mTORC1, the maestro of cell metabolism and growth

In this review, He et al. discuss the cellular functions and regulation of mTORC1, a molecular rheostat that assimilates extracellular signals and controls cellular growth and adaptation by regulating transcription, translation, and cell metabolism. They further highlight the role of mTORC1 in the etiology of cancer, aging, and neurological disorders, as well as its amenability to therapeutic targeting.

Abstract

The mechanistic target of rapamycin (mTOR) pathway senses and integrates various environmental and intracellular cues to regulate cell growth and proliferation. As a key conductor of the balance between anabolic and catabolic processes, mTOR complex 1 (mTORC1) orchestrates the symphonic regulation of glycolysis, nucleic acid and lipid metabolism, protein translation and degradation, and gene expression. Dysregulation of the mTOR pathway is linked to numerous human diseases, including cancer, neurodegenerative disorders, obesity, diabetes, and aging. This review provides an in-depth understanding of how nutrients and growth signals are coordinated to influence mTOR signaling and the extensive metabolic rewiring under its command. Additionally, we discuss the use of mTORC1 inhibitors in various aging-associated metabolic diseases and the current and future potential for targeting mTOR in clinical settings. By deciphering the complex landscape of mTORC1 signaling, this review aims to inform novel therapeutic strategies and provide a road map for future research endeavors in this dynamic and rapidly evolving field.

mTOR is an evolutionarily conserved serine/threonine kinase that integrates extracellular and intracellular signals to regulate cellular homeostasis and metabolism as a molecular rheostat (Laplante and Sabatini 2012; Saxton and Sabatini 2017). Its discovery traces back to the isolation of rapamycin from the soil bacterium Streptomyces hygroscopicus found on Rapa Nui, also known as Easter Island, in 1965 (Sehgal et al. 1975). Initially recognized for its antifungal properties, rapamycin's potential as an immunosuppressant was later unveiled, followed by the discovery of its ability to inhibit cell growth (Martel et al. 1977; Eng et al. 1984). In the pursuit of understanding rapamycin's mechanisms, genetic screens in yeast revealed that mutations in the TOR gene and a cis–trans prolyl isomerase, FKBP12 (FK506-binding protein), rendered cells resistant to rapamycin (Heitman et al. 1991; Cafferkey et al. 1993; Kunz et al. 1993; Helliwell et al. 1994). This seminal discovery paved the way for further investigations into TOR's role in cellular signaling. First, rapamycin:FKBP12 was shown to potently inhibit activation of S6K1 by multiple agonists using different downstream effectors (Calvo et al. 1992; Chung et al. 1992; Kuo et al. 1992; Price et al. 1992). Second, it was then shown that phosphatidylinositol-3-kinase (PI3K) mediated signaling to 70 kDa ribosomal S6 kinase (S6K1), which was inhibited by rapamycin (Cheatham et al. 1994; Chung et al. 1994; Monfar et al. 1995). Third, affinity purification experiments revealed that the FKBP12–rapamycin complex binds to mTOR, unraveling the intricate molecular interactions underlying rapamycin's inhibitory effects (Brown et al. 1994; Sabatini et al. 1994; Sabers et al. 1995). These studies established the framework of what we now refer to as the mitogen-regulated mTOR signaling pathway.

mTOR belongs to the PI3K-related protein kinase (PIKK) family, which includes ataxia-telangiectasia mutated (ATM), ATM- and Rad3-related (ATR), DNA-dependent protein kinase (DNA-PK), and suppressor of morphogenesis in genitalia-1 (SMG-1). These large (∼300–500 kDa) proteins contain a conserved kinase catalytic domain and other regions, such as a 601 amino acid FRAP–ATM–TRRAP (FAT) domain followed by a 313 amino acid PIKK domain and a short (32 amino acid) FAT C-terminal (FATC) domain. Additionally, mTOR possesses a Huntingtin–EF3A–ATM–TOR (HEAT) domain at its N terminus. Functionally, mTOR exists in two distinct complexes: mTORC1 and mTORC2 (Loewith et al. 2002; Jacinto et al. 2004; Sarbassov et al. 2004). mTORC1 consists of essential components that form its functional core. mTOR provides catalytic activity, Raptor (regulatory-associated protein of mTOR) recruits substrates and stabilizes the complex, and mLST8 (mammalian lethal with SEC13 protein 8) helps stabilize the mTOR kinase domain, making these elements crucial for mTORC1's assembly and function. Nonessential components like DEPTOR (DEP domain-containing mTOR-interacting protein) and PRAS40 (the 40 kDa proline-rich Akt substrate) act as negative regulators by binding to mTOR and modulating its activity (Oldham et al. 2000; Hara et al. 2002; Kim et al. 2002; Loewith et al. 2002; Long et al. 2002; Fingar et al. 2004; Vander Haar et al. 2007). Conversely, mTORC2 consists of essential components such as mTOR; Rictor (rapamycin-insensitive companion of mTOR), which is necessary for complex assembly, substrate recruitment, and stability; mLST8; and mSIN1 (mammalian stress-activated protein kinase-interacting protein 1), which is critical for mTORC2 assembly and kinase activity. DEPTOR is nonessential in mTORC2 and mainly modulates complex assembly and substrate recruitment (Jacinto et al. 2004; Sarbassov et al. 2004; Yang et al. 2006; Pearce et al. 2007). mTORC1 is a central regulator in cellular physiology that integrates and coordinates diverse environmental and intracellular cues such as growth factors, amino acids, glucose and energy availability, hypoxia, and stress to regulate cell growth, proliferation, and adaptation to changing environmental conditions. Building on the core pathway first described in the early 1990s, discoveries in both model organisms and human cells have revealed key conserved regulators of mTORC1. Its dysregulation has profound implications for human health, contributing to the pathogenesis of various diseases, including cancer, metabolic disorders, neurodegenerative conditions, and aging.

mTORC1 signaling network

The TSC–Rheb–mTORC1 signaling axis

Tuberous sclerosis complex (TSC) is a rare autosomal dominant genetic disease that can lead to the growth of benign tumors in the brain (giant cell astrocytoma, cortical tumors, and subependymal nodules) that contribute to a range of neurological symptoms, including epilepsy, as well as intellectual, behavioral, and developmental problems. These patients also suffer from hypomelanic macules and facial angiofibromas, kidney angiomyolipomas, heart rhabdomyomas, lung cysts associated with the disease, lymphangioleiomyomatosis (LAM), and/or other less common symptoms. An important early observation was that the cerebral cortical tumors exhibited a large cell size phenotype (Bender and Yunis 1982; Richardson 1991; Mizuguchi and Takashima 2001). The discovery of the TSC protein complex represents a prime example of how the convergence of multiple investigations facilitated our understanding of mTORC1 signal transduction and biology. TSC1 was originally identified by positional cloning as a gene whose loss contributed to TSC (van Slegtenhorst et al. 1997). TSC2 was identified to be lost in TSC and in LAM (Nellist et al. 1993; Carsillo et al. 2000) and also was found to regulate Drosophila cell size (Ito and Rubin 1999). The TSC protein complex is composed of TSC1 (hamartin), TSC2 (tuberin), and TBC1 domain family member 7 (TBC1D7) in a 2:2:1 stoichiometry (Yang et al. 2021). Sequencing of TSC2 from humans to Drosophila revealed a large conserved putative GTPase-activating domain. Early studies suggested that TSC2 might act as a GTPase-activating protein (GAP) toward Ras-related protein 1 (Rap1a) (Wienecke et al. 1995); however, loss of the Rheb GTPase in Schizosaccharomyces pombe, which yielded a G0/G1 arrest and small cell size (Tabancay et al. 2003), suggested that the mammalian equivalent might be the target of TSC2. Indeed, the Ras homolog enriched in brain (RHEB) GTPase was shown to be the in vitro and in vivo target of TSC2, which converted RHEB from its GTP-bound active form to its GDP-bound inactive form (Garami et al. 2003; Tee et al. 2003b; Zhang et al. 2003). The TSC complex was also shown to suppress eukaryotic initiation factor 4E-binding protein 1 (4E-BP1) phosphorylation, enhance 4E-BP1's translation inhibitory function, and suppress S6K1 phosphorylation and activation (Inoki et al. 2002; Tee et al. 2002, 2003a; Roux et al. 2004). These and other studies in a variety of model systems suggested that the TSC protein complex was a negative regulator of cell size and cell growth, acting as a tumor suppressor. How this tumor suppressor was inactivated to promote downstream mitogenic signaling was soon defined (Fig. 1).

Figure 1.

Major upstream regulators of mTORC1. Growth factors, amino acids, energy levels, phosphatidic acid (PA), and cellular stresses, including DNA damage and reactive oxygen species (ROS), modulate mTORC1 activity, which in turn regulates numerous biological processes, including gene expression, nucleotide synthesis, protein translation, lipid synthesis, and autophagy. Negative regulators of mTORC1 are highlighted in red. (Figure created with BioRender.com.)

Regulation of mTORC1 by growth factors

How growth factors regulated mTORC1 signaling began to be defined when rapamycin was shown to potently inhibit PI3K-dependent activation of S6K1 and S6K2 downstream from growth factors and insulin (Chung et al. 1994; Cheatham et al. 1994; Monfar et al. 1996; Martin et al. 2001a,b). Subsequently, another PI3K effector, AKT/PKB (protein kinase B), was identified (Coffer et al. 1998; Vanhaesebroeck and Alessi 2000) and then shown to phosphorylate TSC2 (Ser939, Ser981, Ser1130/1132, and Thr1462) (Inoki et al. 2002; Manning et al. 2002; Potter et al. 2002). The association of the TSC complex with the lysosome where mTORC1 activation predominantly occurs (Saxton and Sabatini 2017) was shown to be mitogen-regulated (Demetriades et al. 2014; Menon et al. 2014). The mechanism was shown to be via TSC2 phosphorylation, which promotes its dissociation from the lysosome, thereby resulting in GTP loading and activation of Rheb and mTORC1 (Menon et al. 2014). The Ras/ERK (extracellular signal-regulated kinase)–MAP (mitogen-activated protein) kinase pathway was also shown to contribute to mTORC1 activation through TSC2 phosphorylation by ERK at Ser664, which leads to TSC1/TSC2 complex dissociation (Ma et al. 2005), and through its downstream effector, p90 ribosomal S6 kinase 1 (RSK1), which suppresses TSC2 function by phosphorylating it at Ser939, Ser981, Thr1462, and Ser1798 (Roux et al. 2004). Additionally serum/glucocorticoid-regulated kinase 1 (SGK1) and other kinases phosphorylate TSC2 to promote mTORC1 activation (Fig. 1; Castel et al. 2016).

Regulation of mTORC1 signaling by amino acids

Lysosomal recruitment of mTORC1 is critical for its activation by Rheb. This process is regulated via the action of the Rag family of small GTPases comprised of RagA/B and RagC/D heterodimers. Changing the nucleotide-bound state of these Rag GTPases is a key mechanism by which amino acids and various stimuli modulate mTORC1 activity (Kim et al. 2008; Sancak et al. 2008). Under low-amino-acid conditions, RagA/B is GDP-bound and RagC/D is GTP-bound, a configuration that does not interact with mTORC1. When amino acids are present, the Rags switch their nucleotide-binding status, enabling interaction with mTORC1 and facilitating its recruitment to the lysosomal surface (Sancak et al. 2008). Additionally, upon amino acid deprivation, Rag GTPases recruit TSC2 to the lysosome, where it can act on Rheb to inhibit mTORC1 activity (Demetriades et al. 2014). RagA/B and RagC/D traditionally were deemed to possess redundant functions due to their high sequence similarity (Sekiguchi et al. 2001); however, recent findings suggest that RagD may also be responsible for the regulation of mTORC1 on lysosomes, whereas RagC is more relevant for the phosphorylation of cytoplasmic mTORC1 substrates (Gollwitzer et al. 2022). The Ragulator complex, comprised of MP1 (LAMTOR3), p14 (LAMTOR2), p18 (LAMTOR1), HBXIP (LAMTOR5), and C7ORF59 (LAMTOR4), serves as an anchor for Rag GTPases on the lysosome and also acts as a guanine nucleotide exchange factor (GEF) for RagA, RagB (Bar-Peled et al. 2012), and RagC (Shen and Sabatini 2018). The nucleotide states of RagA and RagB are also regulated by GATOR1, serving as a GAP for RagA/B (Bar-Peled et al. 2013). Meanwhile, GATOR2, another protein complex, interacts with and inhibits GATOR1's activity (Bar-Peled et al. 2013). Both GATOR1 and GATOR2 are pivotal in modulating amino acid and glucose signals to mTORC1 (Fig. 1). SLC38A9 (solute carrier family 38 member 9), initially identified as another GEF for RagA (Shen and Sabatini 2018), has more recently been recognized for its function as a GAP for RagC (Fromm et al. 2020) and senses multiple amino acids, including asparagine, arginine, glutamine, histidine, and lysine, in the lysosome to regulate mTORC1 signaling (Jung et al. 2015; Rebsamen et al. 2015; Wang et al. 2015). Interestingly, SLC38A9 also interacts with the lysosomal cholesterol transporter Niemann–Pick C1 (NPC1), which is essential for cholesterol-induced mTORC1 activation, highlighting a unique interplay between lipid-sensing and cellular growth signaling mechanisms (Fig. 1; Castellano et al. 2017).

SLC38A9 can also sense arginine in the lysosome, and CASTOR1/2 serves as an arginine detector in the cytoplasm (Chantranupong et al. 2016). Under low-arginine conditions, CASTOR1/2 binds to and inhibits GATOR2, which in turn suppresses mTORC1 activity. Recently, RNF167, an E3 ubiquitin ligase, has been found to degrade CASTOR1, a process facilitated by AKT-mediated phosphorylation at Ser14 (Li et al. 2021). Sestrin2 is a leucine sensor that interacts with and inhibits GATOR2 in the absence of leucine (Wolfson et al. 2016). When leucine binds to Sestrin2, this inhibition is lifted, enabling the recruitment of mTORC1 to the lysosome. Recent findings emphasize the role of secretion-associated Ras-related GTPase 1B (SAR1B) as another leucine sensor to regulate mTORC1 signaling. Likewise, when leucine is low, SAR1B interacts with and inhibits GATOR2, and this interaction is reversed when leucine is high (Chen et al. 2021). Moreover, the multifaceted role of leucine extends to facilitating the translocation of leucyl-tRNA synthetase (LARS) to lysosomes, where it acts as a GAP for RagD, resulting in activation of mTORC1 (Han et al. 2012). The methionine-sensing mechanism primarily revolves around S-adenosylmethionine (SAM), a critical metabolite of methionine. SAMTOR is a sensor of SAM that, in the absence of methionine and SAM, interacts with and activates GATOR1, thereby inhibiting mTORC1 signaling (Gu et al. 2017). Recently, it was shown that SAM-loaded protein arginine methyltransferase 1 (PRMT1) methylates nitrogen permease regulator 2-like protein (NPRL2), the catalytic subunit of GATOR1, which suppresses its GAP activity and positively regulates mTORC1 signaling (Jiang et al. 2023). The molecular mechanism underlying mTORC1's threonine sensing and activity involves mitochondrial threonyl-tRNA synthetase 2 (TARS2) (Kim et al. 2021). Threonine binding prompts TARS2 association with GTP-bound RagC, where it facilitates GTP loading of RagA, though TARS2 itself does not function as a GEF (Fig. 1).

Although Rag GTPases play a crucial role in detecting external amino acids and glucose, regulating mTORC1 involves even more complexity, with mechanisms that operate independently of Rag GTPases. Activation of mTORC1 by amino acids released from lysosomal protein degradation involves the homotypic fusion and protein-sorting (HOPS) complex, showcasing a mechanism that operates independently of Rag GTPases (Hesketh et al. 2020). Furthermore, under amino acid scarcity, Rap1 GTPases relocate lysosomes to the perinuclear area of the cell and reduce lysosome abundance, resulting in downregulation of mTORC1 signaling (Mutvei et al. 2020). When stimulated by glutamine or asparagine, the adenosine diphosphate ribosylation factor 1 (Arf1) GTPase promotes the localization of mTORC1 to the lysosome independently of the Rag GTPases and Ragulator, suggesting a distinct pathway for mTORC1 activation (Jewell et al. 2015; Bernfeld et al. 2018; Meng et al. 2020). In addition, during prolonged amino acid deprivation, mTORC1 suppression remains sustained due to GCN2 (general control nonderepressible 2)- and activating transcription factor 4 (ATF4)-mediated upregulation of Sestrin2 (Ye et al. 2015).

It is important to note that although most studies show a convergence of regulatory inputs to regulate mTORC1 signaling at the lysosome, more recent evidence indicates that mTORC1 is also activated at the Golgi (Buerger et al. 2006; Fan et al. 2016; Hao et al. 2018), and that activation requires reversible Rheb association with internal membranes (Angarola and Ferguson 2019). mTORC1 is also activated at the peroxisome through peroxisomal localization of the TSC via interactions with PEX19 and PEX5 (peroxisomal biogenesis factors 19 and 5), where it suppresses mTORC1 and induces autophagy in response to ROS (Zhang et al. 2013). Finally, nonfarnesylated Rheb regulates mTORC1 in the nucleus (Zhou et al. 2015, 2020; Zhong et al. 2022).

Cellular energy status modulates mTORC1 activation

Under low-energy conditions such as reduced glucose availability, the cell must reduce its energy consumption to survive. Under these conditions, AMP-activated protein kinase (AMPK) is activated and suppresses mTORC1 by phosphorylating TSC2 (Ser1345) and heightening its GAP function (Inoki et al. 2003). Moreover, AMPK prompts Raptor and 14-3-3 interaction through phosphorylation of Raptor (Ser722 and Ser792) and suppresses mTORC1 activity (Gwinn et al. 2008). Recent findings also suggest that upon glucose deprivation, AMPK phosphorylates WDR24 at Ser155, which disrupts GATOR2 complex formation and suppresses mTORC1 activity (Dai et al. 2023). In addition, AMPK phosphorylates FNIP1 (folliculin-interacting protein 1), inhibiting the GAP activity of FNIP1–FLCN (folliculin) toward RagC and leading to GTP-bound RagC accumulation and lysosomal detachment of mTORC1. This selectively prevents mTORC1-mediated phosphorylation of TFEB/TFE3 but not other substrates like S6K1 and 4E-BP1, thereby orchestrating mitochondrial and lysosomal biogenesis (Malik et al. 2023). Energy deprivation also sets off a series of events culminating in mTORC1 suppression in addition to AMPK-dependent TSC2 activation. Glucose starvation triggers O-GlcNAcylation of leucyl-tRNA synthetase 1 (LARS1) at Ser1042, which reduces leucine binding affinity and interaction with RagD, thereby suppressing mTORC1 activity (Kim et al. 2022). In addition, the metabolic byproduct dihydroxyacetone phosphate (DHAP), aside from its glycolytic role, independently regulates mTORC1. Although the exact sensor remains unidentified, it is known that DHAP triggers a response reliant on both GATOR complexes to regulate mTORC1 activity (Orozco et al. 2020). In addition, galectin-3, a β-galactoside-binding lectin, senses lipopolysaccharide (LPS) to promote the association of Rag GTPases and Ragulator, leading to the activation of mTORC1 (Chen et al. 2022). Energy depletion also leads to the disassembly and repression of the ATP-dependent TTT (Tel2, TTI1, and TTI2)–RUVBL (RuvB-like AAA ATPase) 1/2 complex, which in turn suppresses the assembly and dimerization of mTORC1, thereby inhibiting mTORC1 signaling (Fig. 1; Kim et al. 2013).

Additional inputs into regulation of mTORC1 signaling

Hypoxia suppresses mTORC1 activity through the expression of DNA damage-induced expression of development 1 (REDD1), which either sequesters 14-3-3 from TSC2, allowing for its dephosphorylation (Thr1462 and Ser939), or dephosphorylates AKT in a protein phosphatase 2A (PP2A)-dependent manner (Thr308). These events enhance TSC2 activity and subsequent mTORC1 suppression (Brugarolas et al. 2004; Reiling and Hafen 2004; Dennis et al. 2014).

The Hippo and mTOR signaling pathways, both critical regulators of organ size and tumorigenesis, exhibit significant cross-talk (Honda et al. 2023). Hippo pathway activation leads to LATS1/2-mediated phosphorylation and cytoplasmic retention of YAP and TAZ, key transcriptional coactivators. This cross-talk manifests in several interconnected ways. At the core of this interaction, LATS1 and LATS2, key kinases in the Hippo pathway, phosphorylate Raptor at Ser606, attenuating mTORC1 activation by disrupting the Raptor–Rheb interaction (Gan et al. 2020). This phosphorylation event serves to attenuate mTORC1 activation by disrupting the interaction between Raptor and Rheb, thus establishing a direct connection between the Hippo and mTORC1 pathways in growth regulation. Furthermore, the Hippo pathway influences mTOR signaling by modulating insulin signaling. TAZ induces IRS1 expression, enhancing insulin signaling and activating mTOR (Hwang et al. 2019). Concurrently, YAP promotes PIP3 production by suppressing PTEN expression, activating the AKT/mTOR pathway and stimulating cell growth and proliferation across various tissues (Tumaneng et al. 2012). Additionally, YAP/TAZ enhance mTORC1 activity by upregulating the expression of amino acid transporter genes SLC7A5 (Hansen et al. 2015) and SLC38A1 (Park et al. 2016), thereby increasing amino acid uptake. Interestingly, mTOR hyperactivity in TSC1/2 mutant cells leads to impaired autophagy, resulting in YAP accumulation (Liang et al. 2014). These mechanisms collectively underscore the intricate bidirectional interplay between Hippo and mTOR pathways in regulating cellular growth and metabolism, highlighting the complex network of signaling cascades that govern these fundamental biological processes.

Wnt also activates mTOR independently of β-catenin-dependent transcription. Notably, inhibition of mTOR by rapamycin impedes Wnt-induced cell growth and tumor development, indicating rapamycin's potential therapeutic value for cancers driven by activated Wnt signaling (Inoki et al. 2006).

Interestingly, exogenously provided phosphatidic acid (PA) vesicles recruit mTORC1 to the lysosome in the absence of amino acids or Rag GTPases to activate mTORC1 signaling (Fang et al. 2001; Frias et al. 2020, 2023). Recent findings also emphasize the regulatory impact of lysosomal cholesterol signaling (LYCHOS) on mTORC1 activity, influencing lysosomal cholesterol levels and enhancing the recruitment of mTORC1 to lysosomes in a Rag-dependent manner (Shin et al. 2022). Additionally, although acetyl-CoA positively regulates mTORC1 activity by facilitating EP300-mediated acetylation of Raptor at Lys1097 (Son et al. 2019), malonyl-CoA, an intermediate in fatty acid biosynthesis, binds directly to the mTOR catalytic pocket and inhibits it (Nicastro et al. 2023).

Downstream from mTORC1

Comparisons of the structural features of mTOR's catalytic domain within the mTORC1 complex with those of other PIKKs (such as SMG1, ATM, and ATR) reveal both similarities and notable differences in their interactions with substrate phosphorylation sites. One significant difference is that unlike SMG1, ATM, and ATR, mTOR does not always establish specific contacts with the amino acid residues immediately flanking the phosphorylation site (+1 and −1 positions) (Sturgill and Hall 2009; Yang et al. 2013; Langer et al. 2020, 2021; Battaglioni et al. 2022). This flexibility enables mTOR to accommodate substrates with amino acids of varying sizes at these positions. For example, most of mTORC1's reported substrates are proline-directed, whereas S6K1 phosphorylation occurs at its hydrophobic motif (Thr389), with Phe in the −1 and Tyr in the +1 positions (Burnett et al. 1998). In contrast, SMG1, ATM, and ATR show a strong preference for glutamine at the +1 position of their substrates (Johnson et al. 2023). 4EBPs block translation by attaching to the cap-binding protein eIF4E, preventing formation of the eIF4F complex. mTORC1 phosphorylates 4EBPs at multiple proline-directed sites (Fig. 2). Phosphorylation of Thr37–Pro and Thr46–Pro primes further phosphorylation at Thr70–Pro and then Ser65–Pro, leading to its release from eIF4E (discussed in greater detail below; Gingras et al. 1999a, 2001). Phosphorylation of these sites is dependent on the mTOR signaling (TOS) motif, which recruits mTORC1 via Raptor binding (Schalm and Blenis 2002; Schaim et al. 2003). Another substrate of mTORC1, growth factor receptor-bound protein 10 (Grb10), plays a role in feedback inhibition of receptor tyrosine kinase–PI3K signaling. mTORC1 phosphorylates Grb10 at Ser501–Pro and Ser503–Pro, which stabilizes the protein and thereby increases its inhibition of receptor signaling (Hsu et al. 2011; Yu et al. 2011). mTORC1 also regulates the phosphorylation of UNC-51-like kinase 1 (ULK1) (Kim et al. 2011), transcription factor EB (TFEB) (Martina et al. 2012; Roczniak-Ferguson et al. 2012; Settembre et al. 2012; Vega-Rubin-de-Celis et al. 2017), protein associated with UV radiation resistance-associated gene protein (UVRAG), autophagy enhancer (PACER [protein associated with UVRAG as autophagy enhancer]) (Cheng et al. 2019), and WD repeat domain phosphoinositide-interacting protein 2 (WIPI2) (Wan et al. 2018) to control autophagy and lysosome biogenesis, a process detailed further in the next section. These are also phosphorylated at proline-directed sites (X-pS/T-Pro). It remains unclear what additional determinants are required for regulating these unique substrate specificities.

Figure 2.

Translation regulation by mTORC1 signaling. mTORC1 promotes translation of terminal oligopyrimidine (TOP) mRNAs via LARP1 phosphorylation. mTORC1 also enhances cap-dependent translation initiation through phosphorylation of 4E-BP and components of the translation initiation complex. S6K stimulates translation initiation by phosphorylating eIF4B, which promotes its recruitment into the translation preinitiation complex while also phosphorylating PDCD4 and inducing its ubiquitin-mediated turnover, thereby boosting eIF4A RNA helicase activity. S6K also promotes translation elongation through phosphorylation of eEF2K. (Figure created with BioRender.com.)

Emerging evidence suggests that the phosphorylation of microphthalmia/transcription factor E (MiT-TFE) members such as TFEB and TFE3, well known mTORC1 substrates that lack TOS or RAIP motifs, differs markedly from the phosphorylation of other established mTORC1 downstream targets that contain TOS motifs (Wada et al. 2016; Bartolomeo et al. 2017; Lawrence et al. 2019; Li et al. 2019; Napolitano et al. 2022). Notably, under specific conditions such as FLCN disruption, the phosphorylation of MiT-TFE factors (TFEB and TFE3) is inhibited, whereas the phosphorylation of S6K and 4E-BP1 remains unaffected (Wada et al. 2016; Lawrence et al. 2019). Similarly, in Birt–Hogg–Dubé (BHD) syndrome and tuberous sclerosis complex diseases, MiT-TFE factors are dephosphorylated and remain active, promoting increased lysosomal activity despite hyperactivated mTORC1 activity (Hasumi et al. 2009; Alesi et al. 2021). This phenomenon is also observed in various cancers, including renal cell carcinoma, pancreatic ductal adenocarcinoma, and melanoma, where MiT-TFE factors remain active even with the hyperphosphorylation of S6K and 4E-BP1 (Hasumi et al. 2009). TFEB and TFE3 contain an N-terminal Rag-binding region (RBR) domain, and recent structural studies have elucidated the architecture of the lysosomal mTORC1–TFEB–Rag–Ragulator megacomplex. In this complex, residues 2–18 of TFEB form an extended α helix (α1) that interacts directly with RagC^GDP and the Ragulator complex (Cui et al. 2023). This regulatory mechanism demonstrates that TOS motif-containing substrates are phosphorylated in response to stimuli that modulate nutrient-driven Rag activity and growth factor-dependent Rheb activation. In contrast, signals specifically affecting RagC/D activity, such as those arising from FLCN disruption, lysosomal damage, mitophagy, or xenophagy, selectively influence TFEB phosphorylation without altering the phosphorylation of TOS motif-containing substrates. In summary, the phosphorylation and subcellular localization of MiT-TFE factors show a distinct insensitivity to growth factor-induced activation of RagA/B, setting them apart from other mTORC1 substrates like S6K and 4E-BP1. This unique characteristic represents an unprecedented aspect of mTORC1 signaling. Moving forward, advancing proteomic and phosphoproteomic techniques, along with structural studies, will be essential for identifying novel mTORC1 substrates and developing specific inhibitors targeting MiT-TFE factors. These efforts will facilitate the creation of new pharmacological strategies aimed at selectively modulating mTORC1 activity, particularly by targeting the FLCN–RagC/D–Ragulator axis in disease contexts such as BHD syndrome and TSC.

It is worth noting that in the case of S6K1, there is also evidence that mTORC1 phosphorylates several proline-directed sites within its pseudosubstrate autoinhibitory domain, exposing Thr389 for substoichiometric phosphorylation by mTORC1. This limited phosphorylation, however, creates a docking site for PDK1, which then phosphorylates the activation loop site (Thr229) (Alessi et al. 1998; Vanhaesebroeck and Alessi 2000). S6K1 then autophosphorylates at Thr389, promoting its full phosphorylation and activation. In support of this model, expression of multiple kinase-inactive S6K mutants displays limited Thr389 phosphorylation upon mitogen stimulation, in contrast to overexpressed wild-type S6K1 (Romanelli et al. 2002).

Once activated, S6K1 regulates various cellular processes by phosphorylating eukaryotic translation initiation factor 4B (eIF4B) (Raught et al. 2004), the S6K–Aly/REF-like substrate (SKAR) (Richardson et al. 2004), cap-binding protein 80 kDa (CBP80) (Wilson et al. 2000), programmed cell death 4 (PDCD4) (Dorrello et al. 2006), eukaryotic elongation factor 2 kinase (eEF2K) (Browne and Proud 2002), SR protein kinase (SRPK2) (Lee et al. 2017), and others (Fig. 2; Artemenko et al. 2022; Fumagalli and Pende 2022). Through 4EBP1/eIF4E and S6K1, mTORC1 regulates translation or processing of multiple transcription factors such as hypoxia-inducible factor 1α (HIF1α) (Dodd et al. 2015), c-Myc (Balakumaran et al. 2009; Pourdehnada et al. 2013), and sterol regulatory element-binding protein 1 (SREBP1) (Owen et al. 2012).

S6K1 also phosphorylates glycogen synthase kinase 3 (GSK3) at its inhibitory site (Zhang et al. 2006). Interestingly, we and others have found that mTORC1 regulates the cytoplasmic/nuclear distribution of GSK3 (Bautista et al. 2018; He et al. 2019; Wada et al. 2023). Its nuclear accumulation following mTORC1 inhibition offers novel insights into mTORC1's nuclear functions, highlighting a novel aspect of its regulatory impact within the cell nucleus. Several proteins, including Forkhead box K1/2 (FOXK1/2), c-Myc, and PBX homeobox 2 (PBX2), are known to be phosphorylated and regulated by nuclear GSK3 upon mTORC1 suppression (Bautista et al. 2018; He et al. 2019; Wada et al. 2023).

Role of mTORC1 in mRNA metabolism and protein translation

mRNA splicing and post-transcriptional modification

mTORC1 signaling has been reported to regulate RNA biogenesis through mRNA splicing or mRNA N⁶-methyladenosine (m⁶A) modification (Lee et al. 2017; Cho et al. 2021, 2023; Tang et al. 2021; Villa et al. 2021). We recently discovered a mechanism linking mTOR signaling to lipid metabolism through alternative splicing of lipogenic mRNAs (Cho et al. 2023). mTORC1 activates S6K1, which phosphorylates SRPK2, promoting its nuclear translocation. Nuclear SRPK2 enhances interactions between splicing factors (e.g., SRSF1) and spliceosomal protein U1-70K while also facilitating the assembly of a multiprotein complex involving SREBP1, family with sequence similarity 120 member A (FAM120A), and RNA Pol2. This complex coordinates nutrient and mitogenic regulation of lipogenic gene transcription and splicing. When mTORC1 signaling is suppressed, reduced nuclear SRPK2 leads to inefficient mRNA splicing, resulting in intron retention and nonsense-mediated decay (Fig. 3).

Figure 3.

mTORC1–S6K and the central dogma. mTORC1–S6K modulates multiple stages of gene expression, including mRNA transcription, processing (such as mRNA splicing), the pioneer round of translation for quality control, and post-transcriptional (m⁶A) and post-translational modifications. For instance, mRNA splicing of lipogenic mRNA, the pioneer round of translation, and m⁶A modification are promoted by mTORC1 activation. This regulation ensures precise control of gene and protein expression. (Figure created with BioRender.com.)

Newly spliced mRNA is capped by the CBP20/CBP80 (nuclear cap-binding protein) complex and is still bound to many of the proteins involved in its splicing, including exon junction complexes (EJCs). The first round of translation, referred to as the pioneer round of translation, appears to be modulated by recruitment of activated S6K1 to SKAR, a protein bound to the EJC and TREX (essential transcription–export complex, consisting of THO, DDX39B, and ALYREF) complexes. Here, S6K1 phosphorylates CBP80 (Wilson et al. 2000), SKAR (Richardson et al. 2004), and other proteins associated with newly spliced CBP20/CBP80-capped mRNAs, though the mechanistic consequences of these events require further investigation. Importantly, S6K1 recruitment is necessary for efficient pre-mRNA splicing to occur, allowing the pioneer round of translation to proceed (Max et al. 2008), which is required for subsequent conversion to eIF4E-dependent translation (Fig. 3; Ma and Blenis 2009). As translation is an energy-consuming process and mTORC1 signaling is responsive to energy sufficiency, these findings suggest a go/no go energy (ATP availability) checkpoint mechanism for determining whether a newly transcribed mRNA should move forward to energy-consuming steady-state translation.

Recent evidence reveals mTORC1's role in regulating RNA modification, particularly m⁶A methylation, which is crucial in various physiological and oncogenic processes (Yang et al. 2020; Zhang et al. 2022). Recently, m⁶A mRNA modifications were shown to be regulated by mTORC1 signaling (Fig. 3). mTORC1 signaling regulates the m⁶A–writer complex (METTL3–METTL14–WTAP) through S6K1-dependent recruitment and activation of eIF4A/B, increasing WTAP translation, which possesses a highly structured 5′ untranslated region (UTR). This mTORC1-mediated m⁶A modification decreases mRNA stability of tumor suppressor MXD2 (MAX-interacting protein), enhancing cMyc activity (Cho et al. 2021), and also regulates autophagy-related transcripts (Tang et al. 2021), thus influencing cell growth and survival (Fig. 4B).

Figure 4.

The complex regulation of cMyc downstream from mTORC1. (A) Upregulation of cMyc translation via increased eIF4A RNA helicase activity through the eIF4B–S6K pathway to resolve highly structured 5′ UTRs of Myc mRNA. (B) Enhanced cMyc activity through destabilization of MXD2 mRNA by mTORC1-mediated m⁶A modification. (C) Increased cMyc activity through the degradation of MAD1, the negative regulator of Myc, via mTORC1–S6K-mediated phosphorylation. (D) Induction of cMyc stability through the suppression of GSK3 nuclear translocation by mTORC1. (E) Growth factor or nutrient activation of AKT suppresses GSK3-mediated cMyc degradation through GSK3 phosphorylation. (Figure created with BioRender.com.)

Interestingly, mTORC1 regulates c-Myc through multiple mechanisms. mTORC1 enhances c-Myc protein translation by increasing eIF4A helicase activity via the mTORC1–S6K1–eIF4B pathway, which helps resolve the highly structured 5′ UTR of c-Myc mRNA (Fig. 4A; Wolfe et al. 2014). As noted above, mTORC1 promotes c-Myc protein stability by inhibiting GSK-3β nuclear location (Fig. 4D; Bautista et al. 2018). GSK-3β phosphorylates c-Myc at amino acid Thr58, thereby targeting it for degradation (Fig. 4D; Gregory et al. 2003). Insulin or other growth factors also inhibit GSK-3β by phosphorylation at Ser9 by several AGC kinases, including S6K1 (Zhang et al. 2006), thus stabilizing c-Myc (Fig. 4E; Fang et al. 2000). Additionally, mTORC1 boosts c-Myc's transcriptional activity by modulating its negative regulators through post-translational modifications. For instance, phosphorylation of MAD1 (MXD1), a cMyc antagonist, at Ser145 by mTORC1–S6K leads to MAD1's ubiquitination and degradation, thereby increasing c-Myc's transcriptional activity (Fig. 4C; Zhu et al. 2008). Moreover, the mTORC1–S6K1–WTAP axis facilitates the m⁶A modification of MXD2 mRNA, another cMyc antagonist, resulting in the destabilization of MXD2 mRNA and further enhancing Myc's transcriptional activity (Cho et al. 2021). These multifaceted regulatory inputs highlight the pivotal role of mTORC1 in orchestrating cancer cell growth and proliferation through the fine-tuning of Myc expression and activity (Fig. 4A–E).

5′ cap-dependent translation regulation by mTORC1

mTORC1 signaling stands as a central regulator of 5′ cap-dependent translation, a cornerstone process essential for cell growth, proliferation, and survival. Within this intricate regulatory network, pivotal players such as 4E-BP1, eIF4E, S6K1, PDCD4, and eIF4B act in concert to fine-tune protein synthesis in accordance with cellular demands (Hay and Sonenberg 2004; Ma and Blenis 2009). Critical to this process is 4E-BP1, a repressor of translation that binds to the 5′ cap-binding protein eIF4E and inhibits the formation of the 48S translation preinitiation complex (Gingras et al. 1999b). Evidence from our laboratory suggests that in growth factor- and nutrient-deprived conditions, inactive S6K1 is bound to the eIF3 complex (made up of 13 subunits, eIF3a–m). Upon mTORC1 activation, it binds to eIF3 and phosphorylates S6K1 at its hydrophobic motif site, which causes its dissociation (Holz et al. 2005) from the 43S complex. PDK1 then binds to this phosphorylated site, which facilitates PDK1-dependent phosphorylation of the activation loop, yielding an active S6K1 (Williams et al. 2000). eIF3, which is also bound to the 40S ribosome small subunit as part of the 43S preinitiation complex, is recruited to the 5′ cap complex, where mTORC1 is proposed to be in proximity to phosphorylate 4E-BP1 (phosphorylation at Thr37/46 and priming for phosphorylation of Thr70 followed by phosphorylation of Ser65) (Gingras et al. 1999a, 2001). Phosphorylation of 4E-BP1 promotes its release from eIF4E. This exposes a binding site for eIF4G (which is associated with additional translation initiation factors), eIF4A, and polyA binding proteins to form the eIF4F complex, which facilitates mRNA circularization and enhances the 40S:eIF3 complex association (Gingras et al. 2001; Hay and Sonenberg 2004). The resulting formation of the 48S translation preinitiation complex (PIC) allows ribosome scanning to find the translation initiation start codon, resulting in recruitment of the 60S subunit and additional translational initiation and elongation factors. (Gingras et al. 2001; Hay and Sonenberg 2004). Active S6K also phosphorylates eIF4B (Raught et al. 2004), a regulatory subunit of the mRNA helicase eIF4A, resulting in its recruitment into the PIC (Holz et al. 2005). However, the eIF4A/B helicase is suppressed by PDCD4, a translational inhibitor (Yang et al. 2003). S6K1 also phosphorylates PDCD4 (Dorrello et al. 2006), leading to its degradation through the ubiquitin E3 ligase β-TrCP (transducin repeat-containing protein), alleviating its inhibitory impact on translation initiation and augmenting protein synthesis efficiency, especially of mRNAs with highly structured 5′ UTRs such as c-Myc, ODC (ornithine decarboxylase), BCL2 (B-cell lymphoma 2), WTAP, and others (Fig. 2; Holz et al. 2005; Dorrello et al. 2006; Wolfe et al. 2014; Cho et al. 2021). Additionally, RSK, a kinase downstream from ERK in the Ras–ERK/MAP kinase pathway, can also activate mTORC1 and translation initiation via RSK-dependent phosphorylation of TSC2 (Roux et al. 2004; Ballif et al. 2005), eIF4B, and PDCD4, suggesting convergent pathways regulating mTORC1 and protein synthesis (Shahbazian et al. 2006).

Translation upregulation of the 5′ TOP motif containing mRNA by mTORC1

mTORC1 signaling is also pivotal for facilitating the translation of mRNAs harboring the 5′-terminal oligopyrimidine (TOP) motif, characterized by a unique pyrimidine sequence adjacent to the mRNA cap. La-related protein 1 (LARP1), identified as a key repressor of TOP mRNA translation, exhibits a dual binding capacity to both the m7Gppp cap moiety and the adjacent TOP motif within these transcripts (Levy et al. 1991; Hsieh et al. 2012; Thoreen et al. 2012; Philippe et al. 2020). Through this interaction, LARP1 obstructs the assembly of the eIF4F complex. Activation of mTORC1 initiates the translation of TOP motif-containing mRNAs by phosphorylating LARP1 (Ser689 and Ser692), thereby relieving its inhibitory influence on TOP mRNA translation (Fig. 2; Hong et al. 2017; Jia et al. 2021). Notably, LARP1 competes with eIF4G for TOP mRNA binding, further modulating the translation process. The association of LARP1 with mTORC1 via RAPTOR (Philippe et al. 2020) and its interaction with TOP mRNAs in an mTORC1-dependent manner underscore its significance as a critical repressor of TOP mRNA translation downstream from mTORC1 (Philippe et al. 2018). The complexity of the mTORC1–LARP1 relationship has recently been discussed (Berman et al. 2021).

mTORC1 modulates the 3′ UTR to enhance translation

mTORC1 activation also leads to the shortening of 3′ UTRs, promoting the formation of polysomes and increased protein production (Chang et al. 2015). Notably, mTORC1-mediated shortening of the 3′ UTRs appears to have a pronounced effect on the expression of genes involved in distinct cellular processes, such as protein folding and modification within the endoplasmic reticulum (ER), as well as protein degradation through the ubiquitin–proteasome system (Chang et al. 2015, 2018).

Ribosome biogenesis regulation by mTORC1 signaling

mTORC1 signaling also governs the intricate process of ribosome biogenesis, overseeing a cascade of events crucial for the synthesis and assembly of ribosomes, the cellular machinery responsible for protein production. This regulatory network spans various steps, including the transcription of ribosomal RNA (rRNA), the synthesis of ribosomal proteins, and the assembly of functional ribosomes (Chauvin et al. 2014). The TOP motif is present in the mRNA transcripts of all 79 human ribosomal protein genes, whose translation is tightly regulated by and heavily reliant on the activity of the mTORC1 complex (Iadevaia et al. 2008; Thoreen et al. 2012). Mechanistically, the mTOR–S6K1 axis facilitates the interaction of TIF1A with Pol I, a crucial enzyme responsible for rRNA transcription (Mayer et al. 2004). Moreover, mTORC1–S6K1 signaling also influences rRNA expression by facilitating the interaction between upstream binding factor (UBF) and selectivity factor 1 (SL1), a multiprotein complex involved in the initiation of rRNA transcription by Pol I (Tuan et al. 1999; Hannan et al. 2003). This coordinated regulation further enhances the activation of Pol I-mediated rRNA transcription, contributing to ribosome assembly. Additionally, mTORC1 regulates Pol III-mediated transcription by phosphorylating and inhibiting MAF1 (S60, S68, and S75), a key repressor of Pol III activity (Kantidakis et al. 2010; Michels et al. 2010; Shor et al. 2010). Inhibition of MAF1 leads to the enhanced transcription of 5S rRNA and transfer RNA (tRNA), essential components of the protein synthesis machinery (Wei et al. 2009). Notably, mTORC1's regulatory influence extends to both Pol I transcribed 45S rDNA and Pol III transcribed genes, including those critical for cellular growth and protein synthesis (Tsang et al. 2010).

Role of mTORC1 in metabolism

Glucose metabolism

As a master coordinator and integrator of processes dedicated to cell growth and proliferation, mTORC1-dependent regulation of RNA metabolism and protein synthesis is tightly coupled to its control of various metabolic pathways that provide the energy, redox state, and building blocks necessary to support normal cell growth. Glucose metabolism is a well conserved metabolic process that serves as a link to other anabolic pathways, supplying precursors for the synthesis of nucleotides, lipids, and amino acids. mTORC1 signaling enhances glucose consumption and aerobic glycolysis, largely through the action of two transcription factors: HIF1α and FOXK1/2. HIF1α is a short-lived protein under normal oxygen concentration (normoxia) but is stabilized in low-oxygen environments (hypoxia) and is predominantly responsible for glycolysis during oxygen scarcity (Hu et al. 2003; Lum et al. 2007). However, sustained levels of HIF1α are observed in cells with high mTORC1 activity, even under normoxic conditions. mTORC1 increases HIF1α expression by promoting the selective increase in translation of its mRNA owing to the presence of upstream regulatory signals in its 5′ UTR regulated by mTOR (Thomas et al. 2006; Dodd et al. 2015). Additionally, mTORC1 promotes mRNA expression of HIF1α via the transcription factor FOXK1. Upon mTORC1 activation, GSK3 is predominantly cytoplasmic and therefore does not efficiently phosphorylate and expel nuclear FOXK1/2, allowing it to promote HIF1α transcription. Additionally, FOXK1 binds to the promoter regions of many enzymes involved in glycolysis, thereby regulating aerobic glycolysis independently of, as well as together with, HIF1α (Fig. 5; He et al. 2018).

Figure 5.

The mTORC1 signaling pathway extensively controls cellular metabolism. This schematic illustrates the metabolic genes induced by mTORC1 within their respective pathways, based on our current understanding of mTORC1's regulation of glycolysis, the pentose phosphate pathway (PPP), serine/one carbon metabolism, and nucleotide metabolism. mTORC1 orchestrates this metabolic rewiring by regulating gene expression through multiple transcription factors, including Hif1a (yellow), FOXK (green), ATF4 (red), and SREBP (blue). (Figure created with BioRender.com.)

Increased glycolysis allows for the utilization of glycolytic by-products for de novo serine synthesis. HIF1α or ATF4 regulate multiple enzymes involved in the de novo serine synthesis pathway in response to mTORC1 signaling (Adams 2007; Ben-Sahra et al. 2016; He et al. 2018; Torrence et al. 2021), including PHGDH (phosphoglycerate dehydrogenase), PSAT1 (phosphoserine aminotransferase 1), and PSPH (phosphoserine phosphatase). The TCA cycle is one major destination of products of glycolysis. Although there is minimal evidence to suggest that mTORC1 signaling directly supports any enzymes of the TCA cycle, it is known to promote the flow of glutamine carbon into the TCA cycle via c-Myc, another transcription factor that broadly regulates metabolism in many cancer types (Dang 2016). The activation of mTORC1 leads to an increase in c-Myc levels by multiple mechanisms, as described in Figure 4. Given that c-Myc is a major regulator of glucose and glutamine metabolism (Dang 2016; Dejure and Eilers 2017), it is another crucial regulator of cellular metabolism in response to mTORC1 signaling. For example, c-Myc has been linked to upregulating the expression of glucose transporter GLUT1 (glucose transporter 1) and many glycolytic enzymes that contribute to the regulation of glycolysis (Osthus et al. 2000; Kim et al. 2004; Barfeld et al. 2015; Stine et al. 2015; Hu et al. 2017; Dong et al. 2020; Hoxhaj and Manning 2020; Fernandez et al. 2022).

Nucleotide synthesis

The mTORC1 pathway plays a pivotal role in facilitating the synthesis of both pyrimidines and purines, needed for DNA and RNA synthesis. Once glucose enters a cell, it undergoes phosphorylation and may either continue to break down through glycolysis or divert metabolic precursors into the pentose phosphate pathway (PPP). The PPP produces several metabolites, including ribose 5-phosphate, which is a crucial building block for nucleotides. mTORC1 enhances nucleotide synthesis by upregulating glucose-6-phosphate dehydrogenase (G6PD), a key enzyme that transforms glucose 6-phosphate into 6-phosphogluconate, which is regulated by the transcription factors SREBP1, HIF1α, or FOXK1 (Düvel et al. 2010; He et al. 2018). The production of nucleotides is essential for the growth of cells, particularly those that divide quickly and thus require an increased synthesis of these molecules.

The enzyme complex carbamoyl-phosphate synthase 2, aspartate transcarbamylase, and dihydroorotase, collectively known as CAD, is responsible for initiating the first three steps in the de novo synthesis of pyrimidine. Studies have indicated that the mTORC1/S6K1 pathway is instrumental in the direct phosphorylation of CAD at Ser1859, which either enhances dihydroorotase activity or promotes oligomerization of CAD, thereby enhancing the production of pyrimidines (Ben-Sahra et al. 2013; Robitaille et al. 2013).

One carbon metabolism is also essential for synthesizing de novo purines and pyrimidines by contributing carbon units. mTORC1 enhances one carbon metabolism through the activation of multiple transcription factors, thereby supporting the synthesis of nucleotides. The expression of methylenetetrahydrofolate dehydrogenase (NADP⁺-dependent) 2 (MTHFD2), an enzyme crucial for the mitochondrial one carbon metabolic pathway, is increased via mTORC1 signaling through ATF4 (Ben-Sahra et al. 2016) or FOXK1 (He et al. 2022). In addition, upon mTORC1 activation, HIF1α and FOXK1 also support the elevation of two additional key enzymes involved in mitochondrial one carbon metabolism: serine hydroxymethyltransferase 2 (SHMT2) and dehydrogenase (NADP⁺-dependent) 1-like (MTHFD1L) (He et al. 2018).

Additionally, mTORC1 has been shown to stimulate the formation and aggregation of “purinosomes” on the mitochondrial surface (French et al. 2016b). These purinosomes, which are sizable complexes of enzymes dedicated to the new synthesis of purines, are thought to improve the efficiency of this metabolic pathway. Upon mTORC1 activation, FOXK1 binds directly to the promoter region and facilitates the expression of several enzymes essential for purine synthesis (He et al. 2018). Finally, mTORC1 increases the eIF4E-mediated translation of phosphoribosyl pyrophosphate synthetase 2 (PRPS2), while cMyc also promotes its transcription and translation, with PRPS2 playing a pivotal role in converting ribose-5-phosphate (R5P) to 5-phosphoribosyl 1-pyrophosphate (PRPP), an essential step in the synthesis of both purines and pyrimidines (Fig. 5; Mannava et al. 2008; Cunningham et al. 2014).

Lipid synthesis

Lipid synthesis is crucial for energy storage, hormone production, and maintaining cell membrane integrity. mTORC1 plays a critical role by using various mechanisms to regulate this essential biological process. SREBPs are key players in lipid metabolism, linking mTORC1 signaling to lipid biosynthesis and metabolism (Porstmann et al. 2008). mTORC1 activates SREBPs through multiple mechanisms (Düvel et al. 2010; Owen et al. 2012). First, mTORC1 suppresses endosomal recycling that prevents cholesterol from reaching the ER. This leads to a reduction in ER cholesterol levels and triggers the activation of SREBP2, which subsequently moves to the nucleus and initiates the transcription of genes involved in cholesterol metabolism (Eid et al. 2017). Second, mTORC1 also regulates SREBP activity by managing the nuclear access of Lipin 1, which acts as a suppressor of SREBPs. Upon mTORC1 suppression, phosphorylation on Lipin 1 is decreased, leading to its entry into the nucleus, where it suppresses SREBP activity (Peterson et al. 2011). Third, mTOR phosphorylates CREB-regulated transcription coactivator 2 (CRTC2) at Ser136 and suppresses its inhibitory effect on SREBP1 maturation, which regulates lipid metabolism in the fed state and obesity (Han et al. 2015). Finally, mTORC1 signaling promotes the interaction between FAM120A and SREBP1 at the active gene promoters, which effectively connects the newly synthesized lipogenic gene transcripts from RNA polymerase II to the RNA-splicing machinery, streamlining the pathway from transcription to splicing of lipogenic genes (Lee et al. 2017; Cho et al. 2023). This process is mediated by the phosphorylation and activation of SR protein kinase 2 (SRPK2), a key regulator of RNA-binding splicing factors (Lee et al. 2017). Additionally, mTORC1 phosphorylates jumonji domain containing 1C (JMJD1C) at Thr505, enhancing its capacity to demethylate H3K9me2 at the promoter regions of lipogenic genes like fatty acid synthase (FASN) and acetyl-CoA carboxylase (ACC) and thereby also contributing to lipid synthesis (Viscarra et al. 2020).

Catabolic processes

Although mTORC1 enhances the synthesis of macromolecules, it also reduces their breakdown by suppressing autophagy. When mTORC1 is inhibited due to a lack of nutrients or through pharmacological means, autophagy is quickly initiated and serves as a compensatory mechanism, allowing the cell to recover nutrients from its own components during times of scarcity. mTORC1 regulates autophagy by several methods, notably through the post-translational modification and deactivation of ULK1, a key enzyme that initiates autophagy. The rapid phosphorylation of ULK1 at Ser757 by mTORC1 disrupts the ability of ULK1 to interact with AMP-activated protein kinase (AMPK), effectively inhibiting autophagy (Egan et al. 2011; Kim et al. 2011). The role of AMPK in autophagy regulation has recently been reviewed (Wang et al. 2022). mTORC1 also suppresses autophagosome maturation by directly phosphorylating PACER and WIPI2 at Ser157 and Ser395, respectively (Wan et al. 2018). Furthermore, rapamycin-induced acetylation of both ULK1 and PACER by lysine acetyltransferase 5 (KAT5) has been demonstrated to significantly enhance their capacity to initiate autophagy, a process that is reliant on the mTORC1/GSK3 signaling pathway, where GSK3 enhances the activity of KAT5 by directly phosphorylating it (Lin et al. 2012; Cheng et al. 2019). Finally, as mentioned above, mTORC1 can regulate autophagy through m⁶A modification of transcripts that regulate autophagy (Fig. 6; Tang et al. 2021).

Figure 6.

mTORC1 regulation of autophagy. mTORC1 regulates autophagy through both transcriptional and post-translational mechanisms. (Figure created with BioRender.com.)

mTORC1 also inhibits autophagy by controlling the expression of genes involved in the autophagy machinery and lysosome function. It has been reported that FOXK1/2 selectively attracts Sin3A–HDAC complexes to limit the acetylation of histone H4 and the expression of essential autophagic genes during nutrient deprivation or upon mTORC1 inhibition (Bowman et al. 2014). mTORC1 directly phosphorylates TFEB at multiple sites, including Ser122, Ser142, and Ser211 (Martina et al. 2012; Settembre et al. 2012; Vega-Rubin-de-Celis et al. 2017). Under nutrient-deprived conditions or when mTORC1 is pharmacologically inhibited, TFEB becomes dephosphorylated and translocates into the nucleus, where it promotes the expression of genes related to autophagy or lysosome formation. Pexophagy is a specific type of autophagy that involves the degradation of peroxisomes, cellular organelles involved in various metabolic functions such as fatty acid oxidation and the detoxification of reactive oxygen species. Upon nutrient deprivation or inhibition of mTORC1, the expression of peroxisomal biogenesis factor 2 (PEX2) is reported to increase and contributes to the regulation of pexophagy, though the mechanisms remain unclear (Aranovich et al. 2014). Ribophagy, which degrades ribosomes through autophagy, allows cells to recycle ribosomal components under certain conditions. The nuclear fragile X mental retardation-interacting protein 1 (NUFIP1) functions as a specific receptor for ribosome targeted autophagy that moves from the nucleus to autophagosomes and lysosomes to facilitate ribophagy upon mTORC1 suppression (Wyant et al. 2018).

Under some circumstances, mTORC1 inhibition can promote scavenging of extracellular proteins as a source of amino acids and other nutrients through macropinocytosis (Palm et al. 2015). This mechanism allows for the nonselective uptake of extracellular fluid, providing a way for cells to acquire nutrients and other substances from the microenvironment. The ubiquitin–proteasome system (UPS) serves as an essential cellular process for protein breakdown and regulation, crucial for preserving cellular balance. It has been observed that inhibiting mTOR activates UPS-mediated proteolysis. This activation facilitates the release of free amino acids in conditions of nutrient scarcity, suppressing the deceleration of cellular growth (Zhao et al. 2015). However, another group has found that activation of mTORC1 enhances UPS-dependent protein degradation in their experimental setting (Zhang et al. 2014), indicating that more work in this area is needed.

mTORC1 and disease therapy

The mTORC1 pathway acts as a pivotal signaling rheostat, assimilating a variety of extracellular signals (including the availability of growth factors, nutrients, energy, and oxygen) to ensure the maintenance of proper homeostasis and enable adaptation to the constantly changing microenvironment. Thus, it comes as no surprise that the dysregulation or improper activation of mTORC1 signaling, which can occur through environmental or genetic means, is associated with a range of diseases, including diabetes, obesity, cancer, aging, and neurodegenerative disorders (Saxton and Sabatini 2017; Liu and Sabatini 2020).

The mTORC1 signaling pathway is a major convergence point of numerous oncogenes and tumor suppressors, such as RAS, RAF, PI3KCA, AKT, PTEN, TSC1, and TSC2, which result in the chronic activation of mTORC1 signaling regardless of external growth stimuli (Ilagan and Manning 2016; Saxton and Sabatini 2017). Mutations in various components of the mTORC1 nutrient-sensing mechanism have also been identified in cancer progression. For example, a minority of glioblastomas exhibit loss-of-function mutations in GATOR1 elements (Saxton and Sabatini 2017), whereas mTORC1-activating RagC mutations have been notably prevalent in follicular lymphoma (Okosun et al. 2016). Nevertheless, high mTORC1 activity has been detected in >70% of cancer cases, making it a compelling target for the development of anticancer medications (Dobashi et al. 2011; Forbes et al. 2011; Li et al. 2014; Rusquec et al. 2020). In 2007 and 2009, the FDA approved the use of mTORC1 inhibitors or “rapalogs,” including temsirolimus and everolimus, respectively, for treating advanced kidney cancer. Unfortunately, rapalog monotherapy often results in the development of drug resistance and has not achieved the clinical success expected based on preclinical cancer models. Several combination therapy approaches are now being investigated.

Several hypotheses have been proposed to explain this resistance and the observed lack of efficacy. The first hypothesis emerged upon recognizing that, as allosteric inhibitors, rapalogs potently inhibit the phosphorylation of certain mTORC1 substrates but not all (Choo et al. 2008). More specifically, the phosphorylation of 4EBP1 remains largely unaffected by rapamycin in many cancer cells, unlike that of S6K1. Second, the reactivation of AKT signaling by rapalog treatment occurs by releasing the negative feedback on PI3K signaling, which enhances cell survival and contributes to resistance (Wang et al. 2007; Hsu et al. 2011; Yu et al. 2011). Third, suppression of nuclear GSK3 signaling decreases rapamycin sensitivity. Multiple groups have shown that either genetic or pharmaceutic inhibition of GSK3 confers resistance to rapamycin (Dong et al. 2005; Koo et al. 2015; He et al. 2019).

In an effort to address rapamycin resistance and to boost its therapeutic effectiveness, second- and third-generation mTOR inhibitors were developed. The “second-generation” compounds involve ATP-competitive catalytic inhibitors that target mTOR, effectively inhibiting both mTORC1 and mTORC2 pathways (Feldman et al. 2009; Thoreen et al. 2009). We have found that mTORC1/2 inhibition by ATP-competitive inhibitors in many cell lines promotes reorganization of integrin/focal adhesion kinase-mediated adhesomes as well as PI3K and Akt reactivation via an integrin/FAK/IGFR-dependent process (Yoon et al. 2017). The “third-generation” compound, known as “Rapalink,” combines a catalytic inhibitor with rapamycin, creating a linked compound designed to address resistance issues (Rodrik-Outmezguine et al. 2016). Although these inhibitors more effectively suppress the activity of mTOR complexes, their potential toxicity is currently being scrutinized in clinical studies (Wei et al. 2017; Graham et al. 2018).

TOR signaling plays a crucial role in the aging process across a wide range of organisms, such as yeast (Kaeberlein et al. 2005), worms (Vellai et al. 2003), flies (Kapahi et al. 2004), and mammals (Wu et al. 2013). It is important to note that while mTOR hypomorphic allele mice exhibited extended longevity in both genders, only female mtor^+/− mlst8^+/− or s6k1^−/− mice were long-lived, with a significant increase in mean life span relative to wild type (Selman et al. 2009; Lamming et al. 2012). Consistently, rapamycin stands out as a small molecule conclusively demonstrated to prolong life span across all these model organisms (Powers et al. 2006; Harrison et al. 2009; Bjedov et al. 2010; Robida-Stubbs et al. 2012). Although it is widely accepted that mTOR signaling is crucial in the aging process, the mechanisms by which it influences aging remain to be fully understood. One potential explanation is that suppressing mTORC1 activity may decelerate aging through the enhancement of autophagy. This process aids in the removal of damaged proteins and organelles like old or poorly functioning mitochondria, whose buildup is linked to aging and age-associated diseases. Another possible explanation involves mitochondrial respiration and oxygen consumption. It has been reported that in skeletal muscle tissues and cells, the mTOR inhibitor rapamycin decreases the expression of mitochondrial transcriptional regulators such as PGC-1α, estrogen-related receptor α, and nuclear respiratory factors, leading to a reduction in mitochondrial gene expression and oxygen consumption (Cunningham et al. 2007). However, it is worth noting that increased mitochondrial respiration has been observed in adipocyte-specific raptor knockout models (Polak et al. 2008). Beyond pharmaceutical suppression, dietary interventions such as caloric restriction (CR) also prolong life span across numerous organisms, likely by reducing mTOR activity (Kapahi et al. 2004; Kaeberlein et al. 2005; Hansen et al. 2007). Interestingly, RNA splicing factor (SFA-1), when overexpressed, is sufficient to extend life span and is required for life span extension by both dietary restriction and mTORC1 suppression (Heintz et al. 2017).

mTOR signaling also plays a pivotal role across various levels of brain functionality, from neural stem cell proliferation and circuit formation and maintenance to experience-driven plasticity and the control of complex behaviors (Lipton and Sahin 2014). Evidence points to dysfunctional mTORC1 signaling being linked to various neurological disorders. For example, in individuals with TSC due to mutations in TSC1 or TSC2, epilepsy is the prevalent symptom that is observed in 80%–90% of cases (Henske et al. 2016). Clinical studies have shown that treatment with rapalogs can effectively improve these symptoms (French et al. 2016a). Interestingly, abnormal mTORC1 activation has been implicated in autism (Winden et al. 2018), and proper regulation of mTOR signaling is crucial for the efficacy of certain treatments for depression (Duman 2018).

The role of mTORC1 in controlling autophagy is crucial, as malfunctioning autophagy also is a key factor in the development of neurodegenerative diseases such as Alzheimer's and Parkinson's diseases (Frake et al. 2015). Rapamycin is beneficial in the early phases of Alzheimer's disease for its protective effects. However, its use in later stages may worsen the condition, as the brain's lysosomal system, which is already significantly impaired, could be further compromised by rapamycin treatment (Carosi and Sargeant 2019). A diet limiting protein intake, which in turn reduces mTORC1 activity, has been demonstrated to decelerate Alzheimer's disease progression in mouse studies (Babygirija et al. 2024). Rapamycin can also effectively alleviate symptoms of PD MPTP-induced acute Parkinson's disease in a mouse model (Zhang et al. 2017)

Perspective

The mTOR pathway is now recognized as a critical regulator in sensing environmental conditions and then integrating and coordinating metabolism across all levels, from individual cells to the entire organism. Over the last 10 years, there has been remarkable progress in the comprehension of how mTOR signaling regulates physiological and metabolic functions either directly or through various effectors, as well as how mTOR signaling is cooperatively influenced by nutrients, metabolites, and growth factors. However, many aspects of its regulation, downstream effects, and role in human health and disease require more extensive investigation.

Although the role of the lysosome in sensing specific amino acids and metabolites is established, the mechanism by which other nutrients such as saturated or unsaturated fatty acids differentially activate lysosomal mTORC1 remains to be clarified. Furthermore, although significant advances have been made in understanding how mTORC1 is regulated by nutrients at the lysosome, our knowledge of regulation of mTORC1 activity at sites other than the lysosome is still limited. For example, signaling at the Golgi has been reported (Buerger et al. 2006; Fan et al. 2016; Hao et al. 2018). In one study, it was shown that the Golgi-resident amino acid transporter PAT4 recruits mTORC1 to Golgi membranes, which is crucial for sensing the availability of glutamine and serine in rapidly growing colorectal cancer cells (Fan et al. 2016), and the ER-resident protein protrudin can indirectly influence mTORC1 activity by controlling the localization of lysosomes (Raiborg et al. 2015). Additionally, there remain unanswered questions regarding differential signaling and the unique biological consequences that may occur depending on whether mTORC1 is signaling at the lysosome or from other cellular locations such as within the nucleus (Zhong et al. 2022).

Advancements in various biological techniques such as proteomics, metabolomics, gene expression profiling, and gene silencing have significantly broadened our knowledge of mTOR substrates and effectors, leading to a rapid expansion in this area of research. Nevertheless, despite mTORC1's vast influence on cellular processes, the number of identified rapamycin-sensitive mTORC1 direct substrates is relatively small compared with other major kinase signaling systems. This suggests the significance of its downstream effector kinases (S6K1, SRPK2, ULK1, and GSK3) in explaining mTORC1's extensive cellular impact. It is also important to recognize that most of mTORC1 substrates identified to date primarily operate or manifest their effects within the cytoplasm. Given that a significant portion of mTORC1's effects is linked to nuclear processes, identifying both direct targets of mTORC1 and their downstream effectors within the nucleus is key to enhancing our understanding of mTORC1 signaling dynamics. In addition to direct signaling by mTORC1 and S6K1 in the nucleus, uncovering the process by which mTORC1 regulates the nuclear translocation of GSK3 is anticipated to be a significant finding, shedding light on mTORC1's influence within the nucleus, given GSK3's broad involvement in cellular functions. Additionally, the phosphorylation by S6K1 and nuclear translocation of SRPK2 are likely to yield additional links between mTORC1 and mRNA processing.

As we look toward the next decade, the growing need for strategies to enhance health in an aging and increasingly obese population is becoming evident. Further research into mTORC1 regulation and signaling will lead to new therapeutic approaches for treating obesity, diabetes, cardiovascular disease, muscle wasting, aging, and other aging-associated diseases such as Alzheimer's and Parkinson's diseases. We anticipate that in addition to the emergence of new pharmacological strategies and targets, the role of dietary intervention is expected to provide exciting new information focusing on their potential to foster healthy aging.

Competing interest statement

J.B. is a cofounder of the Elikia, Inc., which targets aging-related metabolic dysfunction for the treatment of aging-related diseases. The other authors declare no competing interests.