The translational response of the human mdm2 gene in HEK293T cells exposed to rapamycin: a role for the 5'-UTRs

Abstract

Polysomal messenger RNA (mRNA) populations change rapidly in response to alterations in the physiological status of the cell. For this reason, translational regulation, mediated principally at the level of initiation, plays a key role in the maintenance of cellular homeostasis. In an earlier translational profiling study, we followed the impact of rapamycin on polysome re-seeding. Despite the overall negative effect on transcript recruitment, we nonetheless observed that some mRNAs were significantly less affected. Consequently, their relative polysomal occupancy increased in the rapamycin-treated cells. The behaviour of one of these genes, mdm2, has been further analysed. Despite the absence of internal ribosome entry site activity we demonstrate, using a dual reporter assay, that both the reported mdm2 5'-UTRs confer resistance to rapamycin relative to the 5'-UTR of β-actin. This relative resistance is responsive to the downstream targets mTORC1 but did not respond to changes in the La protein, a reported factor acting positively on MDM2 translational expression. Furthermore, extended exposure to rapamycin in the presence of serum increased the steady-state level of the endogenous MDM2 protein. However, this response was effectively reversed when serum levels were reduced. Taken globally, these studies suggest that experimental conditions can dramatically modulate the expressional output during rapamycin exposure.

Figure 1.

Controls for IRES activity within the mdm2 5′-UTRs. (A) A schematic representation of the organization of the human L-mdm2 and S-mdm2 5′-UTRs. The rectangles refer to exons, P1/P2 are the alternative promoters and AUG indicates the MDM2 start codon. (B) Semi-quantitative RT–PCR that follows L-mdm2 and S-mdm2 levels in HEK293T cells in the absence (DMSO, 24 h) or the presence of rapamycin (100 nM, 24 h). In each case the value of the DMSO control was set as 1. (C) Bicistronic assays were performed using the two alternative mdm2 5′-UTRs. (D) Cryptic promoter activity was assayed in the promoter-less pGL3-Basic monocistronic background. Values for the VCME control were set at 1. (E) As a further control the bicistronic constructs were co-transfected with a plasmid expressing either the siRNA against RLuc (pBS/U6-RLi) or the empty vector control (pBS/U6ApaI). Reporter activity in the cell extracts was measured 24 h post-transfection. The graph plots reporter activities in the presence of the siRNA RLuc relative to the empty vector control. A value of 1 indicates no change. (F–H) Bicistronic clones carrying the EMCV, L-mdm2 or Dap5 intercistronic regions were expressed in HEK293T cells in the presence of either DMSO (CTRL), rapamycin (100 nM added to the growth medium 30 min before transfection), or a second vector expressing a His/myc tagged 4EBP1 (co-transfected with the bicistronic construct). Cell extracts were prepared 24 h post-transfection. Normalized reporter activities were plotted (F) with the control (bicistronic alone) for each construct being set at 100. Rapamycin activity was confirmed by immunoblotting with an anti-4EBP1 antibody (G), as was expression of the tagged 4EBP1 transgene (H). Bars indicate the SEM from triplicate transfections.

Figure 2.

Both mdm2 5′-UTRs confer relative rapamycin resistance to a monocistronic construct. (A) pGL3-Enhancer constructs carrying either the L-mdm2 or S-mdm2 5′-UTRs fused to FLuc were expressed in HEK293T cells in the presence of rapamycin (100 nM added to the growth medium 30 min before transfection) or a second vector expressing ^His/myc4EBP1. As an internal control, all cells were co-transfected with a pGL3-Enhancer vector that carried the 5′-UTR of β-actin fused to an RLuc reporter. Cells were harvested 24 h post-transfection. Normalized reporter activities were plotted with the CTRL (the two reporter constructs alone) set at 100 (upper panels). The effect of rapamycin and the expression of the 4EBP1 transgene were confirmed by immunoblotting with an anti-4EBP1 antibody. An immunoblot for actin provided a loading control (lower panels). (B) HEK293T cells were co-transfected with the S-mdm2 FLuc and β-actin RLuc reporter plasmids 30 min after addition of rapamycin at the concentrations indicated (0 indicates a DMSO control). Cells were harvested 24 h post-transfection and the normalized reporter activities plotted. The value for the rapamycin minus control was set at 100 (upper panel). The rapamycin response was confirmed by immunoblotting using an anti-4EBP1 antibody (lower panel). Bars indicate the SEM from triplicate transfections.

Figure 3.

The relative rapamycin resistance conferred by the mdm2 5′-UTRs responds to the over-expression of the downstream mTORC1 effectors. (A) The S-mdm2 FLuc and β-actin RLuc reporter plasmids were transfected into HEK293T cells with combinations of eIF4E, S6K1T389E and S6K1T389A as indicated in the figure. Rapamycin was applied to the cells 24 h post-transfection. A series of non-treated cells provided the control (RAP-ve). Cell extracts were prepared 24 h after drug treatment and reporter ratios were normalized to the control which was fixed at 100 (upper panel). S6K1 and eIF4E expression were confirmed by immunoblotting and an anti-actin blot provided a loading control (lower panels). (B) The assays outlined above were repeated with the L-mdm2 reporter and the constructs as indicated. The bars in both graphs indicate the SEM from triplicate transfections.

Figure 4.

The endogenous MDM2 protein levels. (A) HEK293T cells were transfected with a pEBS-PL vector expressing a GST-tagged La226-348 (a dominant negative deletion mutant) or an empty plasmid vector. At 48 h post-transfection cells were either treated with DMSO (CTRL) or rapamycin (100 nM). Extracts were prepared 24 h later and analysed by immunoblotting using the antibodies indicated. (B) Semi-quantitative RT–PCR was performed on cells that had been treated or non-treated with rapamycin (lanes 1 and 3 in A) to monitor changes in the levels of the two mdm2 5′-UTRs. Actin provided the internal control. (C) Pulse-chase experiments were performed in HEK293T cells to monitor MDM2 stability after 24 h exposure to rapamycin. DMSO provided the negative control. Radiolabelled cell extracts were immunoprecipitated with an anti-MDM2 antibody and immune complexes were recovered on protein-A sepharose beads. Proteins were resolved by SDS–PAGE. The times of the chase and the estimated protein half-life are indicated.

Figure 5.

What modulates the mdm2 response? (A) HEK293T cells were treated with either DMSO or rapamycin (100 nM) for 24 h in the presence of either 2% or 10% fetal calf serum. Changes in the intracellular levels of the indicated proteins were determined by immunoblotting. (B) The FLuc/RLuc dual reporter assay was performed as outlined in Figure 3. In experiments performed in 2% serum, cells were grown at this serum concentration overnight prior to addition of DMSO/rapamycin (100 nM) and transfection. In the serum-free experiment, cells were grown overnight in 10% serum before starving in serum-free medium for 8 h before DMSO/rapamycin addition and transfection. In both cases the cells were harvested 24 h post-rapamycin addition. Reporter activities were normalized with the DMSO control in each experiment being set at 100. The vertical arrow in the S-mdm2 panel indicates the normalized 2.5-fold increase observed in the presence of 2% serum and rapamycin. (C) Semi-quantitative RT–PCR was performed on total RNA (500 ng) isolated from cells transfected with the reporter plasmid sets in the presence and absence of rapamycin (indicated as + and –) under the serum concentrations indicated at the bottom of each panel. Products were resolved on a 6% polyacrylamide gel (for the L-mdm2 FLuc transcript), an 8% polyacrylamide gel (for the S-mdm2 FLuc transcript) and a 1% agarose gel (for the β-actin RLuc transcript). The values indicated in brackets represent the change in transcript levels observed as a consequence of rapamycin exposure. In each gel, the lane 1 is the RT minus control. (D) The non-normalized values for the S-mdm2 Fluc and β-actin Rluc reporters [right-hand side of (B)] are depicted graphically with the 2% serum, rapamycin negative value being set at 100. The X2.5-fold value reflects the difference in the normalized reporter values observed in the right-hand side of (B). This difference effectively vanishes when the drug is applied in the absence of serum (indicated as the lower hatched line). Note that both rapamycin and the removal of serum are negative for expression of both reporters. The right-hand panel is an immunoblot using the antibodies indicated. The bars in all graphs indicate the SEM from triplicate transfections.

Figure 6.

(A) HEK293T cells were transferred to DMEM containing the serum concentrations indicated at the top of each column 30 min prior to transfecting with vectors expressing either L-mdm2/Fluc, S-mdm2 Fluc or an L-mdm2/Fluc construct in which the two uAUGs were changed to GCG codons. As an internal control cells were co-transfected with the second Rluc reporter carrying the 5′-UTR of β-actin. Extracts were prepared 24 h later and the dual reporter activities measured. The FLuc/RLuc ratio obtained in the presence of 10% serum was then set at 100 and the values obtained at 2 and 0% serum were normalized relative to this. The upper panel is a series of immunoblots that follow eIF2α and phospho-eIF2α levels under the different serum concentrations employed in the assay. (B) The normalized reporter ratios were plotted for cells co-transfected with the β-actin RLuc and L-mdm2 uAUG1/2-ve FLuc reporter constructs in the absence (CTRL) and presence of rapamycin (100 nM) at the serum concentrations indicated at the bottom of each column set. Bars indicate the SEM from triplicate transfections.

Figure 7.

Polysomal re-recruitment assays in the presence of rapamycin. (A) HEK293T cells were hypertonically shocked to induce polysomal disaggregation. Polysomes were then allowed to reform under normal growth conditions in the absence (DMSO) or the presence of rapamycin (100 nM). The gradients were fractionated and RNA isolated from each fraction. The positions of the RNP, monosomal and polysomal regions of each gradient are indicated. RNA from each fraction was analysed on a denaturing agarose gel (lower panels). To have sufficient material for the subsequent RT–PCR analysis gradient fractions containing rRNA (this excluded the RNP fractions) were pooled to generate five aliquots (indicated as 1 to 5 below the agarose gel). (B) Real time RT–PCR was performed starting with 1 µg of total RNA from each aliquot with primer sets specific for L-mdm2, S-mdm2 and β-actin. The results were then plotted as a percentage of the total in each of the five aliquots. The right-hand panel is a table that indicates the percentage of each transcript found in heavy polysomes (≥5 ribosomes). The bars in the graphs indicate the SEM from triplicate assays.