AMPK Restricts ULK1-Driven Autophagy Under Energy Stress

The Nature Communications study Redefining the role of AMPK in autophagy and the energy stress response revisits one of the most established models in autophagy biology. AMPK is commonly described as an energy-sensing kinase that activates ULK1, the kinase that initiates autophagy when nutrients are scarce. Park, Lee, and Kim show that this model is incomplete: in the cellular energy-stress conditions they examined, AMPK inhibited ULK1 activity and restrained autophagy induction.

The study does not define AMPK as a universal autophagy-off switch. Instead, it identifies a dual regulatory function. AMPK limits abrupt activation of the ULK1-Atg14-Vps34 initiation machinery during severe energy deficiency, yet protects components of that machinery from caspase-mediated degradation. This balance may prevent an energetically expensive process from starting when cells cannot support it, while retaining the capacity to restore autophagy and homeostasis after stress subsides.

Study Background and Research Question

Macroautophagy can recycle intracellular macromolecules and potentially provide substrates for energy production during glucose starvation. However, autophagy is not energetically free. Membrane remodeling, vesicle trafficking, autophagosome formation, and lysosomal degradation all require cellular resources. A cell facing an acute energy crisis may therefore need to secure a minimum energy supply before it can execute a complete autophagic response.

The conventional pathway places AMPK upstream of ULK1 activation. In this view, glucose depletion raises the AMP-to-ATP ratio, activates AMPK, and promotes ULK1 phosphorylation at sites associated with autophagy. mTORC1 is generally positioned as the opposing nutrient-sensitive kinase. Yet several observations have challenged this simplified scheme. mTOR inhibition can reduce AMPK-dependent phosphorylation of ULK1 Ser556, nutrient withdrawal can weaken AMPK-ULK1 binding, and pharmacological AMPK activation has sometimes inhibited or failed to induce autophagy.

The central question was therefore whether AMPK activates ULK1 inside cells during energy stress, or whether it performs a different form of pathway control. The authors addressed this question by measuring ULK1 activity directly through cellular substrates and by separating glucose deprivation, amino acid deprivation, and mitochondrial energy failure rather than treating them as interchangeable stimuli.

Key Innovation from the Reference Study

The main innovation is a shift from inferring ULK1 activity from AMPK phosphorylation sites to measuring the functional output of the ULK1 signaling pathway. This distinction is important because a phosphorylation event historically interpreted as activating ULK1 does not necessarily track with autophagy initiation under every nutrient condition.

The authors found that AMPK suppresses ULK1 signaling to the autophagy initiation machinery. Two AMPK-mediated ULK1 phosphorylations were identified as important for this inhibitory effect, including the regulation of the site commonly referred to as ULK1 Ser556 in human cells. The reference study shows that Ser556 phosphorylation depends on mTORC1 activity rather than simply reporting AMPK activation: it was reduced by Torin1, diminished by rapamycin, and disappeared during amino acid starvation.

A second conceptual advance is the separation of pathway inhibition from machinery preservation. During energy deficiency, AMPK can restrain ULK1 activation while protecting ULK1-associated components from caspase-dependent destruction. The result is a temporary state of autophagy restraint rather than irreversible loss of autophagy competence. This model explains why AMPK activation can correlate with energy stress without necessarily producing immediate autophagosome formation.

Methods and Experimental Design Insights

The experimental design combined nutrient manipulations, kinase perturbations, protein-interaction analysis, and functional autophagy measurements. The authors compared amino acid starvation with glucose starvation and examined conditions in which mitochondrial dysfunction generated an energy crisis. This comparison was essential because amino acid withdrawal strongly affects mTORC1 and the autophagy initiation complex, whereas glucose depletion primarily changes cellular energy availability.

Pharmacological experiments used mTORC1 inhibition and AMPK-directed interventions to test the direction of signaling. Torin1 and rapamycin were used to examine how mTORC1 activity influenced ULK1 phosphorylation and the AMPK-ULK1 interaction. Other experiments considered AMPK activation through agents such as A769662, AICAR, and metformin. These perturbations were interpreted together with cellular measurements rather than as stand-alone proof of pathway activation, an important precaution because kinase activators and metabolic drugs can have effects beyond their nominal targets.

Biochemical and cell-based readouts focused on the functional state of the initiation machinery. Immunoblotting and related phosphoprotein analyses tracked AMPK, mTORC1, ULK1, and downstream substrates. Measurements of ULK1 substrate phosphorylation, including ATG13-associated signaling where applicable, provided a more direct estimate of ULK1 activity than measurement of AMPK phosphorylation alone. Co-immunoprecipitation experiments tested whether AMPK and ULK1 remained physically associated after mTORC1 inhibition or nutrient stress.

The study also examined the ULK1-Atg14-Vps34 axis, a core signaling module that connects ULK1 activation to phosphatidylinositol 3-kinase activity and autophagosome initiation. Finally, the authors assessed the stability of autophagy machinery during energy deficiency and evaluated whether preserving these components supported recovery after stress. Together, these experiments link kinase signaling, protein complex organization, autophagy output, and stress adaptation.

Protocol Parameters

• Stress comparison: Compare amino acid starvation, glucose starvation, and mitochondrial energy stress in parallel. This separates nutrient-signaling effects from an absolute shortage of cellular energy.
• Pathway controls: Monitor both AMPK and mTORC1 activity when interpreting ULK1 phosphorylation. Increased AMPK activity alone should not be taken as evidence of increased ULK1 activity.
• Direct ULK1 readouts: Include ULK1 substrate phosphorylation and signaling through the ULK1-Atg14-Vps34 module rather than relying only on canonical AMPK sites.
• Flux assessment: Use an LC3 flux measurement strategy with appropriate lysosomal controls. Reduced LC3-II accumulation can reflect reduced autophagosome production or altered degradation, so static LC3 abundance is insufficient.
• Recovery design: If the question concerns adaptation, include a stress-removal or recovery phase and measure preservation of ULK1-associated proteins as well as autophagy output.

Core Findings and Why They Matter

First, AMPK activation did not produce the expected increase in ULK1 signaling during glucose starvation. Instead, glucose starvation suppressed the ULK1-Atg14-Vps34 response that was stimulated by amino acid starvation, and this suppression depended on AMPK. Thus, a cell can experience two simultaneous nutrient signals with different consequences: amino acid loss can favor autophagy initiation, while glucose loss can activate an AMPK-dependent brake.

Second, mTORC1 inhibition disrupted the interaction between AMPK and ULK1. This result helps explain why Torin1 reduced ULK1 Ser556 phosphorylation. Rather than mTORC1 inhibition simply releasing AMPK to bind and activate ULK1, the data support a model in which mTORC1 status changes the physical organization and phosphorylation state of the AMPK-ULK1 complex.

Third, mitochondrial dysfunction activated the LKB1-AMPK axis but inhibited ULK1 activation and autophagy induction, even under amino acid starvation. This finding is especially relevant to experimental interpretation. A strong energy-stress signal can coexist with low autophagic flux because the cell may prioritize ATP conservation and survival over initiating a resource-intensive degradation program.

Fourth, AMPK had a protective function despite inhibiting autophagy initiation. By limiting caspase-mediated degradation of ULK1-associated machinery, AMPK preserved the cellular capacity to restart autophagy and restore homeostasis after the acute stress ended. This dual action provides a mechanistic explanation for how AMPK can be beneficial during energy deficiency without directly promoting immediate autophagic flux.

These findings affect the interpretation of autophagy inhibition experiments. Blocking ULK1 can reduce initiation, but the downstream phenotype will depend on whether the cells are amino acid limited, glucose limited, mitochondrially impaired, or recovering from stress. Likewise, ATG13 phosphorylation blockade and LC3 flux changes should be interpreted alongside energy-state markers and ULK1 activity measurements. The paper therefore argues for a systems-level view of autophagy signaling rather than a linear AMPK-to-ULK1 activation model.

Comparison with Existing Internal Articles

The internal article AMPK Restricts, Not Triggers, ULK1-Driven Autophagy Under Energy Stress closely follows the reference study’s central conclusion. It is useful as a concise conceptual summary, whereas the Nature Communications paper provides the primary experimental basis for the claims about ULK1 phosphorylation, AMPK-ULK1 interaction, mitochondrial dysfunction, and preservation of autophagy machinery.

For assay planning, Scenario-Driven Solutions for Autophagy Assays: MRT68921 extends the discussion toward pharmacological workflow design. Its relevance here is methodological rather than evidentiary: a ULK1/2 inhibitor can help test whether an observed autophagy phenotype depends on ULK activity, but it cannot by itself establish that AMPK is the upstream cause of that phenotype.

Limitations and Transferability

The study’s conclusions should be transferred cautiously across cell types and stress durations. AMPK signaling, LKB1 status, mitochondrial reserve, caspase activity, and basal autophagy can vary substantially among models. A short nutrient pulse may produce a different response from prolonged glucose deprivation, and combined starvation may engage competing signals that are not present in a single-stress experiment.

Pharmacological perturbations also require orthogonal validation. Torin1, rapamycin, A769662, AICAR, and metformin alter signaling networks beyond one intended node, so genetic depletion, rescue, kinase-substrate measurements, and pathway-specific controls strengthen causal interpretation. Similarly, LC3-based measurements do not distinguish increased autophagosome formation from impaired lysosomal clearance unless flux is measured directly.

The reference evidence is primarily mechanistic and cell-based. It establishes a revised model for AMPK-ULK1 regulation but does not demonstrate that every tissue responds identically during whole-organism energy stress. Future work should therefore test how this regulatory balance behaves in physiologically distinct tissues and in disease-relevant settings without assuming that AMPK activation predicts autophagy induction.

Research Support Resources

For experiments that require pharmacological interruption of ULK-dependent autophagy initiation, researchers can use MRT68921 dual autophagy kinase ULK1/2 inhibitor (SKU B6174) as a research-use tool. The product information reports biochemical IC₅₀ values of 2.9 nM for ULK1 and 1.1 nM for ULK2, with reduced ATG13 phosphorylation and LC3 flux in wild-type but not ULK1 M92T cells. These data support target-engagement controls, while AMPK perturbation and direct ULK1 readouts remain necessary for testing the specific mechanism proposed by the reference study. MRT68921 is preclinical, intended for scientific research only, and has no reported in vivo animal or clinical-trial data.