Rheb-T23M and E40K drive mild endoplasmic reticulum stress through increased protein synthesis
In a recent study (Jianling Xie et al., 2020), we showed that certain mutations in Rheb (which arise in human cancers) are able to drive hyperactive mTORC1 signalling in mouse or human cells.
As mTORC1 is well known to promote multiple steps in mRNA translation (protein synthesis), we assessed whether these Rheb mutants can drive increased protein synthesis. To do this we employed surface sensing of translation western blot assay (SUnSET-WB (Schmidt, Clavarino, Ceppi, & Pierre, 2009)). In the SUnSET-WB technique cells are treated with a low concentration of puromycin (1 µM), a compound which acts as a structural analogue of aminoacyl-tRNA (specifically tyrosyl-tRNA) and can therefore be incorporated into nascent polypeptides. Incorporated puromycin can then be detected by western blot with puromycin-specific antibodies. Thus SUnSET-WB is a radioactive-free assay to measure rates of protein synthesis. For our initial experiments, we elected to utilise HEK293 cells as they are both a common cell line used for cell signalling studies, and ones in which we have previously shown Rheb mutations drive constitutive mTORC1 signalling. HEK293 cells transiently transfected with vectors encoding FLAG-Rheb[WT], [T23M], [Y35N] or [E40K] or an empty vector (EV) were transferred to Dulbecco’s phosphate buffered saline (D-PBS) for 30 min prior to the addition of 1 µM puromycin for an additional 30 min. One well of untransfected cells was pre-treated with 50 µg/mL of cycloheximide, a potent inhibitor of protein synthesis, for 30 minutes prior to the addition of puromycin to provide a ‘negative control’ for any immunostaining that is not due to ongoing protein synthesis. Cell lysates were then harvested for western blot analysis. Rheb[T23M] and [E40K] each stimulated a large increase in puromycin incorporation compared to either Rheb[WT] or EV (Fig. 1a; quantified in Fig. 1b). Interestingly, despite Rheb[Y35N] being known to drive hyperactive mTORC1 signalling, it did not increase puromycin incorporation, in line with the fact that Rheb mutants differ in their downstream effects (Jianling Xie et al., 2020).
It has been shown that increases in overall protein synthesis can overload the protein folding capacity of the endoplasmic reticulum (ER) resulting in ER stress and activation of the unfolded protein response (Sriburi et al., 2004). This process has been observed in response to increased mTORC1 activity (Appenzeller-Herzog & Hall, 2012; Dong et al., 2015) and results in an expansion of the ER and therefore increased protein folding capacity (Shaffer et al., 2004; Sriburi et al., 2004; M. Wang & Kaufman, 2016). We hypothesised that Rheb-mutants may drive mild ER stress resulting in increased ER volume and protein folding capacity. To test this, we first studied several proteins involved in both the ATF4 arm of the UPR and proteins involved in protein folding. Cells stably expressing plasmids encoding Rheb[WT] or mutants of Rheb showed increased expression of ATF4 compared to the empty vector (Fig. 1c). Interestingly, there was no significantly greater change in ATF4 protein expression in cells expressing Rheb mutants compared to WT. However, Rheb[T23M] and [E40K] did promote an increase, or tended to cause an increase, in classical UPR markers or ER resident proteins including PERK, BiP/Grp78, IRE-1α, PDI and ERO1-1α (Fig. 1c; quantified in Supplementary Figure 1). Calnexin did not change. There was also an increase in the protein folding markers ERO1-Lα (Fig. 1c; quantified in Supplementary Fig. S1).
To assess whether these changes reflected increased levels of the corresponding mRNAs, we performed RT-qPCR for the mRNAs encoding BiP (HSPA5 ; Fig. 1d), PDI (PDI ; Fig. 1e), IRE1α (ERN1 ; Fig. 1f), and ATF4 (ATF4 ; Fig. 1g) whose levels were increased by mutant Rheb expression. Increases in mRNA encoding both UPR and protein folding proteins correlated with protein increases with the notable exception of ATF4 mRNA which was significantly higher in cells expressing Rheb mutants compared to both the EV and Rheb[WT] (Fig. 1d-g). To determine if these observed changes are associated with an increase in ER volume, we performed immunofluorescence on HEK293 cells stably expressing Rheb mutants or WT as well as an empty vector. To image the ER, we chose to probe with an anti-calnexin antibody as calnexin is an ER surface protein and there was no significant change in the level of calnexin protein with the different Rheb mutants (Fig. 1b) (so that any alterations seen in the extent of the ER would be independent of changes in its overall levels). Volume was calculated based on a β-actin counter stain. Both Rheb[T23M] and [E40K] promote a significant increase in ER volume compared to Rheb[WT] and EV. Rheb[Y35N] did not promote an increase in ER volume (Fig. 2a; quantified Fig.2b). These data suggest the Rheb[T23M] and [E40K] mutants can each promote increase protein synthesis which, in turn, drives a mild ER stress resulting in increased ER volume and concomitant protein folding capacity.
These data for HEK293 cells prompted us to extend our studies to CHO cells, the dominant type of cells used for in industry for producing recombinant proteins, particularly as an expanded ER is reported to enhance the ability of CHO cells to produce secretory recombinant proteins (Budge et al., 2020).