT values from work flow above depicting expected eIF4F-associated proteins, including 4E-T, DDX3, and eIF3A (C), and novel eIF4F-associated proteins, including LARP1, DDX6, hnRNPU, FAM98A, DHX9, and HSPA5 (D). (*) P < 0.01 using two-way ANOVAs. (E) Several candidate proteins were also validated by immunoblotting, and all were confirmed to have mTOR-regulated association to the mRNA 5′ cap.

Figure 4.

mTOR regulates LARP1 association to polysomes via PABP. (A) Ribosome sedimentation profiling from HEK293 cell extracts. LARP1 cosedimented with 40S, 60S, and 80S subpolysomal fractions as well as with polysomal fractions. Control proteins were distributed as expected from previous studies: S6 in 40S, 80S, and polysomal fractions; L5 in 60S, 80S, and polysomal fractions; eIF4F (eIF4E, eIF4A, and eIF4G) primarily in the 40S fractions; and PABP in all four fractions, as measured by immunoblotting. (B) Treatment of cells with 1 μM PI-103 for 1 h significantly decreased polysome assembly and concomitantly displaced polysome-associated components L5, S6, PABP, and LARP1 to subpolysomal fractions. (C–E) PABP coimmunoprecipitates with LARP1 independently of mTOR activation (100 nM insulin, 30 min) or inhibition (PI-103 and Ku-0063794 for 1 h). The bottom panels show loading and treatment controls by measuring S6 phosphorylation on Ser240/244. (F,G) LARP1 contains two putative RNA-binding domains (LAM and RRM) and a C-terminal stretch of DM15 (LARP1) motifs with unknown functions. Partial (ΔC150) and complete (ΔC300) LARP1 C-terminal deletion of the DM15 stretch drastically reduced association with PABP (G) and cap binding (H). (I,J) Sucrose gradient velocity sedimentation showing distribution of wild-type myc-LARP1 (I) in both subpolysomal and polysomal fractions, while the C-terminal mutant (ΔC150) (J) primarily cosedimented with subpolysomal fractions. (K) The bar graph shows densitometry analysis of the abundance of wild-type LARP1 and the ΔC150 mutant LARP1 in polysomal versus subpolysomal fractions.

Figure 5.

LARP1 regulates protein synthesis, cell cycle progression, and proliferation. (A) HEK293 stably expressing different shRNA constructs targeted against LARP1 (shLARP1.1 and shLARP1.3) show decreased global protein synthesis as measured by [³H]leucine incorporation, compared with control cells (shNT). (B,C) HEK293 cells stably expressing different shRNA constructs targeted against LARP1 (LARP1.1, LARP1.2, and LARP1.3) display significantly decreased cell proliferation as measured by cell counting at 24 h, 48 h, and 72 h. (B) DIC images show representative examples of HEK293 cells stably expressing shRNAs against LARP1. (D) Comparable with the inhibition of mTOR using Ku-0063794, knockdown of LARP1 with three different shRNAs results in a cell cycle arrest and accumulation of cells at G0/G1. (E) Endogenous LARP1 coimmunoprecipitates with HA-tagged Raptor in HEK293 cells. (F) Endogenous Raptor, but not Rictor, coimmunoprecipitates with myc-tagged LARP1. (G) The interaction between LARP1 and Raptor is not significantly affected by the presence of detergents that disrupt mTORC1. (H) Immunofluorescence showing myc-tagged LARP1 (in green) colocalizing with HA-tagged Raptor (in red), notably in puncta (insets). Nuclei were stained with DAPI.

Figure 6.

LARP1 associates with and regulates 5′TOP mRNA translation. (A) Ribosome profiling of HEK293 cells stably expressing a nontarget shRNA (shNT) or shRNAs against LARP1 (shLARP1.1 and shLARP1.2). A decrease in polysome assembly with a concomitant increase in 80S monosomes (polysome/subpolysome ratios of 0.37 and 0.38) was found in LARP1-depleted cells. (Bottom right panels) Immunoblot analysis reveals a modest effect on 4E-BP phosphorylation and S6 ribosomal protein expression. The bottom histogram shows ribosome mRNA profiling of LARP1-depleted cells (shL1.1 and shL1.2) with a significant reduction (∼50%) for 5′TOP mRNAs (average of 10 TOP mRNAs; bottom right panel; see the Materials and Methods) with no significant effects on control mRNAs (average of 10 NON-TOP mRNAs; bottom left panel; also see the Materials and Methods) when compared with a nontarget shRNA (shNT). (B) Treatment of cells with mTOR inhibitors (PI-103 and Ku-0063794 for 1 h) had a similar but stronger effect on polysome disassembly and 4E-BP phosphorylation. The bottom histogram shows ribosome mRNA profiling of cells treated with mTOR inhibitors with a significant reduction (∼90%) for 5′TOP mRNAs (average of 10 TOP mRNAs; bottom right panel; see the Materials and Methods) with no significant effects on control mRNAs (average of 10 NON-TOP mRNAs; bottom left panel; see the Materials and Methods) when compared with untreated cells (DMSO). Absorbance of polysomes and subpolysomal particles was continuously monitored at 260 nm. Representative A₂₆₀-nm traces are shown (n = 3). The areas under the curves were calculated, and the polysome/subpolysome ratio in the histograms refers to the percentage of ribosomes engaged in translation. The data are normalized to polysome/subpolysome ratio of control condition (DMSO) and presented as a mean ± SE (n = 3). (***) P < 0.0001 using two-way ANOVAs. (C) Diagram of the experimental work flow for the RNA immunoprecipitation (RIP). (D) RIP of endogenous LARP1 (left panel) and exogenous LARP1 (right panel) showing enrichment of TOP mRNAs (average of five TOP mRNAs; bottom right panel; also see the Materials and Methods) versus non-TOP (average of 10 non-TOP mRNAs; bottom left panel; also see the Materials and Methods) compared with control IgG or empty vector. (E) Immunoblot analysis showing a significant decrease in the expression of protein encoded by 5′TOP mRNAs in HEK293 cells stably expressing an shRNA construct targeted against LARP1 (shL1.3).

Figure 7.

Schematic representation of the role of LARP1 in mRNA translation. Proposed model in which mTORC1 mediates cell growth and proliferation by promoting 5′TOP mRNA translation through known mechanisms involving inhibition of the 4E-BPs but also by facilitating the recruitment of 5′TOP mRNAs via their interaction with LARP1.