The 3R/4R Isoform Divide: Why PSP and AD Require Different Strategies

PSP is defined by the predominant accumulation of 4-repeat (4R) tau isoforms — forming globose neurofibrillary tangles, tufted astrocytes, oligodendroglial coiled bodies, and neuropil threads — in contrast to the mixed 3R/4R tau found in Alzheimer’s disease. This isoform-selective distinction, explicitly identified in literature from the Center for Systems Neuroscience (Hanover), is not a subtle biological nuance: it has already produced at least one high-profile clinical failure and is now reshaping how the field designs tau-directed therapeutics for each condition.

The clearest demonstration of the consequences of ignoring isoform biology is the Phase II failure of gosuranemab (C2N-8E12) in PSP. This humanized IgG4 antibody, designed to recognize aggregated tau, showed no therapeutic benefit in PSP patients — a result documented in the 2024 Genentech patent filing as strategic context for developing improved tau antibody strategies. The mechanistic explanation is straightforward: an antibody optimized against the mixed 3R/4R tau conformers of AD may not adequately engage the pure 4R aggregates that define PSP pathology.

A patent from Tel Aviv University (Ramot at Tel Aviv University Ltd.) introduces a further layer of complexity: ADNP interactions with the splicing machinery may suppress exon 10 inclusion, modulating the 3R/4R ratio in a sex-dependent and dose-dependent manner. The davunetide (NAP) PSP trial retrospective analysis in this filing suggests that a subpopulation with dynamic 3R tau may have been underrepresented, partially explaining the negative result — and points toward sex-stratified dosing as an emerging design consideration for future PSP trials.

Across the retrieved patent dataset, hyperphosphorylated tau at specific residues — particularly pS396, pS404, pT217, pT231, and pT181 — is the most frequently cited molecular target. Multiple assignees from Axon Neuroscience SE describe antibodies binding “regions of the tau protein that contribute to the initiation and propagation of pathological tau-tau interactions.” A 2024 patent from Adel Company takes a distinct approach, describing an anti-tau antibody binding acetylated tau at lysine 280 (acK280) — showing inhibition of aberrant tau aggregation and improved motor and cognitive function in dementia-induced animal models, separate from the phosphorylation-targeting paradigm.

The mechanistic rationale for tau-directed therapy across both diseases is grounded in tau’s dual role: under normal conditions it stabilizes microtubules and supports axonal transport; upon hyperphosphorylation, it dissociates from microtubules, misfolds, and seeds prion-like cell-to-cell spread of pathological conformers. According to WIPO-registered filings in this dataset, Genentech frames the therapeutic rationale for anti-tau antibodies specifically around intercepting this extracellular spread in the brain environment — citing evidence that tau pathology in AD correlates spatially with affected cortical networks and associated cognitive decline.

Passive and Active Immunotherapy: Phosphoepitope Antibodies and Peptide Vaccines

Passive immunotherapy with phosphoepitope-selective monoclonal antibodies is the most densely represented tau clearance modality in the retrieved patent dataset, with evidence from both commercial and academic assignees spanning multiple jurisdictions. The mechanistic logic proposed across these filings includes antibody penetration via altered blood-brain barrier permeability in tauopathy, peripheral sink effects, antibody-secreting cells trafficking from the periphery to the CNS, and intracellular and transcytotic IgG transport.

Axon Neuroscience SE is the most prolific assignee in this dataset for tau-targeted therapeutic antibodies, with patent families spanning Singapore, Hungary, Israel, and China. Their filings describe antibodies binding pathological tau-tau interaction regions — specifically phosphorylated epitopes including pTau379-408 (P-Ser396,404). H. Lundbeck A/S patents specifically describe monoclonal antibodies with improved affinity for the pS396 residue on pathological hyperphosphorylated paired helical filament (PHF)-tau. New York University filings describe antibodies selective for the pSer404 epitope and methods for clearing tau from brain via immunogenic tau peptides.

The 2024 Genentech patent (Brazil) describes semorinemab as a “first-in-class immunotherapy” for mild-to-moderate AD, with clinical endpoints including CDR-SB, RBANS, and ADAS-Cog13 — explicitly describing clinically meaningful reduction in cognitive decline rate and retention of memory. This constitutes the strongest clinical translation signal in the retrieved dataset. Semorinemab’s mechanism centers on intercepting cell-to-cell spread of pathological tau in the extracellular brain environment, a rationale supported by preclinical data showing that tau knockdown in transgenic mouse models prevented cognitive deficits.

“Tau antibodies effective against mixed 3R/4R tauopathy in AD may not translate to pure 4R tauopathies — the Phase II failure of gosuranemab in PSP is the clearest demonstration that isoform biology cannot be ignored at the clinical design stage.”

Active immunotherapy represents a complementary strategy, with Janssen Pharmaceuticals (2024, Brazil) describing a liposomal delivery system incorporating a toll-like receptor 4 (TLR4) agonist, a helper T-cell epitope, a lipidated CpG oligonucleotide, and a surface-displayed tau phosphopeptide — specifically designed to maximize sustained humoral immunity in humans. Axon Neuroscience’s active vaccination filings cite preclinical data showing that active immunization with phospho-tau epitopes reduced tau aggregation and slowed behavioral phenotype progression in tau tangle mouse models. Clinical endpoints referenced in one Axon Neuroscience China filing (2022) include CDR-SB, RBANS, and ADAS-Cog scores, signaling preparation for human trial design.

An indirect immunotherapy approach is also represented in the dataset: Alector and Amgen filings target TREM2 (triggering receptor expressed on myeloid cells 2) on microglia as a strategy to activate innate immune clearance machinery. TREM2 agonism is presented as enhancing microglial phagocytic activity relevant to both amyloid and tau pathology — addressing the cellular clearance machinery rather than the tau protein itself, and complementary to direct tau-targeting antibodies. This signal aligns with broader trends in neuroinflammation research tracked by institutions including the NIH.

Proteasome-Targeting Degraders and ASOs: Reaching Intracellular Tau

Antibody-based approaches are largely limited to extracellular tau — a significant constraint given that the most toxic tau species, including hyperphosphorylated monomers and early oligomers, accumulate intracellularly before spreading. Two emerging modalities in the retrieved dataset directly address this gap: proteasome-targeting degraders mechanistically analogous to PROTACs, and antisense oligonucleotides that reduce tau at the transcript level.

PROTAC-Adjacent Degraders: PEST Motifs and TARBOD Molecules

The Regeneration Research Foundation (2023, China) describes multifunctional polypeptides comprising an antigen-binding domain — selective for tau, huntingtin, or alpha-synuclein — fused to a programmable PEST motif that directs the target protein to the proteasome. This peptide-based intracellular degradation strategy is applicable to AD, PSP, frontotemporal dementia (FTD), and chronic traumatic encephalopathy (CTE). It shares the bifunctional targeting logic of classical small-molecule PROTACs — a target-binding arm linked to a degradation-recruiting arm — but operates through a distinct structural implementation using polypeptide rather than small-molecule chemistry.

The Children’s Medical Center Corporation filing (2023, Canada) describes modulation of the immunoproteasome and deubiquitinase activity as a tau clearance mechanism in tauopathy, with modulation of inflammatory cytokine signaling as a secondary effect — linking proteasomal tau clearance to neuroinflammation biology. All three proteasome-targeting approaches are at preclinical stage, with in vitro and animal model data described. IP positioning in this space appears early-stage, representing a potentially contestable white space for organizations with ubiquitin-proteasome system biology expertise.

ASOs Targeting MAPT: BIIB080 and Gapmer Chemistry

BIIB080 (IONIS-MAPTRx) — an intrathecally administered antisense oligonucleotide targeting MAPT mRNA — is referenced in the Quris Technologies gapmer ASO patent (2025, China) as a named combination partner, signaling awareness of its clinical investigational status. The Quris filing describes backbone chemistry modifications in gapmer ASO design — specifically spacers improving pharmacokinetic and pharmacodynamic properties relative to unmodified ASOs — with neurological disease generalizability beyond its primary ALS and FTD application context.

The co-referencing of BIIB080 and AAV-vectorized anti-tau antibody strategies (Voyager Therapeutics, 2025) in the same dataset signals that gene therapy and oligonucleotide approaches are being positioned as complementary to antibody immunotherapy — particularly for patients who may need sustained, CNS-localized tau suppression that periodic antibody infusions cannot provide. This convergence is consistent with broader trends in neurological gene therapy documented by the FDA and academic institutions including Nature-published research on CNS oligonucleotide delivery.

A mechanistically distinct approach from Macquarie University (2019, Japan; 2022, China) adds further complexity to the intracellular tau biology picture. Rather than clearing tau, these filings describe promoting phosphorylation of tau at T205 via p38γ kinase (MAPK12) activation as a protective mechanism — counterintuitively using phosphorylation to disrupt toxic tau–PSD-95 signaling complexes. Knockout of p38γ in Alzheimer model mice (Alz17 line) worsened memory formation and consolidation, confirming in vivo protective relevance. This represents a mechanistic inversion of the conventional kinase-inhibition paradigm and suggests that not all tau phosphorylation events are pathogenic.

Biomarker-Guided Sequencing: From pT217 to MTBR-tau243

The most strategically significant emerging direction in the retrieved dataset is not a new therapeutic modality but an architectural shift in how tau-directed therapies are selected, sequenced, and monitored. Washington University filings describe quantitative mass spectrometry-based measurement of site-specific tau phosphorylation in CSF and blood, with pT217/T217 × Aβ42/40 ratios showing significant separation between AD and 4R tauopathy groups including PSP, CBS, and FTD-MAPT P301L — providing the patient stratification framework needed to rationally exclude 4R-enriched patients from AD-optimized trials.

MTBR-tau243 (microtubule-binding region fragment) is identified in Washington University patents as both a biomarker and a therapeutic monitoring tool: blood and CSF MTBR-tau243 levels correlate with tau PET signals and track tau deposition across the AD disease spectrum. The Wildsmith et al. patent (WO, 2026) formalizes this into a two-step treatment decision algorithm — anti-amyloid protofibril antibody as first-line based on p-tau217 levels, followed by anti-tau therapy guided by MTBR-tau243 monitoring. This sequential p-tau217/MTBR-tau243 architecture represents the most operationally detailed biomarker-integrated treatment paradigm in the retrieved dataset.

The Amylyx Pharmaceuticals filing (US, 2025) for AMX0035 — a combination of TURSO and sodium phenylbutyrate — specifically for AD and PSP uses CSF total tau >300 μg/mL and phospho-tau >70 μg/mL as biomarker thresholds for patient selection. This non-antibody, small-molecule combination approach to tau-associated neurodegeneration signals emerging interest in metabolic and proteostatic combination strategies alongside immunotherapy — and demonstrates that biomarker thresholds are being embedded in patent claims rather than left to clinical protocol design.

A retrieved paper from the Yale PET Center (2023) implicates cellular prion protein (PrPC) as a mediator of Aβ oligomer–driven tau accumulation and synapse loss — suggesting PrPC as a potential upstream combination target in the tau clearance architecture. This signal, alongside the NYU and Axon Neuroscience filings proposing simultaneous targeting of tau and Aβ as more effective than monotherapy, suggests that the biomarker sequencing question (Aβ-first vs. tau-first) will require resolution before rational combination immunotherapy trials can be designed. Regulatory guidance from agencies including the EMA on combination therapy development pathways will be relevant to this design challenge.

Strategic Implications for Tau Drug Development

The retrieved patent and literature dataset reveals a field in transition: from monotherapy antibody programs targeting a single tau epitope, toward multi-modal, biomarker-stratified, combination approaches that account for isoform biology, compartment access, and patient selection from the outset. Several strategic implications emerge directly from the evidence.

PSP immunotherapy requires confirmed 4R-tau binding and biomarker-stratified enrollment. The documented failure of gosuranemab in PSP — despite preclinical promise — underscores that tau antibodies effective against mixed 3R/4R tauopathy in AD may not translate to pure 4R tauopathies. Retrieved results suggest that PSP clinical programs require antibodies with confirmed 4R-tau binding, and ideally biomarker-stratified patient selection using CSF pT217 profiles to exclude 4R-enriched patients from AD-optimized trials.

Proteasome-targeting tau degraders represent a white space with growing patent activity. The PEST-motif multifunctional polypeptide and TARBOD approaches in this dataset are at preclinical stage but address intracellular tau — a compartment largely inaccessible to antibody-based approaches. IP positioning in this space appears early-stage and potentially contestable for organizations with ubiquitin-proteasome system biology expertise. The EPO patent landscape for bifunctional protein degraders in neurodegeneration remains less crowded than the equivalent oncology space.

Biomarker-guided sequential combination therapy is becoming an architectural standard. The Wildsmith et al. sequential p-tau217/MTBR-tau243 algorithm signals that future tau clearance programs will require companion diagnostic integration from the outset — a regulatory and IP co-development consideration for drug developers. Biomarker thresholds are already appearing in patent claims (Amylyx, 2025), not just clinical protocols.

Sex-stratified dosing in PSP deserves attention. The Tel Aviv University patent signals emerging recognition that tau-directed therapies may require sex-specific dose optimization in PSP, driven by differential ADNP expression and 3R/4R tau isoform responsiveness. This is an underexplored variable in current PSP trial design.

ASO and AAV gene therapy convergence positions these modalities as complements to antibody immunotherapy. For patients requiring sustained, CNS-localized tau suppression that periodic antibody infusions cannot provide, MAPT-targeting ASOs (BIIB080) and AAV-vectorized anti-tau constructs (Voyager Therapeutics) are being positioned as the long-acting backbone of combination regimens. The IP landscape for CNS-delivered oligonucleotides and viral vectors in tauopathy remains a high-priority monitoring space for R&D strategy teams.