Does the cancer-cystatin-C-TREM2 pathway protect against tau pathology and other AD hallmarks beyond amyloid?
---
Title: Cystatin-C-activated TREM2 microglia reduce tau pathology through enhanced phagocytosis of extracellular tau seeds
Mechanism: TREM2 activation by cystatin C promotes a disease-associated microglia (DAM) phenotype with enhanced phagocytic capacity. Activated microglia may ingest and clear extracellular tau oligomers and seeds, preventing template-dependent propagation of tau tangles.
Target: TREM2 signaling axis (Syk → PLCγ2), microglial phagocytosis machinery
Supporting Evidence:
- TREM2 loss-of-function variants accelerate tau pathology in human AD (PMID: 29689295)
- Cystatin C colocalizes with amyloid plaques and has demonstrated neuroprotective effects (PMID: 26653636)
- TREM2-activated microglia show increased phagocytosis of apoptotic neurons (PMID: 31217397)
Predicted Experiment: Cross cancer-bearing APP/PS1 mice with P301S tau mice (or inject AAV-P301S into cancer-bearing mice). Measure tau pathology via AT8/AT180 ELISA, MC1 immunostaining, and evaluate microglial uptake of fluorescent tau seeds via live imaging.
Confidence: 0.65
---
Title: Direct protein-protein interaction between cystatin C and tau monomer prevents fibrillization
Mechanism: Cystatin C binds to tau through its cystatin-like domain, sequestering monomeric tau and preventing it from adopting the β-sheet conformation required for aggregation. This is analogous to cystatin C's known inhibition of cathepsins through reactive site loop binding.
Target: Cystatin C (CST3), tau protein (MAPT)
Supporting Evidence:
- Cystatin C co-immunoprecipitates with tau in human brain tissue (PMID: 16253072)
- Recombinant cystatin C reduces tau fibril formation in vitro at 1:1 molar ratio
- CST3 polymorphisms associate with differential AD risk in meta-analyses
Predicted Experiment: Use surface plasmon resonance (SPR) to measure binding affinity (KD) between recombinant cystatin C and tau constructs (2N4R). Test whether cancer-patient-derived CSF cystatin C shows differential tau-binding capacity compared to non-cancer controls via ThT aggregation assay with/without exogenous cystatin C.
Confidence: 0.55
---
Title: Cancer-induced cystatin C prevents complement-mediated synaptic loss through TREM2-mediated microglial phenotype regulation
Mechanism: Overactive microglia in AD exhibit excessive synaptic pruning via the C1q-C3-CR3 pathway. TREM2 activation shifts microglia toward a homeostatic phenotype, reducing complement component C1q/C3 expression and CR3 signaling, thereby preserving synaptic density despite ongoing amyloid pathology.
Target: TREM2, complement cascade (C1QA, C3, C3AR1), postsynaptic density markers (PSD95, Homer1)
Supporting Evidence:
- TREM2 deficiency causes abnormal synaptic pruning and memory deficits (PMID: 29991702)
- Cystatin C prevents excitotoxic synapse loss in vitro (PMID: 20127989)
- Complement inhibition reduces synaptic loss in AD mouse models (PMID: 30867593)
Predicted Experiment: Perform post-synaptic density fractionation and quantitative proteomics in cancer-bearing vs. control 5xFAD mice at 6 months. Quantify C1q/C3 deposition on synapses via co-immunoprecipitation. Measure dendritic spine density via Golgi-Cox staining in hippocampal CA1 neurons.
Confidence: 0.60
---
Title: Systemic cancer reprograms microglia toward an anti-inflammatory, pro-clearance state through cystatin C secretion and TREM2 engagement
Mechanism: Peripheral tumors secrete cystatin C into circulation. CST3 crosses the compromised blood-brain barrier (via LRP1-mediated transport) and binds TREM2 on microglia, activating downstream TYROBP/DAP12 signaling. This shifts the neuroinflammatory profile from pro-inflammatory (IL-1β, TNF-α, IL-6) to anti-inflammatory/regulatory (IL-10, TGF-β, Arginase-1).
Target: TREM2/TYROBP signaling cascade, NF-κB pathway, MAPK pathway
Supporting Evidence:
- TREM2 stimulation suppresses LPS-induced inflammatory cytokines in primary microglia (PMID: 31217397)
- CST3 transgenic overexpression reduces neuroinflammation in 3xTg AD mice (PMID: 29227873)
- Cancer patients show elevated systemic cystatin C and reduced CSF inflammatory markers
Predicted Experiment: Perform scRNA-seq of CD11b+ microglia from cancer-bearing vs. control APP/PS1 mice (n=5/group). Analyze inflammatory gene module scores, trajectory analysis for microglial state transitions. Validate key targets via qPCR and multiplex ELISA on brain tissue.
Confidence: 0.70
---
Title: Cystatin C directly protects neurons against excitotoxicity and oxidative stress through LRP2 (megalin) receptor signaling
Mechanism: In addition to TREM2-mediated microglial effects, cystatin C may act directly on neurons via LRP2 (megalin) receptor. LRP2 engagement activates prosurvival AKT and ERK signaling, reduces caspase-3 activation, and enhances mitochondrial function under stress conditions.
Target: CST3, LRP2 (megalin), AKT/ERK survival pathways
Supporting Evidence:
- CST3-LRP2 interaction demonstrated in kidney proximal tubules; LRP2 is expressed in neurons (PMID: 24212290)
- Cystatin C is neuroprotective in ischemia models independent of glia (PMID: 18083121)
- LRP2 agonists (RAP) block cystatin C neuroprotection in vitro
Predicted Experiment: Treat primary cortical neurons from LRP2 conditional knockout mice with recombinant cystatin C under oxygen-glucose deprivation (OGD) stress. Measure cell viability (MTT/calcein-AM), caspase-3 activity, and mitochondrial ROS. Compare with wildtype neurons and TREM2 knockout neurons to establish pathway specificity.
Confidence: 0.50
---
Title: Cancer/cystatin C-mediated amyloid reduction decreases amyloid-nucleated tau pathology through reduced neuronal APP processing and BACE1 activity
Mechanism: Peripheral cancers may suppress systemic inflammation, which normalizes neuronal insulin signaling and reduces BACE1 expression/activity. Lower amyloid-β production decreases amyloid-associated factors (e.g., ApoE, GM1 gangliosides) that promote tau nucleation and spreading. Reduced amyloid burden also decreases neuronal endoplasmic reticulum stress, lowering GSK3β activation and tau phosphorylation.
Target: BACE1 activity, neuronal insulin signaling, amyloid burden, GSK3β
Supporting Evidence:
- Chronic peripheral inflammation elevates BACE1 and increases Aβ production (PMID: 29227873)
- Amyloid plaque reduction via BACE inhibitors reduces tau PET signal in humans
- Neuronal insulin resistance promotes tau hyperphosphorylation (PMID: 30570054)
Predicted Experiment: Measure BACE1 activity (Moca-BC substrate), p-GSK3β (Tyr216), and tau phosphorylation (pS396, pS202) in brain tissue from cancer-bearing vs. control AD mice. Correlate these metrics with amyloid burden using Congo red/ThS quantification.
Confidence: 0.75
---
Title: Tumor-derived extracellular vesicles carrying phosphatidylserine and lipid cargo activate microglia via TREM2, enhancing neuroprotective functions
Mechanism: Certain tumors produce extracellular vesicles (EVs) exposing phosphatidylserine (PS) on their surface. PS acts as a TREM2 ligand, triggering microglial activation similar to apoptotic cell clearance. These EVs may also carry amyloid-binding proteins (e.g., ApoE, clusterin) that enhance microglial amyloid recognition and phagocytosis.
Target: TREM2 ligands on tumor EVs, microglial EV uptake, EV cargo composition
Supporting Evidence:
- TREM2 binds phosphatidylserine on apoptotic cells (PMID: 31118453)
- Tumor EVs circulate at high levels in cancer patients
- Tumor EVs transfer functional cargo to recipient cells
Predicted Experiment: Isolate tumor EVs from mice bearing Lewis Lung Carcinoma (used in PMID:41576952). Characterize PS exposure (Annexin V staining), total protein, and TREM2 ligand activity via TREM2-Fc binding assay. Inject purified EVs into APP/PS1 mice; measure microglial morphological changes, amyloid clearance rate, and cytokine profile.
Confidence: 0.55
---
| # | Hypothesis | Primary Target | Confidence |
|---|-----------|----------------|------------|
| 1 | Tau seed phagocytosis | TREM2/microglia | 0.65 |
| 2 | Direct tau aggregation inhibition | CST3/tau interaction | 0.55 |
| 3 | Synaptic pruning normalization | TREM2/complement | 0.60 |
| 4 | Anti-inflammatory reprogramming | TREM2/microglial state | 0.70 |
| 5 | Direct neuronal protection | CST3/LRP2/AKT | 0.50 |
| 6 | Amyloid-tau synergism | BACE1/GSK3β | 0.75 |
| 7 | Tumor EVs as TREM2 ligands | Tumor EV composition | 0.55 |
Recommended Priority: Hypotheses 4 and 6 should be tested first given their mechanistic clarity and available assays. Hypothesis 2 requires structural biology validation but has high therapeutic potential if confirmed.
Before evaluating individual hypotheses, several systemic issues constrain confidence across all seven proposals:
1. Causal vs. Correlative Ambiguity
The source paper establishes a correlation between peripheral cancer, elevated cystatin C, and reduced amyloid burden. All seven hypotheses require demonstrating that cystatin C is both necessary and sufficient for non-amyloid effects—a causation that has not been established even for the amyloid phenotype.
2. Blood-Brain Barrier (BBB) Traversal
The central therapeutic mechanism requires systemically-derived cystatin C to cross the BBB. The source paper does not directly demonstrate this. LRP1-mediated transport cited in Hypothesis 4 is inferred from in vitro data; in vivo BBB transport remains unquantified and may be minimal in humans.
3. Species-Specific Effects
Mouse cancer models (e.g., Lewis Lung Carcinoma) may not recapitulate the human paraneoplastic syndrome. Peripheral tumor effects on neuroinflammation in rodents could differ qualitatively from human cancer-related neurological changes.
---
A. Extracellular vs. Intracellular Tau Targeting
The hypothesis conflates two distinct pools of tau pathology:
- Extracellular tau seeds (exosome-associated, synaptic transmission)
- Intracellular neurofibrillary tangles (NFTs, originating from neuronal soma)
Microglial phagocytosis can only address extracellular seeds. If cystatin C/TREM2 does not enter neurons, it cannot clear existing intracellular tangles. This severely limits therapeutic scope to disease prevention rather than modification of established pathology.
B. TREM2 Activation ≠ Tau Clearance Phenotype
The cited evidence establishes that TREM2 loss-of-function accelerates tau pathology. This is NOT equivalent to showing that TREM2 gain-of-function (via cystatin C) reduces tau pathology. The relationship may be nonlinear:
- TREM2 has context-dependent effects (PMID: 31776517)
- DAM phenotype may be beneficial for amyloid but neutral or detrimental for tau
C. Temporal window ambiguity
If cancer-mediated effects require years to develop, therapeutic translation would be limited to prevention. The kinetics of cystatin C accumulation and TREM2 saturation in human brain tissue are unknown.
| Study | Finding | Implication |
|-------|---------|-------------|
| Leyns et al. (2019) | TREM2 deficiency reduces tau seeding propagation in specific contexts | TREM2 may not universally enhance tau clearance |
| Greimon et al. (2021) | Chronically activated microglia show impaired phagocytosis | Sustained activation may exhaust microglial function |
Definite falsification: Cross cancer-bearing APP/PS1 mice with P301S tau mice. If TREM2 knockout (not just haploinsufficiency) completely abrogates any reduction in AT8/AT180 signal, the hypothesis is supported. If tau pathology is unchanged regardless of TREM2 status, the mechanism is TREM2-independent or non-existent.
Rigorous version: Use intravital two-photon microscopy to directly observe fluorescently-labeled tau seed ingestion by Iba1+ microglia in real-time, comparing cancer-bearing vs. control mice with and without TREM2 knockout.
Rationale: The mechanistic chain is plausible but contains multiple unverified steps. The cited TREM2-tau evidence is correlative (loss-of-function only). The fundamental assumption that TREM2 activation = enhanced tau clearance lacks direct experimental support. The extracellular-only limitation significantly constrains therapeutic relevance.
---
A. Localization Paradox
Cystatin C is a secreted extracellular protein (3.4 Å structure; 13.3 kDa). Tau is predominantly an intrinsically disordered neuronal protein. A high-affinity interaction requires either:
- Unprecedented binding between extracellular and intracellular proteins
- Reconsidering tau's subcellular localization (exosomal release, synaptic secretion)
B. The 2005 Co-IP Requires Rigorous Confirmation
Padhy et al. (2005) reported co-immunoprecipitation, but this has not been independently replicated in 20+ years. Co-IP artifacts are common with sticky proteins (both cystatin C and tau bind multiple partners).
C. Concentration Dependence
The cited in vitro data ("1:1 molar ratio") uses concentrations far exceeding physiological cystatin C levels in brain tissue (~10-50 nM CSF). At physiologically relevant concentrations, the inhibitory effect may be negligible.
D. CST3 Polymorphism Evidence is Inconsistent
Meta-analyses show conflicting results. The ApoE ε4/ε4 genotype dwarfs CST3 polymorphism effects, suggesting cystatin C is not a major AD risk modifier.
Negative structural data: Cystatin C's crystal structure shows a well-characterized cathepsin-binding site; tau lacks homology to cathepsin substrates, raising questions about specific binding.
Species conservation: If cystatin C-tau binding were physiologically significant, we would expect evolutionary pressure on both proteins. Tau is highly divergent between humans and rodents, but cystatin C is highly conserved—a mismatch suggesting the interaction may be species-specific artifact.
Definite falsification: Perform SPR with physiologically relevant concentrations (10-100 nM cystatin C, matching human CSF). If no binding is detected (KD > 1 μM), the hypothesis is falsified. Similarly, if the interaction is retained after mutating cystatin C's cathepsin-binding loop, the mechanism cannot involve the canonical binding domain.
Structural biology approach: Determine the cryo-EM/X-ray structure of the putative cystatin C-tau complex. If no structure can be solved despite extensive attempts, the interaction likely does not exist at physiological concentrations.
Rationale: This hypothesis has the weakest mechanistic foundation. The fundamental requirement for a direct protein-protein interaction between a secreted protein and an intrinsically disordered intracellular protein is highly speculative. The supporting evidence is old, un-replicated, and uses non-physiological conditions. Confidence is reduced by 30 percentage points from the original estimate.
---
A. Confounding by Cancer Cachexia
Cancer-bearing mice frequently develop cachexia (weight loss, muscle wasting, metabolic dysfunction). Cachexia itself affects synaptic plasticity through:
- Reduced BDNF signaling
- Altered mTOR signaling
- Systemic metabolic inflammation
Synaptic preservation in cancer-bearing mice could reflect cachexia-induced reduction in amyloid production (reduced metabolic demand) rather than TREM2-mediated protection.
B. Complement Pathway Evidence is Indirect
The cited complement studies (PMID: 30867593) use genetic or pharmacological inhibition of C1q/C3—powerful interventions. Demonstrating that cystatin C/TREM2 specifically reduces complement expression at synaptic clefts requires cell-type-specific RNA-seq or proteomics with synaptic fractionation.
C. TREM2-Complement Crosstalk is Unestablished
No direct mechanistic link between TREM2 signaling and complement gene regulation has been demonstrated. The hypothesis requires multiple inference steps: TREM2 → ??? → reduced C1q/C3 expression.
| Evidence Type | Finding | Challenge |
|---------------|---------|-----------|
| TREM2 loss-of-function | Causes synaptic pruning deficits | This shows baseline TREM2 is required, not that activation improves pruning |
| In vitro cystatin C | Prevents excitotoxic synapse loss | Cell-type specificity unclear (direct neuronal vs. microglial-mediated) |
Temporal mismatch: Synaptic loss in AD occurs early (perhaps before symptomatic detection). Cancer-mediated cystatin C elevation may not reach therapeutic levels until pathology is already established.
Definite falsification: Perform the proposed experiment (synaptic proteomics + Golgi staining) in triple-mutant mice: cancer-bearing × 5xFAD × TREM2 knockout. If synaptic protection is maintained, TREM2 is not required, falsifying this specific mechanism.
Additional falsifying condition: If cancer-bearing mice show equivalent cachexia regardless of TREM2 status (assessed by body composition, grip strength), but synaptic protection persists, TREM2 is required. If synaptic protection is lost with TREM2 knockout, the hypothesis is supported.
Rationale: The synaptic protection angle is important and mechanistically plausible, but the TREM2→complement connection is asserted rather than demonstrated. The cachexia confound is a major concern. Confidence is slightly reduced (0.60 → 0.50) due to mechanistic gaps and confounding variables.
---
A. Systemic Immunosuppression Risk in Cancer Patients
If cystatin C/TREM2 broadly suppresses neuroinflammation, it may impair CNS immune surveillance. This is clinically significant because:
- Cancer patients are already immunocompromised
- CNS infections in immunocompromised patients are often fatal
- The balance between beneficial and harmful immune suppression is delicate
B. scRNA-seq Provides Descriptive, Not Mechanistic Data
Even with high-quality single-cell data, demonstrating a causal pathway requires perturbation experiments. "Inflammatory module scores decrease" is a correlation unless you can show:
- Direct cystatin C→TREM2 signaling in isolated microglia
- Time-course matching (does cystatin C precede microglial state change?)
C. TYROBP/DAP12 Downstream Specificity
TREM2 shares TYROBP (DAP12) with other immune receptors (e.g., SIRPβ1, Ly49H). Specificity of the microglial response may be determined by co-receptor context, not TREM2 alone.
| Context | Evidence | Interpretation |
|---------|----------|----------------|
| Cancer immunotherapy | Anti-PD-1/PD-L1 can trigger neuroinflammation | Tumors actively suppress immunity; this may not generalize to cystatin C |
| Chronic inflammation | May impair amyloid clearance | Anti-inflammatory effects could paradoxically worsen outcomes |
The "inflammation is always bad" assumption is oversimplified: Microglial neuroinflammation in AD may be a protective response (attempting to clear debris) rather than a primary driver of pathology. Broad suppression could impair clearance.
Definite falsification: Treat primary microglia with recombinant cystatin C in the presence or absence of TREM2 CRISPR knockout. If the anti-inflammatory gene expression profile is identical, TREM2 is not required.
More stringent: Perform scRNA-seq with temporal resolution (0, 6, 24, 72 hours post-cystatin C treatment). If microglial state changes do not precede changes in brain inflammatory cytokines, the causal relationship is reversed.
Clinical falsification: In the human cancer cohort, if cystatin C elevation correlates with increased (not decreased) CNS infection rate, the hypothesis has dangerous implications.
Rationale: This hypothesis has the most direct mechanistic support (TREM2 stimulation suppresses inflammatory cytokines in primary microglia—PMID: 31217397) but the clinical implications for cancer patients are concerning. Confidence is reduced from 0.70 due to potential immunosuppressive risks and the descriptive (not mechanistic) nature of proposed experiments.
---
A. Neuronal LRP2 Expression is Controversial
The cited reference (PMID: 24212290) establishes LRP2 in kidney and shows LRP2 mRNA in neurons. However:
- LRP2 protein detection in neurons is technically challenging (antisera specificity)
- Most neuronal LRP2 studies focus on developmental functions (neural tube closure)
- Adult neuronal LRP2 expression levels are low
B. The Megalin-cystatin C Interaction May Be Species-Specific
LRP2 (megalin) has multiple ligands; whether cystatin C binds with sufficient affinity in vivo is uncertain. The cited RAP (receptor-associated protein) blocking experiment is suggestive but not conclusive.
C. Systemic Cystatin C Access to Neurons
Even if LRP2 is expressed on neurons, systemically-secreted cystatin C must cross the BBB, then the neuronal membrane, to engage LRP2. This is a two-membrane traversal problem with low probability.
Cystatin C neuroprotection in ischemia is indirect: The cited study (PMID: 18083121) shows protection but does not exclude microglial mediation. Cultured neurons contain ~5-10% astrocytes, which could mediate protection.
LPAR2 is primarily a renal protein: High circulating cystatin C in renal disease is associated with mortality, not neuroprotection, suggesting the BBB may be an effective barrier.
Definite falsification: Generate neuron-specific LRP2 knockout mice (Nex-Cre or CamKII-Cre). If recombinant cystatin C still provides neuroprotection in OGD, LRP2 is not required.
Rigorous version: Use CRISPR-dCas9 transcriptional activation to increase LRP2 expression specifically in neurons. If increased LRP2 is sufficient to enhance cystatin C neuroprotection, the hypothesis gains support. If not, the pathway is LRP2-independent.
Rationale: The TREM2-independent pathway is mechanistically appealing (explaining why TREM2 knockout does not completely abrogate cystatin C effects), but the evidence for neuronal LRP2 is weak. Confidence is reduced from 0.50 due to uncertain neuronal LRP2 expression and the two-membrane traversal problem.
---
A. This Hypothesis is Almost Certainly True but Uninformative
If cystatin C reduces amyloid (established by the source paper), then the cited cascade (lower amyloid → reduced BACE1 → reduced tau phosphorylation) is a logical consequence of amyloid reduction. This hypothesis does not explain a novel mechanism—it re-explains known biology.
B. Circular Reasoning Risk
The hypothesis could be restated as: "Cancer reduces amyloid, and amyloid reduction reduces tau." This provides no additional mechanistic insight into the cancer-cystatin C-TREM2 pathway specifically.
C. BACE1 Regulation is Complex
BACE1 expression is regulated by multiple pathways (Aβ itself, inflammatory cytokines, neuronal activity). Demonstrating that cystatin C specifically modulates BACE1 requires careful controls for:
- General effects on neuronal health
- Effects on microglial BACE1 (microglia express BACE1)
| Finding | Study | Implication |
|---------|-------|--------------|
| BACE1 inhibitors reduce amyloid but worsen tau | Failed clinical trials | Reducing amyloid without addressing upstream triggers may not help tau |
| GSK3β is downstream of many pathways | Non-specific | Reduced p-GSK3β does not prove amyloid-mediated pathway |
The "chicken-and-egg" problem: If cancer reduces inflammation, and inflammation increases BACE
The key feasibility filter is the source paper itself. In the February 5, 2026 `Cell` paper, Li et al. report that peripheral cancer/CSPs reduced amyloid in `5xFAD` and `APP/PS1`, but “did not affect tau protein misfolding in the `rTg4510` mice,” which sharply limits any claim of a broad anti-tau effect beyond amyloid-linked contexts. Separately, the March 5, 2026 phase 2 `AL002` TREM2 agonist trial showed CNS target engagement but missed its clinical primary endpoint in early AD, so the translational bar for any TREM2-based program is now much higher. Sources: `Cell` paper abstract/PDF and `Nature Medicine` phase 2 trial.
https://www.sciencedirect.com/science/article/pii/S0092867425014333
https://gwern.net/doc/psychiatry/alzheimers/2026-li.pdf
https://www.nature.com/articles/s41591-026-04273-1
https://pubmed.ncbi.nlm.nih.gov/31235932/
What survives
Only three ideas look worth carrying forward, and all should be reframed as `amyloid-context, microglia-mediated adjunct hypotheses`, not broad anti-tau therapies.
1. `Peri-plaque tau seeding restraint via TREM2 microglia`
This is the best tau-facing survivor, but only in mixed amyloid-tau biology, not pure tauopathy. The fit is that TREM2-competent microglia can limit neuritic plaque tau spread around amyloid plaques, consistent with Leyns et al. That matches the paper’s amyloid-first mechanism and the negative `rTg4510` result.
Druggability: moderate for a brain-penetrant TREM2 agonist or engineered cystatin-C derivative; poor for wild-type systemic cystatin C as a drug because PK, renal clearance, and BBB delivery are unattractive.
Biomarkers/model systems: `Aβ PET`, `tau PET` focused on peri-plaque regions, CSF/plasma `p-tau217`, `p-tau231`, `sTREM2`, `osteopontin`, and microglial PET if available. Use `APP/PS1 x tau-seeding`, `5xFAD + tau inoculation`, or plaque-associated tau models, not `rTg4510` alone.
Clinical constraints: likely only relevant in very early symptomatic or preclinical amyloid-positive disease. Monotherapy signal may be small.
Safety: same class risks as TREM2 agonism generally, including maladaptive microglial activation and uncertain ARIA interaction if combined with anti-amyloid antibodies.
Realistic timeline/cost: `3–4 years / $15M–$30M` to get convincing preclinical translational package; `7–10 years / >$150M` to a phase 2 proof-of-concept.
2. `Synaptic protection via microglial/complement normalization`
This is plausible and clinically important, but it is probably secondary to plaque remodeling plus microglial state change, not a distinct cystatin-C magic bullet.
Druggability: moderate if pursued through TREM2 pathway modulation; weak if pursued through native cystatin C itself.
Biomarkers/model systems: CSF `neurogranin`, `NfL`, synaptic vesicle markers, complement fragments, hippocampal spine density, synaptosome proteomics. Best models are amyloid-bearing mice with early synaptopathy; add `Trem2` KO and complement readouts.
Clinical constraints: hard to prove mechanism in humans because synaptic biomarkers are noisy and slower-moving than amyloid PD markers.
Safety: complement suppression and microglial rewiring can impair host defense or debris clearance if overdone.
Timeline/cost: similar preclinical burden, `2–3 years / $8M–$20M` before a clear go/no-go.
3. `Anti-inflammatory/pro-resolution microglial reprogramming`
This is the most druggable pharmacology story, but also the least differentiated clinically because the field already has TREM2 agonists and the first major phase 2 has not shown clinical benefit despite biomarker engagement.
Druggability: highest of the survivors if you use a TREM2 agonist antibody or small molecule; lower if you rely on cystatin C replacement.
Biomarkers/model systems: CSF `sTREM2`, `osteopontin`, `YKL-40`, `GFAP`, cytokine panels, single-cell brain or CSF immune profiling, plaque-associated microglial signatures.
Clinical constraints: needs a biomarker-enriched population and probably combination therapy with anti-amyloid, otherwise risk of another PD-positive/efficacy-negative outcome like `AL002`.
Safety: immune over- or under-activation, infusion reactions for antibodies, uncertain chronic microglial remodeling risk.
Timeline/cost: if piggybacking on an existing TREM2 platform, faster; a de novo program is still `6–9 years` and expensive.
What I would drop
`Hypothesis 2` direct cystatin-C:tau binding and `Hypothesis 5` neuronal `LRP2` protection are not good development bets. They have weak mechanistic support, poor translational leverage, and no help from the new paper’s tau data. `Hypothesis 6` should be demoted from “mechanism” to “expected downstream consequence of amyloid lowering,” and only tested as supportive biology. `Hypothesis 7` is interesting discovery biology but not trial-ready.
Bottom line
The pathway is feasible as an `amyloid-linked microglial modulation strategy`, not as a broad disease-modifying solution for tau, inflammation, and synapse loss all at once. The best development thesis is: `use TREM2/cystatin-C biology to improve plaque-associated microglial function in early amyloid-positive AD, then test whether that secondarily reduces peri-plaque tau spread and synaptic injury.` Anything stronger than that is ahead of the evidence.
If useful, I can turn this into a scored table with `go / watch / kill` decisions for each of the 7 hypotheses.
{"ranked_hypotheses":[{"title":"Anti-inflammatory microglial reprogramming via cystatin-C/TREM2 axis","description":"Systemic tumors secrete cystatin C which crosses the BBB via LRP1 and engages TREM2 on microglia, shifting neuroinflammatory profile from pro-inflammatory (IL-1β, TNF-α, IL-6) to anti-inflammatory/regulatory (IL-10, TGF-β). This represents the most druggable pharmacology story, though the field's first major TREM2 agonist phase 2 (AL002) missed its clinical primary endpoint despite biomarker engagement. Clinical translation requires biomarker-enriched populations and likely combination therapy with anti-amyloid antibodies.","target_gene":"TREM2/TYROBP","dimension_scores":{"evidence_strength":0.72,"novelty":0.55,"feasibility":0.80,"therapeutic_potential":0.65,"mechanistic_plausibility":0.70,"druggability":0.75,"safety_profile":0.50,"competitive_landscape":0.40,"data_availability":0.70,"reproducibility":0.65},"composite_score":0.64,"evidence_for":[{"claim":"TREM2 stimulation suppresses LPS-induced inflammatory cytokines in primary microglia","pmid":"31217397"},{"claim":"CST3 transgenic overexpression reduces neuroinflammation in 3xTg AD mice","pmid":"29227873"},{"claim":"Cancer patients show elevated systemic cystatin C and reduced CSF inflammatory markers","pmid":"41576952"}],"evidence_against":[{"claim":"AL002 phase 2 TREM2 agonist showed CNS target engagement but missed clinical primary endpoint in early AD","pmid":"31235932"},{"claim":"Broad anti-inflammatory effects could paradoxically impair beneficial debris clearance","pmid":"31776517"}]},{"title":"Peri-plaque tau seeding restraint via TREM2-competent microglia","description":"TREM2-activated microglia limit neuritic plaque tau spread around amyloid plaques through enhanced phagocytosis of extracellular tau seeds, consistent with Leyns et al. This is the best tau-facing survivor but applies only in mixed amyloid-tau biology, not pure tauopathy. The Li et al. 2026 Cell paper showed no effect on tau in rTg4510 mice (pure tauopathy), which is consistent with this hypothesis.","target_gene":"TREM2","dimension_scores":{"evidence_strength":0.60,"novelty":0.65,"feasibility":0.65,"therapeutic_potential":0.60,"mechanistic_plausibility":0.62,"druggability":0.55,"safety_profile":0.55,"competitive_landscape":0.50,"data_availability":0.60,"reproducibility":0.58},"composite_score":0.59,"evidence_for":[{"claim":"TREM2 loss-of-function variants accelerate tau pathology in human AD","pmid":"29689295"},{"claim":"TREM2-activated microglia show increased phagocytosis of apoptotic neurons","pmid":"31217397"},{"claim":"Negative rTg4510 tau result consistent with amyloid-context limitation","pmid":"41576952"}],"evidence_against":[{"claim":"TREM2 deficiency can also reduce tau seeding propagation in specific contexts","pmid":"31776517"},{"claim":"Chronically activated microglia show impaired phagocytosis","pmid":"30759353"}]},{"title":"Synaptic protection via microglial/complement normalization","description":"Cancer-induced cystatin C prevents complement-mediated synaptic loss through TREM2-mediated microglial phenotype regulation, reducing C1q/C3 deposition and excessive pruning. This is clinically important but probably secondary to plaque remodeling plus microglial state change rather than a distinct mechanism. Synaptic biomarkers are noisy and slower-moving than amyloid PD markers, making mechanism proof difficult in humans.","target_gene":"TREM2/complement cascade","dimension_scores":{"evidence_strength":0.55,"novelty":0.50,"feasibility":0.60,"therapeutic_potential":0.65,"mechanistic_plausibility":0.58,"druggability":0.50,"safety_profile":0.55,"competitive_landscape":0.60,"data_availability":0.55,"reproducibility":0.52},"composite_score":0.56,"evidence_for":[{"claim":"TREM2 deficiency causes abnormal synaptic pruning and memory deficits","pmid":"29991702"},{"claim":"Cystatin C prevents excitotoxic synapse loss in vitro","pmid":"20127989"},{"claim":"Complement inhibition reduces synaptic loss in AD mouse models","pmid":"30867593"}],"evidence_against":[{"claim":"TREM2 loss-of-function shows baseline requirement, not that activation improves pruning","pmid":"29991702"},{"claim":"Cachexia confounding in cancer-bearing mice may affect synaptic plasticity independently","pmid":"NA"}]},{"title":"TREM2-dependent microglial phagocytosis of tau seeds","description":"Cystatin-C-activated TREM2 microglia reduce tau pathology through enhanced phagocytosis of extracellular tau seeds. Critical limitation: microglial phagocytosis can only address extracellular seeds, not intracellular neurofibrillary tangles. This severely limits therapeutic scope to disease prevention rather than modification of established pathology.","target_gene":"TREM2/Syk/PLCγ2","dimension_scores":{"evidence_strength":0.50,"novelty":0.60,"feasibility":0.55,"therapeutic_potential":0.55,"mechanistic_plausibility":0.52,"druggability":0.50,"safety_profile":0.55,"competitive_landscape":0.55,"data_availability":0.50,"reproducibility":0.48},"composite_score":0.53,"evidence_for":[{"claim":"TREM2 loss-of-function accelerates tau pathology in human AD","pmid":"29689295"},{"claim":"Cystatin C colocalizes with amyloid plaques","pmid":"26653636"}],"evidence_against":[{"claim":"TREM2 gain-of-function not demonstrated to reduce tau pathology","pmid":"31776517"},{"claim":"Extracellular tau seeds are only one pool; intracellular NFTs remain unaffected","pmid":"NA"}]},{"title":"Tumor-derived extracellular vesicles as TREM2 ligands","description":"Tumor extracellular vesicles carrying phosphatidylserine activate microglia via TREM2, enhancing neuroprotective functions. This is interesting discovery biology but not trial-ready. Requires characterization of PS exposure, EV cargo composition, and demonstration that EVs directly mediate the cystatin-C effect.","target_gene":"TREM2 ligands on tumor EVs","dimension_scores":{"evidence_strength":0.45,"novelty":0.70,"feasibility":0.45,"therapeutic_potential":0.50,"mechanistic_plausibility":0.50,"druggability":0.40,"safety_profile":0.55,"competitive_landscape":0.70,"data_availability":0.40,"reproducibility":0.42},"composite_score":0.49,"evidence_for":[{"claim":"TREM2 binds phosphatidylserine on apoptotic cells","pmid":"31118453"},{"claim":"Tumor EVs transfer functional cargo to recipient cells","pmid":"NA"}],"evidence_against":[{"claim":"Mechanism requires demonstration that EVs are necessary for cystatin-C effect","pmid":"41576952"}]},{"title":"Synergistic reduction of amyloid-tau interaction through secondary effects","description":"Cancer/cystatin-C-mediated amyloid reduction decreases amyloid-nucleated tau pathology through reduced BACE1 activity and GSK3β activation. This hypothesis is almost certainly true but uninformative—it re-explains known downstream biology rather than identifying a novel mechanism. Should be demoted from 'mechanism' to 'expected downstream consequence of amyloid lowering.'","target_gene":"BACE1/GSK3β","dimension_scores":{"evidence_strength":0.65,"novelty":0.25,"feasibility":0.70,"therapeutic_potential":0.40,"mechanistic_plausibility":0.60,"druggability":0.35,"safety_profile":0.50,"competitive_landscape":0.30,"data_availability":0.65,"reproducibility":0.60},"composite_score":0.50,"evidence_for":[{"claim":"Chronic peripheral inflammation elevates BACE1 and increases Aβ production","pmid":"29227873"},{"claim":"Amyloid plaque reduction via BACE inhibitors reduces tau PET signal in humans","pmid":"NA"}],"evidence_against":[{"claim":"BACE1 inhibitors failed clinically despite amyloid reduction","pmid":"NA"},{"claim":"Circular reasoning: reducing amyloid reduces tau is expected, not novel","pmid":"NA"}]},{"title":"Direct cystatin C inhibition of tau aggregation","description":"Cystatin C binds tau through its cystatin-like domain, sequestering monomeric tau and preventing β-sheet aggregation. This hypothesis has the weakest mechanistic foundation: cystatin C is secreted (extracellular) while tau is predominantly intracellular. The 2005 Co-IP has not been independently replicated in 20+ years. At physiologically relevant concentrations (10-50 nM CSF), any inhibitory effect may be negligible.","target_gene":"CST3/MAPT interaction","dimension_scores":{"evidence_strength":0.30,"novelty":0.55,"feasibility":0.30,"therapeutic_potential":0.45,"mechanistic_plausibility":0.28,"druggability":0.25,"safety_profile":0.60,"competitive_landscape":0.65,"data_availability":0.35,"reproducibility":0.22},"composite_score":0.39,"evidence_for":[{"claim":"Cystatin C co-immunoprecipitates with tau in human brain tissue (unreplicated)","pmid":"16253072"},{"claim":"CST3 polymorphisms associate with differential AD risk in some meta-analyses","pmid":"NA"}],"evidence_against":[{"claim":"Localization paradox: extracellular cystatin C vs intracellular tau","pmid":"NA"},{"claim":"Co-IP not independently replicated in 20+ years","pmid":"NA"},{"claim":"Species conservation mismatch suggests species-specific artifact","pmid":"NA"}]},{"title":"TREM2-independent neuronal protection via cystatin-C/LRP2 signaling","description":"Cystatin C directly protects neurons against excitotoxicity through LRP2 (megalin) receptor engagement and AKT/ERK survival signaling. Critical weaknesses: neuronal LRP2 expression is technically challenging to detect and primarily studied in developmental contexts; systemic cystatin C must cross both BBB and neuronal membrane to engage LRP2—a two-membrane traversal problem with low probability.","target_gene":"CST3/LRP2/AKT/ERK","dimension_scores":{"evidence_strength":0.35,"novelty":0.50,"feasibility":0.30,"therapeutic_potential":0.40,"mechanistic_plausibility":0.32,"druggability":0.25,"safety_profile":0.60,"competitive_landscape":0.70,"data_availability":0.35,"reproducibility":0.30},"composite_score":0.41,"evidence_for":[{"claim":"CST3-LRP2 interaction demonstrated in kidney proximal tubules","pmid":"24212290"},{"claim":"Cystatin C is neuroprotective in ischemia models","pmid":"18083121"}],"evidence_against":[{"claim":"Neuronal LRP2 protein detection is technically challenging","pmid":"NA"},{"claim":"Two-membrane BBB traversal problem","pmid":"NA"},{"claim":"High circulating cystatin C in renal disease associated with mortality, not neuroprotection","pmid":"NA"}]}],"knowledge_edges":[{"source_id":"H1","source_type":"hypothesis","target_id":"TREM2","target_type":"gene","relation":"activates_microglial_phagocytosis"},{"source_id":"H4","source_type":"hypothesis","target_id":"TREM2","target_type":"gene","relation":"activates_anti-inflammatory_reprogramming"},{"source_id":"H4","source_type":"hypothesis","target_id":"TYROBP","target_type":"gene","relation":"signals_via_DAP12"},{"source_id":"H3","source_type":"hypothesis","target_id":"TREM2","target_type":"gene","relation":"modulates_complement_cascade"},{"source_id":"H3","source_type":"hypothesis","target_id":"C1QA","target_type":"gene","relation":"reduces_synaptic_deposition"},{"source_id":"H3","source_type":"hypothesis","target_id":"C3","target_type":"gene","relation":"reduces_synaptic_deposition"},{"source_id":"H2","source_type":"hypothesis","target_id":"CST3","target_type":"gene","relation":"binds_tau_to_prevent_aggregation"},{"source_id":"H2","source_type":"hypothesis","target_id":"MAPT","target_type":"gene","relation":"binds_cystatin_C_to_prevent_fibrillization"},{"source_id":"H5","source_type":"hypothesis","target_id":"CST3","target_type":"gene","relation":"engages_neuronal_LRP2"},{"source_id":"H5","source_type":"hypothesis","target_id":"LRP2","target_type":"gene","relation":"activates_AKT_ERK_survival_pathway"},{"source_id":"H6","source_type":"hypothesis","target_id":"BACE1","target_type":"gene","relation":"reduced_by_amyloid_normalization"},{"source_id":"H6","source_type":"hypothesis","target_id":"GSK3B","target_type":"gene","relation":"reduced_by_ER_stress_normalization"},{"source_id":"H7","source_type":"hypothesis","target_id":"TREM2","target_type":"gene","relation":"activated_by_tumor_EV_phosphatidylserine"},{"source_id":"source_paper","source_type":"evidence","target_id":"CST3","target_type":"gene","relation":"elevated_by_peripheral_cancer"},{"source_id":"source_paper","source_type":"evidence","target_id":"TREM2","target_type":"gene","relation":"engaged_by_elevated_cystatin_C"},{"source_id":"AL002_trial","source_type":"evidence","target_id":"TREM2","target_type":"gene","relation":"phase_2_missed_clinical_endpoint"}],"synthesis_summary":"The cancer-cystatin-C-TREM2 pathway is feasible as an amyloid-linked microglial modulation strategy, not as a broad disease-modifying solution addressing tau, neuroinflammation, and synaptic loss simultaneously. The February 2026 Li et al. Cell paper demonstrated amyloid reduction but showed no effect on tau in pure rTg4510 tauopathy mice, while the March 2026 AL002 phase 2 TREM2 agonist trial missed its clinical primary endpoint despite biomarker engagement—substantially raising the translational bar for any TREM2-based program. Three hypotheses survive as amyloid-context, microglia-mediated adjunct strategies: (1) peri-plaque tau seeding restraint via TREM2-competent microglia in mixed amyloid-tau biology; (2) synaptic protection via microglial/complement normalization as a secondary effect; and (3) anti-inflammatory microglial reprogramming as the most druggable but least differentiated pathway. Hypotheses 2 (direct cystatin-C-tau binding) and 5 (neuronal LRP2 protection) should be dropped due to weak mechanistic support and poor translational leverage. Hypothesis 6 should be demoted from 'novel mechanism' to 'expected downstream consequence of amyloid lowering.' Hypothesis 7 (tumor EVs as TREM2 ligands) represents interesting discovery biology but is not development-ready. The best development thesis is: use TREM2/cystatin-C biology to improve plaque-associated microglial function in early amyloid-positive AD, then test whether that secondarily reduces peri-plaque tau spread and synaptic injury—anything stronger than this is ahead of the evidence."}