Oliver's mTOR Atlas Evidence Platform
Mode
Reading level

mTOR Core · Lesson 02 · Core · 20 min

mTORC1 vs mTORC2

Why one kinase works inside two different machines

The question

Why does the cell use two mTOR complexes?

What you should be able to do

After this lesson you should be able to

  • Distinguish mTORC1 from mTORC2 by their defining subunits and by the substrates each one acts on.
  • Explain why a drug's failure to block part of mTOR signalling is what split the field into two complexes.
  • Interpret a claim that a treatment is ‘mTORC1-selective’ in terms of dose and duration rather than of the molecule alone.
  • Predict what a readout of one complex does when the other complex is manipulated.

Research skill Comparing mechanisms

The core idea

The same mTOR kinase is found in at least two distinct complexes, built around different partner proteins. mTORC1 contains RAPTOR; mTORC2 contains RICTOR and SIN1. Both also contain mLST8.

The partner protein is not decoration. RAPTOR helps recruit the substrates mTORC1 acts on, which is a large part of why the two complexes phosphorylate different targets even though the catalytic machinery is the same. Changing the partner changes what the kinase can reach.

The distinction has an experimental origin worth knowing: mTORC2 was found because a part of mTOR signalling did not behave the way rapamycin said it should. Acute rapamycin treatment inhibits mTORC1 readily; the RICTOR-containing complex was described as insensitive to that acute inhibition. A drug's failure to explain everything is what split the field into two complexes.

Worth noting how the discoveries actually landed, because single-name stories are usually wrong: RAPTOR was reported in 2002 by Kim and colleagues and, independently and in the same issue of Cell, by Hara and colleagues. In the same year Loewith and colleagues described two distinct TOR complexes in yeast, only one of them rapamycin-sensitive — the TORC1/TORC2 paradigm arrives from yeast genetics before the mammalian RICTOR paper in 2004. The Atlas's Timeline records the same parallel attribution.

The mTOR protein is found in two different assemblies. mTORC1 contains a partner called RAPTOR; mTORC2 contains RICTOR instead.

The partner is not decoration. It decides which other proteins mTOR can reach, so the two assemblies act on different targets even though the enzyme in the middle is the same.

The split was discovered because of a drug. Rapamycin blocks one assembly quickly and left the other one running — and that leftover activity is what told the field there were two.

The mechanism, in outline

mTORC1mTOR + RAPTOR + mLST8S6K1 · 4E-BP1 · ULK1translation, autophagy restraintmTORC2mTOR + RICTOR + SIN1 + mLST8AKT (Ser473) · SGK · PKCsurvival, metabolism, cytoskeletonshared subunits & feedback between them

mTORC1 — mTOR + RAPTOR + mLST8. Activated downstream of amino-acid availability and growth-factor signalling, positioned at the lysosome. Substrates include S6K1 and 4E-BP1, which is why its readouts are usually about translation, and ULK1, which is how it restrains autophagy.

mTORC2 — mTOR + RICTOR + SIN1 + mLST8. Its best-known activity is phosphorylation of AKT at Ser473, alongside other AGC-family kinases such as SGK and PKC. Through those substrates it is linked to survival signalling, aspects of metabolism, and cytoskeletal organisation.

The genetic dissection is the cleanest way to see that these are separate functional units: in mice, removing the mTORC1-defining subunit and removing the mTORC2-defining subunit produce different phenotypes, and loss of the mTORC2 subunit affects signalling to AKT while leaving parts of mTORC1 output in place.

mTORC1 responds to food and hormone signals, sits at the lysosome, and acts on proteins that control protein-making and recycling.

mTORC2 acts on a different set, including AKT, and is linked to survival signalling and to the shape of the cell.

The clearest evidence that they are separate came from mice: removing the protein that defines one assembly gives a different result from removing the protein that defines the other.

Interactive model

What each drug actually leaves running

Rapamycin
ATP-competitive inhibitor
mTORC1mTORC2S6K14E-BP1Akt/PKB

The two complexes and the three readouts that distinguish them, under two different kinds of inhibitor. Substrate links and effects are taken from the Atlas pathway model; the drug effects are the enumerated states, not a simulation.

p-S6K1 high · 4E-BP1 phosphorylated · AKT Ser473 present

The untreated cell. Both complexes are working, and the three readouts below them are the ones used to tell the complexes apart.

mTORC1
mTOR–Raptor–mLST8 (+PRAS40, DEPTOR); switched on only at the lysosome.The growth decision itself.Coincidence detector: nutrients supply location via the Rags, growth factors supply activation via Rheb. Neither alone is sufficient — the single most important idea in the pathway.
mTORC2
mTOR–Rictor–SIN1–mLST8; phosphorylates Akt, SGK1 and PKC.mTOR's second, less famous complex.Largely plasma-membrane associated and PI3K-responsive via the SIN1 PH domain; acutely rapamycin-insensitive, which is the cleanest way to separate mTORC1 from mTORC2 biology experimentally.
S6K1
Ribosomal protein S6 kinase; the standard readout of mTORC1 activity.mTORC1's best-known output kinase.T389 phosphorylation by mTORC1 is rapamycin-sensitive, which is why S6K1 became the field's default assay — and why the field long over-read rapamycin as a complete mTORC1 inhibitor.
4E-BP1
Translational repressor released from eIF4E upon multi-site phosphorylation.A cap on protein-making that mTORC1 removes.Only partially rapamycin-sensitive. This single fact explains the rapalog/Torin discrepancy and drove the whole ATP-competitive inhibitor programme.
Akt/PKB
AGC kinase requiring PIP3 recruitment plus two phosphorylations to be fully active.The main 'grow' relay from growth factors.T308 by PDK1 and S473 by mTORC2. Recruitment and activation are separate events — a distinction the older Atlas diagram blurred.

This is a simplified teaching model with a fixed set of states, not a simulation. It shows the direction each control pushes the pathway, not how much, how fast, or what any particular cell would do.

Every state of this model
RapamycinATP-competitive inhibitorReadoutWhat it means
noneabsentp-S6K1 high · 4E-BP1 phosphorylated · AKT Ser473 presentThe untreated cell. Both complexes are working, and the three readouts below them are the ones used to tell the complexes apart.
acuteabsentp-S6K1 low · 4E-BP1 largely still phosphorylated · Ser473 presentAcute rapamycin. S6K1 phosphorylation collapses, which is why rapamycin reads as an mTORC1 inhibitor — but 4E-BP1 phosphorylation is much less affected. mTORC1 is not simply off; part of its output continues.
prolongedabsentp-S6K1 low · 4E-BP1 partly phosphorylated · Ser473 reducedProlonged rapamycin. In many cell types the drug now also reduces mTORC2 signalling, apparently by interfering with assembly of new complexes — so a statement about selectivity that was true at six hours can stop being true at six days.
nonepresentp-S6K1 low · 4E-BP1 unphosphorylated · Ser473 lostAn ATP-competitive inhibitor blocks the kinase site itself, so both complexes stop and the rapamycin-resistant part of mTORC1 output stops with them. Comparing this column with the one above it is how the field learned that rapamycin was never the whole story.
acutepresentp-S6K1 low · 4E-BP1 unphosphorylated · Ser473 lostAdding rapamycin changes little once the active site is already blocked. The combination is drawn here to make the point that the two drugs are not two strengths of the same thing.
prolongedpresentp-S6K1 low · 4E-BP1 unphosphorylated · Ser473 lostBoth routes to loss of mTORC2 signalling at once. A result obtained under these conditions cannot attribute anything to either drug on its own.

Where the two-box model leaks

Three leaks are worth carrying with you.

Rapamycin sensitivity is not a clean partition. Acute treatment mainly inhibits mTORC1, but prolonged treatment can reduce mTORC2 signalling in some cell types, apparently by interfering with assembly of new complexes. 'mTORC2 is rapamycin-insensitive' is shorthand for an acute experiment, not a permanent property.

The complexes are wired to each other. mTORC1 output feeds back onto upstream signalling — inhibiting mTORC1 can relieve that feedback and change AKT signalling, which is an mTORC2 readout. So a manipulation aimed at one complex can move the other without ever touching it directly.

Substrate lists are not sealed. They come from the experiments done so far, in the cell types studied so far. Treat them as the current state of a catalogue, not as a definition.

Three things spoil the neat two-box picture, and all three matter later.

Rapamycin blocks mTORC1 quickly, but given for long enough it can reduce mTORC2 as well — so 'this drug only hits mTORC1' describes an experiment, not the drug.

The two assemblies also affect each other, so pushing on one can move a measurement of the other. And the lists of what each one acts on are a record of experiments done so far, not a finished catalogue.

Reading the tiers on this lesson's evidence

Every study cited below sits at tier D or C — mechanistic, in-vitro or animal work. That is the honest state of the complex-level evidence: the architecture of mTORC1 and mTORC2 is described from biochemistry and mouse genetics, not from human trials. Nothing in this lesson is a claim about what happens in a person.

What does the evidence say?

These are Atlas studies, with the Atlas's own evidence tier. Each card links to the full record — nothing here restates it.

How cleanly mTORC1 can be inhibited without touching mTORC2 over long periods, in a whole animal, is an open question rather than a solved engineering problem — and it is the question standing between the mouse lifespan results and any human use.

Work through it

Each of these asks you to commit to something — a prediction, a reading of two studies, a design — before it answers. Everything here is in the page, so nothing is lost if you would rather just read it.

Predict → observe → explain

Rapamycin for six days, not six hours

Predict — commit before you look

A cell line is treated with rapamycin continuously for several days and you measure AKT phosphorylation at Ser473 — an mTORC2 site. Predict what you see, given that rapamycin is usually described as mTORC1-selective.

Observe — what was actually measured

Multiple human and mouse cell lines treated with rapamycin over short and long periods, with mTORC2 assembly and AKT phosphorylation followed over time.

Short treatment affected mTORC1 signalling only. Prolonged treatment reduced mTORC2 in many of the cell lines and cut AKT signalling with it.

Study page →
Explain

Both of the first two answers describe something real, which is what makes this worth doing rather than reading. Feedback relief can raise AKT signalling after acute mTORC1 inhibition (Lesson 08), and prolonged exposure can lower it by a different mechanism entirely. Which one you observe depends on the duration, the cell type and when you look — so 'rapamycin spares mTORC2' describes a protocol rather than the molecule, and two papers can disagree without either being wrong.

What this evidence supports

  • That the selectivity of rapamycin depends on how long it is present, in the cell lines tested
  • That an assembly effect is a distinct mechanism from active-site inhibition, with a different time course

What it does not establish

  • That the same happens in every cell type — the effect was not uniform across the lines tested
  • What dose and duration in a person would do, which is a separate literature with its own evidence tiers
Show the expected answer

C. Reduced in many cell types — prolonged exposure interferes with the assembly of new mTORC2

Observe — what was actually measured

Multiple human and mouse cell lines treated with rapamycin over short and long periods, with mTORC2 assembly and AKT phosphorylation followed over time.

Short treatment affected mTORC1 signalling only. Prolonged treatment reduced mTORC2 in many of the cell lines and cut AKT signalling with it.

Study page →
Explain

Both of the first two answers describe something real, which is what makes this worth doing rather than reading. Feedback relief can raise AKT signalling after acute mTORC1 inhibition (Lesson 08), and prolonged exposure can lower it by a different mechanism entirely. Which one you observe depends on the duration, the cell type and when you look — so 'rapamycin spares mTORC2' describes a protocol rather than the molecule, and two papers can disagree without either being wrong.

What this evidence supports

  • That the selectivity of rapamycin depends on how long it is present, in the cell lines tested
  • That an assembly effect is a distinct mechanism from active-site inhibition, with a different time course

What it does not establish

  • That the same happens in every cell type — the effect was not uniform across the lines tested
  • What dose and duration in a person would do, which is a separate literature with its own evidence tiers
Compare the evidence

A drug and a knockout, asking the same question

D An ATP-competitive mammalian target of rapamycin inhibitor reveals rapamycin-resistant functions of mTORC1D Ablation in mice of the mTORC components raptor, rictor, or mLST8 reveals that mTORC2 is required for signaling to Akt-FOXO and PKCalpha, but not S6K1
Model systemMouse/human cellsKnockout mice (raptor/rictor/mLST8)
PerturbationChemistry in mouse and human cells: an ATP-competitive inhibitor that blocks the mTOR active site, compared side by side with rapamycin.Genetics in mice: deletion of raptor, rictor or mLST8, removing a complex rather than inhibiting a kinase.
ReadoutWhich mTORC1 outputs stop. Rapamycin left several running — notably 4E-BP1 phosphorylation — that the active-site inhibitor did not.Viability and which signalling survives — loss of rictor or mLST8 removed signalling to AKT while sparing S6K1.
What do both studies support?

That mTORC1 and mTORC2 are genuinely separable functional units rather than two names for one activity, and that neither the drug nor the gene deletion divides the pathway exactly where the textbook diagram does.

Where do they differ?

In what counts as 'inhibited'. The chemical experiment leaves the complex intact and blocks catalysis, so it can distinguish outputs by how sensitive they are — which is how the rapamycin-resistant functions became visible at all. The genetic experiment removes a subunit, so the complex never exists, and it is answering a question about requirement over a lifetime rather than about activity right now. A function that is compensated during development will look dispensable in the knockout and essential under the drug.

What experiment would help next?

The two designs converge when the genetic removal is made acute and reversible in an adult animal, and the drug is used at a dose whose target engagement has been measured rather than assumed — then 'required' and 'inhibited' are being asked on the same timescale.

Scientific caution

What ‘mTORC1-selective’ can and cannot mean

What this evidence supports

  • That the two complexes differ in composition and in substrates, and that the difference is visible both chemically and genetically
  • That rapamycin blocks some mTORC1 outputs much more completely than others

What it does not establish

  • That any compound is selective as a property independent of dose, duration and cell type
  • That a change in one complex's readout was caused by acting on that complex, since the two are wired to each other
  • That the substrate lists in this lesson are complete — they are the catalogue of experiments done so far
Why?

Every study cited in this lesson sits at tier D or C: biochemistry, cell work and mouse genetics. That is the honest state of complex-level evidence. Selectivity in particular is a claim about an experiment, and the experiment has a duration — which is exactly the thing summaries drop.

Think

If mTORC1 and mTORC2 are distinct complexes, why might manipulating one still change the other?

Look for something they share, and something one of them sends backwards.

Think first, then reveal

At least three routes. (1) They share subunits — mTOR itself and mLST8 — so a limited pool can be redistributed between complexes. (2) mTORC1 output feeds back onto upstream insulin/IGF-1 signalling; relieving mTORC1 activity relieves that feedback and raises signalling that mTORC2 reads out through AKT. (3) A drug is not a scalpel: prolonged rapamycin can affect assembly of new mTORC2. This is why 'mTORC1-selective' is a claim about a dose and a duration, not about a molecule.

You read a paper reporting that a compound is 'mTORC2-sparing'. What would you want to know before accepting that?

Think about what was measured, in what, for how long.

Think first, then reveal

At minimum: which readout was used for mTORC2 (usually AKT Ser473, which is one substrate among several), in which cell type or tissue, at what dose, and after how long. Sparing at 24 hours in one cell line is a different claim from sparing after weeks of dosing in an animal — and the second is what matters for anything longevity-related.

Check yourself

Three questions, easiest first. Nothing is recorded and there is no pass mark — the explanation after each answer is the part worth reading.

Question 1 · Warm-upWhich subunit is the one that defines mTORC1 rather than mTORC2?

Show the answer

C — RAPTOR. RAPTOR is the mTORC1-defining partner and RICTOR and SIN1 belong to mTORC2. mLST8 is the distractor worth remembering, because it sits in both complexes.

Question 2 · Step up‘mTORC2 is rapamycin-insensitive.’ What is that sentence actually describing?

Show the answer

B — The result of acute treatment — prolonged rapamycin can still reduce mTORC2 signalling in some cell types by interfering with assembly of new complexes. It is shorthand for one experimental design: short exposure. Change the duration and the statement can stop holding, which is why selectivity claims always need a dose and a time attached.

Question 3 · HarderYou inhibit mTORC1 in a cell line and see AKT Ser473 phosphorylation rise. Which reading fits this lesson best?

Show the answer

C — Removing mTORC1 output also removes the feedback it exerts on upstream insulin/IGF-1 signalling, and that upstream signalling is read out through AKT. This is the practical consequence of the two complexes being wired together: a manipulation aimed at one of them can move a readout of the other without ever touching it. Ser473 is an mTORC2 site, not an mTORC1 site.

Go deeper

Follow a guided route through the mechanism:

Concepts introduced in this lesson

Protein complex composition · Substrate recruitment · Acute vs chronic drug effect · Genetic dissection (subunit knockout)