CoChat
Sign Up
PreprintCited by 0Alzheimer's disease research and treatmentsNeuroinflammation and Neurodegeneration MechanismsNeuroscience and Neuropharmacology Research

NAD+ hydrolase Sarm1 is a key driver of synapse degeneration and memory loss in Alzheimer’s disease

Open access
Fan, Fan, Joshi, Prince, Liu, Xiaojie, Kotturu, Sandeep Kumar, Abiola, Michael, Liu, Qingsong, Lou, Xuelin·2025·bioRxiv (Cold Spring Harbor Laboratory)
DOI 10.64898/2025.12.21.695860PMC PMC12776179
Open this paper in CoChat — free accountOpen PDF

Abstract

Synapse degeneration is a hallmark of neurodegenerative diseases 1–3 , including Parkinson’s and Alzheimer’s disease (AD). Synapse loss has been known for decades as the strongest correlate with cognition and disease progression in AD patients 4–8 , but the molecular mechanisms that drive synapse degeneration remain elusive. Here, we identify Sarm1, a new class of NAD+ hydrolase required for Wallerian degeneration of periphery nerves after injuries 9–12 , as the key mediator of synaptic degeneration in AD brains. Sarm1 knockout largely reversed synapse loss, amyloid-β (Aβ) burden, and cognitive decline in 5XFAD mice. We found that Sarm1 is enriched in synaptic terminals and becomes activated in synaptic dystrophies adjacent to Aβ plaques, leading to synapse degeneration and subsequent neuroinflammation. Sarm1 deletion in the AD mice prevented synaptic dystrophies and rescued short-term and long-term synaptic plasticity. Further, Sarm1 deletion protected synapses from C1q tagging and phagocytosis, and C1q–MERTK signaling in complement cascades 7,13–15 was significantly reduced. The reduced synapse degeneration, in turn, broke the feed-forward loop of “glia activation–neuroinflammation–Aβ deposition”. These data suggest that Sarm1 plays a key role in driving synapse degeneration in AD before C1q-tagging and phagocytic clearance, and targeting Sarm1 may offer a novel intervention to attenuate synapse degeneration and memory loss in AD. Highlight: Sarm1 is enriched at presynaptic dystrophies and correlated with Aβ. Genetic deletion of Sarm1 prevents synapse degeneration in AD. Sarm1 depletion is sufficient to reverse synaptic dysfunction and memory loss. Sarm1 depletion reduces Aβ burden and neuroinflammation C1q--MERTK axis acts downstream of synaptic Sarm1 activation.

AI snapshot

Written from the paper’s full text.

TL;DR

In 5xFAD mice, Sarm1 is enriched and activated at presynaptic dystrophies, and genetic Sarm1 deletion prevents synapse degeneration, reverses synaptic plasticity and memory deficits, and reduces Aβ burden and neuroinflammation. The proposed mechanism is Aβ stress → Sarm1 activation/NAD+ depletion → synapse destruction → C1q–MERTK-dependent phagocytic clearance.

Key findings

  • Sarm1 was enriched at presynaptic terminals and dystrophic synapses in 5xFAD hippocampus, with active p-Sarm1 (Ser548) accumulating around Aβ plaques. Sarm1 deletion in 5xFAD mice reduced axonal dystrophy number by over 50% and reduced presynaptic dystrophy abundance and size.
  • Sarm1 KO reversed AD-related synaptic physiology deficits: impaired fEPSP input/output, paired-pulse ratio (p < 0.005), and LTP (quantified 50–60 min after TBS; n = 11–15 recordings/group; F3,46 = 6.5, p < 0.001). Sarm1 KO alone did not alter neurotransmission.
  • Sarm1 KO rescued context-associated fear memory and elevated plus maze behavior, reduced Aβ plaques across hippocampus, cortex, and thalamus, and reduced glia activation/neuroinflammation. AD-associated increases included TNFα 3.2-fold, CCL3 25-fold, CCL4 10-fold, and CCL5 7-fold; CCL3 was reduced by 75% in rescue mice.
  • Sarm1 KO reduced CD68+ lysosomes containing PSD95 or gephyrin, reduced C1q- and C3-tagged synapses, and reversed increases in C1qA, C1qB, C1r, and MERTK mRNAs, supporting a C1q–MERTK axis downstream of synaptic Sarm1 activation.

Methods

  • Sarm1-deficient AD mice were generated by crossing 5xFAD with Sarm1 KO mice on a congenic C57BL/6J background. Age-matched male and female mice (6–7 months old) were used in WT, 5xFAD, and 5xFAD-Sarm1KO groups, with Sarm1 KO alone as an additional control.
  • Analyses included confocal/3D imaging and morphometrics, acute hippocampal slice electrophysiology (fEPSP, PPR, LTP with theta burst stimulation), fear conditioning and elevated plus maze behavior, qPCR for cytokines/chemokines/complement factors, and Western blot/immunofluorescence.
  • Statistics used mean ± SEM, two-tailed Student’s t-test, and one-way or two-way ANOVA with Tukey’s or Dunnett’s post hoc tests; p < 0.05 was considered significant.

Limitations

  • The paper states that Sarm1 is also expressed in astrocytes, microglia, vascular cells, and macrophages at lower levels, with reported positive or negative effects on inflammation; although predominant neuronal expression favors a major neuron-intrinsic role, contributions from other cell types are not excluded.
  • Text-evident: all experiments are in the 5xFAD mouse model at 6–7 months, with no human AD tissue or clinical validation reported.
  • Text-evident: the preprint is not certified by peer review, and the NAD+ depletion step in the proposed mechanism is inferred from Sarm1 biology rather than directly measured in this study.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.64898/2025.12.21.695860

Cites

143 of 143 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

19 of 35 cited sources checked · 8 supported · 0 flagged · accuracy not yet measured

  • Findingown data

    p-Sarm1 (Ser548) accumulates around Aβ plaques in 5xFAD hippocampus.

    “We further detected strong accumulation of p-Sarm1 (Ser548), an active form of Sarm1 crucial for NAD + cleavage activity 69 , around Aβ plaques (Fig. 1i), suggesting Sarm1 activation in those sites.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.
  • Premisecitation

    Impairment of fear conditioning in 5xFAD mice starts at 6 months of age.

    “Impairment of fear conditioning in 5xFAD mice starts at 6-m of age 83 ; thus, we examined this in 6-month-old mice.”
    Quote in the paperSubstantial claim1 of 1 sources checked
  • Premisecitation

    Sarm1 acts as a metabolic sensor to monitor local NAD+ levels.

    “Along with NAD + synthase NMNAT2 at the upstream 54 , Sarm1 acts as a metabolic sensor to monitor local NAD + levels 12 .”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [12] source

      The Discussion identifies SARM1 as a metabolic sensor, but specifies that it responds to the NMN/NAD+ ratio rather than NAD+ levels alone. The Results section on SARM1 activation in neurons likewise says the ratio predicts activation; the claim omits that qualification. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
  • Findingown data

    The LTP reduction in AD is fully rescued by Sarm1 KO.

    “Importantly, the LTP reduction in AD was fully rescued by sarm1 KO (Fig. 2i).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.
  • Findingown data

    Sarm1 KO causes a consistent reduction of Aβ plaques across hippocampus, cortex, and thalamus.

    “Upon Sarm1 KO, regionspecific quantification revealed a consistent reduction of Aβ plaques across hippocampus, cortex, and thalamus (Fig. 3h)”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.
  • Classified as probably not a substantial claim. 19
  • May combine two claims. 42
  • The quote only partly states this claim. 7

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Synapse degeneration is widely observed across multiple neurodegenerative diseases and is increasingly recognized as a central cause of dementia.

    “Synapses are the hub of neuron communications, and their degeneration is widely observed across multiple neurodegenerative diseases and increasingly recognized as a central cause of dementia 2,4,5,8,16 .”
    Quote in the paperNot a substantial claim1 of 5 sources checked
    • cites [2] source
    • cites [4] source
    • cites [5] source
    • cites [8] source
    • cites [16] source

      The Introduction (p. 4) describes synaptic loss as a seemingly universal feature of dementia pathology and links it to cognitive deficits across Alzheimer’s and Lewy body dementias. The Discussion (pp. 20–21) likewise reports synaptic changes across these diseases, but says declining synaptic function, rather than synaptic loss, may play the more relevant role in dementia progression. The claim’s description of synapse degeneration as a central cause is stronger than the paper’s qualified conclusion. (LLM, conf=0.83)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data

    There were no significant changes in TMEM106B and MLKL in AD brains.

    “we examined TMEM106B 101-103 and necroptosis effector MLKL 104,105 (Fig. 5c-d), two molecules previously implicated in neurodegeneration of AD, and detected no significant changes.”
    Quote in the paperNot a substantial claim3 of 5 sources checked
    • cites [101] source

      The Discussion says TMEM106B fibrils were not observed in the authors’ previous Alzheimer’s disease studies, but it does not report a statistical comparison of TMEM106B levels in AD brains. Searches for significant changes and MLKL found no relevant results in our copy; the cited paper therefore does not establish the claimed lack of significant changes. (LLM, conf=0.77)

      Experimental — accuracy not yet measured
    • cites [102] 2023, C-terminal TMEM106B fragments in human brain correlate with disease-associated TMEM106B haplotypes, Brain source
    • cites [103] source
    • cites [104] 2021, TNF-mediated neuroinflammation is linked to neuronal necroptosis in Alzheimer’s disease hippocampus, Acta neuropathologica communications

      Not checked in the free run.

    • cites [105] source
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Premisecitation

    Synapse degeneration occurs before neuron loss and long before disease diagnosis.

    “Synapse degeneration is an insidious process, occurring before neuron loss and long before disease diagnosis 4,17 .”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [4] source
    • cites [17] source

      The Results section reports changes in proteins associated with synaptic and neuronal coexpression approximately 26–22 years before expected symptom onset, preceding later markers of neuronal loss. But these early protein changes do not establish synapse degeneration; the paper describes frank synaptic and neuronal loss together at a later stage. It also measures time relative to symptom onset, not disease diagnosis. (LLM, conf=0.84)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    Aβ disrupts synaptic transmission and induces synapse loss.

    “While Aβ disrupts synaptic transmission 32,34,35 and induces synapse loss 17”
    Quote in the paperSubstantial claim2 of 4 sources checked
    • cites [32] source
    • cites [34] source
    • cites [35] source
    • cites [17] source

      The Results section reports decreases in neuronal and neurosecretory proteins during the course of autosomal dominant Alzheimer’s disease, which the authors interpret as suggesting synaptic and neuronal loss. It does not establish that Aβ induces that loss. The cited paper is attached to the synapse-loss clause; the separate synaptic-transmission clause has other citations. (LLM, conf=0.82)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    5xFAD mice (6 months or older) prefer open arms due to anxiety or disrupted vibrissa inhibition.

    “5xFAD mice (6-m or older) prefer open arms 134 due to anxiety or disrupted vibrissa inhibition 135 .”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites [134] 2012, Motor deficits, neuron loss, and reduced anxiety coinciding with axonal degeneration and intraneuronal Aβ aggregation in the 5XFAD mouse model of Alzheimer’s disease, Neurobiology of aging

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites [135] source

      The Discussion supports disrupted vibrissal inhibition as a possible explanation for closed-arm avoidance, but explicitly says the open-arm behavior cannot be attributed to decreased anxiety. The Methods > Behavior section reports testing mice at 9–11 months, so this paper does not establish the claim for all mice aged 6 months or older. (LLM, conf=0.88)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    Sarm1 remains inactive in physiological conditions but becomes activated in pathological conditions.

    “For this reason, Sarm1 remains inactive in physiological conditions but becomes activated in pathological conditions 12,48 .”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [12] source
    • cites [48] source

      The “Insights into SARM1 activation from structural biology” section (page unavailable) proposes that SARM1 is stabilized in an inactive state under normal physiological conditions and may become activated after injury. The claim presents this proposed mechanism as established and extends injury-related activation to pathological conditions generally. (LLM, conf=0.85)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Sarm1 knockout mice show few phenotypes.

    “Accordingly, Sarm1 knockout (KO) mice show few phenotypes 49,50”
    Quote in the paperNot a substantial claim2 of 2 sources checked
    • cites [49] source
    • cites [50] source

      The Results sections on RGC degeneration and function examine Sarm1 knockout in mice carrying an Opa1 mutation; their controls are Sarm1 heterozygotes. Our copy reports protection in that disease model but does not assess whether Sarm1 knockout mice generally show few phenotypes. Searches for normal development and baseline phenotypes found no such assessment. (LLM, conf=0.72)

      Experimental — accuracy not yet measured

  • Classified as probably not a substantial claim.

    Premisecitation

    Sarm1 TIR domain cleaves NAD+ into cyclic ADP-ribose (cADPR) as a new class of NAD+ hydrolases.

    “Subsequent work showed that Sarm1 TIR domain cleaves NAD + into cyclic ADP-ribose (cADPR) as a new class of NAD + hydrolases 10,44”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [10] source

      The passage on NAD+ consumption (section labeled Abstract; no page available) reports that SARM1 TIR dimerization causes NAD+ consumption and nicotinamide release. It does not identify cyclic ADP-ribose as a product or establish that the TIR domain itself is a new class of NAD+ hydrolases. (LLM, conf=0.87)

      Experimental — accuracy not yet measured
    • cites [44] source

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Context-associated memory deficit is typical in AD patients and manifested as dysfunction of fear conditioning involving neural circuits of the amygdala, hippocampus, and prefrontal cortex.

    “Context-associated memory: This deficit is typical in AD patients [128][129][130] and manifested as dysfunction of fear conditioning 82 involving neural circuits of the amygdala, hippocampus, and prefrontal cortex 131,132 .”
    Quote in the paperNot a substantial claim2 of 6 sources checked
    • cites [128] source

      The Discussion reports impaired fear conditioning in Alzheimer’s disease, and the Introduction identifies the amygdala and ventromedial prefrontal cortex as regions involved in fear conditioning. However, the paper does not identify the hippocampus as part of the fear-conditioning circuitry studied or establish that this impairment manifests a context-associated memory deficit in Alzheimer’s disease. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites [129] source
    • cites [130] source
    • cites [82] source
    • cites [131] source
    • cites [132] source

      The paper’s Abstract (page unavailable) links fear conditioning and fear-memory reconsolidation to the lateral amygdala in rats. It does not establish the involvement of the hippocampus or prefrontal cortex in the stated circuit. The AD-patient deficit is attached to different citations in the citing sentence, so it is not assessed against this source. (LLM, conf=0.87)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data

    Sarm1 deletion rescued synapse destruction and memory loss.

    “Sarm1 deletion rescued synapse destruction and memory loss.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    Sarm1 depletion in AD mice strongly reduces dystrophic synapses and aberrant phagocytosis.

    “At the cellular level, Sarm1 depletion in AD mice strongly reduces dystrophic synapses and aberrant phagocytosis.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    Sarm1 KO rescues the behavioral deficits in AD to control levels.

    “Again, Sarm1 KO rescued these behavioral deficits in AD to control levels.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    The levels of C1q receptor MERTK were elevated moderately in AD mice but reduced significantly in the rescue mice.

    “Interestingly, the levels of C1q receptor MERTK were elevated moderately in AD mice but reduced significantly in the rescue mice (Fig. 6n).”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    Sarm1 is prominently present and activated at synapses in AD.

    “our data provide direct evidence that Sarm1 is prominently present and activated at synapses in AD (Fig. 1)”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    Sarm1 KO fully rescues context-associated memory impairment to control levels in AD mice.

    “However, this impairment was fully rescued to control levels upon sarm1 KO, implying a strong protection against context-associated memory loss.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims. The quote only partly states this claim.

    Findingown data

    CCL3, CCL4, and CCL5 are downregulated dramatically in AD mice after Sarm1 KO.

    “Again, they were downregulated dramatically in AD mice after Sarm1 KO.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    Sarm1 KO reverses AD synapse deficits, including synaptic dystrophies, LTP, and synapse loss.

    “Sarm1 KO reverses AD synapse deficits, including synaptic dystrophies, LTP, and synapse loss (Fig. 2, 5, and 6).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    These changes are significantly reduced in the rescue group, and the number of PSD95-containing lysosomes is much lower.

    “However, these changes were significantly reduced in the rescue group, and the number of PSD95-containing lysosomes was much lower (Fig. 5g,h).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    AD mice exhibit a greater number of C1q-tagged synapses than controls, and this change is reversed by Sarm1 KO.

    “AD mice exhibited a greater number of C1q-tagged synapses than controls, and this change was reversed by Sarm1 KO in the rescue group (Fig. 6c).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    Sarm1 KO significantly reduces levels of neuroinflammation and glia activation in AD brains.

    “Interestingly, sarm1 KO significantly reduced the levels of neuroinflammation and glia activation in AD brains (Fig. 4).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    In 5xFAD hippocampus, Sarm1 is highly enriched in dystrophic axons and presynaptic boutons, frequently around or mixed with Aβ plaques.

    “In 5xFAD hippocampus, Sarm1 was highly enriched in dystrophic axons and presynaptic boutons (Fig. 1g-h).”
    • Also stated in: § Sarm1 is enriched at presynaptic terminals and activated in dystrophic synapses. · plain
      “The synaptic dystrophies with Sarm1 puncta are frequently present around or mixed with Aβ plaques (Fig. 1h, Supp.Fig. 2).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    CD68 and GSDMD are significantly increased in AD brains and these increases are reversed by Sarm1 KO.

    “Interestingly, CD68 and inflammatory pyroptosis factor Gasdermin D (GSDMD) were significantly increased in AD brains, and these aberrant increases were reversed in the rescue group by Sarm1KO (Fig. 5c, d).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    AD brain exhibits increased CD68 puncta containing gephyrin, and this increase is reduced significantly by Sarm1 KO.

    “AD brain exhibited increased CD68 puncta containing gephyrin, an inhibitory synapse marker (not shown), and this aberrant increase was reduced significantly by Sarm1 KO in the rescue group (Fig. 5i).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    PPR is impaired in AD mice and this deficit is reversed in the rescue group by Sarm1 KO.

    “While PPR was impaired in AD mice, this deficit was reversed in the rescue group by sarm1 KO (p < 0.005, two-way ANOVA with Tukey's post hoc tests).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingreasoning

    Sarm1 KO in AD mice not only reduces synapse loss but also attenuates astrocyte and microglia activation.

    “Sarm1 KO in AD mice not only reduces synapse loss but also attenuates astrocyte and microglia activation”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Findingown data

    Presynaptic terminals contain abundant Sarm1 in healthy brains, which is absent in Sarm1 KO.

    “Arrowheads indicate that presynaptic terminals (vGlut-1, red) contain abundant Sarm1 (green) in healthy brains, which was absent in Sarm1 KO.”
    Quote in the paperSubstantial claim

  • The quote only partly states this claim.

    Findingown data

    Sarm1 deletion significantly reduces axonal dystrophy number per field of view.

    “b, Significant reduction of axonal dystrophy number in each field of view (FOV) in the rescue group.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

  • May combine two claims.

    Findingown data

    Sarm1KO reduces microglia and astrocyte activation.

    “b-c, quantification of microglia (anti-Iba-1) and astrocytes (anti-GFAP) in three groups of mice, showing the rescue effect of Sarm1KO.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

  • May combine two claims.

    Findingown data

    Sarm1KO prevents microglia dystrophy and restores ramified morphology.

    “e, Sarm1KO prevented microglia dystrophy and restored ramified morphology, as measured by filament coverage and branch numbers of each microglia.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

  • The quote only partly states this claim.

    Findingown data

    Sarm1 deletion significantly reduces axonal dystrophy sizes.

    “c, significant reduction of axonal dystrophy sizes.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Long-term potentiation (LTP) is similar in wild-type (WT) and Sarm1 knockout (KO) mice.

    “C) Similar LTP in WT and Sarm1 KO mice.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    The expression levels of IFN-γ, IL-1β, IL-10, CXCR3, and TGF-β were not changed across WT, Sarm1 KO, AD, and Rescue groups.

    “The expression levels of IFN-γ, IL-1β, IL-10, CXCR3, and TGF-β were not changed across the WT, Sarm1 KO, AD, and Rescue groups of mice.”
    Quote in the paperNot a substantial claim

  • The quote only partly states this claim.

    Findingown data

    Sarm1 deletion significantly reduces synaptic dystrophy number.

    “e, significant reduction of synaptic dystrophy number in the rescue mice.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

  • May combine two claims.

    Findingown data

    Sarm1 KO reverses the increase of C3-tagged synapses and reduction of synaptic spine density in the hippocampus.

    “g-h, Sarm1 KO reverses the increase of C3-tagged synapses (g) and reduction of synaptic spine density (h) in the hippocampus.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

  • May combine two claims.

    Findingown data

    Sarm1 KO reduces CD68 expression and prevents synapse degeneration in both excitatory and inhibitory terminals.

    “g-i, Sarm1 KO reduces CD68 expression and prevents synapse degeneration in both excitatory and inhibitory terminals.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

  • The quote only partly states this claim.

    Findingown data

    Sarm1 KO reverses mRNA changes in TMEM106B, MLKL, CD68, and GSDMD.

    “c-f, mRNA changes in TREM106B, MKLK, CD68, and GSDMD.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

  • May combine two claims. The quote only partly states this claim.

    Findingown data

    Sarm1 KO reverses changes of C3-tagged presynaptic terminals and total presynaptic density.

    “i-j, changes of C3-tagged presynaptic terminals and total presynaptic density in WT, AD, and Rescue group.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

  • May combine two claims. The quote only partly states this claim.

    Findingown data

    Sarm1 deletion reverses impaired synaptic transmission and LTP.

    “Sarm1 deletion reduces Aβ-associated synaptic dystrophy and reverses the impaired synaptic transmission and LTP.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

  • May combine two claims.

    Findingown data

    Sarm1 in dystrophic synapses is phosphorylated at Ser-548, representing an activated form, and is enriched around Aβ plaques.

    “i, Sarm1 in dystrophic synapses was phosphorylated (at Ser-548), which represents an activated form of Sarm1 and was enriched around Aβ plaques.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Sarm1 KO rescues memory loss and cognitive deficits of AD mice.

    “These data demonstrate that Sarm1 KO rescues the memory loss and cognitive deficits of AD mice.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • Classified as probably not a substantial claim.

    Study design and conductown data

    Age-matched mice (6-7 months old) were used in three groups: wild-type control, 5xFAD AD, and 5xFAD-Sarm1KO rescue.

    “We tested this with the age-matched mice (6-7 months old) in three groups: wild-type as control, 5xFAD as AD, and 5xFAD-Sarm1KO as rescue group.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    The anti-Sarm1 antibody specificity was validated in brain slices.

    “We applied the anti-Sarm1 antibody (Fig. 1a-b), whose specificity was validated in brain slices”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Sarm1 is activated by local synaptic build-up of Aβ stress and subsequently drives synaptic degeneration as an executor.

    “Together, these data support the model in which Sarm1 is activated by local synaptic build-up of Aβ stress and subsequently drives synaptic degeneration as an executor (Fig. 1j).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Interpretationreasoning

    Synaptic Sarm1 is a sensor to monitor synaptic stress and a mediator to drive synapse degeneration for phagocytic clearance.

    “Thus, synaptic Sarm1 is a sensor to monitor synaptic stress and also a mediator to drive synapse degeneration for phagocytic clearance.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Interpretationown data

    Sarm1 mediates synapse degeneration in response to Aβ accumulation, which in turn initiates C1q tagging and subsequent complement cascades to recruit glia cells for phagocytosis clearance.

    “Overall, these data support a model that Sarm1 mediates synapse degeneration in response to Aβ accumulation, which in turn initiates C1q tagging and subsequent complement cascades to recruit glia cells for phagocytosis clearance (Fig. 6o).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • May combine two claims.

    Interpretationreasoning

    Accumulated stress from Aβ oligomers in synapses induces Sarm1 activation, which triggers local NAD+ depletion, energy collapse, and eventually synapse destruction, followed by C1q tagging and subsequent phagocytosis.

    “In this model, accumulated stress from Aβ oligomers in synapses induces sarm1 activation, which in turn triggers local NAD+ depletion, energy collapse, and eventually synapse destruction, followed by C1q tagging and subsequent phagocytosis (Fig. 6o).”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Both 5xFAD and Sarm1 KO strains and their controls have a congenic C57BL/6J background.

    “Both strains and their controls have a congenic C57BL/6J (B6J) background.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    Sarm1 KO protects synapses against Aβ-induced stress and fully reverses impaired synapse functions regarding basal transmission, short-term plasticity, and LTP in AD.

    “Together, these structural and functional data support that Sarm1 KO protects synapses against Aβ-induced stress and fully reverses the impaired synapse functions regarding basal transmission, short-term plasticity, and LTP, uncovering a crucial role of sarm1 in mediating Aβ-induced synaptic dysfunction and destruction in AD.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    TBS is designed to mimic the in vivo firing patterns of hippocampal neurons during exploratory behavior.

    “TBS is designed to mimic the in vivo firing patterns of hippocampal neurons during exploratory behavior 127 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [127] source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Electrophysiological recordings were made blind to mouse genotype.

    “The recordings were made blind to mouse genotype.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Melt curve confirmed amplification specificity.

    “Melt curve confirmed amplification specificity.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisereasoning

    Upon axon injury, NMNAT2 insufficiency and lower NAD+/NMN ratio activate Sarm1 to further cleave NAD+, resulting in energy collapse, Ca2+ overload, and axon destruction.

    “Upon axon injury, NMNAT2 insufficiency and lower NAD+/NMN ratio activate Sarm1 to further cleave NAD + , resulting in energy collapse, Ca 2+ overload, and axon destruction.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationnone given

    The long travel distance of presynaptic terminals from soma, high energy demand on NAD+, and enriched Sarm1 levels make axon nerve terminals more vulnerable to pathological insult.

    “a, the long travel distance of presynaptic terminals from soma, high energy demand on NAD+, and enriched Sarm1 levels make axon nerve terminals more vulnerable to pathological insult.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Sarm1's primary role is as a prodegenerative molecule in pathological conditions.

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    For the same set of experiments, images were acquired with the same confocal microscope with identical settings.

    “For the same set of experiments, images were acquired with the same confocal microscope with identical settings, including laser intensity, exposure time, and EM gain, as well as brain subregions.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Interpretationreasoning

    Upon phagocytosis, glia become more active for phagocytosis and neuroinflammation, and thus promote further Aβ production and synapse degeneration.

    “Upon phagocytosis, glia become more active for 685 phagocytosis and neuroinflammation, and thus promote further Aβ production and synapse 686 degeneration.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Sarm1's primary role is as a prodegenerative molecule in pathological conditions.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    All strains have a congenic B6J background to minimize the potential non-specific impact of genetic backgrounds.

    “All the strains have a congenic B6J background to minimize the potential non-specific impact of genetic backgrounds.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    Antibody specificity was validated by vendors or in the lab.

    “The specificity of antibodies was validated by vendors or in the lab.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Aβ plaques were detected with a thresholding function followed by a binary mask, and the same threshold was applied to each group.

    “Aβ plaques were detected with a thresholding function followed by a binary mask, and the same threshold was applied to each group.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    In AD synapses, accumulated Aβ stress activates Sarm1, which drives NAD+ depletion and energy collapse and thus induces synapse degeneration.

    “c, in the AD synapses, accumulated Aβ stress activates Sarm1, which drives NAD+ depletion and energy collapse and 683 thus induces synapse degeneration.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Sarm1's primary role is as a prodegenerative molecule in pathological conditions.
59 of 59

  • The paper relies on this without support.

    The molecular mechanisms mediating Aβ-induced synapse degeneration remain unclear.

    “While Aβ disrupts synaptic transmission 32,34,35 and induces synapse loss 17 , the molecular mechanisms mediating Aβ-induced synapse degeneration remain unclear.”

  • The paper relies on this without support.

    What happens in synapses before C1q tagging and what are neuron-intrinsic mechanisms of synaptic degeneration remain unclear.

    “However, what happens in synapses before C1q tagging and what are neuron-intrinsic mechanisms of synaptic degeneration remain unclear.”

  • The paper relies on this without support.

    The function of Sarm1 in synaptic degeneration of AD remains unclear.

    “However, the function in synaptic degeneration of AD remains unclear.”

  • The paper relies on this without support.

    The 5xFAD mouse model recapitulates human AD synapse degeneration and memory loss sufficiently to draw conclusions about AD.

    “The Sarm1-deficient AD mouse model was generated by crossing 5xFAD with sarm1 KO mice.”

  • The paper relies on this without support.

    Sarm1KO mice are indistinguishable from littermate controls, with normal lifespan and no overt phenotypes.

    “Sarm1KO mice were indistinguishable from littermate controls; they exhibited normal lifespan and no overt phenotypes, consistent with their primary role as a prodegenerative molecule in pathological conditions.”

  • The paper relies on this without support.

    Sarm1's primary role is as a prodegenerative molecule in pathological conditions.

    “consistent with their primary role as a prodegenerative molecule in pathological conditions.”
6 of 6

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Behavioral sample sizes fall outside the stated range

    Minor

    The methods state that all behavioral assays used 8-12 mice per group, but the fear-conditioning legend reports 6 WT mice and 14 rescue mice. The elevated-plus-maze legend also reports 14 rescue mice.

1 of 1

Cites

Flagged first, then by citations

Research works printed but not linked: 2

  1. 10.1126/science.aax1911 · DOI
  2. 10.1056/NEJ-Moa2212948 · DOI

Related

Authors
Fan, F., Joshi, P., Liu, X., Kotturu, S. K., Abiola, M., Liu, Q., & Lou, X.
Affiliations
Medical College of Wisconsin (US)
Corresponding author
Lou, Xuelin
ORCID
Venue
bioRxiv (Cold Spring Harbor Laboratory)
Publisher
openRxiv
Published
2025-12-23
DOI
10.64898/2025.12.21.695860
PMC
PMC12776179
OpenAlex
W7117136203
Semantic Scholar
7c7d4a4ef9277cb30e27af2d60572b3caa1aabbd
Cited by
0
References
127
Type
posted content
Language
en
Open access
Licence
www.biorxiv.org/about/FAQ#license (vor)
Locations
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion)
  • PubMed (publishedVersion)
  • PubMed Central (submittedVersion)
Funders
Advancing a Healthier Wisconsin Endowment (FP00028084)
Topics
  • Health Sciences › Medicine › Physiology › Alzheimer's disease research and treatments
  • Life Sciences › Neuroscience › Neurology › Neuroinflammation and Neurodegeneration Mechanisms
  • Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Neuroscience and Neuropharmacology Research
Keywords
SARM1, NAD+ hydrolase, synaptic degeneration, Alzheimer's disease, amyloid-beta, neuroinflammation, complement system, synaptic plasticity, memory loss, 5xFAD mice, phagocytosis

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-30.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-30.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review