Cytosolic acetyl-coenzyme A is a signalling metabolite to control mitophagy

#1

https://www.nature.com/articles/s41586-025-09745-x

These people have their tanks on my lawn. That makes it a bad time to be drunk. However, one needs to deal with reality as it is. Hence I will probably say more about this tomorrow.

3 Likes
#2

I asked Claude for a summary:

I’ll fetch and summarize that Nature article for you.This Nature paper describes a mechanism where cytosolic acetyl-coenzyme A (AcCoA) acts as a signaling metabolite that controls mitophagy (selective degradation of mitochondria) through its receptor NLRX1.

Key Findings

The AcCoA-NLRX1 signaling axis:
When cytosolic AcCoA levels decrease—through short-term fasting or inhibition of ATP-citrate lyase, mitochondrial citrate/malate antiporter, or acyl-CoA synthetase—mitophagy is triggered. This can be reversed by acetate supplementation.

NLRX1 as the AcCoA sensor:
NLRX1, a mitochondria-localized NOD-like receptor, directly binds cytosolic AcCoA within a conserved pocket on its leucine-rich repeat domain. When AcCoA binds, it enhances the interaction between NLRX1’s LRR and NACHT domains, maintaining NLRX1 in an autoinhibited state that prevents its association with LC3. When AcCoA levels drop, NLRX1 becomes activated, oligomerizes, and recruits LC3 to initiate mitophagy.

Therapeutic relevance:
The study reveals that this AcCoA-NLRX1 axis underlies KRAS-inhibitor-induced mitophagy and promotes drug resistance in KRAS-mutant cancers. NLRX1 deficiency enhanced the anti-tumor effects of KRAS inhibitors by preventing compensatory mitophagy and increasing oxidative stress.

This represents a novel non-canonical function for AcCoA beyond its traditional role as a metabolic substrate or acetylation donor—it acts as a direct signaling molecule linking cellular metabolism to mitochondrial quality control.

I also asked a question of chatGPT

Short version:
NLRX1-dependent mitophagy looks genuinely selective, but in a narrow, signal-defined way rather than being the universal selector for all mitophagy.

1. Is NLRX1 mitophagy “selective” in the classical sense?

Yes. Several lines of evidence now treat NLRX1 as a bona fide mitophagy receptor:

  • NLRX1 has a LC3-interacting region (LIR) and can recruit LC3-positive autophagosomes specifically to mitochondria, which is the definition of a selective receptor in this context.(MDPI)
  • In Listeria infection, NLRX1 is required for targeting LC3 to damaged mitochondria, reducing mtROS and promoting mitochondrial clearance, with little evidence for a global increase in non-mitochondrial autophagy.(MDPI)
  • A 2025 Nature paper using a genome-wide CRISPR screen identified NLRX1 as the top mitophagy receptor for mitophagy induced by reduced cytosolic acetyl-CoA. Knockout of NLRX1 largely abolished mitophagy by multiple readouts (mt-Keima, mtDNA/nDNA, mitochondrial proteins, LC3 on mitochondria), but did not prevent general LC3 lipidation or p62 degradation. The authors explicitly state that **“cytosolic AcCoA-reduction-induced mitophagy is highly selective and requires the mitophagy receptor NLRX1.”*(Nature)

So at the level of cargo (mitochondria vs everything else) and machinery (LIR-LC3 interaction), NLRX1 participates in classically selective mitophagy.

2. What is it selective for?

Current data suggest several layers of selectivity:

a) Selective for particular mitochondrial “danger signals”

  • Mitochondrial protein import stress (MPIS): NLRX1 has been proposed to act as a pattern-recognition receptor for defective mitochondrial protein import. MPIS (e.g. MIA40/ERV1 inhibition, CCCP or MB-6) triggers NLRX1-dependent LC3 lipidation and mitophagy.(Cell)

  • Cytosolic Ac-CoA drop without classic damage: In the Nature paper, inhibiting ACLY/SLC25A1 or fasting lowered cytosolic Ac-CoA and induced mitophagy while:

    • Δψm, ROS, ATP and OXPHOS remained largely intact
    • PINK1 and Parkin were not stabilized or recruited
    • General autophagy proceeded normally in the absence of NLRX1
      Yet mitophagy was almost fully NLRX1-dependent, indicating NLRX1 selects mitochondria based on a metabolic signal (Ac-CoA/MPIS axis) rather than gross damage alone.(Nature)

So NLRX1 appears tuned to specific mitochondrial stress states (import/Ac-CoA–linked) rather than “any depolarized mitochondrion”.

b) Selective at the organelle level (subset of mitochondria)

  • In multiple systems (Listeria infection, TNF-α-stimulated breast cancer cells, MPIS, Ac-CoA reduction), LC3 puncta and mt-Keima signal localize to a fraction of mitochondria, not the entire network, consistent with subpopulation-level mitophagy rather than wholesale mitochondrial removal.(MDPI)
  • The NLRX1–SLC39A7 complex in nucleus pulposus cells promotes selective mitochondrial fission plus mitophagy to maintain mitochondrial morphology; loss of NLRX1 causes mitochondrial collapse and then a secondary PINK1-PRKN-driven, excessive mitophagy.(PubMed)
    That implies NLRX1 normally gates a more targeted, adaptive mitophagy mode, while its absence lets a more indiscriminate backup pathway take over.

c) Selective for certain stimuli and cell types

  • Microglia vs astrocytes, morphine model: chronic morphine triggers NLRX1-dependent mitophagy in microglia, with direct NLRX1–LC3 interaction via the LIR motif; Δψm is unchanged and PINK1-Parkin is not involved. In astrocytes, neither NLRX1 nor mitophagy is significantly engaged under the same conditions.(BioMed Central)
  • Reviews now emphasize that NLRX1’s effects on autophagy/mitophagy are strongly cell-type specific, with immune cells, tumor cells, and NP cells showing different NLRX1–mitophagy wiring.(MDPI)

So NLRX1-mediated mitophagy is stimulus- and cell-type-restricted, not a universal housekeeping pathway.

d) Selective via additional receptor axes

  • A cardiac/vascular–oriented study describes an NLRX1–FUNDC1–NIPSNAP1/2 axis, where NLRX1 modulates the outer-membrane receptor FUNDC1 and “eat-me” signals NIPSNAP1/2 to regulate mitophagy.(Wiley Online Library)
  • NLRX1 agonist NX-13 enhances its interaction with the ZnÂČâș transporter SLC39A7, adjusting mitochondrial ZnÂČâș flux, dynamics, and mitophagy; NLRX1-deficient cells instead flip into PINK1-PRKN-driven excessive mitophagy and senescence.(BioMed Central)

These axes reinforce that NLRX1 is coupled to specific receptor complexes and mitochondrial states, not to generic bulk autophagy.

3. What it is not

  • NLRX1 is not required for all mitophagy. Classical PINK1–Parkin mitophagy proceeds in NLRX1-deficient settings, and several other receptors (BNIP3, NIX, FUNDC1 itself, PHB2, cardiolipin, FKBP8, BCL2L13, NIPSNAPs, etc.) mediate mitophagy independently.(rupress.org)
  • NLRX1 is not a general on/off switch for autophagy: in the Ac-CoA/fasting paper, NLRX1 knockout did not block bulk autophagosome formation or p62 turnover; it mainly prevented mitochondrial cargo capture.(Nature)
  • Even for mitochondrial damage, some contexts (e.g. strong CCCP in microglia) can engage mitophagy without major changes in NLRX1, suggesting parallel or alternative pathways.(BioMed Central)

4. Bottom line

If you phrase it as a spectrum:

  • Bulk autophagy (non-selective) ⟷ PINK1-Parkin mitophagy (damage-selective) ⟷ receptor- and signal-defined mitophagy

NLRX1 sits toward the signal-defined, receptor-mediated end:

  • It’s selective for mitochondria (via LIR/LC3 and mitochondrial localization).
  • It’s selective for specific stress signatures (MPIS / Ac-CoA drop, certain toxins, TNF-α, Listeria, morphine) rather than all forms of mitochondrial impairment.
  • It shows cell-type and receptor-complex specificity (microglia, NP cells, tumor cells; FUNDC1/NIPSNAP, SLC39A7, RRBP1).

So: NLRX1-mitophagy is quite selective, but only for a defined subset of mitochondrial stress states and cell contexts, not a universal selector for all mitophagy events.

#3

I think what is key here is that NLRX1 identifies mitochondria to be recycled. Localised high acetyl-CoA concentration inhibits this process. To what extent the citrate efflux from an individual mitochondrion drives the localised acetyl-CoA concentration and to what extent it drives it more generally in the cytosol is unclear. You would expect some effect, but although I think it is reasonable to consider nuclear flows separate to cytosolic (because ACLY is expressed in the nucleus and the nuclear flows probably mainly come from that) I don’t know whether this picks up particularly an inefficient mitochondrion or just generally gives a panmitochondrial benefit from fasting. It is also not clear how short a fast can be to have this effect. It may be that the normal sleeping fast has this effect if glucose levels drop low enough. Whatever way you look at it, however, fasting glucose levels will be really important for not inhibiting NLRX1.

I asked another question of chatGPT

“what are the factors that affect NLRX1”?

Short version: NLRX1 is highly context-dependent. Its expression, stability, localization and signalling outputs are all tuned by (i) inflammatory and pathogen cues, (ii) mitochondrial stress and metabolism, (iii) post-translational modifications, and (iv) binding partners / small-molecule agonists.

I’ll break it down by “what” is being affected:

1. Factors that affect NLRX1 expression levels

a. Cell type & basal metabolic programme

  • NLRX1 is expressed pretty ubiquitously but its abundance and functional role differ a lot between immune vs non-immune cells and by metabolic profile (glycolytic vs OXPHOS-biased). (MDPI)
  • Fekete et al. explicitly argue that discrepant NLRX1 phenotypes likely arise from this cell-type/metabolic dependence (macrophages, DCs, T cells, epithelial cells, fibroblasts, tumour cells all behave differently). (MDPI)

b. Inflammatory cues (TLR ligands, TNF-α, etc.)

  • LPS / TLR stimulation: NLRX1 is part of a negative-feedback loop on NF-ÎșB. NLRX1-deficient macrophages exposed to LPS produce more IFN-ÎČ and IL-6, implying that PRR signalling both induces and then is restrained by NLRX1. (MDPI)
  • TNF-α in cancer cells: in breast cancer lines, TNF-α triggers mitochondria-lysosome crosstalk where NLRX1 supports invasive behaviour and modulates mitophagy; NLRX1 levels and its mitochondrial association are tuned by TNF-α exposure. (PubMed)

c. Tissue injury / disease state

  • NLRX1 expression is downregulated in several human pathologies. For example, in brain tissue from ruptured aneurysm patients and in COPD airway epithelium, NLRX1 levels are reduced and this correlates with heightened NF-ÎșB signalling and inflammation. (MDPI)

So: PRR signalling, cytokines, injury and disease-specific transcriptional programmes all move NLRX1 expression up or down.

2. Factors that affect NLRX1 post-translationally (stability & activity)

a. O-GlcNAcylation and ubiquitination (nutrient / glucose status)

  • Chen et al. show that NLRX1 is O-GlcNAc-modified in M1 macrophages. High O-GlcNAcylation increases NLRX1 ubiquitination and decreases its stability, whereas inhibiting O-GlcNAc transferase (OGT) with OSMI-1 raises NLRX1 protein levels. (PubMed)
  • Because O-GlcNAc is a classic glucose/hexosamine pathway read-out, this effectively links NLRX1 abundance to cellular nutrient status.

b. Polyubiquitination switching binding partners

  • In resting cells NLRX1 can bind TRAF6 and dampen NF-ÎșB; upon activation, polyubiquitination promotes its dissociation from TRAF6 and re-binding to the IKK complex, altering how it regulates NF-ÎșB. (MDPI)

So: glucose flux → O-GlcNAc → ubiquitination → NLRX1 half-life and interaction profile.

3. Factors that affect NLRX1 localization and mitophagy function

a. Mitochondrial protein import status / “import stress”

  • NLRX1 has an N-terminal mitochondrial targeting sequence and can be in outer/inner membrane, matrix, or cytosol. (MDPI)
  • Under mitochondrial protein import stress (e.g. misfolded preproteins), NLRX1 is stabilized in the cytosol, where it can bind the ER protein RRBP1 and together they promote LC3 lipidation and mitophagy. (ScienceDirect)
  • So the efficiency of mitochondrial import machinery directly changes where NLRX1 sits and whether it acts as a mitophagy driver.

b. Pathogen virulence factors (e.g. Listeriolysin O)

  • In Listeria monocytogenes infection, the toxin LLO induces NLRX1 oligomerization. NLRX1’s LC3-interacting region (LIR) then binds LC3 and drives Listeria-induced mitophagy. (PubMed)
  • Here, a specific bacterial toxin is the “factor” that flips NLRX1 into an oligomeric, LC3-binding mitophagy receptor.

c. Mitochondrial ROS and oxidative damage

  • Various models (e.g. TBHP treatment in NP cells) show that oxidative mitochondrial damage is a key upstream trigger that recruits NLRX1 to mediate “beneficial” mitophagy and support mitochondrial quality control. (PubMed)

d. Cytokine-driven, context-specific mitophagy

  • In aggressive breast cancer cells, NLRX1 actually inhibits TNF-α-induced mitophagy while altering mitochondrial dynamics, favouring invasive behaviour. (PubMed)

Net: whether NLRX1 promotes or restrains mitophagy depends on what upstream stressor you’re dealing with (bacterial toxin vs TNF-α vs import stress) and on cell type.

4. Factors that affect NLRX1 binding partners and signalling outputs

From the big reviews and mechanistic papers, NLRX1’s behaviour is heavily shaped by who it’s bound to at a given moment: (MDPI)

  • MAVS / TRAF6 / STING – pathogen-sensing axis (RIG-I, cGAS–STING)
    NLRX1 binding here suppresses type I IFN and NF-ÎșB signalling. Viral RNA, dsRNA mimetics, and specific viral proteins all change this interaction landscape. (PMC)

  • TUFM, Beclin1-UVRAG, LC3 – macroautophagy & LC3-associated phagocytosis
    Depending on pathogen (virus, bacteria, fungi) NLRX1–TUFM can promote or restrain autophagy and LC3 loading. (MDPI)

  • SLC39A7 (ZIP7) – ZnÂČâș trafficking and mitochondrial dynamics
    In nucleus pulposus cells, NLRX1 forms a complex with the zinc transporter SLC39A7 on mitochondria; this modulates mitochondrial ZnÂČâș, fission/fusion proteins, and mitophagy, with clear effects on disc cell senescence. (PubMed)

  • FASTKD5, UQCRC2, DRP1, SARM1, etc. – ETC function, ROS, cell death
    Interaction with these partners toggles NLRX1’s effects on ROS generation, ETC complex maturation, and apoptosis vs necrosis, again in a cell-type-dependent manner. (MDPI)

So: available binding partners (which are themselves regulated by cell type, stress, and transcriptional status) strongly “bias” what NLRX1 actually does.

5. Small-molecule and exogenous modulators

a. NX-13 and related agonists

  • NX-13 is a gut-restricted small-molecule NLRX1 agonist that binds and stabilizes NLRX1 (thermal shift assays) and skews CD4 T-cell metabolism toward OXPHOS with reduced NF-ÎșB and ROS, ameliorating IBD in mice and ex vivo UC PBMCs. (PubMed)
  • In disc cells, NX-13 restores NLRX1 function and rescues mitochondrial dynamics and mitophagy, reducing NP cell senescence. (PubMed)

b. Other drugs / toxins

  • Morphine and other stimuli in microglia have been reported to act through NLRX1-dependent mechanisms (e.g. via its LIR motif and LC3 binding), indicating that some pharmacological agents can indirectly modulate NLRX1-mediated autophagy/mitophagy. (BioMed Central)

6. Genetic variation and disease-linked regulation

  • Human genetic variation in NLRX1 is being linked to susceptibility or severity in COPD, autoimmune disease, and cardiovascular injury; these variants can alter expression, localization, or interaction motifs, but this field is still relatively sparse. (MDPI)

If you want to go deeper

If you tell me the specific context you care about (e.g. NP cells under oxidative stress vs CD4 T cells in autoimmune disease vs tumour cells under TNF-α), I can map out a context-specific “regulation diagram” for NLRX1: upstream factors → modifications/localization → binding partners → mitophagy/autophagy vs signalling outcomes.