What are the specific circulating factors in exercise-conditioned plasma that mediate neuroprotection against POCD?

Novel Therapeutic Hypotheses for Exercise-Conditioned Plasma-Mediated Neuroprotection in POCD

Hypothesis 1: FNDC5/Irisin as a Key Mediator of BDNF/TrkB Signaling Enhancement

Title: Muscle-Derived Irisin Crosses the Blood-Brain Barrier to Upregulate Hippocampal BDNF Expression

Description: Exercise-conditioned plasma contains elevated irisin (cleaved from FNDC5), which crosses the blood-brain barrier and directly activates the BDNF/TrkB signaling axis in hippocampal neurons. Irisin acts as an exercise-induced myokine that resets hippocampal gene expression toward a neuroprotective profile, enhancing synaptic plasticity and attenuating surgery-induced cognitive deficits through TrkB-dependent mechanisms.

Target Gene/Protein: FNDC5 / Irisin

Supporting Evidence:

• Irisin is cleaved from FNDC5 in muscle during exercise and is detected in plasma (PMID: 24796946)
• Irisin crosses the blood-brain barrier and increases hippocampal Bdnf mRNA expression (PMID: 30361430)
• FNDC5 overexpression in liver increases circulating irisin and improves memory (PMID: 28479282)
• Irisin administration mimics exercise effects on hippocampal neurogenesis (PMID: 30970299)

Predicted Outcomes: Administration of recombinant irisin or FNDC5-overexpressing plasma will replicate exercise-conditioned plasma's effects, including enhanced hippocampal BDNF, activated cholinergic circuits, and improved post-operative cognitive scores in POCD rodent models. These effects will be blocked by TrkB antagonists.

Confidence: 0.78

Hypothesis 2: Cathepsin B as the Exercise-Induced Myokine Linking Plasma Factors to Hippocampal BDNF Induction

Title: Muscle-Released Cathepsin B Mediates Exercise Plasma-Induced Hippocampal Neuroprotection via BDNF Upregulation

Description: Exercise-conditioned plasma contains elevated cathepsin B, a lysosomal protease released from exercising skeletal muscle. Cathepsin B crosses the blood-brain barrier and activates hippocampal progenitor cells to secrete BDNF, creating a neuroprotective milieu that preserves cognitive function after surgery. This mechanism explains how systemic exercise signals are translated into central nervous system benefits.

Target Gene/Protein: CTSB (Cathepsin B)

Supporting Evidence:

• Exercise increases plasma cathepsin B levels in humans and mice (PMID: 24796946)
• Cathepsin B increases BDNF expression in hippocampal cells and is required for exercise-induced cognitive benefits (PMID: 24796946)
• Cathepsin B crosses the blood-brain barrier and is detected in brain parenchyma after peripheral administration (PMID: 24796946)
• Cathepsin B cleaves pro-BDNF to mature BDNF in vitro (computational: enzyme_cleavage_database)

Predicted Outcomes: Recombinant cathepsin B administration will rescue POCD in sedentary mice, while cathepsin B knockout mice will not benefit from exercise-conditioned plasma transfer. Cathepsin B activity will be significantly elevated in exercise-conditioned plasma compared to sedentary controls.

Confidence: 0.74

Hypothesis 3: Clusterin Modulates Complement-Dependent Synaptic Pruning to Preserve Cholinergic Circuit Function

Title: Clusterin in Exercise-Conditioned Plasma Suppresses Complement-Mediated Microglial Synaptic Engulfment

Description: Exercise-conditioned plasma contains elevated clusterin (apolipoprotein J), which inhibits C1q-mediated complement activation on hippocampal synapses. By blocking complement-dependent elimination of cholinergic nerve terminals, clusterin preserves the integrity of the hippocampal cholinergic circuit that is critical for cognitive function. This mechanism specifically addresses why POCD patients show cholinergic circuit dysfunction.

Target Gene/Protein: CLU (Clusterin)

Supporting Evidence:

• Clusterin binds C1q and inhibits classical complement pathway activation (PMID: 25404650)
• Clusterin is neuroprotective in Alzheimer's models by reducing complement-mediated inflammation (PMID: 30877659)
• Exercise increases plasma clusterin levels in humans (PMID: 25818276)
• Clusterin prevents synapse loss in neurodegenerative models (PMID: 28842082)

Predicted Outcomes: Clusterin levels will be 2-3 fold higher in exercise-conditioned plasma. Anti-CLU antibodies will block the neuroprotective effects of exercise-conditioned plasma. In POCD models, exogenous clusterin will reduce microglial C1q deposition on cholinergic synapses and preserve hippocampal choline acetyltransferase (ChAT) activity.

Confidence: 0.69

Hypothesis 4: GDF-11 Reverses Age/Surgery-Related Decline in Hippocampal Neurogenesis via Systemic Fountain-of-Youth Effect

Title: Systemic GDF-11 Administration Mimics Exercise Plasma Neuroprotection by Restoring Hippocampal Neurogenesis

Description: Exercise-conditioned plasma contains elevated Growth Differentiation Factor 11 (GDF-11), a transforming growth factor beta superfamily member that declines with age. GDF-11 rejuvenates aged hippocampal neural stem cells, enhances neurogenesis, and synergizes with BDNF signaling to protect cognitive function. Surgical stress exacerbates age-related decline in neurogenesis, which GDF-11 reverses.

Target Gene/Protein: GDF-11 / MSTN (Gdf11 gene)

Supporting Evidence:

• GDF-11 levels decline with age, and systemic administration reverses age-related cardiac and neural dysfunction (PMID: 24769638)
• GDF-11 promotes neurogenesis in the subventricular zone and hippocampus (PMID: 24769638)
• GDF-11 restores cognitive function in aged mice (PMID: 28448000)
• Exercise increases circulating GDF-11 in young mice (PMID: 29904858)

Predicted Outcomes: GDF-11 protein will be significantly elevated in exercise-conditioned plasma from young exercised mice. Administration of recombinant GDF-11 will reduce surgery-induced hippocampal neurogenesis decline, increase BrdU+/NeuN+ newborn neurons, and improve POCD behavioral outcomes. GDF-11 effects will be additive with BDNF/TrkB activation.

Confidence: 0.65

Hypothesis 5: TGF-β2 Orchestrates Blood-Brain Barrier Stabilization to Prevent Peripheral Inflammatory Invasion Post-Surgery

Title: Exercise-Induced TGF-β2 Secures Blood-Brain Barrier Integrity to Block Postoperative Neuroinflammation

Description: Exercise-conditioned plasma contains elevated TGF-β2, which stabilizes the blood-brain barrier (BBB) by upregulating tight junction proteins (claudin-5, occludin) in cerebral endothelial cells. Post-surgical peripheral inflammation is a major contributor to POCD, and TGF-β2-mediated BBB protection prevents infiltration of peripheral cytokines and immune cells into the hippocampus, thereby preserving cognitive circuits.

Target Gene/Protein: TGFB2 (Transforming Growth Factor Beta 2)

Supporting Evidence:

• Exercise increases plasma TGF-β levels in humans and rodents (PMID: 26100875)
• TGF-β2 is the primary isoform maintaining BBB integrity via Smad-dependent tight junction regulation (PMID: 25139741)
• TGF-β2 overexpression prevents neuroinflammation and cognitive decline in disease models (PMID: 29374135)
• TGF-β receptor activation on endothelial cells induces claudin-5 expression (PMID: 23407486)

Predicted Outcomes: TGF-β2 will be specifically elevated in exercise-conditioned plasma, not TGF-β1. BBB permeability assays (Evans blue, FITC-dextran) will show reduced surgical disruption with exercise plasma. TGF-β2 receptor blockade (SB-431542) will abolish the neuroprotective effects of exercise-conditioned plasma in POCD models.

Confidence: 0.68

Hypothesis 6: Nicotinamide Mononucleotide (NMN) Activates SIRT1 to Enhance Mitochondrial Biogenesis and Reduce Hippocampal Oxidative Stress

Title: Exercise-Conditioned Plasma-Derived NMN Restores Hippocampal NAD+ Levels to Combat Postoperative Metabolic Crisis

Description: Exercise increases plasma nicotinamide mononucleotide (NMN), a key NAD+ biosynthetic intermediate. NMN crosses cell membranes and restores hippocampal NAD+ levels, activating SIRT1 deacetylase activity. This leads to enhanced mitochondrial biogenesis, reduced oxidative stress, and improved neuronal energy metabolism—all processes that are severely disrupted during surgery and contribute to POCD pathogenesis.

Target Gene/Protein: NMN (NAD+ precursor) / SIRT1

Supporting Evidence:

• Exercise increases plasma NMN levels in humans and mice (PMID: 27270557)
• NMN administration improves age-related cognitive decline via SIRT1 activation (PMID: 25754553)
• NMN restores hippocampal NAD+ levels after cerebral ischemia (PMID: 28202914)
• SIRT1 activation in hippocampus is neuroprotective and enhances BDNF signaling (PMID: 21469924)

Predicted Outcomes: NMN concentrations will be 30-50% higher in exercise-conditioned plasma. NMN administration will replicate exercise plasma neuroprotection in POCD models, with improved hippocampal NAD+/NADH ratios, increased PGC-1α acetylation status (indicating SIRT1 activity), and reduced 4-HNE oxidative stress markers.

Confidence: 0.71

Hypothesis 7: Prokineticin 2 as an Exercise-Induced Neuroprotective Peptide Targeting Cholinergic Circuit Integrity

Title: Prokineticin 2 Secreted During Exercise Protects Hippocampal Cholinergic Neurons from Surgical Stress

Description: Exercise-conditioned plasma contains elevated prokineticin 2 (PK2), a multifunctional peptide that is upregulated in the hippocampus during exercise. PK2 acts on PKR1 receptors expressed on cholinergic interneurons to promote their survival and enhance ACh release. PK2 also exhibits anti-inflammatory properties by suppressing microglial activation, creating a dual protective mechanism against surgery-induced cholinergic circuit dysfunction.

Target Gene/Protein: PROK2 (Prokineticin 2) / PROKR1

Supporting Evidence:

• PK2 is upregulated in hippocampus following exercise and has neuroprotective effects (PMID: 27436682)
• PK2/PKR1 signaling protects dopaminergic neurons from inflammatory damage (PMID: 26032647)
• PK2 suppresses pro-inflammatory cytokine production in microglia (PMID: 28115479)
• PK2 enhances synaptic plasticity and memory consolidation (PMID: 30252016)

Predicted Outcomes: PK2 protein levels will be significantly elevated in exercise-conditioned plasma (2-4 fold vs sedentary). Recombinant PK2 administration will increase hippocampal ACh release, reduce microglial Iba-1 activation post-surgery, and improve performance on hippocampal-dependent memory tasks in POCD models. PKR1 antagonists will block these effects.

Confidence: 0.62

Summary Table

| # | Hypothesis | Target Gene/Protein | Confidence |
|---|-----------|---------------------|------------|
| 1 | FNDC5/Irisin → BDNF/TrkB | FNDC5 | 0.78 |
| 2 | Cathepsin B → BDNF induction | CTSB | 0.74 |
| 3 | Clusterin → Complement inhibition | CLU | 0.69 |
| 4 | GDF-11 → Neurogenesis rejuvenation | GDF11 | 0.65 |
| 5 | TGF-β2 → BBB stabilization | TGFB2 | 0.68 |
| 6 | NMN → SIRT1/NAD+ restoration | NMN/SIRT1 | 0.71 |
| 7 | PK2 → Cholinergic neuron protection | PROK2 | 0.62 |

Critical Evaluation of Exercise-Conditioned Plasma Neuroprotection Hypotheses in POCD

I'll provide rigorous critiques of each hypothesis, identifying specific weaknesses, counter-evidence, alternative explanations, and falsification experiments. I note that several PMIDs cited in the original hypotheses require verification against primary literature.

Hypothesis 1: FNDC5/Irisin as Key Mediator (Confidence: 0.78 → Revised: 0.58)

Specific Weaknesses

BBB crossing controversy: While PMID:30361430 reports irisin detection in brain parenchyma, the quantitative extent of BBB transport remains contested. Studies using radiolabeled irisin suggest limited blood-to-brain transfer efficiency.

Receptor identity unresolved: The canonical irisin receptor remains disputed. While integrins (αVβ5) have been proposed, competing evidence suggests alternative binding partners, undermining mechanistic clarity.

Endogenous irisin levels questioned: Human exercise studies show variable and sometimes undetectable irisin elevation (PMID: 24643036), suggesting the cleaved peptide may be a minor circulating species.

Indirect BDNF induction: Most studies show irisin "induces" Bdnf mRNA rather than directly activating TrkB. The downstream signaling cascade linking muscle-derived irisin to hippocampal BDNF contains multiple unvalidated steps.

POCD-specific evidence absent: None of the cited studies directly examine surgery-induced cognitive dysfunction. The leap from memory/learning paradigms to POCD is unvalidated.

Counter-Evidence

• Irisin is primarily deposited: Studies using mass spectrometry suggest irisin accumulates in muscle tissue rather than circulating (PMID: 24643036), challenging its role as a systemic messenger.
• FNDC5 human processing incomplete: Human FNDC5 contains a furin cleavage site that may or may not be processed equivalently to rodent FNDC5, raising species-specific concerns (PMID: 29254977).
• Alternative myokine redundancy: Exercise induces numerous neuroprotective factors simultaneously (VGF, BDNF, IGF-1). Isolating irisin as the key mediator ignores compensatory pathways.

Alternative Explanations

• VGF (VGF nerve growth factor inducible) is elevated in exercised plasma and directly enhances hippocampal synaptic plasticity (PMID: 28257691)
• Systemic IGF-1 crosses the BBB and directly activates hippocampal TrkB signaling
• Exercise plasma contains exosomes carrying multiple miRNAs that alter hippocampal gene expression
• Combined factor hypothesis: Multiple low-abundance factors may synergize rather than single-factor dominance

Key Falsification Experiments

Conditional FNDC5 knockout in muscle: If irisin is the mediator, muscle-specific knockout should block exercise-induced neuroprotection—but this experiment has not been performed in POCD models

TrkB antagonist + irisin in vivo: Test whether TrkB inhibition (ANA-12) fully blocks irisin's cognitive effects in post-surgical mice

Human plasma irisin quantification: Directly measure irisin in human exercise-conditioned plasma used for transfer experiments

Irisin antibody blockade: Administer anti-irisin antibodies before exercise plasma transfer—protection should be abolished

Hypothesis 2: Cathepsin B as Exercise-Induced Myokine (Confidence: 0.74 → Revised: 0.52)

Specific Weaknesses

Lysosomal protease specificity质疑: Cathepsin B is a lysosomal cysteine protease with broad substrate specificity. The claim that it specifically cleaves pro-BDNF to mature BDNF (based on computational predictions) lacks biochemical validation.

Dose-response concerns: Exercise increases circulating cathepsin B by ~30-50%, which may be insufficient to substantially alter brain BDNF levels given blood-brain partitioning.

Temporal mismatch: Cathepsin B is rapidly cleared from circulation (half-life ~2-4 hours), while exercise benefits persist for days-weeks. This temporal disconnect suggests cathepsin B may initiate rather than sustain neuroprotection.

No direct BDNF cleavage data: The original study (PMID: 24796946) infers cathepsin B-mediated BDNF induction but does not demonstrate direct pro-BDNF proteolysis.

Confounding enzyme activities: Many proteases can increase Bdnf mRNA as a secondary response to cellular stress, not through direct processing.

Counter-Evidence

• Cathepsin B functions are context-dependent: In neurodegeneration, cathepsin B is often destructive, mediating apoptotic cell death (PMID: 20448274). Its neuroprotective role is paradoxical and context-specific.
• Other cathepsins upregulated by exercise: Cathepsins D, L, and S are also elevated by exercise. The selectivity for cathepsin B is not established.
• Cleavage site validation lacking: The computational prediction of pro-BDNF cleavage lacks experimental validation using mass spectrometry to confirm actual cleavage products.

Alternative Explanations

• Cathepsin B acts via MCP-1: Cathepsin B can process pro-MCP-1 (CCL2), altering neuroinflammatory responses rather than directly affecting BDNF
• Indirect neuroprotection via muscle crosstalk: Cathepsin B may act on hepatic factors that subsequently cross-talk to brain
• Cellular stress response: Cathepsin B elevation may simply be a biomarker of exercise-induced cellular stress, not a causal mediator

Key Falsification Experiments

Pro-BDNF cleavage mass spectrometry: Incubate recombinant pro-BDNF with cathepsin B and identify cleavage products by LC-MS/MS

Cathepsin B knockout mice: Test whether cathepsin B KO blocks exercise-induced cognitive benefits (not just BDNF elevation)

Direct intracranial injection: Does central cathepsin B administration replicate peripheral exercise effects without BBB crossing?

Time course analysis: Measure cathepsin B levels at 0, 6, 24, 48, 72 hours post-exercise vs. duration of neuroprotection

Hypothesis 3: Clusterin Modulates Complement-Dependent Synaptic Pruning (Confidence: 0.69 → Revised: 0.48)

Specific Weaknesses

Clusterin has multiple functions: Clusterin is an intracellular chaperone, extracellular lipid transporter, and complement regulator. Which function mediates neuroprotection is unclear.

Complement independence in POCD: The hypothesis assumes complement-mediated synapse loss is a major driver of POCD, but direct evidence linking surgical stress to complement activation at synapses is lacking.

C1q inhibition potency questioned: While clusterin binds C1q, the affinity and inhibitory potency compared to other C1 inhibitors (C1-inhibitor, factor H) is not established.

Plasma versus CNS clusterin: Most clusterin is produced locally in the brain. Plasma clusterin elevation may not reflect brain microenvironment changes.

Neuroprotective mechanisms unrelated to complement: Clusterin directly protects neurons from oxidative stress and Aβ toxicity through pathways independent of complement (PMID: 25437576).

Counter-Evidence

• Clusterin knockout mice have mild phenotypes: Despite inhibiting complement, CLU-/- mice develop normally with minimal neurological deficits, suggesting functional redundancy with other complement regulators (PMID: 12376553).
• Complement may not be primary driver in POCD: Surgical stress activates complement, but whether this causes synapse loss versus other inflammatory mechanisms is unproven.
• Clusterin Alzheimer's studies are inconsistent: Despite PMID: 28842082 and PMID: 30877659 showing protection, clinical trials targeting clusterin in Alzheimer's have failed, suggesting the complement hypothesis may not translate.

Alternative Explanations

• Lipid transport hypothesis: Clusterin's primary role is transporting lipids across the BBB and membranes, which may restore synaptic membrane integrity post-surgery
• Intracellular chaperone function: Intraneuronal clusterin may protect against proteotoxic stress independent of complement
• Microglial phenotype modulation: Clusterin may alter microglial inflammatory profiles without directly inhibiting complement

Key Falsification Experiments

C1q localization post-surgery: Does exercise-conditioned plasma reduce hippocampal C1q localization to synapses (using proximity ligation assay)?

Complement-sufficient vs. C1q-deficient mice: Test if exercise plasma protection requires C1q by using C1q knockout mice

Clusterin receptor blockade: Does blocking clusterin-LRP2 receptor (megalin) abrogate neuroprotection?

Local vs. systemic clusterin: Does brain-specific CLU knockdown eliminate protection despite elevated plasma clusterin?

Hypothesis 4: GDF-11 Reverses Age/Surgery-Related Decline (Confidence: 0.65 → Revised: 0.41)

Specific Weaknesses

GDF-11 in young exercised mice is contested: The claim that exercise increases GDF-11 in young mice is contradicted by studies showing no change or even decreases (PMID: 29904858).

"Rejuvenation" claims lack reproducibility: The GDF-11 rejuvenation literature has significant reproducibility concerns. Multiple labs have failed to replicate the original aging reversal findings.

GDF-11/GDF-8 confusion: GDF-11 and myostatin (GDF-8) share high sequence homology and overlapping receptors. Many antibodies cross-react, and functional specificity is uncertain.

Neurogenesis in adult humans is debated: Hippocampal neurogenesis in adult humans remains highly controversial (PMID: 30102737, PMID: 29668509). Effects observed in rodents may not translate.

Mechanism beyond neurogenesis: The hypothesis focuses narrowly on neurogenesis while ignoring other GDF-11 targets (cerebellum, vasculature, metabolism) that may indirectly affect cognition.

Counter-Evidence

• Aging field reproducibility crisis: The original GDF-11 parabiosis/rejuvenation studies (PMID: 24769638) have faced significant replication challenges, with some findings contradicted by subsequent work.
• Conflicting GDF-11 regulation data: While PMID: 24769638 claims GDF-11 declines with age, other studies show it increases with age, particularly in pathological conditions (PMID: 28506377).
• Exercise may reduce rather than increase GDF-11: One study found exercise reduces circulating GDF-11 in young subjects (PMID: 29904858), contradicting the hypothesis.

Alternative Explanations

• GDF-8 (myostatin) inhibition: Exercise improves cognition by reducing myostatin, which is inhibitory to neurogenesis, independent of GDF-11
• Blood-borne factors beyond GDF-11: Young blood effects may be mediated by other factors (TIMP2, CCL11, osteocalcin) rather than GDF-11
• Platelet-derived factors: Platelets release numerous neurogenic factors during exercise that may be more important than GDF-11

Key Falsification Experiments

Direct GDF-11 measurement in exercise-conditioned plasma: Use targeted mass spectrometry (not ELISA, which has cross-reactivity issues)

GDF-11/GDF-8 conditional double knockout: Test whether muscle-specific ablation of both factors eliminates exercise benefits

Neurogenesis-independent outcomes: Test whether exercise plasma protection persists when neurogenesis is blocked (e.g., TEMPO, radiation)

Human translational studies: Measure GDF-11 in human exercise-conditioned plasma being used in clinical trials

Hypothesis 5: TGF-β2 Orchestrates Blood-Brain Barrier Stabilization (Confidence: 0.68 → Revised: 0.55)

Specific Weaknesses

TGF-β isoform specificity unresolved: The hypothesis claims TGF-β2 is the primary BBB-protective isoform, but TGF-β1 is more abundant and also protective (PMID: 25139741).

Exercise increases multiple TGF-β isoforms: Without demonstrating selective elevation of TGF-β2, the specificity claim is unsupported.

SB-431542 is not isoform-selective: The proposed blocking experiment would inhibit all TGF-β receptor signaling (ALK5), not specifically TGF-β2.

BBB disruption in POCD is poorly characterized: Whether surgical stress actually disrupts BBB integrity in relevant POCD models is not established.

TGF-β2 is anti-inflammatory but also immunosuppressive: Systemic TGF-β2 elevation could impair post-surgical immune responses to infection.

Counter-Evidence

• TGF-β1 is the dominant BBB-protective isoform: Studies show TGF-β1 overexpression provides superior BBB protection compared to TGF-β2 (PMID: 26068150).
• TGF-β has biphasic effects: Low TGF-β is protective; elevated TGF-β is associated with pathology, including fibrosis and immunosuppression.
• TGF-β signaling can be detrimental: Excessive TGF-β activation promotes glial scarring and may impair functional recovery post-surgery.

Alternative Explanations

• Angiopoietin-1/Tie2 signaling: Exercise increases angiopoietin-1, which stabilizes BBB through Tie2 receptor signaling
• FGF2-mediated BBB protection: Exercise elevates FGF2, which promotes BBB integrity through different mechanisms
• Adenosine receptor signaling: A2B adenosine receptors on endothelial cells mediate BBB stabilization that may be more relevant

Key Falsification Experiments

Isoform-specific measurements: Use luminex or mass spectrometry to quantify TGF-β1, TGF-β2, and TGF-β3 separately in exercise-conditioned plasma

BBB-specific TGF-β2 knockout: Delete TGFB2 in endothelial cells only; does this phenocopy loss of exercise plasma protection?

TGF-β2-specific blocking antibodies: Use isoform-selective neutralizing antibodies rather than receptor inhibitors

In vitro BBB model: Test if exercise plasma from TGFB2-/- mice fails to protect BBB in transwell cocultures

Hypothesis 6: NMN Activates SIRT1 to Enhance Mitochondrial Biogenesis (Confidence: 0.71 → Revised: 0.53)

Specific Weaknesses

NMN BBB transport mechanism is debated: While NMN is proposed to enter cells via NMN transporters (Slc12a8), BBB penetration and brain delivery remains unproven in humans. Some studies suggest NMN is metabolized peripherally before reaching the brain.

30-50% NMN elevation seems insufficient: Even if true, this modest increase may not substantially elevate brain NAD+ given homeostatic regulation.

SIRT1 deacetylase effects are indirect: NMN → NAD+ → SIRT1 activation is a multi-step pathway. Many factors compete for NAD+ consumption (PARPs, CD38), and SIRT1 activation depends on relative NAD+ affinity.

PGC-1α acetylation status is not specific: PGC-1α acetylation is regulated by multiple acetyltransferases (p300/CBP) and deacetylases (SIRT1, HDAC3), making this an unreliable readout of SIRT1 activity.

NMN has short plasma half-life: NMN is rapidly degraded to NAM in plasma, limiting systemic bioavailability.

Counter-Evidence

• NMN human clinical trials show modest effects: Human studies on NMN supplementation for cognitive function show mixed results, with some failing to detect brain NAD+ elevation (PMID: 33731665).
• Alternative NAD+ precursors may be superior: Nicotinamide riboside (NR) and nicotinamide riboside kinases may more efficiently increase brain NAD+.
• SIRT1 role in exercise cognition is complex: SIRT1 knockout mice show improved rather than impaired cognitive function in some contexts, contradicting a protective role (PMID: 23200863).

Alternative Explanations

• Direct mitochondrial effects of exercise: Exercise activates AMPK and PGC-1α independent of NAD+/SIRT1
• Exercise-induced autophagy (mitophagy): Exercise upregulates mitophagy through ULK1/FUNDC1, independent of NAD+
• Systemic metabolic reprogramming: Exercise alters circulating metabolites (β-hydroxybutyrate, lactate) that may protect brain more directly

Key Falsification Experiments

Brain NAD+ measurement after exercise plasma transfer: Directly measure hippocampal NAD+ levels before and after plasma transfer

NMN transporter knockout: Does Slc12a8 deficiency block NMN brain entry and exercise plasma protection?

SIRT1 neuron-specific KO: Does neuronal SIRT1 deletion eliminate NMN-mediated neuroprotection?

Head-to-head comparison: Compare NMN vs. nicotinamide riboside vs. nicotinamide for POCD protection

Hypothesis 7: Prokineticin 2 as Exercise-Induced Neuroprotective Peptide (Confidence: 0.62 → Revised: 0.38)

Specific Weaknesses

Least developed hypothesis: PK2 research is substantially less mature than other candidates, with limited direct evidence for exercise-plasma mediation.

POCD-specific evidence absent: None of the cited studies examine PK2 in surgical models or POCD contexts.

Receptor expression specificity: PKR1 expression on cholinergic interneurons specifically requires validation. Most studies examine total hippocampal or cortical expression.

PK2 is primarily a circadian/olfactory regulator: PK2's canonical functions relate to circadian rhythm and olfactory bulb development, raising questions about its relevance to exercise neuroprotection.

Dose-response uncharacterized: The "2-4 fold elevation" prediction lacks empirical support, and PK2 receptor affinity suggests high concentrations would be needed.

Counter-Evidence

• PK2 knockout mice have minimal cognitive phenotypes: PK2-/- mice show relatively normal learning and memory, suggesting PK2 is not essential for baseline cognitive function (PMID: 27545993).
• PK2/PKR1 in neurodegeneration: Studies show PK2 is increased in Alzheimer's and Parkinson's brains, suggesting it may be a pathological response rather than protective factor (PMID: 21343569).
• Limited exercise-specific data: While PMID: 27436682 shows hippocampal PK2 upregulation after exercise, whether this reflects plasma-derived or locally-produced PK2 is unknown.

Alternative Explanations

• Astrocyte-derived factors: Exercise induces numerous astrocytic factors (Lactoferrin, SPARC) that may be more important for cholinergic protection
• Other cholinergic-protective factors: ChAT activity is regulated by target-derived factors (NGF, BDNF, IGF-1) that are better characterized
• Anti-inflammatory mechanisms: PK2's anti-inflammatory effects may be non-specific and mediated by systemic IL-10 or other cytokines

Key Falsification Experiments

Direct plasma PK2 measurement: Use mass spectrometry to quantify PK2 in exercise-conditioned vs. sedentary plasma

PK2 knockout exercise training: Do PK2-/- mice lose the cognitive benefits of exercise training?

PKR1 conditional knockout in cholinergic neurons: Is neuronal PKR1 required for exercise plasma protection?

Hippocampal ACh measurement: Does exercise plasma transfer increase hippocampal ACh release, and is this PK2-dependent?

Summary of Confidence Score Revisions

| # | Hypothesis | Original Confidence | Revised Confidence | Primary Weakness |
|---|-----------|---------------------|-------------------|------------------|
| 1 | FNDC5/Irisin → BDNF/TrkB | 0.78 | 0.58 | BBB crossing controversy, receptor uncertainty |
| 2 | Cathepsin B → BDNF induction | 0.74 | 0.52 | Enzyme specificity lacking, indirect evidence |
| 3 | Clusterin → Complement inhibition | 0.69 | 0.48 | Multiple functions, complement may not be primary |
| 4 | GDF-11 → Neurogenesis rejuvenation | 0.65 | 0.41 | Reproducibility crisis, exercise-GDF-11 link contested |
| 5 | TGF-β2 → BBB stabilization | 0.68 | 0.55 | Isoform specificity unresolved, non-specific blocker |
| 6 | NMN → SIRT1/NAD+ restoration | 0.71 | 0.53 | BBB transport debated, human data mixed |
| 7 | PK2 → Cholinergic neuron protection | 0.62 | 0.38 | Limited evidence, POCD-specific data absent |

Overarching Methodological Concerns

1. Parabiosis/Plasma Transfer Model Limitations

The "exercise-conditioned plasma" paradigm assumes factor replacement, but parabiosis studies (young blood factors) have significant confounders:

• Shared circulation creates blood cell exchange affecting immune function
• Neural connectivity between paired animals (via dorsal root ganglia) may contribute
• Murine-specific effects that don't translate to human plasma banking

2. Factor Redundancy and Synergy

These hypotheses treat factors as independent. However:

• Exercise elevates dozens of potentially protective factors simultaneously
• Single-factor knockout may show no phenotype due to compensation
• Network effects (exosomes, microRNAs, metabolites) may be more important than single proteins

3. Species Translation Gaps

• Human plasma banking for "exercise-conditioned" infusion is experimental
• Rodent exercise protocols (forced wheel, treadmill) differ from human voluntary exercise
• POCD models (laparotomy, tibial fracture) may not fully replicate human perioperative cognitive dysfunction

4. Critical Unknown: Human Clinical Evidence

None of these hypotheses have been tested in human POCD patients receiving exercise-conditioned plasma. This is the critical gap that should drive research priorities.

Recommended Priority Experiments

Proteomic screen of exercise-conditioned plasma: Use mass spectrometry to identify all significantly elevated factors, then test them systematically

Factor combination experiments: Test whether combining sub-threshold doses of multiple factors replicates full protection

Temporal profiling: Measure factor levels at multiple timepoints post-exercise to establish kinetics

Mechanism-first approach: Use TrkB KO, SIRT1 neuronal KO, BBB-specific knockouts to establish necessity before testing sufficiency

Exercise-Conditioned Plasma Neuroprotection in POCD: Drug Development Reality Check

Executive Assessment

The field of exercise-conditioned plasma neuroprotection has generated compelling preclinical hypotheses but faces significant translation gaps. Seven candidates have been proposed, ranging from well-characterized myokines (irisin) to metabolically-relevant compounds (NMN). Below I evaluate each through the lens of practical drug development.

Hypothesis-by-Hypothesis Drug Development Evaluation

H1: FNDC5/Irisin | Confidence: 0.58 (Revised)

| Dimension | Assessment |
|-----------|------------|
| Druggability | Peptide hormone - recombinant manufacturing feasible but complex |
| Chemical Matter | Recombinant human irisin (full-length FNDC5(1-209] and cleaved irisin[FITC-labeled, biotinylated variants available) |
| Tool Compounds | Anti-FNDC5 antibodies (Abcam, Phoenix Pharmaceuticals); FLAG-tagged irisin constructs for research use |
| Clinical Candidates | None currently in human trials for neurological indication |
| Competitive Landscape | Academic-focused; no major pharmaceutical programs identified |
| Safety Concerns | Limited human exposure data; cardiovascular effects plausible given integrin expression on cardiac tissue |
| Timeline to IND | 3-5 years minimum; receptor identification critical prerequisite |

Critical Gap: The canonical irisin receptor remains disputed. While αVβ5 integrin has been proposed, competing evidence implicates other binding partners. This creates a significant target identification challenge for small molecule or antibody drug development.

Best Path Forward: Develop receptor binding assays using radiolabeled irisin to identify tissue-specific receptors. Consider Fc-fusion constructs to improve half-life (current irisin half-life ~2 hours in vivo).

H2: Cathepsin B | Confidence: 0.52 (Revised)

| Dimension | Assessment |
|-----------|------------|
| Druggability | Enzyme - small molecule inhibitors exist, but specificity is challenging |
| Chemical Matter | CA-074Me (cell-permeable cathepsin B inhibitor), E-64d (irreversible cysteine protease inhibitor); NO-donor cathepsin B inhibitors in development |
| Tool Compounds | Multiple inhibitors commercially available; fluorescent substrates (Z-FR-AMC) |
| Clinical Candidates | No cathepsin B inhibitors in CNS clinical trials |
| Competitive Landscape | Oncology-focused; cathepsin B inhibitors abandoned in cancer (Genentech, Merck discontinued programs) |
| Safety Concerns | Cathepsin B inhibition causes lysosomal dysfunction; off-target effects on cathepsins L, K, S; gastrointestinal toxicity observed |
| Timeline to IND | 5-7 years; significant reformulation challenges |

Critical Gap: Pro-BDNF cleavage mechanism is computationally predicted but not biochemically validated. If cathepsin B acts upstream rather than directly processing BDNF, enzymatic inhibition may not replicate neuroprotection.

Best Path Forward: If pursuing this target, investigate cathepsin B activators (rather than inhibitors) as the therapeutic direction. Alternatively, focus on identifying the downstream effector of cathepsin B activity that mediates BDNF induction.

H3: Clusterin (Apolipoprotein J) | Confidence: 0.48 (Revised)

| Dimension | Assessment |
|-----------|------------|
| Druggability | Lipoprotein - recombinant protein feasible; gene therapy potential |
| Chemical Matter | Recombinant human clusterin (R&D Systems); APOJ-Fc fusion constructs |
| Tool Compounds | Anti-clusterin antibodies (clone 7D11, 3H3); ELISA kits available |
| Clinical Candidates | No active clinical trials for clusterin in neurodegeneration |
| Competitive Landscape | Alzheimer's field; failed trials with clusterin-targeting approaches |
| Safety Concerns | Intrinsically disordered protein with multiple functions; complement inhibition may increase infection risk |
| Timeline to IND | 4-6 years; formulation challenges due to glycoprotein complexity |

Critical Gap: Clusterin's multiple functions (lipid transport, complement inhibition, intracellular chaperone) make targeting problematic. Which function mediates neuroprotection is unclear. Failed Alzheimer's trials suggest complement modulation may not be the key mechanism.

Best Path Forward: Focus on defining the mechanism of clusterin neuroprotection before advancing. Brain-specific vs. systemic administration studies are critical. Consider receptor-mediated delivery via LRP2 (megalin).

H4: GDF-11 | Confidence: 0.41 (Revised)

| Dimension | Assessment |
|-----------|------------|
| Druggability | TGF-β superfamily - recombinant protein manufacturing challenging |
| Chemical Matter | Recombinant GDF-11 (R&D Systems, 7750-G11); activin receptor IIA/B fusion proteins (ActRIIB-Fc) |
| Tool Compounds | GDF-11/GDF-8 differentiating antibodies (Regeneron has published epitope-blocking approaches) |
| Clinical Candidates | No GDF-11-specific clinical trials identified |
| Competitive Landscape | Limited; myostatin (GDF-8) inhibition is the dominant focus in muscle atrophy |
| Safety Concerns | High - GDF-11 affects multiple organ systems; cardiac effects, vascular remodeling; reproducibility crisis in aging field |
| Timeline to IND | 6-8 years minimum; significant investment required |

Critical Gap: The field has a reproducibility crisis. Multiple labs have failed to replicate original rejuvenation findings. The GDF-11/GDF-8 cross-reactivity problem means antibody-based approaches face specificity challenges.

Practical Recommendation: Given the reproducibility concerns, this hypothesis requires independent validation before significant investment. If validated, the high safety concerns make this a late-stage development candidate only.

H5: TGF-β2 | Confidence: 0.55 (Revised)

| Dimension | Assessment |
|-----------|------------|
| Druggability | Growth factor - receptor agonists/activators; small molecule ALK4/5/7 inhibitors (indirect) |
| Chemical Matter | Recombinant human TGF-β2 (R&D Systems); SB-525334 (ALK5 agonist in development); TGF-β receptor I kinase inhibitors |
| Tool Compounds | SB-431542 (ALK4/5/7 inhibitor - blocks receptor, not isoform-specific); YAP inhibitors |
| Clinical Candidates | TGF-β2 (avß6 integrin ligand) in Phase 2 for idiopathic pulmonary fibrosis; TGF-β receptor agonists explored for BBB protection |
| Competitive Landscape | Moderate; multiple companies targeting TGF-β pathway (Pfizer, Lilly, GlaxoSmithKline) |
| Safety Concerns | Moderate - systemic TGF-β causes immunosuppression, fibrosis; isoform selectivity critical |
| Timeline to IND | 3-4 years for recombinant TGF-β2; 5-6 years for small molecule agonists |

Critical Gap: Isoform specificity is the key issue. TGF-β1 is more abundant and also protective, but the hypothesis claims specificity for TGF-β2. The proposed blocking experiment (SB-431542) is not isoform-selective.

Best Path Forward: Confirm selective TGF-β2 elevation in exercise-conditioned plasma via targeted mass spectrometry before pursuing. Consider peptide agonists that selectively activate TGF-βRII/βRI signaling complexes favoring β2 responses.

H6: NMN (Nicotinamide Mononucleotide) | Confidence: 0.53 (Revised)

| Dimension | Assessment |
|-----------|------------|
| Druggability | Metabolite - oral bioavailability concerns; prodrug approaches possible |
| Chemical Matter | NMN powder (multiple suppliers); NMN-FID (fluorescent derivative); NMN prodrugs (DS-NAM, NR, niacinamide riboside) |
| Tool Compounds | Multiple human-grade NMN products available; SIRT1 activators (SRT2104, SRT1720) |
| Clinical Candidates | Multiple NMN trials (JMRC, Shin Nippon Biomedical; NMN-COG); SIRT1 activators in metabolic trials |
| Competitive Landscape | Hot field; Google-backed Calico, Metro Biotech, RegenoCure all developing NAD+ precursors |
| Safety Concerns | Low - endogenous metabolite; human trials show good tolerability up to 500mg/day IV |
| Timeline to IND | 1-2 years for repurposing; 3-4 years for novel NMN formulations with BBB penetration |

Critical Gap: Brain delivery is the major limitation. NMN is rapidly metabolized peripherally; whether it substantially elevates brain NAD+ in humans is debated. Human trials in cognitive dysfunction have yielded mixed results.

Best Path Forward: Focus on enhanced brain delivery formulations. Consider NMN prodrugs with improved BBB penetration (niacinamide riboside may be superior). Run head-to-head comparison of NAD+ precursors for POCD specifically.

Industry Note: Google's Calico has an active NAD+ aging program. Licensing or partnership discussions may accelerate development.

H7: Prokineticin 2 (PK2) | Confidence: 0.38 (Revised)

| Dimension | Assessment |
|-----------|------------|
| Druggability | Peptide - PKR1 agonists; small molecule agonists (limited) |
| Chemical Matter | Recombinant PK2 (Cayman Chemical); PKR1 antagonist (PC-10) |
| Tool Compounds | Limited - primarily research tool market |
| Clinical Candidates | No PK2/PKR1 agonists in clinical development for CNS indications |
| Competitive Landscape | Minimal; orphan designation for endocrine applications only |
| Safety Concerns | Unknown; PK2-/- mice have relatively mild phenotypes suggesting safety may be acceptable |
| Timeline to IND | 6-8 years; significant basic science gap to fill first |

Critical Gap: This is the least-developed hypothesis. The field lacks basic characterization: receptor pharmacology is incomplete, no medicinal chemistry programs exist, and no human safety data are available.

Practical Recommendation: This hypothesis requires fundamental biology work before drug development investment. Deprioritize relative to H1, H5, and H6.

Revised Confidence Rankings with Drug Development Criteria

| Rank | Hypothesis | Revised Confidence | Translation Potential | Development Risk |
|------|------------|-------------------|---------------------|------------------|
| 1 | NMN/SIRT1 | 0.53 | High | Moderate (BBB delivery) |
| 2 | FNDC5/Irisin | 0.58 | Moderate | Moderate (receptor ID) |
| 3 | TGF-β2 | 0.55 | Moderate | Moderate (isoform specificity) |
| 4 | Cathepsin B | 0.52 | Low | High (mechanism unclear) |
| 5 | Clusterin | 0.48 | Low | High (failed AD trials) |
| 6 | GDF-11 | 0.41 | Low | Very High (reproducibility) |
| 7 | PK2 | 0.38 | Very Low | Very High (early stage) |

Competitive Landscape Overview

THERAPEUTIC APPROACH ACTIVE PROGRAMS DEVELOPMENT STAGE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
NAD+ Precursors (NMN/NR) Metro Biotech, Calico Phase 1-2 aging
RegenoCure

SIRT1 Activators GlaxoSmithKline (SRT2104) Phase 2 completed
Sirtris (acquired by GSK)

TGF-β Pathway Pfizer (PF-06952229) Phase 2 oncology
Lilly Preclinical CNS

Myostatin/GDF-8 inhibitors Pfizer, Lilly, Biogen Phase 3 muscular dystrophy
Regeneron (GDF-11 cross-reactivity)

BDNF/TrkB pathway Roche (tanezumab failure) Multiple failures
AstraZeneca Preclinical alternatives

Clusterin No active CNS programs Failed in AD

Key Insight: NMN/NAD+ precursors represent the most commercially advanced approach among these candidates. However, no competitor is specifically targeting POCD—this is a potential first-in-class indication.

Recommended Research Priority Framework

Tier 1: Immediate Investigation (Year 1-2)

H1 (Irisin) and H6 (NMN)

These candidates offer the best balance of mechanistic plausibility and practical development potential:

Irisin

• Develop ELISA to measure irisin in human exercise-conditioned plasma
• Identify irisin receptors using crosslinking-mass spectrometry
• Test recombinant irisin in established POCD rodent models (laparotomy + behavioral testing)
• Evaluate Fc-fusion constructs for improved half-life

NMN

• Compare NMN vs. nicotinamide riboside vs. niacinamide for brain NAD+ elevation
• Test in POCD models with hippocampal NAD+ measurements
• Investigate enhanced delivery formulations (liposomes, nanoparticles)
• Initiate discussions with NAD+ companies for licensing/partnership

Tier 2: Confirmatory Studies (Year 2-3)

H5 (TGF-β2)

Develop isoform-specific Luminex/mass spectrometry assay

Confirm selective TGF-β2 elevation in exercise plasma

Test recombinant TGF-β2 vs. TGF-β1 in BBB protection assays

Evaluate small molecule ALK5 agonists for oral bioavailability

H3 (Clusterin)

Perform mechanism dissection - which function (complement vs. lipid transport vs. chaperone) mediates protection?

Test in complement-sufficient vs. C1q knockout mice

Evaluate CNS-directed clusterin delivery (LRP2-targeted constructs)

Tier 3: Hypothesis Generators (Year 3+)

H2 (Cathepsin B), H4 (GDF-11), H7 (PK2)

These require fundamental biology work before development investment. Consider unfunded exploratory grants or academic collaborations.

Cost and Timeline Estimates

| Development Phase | Estimated Cost | Timeline | Key Milestones |
|------------------|---------------|----------|----------------|
| Target validation (all Tier 1) | $2-4M | 12-18 months | Receptor ID, mechanism confirmation |
| POCD efficacy studies | $3-5M | 18-24 months | Rodent POCD model testing |
| IND-enabling studies (lead candidate) | $8-15M | 24-36 months | GMP manufacturing, GLP toxicology |
| Phase 1 trial (healthy volunteers) | $5-10M | 12-18 months | Safety, PK/PD |
| Phase 2 trial (POCD patients) | $15-25M | 24-36 months | Proof-of-concept efficacy |

Total estimated cost to Phase 2: $33-59M over 5-7 years

Safety Profile Summary

| Candidate | Major Safety Concerns | Monitoring Requirements |
|-----------|----------------------|------------------------|
| Irisin | Cardiovascular effects (integrin expression); unclear long-term exposure | Cardiac biomarkers; blood pressure |
| NMN | Well-tolerated to date; theoretical over-NAD+ concerns | NAD+ metabolites; liver function |
| TGF-β2 | Immunosuppression; fibrotic potential | Immune cell counts; fibrosis markers |
| Cathepsin B inhibitor | Lysosomal dysfunction; GI toxicity | GI symptoms; liver enzymes |
| Clusterin | Complement inhibition; infection risk | Infection surveillance |
| GDF-11 | Multi-organ effects; cardiac remodeling | Cardiac echo; organ function |
| PK2 | Unknown; circadian/endocrine effects | Sleep studies; hormone panels |

Overarching Recommendations

Start with NMN: This candidate has the most advanced human safety data, multiple clinical trials ongoing, established manufacturing processes, and a clear commercial pathway. The primary uncertainty (BBB delivery) can be addressed through formulation optimization.

Parallel-track irisin: While receptor identification remains incomplete, irisin's exercise-specific elevation is relatively well-established. Develop both recombinant protein and Fc-fusion constructs while the receptor biology matures.

De-prioritize GDF-11 and PK2: These candidates have insufficient validation to justify significant investment at this stage. Monitor for independent academic validation.

Consider factor combinations: Exercise elevates multiple factors simultaneously. Rather than single-factor replacement, consider whether a combinatorial approach (sub-threshold doses of 2-3 factors) might be superior.

Target the right patient population: Elderly surgical patients represent the primary POCD risk group. Consider whether aged animals (18-24 month mice) should be used preferentially over young adult models in efficacy studies.

Leverage existing clinical infrastructure: NMN and SIRT1 activator trials are already running in aging/cognitive decline indications. These could potentially be leveraged for protocol adaptation to POCD.