Alkaline intracellular pH (pHi) increases PI3K activity to promote mTORC1 and mTORC2 signaling and function during growth factor limitation

The conserved protein kinase mTOR (mechanistic target of rapamycin) responds to diverse environmental cues to control cell metabolism and promote cell growth, proliferation, and survival as part of two multiprotein complexes, mTOR complex 1 (mTORC1) and mTORC2. Our prior work demonstrated that an alkaline intracellular pH (pHi) increases mTORC2 activity and cell survival in complete media in part by activating AMP-activated protein kinase, a kinase best known to sense energetic stress. It is important to note that an alkaline pHi represents an underappreciated hallmark of cancer cells that promotes their oncogenic behaviors. In addition, mechanisms that control mTORC1 and mTORC2 signaling and function remain incompletely defined, particularly in response to stress conditions. Here, we demonstrate that an alkaline pHi increases phosphatidylinositide 3-kinase (PI3K) activity to promote mTORC1 and mTORC2 signaling in the absence of serum growth factors. Alkaline pHi increases mTORC1 activity through PI3K–Akt signaling, which mediates inhibitory phosphorylation of the upstream proteins tuberous sclerosis complex 2 and proline-rich Akt substrate of 40 kDa and dissociates tuberous sclerosis complex from lysosomal membranes, thus enabling Rheb-mediated activation of mTORC1. Thus, alkaline pHi mimics growth factor–PI3K signaling. Functionally, we also demonstrate that an alkaline pHi increases cap-dependent protein synthesis through inhibitory phosphorylation of eIF4E binding protein 1 and suppresses apoptosis in a PI3K- and mTOR-dependent manner. We speculate that an alkaline pHi promotes a low basal level of cell metabolism (e.g., protein synthesis) that enables cancer cells within growing tumors to proliferate and survive despite limiting growth factors and nutrients, in part through elevated PI3K–mTORC1 and/or PI3K–mTORC2 signaling.

The conserved protein kinase mTOR (mechanistic target of rapamycin) responds to diverse environmental cues to control cell metabolism and promote cell growth, proliferation, and survival as part of two multiprotein complexes, mTOR complex 1 (mTORC1) and mTORC2.Our prior work demonstrated that an alkaline intracellular pH (pHi) increases mTORC2 activity and cell survival in complete media in part by activating AMPactivated protein kinase, a kinase best known to sense energetic stress.It is important to note that an alkaline pHi represents an underappreciated hallmark of cancer cells that promotes their oncogenic behaviors.In addition, mechanisms that control mTORC1 and mTORC2 signaling and function remain incompletely defined, particularly in response to stress conditions.Here, we demonstrate that an alkaline pHi increases phosphatidylinositide 3-kinase (PI3K) activity to promote mTORC1 and mTORC2 signaling in the absence of serum growth factors.Alkaline pHi increases mTORC1 activity through PI3K-Akt signaling, which mediates inhibitory phosphorylation of the upstream proteins tuberous sclerosis complex 2 and proline-rich Akt substrate of 40 kDa and dissociates tuberous sclerosis complex from lysosomal membranes, thus enabling Rheb-mediated activation of mTORC1.Thus, alkaline pHi mimics growth factor-PI3K signaling.Functionally, we also demonstrate that an alkaline pHi increases cap-dependent protein synthesis through inhibitory phosphorylation of eIF4E binding protein 1 and suppresses apoptosis in a PI3K-and mTOR-dependent manner.We speculate that an alkaline pHi promotes a low basal level of cell metabolism (e.g., protein synthesis) that enables cancer cells within growing tumors to proliferate and survive despite limiting growth factors and nutrients, in part through elevated PI3K-mTORC1 and/or PI3K-mTORC2 signaling.
Activation of mTORC1 but not mTORC2 by insulin-IGF requires sufficient levels of cellular amino acids (2,4).As such, mTORC1 functions as an amino acid sensor that fine-tunes metabolic activity in accordance with environmental cues (2,4).mTORC1 senses intracellular amino acids through several multisubunit complexes (e.g., the GATOR2, GATOR1, and Ragulator-LAMTOR complexes), which converge on heterodimeric Rag GTPases (Rag A or B bound to Rag C or D) tethered to lysosomal membranes.Amino acid stimulation of amino acid-starved cells increases GTP loading of RagA or RagB, which recruits mTORC1 to lysosomal membranes through its Raptor subunit (27,28).In simple terms, the insulin-PI3K-Akt pathway and the amino acid-Rag pathway signal in parallel to activate mTORC1 on lysosomal membranes.Emerging data indicate greater complexity in how growth factors and amino acids cooperate to regulate mTORC1 appropriately, however.For example, TSC localizes to lysosomes in amino acid-starved cells cultured with serum growth factors (i.e., fetal bovine serum [FBS]), and the readdition of amino acids induces the rapid dissociation of TSC from lysosomal membranes (29,30).Thus, TSC only fully dissociates from lysosomes in cells cultured with sufficient levels of growth factors and amino acids (29,30).In addition, growth factor signaling increases the mTORC1-mediated phosphorylation of RagC, which promotes mTORC1 signaling in an autoregulatory manner (31), and amino acids increase polyubiquitination of Rheb, which facilitates the mTORC1-Rheb interaction (32).Thus, crosstalk occurs between growth factor and amino acid signaling pathways.Various types of cell stress including hyperosmotic stress, energetic stress, hypoxia, and high pH (i.e., pH 9.4) also recruit TSC to lysosomes to inhibit mTORC1 (30).Thus, growth factor, amino acid, and cell stress signaling converge on TSC to control mTORC1 activity through Rheb (29,30).Taken together, these findings indicate that complex pathways and mechanisms work together to fine-tune mTORC1 activity.
Cap-dependent protein synthesis represents a fundamental cellular process promoted by mTORC1 (33,34).mTORC1 mediates the sequential phosphorylation of the translational inhibitor 4EBP1 on multiple sites (first at T37 and T46 and subsequently at T70 and S65) (33)(34)(35).In the hypophosphorylated state, 4EBP1 binds to and inhibits eIF4E (eukaryotic initiation factor 4E), a rate-limiting translation initiation factor that binds to the cap structure on the 5 0 ends of mRNA transcripts.mTORC1-mediated phosphorylation of 4EBP1 causes its dissociation from eIF4E, enabling the binding of eIF4G and the recruitment of other translation initiation factors to assemble the eIF4F translation preinitiation complex (33,34).mTORC1 signaling to S6K1 also promotes protein synthesis by upregulating rRNA transcription (aka, ribosome biogenesis) and through the phosphorylation of translation factors (e.g., eIF4B, PDCD4) and likely other translationrelated factors (e.g., the ribosomal protein S6) (33,34).
Cell survival and metabolism represent fundamental cellular processes controlled by mTORC2 (5,11,12).Our prior work demonstrated that AMPK (AMP-activated protein kinase), a kinase best known to sense and respond to low levels of cellular energy (i.e., elevated AMP or ADP), phosphorylates mTOR and Rictor to activate mTORC2 directly in response to energetic stress, which suppresses apoptosis (36).This work not only identified energetic stress as a previously unknown activator of mTORC2 but provided a molecular explanation for the AMPK-cancer conundrum in which AMPK functions as either a tumor suppressor (at early stages of oncogenesis through suppression of mTORC1) or a tumor promoter (at later stages of oncogenesis through poorly defined mechanisms) (36)(37)(38)(39)(40).We proposed that the AMPK-mTORC2 axis may promote tumorigenesis by increasing the survival of cancer cells experiencing energetic stress within the core of growing tumors starved of essential factors (e.g., growth factors, nutrients, oxygen).Our subsequent work demonstrated that another type of cell stress, that is, elevated intracellular pH (pHi), increases AMPK activity to promote mTORC2 signaling in cooperation with an unknown input to suppress apoptosis during growth factor limitation (41).Interestingly, tumor tissue often displays a reversal of the pH gradient whereby cancer cells possess a slightly alkaline pHi within an acidic extracellular space, the reverse of normal tissue (42)(43)(44)(45).Moreover, we now appreciate that elevated pHi promotes diverse cancer cell behaviors, including metabolic adaptation, cell proliferation, evasion from apoptosis, and cell migration (42)(43)(44)(45)(46).
Here, we identify pHi as a previously unknown regulator of PI3K.We found that alkaline pHi increases PI3K activity, which promotes PI3K-dependent mTORC1 and mTORC2 signaling, cap-dependent translation, and cell survival during growth factor limitation.Thus, alkaline pHi mimics the growth factor-PI3K pathway.This study also identifies PI3K as the potential elusive input that increases mTORC2 signaling in response to alkaline pHi in cooperation with AMPK from our previous work (41).Taken together, these results suggest that elevated PI3K-mTORC1 and/or PI3K-mTORC2 signaling may drive oncogenic cell behaviors mediated by increased pHi in cancer cells, possibly by promoting the growth and survival of cancer cells within growing tumors deprived of essential growth factors.

Alkaline extracellular pH (pHe) increases mTORC1 and mTORC2 signaling in the absence of serum growth factors
While investigating the activation of the AMPK-mTORC2 axis by alkaline pHi in our prior work (41), we noted that refeeding mouse embryonic fibroblasts (MEFs) cultured in complete media (CM; i.e., Dulbecco's modified Eagle's medium [DMEM] containing FBS and amino acids) at an alkaline pH 8.3 increased the phosphorylation of Akt S473 but not S6K1 T389 in a manner sensitive to Torin1 (an ATP-competitive mTOR inhibitor) (41) (Fig. 1A).Thus, an alkaline pHe is sufficient to increase mTORC2 but not mTORC1 signaling in the presence of growth factors.It is important to note that Demetriades et al. (30) also found that raising the pH of CM (i.e., to pH 8.4) had no effect on mTORC1 signaling to S6K1 in MEFs.To better understand this finding, we found that inhibition of class I PI3Kα with the drug BYL719 (aka, alpelisib) blocked the increase in Akt P-S473 mediated by alkaline pHe in CM (Fig. 1A).As growth factors represent a strong activating input to mTORC1, we reasoned that we should analyze serum-starved MEFs in order to probe a potential role for alkaline pH in the regulation of mTORC1.Refeeding serumstarved immortalized (Fig. 1B) or primary MEFs (Fig. 1C) with serum-free DMEM at pH 8.3 for 5 to 60 min increased the phosphorylation of several mTORC1 substrates, that is, S6K1-T389, 4EBP1-S65, and ULK1-S757, relative to MEFs refed with serum-free DMEM at physiological pH 7.4.Importantly, Torin1 ablated the pHi-induced phosphorylation of these targets (Fig. 1, B and C).These results demonstrate that an alkaline pHe is sufficient to increase mTORC1 signaling in the absence of serum growth factors.In these experiments, we noted that serum-free media (SFM) at alkaline pH also increased the phosphorylation of Akt T308 (a direct target of PI3K-dependent PDK1) and Akt S473 (a direct target of mTORC2) (Fig. 1, A-C).Consistent with mTORC2mediated Akt S473 phosphorylation stabilizing Akt T308 phosphorylation (23,36,47), media at alkaline pH increased Akt P-T308 in a Torin1-sensitive manner (Fig. 1, A-C).We also found that refeeding serum-starved MEFs with media adjusted to increasingly basic or acidic pH values increased (Fig. 1D) or decreased (Fig. 1E) mTORC1 and mTORC2 signaling, respectively.Finally, we analyzed how SFM at alkaline pH controls mTORC1 and mTORC2 signaling in other serum-starved cell lines, including C2C12, U2OS, HeLa, and human embryonic kidney 293 (HEK293) cells (Fig. S1, A-D).In C2C12 and U2OS cells, SFM at an alkaline pH increased both mTORC1 (S6K1 P-T389) and mTORC2 (Akt P-S473) signaling (Fig. S1, A and B).In serum-starved HEK293 cells, SFM at an alkaline pH increased mTORC1 signaling, albeit modestly, and mTORC2 signaling (Fig. S1C), whereas in serum-starved HeLa cells, SFM at an alkaline pH increased mTORC1 but not mTORC2 signaling (Fig. S1D).Taken together, these results indicate that media at an alkaline pH increases mTORC1 and/or mTORC2 signaling in several cell lines, including primary MEFs, during growth factor limitation, albeit to different degrees for reasons unclear at this time.

Alkaline pHi upregulates mTORC1 and mTORC2 signaling in the absence of serum growth factors
Prior work by other groups and us demonstrated that refeeding cells with serum-containing CM at an alkaline pH raises pHi rapidly (41,48,49).To confirm that culturing serum-starved MEFs in SFM at an alkaline pH also raises pHi rapidly, we used live-cell imaging coupled with a cellpermeable and pH-sensitive fluorescent dye (cSNARF1-AM), which monitors changes in pHi between 7 and 8. cSNARF1-AM undergoes a pH-sensitive shift in fluorescence wavelength emission depending on its protonation state.With 488 nm excitation, peak emission occurs at 580 nm in the protonated state (more acidic conditions) and 640 nm in the deprotonated state (more alkaline conditions).Thus, ratiometric imaging (640:580 nm) enables the detection of changes in pHi within the physiological range of normal cells (pH 7.05-7.2) and cancer cells (pH 7.3-7.6)(42)(43)(44)(45).Serumstarved MEFs were loaded with cSNARF1-AM, washed, refed with SFM at pH 7.4 or pH 8.3, and imaged at 0, 0.67 (40 s), 5, and 10 min.Visualization of the acquired pseudocolored ratiometric images (640:580 nm) and quantitation of the signal ratios revealed a rapid and significant increase in pHi upon shift to serum-free DMEM at alkaline pH (green cells) relative to those maintained in media at physiological pH (blue cells) (Fig. 2, A and B).These results demonstrate that shifting cells to SFM at alkaline pH leads to a rapid rise in pHi.
As an alternative method to raising pHi, we used ammonium chloride (NH 4 Cl).NH 4 Cl increases cytosolic pH upon rapid diffusion of NH 3 (a weak base) into the cell and conversion to ammonium ion (NH 4 + ) by scavenging protons from the cytosol, thus increasing cytosolic pH (50,51).Slower and subsequent entry of NH 4 + into the cell attenuates the rise in pH upon its dissociation into NH 3 and H + .As NH 4 Cl raises the lumenal pH of organelles such as lysosomes, it is often employed in autophagy research to impair autophagic flux/ proteolysis, which requires an acidic pH in autolysosomes (50,51).The addition of NH 4 Cl to SFM at pH 7.4 induced a rapid but transient rise in pHi (Fig. 2, C and D), as measured microscopically with cSNARF1-AM.Biochemically, treatment of serum-starved MEFs with NH 4 Cl increased mTORC1 signaling weakly (S6K1 P-T389) but increased PI3K (Akt P-T308) and mTORC2 (Akt P-S473) signaling strongly (Fig. 2E).
We next sought to confirm that culturing cells in SFM at an alkaline pH increases mTORC1 and mTORC2 signaling by raising pHi.To do so, we used the drug cariporide (a morespecific derivative of amiloride) to reduce pHi pharmacologically.Cariporide inhibits NHE1 (sodium-hydrogen exchanger isoform 1), a sodium-proton (Na + /H + ) antiporter on the plasma membrane that pumps hydrogen ions (H + ) from the cytosol to the extracellular space.Thus, cariporide decreases pHi by blunting H + efflux (52)(53)(54).We, therefore, pretreated serum-starved MEFs with cariporide before refeeding cells with media at alkaline pH in the absence or the presence of drug.Cariporide blunted the ability of media at pH 8.3 to increase the phosphorylation of S6K1 T389, Akt T308, and Akt S473 (Fig. 2F).These results indicate that attenuating H + efflux from the cell suppresses mTORC1 and mTORC2 signaling likely by acidifying the cytosol, thus countering the alkalizing effect of media at pH 8.3.Note that NHEI represents one of many hydrogen ion transporters in cells, thus explaining why cariporide only partly blunts the activation of mTORC1 signaling in response to alkaline pHi.Taken together, these results indicate that altering pHi by three independent methods (treating cells with media at alkaline pH, NH 4 Cl, or cariporide) modulates mTORC1 and mTORC2 signaling in the absence of serum growth factors, with alkaline pHi or acidic pHi increasing or decreasing mTORC1/2 signaling, respectively.
Alkaline pHi mimics growth factor signaling to activate mTORC1 in parallel to amino acid signaling Growth factors and amino acids activate mTORC1 cooperatively through parallel pathways, although some crosstalk occurs.We, thus, next investigated the relationship between mTORC1 signaling induced by alkaline pHi versus insulin or amino acids.Treatment of serum-starved MEFs with SFM at pH 8.3 or with a maximal dose of insulin (100 nM) each increased S6K1 P-T389, whereas the addition of both together failed to increase S6K1 P-T389 further (Fig. 3A).When we stimulated cells with a submaximal dose of insulin (5 nM), however, alkaline pHi and insulin increased S6K1 P-T389 in an additive manner (Fig. 3B).Taken together, these data are consistent with the idea that alkaline pHi mimics growth factor signaling.As insulin and growth factors were shown several decades ago to increase pHi in fibroblasts and frog skeletal muscle by activating NHEI (55-61), we next asked whether blunting H + extrusion from the cell through inhibition of NHEI with cariporide would blunt the activating effect of insulin on mTORC1 and mTORC2 signaling.We found that cariporide pretreatment blunted the ability of insulin to increase mTORC1 signaling, albeit modestly.Cariporide had minimal effect on mTORC2 signaling, however (Fig. 3C).
We next investigated the relationship between mTORC1 signaling induced by alkaline pHi versus amino acids.We first tested how the addition of the commercial amino acid solution adjusted to pH 7.4 increases mTORC1 signaling relative to the addition of the original, unadjusted pH 10 commercial amino acid solution.As expected, after starving MEFs of both growth factors and amino acids, amino acids at pH 7.4 increased S6K1 P-T389 but not Akt P-S473; amino acids at pH 10 increased S6K1 P-S6K1 much more strongly, however, and also increased Akt P-S473 (Fig. 3D).These results suggest that alkaline pHi and amino acid signaling increase mTORC1 signaling additively by parallel pathways.Moreover, they demonstrate that alkaline pHi but not amino acids increase mTORC2 signaling.To analyze the effects of amino acids versus an alkaline pHi on mTORC1 signaling more carefully, we tested their ability alone and together to increase mTORC1 signaling.Refeeding serum-and amino acidstarved MEFs with media at pH 8.3 in the presence of amino acids increased mTORC1 signaling to a greater extent than either stimulus alone (Fig. 3E) (lane 4 versus lanes 2 and 3).These results suggest that amino acids and alkaline pHi signal in parallel to promote mTORC1 signaling.Interestingly, amino acid stimulation attenuated the ability of an alkaline pHi to increase in Akt P-T308 and P-S473 (lane 4) (Fig. 3E).Perhaps the well-established ability of mTORC1 to mediate negative feedback on PI3K flux (62-64) accounts for this effect, although 10 min would be quite fast for such a mechanism.Taken together, these results indicate that alkaline pHi mimics growth factor-PI3K signaling and therefore cooperates with amino acid signaling to fully activate mTORC1.In addition, they underscore the importance of controlling the pHe in cell cultures, as the results in Figure 3D could be interpreted to mean, incorrectly, that amino acids increase mTORC2 signaling.Failure to control the pH of commercial amino acid solutions likely explains why some studies concluded that amino acids activate mTORC2 in certain contexts (65,66).

mTORC1 and mTORC2 signaling induced by alkaline pHi requires PI3K activity
To ask whether the ability of alkaline pHi to increase mTORC1 signaling requires PI3K, we treated serum-starved MEFs with BYL719 prior to incubation in SFM at pH 8.3.BYL719 blocked the ability of alkaline pHi to increase the phosphorylation of S6K1-T389, 4EBP1-S65, and Akt-S473 (Fig. 4A).BYL719 also blocked the PDK1-dependent phosphorylation of the activation loop sites on S6K1 (P-T229) and Akt (P-T308) as well as the phosphorylation of the Akt substrates TSC2-S939 and proline-rich Akt substrate of 40 kDa (PRAS40)-T246 (Fig. 4A).The Akt inhibitor MK2206 blocked the ability of alkaline pHi to increase the phosphorylation of S6K1-T389, 4EBP1-S65, Akt-T308, TSC2-S939, and PRAS40- Alkaline intracellular pH increases PI3K-mTORC1/2 signaling T246 (Fig. 4B), thus phenocopying the effect of PI3K inhibition.Collectively, these results indicate that PI3K-Akt signaling is required for the activation of mTORC1 and mTORC2 signaling by alkaline pHi in the absence of serum growth factors.
Phosphorylation of Akt S473 by mTORC2 often increases Akt T308 phosphorylation, possibly by recruiting PDK1 and/ or stabilizing Akt T308 phosphorylation (23,47), whereas cotranslational phosphorylation of Akt T450 by mTORC2 increases Akt stability (67,68).As expected, Torin1 not only ablated Akt P-S473 in cells cultured in media at both pH 7.4 and pH 8.3 but also reduced alkaline pHi-induced Akt P-T308 as well as PRAS40 P-T246, an Akt substrate (Figs.4, A-C and  1, A and B).It was therefore important to determine whether the ability of alkaline pHi to increase mTORC1 signaling occurs independently of the activating effect of alkaline pHi on mTORC2 signaling or not.We thus studied Rictor −/− MEFs stably expressing either vector control or rescued with hemagglutinin (HA)-tagged Rictor.Restoring mTORC2 function boosted S6K1 P-T389 overall in rescued Rictor −/− MEFs refed with SFM at either pH 7.4 or 8.3 (Fig. 4C) (likely because of the combined effect of elevated Akt P-T308 and total Akt protein levels).In Rictor −/− MEFs expressing vector control (V) and lacking mTORC2-dependent Akt P-S473, alkaline pHi still increased PI3K signaling to Akt (P-T308), TSC2 (P-S939), and PRAS40 (P-T246) as well as mTORC1 signaling to S6K1 (P-T389) (Fig. 4C).In this context, Torin1 had no inhibitory effect on Akt P-T308 or TSC2 P-S939 (Fig. 4C).In Rictor −/− (V) MEFs, BYL719 blocked the ability of media at pH 8.3 to increase S6K1 P-T389 and Akt P-T308 (Fig. 4D).Collectively, these results demonstrate that alkaline pHi increases the PI3Kdependent activation of Akt and mTORC1 in the absence of mTORC2 function.
As driver mutations in PIK3Cα, which encodes the catalytic p110 subunit of class I PI3Kα, promote the growth of many cancers including breast cancer because of elevated PI3K activity, we tested whether breast cancer MCF7 cells bearing an oncogenic mutation in PIK3CA (E545K) would be resistant to the ability of an alkaline pHi to increase mTORC1 and/or mTORC2 signaling.As expected, serum-starved MCF7 cells displayed an easily detectable level of BYL719-sensitive mTORC1 (S6K1 P-T389) and mTORC2 (Akt P-S473) signaling because of elevated PI3K activity.In this context, an alkaline pHi failed to increase mTORC1 or mTORC2 signaling (Fig. S2).These results support the idea that an alkaline pHi mimics PI3K signaling further.

Alkaline pHi dissociates TSC2 from lysosomal membranes and increases mTORC1 catalytic activity by increasing PI3K activity
We next sought to better define the mechanism by which alkaline pHi increases mTORC1 signaling.As activation of Akt by PI3K dissociates the TSC complex from lysosomal membranes through Akt-mediated inhibitory phosphorylation of TSC2 (15), we asked whether alkaline pHi is sufficient to dissociate TSC2 from lysosomal membranes in the absence of growth factors.Refeeding serum-starved MEFs with media at pH 8.3 decreased colocalization of TSC2 with lysosomalassociated membrane protein 1 (LAMP1), a lysosomal membrane marker, in a BYL719-sensitive manner (Fig. 5, A and B).As expected, insulin stimulation of serum-starved MEFs at physiological pH reduced the colocalization of TSC2 with LAMP1, as reported previously (Fig. 5, A and B) (15).Thus, alkaline pHi is sufficient to dissociate TSC2 from lysosomal membranes in the absence of serum growth factors.We next investigated whether alkaline pHi increases the intrinsic catalytic activity of mTORC1.To do so, we monitored levels of Raptor-associated mTOR S2481 autophosphorylation, a simple method to monitor the catalytic activity of mTORC1 in intact cells (69).Refeeding MEFs with media at pH 8.3 increased P-S2481 on Raptor-associated mTOR, which occurred in a BYL719-and Torin1-sensitive manner (Fig. 5B).As the ability of alkaline pHi to dissociate TSC2 from lysosomes and increase mTORC1 activity and signaling was PI3K dependent, we next asked whether alkaline pHi increases PI3K activity.We refed serum-starved MEFs with SFM at pH 7.4 or pH 8.3 and measured the levels of PIP 3 in purified lipids by an ELISA.As a positive control, we stimulated cells in media at pH 7.4 with insulin.Media at pH 8.3 increased PIP 3 levels relative to media at pH 7.4 in a statistically significant and BYL719-sensitive manner (Fig. 5, C and D).As expected, insulin increased PIP 3 levels.These results indicate that alkaline pHi is sufficient to increase PI3K activity in the absence of growth factors.Taken together, they reveal that by increasing PI3K activity and Akt signaling, alkaline pHi dissociates TSC2 from the lysosomal surface, thus enabling Rheb to increase mTORC1 catalytic activity.

Alkaline pHi increases cap-dependent protein synthesis and suppresses apoptosis in a PI3K-and mTOR-dependent manner in the absence of serum growth factors
As protein synthesis represents a major cellular function promoted by mTORC1 (33,34), we next asked whether alkaline pHi is sufficient to increase cap-dependent protein synthesis in the absence of serum growth factors through PI3K and mTOR.We first pulled down eIF4E using m 7 GTP-coupled Sepharose beads to assess the ability of alkaline pHi to dissociate 4EBP1 from eIF4E through mTORC1-mediated 4EBP1 phosphorylation.After serum starvation, refeeding MEFs with serum-free DMEM at pH 8.3 was sufficient to increase 4EBP1 phosphorylation and dissociate 4EBP1 from eIF4E, which occurred in a Torin1-sensitive manner (Fig. 6A).Pretreating cells with cariporide to suppress H + efflux and thus reduce pHi blunted the ability of alkaline pHi to dissociate 4EBP1 from eIF4E.As the pH of DMEM at pH 8.3 falls quickly because of the strong buffering capacity of DMEM incubated in an atmosphere of 7.5% CO 2 , we incubated cells at low CO 2 (i.e., 0.1%) to maintain an elevated pH for a longer period.This modification to our method for raising pHi enabled us to study protein synthesis (see later), a process requiring a longer amount of time.In cells cultured at low CO 2 , alkaline pHi dissociated 4EBP1 from eIF4E and increased the association of eIF4G with eIF4E in a statistically significant and PI3K-and mTOR-dependent manner (Fig. 6, B and C).As in other experiments, alkaline pHi increased the PI3K-and mTORdependent phosphorylation of 4EBP1 and S6K1 in cells cultured at low CO 2 in this context (Fig. 6B).These results indicate that elevated pHi inhibits 4EBP1 function and promotes the assembly of translation preinitiation complexes.
We next asked whether increased assembly of eIF4E-based translation preinitiation complexes leads to increased protein synthesis.To do so, we employed the nonradioactive SUnSET assay, which measures the incorporation of the tRNA-mimetic compound puromycin into nascent polypeptide chains by simple Western blotting using an antipuromycin antibody (70).Upon refeeding serum-starved MEFs incubated at either normal CO 2 (i.e., 7.5%) or low CO 2 (i.e., 0.1%) with serum-free DMEM at pH 7.4 or pH 8.3 for 60 min that lacked or contained a pulse of puromycin for the last 10 min, we found that alkaline pHi was sufficient to increase protein synthesis in a statistically significant manner at low CO 2 , which was PI3K and mTOR dependent (Fig. 6, D and E).As expected, stimulation of serum-starved cells cultured at physiological pH 7.4 with insulin also increased protein synthesis (Fig. 6, D and E).Consistently, alkaline pHi increased the PI3K-and mTORdependent phosphorylation of 4EBP1 and S6K1 T389 in this experimental context.
Cell survival represents an important cellular function mediated by mTORC2 and Akt (5,12).Our previous work demonstrated that an AMPK-mTORC2 signaling axis suppresses apoptosis in response to energetic stress (36) or alkaline pHi (41).We, therefore, tested whether activation of PI3K-mTOR signaling by alkaline pHi promotes cell survival during growth factor deprivation.We incubated MEFs in either CM (i.e., DMEM containing FBS) (lane 1) or in SFM (i.e., DMEM lacking FBS) (lane 2) overnight (16 h) and then refed the cells for 1 h with either CM (lane 1) or SFM at pH 7.4 or pH 8.3 (lanes 2-7).As expected, cells cultured in SFM at pH 7.4 for the full 17 h displayed increased apoptosis (lane 2) relative to those kept in CM for 17 h (lane 1), as monitored by Western blotting of cleaved caspase 3 (cCasp3) and cleaved poly(ADP-ribose) polymerase (cPARP), well-established apoptotic markers (Fig. 7, A and B).Culturing MEFs in SFM at pH 8.3 and 7.5% CO 2 reduced levels of cCasp3 and cPARP slightly (lane 3).Culturing MEFs in SFM at pH 8.3 and 0.1% CO 2 , however, reduced the levels of cCasp3 and cPARP significantly (lane 4) relative to physiological pH 7.4 (lane 2) and did so in a BYL719-and Torin1-sensitive manner (lanes 5 and 7) (Fig. 7, A and B).While the mTORC1-specific inhibitor rapamycin blunted the ability of alkaline pHi to suppress apoptosis modestly (lane 6), the effect was not statistically significant (Fig. 7B).Consistent with other experiments, alkaline pHi increased PI3K signaling (Akt P-T308), mTORC1 signaling (S6K1 P-T389), and mTORC2 signaling (Akt P-S473) in a PI3K-and mTOR-dependent manner in this experimental context (Fig. 7A).Collectively, these results indicate that alkaline pHi protects against apoptosis during growth factor limitation by increasing PI3K-mTOR signaling.
higher overall levels of cCasp3 and cPARP than rescued Rictor −/− (HA-Rictor) MEFs (Fig. 7C) (lanes 2 versus 8).As expected, levels of Akt P-S473 were extremely low in Rictor −/− (V) MEFs, and reconstitution with HA-Rictor rescued Akt S473 phosphorylation and boosted Akt T308 phosphorylation with little effect on S6K1 T389 phosphorylation (Fig. 7C).In Rictor −/− (V) MEFs cultured in SFM at low CO 2 (0.1%), alkaline pHi 8.3 increased S6K1 P-T389 and reduced levels of cCasp3 and cPARP relative to cells cultured at physiological pHi 7.4 (lane 4 versus lane 2).These effects were both rapamycin and Torin1 sensitive (lanes 5 and 6 versus lane 4), with Torin1 having a stronger effect than rapamycin (Fig. 7C).These results suggest that in the absence of mTORC2 function, increased mTORC1 signaling in growth factor-starved cells contributes to the evasion of apoptosis mediated by alkaline pHi.Collectively, we propose that activation of both mTORC1 and mTORC2 by alkaline pH promotes cell survival during growth factor withdrawal, with mTORC2 playing a more major role in cell survival than mTORC1.

Discussion
Normal cells and those associated with various disease states such as cancer display changes in pHi dynamics (42)(43)(44)(45).Normal cells maintain pHi at 7.05 to 7.2 through regulation of hydrogen ion (H + ) dynamics at the plasma membrane and possibly internal membranes (43,44).Original thought presumed that rapidly proliferating cancer cells possess an acidic cytosol because of the Warburg effect and production of lactic acid generated by aerobic glycolysis.More recent work reveals a reversal of the pH gradient in many cancer cells because of upregulated efflux of lactate and protons across the plasma membrane, however, resulting in a slightly alkaline pHi (7.3-7.6) and a slightly acidic pHe (6.8-7.0), the reverse of normal cells (42)(43)(44)(45)(46).These changes in pHi and pHe because of altered proton (H + ) flux drive diverse oncogenic cell behaviors, including cell proliferation, evasion from apoptosis, and cell migration through pathways and mechanisms that remain incompletely defined (42)(43)(44)(45)(46).
This study demonstrates that an alkaline pHi mimics growth factor-PI3K signaling to mTORC1 and mTORC2, pathways hyperactivated in many human tumors that contribute to their aberrant growth, proliferation, and survival (4, 71).We found that an elevated pHi is sufficient to increase mTORC1 and mTORC2 signaling in a PI3K-dependent manner in the absence of serum growth factors (Fig. 7D).Mechanistically, alkaline pHi increases PI3K activity and signaling to Akt, which leads to inhibitory phosphorylation of TSC2 and PRAS40 and dissociation of TSC from lysosomal membranes.These events enable Rheb to activate mTORC1 during growth factor limitation. Activation of PI3K by pHi explains why the activation of mTORC2 by alkaline pHi was only partially dependent on AMPK, as reported in our prior work (41).By mimicking PI3K action, alkaline pHi cooperates in parallel with amino acids to activate mTORC1 and cooperates in parallel with AMPK to activate mTORC2 (Fig. 7D).Functionally, alkaline pHi increases initiation of cap-dependent translation in a PI3K-and mTOR-dependent manner in the absence of growth factors through inhibitory phosphorylation of 4EBP1, thus enabling assembly of the eIF4F translation preinitiation complex (Fig. 7D).Alkaline pHi also protects against apoptosis caused by the withdrawal of growth factors in a PI3K-and mTOR-dependent manner.
In addition to finding that an alkaline pHi increases mTORC1 signaling, we found that an acidic pHi suppresses mTORC1 signaling, which agrees with the work of other groups (72)(73)(74).Balgi et al. (72) found that acidic pHi reduces mTORC1 signaling in a TSC-dependent manner, resulting in reduced protein synthesis.Walton et al. (73) found that acidic pHi induces a lysosomal pool of mTORC1 to move to a more peripheral subcellular localization, away from Rheb, thus limiting mTORC1 activity.Erra Diaz et al. (74) found that acidic pHi reduces mTORC1 signaling to direct the differentiation of dendritic cells into monocytes rather than macrophages.Physiologically, high-intensity exercise leads to rapid acidification of the cytosol in skeletal muscle, whereas pathologically, hypoxia/ischemia acidifies the cytosol.Downregulated mTORC1 signaling in both instances would ensure that cells within affected tissues do not engage in energetically costly anabolic metabolism.
Cancer cells often bear an alkaline pHi through the upregulated expression of transporters that actively extrude cytosolic acid (42)(43)(44)(45).A recent study demonstrated that increased expression of the sodium-proton antiporter NHEI or the lactate-proton symporter MCT4, both of which extrude H + from the cytosol, increased the pHi of hematopoietic cells, which promoted their growth and proliferation in culture and in vivo (46).We speculate that elevated mTORC1 and/or mTORC2 signaling may have contributed to the hyperproliferative phenotype of the hematopoietic cells in this study, as we found that inhibition of NHEI with cariporide reduces mTORC1 signaling (this study) and mTORC2 signaling (our prior work) (41).Interestingly, Man et al. (46) found that elevated pHi mediated by overexpression of H + transporters also increased the activity of several metabolic gatekeeper enzymes (e.g., hexokinase 1, pyruvate kinase M2, glucose 6 phosphate dehydrogenase) that promote glycolytic and pentose phosphate carbon flux.Thus, an alkaline pHi appears to influence metabolic adaptation as well as classical oncogenic cell behaviors.
Cancer cells also upregulate glutaminolysis as part of a metabolic reprogramming response that not only provides fuel for growth and proliferation but also liberates ammonium (NH 4 + ) as a waste byproduct.Merhi et al. (75) found that delivering ammonium to cells as either NH 4 Cl or NH 4 OH increased mTORC1 signaling weakly (4EBP1 P-T37/46) but increased PI3K (PRAS40 P-T246) and mTORC2 (Akt P-S473) signaling strongly.In addition, NH 4 Cl increased the proliferation of cells deprived of glutamine (75).Interestingly, we used NH 4 Cl as an alternative method of increasing cytosolic pH here and in our prior work (41).Similar to Merhi et al. (75), we found that NH 4 Cl increases PI3K and mTORC2 signaling strongly but mTORC1 signaling modestly, likely by increasing pHi.
Evasion of apoptosis represents an established hallmark of cancer cells (76).We found that increased pHi attenuated the induction of apoptosis caused by growth factor withdrawal.The ability of alkaline pHi to protect against apoptosis was blunted in a statistically significant manner in cells treated with the PI3K inhibitor BYL719 and the mTOR inhibitor Torin1.These data suggest that an alkaline pHi in cancer cells may contribute to tumorigenesis through increased cell survival mediated by PI3K-mTOR signaling.While the allosteric mTORC1-specific inhibitor rapamycin blunted the evasion of apoptosis caused by alkaline pHi modestly during growth factor withdrawal, the effect was not statistically significant.In Rictor knockout MEFs lacking mTORC2 function, however, alkaline pHi and low CO 2 (i.e., 0.1%) increased S6K1 phosphorylation and attenuated apoptosis concomitantly in a rapamycin-sensitive manner relative to physiological pH and normal CO 2 (i.e., 7.5%) during growth factor withdrawal.It is important to note that as some mTORC1 substrates display only partial sensitivity to rapamycin (e.g., 4EBP1) (77,78), a rapamycin-insensitive mTORC1 substrate may contribute to the attenuation of apoptosis.While limited reports have documented a role for mTORC1 in promoting cell survival, knockout of TSC2 and activation of mTORC1 in a mouse model of lymphoma accelerated oncogenesis, suppressed tumor apoptosis, and decreased animal survival in vivo (79).Treatment of these mice with rapamycin increased apoptosis in the lymphomas and prolonged animal survival (79), suggesting that mTORC1 may promote tumor cell survival in certain oncogenic contexts.Taken together, our results suggest that while mTORC1 may contribute to the evasion of apoptosis by alkaline pHi during growth factor limitation, mTORC2 likely represents the major mTOR complex that suppresses apoptosis in response to alkaline pHi.
It will be important in the future to determine how an alkaline pHi increases PI3K activity, leading to elevated mTORC1 and mTORC2 signaling.Recent research indicates that changes in pHi alter the structure and function of pHsensitive proteins (aka, pH sensors) through dynamic protonation and deprotonation, a post-translational modification akin to phosphorylation, ubiquitination, acetylation, and so forth (45).In fact, recurring charge-changing mutations (e.g., Arg to His) have been documented in the epidermal growth factor receptor, the tumor suppressor protein p53, Ras-GRP1, and β-catenin, which alters their activities in cancer cells bearing an alkaline pHi and contributes to the transformed behaviors of the cells bearing these mutations (44,(80)(81)(82).Theoretically, altered protonation of PI3K subunits, either the catalytic p110 subunit or the regulatory p85 subunit, could increase PI3K activity.In addition, certain recurring mutations in PI3K subunits that drive oncogenesis could render PI3K activity more sensitive to the increased pHi of cancer cells, such as occurs in documented pH sensors (e.g., epidermal growth factor receptor, p53, β-catenin) bearing chargechanging Arg to His mutations (44,80,81).Altered protonation mediated by alkaline pHi may also act upstream of PI3K, possibly at the level of the lipid phosphatase PTEN, growth factor receptors, or their adaptors.
Collectively, these findings reveal that pHi functions as an internal rheostat that fine-tunes the level of PI3K-mTOR signaling, likely in response to physiological and pathological cues.In the context of cancer, we propose that elevated PI3K-mTORC1 and/or PI3K-mTORC2 signaling may contribute to the growth, proliferation, and survival of cancer cells bearing an alkaline cytosolic pH in growing tumors deprived of essential growth factors, prior to the onset of angiogenesis.As a final note, this study provides a cautionary tale highlighting the importance of controlling pH appropriately in cell cultures, as inadvertent alteration of the pHe affects pHi, cell signaling, and cell function (48).

Cell culture and treatments
Cell lines (MEFs, C2C12, U2OS, HEK293, HeLa, and MCF7) were cultured in DMEM containing high glucose (4.5 g/l), glutamine (584 mg/l), and sodium pyruvate (110 mg/l) (Life Technologies; catalog no.: 11995-065) supplemented with 10% FBS (Life Technologies; catalog no.: 10347-028) and incubated at 37 C in a humidified atmosphere with 7.5% CO 2 .To serum-starve cells, cells were washed once in DMEM containing 20 mM Hepes (pH 7.2) and then refed with this medium overnight (16 h).To amino acid-starve cells, cells were washed once in PBS containing Ca 2+ and Mg + and then refed with amino acid-free DMEM (US Biologicals; catalog no.: D9800-13) lacking or containing dialyzed FBS (10%) (Life Technologies; catalog no.: A33820-01) for 50 min.After amino acid deprivation, cells were stimulated with amino acids by adding an amino acid solution (RPMI1640 Amino Acid Solution [50×]; Sigma; catalog no.: R7131) supplemented with L- glutamine (Sigma; catalog no.: G8540) to a final concentration of (1×) (roughly the concentration of amino acids found in RPMI media).The pH of the commercial amino acid solution ( pH 10) was adjusted to physiological pH 7.4 with 1 N HCl prior to addition to cells.To culture cells at an alkaline pHe, and thus to raise pHi, the pH of the DMEM was adjusted to pH 8.3 (or the pH indicated) using NaOH.To maintain DMEM at an alkaline pH (e.g., pH 8.3) for a longer period, cells cultured in DMEM (pH 8.3) were incubated at 37 C in a humidified atmosphere set to low CO 2 (0.1%).Where indicated, cells were pretreated with various drugs or chemicals for 30 min: Torin1 (100 nM), BYL719 (10 μM), MK2206 (10 μM), cariporide (10 μM), and NH 4 Cl (10 mM).

Cell lysis, immunoprecipitation, and immunoblotting
Unless indicated otherwise, cells were washed twice with ice-cold PBS, lysed in ice-cold buffer containing NP-40 (0.5%) and Brij35 (0.1%), and incubated on ice (15 min), as described (83).To maintain the detergent-sensitive mTOR-Raptor interaction, cells were lysed in ice-cold buffer containing CHAPS (0.3%) detergent.Lysates were clarified by spinning at 13,200 rpm for 5 min at 4 C, and the postnuclear supernatants were collected.Bradford assay was used to normalize protein levels for immunoprecipitation and immunoblot analyses.For immunoprecipitation, whole-cell lysates were incubated with antibodies for 2 h at 4 C, followed by incubation with Protein A-Sepharose beads for 1 h.For 7-methyl-GTP Sepharose pull-down assays, cells lysed in ice-cold buffer containing NP-40 (0.5%) and Brij35 (0.1%) were incubated with 7-methyl-GTP Sepharose beads for 2 h.Sepharose beads were washed three times in lysis buffer and resuspended in 1× sample buffer.Samples were resolved on SDS-PAGE gels and transferred to PVDF membranes in Towbin transfer buffer containing 0.02% SDS, as described (83).Immunoblotting was performed by blocking PVDF membranes in Tris-buffered saline (TBS) pH 7.5 with 0.1% Tween-20 (TBST) containing 3% nonfat dry milk, as described (83) and incubating the membranes in TBST containing 2% bovine serum albumin containing primary antibodies or secondary HRP-conjugated antibodies.Blots were developed by enhanced chemiluminescence and detected digitally with a UVP ChemStudio Imaging System from Analytik Jena.

Isolation of primary MEFs
Primary MEFs were isolated from day 12.5 mouse embryos, as described (84,85).

Quantification of PIP 3 to measure PI3K activity
Levels of PIP 3 were measured using an ELISA-based kit according to the manufacturer's instructions (Echelon; catalog no.: K2500s).MEFs were cultured on 15 cm plates and treated with media at pH 7.4 or 8.3, drugs, and/or insulin.The media were aspirated, 10 ml of ice-cold 0.5 M TCA was added, and the plate was incubated on ice for 5 min.Cells were scraped and collected into a 15 ml tube on ice and centrifuged at 950g for 7 min at 4 C.The pellet was resuspended in 3 ml of 5% tricarboxylic acid/1 mM EDTA, vortexed, centrifuged at 3000 rpm for 5 min, and washed once again in 3 ml 5% trichloroacetic acid/1 mM EDTA.Neutral lipids were next extracted by adding 3 ml of MeOH:CHCl 3 (2:1) to the pellet and vortexing 10 min followed by centrifugation at 950g for 5 min.The supernatant was discarded, and this extraction step was repeated once more.To extract acidic lipids, 2.25 ml of MeOH:CHCl3:HCl (12 N) (80:40:1) was added, and the solution was vortexed continuously for 25 min.Extracts were centrifuged at 950g for 5 min, and the supernatant was transferred to a new 15 ml tube.About 0.75 ml of CHCl 3 and 1.35 ml of 0.1 M HCl was added to the supernatant, vortexed, and centrifuged at 950g for 5 min to separate the lower organic and upper aqueous phases.The organic phase (1.45 ml) was transferred to a new tube for the measurement of PIP 3 .All samples were dried in a vacuum dryer for 1 h.PIP 3 samples were resuspended in 140 ml of PBS-Tween +3% Protein Stabilizer (provided in the Echelon kit).Samples were sonicated in an ice-water bath for 10 min and spun down.The experiment was performed four times with technical duplicates.PIP 3 levels were measured by ELISA, according to the manufacturer's instructions (Echelon; catalog no.: K2500s).

cSNARF-1AM microscopy
MEFs were cultured on 8-well cover glass bottom chambers (Lab-Tek; catalog no.: 155409) and loaded with cSNARF1-AM (10 μM) (Thermo Fisher; catalog no.: C1272) for 30 min at 37 C in serum-free DMEM (pH 7.4).The cells were then rinsed with serum-free DMEM (pH 7.4) and refed with this medium.Live-cell imaging (0-10 min) of cells cultured in serum-free DMEM (pH 7.4 or 8.3) was performed using a Nikon A1 confocal microscope equipped with a stage-top incubator that maintains temperature (37 C) and CO 2 (7.5%).Images were acquired using a 40× oil objective (1.3 numerical aperture) at a resolution of 1024 × 1024 and an optical thickness of 1.18 μM (confocal aperture set at three airy units).The cSNARF1-AM signal was excited with an argon laser at 488 nm, and images sets were collected simultaneously in two emission bandpass filters at 553 to 618 nm (for 580 nm) and 663 to 738 nm (for 640 nm).Nikon Elements software was used to create and pseudo-color the 640:580 nm ratiometric images with values ranging from 0 (violet) to 2.0 (red).MetaMorph software (from MDS Analytical Technologies) was used to quantify the ratiometric images by measuring the average ratio of a region of interest within one cell, and 25 cells per image were quantified.A change in the 640:580 nm ratio and accompanying pseudo-colored image reflects a change in pHi.

Measurement of protein synthesis
Protein synthesis was measured by SUnSET assay, as described (87).Briefly, MEFs seeded to 6-well plates were serum-starved overnight (16 h) and refed with serum-free DMEM at pH 7.4 or 8.3 for 60 min.To maintain DMEM at an alkaline pH (i.e., pH 8.3) for a longer period, the cells were incubated at 37 C in a humidified atmosphere at low CO 2 (i.e., 0.1% CO 2 ).Puromycin (10 μg/ml) was added for the last 10 min of refeeding.Cells were lysed with ice-cold lysis buffer containing NP-40 (0.5%) and Brij35 (0.1%), and whole-cell lysates were immunoblotted with an antipuromycin antibody.Puromycin incorporation was analyzed using ImageJ (aka, Fiji) software.Equal protein loading of SDS-PAGE gels was confirmed by staining with SimplyBlue SafeStain (Invitrogen; catalog no.: LC6060), according to the manufacturer's instructions.

Image editing
Adobe Photoshop was used for image preparation, using only the levels, brightness, and contrast controllers equivalently over the entire image.

Figure 2 .
Figure 2. Incubation of cells in media at alkaline extracellular pH or containing NH 4 Cl increases intracellular pH.A, MEFs were serum-starved overnight (o/n), incubated in serum-free DMEM (pH 7.4) supplemented with cSNARF1-AM (10 mM) (30 min), and refed with serum-free DMEM (pH 7.4).One image was acquired immediately (preimage #1), and three additional images were acquired at 40 s, 5 min, and 10 min.Cells were next refed with serum-free DMEM (pH 7.4), and one image was acquired immediately (preimage #2).Cells were refed again with serum-free DMEM (pH 8.3), and three images were acquired at 40 s, 5 min, and 10 min.Scale bar represents 50 μm.B, quantitation of the results in A. Data represent mean ± SD of n = 100 cells from one representative experiment; one-way ANOVA, ***p < 0.001.C, MEFs were loaded with cSNARF1-AM as in C.After refeeding with serum-free DMEM pH 7.4, one image was acquired immediately (preimage), at which point NH 4 Cl (10 mM) was added, and three additional images were acquired at 40 s, 5 min, and 10 min.Scale bar represents 50 μm.D, quantitation of the results in C. Data represent mean ± SD (n = 20 cells from one representative experiment); one-way ANOVA, ***p < 0.001.All panels show representative results from at least three independent experiments.E, MEFs were serumstarved o/n, pretreated without or with Torin1 (T) (100 nM) (30 min), and treated without or with NH 4 Cl (10 mM) (5 and 10 min).F, MEFs were serumstarved o/n, pretreated without (DMSO) or with cariporide (10 μM) (30 min), and refed with serum-free DMEM at pH 7.4 or set to pH 8.3 in the absence or the presence of cariporide (10 min).All panels show representative results from at least three independent experiments.Molecular weight markers are provided in kilodalton.DMEM, Dulbecco's modified Eagle's medium; DMSO, dimethyl sulfoxide; MEF, mouse embryonic fibroblast.

Figure 3 .
Figure 3. Alkaline intracellular pH (pHi) mimics growth factor signaling to activate mTORC1 in parallel to amino acid signaling.A, MEFs were serumstarved overnight (o/n) and pretreated without or with Torin1 (T) (100 nM) (30 min), refed with serum-free DMEM at pH 7.4 or pH 8.3 without (−) or with (+) 100 nM insulin (10 min) in the absence or the presence of Torin1 (T).Whole-cell lysates were immunoblotted as indicated.B, MEFs were treated as in A, except that cells were stimulated with a submaximal dose of insulin, 5 nM.C, MEFs were serum-starved o/n, amino acid-starved for 50 min in DMEM lacking amino acids and FBS, and then stimulated without (−) or with (+) an amino acid solution (1× final) whose pH was adjusted to 7.4 or not adjusted (Ph = 10) (0-30 min).Note: The concentration of amino acids added to (1×) is similar to those found in RPMI media.D, MEFs were deprived of serum growth factors and amino acids as in C, pretreated without or with Torin1 (100 nM) (30 min), and refed with serum-and amino acid-free DMEM at pH 7.4 or pH 8.3 without (−) or with (+) AAs (1×) (10 min).E, MEFs were serum-starved o/n and pretreated without or with cariporide (C) (10 μM) and Torin1 (T) (100 nM) (30 min), refed with serum-free DMEM at pH 7.4 or pH 8.3 without (−) or with (+) 100 nM insulin (10 min) in the absence or the presence of drugs.Whole-cell lysates were immunoblotted as indicated.All panels show representative results from at least three independent experiments.Molecular weight markers are provided in kilodalton.AA-sol, amino acid solution; DE, dark exposure; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum; LE, light exposure; MEF, mouse embryonic fibroblast; mTORC1, mechanistic target of rapamycin complex 1.

Figure 6 .
Figure 6.Alkaline pHi increases cap-dependent protein synthesis in the absence of growth factors through PI3K-mTOR signaling.A, MEFs were serum-starved overnight (o/n), pretreated without or with cariporide (C) (10 μM) or Torin1 (T) (100 nM) (30 min), and refed with serum-free DMEM at pH 7.4 or 8.3 in the absence or th presence of drugs(30 min).eIF4E was pulled down with 7-methyl GTP-Sepharose beads.Pull-downs and whole-cell lysates were immunoblotted as indicated.B, MEFs were treated as in A, except they were pretreated with BYL719 (BYL) (10 μM) or Torin1 (100 nM) (30 min) and incubated at either normal (7.5%) or low CO 2 (0.1%).eIF4E was pulled down with 7-methyl GTP-Sepharose beads.Pull-downs and whole-cell lysates were immunoblotted as indicated.Note: Low CO 2 maintains the media at an alkaline pH for a longer period.C, quantitation of the results shown in B, that is, the amount of 4EBP1 (left graph) or eIF4G (right graph) that coprecipitates with eIF4E at pH 7.4 versus pH 8.3.Data represent mean ± SD (n = 4 independent experiments; one-way ANOVA; **p < 0.01; ***p < 0.001; ****p < 0.0001.D, MEFs were treated as in A, except they were refed with DMEM at pH 7.4 or pH 8.3 for 60 min (instead of 10 min) and pulsed with puromycin (10 μg/ml) for the last 10 min.Cells were cultured in incubators set to either 7.5% CO 2 or 0.1% CO 2 .Note: Low CO 2 maintains the media at an alkaline pH for a longer period.E, quantitation of puromycin incorporation.Data represent mean ± SD (n = 4 independent experiments); one-way ANOVA; ***p < 0.001; ****p < 0.0001.All panels show representative results from at least three independent experiments.Molecular weight markers are provided in kilodalton.4EBP1, eIF4E binding protein 1; DMEM, Dulbecco's modified Eagle's medium; eIF4E, eukaryotic initiation factor 4E; MEF, mouse embryonic fibroblast; mTOR, mechanistic target of rapamycin; pHi, intracellular pH; PI3K, phosphatidylinositide 3kinase.