mTOR Core · Lesson 03 · Core · 25 min
Rheb — The Critical Regulator
The question
How does a small GTPase help control mTORC1?
What you should be able to do
After this lesson you should be able to
- Trace the growth-factor branch from AKT to mTORC1 and state the sign of every step.
- Predict what mTORC1 output does when Rheb is held in its GDP-bound state, and justify the prediction from the mechanism rather than from memory.
- Distinguish what the amino-acid branch decides (where mTORC1 is) from what the growth-factor branch decides (whether the activator it meets is loaded).
- Evaluate what an overexpression experiment can and cannot establish about a protein at its normal level.
- Propose an experiment that would separate nucleotide state from localisation as explanations of the same result.
Research skill Causal reasoning
The core idea
Rheb is a small switch protein. It carries either GTP or GDP, and it behaves differently depending on which. Only the GTP form switches mTORC1 on.
What controls the switch is the TSC complex. It speeds up the step that turns Rheb from GTP to GDP — so it works as a brake.
Growth-factor signals do not switch mTORC1 on directly. They weaken the brake, which leaves more Rheb in its GTP form. Read that twice: it is a double negative, and it is easy to get backwards.
The mechanism, in outline
The chain is: growth factors → AKT ⊣ TSC complex ⊣ Rheb-GTP → mTORC1. Two of those steps are inhibitions, one after the other.
Said slowly: AKT holds back TSC, TSC holds back Rheb, so more AKT means more active Rheb and more mTORC1.
Other signals reach the same brake, including the one that reports low energy. That makes the TSC complex a meeting point rather than a simple relay.
The switchboard: three controls, one output
A simplified model of the growth-factor branch: AKT, the TSC complex, Rheb and mTORC1, with S6K1 and 4E-BP1 as the readouts. The Atlas model also carries an indirect TSC-to-mTORC1 link, which is left out here because it runs through the very step this lesson is about.
p-S6K1 high · 4E-BP1 phosphorylated
The coherent fed state. Growth factors activate AKT, AKT weakens the TSC brake, Rheb stays GTP-loaded and mTORC1 is active — so S6K1 is phosphorylated and 4E-BP1 releases its grip on eIF4E.
- 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.
- TSC1/TSC2
- TSC1–TSC2–TBC1D7 complex; a GAP that switches Rheb off.The pathway's master brake.Integrates Akt, AMPK, ERK/RSK, GSK3 and REDD1 inputs. Regulation is substantially about lysosomal recruitment, not only phosphorylation-driven activity change.
- Rheb
- Small GTPase; GTP-loaded Rheb allosterically activates mTORC1.The one thing that actually switches mTORC1 on.Realigns the mTOR active site; the convergence point of the entire growth-factor arm and the reason localisation alone is not activation. Note a declared simplification in this map: Rheb is farnesylated and distributes across the endomembrane system, with a substantial ER and Golgi pool, and which pool supplies the activating Rheb is still debated. It is drawn on the lysosomal band because that is where it meets mTORC1.
- 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.
- 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.
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
| Growth factors | TSC complex | Rheb | Readout | What it means |
|---|---|---|---|---|
| ON | inhibited | GTP | p-S6K1 high · 4E-BP1 phosphorylated | The coherent fed state. Growth factors activate AKT, AKT weakens the TSC brake, Rheb stays GTP-loaded and mTORC1 is active — so S6K1 is phosphorylated and 4E-BP1 releases its grip on eIF4E. |
| ON | inhibited | GDP | p-S6K1 low · 4E-BP1 unphosphorylated | This is the combination worth sitting with. The hormone is present, the brake is off — and output is still low, because the thing mTORC1 actually reads is the nucleotide state of Rheb. Everything upstream matters only through this step. |
| ON | active | GTP | p-S6K1 high · 4E-BP1 phosphorylated | A brake that is engaged while Rheb stays loaded. A cell does not hold this for long, and that is the point: the TSC complex is a GAP, so it sets the rate at which Rheb unloads rather than switching it off instantly. |
| ON | active | GDP | p-S6K1 low · 4E-BP1 unphosphorylated | Growth-factor signalling arrives, the brake holds anyway, Rheb unloads and output stops. In an intact cell this is the transient you would see on the way down after AKT signalling falls. |
| OFF | inhibited | GTP | p-S6K1 high · 4E-BP1 phosphorylated | No hormone, no brake, full output. This is close to what a cell lacking a working TSC complex looks like — mTORC1 active without the permission signal that normally licenses it, which is the situation in tuberous sclerosis (Lesson 04). |
| OFF | inhibited | GDP | p-S6K1 low · 4E-BP1 unphosphorylated | Nothing arriving and nothing loaded. Removing the brake changes nothing here, which is another way of seeing that the brake acts on Rheb, not on mTORC1. |
| OFF | active | GTP | p-S6K1 high · 4E-BP1 phosphorylated | Loaded Rheb overrides an engaged brake for as long as it stays loaded. Read this together with the state above it: the brake and the hormone are both upstream of the same single question. |
| OFF | active | GDP | p-S6K1 low · 4E-BP1 unphosphorylated | The resting or starved state: no permission signal, brake engaged, Rheb unloaded, mTORC1 quiet and autophagy no longer restrained. |
Why location matters as much as state
Being in the right place matters as much as being switched on. Rheb sits on membranes, especially the lysosome. mTORC1 has to be brought there, and amino acids are what bring it.
So the two branches do different jobs: food decides where mTORC1 is, hormones decide whether the switch it meets there is on. Neither is enough alone.
What is still uncertain
How much of this is about the switch, how much about the location, and how much about how many copies of each protein a cell has, is not settled.
Many of the classic experiments used far more of a protein than a cell normally makes. That shows what a protein can do, which is weaker evidence than showing what it usually does.
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.
TSC2 is a GAP for Rheb; loss of TSC2 raises Rheb-GTP and constitutively activates mTOR.
Study page →Insulin activates Rheb-GTP, inhibited by TSC1/2, positioning Rheb as the direct upstream activator of TOR.
Study page →Rheb promotes cell growth as a component of the insulin/TOR network in Drosophila.
Study page →Akt directly phosphorylates and inactivates TSC2, disrupting the TSC1-TSC2 complex and releasing its inhibition of mTOR - the link between growth-factor/insulin signaling and mTORC1 activation.
Study page →Identifies the Rag GTPase family as the missing link that lets mTORC1 sense amino acids by controlling whether mTOR is positioned near its activator Rheb.
Study page →Lysosomal recruitment of TSC2 is a universal response to cellular stress that inhibits mTORC1.
Study page →The relative contribution of nucleotide state, localisation and protein abundance to mTORC1 activation is unresolved, and several foundational experiments relied on overexpression rather than endogenous levels.
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.
Rheb is held in its GDP-bound state. What happens downstream?
Amino acids are plentiful, so mTORC1 is recruited to the lysosome as usual. But Rheb is held predominantly in its GDP-bound state. Commit to a prediction for S6K1 and 4E-BP1 phosphorylation before you look.
Biochemistry in mammalian cells: TSC2's GAP activity assayed directly on Rheb, with the mTOR pathway read out through the phosphorylation of its substrates.
Raising TSC2 GAP activity shifts Rheb toward its GDP-bound state and lowers mTOR signalling; Rheb is the direct target of that GAP activity rather than a bystander.
Study page →Recruitment and activation are two different requirements. The amino-acid branch delivers mTORC1 to the membrane where Rheb lives; the nucleotide state of Rheb decides whether the kinase it meets there is switched on. Hold Rheb in the GDP state and the first requirement is still satisfied while the second is not — so phosphorylation of S6K1 and 4E-BP1 falls even though nothing about the nutrient supply changed.
What this evidence supports
- That Rheb-GTP is the proximal input to mTORC1 in these systems, and that TSC2 acts on Rheb directly rather than on mTORC1
- That being in the right place is not sufficient for mTORC1 activity
What it does not establish
- How much of mTORC1 regulation in an intact tissue runs through nucleotide state rather than localisation or protein abundance
- That the relationship is quantitatively the same in every cell type, or over a chronic rather than an acute timescale
Show the expected answer
B. Output falls — mTORC1 arrives at the lysosome but the activator waiting there is in its off state
Observe — what was actually measuredBiochemistry in mammalian cells: TSC2's GAP activity assayed directly on Rheb, with the mTOR pathway read out through the phosphorylation of its substrates.
Raising TSC2 GAP activity shifts Rheb toward its GDP-bound state and lowers mTOR signalling; Rheb is the direct target of that GAP activity rather than a bystander.
Study page →Recruitment and activation are two different requirements. The amino-acid branch delivers mTORC1 to the membrane where Rheb lives; the nucleotide state of Rheb decides whether the kinase it meets there is switched on. Hold Rheb in the GDP state and the first requirement is still satisfied while the second is not — so phosphorylation of S6K1 and 4E-BP1 falls even though nothing about the nutrient supply changed.
What this evidence supports
- That Rheb-GTP is the proximal input to mTORC1 in these systems, and that TSC2 acts on Rheb directly rather than on mTORC1
- That being in the right place is not sufficient for mTORC1 activity
What it does not establish
- How much of mTORC1 regulation in an intact tissue runs through nucleotide state rather than localisation or protein abundance
- That the relationship is quantitatively the same in every cell type, or over a chronic rather than an acute timescale
What this lesson's evidence supports, and what it does not
What this evidence supports
- Rheb contributes to mTORC1 activation, and its GTP-bound state is the form that does it
- The TSC complex acts as a GAP for Rheb, which is how growth-factor and stress signals reach the pathway
- Localisation is part of the mechanism rather than a detail drawn on top of it
What it does not establish
- That Rheb is the only determinant of mTORC1 activity
- That the mechanism has the same weight in every cell type, or at every stage of the cell cycle
- That what was measured after overexpression describes what the endogenous protein normally does
Why?
Several of the foundational experiments in this lesson raised the amount of a protein far above its normal level in order to see its effect. That design is informative about what a protein can do and much weaker evidence about what it normally does — and the studies below sit at tier D and C, which marks the design, not the quality.
Where this stops being settled
Sensor-selective geroprotection without the metabolic penalty
What we know
The inputs reaching mTORC1 are separable at the molecular level: the amino-acid branch acts through the Rag GTPases and recruitment, the growth-factor branch through the TSC complex and the nucleotide state of Rheb.
Lowering mTORC1 signalling pharmacologically extends lifespan in several laboratory species, and the same treatment carries metabolic costs.
What we don't know
Whether the geroprotective part of that effect can be obtained by acting on one input branch rather than on the kinase itself.
How much of mTORC1 regulation in an intact mammalian tissue is nucleotide state, how much is localisation, and how much is the amount of protein present — the three are hard to separate in a living animal.
Competing interpretations
One reading: the branches are genuinely separable, so a sensor-directed intervention could reproduce the benefit without the metabolic penalty, which is the hypothesis the Atlas carries.
The other: the branches converge on one output, and anything that lowers that output enough to matter for ageing will carry the same costs, whichever door it comes through.
What would resolve this?
A mammalian experiment that lowers signalling through one input branch alone — genetically rather than with a drug, to avoid changing several things at once — and then measures both a lifespan or healthspan endpoint and the metabolic readouts in the same animals, in both sexes.
Supporting studies
TSC2 is a GAP for Rheb; loss of TSC2 raises Rheb-GTP and constitutively activates mTOR.
Study page →Lysosomal recruitment of TSC2 is a universal response to cellular stress that inhibits mTORC1.
Study page →Identifies the Rag GTPase family as the missing link that lets mTORC1 sense amino acids by controlling whether mTOR is positioned near its activator Rheb.
Study page →Design an experiment: does Rheb localisation influence mTORC1 activation?
The question is whether where Rheb is matters on top of what nucleotide it carries. Choose one option in each row, then submit. There is no single right design — the feedback tells you what each choice buys you and what it costs.
Feedback on your design
See the feedback on every choice
Model system
Cultured cell line. Fast, cheap and easy to perturb, which is why most of this lesson's evidence comes from one. The cost is that a line has been selected for growth in a dish, its signalling often runs harder than a primary cell's, and nothing you measure is automatically a statement about tissue.
Primary cells. Closer to normal physiology and still tractable. Expect more variability between preparations, so plan for more replicates; and be explicit about the donor or animal, because primary cells from different sources are not interchangeable.
Animal model. The only option that lets a tissue-level answer emerge, and the only one where the endocrine context is real. It is also the slowest, and a whole-animal manipulation of Rheb changes many cells at once — so pair it with a tissue-specific approach if you want to attribute the effect.
Perturbation
Rheb-GDP (GAP-insensitive off state). Directly tests the nucleotide arm of the question. Its weakness is that it answers the state question, not the location question — a GDP-locked Rheb sitting in the right place still tells you nothing about whether the place mattered.
Rheb-GTP (constitutively loaded). Useful as the mirror image, and the classic way the field showed sufficiency. Watch the level: if the loaded protein is also overexpressed, you are testing two things at once and the result speaks to what Rheb can do rather than what it does.
Lysosomal targeting of Rheb. This is the perturbation that actually addresses the question you asked, because it changes location while leaving the nucleotide cycle intact. Add the reciprocal version — targeting Rheb away from the lysosome — or a positive result stays ambiguous between 'location matters' and 'more Rheb near mTORC1 matters'.
No perturbation (observation only). Correlational by construction. You can describe where Rheb is and how active mTORC1 is, and you cannot separate cause from consequence — useful as a first description, not as an answer.
Readout
p-S6K1. The standard readout and the one most easily over-read. It reports one substrate of one complex at one moment, and it is the substrate that responds most readily to partial inhibition — so it can move while other mTORC1 outputs do not.
p-4E-BP1. A more demanding readout, and a better one if you want to claim the complex as a whole changed: 4E-BP1 phosphorylation is less sensitive to partial inhibition than S6K1. Reading both is better than choosing between them.
mTOR localisation by imaging. The right readout if the claim is about place, and the one that catches a result the blots would miss. It measures where the kinase is, not whether it is working — so pair it with a phospho-readout or the design proves location without output.
Control
Wild-type Rheb, same expression level. The strongest control here, because it holds the amount of protein constant and varies only the thing you meant to vary. Say how you matched the level — a claim about equal expression needs a measurement.
Untreated cells. Necessary but not sufficient: it controls for the baseline and not for the manipulation itself, so any effect of the tag, the vector or the handling lands in your result.
Vehicle only. The right control when the perturbation is a compound, and beside the point when it is genetic. Matching the control to the kind of perturbation is half of what a reviewer checks.
What no version of this design can establish
- No version of this design separates the amount of Rheb from its location unless expression level is measured and matched
- A phosphorylation readout at a single time point cannot distinguish a lower steady state from a delayed one — a time course can
- None of these versions establishes what the same manipulation would do in an intact tissue with an intact endocrine system
TSC2 is a GAP for Rheb; loss of TSC2 raises Rheb-GTP and constitutively activates mTOR.
Study page →Identifies the Rag GTPase family as the missing link that lets mTORC1 sense amino acids by controlling whether mTOR is positioned near its activator Rheb.
Study page →Lysosomal recruitment of TSC2 is a universal response to cellular stress that inhibits mTORC1.
Study page →Think
What would you expect if Rheb were held predominantly in its GDP-bound state?
Work forwards from the arrow, then ask what else in the cell would notice.
Think first, then reveal
The direct expectation is low mTORC1 activity even when amino acids are plentiful — mTORC1 would be recruited to the lysosome and find no active activator waiting. Downstream, reduced S6K1 and 4E-BP1 phosphorylation, reduced cap-dependent translation, and released restraint on ULK1 and autophagy. The second-order expectation is more interesting: because mTORC1 output feeds back onto upstream insulin signalling, chronically low mTORC1 would also change AKT signalling — so the cell would not look like 'the same cell with one arrow deleted'.
What experiment could distinguish whether mTORC1 regulation here depends mainly on nucleotide state, on localisation, on protein abundance, or on a combination?
Each candidate implies a different thing you would have to hold constant.
Think first, then reveal
There is no single decisive experiment, which is itself the answer worth taking away. A serious attempt separates the variables: nucleotide-state mutants of Rheb locked toward GTP or GDP at endogenous expression levels (knock-in rather than overexpression, so abundance is held constant); forced targeting of mTORC1 to the lysosome independently of amino acids, to ask whether localisation alone is sufficient; and quantitative measurement of Rheb, TSC2 and mTOR protein amounts in the same cells, since a change in any of the three could produce the same readout. Combining them under matched conditions is what the field has actually done piecemeal over two decades — and the pieces were done in different cell types, which is part of why the question is still live.
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 form of Rheb is the direct positive input to mTORC1 in the standard model?
Rheb is a switch protein whose two nucleotide states behave differently, and it is the GTP-bound state that activates the kinase. Everything the TSC complex does is about how much Rheb-GTP there is.
Show the answer
B — Rheb-GTP. Rheb is a switch protein whose two nucleotide states behave differently, and it is the GTP-bound state that activates the kinase. Everything the TSC complex does is about how much Rheb-GTP there is.
Question 2 · Step upTSC2 is a GAP for Rheb, and AKT inhibits the TSC complex. What is the net effect of AKT activity on mTORC1?
A double negative is easy to misread, so it is worth saying slowly: AKT inhibits TSC, TSC inhibits Rheb, so AKT raises Rheb-GTP. Note also that AKT acts on TSC2, not on Rheb.
Show the answer
B — Higher activity, because inhibiting the brake leaves more Rheb in its GTP-bound state. A double negative is easy to misread, so it is worth saying slowly: AKT inhibits TSC, TSC inhibits Rheb, so AKT raises Rheb-GTP. Note also that AKT acts on TSC2, not on Rheb.
Question 3 · HarderWhich pairing matches the division of labour this lesson describes at the lysosome?
The amino-acid branch is about recruitment through the Rag GTPases; the growth-factor branch is about the nucleotide state of Rheb. That split is a tidier account of the AND-gate behaviour than saying the two signals simply add.
Show the answer
B — Amino acids decide where mTORC1 sits; growth factors decide whether the activator it meets there is in its active state. The amino-acid branch is about recruitment through the Rag GTPases; the growth-factor branch is about the nucleotide state of Rheb. That split is a tidier account of the AND-gate behaviour than saying the two signals simply add.
Go deeper
Follow a guided route through the mechanism:
- How does a cell learn that it is allowed to grow?The full growth-factor route: how a cell learns that it is allowed to grow.
- How does a cell know it has enough raw material to grow?The amino-acid route, which decides where mTORC1 sits before Rheb ever meets it.
Concepts introduced in this lesson
Small GTPase and the GTP/GDP cycle · GAP (GTPase-activating protein) · Double-negative regulation · Subcellular localisation as regulation · Overexpression as an experimental caveat