{"artifact":{"id":"experiment_proposal-59785671-1a01-4297-a6fa-8d0664feb04c","artifact_type":"experiment_proposal","entity_ids":"[\"h-var-e95d2d1d86\"]","title":"Experiment Proposal: h-var-e95d2d1d86","quality_score":0.75,"created_by":"forge_experiment_proposal_generator","provenance_chain":"[]","content_hash":"607dce6f90cfb4b211bdffe38d4226c918b88510f758d85eaf9497d19496e7b2","metadata":{"citations":["31076275","35151204","36450248","37384704","38642614","39964974","27929004","31578527","36211804","33127896","28714589","30936556","34982715"],"generated_at":"2026-04-28T19:58:59.205388+00:00","intervention":"Stereotaxic injection of AAV9-CaMKIIα-DIO-ChR2-eYFP (Addgene #27056) into bilateral CA1 stratum pyramidale (AP -2.0, ML ±1.5, DV -1.4 mm from bregma) in PV-Cre mice (Jax #008069); control groups receive AAV9-CaMKIIα-DIO-eYFP (no ChR2) or equivalent volume PBS. Four weeks post-injection for viral expression (validated by eYFP fluorescence and ChR2 photocurrent verification in acute slices, holding current step of 200 pA at -70 mV under 470 nm, 5 ms pulse). Wireless micro-LED arrays (custom fabricated, 2 mm × 0.5 mm, 16 emitters, λ = 470 nm, 3 mW/mm² output) implanted epidurally over dorsal hippocampus. Closed-loop stimulation: real-time LFP acquisition at 1 kHz, theta phase detection via Hilbert transform (6-10 Hz bandpass), gamma amplitude calculation (25-100 Hz), 10 ms blue light pulses delivered at 40 Hz during detected theta troughs (180-270° phase window). Stimulation suspended if gamma power exceeds 3 standard deviations above baseline (seizure safeguard) or if local temperature rises >1°C. Protocol: 2 hours daily stimulation during active wakefulness (dark cycle) for 6 weeks. Non-specific stimulation group receives identical LED pulses in continuous 40 Hz mode without theta phase gating, controlling for direct photic effects on neurons.","key_controls":["AAV9-CaMKIIα-DIO-eYFP injected PV-Cre mice ( specificity control for viral transduction without channelrhodopsin)","APP/PS1 mice with PBS injection ( vehicle control for surgical procedure)","Non-PV-Cre (wild-type) mice injected with AAV9-CaMKIIα-DIO-ChR2-eYFP ( off-target expression control)","Continuous 40 Hz stimulation APP/PS1 mice ( non-specific photic/electrical stimulation control)","Age-matched wild-type C57BL/6J mice with closed-loop ChR2 stimulation ( normative reference)","Age-matched wild-type C57BL/6J mice with no surgery/no stimulation ( naive control)","AAV9-CaMKIIα-DIO-ChR2-eYFP injected APP/PS1 mice without LED implantation (no-stimulation ChR2 control for viral effects independent of light)"],"model_system":"APP/PS1dE9 mice (Jax #005864) on C57BL/6J background; both sexes; experiment onset at 3 months of age (pre-plaque stage based on PMID 31076275); Cre-negative littermates serve as controls for AAV-PV-Cre specificity. Power analysis (G*Power 3.1) for two-way ANOVA interaction effect (genotype × stimulation) with Cohen's d = 0.90 (medium-to-large, derived from PMID 35151204 showing ~70% theta-gamma rescue), α = 0.05, power = 0.85, four groups, six repeated measurements: n = 14 per group (56 total) accounting for 15% attrition; sex-balanced with equal representation (7 M + 7 F per group). Wild-type C57BL/6J mice (Jax #000664) age- and sex-matched serve as normative reference. Separate cohort of 5XFAD mice (Jax #00641) used for cross-validation of amyloid biomarkers (n = 8 per group).","hypothesis_id":"h-var-e95d2d1d86","timeline_weeks":14,"_schema_version":1,"null_hypothesis":"Closed-loop optogenetic activation of PV interneurons in APP/PS1 mice does not restore theta-gamma coupling, prevent amyloid-induced synaptic dysfunction, or improve spatial memory beyond the level achieved by non-specific neural stimulation. Specifically, the improvement in theta-gamma modulation index attributable to PV interneuron-specific activation will be indistinguishable from sham stimulation controls (Δ modulation index < 0.10).","primary_readout":"Theta-gamma coupling modulation index calculated as the phase-amplitude coupling metric from Iac选 et al. (2009) Nature Methods, computed from 30 minutes of LFP recorded during exploration of novel spatial environment (Y-maze, 10 cm arm width, 40 cm length, distal cues). Coupling index expressed as normalized modulation strength (0 = absent, 1 = maximum). Primary endpoint: modulation index at 6 weeks post-stimulation onset. Secondary timepoints at 2 weeks, 4 weeks, and 1 week post-stimulation cessation to assess durability.","estimated_cost_usd":287500.0,"secondary_readouts":["Theta frequency power (6-10 Hz) and gamma frequency power (25-100 Hz) spectral density via Welch's method","CA1 PV interneuron firing rate and action potential waveform parameters from juxtacellular recordings in acute slices (48 hours post-final stimulation session)","Synaptic markers: synaptophysin and PSD-95 protein levels by quantitative western blot (normalized to β-actin) in microdissected CA1","Soluble Aβ42 oligomer concentration by electrochemiluminescence (MSD V-PLEX) in hippocampal homogenates","Golgi-Cox spine density in CA1 pyramidal neuron apical dendrites (oblique branches 50-150 μm from soma)","Morris water maze spatial memory: hidden platform acquisition (4 trials/day, 60 s max, 5 days), probe trial (60 s, platform removed) at 24 hours post-acquisition, and 72-hour retention","CSF phospho-tau181 by Simoa (Quanterix) collected via cisterna magna puncture at terminal harvest","Hippocampal volume by MR imaging (7T, T2-weighted) at baseline and 6-week endpoint for volumetric comparison","Microglial morphometric analysis (Iba1+ cells): Sholl analysis for process ramification index as proxy for activation state","Resting-state fMRI theta coherence between hippocampus and prefrontal cortex at 7T"],"expected_effect_size":"Primary outcome (theta-gamma modulation index): Cohen's d = 0.90 (0.90) between APP/PS1-PV-ChR2 closed-loop and APP/PS1-sham groups, representing restoration to 80% of wild-type values. This effect size is supported by pilot data from PMID 35151204 showing 85% recovery of coupling indices. Power calculation: with n = 14 per group, α = 0.05 (two-tailed), the detectable effect size at 85% power is d = 0.90; smaller effects (d < 0.60) would not reach significance. Effect size for secondary outcomes: synaptic markers (Cohen's d = 0.75), spine density (d = 0.80), Morris water maze probe trial (d = 0.70). Negative control comparison (continuous vs closed-loop stimulation) expected to show d = 0.35, below clinical relevance threshold but above noise.","statistical_approach":"Linear mixed-effects model (LMM) for primary readout with fixed effects of genotype (APP/PS1 vs WT), stimulation condition (closed-loop ChR2, continuous, sham), timepoint (2, 4, 6 weeks, post-stimulation), and all interactions; mouse ID as random effect to account for repeated measures; Kenward-Roger degrees of freedom approximation. Primary contrast: genotype × stimulation interaction at 6-week timepoint. Alpha level 0.05 for primary test. Multiplicity correction for secondary readouts (24 tests): Bonferroni-Holm family-wise error rate control, α = 0.0021 per test. Effect sizes reported as Cohen's d with 95% CI. Assumption checks: Shapiro-Wilk for normality, Levene's test for homoscedasticity, visual inspection of residual distributions. Sensitivity analysis excluding outliers (>3 MAD from median). Permutation-based null distribution (10,000 iterations) for primary test as robustness check. Pre-registered at ClinicalTrials.gov prior to first animal enrollment. All analyses in R 4.3.2 using lme4 and emmeans packages.","sample_size_rationale":"Power analysis performed in G*Power 3.1 for mixed ANOVA (within-between interaction). Assumptions: medium-to-large effect size (f = 0.35, derived from literature), α = 0.05, power = 0.85, 4 groups, 6 repeated measurements, correlation among repeated measures r = 0.5 (conservative), nonsphericity correction ε = 0.75. Total N = 56 yields actual power = 0.87 for primary interaction. Sex as biological variable: balanced design with separate allocation to groups, with sex included as a priori fixed effect in LMM to detectSex × Treatment interactions; sample size provides 80% power to detect sex-by-treatment interaction of f = 0.30. Cross-validation cohort (5XFAD, n = 8/group) powered at 0.80 for amyloid biomarker endpoint (d = 1.0 expected). Sample size exceeds 3R guidelines for refinement, reduction, replacement.","falsification_criterion":"The hypothesis is falsified if: (1) Theta-gamma coupling modulation index in APP/PS1-PV-ChR2 closed-loop group fails to exceed sham controls by ≥0.10 at 6 weeks (primary endpoint), even if wild-type comparison shows significant difference between genotypes (would indicate off-target or non-specific effect); (2) PV interneuron firing rate in acute slices does not exceed 140 Hz (50% restoration threshold from PMID 35151204 baseline of 95 Hz toward wild-type 180 Hz), indicating failed circuit-level rescue; (3) Soluble Aβ42 oligomer levels show no reduction (≤10%) in PV-ChR2 group vs sham, indicating amyloid-independent mechanism that cannot prevent synaptic dysfunction; (4) Spine density remains ≤2.0 spines/μm (below the 2.8 spines/μm threshold associated with preserved memory), indicating functional non-rescue despite oscillatory improvement; (5) Morris water maze probe trial time in target quadrant does not exceed 25 seconds (40% of trial), indicating absence of spatial memory consolidation. If any criterion met, hypothesis rejected regardless of other positive findings."},"created_at":"2026-04-28T12:58:59.217143-07:00","updated_at":"2026-04-28T12:58:59.217143-07:00","version_number":3,"parent_version_id":null,"version_tag":null,"changelog":null,"is_latest":1,"lifecycle_state":"active","superseded_by":null,"deprecated_at":null,"deprecated_reason":null,"dependencies":null,"market_price":0.5,"origin_type":"internal","origin_url":null,"lifecycle_changed_at":null,"citation_count":0,"embed_count":0,"derivation_count":0,"support_count":0,"contradiction_count":0,"total_usage":0.0,"usage_score":0.5,"usage_computed_at":null,"quality_status":null,"contributors":[],"answers_question_ids":null,"deprecated_reason_detail":null,"deprecated_reason_code":null,"commit_sha":null,"commit_submodule":null,"last_mutated_at":"2026-05-16T14:51:34.657673-07:00","disputed_at":null,"gap_id":null,"mission_id":null,"intrinsic_priority":null,"effective_priority":null,"artifact_id":"6145baf5-462a-4dbd-b1f7-4d9a1e70332f","artifact_dir":null,"primary_filename":null,"accessory_filenames":null,"folder_layout_version":1,"migrated_to_folder_at":null,"hypothesis_id":null,"authorship":{"kind":"human","contributors":[{"role":"author","actor_ref":"forge_experiment_proposal_generator"}]},"epistemic_tier":"T3_provisional","created_by_agent_id":null},"outgoing_links":[{"target_artifact_id":"h-var-e95d2d1d86","link_type":"derives_from","strength":1.0,"evidence":"{\"source\": \"forge_experiment_proposal_generator\", \"method\": \"generate_experiment_proposal\"}"}],"incoming_links":[],"current_artifact_id":"experiment_proposal-59785671-1a01-4297-a6fa-8d0664feb04c","is_canonical":true,"supersede_chain":["experiment_proposal-59785671-1a01-4297-a6fa-8d0664feb04c"]}