mTOR Explained: The Cell’s Growth Switch

If AMPK is the cell’s fuel gauge, mTOR is its building permit. It integrates whether nutrients, growth factors and energy are all present, and if they are it licenses the expensive business of growth: making protein, making ribosomes, making more cell. If any one of them is missing, it stands down and the cell switches to recycling. That logic made mTOR one of the most studied enzymes in biology, and it also made it the centre of some of the most overstated claims in longevity writing.

This guide covers what mTOR is, the unlikely route by which it was found, the two complexes it forms, what switches it on and off, and what the famous mouse lifespan experiment actually reported — which is not what it is usually quoted as reporting.

What mTOR is

mTOR is an enzyme that attaches phosphate groups to other proteins. UniProt’s curated entry gives its recommended name as serine/threonine-protein kinase mTOR, with the enzyme classification EC 2.7.11.1, and HGNC places the gene in the phosphatidylinositol 3-kinase-related kinase family — a group of large kinases that look like lipid kinases but phosphorylate proteins.

It is worth stating plainly that mTOR is not a peptide, because it is routinely discussed alongside research peptides and the distinction gets blurred. Nothing in a peptide catalogue is mTOR; a handful of compounds sit upstream or alongside it. The same point applies to AMPK, mTOR’s opposite number, which is also an enzyme rather than a peptide.

The useful way to think about mTOR is as an AND gate. Growth is expensive and irreversible, so a cell should only commit to it when several independent conditions hold at once: amino acids available to build with, growth-factor signalling saying the tissue wants more cells, and enough ATP to pay for it. mTORC1 is the node where those signals are required to converge, and the elegance of the system is that failing any one input is enough to shut it down.

From a soil sample to a kinase

The whole field begins with a soil bacterium. In 1975, Vézina, Kudelski and Sehgal published the taxonomy and isolation of a new antifungal antibiotic they called rapamycin, with a companion paper on its fermentation and characterisation. The compound came from a soil sample collected on Rapa Nui — Easter Island — and is produced by Streptomyces hygroscopicus; as a 2016 review in Cold Spring Harbor Perspectives in Medicine puts it, rapamycin was discovered in the soil of Easter Island in the early 1970s and is a polyketide produced by that organism.

Rapamycin turned out to be more interesting as an immunosuppressant and growth inhibitor than as an antifungal, which raised the obvious question: what does it bind? The answer came from yeast genetics. In 1991 Heitman, Movva and Hall published “Targets for cell cycle arrest by the immunosuppressant rapamycin in yeast” in Science. As Heitman described the work in a 2015 retrospective, they isolated a large collection of rapamycin-resistant yeast mutants, and those studies revealed two novel proteins which they named TOR1 and TOR2, for target of rapamycin.

The mammalian counterpart followed three years later, found independently by several groups: Brown and colleagues in Nature in 1994, describing a mammalian protein targeted by the G1-arresting rapamycin-receptor complex; Sabatini and colleagues in Cell the same year, naming it RAFT1; and Sabers, Abraham and colleagues in the Journal of Biological Chemistry in 1995. Same protein, three names, which is a large part of why the nomenclature is still untidy.

A note on the name

You will see mTOR expanded two ways: mammalian target of rapamycin and mechanistic target of rapamycin. The usual story is that an official body renamed it from the first to the second. That story does not survive a look at the record.

HGNC’s entry for MTOR lists the previous approved symbols as FRAP, FRAP2 and FRAP1, and the previous approved name as “FK506 binding protein 12-rapamycin associated protein 1”. The word “mammalian” does not appear anywhere in HGNC’s list of previous names. The symbol changed to MTOR on 29 May 2009, and the current name, “mechanistic target of rapamycin kinase”, dates from 4 August 2017. We could not identify any publication that coined “mechanistic”; the only dedicated source we found on the question is a 2013 note by Michael Hall in Biochemical Society Transactions, which attributes the newer name to database curators rather than to the TOR research community, calls it needless confusion, and argues for keeping mTOR.

UniProt, pragmatically, lists both expansions as alternative names. For writing purposes the abbreviation is unambiguous and the expansion is not, which is a small lesson in checking even the parts of a sentence that look like background.

Two complexes, one kinase

mTOR does not act alone. The same kinase sits in two different protein complexes with different partners, different substrates and different drug sensitivities, and almost every confusion about mTOR traces back to treating them as one thing.

The split was established in yeast first, by Loewith and colleagues in 2002 in a paper whose title states the finding outright: two TOR complexes, only one of which is rapamycin sensitive, have distinct roles in cell growth control. Jacinto and colleagues showed the same for the mammalian system in 2004, reporting that mTOR complex 2 controls the actin cytoskeleton and is rapamycin insensitive.

The word doing the heavy lifting in all of this is acutely. mTORC2’s rapamycin insensitivity is a short-term property, and the fact that it does not survive chronic exposure is the single most consequential caveat in the whole mTOR-and-ageing literature.

What switches it on

mTORC1 is where three separate classes of information converge. The requirement that all of them be satisfied is what makes it a checkpoint rather than a simple relay.

Growth factors. Hay and Sonenberg’s 2004 review in Genes & Development sets out the route: growth-factor-induced activation of mTOR is mediated by PI3K, TSC2 is directly phosphorylated by Akt both in vitro and in vivo, and Akt activates mTOR at least in part through phosphorylating and inactivating TSC2. Insulin and IGF-1 are the archetypal inputs here — a 2020 review of IGF-1 in skeletal muscle describes IGF-1 increasing protein synthesis via the PI3K/Akt/mTOR pathway, with mTOR as a downstream target of Akt and Akt inhibiting the TSC1/TSC2 complex to relieve the brake on Rheb.

Amino acids. This is the arm worked out most recently and most thoroughly, largely by Sabatini’s laboratory. Sancak and colleagues showed in 2008 that the Rag GTPases bind raptor and mediate amino acid signalling to mTORC1, and in 2010 that the Ragulator-Rag complex targets mTORC1 to the lysosomal surface, which is necessary for its activation by amino acids. The sensors themselves came later: GATOR1 as a tumour suppressor complex with GAP activity for the Rag GTPases in 2013, Sestrin2 identified as a leucine sensor in 2016, and CASTOR1 as an arginine sensor the same year.

Energy. This is where AMPK enters, through two independent mechanisms. Inoki, Zhu and Guan showed in 2003 that TSC2 mediates the cellular energy response to control cell growth and survival, and Gwinn and colleagues showed in 2008 that AMPK phosphorylation of raptor mediates a metabolic checkpoint. The two kinases are therefore reciprocal: the fuel gauge directly restrains the building permit. Our AMPK guide covers the other side of that relationship.

What it does when active

Active mTORC1 pushes the cell towards making things and away from breaking things down.

On the synthesis side, a 1998 paper in PNAS established that the mammalian kinase directly phosphorylates the translational regulators p70 S6 kinase and 4E-BP1, with the rapamycin-sensitive S6K site at threonine 389 required for activity, and phosphorylation of 4E-BP1 disrupting its binding to the cap-binding factor eIF-4E. The net effect is more translation of the messages that build a growing cell.

On the degradation side, mTORC1 holds autophagy off. The 2011 paper by Kim, Kundu, Viollet and Guan — whose title is simply that AMPK and mTOR regulate autophagy through direct phosphorylation of ULK1 — is the reference point, and it captures the symmetry nicely: both kinases hit the same substrate, from opposite directions. mTORC1 also regulates TFEB, the transcription factor controlling lysosomal biogenesis, through a lysosome-to-nucleus signalling mechanism described in 2012.

What rapamycin actually inhibits

Rapamycin is usually described as a specific mTORC1 inhibitor. Two bodies of evidence say that description is too generous, and both matter for interpreting any claim built on rapamycin experiments.

It does not fully inhibit mTORC1. Choo and colleagues reported in 2008 that rapamycin suppresses S6K throughout treatment, but that 4E-BP1 phosphorylation returns within about six hours and the returned phosphorylation is rapamycin-resistant while remaining mTOR- and raptor-dependent. ATP-competitive mTOR inhibitors do block it, which is the point of the next generation of tools; Thoreen and colleagues’ 2009 paper is titled, in effect, for that finding — an ATP-competitive inhibitor reveals rapamycin-resistant functions of mTORC1. So “rapamycin inhibited mTORC1” and “mTORC1 was inhibited” are different statements.

It does not spare mTORC2 indefinitely. Sarbassov and colleagues showed in 2006 that prolonged rapamycin treatment inhibits mTORC2 assembly and Akt. The consequences in a whole animal were spelled out by Lamming and colleagues in Science in 2012, in a paper whose title carries the result: rapamycin-induced insulin resistance is mediated by mTORC2 loss and uncoupled from longevity. Chronic administration substantially impaired glucose tolerance and insulin action; the authors demonstrated that rapamycin disrupted mTORC2 in vivo; mTORC2 was required for the insulin-mediated suppression of hepatic gluconeogenesis; and decreased mTORC1 signalling was sufficient to extend lifespan independently of the changes in glucose homeostasis. A 2013 review by the same group lists hyperlipidaemia, decreased insulin sensitivity and glucose intolerance among the metabolic changes rapamycin produces.

The lifespan data, precisely

The most cited result in this field is Harrison and colleagues’ 2009 Nature paper, “Rapamycin fed late in life extends lifespan in genetically heterogeneous mice”, carried out within the US National Institute on Aging’s Interventions Testing Program — a multi-institutional programme run at three sites that tests candidate interventions in mice.

What the abstract reports is this: rapamycin extends median and maximal lifespan of both male and female mice when fed beginning at 600 days of age, and — the quantified claim — on the basis of age at 90% mortality, rapamycin led to an increase of 14% for females and 9% for males.

Those two figures are the ones everyone quotes, almost always as the increase in median lifespan. They are not. They are the increase in age at 90% mortality, and the abstract gives no percentage figure for median lifespan at all. If you see “rapamycin increased median lifespan by 14% in mice”, the number has been attached to the wrong endpoint. It is a small error that tells you something about how carefully the surrounding text was assembled.

The honest summary is that rapamycin extends lifespan in mice on more than one endpoint, including when started late in life, and that this is one of the better-replicated interventions in the ageing literature. It is also, per the 2012 Science paper above, achieved alongside metabolic side effects, and the authors of a 2023 review in Nature Aging put the translational position bluntly: excitement about rapamycin has outpaced rigorous evidence that rapalogs are both safe and efficacious for diseases of ageing in humans.

Where research peptides fit

mTOR comes up constantly in research-compound writing, and it is worth being exact about which compounds actually touch the pathway and how.

IGF-1 analogues sit directly upstream

This is the clearest connection in the catalogue. IGF-1 signals through the IGF-1 receptor to PI3K and Akt, and Akt relieves the TSC1/TSC2 brake on Rheb, activating mTORC1 — the route the 2020 skeletal-muscle review lays out. IGF-1 LR3 is a modified IGF-1 analogue, so anything studied with it is operating on the growth-factor arm of the diagram above rather than on mTOR itself. Our IGF-1 LR3 research guide covers what has and has not been shown with it.

MOTS-c acts on the other arm

MOTS-c is a mitochondrial-derived peptide whose published mechanism runs through the folate cycle to AMPK. Because AMPK restrains mTORC1 through raptor and TSC2, anything that activates AMPK indirectly restrains mTORC1 — but that is a downstream consequence, not an mTOR-targeting mechanism, and the MOTS-c evidence is cell and mouse work. See our MOTS-c guide and the mitochondrial peptides overview.

What is not an mTOR compound

Rapamycin and the rapalogs are prescription medicines and are not research reagents we supply. NAD⁺ and its precursors act on a different axis entirely — the coenzyme pool, covered in our NAD+ guide — and no source we reached establishes a direct NAD⁺-to-mTOR mechanism. Autophagy-related claims are a particular trap: mTORC1 inhibition does release autophagy, but a compound said to “boost autophagy” has not thereby been shown to act on mTOR.

The cancer context, briefly

Because mTORC1 licenses growth, the pathway above it is one of the most frequently altered in human tumours. A 2023 review in Signal Transduction and Targeted Therapy notes that PIK3CA mutations are frequently observed in a variety of cancers and result in activation of the PI3K/AKT/mTOR cascade. That is why the approved mTOR inhibitors are oncology and transplant drugs: the pathway was a cancer target long before it was an ageing target.

What’s established and what isn’t

Sourcing signalling reagents in Canada

If you are buying reagents for growth-signalling work, the documentation questions are the usual ones, with one addition: several of the molecules in this area are not peptides, and the right analytical evidence differs accordingly.

• Match the analysis to the molecule. A peptide or protein analogue such as IGF-1 LR3 is characterised by HPLC purity and a mass-spectrometry identity check; a coenzyme like NAD⁺ is a different analytical problem. See reading an HPLC chromatogram and reading a mass spectrum.
• Match the batch number on the certificate to the vial. Our COA guide covers what else a useful certificate contains.
• Account for net peptide content. Labelled mass includes counter-ions and water, so it is not peptide mass — our guide to net peptide content works through it.
• Store cold, dry and dark, lyophilized until needed. Longer polypeptides are the least forgiving of freeze-thaw cycling. See the storage guide, what lyophilized means and how peptides degrade.

IGF-1 LR3An IGF-1 analogue — the growth-factor arm upstream of mTORC1View product MOTS-cPublished mechanism runs through AMPK, mTORC1’s opposite numberView product NAD+A separate axis; no direct NAD+-to-mTOR mechanism is establishedView product

Frequently asked questions