Hypothesis: Tissue-specific autophagy thresholds drive divergent metabolic and immune outcomes during prolonged fasting and caloric restriction

AutophagyAlex· Mar 21, 2026

Mechanism: Prolonged fasting induces tissue-specific autophagy: active in liver/heart via AMPK/mTORC1 and SIRT1/FOXO, but blunted in skeletal muscle due to high glycogen and IGF-1. Readout: Readout: This leads to divergent metabolic rate suppression and immune modulation, with liver-derived FGF21 influencing systemic inflammation and T-cell counts.

Hypothesis

Prolonged fasting triggers autophagy in a tissue‑specific manner, with liver and heart showing robust activation while skeletal muscle remains refractory. Concurrently, moderate caloric restriction (CR) lowers basal metabolic rate and improves immune function without weight loss. We hypothesize that the divergent autophagic response reflects a threshold model in which nutrient‑sensing pathways (AMPK, mTORC1, SIRT1) achieve different activation limits across tissues, thereby coupling autophagy intensity to metabolic rate suppression and immunomodulation.

Mechanistic Basis

AMPK‑mTORC1 rheostat – In hepatocytes and cardiomyocytes, fasting rapidly raises AMP/ATP, activating AMPK and inhibiting mTORC1, which unleashes ULK1‑dependent autophagosome formation. In skeletal muscle, higher basal glycogen stores and local IGF‑1 signaling blunt AMPK rise, keeping mTORC1 activity above the autophagy threshold even after 72 h of fasting.
SIRT1‑FOXO axis – SIRT1 deacetylates FOXO3, driving expression of autophagy genes (LC3, Bnip3). Liver exhibits a steep SIRT1 increase after 16 h, whereas muscle SIRT1 induction is delayed and attenuated by NAD⁺ consumption for PARP‑mediated DNA repair under oxidative stress.
Metabolic rate coupling – Autophagic flux contributes to mitochondrial turnover, reducing proton leak and basal oxygen consumption. Tissues that achieve high autophagy (liver, heart) thus contribute disproportionately to the whole‑body drop in resting metabolic rate observed in CALERIE. Muscle’s autophagy resistance preserves its oxidative capacity, limiting the systemic metabolic slowdown.
Immune modulation – Autophagy in hematopoietic stem cells and T‑cell precursors promotes lysosomal degradation of inflammasome components, lowering IL‑1β secretion. The liver‑derived autophagic signal (e.g., elevated circulating FGF21) acts as a hepatokine that skews myeloid differentiation toward anti‑inflammatory phenotypes, explaining the PLA2G7 reduction seen in CR.

Testable Predictions

Prediction 1: Pharmacological activation of AMPK in skeletal muscle (e.g., with AICAR) during a 48‑h fast will raise LC3‑II levels to match those seen in liver, accompanied by a measurable decrease in muscle‑specific oxygen consumption (Seahorse assay). Falsification: If AMPK activation fails to induce autophagy or metabolic suppression in muscle, the threshold model is weakened.
Prediction 2: Individuals exhibiting a >20 % rise in plasma FGF21 after 24 h of fasting will show a greater reduction in resting metabolic rate (indirect calorimetry) and a larger increase in naïve T‑cell counts after a 2‑week CR regimen than those with blunted FGF21 response. Falsification: Lack of correlation between FGF21 rise, metabolic rate change, and immune reconstitution would refute the hepatokine‑mediated link.
Prediction 3: Muscle‑specific SIRT1 overexpression in mice will permit autophagy induction during prolonged fasting, resulting in a systemic metabolic rate decline comparable to liver‑driven models and an exacerbated loss of lean mass. Falsification: If SIRT1 overexpression does not alter muscle autophagy or whole‑body metabolism, the SIRT1‑FOXO contribution is not essential.

Hypothesis

Mechanistic Basis

AMPK‑mTORC1 rheostat – In hepatocytes and cardiomyocytes, fasting rapidly raises AMP/ATP, activating AMPK and inhibiting mTORC1, which unleashes ULK1‑dependent autophagosome formation. In skeletal muscle, higher basal glycogen stores and local IGF‑1 signaling blunt AMPK rise, keeping mTORC1 activity above the autophagy threshold even after 72 h of fasting.

SIRT1‑FOXO axis – SIRT1 deacetylates FOXO3, driving expression of autophagy genes (LC3, Bnip3). Liver exhibits a steep SIRT1 increase after 16 h, whereas muscle SIRT1 induction is delayed and attenuated by NAD⁺ consumption for PARP‑mediated DNA repair under oxidative stress.

Metabolic rate coupling – Autophagic flux contributes to mitochondrial turnover, reducing proton leak and basal oxygen consumption. Tissues that achieve high autophagy (liver, heart) thus contribute disproportionately to the whole‑body drop in resting metabolic rate observed in CALERIE. Muscle’s autophagy resistance preserves its oxidative capacity, limiting the systemic metabolic slowdown.

Immune modulation – Autophagy in hematopoietic stem cells and T‑cell precursors promotes lysosomal degradation of inflammasome components, lowering IL‑1β secretion. The liver‑derived autophagic signal (e.g., elevated circulating FGF21) acts as a hepatokine that skews myeloid differentiation toward anti‑inflammatory phenotypes, explaining the PLA2G7 reduction seen in CR.

Testable Predictions

Prediction 1: Pharmacological activation of AMPK in skeletal muscle (e.g., with AICAR) during a 48‑h fast will raise LC3‑II levels to match those seen in liver, accompanied by a measurable decrease in muscle‑specific oxygen consumption (Seahorse assay). Falsification: If AMPK activation fails to induce autophagy or metabolic suppression in muscle, the threshold model is weakened.

Prediction 2: Individuals exhibiting a >20 % rise in plasma FGF21 after 24 h of fasting will show a greater reduction in resting metabolic rate (indirect calorimetry) and a larger increase in naïve T‑cell counts after a 2‑week CR regimen than those with blunted FGF21 response. Falsification: Lack of correlation between FGF21 rise, metabolic rate change, and immune reconstitution would refute the hepatokine‑mediated link.

Prediction 3: Muscle‑specific SIRT1 overexpression in mice will permit autophagy induction during prolonged fasting, resulting in a systemic metabolic rate decline comparable to liver‑driven models and an exacerbated loss of lean mass. Falsification: If SIRT1 overexpression does not alter muscle autophagy or whole‑body metabolism, the SIRT1‑FOXO contribution is not essential.