mTOR Core · Lesson 09 · Intermediate · 25 min
mTOR and Autophagy
The question
How does one kinase decide between building and recycling?
What you should be able to do
After this lesson you should be able to
- Explain why autophagy is described as restrained rather than switched off.
- Distinguish the fast and slow arms of that restraint and say what each one can do that the other cannot.
- Interpret a count of autophagosomes, and name the measurement that would make it a claim about flux.
- Evaluate how far two genetic experiments in a worm and a mouse support a hypothesis about mTOR and lifespan.
Research skill Mechanistic interpretation
The core idea
Autophagy is the cell eating parts of itself on purpose: material is wrapped up, delivered to the lysosome, broken down, and the pieces reused.
It is not simply what happens when mTORC1 is off. mTORC1 actively holds it back, and letting go is a different thing from being absent — a held-back program can be released quickly and by degrees.
There are two holds, on two timescales. A fast one, on the kinase that starts autophagy, works in minutes. A slow one, on a protein that switches on genes, takes hours.
The mechanism, in outline
When food is plentiful, mTORC1 physically sits on the complex that starts autophagy and marks it to keep it off.
The same target gets an opposite mark from AMPK, the low-energy sensor — so this one protein is reading growth signals and energy at the same time.
The slow arm works through TFEB: while mTORC1 is active, TFEB is kept out of the nucleus; when it falls, TFEB goes in and switches on genes for more lysosomes.
mTOR is also involved in ending the process and rebuilding lysosomes afterwards, so it is on both ends of the cycle.
Two clocks on the same decision
The fast arm (mTORC1 and AMPK acting on ULK1) and the slow arm (mTORC1 holding TFEB out of the nucleus) side by side. The second control is time, because the difference between these two arms is entirely a difference in how long they take.
ULK1 phosphorylated and inactive · TFEB cytosolic
Nutrients present: mTORC1 sits on the ULK1 complex and keeps it off, and TFEB is held out of the nucleus. Autophagy is restrained rather than absent — the machinery is there and being held.
- 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.
- AMPK
- Energy-stress kinase that both activates TSC2 and directly inhibits Raptor.The low-fuel sensor.αβγ heterotrimer. Two independent arms onto mTORC1 plus a direct activating arm onto ULK1 — the reason energy stress switches growth off and recycling on in one move.
- ULK1
- Autophagy-initiating kinase, inhibited by mTORC1 and activated by AMPK.The switch that starts self-digestion.mTORC1 phosphorylates S757 to block the AMPK–ULK1 interaction; ULK1 also feeds back to phosphorylate and dampen AMPK, making this a closed loop rather than a switch.
- TFEB
- Transcription factor for lysosomal and autophagy genes; excluded from the nucleus when phosphorylated by mTORC1.The master switch for recycling genes.Phosphorylated on S211 in a Rag- and FLCN-dependent, substrate-selective manner; this is the clearest case where mTORC1 substrate choice — not overall activity — is the regulated variable.
- Autophagy
- Bulk degradative recycling, held off by mTORC1 and switched on by AMPK.The cell eating its own worn-out parts.Regulated at initiation (ULK1), at transcription (TFEB) and at fusion; most 'autophagy is required' claims rest on flux measurements that are hard to do in tissue.
- Lysosomal biogenesis
- TFEB-driven expansion of the lysosomal compartment.Making more recycling plants.The return arm of the lysosome-to-nucleus circuit (SET2012): mTORC1 phosphorylation keeps TFEB out of the nucleus, and when it is released TFEB expands the very organelle on which mTORC1 is regulated. This closes a genuine homeostatic loop that a linear diagram cannot show.
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
| Nutrients | Time | Readout | What it means |
|---|---|---|---|
| fed | minutes | ULK1 phosphorylated and inactive · TFEB cytosolic | Nutrients present: mTORC1 sits on the ULK1 complex and keeps it off, and TFEB is held out of the nucleus. Autophagy is restrained rather than absent — the machinery is there and being held. |
| fed | hours | ULK1 inactive · TFEB cytosolic · lysosomal genes quiet | Nothing new happens over hours in a fed cell, which is the flat comparison the starved column needs to be read against. |
| starved | minutes | ULK1 released and activated by AMPK · autophagosomes forming · TFEB just entering the nucleus | The fast arm. Removing a phosphorylation is quick, and AMPK pushes the same kinase from the other side, so autophagy can start within minutes — long before any new protein has been made. |
| starved | hours | Autophagy running · TFEB nuclear · lysosomal and autophagy genes transcribed | The slow arm arrives. TFEB in the nucleus means the cell is building more of the apparatus it has been using — and more functional lysosomes eventually release more amino acids, which is how this becomes a loop rather than a one-way street. |
Is the autophagy connected to the lifespan effect?
Lowering mTOR signalling lengthens life in several laboratory species, and it also releases autophagy — so the obvious guess is that autophagy is how the lifespan effect works.
Two results support the guess. Worms without a key autophagy gene lost the lifespan benefit of reduced insulin-like signalling. Mice given extra copies of an autophagy gene lived about 17% longer.
Neither experiment touched mTOR. So they make the idea serious and leave the actual question open — which is why the Atlas still lists it as unresolved.
A measurement warning worth carrying
Autophagy is a flux, not a quantity. Counting autophagosomes at one time point cannot distinguish 'more are being made' from 'they are not being cleared', and those are opposite conclusions. Papers that measure flux — usually by comparing conditions with and without a block on lysosomal degradation — are making a much stronger claim than papers that count structures or report a single marker. This is one of the most common places where a reader can be misled without anything in the paper being wrong.
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.
Showed the DIRECT brake mTORC1 uses on autophagy: when nutrients are plentiful, mTORC1 physically joins the ULK1-Atg13-FIP200 complex (the autophagy-starter kinase) and phosphorylates ULK1 to keep it off.
Study page →Revealed the tug-of-war over ULK1: the energy sensor AMPK phosphorylates ULK1 at activating sites to turn autophagy ON when energy is low, while mTORC1 phosphorylates a different site (Ser757) to keep it OFF and even blocks AMPK from rea…
Study page →A lysosome-to-nucleus mechanism uses mTOR-dependent TFEB phosphorylation to sense lysosomal state.
Study page →Reactivation of mTOR terminates autophagy and drives autophagic lysosome reformation (ALR).
Study page →Worms lacking the autophagy gene bec-1 lost the lifespan-extending benefit of reduced insulin-like signaling - autophagy is mechanistically required for longevity, not just correlated.
Study page →Mice engineered with extra copies of the autophagy gene Atg5 lived 17% longer and were leaner and more insulin-sensitive.
Study page →Whether autophagy is required for the lifespan effects associated with reduced mTOR signalling in mammals has not been tested directly in the form in which the claim is usually made. The supporting studies alter autophagy genetically rather than through mTOR, and in a worm and a mouse respectively.
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.
Necessity in a worm, sufficiency in a mouse
| C Autophagy genes are essential for dauer development and life-span extension in C. elegans | C Overexpression of Atg5 in mice activates autophagy and extends lifespan | |
| Model system | Caenorhabditis elegans | Mouse (Atg5-overexpressing transgenic) |
| Perturbation | C. elegans with reduced insulin-like signalling, in which the autophagy gene bec-1 was removed. | Mice engineered to carry extra copies of the autophagy gene Atg5. |
| Readout | Lifespan. The extension normally conferred by reduced insulin-like signalling was lost without the autophagy gene. | Lifespan and metabolic measures: about 17% longer median lifespan, leaner animals, greater insulin sensitivity. |
What do both studies support?
That autophagy is not merely correlated with longevity in these models but does causal work — required in one direction, and enough to move the outcome in the other.
Where do they differ?
In the direction of the argument and in the organism. Necessity in a worm and sufficiency in a mouse are different claims, and neither manipulated mTOR: both change autophagy by other means. So together they make the autophagy hypothesis serious and leave the specific question — whether autophagy is the route by which reduced mTOR signalling extends mammalian lifespan — untested in the form it is usually stated.
What experiment would help next?
The experiment the field still owes itself: block autophagy in a mammal receiving an mTOR-directed intervention, and see whether the lifespan effect survives. The Atlas carries this as an open question for exactly that reason.
Counting autophagosomes
What this evidence supports
- That mTORC1 restrains autophagy directly through ULK1 and transcriptionally through TFEB
- That autophagy contributes causally to lifespan in a worm and in a mouse, by manipulations that did not involve mTOR
What it does not establish
- That more autophagosomes means more autophagy — blocked clearance produces the same picture
- That autophagy is the mechanism of mTOR-related lifespan effects in mammals
- That a marker measured at one time point reports a flux
Why?
Autophagy is a rate, and most of the disagreement in this literature comes from measurements that report a standing quantity instead. A study that compares conditions with and without a block on lysosomal degradation is making a much stronger claim than one that counts structures.
Think
mTORC1 restrains autophagy on a timescale of minutes through ULK1 and on a timescale of hours through TFEB. Why would a cell need both?
Ask what each one can do that the other cannot.
Think first, then reveal
The fast arm handles the immediate problem: a cell that has just run out of amino acids needs to start recycling now, and post-translational modification of an existing complex is the only mechanism quick enough. The slow arm handles capacity: sustained recycling needs more lysosomes and more of the machinery, which requires transcription and takes hours. A cell with only the fast arm would start well and run out of apparatus; a cell with only the slow arm would respond too late to survive an acute shortage. The pairing also predicts something testable — a brief nutrient dip and a prolonged one should produce qualitatively different cellular states, not simply different magnitudes of the same one.
A study reports more autophagosomes in treated cells and concludes that the treatment increased autophagy. What alternative explanation must be excluded, and how?
An autophagosome is made and then consumed.
Think first, then reveal
The alternative is blocked clearance: if autophagosomes are not being degraded by the lysosome, they accumulate, and a snapshot shows more of them while the process has actually slowed. Distinguishing the two requires a flux measurement — comparing the treated condition with and without an inhibitor of lysosomal degradation, so that the rate of production can be separated from the rate of consumption — or following a marker whose turnover reports degradation. This matters far beyond one paper: it is a systematic reason why the autophagy literature contains apparently contradictory results, and why the question of what autophagy contributes to mTOR-related lifespan effects is harder to answer than it first looks.
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-upHow does mTORC1 keep autophagy initiation switched off when nutrients are plentiful?
mTORC1 associates with the initiating complex and phosphorylates ULK1 at an inhibitory site. Release that phosphorylation and the complex becomes competent — restraint lifted, not machinery rebuilt.
Show the answer
B — By phosphorylating ULK1 in the ULK1–ATG13–FIP200 complex. mTORC1 associates with the initiating complex and phosphorylates ULK1 at an inhibitory site. Release that phosphorylation and the complex becomes competent — restraint lifted, not machinery rebuilt.
Question 2 · Step upULK1 receives opposing phosphorylations from mTORC1 and AMPK. What does that arrangement let the cell do?
It is the same integration architecture seen at TSC2, one layer further down: two inputs, two sites, one output. The cell arrives at a level of ULK1 activity rather than a verdict.
Show the answer
A — Respond to growth signalling and energy state at the same node, through different sites. It is the same integration architecture seen at TSC2, one layer further down: two inputs, two sites, one output. The cell arrives at a level of ULK1 activity rather than a verdict.
Question 3 · HarderWorms lacking bec-1 lose the lifespan extension from reduced insulin-like signalling, and mice overexpressing Atg5 live about 17% longer. What do these two results together support?
Both studies change autophagy by means other than mTOR. They make the hypothesis serious and they do not test it in the form it is usually stated — which is exactly why the Atlas still carries it as an open question.
Show the answer
B — That autophagy matters for lifespan in these models — while leaving open whether it is the route by which reduced mTOR signalling acts, since neither study manipulated mTOR. Both studies change autophagy by means other than mTOR. They make the hypothesis serious and they do not test it in the form it is usually stated — which is exactly why the Atlas still carries it as an open question.
Go deeper
Follow a guided route through the mechanism:
- What does a cell actually do when mTORC1 fires?What else changes downstream when mTORC1 activity falls.
- What happens, and in what order, when you stop eating?The nutrient-withdrawal route that releases this program.
Concepts introduced in this lesson
Restraint versus absence · Flux versus steady-state measurement · Necessity and sufficiency · Fast and slow regulation