Discussion
It is clear that any increase in IgG titre in medium is beneficial to the biopharmaceutical industry in terms of enhancing the production of the desired product and particularly its levels in the cell medium from which is it harvested. Here we show that Rheb mutations we previously identified are capable of increasing IgG titres in CHO cells grown in suspension in chemically-defined, serum-free medium replicating industrial growth conditions.
Consistent with our previous study (Jianling Xie et al., 2020), we found that the exogenous expression of the Rheb mutants Rheb[T23M], [Y35N] and [E40K] drive constitutive mTORC1 signalling through insensitivity to TSC-GAP activity and thus rendering mTORC1 insensitive to removal of upstream activators. We show that in HEK293 cells, Rheb[T23M] and [E40K] drive an increased rate of protein synthesis consistent with previous studies (Jianling Xie et al., 2020) as well as the known function of mTORC1 as a key regulator of protein synthesis (Proud, 2019) and ribosome biogenesis (Iadevaia et al., 2014). mTORC1 also promotes mitochondrial function (de la Cruz López, Toledo Guzmán, Sánchez, & García Carrancá, 2019). A recent paper has also shown the mTORC1 signaling can stimulate ATF4 expression independently of the UPR and thereby upregulate a subset of ATF4 target genes (Torrence et al., 2020). As a key activator of mTORC1 signalling, the expression of an active Rheb mutant such as T23M can drive multiple anabolic pathways, so that this single manipulation can lead to enhanced cell growth and proliferation and faster protein synthesis. This obviates the need to modify multiple genes thereby offering a substantial advantage for cell engineering.
We unexpectedly found that Rheb[Y35N], despite driving constitutive mTORC1 signalling, did not increase protein synthesis. This is consistent with our finding that, while Rheb[T23M and [E40K] increased the levels of ATF4, Rheb[Y35N] did not. Rheb[T23M] and [E40K] also resulted in an increase in markers of the integrated stress response and expansion of the endoplasmic reticulum (ER).
These data suggest that the increase in protein synthesis driven by constitutive activation of mTORC1 increases the protein folding load on the ER. As unfolded proteins accumulate in the ER, the integrated stress response becomes activated; in the short term, this drives increased expression of proteins required for protein folding and expansion of the ER. If this expansion is enough to accommodate the increased protein folding requirements, cells can maintain increased protein synthesis without progressing into apoptosis which can occur during prolonged activation of the UPR (Shaffer et al., 2004; Sriburi et al., 2004; M. Wang & Kaufman, 2016). This mechanism allows cells to produce more protein without adverse effects. The fact we only observe this phenomenon in cells expressing Rheb[T23M] and [E40K] but not Rheb[Y35N] is surprising and suggests that mTORC1 activation alone is not sufficient to support a long-term increase in protein synthesis. We previously showed through proteomic analysis of NIH3T3 cells stably expressing Rheb[T23M] and [E40K] that additional pathways are activated by these mutations. Cells expressing Rheb[T23M] showed increased reliance on anerobic glycolysis while cells expressing Rheb[E40K] showed increased autophagic flux (Jianling Xie et al., 2020). It is possible that activation of these pathways does not occur in Rheb[Y35N] so that cells expressing Rheb[Y35N] are not able to sustain the energy or nutrient requirements to sustain prolonged increases in protein synthesis.
We show that this mechanism of mTORC1-driven activation of the UPR is required for increased secretion of the reporter Gaussialuciferase from CHO cells stably expressing GLuc when grown in monolayer. This hypothesis is supported by the observation that both inhibition of mTOR and inhibition of the ATF4 arm of the UPR decreased GLuc secretion. We also observed a significant upregulation of ATF4 in cells expressing Rheb[T23M] or [E40K]. We show that this is also true for CHO cells grown in suspension in chemically-defined, serum-free medium. We found that stable expression of Rheb[T23M] or [E40K] resulted in around a 2-3 fold increase in IgG1 secretion compared to Rheb[WT] overexpression or cells expressing only their endogenous Rheb.
These results clearly point to a potential approach to significant increasing the efficiency of production of commercially valuable proteins. A further development of this approach would be to generate cell lines stably expressing Rheb[T23M] or with the chromosomal copies of the Rheb gene mutated appropriately. Since all cells in the population would then be expressing mutant Rheb which is not the case for transient transfection), an even greater augmentation of protein output would be expected. The ability to double product output over the same time-frame without increasing nutrient input promises to make production of currently expensive treatments more efficient and may therefore lead to significant reductions in the cost of treatment, thus allowing more people being able to benefit from these highly effective treatments and/or at a lower cost. Our findings may thus be of direct value to the burgeoning ‘biological drug’ sector.