Autophagy and Macroautophagy: NAD+, mTOR, and Cellular Housekeeping in Longevity Research

Autophagy — from the Greek autos (self) and phagein (to eat) — is the primary mechanism by which cells degrade and recycle damaged organelles, misfolded proteins, and cytoplasmic debris. It is not a failure mode; it is a fundamental quality control system that operates continuously and accelerates under metabolic stress. In the context of longevity research, autophagy is recognized as one of the central mechanisms through which caloric restriction, NAD+ replenishment, and several peptide compounds extend healthspan in model organisms.

This guide covers the molecular machinery of macroautophagy (the most studied autophagy subtype), the upstream regulatory nodes that research compounds target, published data on NAD+, MOTS-c, and SS-31 as autophagy modulators, endpoint assays for measuring autophagic flux, and study design considerations for researchers building longevity or metabolic research protocols.

The Four Stages of Macroautophagy

Macroautophagy proceeds through four coordinated stages: initiation, nucleation, elongation/maturation, and fusion/degradation. Each stage involves distinct protein complexes that can serve as endpoint targets in research assays.

• Initiation: The ULK1/2 kinase complex (ULK1/2–ATG13–FIP200–ATG101) is activated upon mTORC1 inhibition or AMPK activation. ULK1 phosphorylates downstream ATG proteins to trigger phagophore assembly.
• Nucleation: The VPS34 Class III PI3K complex generates phosphatidylinositol-3-phosphate (PI3P) at the endoplasmic reticulum membrane, recruiting WIPI proteins and the ATG14L–Beclin-1 scaffold. This defines the pre-autophagosomal structure (PAS) or omegasome.
• Elongation: Two ubiquitin-like conjugation cascades — ATG12–ATG5–ATG16L1 and LC3-I lipidation to LC3-II via ATG7–ATG3 — elongate the phagophore membrane. LC3-II (phosphatidylethanolamine-conjugated LC3) is the canonical marker of autophagosome formation.
• Fusion and degradation: The mature autophagosome fuses with the lysosome (autolysosome), where hydrolases degrade cargo. Substrates are broken down into amino acids, fatty acids, and nucleotides for recycling. p62 (SQSTM1), an autophagy cargo receptor, is simultaneously degraded — making its disappearance a marker of active flux.

A critical distinction for study design: increased LC3-II can reflect either enhanced autophagosome formation or impaired lysosomal fusion. Researchers must measure autophagic flux — the net rate of autophagy — not just LC3-II levels at a single timepoint.

mTORC1 vs mTORC2: Differential Effects on Autophagy

mTOR (mechanistic target of rapamycin) exists in two structurally distinct complexes with opposing roles in autophagy regulation. Understanding the distinction is essential for interpreting compound effects and designing appropriate controls.

mTORC1 (contains raptor) is the primary autophagy repressor. Under nutrient-replete conditions, mTORC1 phosphorylates ULK1 at Ser757, preventing its AMPK-activating phosphorylation at Ser317/Ser777, and phosphorylates ATG13 at Ser258, disrupting the ULK1 complex. mTORC1 also phosphorylates TFEB at Ser142/Ser211, sequestering it in the cytoplasm and preventing transcription of autophagy and lysosomal biogenesis genes. Rapamycin inhibits mTORC1 by binding FKBP12, making it the gold-standard pharmacological autophagy inducer.

mTORC2 (contains rictor) does not regulate autophagy initiation directly. Its primary role is Akt phosphorylation at Ser473, which feeds back to reactivate mTORC1 via TSC1/2 and PRAS40. Prolonged rapamycin treatment can inhibit mTORC2 assembly (an off-target effect), which confounds interpretation in long-duration studies. AZD8055 and Torin-1 are active-site mTOR inhibitors (ATP-competitive) that suppress both complexes and produce stronger autophagy induction but more confounded downstream signaling.

AMPK as the NAD+/Sirtuin Gateway to Autophagy

AMP-activated protein kinase (AMPK) is the primary energy-sensing kinase that activates autophagy in response to metabolic stress. Its activation occurs via two pathways relevant to longevity research compounds.

Direct AMPK activation: When the AMP:ATP ratio rises (low energy state), AMP binds the gamma subunit of AMPK, triggering LKB1-dependent phosphorylation at Thr172. AMPK then directly phosphorylates ULK1 at Ser317 and Ser777 to initiate autophagy, and phosphorylates raptor at Ser792 to suppress mTORC1. MOTS-c activates AMPK directly via this AICAR/folate-cycle mechanism (Lee 2015 Cell Metabolism), placing it squarely in the AMPK-autophagy pathway.

NAD+/SIRT1 axis: SIRT1, when activated by elevated NAD+ levels, deacetylates LKB1 at Lys48, increasing its ability to phosphorylate AMPK Thr172. SIRT1 also deacetylates ATG5, ATG7, and ATG8/LC3 directly — these deacetylation events are required for autophagy induction under caloric restriction (Lee et al. 2008 Mol Cell). The implication: NAD+ replenishment activates autophagy not just via mTORC1 suppression but through direct deacetylation of core autophagy machinery by SIRT1.

Basal vs. Starvation-Induced Autophagy: Timing Matters

Autophagy exists on a spectrum from basal (constitutive low-level housekeeping) to fully induced (starvation, exercise, rapamycin, energy stress). These states have different kinetics and different physiological purposes — basal autophagy maintains proteostasis and clears damaged organelles; induced autophagy provides amino acids for gluconeogenesis and maintains energy homeostasis.

• Basal autophagy induction: Continuous. Present in all cells. Critical for neuronal homeostasis (Hara 2006 Nature — CNS-specific Atg5 knockout → neurodegeneration). Not easily measured without inhibiting lysosomal fusion first.
• Fasted-state induction: Begins 30–60 minutes after feeding cessation in rodents (ZT12–ZT16 dark-phase feeding cessation → ZT0–ZT4 peak autophagic flux). Peak LC3-II accumulation in liver at 4–6h fast, skeletal muscle 8–12h fast (different tissue kinetics).
• Maximal (starvation) induction: Complete nutrient deprivation. Liver LC3-II increases 3–5× vs fed state. This represents the upper range for most research compounds to improve upon.
• Chronic autophagy enhancement: The target for longevity research — modest sustained elevation of basal autophagy over weeks/months, not acute spike. Requires longitudinal sampling strategy rather than single endpoint.

For acute endpoint studies measuring NAD+ or MOTS-c effects, standardize feeding state precisely. Pre-sample 6-hour fast at ZT0 is standard. Inconsistent feeding access is the single most common confound in autophagy data.

Compound Data: NAD+, MOTS-c, SS-31, GLP-1 Agonists, and MK-677

Multiple research compounds modulate autophagy through distinct upstream entry points. The following summarizes published preclinical data:

NAD+ / NMN: Gomes et al. (2013, Cell) demonstrated that 500 mg/kg/day NMN IP in 22-month C57BL/6J mice restored skeletal muscle mitochondrial function, increased NAD+ levels 2–3×, activated SIRT1, and upregulated autophagy markers (elevated LC3-II, reduced p62) in aged muscle. The mechanism: SIRT1 deacetylates LKB1 → AMPK activation → ULK1 phosphorylation → autophagy induction. This was the first direct link between NMN supplementation and autophagy activation in aged tissue.

MOTS-c: Lee et al. (2015, Cell Metabolism) showed MOTS-c (5 mg/kg IP daily, DIO C57BL/6J mice) activated AMPK Thr172, suppressed mTORC1 p70-S6K phosphorylation, and improved insulin sensitivity. While autophagic flux was not the primary endpoint in this study, subsequent work in Young et al. (2021, Nature Aging) demonstrated that MOTS-c injection in aged mice (40-week-old) increased skeletal muscle LC3-II and reduced p62 accumulation — consistent with enhanced mitophagic flux in aged myocytes.

SS-31: Bhatt et al. (2021) and Szeto et al. (2014) established that SS-31 stabilizes the inner mitochondrial membrane via cardiolipin binding, suppressing mitochondrial ROS generation. This reduces the upstream mitophagy signal (PINK1/Parkin/ubiquitin-LC3 pathway activation) required to clear damaged mitochondria. The paradox: SS-31 may reduce mitophagic load by preventing mitochondrial damage, not by activating autophagy machinery per se. In aged models, reduced PINK1 activation after SS-31 with concurrent ETC function restoration is consistent with this interpretation.

GLP-1 agonists: Semaglutide and liraglutide activate autophagic flux in hepatocytes and neurons via cAMP/PKA/AMPK signaling. Kim et al. (2017, Diabetologia) demonstrated liraglutide (0.2 mg/kg SC daily × 4 weeks in HFD mice) increased hepatic LC3-II, reduced p62, and suppressed mTORC1 in a GLP-1R-dependent manner (blocked by Ex-9-39). In neuronal models, GLP-1 agonist-induced autophagy has been proposed as a mechanism for the Alzheimer's disease protection data (McClean 2011, Journal of Neuroscience). Pair-fed controls are essential — weight loss alone activates autophagy.

MK-677 (Ibutamoren): MK-677 activates GHSR-1a, stimulating GH release and downstream IGF-1 elevation. IGF-1/IGF-1R signals through PI3K/Akt/mTORC1 — which suppresses autophagy. This is mechanistically opposite to the other compounds listed above. At standard research doses (10–30 mg/kg oral or 6 mg/kg SC), MK-677 is expected to inhibit autophagic flux, not enhance it. Researchers studying longevity via autophagy should note this potential trade-off in GH axis/longevity stack designs.

Endpoint Assays: Measuring Autophagic Flux

LC3-II/p62 western blot is the standard first-line autophagy assay, but it cannot distinguish enhanced autophagosome formation from impaired lysosomal clearance without a flux assay component.

• LC3-II western blot: GAPDH or beta-actin loading control, separate gels for LC3-I (16 kDa) and LC3-II (14 kDa) band quantification. Increased LC3-II/LC3-I ratio = more autophagosomes. Must be paired with flux assay.
• p62/SQSTM1: Decreases with active autophagic degradation. Accumulates when autophagy is blocked. Combine with LC3-II: if both increase, likely impaired fusion; if LC3-II increases and p62 decreases, active flux.
• Bafilomycin A1 / chloroquine flux assay: Add lysosomal inhibitor (bafilomycin A1 10–40 nM in vitro; chloroquine 50 mg/kg IP in vivo) to block lysosomal acidification. Compare LC3-II with vs without inhibitor. The difference (ΔLCII) = autophagic flux rate. This is the gold standard for flux measurement.
• Tandem mCherry-GFP-LC3 assay (pH-sensitive): The most rigorous method. GFP is quenched at lysosomal pH (~4.5); mCherry is not. Autophagosome = yellow (GFP+mCherry), autolysosome = red only (mCherry). Ratio of red:yellow puncta = flux completion rate. Requires stable or transient transfection — feasible in cell culture but challenging in vivo (requires AAV or transgenic mouse).
• Electron microscopy: Gold standard for morphological confirmation — autophagosomes appear as double-membrane vesicles 0.5–1.5 μm diameter, autolysosomes as single-membrane electron-dense organelles with degraded cargo. Required for publication in high-impact journals but impractical as primary endpoint.

ATG Knockout Controls

Genetic autophagy controls are essential for mechanistic validation. The most commonly used are:

• Atg5-KO / Atg7-KO: Complete macroautophagy block. Atg5 and Atg7 are required for LC3 lipidation. Any compound effect on LC3-II is abolished in these knockouts, confirming the endpoint is autophagy-specific.
• Beclin-1 heterozygous knockout (Becn1+/-): Partial autophagy impairment. Reduces autophagic flux ~50% without complete block. Useful for dose-response study designs.
• FIP200-KO: Blocks ULK1 complex-dependent autophagy specifically (not mTOR-independent pathways). Useful for distinguishing canonical vs non-canonical autophagy.
• SIRT1-KO or EX-527 SIRT1 inhibitor (10 μM in vitro / 10 mg/kg IP in vivo): Blocks SIRT1-mediated deacetylation. If NAD+ compound effect on autophagy is abolished by EX-527, it confirms SIRT1 dependence.
• Atg16L1 T300A knock-in: Disease-relevant human variant that impairs xenophagy (pathogen-targeted autophagy) without blocking bulk macroautophagy — useful for immunology-focused autophagy studies.

Autophagic Flux Timeline: When to Measure

Autophagy induction is rapid — autophagosomes begin forming within 30–60 minutes of mTORC1 suppression. However, chronic changes in basal autophagy relevant to longevity research require weeks to detect reliably in tissue.

• Acute endpoint (in vitro, cell culture): 2–6 hours post-treatment. Measure LC3-II ± bafilomycin A1.
• Acute endpoint (in vivo, bolus injection): 30–120 minutes post-injection for liver; 4–8 hours for muscle.
• Chronic endpoint (longevity protocol): 4–8 weeks minimum for measurable changes in baseline LC3-II and p62 in aged animals. Sample at consistent ZT and feeding state.
• Clearance rate estimation: Use pulse-chase with photoactivatable GFP-LC3 (PA-GFP) — photoactivate a cohort of autophagosomes, measure fluorescence decay over 30–60 minutes. Clearance half-time ~15–25 minutes under active flux.

Pre-Analytical Standardization

Autophagy is acutely sensitive to three pre-analytical variables that confound data if not controlled:

• Feeding state: Remove food at consistent ZT (light phase onset, ZT0). Sample at consistent interval (6h fast for liver; 6–8h for muscle). Free feeding vs 6h fast can shift LC3-II 2–3-fold.
• Anesthesia: Isoflurane inhalation suppresses AMPK and activates mTORC1 acutely (within 10–15 minutes). Use rapid decapitation for liver and muscle samples when autophagic endpoints are primary. If anesthesia is unavoidable, standardize dose and time to death.
• Temperature: Hypothermia activates autophagy. Maintain consistent body temperature throughout protocol. Rectal probe monitoring is best practice for all animals in the same study.

Compound Comparison Table

The following summarizes primary autophagy mechanisms and expected endpoint effects for key research compounds at standard preclinical doses:

• NAD+ (NMN 500 mg/kg IP or direct IP): SIRT1→LKB1→AMPK→ULK1; LC3-II increases, p62 decreases, enhanced flux in aged tissue. mTORC1 not required to be suppressed (SIRT1 bypasses mTORC1 via ATG protein deacetylation).
• MOTS-c (5 mg/kg IP daily): AICAR/folate-cycle AMPK activation → ULK1/mTORC1 suppression. LC3-II increases, p62 decreases, particularly in aged skeletal muscle.
• SS-31 (3 mg/kg IP daily): Cardiolipin/IMM protection → reduces PINK1/Parkin mitophagy signal. May reduce LC3-II and p62 (less mitophagy substrate). Interpret carefully — SS-31 improves mitochondrial quality, reducing demand for mitophagic clearance.
• GLP-1 agonists (semaglutide 0.1–0.3 mg/kg SC 1×/wk; liraglutide 0.2 mg/kg SC daily): cAMP/PKA/AMPK → mTORC1 suppression. LC3-II increases, p62 decreases in liver and neurons. Must use pair-fed controls.
• MK-677 (6 mg/kg SC daily): GH/IGF-1/Akt/mTORC1 activation → autophagy suppression expected. LC3-II likely decreases or unchanged. Not recommended for protocols where autophagy enhancement is a primary endpoint.
• Ipamorelin (100–200 mcg/kg SC 1–3×/day): GH pulse → indirect IGF-1 rise → modest mTORC1 activation. Expected effect: small autophagy suppression, but pulsatile pattern may be insufficient for sustained mTORC1 elevation. Data sparse; pair-fed design needed to dissociate from weight/metabolic effects.

Study Design for Autophagy-Focused Longevity Protocols

A well-powered autophagy study requires attention to genetic controls, flux assay components, and tissue selection. The following is a minimum viable design for a chronic longevity protocol testing NAD+ + MOTS-c combination autophagy effects in aged mice:

• Animals: 18–22 month C57BL/6J, equal male/female per group, n=8–10 per group (power calculation at 25–30% LC3-II change, CV=20%, α=0.05, β=0.20).
• Groups: (1) Vehicle; (2) NMN 500 mg/kg IP 5×/wk; (3) MOTS-c 5 mg/kg IP daily; (4) NMN + MOTS-c; (5) Rapamycin 2 mg/kg IP 3×/wk (positive control).
• Endpoint: Week 8. Liver, skeletal muscle (gastrocnemius), and brain cortex snap-frozen at ZT0 after 6h fast. Rapid decapitation, no isoflurane.
• Assays: LC3-II/LC3-I western blot; p62 western blot; chloroquine flux arm (50 mg/kg IP chloroquine 4h pre-harvest in n=4 per group); SIRT1 deacetylase activity assay (fluorescent substrate); AMPK Thr172 phosphorylation; mTORC1 p-S6K Thr389.
• Optional: Tandem mCherry-GFP-LC3 in MEF cell culture as parallel mechanistic validation.
• Reporting: Separate male/female data; include individual data points and SEM; report ΔLCII for flux quantification.

Research Design Considerations

• Feeding state is the strongest confounder — standardize ZT and fast duration before any other optimization step. Inconsistent feeding windows make LC3-II data uninterpretable.
• Always include a lysosomal inhibitor flux arm (bafilomycin or chloroquine) in at least a subset of animals. Without flux data, increased LC3-II is ambiguous.
• For in vivo SIRT1-autophagy dissection, use EX-527 (10 mg/kg IP) or Atg5 tissue-specific KO (AAV-Cre in floxed Atg5 mice). Systemic Atg5-KO is lethal postnatally — use conditional models.
• MK-677 combinations: If designing a stack with a GH secretagogue + NAD+/MOTS-c for muscle hypertrophy and autophagy simultaneously, be aware that IGF-1/mTORC1 elevation from GH may partially counteract AMPK-driven autophagy. This is a therapeutic window design challenge, not necessarily a failure of either mechanism.
• Aged animals: 18+ month C57BL/6J from NIA Aged Rodent Colonies are preferred. Younger animals have higher basal autophagy and narrower dynamic range for enhancement.
• Sex differences: Female rodents show higher basal AMPK activity and SIRT1 expression in some tissues (estrogen upregulates both). Sex-stratified autophagy analysis is recommended — SIRT1-dependent autophagy may show greater fold-change in females.

Reconstitution and Storage Notes

For NAD+ injection: Use sterile saline or PBS — never BAC water (BAC is an alcohol that can oxidize NAD+). Prepare fresh before each injection. If batching injections, store aliquots at -80°C and thaw immediately before use. Protect from light at all times (NAD+ absorbs UV at 259 nm).

For MOTS-c: Reconstitute in sterile water or saline. 200 mcg/vial typical lyophilized format; add 1 mL sterile water for 200 mcg/mL stock. Stable 14 days at 4°C, 6 months at -20°C lyophilized. MOTS-c is a small 16-amino-acid peptide — stable in solution compared to larger compounds.

For SS-31: Sterile saline is the recommended vehicle — no BAC water (benzyl alcohol can affect mitochondrial membrane potential, which is a primary endpoint in SS-31 studies). Reconstitute at 1 mg/mL in normal saline; inject SC or IP at 0.1–0.3 mL/mouse.

Autophagy research is one of the most technically demanding areas in longevity biology. The distinction between flux and static marker accumulation, the acute sensitivity of autophagy to feeding state and anesthesia, and the opposing mTOR effects of different research compounds require careful protocol architecture. For researchers building multi-compound longevity stacks, autophagy represents a convergence point where NAD+, MOTS-c, and GLP-1 agonists all operate — making it one of the most informative mechanistic endpoints available.