In plain terms

The claim being assessed: Amyloid-beta is an innate antimicrobial effector, and amyloid plaque deposition in Alzheimer disease is a protective pathogen-entrapment response rather than a primary pathological lesion.

Where the evidence stands: Β — Association only. The pattern is real and repeatable, but nobody has shown it causes anything.

What this mechanism does: This is a result of the disease process, not a cause of it.

Strongest evidence for it: The preregistered weakest-link rule yields Β from SC6. Direct human support for net protective pathogen-centered plaques is localization and ex vivo activity, without functional necessity, temporal demonstration or infection-stratified benefit. SC1–5 have mechanistic or simplified-model support (Γ); neither APP receptor immunity nor clinical antibody efficacy proves the nominal protective mechanism. This grade does not deny in vitro activity or refute all pathogen contributions.

What decisive evidence would look like

In plain terms, researchers would change the proposed cause, confirm it changed as intended, and measure a meaningful health outcome while checking other explanations. The result should be able to support or count against the claim.

Relative explanatory fit: Alternative ranked ahead

What this review means

Amyloid-beta can interact with microbes in experimental systems. This review asks whether that activity operates in normal brain defense and whether human Alzheimer plaques are a protective response to infection.

Implications for health and treatment: The preregistered weakest-link rule yields Β from SC6. Direct human support for net protective pathogen-centered plaques is localization and ex vivo activity, without functional necessity, temporal demonstration or infection-stratified benefit. SC1–5 have mechanistic or simplified-model support (Γ); neither APP receptor immunity nor clinical antibody efficacy proves the nominal protective mechanism. This grade does not deny in vitro activity or refute all pathogen contributions.

Limits for individual care: 61 primary reports with 6 selected partial-body examinations; density, heterogeneous controls and incomplete access set the confidence, not correctness. The grade applies to the exact registered claim; stronger findings for narrower subclaims remain separate. No complete-paper audit or blinded calibration is claimed. This educational review does not establish an individual treatment recommendation.

Medical information notice: This review provides general health information, not medical advice. It cannot diagnose a condition or replace an individual discussion with a qualified clinician.

Source checking used complete abstracts/status records and selected original partial-body passages. No complete-paper audit or independent scientific peer review is claimed. Detailed access counts and limitations appear in the review.

A low grade is not an accusation. It means the causal work has not been done yet, which is the ordinary state of most actively studied mechanisms in biology.

At a glance

Sub-claims assessed separately

Figures

Explanatory figures summarizing this review. They carry no evidential weight of their own; where a figure and the review differ, the review governs.

Full assessment

From here the review is written for a technical reader: caps and thresholds, competing models, the independence basis, and the citation ledger.

Sources: 61 primary reports; claims: 68; verified: 68; unverified: 0. Verification is bounded attribution. Access: 6 partial-body and 55 complete abstract/status examinations. Literature cutoff: 2026-10-08; source examination UTC date: 2026-10-09. No complete-paper audit or blinded calibration claimed. Adversarial review: Generated by Sol 6.1; adversarially reviewed by Claude Opus 5.5 (Anthropic) on 2026-10-14. 3 corrections applied (see ADVERSARIAL REVIEW AND APPROVAL). Status: Approved as revised. This supersedes the 2026-10-06 approval of the 2026-10-03 edition, which the 2026-10-09 revision did not carry forward.

AT A GLANCE

Aβ antimicrobial activity is demonstrable in assays and some engineered infection models; net protective human Alzheimer plaques remain an association-level proposition.

Deciding experiment: Use adult wild-type mice with inducible Aβ-selective production/neutralization preserving APP ectodomain functions, crossed with CNS infection and matched sterile inflammation; restore the measured native peptide dose using aggregation- and binding-matched variants. Pair with prospectively ascertained pathogen-positive human tissue/outcome cohorts.

Belief versus evidence: The preregistered weakest-link rule yields Β from SC6. Direct human support for net protective pathogen-centered plaques is localization and ex vivo activity, without functional necessity, temporal demonstration or infection-stratified benefit. SC1–5 have mechanistic or simplified-model support (Γ); neither APP receptor immunity nor clinical antibody efficacy proves the nominal protective mechanism. This grade does not deny in vitro activity or refute all pathogen contributions.

Best-supported rival: Sterile innate alarm and injury amplification. Aβ-selective conditional manipulation crossed with live infection versus sterile injury, with native-dose rescue and interaction/aggregation-matched controls, would separate protective immunity from generic aggregation and physiological signalling. No examined study completes that discrimination.

COMPETING MECHANISTIC MODELS

Rank 1: Sterile innate alarm and injury amplification (Ε). Aβ-associated inflammation and deposition can participate in injury without microbial necessity. NLRP3/Casp1 loss protects memory and increases clearance; Dectin1 mediates infused Aβ inflammation (PMID23254930,37416769). Engineered disease/infused-peptide contexts do not establish every sporadic case; antimicrobial effects remain real.

Rank 2: Physiological neuromodulation with aggregation-dependent dysfunction (Δ). Native Aβ can support synapses while aggregate states are injurious. Picomolar memory effects and Aβ rescue of APP-null synaptic plasticity discriminate physiology from total peptide loss (PMID19318468,33926999); controlled human neurons show aggregate-dependent duality (PMID41026542). Most tests are ex vivo or applied peptide; not a complete human AD initiating model.

Rank 3: Pathogen-triggered pathology with harmful amyloid (Γ). Some infections induce Aβ and injury rather than necessarily protective plaques. Recurrent HSV, oral P.gingivalis and C.pneumoniae models show pathology after infection (PMID30870531,30746447,15013562). Null HSV/MRV and Toxoplasma clearance constrain universal microbial triggering (PMID38143354,35033173,26984535).

Rank 4: Nonspecific amphipathic aggregation (Γ). Concentrated aggregation can bind complex surfaces without an evolved native defence role. Concentration/state dependence and engineered assays leave a physical aggregation explanation open (PMID27225182,37594661). Scrambled/reverse controls, immunodepletion and curli-null differences constrain a wholly nonspecific account (PMID20209079,38438446).

Rank 5: Protective antimicrobial amyloid in human AD (Β). Aβ is an endogenous CNS antimicrobial effector and plaques are net protective pathogen traps. Specific assays, low-dose medium, engineered survival and human DNA localization form its strongest case (PMID20209079,27225182,18973185). Native necessity and human net protection lack matched loss/rescue; conflicting HSV assays and clinical benefit from amyloid removal remain unresolved (PMID34052228,37459141,36449413).

The tournament does not alter the nominal level. Compatible pathways and flexible alternatives are not automatically proved by explaining observations.

FULL REVIEW

Registered claim and scope

Amyloid-beta is an innate antimicrobial effector, and amyloid plaque deposition in Alzheimer disease is a protective pathogen-entrapment response rather than a primary pathological lesion.

Register all six index subclaims; the full conjunction is gated by its weakest link. Forced infection and APP transgenes are not evidence that endogenous A-beta is necessary for human CNS immunity or that sporadic AD plaques are net protective. This is a fresh rank 42 review; earlier antimicrobial-amyloid work is preserved and overlap will be disclosed.

Selection: public-interest rank 42 from the editorial ranking compiled 2026-09-23; categorical prior verdicts were visible. Private fresh review; earlier documents are preserved.

Evidence and causal interpretation

The narrow antimicrobial result deserves credit. Soscia reports peptide-sequence controls, activity against several organisms and diminished anti-Candida activity after Aβ immunodepletion of AD homogenates (PMID20209079). Examined Methods show pooled, disrupted postmortem tissue diluted into assay medium; this is ex vivo activity, not an observation of plaque protection in living patients.

Kumar advances beyond a conventional high-dose MIC assay: original Results place transformed-cell conditioned medium at reported CSF-range 2–20 ng/ml, with anti-adhesion and agglutination diminished after immunodepletion (PMID27225182). The authors distinguish this from fungicidal MIC concentrations about two orders higher. Synthetic material and cell-derived oligomers differ, so dismissing every assay as high-dose killing would be inaccurate.

The native necessity bridge nevertheless remains open. In Kumar, APP-null mortality is a trend with P=0.10, whereas5XFAD Salmonella survival is significant. No native-dose infection rescue occurs in the examined Results. APP loss also removes trophic and receptor functions; Lingel directly identifies APP as a Zika decoy receptor, and APP heparin-binding fragments are themselves antimicrobial (PMID33028637,22249992). Neither is evidence that cleaved Aβ is necessary.

Bourgade provides applied-peptide HSV entry inhibition and Aβ-dependent protection from transformed-cell conditioned medium; Lukiw finds related primary neural culture effects (PMID25376108,26836158,20683212). EV-A71 entry inhibition extends activity to a nonenveloped virus but does not erase adenovirus nonresponse or timing specificity (PMID36127711). These are functionally informative culture results with explicit exposure limits.

Eimer reports improved survival in young female5XFAD and antiherpetic low-nanomolar cell-derived peptide activity (PMID30001512). Its PMC Methods contain injection-unit conflicts; correction PMID30571943 is linked but the body is unavailable in the attempted routes. Exact dose comparisons are therefore withheld. Original antibody colocalization and 48-hour deposition are model observations, not a human functional plaque audit.

Independent challenge is substantial. Bocharova tests two HSV strains, multiple doses and ages, including female-only analyses, without significant survival protection (PMID34052228). Original Results show anti-gB T111 cross-reactivity with uninfected plaques and absent plaque virus with anti-gD/H controls. Lapeyre finds no increased amyloid after intranasal clinical-strain HSV in APP/PS1; murine roseolovirus can interact with Aβ in vitro without driving plaques in vivo (PMID38143354,35033173). Assay interaction and organismal defence must remain distinct.

Other infection models map further boundaries. Recurrent HSV and HHV6A culture studies increase AD-like markers through stress/autophagy pathways; P.gingivalis inhibition reduces bacterial load, Aβ and neuronal injury (PMID30870531,31866416,30746447). AcuteE.coli worsens learning and necrosis without changing Aβ; Toxoplasma recruits plaque-clearing monocytes with an ablation experiment and follow-up lineage tracing (PMID35530117,26984535,41547853). Infection does not have one invariant deposition response.

NontransgenicC.pneumoniae models partly address transgene dependence. An AD-brain isolate induces deposits, whereas laboratory-strain deposits resolve and a replication finds similar immunoreactivity in mock or untouched animals (PMID15013562,25538615,17581756). Nerve-path invasion is experimentally shown, but adjacent Aβ is not by itself protective trapping (PMID35177758). Recent retina/brain publication adds infection, inflammatory and model data; its two preprints are linked updates, not extra independent reports (PMID41571675).

Human localization is heterogeneous. Wozniak detects plaque-associated HSV DNA in AD and aged normals; this is DNA, not demonstrated viable virus or better tissue outcome (PMID18973185). Chlamydia PCR/culture/microscopy series report organisms, but another histological study finds regional apposition without direct plaque colocalization and independent series are negative (PMID9749980,18829386,15998578,20863379,10655406,10878049). Treponema detection is also an association (PMID11929559).

Sequencing cannot silently resolve the conflict. Readhead multiomics reportsHHV6A/7 networks; a reanalysis of the same data identifies specificity problems, and multi-repository PCR/RNA and subsequent genomic studies do not yield a general AD-specific signal (PMID29937276,32336252,31983538,41853978). The 584-brain TASTY series finds HSV in only a small minority, while brain16S patterns depend on cohort/extraction/postmortem interval (PMID27664135,32310171). These are detection and contamination boundaries, not universal claims of a sterile brain.

Recent bacterial amyloid work enriches rather than completes the model. Ali shows microbial fibril remodeling and reduced adhesion, with neuronal toxicity of remodeledFapC; Prosswimmer uses 7.5 µM peptide to disperse biofilm without significant total-density loss and enhance gentamicin susceptibility (PMID37594661,38438446). The neuronal oligomer decrease with curli deletion is P = 0.07. Amylin synergy broadens host peptide activity but remains largely culture-level and is not an adult native Aβ necessity experiment (PMID40731182).

Human clinical perturbation challenges blanket protective deposition. Donanemab and lecanemab yield less decline with ARIA and other harms (PMID37459141,36449413). Phase 2 donanemab is mixed across endpoints; phase 2 lecanemab misses its primary threshold; aducanumab has discordant pivotal trials, and gantenerumab lowers plaques without significant slowing (PMID33720637,33865446,35542991,37966285). These clinical differences do not warrant treating every aggregate as equally harmful or every removal as equally beneficial.

Solanezumab preclinical and mild-dementia trials are negative; BACE inhibition fails and can worsen selected cognition measures (PMID37458272,29365294,29719179,33049114). Enzyme pleiotropy and Aβ target species prevent assigning those harms to antimicrobial loss. Physiological Aβ has memory and synaptic roles, with Aβ-specific organotypic rescue, and controlled human neurons show an aggregation-dependent switch (PMID19318468,33926999,41026542). These furnish a serious non-antimicrobial physiological alternative.

Selective oligodendrocyte Aβ suppression, NLRP3/Casp1 loss and Dectin1 perturbation support bounded harmful or sterile immune contributions (PMID39042667,23254930,37416769). They do not prove the whole amyloid cascade or exclude infection in subgroups. The full antimicrobial-plaque conjunction rests atΒ because its human protective leg remains association-level, while its biochemical and model legs retain Γ.

HYPOTHESIS STATE EVALUATION FRAMEWORK (HSEF v5.3)

Subclaim matrix and aggregation

Preregistered aggregation: weakest-link. The preregistered weakest-link rule yields Β from SC6. Direct human support for net protective pathogen-centered plaques is localization and ex vivo activity, without functional necessity, temporal demonstration or infection-stratified benefit. SC1–5 have mechanistic or simplified-model support (Γ); neither APP receptor immunity nor clinical antibody efficacy proves the nominal protective mechanism. This grade does not deny in vitro activity or refute all pathogen contributions. Sensitivities: weakest-link Β; primary-subclaim Β; scope-matched Β.

Decisive-perturbation inventory and causal tally

[2] 5XFAD resists Salmonella; transformed-cell conditioned medium at reported CSF-range 2–20 ng/ml inhibits adhesion and agglutinates Candida, reduced by Aβ immunodepletion. APP-null mortality is only a trend (P=0.10), with no native-dose rescue in examined Results. Engineered secretion, inoculation and APP pleiotropy do not establish adult CNS Aβ necessity.

[3] Herpesvirus glycoprotein binding accelerates deposition and protective entrapment in model systems. Female 5XFAD overexpression and engineered cultures constrain inference. Examined PMC Methods contain conflicting injection units; correction PMID30571943 metadata is verified but body unavailable, so exact inoculum is not asserted. HSV protection is contradicted elsewhere.

[12] APP loss or blockade increases Zika replication and neonatal mouse mortality; APP acts as a decoy receptor. APP extracellular receptor function is distinct from cleaved Aβ antimicrobial activity; neonatal Zika cannot establish Aβ necessity in adult AD.

[13] 5XFAD fails to protect against HSV; abundant plaques lack viral particles, and aged survival differences are nonsignificant. Anti-gB T111 cross-reacts with plaques in uninfected controls; anti-gD/H do not show viral entrapment. Model and acute route constrain scope, but sex-only and dose/age analyses also fail significant survival protection.

[14] HSV infection does not increase plaque number or volume, with similar disease and survival across genotypes. One strain route and baseline model do not refute all infection-triggered amyloidosis; negative result constrains generality.

[28] Infection reduces plaque burden through recruited monocytes; monocyte ablation increases Aβ. Immune clearance can oppose deposition, showing pathogen specificity; plaque reduction alone does not measure net cognitive protection.

[46] Donanemab clears plaque and slows decline at 76 weeks with ARIA and treatment-related deaths. Net benefit is bounded to enrolled early AD; immune effects and no infection stratification prevent inferring every plaque is harmful or AMP function absent.

[47] Lecanemab lowers amyloid and yields less cognitive/functional decline with infusion reactions and ARIA. Targets soluble protofibrils as well as aggregates; clinical improvement is modest, not prevention or proof against pathogen-specific protection.

Tally: 8 discriminating contrasts. Null and negative contrasts are discriminators, not positive threshold crossings. Reports and replications are not independent causal lines. Eight especially discriminating experimental contrasts are flagged, including negative tests and two clinical countertests. Zero qualifies as positive human net-protective-plaque perturbation. Engineered infection survival, host APP loss, immune-cell ablation and clinical antibodies test distinct targets; their tally is not eight independent confirmations of antimicrobial amyloid.

Cap audit, precedence and Β/Γ/Δ boundary

Constitutive overexpression protection is credited as a model result, but cannot be transferred to endogenous CNS necessity. Native APP genetic protection does not clear Aβ-selective SC4. Infection-driven staining without clinical trajectory remains Γ for the disease-mechanistic inference. Human antibody RCTs clear functional therapeutic thresholds for the tested intervention; they are counterevidence, not positive threshold crossings for protective SC6. The human protective leg remains Β under the association rule.

Gamma hard-rule audit: Purified peptide binding, membrane disruption and curli assays alone do not establish endogenous CNS defence. Overexpression survival can be a genuine result inside that model while the inference to native peptide necessity remains capped. Plaque or Aβ biomarker movement without functional disease trajectory does not cross Δ. APP loss is not Aβ-selective loss: receptor, trophic and processing functions must be separated.

SC1 cap: “Maximum if mechanism is biologically plausible, supported by in vitro or simplified animal paradigms, but no disease-relevant necessity is shown in vivo.” Sequence controls, immunodepletion, binding and structural assays support a real peptide-associated interaction. These experiments remain concentrated or engineered assays; the native CNS mechanism is not established.

SC2 cap: “Maximum if mechanism is biologically plausible, supported by in vitro or simplified animal paradigms, but no disease-relevant necessity is shown in vivo.” Salmonella and one HSV study show survival advantages with 5XFAD expression, whereas two HSV approaches and roseolovirus challenge limit generalization. Constitutive transgenes and APP changes do not isolate native peptide causality.

SC3 cap: “Maximum if mechanism is biologically plausible, supported by in vitro or simplified animal paradigms, but no disease-relevant necessity is shown in vivo.” Transformed-cell media show low-nanomolar or reported CSF-range activity, stronger than synthetic MIC alone. No selective wild-type in vivo loss with native-dose rescue establishes the corresponding physiological CNS function.

SC4 cap: “Maximum if mechanism is biologically plausible, supported by in vitro or simplified animal paradigms, but no disease-relevant necessity is shown in vivo.” APP-null Salmonella mortality is nonsignificant and APP-null Zika mortality reflects a decoy-receptor function. APP-fragment activity and synaptic Aβ rescue cannot be transferred to Aβ-specific infection necessity.

SC5 cap: “Maximum if mechanism is biologically plausible, supported by in vitro or simplified animal paradigms, but no disease-relevant necessity is shown in vivo.” Some infection experiments increase amyloid, while HSV, roseolovirus and BALB/c replication studies are null and Toxoplasma promotes clearance. Biomarker deposition alone does not establish disease-relevant protective dependency.

SC6 cap: “association without temporal or perturbational proof caps at Β.” Human pathogen/DNA colocalization and culture observations provide association, with null detections and processing effects. No examined human study establishes that plaque entrapment improves clinical or tissue outcomes; anti-amyloid clinical benefit challenges uniformly net-protective deposition.

SC1 field-valid definition: “ - Endogenous genetic deletion, suppression, or degron depletion of the proposed mediator in vivo - Selective pathway inhibition that changes onset/progression while preserving broader viability - Vagotomy-type natural experiment, only if confounding is rigorously controlled AND effect is disease-initiation relevant rather than merely associated” Positive full-scope result: no. Sequence controls, immunodepletion, binding and structural assays support a real peptide-associated interaction. These experiments remain concentrated or engineered assays; the native CNS mechanism is not established.

SC2 field-valid definition: “ - Endogenous genetic deletion, suppression, or degron depletion of the proposed mediator in vivo - Selective pathway inhibition that changes onset/progression while preserving broader viability - Vagotomy-type natural experiment, only if confounding is rigorously controlled AND effect is disease-initiation relevant rather than merely associated” Positive full-scope result: no. Salmonella and one HSV study show survival advantages with 5XFAD expression, whereas two HSV approaches and roseolovirus challenge limit generalization. Constitutive transgenes and APP changes do not isolate native peptide causality.

SC3 field-valid definition: “ - Endogenous genetic deletion, suppression, or degron depletion of the proposed mediator in vivo - Selective pathway inhibition that changes onset/progression while preserving broader viability - Vagotomy-type natural experiment, only if confounding is rigorously controlled AND effect is disease-initiation relevant rather than merely associated” Positive full-scope result: no. Transformed-cell media show low-nanomolar or reported CSF-range activity, stronger than synthetic MIC alone. No selective wild-type in vivo loss with native-dose rescue establishes the corresponding physiological CNS function.

SC4 field-valid definition: “ - Endogenous genetic deletion, suppression, or degron depletion of the proposed mediator in vivo - Selective pathway inhibition that changes onset/progression while preserving broader viability - Vagotomy-type natural experiment, only if confounding is rigorously controlled AND effect is disease-initiation relevant rather than merely associated” Positive full-scope result: no. APP-null Salmonella mortality is nonsignificant and APP-null Zika mortality reflects a decoy-receptor function. APP-fragment activity and synaptic Aβ rescue cannot be transferred to Aβ-specific infection necessity.

SC5 field-valid definition: “ - Endogenous genetic deletion, suppression, or degron depletion of the proposed mediator in vivo - Selective pathway inhibition that changes onset/progression while preserving broader viability - Vagotomy-type natural experiment, only if confounding is rigorously controlled AND effect is disease-initiation relevant rather than merely associated” Positive full-scope result: no. Some infection experiments increase amyloid, while HSV, roseolovirus and BALB/c replication studies are null and Toxoplasma promotes clearance. Biomarker deposition alone does not establish disease-relevant protective dependency.

SC6 field-valid definition: “ - Endogenous genetic deletion, suppression, or degron depletion of the proposed mediator in vivo - Selective pathway inhibition that changes onset/progression while preserving broader viability - Vagotomy-type natural experiment, only if confounding is rigorously controlled AND effect is disease-initiation relevant rather than merely associated” Positive full-scope result: no. Human pathogen/DNA colocalization and culture observations provide association, with null detections and processing effects. No examined human study establishes that plaque entrapment improves clinical or tissue outcomes; anti-amyloid clinical benefit challenges uniformly net-protective deposition.

Upward threshold ladder

SC1 Δ: “Requires at least one genuine causal perturbation in a disease-relevant in vivo system, or equivalent human natural experiment that materially supports the mechanism.” Satisfied: no. Binding-defective versus binding-intact Aβ matched for aggregation and membrane interaction, with endogenous host infection outcomes.

SC2 Δ: “Requires at least one genuine causal perturbation in a disease-relevant in vivo system, or equivalent human natural experiment that materially supports the mechanism.” Satisfied: no. Inducible Aβ-selective host manipulation with native-level rescue, survival and pathogen load across predefined strains/routes.

SC3 Δ: “Requires at least one genuine causal perturbation in a disease-relevant in vivo system, or equivalent human natural experiment that materially supports the mechanism.” Satisfied: no. Wild-type extracellular dose measurement and peptide-selective loss/rescue under infection, preserving APP receptor and synaptic functions.

SC4 Δ: “Requires at least one genuine causal perturbation in a disease-relevant in vivo system, or equivalent human natural experiment that materially supports the mechanism.” Satisfied: no. Aβ-specific conditional loss with physiologic rescue that restores infection survival without restoring unrelated APP functions.

SC5 Δ: “Requires at least one genuine causal perturbation in a disease-relevant in vivo system, or equivalent human natural experiment that materially supports the mechanism.” Satisfied: no. Infection/sham plus pathogen clearance and Aβ-specific perturbation that links deposition to meaningful host outcome rather than staining alone.

SC6 Δ: “Requires at least one genuine causal perturbation in a disease-relevant in vivo system, or equivalent human natural experiment that materially supports the mechanism.” Satisfied: no. Contamination-controlled, prospective human pathogen ascertainment and spatial validation combined with infection-stratified plaque modulation and tissue/clinical outcomes.

Upper ladder: Human clinical antibody trials, cell-source suppression and infection/pathway genetic experiments were considered. They can establish harmful or therapeutic contributions, but no set covers native Aβ-specific loss/rescue plus protective human plaque function. Independence is assessed for the same proposition; convergence on other mechanisms cannot carry this conjunction upward.

Independence: No Ζ claim for the registered conjunction. H4/CHO and 5XFAD results share platform vulnerabilities; Readhead reanalysis shares data; Chlamydia preprint updates and laboratory series are not new independent lines. Human trials and postmortem methods fail differently, but clinical benefit points away from blanket protective deposition.

Peptide-sequence and immunodepletion assays [1, 2]: Reduce generic peptide/medium explanations but retain concentration, aggregate and antibody specificity errors.

Controlled acute infection expression contrasts [3, 13, 14]: Test survival and deposition with relevant negative results; share engineered APP biology rather than establish independent native necessity.

Human clinical amyloid manipulation [46, 47]: Provides functional counterevidence with different experimental error from postmortem spatial assays; antibody immune actions and target species still differ.

Falsification and pitfalls

Expression does not consistently protect against infection. HSV dose/sex/age study and intranasal APP/PS1 infection are negative; MRV does not increase plaques (PMID34052228,38143354,35033173). Additional downgrade: False.

Plaque colocalization can reflect assay artifact or lack pathogen specificity. Uninfected anti-gB cross-reactivity and absent anti-gD/H plaque signals; human sequencing specificity challenge and null detection (PMID34052228,32336252,31983538). Additional downgrade: False.

Removing amyloid can improve human clinical trajectory. Donanemab and lecanemab reduce decline with harms (PMID37459141,36449413); this challenges blanket protective deposition without proving every plaque harmful. Additional downgrade: False.

Applied peptide activity treated as evolved physiological necessity. µg/ml or µM assays are separated from low-dose conditioned medium and native in vivo rescue.

APP loss attributed exclusively to Aβ. Zika decoy receptor and non-Aβ APP fragments are scored as alternatives.

DNA or antigen presence treated as viable protective entrapment. Human colocalization and culture are separated from plaque functional outcomes.

Biofilm redistribution treated as microbial killing. Prosswimmer total density is not significantly reduced; curli-dependent oligomer decrease is P=0.07.

Clinical nulls or harms treated as proof of AMP loss. BACE pleiotropy, target species, clearance and disease stage remain explicit.

Provider full-text claim treated as source access. Only Soscia/Kumar/Bocharova returned original aspects are compared; missing originals and correction remain unavailable.

Adversarial steelman: above, then below

One rung above Β is Γ: the low-dose conditioned medium, immunodepletion, transgenic survival and spatial human findings jointly make a plausible functional protective model, stronger than a bare association.

One rung below Β is Α: human localization may reflect antibody/sequence artifact, late barrier failure or opportunistic microbes, leaving protective plaque function mostly conceptual.

Γ is appropriate to SC1–5 but rejected for the full weakest-link conjunction because human SC6 has no temporally or perturbationally demonstrated positive protective outcome. Α understates reproducible specific assay activity and substantive human localization reports. The original positive and negative functional data remain separate from grade.

Final score, causal direction and M/S/D

The preregistered weakest-link rule yields Β from SC6. Direct human support for net protective pathogen-centered plaques is localization and ex vivo activity, without functional necessity, temporal demonstration or infection-stratified benefit. SC1–5 have mechanistic or simplified-model support (Γ); neither APP receptor immunity nor clinical antibody efficacy proves the nominal protective mechanism. This grade does not deny in vitro activity or refute all pathogen contributions.

For the registered infection-response mechanism, deposition follows microbial challenge in the positive models and can resolve with pathogen burden. This identifies a downstream response within those systems, not the primary infection initiator or a proved protective human function. Aβ can separately amplify sterile injury and support normal synapses; no role is assigned from tournament rank.

Momentum: Mechanistic only. Prosswimmer 2024 curli assays, Kumar 2025 amylin synergy and 2026 retina pathogen/inflammasome work extend mechanism and localization. None supplies the missing Aβ-selective native infection rescue plus human plaque outcome.

Stability: Stable. No named pending result is verified that would imminently cross human protective SC6; ongoing mechanistic publication is not a directional score trigger.

Divergence: Contested; no single implied-level arithmetic. These dimensions do not change level or confidence.

PROPOSED DECISIVE EXPERIMENTS

Use adult wild-type mice with inducible Aβ-selective production/neutralization preserving APP ectodomain functions, crossed with CNS infection and matched sterile inflammation; restore the measured native peptide dose using aggregation- and binding-matched variants. Pair with prospectively ascertained pathogen-positive human tissue/outcome cohorts.

Readout: Pathogen viable load, survival, synaptic injury and cognition alongside free/aggregate Aβ concentrations and spatial localization. Controls: APP receptor competence, noninfected and sterile-injury cohorts, sequence/aggregate controls, uninfected staining controls, extraction blanks and prespecified strain/sex/age. Supporting: Selective loss worsens infection outcome and native-dose binding-intact rescue restores protection without increased neural injury; human pathogen-trapping predicts better tissue outcomes. Falsifying: Loss has no infection consequence, rescue works through non-Aβ APP or generic aggregation, or deposits track injury without net outcome benefit.

RESEARCH TIMELINE

KNOWLEDGE GAPS

Selective adult Aβ loss with native-dose infection rescue preserving other APP functions.

Measured free and aggregate Aβ exposure in living infected wild-type CNS, not only CSF comparison to transformed medium.

Mechanistic reconciliation of HSV strains/routes, anti-gB cross-reactivity and positive versus negative survival studies.

Contamination-controlled viable pathogen localization inside human plaques with host injury and longitudinal outcomes.

Infection-stratified anti-amyloid intervention and immune safety; current RCT abstracts do not supply this test.

Direct membrane killing versus adhesion inhibition, microbial dispersion and antibiotic synergy under native ionic conditions.

Pathogen-specific deposition versus recruited immune clearance and sterile inflammatory induction.

Full correction-body audit for Eimer and lecanemab phase 2, plus supplements and inaccessible trial/colocalization originals.

METHODS AND CONSENSUS CONTRIBUTION

Consensus was used through the signed-in browser for discovery, counterevidence, research-gap and timeline suggestions, and attempted checks of specific indexed full-text passages. Provider statements about Methods/Results and any returned passages were compared with original-paper passages where available; failed retrievals, source substitutions and unverified aspects are recorded in the case reconciliation. Successful original-passage comparisons are distinguished from provider summaries and assertions of access. Provider screening counts, citation counts and meters were not used as primary counts or HSEF scores. This source log records access depth; it does not claim complete-paper or supplement review. The monthly API allowance was conserved: zero Consensus API calls in this batch.

REVIEWER NOTES

Fresh rank 42 review preserving earlier antimicrobial-amyloid work; foundational overlap is expected and not independent replication. All 61 primary abstracts/status records were examined, six original bodies partially examined. PubMed links verify Eimer/lecanemab correction metadata and final retina publication/preprint updates; correction bodies were unavailable, so no complete correction audit. Consensus 2 Pro, 0 Deep, 0 API; SciSpace leads resolved to canonical originals, duplicated/wrong-title commentary and cover metadata excluded. Literature cutoff remains 2026-10-08 despite UTC examination 2026-10-09.

Search leads, reviews, protocols and correction notices do not pad the primary count. Downloaded XML does not earn examination credit. Linked notices remain attached; unexamined correction details limit inference. This is a coordinator source/claim consistency assessment, not independent blinded calibration.

ADVERSARIAL REVIEW AND APPROVAL

Authoring model: Sol 6.1 (HSEF v5.3 pipeline). Adversarial reviewer: Claude Opus 5.5 (Anthropic). Date: 2026-10-14.

Standing of the earlier approval. The 2026-10-03 edition of this review (deliverables/hsef_v53_neurodegeneration_2026-10-03/antimicrobial-amyloid/) was adversarially reviewed and approved on 2026-10-06 with 0 corrections. The 2026-10-09 revision is a re-authored edition, not a patch: it changed the title, replaced 37 of 60 references with 38 new ones (23 shared), moved the literature cutoff from 2026-10-03 to 2026-10-08, added a sixth subclaim, and carried neither the earlier approval banner nor its approval section. That earlier approval therefore did not transfer. The present section is a fresh adjudication of the current edition and supersedes it.

What the revision changed, and whether any change is a regression.

1. Headline letter: Γ (Mechanistically plausible, partial system support) to Β (Associational). Not a regression; better grounded. Both editions preregister weakest-link aggregation. The 2026-10-03 edition graded its human-lesion-identity subclaim (SC5) Γ; this edition splits that leg into an explicit human subclaim SC6 — "Human Alzheimer plaques form around real pathogens as a net protective response", scoped to biomarker-defined sporadic AD with infection ascertainment and a clinical horizon — and caps it at Β by quoting the rule "association without temporal or perturbational proof caps at Β". The examined human evidence for that leg is pathogen/DNA colocalization and culture detection (Wozniak PMID 18973185; the Chlamydia series), with null detections and processing/contamination boundaries (the 584-brain TASTY series PMID 27664135, the brain 16S work, the HHV6A/7 reanalysis PMID 32336252), and no study in which plaque entrapment improves a clinical or tissue outcome. The framework's Β definition explicitly covers "pathology-correlation" claims, so the cap is correctly applied, and under weakest-link the conjunction follows it. The letter was therefore not changed back.
2. Causal role: epiphenomenon (step 2) to downstream consequence (step 1). Not a regression; corrected. Step 2 requires that no causal dependency be demonstrated in either direction. In the positive infection models deposition demonstrably follows microbial challenge and can resolve with pathogen burden, which is a demonstrated dependency and so stops the derivation at step 1 rather than step 2.
3. Subclaim count 5 to 6; sources 60 to 61; load-bearing claims 66 to 68. Internally consistent: 61 ledger rows, 61 evidence items, 6 partial-body plus 55 abstract/status examinations, and the provenance header all agree with the banner.
4. Momentum, Stability and Divergence (Mechanistic only / Stable / Contested) are unchanged and the Momentum artifact names real mechanistic work (Prosswimmer 2024 curli assays, Kumar 2025 amylin synergy, the 2026 retina pathogen/inflammasome report) while stating that none supplies the missing Aβ-selective native infection rescue. Correct under the framework's definition.
5. Detail dropped without error. Two pieces of precision that the earlier edition carried and the 2026-10-06 review verified are absent here: the lanabecestat DAYBREAK amyloid-PET sign correction (+2.2 rather than −2.2 centiloids) and Soscia's "eight of twelve" organism count taken from the Discussion. Neither omission makes any statement in this edition wrong — the BACE-inhibitor evidence is retained through PMID 33049114, and "activity against several organisms" is accurate — so this is recorded as a loss of detail rather than a correction.

Scope of checks on the current edition. All 61 REFERENCES PMIDs were re-verified against PubMed metadata: 61/61 resolve to the stated title, first author, journal and year, and 61/61 printed DOI strings match the PubMed DOI field exactly. No reference carries a retraction or expression of concern, and no reference has a consortium first author. Ten load-bearing claims were checked against the original abstracts: Soscia (eight clinically relevant organisms, higher antimicrobial activity in AD than age-matched non-AD homogenates, activity ablated by Aβ immunodepletion); Kumar (5XFAD resistance to Salmonella with rapid seeding and colocalization, oligomerization required for activity); Bocharova (two HSV-1 strains at three doses, 5XFAD failing to protect, a nonsignificant survival difference at 7-10 months, Aβ aggregates free of viral particles and anti-gB cross-reactivity with uninfected plaques); Lapeyre (no increase in plaque number or volume after intranasal HSV in APP/PS1); the Toxoplasma study (reduced Aβ and plaque load with recruited CCR2 monocytes, and increased Aβ after selective monocyte ablation); Wozniak (90% of AD plaques containing HSV-1 DNA and 80% in aged normals, with the plaque-associated fraction differing); the TASTY series (HSV in 11 of 584 brains, 1.9%); the HHV6A/7 reanalysis (no HHV7 reads, HHV6A in 2 of the top 15 samples); Wessels (verubecestat and lanabecestat worsening selected cognitive measures); and the donanemab and lecanemab phase-3 outcomes with ARIA. All ten were supported as stated.

Rejected findings (checked, no change).

1. Aggregation. Weakest-link is preregistered and applied as written; all three alternatives are reported and all return Β. SC1-SC5 remain Γ and are reported separately, so no narrower subclaim is being used to set a broader headline.
2. Β/Γ boundary for the conjunction. The biochemical and model legs genuinely sit at Γ, but the registered claim is the conjunction including net human protection, and the conjunction is gated by its weakest link by preregistration. Raising the headline to Γ would require either abandoning the preregistered rule or lifting the quoted association cap on SC6.
3. Treating PMID 32336252 as a primary report although PubMed types it "Comment". It is a reanalysis of the primary sequencing data with original computation, and the review describes it as such rather than as an independent cohort. Defensible.
4. The clinical anti-amyloid trials used as counterevidence rather than as positive threshold crossings for the nominal claim. Correct: they test an intervention, not the protective-deposition proposition, and the review says so explicitly.
5. PubMed title fields for references 8 and 21 drop the italicised "Aβ1-42" spacing and the species name "E. coli"; the review prints the full correct titles. Standing rejected-finding convention.
6. Validator warning v5.claim_form ("entry_form is 'phenomenon'"). The registered claim states both a presence/function and its consequence, which is the documented exception. Rejected.
7. The 9 October 2026 provenance-reconciliation block that sits inside the framework section. It states that it changes no score or source count and that the earlier adversarial approval refers to the original edition - accurate, and retained unchanged. The self-contradiction reported elsewhere in this audit batch, where a spliced consensus block sits beside preserved notes asserting that no Consensus work was done, does not occur here: the preserved METHODS AND CONSENSUS CONTRIBUTION section already records signed-in Consensus browser use for discovery, counterevidence and attempted indexed full-text checks, so the two blocks agree.
8. Confidence band. Re-derived independently of access depth (see finding C3); on evidence density and reproducibility alone the band remains Moderate, and the framework explicitly allows a low letter to carry a confident ceiling.
9. "Best-supported rival" renderer defect. Checked against the scorecard: models_evaluated[0] is rank 1 with is_nominal false and the review prints that rank-1 rival; the nominal ("Protective antimicrobial amyloid in human AD") is rank 5. The defect does not occur here.

Residual limitations. Six sources were examined at partial-body depth and 55 at abstract/status depth, so Methods-level figures (the reported CSF-range 2-20 ng/ml conditioned-medium activity, the APP-null mortality trend at P = 0.10, the 7.5 µM biofilm-dispersal concentration, the P = 0.07 curli-deletion oligomer difference) rest on the package's own retrieval receipts. The correction linked to PMID 30001512 (PMID 30571943) remains metadata-verified but body-unavailable, so the review's refusal to assert exact inoculum figures is retained. The scorecard and provenance header name the generating system "Codex"; this adversarial review does not alter package-level provenance strings. A divergent copy of this slug exists at deliverables/hsef_v53_popularity_ranks_41_45_2026-10-08/antimicrobial-amyloid/ and was deliberately not modified.

Validator. The frozen HSEF v5.3 validator was re-run on the revised scorecard and prose: 33 pass, 1 warn, 1 fail. The warning is v5.claim_form, rejected above by standing convention. The failure is v5.lay_summary, which the validator itself marks "mandatory from v5.4" and which therefore does not apply to this v5.3 package. All v5.3 checks pass, including level.glyph (Β), level.descriptor, logic.aggregation, logic.cap_ceiling, v5.ledger (61 sources / 68 claims) and v5.provenance.

Approval: Following adversarial review and the corrections listed above, the revised document is approved. - Claude Opus 5.5 (Anthropic), 2026-10-14.

REFERENCES

1. Soscia SJ, Kirby JE, Washicosky KJ et al. The Alzheimer's disease-associated amyloid beta-protein is an antimicrobial peptide. PloS one. 2010. PMID 20209079. DOI: 10.1371/journal.pone.0009505. Access: full_text_partial. Study: PMID20209079.

2. Kumar DK, Choi SH, Washicosky KJ et al. Amyloid-β peptide protects against microbial infection in mouse and worm models of Alzheimer's disease. Science translational medicine. 2016. PMID 27225182. DOI: 10.1126/scitranslmed.aaf1059. Access: full_text_partial. Study: PMID27225182.

3. Eimer WA, Vijaya Kumar DK, Navalpur Shanmugam NK et al. Alzheimer's Disease-Associated β-Amyloid Is Rapidly Seeded by Herpesviridae to Protect against Brain Infection. Neuron. 2018. PMID 30001512. DOI: 10.1016/j.neuron.2018.06.030. Access: full_text_partial. Study: PMID30001512.

4. Bourgade K, Garneau H, Giroux G et al. β-Amyloid peptides display protective activity against the human Alzheimer's disease-associated herpes simplex virus-1. Biogerontology. 2015. PMID 25376108. DOI: 10.1007/s10522-014-9538-8. Access: abstract_metadata. Study: PMID25376108.

5. Bourgade K, Le Page A, Bocti C et al. Protective Effect of Amyloid-β Peptides Against Herpes Simplex Virus-1 Infection in a Neuronal Cell Culture Model. Journal of Alzheimer's disease : JAD. 2016. PMID 26836158. DOI: 10.3233/JAD-150652. Access: abstract_metadata. Study: PMID26836158.

6. Lukiw WJ, Cui JG, Yuan LY et al. Acyclovir or Aβ42 peptides attenuate HSV-1-induced miRNA-146a levels in human primary brain cells. Neuroreport. 2010. PMID 20683212. DOI: 10.1097/WNR.0b013e32833da51a. Access: abstract_metadata. Study: PMID20683212.

7. Papareddy P, Mörgelin M, Walse B et al. Antimicrobial activity of peptides derived from human ß-amyloid precursor protein. Journal of peptide science : an official publication of the European Peptide Society. 2012. PMID 22249992. DOI: 10.1002/psc.1439. Access: abstract_metadata. Study: PMID22249992.

8. Zhong M, Wang H, Yan H et al. Effects and mechanism of Aβ1-42 on EV-A71 replication. Virology journal. 2022. PMID 36127711. DOI: 10.1186/s12985-022-01882-3. Access: abstract_metadata. Study: PMID36127711.

9. Ali SA, Chung KHK, Forgham H et al. Alzheimer's Progenitor Amyloid-β Targets and Dissolves Microbial Amyloids and Impairs Biofilm Function. Advanced science (Weinheim, Baden-Wurttemberg, Germany). 2023. PMID 37594661. DOI: 10.1002/advs.202301423. Access: abstract_metadata. Study: PMID37594661.

10. Prosswimmer T, Heng A, Daggett V. Mechanistic insights into the role of amyloid-β in innate immunity. Scientific reports. 2024. PMID 38438446. DOI: 10.1038/s41598-024-55423-9. Access: full_text_partial. Study: PMID38438446.

11. Vijaya Kumar DK, Mitchell TA, Tailor BA et al. Human amylin is a potent antimicrobial peptide that exhibits antimicrobial synergism with the amyloid beta protein. Alzheimer's & dementia : the journal of the Alzheimer's Association. 2025. PMID 40731182. DOI: 10.1002/alz.70490. Access: abstract_metadata. Study: PMID40731182.

12. Lingel A, Lin H, Gavriel Y et al. Amyloid precursor protein is a restriction factor that protects against Zika virus infection in mammalian brains. The Journal of biological chemistry. 2020. PMID 33028637. DOI: 10.1074/jbc.RA120.015165. Access: full_text_partial. Study: PMID33028637.

13. Bocharova O, Pandit NP, Molesworth K et al. Alzheimer's disease-associated β-amyloid does not protect against herpes simplex virus 1 infection in the mouse brain. The Journal of biological chemistry. 2021. PMID 34052228. DOI: 10.1016/j.jbc.2021.100845. Access: full_text_partial. Study: PMID34052228.

14. Lapeyre L, Piret J, Rhéaume C et al. Herpes Simplex Virus 1 Infection Does Not Increase Amyloid-β Pathology in APP/PS1 Mice. Journal of Alzheimer's disease : JAD. 2024. PMID 38143354. DOI: 10.3233/JAD-230746. Access: abstract_metadata. Study: PMID38143354.

15. Voth S, Gwin M, Francis CM et al. Virulent Pseudomonas aeruginosa infection converts antimicrobial amyloids into cytotoxic prions. FASEB journal : official publication of the Federation of American Societies for Experimental Biology. 2020. PMID 32413239. DOI: 10.1096/fj.202000051RRR. Access: abstract_metadata. Study: PMID32413239.

16. Cao J, Wang M, Gong C et al. Identification of Microbiota within Aβ Plaque in APP/PS1 Transgenic Mouse. Journal of molecular neuroscience : MN. 2021. PMID 33098544. DOI: 10.1007/s12031-020-01715-4. Access: abstract_metadata. Study: PMID33098544.

17. Wang M, Peng IF, Li S et al. Dysregulation of antimicrobial peptide expression distinguishes Alzheimer's disease from normal aging. Aging. 2020. PMID 31907335. DOI: 10.18632/aging.102650. Access: abstract_metadata. Study: PMID31907335.

18. Dominy SS, Lynch C, Ermini F et al. Porphyromonas gingivalis in Alzheimer's disease brains: Evidence for disease causation and treatment with small-molecule inhibitors. Science advances. 2019. PMID 30746447. DOI: 10.1126/sciadv.aau3333. Access: abstract_metadata. Study: PMID30746447.

19. De Chiara G, Piacentini R, Fabiani M et al. Recurrent herpes simplex virus-1 infection induces hallmarks of neurodegeneration and cognitive deficits in mice. PLoS pathogens. 2019. PMID 30870531. DOI: 10.1371/journal.ppat.1007617. Access: abstract_metadata. Study: PMID30870531.

20. Romeo MA, Gilardini Montani MS, Gaeta A et al. HHV-6A infection dysregulates autophagy/UPR interplay increasing beta amyloid production and tau phosphorylation in astrocytoma cells as well as in primary neurons, possible molecular mechanisms linking viral infection to Alzheimer's disease. Biochimica et biophysica acta. Molecular basis of disease. 2020. PMID 31866416. DOI: 10.1016/j.bbadis.2019.165647. Access: abstract_metadata. Study: PMID31866416.

21. Schütze S, Döpke A, Kellert B et al. Intracerebral Infection with E. coli Impairs Spatial Learning and Induces Necrosis of Hippocampal Neurons in the Tg2576 Mouse Model of Alzheimer's Disease. Journal of Alzheimer's disease reports. 2022. PMID 35530117. DOI: 10.3233/ADR-210049. Access: abstract_metadata. Study: PMID35530117.

22. Hammond CJ, Hallock LR, Howanski RJ et al. Immunohistological detection of Chlamydia pneumoniae in the Alzheimer's disease brain. BMC neuroscience. 2010. PMID 20863379. DOI: 10.1186/1471-2202-11-121. Access: abstract_metadata. Study: PMID20863379.

23. Dreses-Werringloer U, Bhuiyan M, Zhao Y et al. Initial characterization of Chlamydophila (Chlamydia) pneumoniae cultured from the late-onset Alzheimer brain. International journal of medical microbiology : IJMM. 2009. PMID 18829386. DOI: 10.1016/j.ijmm.2008.07.002. Access: abstract_metadata. Study: PMID18829386.

24. Gérard HC, Wildt KL, Whittum-Hudson JA et al. The load of Chlamydia pneumoniae in the Alzheimer's brain varies with APOE genotype. Microbial pathogenesis. 2005. PMID 15998578. DOI: 10.1016/j.micpath.2005.05.002. Access: abstract_metadata. Study: PMID15998578.

25. Little CS, Hammond CJ, MacIntyre A et al. Chlamydia pneumoniae induces Alzheimer-like amyloid plaques in brains of BALB/c mice. Neurobiology of aging. 2004. PMID 15013562. DOI: 10.1016/S0197-4580(03)00127-1. Access: abstract_metadata. Study: PMID15013562.

26. Balin BJ, Gérard HC, Arking EJ et al. Identification and localization of Chlamydia pneumoniae in the Alzheimer's brain. Medical microbiology and immunology. 1998. PMID 9749980. DOI: 10.1007/s004300050071. Access: abstract_metadata. Study: PMID9749980.

27. Gaire BP, Koronyo Y, Vit JP et al. Identification of Chlamydia pneumoniae and NLRP3 inflammasome activation in Alzheimer's disease retina. Nature communications. 2026. PMID 41571675. DOI: 10.1038/s41467-026-68580-4. Access: abstract_metadata. Study: Chlamydia-retina-final-publication.

28. Möhle L, Israel N, Paarmann K et al. Chronic Toxoplasma gondii infection enhances β-amyloid phagocytosis and clearance by recruited monocytes. Acta neuropathologica communications. 2016. PMID 26984535. DOI: 10.1186/s40478-016-0293-8. Access: abstract_metadata. Study: PMID26984535.

29. Yanes KJO, Bui CT, Tomasello J et al. Toxoplasma gondii drives myeloid immune cell recruitment to amyloid plaques in Alzheimer's model mice. Journal of neuroinflammation. 2026. PMID 41547853. DOI: 10.1186/s12974-025-03666-2. Access: abstract_metadata. Study: PMID41547853.

30. Jamieson GA, Maitland NJ, Wilcock GK et al. Latent herpes simplex virus type 1 in normal and Alzheimer's disease brains. Journal of medical virology. 1991. PMID 1649907. DOI: 10.1002/jmv.1890330403. Access: abstract_metadata. Study: PMID1649907.

31. Wozniak MA, Mee AP, Itzhaki RF. Herpes simplex virus type 1 DNA is located within Alzheimer's disease amyloid plaques. The Journal of pathology. 2009. PMID 18973185. DOI: 10.1002/path.2449. Access: abstract_metadata. Study: PMID18973185.

32. Olsson J, Lövheim H, Honkala E et al. HSV presence in brains of individuals without dementia: the TASTY brain series. Disease models & mechanisms. 2016. PMID 27664135. DOI: 10.1242/dmm.026674. Access: abstract_metadata. Study: PMID27664135.

33. Allnutt MA, Johnson K, Bennett DA et al. Human Herpesvirus 6 Detection in Alzheimer's Disease Cases and Controls across Multiple Cohorts. Neuron. 2020. PMID 31983538. DOI: 10.1016/j.neuron.2019.12.031. Access: abstract_metadata. Study: PMID31983538.

34. Bigley TM, Xiong M, Ali M et al. Murine roseolovirus does not accelerate amyloid-β pathology and human roseoloviruses are not over-represented in Alzheimer disease brains. Molecular neurodegeneration. 2022. PMID 35033173. DOI: 10.1186/s13024-021-00514-8. Access: abstract_metadata. Study: PMID35033173.

35. Chorlton SD. Reanalysis of Alzheimer's brain sequencing data reveals absence of purported HHV6A and HHV7. Journal of bioinformatics and computational biology. 2020. PMID 32336252. DOI: 10.1142/S0219720020500122. Access: abstract_metadata. Study: Readhead-sequencing-series-reanalysis.

36. Readhead B, Haure-Mirande JV, Funk CC et al. Multiscale Analysis of Independent Alzheimer's Cohorts Finds Disruption of Molecular, Genetic, and Clinical Networks by Human Herpesvirus. Neuron. 2018. PMID 29937276. DOI: 10.1016/j.neuron.2018.05.023. Access: abstract_metadata. Study: Readhead-sequencing-series.

37. Piotrowski SL, Allnutt MA, Johnson K et al. Herpesvirus genome integration in whole-genome sequences of dementia and control cohorts. Alzheimer's & dementia : the journal of the Alzheimer's Association. 2026. PMID 41853978. DOI: 10.1002/alz.71047. Access: abstract_metadata. Study: PMID41853978.

38. Itzhaki RF, Lin WR, Shang D et al. Herpes simplex virus type 1 in brain and risk of Alzheimer's disease. Lancet (London, England). 1997. PMID 9014911. DOI: 10.1016/S0140-6736(96)10149-5. Access: abstract_metadata. Study: PMID9014911.

39. Chacko A, Delbaz A, Walkden H et al. Chlamydia pneumoniae can infect the central nervous system via the olfactory and trigeminal nerves and contributes to Alzheimer's disease risk. Scientific reports. 2022. PMID 35177758. DOI: 10.1038/s41598-022-06749-9. Access: abstract_metadata. Study: PMID35177758.

40. Boelen E, Stassen FR, van der Ven AJ et al. Detection of amyloid beta aggregates in the brain of BALB/c mice after Chlamydia pneumoniae infection. Acta neuropathologica. 2007. PMID 17581756. DOI: 10.1007/s00401-007-0252-3. Access: abstract_metadata. Study: PMID17581756.

41. Little CS, Joyce TA, Hammond CJ et al. Detection of bacterial antigens and Alzheimer's disease-like pathology in the central nervous system of BALB/c mice following intranasal infection with a laboratory isolate of Chlamydia pneumoniae. Frontiers in aging neuroscience. 2014. PMID 25538615. DOI: 10.3389/fnagi.2014.00304. Access: abstract_metadata. Study: PMID25538615.

42. Gieffers J, Reusche E, Solbach W et al. Failure to detect Chlamydia pneumoniae in brain sections of Alzheimer's disease patients. Journal of clinical microbiology. 2000. PMID 10655406. DOI: 10.1128/JCM.38.2.881-882.2000. Access: abstract_metadata. Study: PMID10655406.

43. Ring RH, Lyons JM. Failure to detect Chlamydia pneumoniae in the late-onset Alzheimer's brain. Journal of clinical microbiology. 2000. PMID 10878049. DOI: 10.1128/JCM.38.7.2591-2594.2000. Access: abstract_metadata. Study: PMID10878049.

44. Riviere GR, Riviere KH, Smith KS. Molecular and immunological evidence of oral Treponema in the human brain and their association with Alzheimer's disease. Oral microbiology and immunology. 2002. PMID 11929559. DOI: 10.1046/j.0902-0055.2001.00100.x. Access: abstract_metadata. Study: PMID11929559.

45. Westfall S, Dinh DM, Pasinetti GM. Investigation of Potential Brain Microbiome in Alzheimer's Disease: Implications of Study Bias. Journal of Alzheimer's disease : JAD. 2020. PMID 32310171. DOI: 10.3233/JAD-191328. Access: abstract_metadata. Study: PMID32310171.

46. Sims JR, Zimmer JA, Evans CD et al. Donanemab in Early Symptomatic Alzheimer Disease: The TRAILBLAZER-ALZ 2 Randomized Clinical Trial. JAMA. 2023. PMID 37459141. DOI: 10.1001/jama.2023.13239. Access: abstract_metadata. Study: PMID37459141.

47. van Dyck CH, Swanson CJ, Aisen P et al. Lecanemab in Early Alzheimer's Disease. The New England journal of medicine. 2023. PMID 36449413. DOI: 10.1056/NEJMoa2212948. Access: abstract_metadata. Study: PMID36449413.

48. Mintun MA, Lo AC, Duggan Evans C et al. Donanemab in Early Alzheimer's Disease. The New England journal of medicine. 2021. PMID 33720637. DOI: 10.1056/NEJMoa2100708. Access: abstract_metadata. Study: PMID33720637.

49. Swanson CJ, Zhang Y, Dhadda S et al. A randomized, double-blind, phase 2b proof-of-concept clinical trial in early Alzheimer's disease with lecanemab, an anti-Aβ protofibril antibody. Alzheimer's research & therapy. 2021. PMID 33865446. DOI: 10.1186/s13195-021-00813-8. Access: abstract_metadata. Study: PMID33865446.

50. Budd Haeberlein S, Aisen PS, Barkhof F et al. Two Randomized Phase 3 Studies of Aducanumab in Early Alzheimer's Disease. The journal of prevention of Alzheimer's disease. 2022. PMID 35542991. DOI: 10.14283/jpad.2022.30. Access: abstract_metadata. Study: EMERGE-and-ENGAGE-report.

51. Bateman RJ, Smith J, Donohue MC et al. Two Phase 3 Trials of Gantenerumab in Early Alzheimer's Disease. The New England journal of medicine. 2023. PMID 37966285. DOI: 10.1056/NEJMoa2304430. Access: abstract_metadata. Study: GRADUATE-I-and-II-report.

52. Sperling RA, Donohue MC, Raman R et al. Trial of Solanezumab in Preclinical Alzheimer's Disease. The New England journal of medicine. 2023. PMID 37458272. DOI: 10.1056/NEJMoa2305032. Access: abstract_metadata. Study: PMID37458272.

53. Egan MF, Kost J, Tariot PN et al. Randomized Trial of Verubecestat for Mild-to-Moderate Alzheimer's Disease. The New England journal of medicine. 2018. PMID 29719179. DOI: 10.1056/NEJMoa1706441. Access: abstract_metadata. Study: PMID29719179.

54. Honig LS, Vellas B, Woodward M et al. Trial of Solanezumab for Mild Dementia Due to Alzheimer's Disease. The New England journal of medicine. 2018. PMID 29365294. DOI: 10.1056/NEJMoa1705971. Access: abstract_metadata. Study: PMID29365294.

55. Wessels AM, Lines C, Stern RA et al. Cognitive outcomes in trials of two BACE inhibitors in Alzheimer's disease. Alzheimer's & dementia : the journal of the Alzheimer's Association. 2020. PMID 33049114. DOI: 10.1002/alz.12164. Access: abstract_metadata. Study: PMID33049114.

56. Garcia-Osta A, Alberini CM. Amyloid beta mediates memory formation. Learning & memory (Cold Spring Harbor, N.Y.). 2009. PMID 19318468. DOI: 10.1101/lm.1310209. Access: abstract_metadata. Study: PMID19318468.

57. Galanis C, Fellenz M, Becker D et al. Amyloid-Beta Mediates Homeostatic Synaptic Plasticity. The Journal of neuroscience : the official journal of the Society for Neuroscience. 2021. PMID 33926999. DOI: 10.1523/JNEUROSCI.1820-20.2021. Access: abstract_metadata. Study: PMID33926999.

58. Siddu A, Natale S, Wong CH et al. Aggregation shifts amyloid-β peptides from synaptogenic to synaptotoxic. The Journal of clinical investigation. 2025. PMID 41026542. DOI: 10.1172/JCI193407. Access: abstract_metadata. Study: PMID41026542.

59. Rajani RM, Ellingford R, Hellmuth M et al. Selective suppression of oligodendrocyte-derived amyloid beta rescues neuronal dysfunction in Alzheimer's disease. PLoS biology. 2024. PMID 39042667. DOI: 10.1371/journal.pbio.3002727. Access: abstract_metadata. Study: PMID39042667.

60. Heneka MT, Kummer MP, Stutz A et al. NLRP3 is activated in Alzheimer's disease and contributes to pathology in APP/PS1 mice. Nature. 2013. PMID 23254930. DOI: 10.1038/nature11729. Access: abstract_metadata. Study: PMID23254930.

61. Zhao X, Sun J, Xiong L et al. β-amyloid binds to microglia Dectin-1 to induce inflammatory response in the pathogenesis of Alzheimer's disease. International journal of biological sciences. 2023. PMID 37416769. DOI: 10.7150/ijbs.81900. Access: abstract_metadata. Study: PMID37416769.

Source-specific evidence cards

[1] The Alzheimer's disease-associated amyloid beta-protein is an antimicrobial peptide.

Shows: Synthetic Aβ inhibits multiple microbes; homogenate activity correlates with Aβ and falls after immunodepletion.

Does not show: Disrupted postmortem homogenate is not native extracellular concentration, living CNS defence or net protective plaques.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 3:Materials and Methods:paragraph 5, 3:Materials and Methods:paragraph 10, 3:Materials and Methods:paragraph 12, 1:Results. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3, SC6.

[2] Amyloid-β peptide protects against microbial infection in mouse and worm models of Alzheimer's disease.

Shows: 5XFAD resists Salmonella; transformed-cell conditioned medium at reported CSF-range 2–20 ng/ml inhibits adhesion and agglutinates Candida, reduced by Aβ immunodepletion.

Does not show: APP-null mortality is only a trend (P=0.10), with no native-dose rescue in examined Results. Engineered secretion, inoculation and APP pleiotropy do not establish adult CNS Aβ necessity.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Results:paragraph 1, 1:Results:paragraph 6, 1:Results:paragraph 7, 1:Results:paragraph 9, 1:Results:paragraph 10, 1:Results:paragraph 16. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC4, SC5.

[3] Alzheimer's Disease-Associated β-Amyloid Is Rapidly Seeded by Herpesviridae to Protect against Brain Infection.

Shows: Herpesvirus glycoprotein binding accelerates deposition and protective entrapment in model systems.

Does not show: Female 5XFAD overexpression and engineered cultures constrain inference. Examined PMC Methods contain conflicting injection units; correction PMID30571943 metadata is verified but body unavailable, so exact inoculum is not asserted. HSV protection is contradicted elsewhere.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:RESULTS:paragraph 1, 1:RESULTS:paragraph 2, 1:RESULTS:paragraph 7, 1:RESULTS:paragraph 8, 6:METHOD DETAILS:paragraph 0, 6:METHOD DETAILS:paragraph 1, 6:METHOD DETAILS:paragraph 4, 6:METHOD DETAILS:paragraph 6. No complete-paper or supplement audit claimed. ErratumIn: PMID30571943. Subclaims: SC1, SC2, SC5.

[4] β-Amyloid peptides display protective activity against the human Alzheimer's disease-associated herpes simplex virus-1.

Shows: Aβ inhibits HSV entry with pre- or co-treatment but not later dosing; adenovirus is not inhibited.

Does not show: Applied peptides and timing constrain generalization; no endogenous infection-rescue test or plaque outcome.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[5] Protective Effect of Amyloid-β Peptides Against Herpes Simplex Virus-1 Infection in a Neuronal Cell Culture Model.

Shows: HSV-induced conditioned medium confers Aβ-dependent protection against secondary replication.

Does not show: Transformed cell secretion and conditioned-medium transfer cannot establish wild-type CNS necessity or physiological brain concentration.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[6] Acyclovir or Aβ42 peptides attenuate HSV-1-induced miRNA-146a levels in human primary brain cells.

Shows: Soluble Aβ42 attenuates HSV-associated miR146a and infection readouts.

Does not show: Applied peptide response does not identify native dose, unique sequestration mechanism or protective plaques.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[7] Antimicrobial activity of peptides derived from human ß-amyloid precursor protein.

Shows: Cationic peptides from an APP heparin-binding region kill microbes and disrupt membranes.

Does not show: These APP-derived fragments are not Aβ; APP antimicrobial function cannot automatically be credited to Aβ.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[8] Effects and mechanism of Aβ1-42 on EV-A71 replication.

Shows: Applied Aβ42 inhibits EV-A71 early entry with capsid/receptor binding and agglutination.

Does not show: Culture doses and receptor effects do not establish native mammalian CNS protection; contrast with adenovirus shows pathogen specificity.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[9] Alzheimer's Progenitor Amyloid-β Targets and Dissolves Microbial Amyloids and Impairs Biofilm Function.

Shows: Aβ remodels bacterial amyloids and impairs adhesion; tail-injected fibrils colocalize with Aβ in fish.

Does not show: Injected peripheral fibrils are distinct from CNS plaques; remodeled FapC remains toxic to SH-SY5Y cells.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3, SC5.

[10] Mechanistic insights into the role of amyloid-β in innate immunity.

Shows: At 7.5 µM, Aβ oligomers shift bacteria from biofilm to planktonic phase without significant total-density loss and increase gentamicin susceptibility; curli-null control lacks these effects.

Does not show: The neuronal oligomer reduction with curli deletion is nonsignificant (P=0.07). Redistribution is not direct killing; applied micromolar peptide and antibiotic synergy do not establish native CNS necessity.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 4:Materials and methods:paragraph 3, 4:Materials and methods:paragraph 5, 4:Materials and methods:paragraph 7, 1:Results. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3, SC5.

[11] Human amylin is a potent antimicrobial peptide that exhibits antimicrobial synergism with the amyloid beta protein.

Shows: Amylin and Aβ interact synergistically against bacteria including in 3D neural cultures.

Does not show: Amylin is the main tested effector; no endogenous mammalian CNS Aβ loss/rescue or human plaque outcome.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[12] Amyloid precursor protein is a restriction factor that protects against Zika virus infection in mammalian brains.

Shows: APP loss or blockade increases Zika replication and neonatal mouse mortality; APP acts as a decoy receptor.

Does not show: APP extracellular receptor function is distinct from cleaved Aβ antimicrobial activity; neonatal Zika cannot establish Aβ necessity in adult AD.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: Scite linear original body offset 0 length 8000: Introduction and first Results, truncated at next paragraph. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

[13] Alzheimer's disease-associated β-amyloid does not protect against herpes simplex virus 1 infection in the mouse brain.

Shows: 5XFAD fails to protect against HSV; abundant plaques lack viral particles, and aged survival differences are nonsignificant.

Does not show: Anti-gB T111 cross-reacts with plaques in uninfected controls; anti-gD/H do not show viral entrapment. Model and acute route constrain scope, but sex-only and dose/age analyses also fail significant survival protection.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 0:Results:paragraph 0, 0:Results:paragraph 2, 0:Results:paragraph 13, 0:Results:paragraph 14, 0:Results:paragraph 15, 0:Results:paragraph 16, 0:Results:paragraph 18, 0:Results:paragraph 20, 0:Results:paragraph 23. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC5.

[14] Herpes Simplex Virus 1 Infection Does Not Increase Amyloid-β Pathology in APP/PS1 Mice.

Shows: HSV infection does not increase plaque number or volume, with similar disease and survival across genotypes.

Does not show: One strain route and baseline model do not refute all infection-triggered amyloidosis; negative result constrains generality.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC5.

[15] Virulent Pseudomonas aeruginosa infection converts antimicrobial amyloids into cytotoxic prions.

Shows: Pseudomonas effectors shift host amyloids from antimicrobial to cytotoxic activity.

Does not show: Peripheral endothelial amyloids include Aβ and tau; no CNS Aβ-specific necessity or uniformly protective response.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC5.

[16] Identification of Microbiota within Aβ Plaque in APP/PS1 Transgenic Mouse.

Shows: Aβ-positive tissues contain more bacterial sequence relative abundance alongside barrier permeability changes.

Does not show: Low-biomass contamination and barrier failure can explain association; sequence presence does not show viable entrapment or survival benefit.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[17] Dysregulation of antimicrobial peptide expression distinguishes Alzheimer's disease from normal aging.

Shows: AMP expression and Aβ42 covary with age differently in AD flies.

Does not show: AMP transcription is not Aβ antimicrobial function; correlations do not identify infection initiation or protection.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[18] Porphyromonas gingivalis in Alzheimer's disease brains: Evidence for disease causation and treatment with small-molecule inhibitors.

Shows: Gingipains associate with AD pathology; infection induces Aβ and inhibition reduces bacterial load, inflammation and neuronal injury.

Does not show: Human association is postmortem; lowering Aβ after pathogen treatment does not establish Aβ entrapment as protective or necessary.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5, SC6.

[19] Recurrent herpes simplex virus-1 infection induces hallmarks of neurodegeneration and cognitive deficits in mice.

Shows: Repeated thermal reactivation produces Aβ, tau changes and progressive cognitive deficits.

Does not show: Stress, inflammation and direct viral injury remain alternatives; no Aβ-specific perturbation separates mediator from response.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[20] HHV-6A infection dysregulates autophagy/UPR interplay increasing beta amyloid production and tau phosphorylation in astrocytoma cells as well as in primary neurons, possible molecular mechanisms linking viral infection to Alzheimer's disease.

Shows: Infection disrupts autophagy and UPR with increased Aβ and tau phosphorylation.

Does not show: Stress pathways offer an indirect deposition mechanism; no native host survival or plaque benefit test.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[21] Intracerebral Infection with E. coli Impairs Spatial Learning and Induces Necrosis of Hippocampal Neurons in the Tg2576 Mouse Model of Alzheimer's Disease.

Shows: Infection worsens learning and neuronal necrosis in transgenic mice without altered Aβ concentrations.

Does not show: Acute inoculation plus ceftriaxone is distinct from chronic human infection; deposition is not an obligatory accompaniment.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC5.

[22] Immunohistological detection of Chlamydia pneumoniae in the Alzheimer's disease brain.

Shows: Chlamydial antigens occur near AD pathology but Thioflavin overlays show no direct plaque colocalization.

Does not show: Small sample and antibody specificity limits; regional proximity is not entrapment or protective action.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[23] Initial characterization of Chlamydophila (Chlamydia) pneumoniae cultured from the late-onset Alzheimer brain.

Shows: Culturable PCR-confirmed C. pneumoniae isolates were obtained from two AD brains.

Does not show: Small selected samples and earlier laboratory series do not establish population frequency, timing or plaque protective function.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[24] The load of Chlamydia pneumoniae in the Alzheimer's brain varies with APOE genotype.

Shows: C. pneumoniae load is higher in APOE4-bearing AD tissue.

Does not show: Cross-sectional genotype association cannot distinguish impaired clearance, susceptibility or protective amyloid function.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[25] Chlamydia pneumoniae induces Alzheimer-like amyloid plaques in brains of BALB/c mice.

Shows: C. pneumoniae infection precedes increased Aβ42-immunoreactive deposits, some Thioflavin-positive.

Does not show: Biomarker deposition is distinct from AD neurodegeneration and antimicrobial benefit; strain and histology specificity matter.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[26] Identification and localization of Chlamydia pneumoniae in the Alzheimer's brain.

Shows: Multiple assays detect viable transcriptionally active C. pneumoniae in a small AD series versus controls.

Does not show: Other laboratories report null detections; no prospective chronology or plaque-specific benefit measurement.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[27] Identification of Chlamydia pneumoniae and NLRP3 inflammasome activation in Alzheimer's disease retina.

Shows: C. pneumoniae associates with Aβ42 and inflammation; infection worsens cognition in a model.

Does not show: Disease-amplifying association and model effect do not identify protective Aβ; linked update records are not independent reports.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. UpdateOf: PMID40667156; UpdateOf: PMID40678205. Subclaims: SC5, SC6.

[28] Chronic Toxoplasma gondii infection enhances β-amyloid phagocytosis and clearance by recruited monocytes.

Shows: Infection reduces plaque burden through recruited monocytes; monocyte ablation increases Aβ.

Does not show: Immune clearance can oppose deposition, showing pathogen specificity; plaque reduction alone does not measure net cognitive protection.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[29] Toxoplasma gondii drives myeloid immune cell recruitment to amyloid plaques in Alzheimer's model mice.

Shows: Toxoplasma infection reduces plaques and recruits activated myeloid cells and T cells.

Does not show: Shares infection/5XFAD paradigm with earlier work; lineage tracing is not Aβ-specific necessity or human causation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[30] Latent herpes simplex virus type 1 in normal and Alzheimer's disease brains.

Shows: HSV DNA is found in both AD and normal brains with controls for amplification.

Does not show: Latent DNA presence does not distinguish active infection, etiologic timing or protective plaques.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[31] Herpes simplex virus type 1 DNA is located within Alzheimer's disease amyloid plaques.

Shows: HSV DNA is plaque-associated in AD and aged normal tissue, with greater plaque association in AD.

Does not show: DNA localization does not measure viable trapped viruses, temporal initiation or protective tissue outcomes; aged normals also show plaque-associated DNA.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[32] HSV presence in brains of individuals without dementia: the TASTY brain series.

Shows: HSV is detected in a small minority of brains, and most Aβ-positive brains lack detected HSV.

Does not show: Regional sampling and low burden affect sensitivity; presence or absence alone does not test Aβ antimicrobial function.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[33] Human Herpesvirus 6 Detection in Alzheimer's Disease Cases and Controls across Multiple Cohorts.

Shows: HHV6 lacks AD specificity across multiple brain repositories and assays.

Does not show: Negative detection constrains common pathogen association; cannot exclude other pathogens, remote cleared infection or restricted subgroups.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[34] Murine roseolovirus does not accelerate amyloid-β pathology and human roseoloviruses are not over-represented in Alzheimer disease brains.

Shows: Aβ-virus interaction inhibits infection in vitro but infection does not increase plaques; human roseoloviruses are not overrepresented.

Does not show: In vitro binding need not produce endogenous deposition; murine virus is an analogue and cohort sequencing shares assay limits.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC5, SC6.

[35] Reanalysis of Alzheimer's brain sequencing data reveals absence of purported HHV6A and HHV7.

Shows: Alternative sequence analysis finds very few HHV6A reads and no HHV7, identifying specificity problems.

Does not show: Same underlying data as Readhead; a methodological challenge rather than an independent biological cohort.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[36] Multiscale Analysis of Independent Alzheimer's Cohorts Finds Disruption of Molecular, Genetic, and Clinical Networks by Human Herpesvirus.

Shows: Original pipeline reports increased HHV6A/7 and networks involving APP metabolism in AD.

Does not show: Postmortem networks do not establish infection timing; specificity is directly challenged by reanalysis and subsequent detection studies.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[37] Herpesvirus genome integration in whole-genome sequences of dementia and control cohorts.

Shows: Partial HHV6 integration signals are more prevalent in synucleinopathy cohorts rather than broadly across dementias.

Does not show: Integrated segments do not establish active AD infection or plaque entrapment; cohort and ancestry effects remain possible.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[38] Herpes simplex virus type 1 in brain and risk of Alzheimer's disease.

Shows: HSV-positive APOE4 carriers show an AD association in a small case-control series.

Does not show: Large odds ratio with wide interval is not a plaque protection measurement, prospective cause or population necessity.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[39] Chlamydia pneumoniae can infect the central nervous system via the olfactory and trigeminal nerves and contributes to Alzheimer's disease risk.

Shows: Live organisms traverse olfactory/trigeminal paths and Aβ accumulates near inclusions.

Does not show: Adjacency after experimental exposure is not entrapment-mediated protection or spontaneous human AD.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[40] Detection of amyloid beta aggregates in the brain of BALB/c mice after Chlamydia pneumoniae infection.

Shows: Extracellular Aβ immunoreactivity also occurs in mock and untouched animals and differs from transgenic plaques.

Does not show: Assay and strain differences require reconciliation; these data do not unequivocally support infection-induced AD-like plaques.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[41] Detection of bacterial antigens and Alzheimer's disease-like pathology in the central nervous system of BALB/c mice following intranasal infection with a laboratory isolate of Chlamydia pneumoniae.

Shows: Transient Aβ deposits follow predominantly olfactory infection and resolve as antigen burden falls.

Does not show: Small cultures and strain dependence constrain inference; no demonstrated Aβ necessity or sustained AD phenotype.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[42] Failure to detect Chlamydia pneumoniae in brain sections of Alzheimer's disease patients.

Shows: Neither C. pneumoniae DNA nor antigens are detected in an independent AD series.

Does not show: Paraffin material and sensitivity may limit detection; contradicts common presence without proving universal absence.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[43] Failure to detect Chlamydia pneumoniae in the late-onset Alzheimer's brain.

Shows: Repeated nested PCR and culture fail to unequivocally detect C. pneumoniae in AD or controls.

Does not show: Stored-tissue sampling and organism burden constrain negative inference; no direct plaque functional experiment.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[44] Molecular and immunological evidence of oral Treponema in the human brain and their association with Alzheimer's disease.

Shows: Oral Treponema DNA and antigens are more frequent in small AD samples than controls.

Does not show: Postmortem presence with possible nerve route does not establish disease initiation, viable plaque trapping or host protection.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[45] Investigation of Potential Brain Microbiome in Alzheimer's Disease: Implications of Study Bias.

Shows: Bacterial sequence patterns vary strongly with cohort, extraction, region and postmortem interval; AD differences are limited.

Does not show: Low biomass and processing biases can generate apparent brain microbial signals; no viable organism or plaque function proof.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[46] Donanemab in Early Symptomatic Alzheimer Disease: The TRAILBLAZER-ALZ 2 Randomized Clinical Trial.

Shows: Donanemab clears plaque and slows decline at 76 weeks with ARIA and treatment-related deaths.

Does not show: Net benefit is bounded to enrolled early AD; immune effects and no infection stratification prevent inferring every plaque is harmful or AMP function absent.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[47] Lecanemab in Early Alzheimer's Disease.

Shows: Lecanemab lowers amyloid and yields less cognitive/functional decline with infusion reactions and ARIA.

Does not show: Targets soluble protofibrils as well as aggregates; clinical improvement is modest, not prevention or proof against pathogen-specific protection.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[48] Donanemab in Early Alzheimer's Disease.

Shows: Phase 2 donanemab improves primary iADRS but most secondary outcomes lack substantial differences.

Does not show: Small early trial and ARIA limit generalization; plaque reduction does not uniquely identify a microbial mechanism.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[49] A randomized, double-blind, phase 2b proof-of-concept clinical trial in early Alzheimer's disease with lecanemab, an anti-Aβ protofibril antibody.

Shows: Phase 2 lecanemab misses its 12-month primary probability threshold despite positive 18-month measures.

Does not show: Primary versus later secondary outcomes must remain distinct; correction metadata noted and full correction not audited.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID35598024. Subclaims: SC6.

[50] Two Randomized Phase 3 Studies of Aducanumab in Early Alzheimer's Disease.

Shows: Aducanumab meets EMERGE primary endpoint but not ENGAGE despite target engagement in both.

Does not show: Futility halt and discordant efficacy prevent counting two convergent successful causal lines; no infectious subgroup test.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[51] Two Phase 3 Trials of Gantenerumab in Early Alzheimer's Disease.

Shows: Gantenerumab lowers plaques without significant slowing in either GRADUATE trial; ARIA occurs.

Does not show: Incomplete clearance and antibody-specific effects constrain interpretation; null benefit does not prove plaques protective against infection.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[52] Trial of Solanezumab in Preclinical Alzheimer's Disease.

Shows: Solanezumab fails to slow decline in amyloid-positive cognitively unimpaired participants over 240 weeks.

Does not show: Monomer-targeting differs from plaque clearance; no infection stratification or necessity test.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[53] Randomized Trial of Verubecestat for Mild-to-Moderate Alzheimer's Disease.

Shows: Verubecestat fails to reduce cognitive or functional decline and increases adverse events.

Does not show: BACE1 has other substrates and broad processing functions; drug harms cannot be assigned specifically to loss of Aβ antimicrobial defence.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC6.

[54] Trial of Solanezumab for Mild Dementia Due to Alzheimer's Disease.

Shows: EXPEDITION3 fails its solanezumab primary endpoint; secondary results are descriptive after hierarchy failure.

Does not show: Target species and exposure differ from successful antibodies; no infectious-protection mechanism measurement.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[55] Cognitive outcomes in trials of two BACE inhibitors in Alzheimer's disease.

Shows: Several cognitive measures worsen with two BACE inhibitors while verbal fluency improves.

Does not show: Secondary endpoints and enzyme pleiotropy restrict causation; physiological Aβ functions need not be antimicrobial.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC6.

[56] Amyloid beta mediates memory formation.

Shows: Endogenous Aβ participates in memory and picomolar applied Aβ enhances consolidation.

Does not show: Supports a non-antimicrobial physiological alternative; no microbial challenge or plaque protective function.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3.

[57] Amyloid-Beta Mediates Homeostatic Synaptic Plasticity.

Shows: Aβ rescues homeostatic synaptic strengthening in APP-null tissue whereas sAPPα does not.

Does not show: A native peptide rescue exists for synaptic plasticity, distinct from infection rescue; ex vivo culture does not settle adult CNS immunity.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC4.

[58] Aggregation shifts amyloid-β peptides from synaptogenic to synaptotoxic.

Shows: Free Aβ can be synaptogenic while aggregation shifts it toward synaptotoxicity; scrambled/inverted controls are inactive.

Does not show: Chemically controlled culture is distinct from native pathogen defence; aggregation-state duality argues against assuming all deposition protective.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3, SC6.

[59] Selective suppression of oligodendrocyte-derived amyloid beta rescues neuronal dysfunction in Alzheimer's disease.

Shows: Suppressing oligodendrocyte Aβ improves neuronal function and pathology in vivo.

Does not show: Secretase/cell-source perturbation is not an infection-stratified human plaque experiment; supports bounded harmful contribution.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC6.

[60] NLRP3 is activated in Alzheimer's disease and contributes to pathology in APP/PS1 mice.

Shows: Inflammasome loss protects memory and enhances Aβ clearance in mice; human caspase activation is observed.

Does not show: Germline immune-pathway effects do not isolate plaque toxicity or require infection; sterile amplification remains a credible alternative.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5, SC6.

[61] β-amyloid binds to microglia Dectin-1 to induce inflammatory response in the pathogenesis of Alzheimer's disease.

Shows: Dectin1 loss reduces Aβ-triggered inflammation and functional deficits through Syk/NFκB.

Does not show: Infused peptide is a forced model; receptor-dependent sterile injury differs from endogenous AD initiation and antimicrobial defence.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5, SC6.

Documentary record

Every grade on this site is reproducible from its record. The files below are the posted edition exactly as assessed.

• ADVERSARIAL_REVIEW.md
• SHA256.txt
• adversarial-findings.json
• antimicrobial-amyloid.review.md
• antimicrobial-amyloid.scorecard.json
• codex-correction-provenance.json
• consensus-addendum.md
• consensus-validation.json
• edition-approval-audit.json
• reader-summary.json
• validation-report.json
• validation-round2-2026-10-14.json
• validation.txt

Sources: 61 · load-bearing claims: 68 · verified: 68 · unverified: 0 · contradicted: 0 · schema 5.3